Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Reservoir characterization of the Duvernay Formation, Alberta : a pore- to basin-scale investigation Munson, Erik Ole 2015

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2015_november_munson_erik.pdf [ 36.14MB ]
Metadata
JSON: 24-1.0166722.json
JSON-LD: 24-1.0166722-ld.json
RDF/XML (Pretty): 24-1.0166722-rdf.xml
RDF/JSON: 24-1.0166722-rdf.json
Turtle: 24-1.0166722-turtle.txt
N-Triples: 24-1.0166722-rdf-ntriples.txt
Original Record: 24-1.0166722-source.json
Full Text
24-1.0166722-fulltext.txt
Citation
24-1.0166722.ris

Full Text

RESERVOIR CHARACTERIZATION OF THE DUVERNAY FORMATION, ALBERTA: A PORE- TO BASIN-SCALE INVESTIGATIONbyERIK OLE MUNSONB.Sc., Illinois State University, 2011A DISSERTATION SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFDOCTOR OF PHILOSOPHYinTHE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES(Geological Sciences)THE UNIVERSITY OF BRITISH COLUMBIA(Vancouver)September 2015©Erik Ole Munson 2015iiABSTRACTThe reservoir properties of the Duvernay Formation mudrock gas and oil reservoir (“shale gas”) in Alberta were investigated. The investigation included an assessment of current methodologies utilized to study mudrocks, development of new methodologies, pore- to basin-scale characterization and integration of core data with wireline logs. The Duvernay exists over multiple thermal maturity boundaries and provides a laboratory to investigate numerous pertinent research questions. Deposition of organic-rich Duvernay mudrocks was controlled by the spatial relationship to Leduc reef complexes. Greater thicknesses (> 70 m) of Duvernay mudrocks are found within embayments where oxygenated water circulation was most restricted. The Duvernay progressively thins (< 5 m) to the basin center where organic-lean lime mudstones were deposited instead of Duvernay mudrocks. Core samples were taken from eight wells to form a high-resolution database of rock properties including mineralogy, total organic carbon (TOC) content, and total porosity. Duvernay mudrocks average between 2.4 % and 5.6 % total porosity and between 3.0 % and 4.4 % TOC per well. A regional lime mudstone within the Duvernay averages less than 2 % porosity and 0.5 % TOC and therefore is considered to be of poor reservoir quality. Artificial neural network models were used to successfully integrate laboratory data with wireline logs to predict rock properties in wells without laboratory data, providing enhanced correlation coefficients over linear regressions (R=0.82 versus R=0.67). The pore structure of Duvernay mudrocks varies systematically with thermal maturity. Fine pore sizes (micro- and fine mesopores) hosted within organic matter progressively increase in volume as thermal maturity increases. Coarse pore sizes (coarse meso- to macropores) progressively decrease in volume with increasing burial depth due to compaction. Wet and dry gas window samples have average pulse-decay permeabilities (PDP) an order of magnitude higher (1.8 x 10-4 mD) than oil window samples (1.6 x 10-5 mD), despite the shift in pore modality to finer sizes. The network of fine pores developed with maturity and associated with organic matter is sufficiently connected to contribute to higher PDP values. Gas expansion permeability experiments further indicate fine pores are connected and yield increased matrix permeabilities.iiiPREFACEThe research presented is an original study by the author, Erik O. Munson. The research was designed by Erik O. Munson and supervisor R. Marc Bustin with input from committee members Lori Kennedy, Stuart Sutherland and Kurt Grimm. All work presented was conducted within the Unconventional Reservoir Research Laboratory at The University of British Columbia, Vancouver unless otherwise stated. The research was funded by Shell Canada Ltd., Trilogy Energy Corp., Talisman Energy Inc., Athabasca Oil Corp., Husky Energy Inc., Encana Corp. and Penn West Petroleum Ltd. Funding was also provided in part by a Geoscience BC grant to R. Marc Bustin. All core samples were taken by the author from cores at the Alberta Energy Regulator Core Research Facility in Calgary, Alberta. Pyrolysis analyses were performed by Trican Geological Solutions, Calgary. Built-for-purpose permeameters were constructed at UBC by the author; built-for-purpose parts were designed by the author and constructed in the machine shop located in Earth, Ocean and Atmospheric Sciences at UBC. Laboratory assistants Scott Malo, Maryam Alsadiq, Benjie Friedman, Melissa Friend and Kristal Li helped in various capacities to prepare samples. Wireline log suites and additional pyrolysis data were compiled from public sources. Research results are subdivided into 5 chapters which comprise the thesis. The chapters are designed as standalone manuscripts. Two manuscripts have been submitted for publication. Co-authorship details are summarized below. Chapter 2 is submitted as: Munson, E. O., G. R. L. Chalmers, R. M. Bustin, and K. Li, Utilizing smear mounts for X-ray diffraction as a fully quantitative approach in rapidly characterizing the mineralogy of shale gas reservoirs. Munson was responsible for 95 % of the research and 95 % of the writing. Chalmers provided research ideas and editorial comments on the composition of the manuscript. Bustin provided editorial comments. Li provided laboratory assistance throughout the research phase. Chapter 3 is submitted as: Munson, E. O., A. M. M. Bustin, and R. M. Bustin, Correcting for compressibility and closure in mercury intrusion porosimetry: Calculations for determining pore volume compressibility in fine grained reservoir rocks. Munson was responsible for 95 % of the research and 95 % of the writing. Bustin, A. M. M. provided significant editorial comments which improved the focus, structure and composition of the manuscript. Bustin, R. M. provided editorial ivcomments. Thompson, W. F. provided assistance with portions of the mathematics (declined co-authorship). Chapter 4 comprises research results written by Munson, E. O. supervised by R. M. Bustin. Bustin provided guidance which improved the readability of the manuscript. Portions of Chapter 4 were presented as an oral presentation: Munson, E. O., and R. M. Bustin, Visual and quantitative investigation of porosity across thermal maturity boundaries in the Duvernay Formation, Alberta: CSPG GeoConvention 2014, Calgary, Alberta. Chapters 5 and 6 comprise research results written by Munson, E. O. under supervision of R. M. Bustin. Bustin provided editorial comments which improved the readability and structure of the manuscripts.vTABlE oF ConTEnTSABSTRACT ................................................................................................................................... iiPREFACE ..................................................................................................................................... iiiTABlE oF ConTEnTS ...............................................................................................................vlIST oF TABlES ........................................................................................................................ ixlIST oF FIGURES .......................................................................................................................xlIST oF ABBREVIATIonS, SYMBolS AnD ACRonYMS ............................................. xixACKnoWlEDGMEnTS ........................................................................................................ xxiiDEDICATIon .......................................................................................................................... xxiiiChapter 1: Introduction ............................................................................................................11.1 InTRoDUCToRY STATEMEnTS ........................................................................11.2 ThESIS oBjECTIVES ............................................................................................31.2.1	 Laboratory	analyses	and	fine-grained	lithologies ...................................41.2.2 Thermal maturity, compaction and pore size distribution ....................41.2.3 Pore structure evolution and permeability ..............................................41.2.4 Pore structure wettability and permeability ...........................................41.2.5	 Regional	mudrock	reservoir	characterization .........................................41.3 ThESIS STRUCTURE ............................................................................................5Chapter	 2:	 Utilizing	smear	mounts	for	X-ray	diffraction	as	a	fully	quantitative	approach	in	rapidly	characterizing	the	mineralogy	of	shale	gas	reservoirs ..............................................62.1 InTRoDUCTIon.....................................................................................................62.2	 XRD	MethoDoLogy	BaCkgRoUnD ............................................................62.3 MoDIFIED SMEAR MoUnT METhoD .............................................................92.4 MATERIAlS AnD METhoDoloGY ...............................................................102.4.1	 Whole	rock	sample	preparation .............................................................102.4.2	 artificial	sample	preparation ..................................................................112.4.3	 XRD	and	quantitative	phase	analysis ....................................................122.4.4	 Scanning	electron	microscope .................................................................122.4.5 Particle size analysis ................................................................................122.5 RESUlTS .................................................................................................................132.5.1	 Whole	rock	analysis .................................................................................132.5.2	 artificial	sample	analysis .........................................................................132.5.3 BSEM ........................................................................................................142.5.4 Particle size analysis ................................................................................142.6 DISCUSSIon ..........................................................................................................152.6.1	 Whole	rock	analysis .................................................................................152.6.2	 artificial	sample	analysis .........................................................................162.7 ConClUSIonS AnD RECoMMEnDATIonS ................................................16viChapter	 3:	 Correcting	for	compressibility	and	closure	in	mercury	intrusion	porosimetry:	Calculations	for	determining	pore	volume	compressibility	in	fine	grained	reservoir	rocks ..............................................................................................................................303.1 InTRoDUCTIon...................................................................................................303.2	 eXPeRIMentaL	MethoDoLogy ................................................................323.2.1 Mercury injection porosimetry ...............................................................323.2.2 Closure correction ....................................................................................333.2.3 Compression correction ...........................................................................343.2.4 Stressed porosity calculation ...................................................................353.2.5 Assumptions..............................................................................................363.3 RESUlTS AnD DISCUSSIon ..............................................................................373.3.1	 Sample	size	effect	on	closure ...................................................................373.3.2 Commercial lab closure correction .........................................................383.3.3	 Compression	modeling	and	correction ..................................................383.3.4	 Closure	and	compression	effect	on	permeability	estimation ...............393.4 ConClUSIonS AnD RECoMMEnDATIonS ................................................40Chapter	 4:	 Impact	of	thermal	maturity	and	compaction	on	the	pore	size	distribution	and	matrix	permeability	in	the	shale	gas	and	shale	oil	producing	Duvernay	Formation,	Alberta ..........................................................................................................................................534.1 InTRoDUCTIon...................................................................................................534.2 AnAlYTICAl ConSIDERATIonS ...................................................................564.3 MATERIAlS, METhoDoloGY AnD lIMITATIonS ..................................574.3.1 Core samples .............................................................................................574.3.2	 Mineralogy ................................................................................................584.3.3	 organic	geochemistry ..............................................................................584.3.4 Pore size distribution ...............................................................................584.3.5	 FIB/Fe-SeM .............................................................................................594.3.6	 Confined	matrix	permeability .................................................................594.3.7 Wettability ................................................................................................604.4 RESUlTS .................................................................................................................614.4.1 Sample composition and maturity ..........................................................624.4.2	 nitrogen	LPgS .........................................................................................624.4.3 Carbon dioxide lPGS..............................................................................654.4.4	 FIB/Fe-SeM .............................................................................................664.4.5 Matrix permeability .................................................................................674.4.6 Wettability ................................................................................................684.5 DISCUSSIon ..........................................................................................................694.5.1 Duvernay microstructure and maturity ................................................69vii4.5.2	 Impact	of	maturity	on	permeability .......................................................724.5.3 GEPP permeability and microstructure ................................................744.5.4	 Wettability,	liquid	permeability	and	pore	structure .............................764.6 ConClUSIonS .....................................................................................................78Chapter	 5:	 Regional	reservoir	characterization	model	for	the	shale	gas	and	shale	oil	producing	Duvernay	Formation,	alberta,	Part	I:	Regional	reservoir	distribution	and	reservoir	properties	using	wireline	log	signatures	and	high-resolution	laboratory	data ....1105.1 InTRoDUCTIon.................................................................................................1105.2 REGIonAl GEoloGIC BACKGRoUnD .....................................................1115.2.1 overview .................................................................................................1115.2.2 Depositional environment .....................................................................1125.3 STUDY AREA, InFoRMAl UnITS AnD loG SIGnATURES ....................1155.4 MATERIAlS, METhoDoloGY AnD lIMITATIonS ................................1175.4.1 Core samples ...........................................................................................1175.4.2	 Well	log	suites .........................................................................................1175.4.3	 Mineralogy ..............................................................................................1175.4.4	 organic	geochemistry ............................................................................1185.4.5	 Unconfined	porosity ...............................................................................1185.5 RESUlTS ...............................................................................................................1185.5.1 Structure .................................................................................................1195.5.2 Thermal maturity ...................................................................................1205.5.3 Reservoir pressure and temperature ....................................................1215.5.4	 Lithostratigraphy ...................................................................................1225.5.5	 Majeau	Lake ...........................................................................................1225.5.6 lower Duvernay .....................................................................................1235.5.7 Middle carbonate ...................................................................................1245.5.8 Upper Duvernay .....................................................................................1265.5.9 Ireton .......................................................................................................1305.5.10	 Regional	comparison	of	non-reservoir	to	reservoir	lithologies ..........1305.6 DISCUSSIon ........................................................................................................1315.6.1	 Pattern	of	basin	fill .................................................................................1315.6.2	 Reef	debris	and	reservoir	quality .........................................................1335.6.3 Controls on reservoir pressure and temperature ................................1345.7 ConClUSIonS ...................................................................................................135Chapter	 6:	 Regional	reservoir	characterization	model	for	the	shale	gas	and	shale	oil	producing	Duvernay	Formation,	alberta,	Part	II:	Integrating	high-resolution	laboratory	data	and	wireline	logs	using	artificial	neural	networks	to	quantify	regional	reservoir properties ....................................................................................................................180viii6.1 InTRoDUCTIon.................................................................................................1806.2 MATERIAlS .........................................................................................................1816.2.1 Core samples ...........................................................................................1816.2.2	 Well	log	suites .........................................................................................1816.3 ARTIFICIAl nEURAl nETWoRKS ...............................................................1816.3.1 ToC model .............................................................................................1836.3.2 Quartz model ..........................................................................................1846.3.3 Carbonate model ....................................................................................1856.3.4 Porosity model ........................................................................................1866.4 DISCUSSIon ........................................................................................................1866.4.1	 Controls	on	toC,	mineralogy	and	porosity	distribution ...................1866.4.2	 Dynamic	reservoir	quality	assessment .................................................1886.4.3 Vertical variation in reservoir properties ............................................1896.5 ConClUSIonS ...................................................................................................191Chapter 7: Conclusions .........................................................................................................2127.1 oVERVIEW ..........................................................................................................2127.2 KEY FInDInGS ...................................................................................................2137.3 FUTURE RESEARCh .........................................................................................214REFEREnCES ...........................................................................................................................216aPPenDIX	 a:	 MIneRaLogy	XRD	ReSULtS ..............................................................231aPPenDIX	 B: PETRoPhYSICAl RESUlTS ..................................................................239aPPenDIX	 C: InSTRUMEnTS AnD SoFTWARE ........................................................247ixlIST oF TABlESTable 2.1 Mineral composition of artificial mudrocks (ES1 – ES4). ...............................................19Table 2.2 Calculated Rietveld weight percent for the phases present in each formation using different methods of sample preparation. Spray = Micronized, spray dried front drift preparation; Smear = Hand ground, smear mount preparation; Back = Micronized, back mount preparation. .............................................................................20Table 2.3 Absolute weight percent error statistics for each phase. ..................................................22Table 2.4 Quantitative results for the artificial samples using the spray dried, smear mount and back mount preparation methods. Bias is the absolute difference between weighed mineral percent and measured weight percent. ................................................24Table 2.5 Particle size statistics for three whole rock samples prepared by hand grinding and micronizing. Samples were analyzed before and after ultrasonic treatment. d(0.1) = Diameter in which 10% of particles are less than. ..........................................................28Table 3.1 Raw and closure corrected (denoted by subscript c) density (ρ) and porosity (φ) data for the four samples crushed to various sizes. .........................................................49Table 4.1 Well location and averaged data for wells in the study. UWI = unique well identifier, Well ID is the identifier used in this study. SSTVD = sub-surface true vertical depth to the top of the Duvernay. HI = hydrogen index, calculated as S2*TOC/100. OI = oxygen index, calculated as S3*TOC/100. Average Tmax is not valid and therefore omitted for overmature samples. See text for discussion of qualitative maturity descriptor. .......................................................................................82Table 4.2 Samples and properties selected for PDP and GEPP analysis. ........................................95Table 5.1 Well location and averaged data for all wells in the study. UWI = unique well identifier, Well ID is the identifier used in this study. SSTVD = sub-surface true vertical depth to the top of the Duvernay. HI = hydrogen index, calculated as S2*TOC/100. OI = oxygen index, calculated as S3*TOC/100. See text for discussion of qualitative maturity descriptor. ...............................................................139Table 6.1 Detailed analysis wells used for developing ANN models. ...........................................193Table 6.2 Criteria to define reservoir quality models. ....................................................................207Table A.1 XRD results for detailed analysis wells ........................................................................231Table B.1 Petrophysical results from detailed analysis wells ........................................................239xlIST oF FIGURESFigure 2.1 Relation of the 10 µm recommended crystallite size to the particle size of fine grained rocks (Wentworth, 1922). .................................................................................17Figure 2.2 General geographical subcrop locations of the Barnett, Woodford, Haynesville, Eagle Ford, Duvernay and Muskwa shales. ..................................................................18Figure 2.3 Graphs of the quantitative results from each methodology by formation. .....................21Figure 2.4 Graph of spray dry weight percent versus smear mount weight percent by individual phases. The line is a 1:1 ratio. ......................................................................23Figure 2.5 (A1, A2, A3) Crushed Duvernay whole-rock sample passed through 60 mesh. The large particles are aggregates of smaller crystals. Particles can be seen in excess of 250 µm. (B1, B2, B3) Duvernay sample hand ground with agate mortar and pestle in ethanol for two minutes. The larger particles are disaggregated into smaller crystallites, many particles have been reduced to less than 10 µm. Large particles in excess of 30 µm can remain. (C1, C2, C3) Duvernay sample micronized with agate elements in ethanol for seven minutes. Most particles have been reduced to less than 10 µm, some larger crystallites remain in excess of 30 µm. The phyllosilicate minerals (muscovite) are notably difficult to grind. .................26Figure 2.6 Backscatter SEM images of the smear mounted Duvernay slide that was analyzed by XRD. Overall there is no obvious preferred orientation of clay mineral grains in whole rock samples. ..................................................................................................27Figure 2.7 Twelve total mounts from 3 users illustrate the reproducibility of the smear mount method. ..........................................................................................................................29Figure 3.1 Incremental volume (mL) versus injection pressure (psi) from an MIP analysis on a typical fine-grained mudstone of the Duvernay Fm. The closure volume is subtracted from the uncorrected raw data (blue curve in A) to obtain the closure corrected incremental volume curve (red curve in A). Subtracting the volume intruded due to compression (blue curve in B) from the closure corrected incremental curve (red curve in B) gives the closure and compression corrected incremental curve (red curve in C). ...............................................................................42Figure 3.2 Semi-log cumulative intrusion plot for the typical mudstone sample used in Figure 3.1. Red line is a power law fit over the interval attributed to compression. Where the fit deviates is closure (at low pressure) or intrusion (at high injection pressure). ................................................................................................43Figure 3.3 Compressibility (Cp, psi-1) versus injection pressure (P, psi). A power law function is fit (red) to the range attributed to compression. Closure and intrusion are apparent where Cp (black) deviates from the power law fit. ....................................44Figure 3.4 Porosity at stress calculated for the typical mudstone sample used in previous figures (A). The relative percent reduction at any NCS can also be calculated and xiused to correct porosities measured at ambient conditions (B). ....................................45Figure 3.5 Porosity reduction versus TOC (A) and clay (B) for mudstones of the Duvernay Fm. analyzed for this study. Poor correlations for both TOC and total clay content with porosity reduction, Rφ, suggest that the more ductile matrix components do not significantly contribute to compression. Red lines are linear least squares fits, R2 is the coefficient of determination and n is the number of samples. .........................................................................................................................46Figure 3.6 Helium porosity (%) versus MIP porosity (%) measured on mudstones of the Duvernay Fm. Helium porosity is consistently higher than MIP porosity. Black line is 1:1 ratio, red line is a linear least squares fit, R2 is the coefficient of determination and n is the number of samples. .............................................................47Figure 3.7 Incremental volume intrusion plot for the four mudstone samples of different sizes tested in this study, as well as stainless steel ball bearings for comparison. Closure increases as sample size and sorting increase. Quarter-inch stainless steel ball bearings (dark blue) do not exhibit significant closure in comparison to the rock samples. .................................................................................................................48Figure 3.8 Raw and closure corrected intrusion curves analyzed in this study compared with corrected data supplied by a commercial laboratory for a single mudstone sample from the Duvernay. ........................................................................................................50Figure 3.9 Porosity reduction (Rφ) versus injection pressure (psi) calculated for the four samples of various sizes. Porosity reduction at 3,500 psi NCS was chosen to illustrate the general trend of increasing porosity reduction and decreasing sample size. See text for discussion. .............................................................................51Figure 3.10 (A) Bulk mercury saturation versus injection pressure with uncorrected raw data (grey curves), closure corrected (blue curves) and compression corrected (orange curves) data for the typical mudstone used in Figures 3.1 - 3.4. The black dots represent the point at which Sb/Pc is a maximum, which is the parameter used in Swanson’s (1981) model to calculate permeability. (B) Swanson permeability calculated for all values of Sb/Pc. The black dots represent the Sb/Pc value used to derive Swanson permeability. The red line is confined permeability to helium at various net confining stresses. .......................................................................................52Figure 4.1 Pore size classification scheme used in this study. Modified from Sing et al. (1985). ...80Figure 4.2 Study location in central Alberta. Samples were taken from 19 wells total. Leduc Fm. reefs are light grey. Vitrinite reflectance lines are modified from Stasiuk and Fowler (2002). Well identifiers are given in Table 4.1. ................................................81Figure 4.3 Bulk mineral composition for LPGS samples normalized to total quartz and feldspar content, total clay content, and total carbonate content. ..................................83Figure 4.4 Modified Van Krevelan diagram (Tissot and Welte, 1984) for all wells analyzed in xiithe study. Duvernay kerogen is dominantly Type II. .....................................................84Figure 4.5 Typical N2 (A) and CO2 (B) isotherm for Duvernay samples within this study. ............85Figure 4.6 Nitrogen BJH pore size distribution for Duvernay samples by maturity level. (A) immature wells; (B) oil window ATH 13-18-64; (C) wet gas well ECA 11-8-26; (D) dry gas window wells. Samples from dry gas maturity generally have the least volume or coarse mesopores and macropores with the exception of HSK 10-33-56. .......................................................................................................................86Figure 4.7 Nitrogen BJH pore body distributions for immature wells colored by (A) TOC, (B) total quartz and feldspar content, (C) total carbonate content, and (D) total clay content. Note y-axis scale for immature samples is expanded compared to other maturity groups. ...................................................................................................87Figure 4.8 Nitrogen BJH pore body distributions for oil window well ATH 13-18-64 colored by (A) TOC, (B) total quartz and feldspar content, (C) total carbonate content, and (D) total clay content. .............................................................................................88Figure 4.9 Nitrogen BJH pore body distributions for wet gas window well ECA 11-8-62 colored by (A) TOC, (B) total quartz and feldspar content, (C) total carbonate content, and (D) total clay content. ...............................................................................89Figure 4.10 Nitrogen BJH pore body distributions for dry gas window wells colored by (A) TOC, (B) total quartz and feldspar content, (C) total carbonate content, and (D) total clay content. ..........................................................................................................90Figure 4.11 Texture and porosity associations from FE-SEM for early oil window well FB 10-4-51. Organic matter is generally non-porous, however fine mesopores (C) and coarse micropores (D) were imaged in two separate locations. Macropores exist in crack-like distributions within or at the boundaries of OM particles (D). Intra-particle clay hosted porosity is also evident (E). ..................................................91Figure 4.12 Organic matter-hosted porosity for oil window well ATH 13-18-64. Inter-crystalline porosity was found to be limited in two imaging locations. Organic-matter is both displays varying porosity, with pores generally within the mesopore size fraction. ..................................................................................................92Figure 4.13 Porosity and texture for wet gas window well ECA 11-8-62. The macropore in (A) and (B) is the largest pore observed in the study. Variable non-porous OM and porous OM was observed in multiple imaging locations (A, E). ...........................93Figure 4.14 Porosity and texture for dry gas window well HSK 10-33-56. Organic matter is meso- to macroporous (A, E) and exhibits the largest OM pores of all wells. Development of OM porosity is less variable than other maturity windows. Inter-crystalline porosity is associated with clay minerals (D) and carbonate grains (C). .....94Figure 4.15 Helium PDP versus decane PDP, performed on the on the same sample, from the oil window (ATH 13-18-64) and wet gas window (ECA 11-8-62). ..............................96xiiiFigure 4.16 The contact angle of water in air versus the difference between helium PDP and decane PDP. Black line is a power fit. ...........................................................................97Figure 4.17 Contact angle of water in air on Duvernay mudrocks vs TOC content. .......................98Figure 4.18 Evolution of mudrock microstructure for Duvernay samples analyzed in this study. Red circles are average value for the well. Levels of thermal maturation adapted from Peters (1986) and Peters and Cassa (1994). Organic matter and interparticle porosity abundance is estimated from FE-SEM imaging and correlations of TOC with LPGS values (see text for discussion).  ................................99Figure 4.19 Box and whisker plot of N2 BET surface area for all wells, sorted by increasing maturity. The fliers are data range, upper and lower limits of the box are 1st and 3rd quartiles, red line is the median and red square is the average. The wide range of values (e.g. SCL 11-1-38) shows the significant sedimentary variability. ....100Figure 4.20 Box and whisker plot of CO2 BET surface area for all wells, sorted by increasing maturity. The fliers are data range, upper and lower limits of the box are 1st and 3rd quartiles, red line is the median and red square is the average. ............................101Figure 4.21 Average CO2 BET surface area (A) and average CO2 total pore volume (B) versus HI colored by average total clay content. Average N2 BET surface area (C) and average N2 total pore volume (D) colored by average clay content. Only samples that have undergone significant thermal maturity (oil window and higher) are shown to highly effects of maturity. ..........................................................102Figure 4.22 Nitrogen BJH pore body distribution for oil window ATH 13-18-62 colored by He PDP. .......................................................................................................................103Figure 4.23 Nitrogen BJH pore body distributions only for samples with He PDP measurements from the oil window well ATH 13-18-64, colored by (A) TOC, (B) total quartz and feldspar content, (C) total carbonate content, and (D) total clay content. ........................................................................................................................104Figure 4.24 Nitrogen BJH pore body distribution for wet gas window well ECA 11-8-62 colored by He PDP. .....................................................................................................105Figure 4.25 Nitrogen BJH pore body distributions only for samples with He PDP measurements from the wet gas window well ECA 11-8-62, colored by (A) TOC, (B) total quartz and feldspar content, (C) total carbonate content, and (D) total clay content. .................................................................................................................106Figure 4.26 Nitrogen BJH pore body distribution for dry gas window wells colored by He PDP. .............................................................................................................................107Figure 4.27 Nitrogen BJH pore body distributions only for samples with He PDP measurements from the dry gas window wells HSK 10-33-56 and SCL 11-1-38, colored by (A) TOC, (B) total quartz and feldspar content, (C) total carbonate content, and (D) total clay content. .............................................................................108xivFigure 4.28 Pore size distributions colored by GEPP permeability. ..............................................109Figure 5.1 Stratigraphic table of formations within the central plains, Alberta. Adapted from ERCB (2009). ..............................................................................................................137Figure 5.2 Sample locations and Late Devonian paleogeography of the Leduc Fm. reefs. Mudstones of the Majeau Lake, Duvernay and Ireton fms. were deposited off-reef in the East and West Shale Basins. .......................................................................138Figure 5.3 Representative wireline section through the Duvernay Formation in the Kaybob region. Interbedded carbonates (e.g. 2840 m) are locally occurring within the upper Duvernay and can decrease overall reservoir thickness as well as inhibit hydraulic fracture growth. MD = measured depth; GR = total gamma radiation; DT = P-wave interval transit time; RHOB = bulk density, gamma; RES = resistivity; ILD = deep induction; ILM = medium induction; SFL = laterolog resistivity......................................................................................................................140Figure 5.4 Structure to the top of the Duvernay within the WSB and western East Shale Basin. Depths are meters below sea level. Surface dips are calculated by change in depth over change in surface distance perpendicular to contour lines. ...................141Figure 5.5 Majeau Lake residual structure map for the Kaybob region. A slope angle profile from A-A’ is included to illustrate deviations. Red line is a structural lineament from Lyatsky et al. (2005). ..........................................................................................142Figure 5.6 Lower Duvernay residual structure map for the Kaybob region. Red line is a structural lineament from Lyatsky et al. (2005). .........................................................143Figure 5.7 Middle carbonate residual structure map for the Kaybob region. Red line is a structural lineament from Lyatsky et al. (2005). .........................................................144Figure 5.8 Upper Duvernay residual structure map for the Kaybob region. Red line is a structural lineament from Lyatsky et al. (2005). .........................................................145Figure 5.9 Maturity map from vitrinite reflectance measurements made on Duvernay Fm. samples. Data adapted from Stasiuk and Fowler (2002). Vitrinite reflectance values of 0.60 % approximate the beginning of hydrocarbon generation. ..................146Figure 5.10 Average Duvernay Tmax from publicly available data and data generated for this study. ............................................................................................................................147Figure 5.11 Average Duvernay HI from publicly available data and data generated for this study. ............................................................................................................................148Figure 5.12 Average well Tmax in the greater Kaybob region for detailed analysis wells and wells in the public domain. Refer to Figure 5.2 for well identifiers from this study. .149Figure 5.13 Average well HI in the greater Kaybob region for detailed analysis wells and wells in the public domain. Refer to Figure 5.2 for well identifiers from this study. .150Figure 5.14 Average hydrogen index per well versus SSTVD (subsurface true vertical depth) to the top of the Duvernay. ..........................................................................................151xvFigure 5.15 Modified van Krevelen diagram (Tissot and Welte, 1984) for Duvernay source rocks from this study. ..................................................................................................152Figure 5.16 Modified van Krevelen (Tissot and Welte, 1984) of the Kaybob region wells. Two Pembina region wells are included for comparison (PWT 10-17-45 and TLM 12-32-41). ...........................................................................................................153Figure 5.17 Reservoir pore pressure gradient for wells which have tested the Duvernay section. Higher pore pressures positively enhance reservoir properties including increased storage and delivery. ....................................................................................154Figure 5.18 Present day temperature for the Duvernay from shut-in pressure test data. Reservoir temperature correlates poorly with depth for this region. ...........................155Figure 5.19 Reservoir pressure versus depth for wells which have tested the Duvernay. In general pressure correlates with present day burial depth, however notable exceptions can increase reservoir storage and delivery locally (see Figure 5.17). The dashed line is the normal hydrostatic gradient of 9.79 kPa/m (0.433 psi/ft), the black line is linear line of best fit for the wells. N is the number of samples. .......156Figure 5.20 Index map for cross sections constructed for this study. ............................................157Figure 5.21 Cross section A-A’ showing relationship of Leduc reef embayment on Duvernay thickness. GR = gamma radiation, DRES = deep resistivity. Interfingering mudstones with Leduc reef is for representation only. Dashed lines are approximate lithofacies subdivisions based on log signature. ....................................158Figure 5.22 Cross section B-B’ showing north-south stratal relationships. The middle carbonate unit thins to the south while the Duvernay mudstones thicken toward the Leduc reef buildups. Dashed lines are approximate lithofacies subdivisions based on log signature. ................................................................................................159Figure 5.23 Cross section C-C’ illustrating east to west stratal relationships. Organic-rich mudstones of the upper Duvernay thin markedly toward basin center away from Leduc reef influence. The middle carbonate unit progressively constitutes a larger portion of the section toward the basin center. Dashed lines are approximate lithofacies subdivisions based on log signature. .........................................................160Figure 5.24 Majeau Lake isopach map in the WSB. Reefs are zero contoured. CI = contour interval. ........................................................................................................................161Figure 5.25 Majeau Lake isopach map in the Kaybob region. Reefs are zero contoured. CI = contour interval. Color scale maximum is 20 m in order to highlight variability within the Kaybob region. ...........................................................................................162Figure 5.26 Bulk mineralogy normalized to total quartz and feldspars, total clays, and total carbonate as quantified by XRD. Feldspars include Na-feldspars (albite) and K-feldspars (orthoclase). Total clays include muscovite/illite, chlorite (clinochlore), and kaolinite. Total carbonate includes calcite, dolomite and Fe-xvidolomite (ankerite). N is the number of samples. .......................................................163Figure 5.27 Lower Duvernay isopach map in the WSB. Reefs are zero contoured. CI = contour interval. ...........................................................................................................164Figure 5.28 Lower Duvernay isopach map in the Kaybob region. Reefs are zero contoured. CI = contour interval. ..................................................................................................165Figure 5.29 Middle carbonate isopach map in the WSB. Reefs are zero contoured. CI = contour interval. ...........................................................................................................166Figure 5.30 Middle carbonate unit isopach map in the Kaybob region. Reefs are zero contoured. CI = contour interval. ................................................................................167Figure 5.31 Examples of upper Duvernay reservoir lithofacies A. Sample depths and properties are given in Appendix A and B. Variation in color contrast between samples is largely due to core storage and handling effects and may not be indicative of lithology or reservoir properties (i.e. TOC content). Yellow line in G is approximate division of reservoir lithofacies A and B. ...........................................168Figure 5.32 Examples of upper Duvernay reservoir lithofacies B. Sample depths and properties are given in Appendix A and B. Variation in color contrast between samples is largely due to core storage and handling effects and may not be indicative of lithology or reservoir properties (i.e. TOC content). ..............................169Figure 5.33 Upper Duvernay isopach map in the WSB. Reefs are zero contoured. Refer to structure map for well control locations (Figure 5.4). CI = contour interval. ............170Figure 5.34 Upper Duvernay isopach in the Kaybob region. Reefs are zero contoured. CI = contour interval. Color scale maximum is 70 m in order to highlight variability in the 15-70 m range, which dominantly characterizes the upper Duvernay thickness. .171Figure 5.35 Total reservoir thickness (lower and upper Duvernay) in the Kaybob region. Distribution is similar to upper Duvernay due to low total thickness of lower Duvernay. However, the lower Duvernay does constitute a significant proportion of the total reservoir in some areas (e.g. 45 m contour in Township 63, Range 16). ..172Figure 5.36 (A) Quartz versus TOC for wells with detailed analysis. The correlation is moderate at low concentrations of quartz and TOC. At high concentrations, there is no correlation. (B) Poor correlation of quartz versus TOC for immature wells only. .............................................................................................................................173Figure 5.37 Total carbonate versus TOC content for wells with detailed analysis. .......................174Figure 5.38 Total porosity to helium by well compared with quartz (A), calcite (B), illite (C) and TOC (D). ...............................................................................................................175Figure 5.39 (A) Total organic carbon versus total porosity colored by average well Tmax. (B) Total organic carbon versus quartz content colored by porosity. ................................176Figure 5.40 Controls and associations on reservoir thickness within the WSB and the greater Kaybob region. ............................................................................................................177xviiFigure 5.41 Gamma-ray log and TOC for ATH 1-24-61. Reef-derived breccia debris flows (1, 2, 4) occur interbedded with typical organic-rich mudstones. Debris flows fine upwards (1, 2) and are organic lean with low porosity. ...............................................178Figure 5.42 Horizontal-gradient gravity map with reefs and reservoir pressure contours (white) superimposed. A zone of high gravity anomaly (hot colors), which may be related to basement faulting, is aligned along a spatially similar trend in reservoir pressure. Refer to Figure 5.17 for reservoir pressure contour values. Gravity map adapted from Lyatsky et al. (2005). ........................................................179Figure 6.1 General framework of an artificial neural network model. A variety of logs can be included in the input layer (denoted by (…)). .............................................................194Figure 6.2 Wireline logs and ANN models (black curves) compared with laboratory data (blue circles) for the TET 10-18-64 well. ....................................................................195Figure 6.3 Wireline logs and ANN models (black curves) compared with laboratory data (blue circles) for the ATH 13-18-64 well. ...................................................................196Figure 6.4 Wireline logs and ANN models (black curves) compared with laboratory data (blue circles) for the HSK 5-11-60 well. .....................................................................197Figure 6.5 Wireline logs and ANN models (black curves) compared with laboratory data (blue circles) for the HSK 8-25-60 well. The middle carbonate unit and Majeau Lake are not present within the section. ......................................................................198Figure 6.6 Wireline logs and ANN models (black curves) compared with laboratory data (blue circles) for the ATH 4-2-62 well. The ANN models are uncalculated where the RHOB and PEF logs are missing (e.g. top and base of section in this example). .199Figure 6.7 Wireline logs and laboratory-data-trained ANN models for a well without laboratory data within the study area. The example presented here represents the ultimate result of training the ANN models on wells with laboratory data and then applying the ANN models to a wider distribution of wells in the basin. .............200Figure 6.8 (A) Laboratory TOC versus ANN TOC shows a better correlation than TOC derived from a simple linear correlation with RHOB (B). Red lines are linear lines of best fit, black line is 1 to 1 ratio, and R is the correlation coefficient. ...........201Figure 6.9 Average TOC content for the upper and lower Duvernay. ...........................................202Figure 6.10 (A) Quartz content determined by XRD versus ANN model quartz content. (B) Carbonate content determined by XRD versus ANN model carbonate content. (C) Unconfined porosity to helium versus ANN model porosity. Red lines are linear line of best fit, black line is 1 to 1 ratio, and R is the correlation coefficient. ...203Figure 6.11 Average quartz content for the upper and lower Duvernay. Reefs are zeroed contoured; white is < 10 % quartz content or areas of no data. ..................................204Figure 6.12 Average carbonate content for the upper and lower Duvernay. Reefs are zeroed contoured; white is < 10 % carbonate content or areas of no data. .............................205xviiiFigure 6.13 Average porosity for the upper and lower Duvernay. Reefs are zeroed contoured; white is < 1 % porosity or areas of no data. ................................................................206Figure 6.14 ANN models and thickness of rocks (shaded) which meet the criteria for the reservoir quality models defined in Table 6.2. The net thickness of rocks which meet the criteria for each model is then mapped over the study area..........................208Figure 6.15 Total thickness of Duvernay mudrocks which satisfy the criteria for reservoir quality model #1 (Table 6.2). .....................................................................................209Figure 6.16 Total thickness of Duvernay mudrocks which satisfy the criteria for reservoir model # 2 (Table 6.2). .................................................................................................210Figure 6.17 Total thickness of Duvernay mudrocks which satisfy the criteria for reservoir model #3 (Table 6.2). ..................................................................................................211Figure C.1 Mantis-Perm schematic. ...............................................................................................258Figure C.2 Perm-In-A-Box schematic. ..........................................................................................259Figure C.3 LiquiPerm X7.9 Archangel schematic. ........................................................................260Figure C.4 Mini-Pycno schematic. ................................................................................................261Figure C.5 Load frame technical drawing. ....................................................................................262Figure C.6 The dashboard for the QuickPerm permeability data collection software. ..................263Figure C.7 Configuration page for the QuickPerm permeability data collection software. ..........264Figure C.8 Pressure test page for the QuickPerm permeability data collection software. .............265xixlIST oF ABBREVIATIonS, SYMBolS AnD ACRonYMSMathematical variablesφ = sample porosity, L3/L3, %φo = initial, unstressed sample porosity at ambient pressure, L3/L3, %φP = sample porosity at pressure P, L3/L3, %β = compressibility, Lt2/m, psi-1ρbc = sample bulk density corrected for closure, m/L3, g/cm3ρg = sample grain density, m/L3, g/cm3ρHg = bulk density of Hg at analysis temperature, m/L3, g/cm3Cb = bulk volume compressibility, Lt2/m, psi-1Cp = pore volume compressibility, Lt2/m, psi-1Cpfit = pore volume compressibility calculated from the fitted power-law function, Lt2/m, psi-1Cpo = intercept from fitted power-law function, Lt2/m, psi-1m = slope or exponent from fitted power law function, (Lt2/m)2, psi-2P = intrusion pressure or confining pressure, m/Lt2, psi [MPa]Po = ambient pressure, m/Lt2, psi [MPa]Rφ = reduction in porosity from unstressed to stressed state, L3/L3, %V = sample volume, L3, mLVo = initial sample bulk volume at ambient pressure, L3, mLVb = sample bulk volume, L3, mLVbc = closure corrected sample bulk volume from MIP, L3, mLVclosure = closure volume, L3, mLVi = total volume of mercury intruded including closure, L3, mLVic = closure corrected intrusion volume, L3, mLVicc = closure and compressibility corrected intrusion volume, L3, mLVp = remaining sample air-filled pore volume, L3, mLVpc = closure corrected sample pore volume, L3, mLVpen = empty penetrometer volume, L3, mLVpfit = volume intruded due to compression, L3, mLxxWa = assembly weight of MIP penetrometer filled with Hg and sample, m, gWp = weight of empty penetrometer, m, gWs = dried sample weight before MIP analysis, m, gAcronymsANN = Artificial neural networkBET = Brunauer-Emmett-TellerBIB = Broad ion beamBJH = Barrett-Joyner-HalendaBSEM = Back-scattered scanning electron microscopyCI = Contour intervalCO2 = Carbon dioxideDFT = Density functional theoryD-R = Dubinin-RadushkevichDRES = Deep resistivityDT = P-wave interval transit timeFE-SEM = Field-emission scanning electron microscopyFIB = Focused ion beamGEPP = Gas expansion porosity and permeabilityGR = Total gamma radiationHe = HeliumHg = MercuryHI = Hydrogen indexILD = Deep inductionILM = Medium inductionLPGS = Low pressure gas sorptionMD = Measured depthMIP = Mercury intrusion porosimetryN = Number of samplesN2 = NitrogenxxiNCS = Net confining stressOM = Organic matterP = LPGS vapor pressureP0 = LPGS saturation vapor pressurePDP = Pulse-decay permeabilityPEF = Photoelectric factorR = Pearson product-momentum correlation coefficientR2 = Coefficient of determinationRES = ResistivityRHOB = Bulk densityRHOS = Skeletal densityRIR = Reference intensity ratioSEM = Scanning electron microscopySFL = Laterolog resistivitySSTVD = Subsurface true vertical depthTEM = Transmission electron microscopyTOC = Total organic carbonWCSB = Western Canada Sedimentary BasinWSB = West Shale BasinXRD = X-Ray diffractionxxiiACKnoWlEDGMEnTSThe journey through this project over the past years would not have been successful without the constant support, guidance and understanding from numerous individuals. Firstly, I would like to thank my advisor Dr. R. Marc Bustin, who gave me the opportunity and assistance to pursue my interests to levels formerly unimaginable. It was a privilege to work in the laboratory he has constructed for (ab)use by his students to pursue any research avenue. I am grateful to the companies that provided real-world applicability, core samples and monetary assistance, including Shell Canada Ltd., Trilogy Energy Corp., Talisman Energy Inc., Athabasca Oil Corp., Husky Energy Inc., Encana Corp. and Penn West Petroleum Ltd. I would also like to acknowledge monetary support from Geoscience BC for the study. I am appreciative of the array of technical staff who provided assistance throughout the research project. Specifically, thank you to Dave Jones, who showed calm in a sea of seemingly never-ending questions, badgering, hard drive failures and errant code (of my own design). Also to Joern Unger, whose technical expertise at machining custom-designed equipment often enabled potentially catastrophic failures to succeed uninhibited. The willingness by all to drop your work and assist immediately did not go unnoticed. I owe the most sincere thank you to Kristal Li, for years of patience, support and attention. Through all stages of the project you continually re-defined the meaning of “to help”. In the process, I acquired a good friend. I also thank Eric Letham, for numerous brainstorming sessions, enduring the burdens of my rough drafts and the daily, financially-taxing Americano runs. I would have likely finished sooner in absence of you both, but the lab would have been less fun. I am further humbled by the countless friendships forged at UBC along the way. Specifically to Jacquie Colborne, for the perpetual energy and countless adventures. You bring joy to all who know you. Finally, I thank my parents, Steve and Cindy Munson, and my brother, John. Unknown to me, you instilled a relentless drive to succeed and a desire for perfection. Thank you for the 26 years (and counting) of support.xxiiiTo my family1Chapter 1: Introduction1.1 InTRoDUCToRY STATEMEnTSThe increase in exploration for, and production from, mudrock (“shale”) reservoirs has outpaced our understanding of the controls on the distribution and production of hydrocarbons. Recent global interest in fine-grained rocks, such as mudrocks, as hydrocarbon reservoirs have shown fine-grained lithologies to be a significant petroleum resource capable of economic production. Since 2007, natural gas production from shale reservoirs in the United States increased over 700 %, from 1,293 billion-cubic-feet per year (BCF/yr) to 11,415 BCF/yr in 2013 (EIA, 2015). Production trends from Canadian reservoirs follow similar, albeit slower, trends as those in the United States. Fine-grained reservoirs in Western Canada produced approximately 4.7 BCF in 2012, approximately 15 % of total Canadian natural gas production (Chong and Simikian, 2014). In response to the heightened industry interest in mudrocks as hydrocarbon reservoirs, numerous studies have been published examining the geologic controls on hydrocarbon generation potential (organic-matter richness), hydrocarbon storage potential (total porosity and surface area), and hydrocarbon production potential (matrix permeability) of mudrock reservoirs, among others. Lithologies which may act as mudrock reservoirs are diverse (e.g. Chalmers et al., 2012b) and the evaluation of mudrock reservoir characteristics requires a multi-disciplinary approach tailored to a particular reservoir. As such, equally numerous (and sometimes contradictory) conclusions have been reached within the mudrock reservoir characterization literature. In some cases, misleading data and hence conclusions have been reported due to methodologies with inherent analytical errors when applied to characterize the properties of mudrocks. Conventional methods cannot investigate all aspects of the mudrock pore structure since mudrock pore sizes approach or exist beyond the analytical limits of many techniques. Conventional methodologies, developed for compositionally and mechanically distinct reservoir rocks, may have varying applicability for characterizing mudrocks. For example, mercury intrusion porosimetry (MIP) has been used to study the pore throat distributions of mudrocks (e.g. Chalmers and Bustin, 2008a, 2008b; Chalmers et al., 2012a; Kuila and Prasad, 2013; Mastalerz et al., 2013). Some common analytical errors associated with MIP and mudrocks have been identified (e.g. Comisky et al., 2011) but corrections to analytical errors are not applied en masse within geological studies. In 2addition to identifying analytical errors, novel workflows which generate additional rock properties, such as mudrock pore compressibility from MIP, present avenues of research for characterizing mudrocks which are not utilized on conventional reservoir rocks. Other methodologies, such as low-pressure gas sorption (LPGS), have been utilized without a thorough review of potential pitfalls and best practices which may lead to discrepancies between studies. Therefore, it is necessary to establish consistent, error-limited results by highlighting common experimental inaccuracies for methodologies applied to characterize fine-grained reservoir rocks. Within mudrock reservoir characterization studies, there is a need to analyze many samples due to the heterogeneous nature of mudrocks which is difficult to appreciate by visual inspection alone. Determining the lateral and vertical variations in mineralogy is a cornerstone of reservoir characterization studies and as such requires accurate, high-throughput methods to effectively identify zones of most prospective reservoir. Numerous reservoir properties can be correlated to bulk mineralogy, such as porosity (Bustin et al., 2008; Chalmers and Bustin, 2012a; Chalmers et al., 2012b), brittleness (Jarvie et al., 2007), total organic carbon (TOC) content (Ross and Bustin, 2007, 2008) and permeability (Bustin et al., 2008). X-Ray diffraction (XRD) is most frequently employed in reservoir characterization studies (and the wider geological literature in general) to determine mineralogy. Conventional XRD sample preparation methods considered to be “fully quantitative” are time consuming, with analysis time on the order of days. Time constraints may limit the number of analyses performed and therefore the level of characterization detail, which in turn may not capture the sedimentary variability of the reservoir. Accurately determining the mineralogy can be accomplished by more efficient methodologies by exploiting the fine grain size inherent to mudrocks, which enables a greater characterization detail to be achieved. Mudrock reservoir characterization studies have ranged in scope from pore-scale to basin-scale. Pore-scale studies have focused on determining the textural occurrence of porosity (mineral- or organic matter-hosted), the evolution of porosity with thermal maturity, the contribution of various pore types to the total pore volume, the pore size distribution and the control of various pore types on matrix permeability (Loucks et al., 2009; Ross and Bustin, 2009; Milner et al., 2010; Curtis et al., 2011b; Slatt and O’Brien, 2011; Chalmers et al., 2012a; Curtis et al., 2012a, 2012b; Loucks et al., 2012; Clarkson et al., 2013; Kuila and Prasad, 2013; Mastalerz et al., 2013; Milliken et al., 2013; 3Reed et al., 2013; Kuila et al., 2014b; Klaver et al., 2015; Lu et al., 2015). Despite the multitude of studies conducted to investigate the pore system of organic-rich mudrocks, numerous questions remain. The development of mudrock pore structure with thermal diagenesis (maturity) has been investigated by a number of studies and a variety of conclusions have been reported, often because of the diverse methodologies employed. Mudrock reservoirs commonly span multiple hydrocarbon boundaries and the geologic controls on porosity distribution vary between maturity windows. The consequent impact of mudrock pore structure evolution with maturity on matrix permeability remains unknown. Basin-scale studies have focused on determining the areal and stratigraphic occurrence of net favorable reservoir by examining the regional distribution, thickness trends, reservoir hazards, organic matter content and mineralogical properties of the target reservoir. Combining core-measured properties with wireline log suites is an integral part of developing a reservoir characterization model, yet is infrequent within the literature. Methods have been developed to predict core properties from wireline logs, such as organic matter content, which are effective for general characterization purposes (e.g. Schmoker, 1981; Schmoker and Hester, 1983; Meyer and Nederlof, 1984; Passey et al., 1990). Better correlations and therefore more accurate predictions can be achieved through artificial neural network (ANN) models (e.g. Rezaee et al., 2007). Studies utilizing ANN models have demonstrated the effectiveness of this method by predicting a variety of properties from wireline logs, including TOC content (Huang and Williamson, 1996; Yang et al., 2004) and porosity and permeability (Rogers et al., 1995; Wong et al., 1995; Helle et al., 2001; Saemi et al., 2007), among others. However, demonstrating the capabilities of ANN models to predict core properties from wireline logs is only the first step in the greater goal of developing a quantitative framework of regional reservoir properties. Applying ANN models on a basin scale presents an opportunity which has wide-ranging implications for exploration and production efforts.1.2 ThESIS oBjECTIVESIn this thesis, currently producing mudrocks and limestones of the Duvernay Formation, Alberta are used to explore a number of topics concerning laboratory analyses of mudrocks and regional reservoir development. Topics and questions addressed within the study include:41.2.1 Laboratory	analyses	and	fine-grained	lithologies• Can new XRD methodologies be developed to improve analysis time and therefore sampling density for stratigraphically heterogeneous mudrocks?• Can new XRD methods be considered “quantitative” when compared to “standard” methodologies?• What effect do experimental errors have on MIP analyses of mudrocks?• To what degree do MIP experimental errors impact conclusions derived from utilizing erroneous MIP data?1.2.2 Thermal maturity, compaction and pore size distribution• How does thermal maturity alter the pore structure of mudrocks?• How does compaction alter the pore structure of mudrocks?• How does the distribution of porosity within mudrocks vary with maturity?1.2.3 Pore structure evolution and permeability• How does the variation in pore structure with thermal maturity and compaction impact matrix permeability?• Is it possible to determine the contribution of various pore types by utilizing multiple methods of permeability analysis?1.2.4 Pore structure wettability and permeability• What is the wettability of various matrix components?• How does the matrix permeability to hydrocarbon liquids vary with maturity and the contribution of various pore types?1.2.5 Regional	mudrock	reservoir	characterization• What are the geologic controls on the distribution of Duvernay mudrocks?• Where are stratigraphic and areal zones of organic-richness?• Which properties are correlated to better quality reservoir rocks?• Can laboratory data be integrated with wireline log suites and applied at basin-scale?• Which regions are most prospective for exploration and production efforts?The Duvernay Formation provides an excellent laboratory to investigate the research questions presented for this study. Industry focus on the Duvernay has enabled financial support to be secured 5along with connections to industry geologists and hence, industry concerns. Research goals could therefore be aimed at problems pertinent to exploration and development efforts. Recent interest in the Duvernay has enabled fresh core to be secured which is necessary for obtaining detailed datasets. The Duvernay spans multiple thermal maturity boundaries and allows for a comparison of pore structure attributes from rocks with similar organic matter type and compositional properties.1.3 ThESIS STRUCTUREThe results of the study are presented as 5 standalone manuscripts which address the questions presented previously. Minor overlap of results occur between studies due to the multiple topics discussed from a single dataset compiled for the thesis. Chapters are often complementary to each other. For example, methods developed in chapter 2 are applied in chapters 3 through 5. The chapters are discussed briefly below, placing each within the context of mudrock reservoir development. Chapter 2 examines the utility of new XRD sample preparation methods to quickly, accurately and precisely determine the mineralogy of mudrocks. Samples from the Duvernay and five other producing mudrock reservoirs from North America demonstrate the applicability of the new method. The accuracy of the new method is highlighted by preparing artificial samples of known phase concentration. Chapter 3 documents experimental errors common to MIP analyses on mudrocks. Samples from the Duvernay are utilized to determine the impact of experimental errors on MIP-derived properties such as porosity and pore size distribution. The impact of new workflows on the presentation of MIP data is investigated. Chapter 4 addresses the evolution of mudrock pore structure with thermal maturity and compaction. Duvernay samples from multiple maturity zones, varying from immature to the dry gas window, are utilized to examine the change in pore structure. The change in pore structure with maturity is then compared with matrix permeability to determine the impact of maturity on matrix permeability. Chapters 5 and 6 evaluate the regional reservoir potential of the Duvernay. Data gathered using new XRD methodologies developed in chapter 2 is combined with wireline log suites to define a regional model of reservoir development. The distributions and associations of favorable reservoir are delineated through the confluence of thickness distributions and favorable reservoir properties. 6Chapter 2: Utilizing	smear	mounts	for	X-ray	diffraction	as	a	fully	quantitative	approach	in	rapidly	characterizing	the	mineralogy	of	shale	gas	reservoirs2.1 InTRoDUCTIonThe emergence of fine grained lithologies as economical reservoirs for hydrocarbons (i.e. shale gas and shale oil) has prompted a revision of methodologies for measuring the geological properties of rocks with structures and particle sizes in the sub-micrometer range. An important part of reservoir characterization is identifying and quantifying mineral phases present. Reservoir properties such as matrix permeability, rock moduli, porosity and texture are dependent on mineralogy. These important reservoir properties have been correlated with mineralogy (e.g. Clarkson et al., 2013; Kuila and Prasad, 2013) and lithologic groups to identify the associations which lead to a more favorable reservoir, such as higher permeability and greater porosity. In fine grained reservoirs where stimulation by hydraulic fracturing is necessary, small changes in mineralogy can have large impacts on overall well performance, as mineralogy along with texture will define the reservoir zone versus a potential fracture barrier. Therefore, a method to quickly and accurately characterize the mineralogy of a reservoir rock is an extremely valuable exploration tool.2.2 XRD	MethoDoLogy	BaCkgRoUnDThere are numerous methodologies for preparing, analyzing and quantifying the mineralogy of rocks by XRD (Bish and Post, 1989). Each methodology can be broken into four parts: (1) sample grinding; (2) sample drying; (3) sample mounting; and (4) diffraction pattern analysis. Artificial mixtures serve as a means to test the accuracy and precision of a methodology (e.g. Hillier, 2000; Środoń et al., 2001; McCarty, 2002; Kleeberg, 2005; Omotoso et al., 2006; Ufer et al., 2008). Certain methodologies have been shown to produce accurate and precise quantitative results (e.g. within ± 3 wt. % of known mixtures, see discussion in Hillier, 2000), while others may be more suited for semi-quantitative or qualitative assessment (Moore and Reynolds, 1997). Sample grinding involves an initial crushing of the whole rock sample, sieving of the sample, and further grinding in order to produce a sufficiently small crystallite-particle size. The recommended average crystallite size for accurate diffraction intensities is 5 – 15 µm or smaller (see discussion in Bish and Reynolds, 1989; Klug and Alexander, 1974), which reflects a compromise between grinding time, crystallite size and repeatability. Diffraction errors increase significantly 7for particles larger than 5 – 15 µm (Bish and Reynolds, 1989). Phases with good cleavage planes and a platy habit (i.e. clays) have a stronger tendency for preferred orientation when the crystallite size is larger than 30 µm compared to a crystallite size that is < 5 µm (Bish and Reynolds, 1989). There is an assortment of equipment available to reduce the average crystallite size, ranging from simple hand grinding in a mortar with pestle to automatic grinding machines such as the McCrone™ micronizing mill. The McCrone™ mill has been used as the primary grinding method in experiments investigating the accuracy of quantitative analysis on artificial mixtures (Hillier, 2000; Środoń et al., 2001; Omotoso et al., 2006; Ufer et al., 2008), within general earth science literature (Ugolini et al., 2008; Day-Stirrat et al., 2008; Jeong et al., 2011) and the industry. However, a McCrone™ mill is an expensive and time consuming method that does not always add accuracy to diffraction pattern analysis of certain rock types. Published grinding times for clay rich rocks are anywhere from five minutes (Środoń et al., 2001; Omotoso et al., 2006) to 12 minutes (Hillier, 2000; Omotoso et al., 2006) compared to two minutes for a hand ground method. To “micronize” and recover the sample from the slurry requires significant processing time, commonly on the order of 24 hours depending on the milling fluid used. Methods for recovering the powder from the ground slurry are numerous. Assuming the sample has been ground in a McCrone™ mill, the powder must be recovered by settling, decanting, filtering, or evaporating the milling fluid. The recovered dried powder must then be disaggregated, either by hand grinding in a mortar, sieving, shaker-milling with balls, or using a vibrating mill (Kleeberg, 2008). Each of these requirements to recover the powder increases processing time, possibility of contamination, and requires specialized equipment and laboratory experience. A method of spray drying the slurry to produce spherical, randomly oriented agglomerates has been experimented with for a number of years (Jonas and Kuykendall, 1966; Hughes and Bohor, 1970; Smith et al., 1979a,b) but did not gain wide spread usage until the innovations by Hillier (1999; 2002). The spray drying technique is faster than evaporation; however, the drying chamber must be heated to 150°C, with lower temperatures resulting in insufficient drying of the slurry before reaching the chamber floor (Hillier, 1999). At these temperatures alteration of minerals may occur. For example, some common sulfates such as gypsum can dehydrate to bassanite (Hillier, 2002). Lower chamber temperatures of 60°C along with fast drying ethanol as the slurry liquid has been 8successfully used to spray-dry specimens with minerals susceptible to alteration at low temperatures (Jeong et al., 2008). However, as with most specialized methodologies, the spray drying technique adds additional cost, time and experience required to operate the laboratory setup. Commercial laboratories commonly employ one of the simpler sample preparation techniques (Kleeberg et al., 2008). The most utilized mounting techniques for generating randomly oriented specimens are dry powder cavity mounts. For powder cavity mounts, the specimen powder is packed into the mount from the back, side or front. These methods provide good results (Bish and Reynolds, 1989), with the side- and top- mount methods used to accurately and precisely quantify artificial mixtures (Hillier, 2000; Środoń et al., 2001; Omotoso et al., 2006; Ufer et al., 2008). However, if not loaded properly, the powder in the cavity mount can be deformed during long analyses times and movement in the sample chamber. If samples are packed too loosely, the sample can slump before or during analysis, especially when automatic sample changers are used and the time between packing and analyzing is long. On the other hand, if samples are packed too densely, preferred orientation of crystallites will increase as they align perpendicular to the direction of packing. This requires experience with the mounting technique and may not be reproducible for different users. Quantitative X-ray diffraction phase analysis is well established within the literature (Klug and Alexander, 1974; Zevin and Kimmel, 1995; Jenkins and Snyder, 1996; Cullity and Stock, 2001; Madsen and Scarlett, 2008). Phase analysis is most commonly done by one of two methods, the reference intensity ratio (RIR) method (Snyder and Bish, 1989) or the Rietveld method (Rietveld, 1967). The RIR method is based on the comparison of selected observed diffraction intensities versus reference intensities from added internal or external reference phases (Snyder and Bish, 1989). The RIR method requires a large collection of reference mineral patterns and calibration may be different from instrument to instrument and therefore may only be applicable to the laboratory in which the standards were developed (Hillier, 2000; Ufer et al., 2008). The Rietveld method uses a whole pattern multiphase calculation of the observed data versus ideal structure models for individual phases. The difference between the calculated pattern and the observed data is minimized by structure refinement. Preferred orientation correction by the March-Dollase method (March, 1932; Dollase, 1986) can be further applied to refine mineral weighting. Modern Rietveld 9quantitative analysis is easily accessible (i.e. commercial software) and can be successfully operated by researchers having a practical knowledge of the theory. The Rietveld method has proven to be accurate and precise in tests on artificial mixtures (Hillier, 2000; Omotoso et al., 2006), while still being accessible to infrequent users.2.3 MoDIFIED SMEAR MoUnT METhoDThe purpose of this study is to investigate the ability of a modified hand ground, smear mount method to quantify mineralogy of fine grained rocks compared to other more laborious methods. The published smear mount method (Theisen and Harward, 1962; Gibbs, 1965; Poppe et al., 2000) has been generally reserved as a method to produce oriented clay mounts to increase the detection limit of clays (Cody and Thompson, 1976) and for illite crystallinity tests (Robinson et al., 1990; Kisch, 1991). By definition, fine grained lithologies such as siltstones and mudrocks contain at least 50% of particles less than 62.5 µm (Folk, 1974). Potter et al. (1980) further subdivide this definition into siltstones (0 – 32 % clay-sized particles), mudstones (33 – 65 % clay-sized particles), and claystones (66 – 100 % clay-sized particles). The recommended XRD crystallite size of 10 µm is within the lower range of fine silt (15.6 µm – 7.8 µm). Since the boundary between clay and silt is 3.9 µm (Wentworth, 1922), mudstones, by definition, inherently have at least 40 – 69 % of constituent grains below the 10 µm recommendation (Figure 2.1). Due to the small crystallite size of fine grained lithologies, there is potential to reduce sample processing time while maintaining accurate quantitative results. The published smear mount method involves centrifuging the sample and smearing a small amount of slurry on a petrographic slide to provide a smooth surface (Theisen and Harward, 1962) or spreading a clay paste on a glass slide in a thin layer with a single stroke of a spatula (Gibbs, 1965). The method as described by Poppe et al. (2000) involves grinding a dried powder in a mortar, placing a small amount of dry sample on a glass slide, wetting the sample with water or acetone, and spreading it over the slide with a glass rod. The modified smear mount method presented here (smear mount method herein) involves grinding the sample powder with ethanol in an agate mortar, transferring the slurry with a rounded micro spatula to evenly cover a round glass slide and allowing the sample to air dry. The modified method expands the capability of the published smear mount method by being able to quantify 10minerals other than clay and analyze particles initially larger than clay-sized. In contrast to a micronized, evaporated cavity mount method, the smear mount method combines the mounting and drying steps reducing the preparation time for a single sample to ~ 5 minutes compared to over 24 hours. This is an obvious advantage when there is a need to analyze many samples in a short time frame. In addition, the method eliminates the requirement for expensive specialized equipment which add required skill and laboratory expenses. Six major shale gas formations (Figure 2.2) were chosen to highlight the precision of the smear mount method as it compares to more widely used and specialized techniques. The formations represent a range of mineralogy with varying proportions of typical sedimentary minerals. Artificial mixtures were also prepared to reflect varying mineralogy of whole rock shale gas formations. The artificial mixtures were analyzed by the smear mount method and widely used techniques to assess the accuracy of all methods and highlight minerals that may be a source of error when quantifying whole rock samples. In addition, diffraction patterns from users with varying sample preparation experience were compared to assess the reproducibility of the smear mount method.2.4 MATERIAlS AnD METhoDoloGY2.4.1 Whole	rock	sample	preparationA representative whole rock core sample was taken from the following shale gas plays: the Barnett Shale, Eagle Ford Shale, Woodford Shale, Haynesville Shale, Duvernay Formation, and Muskwa Formation (Figure 2.2). The bulk samples were reduced in size by hand crushing in a cast iron mortar to pass through a 0.84 mm sieve (20 mesh). Fifteen gram aliquots were then taken from the < 0.84 mm crushed rock and placed in a vial. To assess the effect of initial sample size on quantitative results, additional rock from the Duvernay Formation was passed through a 0.25 mm sieve (60 mesh) and sampled. The aliquots were first analyzed for mineralogy by the hand ground, smear mount method. Approximately 2.0 g of sample were taken from the vial and mixed with 3.0 mL of ethanol in an agate mortar. The mixture was then hand ground for two minutes and a small amount of slurry scooped and transferred to a round 2.54 cm diameter glass slide. The slide was then tapped gently on the side with a micro spatula to distribute the slurry in an even, thin layer and left to air dry. The slurry was visually confirmed to be level and distributed evenly over the slide during smear-11mounting. Four grams were then sampled from the aliquots and mixed with 10.0 ml of ethanol and micronized with agate pellets in a McCrone™ micronizing mill for seven minutes. The slurry was then transferred to a petri dish and allowed to dry by evaporation. The dry powder was scraped into an agate mortar, disaggregated, and back mounted into circular 2.54 cm diameter cavity holders. The samples were then prepared using the spray dry method (Hillier, 1999). Approximately 2.5 g of micronized powder from each formation was mixed with 5 mL of ethanol in a graduated cylinder and spray dried at 60°C. The spherical aggregates were then collected and front mounted into a 2.54 cm circular cavity holder.2.4.2 artificial	sample	preparationArtificial mixtures reflecting the varying proportions of typical sedimentary minerals in fine grained rocks (e.g. quartz, calcite, illite, ankerite, chlorite, albite, orthoclase and pyrite) were prepared from weighed amounts of pure minerals (Table 2.1). The pure minerals came as crystals with the exception of chlorite, all sourced from a private collection. The crystals were first crushed using a cast iron mortar and passed through a 0.25 mm sieve (60 mesh). The pure samples were then analyzed by XRD to assess purity. Samples with minor contamination (e.g. albite exsolution in orthoclase) were allowed if the contaminant mineral was to be included in the artificial mixture and able to be quantified. For example, albite impurity in the orthoclase sample was used because albite was to be included as a phase in the artificial mixture. Four 2.0 g mixtures (Samples E1 – E4, Table 2.1) were prepared from the < 60 mesh pure samples. The weighed mixtures were then ground with 5.0 mL of ethanol in a McCrone™ mill with agate pellets for 7 minutes. The artificial mixtures were not hand ground because the crystallite size is not representative of the crystallite size observed in mudrocks. Due to their platy habit phyllosilicate minerals, especially muscovite, are historically difficult to grind, even when using a McCrone™ mill (Bish and Reynolds, 1989). Therefore, only one sample was prepared with muscovite to demonstrate the limitations of the smear mount technique when the particle size of platy minerals is large. The artificial samples were then analyzed using the smear mount, back mount, and spray-dry methods.122.4.3 XRD	and	quantitative	phase	analysisThe XRD data was collected using normal-focus CoKα radiation on a Bruker® D8 Focus at 35 kV and 40 mA. Sample diffraction patterns were obtained over the range of 3-70° 2θ at a step size of 0.03° and counting for 0.8 s per step while continuously spinning at 60 rpm. At this setting the samples took 33.5 minutes to analyze. The scan settings chosen for the experiment represent those routinely used for quantitative phase analysis. The scan settings are a compromise between data quality and time which reflect the need to analyze many samples on a routine basis. Phases were quantified over 3-70° 2θ using the Rietveld method of full-pattern fitting using Bruker® AXS Topas V3.0 software.2.4.4 Scanning	electron	microscopeWhole rock sample powders were examined using back scattered scanning electron microscopy (BSEM) to investigate initial crystallite and particle sizes, compare the effects of hand grinding and micronizing on particle size reduction and homogenization, and assess preferred orientation of smear mount slides. Powder from Duvernay Formation sample was examined: (1) after passing through the 60 mesh sieve; (2) after hand grinding in ethanol for two minutes; and (3) after micronizing for seven minutes. The prepared smear mount from the Duvernay Formation was also analyzed by BSEM to determine if there was visible alignment or preferred orientation of grains. The artificial sample powder was examined after micronizing to assess the crystallite size. BSEM images were taken using a Philips XL-30 SEM with an accelerating voltage of 15 keV at a working distance of 10 to 11 mm.2.4.5 Particle size analysisDispersed powders were analyzed by laser diffraction to assess the average particle size after hand grinding and after micronizing. Approximately 0.5 grams of sample was mixed with 15.0 mL of water in a beaker and continuously stirred. Using a pipette, approximately 2.0 mL of slurry was sampled from the beaker and analyzed for particle size using a Malvern™ Mastersizer 2000. Thirty seconds of ultrasonic treatment was applied to the sample in 15 second increments during analysis to disaggregate any particle agglomerates formed during mixing.132.5 RESUlTS2.5.1 Whole	rock	analysisThe quantitative results are presented in Table 2.2 and illustrated in Figure 2.3 by graphs for each formation. To compare quantitative results, the smear mount and back mount results were compared to the spray dried results. For a given sample, the absolute difference in weight percent were calculated for each phase, and then the mean of these absolute differences were taken to give the mean absolute difference for the sample. The quantitative results were then compared on a phase by phase basis to highlight phases of higher uncertainty. When the smear mount method is compared to the spray dried method, the smear mount method has a mean absolute difference in calculated weight percent of 0.77 % for all samples, with a mean difference maximum of 1.3 wt. % for the Duvernay sample and a mean difference minimum of 0.2 wt. % for the Barnett sample. At the 95 % confidence level the spray dried to smear mount mean difference is 0.9 wt. % for all phases within the sample suite. When the back mount method is compared to the spray dried method, the back mount method has a mean absolute difference of 0.65 wt. % for all phases within all samples, with a mean difference maximum of 1.0 wt. % for the Duvernay and Haynesville samples and a mean difference minimum of 0.1 wt. % for the Muskwa sample. At the 95 % confidence level the spray dried to back mount mean difference is 0.7 wt. % for all phases within the sample suite. The statistical analyses of absolute errors for individual phases are presented in Table 2.3 and illustrated in Figure 2.4. The mean absolute error is lowest for pyrite at 0.1 wt. % and highest for muscovite at 1.4 wt. %. Muscovite and illite are reported as muscovite due to the variability in layer stacking and lack of a well-defined crystal structure model of illite. Muscovite has the largest range of error at 3.7 wt. % and pyrite has the smallest at 0.3 wt. %. The average range for all phases within the sample suite is 1.6 wt. %. At the 95 % confidence level the mean error for all phases is 1.3 wt. %.2.5.2 artificial	sample	analysisThe quantitative results for the artificial samples are presented in Table 2.4. For the three samples without muscovite, the back mounted sample produced the lowest total mean error of 2.5 wt. %, while the smear and spray dried samples produced similar total mean errors of 3.5 and 3.7 14wt. %, respectively. For the sample containing muscovite, the spray dry and back mount sample produced similar total errors of 10.9 and 8.0 wt. % respectively, while the smear mount sample had a total error of 21.4 wt. % due to a 9.0 wt. % overestimation of the muscovite content because of insufficient particle size reduction during micronizing. 2.5.3 BSEMImages from the crushed, hand ground and micronized Duvernay sample are shown in Figure 2.5. A wide particle size range exists in the < 60 mesh sieved fraction (A1, A2, A3, Figure 2.5). The long dimension of the largest particles measure over 200 µm while the smallest particles can be less than 10 µm. The large particles are observed to be composed of smaller crystallites, commonly less than 10 µm (A3, Figure 2.5). The hand ground sample is more homogeneous and the particle size range is smaller compared to the sieved sample, with the largest observed particles measuring around 50 µm (B1, B2, Figure 2.5). The average observed particle size of the hand ground sample has been reduced to less than 10 µm (B3, Figure 2.5). In the micronized sample, the particle distribution appears very narrow and homogeneous (C1, C2, C3, Figure 2.5). The average observed particle size has been reduced to less than 10 µm, however large particles in excess of 50 µm remain (C3, Figure 2.5).  The smear mount of the Duvernay sample is shown in Figure 2.6. Overall, no preferred orientation is obvious. Rarely, an oriented phyllosilicate grain can be seen (F, Figure 2.6) in the prepared smear mount. However, in the same micrograph, a randomly oriented phyllosilicate grain can be seen adjacent to the preferably oriented grain (F, Figure 2.6).2.5.4 Particle size analysisThe Duvernay, Woodford and Eagle Ford samples were analyzed for particle size statistics. Results of the analysis are presented in Table 2.5. The samples were first analyzed without ultrasonic treatment, after 15 seconds of ultrasonic treatment, and after a final 15 seconds of ultrasonic treatment, totaling 30 seconds of ultrasonic treatment. The median particle size of the hand ground samples is 30.5 µm before ultrasonic treatment and is reduced to 19.9 µm after treatment. The median particle size of the micronized samples is 6.4 µm before ultrasonic treatment and is reduced to 4.6 µm after treatment.152.6 DISCUSSIon2.6.1 Whole	rock	analysisWhile commonly referred to as ‘semi-quantitative’ in the literature, the precision of a hand ground, smear mount method has been validated by a comparison to published methodologies regarded as ‘quantitative’ on whole rock samples. Based on the compilation of many studies investigating methodologies on known compositions, results that fall within ± 3 wt. % absolute of a known composition at the 95% confidence level are considered to be ‘highly accurate’ (see discussion in Hillier, 2000). We have applied the ± 3 wt. % guideline to characterize the precision of the two methods investigated in this study. It is evident from the results that the hand ground, smear mount method is ‘highly precise’ when compared to both the spray dry and the back mount methods in quantifying the mineralogy of fine grained rocks. At the 99.6 % confidence level (3 standard deviations), the total absolute error between the spray dry method and the smear mount method for fine grained reservoir rocks is ± 2.6 wt. %, well within the ± 3 wt. % guideline.  The success of the smear mount method is due in large part to the fact that the whole-rock samples are visually observed to be composed mostly of crystallites less than 10 µm. As observed by BSEM, hand grinding in ethanol for two minutes effectively disaggregates larger particles to their smaller crystallite constituents (Figure 2.5). Even if the large particles are not fully disaggregated, they are still composed of smaller crystallites which do not cause grain size diffraction errors. For fine grained lithologies such as shale gas reservoirs it may not be necessary to micronize a sample to produce a narrow and sufficiently small crystallite distribution for quantitative XRD. The Duvernay sample further illustrates the precision of the smear mount method. The Duvernay sample was passed through a 60 mesh sieve before grinding while all other samples were passed through a 20 mesh sieve. It would be expected that given the smaller starting particle size of the Duvernay sample that the quantitative results would be more precise compared to samples with a larger starting particle size. However, the average error of 0.6 wt. % for the Duvernay sample is greater than the average errors of 0.4, 0.5 and 0.5 wt. % for the Barnett, Haynesville, and Muskwa samples respectively. The average Duvernay phase error is only 0.04 wt. % less than the average error for all the formations investigated. Given these relationships, initial sample size up to 0.84 mm has no obvious impact on the precision of the hand ground, smear mount method.16 The smear mount method was also tested for reproducibility. Three users of varying experience with the method prepared 4 separate mounts of the Duvernay sample each totaling 12 patterns (Figure 2.7). Smear mounts from all users produce indistinguishable diffraction patterns and similar quantitative results. This is important for laboratories such as those in universities where experience level fluctuates and some users only need to occasionally analyze a few samples.2.6.2 artificial	sample	analysisThe smear mount method was further validated by quantifying samples of known mineralogy. In samples without muscovite, the smear mount method had similar accuracy to the spray dry and back mount methods. In the sample containing muscovite the smear mount method had a total error of 21.4 wt. %, compared to 8.0 and 10.9 wt. % error for the back mount and spray dry methods respectively. This error is in part due to preferred orientation of large muscovite crystallites which could not be sufficiently reduced by micronizing (C3, Figure 2.5). The large muscovite particle size and error due to the ineffective micronizing of the artificial sample is not representative of the particle size and thus error found in fine grained rocks. We note that for the sample containing muscovite the total error for each method was higher overall than the samples without muscovite. At the 99.6 % confidence level for samples without muscovite, the individual phase error is ± 1.3 wt. % for the smear mount method. It is evident from the results of the artificial sample analysis that the smear mount method is both accurate and precise.2.7 ConClUSIonS AnD RECoMMEnDATIonSBased on the results of this study, we recommend the method be considered a quantitative approach in characterizing the mineralogy of fine grained lithologies such as shale gas reservoirs. A single smear mount requires 5 - 15 minutes to be fully prepared, compared to over 8 - 24 hours for more laborious methods. Therefore, the smear mount method is significantly faster than all other published methodologies while still maintaining both a high degree of precision when compared to whole rock samples and a high degree of precision and accuracy when compared to artificial mixtures of known composition. In reservoirs where the vertical heterogeneity is visually subtle and often exceeds the sampling density, the ability to analyze more samples faster will add resolution to geologic interpretations.17Figure 2.1 Relation of the 10 µm recommended crystallite size to the particle size of fine grained rocks (Wentworth, 1922).18Figure 2.2 General geographical subcrop locations of the Barnett, Woodford, Haynesville, Eagle Ford, Duvernay and Muskwa shales.19ES1 ES2 ES3 ES4 ES1 ES2 ES3 ES4Phase (g) (g) (g) (g) (%) (%) (%) (%)Quartz 0.6003 1.2001 0.8002 0.3999 30.0210 60.0110 40.0040 19.9930Calcite 0.4001 0.4001 0.6002 1.2000 20.0090 20.0070 30.0055 59.9940Muscovite 0.1998 - - - 9.9920 - - -Chlorite 0.1198 0.0198 0.1200 0.2402 5.9912 0.9901 5.9991 12.0088Pyrite 0.0599 0.0200 0.1001 0.0600 2.9956 1.0001 5.0042 2.9997Albite 0.3201 0.2000 0.3200 0.0200 16.0082 10.0010 15.9976 0.9999Ankerite 0.1597 0.1598 0.0598 0.0801 7.9866 7.9908 2.9896 4.0046Orthoclase 0.1399 - - - 6.9964 - - -Total (g, %) 1.9996 1.9998 2.0003 2.0002 100 100 100 100Table 2.1 Mineral composition of artificial mudrocks (ES1 – ES4).20Spray Smear Back Spray Smear BackFormation Phase Wt. % Wt. % Wt. % Formation Phase Wt. % Wt. % Wt. %Duvernay Quartz 47.1 46.9 45.7 Haynesville Quartz 27.1 27.8 27.7Calcite 13.2 12.8 12.9 Calcite 22.3 22.3 23.6Muscovite 19.0 14.9 18.2 Muscovite 34.4 33.8 31.0Pyrite 3.1 3.0 3.0 Clinochlore 0.6 0.4 0.7Clinochlore 1.4 1.8 1.0 Pyrite 3.6 3.7 3.8Ankerite 4.7 4.4 4.3 Albite 7.1 6.8 7.8Albite 3.1 2.8 2.6 Ankerite 4.9 5.1 5.4Orthoclase 8.4 13.3 12.2Barnett Quartz 64.0 63.7 63.7 Muskwa Quartz 65.3 61.7 65.1Calcite 3.3 3.3 3.2 Muscovite 26.0 28.0 26.0Muscovite 22.6 22.5 24.8 Clinochlore 1.2 2.6 1.2Pyrite 2.8 2.6 2.8 Pyrite 2.9 2.7 3.0Dolomite 1.5 1.8 1.3 Albite 3.4 3.7 3.7Albite 5.9 6.0 4.2 Ankerite 1.1 1.2 1.0Eagle Ford Quartz 21.8 21.7 22.4 Woodford Quartz 38.9 38.5 37.9Calcite 61.7 62.3 62.9 Calcite 5.8 5.9 5.9Muscovite 7.1 8.0 5.2 Muscovite 37.7 36.5 37.8Pyrite 4.2 4.3 4.4 Clinochlore 2.3 3.8 2.8Clinochlore 2.7 2.1 3.1 Pyrite 3.4 3.4 3.5Albite 2.4 1.6 2.0 Albite 8.5 7.9 8.6Ankerite 3.4 4.0 3.6Table 2.2 Calculated Rietveld weight percent for the phases present in each formation using different methods of sample preparation. Spray = Micronized, spray dried front drift preparation; Smear = Hand ground, smear mount preparation; Back = Micronized, back mount preparation.21Figure 2.3 Graphs of the quantitative results from each methodology by formation.22Phase: Quartz Calcite Dolomite Muscovite Chlorite Pyrite AlbiteMaximum error: 1.1 1.5 1.6 4.2 1.4 0.3 2Minimum error: 0.1 0.1 0.0 0.5 0.1 0.0 0.0Mean error: 0.8 0.5 0.6 1.4 0.5 0.1 0.72 St. Dev. 0.9 1.1 1.3 2.8 1.0 0.2 1.5Table 2.3 Absolute weight percent error statistics for each phase.23Figure 2.4 Graph of spray dry weight percent versus smear mount weight percent by individual phases. The line is a 1:1 ratio.24Spray Dry Smear Mount Back MountRietveld Bias Rietveld Bias Rietveld BiasSample Phase Weighted % Wt. % Wt. % Wt. % Wt. % Wt. % Wt. %ES1 Quartz 30.0 32.7 2.7 31.8 1.7 33.2 3.1Calcite 20.0 19.1 1.0 17.9 2.2 20.3 0.3Muscovite 10.0 10.7 0.8 19.0 9.0 9.7 0.3Chlorite 6.0 6.5 0.5 2.1 3.9 4.1 1.9Pyrite 3.0 2.5 0.5 2.1 0.9 2.7 0.3Albite 16.0 17.5 1.5 13.4 2.7 16.6 0.5Ankerite 8.0 6.2 1.8 7.0 1.0 7.5 0.5Orthoclase 7.0 4.8 2.2 7.0 0.0 6.0 1.0Total bias (%) 10.9 21.4 8.0ES2 Quartz 60.0 60.8 0.8 60.4 0.4 60.7 0.7Calcite 20.0 20.2 0.1 20.3 0.3 20.6 0.6Chlorite 1.0 0.9 0.1 0.5 0.4 0.3 0.7Pyrite 1.0 0.7 0.3 0.7 0.3 0.8 0.2Albite 10.0 10.7 0.7 10.4 0.4 10.5 0.5Ankerite 8.0 6.8 1.2 7.6 0.4 7.2 0.8Total bias (%) 3.1 2.3 3.5ES3 Quartz 40.0 41.3 1.3 41.0 1.0 41.0 1.0Calcite 30.0 28.6 1.4 30.7 0.7 30.4 0.4Chlorite 6.0 6.8 0.8 5.8 0.2 6.1 0.1Pyrite 5.0 4.0 1.0 4.0 1.0 4.4 0.6Albite 16.0 16.6 0.6 15.4 0.6 15.4 0.6Ankerite 3.0 2.7 0.3 3.1 0.1 2.8 0.2Total bias (%) 5.5 3.6 2.8ES4 Quartz 20.0 20.8 0.8 21.5 1.5 20.6 0.6Calcite 60.0 60.4 0.5 58.5 1.5 59.7 0.3Chlorite 12.0 12.0 0.0 12.8 0.8 11.8 0.2Pyrite 3.0 2.8 0.2 2.7 0.3 2.8 0.2Albite 1.0 0.9 0.1 0.7 0.3 1.0 0.0Ankerite 4.0 3.1 0.9 3.9 0.1 4.1 0.1Total bias (%) 2.5 4.5 1.3Table 2.4 Quantitative results for the artificial samples using the spray dried, smear mount and back mount preparation methods. Bias is the absolute difference between weighed mineral percent and measured weight percent.2526Figure 2.5 (A1, A2, A3) Crushed Duvernay whole-rock sample passed through 60 mesh. The large particles are aggregates of smaller crystals. Particles can be seen in excess of 250 µm. (B1, B2, B3) Duvernay sample hand ground with agate mortar and pestle in ethanol for two minutes. The larger particles are disaggregated into smaller crystallites, many particles have been reduced to less than 10 µm. Large particles in excess of 30 µm can remain. (C1, C2, C3) Duvernay sample micronized with agate elements in ethanol for seven minutes. Most particles have been reduced to less than 10 µm, some larger crystallites remain in excess of 30 µm. The phyllosilicate minerals (muscovite) are notably difficult to grind.27Figure 2.6 Backscatter SEM images of the smear mounted Duvernay slide that was analyzed by XRD. Overall there is no obvious preferred orientation of clay mineral grains in whole rock samples.A500 µm 200 µmB C100 µm50 µm 50 µm 50 µmD E F28Handground Ultrasonic d(0.1) d(0.5) d(0.8) d(0.9) Micronized Ultrasonic d(0.1) d(0.5) d(0.8) d(0.9)Sample (sec) µm µm µm µm Sample (sec) µm µm µm µmDuvernay 0 2.2 19.5 49.9 76.5 Duvernay 0 1.3 6.0 14.7 22.915 1.8 15.1 40.0 57.4 15 1.1 4.9 10.8 14.730 1.7 14.0 38.6 55.8 30 1.0 4.7 10.4 14.2Woodford 0 3.5 27.8 74.0 109.5 Woodford 0 1.7 8.6 18.4 27.415 2.7 23.8 68.7 105.4 15 1.3 6.2 12.4 16.630 2.4 22.7 65.8 100.6 30 1.3 5.7 11.7 15.7Eagle Ford 0 3.6 44.3 123.6 189.7 Eagle Ford 0 1.1 4.6 13.2 24.315 1.9 24.7 68.8 96.0 15 0.9 3.4 7.9 11.530 1.7 23.0 66.6 93.0 30 0.8 3.3 7.6 10.9Table 2.5 Particle size statistics for three whole rock samples prepared by hand grinding and micronizing. Samples were analyzed before and after ultrasonic treatment. d(0.1) = Diameter in which 10% of particles are less than.2920 40 60Counts10,00020,00030,0002Θ CoKα020 30Counts2Θ CoKαFigure 2.7 Twelve total mounts from 3 users illustrate the reproducibility of the smear mount method.30Chapter 3: Correcting	for	compressibility	and	closure	in	mercury	intrusion	porosimetry:	Calculations	for	determining	pore	volume	compressibility	in	fine	grained	reservoir	rocks3.1 InTRoDUCTIonThe nanoscale nature of porosity in shale gas and shale oil rocks requires the use of multiple techniques, both routine and non-routine, to adequately characterize the pore system (Chalmers et al., 2012a; Clarkson et al., 2012, 2013; Labani et al., 2013; Mastalerz et al., 2013; Wang et al., 2014). Since Washburn’s (1921) pioneering research on capillary forces and the development of the high pressure porosimeter (Ritter and Drake, 1949), mercury intrusion porosimetry (MIP) has become a standard procedure for measuring the properties of porous materials. A comprehensive historical review of the development and applicability of MIP is given by León (1998). Within the energy industry, MIP has been used in a multitude of applications, such as studies on the pore network of conventional reservoir rocks (Swanson, 1981; Katz and Thompson, 1986; Pittman, 1992), the sealing efficiency of caprocks for hydrocarbon trapping (Schlömer and Krooss, 1997; Dewhurst et al., 2002; Heath et al., 2011), carbon dioxide sequestration (Okamoto et al., 2005; Wollenweber et al., 2010; Heath et al., 2012), coals (Debelak and Schrodt, 1979; Spitzer, 1981; Friesen and Ogunsola, 1995; Clarkson and Bustin, 1999), and more recently on tight rocks as they relate to shale gas and shale oil reservoirs (Chalmers and Bustin, 2008a, 2008b; Chalmers et al., 2012a; Clarkson et al., 2013; Kuila and Prasad, 2013; Mastalerz et al., 2013).MIP has been well documented to produce misleading raw data due to sample size effects and the high intrusion pressures (15,000 – 60,000 psi, 103.4 – 413.7 MPa) required to intrude the smallest pore throats (Rootare and Prenzlow, 1967; Wagner, 1983; León, 1998; Webb, 2001; Bailey, 2011; Comisky et al., 2011). Despite the widespread use of MIP, data reduction procedures to remove experimental errors are not routinely applied. These errors are magnified when attempting to measure the pore size distribution in ductile materials with average pore diameters approaching the analytical limit of modern porosimeters (tens of nanometers), such as mudstones and tight siltstones. Without corrections to experimental errors, attempts to calculate MIP derived parameters such as total pore volume, bulk density, skeletal density, permeability, Hg saturation, threshold pressure, pore size distributions, and mean pore diameter can be distorted.31Plugs to crushed samples of varying sizes are used in MIP analyses. In some analyses, an excess volume of mercury is intruded into the sample cell at very low pressures (1.5 – 30 psi, Lee and Maskell, 1973; Comisky et al., 2011). This volume at low pressures has been referred to by a variety of terms (e.g. intergranular intrusion, Lee and Maskell, 1973; interparticle voids, Palmer and Rowe, 1974; interparticle volume, Colombo and Carli, 1981; particle density, Mukaida, 1981; interparticle porosity, León, 1998; Webb, 2001; envelope volume, Webb, 2001; conformance, Comisky et al., 2011; closure, Shafer and Neasham, 2000; Bailey, 2011). In the case of crushed samples, this excess volume is interpreted to be a particle packing effect as well as surface roughness. When the sample cell is filled with mercury at low pressures (1.5 psi, 0.01 MPa), crushed sample particle may create small interparticle voids inaccessible to mercury at low pressure. When the pressure is increased slightly, this void space is filled and a volume of intrusion is recorded. For plugs, an excess volume at low pressure is interpreted to be primarily related to surface roughness. The volume of mercury which fills interparticle voids at low pressure is referred to herein as “closure”. Closure has been recognized as a significant source of error in calculating bulk density (i.e. particle density, Mukaida, 1981; Webb, 1993) and total mercury saturations (Comisky et al., 2011). In this study, closure of crushed mudstone samples can account for over 40 % of the total intrusion volume, thereby significantly skewing parameters such as total porosity, average pore diameter, and pore distribution. Despite the early awareness of the closure effect, closure corrected MIP data has only recently appeared in published geological literature (Comisky et al., 2011; Heath et al., 2011; Clarkson et al., 2012) and uncorrected data is still widely published (e.g. Labani et al., 2013; Wüst et al., 2013; Wang et al., 2014). In addition, data containing unreported corrections have also been encountered in the literature, which likely arise from analyses done at commercial laboratories where corrections are applied before data is delivered to the client (Bailey 2011; Heath et al., 2012). The problem is likely compounded by the lack of familiarity with the common experimental errors associated with MIP analyses. Making correlations between published studies may be difficult or impossible when MIP data is variably corrected.At higher pressures, apparent intrusion of mercury due to sample compression has been recognized by many workers (i.e. Rootare and Prenzlow, 1967; Lee and Maskell, 1973; Colombo and Carli, 1981; Wagner, 1983; Suuberg et al., 1995; León, 1998; Webb, 2001). Initially, sample 32compression was noted by powder technologists as a post-intrusion stress, which led to the collapse of the solid matrix or to pores inaccessible to mercury (Rootare and Prenzlow, 1967; Palmer and Rowe, 1974; Colombo and Carli, 1981) and has since been used in studies on the compressibility of clays (Penumadu and Dean, 2000) and coals (Toda and Toyoda, 1972; Suuberg et al., 1995; Li et al., 1999). However, for tight rocks where initial pore intrusion pressures exceed 1,000 psi (6.9 MPa) (commonly up to 8,000 psi, e.g. Comisky et al., 2011), compression of the sample volume prior to pore intrusion can be significant. Bailey (2009, 2011) provides a methodology to identify and fully remove volume intruded due to sample compression before intrusion, which was later applied by Comisky et al. (2011). However, Bailey (2011) points out that intrusion may occur concurrently with sample compression. Therefore, correcting for compression is not as simple as removing the volume assumed to be due to compression from calculations.In this paper we present novel techniques to analyze MIP data that can generate consistent, error-limited results as well as additional rock property data apart from the usual MIP derived parameters, such as pore size distribution and porosity. In this paper, we confirm the closure correction for tight rocks proposed by Bailey (2009) and investigated by Comisky et al. (2011). We expand the workflow based on ideas of Bailey (2011) to: (1) correct incremental intrusion curves for pore volume lost due to compression and (2) account for incremental intrusion that occurs concurrent with compression. In addition, we develop a new calculation for stressed porosity using MIP compression data to correct unconfined porosity measurements for any net confining stress desired. We also examine the effect of sample size on closure and MIP derived compression, investigate commercial laboratory closure corrections, and present examples of the effects closure and compression have on calculating other commonly reported MIP parameters such as “Swanson” permeability. In a companion paper, we analyze a large suite of samples from the shale gas and oil producing Duvernay Formation of Alberta using the calculations presented herein.3.2 eXPeRIMentaL	MethoDoLogy3.2.1 Mercury injection porosimetryTo assess the effect of sample size and sorting on closure and compression, the following samples were prepared: (1) a whole rock piece measuring approximately 1.5 cm by 1.5 cm, (2) two whole rock pieces each approximately 1.0 cm by 1.0 cm, (3) sample crushed and sieved to pass between 334 and 8 mesh (4750 μm to 2000 μm) and (4) sample crushed and sieved to pass between 4 and 20 mesh (4750 μm to 853 μm). The samples were taken as plugs from a whole core piece, crushed using a mortar and pestle and dried at 110° C for 24 hours.Mercury intrusion porosimetry was performed using a Micromeritics® Autopore IV capable of producing 60,000 psi (413.7 MPa) of injection pressure corresponding to a lower limit pore throat diameter of approximately 3 nm. A review of standard MIP operating procedure is given in Comisky et al. (2011).3.2.2 Closure correctionThe closure corrected bulk volume, Vbc, is calculated from the measured weight of the penetrometer after being fully filled with mercury at low pressure, Wa, and using:𝑉𝑉𝑏𝑏𝑏𝑏 = 𝑉𝑉𝑝𝑝𝑝𝑝𝑝𝑝 −(𝑊𝑊𝑎𝑎− 𝑊𝑊𝑠𝑠−𝑊𝑊𝑝𝑝)𝜌𝜌𝐻𝐻𝐻𝐻− 𝑉𝑉𝑏𝑏𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑝𝑝        ……… (1)where Ws is the weight of the sample, Wp is the weight of the empty penetrometer, ρHg is the bulk density of mercury at the analysis temperature, Vpen is the volume of the empty penetrometer, and Vclosure is the closure as determined by the incremental or cumulative injection plot. As an example, Figure 3.1A shows uncorrected (blue) and closure corrected (red) incremental intrusion curves for a typical mudstone of the Duvernay Fm. The closure peak is identified by the sharp increase in mercury volume at low pressure. The total volume of this peak is the closure. The closure corrected bulk density, ρbc, is then calculated by:𝜌𝜌𝑏𝑏𝑏𝑏 =𝑊𝑊𝑠𝑠𝑉𝑉𝑏𝑏𝑏𝑏    ……… (2)Porosity, φ, (Eq. 3) and grain (skeletal) density, ρg, (Eq. 4) are determined once the analysis has completed the high pressure intrusion from:𝜑𝜑 =𝑉𝑉𝑝𝑝𝑝𝑝𝑉𝑉𝑏𝑏𝑝𝑝    ……… (3)𝜌𝜌𝑔𝑔 =𝑊𝑊𝑠𝑠𝑉𝑉𝑏𝑏𝑏𝑏−𝑉𝑉𝑝𝑝𝑏𝑏     ……… (4)where Vpc is the closure corrected total pore volume given by:𝑉𝑉𝑝𝑝𝑝𝑝 = 𝑉𝑉𝑖𝑖 − 𝑉𝑉𝑝𝑝𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐      ……… (5)and Vi is the total volume of mercury intruded. Since the closure corrected bulk volume of the sample (Vbc) is used in the skeletal density calculation (Eq. 4) it is also necessary to subtract the closure from the total volume of mercury intruded (Eq. 5). This calculation is often incorrectly noted (e.g. Eq. 7 in Comisky et al., 2011).343.2.3 Compression correctionCompressibility during MIP analysis was originally noticed as a post-intrusion stress (Rootare and Prenzlow, 1967). After filling of the pore network and interparticle voids with mercury, cumulative intrusion would “plateau” and further small volume changes at high pressures were attributed to the compressibility of mercury or the sample particles (Rootare and Prenzlow, 1967). For many tight rocks such as mudstones, intrusion following a power-law function prior to true intrusion is evident on cumulative intrusion plots. Figure 3.2 presents an example of a typical semi-log cumulative intrusion plot of a mudstone sample from this study. It would not be expected that heterogeneous, natural samples would have a predictable uptake of mercury into a pore structure comprised of varying pore throat diameters and volumes. Therefore, this region of apparent intrusion has been attributed to pore volume compression (Bailey, 2011; Comisky et al., 2011). The bulk volume compressibility of a solid under hydrostatic compression, Cb, is given as (Colombo and Carli, 1981; Zimmerman et al., 1986):𝐶𝐶𝑏𝑏 = −1𝑉𝑉𝑏𝑏𝑖𝑖 (𝑑𝑑𝑉𝑉𝑏𝑏𝑑𝑑𝑑𝑑)     ……… (6)where Vb is the bulk volume, P is the confining (intrusion) pressure, and the superscript i represents the initial, unstressed volume. If the grain volume is assumed to be incompressible, then the compressibility can be considered a function of the pore volume compressibility as (Bailey 2011):𝐶𝐶𝑝𝑝 =1𝑉𝑉𝑝𝑝(𝑑𝑑𝑉𝑉𝑝𝑝𝑑𝑑𝑑𝑑)     ……… (7)where Cp is the pore volume compressibility and Vp is the remaining air-filled pore volume. The negative sign in Eq. 6 is an indicator of compression direction (i.e. if a sample becomes volumetrically smaller with an increase in pressure, the compressibility is negative). This convention is removed from Eq. 7 since negatives cannot be plotted on a log scale, as is required in this workflow. When plotted in log space versus P (Figure 3.3), a power law function exists over the same interval as the semi-log cumulative intrusion plot and is expressed as:𝐶𝐶𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝 = 𝐶𝐶𝑝𝑝𝑝𝑝𝑃𝑃𝑚𝑚     ……… (8)where Cpfit is the compressibility from the fitted function, Cpo is the intercept, and m is the slope (Bailey, 2011; Comisky et al., 2011). By combining Eqs. 7 and 8:𝑑𝑑𝑉𝑉𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝 = 𝐶𝐶𝑝𝑝𝑝𝑝𝑃𝑃𝑚𝑚(𝑑𝑑𝑃𝑃)𝑉𝑉𝑝𝑝      ……… (9)where dVpfit is the incremental volume intruded due to compression. The closure and compressibility 35corrected incremental intrusion, dVicc, can then be calculated from:𝑑𝑑𝑉𝑉𝑖𝑖𝑖𝑖𝑖𝑖 = 𝑑𝑑𝑉𝑉𝑖𝑖𝑖𝑖 − 𝑑𝑑𝑉𝑉𝑝𝑝𝑝𝑝𝑖𝑖𝑝𝑝      ……… (10)where dVic is the closure corrected incremental intrusion.An example of the methodology to correct incremental intrusion for closure and compression on a typical mudstone is presented in Figure 3.1 and Figure 3.3. First, the raw incremental intrusion (Figure 3.1A, blue) is corrected by removing the apparent volume intruded due to closure (Figure 3.1B, red). The compressibility is then calculated at each point using Eq. 7 (Figure 3.3, black). A power function is then fitted to the log-linear portion of the compression curve (Figure 3.3, red). Using the slope and intercept from the fitted power function, dVpfit can be calculated using Eq. 9 (Figure 3.1B, blue). Finally, applying Eq. 10 yields the closure and compression corrected incremental intrusion curve (Figure 3.1C).3.2.4 Stressed porosity calculationColombo and Carli (1981) calculated the bulk volume compressibility of a non-porous rubber, polybutadiene, using MIP. They consider the differential form of the equation for bulk volume compressibility (Eq. 6):∫𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑜𝑜= −𝛽𝛽 ∫ 𝑑𝑑𝑑𝑑𝑃𝑃0      ……… (11)where V is the volume of solid and β is the compressibility coefficient with the specific boundary conditions of V = Vo at P = 0. By integrating, Eq. 11 becomes:ln𝑉𝑉𝑉𝑉𝑜𝑜= −𝛽𝛽𝛽𝛽     ……… (12)where Vo is the initial volume of solid and V is the volume of solid at pressure P. By applying similar mathematics we consider:𝑑𝑑𝑑𝑑𝑑𝑑= −𝐶𝐶𝑝𝑝𝑝𝑝𝑃𝑃𝑚𝑚𝑑𝑑𝑃𝑃     ……… (13)where φ is porosity (or Vp as in Eq. 7). Integrating over the boundary conditions φ = φo at P = Po where φo is the unstressed porosity, Po is ambient pressure or initial test pressure, and φP is the porosity at corresponding pressure P:∫𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑃𝑃𝑑𝑑𝑜𝑜= ∫ −𝐶𝐶𝑝𝑝𝑝𝑝𝑃𝑃𝑚𝑚𝑑𝑑𝑃𝑃𝑃𝑃𝑃𝑃𝑜𝑜      ……… (14)After integrating, Eq. 14 can be written as:ln𝜑𝜑𝑃𝑃𝜑𝜑𝑜𝑜=−𝐶𝐶𝑝𝑝𝑜𝑜(𝑃𝑃𝑚𝑚+1−𝑃𝑃𝑜𝑜𝑚𝑚+1)𝑚𝑚+1      ……… (15)Taking the exponent of both sides and rearranging:36𝜑𝜑𝑃𝑃 = 𝜑𝜑𝑜𝑜𝑒𝑒−𝐶𝐶𝑝𝑝𝑝𝑝(𝑃𝑃𝑚𝑚+1−𝑃𝑃𝑝𝑝𝑚𝑚+1) 𝑚𝑚+1⁄        ……… (16)Eq. 16 defines the porosity versus stress curve for all stresses. In the specific case where m = 0 and Po = 0, Eq. 16 reduces to the form presented by Colombo and Carli (1981). Porosity reduction at any net confining stress (NCS) can then be calculated by:𝑅𝑅𝜑𝜑 = 100(𝜑𝜑𝑜𝑜−𝜑𝜑𝑃𝑃𝜑𝜑𝑜𝑜)     ……… (17)where Rφ is the percentage reduction in porosity from ambient to any pressure P. An example of the stressed porosity correction is shown in Figure 3.4 for a typical mudstone sample from this study. Using a known porosity measured at ambient conditions, such as from helium pycnometry, Eqs. 16 and 17 can be applied using the slope and intercept calculated in Figure 3.3. An example of Eq. 16 is shown in Figure 3.4A, which can be used to determine porosity at any NCS. Alternatively, the percentage of porosity reduction from ambient to any stress (Eq. 17) is shown in Figure 3.4B. In this case, the helium pycnometry porosity at ambient pressure is 5.9 % which is reduced by 22 % at a NCS of 2,500 psi (17.2 MPa). Therefore, the confined porosity at 2,500 psi is 4.6 %.3.2.5 AssumptionsThe compression corrected incremental and stressed porosity calculations (Eq. 10; Eq. 16) require several assumptions. First, the grain volume is assumed incompressible and therefore any incremental intrusion is attributed to pore volume compression or intrusion of pore space. This assumption is supported by a poor correlation between Rφ and total organic carbon (TOC) and clay at 3,000 psi (20.7 MPa) NCS (Figure 3.5). If the more compressible components of the matrix were contributing significantly to pore volume loss some correlation would be expected. Second, it is assumed that once a pore has been intruded it can no longer contribute to compressibility. In theory, once a pore is intruded, the NCS of the pore is reduced to zero, the pore rebounds to the unstressed volume, and the mercury that was injected during compression is displaced into the rebounding pore (Bailey, 2011). The total pore volume corrected for closure (Vpc) is then assumed equal to the total pore volume for all pores with throats greater than approximately 3 nm. However, MIP compression tests on coals (Suuberg et al., 1995) and hydrostatic compression tests on mudstones (Suarez-Rivera and Fjær, 2013) show a significant amount of mercury uptake and volumetric strain hysteresis 37between loading and unloading stages. The pores had not recovered until the sample was reduced to atmospheric pressure, even when samples were held at constant pressure and allowed to equilibrate (Suuberg et al., 1995). Therefore, MIP may be underestimating porosity due to the compression of pores that do not fully rebound once the pore is intruded and NCS is reduced to zero. In addition, the extent to which pores are completely compressed and cut off from the pore system is unknown. This likely leads to distortion of the pore size distribution towards smaller pore throats. Each of these factors, as well as MIP only being able to access pore throats greater than about 3 nm, contribute to the disparity between MIP porosity and helium pycnometry porosity. Figure 3.6 demonstrates this disparity on 115 tight mudstone samples analyzed using both MIP and helium pycnometry from the Duvernay Formation. Porosity values to helium are consistently higher than MIP porosities, and the disparity increases with increasing porosity.3.3 RESUlTS AnD DISCUSSIon3.3.1 Sample	size	effect	on	closureClosure was found to be sample size dependent, which is in agreement with other studies (i.e. Comisky et al., 2011; Figure 3.7). Closure increases as sample size decreases and constitutes a significant percentage of the total intrusion for all samples, but most notably for the crushed samples. The total closure, as a percent of total intrusion, is 9.3 % for the single plug sample, 13.9 % for the two plugs sample, 35.8 % for the 4 – 8 mesh sample, and 43.3 % for the 4 – 20 mesh sample. As seen in the incremental intrusion curves plotted in Figure 3.7, closure is complete by 15 psi (0.1 MPa) for all samples. In order to compare with bulk density measured by mercury immersion and Archimedes’ principle at atmospheric conditions, we apply the closure correction at the nearest data point to 14.7 psi (0.1 MPa). Due to increasing closure volume with decreasing sample size, measured bulk density from MIP decreases from 2.48 to 2.42 g/cm3 with an average of 2.45 g/cm3 and a standard deviation of 0.03 g/cm3 (Table 3.1). Uncorrected porosity increases from 1.79 to 3.71 % for decreasing sample size, with a standard deviation of 0.77 % (Table 3.1). By applying the closure correction, the average bulk density increases to 2.47 g/cm3 and the standard deviation is reduced to 0.01 g/cm3. Bulk density measured on a plug using mercury immersion and Archimedes’ principle is 2.49 g/cm3. The plug and two plugs are in close agreement with immersion bulk density, with MIP bulk densities of 2.48 g/cm3 each. The closure correction does not fully eliminate sample size 38effects because corrected MIP bulk densities of the crushed samples are still lower than the plugs (2.45 and 2.46 g/cm3). The difference in bulk density is likely attributed to the increased surface area and roughness associated with the finely crushed particles. Closure corrected porosity increases from 1.65 to 2.25 % with decreasing sample size and the standard deviation is reduced to 0.24 %. The closure correction reduces standard deviation between raw and corrected porosity values significantly; however, systematic sample size trends still exist in porosity measurements due to pore access as mentioned by other authors (Comisky et al., 2011).3.3.2 Commercial lab closure correctionMercury intrusion on a mudstone sample from this study was performed on identical sample splits by our laboratory and a commercial laboratory to illustrate the corrections and missing data often supplied to researchers and industry alike. A comparison of the results of the two analyses are presented in Figure 3.8. This is not intended to represent all analyses from all commercial laboratories, but serves to demonstrate the magnitudes of corrections that may be applied (but unreported) before data is delivered. As noted by Bailey (2011), commercial laboratories commonly apply a correction to MIP data at the point of true intrusion and eliminate any prior compression or intrusion from reported data (compression corrected herein). Porosity to helium for the example sample is 2.29 %, closure corrected MIP porosity measured by our laboratory is 1.04 % and compression corrected porosity reported by the commercial laboratory is 0.2 %. Clearly, by applying a simple correction the apparent porosity calculated by the commercial laboratory is artificially low. The commercial laboratory intrusion correction removes the volume of mercury injected into the penetrometer due to compression, which in theory is displaced into rebounding pores after intrusion of mercury into the sample. Therefore, the volume of mercury intruded during compression is included in the calculation of porosity. In this instance the commercial lab did not report MIP bulk density; however, it would have been artificially high because of the correction applied which leads to the low porosity value.3.3.3 Compression	modeling	and	correctionThe impact of compression on porosity was modeled using Eq. 16 and 17 on samples of various sizes (Figure 3.9). In general, for any given NCS porosity reduction increases with decreasing sample size. The single plug sample versus two plugs sample show slightly different compression 39trends resulting in slightly less porosity reduction for the two plugs sample than the single plug sample (Figure 3.9). For example, at 3,500 psi (24.1 MPa) NCS porosity is reduced by 28 % in the single plug sample while the two plugs sample is reduced by 26 %. In comparison, the porosity of the 4 – 8 mesh sample and 4 – 20 mesh sample is reduced by 32 % and 36 %, respectively. The trend of increasing compression with decreasing sample size is likely due to sample shape effects; as the samples are crushed to smaller sizes they break dominantly parallel to bedding. Smaller sized samples begin to appear as chips rather than uniformly dimensioned plugs and therefore samples will have slightly different stress distributions. Fine-grained mudstones are known to be anisotropic with respect to mechanical properties. As an example, Suarez-Rivera and Fjær (2013) found the radial direction perpendicular to bedding exhibits nearly twice the strain as the radial and axial directions parallel to bedding in one study (their Figure 2). As crushed samples tend to elongate in a direction parallel to bedding, a significantly greater amount of the total surface area is exposed to stress applied in the direction perpendicular to bedding. Choosing a sample size for MIP analyses will depend on a balance of the properties sought. Smaller samples will allow more porosity to be intruded due to increased surface area; however, smaller samples may be prone to higher compression due to sample shape effects. In contrast, plugs limit crushing induced effects, but expose less surface area to intrusion, and thus consistently give lower MIP porosities. Larger plugs may be more suited to compression tests due to shape dependency.3.3.4 Closure	and	compression	effect	on	permeability	estimationNumerous models exist to estimate permeability from MIP data. A summary can be found in Comisky et al. (2007) who also compared steady-state permeability in tight gas sands to permeability estimated from MIP models. Comisky et al. (2007) found that existing correlations are poor where steady-state permeability is less than 0.2 md and in general become worse as permeability decreases. Despite the poor correlation for tight gas sands, permeability estimates are still commonly reported for tight reservoirs such as siltstones and mudstones (e.g. Chalmers and Bustin, 2008b; Ross and Bustin, 2008; Mastalerz et al., 2013). Shale gas reservoirs commonly have permeabilities a few orders of magnitude below permeabilities reported for tight gas sands (e.g. Chalmers et al., 2012b), thereby increasing uncertainty in the correlations. In addition, the closure and compression of 40samples, as noted in this study and others, will further impact calculations. We highlight the difficulties in correlating permeability with MIP derived parameters using the commonly reported “Swanson permeability” method to estimate permeability of tight reservoirs (Swanson, 1981). The model was developed using air and brine permeability for sandstones and carbonates based on the point at which the ratio of the bulk mercury saturation, Sb, to injection pressure, Pc, takes a maximum value, A, defined as (Swanson parameter herein):(𝑆𝑆𝑏𝑏𝑃𝑃𝑐𝑐) 𝐴𝐴    ……… (18)Calculating the Swanson parameter on samples which exhibit closure and compression is difficult as the Swanson parameter often cannot be clearly identified in experimental data. Figure 3.10 illustrates this problem on the mudstone sample used in Figure 3.1 through Figure 3.4. The Swanson parameter appears in Figure 3.10 as a black dot. In the uncorrected data, the Swanson parameter is within the closure region (Figure 3.10A, grey). In the closure corrected data the Swanson parameter is found within the region of compression (Figure 3.10A, blue) and in the compression corrected data the Swanson parameter is found at the final point of intrusion (Figure 3.10A, orange) When the Swanson parameter is taken at these points the Swanson permeability ranges over 7 orders of magnitude, from 7.52 mD in the uncorrected sample to 3.1 x 10-7 mD in the compression corrected sample. In comparison, confined permeability to helium on this sample measured 2.07 x 10-5 mD at 6,000 psi (41.4 MPa) NCS. Clearly there is need for an improved calculation for tight reservoirs. Interestingly, when the Swanson parameter is taken at the point of intrusion, confined permeability to helium and the Swanson permeability are in close agreement in this sample. This relationship was investigated on other samples and it does not occur frequently. Swanson permeability typically underestimates confined helium permeability by up to 2 orders of magnitude. Since cores are not commonly taken from wells in tight reservoirs, a correlation to derive permeability for stress-sensitive rocks that undergo significant compression before intrusion by MIP is an avenue of research that needs further exploration.3.4 ConClUSIonS AnD RECoMMEnDATIonSThis study highlights common experimental errors associated with MIP on fine-grained reservoir rocks. Necessary corrections are not routinely applied or not explicitly stated within the literature or are applied without the clients’ knowledge by commercial laboratories. Novel methods were 41expanded to account for closure and compression concurrent with intrusion. Examples demonstrate the dependence of closure on sample size and the impact of closure on MIP properties such as bulk density and porosity. Examples demonstrating the workflow for the new calculations were presented using a typical fine-grained mudstone which can be implemented on any fine-grained rock exhibiting sample compression. New calculations were developed to correct porosity measured at ambient conditions to stresses more representative of reservoir conditions which show that unconfined porosity values may be significantly over-estimated compared with stressed porosity values. Compression dependency on sample size was also investigated, with larger samples such as full plugs being more suited to compression tests. However, larger samples do not allow significant pore access and porosity may be underestimated. Smaller samples allow more pore access and thus characterize a greater proportion of the pore space, but may be prone to increased compression due to shape effects. Commercial laboratory corrections were investigated and found to remove required data for properly calculating MIP properties, such as porosity. Further, compression data, which may be of interest to some clients, was removed. More research is needed to develop a permeability estimate from MIP for tight rocks. Commonly used models, such as Swanson permeability, do not account for compression and were developed for less stress sensitive rocks of higher permeabilities and therefore have limited applicability to tight rocks which are stress sensitive. Currently, MIP models do not account for compression of samples or changing permeability with effective stress.4200.0020.0040.0060.0080.011 10 100 1000 10000 100000dVi dVic00.0020.0040.0060.0080.011 10 100 1000 10000 100000Incremental Volume (mL)dVpfit dVic00.0020.0040.0060.0080.011 10 100 1000 10000 100000Injection Pressure P (psi)dViccClosure Intrusion(A)(B)(C)Figure 3.1 Incremental volume (mL) versus injection pressure (psi) from an MIP analysis on a typical fine-grained mudstone of the Duvernay Fm. The closure volume is subtracted from the uncorrected raw data (blue curve in A) to obtain the closure corrected incremental volume curve (red curve in A). Subtracting the volume intruded due to compression (blue curve in B) from the closure corrected incremental curve (red curve in B) gives the closure and compression corrected incremental curve (red curve in C).43R² = 0.9801020304050607080901001 10 100 1000 10000 100000Cumulative Volume (%)Injection Pressure P (psi)Closure CompressionIntrusionFigure 3.2 Semi-log cumulative intrusion plot for the typical mudstone sample used in Figure 3.1. Red line is a power law fit over the interval attributed to compression. Where the fit deviates is closure (at low pressure) or intrusion (at high injection pressure).440.0000010.000010.00010.0010.010.111 10 100 1000 10000 100000Compressibility Cp (psi-1) Injection Pressure P (psi) Cp CpfitClosureCompressionIntrusiony = 0.0394x-0.956Figure 3.3 Compressibility (Cp, psi-1) versus injection pressure (P, psi). A power law function is fit (red) to the range attributed to compression. Closure and intrusion are apparent where Cp (black) deviates from the power law fit.4505101520253035401 10 100 1000 10000 100000Porosity Reduction (%)00.010.020.030.040.050.060.070 10000 20000 30000 40000 50000 60000 70000Pressure P (psi)(A)(B) Pressure P (psi)Porosity (dec)Figure 3.4 Porosity at stress calculated for the typical mudstone sample used in previous figures (A). The relative percent reduction at any NCS can also be calculated and used to correct porosities measured at ambient conditions (B).46Figure 3.5 Porosity reduction versus TOC (A) and clay (B) for mudstones of the Duvernay Fm. analyzed for this study. Poor correlations for both TOC and total clay content with porosity reduction, Rφ, suggest that the more ductile matrix components do not significantly contribute to compression. Red lines are linear least squares fits, R2 is the coefficient of determination and n is the number of samples.R² = 0.07n = 1150.000.200.400.600.801.000 2 4 6 8 10Porosity Reduction (dec)TOC (%)R² = 0.01n = 800 20 40 60 80 100Clay (%)(A) (B)47R² = 0.66n = 1150123456780 1 2 3 4 5 6 7 8He Porosity (%)MIP Porosity (%)Figure 3.6 Helium porosity (%) versus MIP porosity (%) measured on mudstones of the Duvernay Fm. Helium porosity is consistently higher than MIP porosity. Black line is 1:1 ratio, red line is a linear least squares fit, R2 is the coefficient of determination and n is the number of samples.4800.0050.010.0150.020.0250.030.0351 10 100 1000 10000 100000Incremental Volume dVi (mL) Injection Pressure P (psi) 1 Plug 2 Plugs 4-8 Mesh 4-20 Mesh SS Ball BearingsClosureCompressionIntrusionFigure 3.7 Incremental volume intrusion plot for the four mudstone samples of different sizes tested in this study, as well as stainless steel ball bearings for comparison. Closure increases as sample size and sorting increase. Quarter-inch stainless steel ball bearings (dark blue) do not exhibit significant closure in comparison to the rock samples.49ρb ρbc φ φcSample g/cm3 g/cm3 % %Plug 2.48 2.48 1.79 1.652 Plugs 2.47 2.48 1.90 1.684 - 8 Mesh 2.43 2.45 2.69 1.834 - 20 Mesh 2.42 2.46 3.71 2.25Average 2.45 2.47 2.52 1.85St. Dev 0.03 0.01 0.77 0.24Table 3.1 Raw and closure corrected (denoted by subscript c) density (ρ) and porosity (φ) data for the four samples crushed to various sizes.5005101520251 10 100 1000 10000 100000Incremental Volume (%)0204060801001 10 100 1000 10000 100000Cumulative Volume (%)Injection Pressure P (psi)Raw Closure Corrected Commercial Lab(B)(A)Figure 3.8 Raw and closure corrected intrusion curves analyzed in this study compared with corrected data supplied by a commercial laboratory for a single mudstone sample from the Duvernay.51Porosity Reduction R φ (%)Injection Pressure P (psi)01020304050601 10 100 1000 10000 100000Plug 2 Plugs 4 - 8 Mesh 4 - 20 MeshRφ at 3500 psi NCS:Plug:   28%2 Plugs:  26%4 - 8 Mesh:  32%4 - 20 Mesh:  36%Figure 3.9 Porosity reduction (Rφ) versus injection pressure (psi) calculated for the four samples of various sizes. Porosity reduction at 3,500 psi NCS was chosen to illustrate the general trend of increasing porosity reduction and decreasing sample size. See text for discussion.521101001000100001000001E-090.00000010.000010.0010.110Swanson Brine Permeability (mD)He Permeability1101001000100001000000.010.1110Bulk Mercury Saturation Sb (%)Uncorrected Closure CorrectedCompression Corrected (Sb/Pc)AIntrusionInjection Pressure P (psi)Swanson Permeability:Uncorrected:  7.52 mDClosure Corrected: 5.8E-01 mDIntrusion Corrected: 3.1E-07 mDHe Perm @ 6000 NCS: 2.07E-05 mD(A)(B)CompressionClosureIntrusionCompressionClosureFigure 3.10 (A) Bulk mercury saturation versus injection pressure with uncorrected raw data (grey curves), closure corrected (blue curves) and compression corrected (orange curves) data for the typical mudstone used in Figures 3.1 - 3.4. The black dots represent the point at which Sb/Pc is a maximum, which is the parameter used in Swanson’s (1981) model to calculate permeability. (B) Swanson permeability calculated for all values of Sb/Pc. The black dots represent the Sb/Pc value used to derive Swanson permeability. The red line is confined permeability to helium at various net confining stresses.53Chapter 4: Impact	of	thermal	maturity	and	compaction	on	the	pore	size	distribution	and	matrix	permeability	in	the	shale	gas	and	shale	oil	producing	Duvernay Formation, Alberta4.1 InTRoDUCTIonStudies into the pore structure of tight reservoir rocks (mudrocks and tight siltstones) are an important, albeit challenging, aspect of reservoir characterization due to the heterogeneity of fine grained rocks at the nanometer scale. Multiple techniques have been utilized to elucidate the pore structure of tight rocks, each offering a unique perspective into pore morphology, distribution and association. Methods such as nitrogen and carbon dioxide low-pressure gas sorption (LPGS) and mercury intrusion porosimetry (MIP) have been used to quantify pore size distributions and pore throat distributions (e.g. Ross and Bustin, 2009; Chalmers et al., 2012a; Clarkson et al., 2013; Kuila and Prasad, 2013; Kuila et al., 2014b) along with small-angle and ultrasmall-angle neutron scattering (e.g. Mastalerz et al., 2012; Clarkson et al., 2013; Bahadur et al., 2014). Scanning electron microscopy (SEM), field-emission scanning electron microscopy (FE-SEM), or transmission electron microscopy (TEM) coupled with focused ion beam (FIB) or broad ion beam (BIB) milling have also been used to provide a visual link into tight reservoir pore structures (Milner et al., 2010; Curtis et al., 2011b; Slatt and O’Brien, 2011; Chalmers et al., 2012a; Curtis et al., 2012a, 2012b; Loucks et al., 2012; Milliken et al., 2013; Reed et al., 2013; Klaver et al., 2015; Lu et al., 2015). A diverse array of pore structures have been documented (e.g. Loucks et al., 2012; Milliken et al., 2013) which are dependent on grain size, mineralogy, organic matter (OM) content and grain provenance that create a unique reservoir texture which, in turn, varies with burial depth (compaction), thermal maturity and chemical diagenesis. Despite the multitude of studies and techniques applied to characterize the pore structure of tight reservoir rocks, determining the controls on porosity, pore size distribution and permeability within a given tight reservoir is not a trivial task and diverse opinions exist. The documented pore structure is method-dependent and studies have yielded different results because of the methods employed. Therefore, there is a need to analyze many samples utilizing various methods to elucidate the controls on pore structure variation with composition and maturity. Recent studies have made advances in understanding the nature of porosity within tight reservoirs; however, the linkage between the impact of thermal maturity on the 54evolution of tight reservoir pore structure and the consequent impact on matrix permeability remains elusive. Previous studies have investigated mudstones as lithological units within basinal successions to determine the impact of compaction on the pore structure and flow properties (e.g. Vasseur et al., 1995; Dewhurst et al., 1998, 1999; Aplin et al., 2003, 2006; Mondol et al., 2007; Yang and Aplin, 2007; Schneider et al., 2011) or in the context of tight hydrocarbon reservoirs with the aim of determining the controls on hydrocarbon production and distribution (e.g. Ross and Bustin, 2009; Chalmers and Bustin, 2012a; Chalmers et al., 2012b; Mastalerz et al., 2012; Clarkson et al., 2013; Kuila and Prasad, 2013; Mastalerz et al., 2013; Kuila et al., 2014b). Studies investigating mudstones as lithologic units have largely focused on the impact of burial and compaction on the pore structure and hydraulic conductivity of clay minerals and mudstones. Vasseur et al. (1995) investigated the impact of compaction and microstructural changes on kaolinite in the range of 0.1 to 50 MPa and found the MIP modal pore throat size varied as a function of effective stress, shifting from 100 nm to 55 nm over an effective stress range of zero (uncompacted) to 20 MPa. Dewhurst and Aplin (1998) investigated the impact of compaction on pore network collapse and permeability in organic-lean mudstones (< 0.88 % total organics) and found that larger pores were preferentially collapsed up to 33 MPa, shifting modal pore size from 30 nm to 15 nm over an effective stress range of 2 – 33 MPa. The authors also found that permeability decreased systematically with increasing effective stress due to the collapse of larger pores (Dewhurst et al., 1998). Dewhurst and Aplin (1999) investigated the influence of clay content and grain size distribution on compressibility and pore structure of clay-rich mudstones and found compaction to 10 MPa leads to preferential loss of larger pores, decreasing the modal pore size from 60 to 25 nm. Yang and Aplin (2006) investigated the relationship of pore structure and lithology on permeability in natural mudstones and found the reduction of porosity to be primarily the result of large pore compression. Schneider et al. (2011) studied laboratory-prepared mudstones of varying clay- and silt-size fractions compacted to lower pressures of 2.4 MPa and found increasing permeability with decreasing clay fraction due to the preservation of large pores around silt grains. Studies that consider mudrocks (“shale”) in the context of hydrocarbon reservoirs have largely focused on determining the geologic controls on pore size distribution and to a lesser extent 55the impact of varying pore size distributions on permeability. In comparison to studies investigating clay minerals and mudstones as lithologic units, mudrocks in reservoir characterization studies generally have higher OM content and have undergone some degree of thermal maturation, which is necessary for thermogenic hydrocarbon generation. Recently, the thermal generation of OM-hosted porosity has been proposed to contribute a significant percentage of the total porosity of organic rich mudrock reservoirs (Loucks et al., 2009, 2012; Chalmers et al., 2012a; Milliken et al., 2013). However, the development of OM-hosted porosity with maturity is not conclusive (e.g. Bernard et al., 2012; Mastalerz et al., 2013; Reed et al., 2013; Bernard and Horsfield, 2014) and the subsequent evolution of the pore size distribution and impact on matrix permeability has not been evaluated. Chalmers et al. (2012b) investigated the impact of pore size distribution on matrix permeability within a suite of mudstones and found higher permeability samples exhibit more balanced ratios of micro-, meso- and macropores (see Figure 4.1 for pore size classification scheme; Sing et al., 1985). Ross and Bustin (2009) found higher micropore volumes to be associated with thermally mature mudstones which varies with organic content and that clay-rich sediments have a significant percentage of total porosity within the mesopore range. Kuila and Prasad (2013) found that clay minerals (illite-smectite) have pore structures at multiple scales and that compaction of montmorillonite resulted in a decrease in pore volume of macropores between clay aggregates but no change in the micropore size fraction between clay “tachoids”. Kuila et al. (2014b) attempted to quantify the influence of OM-hosted porosity developed during thermal maturation using LPGS and found that OM-hosted pores are only present at high maturities, possibly associated with gas generation. The authors also argued clay hosted micro- and mesoporosity to be a fundamental building block of mudrocks; however, their samples contained 36 - 63 % clay minerals (mix layered illite-smectite), which is not representative of all mudrocks nor many hydrocarbon producing mudrock reservoirs. Previous studies provide the framework by which to compare the impacts of thermal maturity on the pore structure of organic-rich mudrocks of variable composition and the relationship to matrix permeability. This paper extends previous studies to consider the influence of compaction and thermal maturity on the pore structure evolution of organic-rich, shale gas and shale oil producing reservoir rocks. The Duvernay is ideally suited to this purpose, as the Formation spans multiple 56thermal maturity boundaries from immature to overmature and is compositionally similar. Sampling was performed over the entire range of thermal maturities which enables the comparison of pore structure versus maturity from samples with the same depositional environment, mineral provenance and OM type. The study focused on utilizing LPGS to highlight the pore structure attributes and was supplemented with FE-SEM imaging in order to provide a more complete characterization of the pore structure with maturity. The pore structure attributes from these methods were then correlated to matrix flow properties measured using confined pulse-decay permeability and confined gas expansion permeability. Specifically, this study investigates the thermal evolution of mudrock pore structure, the impact of thermal maturity on OM-hosted porosity, its contribution to the pore structure, the impact of pore structure evolution on permeability, the impact of compaction and burial on the pore structure and permeability, the compositional control on mudrock wettability and the impact of pore wettability on capillary pressure and hence permeability. This study furthers the understanding of the relationship between pore size distribution, thermal maturity, texture, and permeability and links thermal diagenesis to the enhancement of permeability, despite the shift in pore modality to smaller pore sizes.4.2 AnAlYTICAl ConSIDERATIonSIn most tight reservoirs, pore sizes approach or are below the analytical limits of many reservoir characterization techniques (Clarkson et al., 2012, their Figure 1). Pore classification schemes for mudstone reservoirs have been proposed which are based on visual porosity associations using techniques where the lower limit of investigation is greater than the smallest pores (e.g. Loucks et al., 2012). From comparisons with other measurements, such as low-pressure gas adsorption and MIP, it is known that porosity exists below the resolution of TEM/SEM instruments (e.g. Chalmers and Bustin, 2012a; Chalmers et al., 2012a; Kuila et al., 2014b). The quantity of porosity below visual observation as measured by other methods is variable and poorly constrained, but may constitute the majority of the total pore volume of tight reservoirs (Kuila et al., 2014b). A recent review of studies which use TEM/SEM methods to investigate pore sizes in fine-grained reservoirs reveals that the smallest reported pore diameter is around 2 – 3 nm (Curtis et al., 2011b), 2.5 nm (Curtis et al., 2012b), 3 nm (Milner et al., 2010), 4 – 7 nm (Milliken et al., 2013), 5 nm (Curtis et al., 2011a; Chalmers et al., 2012a; Loucks et al., 2012; Klaver et al., 2015), 57and 7 nm (Lu et al., 2015). The practical limitations of spatial resolution (lower confidence limit) of SEM analyses are rarely reported (e.g. Chalmers et al., 2012a; Curtis et al., 2012b; Milliken et al., 2013; Lu et al., 2015) which admittedly may be due to varying sample properties which can alter resolution between samples. However, quantifying physical pore properties, such as length, width, area, and equivalent circular diameter from two-dimensional images has become frequent (e.g. Milliken et al., 2013; Lu et al., 2015) despite the fact that the morphology of the pore in three-dimensions is unknown but never spherical. These methods commonly assume a circular pore shape (i.e. equivalent circular diameter, (Lu et al., 2015), whereas gas adsorption isotherm theory implies a dominant slit- or slot-shaped pore size distribution (Clarkson et al., 2013). Attempts to quantify OM-hosted porosity have proven to be difficult, where methods are based on point counting, grey-scale contrast tracing of OM pores on SEM micrographs (Milliken et al., 2013; Lu et al., 2015) or removal of OM before and after LPGS analysis, which is only partially effective (Kuila et al., 2014b). Therefore, efforts to determine the volumetric contribution of various pore sizes or types (mineral- versus OM-hosted) to the total porosity and correlate pore structure to permeability by utilizing a single technique may be incomplete due to fundamental limitations.4.3 MATERIAlS, METhoDoloGY AnD lIMITATIonS4.3.1 Core samplesSamples were selected over a range of thermal maturities and burial depths from west-central Alberta (Figure 4.2). For tight reservoirs, a high sampling density is necessary for a reservoir characterization study to be successful since tight reservoirs are heterogeneous on the millimeter scale and smaller (Chalmers et al., 2012b). The heterogeneity is often difficult to appreciate at hand sample scale, which renders targeted sampling by facies ineffective. Determining the controls on pore structure, permeability and sample composition can be obscured by not capturing the sedimentary variability of the reservoir within the sample suite. A large sample suite also helps to minimize the effects of sampling bias as well as the small sample volumes analyzed in laboratory experiments. This is especially crucial for reservoirs which span multiple thermal maturity boundaries, as reservoir characteristics are known to change with maturity; the variability between samples of similar maturity can easily distort correlations with samples of varying maturities. Therefore, a few “representative samples” from varying maturity windows contribute little to 58understanding the true associations of reservoir quality with facies and thermal maturation. It should be noted that all samples within this study are considered “representative”. In an attempt to capture the variability of the reservoir, 78 samples were selected for LPGS analysis from 15 wells of varying maturity. Plugs for confined permeability were taken parallel to bedding for 26 samples. Bulk samples were taken for the pore size distribution, organic geochemistry and mineralogy analyses.4.3.2 MineralogyCrushed sample (< 250 μm) was mixed with ethanol, hand ground in a mortar and smeared on a glass slide for X-Ray diffraction analysis. Normal-focus CoKα radiation on a Bruker® D8 Focus at 35 kV and 40 mA was used on samples over a range of 3-70° 2θ at a step size of 0.03° and counting for 0.8 s per step while continuously spinning at 60 rpm. Mineral phases were quantified using the Rietveld method (Rietveld, 1967) of full-pattern fitting using Bruker® AXS Topas V3.0 software.4.3.3 organic	geochemistryThe total organic carbon (TOC) contents, organic geochemistry and Tmax values were determined using a SRA-TPH/TOC™ source rock analyzer. Samples were crushed (< 250 μm) and approximately 1.0 g of sample pyrolysed. Common parameters calculated from pyrolysis experiments are discussed elsewhere (Peters, 1986).4.3.4 Pore size distributionStudies which utilize LPGS have adopted the International Union of Pure and Applied Chemistry classification scheme to characterize the porosity of tight reservoirs, where macropores have widths greater than approximately 50 nm, mesopores have widths between approximately 50 and 2 nm, and micropores have widths less than approximately 2 nm (Figure 4.1; Sing et al., 1985). These divisions are based on the dominant process in which pores fill (e.g. micropore filling in micropores, mono- to multilayer adsorption and capillary condensation in mesopores and small macropores) and as such will be influenced by pore shape as well as the interaction between the adsorbent and the adsorbate (Sing et al., 1985). Low-pressure gas sorption for this study was performed on a Quantachrome® Autosorb-1 using N2 and CO2 at 77 K and 273 K, respectively. Adsorption and desorption isotherms were run at 78 pressure steps over a P/P0 range of 0.01 to 0.99. Nitrogen surface areas were calculated using the BET method (Brunauer et al., 1938). Similarly, CO2 equivalent surface areas were calculated using 59the BET and Dubinin-Radushkevich (D-R) equation (Gregg and Sing, 1982). Pore size distributions were determined by the BJH method (Barrett et al., 1951) and the density functional theory (DFT) method (Do and Do, 2003). The DFT models used were the N2 at 77 K on carbon slit pore model and the CO2 at 273 K on carbon slit pore model for N2 and CO2 analyses, respectively. Average fitting error between the experimental and theoretical isotherm for DFT analysis was less than 0.9 % for N2 and less than 0.05 % for CO2. The cross sectional area used for calculations was 0.162 nm2 for N2 (Rouquerol et al., 1994) and 0.21 nm2 for CO2 (Lowell, 2004).4.3.5 FIB/Fe-SeMScanning electron microscopy was performed using a FEI Helios NanoLab 650™ dual beam system. Prior to imaging, samples were milled using gallium ions at an accelerating voltage of 30.0 kV and a current of 9.3 nA. Images were collected using backscattered electrons at an accelerating voltage of 1 kV and a beam current of 0.2 nA. These beam conditions help to minimize the charging effects of samples which have not been prepared with a conductive coating (Curtis et al., 2012b). This study did not seek to quantify pores from SEM micrographs; therefore, images were collected on average at a horizontal field width of 7.46 μm (pixel resolution = 4.9 nm/pixel). This allowed practical observation of pores to approximately 5 - 15 nm. Samples were milled and imaged perpendicular to bedding in two separate locations on the sample.4.3.6 Confined	matrix	permeabilityConfined pulse-decay permeability (PDP) was determined using two custom designed, built-for-purpose permeameters utilizing helium and decane as the probe fluids. Plugs were milled in a milling machine to create smooth, parallel faces on both ends. Plugs were then confined in a hydrostatic Hoek-type core cell holder and subjected to 6,000 psi (41.4 MPa) confining pressure and 1,000 psi (6.9 MPa) pore pressure which approximates the present reservoir net effective stress. All permeability measurements were performed at the same net effective stress in order to compare measurements between samples since permeability is known to be stress sensitive for fine grained rocks (Chalmers et al., 2012b). Before analyses, plugs were dried at 60 °C until weight stabilized. Since plugs were taken as-received, saturations could not be considered to be representative of reservoir saturations. Using helium as a probe gas will yield higher permeability than using a reservoir gas due to the smaller kinetic diameter of helium. Methods from Cui et al. (2009) were 60followed to calculate permeability. Confined permeability was determined using a custom designed Boyle’s Law type apparatus, or gas expansion porosity and permeability (GEPP) apparatus. Plugs were confined to 6,000 psi (41.4 MPa) and 1,000 psi (6.9 MPa) of helium gas was expanded into the sample from a reference chamber of known volume. Once at equilibrium, the permeability could be determined by the pressure-decay curve, the final pressure at equilibrium and the known system volume. Analyses times are significantly longer than unconfined PDP on crushed samples or confined PDP measurements; analysis time to equilibrium for plugs used in this study, depending on the permeability and sample volume, are commonly in excess of 300 hours. Permeability was determined from the expansion decay following the methods of Cui et al. (2010). Combining PDP and GEPP provides a unique opportunity to examine the heterogeneity of the bulk rock matrix under stress. In longitudinally layered samples, the PDP method measures approximately the arithmetic average of the permeability while the GEPP method measures approximately the geometrical average of the permeability (Cui et al., 2010). In transversely layered samples, the PDP and GEPP methods will yield approximately equivalent permeabilities (Cui et al., 2010). The PDP method is biased toward measuring the highest permeability pathways or fractures and can be significantly influenced by microfractures developed during coring, plugging, or drying of the sample. The GEPP method, on the other hand, is influenced by the matrix permeability of the rock and is not as significantly influenced by microfractures. It has been shown that in longitudinally layered samples the GEPP method approximates the average permeability of the sample better due to the insensitivity to a high permeability fracture network (Cui et al., 2010). Permeability anisotropy can be assessed by differences between the PDP and GEPP permeabilities. Prior to this study, GEPP has not been reported in the literature to study the pore system of shale gas and oil reservoir rocks.4.3.7 WettabilityStatic contact angles were measured using the sessile drop method on a Biolin Scientific Theta™ optical tensiometer. The milled plugs collected for PDP were utilized for contact angle measurements since the plug ends are parallel and smooth. Any inclination or roughness of the sample surface will alter the static contact angle. While the wettability of a system is dependent on the interaction of multiple contact angles for various reservoir fluids (brine, oil, and gas), the pore space morphology 61(e.g. OM-hosted or mineral-hosted; Hirasaki, 1991), along with reservoir conditions (pressure and temperature), a single contact angle at atmospheric conditions can provide useful qualitative information about the compositional controls on total system wettability. Multiple fluids were tested to determine the feasibility of measuring the contact angle on Duvernay Fm. mudrocks. Fluids which spontaneously imbibe into the pore system pose difficulties in determining an accurate contact angle as the contact angle is distorted by the fluid entering the pore space. For example, decane sessile drops on the mudrocks within this study commonly have initial contact angles of < 10 degrees. The apparent decane contact angle quickly reduces to angles below measurement threshold due to imbibition. Water in air, however, was found to exhibit contact angles which were amenable to measurement and showed significant variation between samples. Therefore, water in air contact angles were utilized to qualitatively asses the compositional controls on wettability.4.4 RESUlTSPyrolysis data (hydrogen index and Tmax) were averaged for each well in the study and a qualitative descriptor of maturity was assigned (Table 4.1) in order to facilitate correlations between wells of varying maturity. The qualitative descriptors generally align with published pyrolysis maturity classifications (Peters, 1986; Peters and Cassa, 1994). Since the Duvernay is relatively thin (30 – 60 m), it is reasonable to assume that all Duvernay samples from a given well have experienced a similar thermal history. There are known issues associated with pyrolysis maturity parameters, such as indistinct S2 peaks and therefore unreliable Tmax values in high maturity samples (Peters, 1986) and drilling fluid contamination impacts on the S2 peak (Kuila et al., 2014a). While the hydrogen index (HI) is not strictly a maturity parameter and may vary due to organic matter type, all samples in this study come from a similar depositional environment so the HI is effective at describing maturity within this sample suite. A more general classification scheme is also used to refer to wells within the following sections; immature samples refer to wells which have not entered the oil window, mature samples are those within the oil to wet gas windows, and overmature samples are in the dry gas window. Below, we highlight important aspects of the microstructure of Duvernay Fm. mudrocks and present relevant correlations between composition, pore structure, permeability and maturity. 62Nitrogen LPGS, CO2 LPGS, FE-SEM images, PDP and GEPP are discussed separately as each method is biased toward different aspects of the pore structure.4.4.1 Sample composition and maturityMineralogy of Duvernay mudstones consists of varying proportions of quartz, carbonate and clay minerals with minor amounts of feldspars (plagioclase and sodium feldspar) and pyrite (Figure 4.3). Quartz plus feldspar content varies from 2 to 74 wt. % and averages 49 wt. %, total carbonate (calcite and ankerite/dolomite) varies from 3 to 98 wt. % and averages 29 wt. %, and total clay (illite/muscovite and chlorite group) content varies from 1 to 51 % and averages 19 wt. %. High carbonate intervals (e.g. limestones) have low TOC and low total porosity are included in the pore structure comparison of organic-rich Duvernay mudstones as end-member lithofacies. Total organic carbon content varies from 0 to 18 % and averages 4 %. Maturity as measured by Tmax varies from 417 °C to 482 °C for samples where the S2 peak was distinct. Tmax could not be determined for the wells within the dry gas window due to their advanced thermal maturity. The HI varies from 535 mg HC/g TOC to 10 mg HC/g TOC. Maturity generally correlates with present day burial depth for the basin. Organic matter is dominantly Type II (Tissot and Welte, 1984; Figure 4.4).4.4.2 nitrogen	LPgSAll N2 isotherms are hysteretic Type IV without a plateau at high partial pressures (Figure 4.5A), indicative of capillary condensation in mesopores (Sing et al., 1985). The absence of a plateau at high partial pressures is due to unrestricted macropore filling (Sing et al., 1985). Hysteresis profiles are Type H3 which are interpreted to be associated with slot- or slit-like pore geometries (Sing et al., 1985). All isotherms exhibit micropore filling at low P/P0 (< 0.01 P/P0) as well as forced closure of the hysteresis loop between approximately 0.47 and 0.50 P/P0 which is interpreted to be due to the tensile strength effect (Groen et al., 2003). Nitrogen BET surface areas vary from 1.56 to 34.24 m2/g and average 7.79 m2/g. No significant correlation exists between N2 BET surface area and sample composition when comparing all wells. The lack of correlation is interpreted to be due to the effects of maturity which alters the mudrock microstructure. For immature wells, N2 BET surface area averages 6.67 m2/g and shows a weak positive correlation to total clay content (R2 = 0.29) and no correlation to TOC content (R2 = 0.02). For mature samples, N2 BET surface area averages 6.08 m2/g and shows a positive correlation 63with TOC (R2 = 0.46) and no correlation with total clay content (R2 = 0.01). In overmature wells, N2 BET surface area averages 17.96 m2/g and shows a strong positive correlation with TOC (R2 = 0.83) and a moderate correlation with total clay content (R2 = 0.52). Total pore volumes from N2 LPGS show no correlation with sample composition for immature and mature wells. For overmature wells, total pore volume shows a poor correlation with TOC (R2 = 0.13) and a positive correlation with total quartz and feldspar content (R2 = 0.76) and clay (R2 = 0.42) indicating diverse modes of porosity occurrence as measured by N2 LPGS. Based on these relationships, a significant portion of the N2 measured pore volume may be associated with inter-crystalline porosity. Average pore diameter calculated by the D-R method varies from 3.5 to 6.8 nm and averages 5.9 nm. Average pore diameter decreases with increasing maturity. In immature wells, average pore diameter shows a weak positive correlation with TOC (R2 = 0.35). In mature wells, average pore diameter shows a weak negative correlation with TOC (R2 = 0.35). In overmature wells, average pore diameter shows a moderate negative correlation with TOC (R2 = 0.56).4.4.2.1 Bjh pore size distributionsPore size distributions calculated using the N2 BJH method vary systematically with maturity (Figure 4.6). Samples from the oil window (ATH 13-18-64) and wet gas window (ECA 11-8-62) were compared with immature and overmature samples to highlight the pore structure evolution with maturity.Immature samplesWithin immature samples, a higher proportion of pores within the macropore and coarse mesopore (> 10 nm) size fraction than occur in other maturity groups (Figure 4.6A). The pore size distribution peaks in volume around approximately 30 - 60 nm and progressively decreases in volume toward smaller pore diameters. The pore volume distribution indicates a minimal volume of N2-accessible fine mesoporosity within immature samples (approximately 0.0 mL/g near 1 nm). Immature samples also display a modal peak in pore volume at approximately 3 nm which is associated with clay mineral “intra-tachoid” porosity (Kuila and Prasad, 2013). Samples with the greatest volume of pores within the 20 – 70 nm fraction have greater TOC (up to 18.1 %) and total clay contents (34 %) (Figure 4.7A, D). In general, samples with lower pore volumes have correspondingly lower clay and 64TOC contents and variable mineral contents (Figure 4.7B, C)oil window samplesSamples from the oil window exhibit a similar character to the immature wells, but contain less porosity within the fine macropore and coarse mesopore size fraction (Figure 4.6B). These samples do not display the 3 nm peak associated with clay hosted porosity, indicating clay microstructure may not be a significant control on Duvernay porosity unlike other tight reservoirs (Kuila et al., 2014b). Samples from the oil window show a pore size distribution peak around approximately 20 – 70 nm with pore volume progressively decreasing toward smaller pore sizes. In contrast to immature samples, oil window samples contain greater volumes within the fine mesopore size fraction. Total organic carbon contents show a moderate control on pore distribution, with greater TOC samples generally having greater pore volumes while displaying similar pore size distributions to lower TOC samples (Figure 4.8A).Wet	gas	window	samplesSamples from the wet gas window have variable coarse meso- and macopores and greater fine mesopore volumes than oil window or immature samples. These samples do not display a pronounced peak in pore volume within the macropore and coarse mesopore size fraction and pore volume decreases only slightly in the micropore and fine mesopore size fraction (Figure 4.6C). Total organic carbon content shows a pronounced systematic control on pore size distribution, with greater TOC samples having more pore volume within the mesopore size fraction (Figure 4.9A). All samples display similar pore size distribution trends, however organic content shows the most control on the distribution of pores < 10 nm. Wet gas window samples contain more than twice the pore volume within the fine mesopore fraction as compared to oil window samples. Greatest coarse mesopore to macropore volumes are associated with highest quartz and feldspar contents (> 55 %), low total carbonate (< 20 %) and average clay (20 %) content (Figure 4.9B, C, D). These samples are also correlated with greater TOC contents; however, the sample with the greatest TOC content (8.3 %) does not have significant coarse mesopore to macropore volumes (Figure 4.9A)Dry	gas	window	samplesOvermature samples have, in general, the least pore volume within the macropore and coarse mesopore size fraction compared to other samples (Figure 4.6D). Fine mesoporosity within these 65samples is the greatest of the entire sample suite and displays the most pronounced increase in pore volume with decreasing pore size. Pore volumes within the fine mesopore range is commonly twice the volume of pores of wet gas window samples at a given pore size. The pore size distribution varies systematically with composition (Figure 4.10). High mesopore volumes are systematically associated with high TOC contents (Figure 4.10A), low total carbonate contents (< 35 %; Figure 4.10C) and average clay contents (20 %; Figure 4.10D). The sample with the greatest fine mesopore volume is associated with both high TOC and high clay contents.4.4.3 Carbon dioxide lPGSNitrogen lacks the kinetic energy at low temperatures (77 K) to access the finest micropores, whereas CO2 analyses typically operate at higher temperatures (273 K) which enhances the diffusion properties (Gregg and Sing, 1982; Unsworth et al., 1989; Groen et al., 2003; Rouquerol et al., 2013) and therefore CO2 can access pores inaccessible to N2. However, the saturation vapor pressure of CO2 at these analytical temperatures is higher than many sorption apparatuses can accommodate (commonly to 133 kPa), which limits the range of partial pressures which may be measured and therefore pore size (Rouquerol et al., 2013). The result is that in the low P/P0 range where N2 at 77 K struggles to characterize the material, CO2 analyses achieve greater accuracy in this portion of the isotherm but is restricted to a limited range (Rouquerol et al., 2013). In the studied sample suite, CO2 LPGS isotherms are hysteretic over a P/P0 range of 0.001 to 0.030 (Figure 4.5B). The low range of partial pressures measured reflects the analytical temperature of the carbon dioxide sorption analyses. The hysteresis is only partially recovered in all samples even at the lowest analytical pressure, which may be associated with pore structures (throats) with a similar diameter to the adsorbate, swelling of non-rigid pore structures or chemical interactions of the adsorbate with adsorbent (Sing et al., 1985). Carbon dioxide BET surface area for all samples varies from 1.0 m2/g to 13.1 m2/g and averages 5.4 m2/g. For immature samples, CO2 BET surface area averages 7.4 m2/g and shows a strong positive correlation to TOC (R2 = 0.97) and total clay content (R2 = 0.73). These correlations may be inherited from the positive correlation between clay and TOC for immature wells (R2 = 0.61), however immature organic matter has been shown to have a microporous texture resulting from quasi-amorphous stacked aromatic layers (Romero-Sarmiento et al., 2014). For mature 66samples, CO2 BET averages 3.8 m2/g and is positively correlated with TOC (R2 = 0.50) and shows a limited correlation with total clay content (R2 = 0.24). For overmature wells, CO2 BET surface area averages 7.0 m2/g and is positively correlated to TOC (R2 = 0.86) and moderately correlated with total clay content (R2 = 0.44). Carbon dioxide DFT total pore volumes (pore volume < 0.7 nm) for all samples vary from 0.09 mL/100 g to 1.03 mL/100 g and average 0.42 mL/100 g. For all wells, CO2 DFT pore volumes show a moderate correlation with TOC (R2 = 0.50) and a poor correlation with total clay content (R2 = 0.18). For immature wells, CO2 DFT pore volumes show good correlations with both TOC (R2 = 0.96) and total clay content (R2 = 0.75) due to the association of TOC with clay for immature samples as well as the microporous texture of immature organic matter. In mature samples, CO2 DFT pore volumes show a moderate positive correlation to TOC (R2 = 0.48) and a limited positive correlation to total clay content (R2 = 0.19). In overmature wells, CO2 DFT pore volumes show a strong positive correlation to TOC (R2 = 0.85) and a moderate correlation to total clay content (R2 = 0.48).4.4.4 FIB/Fe-SeMSamples from wells FB 10-4-51 (Figure 4.11), ATH 13-18-64 (Figure 4.12), ECA 11-8-62 (Figure 4.13) and HSK 10-33-56 (Figure 4.14) were selected for FIB/FE-SEM analysis to document the variation in mudrock pore structure with increasing maturity from the onset of hydrocarbon generation. All reported pore diameters from FE-SEM are apparent pore diameters since the morphology in three dimensions is unknown. In contrast to LPGS analyses, FE-SEM micrographs show both connected and unconnected porosity which may lead to disconnect between FE-SEM and LPGS analyses. Organic matter-hosted porosity is found to increase systematically in size and occurrence with maturity from the approximate onset of the oil window. The FE-SEM analysis of early oil window well FB 10-4-51 shows limited visible porosity overall. Organic matter and clays are distributed parallel to bedding, which may be an indication of hemipleagic sedimentation of clay particles to basin floor (Slatt and O’Brien, 2011) or compaction effects (Vasseur et al., 1995). Slit-like mesoporosity is associated with clay minerals (Figure 4.11E, F) and within OM particles (Figure 4.11D). Organic matter is generally non-porous at FE-SEM resolution; however, some OM-hosted porosity was observed (Figure 4.11C, F). When visible, 67OM-hosted porosity within the early oil window FB 10-4-51 consists of mesopores with maximum diameter of approximately 30 nm. Fine mesopores and possibly coarse micropores were observed in one particle (Figure 4.11F). Pores within OM in the oil window (ATH 13-18-64) are distributed as mesopores and fine macropores (Figure 4.12B, C, D). Minor slit-like porosity occurs along OM – grain boundaries (Figure 4.12B). Overall, OM-hosted porosity is variable but is significantly more developed than early oil window FB 10-4-51 samples. Inter-crystalline porosity is rare. Samples from the wet gas (ECA 11-8-62) show significantly less variability in OM porosity development than samples of lower maturity. The majority of OM particles contain some mesoporosity (Figure 4.13C, D, E, F). Significant inter-crystalline porosity occurs at quartz grain boundaries (Figure 4.13A, B). Inter-crystalline pores are generally within the coarse mesopore to macropore size fraction and visible OM-hosted porosity is generally within the coarse mesopore size fraction. Overall, OM is significantly more porous than oil window samples. A single, large (apparent 5 μm x 10 μm) OM particle was observed to be non-porous with slit-like mesopores at the grain boundaries (Figure 4.13A, B). Such particles may be derived from higher land plants (Milliken et al., 2013) rather than marine organisms. Samples from the dry gas window (HSK 10-33-56) show the best developed OM-hosted porosity, with pore sizes ranging from meso- to macropores. In some instances OM pores can exceed 200 nm in apparent diameter (Figure 4.14D, E). Inter-crystalline macropores at carbonate grain boundaries were observed (Figure 4.14D) as well as associated with clay minerals (Figure 4.14D).4.4.5 Matrix permeability4.4.5.1 helium PDPHelium PDP parallel to bedding for all samples ranges over two orders of magnitude and varies with maturity. Matrix permeability for all samples varies from 2.3 x 10-6 to 5.3 x 10-4 mD, averages 1.3 x 10-4 mD with a median value of 3.4 x 10-5 mD and shows no direct correlation to sample composition due to the effects of texture and maturity (Table 4.2). Permeability is the lowest in oil window samples, varying from the minimum permeability for all samples of 2.3 x 10-6 mD up to 5.6 x 10-5 mD and averaging 1.6 x 10-5 mD with a median value of 9.8 x 10-6 mD (Table 4.2). Permeability is highest in samples with high TOC (> 4.0 %) and 68high total clay contents (32 %). Permeability in oil window samples show no direct correlation to LPGS or MIP measured properties. Permeability for wet gas window samples varies from 6.1 x 10-6 to 5.3 x 10-4 mD and averages 1.8 x 10-4 mD with a median value of 9.5 x 10-5 mD, an order of magnitude higher than oil window samples (Table 4.2). Permeability for wet gas window samples show no direct correlation to sample composition. High permeability samples can have moderate quartz and total carbonate content (40 % respectively), high quartz and low carbonate content and high or low total clay content. Permeability shows a moderate positive correlation to N2 DFT total pore volume (R2 = 0.50). Permeability for dry gas window samples varies from 1.0 x 10-5 to 3.4 x 10-4 mD and averages 1.1 x 10-4 mD with a median value of 4.4 x 10-5 mD, similar to wet gas window samples, and an order or magnitude higher than oil window samples (Table 4.2). Higher permeability samples have higher total clay contents (> 25 %), low carbonate (< 35 %) and higher quartz and feldspar content (> 40 %). Permeability shows a weak positive correlation to N2 DFT total pore volume (R2 = 0.41).4.4.5.2 GEPPThe helium GEPP varies from 6.1 x 10-7 to 8.5 x 10-7 mD and is up to two orders of magnitude lower than the helium PDP on the same sample (Table 4.2). The GEPP is lowest within the oil window and increases with maturity. The GEPP shows limited correlation to compositional data.4.4.5.3 Decane PDPSamples from the oil window and wet gas window were chosen for liquid permeability measurements due to the presence of liquid hydrocarbons. Decane PDP for the sample suite varies from 2.3 x 10-6 to 5.7 x 10-5 mD and averages 2.0 x 10-5 mD with a median value of 4.5 x 10-6 mD (Table 4.2). The decane PDP values are 55 – 96 % lower than the He PDP value for all samples (Figure 4.15). The difference between the He PDP and decane PDP increases with an increasing contact angle of water in air (Figure 4.16).4.4.6 WettabilityThe contact angle of water in air varies from 8 degrees to 121 degrees in the sample suite and shows no systematic change with maturity. Contact angles for all samples show a strong positive correlation 69to TOC content (Figure 4.17) which masks maturity and other compositional effects. On a per well basis, the contact angle correlation to TOC varies from R2 = 0.77 to R2 = 0.84.4.5 DISCUSSIonThe effect of maturity on mudrock permeability is dependent on the evolution of the pore structure. To highlight these controls, the following sections discuss the influence of: (1) maturity and composition on mudrock microstructure; and (2) mudrock microstructure on permeability. A diagram simplifying the discussion concepts is given in Figure 4.18.4.5.1 Duvernay microstructure and maturityComparing the pore size distribution of samples of varying maturity demonstrates the impact of thermal maturation and burial (Figure 4.6). The progressive development of fine mesoporosity and microporosity with maturity is a dominant feature of the pore structure evolution within mudrocks. Secondarily, the pore structure undergoes a general loss of the macropore size fraction with progressive burial and compaction. The importance of compositional controls on the pore structure varies between maturity windows due to the diversity of porosity origins. Fine pore sizes (< 20 nm) have the largest control on surface area (Chalmers and Bustin, 2012a). Surface area increases with decreasing pore diameter as a power function of the diameter (assuming cylindrical or spherical pores). The lower analytical temperatures (77 K) of N2 LPGS may render N2 inaccessible to microporosity (Gan et al., 1972). Hence, N2 LPGS only records the external particle surface area and meso- to macroporosity. Since variations in surface area are dominantly controlled by fine pore sizes, N2 BET surface area can be used as a proxy for mesoporosity. Carbon dioxide, on the other hand, is able to access the finest microporosity due to higher analytical temperature (273 K). Therefore CO2 BET surface area can be used as a proxy for micro- and mesoporosity development. The narrow range of partial pressures measured by CO2 yields surface areas that are smaller than N2 surface areas despite accessing smaller pore sizes. Therefore, the CO2 BET surface area will not give an accurate value of the total surface area. The CO2 BET surface area is still useful for comparative purposes and elucidating pore structure change with maturity since all samples were measured over the same range of CO2 partial pressure. Nitrogen BET surface area shows a limited negative trend with maturity from the immature to wet gas window which abruptly increases in the dry gas window. Nitrogen BET surface area 70averages 5.9 m2/g within the immature, oil and wet gas windows, while the dry gas window samples average 18.0 m2/g (Figure 4.19). In contrast, CO2 BET surface area shows a progressive increase from the oil window to the dry gas window (Figure 4.20). Escalating CO2 BET surface areas with maturity is indicative of progressive fine meso- to microporosity generation, which is dominantly correlated to TOC content and to a lesser extent total clay content (e.g. Figure 4.9; Figure 4.10). The control of TOC on fine mesoporosity is evident in the pore size distribution curves for wet gas (Figure 4.9A) and dry gas window samples (Figure 4.10A). The control of total clay content on the mesoporous size fraction within wet gas window samples is unclear (Figure 4.9D) as the samples exhibit a narrow range of clay contents (10 – 25 %) with little variation. Within dry gas window samples, the control of total clay content on mesoporosity is moderate (Figure 4.10D) with clay contents varying from 1 to 27 %; however, the correlation is dually associated with increasing TOC contents. Mesopore volumes within dry gas window samples are significant even in samples with lesser clay contents (10 – 15 %). Carbon dioxide BET surface area is high (6 – 12 m2/g) within immature and early oil window samples due to greater clay contents and immature TOC, since TOC can be microporous at low maturities (Romero-Sarmiento et al., 2014). Upon entering the oil window, CO2 surface area decreases abruptly which may be related to oil or bitumen clogging pore space within the OM (Mastalerz et al., 2013). Upon further thermal maturation, CO2 BET surface area progressively increases as more gaseous hydrocarbons are generated, until CO2 BET surface area reaches a maxima in the dry gas window (Figure 4.20). The maxima in CO2 BET surface area within the dry gas window is coincident with the abrupt increase in N2 BET surface area, indicating development of significant volumes of fine meso- to microporosity accessible by both molecules. The combined relationships of progressively increasing CO2 BET surface area and N2 BET surface area suggests that pores within the OM have coalesced and have developed a pore structure in which mesoporosity constitutes a significantly larger percentage of the pore volume and therefore surface area, which is evident in the BJH pore size distribution curves (Figure 4.6). This is substantiated by FE-SEM imaging, where OM-hosted pores are seen to increase in both size and abundance systematically with maturity from the onset of the oil window (Figure 4.11; Figure 4.12; Figure 4.13; Figure 4.14).71 The trend of progressively increasing CO2 surface area with maturity and abruptly increasing N2 surface area within the gas window is best represented by averaging the BET surface area for wells where thermal maturation has reached the oil window and higher. Compositional controls (e.g. clay content) which could impact the surface area are included in order to isolate the effects of maturity (Figure 4.21). The surface area of lower maturity wells is controlled by higher total clay contents or immature TOC microporosity, unrelated to thermal diagenesis. As such, the pore structure correlations of these samples do not contribute to the understanding the role of thermal maturation and hydrocarbon generation and therefore have not been included in the correlations. The average CO2 surface area shows an excellent positive correlation with progressive maturation (R2 = 0.94, Figure 4.21A) and the average total pore volume as measured by CO2 (R2 = 0.83, Figure 4.21B). Total clay content does not exhibit a significant control on CO2 BET surface area since lower maturity (oil and wet gas) have average clay contents similar to dry gas samples while still exhibiting lower BET surface areas which demonstrates the limited role of clay content on CO2 BET surface area variation. Nitrogen BET surface area shows a poorer correlation to increasing maturation (R2 = 0.51, Figure 4.21A) and average total pore volume shows a limited correlation to N2 BET surface area (R2 = 0.14, Figure 4.21D). In addition to the development of fine meso- and microporosity with maturity, the pore volume of large pores within the coarse meso- to macropore size fraction undergoes a progressive decrease in volume with increasing maturity and burial depth due to compaction. Immature Duvernay mudrocks contain the greatest pore volume within the 20 – 70 nm pore mode, while mature mudrocks generally contain the least (Figure 4.6). The subsequent loss of larger pore modes is indicative of progressive burial and compaction. Present day burial depths are the shallowest for immature wells and progressively increase coincident with increasing thermal maturity (Table 4.1). This conclusion is supported by studies on mudrock compaction which found effective stress to control the loss of coarse mudrock pores with progressive compaction (Dewhurst et al., 1998, 1999; Yang and Aplin, 2007). A notable exception to this trend are samples from the HSK 10-33-56 well (Figure 4.6). This overmature well has higher coarse meso- and macropore volumes than other dry gas window samples, possibly due to lower burial depth than other dry gas window wells (4011 m vs 4594 and 5210 m, Table 4.1). Microstructural imaging by FE-SEM for this well reveal significant 72volumes of OM-hosted coarse meso- and macroporosity, which contributes to the increase in macropore volumes measured by N2 LPGS (Figure 4.14).4.5.2 Impact	of	maturity	on	permeabilityMatrix permeability within tight reservoirs is a function of multiple variables such as composition, texture, thermal maturity and compaction which have been shown to control the pore size distribution. Determining precise correlations between these variables and permeability is therefore challenging as the mode of porosity occurrence with mineralogy and organic content changes between samples and with maturity due to numerous factors, including new mineral growth, compaction, and re-orientation of grains. However, more general correlations are possible which highlight the dominant maturity trends and compositional components which influence the pore size distribution and permeability. Compaction tests on mudstones show porosity loss is dominated by the collapse of larger pores, with little effect to smaller pores and a concomitant decrease in permeability (i.e. Dewhurst et al., 1998). Permeability in mudstones has been found to be controlled by the preferential fluid flow through the largest pore sizes, which is supported by the proportional relationship between fluid flow through a cylindrical pipe with the fourth power of the tube radius (Dewhurst et al., 1998), decreasing permeability with increasing effective stress is consistent with the collapse of larger pores (Dewhurst et al., 1998). The extent to which the development of the fine meso- to micropore size fraction with maturity and the impact on permeability can be examined based on previous findings suggesting progressive burial should decrease permeability. Within the oil window the control of pore size distribution on matrix permeability are not clear (Figure 4.22). The highest permeability is generally associated with the greatest TOC (> 4.0 %) and greatest total clay content (> 30 %) (Figure 4.23A, D). The oil window sample with the lowest permeability is associated with moderate carbonate contents (22 %), moderate quartz and feldspar content (59 %), lower clay contents (17 %), and moderate TOC content (3 %). These relationships indicate that within oil window samples, where fine-mesoporosity is less developed than higher maturity windows, clay-hosted mesoporosity along with OM-hosted mesoporosity may contribute significantly to matrix permeability. The control of total quartz and feldspar and total carbonate content on permeability is less clear. Moderate permeability samples can have varying 73carbonate contents (15 – 24 %), varying quartz contents (55 – 70 %) and moderate TOC contents (Figure 4.23A, B, C). The pore size distribution controls on permeability within the wet gas window samples are more distinct than oil window samples (Figure 4.24). Highest permeability samples have the greatest BJH pore volumes in the 10-200 nm pore size range (coarse meso- to macropores, Figure 4.24). The wet gas window sample with the lowest permeability has a moderate amount of coarse mesopores and a low volume of fine mesopores which indicates that in the absence of high volumes of fine mesopores, permeability is lower (as seen in oil window samples). The highest wet gas window permeability samples, which have high volumes of coarse meso- to macropores, can have low or high volumes of fine mesopores. These relationships suggest that the coarse meso- to macropore size fraction is the dominant control on permeability. However, the fine mesopore size fraction may contribute significantly to permeability because wet gas window samples have higher permeabilities overall than oil window samples, despite having lesser volumes of coarse meso- to macropores. The key pore structure change which enables higher permeabilities within wet gas window samples is the greater volumes of fine mesopores. The BJH pore size distribution curves for wet gas window samples show excellent correlation with TOC, especially in the < 10 nm size fraction (Figure 4.9A). The volume of fine mesopores increases with increasing TOC contents. Since fine pores may increase permeability, this suggests that the mesoporous network developed with thermal maturity and correlated to OM content is sufficiently connected to contribute to matrix permeability. Coarse meso- to macropores correlate with relatively higher quartz content (> 55 %) and lower carbonate contents (< 25 %) (Figure 4.25B, C). Higher permeability associated with higher volumes of coarse meso- to macropores may be associated with detrital quartz grains. Alternatively, low carbonate contents could be associated with limited diagenetic carbonate cements which would destroy pore volume. Coarse meso- to macropores were imaged within wet gas window samples associated with quartz grains (Figure 4.13A). Total clay content does not vary significantly within wet gas window samples and shows little variation between low and high permeability samples (Figure 4.25D). This indicates that the clay microstructure may not be a controlling factor on matrix permeability within wet gas window samples.74 The fact that matrix permeability correlates with greater volumes of coarse meso- to macropores within wet gas windows is significant since LPGS is an unstressed analysis, while PDP measurements were measured at 5,000 psi (34.5 MPa) net effective stress. A relationship between larger pore sizes (unstressed) and increased permeability (stressed) indicates that while coarse meso- to macropores may undergo some compression once stress is applied for PDP analysis, the majority of these pores remain open and contribute to matrix permeability. The dry gas window samples show similar permeability trends to the wet gas window samples. Samples with higher BJH pore volumes overall have higher permeability (Figure 4.26). All samples show similar development in the fine mesopore size fraction, so the impact of fine mesopores on permeability is difficult to separate from increased overall total mesopore volume. High permeability samples have varying TOC contents, higher quartz and feldspar contents, lower carbonate contents, and higher total clay contents (Figure 4.27). Conversely, low permeability samples have relatively higher carbonate contents, lower clay contents, and lower quartz and feldspar contents (Figure 4.27) Dry gas window samples have average permeabilities similar to wet gas window samples despite having significantly less coarse meso- to macropore volumes than oil and wet gas window samples. Similar to wet gas window samples, higher average permeability coincident with lower volumes of coarse meso- to macropores in dry gas window samples indicates that the fine mesoporosity developed with thermal maturity provides an interconnected network which contributes to gas flux. The significance of mesoporosity development with thermal maturity on mudrock matrix permeability is clear. Coarse meso- to macroporosity is a fundamental control on matrix permeabilities, whereas fine mesoporosity becomes progressively more important to gas flux as maturity increases. Permeability is enhanced within wet gas window and dry gas window samples compared to oil window samples despite the shift to smaller modal pore sizes with thermal maturity and burial.4.5.3 GEPP permeability and microstructureThe pore size distribution curves colored by helium PDP demonstrates the significant control of larger pore sizes on gas flux through the mudrock matrix. Assessing the impact of finer pore sizes on 75matrix permeability is difficult using PDP measurements; permeability is biased toward larger pore sizes, as seen in the wet gas window samples. Assessing the impact of fine pore sizes is important for determining total gas flow through the rock matrix. Finer pore sizes were shown in this study to develop with maturity and to constitute a significant portion of the total pore volume. Surface area is also correlated to smaller pore sizes which provide the majority of adsorbed gas storage sites (Ross and Bustin, 2009). Samples were chosen with similar PDP permeabilities but various compositional properties to investigate the controls on GEPP permeability (Table 4.2). Utilizing samples with similar PDP permeabilities will highlight the control of fine pore sizes on the GEPP permeability, as larger pore sizes could provide interconnected pathways between fine pores, increasing the total gas flux in GEPP measurements and therefore apparent fine pore permeability. The impact of pore size distribution on matrix permeability as measured by GEPP is shown in Figure 4.28. The oil window sample has the smallest volume of fine pore sizes and the lowest GEPP permeability. In contrast, the wet gas window and dry gas window samples have the greatest volumes of fine pore sizes and the highest GEPP permeabilities. A sample with poorly interconnected fine pores is anticipated to yield a relatively low GEPP permeability. On the other hand, a sample with well-connected fine pores would yield a relatively higher GEPP permeability. Since fine pore sizes are associated with TOC and have been shown to develop with maturity, higher GEPP permeabilities associated with overmature rocks are indicative of increased connectivity of OM-hosted fine pores in comparison to lower maturity samples (Figure 4.28). The dry gas window samples have significantly greater fine pore volumes than the oil window sample and are coincident with higher GEPP permeabilities. Therefore, the fine pore sizes developed with maturity and associated with TOC become increasingly connected and provide significant pathways for gas flux through the matrix. Comparing the oil window (ATH 13-18-64) and dry gas window (HSK 7-9-37) samples highlight the increasing connectivity of fine pore sizes with maturity. The PDP for the oil window sample is 28 % higher than the dry gas window sample (1.5 x 10-5 versus 1.0 x 10-5 mD, respectively; Table 4.2). The GEPP permeability for the dry gas window sample, however, is over 40 % greater than the oil window sample, varying from 6.1 x10-7 mD for the oil window sample to 8.5 x 10-7 mD for the dry gas window sample. Increased GEPP permeability in the dry gas window sample 76compared to the oil window indicates that even though the PDP for the oil window is greater than the dry gas window sample, the higher volume of fine pore sizes form interconnected permeability pathways which yields a higher GEPP permeability for the dry gas sample. The wet gas sample has significant volumes of fine pore sizes and almost twice the TOC content compared to overmature samples (Table 4.2). However, the GEPP permeability of the wet gas sample is similar to the GEPP permeabilities measured in overmature samples. This relationship demonstrates that similar volumes of fine pores yield similar GEPP permeabilities and the fine pores which have developed within the wet gas sample show similar connectivity to the dry gas window samples. It would be reasonable to expect that upon further maturation of the wet gas sample significant volumes of fine pore sizes would develop and subsequently increase the GEPP permeability. The effect of total clay content is difficult to separate from the fine porosity development with maturity correlated to OM content. Duvernay samples have low total clay contents and often have minor variability between samples (e.g. Table 4.2). The dry gas window samples selected for GEPP, however, are amenable to compositional comparison as these samples have a similar fine pore structure (Figure 4.28) and TOC content, while total clay content varies from 9 % for the HSK 7-9-37 sample to 27 % for the SCL 11-1-38 sample (Table 4.2). The N2 BET surface area for the HSK 7-9-37 and SCL 11-1-38 sample is 14.6 and 19.4 m2/g, respectively, which corresponds to the increased total clay content. The GEPP permeability for the HSK 7-9-37 sample is 8.5 x 10-7 mD while the GEPP permeability for the SCL 11-1-38 sample is 7.3 x 10-7 mD, approximately 15 % less. Thus, in the absence of other microstructural factors, fine pore sizes associated with clay minerals may be more poorly connected than OM-hosted porosity. As all samples are natural samples other variations in the microstructural texture which are not distinctly captured within the N2 LPGS pore size distributions and compositional data may impact GEPP permeability measurements, such as clay alignment or flocculation, the spatial distribution of OM within the matrix and the spatial relation of silt grains respective to clay-sized grains.4.5.4 Wettability,	liquid	permeability	and	pore	structurePermeability to liquids for tight reservoirs is an important parameter which is seldom reported within the literature due to measurement difficulties. Tight reservoirs commonly span multiple thermal 77maturity boundaries with the capability to produce liquids, wet gas, or dry gas (e.g. the Duvernay or the Eagle Ford Formation, Texas). Depending on the reservoir to wellbore pressure gradient, condensate may occur within the fracture network and possibly the matrix as hydrocarbons are produced (condensate banking, Devegowda et al., 2012). Wells within the oil window will have liquid hydrocarbons within the pore system, and depending on the pore system wettability and fluid saturation (oil and water content), relative permeability can vary significantly. If exploration moves to target reservoirs of lower thermal maturity, identifying the influence of different pore systems (OM-hosted versus mineral-hosted) on relative permeability will have implications for production. In systems where multiple fluids are present, the wettability of the pore system will affect relative permeability by controlling the distribution and flow of each fluid at a given saturation (Anderson, 1987). In general, in a system of uniform wettability, the wetting phase will occupy the smallest pores and form a coating on all pore surfaces while the non-wetting phase will occupy the centers of larger pores (Anderson, 1987). Therefore, the relative permeability of the non-wetting phase in the presence of a wetting phase is higher at a given saturation because the non-wetting phase occupies the centers of the larger, more permeable pore throats. In contrast, the wetting phase occupies the smaller pores and coats grains. In a system with a single non-wetting phase or at low wetting phase saturations, the relative permeability should approach the absolute permeability (Anderson, 1987). The capillary pressure required for a fluid to enter a pore is dependent on the pore radius, surface tension of the fluid and contact angle of the fluid with the pore body, exemplified by the Washburn equation for ideal cylindrical capillaries (Washburn, 1921). Pores may preferentially contribute to flow depending on the capillary pressure and wettability. It may be possible to assess the contribution of certain pore types to the permeability of liquids. The helium PDP measurements can be used to approximate the absolute permeability of the sample since gas is the non-wetting phase. Helium was the only flowing phase since the helium PDP measurements were performed on dried plugs (i.e. no water or oil saturations). The liquid permeability measured using decane as the probe fluid, which has a varying wettability with the pore structure, will vary depending on the interaction of the fluid with the pores contributing to flow. The contact angle of water in air provides useful qualitative information about the pore 78system wettability (Figure 4.17). The OM is observed to be strongly hydrophobic, since the contact angle of water in air increases systematically with increasing TOC (Figure 4.17). It is reasonable to assume that the OM may be oil-wet, but at very least it is not water-wet. The intercept of best fit line in Figure 4.17, representing zero TOC content, is approximately 14 degrees and indicates that the mineral-hosted pore system is strongly water-wet based on the limited data set. The decane PDP values show no direct correlation to sample composition or the contact angle of water in air. The lack of correlation is attributed to the complex and heterogeneous nature of the pore structure. However, the difference between the helium PDP and decane PDP permeability value decreases with decreasing contact angle (i.e. decane PDP approaches the absolute helium permeability, Figure 4.16). In a system where the pore structure is dominated by fine, OM-hosted pores, decane is the wetting phase and forms a coating on all pore surfaces which may restrict flow. The OM-hosted pore system may be isolated or discontinuous to varying degrees depending on the texture and spatial relationship with the mineral pore system. Therefore, a trend of decane PDP approaching the absolute helium PDP with a decreasing hydrophobicity suggests that OM-hosted pores may not be contributing significantly to decane flux. Since helium and decane are both non-wetting within mineral-hosted pore systems, the decane permeability approaches the absolute permeability as the pore system becomes increasingly water-wet. Decane is likely occupying the centers of the largest, mineral-hosted pores where it can more effectively contribute to flow. In the case of isolated OM-pores surrounded by larger mineral pores, the decane may preferentially flow through the mineral-hosted pores which leads to the poor correlation of TOC with the decane PDP value (R2 = 0.04).4.6 ConClUSIonSThe comparison of pore structures from samples of varying maturity provides insight into the impact of thermal maturity and compaction on mudrock microstructure and permeability. Significant conclusions from this study include:• Organic-matter hosted fine mesoporosity is visible with FE-SEM imaging from the onset of the oil window (HI = 348 mg HC/g TOC) and the development of OM-hosted pores increases in both size and abundance systematically with increasing maturity into the dry gas window.• The pore size distribution as measured by N2 LPGS shows a progressive increase in fine pore 79sizes with thermal maturity from the oil window through the dry gas window and the subsequent loss of larger pore modes. Macropore volumes are highest within the shallow, immature samples and progressively decrease in volume with increasing maturity and burial due to compaction.• Average helium pulse-decay permeabilities for samples from the wet gas and dry gas window are an order of magnitude higher than oil window samples. In general, this is a result of the development of fine pore sizes with maturity associated with OM, indicating OM-hosted pores are sufficiently interconnected to provide increased matrix permeability.• Within the wet gas window, helium PDP is dominantly controlled by the coarse meso- to macropore size fraction which is associated with increased total quartz and feldspar content. The helium PDP for dry gas window samples is controlled by higher fine pore volumes correlated to relatively higher total clay and quartz and feldspar content.• The GEPP permeability, which probes the total gas uptake by the matrix and is more suited to measuring the permeability of finer pore sizes, reveals the fine pore sizes are sufficiently interconnected to yield higher GEPP permeabilities. Oil window samples have lower GEPP permeabilities while wet gas and dry gas window samples have higher GEPP permeabilities and higher fine pore volumes.• The contact angle of water in air increases with increasing OM content since the OM (and OM-hosted pores) is hydrophobic and mineral surfaces are hydrophilic.• The decane PDP is lower than the helium PDP which is interpreted to be due to wettability variations within the pore structure. The decane PDP value shows no direct correlation to composition due to the complex nature of the pore system, which is composed of varying pore sizes, spatial distributions and pore compositions which have varying wettability.800.1 nm1 nm10 nm100 nmMacroporeMesoporeCoarseFineMicroporeFigure 4.1 Pore size classification scheme used in this study. Modified from Sing et al. (1985).81BRITISHCOLUMBIA ALBERTACANADA0 25 50kmCI: 0.40 % VRo NBRITISH COLUMBIAALBERTADeformation Fronto117 00’o118 00’o116 00’o115 00’o113 00’o114 00’0.600.600.601.001.001.401.401.801.802.202.202.402.40HSK 10-25-59SCL 4-7-59HSK 10-16-61LRE 4-34-77MQL 11-18-72HSK 10-33-56ATH 13-18-64TET 10-18-64ECA 11-8-62HSK 8-25-60HSK 5-11-60YO 14-16-62ATH 1-24-61ATH 4-2-62FB 10-4-51PWT 10-17-45TLM 12-32-41HSK 7-9-37SCL 11-1-38o55 00’o54 00’o52 00’o53 00’EdmontonFigure 4.2 Study location in central Alberta. Samples were taken from 19 wells total. Leduc Fm. reefs are light grey. Vitrinite reflectance lines are modified from Stasiuk and Fowler (2002). Well identifiers are given in Table 4.1.82UWI Well ID SSTVD Avg. HI Avg. OI Avg. LPGS Perm Maturitym mg HC/g TOC mg CO2/g TOC Tmax oC samples plugs100/10-25-059-19W4/00 HSK 10-25-59 360 500.8 24.3 417 4 - Immature100/10-16-061-26W4/00 HSK 10-16-61 753 508.5 17.5 419 4 - Immature100/04-07-059-20W4/00 SCL 4-7-59 450 535.3 26.5 419 4 - Immature100/11-18-072-17W5/00 MQL 11-18-72 1636 319.0 24.0 433 3 - Early oil100/04-34-077-23W5/00 LRE 4-34-77 1823 247.5 55.5 438 2 - Early oil100/10-04-051-27W4/00 FB 10-4-51 1189 348.0 4.0 444 1 - Early oil100/13-18-064-17W5/00 ATH 13-18-64 1971 128.4 16.7 445 11 8 Oil104/10-18-064-19W5/00 TET 10-18-64 2133 62.9 12.1 464 13 - Wet gas100/05-11-060-18W5/00 HSK 5-11-60 2190 68.4 7.6 467 - - Wet gas103/08-25-060-18W5/00 HSK 8-25-60 2139 75.1 6.9 467 5 - Wet gas100/12-32-041-05W5/00 TLM 12-32-41 2114 56.7 11.9 468 4 - Wet gas102/11-08-062-24W5/00 ECA 11-8-62 2717 75.4 6.8 469 13 12 Wet gas100/14-16-062-21W5/00 YO 14-16-62 2404 59.3 9.4 470 - - Wet gas100/10-17-045-06W5/00 PWT 10-17-45 2005 38.5 17.9 471 - - Wet gas100/13-02-062-23W5/00 ATH 4-2-62 2584 48.3 7.3 474 - - Wet gas100/01-24-061-23W5/00 ATH 1-24-61 2581 23.2 11.2 482 5 - Wet gas100/07-09-037-10W5/00 SCL 11-1-38 3674 11.3 16.3 - 7 3 Dry gas100/11-01-038-13W5/00 HSK 10-33-56 2844 10.6 9.8 - 4 - Dry gas100/10-33-056-22W5/00 HSK 7-9-37 3197 10.5 9.5 - 2 2 Dry gasTable 4.1 Well location and averaged data for wells in the study. UWI = unique well identifier, Well ID is the identifier used in this study. SSTVD = sub-surface true vertical depth to the top of the Duvernay. HI = hydrogen index, calculated as S2*TOC/100. OI = oxygen index, calculated as S3*TOC/100. Average Tmax is not valid and therefore omitted for overmature samples. See text for discussion of qualitative maturity descriptor.830 20 40 60 80 100100806040200100806040200Qtz + Feld (%)Clay (%) Carbonate (%)N = 78Figure 4.3 Bulk mineral composition for LPGS samples normalized to total quartz and feldspar content, total clay content, and total carbonate content.840.0100.0200.0300.0400.0500.0600.0700.0800.0900.00.0 50.0 100.0 150.0Hydrogen Index (mg HC/g TOC)Oxygen Index (mg CO2/g TOC)HSK 7-9-37SCL 11-1-38HSK 10-33-56ATH 1-24-61PWT 10-17-45ATH 4-2-62TLM 12-32-41YO 14-16-62TET 10-18-64HSK 5-11-60HSK 8-25-60ECA 11-8-62ATH 13-18-64LRE 4-34-77MQL 11-18-72FB 10-4-51HSK 10-25-59HSK 10-16-61SCL 4-7-59IIIIIIFigure 4.4 Modified Van Krevelan diagram (Tissot and Welte, 1984) for all wells analyzed in the study. Duvernay kerogen is dominantly Type II.85Figure 4.5 Typical N2 (A) and CO2 (B) isotherm for Duvernay samples within this study.86Figure 4.6 Nitrogen BJH pore size distribution for Duvernay samples by maturity level. (A) immature wells; (B) oil window ATH 13-18-64; (C) wet gas well ECA 11-8-26; (D) dry gas window wells. Samples from dry gas maturity generally have the least volume or coarse mesopores and macropores with the exception of HSK 10-33-56.87Figure 4.7 Nitrogen BJH pore body distributions for immature wells colored by (A) TOC, (B) total quartz and feldspar content, (C) total carbonate content, and (D) total clay content. Note y-axis scale for immature samples is expanded compared to other maturity groups.88Figure 4.8 Nitrogen BJH pore body distributions for oil window well ATH 13-18-64 colored by (A) TOC, (B) total quartz and feldspar content, (C) total carbonate content, and (D) total clay content.89Figure 4.9 Nitrogen BJH pore body distributions for wet gas window well ECA 11-8-62 colored by (A) TOC, (B) total quartz and feldspar content, (C) total carbonate content, and (D) total clay content.90Figure 4.10 Nitrogen BJH pore body distributions for dry gas window wells colored by (A) TOC, (B) total quartz and feldspar content, (C) total carbonate content, and (D) total clay content.911 μmPanel F200 nm1 μm10 nmBC DE F1 μmPanel CAfine mesopore2 nm microporeclayclayOMclayfine mesoporescarbqtzOMmacroporecarbclayqtz1 μmqtzcarbclayFigure 4.11 Texture and porosity associations from FE-SEM for early oil window well FB 10-4-51. Organic matter is generally non-porous, however fine mesopores (C) and coarse micropores (D) were imaged in two separate locations. Macropores exist in crack-like distributions within or at the boundaries of OM particles (D). Intra-particle clay hosted porosity is also evident (E).921 μm 500 nm200 nmAPanel BPanel DBCmesoporescoarse mesoporesdolo200 nmDPanel CFigure 4.12 Organic matter-hosted porosity for oil window well ATH 13-18-64. Inter-crystalline porosity was found to be limited in two imaging locations. Organic-matter is both displays varying porosity, with pores generally within the mesopore size fraction.93Figure 4.13 Porosity and texture for wet gas window well ECA 11-8-62. The macropore in (A) and (B) is the largest pore observed in the study. Variable non-porous OM and porous OM was observed in multiple imaging locations (A, E).ED1 μm500 nm200 nm500 nm500 nmpyrPanel BOMcarbcarbcarbmacroporemesoporesmesoporesA BC200 nmPanel D50 nm mesoporeF94500 nm200 nm200 nm200 nm 200 nmA BC DEmacroporesmesoporesmacroporesPanel Dclaycoarse mesoporesmacroporeFigure 4.14 Porosity and texture for dry gas window well HSK 10-33-56. Organic matter is meso- to macroporous (A, E) and exhibits the largest OM pores of all wells. Development of OM porosity is less variable than other maturity windows. Inter-crystalline porosity is associated with clay minerals (D) and carbonate grains (C).95Sample Permeability (mD) Composition (%) N2 BET CO2 BETWell No. He PDP Decane PDP GEPP Qtz+Fld Carb Clay TOC m2/g m2/gATH 13-18-64 7 8.6E-6 2.3E-6 - 60 19 17 2.8 - -11 1.5E-5 5.1E-6 6.1E-7 58 24 16 2.8 4.0 2.712 6.7E-6 - - 59 22 17 3.2 5.9 3.216 5.6E-5 - - 46 18 32 4.3 5.9 3.817 2.3E-6 - - 62 14 21 2.8 - -24 1.1E-5 - - 70 15 12 3.1 3.3 1.9ECA 11-8-62 3 1.2E-5 - - 51 22 25 3.4 - -5 6.1E-6 - - 48 27 23 3.1 4.3 2.78 1.4E-4 - - 46 40 12 3.6 5.7 2.619 3.9E-5 - - 53 28 14 6.8 8.7 -21 6.4E-5 2.6E-6 8.5E-7 53 27 16 8.3 10.3 6.223 4.9E-4 5.1E-5 - 58 17 18 5.2 11.0 5.124 3.5E-4 - - 58 18 22 5.6 10.5 5.831 5.3E-4 - - 50 26 17 2.6 5.0 -33 2.9E-5 3.8E-6 - 38 40 19 2.8 4.0 -36 1.3E-4 5.7E-5 - 31 43 23 4.3 9.8 5.3SCL 11-1-38 5 3.4E-4 - - 42 29 26 4.7 18.0 7.07 7.8E-5 - 7.3E-7 40 30 27 4.4 19.4 7.0HSK 7-9-37 1 1.0E-5 - 8.5E-7 25 63 9 4.3 14.6 5.92 1.0E-5 - - 34 50 13 5.8 20.0 8.4Table 4.2 Samples and properties selected for PDP and GEPP analysis.96Figure 4.15 Helium PDP versus decane PDP, performed on the on the same sample, from the oil window (ATH 13-18-64) and wet gas window (ECA 11-8-62).97Figure 4.16 The contact angle of water in air versus the difference between helium PDP and decane PDP. Black line is a power fit.98Figure 4.17 Contact angle of water in air on Duvernay mudrocks vs TOC content.99Figure 4.18 Evolution of mudrock microstructure for Duvernay samples analyzed in this study. Red circles are average value for the well. Levels of thermal maturation adapted from Peters (1986) and Peters and Cassa (1994). Organic matter and interparticle porosity abundance is estimated from FE-SEM imaging and correlations of TOC with LPGS values (see text for discussion). Volume of pore sizes cc/g0.01 0.02 0.03Fine mesopores/microporesCoarse mesoporesMacroporesVolume of pore sizes cc/g0.01 0.02VRo%~TmaxType I/II thermal maturity Stage WindowPorosity typeabundance2LPGS BET (m /g)N2CO2Permeability nD0.20.61.351.1435ImmatureMatureOvermatureImmatureOilEarlyCondensateWet gasThermogenicdry gasBiogenic gas4454804604400 20 0 10 + 1 10 100 1000SamplesCompactionThermal diagenesisOM-hostedInterparticle100Figure 4.19 Box and whisker plot of N2 BET surface area for all wells, sorted by increasing maturity. The fliers are data range, upper and lower limits of the box are 1st and 3rd quartiles, red line is the median and red square is the average. The wide range of values (e.g. SCL 11-1-38) shows the significant sedimentary variability.101Figure 4.20 Box and whisker plot of CO2 BET surface area for all wells, sorted by increasing maturity. The fliers are data range, upper and lower limits of the box are 1st and 3rd quartiles, red line is the median and red square is the average.102Figure 4.21 Average CO2 BET surface area (A) and average CO2 total pore volume (B) versus HI colored by average total clay content. Average N2 BET surface area (C) and average N2 total pore volume (D) colored by average clay content. Only samples that have undergone significant thermal maturity (oil window and higher) are shown to highly effects of maturity.103Figure 4.22 Nitrogen BJH pore body distribution for oil window ATH 13-18-62 colored by He PDP.104Figure 4.23 Nitrogen BJH pore body distributions only for samples with He PDP measurements from the oil window well ATH 13-18-64, colored by (A) TOC, (B) total quartz and feldspar content, (C) total carbonate content, and (D) total clay content.105Figure 4.24 Nitrogen BJH pore body distribution for wet gas window well ECA 11-8-62 colored by He PDP.106Figure 4.25 Nitrogen BJH pore body distributions only for samples with He PDP measurements from the wet gas window well ECA 11-8-62, colored by (A) TOC, (B) total quartz and feldspar content, (C) total carbonate content, and (D) total clay content.107Figure 4.26 Nitrogen BJH pore body distribution for dry gas window wells colored by He PDP.108Figure 4.27 Nitrogen BJH pore body distributions only for samples with He PDP measurements from the dry gas window wells HSK 10-33-56 and SCL 11-1-38, colored by (A) TOC, (B) total quartz and feldspar content, (C) total carbonate content, and (D) total clay content.109Figure 4.28 Pore size distributions colored by GEPP permeability.110Chapter 5: Regional	reservoir	characterization	model	for	the	shale	gas	and	shale	oil	producing	Duvernay	Formation,	alberta,	Part	I:	Regional	reservoir	distribution	and	reservoir	properties	using	wireline	log	signatures	and	high-resolution laboratory data5.1 InTRoDUCTIonThe delineation of potential fracture barriers, stratigraphic and areal zones of organic richness, net reservoir thickness distributions and associations of porosity with mineralogy, total organic carbon (TOC) content, and texture are all paramount to understanding tight reservoir development, risk assessment and high-grading existing assets. Prospective formations need to be examined on an individual basis due to the diverse set of lithologies which can act as fine-grained reservoirs (Chalmers et al., 2012b), as well as more subtle changes in mineralogy, organic content, texture, and maturity within an individual reservoir (Ross and Bustin, 2008). Within fine-grained reservoirs, vertical variability can exceed lateral variability due to the heterogeneity which is known to exist. Therefore, extensive datasets from multiple wells are required to determine the vertical and lateral variation in reservoir properties. Previous multi-disciplinary reservoir evaluation studies include those of the Devonian Barnett Formation (Montgomery et al., 2005; Jarvie et al., 2007), the Cretaceous Buckinghorse Formation (Chalmers and Bustin, 2008a, 2008b), the Cretaceous Shaftesbury Formation (Chalmers and Bustin, 2012b), the Cretaceous Pearsall Formation (Hackley, 2012), the Devonian Horn River Formation, (Ross and Bustin, 2008; Chalmers et al., 2012b), the Haynesville and Bossier Shale (Hammes et al., 2011; Hammes and Frébourg, 2012), the Jurassic Gordondale Formation (Ross and Bustin, 2007), the Woodford Shale (Romero and Philp, 2012), and the Halfway-Doig-Montney hybrid gas shale-tight gas reservoir (Chalmers and Bustin, 2012a) among others. These studies provide a background to compare Duvernay reservoir properties and areal distributions in the context of global fine-grained reservoir development. Current industry interest, drilling and production activity confirm the Late Devonian Duvernay Formation of Alberta contains significant hydrocarbon saturations (Smith Low, 2012; Smith Low et al., 2013) and is capable of producing hydrocarbons over multiple thermal maturity boundaries as a “shale” gas and “shale” oil reservoir. To date, no published regional reservoir 111characterization studies of the Duvernay exist. The unique depositional environment of the Duvernay compared to other mudrock reservoirs provides an opportunity to investigate the localized control of reef development on the distribution of organic-rich mudstones and their reservoir properties within a broader sequence of basinal mudstones. Fresh core is available for sampling due to recent industry interest. The present study is divided into two parts. In Part I, aspects of core and wireline log analysis are combined to determine the regional distribution of reservoir lithologies for the basin, including the examination of mineralogy, total porosity, organic matter (OM) content, net reservoir thickness, potential fracture barriers or baffles, reservoir quality impairments, and the impact of reef development on reservoir quality. In Part II, artificial neural network models are used to integrate laboratory data and wireline logs suites to extend and quantify reservoir properties areally.5.2 REGIonAl GEoloGIC BACKGRoUnD5.2.1 overviewThe sedimentology of the Late Devonian Duvernay Formation and sediments of the Woodbend Group (Figure 5.1) in the Western Canada Sedimentary Basin (WCSB) have been the subject of numerous publications due to the importance of the Leduc Formation as a conventional hydrocarbon reservoir and the Duvernay as an important petroleum source rock (Imperial, 1950; Andrichuk, 1958a, 1958b, 1961; Oliver and Cowper, 1965; Campbell and Oliver, 1968; Stoakes, 1980; Stoakes and Wendte, 1987; Switzer et al., 1994; Weissenberger, 1994; Wendte, 1994; Chow et al., 1995). The Duvernay is considered to have generated most of the hydrocarbons found within the Leduc reefs and other Devonian age accumulations (Stoakes and Creaney, 1984, Creaney and Allan, 1990; Allan and Creaney, 1991; Creaney et al., 1994; Fowler et al., 2001). During the Late Proterozoic to Late Jurassic, the WCSB existed as a platformal, passive margin on the western edge of the North American craton (Price, 1994). During Woodbend time in central Alberta, increasing accommodation space within the WCSB led to a series of basin filling mudstones (shales) and limestones being deposited concurrently with Leduc reefs (Stoakes, 1980; Stoakes, 1992). The Duvernay is a prolific source rock deposited during the initial part of the Woodbend Group basin-filling interval. Eventually, mudstones of the Ireton filled the basin and covered the reefs (Stoakes, 1980). Multiple factors influenced the relationship between basin filling 112shales and reef development, including local basement fault block movements, removal of earlier salt deposits, varying differential subsidence and the volume of clastic sediments deposited in the basin (Switzer et al., 1994). Therefore, the relationships between basin filling mudstones and reefs or carbonate shelf evolution are complex. The Woodbend Group was subsequently buried as a result of the Laramide deformation to depths where temperatures and pressures were sufficient to generate thermogenic hydrocarbons within the Duvernay (Switzer et al., 1994). The Duvernay dips to the southwest as a result of Laramide deformation (Switzer et al., 1994). Major lineaments and landmasses during Duvernay deposition include the Peace River-Athabasca Arch, an emergent landmass during Duvernay time, and the Rimbey Arc, a basement structure which underlies part of the Rimbey-Meadowbrook reef trend dividing the East and West Shale Basins (Switzer et al., 1994; Ross and Stephenson, 1989). The western boundary of the WCSB is difficult to delineate due to accretion of younger terranes (Switzer et al., 1994), but it is generally placed at the eastern-most extent of the deformation belt of the ancestral North American margin (Wright et al., 1994).5.2.2 Depositional environmentThe deposition of Woodbend group sediments marks a change in provenance compared to underlying sediments of the Beaverhill Lake Group. During Beaverhill Lake time, an interbasinal source of carbonate mud was derived from carbonate banks in the east (Stoakes, 1992). In comparison, the Woodbend Group contains a significant amount of terrigenous clay derived from emergent landmasses to the north as well as an interbasinal carbonate component (Stoakes, 1980; Stoakes, 1992). The deposition of organic-rich mudstones of the Duvernay versus organic-lean calcareous mudstones of the overlying Ireton is intricately linked to Cooking Lake platform highs, Leduc reef inception, vertical aggradation and sea level fluctuations (Stoakes, 1980; Stoakes, 1992; Wendte, 1994). Variations in sea-level, basinal currents, terrigenous source component, interbasinal carbonate source component and reef growth have been correlated to the Leduc, Duvernay and Ireton formations (Stoakes, 1980). The interplay of these components govern intervals of most prospective reservoir, high net pay, organic richness and fracture barriers within the Duvernay. The WCSB during Woodbend time has been informally divided into the “East Shale Basin” and “West Shale Basin” in relation to the Rimbey-Meadowbrook reef trend (Belyea, 1964; Stoakes, 1131980; Switzer et al., 1994; Figure 5.2). Previous studies have focused on the Woodbend Group within the East Shale Basin due to core control and the early exploration in this area (Imperial, 1950; Andrichuk, 1958a, 1958b; Stoakes, 1980; Wendte, 1994).Within the East Shale Basin, the Duvernay is underlain by platform carbonates of the Cooking Lake Formation. The Cooking Lake platform records three major incremental sea level rises which gave way to variable platform topography due to selective colonization or accretion on pre-existing highs, which became nucleation sites for future Leduc reef growth (Stoakes, 1992; Wendte, 1994). The western edge of the Cooking Lake platform and East Shale Basin is marked by the Rimbey-Meadowbrook reef trend, whose location may be correlated with basement-scale features (Ross and Stephenson, 1989; Switzer et al., 1994).Within the West Shale Basin, the Duvernay is generally underlain by the Majeau Lake Formation which consists of green-gray and dark brown shales (Stoakes, 1980) and are clay-rich and organic-lean compared to the overlying Duvernay. The Majeau Lake is considered the basinal equivalent of the Cooking Lake platform developed in the east (Stoakes, 1980). The Majeau Lake exceeds 100 m in thickness to the northeast and thins north and westward to a depositional zero edge (Switzer et al., 1994). Leduc reefs and other carbonate banks in the West Shale Basin overlie an areally widespread Leduc platform or rest directly upon Swan Hills carbonate banks of the Beaverhill Lake Group, as opposed to directly upon the Majeau Lake (Stoakes, 1992). Sea level rise, which gave way to localized topographic highs on the Cooking Lake platform, also accounts for backstepping and aggradation of the lower Leduc carbonate platform in the West Shale Basin (Stoakes, 1992).Deposition of organic -rich mudstones of the Duvernay has been attributed to an initial deepening of the basin during Cooking Lake time (Stoakes, 1980; Switzer et al., 1994). Leduc reefs grew vertically from topographic highs and restricted basinal circulation which led to depletion of oxygenated waters (Andrichuk and Wonfor, 1954; Stoakes, 1980). The development of dysoxic to anoxic bottom waters (anoxic herein) due to restricted circulation resulted in enhanced preservation of organic material within Duvernay sediments (Stoakes, 1980; Stoakes and Creaney, 1984; Chow et al., 1995). In shallower reef-proximal locations, organic matter content is lower due to more oxygenated conditions from increased wave agitation as well as increased sediment dilution derived 114from the reefs (Chow et al., 1995). Near Leduc reef complexes, the Duvernay can thicken to over 100 m due to input of Leduc reef detritus (Switzer et al., 1994). The Duvernay is absent where Leduc reef accumulations are present (Stoakes, 1980). The depositional limit of the Duvernay is areally more extensive than the underlying Majeau Lake and the Duvernay overlies portions of the Beaverhill Lake Group in the extreme southern and western portions of the West Shale Basin (Switzer et al, 1994). The Duvernay thins to the north where it onlaps the Peace River Arch (Switzer et al., 1994). The organic-rich units of the Duvernay are equivalent to the Muskwa unit of northwest Alberta and northeast British Columbia (Switzer et al, 1994; Chow et al., 1995). Anoxic bottom waters based on slope to platform paleo-bathymetric profiles (uncorrected for compaction) are thought to have developed in water depths greater than approximately 40 to 50 m (Chow et al., 1995) or greater than 65 m (Stoakes, 1980).The organic-rich mudstones of the Duvernay (Chow et al., 1995) are divided by a grey, micritic organic-lean lime mudstone. The lime mudstone exceeds 35 m in the basin center where the typical high radioactivity and high resistivity organic-rich mudstones of the Duvernay are dominated by low radioactivity, moderate resistivity limestones (Switzer et al., 1994, their Figure 12.14 and 12.20). North of the Peace River Arch, the lime mudstone is absent. The Duvernay is conformably overlain by calcareous green-grey shales of the Ireton. The top of the Duvernay is placed at the highest occurrence of dark brown to black shale (Imperial, 1950; Andrichuk, 1958). Therefore, apparent thickness changes of the Duvernay are due to local environmental conditions and do not necessarily constitute a time boundary; the upper limit is diachronic (Imperial, 1950; Andrichuk, 1958). Consequently, where the Ireton is thickened, the Duvernay is thin (Stoakes, 1980). Numerous episodes of sea level transgression and highstand are recorded within Leduc reef accumulations as well as in the formation of basinal marker beds within the Ireton (Stoakes, 1980). Stoakes (1980) demonstrated the lateral continuity of Ireton log markers within the East Shale Basin which enabled the pattern of basin filled to be interpreted based on the geometry of synchronous log markers and the stratal units they bound. Within the East Shale Basin, sediments show an overall east to west progradation normal to the paleocoastline in the east (Stoakes, 1980). Ireton clinoforms can be traced basinward, where the Duvernay has been correlated to represent the “toe-of-the-slope” of Ireton stratal units (Stoakes, 1980). Consequently, 115the Duvernay represents a condensed section and further correlations within the sediments may be difficult (Stoakes, 1980).Paleocurrent evidence suggests that dominant ocean currents in Duvernay time were from the north passing between the emergent Peace River Arch and the Grosmont shelf, which entered the East Shale Basin between the reefs on the Rimbey-Meadowbrook trend (Newland, 1954; Stoakes, 1980). Based on paleoreconstructions and modern surface wind patterns as analogs, dominant paleo-surface winds may have been from the northeast (Wendte, 1994). Within the East Shale Basin, deposition of reef-derived material is dominant on the southwest to southeast side of reefs (Andrichuk, 1961; McCrossan, 1961). Based on the paleocurrent evidence, Stoakes (1980) suggests the current may have circulated in a clockwise direction.5.3 STUDY AREA, InFoRMAl UnITS AnD loG SIGnATURESThe study area in the West Shale Basin (WSB) of central Alberta (Figure 5.2) covers an area of approximately 81,000 km2. The Duvernay Formation was picked in over 1,600 penetrations and samples were taken from eight wells for detailed analysis from the greater Kaybob region (Table 5.1; Figure 5.2) which covers an area of over 20,000 km2. For regional comparison, one well was sampled for detailed analysis within the greater Pembina region. Supplementary samples were taken from nine wells within the Pembina, Radway, and greater Sturgeon Lake regions and one well near the Golden Spike reef (Figure 5.2). A representative wireline log through the Duvernay section is shown in Figure 5.3.Within the Kaybob study area, the Duvernay consists of laminated dark brown to black, organic-rich, siliceous to calcareous mudstones with variable amounts of reef-derived silt sized calcite and interbedded, discontinuous, and variably pyritized carbonate nodules. An organic-lean lime mudstone ranging in thickness from zero to 30 metres is traceable on log signature within the study area along with other thinner, locally continuous to discontinuous limestone interbeds. For this study, the Duvernay has been informally subdivided into the lower Duvernay mudstone, middle carbonate and upper Duvernay mudstone units in order to map and differentiate reservoir-quality facies from non-reservoir facies (Figure 5.3). The upper and lower Duvernay mudstones constitute the dominant reservoir-quality lithofacies, while the middle carbonate, as will be shown later, is of poor reservoir quality. Where the middle carbonate is absent, the entire section is referred to as the 116Duvernay, without distinction of upper or lower units.Underlying argillaceous mudstones of the Majeau Lake are characterized by consistent low resistivity (ILD: 5 – 20 ohmm) and radioactivity (50 – 150 gAPI) compared to overall high resistivity and high radioactivity of the Duvernay (Figure 5.3). The lower Duvernay contact with the Majeau Lake is identifiable in logs by an abrupt increase in resistivity (3865 m, Figure 5.3) and an increase in radioactivity. The middle carbonate, where developed, is distinguishable by low radioactivity (20 – 60 gAPI), low transit times (165 – 215 μs/m), and high bulk density (2.68 – 2.73 g/cc). The upper Duvernay is generally characterized by higher radioactivity than the lower Duvernay (in some areas exceeding 300 gAPI), lower bulk density (2.40 – 2.60 g/cc) and high resistivity (ILD commonly exceeding 1,200 ohmm). The upper Duvernay contact with the Ireton is marked by an abrupt decrease in resistivity (Figure 5.3). The overlying green-grey mudstones and limestones of the Ireton are characterized by low resistivity (ILD: 5 - 15 ohmm) and lower radioactivity (50 – 130 gAPI). These cutoffs represent average values and may vary locally due to input of reef detritus, clay content and other factors.The Duvernay is a diachronous condensed section and hence regionally correlatable timelines within the Ireton (Stoakes, 1980) or Leduc reefs are not traceable into the Duvernay (Stoakes, 1980; Chow et al., 1995). The lithology of chronostratigraphic units vary based on sequence stratigraphic controls (e.g. progradation, aggradation, or retrogradation of facies in response to sea level); hence, time-significant Duvernay subdivisions (where identifiable) may yield favorable reservoir lithologies in one locality but poor reservoir lithologies in another. Isopach distributions of a given chronostratigraphic unit has limited applicability to determining the distribution and thickness of reservoir units, since reservoirs are inherently lithostratigraphic units.Subdivisions within the Duvernay based on log signatures are difficult to identify and often ambiguous. Lithologic subdivisions and their associated log response are not regionally traceable due to local influences on Duvernay deposition which dramatically affect log signature. Local influences include the impact of reef development on anoxia and hence organic content, which yields a varying log response, or local reef-derived debris flows which disrupts “typical” Duvernay deposition. Few distinct reservoir lithologies can be recognized in hand sample and often the variation in compositional properties is minimal with limited “natural” subdivisions. 117Alternatively, compositional properties measured on core samples may be below log scale resolution and therefore reservoir lithologies and heterogeneity defined in core may not be identifiable in logs. The delineation of favorable reservoir, based on lithology, can be more robustly accomplished by defining log signature cutoffs, which are more dynamic (i.e. can be defined for multiple values, adjusted as new data becomes available, or be defined in multiple gradations of reservoir quality) than defining discrete Duvernay subdivisions and give a better indication of net favorable reservoir lithologies. Regional correlation of net favorable reservoir using this methodology is discussed in Part II.5.4 MATERIAlS, METhoDoloGY AnD lIMITATIonS5.4.1 Core samplesWithin mudrock reservoirs, stratigraphic variability commonly exceeds lateral variability due to their basinal depositional environment. Therefore, this study focused on obtaining extensive datasets from multiple wells (detailed analysis wells), rather than a few samples from numerous wells. Sampling intervals for detailed analysis wells within this study are usually less than a metre, but on average at least every two metres. A total of 283 samples from full diameter core were obtained from the nine detailed analysis wells which penetrated the majority of the Duvernay section within the oil and wet gas maturity windows (Table 5.1). In addition, 41 samples were taken from the 10 supplementary wells which are of varying maturity (Table 5.1).5.4.2 Well	log	suitesRegional mapping and correlation was carried out on a dataset of over 1,600 wells penetrating the Duvernay within the West and East Shale Basins. Publicly available wireline log suites were obtained to generate maps of net reservoir thickness, petrophysical properties, hydraulic fracture barriers and structure.5.4.3 MineralogyCrushed sample (< 250 μm) was mixed with ethanol, hand ground in a mortar and smeared on a glass slide for XRD analysis. Normal-focus CoKα radiation on a Bruker® D8 Focus at 35 kV and 40 mA was used on samples over a range of 3-70° 2θ at a step size of 0.03° and counting for 0.8 s per step while continuously spinning at 60 rpm. Mineral phases were quantified using the Rietveld method (Rietveld, 1969) of full-pattern fitting using Bruker® AXS Topas V3.0 software.1185.4.4 organic	geochemistryThe TOC content, organic geochemistry and Tmax values was determined using a source rock SRA-TPH/TOC™ source rock analyzer. Samples were crushed (< 250 μm) and approximately 100 mg of sample pyrolysed. Discussions of common pyrolysis parameters and their meanings (e.g. S1, S2 and S3) can be found elsewhere (Espitalie et al., 1977; Peters, 1986).5.4.5 Unconfined	porosityUnconfined total porosity was determined by the difference between bulk and skeletal density. Since samples were retrieved from non-preserved core, as-received saturations could not be confidently assumed to be representative of reservoir saturations. Therefore, samples were dried at 110 °C for 48 hours prior to analysis. Since samples were dried, measured porosity is the absolute porosity to helium. Bulk density was determined by combining sample weight and sample volume from mercury immersion utilizing Archimedes’ principle at atmospheric conditions. Skeletal density was determined on samples crushed between 20 and 35 mesh (850 and 500 μm) using helium pycnometry at atmospheric conditions (Luffel et al., 1996).5.5 RESUlTSThe regional structure, thermal maturity, reservoir pressure and reservoir temperature data are presented below, followed by detailed core results from laboratory analyses (mineralogy, organic geochemistry and porosity) and thickness trends by unit to highlight the regional architecture and reservoir properties of the Duvernay. Lastly, the controls on regional reservoir distribution are discussed. The detailed analysis wells are primarily located within the greater Kaybob region, and are referred to here as the ‘Kaybob wells’. The exception is the PWT 10-17-45 well, which is located in the Pembina area. Supplementary wells are from various localities and are referred to by their name, locality or maturity. Analyses from the immediately overlying mudstones of the Ireton and underlying Majeau Lake have been included in the results in order to compare with Duvernay lithologies. Due to sample availability, samples from the Ireton are from the base and always immediately overlie the Duvernay. Therefore, samples from the Ireton do not necessarily represent the Ireton section as a whole, which can commonly exceed 300 m in thickness. The underlying Majeau Lake is included in regional correlations due to its close depositional association with the 119Duvernay.5.5.1 StructureWithin the WSB, the Duvernay dips to the southwest with dip angles increasing toward the deformation front. Present day surface dips average 0.68 degrees (0.012 rad) in the WSB, varying from approximately 0.40 degrees (0.007 rad) in the northeast and increasing to over 1.20 degrees (0.021 rad) closest to the deformation front (Figure 5.4). To the northeast, present day sub-sea depths to the top of the Duvernay are approximately 450 m. In the furthest southwest Pembina region, present day sub-sea depths exceed 3,100 m. Residual structure maps (first derivative terrain slopes1) were created for the underlying Majeau Lake and each Duvernay unit to emphasize local structural trends (Figure 5.5; Figure 5.6; Figure 5.7; Figure 5.8). The prevailing terrain slopes within the Kaybob region are northwest-southeast trending linear zones of uniform slope angles. Terrain slopes within the Kaybob region for all units vary from less than 0.5 degrees (0.009 rad) in the northeast to over 0.8 degrees (0.01 rad) in the southwest, closest to the deformation front (Figure 5.5; Figure 5.6; Figure 5.7; Figure 5.8). Some notable deviations to the overall trend exist. A trend profile line is included in Figure 5.5 to illustrate these deviations. Over the interval of A to B (from northeast to southwest), the terrain slope progressively increase from 0.5 degrees to over 0.7 degrees (Figure 5.5), consistent with increasing slope closer to the deformation front. From B to C, however, the terrain slope shallows from over 0.7 degrees to 0.6 degrees as the slope angle profile passes perpendicular to a northwest to southeast trending zone of decreased slope angle (Figure 5.5). From C to A’ (furthest southwest), slope angles increase again from 0.6 degrees to over 0.8 degrees as the slope profile approaches the deformation front. This general profile is characteristic for all units. The decreased slope near point C (Figure 5.5) is important because it illustrates structural variation and deviation from the overall trend of increasing slope angle to the southwest which may impart control on reservoir properties of Duvernay rocks. Deviations in terrain slope may be due to localized block faulting. A structural lineament 1The first derivative terrain slope calculates the slope of the surface at each grid point in degrees from horizontal and takes the value of steepest slope at that point. Directional derivatives, in contrast to the terrain slope, utilize a constant compass direction (e.g. 90o), whereas the terrain slope takes the value of steepest slope regardless of direction. Contour lines on the residual structure maps are therefore lines of constant steepest slope.120determined from gravity gradient data adapted from Lyatsky et al. (2005) is shown with the terrain slope maps (Figure 5.5; Figure 5.6; Figure 5.7; Figure 5.8). The lineament trends approximately parallel with the deformation front in a north-northwest to south-southeast direction. The lineament is also parallel and coincident with a distinct slope change, which could be indicative of draped sediment over a basement fault block. To the east of the lineament, terrain slopes for the Duvernay are generally less than 0.7 degrees. West of the lineament, terrain slope angles increase to over 0.8 degrees.5.5.2 Thermal maturityDue to the large areal extent of the Duvernay shale gas play, publicly available thermal maturity data was retrieved to supplement data collected from wells in this study and to provide a more complete regional characterization. These measurements include both vitrinite reflectance and pyrolysis maturity parameters (Tmax and hydrogen index) made on Duvernay samples and are mapped separately due to the inherent measurement differences. In general, regional organic maturity trends coincide with current burial depth. In terms of hydrocarbon generation potential, the Duvernay is immature in the northeastern portion of the basin (0.4 – 0.6 %Ro, Stasiuk and Fowler, 2002; Tmax 415 – 420 °C, this study) and increases in maturity to the southwest where it is overmature (1.8 – 2.4 %Ro, Stasiuk and Fowler, 2002). Locally, maturity trends are more complex and may be related to basement heat flow (Davis and Karlen, 2013). More research is needed to precisely delineate boundaries of dry gas, wet gas, condensate and volatile oil hydrocarbon zones. Thermal maturity as measured by vitrinite reflectance and reflectance equivalent (bitumen) for the Duvernay in the WSB is shown in Figure 5.9 using data from 25 wells adapted from Stasiuk and Fowler (2002). Vitrinite reflectance values vary from approximately 0.29 to 2.91 % VRo (Stasiuk and Fowler, 2002). Maturity increases in a broad arcuate band to the west, trending approximately parallel to the deformation front. Spatial distribution of vitrinite reflectance measurements within the WSB are low which may lead to variances with other mapped maturity parameters. Thermal maturity trends as measured by Tmax and the hydrogen index (HI) for the Duvernay in the WSB is determined using data from wells in the public domain and samples collected for 121this study (Figure 5.10; Figure 5.11). The Tmax values show a similar trend to vitrinite reflectance values, where Tmax values increase toward the deformation front. Local variations are significant, however, and may be related to localized heat flow (see 5.6.3). The difference between pyrolysis data and vitrinite data is also attributed to variations in sampling density and inherent methodology differences. Average HI for the Duvernay in the WSB from wells in the public domain and this study is shown in Figure 5.11. Trends are consistent with Tmax data and vitrinite reflectance values. Local deviations in organic matter type could impact the HI value. For example, in the northern region of the WSB, there are local troughs of increased HI values which may be related to variations in OM-type and not maturity related. For study wells within the Kaybob region, Tmax varies from 445 °C to 482 °C for samples with a distinct S2 peak (Figure 5.12). The lowest maturity sampled well in the Kaybob region is ATH 13-18-64 with an average HI of 128 mg hydrocarbon per gram of TOC (mg HC/g TOC) and the highest maturity well is the supplementary HSK 10-33-56 with an average HI of 16 mg HC/g TOC. In general, maturity as measured by the HI coincides with present day depth for the basin, with local deviations (e.g. ECA 11-8-62, Figure 5.13; Figure 5.14). A modified van Krevelen for the wells in this study show that kerogen for the Duvernay is classified as Type II (Figure 5.15; Tissot and Welte, 1984). The Duvernay kerogen type is inferred from supplementary wells of lower maturity since the kerogen type may be obscured in high maturity samples (i.e. Kaybob wells). The Kaybob wells are shown in a modified van Krevelen diagram in Figure 5.16.5.5.3 Reservoir pressure and temperatureWell-bore measured pore pressures within the Duvernay are over-pressured with respect to a hydrostatic gradient of 9.79 kPa/m (0.433 psi/ft) assuming fresh water as the pore fluid (Fertl, 1981). Pore pressure gradients within the Duvernay vary from 12.0 kPa/m to 20.2 kPa/m (0.53 psi/ft to 0.89 psi/ft; Figure 5.17). Present day reservoir temperatures vary from 93 °C to 117 °C for wells with pressure tests2 within the greater Kaybob region (Figure 5.18). In general, pore pressure correlates with present day burial depth (Figure 5.19). However, 2Temperatures from wireline logging devices were not utilized since logging devices measure mud circulating temperature and may not reflect true formation temperatures.122notable exceptions exist. A broad, northeast to southwest arcuate trend of increased pressure gradient is evident (e.g. Township 63, Figure 5.17) which does not follow depth based models. This zone aligns with the deviation in the maturity-depth trend towards lower maturity. Present day reservoir temperature shows a moderate positive correlation with reservoir pressure (R2 = 0.55), and exhibits similar areal trends in the greater Kaybob region (e.g. arcuate northeast to southwest trends). Reservoir temperatures generally show a moderate positive correlation with depth (R2 = 0.43) and may be influenced by other factors. Reservoir temperature in the greater Kaybob region increases from east to west, however well control is currently limited and these relationships may change as new data becomes available. A northeast to southwest trending zone of increased temperature follows a similar geographic trend as Leduc reef complexes within the region (Figure 5.18).5.5.4 LithostratigraphyMultiple litho-stratigraphic cross sections illustrate the stratal relationships of Duvernay and related sediments within the greater Kaybob region (Figure 5.20; Figure 5.21; Figure 5.22; Figure 5.23).5.5.5 Majeau	Lake5.5.5.1 Lithology	and	depositional	environmentThe Majeau Lake consists dominantly of green-grey to brown mudstones. Deposition of the Majeau Lake reflects a gradual deepening of the basin where accommodation space matched or was greater than the sedimentation rate (Switzer et al., 1994). Majeau Lake deposition may have been from the east with circulation patterns governing the pattern of basin filling (Stoakes, 1980; Switzer et al., 1994).5.5.5.2 Regional	IsopachThe Majeau Lake thins from the east to the west within the WSB (Figure 5.24), consistent with the overall pattern of basinal sedimentation for the Woodbend Group (Stoakes, 1980). The Formation varies in thickness from over 70 m in the east to less than a metre in the northwest immediately adjacent to the Peace River Arch in the northwest, averaging 14 m in 1,425 wells penetrating the Majeau Lake section in the WSB for this study. Interpolating from regional contour maps, the Majeau Lake averages 20 m. Within the Kaybob region, the Majeau Lake varies in thickness from zero to 27 m, with an average thickness of 7 m (Figure 5.25). East-to-west linear thickness trends are evident in the Kaybob region (e.g. Townships 63 and 64, Figure 5.24; Figure 5.25) and are 123possibly related to increased accommodation (paleo-channels) due to structure in the underlying units, independent of the overall east to west basinal progradation (Switzer et al., 1994). The Majeau Lake thins near the sites of future Leduc reef accumulation suggesting these areas were high during Majeau Lake deposition due to preexisting topographic relief of the underlying Swan Hills and Beaverhill Lake Group sediments.5.5.5.3 MineralogyThe Majeau Lake is composed of equal parts quartz plus feldspar, carbonate and clays (Figure 5.26). Quartz plus feldspar content varies from 22 to 44 % and averages 34 %. Total clay content is higher than Duvernay mudstones, varying from 31 to 45 % and averaging 38 %. The clay component is dominantly composed of illite with minor amounts of chlorite. Total carbonate content varies from 15 to 43 % and averages 26 %.5.5.5.4 organic	geochemistryTotal organic carbon contents for the Majeau Lake are lower than the Duvernay (Appendix B). The sediments within the Majeau Lake immediately adjacent to the Duvernay still contain TOC, which may be significant for hydrocarbon generation. Average TOC for the Majeau Lake is 2.2 % and varies from 0.8 % to 9.7 %. Total organic carbon contents near the maximum of 9.7 % are rare and not representative of the Majeau Lake; removing this point, the average TOC is 1.3 %.5.5.5.5 PorosityThe Majeau Lake averages 4.7 % porosity. As the Majeau Lake samples were dried and shown to be clay rich, the extent of clay desiccation and impact on total porosity measurements is unknown.5.5.6 lower DuvernayDue to core control, the lower Duvernay could not be sampled for mineralogy, organic geochemistry or porosity analyses. The properties of the lower Duvernay determined using artificial neural network models are shown in Part II.5.5.6.1 Lithology	and	depositional	environmentThe well log signature of the lower Duvernay suggests organic-rich mudstone lithologies, similar to the upper Duvernay discussed in the following sections. The lower Duvernay records the initial deepening of the WSB, which led to vertical aggradation of Leduc reefs and a restriction of basinal water circulation and therefore enhanced preservation of organic material.1245.5.6.2 Regional	IsopachThe lower Duvernay in the WSB is inversely distributed compared to the Majeau Lake (Figure 5.27). The lower Duvernay thins toward the basin center, averaging less than five metres, and thickens to over 15 m toward the basin margins (Figure 5.27). The distribution of the lower Duvernay in the Kaybob region is shown in Figure 5.28. In the central Kaybob region, the lower Duvernay thickens to over five metres and progressively thins to less than a metre adjacent to Leduc reefs. The Duvernay cannot be divided when the middle carbonate unit is absent; therefore, the lower Duvernay is not recognized in some areas (e.g. Pine Creek embayment).5.5.7 Middle carbonate5.5.7.1 LithologyThe middle carbonate is a massive, micritic, grey lime mudstone.5.5.7.2 Regional	IsopachWithin the WSB, the middle carbonate generally thins from east to west, ranging in thickness from zero to over 40 m in the basin center and averaging 11 m (Figure 5.29). The middle carbonate is the dominant unit of the Duvernay section in the basin center (e.g. near C’, Figure 5.23). In the eastern portion of the WSB, the middle carbonate thins from north to south and is less than five metres thick where the lower Duvernay is thickened along the basin margins (Figure 5.29). In some areas, the middle carbonate is divided by a thin mudstone interbed, visible on log signatures. Absence of core did not permit sampling of this interval.Within the Kaybob region, the middle carbonate varies in thickness from zero to 25 m, averaging 7 m (Figure 5.30). The middle carbonate unit has a similar isopach pattern to the Majeau Lake in the Kaybob region, thinning east to west and exhibiting linear thickness trends, with some local exceptions. Along Township 63, where the Majeau Lake is thickened, the middle carbonate is thin, implying an underlying structural or sedimentological control on deposition of middle carbonate facies. However, wells in this region frequently do not penetrate below the Beaverhill Lake Group, which renders further exploration of this hypothesis difficult. Thinning of the middle carbonate where the Majeau Lake is thickened does not characterize the majority of the basin, but may be of local significance for wells in producing areas within the Kaybob region. A trend of thin middle carbonate also coincides with the arcuate trend of increased pore pressure gradient within 125the Duvernay, but the cause is unknown. The middle carbonate is largely absent in localized Leduc reef embayments (e.g. Pine Creek region, Range 18 and Townships 59 to 56, Figure 5.30). Where reef-derived debris flows are present in the Duvernay section, the middle carbonate could not be identified.5.5.7.3 MineralogyThe middle carbonate unit is composed dominantly of calcite (40 to 93 %), averaging 73 % (Figure 5.26). The middle carbonate is locally capped by a quartz- and clay-rich interbeds, with clay contents up to 22 % and quartz contents up to 26 %. The quartz- and clay-rich interbeds are distinguishable on logs by an increase in radioactivity (gAPI approximately 50 to 55) and sonic transit time and a relative decrease in bulk density and resistivity compared to the underlying middle carbonate. The relatively quartz-and clay-rich interbeds are included in the middle carbonate unit due to similar reservoir properties and lithology.5.5.7.4 organic	geochemistryThe middle carbonate member of the Duvernay is organic-lean, with TOC contents varying from 0.2 to 2.1 % and averaging 0.8 %. Higher TOC values occur in the local quartz- and clay-rich bed which caps the middle carbonate.5.5.7.5 PorosityPorosity within the middle carbonate unit varies from 1.5 % to 3.6 % and averages 2.5 %. The higher porosity (> 2.5 %) in the middle carbonate is found within the quartz rich interbed which locally caps the unit.5.5.7.6 Depositional environmentDeposition of the middle carbonate unit resulted from a loss of anoxic bottom waters which resulted in carbonate deposition as opposed to organic-rich mudstones which surround the middle carbonate. This interpretation is supported by the thickening of the middle carbonate toward the basin center, furthest from the influence of reefs on basinal circulation (Figure 5.23). Proximal to Leduc reef growth within local embayments, where reefal influence on circulation would have been more pronounced, the middle carbonate is absent, while on the basinward side of Leduc reefs the middle carbonate is present (Figure 5.21). Deposition of the middle carbonate may have occurred during a highstand, when sediment 126supply was greater than accommodation within the WSB (Dunn et al., 2014). Since the middle carbonate thickens to the east and north, it may represent a prograding carbonate ramp which developed as a result of diminishing accommodation space and a breakdown in the anoxic regime (Dunn et al., 2014).5.5.8 Upper Duvernay5.5.8.1 LithologyDominant reservoir lithologies in the upper Duvernay can be divided into two main groups on the basis of visual identification. Reservoir lithologies are a continuum between dark grey to black, laminated, organic-rich, siliceous calcareous mudstones (Figure 5.31) with variable concentrations of fine, light-grey carbonate material dispersed along bedding planes (Figure 5.32). Reservoir lithologies are defined by favorable reservoir properties (i.e. greater porosity within Duvernay mudstones) compared with other units (e.g. low porosity in the middle carbonate). Discrete subdivisions of reservoir lithofacies are not mappable since the variability of the reservoir facies is often lower than log scale. Local influences on TOC content and mineralogy also impact the interpretation of facies from wireline logs (refer to Part II). Vertical and lateral variation in reservoir properties and the distribution of highest net favorable reservoir will also be addressed in Part II.Reservoir lithofacies AReservoir lithofacies A consists of dark grey to black, laminated, organic-rich, siliceous and calcareous mudstones without carbonate-rich material along laminations (Figure 5.31). Reservoir lithofacies A is homogeneous, with TOC contents varying between 3 – 5 %, quartz content varying between 55 – 75 %, carbonate content varying between 10 – 20 % and total clay content varying between 15 – 25 %. Porosity varies between 4 – 7 %. This lithofacies is characteristic of anoxic to dysoxic bottom waters, with minimal bioturbation (Stoakes and Creaney, 1984).Reservoir lithofacies BReservoir lithofacies B consists of dark grey to black, laminated, organic-rich, siliceous and calcareous mudstones with variable concentrations of carbonate-rich laminations (Figure 5.32). Silt-sized carbonate debris derived from the reefs or other interbasinal sources (carbonate platforms) are evident in some laminations. The bulk composition of reservoir lithofacies B is more variable than reservoir lithofacies A due to the variable density of carbonate-rich laminations. The carbonate 127laminations (or oxygenated conditions present during carbonate lamination deposition) dilute bulk rock TOC contents to 1.5 – 3 %. Total quartz content varies between 40 – 60 %, total carbonate content varies between 20 – 45 % (depending on degree of carbonate laminations) and total clay content between 10 – 15 %. Total porosity varies between 3 – 6 %. The carbonate laminations may be the result of horizontal grazing trails, which could have oxygenated the sediments and destroyed OM preservation (Stoakes and Creaney, 1984).5.5.8.2 Regional	IsopachIn the WSB, the upper Duvernay varies in thickness from zero to over 105 m (e.g. in the Wild River sub-basin) and averages 18 m (Figure 5.33). In the basin center, the upper Duvernay varies in thickness from five to 15 m and thickens proximal to Leduc reefs, which is most pronounced in the Kaybob region. Overall patterns of basinal sedimentation are obscured in the upper Duvernay since deposition is controlled by the development of Leduc reefs (Figure 5.33). In the Kaybob region, the upper Duvernay thickness can vary markedly at reservoir scale in relation to Leduc reefs. The upper Duvernay is thickest in Leduc embayments (e.g. Pine Creek region, Figure 5.34) and decreases in overall thickness basinward from Leduc reefs. For example, in the Pine Creek embayment to the west of the Windfall Leduc reef, the upper Duvernay is over 65 m thick. Immediately east of the Windfall Leduc reef, the upper Duvernay is five to 15 m thick (Figure 5.21). Adjacent to Leduc reefs, reef-derived breccia debris flows can thicken the Duvernay section (e.g. Township 68, Range 25, Figure 5.34). Debris flow units are not predictable since they are a local occurrence. One detailed analysis well, ATH 1-24-61, penetrated a section of Duvernay with multiple debris flow breccias (see 5.6.2). Within the Pine Creek embayment (i.e. proximal to Leduc reefs) the overall thickness is dominated by organic-rich mudstones. The lower and upper Duvernay constitute the dominant reservoir units based on more favorable reservoir properties (porosity, mineralogy, and TOC), compared to the Majeau Lake and middle carbonate units. The total thickness of lower and upper Duvernay is shown in Figure 5.35. The distribution pattern is similar to the upper Duvernay given the greater thickness of the upper Duvernay compared to the lower Duvernay in this region.1285.5.8.3 MineralogyThe bulk mineralogy of the upper Duvernay is dominated by quartz and feldspar-rich mudstones with varying amounts of clay and carbonate minerals (Figure 5.26; Appendix A). In some areas, lithologies grade to limestones (> 50 % calcite) with a lesser amount of silicate minerals. Minor amounts of kaolinite, chlorite and pyrite are present within the samples. Since Rietveld quantitative analysis measures relative weight percent, a relative increase in one mineral leads to a decrease in relative percentage of another. This relationship obscures the correlation between minerals. Total carbonate content within upper Duvernay mudstones varies from 6 to 98 % and averages 27 %. The carbonate component is dominantly micritic to silt-sized reef-derived detrital calcite, which varies from 4 to 95 % and averages 22 %. Skeletal derived carbonate from pleagic fauna such as styliolinids and ostracods are sparse. The dolomite group minerals are dominated by ankerite which varies from 0 to 28 % and averages 6 %. There is no correlation between calcite and ankerite content (R2 = 0.02). Carbonate content is inversely related to quartz and feldspar content due to an inverse relationship between calcite and quartz. The dolomite mineral group shows no correlation to quartz content.The quartz content within upper Duvernay mudstones varies from 2 to 65 % and averages 40 %. The feldspar group minerals are dominated by orthoclase and secondarily by albite. Orthoclase content varies from 0 to 32 % and averages 9 %. Albite occur in lesser abundance, varying from 0 to 7 % and averaging 3 %. There is a positive correlation between quartz and feldspar minerals at lower quartz and feldspar contents (quartz < 20 % and feldspars < 5 %), but this correlation decreases with increasing quartz and feldspar contents.Total clay content within Duvernay mudstones varies from 1 to 38 % and averages 18 %. The clay content is dominated by illite/muscovite which varies from 0 to 33 % and averages 16 %. Chlorite (clinochlore) occurs in lesser abundance, varying from 0 to 9 % and averaging 2 %. Kaolinite is rarely present, varying from 0 to 3 % and averaging 0.2 % within the entire samples suite, but 2 % when present. Illite/muscovite content is positively correlated to chlorite content. Total clay content shows a weak negative correlation with total carbonate content due to a negative correlation between illite/muscovite and calcite. No correlation exists between quartz and clay content or feldspars and clay content. In general, chlorite and illite contents increase toward the top 129of the Duvernay section and into the lower Ireton.5.5.8.4 organic	geochemistryTotal organic carbon contents of upper Duvernay mudstones within the Kaybob region vary from 0.2 to 8.3 % and average between 3.0 % and 4.4 % on a per well basis (Appendix B). In immature samples TOC varies from 4.0 to 18.1 % and averages 8.7 %. Overmature samples still contain significant amounts of TOC, varying from 1.2 to 9.9 % and averaging 4.6 %. For wells with detailed analysis, TOC shows a broad positive correlation with quartz content, perhaps indicating a biogenic source for the silica (Figure 5.36A; Ross and Bustin, 2008). No direct evidence of biogenic silica was seen in scanning electron microscopy, possibly due to recrystallization. For undermature samples in the Radway region TOC and quartz shows no relationship (Figure 5.36B). Total organic carbon shows no correlation with total clay content in Kaybob wells (R2 = 0.02). However, in immature wells there is a moderate positive correlation of TOC with total clay content (R2 = 0.61). Calcite and TOC show a broad negative correlation (R2 = 0.45), with TOC being low (< 1.5 %) where calcite content is greater than 60 % (Figure 5.37). Below 60 % calcite there is no correlation with TOC (R2 = 0.10). Pyrite and TOC show a weak positive correlation (R2 = 0.29).5.5.8.5 PorosityUnconfined total porosity to helium within upper Duvernay mudstones varies from 0.6 to 8.3 % and averages 4.4 % (Appendix B). On an average per well basis, porosity varies between 2.4 and 5.6 %. Porosity has the strongest overall correlation with quartz content of all mineral groups (Figure 5.38A). Porosity has a moderate negative correlation with total calcite content (Figure 5.38B). Above approximately 40 % calcite the negative trend with porosity is more pronounced. Below this level, there is no correlation of total calcite content and porosity. Porosity and clay contents (illite) show a poor positive correlation (Figure 5.38C). Total porosity and TOC for the detailed analysis wells show a poor positive correlation (Figure 5.38D). Since OM-hosted porosity has been shown to develop with maturity, the comparison of TOC versus porosity for wells of varying maturity may lead to inconclusive results. However, when compared with average well maturity, TOC and porosity still have a poor correlation (Figure 5.39A). The interrelationship between quartz content, TOC and porosity is dominantly 130controlled by varying quartz content within the sample suite (Figure 5.39B).5.5.8.6 Depositional environmentDeposition of upper Duvernay lithologies is the direct result of anoxic bottom waters developed due to Leduc reef aggradation in response to sea level rise (Stoakes, 1980). This interpretation is confirmed by the substantially greater thicknesses of organic-rich mudstones deposited within embayments proximal to Leduc reefs, which would have presumably been more sheltered from circulation than basin center deposits. Overall low thicknesses of the Majeau Lake and middle carbonate units in the Kaybob region may have yielded increased accommodation space for deposition of upper Duvernay lithologies. Anoxic conditions may have been more pronounced and remained the longest in the deepest portions of the basin as anoxic regime gradually broke down into Ireton time. A prolonged anoxic regime would have allowed for increased thicknesses of organic-rich mudstones to be deposited in these localities.5.5.9 Ireton5.5.9.1 MineralogyThe basal Ireton is similar in bulk composition to the Majeau Lake. Total clay content varies from 21 to 44 % and averages 35 % (Figure 5.26). The clay minerals are dominantly composed of illite/muscovite and lesser amounts of chlorite (averaging 7 %). Total carbonate varies from 13 to 62 % and averages 34 % and is dominantly composed of calcite. Quartz content varies from 17 to 39 % and averages 23 %.5.5.9.2 organic	geochemistryTotal organic carbon contents for the Ireton is lower than the Duvernay (Appendix B), varying from 0.3 % to 2.3 % and averaging 0.9 %.5.5.9.3 PorosityThe Ireton samples average 4.6 % porosity. As the Ireton samples were dried and shown to be clay rich, the extent of clay desiccation and impact on total porosity measurements is unknown.5.5.10 Regional	comparison	of	non-reservoir	to	reservoir	lithologiesIn addition to the relationships established from regional isopach maps, a quantitative comparison of non-reservoir lithologies (Majeau Lake and middle carbonate) to reservoir lithologies (lower and upper Duvernay) provides insight into the unit thicknesses which occur coincident with highest 131(or lowest) thickness of reservoir lithologies. Function plots3 can be used to establish tripartite associations of reservoir lithologies, non-reservoir lithologies and fracture barriers. Within the WSB, thickness of the Majeau Lake and middle carbonate unit are positively correlated (R = 0.74) due to the east to west thinning of the Majeau Lake and the middle carbonate unit (Figure 5.24; Figure 5.29; Figure 5.40A) and the thickening of both units toward the basin center. Within the Kaybob region, the overall trend is similar (R = 0.69, Figure 5.40B). When compared with reservoir thickness, highest reservoir thicknesses are associated with Majeau Lake thicknesses up to approximately 12 m (blue in Figure 5.40B). An increase in thickness of the middle carbonate unit is associated with thinner upper Duvernay, which comprises the majority of the reservoir-quality lithofacies as demonstrated by core analysis. Within the WSB, the upper Duvernay is thin coincident with thickening of the middle carbonate (Figure 5.40C) due to thinning of the upper Duvernay toward basin center where the middle carbonate unit increases in thickness. Below approximately 20 m net reservoir thickness the middle carbonate exhibits the greatest thicknesses, up to 30 m (Figure 5.40D). Where net reservoir thickness is between 20 and 50 m, the middle carbonate is less than 18 m thick (Figure 5.40D). Highest reservoir thicknesses occur where the middle carbonate is absent. Increased thickness of the middle carbonate unit may act as a fracture barrier and render the majority of reservoir quality lithofacies beneath the middle carbonate inaccessible to the wellbore. Increased middle carbonate thickness associated with thinner upper Duvernay sections may therefore be unattractive exploration targets.5.6 DISCUSSIon5.6.1 Pattern	of	basin	fillThe regional thickness distribution and lithology of each mappable stratigraphic unit enables the pattern of basin fill to be examined based on the framework provided by previous studies (e.g. Stoakes, 1980; Switzer et al., 1994). The regional WSB isopach maps give the 1st order basin sedimentation patterns. Local thickness trends from more detailed mapping in the Kaybob region provide the 2nd order trends, which are useful for production and exploration efforts.3A mapped unit (e.g. middle carbonate) is compared with another unit (e.g. the Majeau Lake) and colored by a third variable (e.g. net mudstone reservoir thickness).132 The Cooking Lake carbonate platform in the East Shale Basin and equivalent Majeau Lake mudstones in the WSB were deposited during a sea level highstand or period of constant accommodation space (Stoakes, 1980; Switzer et al., 1994). A subsequent increase in accommodation space due to sea level transgression resulted in deposition of organic-rich mudstones, likely due to vertical aggradation of Leduc reefs from topographic highs within the ESB and WSB which restricted basinal circulation (Stoakes, 1980; Switzer et al., 1994). However, the inverse thickness relationships of the lower Duvernay and Majeau Lake suggest that the initial sea level transgression during Majeau Lake time may have resulted in the development of anoxic bottom waters and deposition of organic-rich mudstones independent of Leduc reef aggradation (Figure 5.24; Figure 5.27). Bottom water anoxia may have developed in deeper waters toward the basin margins which allowed for preservation of organic matter, which is manifest as a high gamma ray radioactivity of the lower Duvernay compared to the underlying Majeau Lake (e.g. Figure 5.21). Lower Duvernay lithologies generally increase in thickness toward the basin margins as the Majeau Lake thins, which suggests organic-rich mudstones of the lower Duvernay were deposited concurrent with the Majeau Lake at the toe of Majeau Lake clinoforms in the WSB (Figure 5.24; Figure 5.27). A similar depositional pattern of organic-rich and organic-lean mudstones has been identified for the upper Duvernay and Ireton clinoforms (e.g. Stoakes, 1980). Organic-rich mudstones commonly occur in the extreme distal portions of systems tracts (i.e. basin margin, Figure 5.27) and persist in these locations longer than more shelfward locations (e.g. Creaney and Passey, 1993). Transgressive events associated with periods of maximum basinal flooding are also commonly associated with organic-rich mudstones deposits in basins without reef development (e.g. Creaney and Passey, 1993; Wignall, 1994; Bessereau et al., 1995). However, no synchronous stratigraphic marker beds are recognized within the Majeau Lake to confirm this hypothesis due to the relative consistency in depositional environment and therefore lithology (e.g. Majeau Lake log signature in Figure 5.21 and Figure 5.23). The loss of bottom water anoxia and deposition of the middle carbonate may have resulted from a sea-level highstand (Dunn et al., 2014) when creation of accommodation space was lower than the sedimentation rate. Sediments of the Majeau Lake and lower Duvernay progressively filled the WSB to the point where carbonate deposition could occur along the upper slopes of the 133Majeau Lake, which is supported by the similar thickness distributions of the middle carbonate and the Majeau Lake. The carbonate ramp would have been initially deposited in the east, where the Majeau Lake is the thickest, and prograded to the west. Deposition of the middle carbonate occurred concurrent with deposition of the lower Duvernay, as indicated by the inverse depositional patterns of the lower Duvernay and middle carbonate unit (Figure 5.27; Figure 5.29). Inter-fingering of the lower Duvernay with the middle carbonate may be present toward the margins of middle carbonate deposition (Figure 5.22, see HSK 5-11-60 well section in Part II) which supports this hypothesis. Linear troughs of poorly oxygenated waters may have existed throughout middle carbonate deposition, evident by greater thicknesses of lower Duvernay lithologies where the middle carbonate is thin toward the basin center (Figure 5.27; Figure 5.29). The upper Duvernay lithologies are the direct result of Leduc reef aggradation and anoxic conditions developed through restricted circulation (Stoakes, 1980; Chow et al., 1995). First order patterns of basin fill are obscured due to the localized control of Leduc reefs on deposition of the upper Duvernay (Figure 5.33). The upper Duvernay is thickest proximal to Leduc reefs in restricted embayments and progressively thins toward the basin center.5.6.2 Reef	debris	and	reservoir	qualityThe presence of reef-derived debris has a negative impact on reservoir quality. Debris flows thicken the Duvernay section in areas where surrounding deposits are thin. The debris flow units average 1.3 % porosity and < 0.5 % TOC and hence have poor hydrocarbon generation and storage potential. The thickness of individual debris flow units can exceed 7 m and may act as a fracture barrier due to the low porosity of debris flow units. Multiple debris flow units occur within some well sections, which further subdivide interbedded mudrock reservoir units. Interbedded organic-rich mudstones are shown to have significant generation potential (e.g. Sample 3, Figure 5.41), however these units have porosities lower than average which provide limited storage of free hydrocarbons. Debris flow units are identified in core (ATH 1-24-61) and correlate to a diagnostic log signature (Figure 5.41) which is then extrapolated to other wells within the basin. Reef-derived debris flows occur within numerous wells in the Kaybob and surrounding regions. Reef derived debris is found from immediately adjacent to Leduc reef accumulations (< 1 km) up to approximately 4 km from the nearest reef edge. Since these deposits occur locally, wells proximal 134to Leduc reefs (< 4 km) may not contain reef-derived debris while distal wells (> 4 km) may contain debris flow deposits. In general, wells greater than 4 km from the nearest reef edge do not show log signatures indicating reef derived debris.A well log schematic of ATH 1-24-61 is shown in Figure 5.41. The average TOC contents of ATH 1-24-61 mudstones interbedded with debris flow units is higher than the overall sample suite, averaging 4.0 % with TOC contents commonly greater than 5.0 % for individual samples. Porosity within interbedded mudstones from ATH 1-24-61 is approximately 35 % lower than all Duvernay samples, averaging 2.9 % compared with 4.4 % for all Duvernay samples. Low porosity within interbedded mudstones indicates that enhanced carbonate content within the section due to debris flow units may lead to increased calcite cementation which would destroy open porosity. Calcite is negatively correlated to porosity for Duvernay mudstones (Figure 5.38), but the effect is pronounced for mudstones interbedded with debris flow units.5.6.3 Controls on reservoir pressure and temperatureHigher reservoir pressures at a given burial depth is a positive factor hydrocarbon reservoirs due to enhanced hydrocarbon storage, lower net effective stress, and increased hydrocarbon deliverability. Therefore, delineating zones of increased pore pressure can have significant implications on ultimate recovery and production rates. Pore pressures within the Duvernay are generally correlative with current burial depth (e.g. Figure 5.19). There are significant deviations in the pore pressure gradient are aligned along a general northeast to southwest trend (e.g. Township 63, Figure 5.17). Wells in this area have high pressure gradients and hence increased storage and deliverability. The deviation of reservoir pore pressure from depth-based models indicates other factors, such as basement-scale heat flow (e.g. Davis and Karlen, 2013), may be controlling zones of overpressure within the Duvernay. Many hydrocarbon reservoirs are known to be associated with basement scale faults (Lyatsky et al., 2005) and potential-field geophysical data such as aeromagnetic surveys and gravity data have been used to help identify basement-scale structure (Edwards et al., 1998; Lyatsky et al., 1998; Lyatsky et al., 2005). In particular, horizontal-gradient vector maps have been used to delineate basement anomalies (Lyatsky et al., 2005). Overlap of high reservoir pressure and temperature along with an anomalous gravity lineament may be indicative of 135heat flow along a basement scale fault. Within the Kaybob region, a linear gravity anomaly oriented northeast to southwest coincides in part with a similar trending zone of anomalous Duvernay temperatures (Figure 5.42). Near the gravity anomaly, temperatures are approximately 20 °C above background trends (up to 117 °C). Directly east of the gravity anomaly, temperatures are lower (93 – 99 °C). Pore pressure gradients deviate from depth-based models and increase coincident with increasing reservoir temperature (Figure 5.42). Within the Pembina region, a linear gravity anomaly trending northeast to southwest could be indicative of increased pore pressures within this area. No Duvernay pressure tests could be found in the public domain for this locality.5.7 ConClUSIonSThe regional characterization of the Duvernay shale gas and shale oil reservoir provides insight into the distribution of reservoir units, potential hydraulic fracture barriers, and reservoir impairments. Significant conclusions from this study include:• Total porosity shows a positive correlation to total quartz content, a moderate negative correlated to total carbonate content at higher carbonate concentrations (> 40 %), a poor negative correlation with TOC content and no systematic trend with maturity.• The Majeau Lake and lower Duvernay may be in part time equivalent. The organic-rich mudstones of the lower Duvernay were deposited during the initial sea level transgression when organic matter preservation was more pronounced, independent of Leduc reef development. Lower Duvernay thicknesses increase in the more distal portions of the basin, while Majeau lake thicknesses increase toward the basin center. The lower Duvernay may thus represent the toe of Majeau Lake clinoforms.• The middle carbonate unit has poor reservoir production and storage potential due to low porosity and low TOC contents. The thickness of the middle carbonate unit increases basinward from Leduc reef accumulations and may represent a prograding carbonate ramp developed during a highstand. The middle carbonate dominates the Duvernay section in the basin center, reaching thicknesses over 40 metres. Increased thickness of the middle carbonate negatively impacts total reservoir thickness and where thick, is a hydraulic fracture barrier.• Reef development is the dominant control on the distribution of upper Duvernay organic-136rich mudstones. Increased mudstone thicknesses occur in inter-reef embayments where basinal water circulation would have been the most restricted. On the basinward side of reefs, Duvernay lithofacies are thinner, highlighting the circulation-restriction effect of reef complexes.• Reef-derived debris lithologies increase total Duvernay thickness locally but are detrimental to overall reservoir quality. Reef-debris lithofacies have low hydrocarbon production and storage potential due to low TOC and low total porosity. Reef debris flows can occur in multiple stacked units with interbeds of organic-rich mudstones. Where thickened, reef debris lithologies pose hydraulic fracture propagation issues for production from interbedded organic-rich mudstones. Mudstones interbedded with debris flow units have greater TOC but significantly less porosity and in general are of poorer reservoir quality than typical reservoir mudstones, unassociated with debris flow units.• Local variations in Duvernay pore pressure and temperature are significant. Variations in pressure and temperature conditions may be correlated to basement scale heat flow and generally align with structural lineaments determined from gravity anomaly data.137Beaverhill Lake GroupDevonianLateWoodbend GroupWinterburn GroupLeduc Fm.Majeau Lake Fm.Duvernay Fm.Cooking Lake Fm.Duvernay Fm.Ireton Fm. Ireton Fm.PeriodEpochCentral Plains, AlbertaFigure 5.1 Stratigraphic table of formations within the central plains, Alberta. Adapted from ERCB (2009).138BRITISHCOLUMBIA ALBERTACANADABRITISH COLUMBIAALBERTADeformation Fronto117 00’o118 00’o116 00’o115 00’o113 00’o114 00’HSK 10-25-59SCL 4-7-59HSK 10-16-61LRE 4-34-77MQL 11-18-72HSK 10-33-56ATH 13-18-64TET 10-18-64ECA 11-8-62HSK 8-25-60HSK 5-11-60YO 14-16-62ATH 1-24-61ATH 4-2-62FB 10-4-51PWT 10-17-45TLM 12-32-41HSK 7-9-37SCL 11-1-38o55 00’o54 00’o52 00’o53 00’0 25 50kmNEdmontonWEST SHALE BASINEAST SHALEBASINPembinaRadwayKaybobSturgeonKey:Sample locationLeduc Fm. reefFigure 5.2 Sample locations and Late Devonian paleogeography of the Leduc Fm. reefs. Mudstones of the Majeau Lake, Duvernay and Ireton fms. were deposited off-reef in the East and West Shale Basins.139UWI Well ID SSTVD to Top Number of Avg. HI Avg. OI Avg. Tmax Maturitym samples mg HC/g TOC mg CO2/g TOCoC100/07-09-037-10W5/00 HSK 7-9-37 3197 2 10.5 9.5 - Dry gas100/10-33-056-22W5/00 HSK 10-33-56 2844 5 10.6 9.8 - Dry gas100/11-01-038-13W5/00 SCL 11-1-38 3674 7 11.3 16.3 - Dry gas100/01-24-061-23W5/00 ATH 1-24-61 2581 34 23.2 11.2 482 Wet gas100/10-17-045-06W5/00 PWT 10-17-45 2005 21 38.5 17.9 471 Wet gas100/13-02-062-23W5/00 ATH 4-2-62 2584 50 48.3 7.3 474 Wet gas100/12-32-041-05W5/00 TLM 12-32-41 2114 7 56.7 11.9 468 Wet gas100/14-16-062-21W5/00 YO 14-16-62 2404 18 59.3 9.4 470 Wet gas104/10-18-064-19W5/00 TET 10-18-64 2133 39 62.9 12.1 464 Late100/05-11-060-18W5/00 HSK 5-11-60 2190 29 68.4 7.6 467 Late103/08-25-060-18W5/00 HSK 8-25-60 2139 19 75.1 6.9 467 Late102/11-08-062-24W5/00 ECA 11-8-62 2717 24 75.4 6.8 469 Late100/13-18-064-17W5/00 ATH 13-18-64 1971 49 128.4 16.7 445 Oil100/04-34-077-23W5/00 LRE 4-34-77 1823 2 247.5 55.5 438 Early100/11-18-072-17W5/00 MQL 11-18-72 1636 3 319.0 24.0 433 Early100/10-04-051-27W4/00 FB 10-4-51 1189 1 348.0 4.0 444 Early100/10-25-059-19W4/00 HSK 10-25-59 360 4 500.8 24.3 417 Immature100/10-16-061-26W4/00 HSK 10-16-61 753 6 508.5 17.5 419 Immature100/04-07-059-20W4/00 SCL 4-7-59 450 4 535.3 26.5 419 ImmatureTable 5.1 Well location and averaged data for all wells in the study. UWI = unique well identifier, Well ID is the identifier used in this study. SSTVD = sub-surface true vertical depth to the top of the Duvernay. HI = hydrogen index, calculated as S2*TOC/100. OI = oxygen index, calculated as S3*TOC/100. See text for discussion of qualitative maturity descriptor.140GR(gAPI)DT(μs/m)RHOB(g/cc)RES(ohm.m)27602780280028202840286028800 300 2 3 1 100 10kMD(m)ILDILMSFLIretonUpper DuvernayMiddle CarbonateLower DuvernayMajeau LakeBeaverhill LakeTrilogy Wascana Kaybob 100/08-26-063-18W5/00Unit300 100Figure 5.3 Representative wireline section through the Duvernay Formation in the Kaybob region. Interbedded carbonates (e.g. 2840 m) are locally occurring within the upper Duvernay and can decrease overall reservoir thickness as well as inhibit hydraulic fracture growth. MD = measured depth; GR = total gamma radiation; DT = P-wave interval transit time; RHOB = bulk density, gamma; RES = resistivity; ILD = deep induction; ILM = medium induction; SFL = laterolog resistivity.141-4000-3000-3000-3000-3000-2000-2000-2000-2000-2000-1000-1000-1000-1000119°W 118°W 117°W 116°W 115°W 114°W 113°W52°30'N53°00'N53°30'N54°00'N54°30'N55°00'N55°30'N56°00'NKeyWell locationLeduc Fm. reefs-4000 -3000 -2000 -1000CI: 200 m0 25 50kmNFigure 5.4 Structure to the top of the Duvernay within the WSB and western East Shale Basin. Depths are meters below sea level. Surface dips are calculated by change in depth over change in surface distance perpendicular to contour lines.142km0 10 20NSlope angle (degrees)oCI: 0.11.100.900.700.500.300.10KeyWell locationLeduc Fm. reefsGravity lineament0.300.400.400.500.500.500.500.600.600.600.600.600.600.600.600.700.700.700.700.700.70 0.700.700.700.800.800.90Range15 14W51617181920212223242526Township615960585762636465666768117°40'W 117°20'W 117°00'W 116°40'W 116°20'W54°00'N54°08’N54°16'N54°24'N54°32’N54°40'N54°48’NAA’BCAA’0.90.50.60.70.8BCSlope angle profileFigure 5.5 Majeau Lake residual structure map for the Kaybob region. A slope angle profile from A-A’ is included to illustrate deviations. Red line is a structural lineament from Lyatsky et al. (2005).143km0 10 20NSlope angle (degrees)oCI: 0.11.100.900.700.500.300.100.500.500.500.500.500.600.600.600.600.600.600.600.600.600.700.700.700.700.700.700.700.700.700.700.700.800.800.800.801.00KeyWell locationLeduc Fm. reefsGravity lineamentRange15 14W51617181920212223242526Township615960585762636465666768117°40'W 117°20'W 117°00'W 116°40'W 116°20'W54°00'N54°08’N54°16'N54°24'N54°32’N54°40'N54°48’NFigure 5.6 Lower Duvernay residual structure map for the Kaybob region. Red line is a structural lineament from Lyatsky et al. (2005).144km0 10 20NSlope angle (degrees)oCI: 0.11.100.900.700.500.300.10KeyWell locationLeduc Fm. reefsGravity lineament0.500.500.500.600.600.600.600.600.600.600.600.700.700.700.700.700.700.700.700.700.800.800.800.801.00Range15 14W51617181920212223242526Township615960585762636465666768117°40'W 117°20'W 117°00'W 116°40'W 116°20'W54°00'N54°08’N54°16'N54°24'N54°32’N54°40'N54°48’NFigure 5.7 Middle carbonate residual structure map for the Kaybob region. Red line is a structural lineament from Lyatsky et al. (2005).145km0 10 20NSlope angle (degrees)oCI: 0.11.100.900.700.500.300.10KeyWell locationLeduc Fm. reefsGravity lineament0.400.500.500.500.500.500.600.600.600.600.600.700.700.700.700.700.700.700.700.800.800.800.800.900.900.900.901.001.10Range15 14W51617181920212223242526Township615960585762636465666768117°40'W 117°20'W 117°00'W 116°40'W 116°20'W54°00'N54°08’N54°16'N54°24'N54°32’N54°40'N54°48’NFigure 5.8 Upper Duvernay residual structure map for the Kaybob region. Red line is a structural lineament from Lyatsky et al. (2005).146Deformation FrontALBERTABRITISH COLUMBIA0.600.600.600.800.800.800.800.801.001.001.001.001.001.201.201.201.201.201.201.401.401.401.401.401.401.601.601.601.601.601.601.801.801.801.801.801.802.002.002.002.002.002.002.202.202.202.202.202.402.402.402.402.402.602.602.602.601.001.401.802.202.60 0.60Vitrinite Reflectance (% VRo) CI: 0.20118°24'W 117°36'W 116°48'W 116°00'W 115°12'W 114°24'W52°24'N52°48'N53°12'N53°36'N54°00'N54°24'N54°48'NKeyWell locationLeduc Fm. reefs0 25 50kmNFigure 5.9 Maturity map from vitrinite reflectance measurements made on Duvernay Fm. samples. Data adapted from Stasiuk and Fowler (2002). Vitrinite reflectance values of 0.60 % approximate the beginning of hydrocarbon generation.147N0 25 50kmKeyPublic well locationLeduc Fm. reefsStudy well locationDeformation FrontALBERTABRITISH COLUMBIAoAverage Tmax ( C)oCI: 10 C420440460480450450 450450450450450460460460460460460460440460470430430440440440440440440440430430480119°W 118°W 117°W 116°W 115°W 114°W 113°W 112°W119°W 118°W 117°W 116°W 115°W 114°W 113°W52°00'N52°30'N53°00'N53°30'N54°00'N54°30'N55°00'N55°30'N52°00'N52°30'N53°00'N53°30'N54°00'N54°30'N55°00'N55°30'NFigure 5.10 Average Duvernay Tmax from publicly available data and data generated for this study.148N0 25 50kmKeyPublic well locationLeduc Fm. reefsStudy well location50150250550 350450CI: 50 mg HC/g TOCAverage HI (mg HC/g TOC)Deformation FrontALBERTABRITISH COLUMBIA1010606060 60606011010110110110110110110160160160160160210210210260260260310310360360410460510510 560602606021021021026010106060110210260260310310310110560119°W 118°W 117°W 116°W 115°W 114°W 113°W 112°W119°W 118°W 117°W 116°W 115°W 114°W 113°W52°00'N52°30'N53°00'N53°30'N54°00'N54°30'N55°00'N55°30'N52°00'N52°30'N53°00'N53°30'N54°00'N54°30'N55°00'N55°30'NFigure 5.11 Average Duvernay HI from publicly available data and data generated for this study.149km0 10 20NKeyPublic well locationLeduc Fm. reefsStudy well locationoAverage Tmax ( C)oCI: 10 C420440460480450450460440440470Range15 14W51617181920212223242526Township615960585762636465666768117°40'W 117°20'W 117°00'W 116°40'W 116°20'W54°00'N54°08’N54°16'N54°24'N54°32’N54°40'N54°48’NFigure 5.12 Average well Tmax in the greater Kaybob region for detailed analysis wells and wells in the public domain. Refer to Figure 5.2 for well identifiers from this study.150km0 10 20NKeyPublic well locationLeduc Fm. reefsStudy well locationAverage HI (mg HC/g TOC)CI: 25 mg HC/g TOC050100150200356085851101101351601601852101035606060608585110110135135160Range15 14W51617181920212223242526Township615960585762636465666768117°40'W 117°20'W 117°00'W 116°40'W 116°20'W54°00'N54°08’N54°16'N54°24'N54°32’N54°40'N54°48’NFigure 5.13 Average well HI in the greater Kaybob region for detailed analysis wells and wells in the public domain. Refer to Figure 5.2 for well identifiers from this study.151050010001500200025003000350040000.0 200.0 400.0 600.0SSTVD to top (m)Hydrogen Index (mg HC/g TOC)HSK 7-9-37SCL 11-1-38HSK 10-33-56ATH 1-24-61PWT 10-17-45ATH 4-2-62TLM 12-32-41YO 14-16-62TET 10-18-64HSK 5-11-60HSK 8-25-60ECA 11-8-62ATH 13-18-64LRE 4-34-77MQL 11-18-72FB 10-4-51HSK 10-25-59HSK 10-16-61SCL 4-7-59Figure 5.14 Average hydrogen index per well versus SSTVD (subsurface true vertical depth) to the top of the Duvernay.1520.0100.0200.0300.0400.0500.0600.0700.0800.0900.00.0 50.0 100.0 150.0Hydrogen Index (mg HC/g TOC)Oxygen Index (mg CO2/g TOC)HSK 7-9-37SCL 11-1-38HSK 10-33-56ATH 1-24-61PWT 10-17-45ATH 4-2-62TLM 12-32-41YO 14-16-62TET 10-18-64HSK 5-11-60HSK 8-25-60ECA 11-8-62ATH 13-18-64LRE 4-34-77MQL 11-18-72FB 10-4-51HSK 10-25-59HSK 10-16-61SCL 4-7-59IIIIIIFigure 5.15 Modified van Krevelen diagram (Tissot and Welte, 1984) for Duvernay source rocks from this study.1530.020.040.060.080.0100.0120.0140.0160.00.0 5.0 10.0 15.0 20.0 25.0 30.0Hydrogen Index (mg HC/g TOC)Oxygen Index (mg CO2/g TOC)HSK 10-33-56ATH 1-24-61PWT 10-17-45ATH 4-2-62TLM 12-32-41YO 14-16-62TET 10-18-64HSK 5-11-60HSK 8-25-60ECA 11-8-62ATH 13-18-64Figure 5.16 Modified van Krevelen (Tissot and Welte, 1984) of the Kaybob region wells. Two Pembina region wells are included for comparison (PWT 10-17-45 and TLM 12-32-41).15418191515161717Range16 15W5Township56116°12'W53°54'N116°24'W116°36'W116°48'W117°00'W117°12'W117°24'W117°36'W54°00'N54°06'N54°12'N54°18'N54°24'N54°30'N54°36'N54°42'N575859606162636465661718192021222324111315171921CI: 1Reservoir Pressure [kPa/m]0 5 10kmNKeyWell locationLeduc Fm. reefsFigure 5.17 Reservoir pore pressure gradient for wells which have tested the Duvernay section. Higher pore pressures positively enhance reservoir properties including increased storage and delivery.155114116118949698100102104104104106108108110110112112106108909498102106110114118CI: 2oReservoir Temperature [ C]Range16 15W5Township56116°12'W53°54'N116°24'W116°36'W116°48'W117°00'W117°12'W117°24'W117°36'W54°00'N54°06'N54°12'N54°18'N54°24'N54°30'N54°36'N54°42'N5758596061626364656617181920212223240 5 10kmNKeyWell locationLeduc Fm. reefsFigure 5.18 Present day temperature for the Duvernay from shut-in pressure test data. Reservoir temperature correlates poorly with depth for this region.1560 10,000 20,000 30,000 40,000 50,000 60,000 70,000 80,000 90,00001,0002,0003,0004,0005,0006,0007,000y = 0.0238x + 1933.12R  = 0.76N = 29Normal hydrostatic gradient9.79 kPa/m (0.433 psi/ft)Reservoir Pressure [kPa]Depth [TVD m]Figure 5.19 Reservoir pressure versus depth for wells which have tested the Duvernay. In general pressure correlates with present day burial depth, however notable exceptions can increase reservoir storage and delivery locally (see Figure 5.17). The dashed line is the normal hydrostatic gradient of 9.79 kPa/m (0.433 psi/ft), the black line is linear line of best fit for the wells. N is the number of samples.157AA’BB’CC’KeyWell locationLeduc Fm. reefsLine of sectionRange13 12 11 10 9W5141516171819202122232425Township59575856555453526061626364656667117°20'W 117°00'W 116°40'W 116°20'W 116°00'W 115°40'W 115°20'W53°30'N53°40'N53°50'N54°00'N54°10'N54°20'N54°30'N54°40'N117°40'W0 10 20kmNFigure 5.20 Index map for cross sections constructed for this study.1580/05-13-053-10W5/0 0/06-15-054-15W5/0 0/10-14-054-16W5/0 0/06-02-056-17W5/0 0/11-26-056-18W5/0 0/10-11-058-18W5/0 0/11-27-059-19W5/0GR DRES GR DRES GR DRES GR DRES GR DRES GR DRES GR DRES334035653575333032303190275552.5 km 7.5 km 19.7 km 12.0 km 14.5 km 19.9 kmIreton mudstonesOrganic-rich Duvernay mudstonesOrganic-lean middle carbonateOrganic-lean Majeau LakeLeduc Fm. reefA A’010mDatum: Base Majeau LakeDepth: measured depthIretonDuvernayFigure 5.21 Cross section A-A’ showing relationship of Leduc reef embayment on Duvernay thickness. GR = gamma radiation, DRES = deep resistivity. Interfingering mudstones with Leduc reef is for representation only. Dashed lines are approximate lithofacies subdivisions based on log signature.15933503450325032503050300029502950020m06-08-060-19W5 09-32-060-19W5 07-20-061-19W5 02-06-062-19W5 16-22-063-19W5 08-09-064-19W5 15-28-064-19W5 02-04-065-19W5GR DRES GR DRES GR DRES GR GR DRES GR DRES GR DRES GR DRESIretonUpperDuvernayMiddleCarbonateLow. Duv.Majeau LakeBeaverhillLakeGroupB B’7.0 km 6.1 km 4.9 km 16.1 km 5.9 km 5.6 km 2.1 kmDuvernay mudstonesMiddle carbonateMajeau LakeBeaverhill Lake GroupDatum: Base Majeau LakeDepth: Measured depthIreton mudstones??Intersection with C-C’?Figure 5.22 Cross section B-B’ showing north-south stratal relationships. The middle carbonate unit thins to the south while the Duvernay mudstones thicken toward the Leduc reef buildups. Dashed lines are approximate lithofacies subdivisions based on log signature.16002/11-08-062-24W5 07-21-062-23W5 02-02-063-22W5 06-07-063-20W5 16-22-063-19W5 06-18-063-17W5 10-18-063-16W5 04-10-063-15W5 07-07-063-14W5 11-15-063-13W5277528002850305029753000320034003550GR DRES GR DRES GR DRES GR DRES GR DRES GR DRES GR DRES GR DRES GR DRES GR DRES12.3 km 12.1 km 12.6 km 16.3 km 14.0 km 10.4 km 14.0 km 5.7 km 14.4 kmIretonDuvernayMajeauC’Duvernay mudstonesMiddle carbonateMajeau LakeDatum: Base Majeau LakeDepth: Measured depthIreton mudstones??Intersection with B-B’3850025mC?Figure 5.23 Cross section C-C’ illustrating east to west stratal relationships. Organic-rich mudstones of the upper Duvernay thin markedly toward basin center away from Leduc reef influence. The middle carbonate unit progressively constitutes a larger portion of the section toward the basin center. Dashed lines are approximate lithofacies subdivisions based on log signature.1610 25 50kmNKeyWell locationLeduc Fm. reefsCI: 5 m0102030405060708055555555510101010101010510101515151515152020202525253030303535404040404545505055556065701510101550119°W 118°W 117°W 116°W 115°W 114°W 113°W52°30'N53°00'N53°30'N54°00'N54°30'N55°00'N55°30'N56°00'NDeformation FrontALBERTABRITISH COLUMBIAFigure 5.24 Majeau Lake isopach map in the WSB. Reefs are zero contoured. CI = contour interval.1620 10 20kmN 048121620CI: 2 mRange15 14 13 12 11W516171819202122232425261W62345Township6159605857565554626364656667686970717273118°24'W 118°00'W 117°36'W 117°12'W 116°48'W 116°24'W 116°00'W53°48'N54°00'N54°12'N54°24'N54°36'N54°48'N55°00'N55°12'N15151517171919212123232511333557117911133333555577777799999911111111111313131315151551115111111333333555555777777799999999111111131313557KeyWell locationLeduc Fm. reefsFigure 5.25 Majeau Lake isopach map in the Kaybob region. Reefs are zero contoured. CI = contour interval. Color scale maximum is 20 m in order to highlight variability within the Kaybob region.1630 20 40 60 80 100100806040200100806040200CarbonateQuartz + FeldsparClayDuvnerayIretonMajeau LakeMiddle carbonaten = 283Figure 5.26 Bulk mineralogy normalized to total quartz and feldspars, total clays, and total carbonate as quantified by XRD. Feldspars include Na-feldspars (albite) and K-feldspars (orthoclase). Total clays include muscovite/illite, chlorite (clinochlore), and kaolinite. Total carbonate includes calcite, dolomite and Fe-dolomite (ankerite). N is the number of samples.1640 25 50kmNCI: 5 m0510152025KeyWell locationLeduc Fm. reefs055555510101010555551010152000555101015151515005555101015155510119°W 118°W 117°W 116°W 115°W 114°W 113°W52°30'N53°00'N53°30'N54°00'N54°30'N55°00'N55°30'N56°00'NDeformation FrontALBERTABRITISH COLUMBIAFigure 5.27 Lower Duvernay isopach map in the WSB. Reefs are zero contoured. CI = contour interval.165Range15 14 13 12 11W516171819202122232425261W62345Township6159605857565554626364656667686970717273118°24'W 118°00'W 117°36'W 117°12'W 116°48'W 116°24'W 116°00'W53°48'N54°00'N54°12'N54°24'N54°36'N54°48'N55°00'N55°12'N02468101214CI: 2 m0 10 20kmNKeyWell locationLeduc Fm. reefs222222222444444666686888810281012822444444444446666666666688888881046666810101010Figure 5.28 Lower Duvernay isopach map in the Kaybob region. Reefs are zero contoured. CI = contour interval.1660 25 50kmNCI: 10 m01020304055555510105510101010101010151515151515152020202020252530510101015155101520253540119°W 118°W 117°W 116°W 115°W 114°W 113°W52°30'N53°00'N53°30'N54°00'N54°30'N55°00'N55°30'N56°00'NDeformation FrontALBERTABRITISH COLUMBIAKeyWell locationLeduc Fm. reefsFigure 5.29 Middle carbonate isopach map in the WSB. Reefs are zero contoured. CI = contour interval.16753°48'N54°00'N54°12'N54°24'N54°36'N54°48'N55°00'N55°12'N118°24'W 118°00'W 117°36'W 117°12'W 116°48'W 116°24'W 116°00'WTownship6159605857565554626364656667686970717273Range15 14 13 12 11W516171819202122232425261W6234504812162024CI: 2 m0 10 20kmNKeyWell locationLeduc Fm. reefs22222444444666666888888810101010101212121212121414141414161616181818182414161820206814141416182224666610141416161820202224Figure 5.30 Middle carbonate unit isopach map in the Kaybob region. Reefs are zero contoured. CI = contour interval.168HSK 5-11-60Sample No. 231 cmAATH 4-2-62Sample No. 211 cmBATH 13-18-64Sample No. 121 cmCATH 4-2-62Sample No. 171 cmDATH 4-2-62Sample No. 91 cmTET 10-18-64Sample No. 271 cmATH 13-18-64Sample No. 261 cmEFGFigure 5.31 Examples of upper Duvernay reservoir lithofacies A. Sample depths and properties are given in Appendix A and B. Variation in color contrast between samples is largely due to core storage and handling effects and may not be indicative of lithology or reservoir properties (i.e. TOC content). Yellow line in G is approximate division of reservoir lithofacies A and B.169TET 10-18-64Sample No. 191 cmATH 13-18-64Sample No. 81 cmAHSK 5-11-60Sample No. 211 cmBATH 13-18-64Sample No. 71 cmCHSK 8-25-60Sample No. 81 cmDATH 13-18-64Sample No. 191 cmFEFigure 5.32 Examples of upper Duvernay reservoir lithofacies B. Sample depths and properties are given in Appendix A and B. Variation in color contrast between samples is largely due to core storage and handling effects and may not be indicative of lithology or reservoir properties (i.e. TOC content).170KeyLeduc Fm. reefs0 25 50kmN 020406080100CI: 10 mDeformation FrontALBERTABRITISH COLUMBIA1010101010101010203040507001020202020303030401010101010102020303010101010102020404010100101020119°W 118°W 117°W 116°W 115°W 114°W 113°W52°30'N53°00'N53°30'N54°00'N54°30'N55°00'N55°30'N56°00'NFigure 5.33 Upper Duvernay isopach map in the WSB. Reefs are zero contoured. Refer to structure map for well control locations (Figure 5.4). CI = contour interval.171515151515252525353535454555552535455555656565757585955555155545255151515254545455565555555 1551515RangeTownship615960585756555462636465666768697071727315 14 13 12 11W516171819202122232425261W6234553°48'N54°00'N54°12'N54°24'N54°36'N54°48'N55°00'N55°12'N118°24'W 118°00'W 117°36'W 117°12'W 116°48'W 116°24'W 116°00'W010203040506070CI: 10 m0 10 20kmNKeyWell locationLeduc Fm. reefsFigure 5.34 Upper Duvernay isopach in the Kaybob region. Reefs are zero contoured. CI = contour interval. Color scale maximum is 70 m in order to highlight variability in the 15-70 m range, which dominantly characterizes the upper Duvernay thickness.172Range15 14 13 12 11W516171819202122232425261W62345Township6159605857565554626364656667686970717273118°24'W 118°00'W 117°36'W 117°12'W 116°48'W 116°24'W 116°00'W53°48'N54°00'N54°12'N54°24'N54°36'N54°48'N55°00'N55°12'N5555551551515151515152525252535353545454545455535455565656575759515551535455515151515151525252525354545455555656565751545020406080100CI: 10 m0 10 20kmNKeyWell locationLeduc Fm. reefsFigure 5.35 Total reservoir thickness (lower and upper Duvernay) in the Kaybob region. Distribution is similar to upper Duvernay due to low total thickness of lower Duvernay. However, the lower Duvernay does constitute a significant proportion of the total reservoir in some areas (e.g. 45 m contour in Township 63, Range 16).173Figure 5.36 (A) Quartz versus TOC for wells with detailed analysis. The correlation is moderate at low concentrations of quartz and TOC. At high concentrations, there is no correlation. (B) Poor correlation of quartz versus TOC for immature wells only.174Figure 5.37 Total carbonate versus TOC content for wells with detailed analysis.1750 20 40 60 80 10001234567890 20 40 60 80 10001234567890 20 40 60 80 10001234567890 2 4 6 8 100123456789HSK 5-11-60ATH 13-18-64ATH 4-2-62HSK 8-25-60ECA 11-8-62YO 14-16-62PWT 10-17-45TET 10-18-64 ATH 1-24-61Porosity [%]Quartz [wt. %] Calcite [wt. %]Illite [wt. %] TOC [%]Porosity [%]Porosity [%]Porosity [%](A) (B)(C) (D)R = 0.60R = 0.27R = 0.34R = 0.58Figure 5.38 Total porosity to helium by well compared with quartz (A), calcite (B), illite (C) and TOC (D).176Figure 5.39 (A) Total organic carbon versus total porosity colored by average well Tmax. (B) Total organic carbon versus quartz content colored by porosity.0 2 4 6 8 1001234567894454554654750 2 4 6 8 100102030405060708002468TOC [%] TOC [%]Quartz [%]Porosity [%]oAverage well Tmax [ C]10Porosity [%](A) (B)HSK 5-11-60ATH 13-18-64ATH 4-2-62HSK 8-25-60ECA 11-8-62YO 14-16-62PWT 10-17-45TET 10-18-64R = 0.27R = 0.511770102030400 20 40 60Majeau Lake, WSB [m]051015202530354045Middle Carbonate, WSB [m]0 20 40 60 80 100Upper Duvernay, WSB [m]Middle Carbonate, WSB [m]0 4 8 12 16 20 24Majeau Lake, Greater Kaybob [m]048121620242832Middle Carbonate, Greater Kaybob [m]0 40 80 120Upper Duvernay, Greater Kaybob [m]Middle Carbonate, Greater Kaybob [m]R = 0.74 R = 0.69(A) (B)(D)(C)50Duvernay Reservoir Thickness, Greater Kaybob [m]051015202530354045048121620242832Figure 5.40 Controls and associations on reservoir thickness within the WSB and the greater Kaybob region.1783690370037103720373036801 cm ATH 1-24-47TOC: <0.5 %Qtz+Feld: 2.3 %Carb: 97.7 %Clay: 0.0 %Porosity: 1.3 %11 cm2ATH 1-24-46TOC: <0.5 %Qtz+Feld: 4.0 %Carb: 96.0 %Clay: 0.0 %Porosity: 1.8 %1 cm3ATH 1-24-36TOC: 5.3 %Qtz+Feld: 71.7 %Carb: 18.8 %Clay: 7.7 %Porosity: 3.3 %1 cm4ATH1-24-26TOC: <0.5 %Qtz+Feld: 12.0 %Carb: 87.6 %Clay: 0.0 %Porosity: 1.5 %1 cm55ATH 1-24-22TOC: 5.3 %Qtz+Feld: 58.6 %Carb: 20.8 %Clay: 16.4 %Porosity: 2.5 %GR0 300TOC (%)gAPI 0 100/01-24-061-23W5/03740MDm4321Figure 5.41 Gamma-ray log and TOC for ATH 1-24-61. Reef-derived breccia debris flows (1, 2, 4) occur interbedded with typical organic-rich mudstones. Debris flows fine upwards (1, 2) and are organic lean with low porosity.179118°00'W 117°40'W 117°20'W 117°00'W 116°40'W 116°20'W 116°00'W53°40'N53°50'N54°00'N54°10'N54°20'N54°30'N54°40'N54°50'NKeyWell locationLeduc Fm. reefsRange15 14 13 12W516171819202122232425261W62Township616059585756555453626364656667680.060.110.150.180.210.230.260.280.310.330.360.390.410.450.490.530.580.650.77µgal/m0 10 20kmNFigure 5.42 Horizontal-gradient gravity map with reefs and reservoir pressure contours (white) superimposed. A zone of high gravity anomaly (hot colors), which may be related to basement faulting, is aligned along a spatially similar trend in reservoir pressure. Refer to Figure 5.17 for reservoir pressure contour values. Gravity map adapted from Lyatsky et al. (2005).180Chapter 6: Regional	reservoir	characterization	model	for	the	shale	gas	and	shale	oil	producing	Duvernay	Formation,	alberta,	Part	II:	Integrating	high-resolution	laboratory	data	and	wireline	logs	using	artificial	neural	networks	to	quantify	regional	reservoir	properties6.1 InTRoDUCTIonFine-grained reservoir evaluation has struggled to integrate relatively sparse laboratory data with more abundant wireline log suites to delineate zones of most prospective reservoir and develop a regional framework of reservoir development. Wells penetrating the prospective formation with wireline log suites are much more numerous than achieved core and subsequent core analyses. Fine-grained reservoirs commonly have substantial areal distributions, sometimes in excess of 100,000 km2 (Ross and Bustin, 2008) which invariably exceeds the areal extent of core control and can be focused in localized regions (e.g. 20,000 km2). Therefore, fine-grained reservoir evaluation requires a robust characterization model whereby core control can be extrapolated to basin scale. The model can be refined as new data becomes available to aid in risk assessment and identification of sweet spots.Numerous methods exist to calibrate core properties to wireline log signatures, with each individual method offering varying applicability and success at predicting core-measured properties from wireline logs. Methods to determine mudstone properties from wireline logs have largely focused on determining organic matter (OM) richness in order to identify potential source rock intervals (e.g. Schmoker, 1981; Schmoker and Hester, 1983; Meyer and Nederlof, 1984; Passey et al., 1990) and commonly rely on regressions with a single wireline log (e.g. formation bulk density or gamma ray log, Schmoker, 1981; Schmoker and Hester, 1983) or regressions utilizing multiple logs (e.g. Meyer and Nederlof, 1984; Passey et al., 1990; see review in Huang and Williamson, 1996). Methods to determine other reservoir properties within mudstones, such as mineralogy, porosity and permeability are less numerous, but commonly rely on relationships between one or more logs (e.g. sonic response and quartz content; Ross and Bustin, 2008).Artificial neural network (ANN) models have been utilized in a variety of applications to predict rock properties, including total organic carbon (TOC) content (Huang and Williamson, 1996; Yang et al., 2004), porosity and permeability (Rogers et al., 1995; Wong et al., 1995; Helle 181et al., 2001; Saemi et al., 2007), lithofacies classification (Wong et al., 1995), shear wave velocities (Eskandari et al., 2004), and condensate-to-gas ratio prediction (Zendehboudi et al., 2012), among others. The ANN models provide a robust methodology to estimate core properties from wireline logs that is often superior to regression analyses (Rezaee et al., 2007) and is more effective in handling non-linearity between wireline data and the correlating data of interest (i.e. core-measured properties; Huang and Williamson, 1996). However, due to the heterogeneities that exist between reservoirs and facies, wireline to core correlations developed for one reservoir may not prove useful when applied to other reservoirs. Therefore, new correlations between core data and log suites must be developed in order to accurately predict petrophysical properties.In Part I, the distribution and compositional properties of the Duvernay Formation was presented for the West Shale Basin and Kaybob region, Alberta. In this section, wireline log datasets were compiled in order to compare laboratory measured properties with log suites and develop ANN models to predict core-based reservoir properties from wireline logs. The ANN models are then applied on a basin scale to delineate zones of organic richness and dominant mineralogical and porosity trends.6.2 MATERIAlS6.2.1 Core samplesThe core analyses performed on detailed analysis wells presented in Part I comprises the core-based dataset for this section (Table 6.1).6.2.2 Well	log	suitesWell log suites were compiled from the public domain. Regional mapping and correlation was carried out on a dataset of over 1,600 wells penetrating the Duvernay within the West Shale Basin.6.3 ARTIFICIAl nEURAl nETWoRKSIn this study, supervised ANN models were developed to predict and calibrate laboratory properties to wireline log suites. In a supervised ANN, the model is trained using known input (wireline log suites) and known supervising data (laboratory data) to develop non-linear relationships between the input data and supervising data to generate an output which best matches (least root-mean-squared-error) the supervising data. The supervising data is split into two sets; one set (training set) is utilized for training the model, while the other set (external examination set) is used to assess the accuracy of 182the model (i.e. has no part in training the model). Numerous types of ANN learning algorithms are available. The most popular in geological studies (and the overall literature) is the back-propagation method and the architecture behind this method and ANN’s in general is discussed extensively elsewhere (Rogers et al., 1995; Huang and Williamson, 1996; Helle et al., 2001; Eskandari et al., 2004; Yang et al., 2004; Saemi et al., 2007; Zendehboudi et al., 2012). The general framework for a back-propagation neural network is given in Figure 6.1. Input data (wireline logs) are given to individual input layer nodes. Each input layer node is connected by synapses to all hidden layer nodes, which in turn are all connected to an output layer node. The input value of a given hidden layer node is equal to the sum of the synapses connected to the hidden layer node from each input layer node, multiplied by the respective weight of each synapse (Yang et al., 2004). The hidden layer node input value is then transformed using a transfer function (e.g. sigmoidal transfer function) and the output of the hidden layer node is transferred via synapses to the output layer node, which repeats the transfer process and produces an output value (petrophysical property). The model is “trained” by modifying the weights of each synapse and monitoring the error of the ANN model output with the external examination set of laboratory data. In the back-propagation method, synapse weights are changed depending on the error of the previous cycle, the correction gained by the previous cycle, and the input to the node (Yang et al., 2004). The model is repeatedly presented with the input data and synapse weights are adjusted a specified number of times. The model gradually converges to an error equilibrium, at which point the model is considered trained if the outputs accurately predict the training data. Once the model has been developed using known data, the ANN can be applied to a wider dataset of wireline log suites where laboratory data is unavailable. Discrepancies between the model and laboratory data can arise due to the difference in measurement scale (depth of investigation, wireline log measurement interval) between wireline log data and laboratory data. Heterogeneities which may exist within the laboratory data that are not measured by the log suites will render the model less effective in predicting the output, since wireline logs may or may not respond to the variations in laboratory measured properties. For example, core samples are generally averaged over a 5 cm interval, while wireline logs commonly measure over a larger interval of greater than 18315 cm. However, a large sample suite minimizes the effects of differing measurement scale and the correlation between model and laboratory data is good. The models can only be considered valid for the intervals in which they were trained to laboratory data (e.g. the Duvernay) and may provide misleading values for other formations that the well penetrates due to the diverse set of lithologies and relationships to wireline logs in a given sedimentary package.The wireline logs which are input into the ANN model for correlation with individual rock properties vary, as some logs may provide limited enhancement of the models ability to generate the desired output. Once the model has been developed, the wells in the larger dataset without laboratory data require the wireline log suite which was used to develop the ANN model. Incorporating more wireline log input data may provide higher correlation factors between the model and laboratory data (e.g. photoelectric log for mineralogy); however, wells within the basin must then have more extensive wireline log datasets in order to participate in modelling. In basins where well control is limited or legacy wells contain limited wireline suites, this may significantly impact the ability to apply the model on a basinal scale (e.g. legacy wells in this study do not commonly have the photoelectric log). Therefore, a balance is sought between model correlation and areal distribution of wells with the required wireline logs.Wireline logs and ANN models compared to laboratory data for the TET 10-18-64, ATH 13-18-64, HSK 5-11-60, HSK 8-25-60 and ATH 4-2-62 detailed analysis wells are shown to demonstrate the application ANN modeling within the Duvernay and Majeau Lake (Figure 6.2; Figure 6.3; Figure 6.4; Figure 6.5; Figure 6.6). An example of the trained ANN models developed for this study applied to a well without laboratory data is given in Figure 6.7. The properties of the individual ANN models and regional reservoir property maps constructed utilizing the models are explained in the following sections.6.3.1 ToC modelDue to the self-sourced nature of unconventional reservoirs, TOC content has a significant impact on total hydrocarbons in place since it is a controlling factor in hydrocarbon generation. The potential for OM-hosted porosity has been correlated to TOC which may contribute significantly to the total pore volume as storage space for produced hydrocarbons. Multiple models have been proposed to calibrate common wireline log suites to TOC, such as gamma radiation (GR) and formation 184density-type models (e.g. Schmoker, 1981), the “delta log R” technique utilizing a porosity log and a resistivity log (Passey et al., 1990) and simpler correlations to the formation bulk density log (Schmoker and Hester, 1983) have been used on organic-rich rocks. The input logs used for the TOC ANN are the GR, deep resistivity (DRES) and formation bulk density (RHOB). This suite was found to yield an acceptable correlation of laboratory TOC to model TOC (R = 0.82, Figure 6.8A) and a high distribution of wells contained the required logs (527 in the Kaybob region). A simple linear correlation of RHOB to TOC is shown for comparison (R = 0.67, Figure 6.8B). Deviations of laboratory derived TOC to model TOC highlight the significant heterogeneity of mudstone reservoirs which exists below the log scale (i.e. the ANN models cannot predict laboratory properties since the wireline logs are not capturing laboratory scale variability). Average TOC for the Duvernay was calculated for wells in the greater Kaybob region using the ANN model (Figure 6.9). Average TOC content per well varies from 0.3 to 5.2 % and averages 3.2 %. Highest average TOC concentrations occur within localized embayments and along the northeast flanks of Leduc reefs (Figure 6.9). Toward the basin center average TOC contents decrease to 2 % or less.6.3.2 Quartz modelQuartz within unconventional reservoirs, depending on mode of occurrence (biogenic versus detrital), has been positively correlated to porosity (Bustin et al., 2008; Chalmers and Bustin, 2012a; Chalmers et al., 2012b), brittleness (Jarvie et al., 2007), TOC content (Ross and Bustin, 2007, 2008), and permeability (Bustin et al., 2008). Within Duvernay mudstones it has been shown that quartz positively correlates to macropore volumes in wet gas window wells (see Chapter 4), permeability (see Chapter 4) and negatively to pore volume compression (Munson and Bustin, unpublished). Therefore, accurately mapping areas of increased quartz content is important due to the association with favorable reservoir properties. Logs used for the quartz ANN model are the GR, DRES, RHOB, and photoelectric factor (PEF). Similar to the TOC model, this suite yielded a high correlation coefficient (R = 0.85, Figure 6.10A). In the Kaybob region, 147 wells contained the required logs, which is less than the TOC model due to the inclusion of the PEF log. However, inclusion of the PEF log is warranted 185because it significantly increased the ability of the model to predict total quartz content (R = 0.78 without PEF log). The average quartz content of the Duvernay is shown in Figure 6.11. No dominant regional trends are observed, although local variations in average quartz content are evident in the Kaybob region which may be significant to local production and exploration. Average quartz content is generally higher (> 50 %) towards the basin center, furthest from Leduc reef complexes.6.3.3 Carbonate modelHigher carbonate concentrations (e.g. > 40 %) have been shown to negatively impact reservoir properties such as porosity (Chalmers and Bustin, 2012a; Chalmers et al., 2012b), TOC content (see Chapter 5) and pore compressibility (Munson and Bustin, unpublished) within Duvernay mudstones. However, lower concentrations of carbonate (< 25 %) generally do not negatively impact reservoir quality, perhaps due to the silt-sized mode of carbonate occurrence within Duvernay mudstones. Therefore, it is necessary to distinguish between zones of higher carbonate content versus moderate to low carbonate enrichment. Discrete carbonate beds within the Duvernay may act as hydraulic fracture barriers or baffles. Carbonate reef-derived debris flows have been shown to negatively impact reservoir quality and will need to be identified for risk assessment. However, as reef-derived debris flows are localized occurrences, the prediction of such facies is limited to wells which penetrate debris-flow units and cannot be regionally predicted from ANN mapping. The local occurrence of wells which penetrate debris flows will yield important information for risk assessment when planning future wells in the proximity of Leduc reefs and debris flow occurrences. Logs used for the carbonate ANN model are the GR, DRES, RHOB and PEF. This suite yielded good correlation between the laboratory data and ANN models (R = 0.87, Figure 6.10B). The model was applied to the same suite of wells as the quartz model since both models utilized the same log suite. Average carbonate contents of the Duvernay are shown in Figure 6.12. In general, average carbonate content is regionally consistent and only varies from approximately 10 – 25 %. Slightly higher average carbonate contents (20 – 25 %) within the Duvernay occur towards the basin center. Regional trends of carbonate content may be obscured by the occurrence of local limestone beds 186which increase average carbonate content within the section or the small range of average carbonate content observed within Duvernay mudstones (10 – 20 %). High carbonate concentrations proximal to Leduc reefs were predicted by the ANN models in multiple wells (e.g. Township 61, Range 16 and Township 60, Range 17, Figure 6.12). The log signature in these wells was diagnostic of carbonate rich debris flows; therefore, the ANN models accurately predicted debris flow occurrence in these wells.6.3.4 Porosity modelOpen and connected porosity serves as the primary storage space for free hydrocarbons compared with adsorbed hydrocarbons, which depends on surface area. Therefore, delineating areas of enhanced porosity is critical for determining areas of potential increased hydrocarbon saturations. The logs used to develop the ANN porosity model were the GR, DRES and RHOB. This suite yielded moderate model to core correlations (R = 0.72; Figure 6.10C). No additional improvement of the correlation coefficient was gained through including the PEF log (R = 0.71). The model was applied to 527 wells within the Kaybob region. Average porosity for the Duvernay is shown in Figure 6.13. No dominant regional trends are observed. However, porosity generally increases toward the basin center (e.g. wells in Township 60, Range 17) but has many local deviations which is significant for exploration.6.4 DISCUSSIon6.4.1 Controls	on	toC,	mineralogy	and	porosity	distribution6.4.1.1 ToC contentThe TOC content of Duvernay rocks as predicted by ANN modelling is found to be significantly higher within localized embayments of Leduc reef complexes (Figure 6.9) compared with more basinward deposits. While restricted basinal circulation due to Leduc reef development is favored as the mechanism for preservation of organic matter within Duvernay sediments as a whole, the effect is particularly pronounced within the sheltered embayments. This may lead to increased hydrocarbon saturations and greater volumes of OM-hosted porosity in these areas which are positive reservoir characteristics.1876.4.1.2 Quartz contentThe average quartz content as predicted by trained ANN models for this study increases toward the basin center, which is attributed to a reduced input of reef-derived carbonate components (Figure 6.11). No other dominant regional trends in quartz content are observed, which could be due to the varying mode of occurrence of quartz within Duvernay rocks. ANN models cannot distinguish biogenic versus detrital quartz.6.4.1.3 Carbonate contentTotal average carbonate content as predicted by trained ANN models for this study increases slightly toward the basin center (up to 25 % average carbonate content per well). This indicates that a significant component of the carbonate within the Duvernay could be inter-basinal and is being transported from carbonate sources other than the local reef developments. A possible source could be the carbonate platforms developed to the east during Duvernay deposition (Switzer et al., 1994). However, average carbonate content within Duvernay mudstones (excluding reef debris flows) does not vary significantly within the study area (from 10 – 25 % average carbonate content) which may be obscuring regional trends. The presence of local, discontinuous limestone interbeds could also impact the calculated average carbonate content within a given well. Multiple wells containing reef-derived debris flows are identified by increased average carbonate content (typically greater than 50 % average for the section; Figure 6.12). From a regional perspective, the ability to quickly identify potential areas of debris flow units will assist in risk assessment to reservoir hazards.6.4.1.4 PorosityPorosity is controlled by mineralogy and organic matter content which create a unique mudrock reservoir texture that varies with thermal maturity and burial. Therefore, regional trends in porosity are based on the interplay of these components. However, no dominant regional porosity trends are clear as predicted by ANN modelling. Average porosity generally decreases proximal to Leduc reefs (Figure 6.13), perhaps due to the input of reef-derived carbonate and the negative correlation of calcite and total porosity. The distribution of wells containing the required logs for carbonate ANN modelling is lower than that of porosity modelling, rendering a direct test of this hypothesis difficult. Regions of greater porosity generally coincide with greater average quartz content 188(Figure 6.11; Figure 6.13) but there are many exceptions. In general, the distribution of average porosity is controlled by local sedimentological factors along with the complex interplay of thermal maturation and burial which is not directly evident in regional maps.6.4.2 Dynamic	reservoir	quality	assessmentThe relationships of mineralogy, TOC, porosity and mudstone reservoir thickness to reservoir quality creates a spectrum of favorable reservoir rocks based on the interplay of these variables and their individual relationships to reservoir quality. Since reservoir quality within the Duvernay varies with few natural and explicit lithologic cutoff points (e.g. beds of consistent lithology, such as limestone beds), discrete reservoir quality gradations are defined for this study to highlight areas of most-prospective reservoir along with areas of lesser potential (Table 6.2). The reservoir quality gradations are based on metrics considered to be important for Duvernay mudstones as shown within Part I and this study. The gradations can be adjusted to suit a specific exploration purpose or modified for use in another reservoir. Three reservoir quality gradations were defined using cutoff values of the mineralogy, TOC and porosity ANN models created for the Duvernay (Table 6.2). Correlations of reservoir characteristics to mineralogy, porosity and TOC contents within Duvernay mudstones are used to guide the gradations. For example, model 1 represents the best quality reservoir rocks as determined in this study. Reservoir rocks which fulfill the criteria for model 1 must have > 50 % quartz, < 30 % carbonate, > 3 % TOC content, and > 4 % porosity (Table 6.2; Figure 6.14). These parameters were chosen based on the positive relationship of total quartz content to porosity, brittleness, and permeability, the positive correlation of porosity to hydrocarbon storage, the positive correlation of TOC to hydrocarbon generation potential and the negative relationship of carbonate content to porosity, TOC and potential fracture barriers (as discrete beds). The total thickness of Duvernay mudrocks which fulfill the criteria for each model is then mapped to determine the regional distribution of prospective reservoir facies. These models are meant to be dynamic; various criteria may be included, refined as new data becomes available, or tailored to specific goals. The various cutoff values chosen for defining the models in this study represent a gradient of reservoir quality. As the parameters shift to include more “lower quality” reservoir rocks, facies which meet the criteria for model 1 also meet the criteria for models 2 and 3 189(Figure 6.14). Regional maps of reservoir quality models 1, 2 and 3 are given in Figure 6.15, Figure 6.16, Figure 6.17, respectively. Since model 1 was designed to represent the highest quality reservoir, it is the most restrictive and shows the smallest thickness and distribution of facies meeting the criteria (Figure 6.14; Figure 6.15). Model 2 represents all reservoir facies of good quality (and best quality) and the thickness and distribution is significantly greater than model 1 (Figure 6.14; Figure 6.16). Model 3 was designed with the least restrictive criteria and most mudstone facies, exclusive of discrete limestone beds, meet the criteria (Figure 6.14). The distribution of facies for model 3 is similar to model 2; however, the thickness of strata which meet the criteria for model 3 is significantly greater as model 3 is the least restrictive (Figure 6.14; Figure 6.17).6.4.3 Vertical variation in reservoir propertiesThe ANN models enable the vertical variation in reservoir properties to be examined over the entire Duvernay interval despite only having core samples from certain zones (i.e. upper Duvernay) in a limited number of detailed analysis wells within a localized region. The vertical variation in reservoir properties can aid in determining depositional environments by highlighting progressive changes in the composition of the Duvernay and underlying Majeau Lake. The TOC content, for example, has been shown to be controlled by water column anoxia (Stoakes, 1980; Chow et al., 1995) and therefore changes in TOC may be used as a proxy for water column oxygenation. Other attributes, such as total carbonate content, may be used to determine variations in reef-derived components or zones of less favorable reservoir rocks based on the correlation of carbonate to other reservoir properties, such as porosity. For reservoir exploration, identifying the vertical variability in reservoir properties is essential for choosing the prospective landing zones for horizontal wells, hydraulic fracture placement, and determining potential fracture impediments within the section.6.4.3.1 ToC contentIn general, the TOC content of the Majeau Lake, as determined by ANN modelling, progressively increases from the contact with the underlying Beaverhill Lake Group into the lower Duvernay (e.g. Figure 6.4; Figure 6.6). Increasing TOC through the Majeau Lake suggests that water column anoxia may have developed gradually during Majeau Lake deposition in some localities. However, in other wells, TOC contents within the Majeau Lake are consistent throughout the entire interval, 190averaging 1 % TOC (Figure 6.2; Figure 6.3). Local controls, such as basinal circulation patterns during deposition, may have played an important role in governing early preservation of organic matter within the Majeau Lake preceding deposition of Duvernay lithologies. The TOC contents within the lower Duvernay are generally consistent, averaging 3 – 4 % per well as measured by ANN modelling in the Kaybob region (e.g. Figure 6.2; Figure 6.3; Figure 6.4). The increased TOC contents of the lower Duvernay compared to the underlying Majeau Lake highlight the development of bottom water anoxia and enhanced organic matter preservation during this time due to initial sea level transgression or reef development. Where the middle carbonate is present, the decrease in TOC compared to the underlying lower Duvernay is generally sharp (e.g. Figure 6.2; Figure 6.3). Since the Duvernay represents a condensed section (Stoakes, 1980), the timing between deposition of the uppermost lower Duvernay and the middle carbonate is uncertain. The TOC contents in the middle carbonate are consistently low (< 0.5 %), indicating the environment of deposition remained constant. Toward the top of the middle carbonate, TOC contents gradually increase in the relatively TOC- and quartz-rich interbed which caps the middle carbonate (e.g. Figure 6.2; the interbed capping the middle carbonate is described in Part I). This indicates that re-establishment of bottom water anoxia may have occurred progressively, as Leduc reefs began to have a more pronounced effect on basinal circulation. In other localities the TOC-rich interbed is thinner (e.g. Figure 6.3) and the contact between the middle carbonate and upper Duvernay is sharper, indicating transition from middle carbonate deposition to upper Duvernay lithologies may have occurred more abruptly. The TOC content within the upper Duvernay is highest within the bottom half of the section compared to the top half (e.g. Figure 6.6). Compared to late Duvernay sedimentation, anoxic conditions may have been more pronounced during the sea level transgression immediately following a middle carbonate highstand. At the base of the upper Duvernay, TOC contents gradually increase from the contact with the middle carbonate and reach a local maximum approximately 5 – 7 m above the top of the middle carbonate (e.g. Figure 6.2; Figure 6.4; Figure 6.6). Gradually increasing TOC at the base of the upper Duvernay indicates reefal development and subsequent restriction of basinal circulation was not immediate; a slight lag in the development of maximum anoxia may have occurred prior to Leduc reefs becoming fully established.191 Near the top of the upper Duvernay, TOC contents gradually decrease into the overlying Ireton. The gradual decrease in TOC content into the Ireton is suggestive of progressive basin fill and a gradual breakdown in the anoxic regime established by the Leduc reefs (e.g. Stoakes, 1980).6.4.3.2 Mineralogical	variationGeologic controls on the vertical variation in mineralogy are less obvious than the controls on TOC content. Multiple modes of occurrence for many minerals and local influences significantly impact the vertical distribution of dominant mineralogy and render detailed correlation between wells in a large region difficult. Broad trends which are evident in most wells, however, allow for some generalizations which point to the zones within the Duvernay which may be more prospective for exploration and production efforts. Total carbonate content within Duvernay mudstones is generally consistent over the entire interval, exclusive of locally occurring limestone beds or reef-derived components, and displays no discernable vertical trends (e.g. Figure 6.2; Figure 6.3; Figure 6.4; Figure 6.6). Quartz content generally follows the vertical trend in TOC content which may be due to the biogenic source of the silica (e.g. Figure 6.2; see Part I). In general, quartz content is highest at the base of the upper Duvernay and decreases toward the top of the upper Duvernay (e.g. Figure 6.2), but many exceptions exist. In some wells, quartz content is consistent throughout the entire upper Duvernay interval and the only significant deviation is at the contact with the Ireton (e.g. Figure 6.4; Figure 6.6).6.5 ConClUSIonSThe associations of important reservoir properties with composition and the integration of these properties with wireline logs though ANN modelling provides a powerful methodology to highlight regions of favorable reservoir characteristics. These models will have implications for exploration and production by expanding data-rich areas (the detailed analysis wells) into areas with limited core-based data. The development of ANN models for TOC, dominant mineral groups and total porosity demonstrates the effectiveness of ANN models for correlating laboratory data to wireline log suites in the Duvernay. The models show better correlations compared with linear regressions with a single log (Figure 6.8). The models were also able to handle correlations with dominant mineral groups, 192which is generally difficult to do due to non-linear relationships between logs and mineral groups or similar log responses from multiple mineral groups. Minerals which occur in varying concentrations and have similar physical properties (e.g. density of quartz = 2.65 g/cc, density of calcite = 2.71 g/cc) can yield similar responses in common log suites (e.g. formation bulk density) and will be difficult to accurately quantify from single-log correlations. Deploying the ANN models on a regional scale enables the extrapolation of detailed datasets on a few wells to many wells within a producing basin. The models need to be developed using common wireline log suites in order to deploy them on a regional scale and be effective at mapping the distribution of TOC, dominant mineral groups and porosity. Models can be updated and “ground-truthed” as more data becomes available. The viability of ANN modelling for facies prediction and predicting other reservoir properties from logs presents an avenue of research that has the potential to yield significant results as demonstrated by this study. Additional reservoir data measured on core, such as fluid saturations and permeability, can be utilized to develop ANN models if sufficient sample points are collected and further define reservoir characteristics across a basin. Local variations in reservoir attributes determined by utilizing ANN models on wells within the basin are significant. The maps provide additional insight into risk assessment and delineating local enhancements of reservoir quality. The models of reservoir quality represent just a few examples of the capabilities of correlating laboratory data to wireline log suites using ANN models, defining dynamic reservoir quality parameters and mapping the distribution and thickness of these units. This workflow presents implications for basin exploration by enabling the high-grading of certain facies (e.g. > 4 % porosity, etc.) to target areas of enhanced reservoir properties. Gradations can be defined for numerous properties or revised to meet specific goals.193UWI Well ID SSTVD to Top Number of Avg. HI Avg. OI Avg. Tmax Maturitym samples mg HC/g TOC mg CO2/g TOCoC100/01-24-061-23W5/00 ATH 1-24-61 2581 34 23.2 11.2 482 Wet gas100/10-17-045-06W5/00 PWT 10-17-45 2005 21 38.5 17.9 471 Wet gas100/13-02-062-23W5/00 ATH 4-2-62 2584 50 48.3 7.3 474 Wet gas100/14-16-062-21W5/00 YO 14-16-62 2404 18 59.3 9.4 470 Wet gas104/10-18-064-19W5/00 TET 10-18-64 2133 39 62.9 12.1 464 Late100/05-11-060-18W5/00 HSK 5-11-60 2190 29 68.4 7.6 467 Late103/08-25-060-18W5/00 HSK 8-25-60 2139 19 75.1 6.9 467 Late102/11-08-062-24W5/00 ECA 11-8-62 2717 24 75.4 6.8 469 Late100/13-18-064-17W5/00 ATH 13-18-64 1971 49 128.4 16.7 445 OilTable 6.1 Detailed analysis wells used for developing ANN models.194Input ParametersHidden LayerOutput LayerGRRHOBResistivityPEFProperty(...)Figure 6.1 General framework of an artificial neural network model. A variety of logs can be included in the input layer (denoted by (…)).1952915.52925.52935.52945.52955.52965.52975.52985.52995.5TET 10-18-64MD0 % 10ANN PorosityCore PorosityANN CarbonateXRD CarbonateANN QuartzXRD QuartzANN TOCCore TOCRHOBCore RHOBPEFDRESGR0 10%0 100%0 100%0 100%0 100%0 10%0 10%2 3g/cc2 3g/cc2 6b/e0.2 2000ohm.m0 300gAPIIretonUpper DuvernayMid. Carb.L. Duv.Maj.Beaverhill Lake Gp.mUnitFigure 6.2 Wireline logs and ANN models (black curves) compared with laboratory data (blue circles) for the TET 10-18-64 well. RHOB = formation bulk density, ANN = artificial neural network modeled property, Core = core measured property, XRD = x-ray diffraction measured property, PEF = photoelectric factor with units barns per electron, DRES = deep resistivity, GR = gamma radiation.1962756.42766.42776.42786.42796.42806.42816.42826.4ATH 13-18-64MD0 % 10ANN PorosityCore PorosityANN CarbonateXRD CarbonateANN QuartzXRD QuartzANN TOCCore TOCRHOBCore RHOBPEFDRESGR0 10%0 100%0 100%0 100%0 100%0 10%0 10%2 3g/cc2 3g/cc2 6b/e0.2 2000ohm.m0 300gAPImUnitIretonUpper DuvernayMid. Carb.L. Duv.Maj.Beaverhill Lk.Figure 6.3 Wireline logs and ANN models (black curves) compared with laboratory data (blue circles) for the ATH 13-18-64 well. RHOB = formation bulk density, ANN = artificial neural network modeled property, Core = core measured property, XRD = x-ray diffraction measured property, PEF = photoelectric factor with units barns per electron, DRES = deep resistivity, GR = gamma radiation.1973044.73054.73064.73074.73084.73094.73104.73114.73034.7HSK 5-11-60MD0 % 10ANN PorosityCore PorosityANN CarbonateXRD CarbonateANN QuartzXRD QuartzANN TOCCore TOCRHOBCore RHOBPEFDRESGR0 10%0 100%0 100%0 100%0 100%0 10%0 10%2 3g/cc2 3g/cc2 6b/e0.2 2000ohm.m0 300gAPImUnitIretonUpper DuvernayM. C.L. Duv.Maj.Beaverhill Lk.Figure 6.4 Wireline logs and ANN models (black curves) compared with laboratory data (blue circles) for the HSK 5-11-60 well. RHOB = formation bulk density, ANN = artificial neural network modeled property, Core = core measured property, XRD = x-ray diffraction measured property, PEF = photoelectric factor with units barns per electron, DRES = deep resistivity, GR = gamma radiation, M. C. = middle carbonate.1983093.13103.13113.13123.13133.13143.13153.13083.1HSK 8-25-60MD0 % 10ANN PorosityCore PorosityANN CarbonateXRD CarbonateANN QuartzXRD QuartzANN TOCCore TOCRHOBCore RHOBPEFDRESGR0 10%0 100%0 100%0 100%0 100%0 10%0 10%2 3g/cc2 3g/cc2 6b/e0.2 2000ohm.m0 300gAPImUnitIretonDuvernayBeaverhill Lk.Figure 6.5 Wireline logs and ANN models (black curves) compared with laboratory data (blue circles) for the HSK 8-25-60 well. RHOB = formation bulk density, ANN = artificial neural network modeled property, Core = core measured property, XRD = x-ray diffraction measured property, PEF = photoelectric factor with units barns per electron, DRES = deep resistivity, GR = gamma radiation. The middle carbonate unit and Majeau Lake are not present within the section.1993606.63616.63626.63636.63646.63656.63666.63596.61231 Middle carbonate2 Lower Duvernay?3 Majeau LakeATH 4-2-62MD0 % 10ANN PorosityCore PorosityANN CarbonateXRD CarbonateANN QuartzXRD QuartzANN TOCCore TOCRHOBCore RHOBPEFDRESGR0 10%0 100%0 100%0 100%0 100%0 10%0 10%2 3g/cc2 3g/cc2 6b/e0.2 2000ohm.m0 300gAPImUnitIretonUpper DuvernayBvhl. Lk.Figure 6.6 Wireline logs and ANN models (black curves) compared with laboratory data (blue circles) for the ATH 4-2-62 well. RHOB = formation bulk density, ANN = artificial neural network modeled property, Core = core measured property, XRD = x-ray diffraction measured property, PEF = photoelectric factor with units barns per electron, DRES = deep resistivity, GR = gamma radiation. The ANN models are uncalculated where the RHOB and PEF logs are missing (e.g. top and base of section in this example).2002821.82831.82841.82851.82861.82871.82881.8100/08-26-063-18W5/00MD0 % 10ANN PorosityANN CarbonateANN QuartzANN TOCRHOBPEFDRESGR0 100%0 100%0 10%2 3g/cc2 6b/e0.2 2000ohm.m0 300gAPIm UnitIretonMid. Carb.Beaverhill Lk.Upper DuvernayL. Duv.Majeau Lk.2811.8Figure 6.7 Wireline logs and laboratory-data-trained ANN models for a well without laboratory data within the study area. The example presented here represents the ultimate result of training the ANN models on wells with laboratory data and then applying the ANN models to a wider distribution of wells in the basin. RHOB = formation bulk density, ANN = artificial neural network modeled property, PEF = photoelectric factor with units barns per electron, DRES = deep resistivity, GR = gamma radiation.201Figure 6.8 (A) Laboratory TOC versus ANN TOC shows a better correlation than TOC derived from a simple linear correlation with RHOB (B). Red lines are linear lines of best fit, black line is 1 to 1 ratio, and R is the correlation coefficient.202KeyWell locationLeduc Fm. reefs0 10 20kmN11111122222233333333344431223333 344112222333333334 44444444455522Range15 14 13 12 11W516171819202122232425261W62345Township6159605857565554626364656667686970717273118°24'W 118°00'W 117°36'W 117°12'W 116°48'W 116°24'W 116°00'W53°48'N54°00'N54°12'N54°24'N54°36'N54°48'N55°00'N55°12'N0123456CI: 1 %Average TOC content (%)Figure 6.9 Average TOC content for the upper and lower Duvernay.203Figure 6.10 (A) Quartz content determined by XRD versus ANN model quartz content. (B) Carbonate content determined by XRD versus ANN model carbonate content. (C) Unconfined porosity to helium versus ANN model porosity. Red lines are linear line of best fit, black line is 1 to 1 ratio, and R is the correlation coefficient.204km0 10 20NKeyWell locationLeduc Fm. reefsCI: 10 %0204060Average quartz content (%)101010101020202020203030303030303040504040404040404040404040103040404040405050505050505050505050505050Range15 14W51617181920212223242526Township615960585762636465666768117°40'W 117°20'W 117°00'W 116°40'W 116°20'W54°00'N54°08’N54°16'N54°24'N54°32’N54°40'N54°48’NFigure 6.11 Average quartz content for the upper and lower Duvernay. Reefs are zeroed contoured; white is < 10 % quartz content or areas of no data.205KeyWell locationLeduc Fm. reefskm0 10 20NCI: 10 %0204060Average carbonate content (%)101010101010101020202020202020 20202020202020202020303030404050102020202020202020303030303030404020Range15 14W51617181920212223242526Township615960585762636465666768117°40'W 117°20'W 117°00'W 116°40'W 116°20'W54°00'N54°08’N54°16'N54°24'N54°32’N54°40'N54°48’NFigure 6.12 Average carbonate content for the upper and lower Duvernay. Reefs are zeroed contoured; white is < 10 % carbonate content or areas of no data.206KeyWell locationLeduc Fm. reefskm0 10 20NAverage Porosity (%)CI: 1 %468 0211111112222222333333334444444455446555556445555112333444444444555555555555555555566745555Range15 14W51617181920212223242526Township615960585762636465666768117°40'W 117°20'W 117°00'W 116°40'W 116°20'W54°00'N54°08’N54°16'N54°24'N54°32’N54°40'N54°48’NFigure 6.13 Average porosity for the upper and lower Duvernay. Reefs are zeroed contoured; white is < 1 % porosity or areas of no data.207Reservoir Components (%)Model Quality Quartz Carbonate TOC Porosity1 Best > 50 < 30 > 3 > 42 Good > 40 < 35 > 2 > 33 Moderate > 30 < 45 > 2 > 3Table 6.2 Criteria to define reservoir quality models.2082756.42766.42776.42786.42796.42806.42816.42826.4ATH 13-18-64MD0 % 10ANN PorosityCore PorosityANN CarbonateXRD CarbonateANN QuartzXRD QuartzANN TOCCore TOCRHOBCore RHOB0 10%0 100%0 100%0 100%0 100%0 10%0 10%2 3g/cc2 3g/ccmUnitIretonUpper DuvernayMid. Carb.L. Duv.Maj.Beaverhill Lk.Model #1Model #2Model #3Figure 6.14 ANN models and thickness of rocks (shaded) which meet the criteria for the reservoir quality models defined in Table 6.2. The net thickness of rocks which meet the criteria for each model is then mapped over the study area.209KeyWell locationLeduc Fm. reefskm0 10 20NCI: 10 m20 10 030405555555555555555555551515151515251525Range15 14W51617181920212223242526Township615960585762636465666768117°40'W 117°20'W 117°00'W 116°40'W 116°20'W54°00'N54°08’N54°16'N54°24'N54°32’N54°40'N54°48’NFigure 6.15 Total thickness of Duvernay mudrocks which satisfy the criteria for reservoir quality model #1 (Table 6.2).210KeyWell locationLeduc Fm. reefskm0 10 20NCI: 10 m20 10 03040555555555151515151515151525252555555551515151515152525252525252535353535Range15 14W51617181920212223242526Township615960585762636465666768117°40'W 117°20'W 117°00'W 116°40'W 116°20'W54°00'N54°08’N54°16'N54°24'N54°32’N54°40'N54°48’NFigure 6.16 Total thickness of Duvernay mudrocks which satisfy the criteria for reservoir model # 2 (Table 6.2).211KeyWell locationLeduc Fm. reefskm0 10 20NCI: 10 m20 10 0304055555 551515151515151515252525252525 253535555515151515252525253535353535355Range15 14W51617181920212223242526Township615960585762636465666768117°40'W 117°20'W 117°00'W 116°40'W 116°20'W54°00'N54°08’N54°16'N54°24'N54°32’N54°40'N54°48’NFigure 6.17 Total thickness of Duvernay mudrocks which satisfy the criteria for reservoir model #3 (Table 6.2).212Chapter 7: Conclusions7.1 oVERVIEWThe goals of this research were to investigate the physical properties of mudrocks as they relate to hydrocarbon reservoirs. Mudrocks from the Duvernay Formation, Alberta were used to investigate errors encountered in common methodologies, suggest new workflows, document the change in mudrock microstructure with thermal maturity and compaction, investigate the controls on matrix permeability, and predict laboratory data from wireline logs and apply predictive models on a basin scale. The combined effects of this research further the understanding of the controls on hydrocarbon distribution, deliverability and ultimately the economic potential of mudrock reservoirs. Understanding the reservoir characteristics of fine-grained lithologies, such as mudrocks, requires a pore- to basin-scale investigation. Heterogeneity within mudrocks exists at multiple levels which necessitates a multidisciplinary approach to study the geologic factors governing reservoir development. At the pore-scale, some methodologies cannot measure the full variety of matrix structures which exist within mudrocks and therefore the results are methods-dependent. While methods-dependent results are not unknown to many research findings, the conclusions drawn and implications can be misleading if not explicitly discussed. Misleading findings are evident in some studies within the mudrock literature (Chapter 3) and are a hazard to advancing the understanding of mudrocks as hydrocarbon reservoirs. The variation in mudrock properties is wide-ranging and often cannot be appreciated at hand sample scale. Numerous samples from an individual reservoir must be analyzed, as petrophysical properties vary with thermal maturity, compaction and composition (Chapter 2; Chapter 4). When comparing different reservoirs, complexity is introduced to the discussion of mudrock microstructure since reservoirs invariably have distinct organic-matter types, mineral provenance and degrees of diagenesis. Determining the associations of favorable reservoir properties, such as mineral composition and matrix permeability, is the first step towards determining the economic potential of a reservoir. Integrating high-resolution laboratory data with more abundant wireline log suites enables a regional framework of reservoir properties to be established (Chapter 5, Chapter 6). Accurate integration of laboratory-based data and wireline log suites will assist in developing more effective exploration 213strategies, development plans and risk-assessment once the play is established.7.2 KEY FInDInGSExploiting the fine grain size inherent to mudrocks enabled the smear mount sample preparation method for X-Ray diffraction to be shown as “fully quantitative” compared to established methods. The smear mount method is less time consuming, laborious and expensive than conventional methods. Shorter time constraints allows a greater number of samples to be analyzed and therefore a greater detail of reservoir characterization. When applied to mudrocks, mercury intrusion porosimetry (MIP) analyses can contain a number of experimental errors, some of which are more evident than others. An additional volume of mercury around the sample (closure) during low-pressure filling is most pronounced when using crushed particles for MIP analysis. The volume of closure systematically decreases with increasing sample size due to particle packing or sorting effects. The volume of closure was found to constitute a significant percentage of the total intruded mercury and therefore closure has significant impacts on the calculated MIP porosity, bulk density and pore size distribution. Compression of the sample volume is also postulated to occur due to the predictable uptake of mercury into the sample cell during the high-pressure phase of the analyses. New equations are presented which utilizes the compression phenomena to determine the amount of pore volume loss and hence, the porosity at stress. The pore structure of Duvernay mudrocks varies systematically with maturity. Organic matter hosted porosity is evident from the onset of the oil window through the dry gas window. Organic matter hosted porosity increases in both size and abundance with increasing maturity due to thermal diagenesis and transformation of organic matter to hydrocarbons. Fine mesopores (< 10 nm), as measured in N2 low pressure gas sorption analyses, develop systematically with maturity which reflects the generation of fine pores hosted within organic matter. Simultaneously, coarse mesopores (> 10 nm) and macropores (> 50 nm) undergo a progressive loss of pore volume due to compaction effects. For samples in the wet gas window, higher volumes of coarse mesopores generally yield higher permeabilities which is correlated to quartz content. Inter-crystalline coarse meso- to macropores were imaged around quartz grain boundaries using focus ion beam milling and field-emission electron microscopy which substantiates gas sorption findings. In general, wet gas window 214and dry gas window samples have average pulse-decay permeabilities an order of magnitude higher than oil window samples due to increasing connectivity within the fine pore network, which is largely contained within the organic matter. Gas expansion porosity and permeability results, which probes the total matrix uptake of gas, further indicates that the fine pore network becomes increasingly connected as maturity increases. The wettability of the mudrock pore structure is complex due to varying types of porosity. Varying wettability of the pore structure may lead to varying liquid permeabilities, where some pores preferentially contribute to fluid flow. The distribution of organic-rich Duvernay mudstones is controlled by the spatial relationship to Leduc Reefs during deposition. In embayments where circulation was restricted by Leduc Reefs, the thickness of organic-rich units is greater than more basinward deposits. Artificial neural network models are used to successfully integrate wireline log suites and high-resolution laboratory data to predict rock properties in wells without laboratory data. The models, when applied on a basin scale, predict areas of most prospective reservoir. The above findings emphasize the need to predict laboratory properties from more abundance wireline logs, especially for mudrock reservoirs where the areal extent of the play commonly exceeds 100,000 km2.7.3 FUTURE RESEARChFuture studies need to expand and substantiate the MIP pore volume compression findings of Chapter 3. The quantities of various physical components and the unique microstructural texture of mudrocks likely controls the extent of pore volume compression during MIP experiments. Samples of various structure attributes could be compared to determine the precise controls on compression. Mudrocks that are less stress sensitive and therefore undergo less pore volume compression will not be as impacted by changing pressure conditions within the reservoir as hydrocarbons are produced. Brittle rocks generally fracture more easily than ductile lithologies and therefore are a positive factor in hydraulic fracturing efforts. The MIP compression tests could prove to be a quick and accurate way to determine the mechanical properties of mudrocks and would have implications for reservoir evaluation. The control of pore size distribution on the permeability of rocks from varying thermal maturity windows requires additional study. The results from this study indicate significant variation between maturity windows which then necessitates additional studies on other mudrock reservoirs. 215Since the pulse-decay permeability (PDP) is biased toward the highest permeability fractures, the applicability of this measurement to determining total gas flux through the matrix should be examined. The gas expansion porosity and permeability (GEPP) probes the average matrix permeability better than PDP experiments and may estimate total gas flux more accurately than PDP measurements. More accurate assessments of total gas flux in laboratory measurements may be important for predicting the cumulative hydrocarbon production over the life of a well. The variation in GEPP with maturity presents many avenues to assess the connectivity and contribution of various pore sizes and types to total gas flux. Wettability is an important avenue of research which deserves further attention. The wettability of the pore system to various fluids impacts the distribution of fluids. Depending on the location of the fluid, permeability and therefore production may be affected. The interaction of gas and liquid hydrocarbons, water, and hydraulic fracturing fluid at various pressures and temperatures create a dynamic environment which is difficult to recreate in the lab. Upscaling laboratory data to log scale is problematic and requires further study for heterogeneous mudrock reservoirs. Wireline log scale is commonly averaged over 15 cm intervals, while core samples are taken from zones smaller than 5 cm. Laboratory analyses suffer from the small area of investigation and many findings may be difficult to confirm at reservoir scale. Wireline logs provide an enhanced volume of investigation (i.e. are measured over the entire section) and are abundant then core analyses within many producing basins. Artificial neural network models are a promising tool to generate accurate predictions from wireline log suites.216REFEREnCESAllan, J., and S. Creaney, 1991, Oil families of the Western Canada Basin: Bulletin of Canadian Petroleum Geology, v. 39, p. 107-122.Anderson, W., 1987, Wettability literature survey part 5: The effects of wettability on relative permeability: Journal of Petroleum Technology, November, p. 1453–1468.Andrichuk, J. M., 1958a, Cooking Lake and Duvernay (Late Devonian) Sedimentation in Edmonton Area of Central Alberta, Canada: AAPG Bulletin, v. 42, no. 9, p. 2189–2222, doi:10.1306/0BDA5BB5-16BD-11D7-8645000102C1865D.Andrichuk, J. M., 1958b, Stratigraphy and facies analysis of Upper Devonian reefs in Leduc, Stettler, and Redwater areas, Alberta: AAPG Bulletin, v. 42, no. 1, p. 1–93.Andrichuk, J. M., 1961, Stratigraphic evidence for tectonic and current control of Upper Devonian reef sedimentation, Duhamel area, Alberta, Canada: American Association of Petroleum Geologists Bulletin, v. 45, no. 5, p. 612–632.Andrichuk, J. M., and J. S. Wonfor, 1954, Late Devonian geologic history in the Stettler area, Alberta, Canada: Bulletin of the American Association of Petroleum Geologists, v. 54, p. 2500-2636.Aplin, A. C., I. F. Matenaar, and B. A. van der Pluijm, 2003, Influence of mechanical compaction and chemical diagenesis on the microfabric and fluid flow properties of Gulf of Mexico mudstones: Journal of Geochemical Exploration, v. 78-79, p. 449–451, doi:10.1016/S0375-6742).Aplin, A. C., I. F. Matenaar, D. K. McCarty, and B. A. van der Pluijm, 2006, Influence of mechanical compaction and clay mineral diagenesis on the microfabric and pore-scale properties of deep-water Gulf of Mexico mudstones: Clays and Clay Minerals, v. 54, no. 4, p. 500–514, doi:10.1346/CCMN.2006.0540411.Bahadur, J., Y. B. Melnichenko, M. Mastalerz, A. Furmann, and C. R. Clarkson, 2014, Hierarchical pore morphology of Cretaceous shale: A small-angle neutron scattering and ultrasmall-angle neutron scattering study: Energy & Fuels, v. 28, 6336-6344.Bailey, S. 2009. Closure and compressibility corrections to capillary pressure data in shales. Presented at the DWLS 2009 Fall Workshop, Colorado School of Mines, October 19.Bailey, S., 2011, Closure and Compressibility Corrections to Capillary Pressure Data in Shales: presented at a meeting of the Canadian Well Logging Society , October 19.Barrett, E. P., L. G. Joyner, and P. P. Halenda, 1951, The determination of pore volume and area distributions in porous substances. I. Computations from nitrogen isotherms: Journal of the American Chemical Society, v. 73, no. 1948, p. 373–380, doi:10.1021/ja01145a126.Belyea, H. R., 1964, Upper Devonian, Part II, Woodbend, Winterburn and Wabamun Groups, in: Geological History of Western Canada, R. G. McCrossan and R. P. Glaister, (eds), Calgary, Alberta Society of Petroleum Geologists, p. 66-68.Bernard, S., and B. Horsfield, 2014, Reply to comment on “Formation of nanoporous pyrobitumen residues during maturation of the Barnett Shale (Fort Worth Basin)”: International Journal of 217Coal Geology, v. 127, p. 114-115, doi: 10.1016/j.coal.2014.01.005.Bernard, S., R. Wirth, A. Schreiber, H.-M. Shulz, and B. Horsfield, 2012, Formation of nanoporous pyrobitumen residues during maturation of the Barnett Shale (Fort Worth Basin): International Journal of Coal Geology, v. 103, p. 3-11, doi: 10.1016/j.coal.2012.04.010.Bessereau, G., F. Guillocheau, and A.-Y. Huc, 1995, Source rock occurrence in a sequence stratigraphic framework: the example of the Lias of the Paris Basin, in: Paleogeography, Paleoclimate, and Source Rocks, A.-Y. Huc, (ed.), American Association of Petroleum Geologistis, Studies in Geology, 40, p. 273–301.Bish, D. L., and J. E. Post, 1989, Modern powder diffraction: Reviews in mineralogy: p. 369.Bish, D. L., and J. R. C. Reynolds, 1989, Sample preparation for X-ray diffraction, in D. L. Bish, and J. E. Post, eds., Modern powder diffraction: Reviews in mineralogy: p. 73–99.Bottani, E. J., V. Bakaev, and W. Steele, 1994, A simulation/experimental study of the thermodynamic properties of carbon dioxide on graphite: Chemical Engineering Science, v. 49, no. 17, p. 2931–2939, doi:10.1016/0009-2509(94)E0111-3.Brunauer, S., P. H. Emmett, and E. Teller, 1938, Adsorption of gases in multimolecular layers: Journal of the American Chemical Society, v. 60, no. 1, p. 309–319, doi:10.1021/ja01269a023.Bustin, R. M., A. M. M. Bustin, A. Cui, D. J. K. Ross, and V. Pathi, 2008, Impact of shale properties on pore structure and storage characteristics: SPE Shale Gas Production Conference: Fort Worth, Texas 16-18 November, p. 1–28, doi:10.2118/119892-MS.Campbell, F. A., and T. A. Oliver, 1968, Mineralogic and chemical composition of Ireton and Duvernay Formations, Central Alberta: Bulletin of Canadian Petroleum Geology, v. 16, no. 1, p. 40–63.Cascarini de Torre, L. E., and E. J. Bottani, 1996, Amorphous carbons: Surface structure and adsorptive properties: Langmuir, v. 12, no. 3, p. 5399–5406.Chalmers, G. R. L., and R. M. Bustin, 2008a, Lower Cretaceous gas shales in northeastern British Columbia, Part I: Geological controls on methane sorption capacity: Bulletin of Canadian Petroleum Geology, v. 56, no. 1, p. 1–21.Chalmers, G. R. L., and R. M. Bustin, 2008b, Lower Cretaceous gas shales in northeastern British Columbia, Part II: Evaluation of regional potential gas resources: Bulletin of Canadian Petroleum Geology, v. 56, no. 1, p. 22-61.Chalmers, G. R. L., and R. M. Bustin, 2012a, Geological evaluation of Halfway-Doig-Montney hybrid gas shale-tight gas reservoir, northeastern British Columbia: Marine and Petroleum Geology, v. 38, no. 1, p. 53–72, doi:10.1016/j.marpetgeo.2012.08.004.Chalmers, G. R. L., and R. M. Bustin, 2012b, Light volatile liquid and gas shale reservoir potential of the Cretaceous Shaftesbury Formation in northeastern British Columbia, Canada: AAPG Bulletin, v. 96, no. 7, p. 1333–1367, doi:10.1306/10041111041.Chalmers, G. R. L., R. M. Bustin, and I. M. Power, 2012a, Characterization of gas shale pore systems by porosimetry, pycnometry, surface area, and field emission scanning electron 218microscopy/transmission electron microscopy image analyses: Examples from the Barnett, Woodford, Haynesville, Marcellus, and Doig units: AAPG Bulletin, v. 96, no. 6, p. 1099–1119, doi:10.1306/10171111052.Chalmers, G. R. L., D. J. K. Ross, and R. M. Bustin, 2012b, Geological controls on matrix permeability of Devonian Gas Shales in the Horn River and Liard basins, northeastern British Columbia, Canada: International Journal of Coal Geology, v. 103, p. 120–131, doi:10.1016/j.coal.2012.05.006.Chong, J. and M. Simikian, 2014, Shale gas in Canada: Resource potential, current production and economic implications: Library of Parliament, Publication No. 2014-08-E, p. 1-14.Chow, N., J. Wendte, and L. Stasiuk, 1995, Productivity versus preservation controls on two organic-rich carbonate facies in the Devonian of Alberta: sedimentological and organic petrological evidence: Bulletin of Canadian Petroleum Geology, v. 43, no. 4, p. 433–460.Clarkson, C. R., and R. M. Bustin, 1996, Variation in micropore capacity and size distribution with composition in bituminous coal of the Western Canadian Sedimentary Basin: Implications for coalbed methane potential: Fuel, v. 75, no. 13, p. 1483–1498.Clarkson, C. R., and R. M. Bustin, 1999, The effect of pore structure and gas pressure upon the transport properties of coal: a laboratory and modeling study. 1. Isotherms and pore volume distributions: Fuel, v. 78, p. 1333–1344.Clarkson, C. R., J. M. Wood, S. E. Burgis, S. D. Aquino, and M. Freeman, 2012, Nanopore-structure analysis and permeability predictions for a tight gas siltstone reservoir by use of low-pressure adsorption and mercury-intrusion techniques: SPE Reservoir Evaluation & Engineering, v. 15, no. 6, p. 648–661.Clarkson, C. R., N. Solano, R. M. Bustin, A. M. M. Bustin, G. R. L. Chalmers, L. He, Y. B. Melnichenko, A. P. Radliński, and T. P. Blach, 2013, Pore structure characterization of North American shale gas reservoirs using USANS/SANS, gas adsorption, and mercury intrusion: Fuel, v. 103, p. 606–616, doi:10.1016/j.fuel.2012.06.119.Cody, R., and G. Thompson, 1976, Quantitative X-ray powder diffraction analyses of clays using an orienting internal standard and pressed disks of bulk shale samples: Clays and Clay Minerals, v. 24, p. 224–231.Colombo, I., and F. Carli, 1981, Measurement of compressibility coefficient of non-porous polymer powders by mercury porosimetry: Powder Technology, v. 29, p. 285–287.Comisky, J., K. E. Newsham, J. Rushing, and T. Blasingame, 2007, A comparative study of capillary-pressure-based empirical models for estimating absolute permeability in tight gas sands: Society of Petroleum Engineers Annual Technical Conference, Anaheim, California, 11-14 November, p. 1–18.Comisky, J., M. Santiago, B. McCollom, A. Buddhala, and K. E. Newsham, 2011, Sample size effects on the application of mercury injection capillary pressure for determining the storage capacity of tight gas and oil shales: Society of Petroleum Engineers Canadian Unconventional Resources Conference, Calgary, Alberta 15-17 November, p. 1–23.219Creaney, S., and J. Allan, 1990, Hydrocarbon generation and migration in the Western Canada Sedimentary Basin, in: Classic Petroleum Provinces, J. Brooks, (ed.), Special Publication of the Geological Society, Blackwell Scientific, p. 189-202.Creaney, S., and Q. Passey, 1993, Recurring patterns of total organic carbon and source rock quality within a sequence stratigraphic framework: AAPG Bulletin, v. 77, no. 3, p. 386–401.Creaney, S., J. Allan, K. S. Cole, M. G. Fowler, P. W. Brooks, K. G. Osadetz, R. W. Macqueen, L. R. Snowdon, and C. L. Riediger, 1994, Petroleum generation and migration in the Western Canadaian Sedimentary Basin, in: Geological Atlas of the Western Canada Sedimentary Basin, G.D. Mossop and I. Shetsen, (comps.), Canadian Society of Petroleum Geologists and the Alberta Research Council, p. 455-468.Cui, X., A. M. M. Bustin, and R. M. Bustin, 2009, Measurements of gas permeability and diffusivity of tight reservoir rocks: different approaches and their applications: Geofluids, v. 9, no. 3, p. 208–223, doi:10.1111/j.1468-8123.2009.00244.x.Cui, X., R. M. Bustin, R. Brezovski, B. Nassichuk, and K. Glover, 2010, A new method to simultaneously measure in-situ permeability and porosity under reservoir conditions: Implications for characterization of unconventional gas reservoirs: Canadian Unconventional Resources & International Petroleum Conference, Calgary, Alberta 19-21 October, p. 1–8.Cullity, B. D., and S. R. Stock, 2001, Elements of X-ray diffraction: New Jersey, Prentice Hall, p. 664.Curtis, M. E., B. J. Cardott, C. H. Sondergeld, and C. S. Rai, 2012a, Development of organic porosity in the Woodford Shale with increasing thermal maturity: International Journal of Coal Geology, v. 103, p. 26–31, doi:10.1016/j.coal.2012.08.004.Curtis, M. E., C. H. Sondergeld, R. J. Ambrose, and C. S. Rai, 2012b, Microstructural investigation of gas shales in two and three dimensions using nanometer-scale resolution imaging: AAPG Bulletin, v. 96, no. 4, p. 665–677, doi:10.1306/08151110188.Curtis, M. E., R. J. Ambrose, C. H. Sondergeld, and C. S. Rai, 2011a, Investigation of the relationship between organic porosity and thermal maturity in the Marcellus Shale: SPE North American Unconventional Gas Conference and Exhibition, The Woodlands, Texas 14-16 June, p. 1-4.Curtis, M. E., R. J. Ambrose, C. H. Sondergeld, and C. S. Rai, 2011b, Transmission and scanning electron microscopy investigation of pore connectivity of gas shales on the nanoscale: SPE North American Unconventional Gas Conference and Exhibition, The Woodlands, Texas 14-16 June, p. 1-10.Davis, M., and G. Karlen, 2013, A regional assessment of the Duvernay Formation; a world-class liquids-rich shale play: CSPG geoConvention, Calgary, Alberta 6-12 May, p. 1-3.Day-Stirrat, R. J., S. P. Dutton, K. L. Milliken, R. G. Loucks, A. C. Aplin, S. Hillier, and B. A. van der Pluijm, 2010, Fabric anisotropy induced by primary depositional variations in the silt: clay ratio in two fine-grained slope fan complexes: Texas Gulf Coast and northern North Sea: Sedimentary Geology, v. 226, p. 42–53, doi:10.1016/j.sedgeo.2010.02.007.220Debelak, K. A., and J. T. Schrodt, 1979, Comparison of pore structure in Kentucky coals by mercury penetration and carbon dioxide adsorption: Fuel, v. 58, no. 10, p. 732–736, doi:10.1016/0016-2361(79)90071-1.Devegowda, D., K. Sapmanee, F. Civan, and R. Sigal, 2012, Phase Behavior of Gas Condensates in Shales Due to Pore Proximity Effects: Implications for Transport Reserves and Well Productivity: SPE Annual Technical Conference and Exhibition, San Antonio, Texas 8-10 October, p.1-22.Dewhurst, D. N., A. C. Aplin, and J.-P. Sarda, 1999, Influence of clay fraction on pore-scale properties and hydraulic conductivity of experimentally compacted mudstones: Journal of Geophysical Research, v. 104, no. B12, p. 29261-29274, doi:10.1029/1999JB900276.Dewhurst, D. N., A. C. Aplin, J.-P. Sarda, and Y. Yang, 1998, Compaction-driven evolution of porosity and permeability in natural mudstones: An experimental study: Journal of Geophysical Research, v. 103, no. B1, p. 651-661, doi:10.1029/97JB02540.Dewhurst, D., R. Jones, and M. Raven, 2002, Microstructural and petrophysical characterization of Muderong Shale: application to top seal risking: Petroleum Geoscience, v. 8, p. 371–383.Do, D. D., and H. D. Do, 2003, Pore characterization of carbonaceous materials by DFT and GCMC simulations: A Review: Adsorption Science and Technology, v. 21, no. 5, p. 389-423, doi:10.1260/026361703769645753.Dollase, W., 1986, Correction of intensities for preferred orientation in powder diffractometry: application of the March model: Journal of Applied Crystallography, v. 19, p. 267–272.Dunn , L., J. Humenjuk, N. Taylor, and D. McHarg, 2014, Fifty shades freed: from pore to seismic and integrated workflow identifying sweet spots in the liquids rich Duvernay Formation Kaybob, Alberta, Canada: CSPG Technical Luncheon, Calgary, Alberta 23 September, p. 1-46.Edwards, D. J., H. V. Lyatsky, and R. J. Brown, 1998, Regional interpretation of steep faults in the Alberta Basin from public-domain gravity and magnetic data: an update: CSEG Recorder, v. 23, no. 1, p. 15-24.Energy Information Administration, U.S., 2015, Natural gas: Shale gas production: U.S. Energy Information Administration, http://www.eia.gov/dnav/ng/ng_prod_shalegas_s1_a.htm (Accessed June 21st, 2015).Eskandari, H., M. Rezaee, and M. Mohammadnia, 2004, Application of multiple regression and artificial neural network techniques to predict shear wave velocity from wireline log data for a carbonate reservoir, south-west Iran: CSEG Recorder, September, p. 42–48.Espitalie, J., M. Madec, and B. Tissot, 1977, Source rock characterization method for petroleum exploration: Offshore Technology Conference, Houston, Texas 2-5 May, p. 439-444.Fertl, W. H., 1981, Abnormal formation pressures: Implications for exploration, drilling, and the production of oil and gas resources: Amsterdam, Elsevier, 400 p.Folk, R. L., 1974, Petrology of sedimentary rocks: Austin, TX, Hemphill Publishing Co., 182 p..Fowler, M. G., L. D. Stasiuk, M. Hearn, and M. Obermajer, 2001, Devonian hydrocarbon source 221rocks and their deived oils in the Western Canada Sedimentary Basin: Bulletin of Canadian Petroluem Geology, v. 49, no. 1, p. 117-148.Friesen, W., and O. Ogunsola, 1995, Mercury porosimetry of upgraded western Canadian coals: Fuel, v. 74, no. 4, p. 604–609.Gan, H., S. P. Nandi, and P. L. Walker Jr., 1972, Nature of porosity in American coals: Fuel, v. 51, p. 272–277.Gibbs, R. J., 1965, Error due to segregation in quantitative clay mineral X-ray diffraction mounting techniques: The American Mineralogist, v. 50, p. 741–751.Gregg, S. J., and K. S. W. Sing, 1982, Adsorption, surface area, and porosity: New York, Academic Press, 303 p.Groen, J. C., L. A. A. Peffer, and J. Pérez-Ramı́rez, 2003, Pore size determination in modified micro- and mesoporous materials. Pitfalls and limitations in gas adsorption data analysis: Microporous and Mesoporous Materials, v. 60, no. 1-3, p. 1–17, doi:10.1016/S1387-1811(03)00339-1.Groen, J., and J. Perez-Ramirez, 2004, Critical appraisal of mesopore characterization by adsorption analysis: Applied Catalysis A: General, v. 268, p. 121–125, doi:10.1016/j.apcata.2004.03.031.Heath, J. E., T. A. Dewers, B. J. O. L. McPherson, M. B. Nemer, and P. G. Kotula, 2012, Pore-lining phases and capillary breakthrough pressure of mudstone caprocks: Sealing efficiency of geologic CO2 storage sites: International Journal of Greenhouse Gas Control, v. 11, p. 204–220, doi:10.1016/j.ijggc.2012.08.001.Heath, J. E., T. A. Dewers, B. J. O. L. McPherson, R. Petrusak, T. C. Chidsey, A. J. Rinehart, and P. S. Mozley, 2011, Pore networks in continental and marine mudstones: Characteristics and controls on sealing behavior: Geosphere, v. 7, no. 2, p. 429–454, doi:10.1130/GES00619.1.Helle, H. B., A. Bhatt, and B. Ursin, 2001, Porosity and permeability prediction from wireline logs using artificial neural networks: a North Sea case study: Geophysical Prospecting, v. 49, no. 4, p. 431–444, doi:10.1046/j.1365-2478.2001.00271.x.Hillier, S., 1999, Use of an Air Brush to Spray Dry Samples for X-ray Powder Diffraction: Clay Minerals, v. 34, p. 127–135, doi:10.1180/claymin.1999.034.1.14.Hillier, S., 2000, Accurate quantitative analysis of clay and other minerals in sandstones by XRD: comparison of a Rietveld and a reference intensity ratio (RIR) method and the importance of sample preparation: Clay Minerals, v. 35, p. 291–291, doi:10.1180/000985500546666.Hillier, S., 2002, Spray drying for X-ray powder diffraction specimen preparation: IUCR CPD Newsletter, v. 27, p. 7–9.Hirasaki, G., 1991, Wettability: fundamentals and surface forces: SPE Formation Evaluation, p. 217–226.Huang, Z., and M. A. Williamson, 1996, Artificial neural network modelling as an aid to source rock characterization: Marine and Petroleum Geology, v. 13, no. 2, p. 277–290, doi:10.1016/0264-8172(95)00062-3.222Hughes, R., and B. Bohor, 1970, Random clay powders prepared by spray-drying: American Mineralogist, v. 55, p. 1780–1786.Imperial Oil Limited Geological Staff, Western Division, 1950, Devonian nomenclature in Edmonton area, Alberta, Canada: AAPG Bulletin, v. 34, p. 1807-1825.Jarvie, D. M., R. J. Hill, T. E. Ruble, and R. M. Pollastro, 2007, Unconventional shale-gas systems: The Mississippian Barnett Shale of north-central Texas as one model for thermogenic shale-gas assessment: AAPG Bulletin, v. 91, no. 4, p. 475–499, doi:10.1306/12190606068.Jenkins, R., and R. L. Snyder, 1996, Introduction to X-ray powder diffractometry: New York, Wiley-Interscience, p. 432.Jeong, G. Y., S. Hillier, and R. A. Kemp, 2008, Quantitative bulk and single-particle mineralogy of a thick Chinese loess–paleosol section: implications for loess provenance and weathering: Quaternary Science Reviews, v. 27, p. 1271–1287, doi:10.1016/j.quascirev.2008.02.006.Jeong, G. Y., S. Hillier, and R. A. Kemp, 2011, Changes in mineralogy of loess–paleosol sections across the Chinese Loess Plateau: Quaternary Research, v. 75, p. 245–255, doi:10.1016/j.yqres.2010.09.001.Jonas, E. C., and J. R. Kuykendall, 1965, Preparation of montmorillonites for random powder diffraction: Clay Minerals, v. 6, p. 232–235.Katz, A., and A. Thompson, 1986, Quantitative prediction of permeability in porous rock: Physical review B, v. 34, no. 11, p. 2–4.Kisch, H., 1991, Illite crystallinity: recommendations on sample preparation, X-ray diffraction settings, and interlaboratory samples: Journal of Metamorphic Geology, v. 9, p. 665–670.Klaver, J., G. Desbois, R. Littke, and J. L. Urai, 2015, BIB-SEM characterization of pore space morphology and distribution in postmature to overmature samples from the Haynesville and Bossier Shales: Marine and Petroleum Geology, v. 59, p. 451–466, doi:10.1016/j.marpetgeo.2014.09.020.Kleeberg, R., 2005, Results of the second Reynolds Cup contest in quantitative mineral analysis: IUCR CPD Newsletter, v. 30, p. 22–26.Kleeberg, R., T. Monecke, and S. Hillier, 2008, Preferred Orientation of Mineral Grains in Sample Mounts for Quantitative XRD Measurements: How Random are Powder Samples?: Clays and Clay Minerals, v. 56, no. 4, p. 404–415, doi:10.1346/CCMN.2008.0560402.Klug, H. P., and L. E. Alexander, 1974, X-ray diffraction procedures: New York, J. Wiley & Sons, p. 966.Kuila, U., and M. Prasad, 2011, Surface area and pore-size distribution in clays and shales: SPE Annual Technical Conference and Exhibition, Denver, Colorado 30 October – 2 November, p. 1-13.Kuila, U., and M. Prasad, 2013, Specific surface area and pore-size distribution in clays and shales: Geophysical Prospecting, v. 61, p. 341–362, doi:10.1111/1365-2478.12028.223Kuila, U., D. K. McCarty, A. Derkowski, T. B. Fischer, and M. Prasad, 2014a, Total porosity measurement in gas shales by the water immersion porosimetry (WIP) method: Fuel, v. 117, p. 1115–1129, doi:10.1016/j.fuel.2013.09.073.Kuila, U., D. K. McCarty, A. Derkowski, T. B. Fischer, T. Topór, and M. Prasad, 2014b, Nano-scale texture and porosity of organic matter and clay minerals in organic-rich mudrocks: Fuel, v. 135, p. 359–373, doi:10.1016/j.fuel.2014.06.036.Labani, M. M., R. Rezaee, A. Saeedi, and A. Al Hinai, 2013, Evaluation of pore size spectrum of gas shale reservoirs using low pressure nitrogen adsorption, gas expansion and mercury porosimetry: A case study from the Perth and Canning Basins, Western Australia: Journal of Petroleum Science and Engineering, v. 112, p. 7–16, doi:10.1016/j.petrol.2013.11.022.Lee, J., and W. Maskell, 1973, Correction factors involved in mercury porosimetry: Powder Technology, v. 7, p. 259–262.León, C. L. Y., 1998, New perspectives in mercury porosimetry: Advances in colloid and interface science, v. 76-77, p. 341–372.Li, Y., G. Lu, and V. Rudolph, 1999, Compressibility and fractal dimension of fine coal particles in relation to pore structure characterisation using mercury porosimetry: Particle & Particle Systems Characterization, v. 16, p. 25–31.Loucks, R. G., R. M. Reed, S. C. Ruppel, and U. Hammes, 2012, Spectrum of pore types and networks in mudrocks and a descriptive classification for matrix-related mudrock pores: AAPG Bulletin, v. 96, no. 6, p. 1071–1098, doi:10.1306/08171111061.Loucks, R., R. Reed, S. Ruppel, and D. Jarvie, 2009, Morphology, genesis, and distribution of nanometer-scale pores in siliceous mudstones of the Mississippian Barnett Shale: Journal of Sedimentary Research, v. 79, p. 848–861.Lowell, S., J. E. Shields, M. A. Thomas, and M. Thommes, 2004, Characterization of porous solids and powders: surface area, pore size and density: Netherlands, Springer, 350 p.Lu, J., S. C. Ruppel, and H. D. Rowe, 2015, Organic matter pores and oil generation in the Tuscaloosa marine shale: AAPG bulletin, v. 2, no. 2, p. 333–357, doi:10.1306/08201414055.Luffel, D. L., J. B. Curtis, and F. K. Guidry, 1996, Development of laboratory and petrophysical techniques for evaluating shale reservoirs: Gas Research Institute, April, p. 1-44.Lyatsky, H. V., D. I. Pana, and M. Grobe, 2005, Basement structure in central and southern Alberta: insights from gravity and magnetic maps: Alberta Energy and Utilities Board, EUB/AGS Special Report 72, 76 p.Lyatsky, H. V., J. R. Dietrich, and D. J. Edwards, 1998, Analysis of gravity and magnetic horizontal-gradient vector data over the buried Trans-Hudson Orogen and Churchill-Superior Boundary Zone in southern Saskatchewan and Manitoba: Geological Survey of Canada, Open File 3614, 34 p.Madsen, I. C., and N. V. Y. Scarlett, 2008, Quantitative phase analysis, in: Powder diffraction: Theory and practice, R. E. Dinnebier, and S. J. . Billinge, (eds.), Cambridge, RCS Publishing, 224604 p.March, A., 1932, Mathematische Theorie der Regelung nach der Korngestalt bei affiner Deformation: Z. kristallogr, v. 81, p. 285–297.Mastalerz, M., L. He, Y. B. Melnichenko, and J. A. Rupp, 2012, Porosity of coal and shale: Insights from gas adsorption and SANS/USANS techniques, Energy & Fuels, v. 26, p. 5109-5120.Mastalerz, M., A. Schimmelmann, A. Drobniak, and Y. Chen, 2013, Porosity of Devonian and Mississippian New Albany Shale across a maturation gradient: Insights from organic petrology, gas adsorption, and mercury intrusion: AAPG Bulletin, v. 97, no. 10, p. 1621–1643, doi:10.1306/04011312194.McCarty, D. K., 2002, Quantitative mineral analysis of clay-bearing mixtures: IUCR CPD Newsletter, v. 27, p. 12–16.McCrossan, R. G., 1961, Resistivity mapping and petrophysical study of the upper Devonian inter-reef calcareous shales of central Alberta, Canada: Bulletin of the American Association of Petroleum Geologists, v. 45, p. 441-470.Meyer, B. L., and M. H. Nederlof, 1984, Identification of source rocks on wireline logs by density/resistivity and sonic transit time/resistivity crossplots: AAPG Bulletin, v. 68, no. 2, p. 121–129, doi:10.1306/AD4609E0-16F7-11D7-8645000102C1865D.Milliken, K. L., M. Rudnicki, D. N. Awwiller, and T. Zhang, 2013, Organic matter-hosted pore system, Marcellus Formation (Devonian), Pennsylvania: AAPG Bulletin, v. 97, no. 2, p. 177–200, doi:10.1306/07231212048.Milner, M., R. McLin, and J. Petriello, 2010, Imaging texture and porosity in mudstones and shales: Comparison of secondary and ion-milled backscatter SEM methods: Canadian Unconventional Resources & International Petroleum Conference, Calgary, Alberta 19-21 October, p. 1–10.Mondol, N. H., K. Bjørlykke, J. Jahren, and K. Høeg, 2007, Experimental mechanical compaction of clay mineral aggregates—Changes in physical properties of mudstones during burial: Marine and Petroleum Geology, v. 24, no. 5, p. 289–311, doi:10.1016/j.marpetgeo.2007.03.006.Moore, D. M., and J. R. C. Reynolds, 1997, X-ray diffraction and the identification and analysis of clay minerals: Oxford, Oxford University Press, p. 378.Mukaida, K., 1981, Density of small porous particles by mercury porosimetry: Powder Technology, v. 29, no. 1, p. 99–107.Newland, J. B., 1954, Interpretation of Alberta reefs based on experience in Texas and Alberta: Journal of the Alberta Society of Petroleum Geologists, v. 2, p. 1-6.Okamoto, I., X. Li, and T. Ohsumi, 2005, Effect of supercritical CO2 as the organic solvent on cap rock sealing performance for underground storage: Energy, v. 30, no. 11-12, p. 2344–2351, doi:10.1016/j.energy.2003.10.025.Oliver, T. A., and N. W. Cowper, 1965, Depositional environments of Ireton Formation, Central Alberta: p. 1410–1425, doi:10.1306/A6633722-16C0-11D7-8645000102C1865D.225Omotoso, O., D. K. McCarty, S. Hillier, and R. Kleeberg, 2006, Some successful approaches to quantitative mineral analysis as revealed by the 3rd Reynolds Cup contest: Clays and Clay Minerals, v. 54, no. 6, p. 748–760, doi:10.1346/CCMN.2006.0540609.Palmer, H., and R. Rowe, 1974, The application of mercury porosimetry to porous polymer powders: Powder Technology, v. 9, no. 4, p. 181–186.Passey, Q., S. Creaney, J. Kulla, F. J. Moretti, and J. D. Stroud, 1990, A practical model for organic richness from porosity and resistivity logs: AAPG Bulletin, v. 74, no. 12, p. 1777–1794.Penumadu, D., and J. Dean, 2000, Compressibility effect in evaluating the pore-size distribution of kaolin clay using mercury intrusion porosimetry: Canadian Geotechnical Journal, v. 37, no. 2, p. 393–405.Peters, K. E., and M. R. Cassa, 1994, Applied source rock geochemistry, in: The petroleum system: from source to trap, L. B. Magoon and W. G. Dow, (eds.), AAPG Memoir 60, p. 93-120.Peters, K., 1986, Guidelines for evaluating petroleum source rock using programmed pyrolysis: AAPG Bulletin, v. 7, no. 3, p. 318–329.Pittman, E., 1992, Relationship of porosity and permeability to various parameters derived from mercury injection-capillary pressure curves for sandstone: AAPG Bulletin, v. 76, no. 2, p. 191–198.Poppe, L. J., V. F. Paskevich, J. C. Hathaway, and D. S. Blackwood, 2000, A laboratory manual for X-ray powder diffraction: U.S. Geological Survey Open-File Report 01-041.Potter, P. E., J. B. Maynard, and W. A. Pryor, 1980, Sedimentology of shales: New York, Springer-Verlag, Inc., p. 306.Price, R. A., 1994, Cordilleran tectonics and the evolution of the Western Canada Sedimentary Basin, in: Geological Atlas of the Western Canada Sedimentary Basin, G.D. Mossop and I. Shetsen, (comps.), Canadian Society of Petroleum Geologists and the Alberta Research Council, p. 165-202.Ravikovitch, P. I., and A. V. Neimark, 2001, Characterization of nanoporous materials from adsorption and desorption isotherms: Colloids and Surfaces A: Physicochemical and Engineering Aspects, v. 187-188, p. 11–21, doi:10.1016/S0927-7757(01)00614-8.Reed, R. M., R. G. Loucks, and S. C. Ruppel, 2013, Comment on “Formation of nanoporous pyrobitumen residues during maturation of the Barnett Shale (Fort Worth Basin)” by Bernard et al. (2012): International Journal of Coal Geology, v. 127, p. 111–113, doi:10.1016/j.coal.2013.11.012.Rezaee, R., R. M. Slatt, and R. F. Sigal, 2007, Shale gas rock properties prediction using artificial neural network technique and multi-regression analysis, an example from a North American shale gas reservoir: 19th International Geophysical Conference and Exhibition, Perth, Australia 19-22 November, p. 1–4.Rietveld, H. M., 1967, Line profiles of neutron powder-diffraction peaks for structure refinement: Acta Crystallographica, v. 22, p. 151–152, doi:10.1107/S0365110X67000234.226Ritter, H. L., and L. Drake, 1949, Pore-size distribution in porous materials: Industrial & Engineering Chemistry, v. 17, no. 12, p. 782–786.Robinson, D., L. N. Warr, and R. E. Bevins, 1990, The illite crystallinity technique: a critical appraisal of its precision: Journal of Metamorphic Geology, v. 8, p. 333–344, doi:10.1111/j.1525-1314.1990.tb00476.x.Rogers, S. J., H. C. Chen, D. C. Kopaska-Merkel, and J. H. Fang, 1995, Predicting permeability from porosity using artificial neural networks: AAPG Bulletin, v. 79, no. 12, p. 1786–1797, doi:10.1306/7834DEFE-1721-11D7-8645000102C1865D.Romero-Sarmiento, M. F., J. N. Rouzaud, S. Bernard, D. Deldicque, M. Thomas, and R. Littke, 2014, Evolution of Barnett Shale organic carbon structure and nanostructure with increasing maturation: Organic Geochemistry, v. 71, p. 7–16, doi:10.1016/j.orggeochem.2014.03.008.Rootare, H., and C. Prenzlow, 1967, Surface areas from mercury porosimeter measurements: The Journal of Physical Chemistry, v. 71, p. 2733–2736.Ross, D. J. K., and R. M. Bustin, 2007, Shale gas potential of the lower Jurassic Gordondale member, northeastern British Columbia, Canada: Bulletin of Canadian Petroleum Geology, v. 55, no. 1, p. 51–75.Ross, D. J. K., and R. M. Bustin, 2008, Characterizing the shale gas resource potential of Devonian–Mississippian strata in the Western Canada sedimentary basin: Application of an integrated formation evaluation: AAPG Bulletin, v. 92, no. 1, p. 87–125, doi:10.1306/09040707048.Ross, D. J. K., and R. M. Bustin, 2009, The importance of shale composition and pore structure upon gas storage potential of shale gas reservoirs: Marine and Petroleum Geology, v. 26, no. 6, p. 916–927, doi:10.1016/j.marpetgeo.2008.06.004.Ross, G. M., and R. A. Stephenson, 1989, Crystalline basement: the foundations of the Western Canada Sedimentary Basin, in: Western Canada Sedimentary Basin: A Case History, B. D. Ricketts, (ed.), Canadian Society of Petroleum Geologists, p. 33-46.Rouquerol, F., J. Rouquerol, K. S. W. Sing, P. Llewellyn, and G. Maurin, 2013, Adsorption by powders and porous Solids: Poland, Elsevier, 646 p.Rouquerol, J., D. Avnir, C. W. Fairbridge, D. H. Everett, J. H. Haynes, N. Pernicone, J. D. F. Ramsay, K. S. W. Sing, and K. K. Unger, 1994, Recommendations for the characterization of porous solids: Pure & Appl. Chem., v. 66, no. 8, p. 1739–1758, doi:10.1351/pac199466081739.Saemi, M., M. Ahmadi, and A. Y. Varjani, 2007, Design of neural networks using genetic algorithm for the permeability estimation of the reservoir: Journal of Petroleum Science and Engineering, v. 59, no. 1-2, p. 97–105, doi:10.1016/j.petrol.2007.03.007.Schlömer, S., and B. Krooss, 1997, Experimental characterisation of the hydrocarbon sealing efficiency of cap rocks: Marine and Petroleum Geology, v. 14, no. 5, p. 565–580.Schmoker, J. W., 1981, Determination of organic matter content of Appalachian Devonian Shale from gamma ray logs: AAPG Bulletin, v. 65, p. 1285–1298, doi:10.1306/03B5949A-16D1-11D7-8645000102C1865D.227Schmoker, J. W., and T. C. Hester, 1983, Organic Carbon in Bakken Formation, United States Portion of Williston Basin: AAPG Bulletin, v. 67, no. 12, p. 2165–2174, doi:10.1306/AD460931-16F7-11D7-8645000102C1865D.Schneider, J., P. B. Flemings, R. J. Day-Stirrat, and J. T. Germaine, 2011, Insights into pore-scale controls on mudstone permeability through resedimentation experiments: Geology, v. 39, no. 11, p. 1011–1014, doi:10.1130/G32475.1.Shafer, J., and J. Neasham, 2000, Mercury porosimetry protocol for rapid determination of petrophysical and reservoir quality properties: Society of Core Analysis, SCA, p. 1–12.Sing, K., D. H. Everett, R. A. W. Haul, L. Moscou, R. A. Pierotti, J. Rouquerol, and T. Siemieniewska, 1985, Reporting physisorption data for gas/solid systems with special reference to the determination of surface area and porosity: Pure & Appl. Chem., v. 57, no. 4, p. 603–619.Slatt, R. M., and N. R. O’Brien, 2011, Pore types in the Barnett and Woodford gas shales: Contribution to understanding gas storage and migration pathways in fine-grained rocks: AAPG Bulletin, v. 95, no. 12, p. 2017–2030, doi:10.1306/03301110145.Smith Low, W., 2012, The New Millennium Gold Rush: BMO Capital Markets Oil & Gas A&A Advisory, June, p. 1–44.Smith Low, W., J. Kennedy, G. Fisher, and M. Greene, 2013, The Duvernay shale: the new millennium gold rush is condensate: BMO Capital Markets Oil & Gas A&A Advisory, November, p. 1–18.Smith, S. T., R. L. Snyder, and W. E. Brownwell, 1979a, Minimisation of preferred orientation in powders by spray-drying: Advances in X-ray Analysis, v. 22, p. 77–87.Smith, S. T., R. L. Snyder, and W. E. Brownwell, 1979b, Quantitative phase analysis of Devonian shales by computer controlled X-ray diffraction of spray-dried samples: Advances in X-ray Analysis, v. 22, p. 181–191.Snyder, R. L., and D. L. Bish, 1989, Quantitative analysis, in: Modern powder diffraction: Reviews in mineralogy v. 20, D. L. Bish, and J. E. Post, (eds.), p. 101–144.Spitzer, Z., 1981, Mercury porosimetry and its application to the analysis of coal pore structure: Powder Technology, v. 29, p. 177–186.Środoń, J., V. A. Drits, D. K. McCarty, J. C. C. Hsieh, and D. D. Eberl, 2001, Quantitative X-ray diffraction analysis of clay-bearing rocks from random preparations: Clays and Clay Minerals, v. 49, no. 6, p. 514–528.Stasiuk, L. D. and M. G. Fowler, 2002, Thermal maturity evaluation (vitrinite and vitrinite reflectance equivalent) of Middle Devonian, Upper Devonian, and Mississippian strata in the Western Canada Sedimentary basin: Geological Survey of Canada, Open file 4341.Stoakes, F. A., 1980, Nature and control of shale basin fill and its effect on reef growth and termination: Upper Devonian Duvernay and Ireton Formations of Alberta, Canada: Bulletin of Canadian Petroleum Geology, v. 28, no. 3, p. 345–410.Stoakes, F. A., 1992, Winterburn megasequence, in: Devonian-Early Mississippian carbonates of the 228Western Canada Sedimentary Basin: A sequence stratigraphic framework, J. C. Wendte, F. A. Stoakes, and C. V. Campbell, (authors), Society for Sedimentary Geology Short Course No. 28, p. 207-224Stoakes, F. A., and J. C. Wendte, 1987, The Woodbend Group: Devonian Lithofacies and Reservoir Styles in Alberta: 13th CSPG Core Conference and Display, Calgary, Alberta, p. 153-170.Stoakes, F. A., and S. Creaney, 1984, Sedimentology of a carbonate source rock: the Duvernay Formation of central Alberta, in: Carbonates in Subsurface and Outcrop, L. Eliuk, (eds.), CSPG Core Conference, Calgary, Alberta, p. 132-147.Suarez-Rivera, R., and E. Fjær, 2013, Evaluating the poroelastic effect on anisotropic, organic-rich, mudstone systems: Rock Mechanics and Rock Engineering, v. 46, no. 3, p. 569–580, doi:10.1007/s00603-013-0374-y.Suuberg, E., S. Deevi, and Y. Yun, 1995, Elastic behaviour of coals studied by mercury porosimetry: Fuel, v. 74, no. 10, p. 1522–1530.Swanson, B., 1981, A simple correlation between permeabilities and mercury capillary pressures: Journal of Petroleum Technology, v. 33, no. 12, p. 2498–2504.Switzer, S. B., W. G. Holland, D. S. Christie, G. C. Graf, A. S. Hedinger, R. J. McAuley, R. A. Wierzbicki, and J. J. Packard, 1994, Devonian Woodbend-Winterburn strata of the Western Canadian Sedimentary Basin, in: Geological Atlas of the Western Canada Sedimentary Basin, G.D. Mossop and I. Shetsen, (comps.), Canadian Society of Petroleum Geologists and the Alberta Research Council, p. 165-202.Theisen, A. A., and M. E. Harward, 1962, A paste method for preparation of slides for clay mineral identification by X-ray diffraction: Soil Science Society of America Journal, v. 26, p. 90–91.Thomas, Jr., J., and B. Bohor, 1968, Surface area of montmorillonite from the dynamic sorption of nitrogen and carbon dioxide: Clays and Clay Minerals, v. 16, p. 83–91.Tissot, B.P., and D. H. Welte, 1984, Petroleum formation and occurrence: Berlin, Springer-Verlag, 699 p.Toda, Y., and S. Toyoda, 1972, Application of mercury porosimetry to coal: Fuel, v. 51, no. 3, p. 199–201, doi:10.1016/0016-2361(72)90080-4.Ufer, K., H. Stanjek, G. Roth, R. Dohrmann, R. Kleeberg, and S. Kaufhold, 2008, Quantitative phase analysis of bentonites by the Rietveld method: Clays and Clay Minerals, v. 56, no. 2, p. 272–282, doi:10.1346/CCMN.2008.0560210.Ugolini, F. C., S. Hillier, G. Certini, and M. J. Wilson, 2008, The contribution of aeolian material to an Aridisol from southern Jordan as revealed by mineralogical analysis: Journal of Arid Environments, v. 72, p. 1431–1447, doi:10.1016/j.jaridenv.2008.02.014.Unsworth, J. F., C. S. Fowler, and L. F. Jones, 1989, Moisture in coal: Fuel, v. 68, no. 1, p. 18–26, doi:10.1016/0016-2361(89)90005-7.Vasseur, G., I. Djeran-Maigre, D. Grunberger, G. Rousset, D. Tessier, and B. Velde, 1995, Evolution of structural and physical parameters of clays during experimental compaction: Marine and 229Petroleum Geology, v. 12, no. 8, p. 941–954, doi:10.1016/0264-8172(95)98857-2.Wagner, Z., 1983, Empirical correction factors involved in mercury porosimetry: Powder Technology, v. 35, p. 83–87.Wang, Y., Y. Zhu, S. Chen, and W. Li, 2014, Characteristics of the nanoscale pore structure in northwestern Hunan shale gas reservoirs using field emission scanning electron microscopy, high-pressure mercury intrusion, and gas adsorption: Energy & Fuels, v. 28, p. 945–955.Washburn, E., 1921, The dynamics of capillary flow: Physical review, v. 17, no. 3, p. 273–283.Webb, P., 2001, An introduction to the physical characterization of materials by mercury intrusion porosimetry with emphasis on reduction and presentation of experimental data: MicroMeritics Instrument Corp, Norcross, Georgia.Weissenberger, J. a W., 1994, Frasnian reef and basinal strata of west central Alberta: a combined sedimentological and biostratigraphic analysis: Bulletin of Canadian Petroleum Geology, v. 42, no. 1, p. 1–25.Wendte, J. C., 1994, Cooking Lake platform evolution and its control on late Devonian Leduc reef inception and localization, Redwater, Alberta: Bulletin of Canadian Petroleum Geology, v. 42, no. 4, p. 499–528.Wentworth, C., 1922, A scale of grade and class terms for clastic sediments: The Journal of Geology, v. 30, no. 5, p. 377–392.Wignall, P. B., 1994, Black shales: Oxford, Oxford University Press, 130 p.Wollenweber, J., S. Alles, A. Busch, B. M. Krooss, H. Stanjek, and R. Littke, 2010, Experimental investigation of the CO2 sealing efficiency of caprocks: International Journal of Greenhouse Gas Control, v. 4, no. 2, p. 231–241, doi:10.1016/j.ijggc.2010.01.003.Wong, P. M., F. X. Jian, and I. J. Taggart, 1995, A critical comparison of neural networks and discriminant analysis in lithofacies, porosity and permeability predictions: Journal of Petroleum Geology, v. 18, no. 2, p. 191–206.Wright, G. N., M. E. McMechan, D. E. G. Potter, 1994, Structure and architecture of the western Canada Sedimentary Basin, in: Geological Atlas of the Western Canada Sedimentary Basin, G.D. Mossop and I. Shetsen, (comps.), Canadian Society of Petroleum Geologists and the Alberta Research Council, p. 25-40.Wust, R. A. J., B. R. Nassichuk, and R. M. Bustin, 2013, Porosity characterization of various organic-rich shales from the Western Canadian Sedimentary Basin, Alberta and British Columbia, Canada: p. 81–100.Yang, Y., A. C. Aplin, and S. R. Larter, 2004, Quantitative assessment of mudstone lithology using geophysical wireline logs and artificial neural networks: Petroleum Geoscience, v. 10, no. 2, p. 141–151, doi:10.1144/1354-079302-566.Yang, Y., and A. C. Aplin, 2007, Permeability and petrophysical properties of 30 natural mudstones: Journal of Geophysical Research: Solid Earth, v. 112, no. 3, doi:10.1029/2005JB004243.230Zendehboudi, S., M. A. Ahmadi, L. James, and I. Chatzis, 2012, Prediction of condensate-to-gas ratio for retrograde gas condensate reservoirs using artificial neural network with particle swarm optimization: Energy & Fuels, v. 26, p. 3432–3447.Zevin, L. S., and G. Kimmel, 1995, Quantitative X-ray diffractometry: New York, Springer-Verlag, Inc., 972 p.Zimmerman, R., W. Somerton, and M. King, 1986, Compressibility of porous rocks: Journal of Geophysical Research, v. 91, no. B12, p. 12765–12777.231aPPenDIX	 a:	 MIneRaLogy	XRD	ReSULtS232Sample ID Depth Albite Ankerite Calcite Chlorite Dolomite Illite Kaolinite Orthoclase Pyrite Quartzm wt. % wt. % wt. % wt. % wt. % wt. % wt. % wt. % wt. % wt. %ATH1-24-2 3681.18 3.4 6.6 19.2 8.8 - 26.4 - 5.1 2.7 27.8ATH1-24-3 3683.42 5.1 2.5 18.3 6.1 7.6 17.4 - 5.6 1.9 35.5ATH1-24-5 3685.07 2.6 27.7 10.4 2.0 - 12.4 - 3.3 1.2 40.5ATH1-24-6 3686.55 5.0 11.2 5.9 2.6 - 18.4 - 11.6 2.7 42.6ATH1-24-8 3689.66 3.9 4.8 7.2 1.4 - 17.1 - 11.8 2.3 51.5ATH1-24-10 3691.53 - 9.8 79.2 - - 2.1 - - 0.4 8.4ATH1-24-12 3692.86 3.3 - 11.0 5.3 22.6 - - 19.8 3.3 34.8ATH1-24-13 3694.40 0.8 - 94.7 - - 1.7 - - - 2.7ATH1-24-14 3695.48 4.0 6.6 9.2 - - 29.8 - 16.4 2.5 31.4ATH1-24-16 3696.53 3.8 5.1 8.6 - 0.8 24.3 - 17.8 1.9 37.7ATH1-24-17 3697.30 0.2 7.5 44.1 - 0.7 11.4 - 11.1 2.3 22.7ATH1-24-19 3699.50 2.7 4.0 13.8 - - 15.3 - 19.0 3.5 41.7ATH1-24-21 3700.00 3.9 - 15.7 - 5.1 2.1 - 20.6 3.4 49.1ATH1-24-22 3700.62 2.8 3.5 17.3 - - 16.4 - 21.5 4.1 34.4ATH1-24-24 3702.00 3.4 2.2 16.1 - - 14.3 - 18.8 3.5 41.8ATH1-24-25 3703.23 4.3 - 12.6 - 2.9 22.2 - 26.6 3.0 28.4ATH1-24-26 3703.63 - 2.6 84.9 - - - - - 0.4 12.1ATH1-24-27 3704.62 3.7 - 10.8 - 1.8 21.4 - 29.2 5.1 28.1ATH1-24-29 3706.40 6.3 - 9.9 - 2.0 23.7 - 25.0 4.2 28.9ATH1-24-30 3707.15 2.8 1.5 12.2 - - 20.8 - 23.6 3.4 35.8ATH1-24-31 3708.00 4.1 8.2 31.8 - - 12.1 - 9.4 3.3 31.1ATH1-24-33 3709.98 1.6 9.2 40.8 - - 7.6 - 8.1 3.1 29.7ATH1-24-35 3711.00 2.4 4.2 28.6 - 3.2 12.1 - 10.5 3.2 35.7ATH1-24-36 3712.40 1.8 1.9 16.9 - - 7.6 - 10.5 1.9 59.5ATH1-24-37 3712.70 2.3 - 23.5 - 2.1 10.6 - 14.0 2.5 45.0ATH1-24-39 3713.80 3.7 - 4.8 - 1.2 21.5 - 32.4 4.4 31.9ATH1-24-40 3715.32 0.5 5.9 80.0 - - 1.4 - 2.0 0.4 9.7ATH1-24-41 3715.73 0.1 5.9 83.6 - - 2.6 - 1.8 0.7 5.3ATH1-24-43 3715.93 - 8.6 81.8 - - 1.8 - 1.7 0.5 5.5ATH1-24-46 3719.53 - 2.4 93.6 - - - - - - 4.0ATH1-24-47 3720.70 - 3.6 94.1 - - - - - - 2.3ATH1-24-48 3721.30 - 3.3 92.8 - - - - - - 3.9ATH1-24-49 3722.75 1.8 2.5 43.6 - - 11.4 - 10.3 2.1 28.3ATH1-24-50 3724.26 4.1 - 9.6 - 1.1 24.4 - 21.3 4.8 34.7ATH4-2-1 3606.24 3.4 7.8 15.3 7.8 - 33.6 - 3.7 2.0 26.5ATH4-2-2 3606.26 3.6 7.2 11.8 7.3 - 35.8 - 3.6 2.0 28.6ATH4-2-3 3608.24 3.6 6.2 11.5 9.4 - 34.2 - 4.8 1.9 28.4ATH4-2-4 3608.94 3.7 5.9 11.4 9.9 - 34.0 - 4.9 2.3 27.8ATH4-2-5 3608.97 3.3 6.3 13.6 9.3 - 33.1 - 4.8 2.2 27.4ATH4-2-6 3610.33 3.9 7.3 11.0 6.9 - 34.1 - 4.8 2.1 29.9ATH4-2-7 3610.56 2.5 10.5 26.9 6.1 - 25.3 - 3.4 2.2 23.0ATH4-2-8 3612.04 3.4 8.2 11.2 7.4 - 33.1 - 4.6 1.8 30.2ATH4-2-9 3613.37 3.7 10.9 10.9 6.5 - 30.7 - 4.7 2.9 29.8Table A.1 XRD results for detailed analysis wells233Sample ID Depth Albite Ankerite Calcite Chlorite Dolomite Illite Kaolinite Orthoclase Pyrite Quartzm wt. % wt. % wt. % wt. % wt. % wt. % wt. % wt. % wt. % wt. %ATH4-2-10 3613.85 3.5 7.1 20.7 4.6 - 22.9 - 4.0 2.9 34.2ATH4-2-11 3614.95 3.2 5.1 13.2 2.4 - 17.0 - 5.9 1.8 51.1ATH4-2-12 3615.25 2.5 12.5 11.9 2.4 - 15.3 - 4.0 1.8 49.6ATH4-2-13 3615.28 3.0 8.3 13.6 2.3 - 19.2 - 6.7 1.8 45.2ATH4-2-14 3616.84 3.8 9.7 16.7 2.0 - 15.9 - 3.4 2.1 46.4ATH4-2-15 3618.34 3.4 3.3 14.1 1.4 - 16.5 - 6.2 2.0 53.2ATH4-2-16 3618.45 2.8 5.3 14.3 1.2 - 9.1 - 4.9 1.3 61.1ATH4-2-17 3619.02 3.5 4.8 17.7 1.2 - 12.4 - 5.2 2.3 52.9ATH4-2-18 3620.91 5.0 4.9 15.7 2.7 - 25.3 - 7.9 2.2 36.3ATH4-2-19 3621.17 4.8 4.3 13.5 4.2 - 25.5 - 8.0 2.9 36.8ATH4-2-20 3622.80 5.3 5.9 8.0 3.8 - 26.4 - 8.9 2.4 39.3ATH4-2-21 3624.33 4.2 5.6 7.9 3.7 - 23.6 - 8.4 1.6 45.0ATH4-2-22 3624.53 4.7 6.1 9.9 3.3 - 22.2 - 8.4 1.8 43.6ATH4-2-23 3626.10 4.0 5.9 14.6 2.9 - 20.8 - 8.6 1.9 41.4ATH4-2-24 3627.30 3.5 5.4 14.2 2.3 - 20.9 - 8.7 2.0 43.0ATH4-2-25 3627.42 3.8 5.2 13.5 2.2 - 21.5 - 8.4 1.8 43.7ATH4-2-26 3628.75 4.6 7.3 7.0 2.8 - 23.0 - 10.7 1.3 43.4ATH4-2-27 3630.29 4.0 5.3 18.1 1.7 - 12.4 - 8.5 1.4 48.8ATH4-2-28 3630.41 5.1 6.0 21.9 1.0 - 12.7 - 8.4 2.3 42.6ATH4-2-29 3630.59 5.8 5.8 33.1 1.8 - 9.8 - 6.0 2.3 35.5ATH4-2-30 3631.99 4.0 4.8 14.7 1.0 - 15.8 - 11.1 4.0 44.6ATH4-2-31 3632.80 3.8 4.6 12.8 1.0 - 18.8 - 12.1 2.5 44.4ATH4-2-32 3632.94 4.7 5.4 11.7 - - 18.5 2.5 11.8 2.6 42.8ATH4-2-33 3634.23 4.0 4.7 18.0 - - 14.6 1.3 12.5 2.8 42.0ATH4-2-34 3634.79 3.0 5.7 14.8 - - 11.3 1.7 10.4 2.5 50.4ATH4-2-35 3635.56 4.3 4.7 13.9 - - 18.5 1.3 11.0 2.2 44.1ATH4-2-36 3636.10 3.4 4.5 13.3 - - 16.6 1.9 12.1 2.6 45.6ATH4-2-37 3636.62 3.4 3.5 11.4 - - 17.1 2.3 13.3 2.0 47.0ATH4-2-38 3638.41 3.6 4.3 17.9 3.1 - 16.7 - 12.9 2.4 39.1ATH4-2-39 3639.62 4.2 3.6 16.1 1.8 - 19.7 - 12.8 2.4 39.5ATH4-2-40 3639.70 4.0 4.3 11.9 0.6 - 20.1 2.4 16.3 2.6 37.8ATH4-2-41 3641.72 3.6 3.4 7.4 1.3 - 26.5 2.3 17.2 2.5 35.7ATH4-2-42 3641.86 4.3 3.0 5.7 4.5 - 23.3 2.7 15.2 3.1 38.3ATH4-2-43 3642.22 4.5 4.3 14.8 2.7 - 21.8 - 15.2 3.8 32.9ATH4-2-44 3643.95 2.6 13.9 18.5 - - 17.3 1.6 13.1 3.0 30.0ATH4-2-45 3644.14 2.6 23.8 30.7 - - 9.3 - 5.5 1.2 26.9ATH4-2-46 3644.75 3.5 24.4 20.0 - - 15.5 - 8.6 1.8 26.3ATH4-2-47 3645.43 3.2 3.7 7.3 2.1 - 20.0 - 18.8 2.7 42.2ATH4-2-48 3646.61 2.1 2.6 3.6 1.7 - 18.3 - 27.9 2.7 41.1ATH4-2-49 3647.75 3.0 4.6 20.5 - - 12.2 1.7 15.0 2.8 40.2ATH4-2-50 3649.14 4.0 3.9 14.5 0.9 - 18.8 - 11.2 3.9 42.7ATH13-18-1 2754.95 2.1 4.0 44.6 7.2 - 22.0 - 3.2 0.9 16.0234Sample ID Depth Albite Ankerite Calcite Chlorite Dolomite Illite Kaolinite Orthoclase Pyrite Quartzm wt. % wt. % wt. % wt. % wt. % wt. % wt. % wt. % wt. % wt. %ATH13-18-2 2756.10 2.0 3.2 46.1 6.3 1.3 21.4 - 3.5 1.2 15.0ATH13-18-3 2759.27 1.0 3.6 51.2 5.9 - 20.1 - 3.1 1.0 14.0ATH13-18-4 2759.57 2.1 3.7 46.4 7.0 - 23.1 - 2.0 0.8 14.9ATH13-18-5 2761.00 2.1 4.0 26.7 8.0 0.4 32.3 - 4.3 1.3 20.8ATH13-18-6 2762.65 3.1 5.1 8.6 5.6 - 32.2 - 3.9 3.0 38.5ATH13-18-7 2763.30 2.1 4.8 14.6 1.9 - 15.6 - 2.3 2.9 55.8ATH13-18-8 2764.65 2.7 4.5 12.7 1.2 - 13.9 - 3.2 2.3 59.4ATH13-18-9 2766.65 3.4 4.6 16.7 2.4 - 18.2 - 3.4 2.7 48.6ATH13-18-10 2768.00 2.2 3.3 19.4 2.1 0.8 16.4 - 3.4 2.7 49.7ATH13-18-11 2769.00 2.1 4.1 18.2 1.3 2.0 14.6 - 3.0 2.1 52.7ATH13-18-12 2769.50 3.6 3.9 18.0 4.6 - 12.1 - 3.7 2.7 51.3ATH13-18-13 2771.00 2.8 6.7 21.9 2.2 - 16.5 - 4.3 2.8 42.9ATH13-18-14 2771.46 2.0 6.3 24.6 2.7 - 20.2 - 3.6 2.5 38.1ATH13-18-15 2771.50 3.6 6.5 23.5 1.6 - 15.7 - 4.4 2.1 42.5ATH13-18-16 2773.55 4.5 2.4 15.3 4.1 - 28.3 - 4.4 4.0 37.1ATH13-18-17 2774.85 3.0 3.5 10.6 2.4 - 18.7 - 3.1 2.9 55.8ATH13-18-18 2774.97 3.1 4.1 27.6 1.9 - 16.4 - 6.1 3.1 37.7ATH13-18-19 2776.30 3.5 5.8 9.8 - - 13.2 0.9 2.4 2.5 62.0ATH13-18-20 2776.78 3.4 3.5 18.4 2.1 - 17.7 - 4.3 3.6 47.0ATH13-18-21 2777.04 4.1 4.5 14.6 2.1 - 16.3 - 2.9 2.8 52.6ATH13-18-22 2777.20 2.3 5.2 13.2 1.0 - 8.6 - 5.0 1.7 63.0ATH13-18-23 2779.40 3.3 4.0 13.1 1.2 - 10.1 - 1.6 2.3 64.5ATH13-18-24 2780.36 2.5 4.1 11.4 1.3 - 11.1 - 3.8 1.8 64.0ATH13-18-25 2782.30 3.7 6.0 13.2 1.8 - 16.1 - 7.9 2.9 48.3ATH13-18-26 2782.48 2.7 5.0 11.1 1.3 - 11.1 - 5.9 2.2 60.8ATH13-18-27 2783.15 3.9 4.7 10.6 2.3 - 15.6 - 4.6 2.7 55.5ATH13-18-28 2783.55 4.9 5.3 15.9 2.3 - 14.7 - 4.8 3.5 48.6ATH13-18-29 2784.50 4.9 5.1 12.3 2.7 - 16.8 - 3.8 3.0 51.4ATH13-18-30 2785.90 2.2 5.0 61.4 1.9 - 10.8 - 2.6 1.5 14.7ATH13-18-31 2786.11 2.2 5.7 53.6 2.2 - 13.2 - 3.6 1.9 17.7ATH13-18-32 2786.20 1.9 5.5 53.6 2.6 - 13.6 - 2.9 2.0 18.0ATH13-18-33 2787.67 2.8 6.0 46.3 2.7 1.1 15.1 - 5.4 2.3 18.3ATH13-18-34 2787.95 3.9 7.8 34.8 3.6 - 18.3 - 5.3 1.9 24.5ATH13-18-35 2787.98 4.3 8.2 24.7 3.4 - 21.7 - 6.4 2.3 29.0ATH13-18-36 2788.35 3.9 8.5 12.7 4.1 - 24.6 - 6.0 4.6 35.5ATH13-18-37 2790.00 4.3 4.5 13.1 2.4 1.2 17.9 - 5.5 2.8 48.2ATH13-18-38 2790.02 5.2 6.2 11.9 3.1 - 16.0 - 5.7 3.3 48.5ATH13-18-39 2790.30 5.7 6.2 9.8 2.7 - 15.9 - 5.4 6.1 48.2ATH13-18-40 2790.72 4.2 3.0 10.5 3.9 - 21.6 - 3.9 3.6 49.1ATH13-18-41 2791.12 4.8 3.2 12.1 3.4 - 21.5 - 5.6 3.5 45.9ATH13-18-42 2791.66 3.7 3.1 27.9 3.5 - 21.5 - 6.1 3.1 31.1ATH13-18-43 2791.80 1.9 2.7 40.1 3.5 - 16.1 - 3.8 5.5 26.4235Sample ID Depth Albite Ankerite Calcite Chlorite Dolomite Illite Kaolinite Orthoclase Pyrite Quartzm wt. % wt. % wt. % wt. % wt. % wt. % wt. % wt. % wt. % wt. %ATH13-18-44 2792.30 - 1.4 82.0 - - 7.0 - 1.3 1.6 6.7ATH13-18-46 2793.37 0.4 3.0 87.5 0.8 - 3.3 - - - 5.0ATH13-18-48 2794.77 0.7 0.7 92.6 1.5 - 1.8 - 0.7 - 2.0ATH13-18-51 2796.26 0.0 2.5 76.2 2.1 - 8.4 - 1.3 8.7 0.8ATH13-18-52 2797.00 - - 86.2 1.6 - 4.7 - 0.7 1.8 4.9ATH13-18-54 2798.32 - 3.4 68.7 3.3 - 11.5 - 1.9 0.7 10.5ECA11-8-2 3794.35 4.5 2.9 20.0 4.2 - 24.1 - 6.3 2.6 35.4ECA11-8-3 3795.65 4.8 5.2 16.4 3.1 - 21.9 - 4.6 2.4 41.6ECA11-8-5 3797.75 5.0 4.9 22.2 2.6 - 20.1 - 4.2 1.8 39.3ECA11-8-8 3802.40 4.7 4.5 35.9 1.3 - 10.8 - 6.8 1.5 34.4ECA11-8-9 3802.91 3.7 3.0 54.1 - - 7.9 - 3.5 1.6 26.2ECA11-8-11 3804.75 4.5 4.6 9.6 3.1 - 28.7 - 7.8 2.3 39.5ECA11-8-12 3805.10 4.8 3.3 9.0 3.6 - 32.7 - 10.1 1.8 34.7ECA11-8-17 3811.25 4.9 5.7 23.2 - - 13.8 - 6.7 2.4 43.4ECA11-8-19 3812.70 3.9 4.9 22.8 - - 14.3 - 7.0 4.1 42.4ECA11-8-21 3813.35 4.6 4.8 22.6 - - 15.6 - 5.3 4.1 43.1ECA11-8-22 3813.95 4.4 3.6 13.2 - - 17.9 - 8.9 4.7 46.8ECA11-8-23 3814.00 4.5 3.8 12.9 - - 18.2 - 8.3 7.2 45.0ECA11-8-24 3816.20 6.5 4.7 13.4 1.2 - 20.4 - 11.7 2.6 39.5ECA11-8-25 3816.70 3.7 3.0 9.9 3.1 - 18.9 - 8.1 5.4 47.8ECA11-8-26 3819.80 4.6 2.8 5.0 4.1 - 27.5 - 9.9 5.7 40.4ECA11-8-28 3823.40 4.8 4.4 10.5 1.2 - 19.0 - 13.2 1.7 45.2ECA11-8-30 3826.35 - - - - - - - - - -ECA11-8-31 3828.50 3.2 2.8 23.3 2.2 - 14.8 - 9.6 6.3 37.7ECA11-8-32 3830.05 3.1 4.8 22.3 2.7 - 19.4 - 12.1 2.3 33.3ECA11-8-33 3831.15 3.1 1.7 38.2 - - 19.0 - 9.9 3.0 25.1ECA11-8-36 3833.00 3.2 11.2 31.8 3.8 - 19.2 - 8.9 2.6 19.3ECA11-8-37 3833.60 3.2 4.3 12.9 8.7 - 27.3 - 11.5 2.4 29.6ECA11-8-38 3835.75 3.7 - 12.0 6.0 3.5 34.7 - 7.8 3.3 29.0ECA11-8-39 3836.85 3.7 - 11.9 5.2 2.7 39.6 - 8.5 1.4 27.1HSK5-11-1 3049.20 3.4 12.5 15.8 6.2 - 32.7 - 4.4 1.7 23.4HSK5-11-2 3051.95 3.6 4.3 11.4 1.6 - 13.8 - 5.3 2.6 57.4HSK5-11-3 3052.95 2.6 6.8 18.3 1.0 - 12.8 - 3.8 3.0 51.6HSK5-11-4 3054.60 3.2 5.6 12.4 0.7 - 15.7 - 6.7 2.7 53.0HSK5-11-5 3056.13 2.6 8.2 15.5 - - 13.2 1.2 5.7 1.9 51.7HSK5-11-6 3057.69 3.1 8.7 27.4 - - 16.5 1.7 7.1 3.6 31.9HSK5-11-7 3059.35 3.3 10.7 15.2 - - 19.7 - 8.4 4.2 38.5HSK5-11-8 3060.30 2.1 4.9 21.9 - - 4.8 - 4.1 1.7 60.5HSK5-11-9 3062.00 3.4 8.0 28.3 - - 12.2 - 9.1 3.2 35.8HSK5-11-10 3064.15 2.6 5.1 14.1 - - 13.5 - 12.2 5.2 47.3HSK5-11-11 3065.20 3.8 5.7 13.9 - - 20.3 1.8 13.2 3.1 38.1HSK5-11-12 3066.60 4.7 6.9 12.3 0.9 - 24.9 0.2 11.9 3.3 34.7236Sample ID Depth Albite Ankerite Calcite Chlorite Dolomite Illite Kaolinite Orthoclase Pyrite Quartzm wt. % wt. % wt. % wt. % wt. % wt. % wt. % wt. % wt. % wt. %HSK5-11-13 3068.50 3.5 8.8 41.6 - - 16.9 - 7.9 2.1 19.3HSK5-11-14 3069.60 2.8 7.2 51.7 - - 12.1 - 7.5 3.3 15.4HSK5-11-15 3071.25 3.0 5.1 19.0 - - 14.0 - 10.7 2.2 45.9HSK5-11-16 3072.96 2.7 5.3 19.0 - - 13.1 - 11.8 2.5 45.8HSK5-11-17 3074.60 2.5 6.0 19.9 - - 13.7 - 12.7 3.1 42.1HSK5-11-18 3076.25 2.0 7.5 23.5 - - 13.9 - 11.2 2.6 39.4HSK5-11-19 3077.65 1.3 7.7 19.2 - - 12.9 - 9.2 4.3 45.5HSK5-11-20 3078.95 2.7 14.0 19.8 - - 13.1 - 9.4 2.0 39.1HSK5-11-21 3080.80 2.7 23.1 21.6 - - 12.5 - 9.2 2.3 28.8HSK5-11-22 3082.60 2.5 6.4 22.6 1.4 - 15.3 - 15.4 1.8 34.7HSK5-11-23 3085.17 1.6 4.2 26.1 - - 9.8 - 9.3 1.6 47.3HSK5-11-24 3089.20 4.3 4.7 15.8 4.5 - 33.4 - 13.3 3.2 20.8HSK5-11-25 3093.60 2.5 3.6 10.2 - - 20.2 2.5 14.4 2.9 43.8HSK5-11-26 3095.90 3.1 2.2 15.2 - - 23.1 2.4 17.5 3.8 32.8HSK5-11-27 3096.90 1.2 2.3 66.3 - - 4.1 - 3.9 2.1 20.2HSK5-11-28 3099.85 1.7 5.1 46.7 - - 13.8 0.2 9.0 3.7 19.8HSK5-11-29 3100.80 3.7 3.5 18.3 6.8 - 31.1 - 10.7 2.1 23.8HSK8-25-1 3102.70 2.7 12.0 34.6 5.2 - 23.3 - 3.1 1.2 17.9HSK8-25-2 3104.19 1.9 9.1 45.2 4.5 - 20.5 - 2.7 0.9 15.2HSK8-25-3 3106.48 2.1 9.6 52.4 3.9 - 16.5 - 2.0 0.6 12.8HSK8-25-4 3108.17 5.1 5.8 6.8 7.0 - 31.2 - 5.1 2.5 36.6HSK8-25-5 3109.62 2.6 2.9 10.8 2.0 1.4 21.0 - 5.9 3.4 50.1HSK8-25-6 3111.38 3.1 5.1 15.4 1.5 - 20.6 - 5.9 3.4 45.0HSK8-25-7 3113.47 2.8 7.3 23.6 - - 11.9 - 3.6 2.5 48.3HSK8-25-8 3114.69 2.7 7.2 39.3 - - 9.5 - 3.9 2.2 35.2HSK8-25-9 3116.10 2.7 6.1 39.4 1.5 - 15.3 - 4.6 3.1 27.2HSK8-25-10 3117.80 3.3 9.0 21.2 1.2 - 18.3 - 6.9 3.3 36.8HSK8-25-11 3119.20 3.2 6.9 33.2 - - 8.5 - 6.0 2.2 39.9HSK8-25-12 3121.84 2.8 6.5 10.5 - - 11.9 - 8.8 2.1 57.4HSK8-25-13 3122.96 3.2 5.7 12.7 - - 14.3 - 9.2 2.7 52.2HSK8-25-14 3124.20 3.5 4.9 18.6 4.5 - 20.9 - 9.1 4.5 34.1HSK8-25-15 3126.34 3.3 4.7 10.7 - - 30.0 1.0 12.2 3.7 34.4HSK8-25-16 3130.30 2.9 3.6 14.2 - 2.0 15.0 - 11.6 2.5 48.2HSK8-25-17 3132.70 3.1 5.4 15.4 - 0.9 12.9 - 9.9 2.9 49.5HSK8-25-18 3134.38 2.5 8.7 15.7 - - 16.6 - 10.8 2.5 43.2HSK8-25-19 3136.25 2.9 11.8 17.0 - - 15.3 - 10.6 1.7 40.8PWT10-17-1 2982.95 0.7 8.9 39.9 5.8 - 24.1 - 2.2 1.3 17.1PWT10-17-2 2989.80 1.3 3.5 33.6 6.8 - 27.5 - 2.1 2.1 23.3PWT10-17-3 2993.90 3.0 2.7 29.3 3.9 - 21.3 - 3.3 3.1 33.5PWT10-17-4 2996.74 1.6 5.6 46.0 3.2 - 13.7 - 3.0 2.4 24.5PWT10-17-5 3001.73 2.3 4.4 11.8 5.1 - 28.0 - 3.5 2.5 42.4PWT10-17-6 3004.77 1.6 2.5 50.5 4.0 - 14.0 - 2.9 3.9 20.6237Sample ID Depth Albite Ankerite Calcite Chlorite Dolomite Illite Kaolinite Orthoclase Pyrite Quartzm wt. % wt. % wt. % wt. % wt. % wt. % wt. % wt. % wt. % wt. %PWT10-17-7 3006.71 1.3 1.4 78.1 - - 6.0 - - 0.6 12.6PWT10-17-8 3007.42 1.7 2.6 25.6 5.4 - 24.3 - 3.7 4.7 32.0PWT10-17-9 3008.84 2.2 4.3 24.5 2.4 - 13.9 - 2.9 3.8 46.0PWT10-17-10 3009.24 2.3 2.8 20.3 2.7 - 15.4 - 2.5 3.0 51.0PWT10-17-11 3011.26 2.7 1.7 84.2 - - 2.5 - - 0.4 8.5PWT10-17-12 3013.21 2.0 1.4 12.4 4.9 1.2 25.0 - 4.8 5.0 43.3PWT10-17-13 3017.50 3.2 1.2 23.2 6.0 - 16.8 - 3.2 4.5 41.9PWT10-17-14 3019.78 2.0 4.6 28.0 4.6 1.8 18.1 - 3.6 4.9 32.6PWT10-17-15 3020.73 2.3 2.3 16.2 8.0 - 24.6 - 3.8 4.2 38.4PWT10-17-16 3026.20 1.8 1.6 26.6 10.5 - 27.1 - 3.8 1.4 27.3PWT10-17-17 3033.00 1.5 1.4 30.2 13.5 - 25.2 - 3.2 1.3 23.7PWT10-17-18 3036.37 3.5 2.2 19.6 8.7 - 30.8 - 5.1 1.9 28.2PWT10-17-19 3039.77 3.0 1.2 30.6 11.0 - 27.9 - 4.4 1.2 20.7PWT10-17-20 3046.85 1.2 3.5 39.6 7.4 - 23.9 - 3.4 3.2 17.8PWT10-17-21 3049.20 3.0 0.3 33.5 7.4 1.2 23.5 - 5.9 5.2 20.1TET10-18-1 2929.30 1.9 4.0 39.5 1.8 - 11.0 - 1.4 2.2 38.1TET10-18-2 2930.00 3.6 3.9 22.1 2.9 - 17.8 - 1.9 4.2 43.5TET10-18-3 2930.75 3.1 6.9 24.3 3.6 - 16.2 - 2.7 2.8 40.4TET10-18-4 2930.85 3.0 7.2 22.3 2.6 - 17.6 - 2.9 3.2 41.3TET10-18-5 2931.90 3.3 4.6 26.6 2.4 - 18.1 - 2.6 2.6 39.7TET10-18-6 2932.20 2.3 5.7 27.9 2.7 - 18.0 - 1.9 2.9 38.7TET10-18-7 2932.88 3.3 6.5 38.9 3.4 - 15.3 - 2.5 2.2 27.9TET10-18-8 2933.80 3.0 3.4 14.3 4.0 - 19.5 - 2.5 3.5 49.7TET10-18-9 2934.90 2.6 3.3 13.0 2.6 - 20.4 - 3.0 3.1 52.1TET10-18-10 2935.34 2.5 5.2 19.7 2.0 - 15.2 - 2.7 3.4 49.2TET10-18-11 2936.30 2.8 4.4 18.9 1.9 - 20.3 - 2.8 2.7 46.2TET10-18-12 2937.00 2.6 3.6 9.5 3.0 - 23.0 - 3.8 2.0 52.5TET10-18-13 2938.00 3.7 8.9 18.8 1.6 - 15.3 - 3.6 1.0 47.1TET10-18-14 2938.30 3.8 9.5 23.4 1.9 - 13.4 - 3.5 1.1 43.4TET10-18-15 2939.00 2.8 7.8 26.4 1.5 - 11.0 - 2.7 1.4 46.4TET10-18-16 2939.80 2.6 6.6 13.8 0.5 - 8.8 - 2.7 1.5 63.6TET10-18-17 2940.00 2.7 5.1 13.1 1.7 - 11.4 - 2.6 2.2 61.2TET10-18-18 2941.00 2.8 5.8 15.6 2.8 - 12.0 - 3.0 2.6 55.5TET10-18-19 2941.40 3.2 3.8 14.2 2.9 - 12.7 - 3.6 2.4 57.2TET10-18-20 2942.00 2.8 3.9 15.3 1.4 - 15.1 - 4.9 2.3 54.3TET10-18-21 2942.40 2.9 3.7 15.6 2.3 - 16.7 - 4.4 2.7 51.7TET10-18-22 2942.67 2.8 3.3 16.2 1.8 - 13.5 - 3.1 2.8 56.5TET10-18-23 2943.55 2.9 3.9 11.6 1.8 - 12.8 - 2.5 1.9 62.5TET10-18-24 2944.00 3.4 5.5 17.7 2.2 - 15.6 - 3.3 1.8 50.4TET10-18-25 2944.50 3.6 5.6 17.0 2.2 - 15.0 - 4.5 2.1 49.9TET10-18-26 2945.35 3.4 3.9 12.5 2.3 - 17.3 - 4.3 2.3 54.0TET10-18-27 2945.90 3.1 3.6 13.2 2.3 - 19.4 - 6.2 3.0 49.3238Sample ID Depth Albite Ankerite Calcite Chlorite Dolomite Illite Kaolinite Orthoclase Pyrite Quartzm wt. % wt. % wt. % wt. % wt. % wt. % wt. % wt. % wt. % wt. %TET10-18-28 2947.10 3.2 3.0 12.4 2.5 - 17.8 - 4.8 3.5 52.7TET10-18-29 2948.50 2.6 3.3 49.9 2.7 - 8.4 - 2.1 2.8 28.3TET10-18-30 2949.80 4.5 2.5 15.1 7.1 - 27.5 - 8.9 3.6 30.7TET10-18-31 2950.50 3.5 11.5 36.9 3.3 - 13.5 - 4.6 1.8 24.9TET10-18-32 2951.60 4.2 2.3 12.9 8.5 - 25.7 - 6.9 4.1 35.2TET10-18-33 2952.25 1.9 2.3 61.7 3.0 - 12.4 - 2.4 1.4 14.8TET10-18-34 2952.70 1.1 1.9 69.8 2.4 - 11.1 - 1.8 1.1 10.8TET10-18-35 2953.00 2.4 2.7 52.7 2.9 - 18.2 - 3.2 2.6 15.3TET10-18-36 2954.08 1.9 3.3 52.2 4.3 - 18.0 - 2.6 1.4 16.3TET10-18-37 2955.00 0.3 2.4 83.9 1.4 - 5.4 - - 0.6 5.9TET10-18-38 2955.65 - 1.7 91.1 1.1 - 2.8 - - 0.2 3.1TET10-18-39 2956.00 - 2.5 83.9 1.4 - 6.2 - 0.7 0.4 4.8YO14-16-1 3255.00 3.1 3.7 9.5 3.1 - 18.2 0.7 5.7 1.9 54.1YO14-16-2 3257.30 3.5 5.3 14.7 2.9 - 19.7 - 6.0 3.1 44.8YO14-16-3 3260.00 3.9 3.3 16.1 1.7 1.0 15.2 - 8.1 2.4 48.4YO14-16-4 3261.70 3.6 6.2 12.4 2.3 - 23.4 - 8.8 2.2 41.1YO14-16-5 3265.30 3.5 5.6 14.8 2.6 1.5 22.3 - 10.2 2.3 37.3YO14-16-6 3266.30 3.9 8.5 17.0 2.5 - 22.7 - 15.2 2.3 27.9YO14-16-7 3269.20 3.3 3.8 19.6 0.9 - 9.6 - 14.2 2.1 46.5YO14-16-8 3271.50 2.3 5.8 13.8 0.8 - 12.9 - 20.9 2.3 41.2YO14-16-9 3273.70 2.2 4.7 18.3 1.8 - 12.9 - 20.5 3.3 36.4YO14-16-10 3276.20 2.5 3.9 18.9 1.1 1.3 12.9 - 19.3 2.8 37.3YO14-16-11 3279.80 1.9 3.1 15.6 0.8 0.8 15.0 - 21.4 2.7 38.6YO14-16-12 3281.20 1.8 4.7 13.4 1.6 - 16.6 - 22.6 3.0 36.3YO14-16-13 3283.00 3.3 3.1 6.5 1.0 - 22.6 - 30.2 2.4 31.1YO14-16-14 3283.10 1.4 3.4 11.2 1.5 - 11.5 - 26.4 3.4 41.3YO14-16-15 3286.50 1.6 26.7 13.5 - - 12.4 0.8 16.0 2.0 27.0YO14-16-16 3288.50 1.5 1.7 7.2 1.7 - 12.0 - 27.6 3.2 45.1YO14-16-17 3290.10 2.0 1.4 28.2 2.2 - 16.6 - 22.7 2.7 24.2YO14-16-18 3291.00 2.4 3.6 8.0 2.3 - 20.5 - 24.4 3.0 35.7239aPPenDIX	 B:	 PETRoPhYSICAl RESUlTS240Sample ID Depth Unit Hg MIP He MIP He Tmax S1 S2 TOC HI OI(m) RhoB (g/cc) RhoS (g/cc) Porosity (%) (oC) (mg/g) (mg/g) (%)ATH1-24-2 3681.18 Duv 2.65 2.61 2.76 1.5 3.7 - 0.76 0.25 1.46 17 28ATH1-24-3 3683.42 Duv 2.61 2.55 2.70 1.1 3.3 - 0.83 0.24 1.73 14 24ATH1-24-5 3685.07 Duv 2.63 2.60 2.73 1.2 3.7 - 1.32 0.31 1.78 17 18ATH1-24-6 3686.55 Duv 2.58 2.52 2.68 1.4 3.6 452 1.80 0.63 2.46 26 18ATH1-24-8 3689.66 Duv 2.57 2.53 2.64 1.2 2.8 459 1.25 0.75 3.22 23 12ATH1-24-10 3691.53 Duv 2.68 2.66 2.74 0.7 2.1 431 0.30 0.16 0.65 24 42ATH1-24-12 3692.86 Duv 2.57 2.54 2.65 1.2 3.1 494 0.94 0.70 3.31 21 13ATH1-24-13 3694.40 Duv 2.68 2.64 2.72 0.4 1.4 417 0.10 0.08 0.50 16 80ATH1-24-14 3695.48 Duv 2.60 2.57 2.69 1.3 3.4 439 0.67 0.52 2.25 23 27ATH1-24-16 3696.53 Duv 2.57 2.52 2.67 1.1 3.5 445 0.71 0.71 2.67 27 9ATH1-24-17 3697.30 Duv 2.65 2.62 2.71 0.6 2.3 - 0.34 0.39 1.39 28 17ATH1-24-19 3699.50 Duv 2.53 2.48 2.61 1.1 3.0 492 0.99 1.26 5.76 22 3ATH1-24-21 3700.00 Duv 2.53 2.51 2.61 1.1 2.8 498 0.74 1.20 5.42 22 3ATH1-24-22 3700.62 Duv 2.54 2.49 2.61 1.1 2.5 495 0.54 1.17 5.30 22 2ATH1-24-24 3702.00 Duv 2.53 2.51 2.61 1.0 2.7 501 0.53 1.12 5.64 20 5ATH1-24-25 3703.23 Duv 2.56 2.51 2.62 0.9 2.3 474 0.53 1.07 4.27 25 1ATH1-24-26 3703.63 Duv 2.67 2.65 2.71 0.5 1.5 433 0.15 0.23 0.36 63 20ATH1-24-27 3704.62 Duv 2.57 - 2.61 1.2 1.7 500 0.38 1.03 5.40 19 1ATH1-24-29 3706.40 Duv 2.56 2.53 2.62 1.1 2.3 506 0.40 1.01 5.35 19 1ATH1-24-30 3707.15 Duv 2.55 2.50 2.60 1.1 2.0 504 0.41 0.99 5.22 19 0ATH1-24-31 3708.00 Duv 2.58 2.55 2.62 0.9 1.7 507 0.28 0.98 5.55 18 1ATH1-24-33 3709.98 Duv 2.59 2.57 2.64 0.7 2.0 511 0.36 0.93 4.90 19 0ATH1-24-35 3711.00 Duv 2.56 2.53 2.61 0.9 2.0 512 0.47 1.12 6.29 18 1ATH1-24-36 3712.40 Duv 2.53 2.49 2.61 0.8 3.2 515 0.60 1.07 5.27 20 0ATH1-24-37 3712.70 Duv 2.56 2.55 2.63 0.9 2.8 511 0.37 0.78 3.75 21 0ATH1-24-39 3713.80 Duv 2.55 2.54 2.63 0.9 3.2 510 0.33 0.67 3.61 19 1ATH1-24-40 3715.32 Duv 2.67 2.65 2.72 0.5 1.8 435 0.19 0.27 0.74 36 4ATH1-24-41 3715.73 Duv 2.68 2.63 2.71 0.6 1.1 - 0.13 0.22 0.54 41 2ATH1-24-43 3715.93 Duv 2.69 2.66 2.71 0.6 0.6 432 0.20 0.24 0.58 42 5ATH1-24-46 3719.53 Duv 2.68 2.63 2.73 0.6 1.8 435 0.15 0.19 0.32 59 20ATH1-24-47 3720.70 Duv 2.68 2.66 2.72 0.6 1.3 432 0.14 0.18 0.27 67 12ATH1-24-48 3721.30 Duv 2.68 2.63 2.71 0.7 1.2 432 0.12 0.17 0.22 80 31ATH1-24-49 3722.75 Duv 2.61 2.60 2.67 0.6 2.3 533 0.26 0.51 2.31 22 3ATH1-24-50 3724.26 Duv 2.57 2.54 2.64 1.0 2.6 567 0.33 0.84 4.17 20 3ATH4-2-1 3606.24 Ire 2.67 - 2.80 - 4.8 - 0.73 0.48 1.30 37 29ATH4-2-2 3606.26 Ire 2.66 - 2.80 - 5.0 - 0.81 0.40 1.25 32 25ATH4-2-3 3608.24 Ire 2.64 - 2.80 - 5.7 431 0.79 0.44 1.32 33 30ATH4-2-4 3608.94 Ire 2.62 - 2.80 - 6.4 444 0.54 0.43 1.39 31 14ATH4-2-5 3608.97 Ire 2.62 - 2.79 - 6.1 440 0.50 0.38 1.33 29 30ATH4-2-6 3610.33 Ire 2.58 - 2.78 - 7.3 436 0.99 0.65 1.80 36 23ATH4-2-7 3610.56 Ire 2.69 - 2.82 - 4.8 - 0.53 0.27 0.90 30 42ATH4-2-8 3612.04 Ire 2.63 - 2.78 - 5.5 455 0.85 0.45 1.36 33 26Table B.1 Petrophysical results from detailed analysis wells241Sample ID Depth Unit Hg MIP He MIP He Tmax S1 S2 TOC HI OI(m) RhoB (g/cc) RhoS (g/cc) Porosity (%) (oC) (mg/g) (mg/g) (%)ATH4-2-9 3613.37 Duv 2.63 - 2.75 - 4.2 446 1.50 1.04 2.32 45 15ATH4-2-10 3613.85 Duv 2.53 - 2.70 - 6.3 460 2.60 1.56 3.09 51 13ATH4-2-11 3614.95 Duv 2.50 - - - - 471 3.42 1.82 3.00 61 8ATH4-2-12 3615.25 Duv 2.52 - 2.70 - 6.4 469 1.58 1.13 2.89 39 11ATH4-2-13 3615.28 Duv 2.52 - 2.68 - 5.8 466 2.09 1.58 3.14 50 10ATH4-2-14 3616.84 Duv 2.53 - 2.67 - 5.0 474 4.33 2.25 3.94 57 6ATH4-2-15 3618.34 Duv 2.47 - 2.64 - 6.4 475 3.14 1.99 3.74 53 6ATH4-2-16 3618.45 Duv 2.45 - 2.62 - 6.5 485 4.81 1.62 3.23 50 6ATH4-2-17 3619.02 Duv 2.48 - 2.64 - 5.9 480 4.25 1.76 3.55 50 4ATH4-2-18 3620.91 Duv 2.53 - 2.67 - 5.2 470 2.15 1.21 3.10 39 8ATH4-2-19 3621.17 Duv 2.51 - 2.70 - 6.9 476 2.69 1.70 3.51 48 5ATH4-2-20 3622.80 Duv 2.53 - 2.66 - 5.1 464 2.29 1.11 2.68 41 9ATH4-2-21 3624.33 Duv 2.52 - 2.67 - 5.7 473 2.20 1.19 2.76 43 8ATH4-2-22 3624.53 Duv 2.49 - 2.66 - 6.6 473 2.02 1.22 2.72 45 11ATH4-2-23 3626.10 Duv 2.53 - 2.65 - 4.7 468 2.12 1.05 2.62 40 12ATH4-2-24 3627.30 Duv 2.52 - 2.66 - 5.2 467 1.76 0.95 2.34 41 12ATH4-2-25 3627.42 Duv 2.50 - 2.68 - 6.8 465 1.62 1.24 2.74 45 9ATH4-2-26 3628.75 Duv 2.50 - 2.64 - 5.2 476 1.90 1.62 2.84 57 11ATH4-2-27 3630.29 Duv 2.42 - 2.61 - 7.2 484 1.87 2.94 4.35 68 7ATH4-2-28 3630.41 Duv 2.43 - 2.62 - 7.4 480 3.20 2.43 4.39 55 8ATH4-2-29 3630.59 Duv 2.53 - 2.63 - 3.9 482 1.00 1.74 3.67 47 10ATH4-2-30 3631.99 Duv 2.47 - 2.58 - 4.2 480 4.30 3.08 5.65 54 4ATH4-2-31 3632.80 Duv 2.48 - 2.59 - 4.6 481 3.48 2.82 5.02 56 4ATH4-2-32 3632.94 Duv 2.45 - 2.61 - 6.2 481 2.45 2.08 4.66 45 5ATH4-2-33 3634.23 Duv 2.48 - 2.63 - 5.7 482 4.71 2.06 4.11 50 5ATH4-2-34 3634.79 Duv 2.42 - 2.59 - 6.6 480 5.12 2.90 4.88 60 4ATH4-2-35 3635.56 Duv 2.45 - 2.62 - 6.4 481 4.00 1.93 3.78 51 5ATH4-2-36 3636.10 Duv 2.48 - 2.62 - 5.5 480 5.85 1.98 4.09 48 5ATH4-2-37 3636.62 Duv 2.44 - 2.61 - 6.3 476 5.72 2.33 4.14 56 5ATH4-2-38 3638.41 Duv 2.48 - 2.61 - 4.8 481 4.37 2.19 4.60 48 5ATH4-2-39 3639.62 Duv 2.49 - 2.61 - 4.5 483 2.31 1.93 4.76 41 6ATH4-2-40 3639.70 Duv 2.44 - 2.57 - 5.1 479 4.49 2.96 5.34 55 4ATH4-2-41 3641.72 Duv 2.43 - 2.59 - 6.1 478 3.51 2.59 4.92 53 4ATH4-2-42 3641.86 Duv 2.45 - 2.61 - 6.1 483 3.03 1.80 4.07 44 5ATH4-2-43 3642.22 Duv 2.45 - 2.57 - 4.9 476 4.56 2.02 4.13 49 7ATH4-2-44 3643.95 Duv 2.53 - 2.65 - 4.8 457 2.91 0.64 1.85 35 11ATH4-2-45 3644.14 Duv 2.60 - 2.74 - 5.3 465 3.42 1.25 2.81 44 7ATH4-2-46 3644.75 Duv 2.60 - 2.69 - 3.3 463 3.34 1.09 3.17 34 5ATH4-2-47 3645.43 Duv 2.40 - 2.56 - 6.2 477 4.90 2.17 4.06 54 3ATH4-2-48 3646.61 Duv 2.43 - 2.58 - 5.7 479 3.47 1.82 3.79 48 5ATH4-2-49 3647.75 Duv 2.50 - 2.64 - 5.3 479 3.76 2.27 4.70 48 4ATH4-2-50 3649.14 Duv 2.47 - 2.65 - 6.9 451 1.93 0.66 2.33 28 24ATH13-18-1 2754.95 Ire 2.69 - 2.84 - 5.1 436 0.24 0.38 0.42 90 133242Sample ID Depth Unit Hg MIP He MIP He Tmax S1 S2 TOC HI OI(m) RhoB (g/cc) RhoS (g/cc) Porosity (%) (oC) (mg/g) (mg/g) (%)ATH13-18-2 2756.10 Ire 2.70 - 2.82 - 4.2 433 0.21 0.30 0.38 78 153ATH13-18-3 2759.27 Ire 2.69 - 2.79 - 3.5 441 0.16 0.29 0.32 91 139ATH13-18-4 2759.57 Ire 2.68 - 2.79 - 3.8 422 0.40 0.44 0.47 93 134ATH13-18-5 2761.00 Ire 2.68 - 2.80 - 4.3 441 0.37 0.52 0.59 88 61ATH13-18-6 2762.65 Ire 2.66 - 2.79 - 4.5 422 0.90 0.49 0.65 75 56ATH13-18-7 2763.30 Duv 2.56 - 2.67 - 4.3 444 3.96 3.23 2.76 116 14ATH13-18-8 2764.65 Duv 2.52 - 2.67 - 5.5 449 3.37 3.62 2.69 134 10ATH13-18-9 2766.65 Duv 2.54 - 2.66 - 4.6 439 3.37 3.01 2.42 123 20ATH13-18-10 2768.00 Duv 2.55 - 2.67 - 4.3 443 4.39 3.72 2.86 129 13ATH13-18-11 2769.00 Duv 2.55 - 2.68 - 4.7 448 5.07 3.70 2.82 130 10ATH13-18-12 2769.50 Duv 2.54 - 2.64 - 3.9 449 3.75 4.38 3.23 135 9ATH13-18-13 2771.00 Duv 2.57 - 2.66 - 3.6 446 3.81 3.89 3.13 124 12ATH13-18-14 2771.46 Duv 2.55 - 2.64 - 3.4 450 3.09 4.47 3.39 131 11ATH13-18-15 2771.50 Duv 2.54 - 2.64 - 3.8 445 3.25 4.23 3.16 133 15ATH13-18-16 2773.55 Duv 2.51 - 2.61 - 4.2 449 4.61 6.13 4.34 141 9ATH13-18-17 2774.85 Duv 2.53 - 2.65 - 4.7 442 4.25 3.84 2.81 136 13ATH13-18-18 2774.97 Duv 2.53 - 2.62 - 3.5 445 3.80 4.42 3.58 123 15ATH13-18-19 2776.30 Duv 2.51 - 2.64 - 4.8 446 4.87 4.39 3.12 140 10ATH13-18-20 2776.78 Duv 2.56 - 2.64 - 3.2 440 4.02 4.21 3.23 130 24ATH13-18-21 2777.04 Duv 2.51 - 2.63 - 4.8 444 3.94 4.43 3.29 134 14ATH13-18-22 2777.20 Duv 2.50 - 2.63 - 4.9 454 4.92 4.41 2.87 153 15ATH13-18-23 2779.40 Duv 2.51 - 2.65 - 5.3 448 4.99 5.90 4.07 145 9ATH13-18-24 2780.36 Duv 2.48 - 2.64 - 5.8 441 5.54 4.65 3.12 149 15ATH13-18-25 2782.30 Duv 2.53 - - - - 448 3.22 4.02 3.25 123 18ATH13-18-26 2782.48 Duv 2.51 - 2.62 - 4.4 445 3.91 3.95 2.90 136 13ATH13-18-27 2783.15 Duv 2.50 - 2.61 - 4.3 446 3.94 4.44 3.35 132 9ATH13-18-28 2783.55 Duv 2.52 - 2.66 - 5.2 445 3.33 3.49 2.97 117 13ATH13-18-29 2784.50 Duv 2.51 - 2.63 - 4.2 447 4.62 4.62 3.50 131 11ATH13-18-30 2785.90 Duv 2.67 - 2.71 - 1.4 442 0.96 0.94 0.97 96 45ATH13-18-31 2786.11 Duv 2.67 - 2.73 - 2.2 439 1.07 1.18 1.16 101 51ATH13-18-32 2786.20 Duv 2.65 - 2.73 - 2.8 444 1.02 1.17 1.08 107 34ATH13-18-33 2787.67 Duv 2.64 - 2.74 - 3.6 442 1.37 1.88 1.41 132 27ATH13-18-34 2787.95 Duv 2.61 - 2.69 - 2.9 443 1.77 2.28 1.81 126 30ATH13-18-35 2787.98 Duv 2.57 - 2.68 - 3.9 447 2.84 3.54 2.66 133 21ATH13-18-36 2788.35 Duv 2.60 - 2.72 - 4.6 439 3.18 3.06 2.52 121 18ATH13-18-37 2790.00 Duv 2.48 - 2.61 - 5.0 451 4.25 5.95 4.18 142 8ATH13-18-38 2790.02 Duv 2.48 - 2.61 - 4.9 446 4.19 5.83 4.24 137 9ATH13-18-39 2790.30 Duv 2.51 - 2.64 - 4.8 447 4.50 5.52 4.64 118 6ATH13-18-40 2790.72 Duv 2.50 - 2.63 - 5.0 445 3.71 4.71 3.60 130 11ATH13-18-41 2791.12 Duv 2.48 - 2.62 - 5.4 445 4.04 5.13 3.77 136 10ATH13-18-42 2791.66 Duv 2.54 - 2.61 - 2.9 448 3.36 5.21 3.98 131 13ATH13-18-43 2791.80 Mid 2.68 - 2.72 - 1.5 446 2.04 1.73 1.82 95 34ATH13-18-44 2792.30 Mid 2.67 - 2.74 - 2.6 428 0.59 0.54 0.57 94 76243Sample ID Depth Unit Hg MIP He MIP He Tmax S1 S2 TOC HI OI(m) RhoB (g/cc) RhoS (g/cc) Porosity (%) (oC) (mg/g) (mg/g) (%)ATH13-18-46 2793.37 Mid 2.69 - 2.74 - 1.8 414 0.26 0.39 0.37 105 172ATH13-18-48 2794.77 Mid 2.68 - 2.75 - 2.6 416 0.19 0.29 0.25 117 185ATH13-18-51 2796.26 Mid 2.68 - 2.74 - 2.2 427 0.42 0.43 0.48 89 103ATH13-18-52 2797.00 Mid 2.68 - 2.74 - 2.4 420 0.22 0.30 0.38 80 117ATH13-18-54 2798.32 Mid 2.69 - 2.77 - 3.1 420 0.32 0.39 0.45 87 143ECA11-8-2 3794.35 Duv 2.55 2.50 2.64 2.1 3.6 468 2.37 2.92 3.60 81 7ECA11-8-3 3795.65 Duv 2.51 2.44 2.64 2.3 5.1 468 2.43 2.69 3.39 79 7ECA11-8-5 3797.75 Duv 2.52 2.51 2.65 2.0 4.9 470 2.55 2.24 3.06 73 8ECA11-8-8 3802.40 Duv 2.52 2.49 2.64 1.5 4.6 470 1.70 2.67 3.61 74 7ECA11-8-9 3802.91 Duv 2.56 2.57 2.67 1.3 4.2 473 1.18 1.77 2.69 66 8ECA11-8-11 3804.75 Duv 2.50 2.50 2.65 2.5 5.5 467 2.01 2.14 2.92 73 8ECA11-8-12 3805.10 Duv 2.49 2.43 2.64 2.7 5.7 467 2.01 2.15 2.77 78 8ECA11-8-17 3811.25 Duv 2.48 2.33 2.61 2.6 5.1 474 2.05 3.02 4.11 74 6ECA11-8-19 3812.70 Duv 2.46 2.35 2.55 2.7 3.8 473 3.23 5.99 6.79 88 4ECA11-8-21 3813.35 Duv 2.38 2.32 2.52 2.8 5.3 473 3.28 7.39 8.32 89 3ECA11-8-22 3813.95 Duv 2.43 2.42 2.59 2.8 6.2 471 2.45 4.61 5.80 79 4ECA11-8-23 3814.00 Duv 2.57 2.43 2.67 2.6 3.7 470 2.49 4.23 5.20 81 4ECA11-8-24 3816.20 Duv 2.39 2.38 2.55 2.9 6.1 472 4.04 5.62 5.61 100 4ECA11-8-25 3816.70 Duv 2.53 2.47 2.64 2.8 4.4 469 1.90 3.13 4.16 75 5ECA11-8-26 3819.80 Duv 2.52 2.47 2.65 3.0 4.9 470 2.50 3.39 4.51 75 5ECA11-8-28 3823.40 Duv 2.43 2.37 2.60 2.7 6.5 473 2.33 3.57 4.39 81 5ECA11-8-30 3826.35 Duv 2.45 2.42 2.60 3.3 5.9 475 2.36 2.72 3.55 77 6ECA11-8-31 3828.50 Duv 2.57 2.50 2.73 2.4 5.5 461 1.35 1.62 2.60 62 9ECA11-8-32 3830.05 Duv 2.52 2.52 2.66 2.1 5.5 466 1.62 1.55 2.60 59 11ECA11-8-33 3831.15 Duv 2.55 2.49 2.67 1.7 4.4 467 1.24 1.66 2.75 60 10ECA11-8-36 3833.00 Duv 2.56 2.45 2.62 2.2 2.2 470 2.07 3.58 4.30 83 6ECA11-8-37 3833.60 Maj 2.55 2.55 2.73 2.4 6.4 462 0.81 0.98 1.91 51 15ECA11-8-38 3835.75 Maj 2.54 2.50 2.75 2.7 7.4 457 0.73 0.83 1.70 49 15ECA11-8-39 3836.85 Maj 2.57 2.50 2.76 2.7 7.0 460 0.51 0.64 1.52 42 19HSK5-11-1 3049.20 Ire 2.69 - 2.81 - 4.3 459 0.44 0.48 0.67 70 49HSK5-11-2 3051.95 Duv 2.50 - 2.63 - 5.0 465 4.09 2.81 3.69 76 7HSK5-11-3 3052.95 Duv 2.56 - 2.67 - 4.2 467 4.15 2.08 3.09 67 7HSK5-11-4 3054.60 Duv 2.51 - 2.65 - 5.2 467 4.44 2.03 2.89 70 8HSK5-11-5 3056.13 Duv 2.52 - 2.63 - 4.2 463 6.22 2.31 3.39 68 9HSK5-11-6 3057.69 Duv 2.58 - 2.66 - 3.0 469 2.82 2.59 3.93 65 9HSK5-11-7 3059.35 Duv 2.53 - 2.67 - 5.4 466 3.09 2.76 4.42 62 6HSK5-11-8 3060.30 Duv 2.54 - 2.67 - 4.9 460 2.50 1.13 1.78 63 16HSK5-11-9 3062.00 Duv 2.52 - 2.64 - 4.6 469 3.39 3.75 5.23 71 7HSK5-11-10 3064.15 Duv 2.52 - 2.66 - 5.2 461 6.21 2.71 4.12 65 8HSK5-11-11 3065.20 Duv 2.47 - 2.63 - 6.2 466 5.11 3.89 5.27 73 6HSK5-11-12 3066.60 Duv 2.48 - 2.60 - 4.6 467 3.48 3.92 5.24 74 5HSK5-11-13 3068.50 Duv 2.61 - 2.69 - 2.9 466 1.51 1.32 2.06 64 11HSK5-11-14 3069.60 Duv 2.60 - 2.70 - 3.6 459 1.25 1.03 2.04 50 14244Sample ID Depth Unit Hg MIP He MIP He Tmax S1 S2 TOC HI OI(m) RhoB (g/cc) RhoS (g/cc) Porosity (%) (oC) (mg/g) (mg/g) (%)HSK5-11-15 3071.25 Duv 2.45 - 2.59 - 5.5 468 4.13 3.71 5.13 72 6HSK5-11-16 3072.96 Duv 2.45 - 2.60 - 5.7 469 4.88 4.34 5.47 79 4HSK5-11-17 3074.60 Duv 2.44 - 2.59 - 5.8 469 3.93 4.29 5.65 75 4HSK5-11-18 3076.25 Duv 2.47 - 2.61 - 5.5 470 4.14 4.15 5.43 76 5HSK5-11-19 3077.65 Duv 2.47 - 2.61 - 5.4 468 4.57 3.69 5.32 69 5HSK5-11-20 3078.95 Duv 2.51 - 2.64 - 5.0 469 2.92 2.72 3.83 71 7HSK5-11-21 3080.80 Duv 2.57 - 2.70 - 4.6 472 2.56 1.96 3.14 62 7HSK5-11-22 3082.60 Duv 2.50 - 2.62 - 4.5 467 2.36 2.62 3.61 72 7HSK5-11-23 3085.17 Duv 2.47 - 2.62 - 5.5 468 3.59 2.74 3.67 74 9HSK5-11-24 3089.20 Duv 2.59 - 2.68 - 3.3 457 1.82 1.58 2.48 63 10HSK5-11-25 3093.60 Duv 2.46 - 2.60 - 5.4 468 3.56 3.67 5.07 72 6HSK5-11-26 3095.90 Duv 2.47 - 2.57 - 4.1 470 3.06 3.86 5.63 68 4HSK5-11-27 3096.90 Duv 2.61 - 2.69 - 2.9 474 1.51 1.63 2.69 60 13HSK5-11-28 3099.85 Duv 2.55 - 2.62 - 2.8 469 2.19 3.18 4.79 66 6HSK5-11-29 3100.80 Maj 2.65 - 2.70 - 1.9 443 0.82 0.82 1.05 78 21HSK8-25-1 3102.70 Ire 2.70 2.69 2.78 0.9 3.1 451 0.10 0.16 0.61 26 68HSK8-25-2 3104.19 Ire 2.69 2.64 2.78 0.8 3.0 - 0.12 0.14 0.55 25 74HSK8-25-3 3106.48 Ire 2.69 2.69 2.78 0.7 3.2 470 0.12 0.13 0.57 23 55HSK8-25-4 3108.17 Ire 2.58 2.51 2.70 2.2 4.5 453 1.06 1.26 2.25 56 11HSK8-25-5 3109.62 Duv 2.51 2.50 2.63 2.2 4.6 466 2.16 2.76 3.59 77 7HSK8-25-6 3111.38 Duv 2.53 2.51 2.64 1.6 4.1 468 2.14 2.68 3.52 76 6HSK8-25-7 3113.47 Duv 2.59 2.53 2.69 2.2 3.8 468 1.17 1.08 1.99 54 10HSK8-25-8 3114.69 Duv 2.59 2.53 2.69 1.7 3.7 468 1.53 0.99 1.89 53 10HSK8-25-9 3116.10 Duv 2.58 2.58 2.70 1.0 4.2 470 1.15 1.53 2.59 59 8HSK8-25-10 3117.80 Duv 2.53 2.49 2.65 1.6 4.6 465 2.09 2.60 3.29 79 6HSK8-25-11 3119.20 Duv 2.55 2.51 2.65 1.6 3.4 468 1.39 1.63 2.66 61 8HSK8-25-12 3121.84 Duv 2.45 2.42 2.58 3.4 5.1 472 3.35 3.04 3.80 80 5HSK8-25-13 3122.96 Duv 2.47 2.41 2.59 2.8 4.7 469 4.38 3.18 3.87 82 6HSK8-25-14 3124.20 Duv 2.56 2.51 2.66 2.1 3.7 469 1.58 2.31 3.22 72 7HSK8-25-15 3126.34 Duv 2.50 2.43 2.60 2.9 3.9 466 2.43 3.39 4.06 84 6HSK8-25-16 3130.30 Duv 2.44 2.42 2.58 2.5 5.5 470 2.68 4.23 4.55 93 4HSK8-25-17 3132.70 Duv 2.47 2.42 2.58 2.2 4.3 468 3.02 4.01 4.44 90 5HSK8-25-18 3134.38 Duv 2.46 2.44 2.57 2.0 4.3 471 2.88 4.40 4.83 91 5HSK8-25-19 3136.25 Duv 2.48 2.43 2.60 1.7 4.4 470 2.55 4.19 4.42 95 6PWT10-17-1 2982.95 Ire 2.71 - 2.79 - 2.9 - 0.05 0.37 0.58 63 69PWT10-17-2 2989.80 Ire 2.68 - 2.78 - 3.5 - 0.11 0.42 0.62 66 51PWT10-17-3 2993.90 Duv 2.45 - 2.67 - 8.3 465 0.96 1.23 5.17 23 7PWT10-17-4 2996.74 Duv 2.55 - 2.69 - 5.2 474 0.57 0.94 3.04 30 13PWT10-17-5 3001.73 Duv 2.56 - 2.69 - 5.1 474 0.52 0.84 2.16 38 11PWT10-17-6 3004.77 Duv 2.62 - 2.72 - 3.7 476 0.28 0.87 2.37 36 16PWT10-17-7 3006.71 Duv 2.65 - 2.73 - 3.0 - 0.13 0.34 0.65 51 63PWT10-17-8 3007.42 Duv 2.49 - 2.61 - 4.8 473 0.80 2.22 7.12 31 6PWT10-17-9 3008.84 Duv 2.51 - 2.64 - 4.7 465 12.22 5.49 6.95 79 4245Sample ID Depth Unit Hg MIP He MIP He Tmax S1 S2 TOC HI OI(m) RhoB (g/cc) RhoS (g/cc) Porosity (%) (oC) (mg/g) (mg/g) (%)PWT10-17-10 3009.24 Duv 2.48 - 2.62 - 5.2 463 0.70 1.69 6.10 27 7PWT10-17-11 3011.26 Duv 2.66 - 2.73 - 2.7 - 0.10 0.83 0.96 86 71PWT10-17-12 3013.21 Duv 2.45 - 2.65 - 7.5 452 0.82 1.01 5.59 18 5PWT10-17-13 3017.50 Duv 2.44 - 2.58 - 5.3 475 0.85 1.79 7.26 24 6PWT10-17-14 3019.78 Duv 2.55 - 2.67 - 4.4 450 0.59 0.74 4.67 15 10PWT10-17-15 3020.73 Duv 2.50 - 2.67 - 6.3 472 0.60 0.89 4.52 19 8PWT10-17-16 3026.20 Maj 2.64 - 2.79 - 5.3 - 0.08 0.32 0.82 38 63PWT10-17-17 3033.00 Maj 2.67 - 2.79 - 4.3 - 0.09 0.88 0.82 107 73PWT10-17-18 3036.37 Maj 2.59 - 2.76 - 5.9 - 0.22 0.47 1.98 23 28PWT10-17-19 3039.77 Maj 2.67 - 2.78 - 4.2 - 0.10 0.85 0.98 86 53PWT10-17-20 3046.85 Maj 2.67 - 2.78 - 4.0 - 0.14 0.85 1.32 64 50PWT10-17-21 3049.20 Maj 2.50 - 2.53 - 1.3 469 0.53 1.84 9.66 19 6TET10-18-1 2929.30 Duv 2.60 2.57 2.68 1.2 2.9 464 1.19 1.16 2.44 48 15TET10-18-2 2930.00 Duv 2.59 2.52 2.68 1.5 3.3 462 1.59 1.89 3.16 60 14TET10-18-3 2930.75 Duv 2.55 2.53 2.68 1.8 4.9 464 1.63 1.74 3.25 54 11TET10-18-4 2930.85 Duv 2.54 2.51 2.66 1.4 4.6 462 1.73 2.39 3.73 64 10TET10-18-5 2931.90 Duv 2.53 2.52 2.65 1.5 4.3 466 1.60 2.35 3.72 63 12TET10-18-6 2932.20 Duv 2.56 2.50 2.66 1.5 4.0 465 1.60 2.19 3.66 60 12TET10-18-7 2932.88 Duv 2.59 2.56 2.69 1.2 3.5 463 1.23 1.66 2.72 61 20TET10-18-8 2933.80 Duv 2.53 2.48 2.64 1.9 4.3 459 1.90 2.05 3.49 59 14TET10-18-9 2934.90 Duv 2.53 2.50 2.65 2.0 4.6 471 1.82 1.50 3.15 48 16TET10-18-10 2935.34 Duv 2.53 2.50 2.66 1.5 5.1 468 1.63 1.46 2.84 51 20TET10-18-11 2936.30 Duv 2.52 2.54 2.65 1.7 5.0 468 1.60 1.49 2.87 52 20TET10-18-12 2937.00 Duv 2.51 2.47 2.63 2.0 4.3 461 1.76 1.91 3.21 59 20TET10-18-13 2938.00 Duv 2.54 2.51 2.65 1.5 4.1 464 1.54 1.51 2.54 59 21TET10-18-14 2938.30 Duv 2.56 2.49 2.66 1.7 4.0 463 1.51 1.55 2.66 58 24TET10-18-15 2939.00 Duv 2.56 2.53 2.67 1.4 4.1 468 1.46 1.56 2.74 57 21TET10-18-16 2939.80 Duv 2.47 2.43 2.66 3.9 7.3 459 2.63 1.67 2.79 60 18TET10-18-17 2940.00 Duv 2.47 2.44 2.63 2.3 6.2 461 2.38 2.40 3.11 77 17TET10-18-18 2941.00 Duv 2.49 2.48 2.65 2.2 5.8 460 1.96 2.09 3.64 57 14TET10-18-19 2941.40 Duv 2.50 2.47 2.62 2.2 4.8 463 2.05 2.16 3.87 56 16TET10-18-20 2942.00 Duv 2.50 2.46 2.62 1.8 4.6 464 2.12 2.69 3.93 68 15TET10-18-21 2942.40 Duv 2.51 2.48 2.63 1.9 4.6 466 2.08 2.58 4.03 64 6TET10-18-22 2942.67 Duv 2.50 2.46 2.64 2.0 5.4 463 2.14 2.29 3.50 65 6TET10-18-23 2943.55 Duv 2.47 2.45 2.63 2.2 6.0 465 2.42 2.62 3.50 75 5TET10-18-24 2944.00 Duv 2.50 2.46 2.63 2.1 5.0 463 2.08 2.18 3.24 67 3TET10-18-25 2944.50 Duv 2.50 2.48 2.63 1.8 4.8 463 2.11 2.76 3.41 81 3TET10-18-26 2945.35 Duv 2.49 2.44 2.62 2.2 5.2 461 2.34 2.73 3.70 74 4TET10-18-27 2945.90 Duv 2.49 2.47 2.63 2.2 5.1 463 2.18 2.93 3.94 74 2TET10-18-28 2947.10 Duv 2.53 2.46 2.64 1.8 4.2 465 2.19 2.11 3.38 62 3TET10-18-29 2948.50 Duv 2.66 2.63 2.71 0.7 1.9 448 0.71 0.41 1.15 36 27TET10-18-30 2949.80 Duv 2.54 2.48 2.65 2.1 4.4 464 1.78 2.23 3.47 64 4TET10-18-31 2950.50 Duv 2.60 2.57 2.68 1.1 3.0 463 1.19 1.60 2.60 62 9246Sample ID Depth Unit Hg MIP He MIP He Tmax S1 S2 TOC HI OI(m) RhoB (g/cc) RhoS (g/cc) Porosity (%) (oC) (mg/g) (mg/g) (%)TET10-18-32 2951.60 Duv 2.56 2.45 2.59 2.2 1.4 465 2.23 4.10 5.44 75 3TET10-18-33 2952.25 Mid 2.65 2.63 2.73 0.9 3.0 438 0.45 0.26 1.06 24 39TET10-18-34 2952.70 Mid 2.68 2.61 2.74 0.9 2.4 429 0.34 0.20 0.84 23 47TET10-18-35 2953.00 Mid 2.66 2.64 2.76 1.0 3.6 431 0.39 0.22 1.08 20 59TET10-18-36 2954.08 Mid 2.61 2.56 2.71 1.2 3.5 458 0.93 1.14 2.07 55 16TET10-18-37 2955.00 Mid 2.68 2.66 2.74 0.6 2.0 - 0.20 0.06 0.46 13 80TET10-18-38 2955.65 Mid 2.68 2.73 2.73 0.5 1.8 - 0.11 0.02 0.22 10 122TET10-18-39 2956.00 Mid 2.68 - 2.75 - 2.6 - 0.19 0.05 0.42 13 98YO14-16-1 3255.00 Duv 2.47 - 2.63 - 6.3 469 2.64 2.21 3.56 62 13YO14-16-2 3257.30 Duv 2.53 - 2.64 - 4.4 469 3.01 1.90 3.71 51 10YO14-16-3 3260.00 Duv 2.51 - 2.63 - 4.7 471 3.51 2.33 3.70 62 8YO14-16-4 3261.70 Duv 2.54 - 2.64 - 4.0 463 3.65 1.79 3.27 54 16YO14-16-5 3265.30 Duv 2.54 - 2.64 - 3.8 464 3.35 2.03 3.39 59 17YO14-16-6 3266.30 Duv 2.56 - 2.70 - 5.2 469 3.08 1.91 3.10 61 10YO14-16-7 3269.20 Duv 2.48 - 2.58 - 4.0 470 5.28 2.91 4.80 60 9YO14-16-8 3271.50 Duv 2.49 - 2.59 - 3.8 470 5.42 2.79 4.97 56 6YO14-16-9 3273.70 Duv 2.53 - 2.62 - 3.4 473 3.05 2.65 4.70 56 8YO14-16-10 3276.20 Duv 2.50 - 2.61 - 4.4 473 3.83 2.38 4.58 52 10YO14-16-11 3279.80 Duv 2.49 - 2.61 - 4.5 473 4.11 2.45 4.38 55 7YO14-16-12 3281.20 Duv 2.50 - 2.61 - 4.2 472 4.68 7.11 4.93 144 6YO14-16-13 3283.00 Duv 2.46 - 2.59 - 4.7 475 3.43 2.60 4.97 52 5YO14-16-14 3283.10 Duv 2.45 - 2.55 - 3.8 472 4.07 3.04 6.33 47 5YO14-16-15 3286.50 Duv 2.63 - 2.69 - 2.5 462 2.72 1.37 3.17 43 17YO14-16-16 3288.50 Duv 2.49 - 2.60 - 4.5 473 3.48 1.87 4.00 46 6YO14-16-17 3290.10 Duv 2.56 - 2.63 - 2.5 473 1.59 1.86 3.87 48 9YO14-16-18 3291.00 Duv 2.50 - 2.62 - 4.7 474 3.80 2.71 4.56 59 8247aPPenDIX	 C:	 InSTRUMEnTS AnD SoFTWARE248InTRoDUCTIonThis appendix provides the design documentation, technical specifications and operating instructions for instruments and software developed for this study. Two pulse-decay permeameters (PDP), one liquid PDP, one gas expansion porosity and permeability apparatus (GEPP) and one axial load frame were constructed in order to analyze the permeability and other properties of core samples under reservoir conditions. Each instrument is of varying design and are discussed separately. Software to operate the instruments and collect and analyze associated data was designed using National Instruments™ LabVIEW™ 2013. As each instrument has varying valve and pressure transducer configurations, software to operate the instruments was designed to provide compatibility with multiple configurations. In addition, legacy instruments already in operation within the laboratory at UBC can utilize the software. Data analysis software was designed in parallel to collection software to provide an integrated data collection-analysis software suite. The organization of this appendix is as follows:1. Mantis-Perm automated high-pressure PDP technical specs and operating instructions.2. Perm-In-A-Box manual high-pressure PDP technical specs and operating instructions.3. LiquiPerm X7.9 Archangel manual high-pressure liquid-specific PDP technical specs and operating instructions.4. Mini-Pycno manual high-pressure GEPP technical specs and operating instructions.5. Axial load frame technical specs.6. QuickPerm data collection and instrument operating software specs.249Mantis-Perm	automated	high-pressure	PDPThe Mantis-Perm automated high-pressure PDP was designed to accommodate pore pressures representative of many unconventional hydrocarbon reservoirs, which is typically higher than many commercial instruments can accommodate. Valves were fully automated by a relay and solenoid valves in order to provide remote operation. The design schematic is given in Figure C.1.technical	specifications:Pore pressure: 10,000 psi (max).Confining	pressure: 10,000 psi (max).Absolute transducers: Stellar Technology Inc. Analog GT1600, 0 – 10,000 psia range, 12 – 32 VDC input, 0 – 10 VDC output, 0.1 % FSO accuracy.Differential	transducer: Stellar Technology Inc. Analog DT1900, 0 – 2,000 psia range, 0-10 VDC output, 0.1 % FSO accuracy.analog	to	digital	data	acquisition	device: National Instruments™ USB-6001 14-bit multifunction DAQ. 8 analog inputs, 13 digital I/O lines.Relay: Measurement Computing™ USB-ERB08 electromechanical relay.Solenoids: STC™ 3V110-1/8 1/8” NPT three-way directional pneumatic solenoid. Valves: High Pressure Equipment Company™ Hippo Piston Air operators, 1/8” rated to 10,000 psi.Internal	tubing: 1/8” 316 Stainless steel, rated to 15,000 psi.Confining	cell: Hoek-type, biaxial.Core diameter: 1.18”air	supply	required: 100 psi.Temperature: Ambient.Power: 110 - 240 VAC, 50/60 Hz.operating	instructions:1. On software, enter sample name, length, weight, current confining pressure, and stop dP.2. Open Expansion and Bleed #1.3. Fill to desired pore pressure by opening Inlet. Close Inlet.4. Close Bypass Valve.5. Open Bleed #2 to create differential pressure. The Bleed control will regulate gas flow.6. Close Bleed #1 and Bleed # 2 when desired differential pressure is created.7. Press “Analyze” button on software. File creation dialog will appear.8. Name file. Click OK9. Sample will run until the “Stop dP” differential is reached.10. Expansion valve will automatically close and Bypass valve will open once test is complete.250Perm-In-a-Box	manual	high-pressure	PDPThe Perm-In-A-Box manual high-pressure PDP was designed similarly to the Mantis-Perm but provides manual operation (Figure C.2). The instrument was designed using two absolute transducers instead of a differential transducer which enables differential pressures to the limit of the absolute transducers (e.g. 10,000 psi). An expansion valve was incorporated into the design to accommodate GEPP-type measurements. The expansion valve can also be used to separate the system for leak testing.technical	specifications:Pore pressure: 10,000 psi (max).Confining	pressure: 10,000 psi (max).Absolute transducers: Heise® DXD digital pressure transducers, 0 – 10,000 psia range, 0.02 % FS accuracy.Valves: High Pressure Equipment Company™ two-way Trunion style ball valves, 1/16” rated to 15,000 psi.Internal	tubing: 1/16” 316 Stainless steel, rated to 15,000 psi.Confining	cell: Hoek-type, biaxial.Core diameter: 1.18”Temperature: Ambient to 70 °C.operating	instructions:1. On software, enter Sample Name, Length, Weight, Current confining pressure, and Stop dP.2. Starting valve setup: Open bypass, open expansion valve, close bleed.3. Fill to desired pore pressure by opening tank cylinder valve and opening Gas In valve slowly. 4. Close Gas In valve once pore pressure is reached.5. Close Bypass Valve.6. Open Bleed Valve slowly to create 50 psi of differential pressure. Overshoot of ~200 is OK.7. Press “Analyze” button on software. File creation dialog will appear.9. Name file. Click OK10. Sample will run until the “Stop dP” differential is reached.251LiquiPerm	X7.9	archangel	high-pressure	liquid-specific	PDPThe LiquiPerm X7.9 Archangel high-pressure liquid-specific PDP was designed to utilize liquids as the operating fluid. A fluid reservoir is connected to a manual high-pressure generator which creates the desired pore pressure (Figure C.3). A 60 psi air supply attached to the top of the liquid reservoir is used to fill the system with fluid. The differential pressure pulse is created using a needle valve on the upstream side. The variable position of the ‘Create dP’ needle can generate positive (flowing from upstream to downstream) or negative (flowing from downstream to upstream) differential pressures.technical	specifications:Pore pressure: 10,000 psi (max).Confining	pressure: 10,000 psi (max).Absolute transducers: Stellar Technology Inc. Analog GT1600, 0 – 10,000 psia range, 12 – 32 VDC input, 0 – 10 VDC output, 0.1 % FSO accuracy.Differential	transducer: Stellar Technology Inc. Analog DT1900, 0 – 2,000 psia range, 0-10 VDC output, 0.1 % FSO accuracy.analog	to	digital	data	acquisition	device: National Instruments™ IOTech Personal DAQ 50 series 22-bit multifunction DAQ. 10 analog inputs, 8 digital I/O lines.Valves: High Pressure Equipment Company™ two-way straight needle valves, 1/8” rated to 15,000 psi.Internal	tubing: 1/8” 316 Stainless steel, rated to 15,000 psi.Confining	cell: Hoek-type, biaxial.Core diameter: 1.18”air	supply	required: 60 psi.Temperature: Ambient.Power: 110 - 240 VAC, 50/60 Hz.operating	instructionsTo load a sample and pressurize the system:Order of operations:1. Load sample2. Vacuum system3. Fill system with liquid and pressurize4. Create DP252Load a sample into the Hoek cell.To vacuum system:1. Make sure Air Pressure valve on reservoir is closed.2. Open Outlet valve on reservoir to release any pressure that may be in the reservoir.3. Open Liquid In, Reservoir, Bypass and Create dP valves fully.4. Close Outlet valve on reservoir.5. Open Vacuum valve on reservoir.6. Turn on vacuum pump.7. Let run for 15 - 30 minutes. Pressure in system will decrease to approx. -14 psi.8. Close Reservoir valve. This seals the system at -14 psi.9. Turn off vacuum pump.10. Open Outlet valve on reservoir. This increases the pressure in the reservoir to atmospheric11. Close Vacuum valve on reservoir.12. Close Outlet valve on reservoir.13. System is now ready to be filled with fluid.To fill system with fluid and pressurize:1. Pressure in system should be approx. -14 psi.2. Make sure pump is backed out (counter-clockwise) so the Vernier indicator on the pump is between 40 and 50.3. Open Air Pressure valve on reservoir. This will supply air pressure to the top of the liquid in the reservoir.4. Open Reservoir valve. The pressurized liquid will be forced into the pump and the system. Pressure should increase to the air pressure supplied.5. Close Reservoir valve. This seals the system from the reservoir.6. Close Air Pressure valve on reservoir.7. Open Outlet. This reduces the air pressure on the liquid reservoir. System pressure should stay sealed.8. The system is now ready to be pressurized to pore pressure. By turning the pump clockwise, slowly increase the pressure in the system. A quarter turn should yield about 250 psi in pressure increase.9. Once the system is at desired pore pressure, close Liquid In. This seals the system from the pump at desired pore pressure.10. Open Reservoir valve. This reduces the liquid pressure in the pump to atmospheric because the Outlet valve on the reservoir is open.11. Close Reservoir valve.To create a dP using the create dP valve:1. Close the bypass valve.2532. By using the Create dP valve, turn clockwise or counterclockwise to create the desired differential pressure.3. Start Analysis.To create a dP using the pump:1. Open Liquid In valve. This connects the system to the pump. Since the pressure in the pump is 0, the pressure in the system should decrease. This is okay.2. Repressurize the system with the pump to desired pore pressure.3. Close Bypass.4. The dP will be created using the pump. Increase the pressure on the upstream end using the pump.5. Close Liquid In.6. Start Analysis.254Mini-Pycno	manual	high-pressure	gePPThe Mini-Pycno manual high-pressure GEPP was designed to measure the total gas uptake by the sample matrix (Figure C.4). The instrument was designed using a single absolute transducer to measure pressure decline. The system was designed using 1/16” tubing to minimize system volume and therefore maximize sensitivity to the pressure decline as gas enters the pore space of the sample.technical	specifications:Pore pressure: 10,000 psi (max).Confining	pressure: 10,000 psi (max).Absolute transducers: Heise® DXD digital pressure transducers, 0 – 10,000 psia range, 0.02 % FS accuracy.Valves: High Pressure Equipment Company™ two-way straight needle valves, 1/16” rated to 15,000 psi.Internal	tubing: 1/16” 316 Stainless steel, rated to 15,000 psi.Confining	cell: Hoek-type, biaxial.Core diameter: 1.18”Temperature: Ambient to 70 °C.operating	instructions:Order of operations:1. Load sample2. Open Fill valve to fill Reference with gas3. Click “Analyze” on software4. Open Expansion valve to expand gas into the sampleLoad a sample into the Hoek cell.1. Open “MiniPycno.exe” software.2. Pressure in system should be approx. 0 psi. (With the system open to the atmosphere, hit “Zero Pressure” if it is not)3. Enter Sample ID, Length, and Confining Pressure. Diameter should be 2.9972 cm if using 1.18” plugs.4. Make sure Expansion valve and Fill valve are closed.5. Make sure the gas tank line is securely connected to the Mini-pycno.6. Open gas tank valve. This puts pressure into the line connected to the Mini-pycno.7. Open the Fill valve slightly to increase pressure in Reference to desired pressure.8. Close Fill valve. Close gas tank valve.2559. Wait for the system to come to equilibrium (15 - 30 minutes).10. Click “Analyze”. The software will record the fill pressure and sample details.11. Once it has done so, you will be prompted to “Open Expansion” valve.12. Open Expansion valve fully. System was calibrated with the Expansion valve fully open.13. Leave Expansion valve open while test runs. The test needs to be manually stopped (once the porosity vs time graph comes to equilibrium)256axial	load	frameA uniaxial load frame was designed and constructed to more precisely control the load conditions and balance the confining stress on the sample. A schematic is given in Figure C.5. The load frame is equipped with a 10,000 lb hydraulic cylinder which provides the axial load. A 10,000 lb load cell measures the load placed on the platens and rock sample. Specifications are given in the diagram.technical	specifications:Axial load: 10,000 psi (recommended).load cell: Stellar Technology Inc. Analog PNC720 Compression Pancake Load Cell, 0 – 10,000 lbs. 0-10 VDC output, 0.1 % FSO accuracy.hydraulic cylinder: Eagle Pro® ES-50 hydraulic cylinder, 0.63 in stroke, 5 metric ton capacity at 10,000 psi.End plates: 1” steel.Cross bars: 3/4” - 10 B7 Alloy steel stud. Rated to 41,450 lbs max each.257QuickPerm	operating	SoftwareThe QuickPerm data collection software (Figure C.6; Figure C.7; Figure C.8) was designed to accommodate multiple configurations of pore pressure transducers, confining pressure transducers and data acquisition devices. Transducers can have analog or digital output. The QuickPerm software was also designed to integrate a load cell and linear variable differential transformers (LVDT). Various pore fluid properties were included to allow the use of multiple test fluids. The settings for currently operational permeameters at UBC are precoded. The configurations are outlined below.Data	acquisition	devices	supported:• National InstrumentsTM USB DAQ, NI USB-6XXX series• Measurement ComputingTM USB DAQ, USB-200, USB-230, USB-1208FS-PLUS Series, USB-1608FS-PLUS• Measurement ComputingTM IOTech Personal Daq/50 seriesPore	pressure	transducer	configurations:• Analog transducer upstream and analog differential pressure transducer• Analog transducer downstream and analog differential pressure transducer• Analog transducer upstream and analog transducer downstream• Digital transducer upstream and analog differential pressure transducer• Digital transducer downstream and analog differential pressure transducer• Digital transducer upstream and digital transducer downstream• Digital transducer upstream, digital transducer downstream and analog differential transducerConfining	pressure	transducers:• Analog or digital (Heise)load cell:• AnaloglVDT:• Analog258QuickPerm	Dashboard	pageThe Dashboard (Figure C.6) is the front page of the QuickPerm software. Pressure transducer data is displayed in graphical and text form, sample details are recorded and various test parameters are set on the Dashboard. The Dashboard includes a permeability graph where real-time permeability is displayed during an analysis. The valve control is located on the Dashboard for pneumatic-operated valve configurations (i.e. Mantis-Perm). Essential parameters for an analysis are denoted by green text fill (Figure C.6), which include sample name, sample length, sample diameter and stop dP.QuickPerm	Configuration	PageThe Configuration page (Figure C.7) contains all setup information such as the permeameter currently being operated, data acquisition device type, pore pressure transducer configurations, confining pressure transducer type, transducer channels and gain and load cell or LVDT options. Necessary outputs for troubleshooting such as raw voltage (input voltage to the DAQ for analog transducers) and other parameters are also displayed. System volumes can be entered manually.Pressure	test	PageFrequent system pressure tests are required to check for pore fluid leaks, confining pressure leaks, or to monitor the equilibrium state of the sample. A pressure test page was included in the QuickPerm software to accomplish this task (Figure C.8). The pressure test is started by clicking the “Pressure test” button. A file creation dialog box will appear to recored all pressure test details as a text file for later inspection. The “Minutes to wait” text box specifies the duration the software waits before taking a sample. The test is finished when the “Stop” button is pressed. All samples are recorded in the text file and displayed in the graphs on the right side of the pressure test page for visual inspection.259OUTINHoek ConfiningCellPlatensCoreMantis-PermNormally-closed two-way needle valveNormally-open two-way needle valveManual two-way ball valveDPDDP Differential pressure transducerAbsolute Absolute pressure transducerTeeCrossTubingInletBypassExpansionBleed #1Bleed controlBleed #2AbsoluteFigure C.1 Mantis-Perm schematic.260AbsoluteOUTINHoek ConfiningCellPlatensCorePerm-In-A-BoxManual two-way ball valveAbsolute Absolute pressure transducerTeeCrossTubingInletBypassExpansionBleedAbsoluteFigure C.2 Perm-In-A-Box schematic.261OUTINHoek ConfiningCellPlatensCoreLiquiPerm X7.9 ArchangelManual two-way needle valveDP Differential pressure transducerAbsolute Absolute pressure transducerTeeTubingManual high-pressure generatorLiquid reservoirReservoirBypassInletBleedCreate dPDPDAbsoluteFigure C.3 LiquiPerm X7.9 Archangel schematic.262INHoek ConfiningCellPlatensCoreMini-PycnoManual two-way needle valveAbsolute Absolute pressure transducerTeeTubingInlet ExpansionAbsoluteFigure C.4 Mini-Pycno schematic.2633/4” - 10 B7 Alloy Steel stud, rated 41,750 lbs eachHoekTop view (not to scale)PlatenPlaten10KLoad cell10KHydr. cyl.Alu stand10.00"Vent2.50"8.00"12.00"2.29"1.61"4-Bolt Free-standing Load FrameHoek1” Steel plate3/4” - 10 2H Heavy structural nut3/4” - 10 Jam nutSide view (0.5” = 1.0”)10.00"4.00"End viewFigure C.5 Load frame technical drawing.264Figure C.6 The dashboard for the QuickPerm permeability data collection software.265Figure C.7 Configuration page for the QuickPerm permeability data collection software.266Figure C.8 Pressure test page for the QuickPerm permeability data collection software.

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0166722/manifest

Comment

Related Items