International Conference on Gas Hydrates (ICGH) (6th : 2008)


You don't seem to have a PDF reader installed, try download the pdf

Item Metadata


59278-5341_1.pdf [ 723.32kB ]
JSON: 59278-1.0040972.json
JSON-LD: 59278-1.0040972-ld.json
RDF/XML (Pretty): 59278-1.0040972-rdf.xml
RDF/JSON: 59278-1.0040972-rdf.json
Turtle: 59278-1.0040972-turtle.txt
N-Triples: 59278-1.0040972-rdf-ntriples.txt
Original Record: 59278-1.0040972-source.json
Full Text

Full Text

SEEDING HYDRATE FORMATION IN WATER-SATURATED SANDWITH DISSOLVED-PHASE METHANE OBTAINED FROM HYDRATEDISSOLUTION: A PROGRESS REPORTW.F. Waite1?, J.P. Osegovic2, W.J. Winters1, M.D. Max2, D.H. Mason11U.S. Geological Survey 2Marine Desalinization Systems384 Woods Hole Road 1601 3rd St. SouthWoods Hole, MA 02543 St. Petersburg, FL 33701USA USAABSTRACTAn  isobaric  flow  loop  added  to  the  Gas  Hydrate  And  Sediment  Test  Laboratory  Instrument(GHASTLI)  is  being  investigated  as  a  means  of  rapidly  forming  methane  hydrate  in  water-saturated  sand  from  methane  dissolved  in  water.  Water  circulates  through  a  relatively  warmsource  chamber,  dissolving  granular  methane  hydrate  that was pre-made  from seed ice, then en-ters a colder hydrate growth chamber where hydrate can precipitate in a water-saturated sandpack.   Hydrate  dissolution  in the source  chamber  imparts  a known methane  concentration  to thecirculating water, and hydrate particles from the source chamber entrained in the circulating watercan become nucleation sites to hasten the onset of hydrate formation in the growth chamber.   Ini-tial results  suggest  hydrate grows  rapidly near the growth  chamber inlet.   Techniques  for estab-lishing homogeneous hydrate formation throughout the sand pack are being developed.Keywords: gas hydrates, methane, dissolved-phase, solubility                                                       ? Corresponding author: Phone: +1 508 457 2346 Fax +1 508 457 2310 E-mail: wwaite@usgs.govNOMENCLATUREA: cross-sectional area of the sample [cm2]k: permeability [cm2/s]k0: initial permeability [cm2/s]L: sample length [cm]M: moles of methane per cubic centimeter [mol/cc]?P: pressure difference across the sample [MPa]Q: fluid flow rate through the sample [cc/min]Q0:  initial  fluid  flow  rate  through  the  sample[cc/min]Subscripts:  Interface (Interface Chamber), Source(Source Chamber), Growth (Growth Chamber),G  (gas),  H  (hydrate),  IW  (initial,  pre-dissociation  water),  FW  (final,  post-dissociation water) [unitless]T: temperature [?C]V: volume in cubic centimeters [cc]?: dynamic viscosity of the fluid [Pa?s]?: porosity in hydrate-bearing sediment [unitless]?0: porosity in hydrate-bearing sediment [unitless]1. INTRODUCTIONFormation  of  naturally-occurring  gas  hydrate,which  is  most  commonly  methane  hydrate  [1],alters  sediment  properties  when  the  crystallinesolid  replaces  pore  water  [2,  3].   How  sedimentproperties change depends on where hydrate formswithin  the  pore  space,  which  in  turn  depends  onthe formation environment.   For example, in par-tially water-saturated, gas-rich environments, hy-drate  tends  to  cement  sediment  grains  together,and  even  a small  amount  of hydrate  significantlyincreases seismic wave speeds [4, 5].   In contrast,hydrate  formed  in  water-saturated  systems  fromProceedings of the 6th International Conference on Gas Hydrates (ICGH 2008),Vancouver, British Columbia, CANADA, July 6-10, 2008.Figure 1:  Flow loop schematic.  Warm, meth-ane-rich  water  is  pumped  from  the  methaneinterface  chamber  to  the  cooler  source  cham-ber, where it passes through a porous networkof  granular  methane  hydrate  pre-made  fromseed ice.  Hydrate dissolution raises the water?smethane content to the equilibrium solubility ata temperature TSource, which  exceeds  the  equi-librium solubility in the presence of hydrate forthe  water-saturated  sand  held  in  the  growthchamber  at  TGrowth.   Water  entering  the  sandpack  is  therefore  supersaturated  in  methane,acting  as  a methane  source  for additional  hy-drate formation.gas dissolved in pore water does not preferentiallygrow  at grain  contacts  [6].   Relative to cementa-tion,  hydrate  formation  away  from  grain  contactsgenerates  only a limited  wave speed increase  untilthe pore-space hydrate saturation exceeds 40-50%[7, 8].   Because  many  naturally-occurring  hydratereservoirs  are  thought  to  form  in  the  absence  offree  gas  [9],  testing  models  relating  pore-spacehydrate saturation to seismic wave speed requiressamples formed from dissolved-phase methane.Using a glass bead pack, Spangenberg et al. [10]saturated  ~95%  of  the  pore  space  with  hydrateformed from dissolved-phase  methane in ~50 daysby  circulating  water  via  an  interface  chamber  inwhich  water  dripped  through  methane  gas.   Weseek to accelerate the growth rate and reduce thetime  required  for  each  experiment  by  adding  ahydrate source chamber to the system described bySpangenberg et al. [10].The  mechanical  properties  of  hydrate-bearingsediment  depend  on  the  extent  to  which  hydratebinds sediment grains together [11, 12], so a con-cern  with  accelerated  hydrate  growth  is the  shapeof  the  hydrate  crystal  and  the  resulting  contactbetween  hydrate  and  sediment  grains.   As  thegrowth rate for hydrate increases, crystals form ina  more  dendritic  pattern  [13,  14].   When  thegrowth rate is slow, hydrate forms faceted crystalsand binds more strongly to the sediment particles[14].   The faceted crystal  growth  morphology canalso be attained by annealing rapid-growth hydrateat conditions near the hydrate phase boundary [15]however,  so the  primary  goal  in this  work was  todevelop  a  technique  to  form  hydrate  relativelyrapidly utilizing methane dissolved in water.We  have  configured  the  Gas  Hydrate  And  Sedi-ment Test Laboratory Instrument (GHASTLI) [2]as a flow loop.  Water leaving a methane gas/waterinterface  chamber  passes  through  a source  cham-ber  containing  granular  methane  hydrate,  pre-made  from  seed  ice,  before  entering  water-saturated sand in a growth chamber.   Circulatingwater  dissolves  hydrate  in  the  source  chamber,acquiring  a  well-constrained  methane  saturation,as  well  as  entraining  hydrate  micro-crystalssloughed off the source-chamber hydrate.   In thepresence of the hydrate crystals, water entering therelatively  cool  growth  chamber  becomes  super-saturated  with  respect  to  methane  as  it  cools.Methane in excess of the solubility limit can pre-cipitate  as  hydrate,  using  the  hydrate  micro-crystals as nucleation sites.  We describe this tech-nique here, and present results from two prelimi-nary tests.2. APPARATUSThe flow loop consists of three distinct subsystemsdesigned  to  control  the  methane  saturation  andmaintain a bubble-free growth chamber (Fig. 1): 1)a  room-temperature  methane  gas/water  interfacechamber  and  a  constant  flow-rate  circulatingpump,  2)  a  temperature-controlled  methane  hy-drate  source  chamber,  and  3)  a  temperature-controlled,  water-saturated  sand  sample  in  a  hy-drate growth chamber.2.1. Interface chamber and circulating pump:This subsystem is the gas-phase methane reservoir,which  imparts  a  preliminary  degree  of  methanesaturation  to  the  circulating  water  while  storinggas  purged  from  the  rest  of  the  apparatus.    Ini-tially,  the  flow  loop  is  brought  to  its  operatingpressure of 12 MPa using pressurized methane gasintroduced  through  the  flow  loop  "supply"  inlet(Fig. 1).A Quizix-brand  QL-6000  dual-piston  pump deliv-ers water at a constant, user-defined flow rate asdescribed  in the procedure. Once flow through thecomplete  system  begins,  methane-depleted  waterreturning  from  the  growth  chamber  enters  the  topof  the  interface  chamber,  partially  resaturatingwith methane as it drips through free methane gas.A  custom-designed  float  rests  on  the  gas-waterinterface,  monitored  via a  Schaevitz-brand  mag-netic sensor to track free gas consumption. Pres-sure is monitored using a pressure sensor locatedat the top of the interface chamber.   Temperatureis  measured  by  a  thermistor  held  against  thechamber's  outer wall.   A lamp warms the interfacechamber's  downstream  end,  reducing  the  methaneconcentration  in  the  circulating  water  below  thesource  vessel's  equilibrium  methane  solubility  inthe  presence  of  hydrate  (See  "Interface"  and"Source Chamber" circles in Fig. 2).2.2. Source chamber:Pure,  granular  methane  hydrate  is  formed  in  thesource  chamber  in  a  separate,  high-pressure  sys-tem  prior  to  being  connected  to  the  flow  loop.Initially, water ice is mixed with a small amount ofliquid nitrogen, ground with a mortar and pestle,then sieved to obtain the 180-250 ?m  grain  sizefraction.  Following the method of Stern et al. [16,17],  methane  hydrate  is  formed  by  warming  thegranular  water  ice  in  a  pressurized  methane  at-mosphere from -20 to +17?C over the course of 17hours.   The sample is held at 17?C for 24 hoursbefore  being  reduced  to  6?C  in  preparation  fortransfer  into  the  flow  loop.  Once  hydrate  forma-tion  is  complete,  the  source  chamber  contains  aporous  network  of  methane  hydrate  grains,  withmethane gas-filled pore space.The chamber is transferred, under pressure and at6?C,  to  a  temperature-controlled  bath  and  con-nected to the flow loop system. To purge free gasfrom  the  granular  methane  hydrate,  water  ispumped  backward  through  the  source  chamber(from bottom to top) and into the interface cham-ber  (Fig.  1).   Flow  continues  through  the  purgeline until gas no longer enters the interface cham-ber, as indicated by stabilization of the interfacechamber float position.   During  flow-loop  circula-tion, water is pumped through the source chamberfrom  top  to  bottom  to  ensure  residual  gas-phasemethane remains in the source chamber rather thanbeing  transported  to  the  growth  chamber.   Tem-perature  is  measured  via  two  internal  thermocou-ples,  one near  each  end of the chamber.   Pressureis monitored with sensors in the flow line betweenthe  interface  and  source  chambers,  and  betweenthe source and growth chambers.Methane hydrate dissolution in the source chambernot  only  provides  a  means  of  imparting  a  well-constrained  methane saturation  level to the circu-lating  water  that is nearly  independent  of pressure[18],  but  can  also  slough  off  hydrate  particleswhich can then nucleate hydrate formation in thegrowth chamber [13].   Given the 1-5 cc/min flowrates, the chamber geometry and exposed hydratesurface area in the source chamber, water leavingthe  source  vessel  would  be  fully  saturated  withmethane  even  if  the  methane  hydrate  dissolutionrate  in  fresh  water  were  nearly  twelve  orders  ofmagnitude slower than the 0.4 mmol CH4/m2?s ratedetermined  by  Rehder  et  al.  [19]  in  experimentsconducted in Monterey Bay.Figure  2:  Methane  solubility  at  12  MPa  forindividual system components (circles), plottedon the equilibrium solubility curve for methanehydrate and water (solid curve), or methane gasand water  (dashed  curve).   Equilibrium  curvesare calculated  from the online  models  by Duan[27, 30].   The interface chamber must be heldat a temperature  high enough  to keep the solu-bility  below  that  of  the  source  chamber,  orhydrate will form, rather than dissolve, in thesource chamber and clog the flow loop.2.3. Growth chamber:Growth chamber characteristics and measurementcapabilities are described in detail by Winters et al.[2].   The  growth  chamber  contains  a  water-saturated,  Ottawa  sand  with  a  0.25  to  0.5-mmgrain size range.  The cylindrical sample (nominallength: 134 mm, diameter: 72 mm) is jacketed in a0.65-mm-thick  Viton  membrane  and  capped  atboth  ends  with  titanium  endcaps.   The  endcapshouse  transducers  for  measuring  acoustic  wavespeeds  along the sample's  long axis.   Circulatingwater enters (or exits) the sample through annulardiffuser plates in each endcap.Confining and pore pressures are maintained inde-pendently  by  Isco-brand  500D  syringe  pumps.Pressures  are  measured  using  sensors  located  inthe flow and chamber pressurization lines outsidethe growth chamber.Temperature is maintained using an external bathpumping  ethylene  glycol  through  cooling  coilssurrounding  the  chamber  and  through  a  heat  ex-changer  held  against  the  sample's  top  endcap.Temperature is monitored with four thermocouplesand  four  thermistors  within  the  chamber,  heldagainst the sample sides and spaced to cover thefull  length  of  the  sample.   Each  endcap  has  anembedded thermocouple.3. PROCEDUREA flow-loop test consists of three phases: 1) pres-surization, 2) hydrate growth, and 3) dissociation.3.1. Pressurization:The three flow-loop subsystems are initially pres-surized independently.   The hydrate source cham-ber is connected under pressure, and the system isthen  pressure-equilibrated.   The  flow  loop  andpore pressure are increased to 12 MPa by feedingmethane through the supply line near the top of theinterface chamber (Fig. 1).  The confining pressureis raised to 12.25 MPa  using  an Isco-brand  500Dsyringe pump to impart a 0.25 MPa effective stresson  the  water-saturated sand  in  the  growth  cham-ber.   In addition to simulating the confining loadon  buried  sediment,  the  effective  pressure  holdsthe sample jacket against the sample, forcing flowto pass through, rather than around, the sand pack.3.2. Hydrate growth:Hydrate  formation  and  growth  are  regulated  bymanipulating methane solubility with temperature[13].   Two  hydrate  formation  techniques  havebeen  tested:  1)  continuous  flow,  with  a  constanttemperature  difference  between  the  source  andgrowth chamber, and 2) episodic flow with a tem-perature cycle between flow intervals.In the continuous-flow case, the source chamber isheld at 12?C, which, at 12 MPa, is within the hy-drate stability field and near the peak of the solu-bility curve shown in Fig. 2.  This temperature waschosen to maximize the quantity of methane trans-ported  to  the  sand  sample,  held  at  6?C,  whilemaintaining  nearly  a 3?C temperature  window  forfluid  to  warm  during  transport  to  the  growthchamber  without  producing  methane  bubbles.Water  exiting  the  interface  chamber  must  beheated  above  25?C  to  ensure  that  it  is  under-saturated  with  methane  relative  to  water  in  thesource  chamber (Fig.  2).   This  requirement forceswater  entering  the  source  chamber  to  dissolve,rather  than  form,  methane  hydrate.   Circulatingwater entering  the growth  vessel  is cooled  relativeto  the  source  vessel,  and  in  the  presence  of  hy-drate,  this  cooling  produces  a  state  of  methanesupersaturation  in  the  water  entering  the  growthchamber.The  flow  rate  is  initially  set  to  5  cc/min,  thenmanually reduced over time as the sample  perme-ability  decreases  and  the  pressure  required  tomaintain flow increases.  Flow cannot be driven bypressures  exceeding  the  12.25  MPa  confiningpressure  in  the  growth  chamber,  or  circulatingwater  would  be  able  to  expand  the  Viton  jacketand  flow  around,  rather  than  through,  the  sandsample.   The hydrate formation phase ends when a1 cc/min flow rate can no longer be maintained.For the episodic flow case, the source chamber isstill held at 12?C and 12 MPa.   During periods offlow,  however,  the  growth  chamber  is  held  at10?C,  while  a  minimum  of  1800  cc  of  methane-rich  water  is  pumped  through  the  sample  for  ~6hours at 5 cc/min.   This volume is ten times thesample's  total  pore  space.   Between  periods  offlow, the growth chamber is cooled to 6?C to in-crease  the  methane  supersaturation  and  thus  theamount  of  methane  available  for  hydrate  forma-tion.  The system is allowed  to equilibrate for ~43hours before  rewarming  to 10?C and repeating  thecycle.   In the test described here, six flow cycleswere completed.3.3. Hydrate dissociation:With  the  inlet  and  outlet  flow  lines  closed,  thegrowth  chamber  temperature  is  raised  to 20?C at.66 ?C/hour, dissociating any hydrate in the growthchamber.   The total amount of hydrate containedin the growth chamber is estimated from the meth-ane  solubility  and  the  observed  pore  pressure  re-sponse during dissociation, as discussed below.4. RESULTS AND DISCUSSION4.1. Continuous flow test:Open  circles  in  figure  3  show  the  pore  pressureincrease  in  the  growth  chamber  in  response  towarming the sample through the hydrate stabilitytemperature (diamonds and solid curve, data com-bined  from Jhaveri  and Robinson  [20], deRoo etal. [21], Yang et al. [22]).   The initial pore pres-sure  is slightly  below  8.4 MPa because  of a shearstrength  test  performed  prior  to dissociation.   Ax-ial deformation  of the sample  during  shear  causesthe pore space to dilate, reducing the pressure from12 to ~8.4 MPa.   Temperature measurements aremade  on  the  outer  sample  surface,  meaning  thesample interior is slightly cooler, particularly dur-ing the period of active dissociation.  The pressureincrease  along  the  methane  hydrate  equilibriumcurve indicates methane hydrate is breaking down,increasing pressure as free gas forms from dissoci-ating hydrate.   Once  hydrate is  consumed,  a por-tion  of  the  free  gas  produced  goes  into  solution,lowering  the measured  pore pressure  to 9.09 MPafrom its peak of ~9.2 MPa.To calculate the hydrate volume, the sealed growthchamber is assumed to contain a constant numberof  methane  molecules  before  and  after  dissocia-tion.   Prior  to  dissociation,  the  pore  space  is  as-sumed to contain only methane hydrate and water,with free methane gas and methane-saturated wa-ter  being  the  only  post-dissociation  pore  constitu-ents.   The methane  budget  can be cast in terms  ofmoles of methane per cc in each phase, M, and thevolume of each phase, V as:MH ? VH + MIW ? VIW = MG ? VG + MFW ? VFW, (1)where the subscripts  H and G refer to hydrate  andgas,  while  IW and FW  refer  to  the  initial,  pre-dissociation  water  volume,  and  the  final,  post-dissociation water volume, respectively.When hydrate dissociates,  the water  volume pro-duced can be related to the original water  volumeusing the density of water molecules in the hydratephase compared to the density of liquid water.  Atpre-dissociation  conditions  of  8.55  MPa  and11.6?C, point "A" in Fig. 3, the methane hydrateunit cell volume given by  Shpakov et al. [23] is1.71x10-21 cc, meaning the hydrate density consid-ering  water  molecules  alone  is  0.804  g/cc.   Fol-lowing dissociation, the water density at 9.09 MPaand 13.4?C, is 0.99958 g/cc [24], point "B" in Fig.3.   Relative to the initial hydrate volume, VH, theinitially  hydrate-bound  water  occupies  a  volumeVH  ? (0.804 g/cc)/(0.99958 g/cc) = 0.805 ? VH afterdissociation.   Gas takes up the remaining portionof the volume previously  occupied by hydrate: VG=  0.195  ? VH.   For  one  cubic  centimeter  of  porespace, Eq. 1 volumes can be related to the initialhydrate volume through:VIW = 1 - VH, (2a)VFW = VIW + 0.805 ? VH,       (2b)VG = 0.195 ? VH.   (2c)The number  of moles  of methane  per cc, M, mustbe calculated for each phase:Hydrate: In methane hydrate,  a non-stoichiometricmaterial  in  which  the  ratio  of  methane  to  watermolecules varies, we assume a constant ratio of 6water molecules per methane molecule, found byCircone et al. [25] to be representative of methanehydrate held near its phase boundary.   Combiningthis  ratio  with  the  unit  cell  volume  measured  byFigure  3:  Pore  pressure  versus  temperatureduring  hydrate  dissociation  in  the  growthchamber.   Tracking  of  the  measured  pressureand  temperature  (circles)  along  the  hydrateequilibrium  curve  (diamonds  and  solid  curve,from Jhaveri  and Robinson  [20], deRoo et al.[21],  Yang  et  al.  [22])  indicates  hydrate  hasformed in the growth chamber.Shpakov et al. [23] for point "A" in Fig. 3,  yieldsMH = 7.4x10-3 moles of methane per cc of hydrate.Gas: For post-dissociation conditions of 9.09 MPaand 13.4?C, point "B" in Fig. 3, the methane den-sity  is  4.6x10-3 moles of methane per cc gas [26,27].Water:   Prior  to  dissociation  at  8.55  MPa  and11.6?C, point "A" in Fig. 3, MIW = 8.17x10-5 molesof methane per cc in the pore water [28-30].   Fol-lowing dissociation, at 9.09 MPa and 13.4?C, MFW= 1.17x10-4 moles of methane per cc in the porewater [28-30].From  the  volume  of  confining  fluid  surroundingthe sample in the chamber, a ~0.62 cc increase insample  volume  is  observed  during  dissociation.This volume increase must be taken up by the gasphase, because water and sand grains are relativelyincompressible. The largest uncertainty is in esti-mating the pore space increase during dissociation,conservatively  leading  to  an  overall  pore-spacehydrate saturation uncertainty estimate of ?0.1%.With the added pore gas volume, a 0.33% increasein the total pore space, equation 2c becomes VG =0.195 ? VH + 0.0033.  The hydrate saturation VH is0.8 ? 0.1% of the total porosity.   Over the entire189 cc of pore space, an estimated 1.5 ? 0.2 cc ofhydrate was formed.4.2. Hydrate effect on permeability:Given  the circulating  pump's  flow  rate and pres-sure  measurements  made  in  the  inlet  and  outletlines  of  the  growth  chamber,  the  sample  perme-ability can be calculated using Darcy's law:  ? Q = kA? ? ?PL ,  (3)where the flow rate, Q, is given by the permeabil-ity,  k,  cross-sectional  area  of  the  sample,  A, dy-namic  viscosity  of  the  fluid,  ?,  and  the  pressuredifference ?P, driving  flow through  the sample  oflength  L.   Only  the  permeability  relative  to  theinitial permeability, k0,  is  of  interest  here,  so  as-suming  a  constant  cross-sectional  area,  samplelength,  and  viscosity,  the  fractional  permeabilitychange is calculated as follows:    ? kk0 =QQ0 ??P0?P ,  (4)where  Q0  and  ?P0  are  the  initial  flow  rate  andpressure  drop  across  the  sample,  respectively.Hydrate formation causes the permeability to fallto approximately 20% of its original value (Fig. 4).Assuming  hydrate  forms  only  in  the  water-saturated sand and not in the sample endcaps, wecan  use  the  Kozeny-Carman formulation to con-strain the hydrate distribution [31]:    ? kk0 =??0? ? ? ? ? ? 3? 1??01??? ? ? ? ? ? 2,        (5)where  ?0  is  0.335,  the  initial  hydrate-free  sandporosity,  and ? is  the porosity  of a homogeneousdistribution of sand and hydrate.  Given the calcu-lated  hydrate  content  of 1.5 ? 0.2 cc and  the  72-mm sample  diameter,  Eq. 5 suggests  the observedpermeability decrease is due to ~38% pore spacehydrate saturation  in a ~2.7 mm layer of the oth-erwise water-saturated sand sample.This distribution indicates the methane supersatu-ration  may  be  too  high  upon  entering  the  water-saturated sand, causing hydrate to form just as themethane-rich water enters the sample, choking offsubsequent  flow  before  hydrate  can  form  morebroadly within the sample.  To allow methane-richwater  to permeate the sample  prior to hydrate for-mation, an episodic flow technique was employedin a subsequent test as described below.Figure  4:  Permeability  change  in  the  sampleduring  the  continuous  flow  test  (circles),  andthe  25-point  running  average  (solid  curve).Permeability  drops  by  approximately  a  factorof 5 over  the  course  of the  hydrate  formationprocess.4.3. Episodic flow technique:To avoid the plugging  problem  encountered  in thecontinuous-flow test, this technique is intended touniformly  distribute  methane-rich  water  and  hy-drate particles obtained from the source chamberthroughout the sand pack prior to cooling the sam-ple in the absence of flow.   In the presence of hy-drate, cooling increases the water's methane super-saturation  (Fig.  2),  triggering  additional  hydrategrowth nucleated around the hydrate particles.Rewarming  the  sample  reverses  the  process,  in-creasing  the  pore-water  methane  solubility,  andleading  to  hydrate  dissolution.    It  is  hoped  thatsubsequent  warming  of  the  sample  to  10?C  andrestarting flow of methane-rich water to replenishthe methane concentration in the growth chambercan  be  accomplished  quickly  enough  to  avoidcompletely dissolving  hydrate  formed  during  thecooling  cycle.   Permeability  measurements  duringperiods  of flow  show no sign  of a net  increase  inhydrate content from one cooling cycle to the next,however.4.4. Verifying the absence of free gas:Methane bubbles in an otherwise water-saturatedsediment attenuate and slow compressional waves[32].   Figure  5  displays  compressional  wavemeasurements taken following hydrate formationin the continuous flow test (solid curve), comparedto  measurements  taken  in  a  water-saturated  sandunder identical pressure conditions (dashed curve).The hydrate-free sample has slightly less porositythan  the  hydrate-bearing  sample  initially  con-tained,  and  correspondingly,  has  a slightly  fasterand  stronger  signal.   Though  the  amplitude  de-crease  is  suggestive  of  methane  gas,  comparablewave speeds indicate the potential quantity of gas-phase methane is low.4.5. Future direction:To balance the rapid-formation objective with thetendency for rapid hydrate formation near the flowinlet  to  the  growth  chamber,  ongoing  tests  willutilize  a  slow  temperature-cooling  ramp  for  thegrowth chamber.  As with the tests presented here,more than ten total pore-volumes of methane- andhydrate-particle-rich  water  will  pass  through  thegrowth chamber while the chamber is only slightlycooler than the source. Once the growth chamberis  primed,  flow  will  continue  while  the  growthchamber is cooled slowly, starting from the down-stream end of the sample, opposite the flow inlet.By increasing the driving force for hydrate forma-tion slowly, it is hoped methane and hydrate parti-cles  will  initiate  significant  hydrate  growth  onlyafter  permeating  the  sample  and  moving  awayfrom the flow inlet.5. CONCLUSIONSForming  methane  hydrate  from  dissolved-phasemethane is critical for mimicking oceanic hydrateformation in marine sediments.   To provide boththe  methane  supersaturation  and  nucleation  sitesrequired  for  hydrate  formation,  a  flow  loop  hasbeen designed in which circulating water dissolvesgranular  methane  hydrate  in  a  source  chamber,thereby  developing  a  well-constrained  methanesaturation  while  entraining  hydrate  micro-crystalsthat can shorten the induction period for hydrateformation in the growth  chamber.   This  techniquerapidly forms hydrate close to the flow inlet sideof  the  sample,  choking  off  subsequent  flow.  Atemperature ramp approach for slowly building thehydrate  formation  driving  force  is  under  investi -gation.Figure  5:  Compressional  waveform  in  a  hy-drate-bearing  sample  (solid  curve)  closelymatches  the  waveform  through  a  water-saturated  sand in which  no hydrate  or methaneis  present  (dashed  curve).   Effective  stress  isthe same for both samples.   The similarity be-tween wave speeds indicates free methane gas,which  slows  compressional  waves,  is  not  pre-sent in significant quantities.6. ACKNOWLEDGMENTSU.S.  Geological  Survey  contributions  were  sup-ported  by  the  Gas  Hydrate  Project  of  the  U.S.Geological Survey's Coastal and Marine GeologyProgram.   USGS and Marine Desalinization Sys-tems (MDS) contributions were also supported byDepartment  of  Energy  Contract  No.  DE-AI21-92MC29214.   Any use of trade names is for de-scriptive  purposes  only  and  does  not  imply  en-dorsement by the U.S. Government.REFERENCES[1] Kvenvolden,  K.A.,  Natural  gas  hydrate;  in-troduction  and  history  of  discovery,  in  NaturalGas  Hydrate  In  Oceanic  and  Permafrost  Envi-ronments, M.D. Max, Editor. 2000, Kluwer Aca-demic Publishers: Dordrecht,  Netherlands.  p. 9-16.[2] Winters, W.J., W.P. Dillon, I.A. Pecher, andD.H.  Mason,  GHASTLI;  determining  physicalproperties  of  sediment  containing  natural  andlaboratory-formed  gas  hydrate,  in  Natural  GasHydrate  In  Oceanic  and  Permafrost  Environ-ments,  M.D.  Max,  Editor.  2000,  Kluwer  Aca-demic  Publishers:  Dordrecht,  Netherlands.  p.311-322.[3] Hornbach,  M.J.,  W.S.  Holbrook,  A.R.  Gor-man,  K.L.  Hackwith,  D.  Lizarralde,  and  I.Pecher,  Direct  seismic  detection  of methane  hy-drate  on  the  Blake  Ridge.  Geophysics,  2003.68(1): p. 92-100.[4] Waite, W.F., W.J. Winters, and D.H. Mason,Methane  hydrate  formation  in  partially  water-saturated  Ottawa  sand.  American  Mineralogist,2004. 89: p. 1202-1207.[5] Priest,  J.A.,  A.  Best,  and  C.R.  Clayton,  Alaboratory  investigation  into  the  seismic  veloci-ties of methane gas hydrate-bearing sand. Journalof Geophysical Research, 2005. 110: p. B04102,doi:10.1029/2004JB003259.[6] Tohidi, B., R. Anderson, M.B. Clennell, R.W.Burgass, and A.B. Biderkab, Visual observationof  gas-hydrate  formation  and  dissociation  insynthetic  porous  media  by  means  of  glass  mi-cromodels. Geology, 2001. 29(9): p. 867-870.[7] Dvorkin, J., M.B. Helgerud, W.F. Waite, S.H.Kirby,  and  A.  Nur,  Introduction  to  physicalproperties and elasticity models, in Natural GasHydrate  In  Oceanic  and  Permafrost  Environ-ments,  M.D.  Max,  Editor.  2000,  Kluwer  Aca-demic  Publishers:  Dordrecht,  Netherlands.  p.245-260.[8] Yun, T.S.,  F.M. Francisca,  J.C. Santamarina,and  C.  Ruppel,  Compressional  and  shear  wavevelocities  in  uncemented  sediment  containinggas hydrate. Geophysical Research Letters, 2005.32: p. L10609, doi:10.1029/2005GL022607.[9] Buffett,  B.  and  O.  Zatsepina,  Formation  ofgas hydrate  from dissolved  gas in natural  porousmedia. Marine Geology, 2000. 164: p. 69-77.[10] Spangenberg,  E.,  J.  Kulenkampff,  R.  Nau-mann,  and  J.  Erzinger,  Pore  space  hydrate  for-mation in a glass bead sample from methane dis-solved  in  water.  Geophysical  Research  Letters,2005.  32:  p.  L24301,doi:10.1029/2005GL024107.[11] Helgerud, M.B., J. Dvorkin, A. Nur, A. Sakai,and  T.  Collett,  Elastic-wave  velocity  in  marinesediments with gas hydrates: Effective mediummodeling.  Geophysical  Research  Letters,  1999.26(13): p. 2021-2024.[12] Yun, T.S., J.C. Santamarina, and C. Ruppel,Mechanical  properties  of  sand,  silt,  and  claycontaining  tetrahydrofuran  hydrate.  Journal  ofGeophysical  Research,  2007.  112:  p.  B04106,doi:10.1029/2006JB004484.[13] Osegovic, J.P., S.R. Tatro, and S.A. Holman,Physical chemical characteristics  of natural gashydrate,  in  Economic  Geology  of  Natural  GasHydrate,  M.D.  Max,  A.H.  Johnson,  and  W.P.Dillon,  Editors.  2006,  Springer:  Dordrecht,Netherlands. p. 45-104.[14] Katsuki, D., R. Ohmura, T. Ebinuma, and H.Narita,  Formation,  growth  and  ageing  of  clath-rate hydrate crystals in a porous medium. Phili-sophical Magazine, 2006. 86(12): p. 1753-1761.[15] Stern, L.A., S.H. Kirby, S. Circone, and W.B.Durham, Scanning electron microscopy investi-gations  of  laboratory-grown  gas  clathrate  hy-drates  formed  from  melting  ice,  and comparisonto  natural  hydrates.  American  Mineralogist,2004. 89(8-9): p. 1162-1175.[16] Stern,  L.A.,  S.H.  Kirby,  and  W.B.  Durham,Peculiarities  of methane  clathrate  hydrate  forma-tion and solid-state  deformation,  including  possi-ble  superheating  of  water  ice.  Science,  1996.273(5283): p. 1843-1848.[17] Stern,  L.A.,  S.H.  Kirby,  and  W.B.  Durham,Polycrystalline methane hydrate: Synthesis fromsuperheated  ice, and low-temperature  mechanicalproperties. Energy & Fuels, 1998. 12(2): p. 201-211.[18] Lu, W., I.M. Chou, and R.C. Burruss, Deter-mination of methane concentrations  in water  inequilibrium  with  sI  methane  hydrate  in  the  ab-sence of a vapor phase by in situ Raman spec-troscopy.  Geochimica  et  Cosmochimica  Acta,2008. 72: p. 412-422.[19] Rehder, G., S.H. Kirby, W.B. Durham, L.A.Stern, E.T. Peltzer, J. Pinkston, and P.G. Brewer,Dissolution  rates  of  pure  methane  hydrate  andcarbon  dioxide  hydrate  in  undersaturated  sea-water  at  1000-m  depth.  Geochimica  et  Cosmo-chimica Acta, 2004. 68(2): p. 285-292.[20] Jhaveri,  J.  and  D.B.  Robinson,  Hydrates  inthe  methane-nitrogen  system.  The  CanadianJournal  of  Chemical  Engineering,  1965.  43:  p.75-78.[21] deRoo,  J.L.,  C.J.  Peters,  R.N.  Lichtenthaler,and G.A.M. Diepen, The occurrence of methanehydrate  in saturated  and unsaturated  solutions  ofsodium  chloride  and  water  In  dependence  oftemperature and pressure. AIChE Journal, 1983.29(4): p. 651-657.[22] Yang, S.O., S.H. Cho, H. Lee, and C.S. Lee,Measurement  and prediction of phase  equilibriafor water plus methane in hydrate forming con-ditions.  Fluid  Phase  Equilibria, 2001. 185(1-2):p. 53-63.[23] Shpakov,  V.P.,  J.S.  Tse,  C.A.  Tulk,  B.Kvamme,  and  V.R.  Belosludov,  Elastic  modulicalculation and instability in structure I methaneclathrate  hydrate.  Chemical  Physics  Letters,1998. 282(2): p. 107-114.[24] Weast,  R.C.,  CRC  Handbook  of  Chemistryand  Physics.  68  ed.  1987,  Boca  Raton:  CRCPress, Inc.[25] Circone, S., S.H. Kirby, and L.A. Stern, Di-rect measurement of methane hydrate composi-tion along the hydrate equilibrium boundary. J.Phys. Chem. B, 2005. 109: p. 9468-9475.[26] Duan,  Z.,  N.  Moller,  and  J.H.  Weare,  Anequation of state  for the CH4-CO2-H20  system:I. Pure systems from 0 to 1000 ?C and 0 to 8000bar.  Geochimica  et  Cosmochimica  Acta,  1992.56: p. 2605-2617.[27] Duan,  Z.  Pure  phase  systems  -[28] Sun,  R.  and  Z.H.  Duan,  Prediction  of  CH4and CO2 hydrate phase equilibrium  and cage oc-cupancy from ab initio intermolecular potentials.Geochimica  et  Cosmochimica  Acta,  2005.69(18): p. 4411-4424.[29] Duan,  Z.H.  and S.D.  Mao,  A thermodynamicmodel for calculating methane solubility, densityand  gas  phase  composition  of  methane-bearingaqueous fluids from 273 to 523 K and from 1 to2000  bar.  Geochimica  et  Cosmochimica  Acta,2006. 70(13): p. 3369-3386.[30] Duan,  Z.  Methane  clathrate  hydrate  - Liquidequilibrium calculations.   2006[31] Mavko,  G.,  T. Mukerji,  and J.  Dvorkin, Therock physics handbook: Tools for seismic analy-sis  in  porous  media.  1998,  Cambridge:  Cam-bridge University Press. 329.[32] Yun,  T.S.,  G.A.  Narsilio,  J.C.  Santamarina,and  C.  Ruppel,  Instrumented  pressure  testingchamber  for  characterizing  sediment  cores  re-covered  at  in  situ  hydrostatic  pressure.  MarineGeology, 2006. 229: p. 285-293.


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



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"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items