Tailings and Mine Waste Conference

Slurry to soil clay behaviour model : using methylene blue to cross the process / geotechnical engineering… Wells, Patrick Sean; Kaminsky, H. A. W. Oct 31, 2015

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

Item Metadata


59368-Wells_P_et_al_Slurry_soil_clay.pdf [ 357.79kB ]
JSON: 59368-1.0340913.json
JSON-LD: 59368-1.0340913-ld.json
RDF/XML (Pretty): 59368-1.0340913-rdf.xml
RDF/JSON: 59368-1.0340913-rdf.json
Turtle: 59368-1.0340913-turtle.txt
N-Triples: 59368-1.0340913-rdf-ntriples.txt
Original Record: 59368-1.0340913-source.json
Full Text

Full Text

Proceedings Tailings and Mine Waste 2015 Vancouver, BC, October 26 to 28, 2015 Slurry to soil clay behaviour model – using methylene blue to cross the process / geotechnical engineering divide P. S. Wells Suncor Energy H. A.W. Kaminsky Suncor Energy ABSTRACT The models and formulae used to describe the behaviour of materials vary based on the history of the individuals and groups who defined the area of study. In the case of tailings behaviour, two main disciplines describe either end of the processes with very little in the way of common descriptors to bridge the base scientific and engineering principles. Process Engineering dominates the descriptions for tailings materials in their suspended solids or liquid slurry state. Based on various incarnations of rheology, including terms such as viscosity or yield stress, as well as material and chemical descriptors such as clay content, particle specific gravity, pH, and ionic concentration, all contribute to a well-defined behaviour model.  Geotechnical Engineering dominates the descriptions for the same materials, once they have dewatered sufficiently to be described by the Terzaghi principle. Terms such as effective stress, pore water pressure, permeability, and coefficients of consolidation are used to monitor and predict the behaviour of these materials once deposited. Tailings engineering straddles these two disciplines and practitioners need to be fluent in both descriptions. However, these two areas do not readily translate and there are many cases where similar terms actually provide very different meaning (“water content” being one prime example). One area which could bridge the gap for clay-dominated tailings is through the methylene blue index (MBI) measurement and its clear correlation to rheology / strength, as well as correlations to liquid and plastic limits.  This paper uses data from tailings in the oil sands of northeastern Alberta, Canada to show the relationship between MBI and the gradual transition from slurry to soil. 1 INTRODUCTION TO MBI The methylene blue index is a number used to describe the clay content and activity of a sample. The methylene blue test measures how much methylene blue dye can be absorbed by a sample as determined by a titration test. The amount of methylene blue that can be absorbed has been found to correlate with many other tests including water active surface area, and the amount of clay minerals measured by XRD. More importantly the MBI has been found to be a useful predictor of poor recovery in oil sands extraction and more recently as the key indicator of process ability for Suncor’s MFT flocculation process. (Diep et al, 2014), (Omotoso et al., 2014). MBI has also been used in other industries to to characterize the stratigraphy of soil foundations and identify unstable layers (Chiappone et al, 2004). The methylene blue index test most commonly used in oil sands was developed by Amar Sethi at MRRT labs (Sethi,1995), as a modification from the ASTM test method. The test method has since been refined by various groups, most particularly CANMET energy (Omotoso & Morin, 2008) and more recently the Clay Focus Group (Kaminsky, 2014).  The fundamental MB test measures the number of milliliters of MB solution absorbed a given mass of sample. Usually the mass of sample used is small so the end point of the titration is Proceedings Tailings and Mine Waste 2015 Vancouver, BC, October 26 to 28, 2015 often less than 20 mLs. The standard titration step size is often 1mL and so the relative uncertainty of the end point is usually in the order of 5-10% on a relative basis and usually ~0.6meq/100g on an absolute basis. The measured milliliters of MB are then converted to one of several numbers used in report.  For the purpose of this paper three numbers are used:   MB index (MBI) – this is the milliequivalents of MB per 100 g of sample. It is calculated by:  Where the molarity of MB used is 0.006 N for tests using the Sethi, CANMET or clay focus group methods. For ores this number is usually between 0.25-2 meq/100g. For MFT this number is usually between 5 and 15 meq/100g.   % Clay as determined by MB - this generally takes the measured MBI and applies an empirically derived conversion between the MBI and the % clay mineral as measured by Sethi by XRD&PSD on a few samples of oil sands clay:  It is important to note that this equation can and does provide % clays in excess of 100%. This % clay is an index that describes activity and while it is a property that scales with mass (and thus can be used in much the same way as mass % values) does not imply that % clay +% non clay = 100% and so care must be taken in its application.  Surface area as determined by MB – this is a conversion based on the assumption of a monolayer of methylene blue absorbed on all available surfaces and that the surface area of the methylene blue is as described by Hang & Brindley.  The conversions to % clay and surface area create much confusion as the conversions are empirical in nature and thus are subject to revision. The important point is that such empirical correlations can be developed and that they are useful in providing a conceptual basis for understanding behavior. In practical terms the MBI on its own is the more direct measurement and is the value that should be used to report the activity of a given tailings sample. The caveat to the usefulness of MBI testing is that the test must be done properly with good quality control checks and attention paid to dispersion, pH, halo determination etc. This proviso is common for all testing but is often overlooked when procedures are adapted without proper understanding of the first principles behind the test ( Kaminsky, 2014), (Currie et al, 2014).  2 CORRELATION WITH ATTERBERG LIMITS There are several published correlations between Atterberg limits and MBI in the geotechnical literature (Cerato, 2001). Unfortunately, none have been published for oil sands. Published data where Atterberg limits and MBI have both been measured on oil sands fine tailings show a disappointing correlation. Atterberg limit tests are also an index test and as such the results are best compared when the tests have been conducted in the same mine. The 2013 presentation by Gidley highlighted that the test method used for Atterberg limit testing can have a significant impact on the results. As such, the lack of correlation from literature is unsurprising as there is a limited amount of data where MBI and Atterberg limits were tested over a significant range using consistent methods. This is the topic of ongoing investigations.  Proceedings Tailings and Mine Waste 2015 Vancouver, BC, October 26 to 28, 2015 3 STRENGTH AND CLAY ACTIVITY Theory According to Mitchell and Sogi (2005) the total strength of a clay is composed on two distinct parts: a cohesion that depends only a on void ratio (Water content), and a frictional component dependent on normal effective stress. The cohesion is known to be related to the strength of the inter-particle bonding in the sample – which in turn is related to the water chemistry, organic content and the type and quantity of clay minerals in a sample.  Atterberg limits are one method used by geotechnical engineers to understand the cohesion of a material via the liquidity index. Atterberg limits combine a range of factors into a single index of cohesion – these factors include the impact of organics, water chemistry, clay abundance and clay mineralogy. The challenge with these limits in a process environment is two-fold: firstly they combine all the process variables into the single index which makes it difficult for process engineers to understand how they need to change their process without doing many limit tests, secondly the time and material required for Atterberg limit tests is large relative to most process bench tests. It usually takes 300g+ of solid material for the tests and may take a day or two to conduct whereas process samples are usually on the order of 100g and can be generated by the hundreds a day. This disconnect means that Atterberg limits are often not done in the early stage of process development. The time component also means that this limit testing isn’t an effective process control tool as the number of tests would be prohibitive to control in a large process.  The theory we propose is that the clay to water ratio as measured by methylene blue index provides a suitable proxy measurement of the total type and quantity of clay minerals in a sample. In cases where the water chemistry and effective stress conditions are relatively constant this means that the resultant strength variations should be dominated by the clay to water ratio. This scenario is particularly relevant to tailings engineering where large variations occur in water content, clay abundance and clay mineral type of the tailings material filling a structure but that process conditions dictate relatively slow change in water chemistry and tailings deposition methods and containment design dictate relatively constant (or predictable) effective stress conditions. Laboratory Testing Material and rheological characterisations of samples obtained from various tailings ponds were conducted over several campaigns from various source tailings ponds. Samples were selected from this dataset based on the following: 1. Each sample included was analysed for, as a minimum: a. Weight percent mineral (% solids) b. Weight percent solids passing 44 micron (% fines) c. Calculated Sand-to-Fines Ratio (SFR) i. (100% - % fines) / (% fines) d. Weight percent bitumen (% bitumen) e. Peak yield stress f. Methylene blue index test i. Determine CWR and % clay on solids by MBI 2. Samples were distinctly fines dominated, and specifically clay dominated g. SFR < 1.0 AND/OR h. % Clay by MBI > 20% Proceedings Tailings and Mine Waste 2015 Vancouver, BC, October 26 to 28, 2015 In some cases, weight percent fines passing 44 microns were not available, so the weight percent lay value was used, assuming this would equate to an SFR <1.0. The resulting dataset of consisted of 295 individual samples. The programmes involved analysis by rheology labs as well as labs conducting geotechnical testing, but the measurements of yield stress were made under comparable conditions with a Brookfield 5XHB rheometer. Given the common procedures and equipment, these tests offer ideal conditions to identify any trends related to material properties with minimal variations due to testing under field conditions.  Laboratory Results and Correlations Figure 1 shows the relationship between CWR and Yield Stress as measured in the lab. Given that these samples originated from varying ponds, depths, and years, the correlation is excellent.   Figure 1 - Peak Yield Strength vs. CWR - Laboratory Data Material in the low CWR and low yield stress ranges are slurries, while the material in the higher ranges behaves as a high clay Terzaghi material. The relative correlation of this strength to the CWR through these ranges, or from slurry to soil, indicates that CWR is an important factor in predicting soil strength.  It should also be noted that the correlations provided here are for peak yield strength. Similar correlations exist for remoulded and residual strengths, but that work is not yet complete. 4 IN-SITU FIELD TESTING In comparison, the correlation with field measurements is not so clear, likely due to differences in testing equipment, procedures, and uncontrolled field conditions. Regardless, it is of interest to compare how this in-situ testing relates to the controlled lab results. Proceedings Tailings and Mine Waste 2015 Vancouver, BC, October 26 to 28, 2015 Several field datasets spanning over 20 years of operation have been collected and analysed. The field dataset contains some 6,000 in-situ vane shear tests from 1995 to present, 19 continuous ball CPT sites, and 6,173 CWR analyses.  Field samples were selected based on the same criteria as those from the Laboratory tests, with the inclusion that the samples must be correlated to in-situ yield stress testing positions. As shown in Figure 2, most of the vane shear tests were conducted by hand with boat mounted instrumentation. These instruments generally did not provide any reliable measure of the torque or shear rate applied, nor of the inclination of the instrument Given that there was a variety of vane types used, and that it is unknown if rod friction was fully accounted in the calculations, most of these historical strength measurements could not be used for this analysis. In addition, many of the recent vane shear tests were conducted to validate strength measurements from the ball CPT tests, and as such were not directly correlated to sample locations and did not have material properties associated with the test. With the need to eliminate these tests, the total remaining in-situ dataset consists of 66 individual tests, all from correlated flow penetrometer strength tests. These strength tests were correlated to vane tests, but the vane tests themselves were not used.   Figure 2 - Early vane shear strength testing from the sample boat circa 2001, using a large protractor to monitor shear angles 5 IN-SITU TEST RESULTS AND CORRELATIONS Figure 3 shows the relationship between field yield strength tests and the laboratory determined model. As shown the ball data >500Pa seemed to follow the same curve seen in the lab data. Below 500Pa the data was very scattered. The fact that the data was reasonable down to 500Pa is surprisingly good given that commercial ball penetrometers are designed to operate in much higher strength regimes with minimum accuracies required by ISO on the order of 10’s of kPa vs on the order of Pa.  Further evidence that the use of CWR is a relevant parameter is to show the same data re-plotted as a function of geotechnical water content. This is the parameter that is typically used to provide guidance on strength targets to process engineers. As shown in Figure 4 this re-plotting leads to a general scatter plot with no clear target of solids content to use to control a tailings process. On the other hand it appears that CWR can be used to provide useful process targets in ensuring a target strength can be reached given the right environmental conditions post process. Proceedings Tailings and Mine Waste 2015 Vancouver, BC, October 26 to 28, 2015 Figure 3 - Peak Yield Strength vs. CWR - All Data Proceedings Tailings and Mine Waste 2015 Vancouver, BC, October 26 to 28, 2015  Figure 4 - Peak Yield Strength vs. Geotechnical Water content - All Data Based on the analysis of the in-situ data, the following recommendations for in-pond field testing are made: 1. Equipment with downhole instrumentation and torque systems located near the vane to reduce rod friction effects are preferred 2. Appropriate and well controlled ASTM standards must be followed in order produce datasets which are comparable between techniques 3. Properly done Ball CPT test results can be used in pond survey applications to measure yield strengths down to 500Pa. 4. Laboratory equipment is preferred to determine clay behaviour for slurries, or for materials with yield strengths under 500 Pa 6 USE IN FORECASTING DEPOSIT STRENGTH GAIN  Given a high clay material with a known starting CWR, and given a known dewatering rate over time, it is now possible to predict future strengths. This is important for the tailings engineer in predicting production rates for a tailings facility. Several types of dewatering processes are known, including evaporation, freeze/thaw, and self-weight consolidation. To illustrate how this tool could be used, an example is provided here. The example operation is a thin lift deposition site, depositing untreated MFT, 100% clay, at 33.3% solids by weight. The drying season is 180 days (April 1 to October 1), with an average daily evaporation of 2.78mm/day. The table below provides initial conditions and predictions of strength with time under these conditions. Proceedings Tailings and Mine Waste 2015 Vancouver, BC, October 26 to 28, 2015 Table 1 - Example Strength Gain with Time % clay on solidsSolids SG (t/m3)Solids Content (t/m3)Water Content (t/m3)Lift Thickness (m)Surface Area for 1m3 (m2) CWRPeak Yield Strength (Pa)Starting Lift 100% 2.6 0.419 0.839 0.15 6.67 0.50 76Days Drying1 100% 2.6 0.419 0.820          0.15 6.67 0.51 842 100% 2.6 0.419 0.802          0.15 6.67 0.52 933 100% 2.6 0.419 0.783          0.15 6.67 0.54 1034 100% 2.6 0.419 0.765          0.15 6.67 0.55 1145 100% 2.6 0.419 0.746          0.15 6.67 0.56 1276 100% 2.6 0.419 0.728          0.15 6.67 0.58 1417 100% 2.6 0.419 0.709          0.15 6.67 0.59 1588 100% 2.6 0.419 0.691          0.15 6.67 0.61 1779 100% 2.6 0.419 0.672          0.15 6.67 0.62 19910 100% 2.6 0.419 0.654          0.15 6.67 0.64 22515 100% 2.6 0.419 0.561          0.15 6.67 0.75 43525 100% 2.6 0.419 0.376          0.15 6.67 1.12 2457  Assuming a deposit requires a minimum lift strength of 500Pa before a second lift can be placed, this would require a drying time of around 15 days, and provides the basis for production planning. 7 CONCLUSIONS The methylene blue index test provides a quick and easy test method that can provide a single number to represent the clay behavior of a material. In tailings engineering this Clay to water ratio can be used to help predict how a material transitions from slurry to soil and provides a useful process control tool.  8 REFERENCES ASTM Standard: C 837 – 99 (Re-approved 2003) Cerato, A. (2001). Influence of Specific Surface Area on Geotechnical Characteristics of Fine Grained Soils (Unpublished graduate dissertation). Department of Civil & Environmental Engineering, University of Massachusetts. Chiappone, A., Marello, S., Scavia, C., & Setti, M. (2004). Clay mineral characterization through the methylene blue test: comparison with other experimental techniques and applications of the method. Canadian Geotechnical Journal, 41 (6), 1168-1178. Currie, R; Bansal, S.; Khan, I.; & Mian, H. An Investigation of the Methylene Blue Titration Method for Clay Activity of Oil Sands Samples Diep, J., Weiss, M., Revington, A., Moyls, B., & Mittal, K. (2014). In-line mixing of mature fine tailings and polymers. In Jewell, R., Fourie, A., Wells, P.S., van Zyl, D. (Eds.), Proceedings of the 17th International Seminar on Paste and Thickened Tailings (pp. 111-126). Canada: InfoMine Inc. Gildey, I., & Moore, T. (2013, February). Impact of Test Methodolgy on the Atterberg Limits of Mature Fine Tailings. Paper presented at the CONRAD Oilsands Clay Conference, Edmonton AB.  Hang, P.T., & Brindley, G.W. (1970). Methylene Blue absorption by clay minerals – determination of surface areas and cation exchange capacities. Clays and Clay Minerals, 18, 203-212. Kaminsky, H.A.W (2014) Demystifying the Methylene Blue Index, Proceedings of the 4th International oil sands tailings conference, Banff AB.  Mitchell, J.K. & Sogi, K. (2005) Fundamentals of Soil Behavior, Third Edidtion. Wiley & Sons Proceedings Tailings and Mine Waste 2015 Vancouver, BC, October 26 to 28, 2015 Omotoso, O., & Melanson A. (2014). Influence of clay minerals on the storage and treatment of oil sands tailings. In Jewell, R., Fourie, A., Wells, P.S., van Zyl, D. (Eds.), Proceedings of the 17th International Seminar on Paste and Thickened Tailings (pp. 269-280). Canada: InfoMine Inc. Omotoso, O., & Morin, M. (2008). Methylene Blue Procedure: Dean Stark Solids. CanmetENERGY: Devon. OSRIN report retrieved from - http://hdl.handle.net/10402/era.40164 (http://hdl.handle.net/10402/era.40164)  Sethi, A. (1995, January 23). Methylene Blue Test for Clay Activity Determination in Fine Tails. MRRT Procedures.  


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