SPE PP. Abstract

Transcription

1 SPE PP EVALUATION OF THE CLEARWATER FORMATION CAPROCK FOR A PROPOSED, LOW PRESSURE, STEAM-ASSISTED GRAVITY-DRAINAGE PILOT PROJECT IN NORTHEAST ALBERTA M.M.E. Uwiera-Gartner, RPS Energy Canada Ltd., M. R. Carlson, RPS Energy Canada Ltd., and C.T.S. Palmgren, Alberta Oilsands Inc. Copyright 2011, Society of Petroleum Engineers This paper was prepared for presentation at the SPE Annual Technical Conference and Exhibition held in Denver, Colorado, USA, 30 October 2 November This paper was selected for presentation by an SPE program committee following review of information contained in an abstract submitted by the author(s). Contents of the paper have not been reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect any position of the Society of Petroleum Engineers, its officers, or members. Electronic reproduction, distribution, or storage of any part of this paper without the written consent of the Society of Petroleum Engineers is prohibited. Permission to reproduce in print is restricted to an abstract of not more than 300 words; illustrations may not be copied. The abstract must contain conspicuous acknowledgment of SPE copyright. Abstract In northeast Alberta, Canada, steam-assisted gravity-drainage (SAGD) is used to produce bitumen from the Athabasca oil sands. In some regions, bitumen resources are located within 100 m of the ground surface and have average caprock thicknesses of about 50 m. The caprock unit overlying the bitumen reservoir is comprised mainly of mudstones and shales from the Early Cretaceous, Wabiskaw Member and Clearwater Formation. These shales form a baffle, impeding vertical steam migration and confining the stress and strain fields developed during the SAGD process, which are crucial for maintaining caprock integrity at SAGD operations. In May 2006, an unexpected, catastrophic failure of the caprock seal occurred at the Joslyn Creek SAGD project, resulting in a steam release, which caused considerable caprock and ground surface deformation. Because of this failure, the Alberta Energy Resources Conservation Board modified the existing application guidelines and directives to include an assessment of caprock integrity at the maximum operating pressure specified for the thermal project applications. The authors recently completed a core based, caprock integrity assessment for a proposed, shallow, low pressure SAGD, pilot project in northeast Alberta. This paper describes the information reviewed, assessment method and results of this study. From this evaluation, information about the geomechanical characteristics, the potential for containing the induced stress and strain fields and the geologic risks associated with caprock integrity for the proposed pilot project were determined. An assessment of the underburden seal was also performed as this seal can influence caprock properties. Results of the underburden seal assessment are also presented. INTRODUCTION In northeast Alberta, Athabasca bitumen resources are relatively shallow, with about one-fifth of these shallow enough for surface mining. The remaining bitumen resources often require steam-assisted gravity-drainage (SAGD) for its development. As bitumen exploitation continues in the Athabasca oil sands, shallow bitumen resources located within about 100 m of the ground surface are being considered for development. The caprock unit for these assets is, on average, 50 m thick and consists of the Early Cretaceous, Wabiskaw Member and Clearwater Formation mudstones and shales, with silt and sand laminations and beds. Such geological conditions can provide challenges to these in situ thermal developments. For shallow SAGD projects having less caprock and overburden thicknesses when compared to their deeper counterparts, ensuring sustained caprock integrity at the designed, maximum operating pressures (MOP) is crucial. In Alberta, the interest in caprock integrity notably increased following the May 2006 catastrophic failure of the Total E&P Canada Ltd. (Total), Joslyn Creek SAGD project. A loss of caprock containment resulted in a steam release at the ground surface, significant caprock and ground surface deformation and the ultimate abandonment of the project. In light of this failure, the Alberta Energy Resources Conservation Board (ERCB) initiated a variety of response activities to address caprock integrity for

2 2 SPE PP SAGD operations in Alberta, including caprock characterization and modifications to the application process (ERCB, 2010a; ERCB, 2010b). The authors recently completed a core based, evaluation of the caprock and underburden units for the proposed Clearwater West Phase I Pilot Project (the Pilot Project ), which is to be developed using solvent-assisted, low pressure, SAGD (LP-SAGD) technologies. The Pilot Project is owned and operated by Alberta Oilsands Inc. (AOS) and located 8 km southeast of Fort McMurray, Alberta. The results of these evaluations, which are part of the ERCB s requirements for investigating caprock integrity, are published in the AOS (2010) project update. Detailed assessments of caprock core photographs and geological logs were completed by the authors. Quantitative and qualitative evaluations were performed on datasets collected from 26 fully delineated wellbores located at and proximal to the Pilot Project. The information reviewed for this evaluation included core analyses and photographs, x-ray images, wireline logs, Formation MicroImage (FMI) logs and a geotechnical borehole log. Comparisons and correlations of these datasets were also performed. Evaluations of the underburden unit were also completed. The evaluation completed for this study focused on identifying the geological and geomechanical characteristics of the caprock and underburden seals that could impede steam chamber performance and containment at the Pilot Project. Key assessment parameters in this study were geological (lithology, continuity, heterogeneity, unit thickness) and geomechanical (discontinuities, rock anisotropy, deformation features). CLEARWATER WEST PHASE 1 PILOT PROJECT The study area for this evaluation is the Clearwater West property, which is located in Sections 21 and 22, Township 88, Range 8, West of the 4 th Meridian, about 8 km southeast of Fort McMurray, Alberta (Figure 1). AOS proposes to develop this property in stages using solvent-assisted, LP-SAGD. Phase 1 of this development is the Pilot Project, which will consist of six SAGD wellpairs. Each wellpair is expected to have a nominal length of 600 m, spaced 75 m apart and will be operated at a MOP of 1,000 kpa. The wellpairs will be located about 110 m below ground surface (bgs) and within the high quality bitumen resources of the McMurray Formation reservoir. Steam injection and bitumen production are expected to commence in Year The maximum steam injection and bitumen production rates are designed at 1,113 m 3 /d (7,000 bpd) and 695 m 3 /d (4,350 bpd), respectively (AOS, 2010; Palmgren et al., 2011). R9 R8 R7W4M T89 Hwy 89 T88 Hwy 68 Figure 1 Clearwater West Project Area

3 SPE PP 3 Geological and Geomechanical Settings The Clearwater West property is located in the northeastern region of the Alberta Basin which is part of the Western Canadian Sedimentary Basin (WCSB). The WCSB is divided into two major components: a passive margin and a foreland basin, which are separated by the sub-cretaceous, erosional unconformity. The passive margin developed after the late Proterozoic rifting of the North American craton, depositing shield-derived clastics, carbonates and evaporates with intervening shales, between the Middle Cambrian and Middle Jurassic Periods. The foreland basin developed from the collision of the allocthonous terrains with the western margin of the craton between the Middle Jurassic and Tertiary Periods. The foreland basin succession generally consists of Late Jurassic to Early Cretaceous clastic, shale-dominated and sanddominated materials (Porter et al., 1982; Bachu et al., 1993). This succession is overlain by Pleistocene and recent deposits. The stratigraphy and geomechanical units (GMUs) of interest to this caprock integrity study are Devonian to recent in age and are shown in Figure 2. Figure 2 Stratigraphy and Geomechanical Units of Interest The Lower to Middle Devonian aged strata is comprised of the Lower Elk Point Group, red bed evaporite successions, which unconformably overlie the Precambrian basement. The Middle Devonian, Upper Elk Point Group is comprised of clastics, carbonates, and evaporates. The Prairie Evaporite Formation has undergone salt dissolution near the eastern edge of the Alberta Basin during the Devonian Period which can influence the characteristics of the Devonian and Cretaceous strata (Bachu et al., 1993). The Upper Devonian, Beaverhill Lake Group forms the underburden seal at the Pilot Project. This GMU generally consists of calcareous shales and carbonates overlain by the sub-cretaceous unconformity. Karsting events of the Devonian Period have been known to impact the structural characteristics of the Devonian and Cretaceous strata in northeast Alberta. The Mannville Group forms the Lower Cretaceous strata of the foreland basin, which overlies the sub-cretaceous unconformity. At the Pilot Project, the Mannville Group is comprised of the Lower Mannville, McMurray Formation, which is target reservoir. The discontinuous shale located at the base of the McMurray Formation also contributes to the underburden seal. Overlying the McMurray Formation are the marine shales of the Wabiskaw Member and Clearwater Formation, which forms the caprock unit at the Pilot Project. The Clearwater Formation is shale-dominated, consisting primarily of mudstones and shales, with sand and silt laminations and beds. Hydrogeologically, the Clearwater Formation caprock is a baffle, inhibiting fluid flow between the McMurray Formation and the overlying strata (Bachu et al., 1993). Assessment of structure and isopacs maps of the Clearwater Formation caprock at the Pilot Project indicate that salt dissolution and karsting events have not adversely impacted the geomechanical character of the caprock. The Upper Mannville, Grand Rapids Formation and Colorado Group are eroded at the Pilot Project.

4 4 SPE PP Quaternary aged sediments unconformably overlie the Clearwater Formation caprock. The pre-quaternary unconformity separates the Quaternary and Clearwater Formation strata (Andriashek, 2003). At the Pilot Project, the Quaternary sediments are 3 m to 21 m thick and consist primarily of glaciolaustrine clays and silts, which form the non-sealing, overburden GMU (AOS, 2010). JOSLYN CREEK SAGD PROJECT The Joslyn Creek SAGD project is located about 60 km north of Fort McMurray, Alberta. The McMurray Formation reservoir is located about 70 m bgs, and is confined above by 50 m to 60 m of Clearwater Formation shales (caprock) and below by a thick sequence of Devonian aged carbonates (underburden). On May 18, 2006, a catastrophic failure occurred at the Joslyn Creek SAGD project due to a loss of caprock containment. Steam was released at the ground surface after four months of wellpair circulation at well pad 204. This steam release event lasted about 5 minutes and formed a 75 m by 125 m, surface crater (Figure 3). Rocks were thrown up to 300 m away from the release point and a 1 km dust plume was created (ERCB, 2010b). Figure 3 Post-Failure Air photo of the Joslyn Creek SAGD project (from ERCB, 2010b) Post-failure analyses of the causes for caprock failure are not entirely conclusive. Total (2007) stated that a steam chamber, or chimney, developed at the top of the SAGD pay zone, forming a pressurized zone below the Upper McMurray Formation and a shear failure plane at its edges. Upward pressure migration induced by the steam chamber caused a loss of caprock containment and its catastrophic shear failure. The ERCB (2010b) agreed with the Total (2007) interpretation; however, suggested that naturally occurring, horizontal and vertical fractures within the caprock and steam migration along the cement plugs of abandoned, vertical wells may have also contributed to the steam release. An independent review of the Joslyn Creek caprock failure completed by Carlson (2010) and Carlson (2011) did not entirely support the findings of Total (2007) and ERCB (2010b). His investigations indicated that considerable time, in the order of years, were needed to develop shear failure conditions in the caprock, when the operating conditions designed for the Joslyn Creek SAGD project were applied, rather than months as observed in the field. Carlson (2010) and Carlson (2011) hypothesized that condensation induced water hammer (CIWH) may have developed in the wellbore and caused the large, repetitive, transient, pressure spikes that were recorded in the monitoring gauges prior to failure. These pressure transients, developed from CIWH were then transferred to the surrounding material. This induced tensile fracturing within the formation and caused fracture advancement and propagation, which led to the loss of caprock containment (AOS, 2010). After the caprock failure, the ERCB imposed a MOP restriction of 1,250 kpa for the project, a significant decrease from the original operating pressure of 1,800 kpa. Since that time, Total has suspended the SAGD development at Joslyn Creek. The ERCB also initiated the following response activities to address caprock integrity of SAGD developments in the Athabasca oil sands: A joint study between the ERCB and the Alberta Geologic Survey to investigate caprock integrity in Alberta (ERCB, 2010b; Hein, 2010). Modification of Directive 078 to include caprock integrity assessments at the designed MOP (ERCB, 2010a;

5 SPE PP 5 ERCB, 2010b). The rewrite of Directive 051 for addressing injection at in situ thermal operations. INFORMATION REVIEWED Wellbore data collected from the Pilot Project and proximity were reviewed with a focus placed on identifying the geological and geomechanical characteristics of the caprock and underburden GMUs. Information was collected from 22 wellbores located in Sections 21 and 22 of the Pilot Project area and four wellbores located adjacent to it. The information reviewed consisted of core analyses and photographs, geophysical well logs, FMI logs and a geotechnical borehole log. Comparisons and correlations of available datasets were also completed. Core analyses and laboratory reports typically included facies identification and photographs. Total core recovery (TCR) was determined from photographs, along with identifying geological properties, such as lithology, continuity and thickness, as well as geomechanical features, including fracturing and deformation. X-ray scans of recovered cores were also reviewed. Geophysical logs reviewed typically consisted of gamma ray, resistivity, neutron-porosity, density-porosity and calliper. Evaluations of lithology, continuity, thickness and variability were conducted. Above casing log signatures were also reviewed to determine the structure top of the Clearwater Formation caprock and overburden thickness. High resolution, FMI logs of the caprock and underburden evaluated were of good image quality. Geologic features were readily identified from resistivity contrasts. Lithology changes, lamination and bedding plane locations, conductive mineral grains, natural and induced fractures and borehole breakouts were determined from FMI logs. A geotechnical borehole log of the Quaternary and Clearwater Formation caprock was also reviewed. Lithology, discontinuity locations and descriptions, core recovery and rock quality designations (RQDs) recorded on the log were incorporated into the caprock evaluation (AOS, 2010). ASSESSMENT METHOD Both qualitative and quantitative evaluations were conducted for the caprock integrity and underburden seal assessments. Datasets were correlated and evaluated with consideration given to the geologic history of the Pilot Project area as it can impact the geomechanical character, in situ state of stress and pore pressure distribution within the caprock and underburden GMUs. Caprock integrity and underburden investigations focused on evaluating key geological properties and geomechanical features. A rock mass characterization method was developed for evaluating the geological and geomechanical information contained in the core photographs taken of the caprock and underburden. This characterization is similar to empirical rock mass classifications used in early stages of engineering design (Goodman, 1989). Moreover, the characterization method used by the authors for this study was specific to the Pilot Project area and provided a cost effective method for summarizing and evaluating data collected. The key parameters of this characterization were: Geology (lithology, heterogeneity, continuity and thickness) Core recovery Generalized RQD classification Fracture evaluation (type, frequency, spacing and infill material) Deformation and yield Considerations were given to determining whether the information available provided adequate coverage of the caprock and underburden characteristics at the Pilot Project. Information deficiencies and/or locations of inadequate coverage were determined and recommendations for additional data collection and analyses were provided where applicable. It should be noted that the authors considered that the data collected from the Pilot Project was satisfactory for conducting a thorough evaluation of the caprock and underburden GMUs at this location. Geologic Properties Evaluations of the physical character of the Clearwater Formation caprock and Beaverhill Lake Group underburden GMUs were conducted, with attention given to assessing spatial continuity and lithologic variability. All physical characteristic

6 6 SPE PP assessments were completed at the macroscopic level, evaluating colour, texture and composition. Geophysical logs were used to attain information about the stratigraphic contacts, lithology changes and thicknesses of the GMUs. Core Recovery Well information such as but not limited to, elevations and measured depths of the caprock and underburben, were tabulated. The TCR of the caprock and underburden units were determined. The TCR is defined as the percentage of the total length of core recovered to the total core run. Generalized Rock Quality Designation A generalized RQD classification was developed for the Pilot Project for assessing the discontinuity information contained in core photographs of the Clearwater Formation caprock. This generalized RQD is similar to that developed during the 1960s (Goodman, 1989), which is calculated by summing the measured lengths of intact drill core exceeding 0.1 m and dividing it by the total length of the drill core for a given rock type. The classification categories established by the authors were modified from the 1960s categories due to the likely overstating of fractures resulting from core handling and storage. The generalized RQD categories relates to how intact the core samples are. The categories specified for the Pilot Project were defined as Good (> 75% RQD), Fair (50-75% RQD) and Poor (< 50% RQD). Figure 4 provides examples of the generalized RDQ categories observed in core photographs taken from the caprock and underburden at the Pilot Project. Fracture Characterization Figure 4 Core Photographs of the Generalized RQD Categories Characterization of fractures in the Clearwater Formation caprock and Beaverhill Lake Group underburden were completed using available data. Geophysical and FMI logs were correlated to core photographs and the geotechnical borehole log, where applicable, as part of the characterization process. Fracture information from the geotechnical borehole log was assessed relative to a vertical core axis as magnetic orientations of the discontinuities were not recorded. Key parameters for assessing the fracture conditions in the caprock and underburden were fracture type, fracture spacing index (FSI), frequency, aperture and secondary infill materials. Considerations were given to whether the fractures were natural or induced during the characterization process. Difficulties with distinguishing between natural and induced fractures in core are well known and can provide difficulties in characterizing the caprock and underburden rock masses. The main types of induced fractures observed from the Pilot Project core were likely formed during core removal, sample freezing and moisture loss. Fracture type assessments were performed to determine if the discontinuities were structural, contractional or a combination of both. Structural fractures are generally caused by changes in the stress fields while contractional fractures are formed by a reduction in the original bulk volume (Kulander et al., 1990). Figures 5 and 6 provide examples of structural and contractional fractures observed in core samples from the Pilot Project.

7 SPE PP 7 Figure 5 Structural Discontinuities in the Caprock Figure 6 Contractional Discontinuities in the Underburden Fracture spacings observed in core photographs and recorded on the geotechnical borehole log were categorized as Closely (< 15 cm spacing), Moderate (15-60 cm spacing), and Blocky (> 60 cm spacing) for the Pilot Project study (Figure 7). Core photographs from the Clearwater Formation caprock demonstrated considerable fracturing, which are attributed to core handling methods. Because of this, a significant length of the caprock core fell in the Closely category, which is not likely representative of the entire caprock unit at the Pilot Project. The FSI, which is the number of fractures per meter, were also calculated for the caprock. Figure 7 Discontinuity Spacing Categories The occurrences infill materials along fractured surfaces observed in core photographs were categorized as None (0% occurrence), Seldom (< 10% occurrence), Occasional (10-50% occurrence) and Frequent (> 50% occurrence). The lithology of these infill materials were identified from core photographs where possible. Figures 8 and 9 illustrate examples of infill materials observed and its occurrences in core photographs.

8 8 SPE PP Figure 8 No Occurrences of Secondary Infill Material in Caprock Discontinuities Figure 9 Frequent Occurrences of Infill Materials in Underburden Discontinuities Deformation and Yield Deformation features recorded in core photographs and geological logs were identified. The deformation features of interest to this study were slickenside surfaces, soft sediment deformations and borehole breakouts. CLEARWATER WEST PHASE 1 PILOT PROJECT STUDY RESULTS Caprock Assessment Comprehensive, geological and geomechanical assessments of the caprock characteristics at the Pilot Project were completed. Results indicate that the Clearwater Formation caprock unit is continuous with thicknesses ranging from 46 m to 61 m and averages 51 m. Core photographs demonstrate that the caprock is comprised of grey mudstones and shales which are corroborated by gamma-ray signatures, FMI logs and geotechnical log information. FMI logs show that silt and sand laminations and beds are present in the Clearwater Formation caprock (Figure 10). Geotechnical logging describes the caprock as being heterogeneous, consisting of sandstone and siltstone beds within the larger, shale package. Coal, pyrite and shell fragments were observed during geotechnical coring. Figure 10 FMI Log of the Clearwater Formation Caprock

9 SPE PP 9 Figure 11 provides an example of the correlated datasets from a wellbore at the Pilot Project. As shown in this figure, lithologic variability exists in caprock unit; however the main rock type is shale. Clearwater 21m Casing 51 m Core Photos Wabiskaw 67.0 m McMurray 71.8 m Reservoir seal Caprock Reservoir seal Underburden Quaternary Overburden (not sealing) Geotechnical Log Caprock Feature Underburden Discontinuity Bottom Shale m Devonian m Figure 11 Correlation of Available Data Sources The assessment completed by the authors indicated that the data available adequately captured the lateral and vertical characteristics of the Clearwater Formation caprock at the Pilot Project. The total length of caprock core collected from the Pilot Project was 155 m, based on core photograph data. In addition, nearly 50 m of core was recovered and logged during geotechnical drilling, providing a total core return exceeding 200 m in length that was collected from the Pilot Project. This length was notably less than the cumulative length of wireline signatures and FMI logs taken from the caprock unit at the Pilot Project, further supporting the author s opinion of adequate data coverage. Table 1 summarizes the core recovery results determined from core photographs and geotechnical log of the Clearwater Formation caprock. Core lengths obtained on a per wellbore basis ranged between about 1 m and 24 m and averaged about 11 m, with most of the core obtained from the Wabiskaw Member and lower regions of the Clearwater Formation. The TCR from core photographs varied between 73% and 100% and averaged about 91%, which was consistent with values recorded on the geotechnical log. Table 1 Core Recovery Summary from the Clearwater Formation Caprock Source Parameter Maximum Minimum Average Core Length (m) Core Photographs TCR (%) RQD (%) Spacing* (%) Core Length (m) 47.7 Note: Geotechnical Borehole Log TCR (%) RQD (%) FSI (count/m) * - Percentage of core photographs exceeding the closely spacing category. Discontinuity assessments from core photographs of the caprock show some broken-up and fissile regions. Most discontinuities were sub-horizontal and followed bedding and lamination planes. The generalized RQDs determined from

10 10 SPE PP core photographs were under 37% and averaged 13%, notably lower than those recorded during geotechnical drilling (67% average, Table 1). Fracture spacing assessment of the caprock core photographs suggested that spacings were mainly in the Closely category and to a lesser extent, in the Moderate category. Furthermore, only 24% of the drill core lengths evaluated from the caprock reported fracture spacings exceeded the Closely category (Table 1), corresponding to a FSI of about 5/m. The majority of the fractures identified were structural, having sub-horizontal orientations that followed lamination planes (Figures 2 and 5). Infill materials were not observed along the discontinuity surfaces captured in the core photographs. Fracture spacings from geotechnical logging indicated a considerably lower fracture frequency than those recorded in core photographs. The geotechnical log indicated that for every 5 m of caprock core recovered, 1 to 9 fractures were present, hence corresponding to FSIs of 0.2/m to 1.8/m, respectively and averaged about 1.2/m (Figure 12) Frequency ( ) Depth Inveral (m) Figure 12 Discontinuity Frequency from the a Geotechnical Borehole Log The predominate fracture orientation recorded during geotechnical logging was sub-horizontal, ranging between 75 and 90 relative to the vertical core axis. A secondary, sub-vertical orientation was also observed with fracture angles varying from 0 and 30 relative to the vertical core axis (Figure 13). 20 Frequency ( ) Orientation to Core Axis (degress) Figure 13 Frequency of Discontinuity Orientations Relative to the Vertical Core Axis

11 SPE PP 11 Comparisons of the generalized RDQs and fracture spacings recorded in core photographs and the geotechnical log indicates that the information portrayed in core photographs likely overstated the fracture conditions in the caprock. The authors attributed this to induced fractures caused by core handling. It should be noted that FMI logs indicated even fewer fracture occurrences, when compared to core photographs and geotechnical logging, thereby further corroborating the occurrences of induced fracturing in core photographs. Even though the core photographs and geotechnical logging likely overstated in situ conditions, the sub-horizontal and subvertical discontinuity orientations were likely representative. Such orientations, especially sub-vertical, can create some risk to caprock integrity and impede its ability to contain the steam chamber developed during the LP-SAGD operation. Based on the data reviewed, it is unlikely that these fracture sets formed a pervasive fracture system through the caprock. Furthermore, sub-horizontal fractures can provide a horizontal direction for stress and temperature dissipations, which in turn may assist with caprock integrity. Soft sediment deformations were identified on two of the five FMI logs of the caprock from the Pilot Project. These deformation features generally coincide with the Clearwater Formation-Wabiskaw Member stratigraphic contact and its associated lithologic change (Figure 10). Borehole breakouts were recorded in one FMI log taken of the caprock. These breakouts were relatively small and not attributed to an anisotropic state of stress and rock strength, but rather to the anomalous drilling and completion history of that particular wellbore. In this instance, the wellbore was drilled and cored to the base of the caprock, prior to setting in casing a few days later at a shallower depth than the base of the cored interval. After this time the wellbore was reamed to the base of the caprock and then subsequently drilled and cored to its final depth within the underburden unit. Borehole breakouts were recorded on the FMI log coinciding with the reamed interval. Underburden Assessment A geological assessment of the underburden seal was performed since dissolution and karsting events are known to impact the characteristics of the Devonian and Cretaceous strata in northeast Alberta. Evaluations of the underburden at the Pilot Project indicated that these events had little influence on the characteristics of the caprock at this location. At the Pilot Project, the underburden consists of up to 5 m of discontinuous, bottom shales from the McMurray Formation and also a thick sequence of intact, fractured and/or argillaceous limestone from the Beaverhill Lake Group. Core returns obtained from the underburden typically ranged from 1 m to 16 m and averaged 8 m per wellbore, where collected. The total length of core obtained from the underburden at the Pilot Project was nearly 215 m. The TCR ranged between 77% and 100% and averaged 97%. Structural and contractional fractures were observed in core photographs and FMI logs. Occasional to frequent occurrences of secondary infill materials, likely consisting of clay and/or carbonate, were observed on photographs. Such fractures in the underburden can impact fluid losses at the Pilot Project. Lost circulation instances were reported by AOS while drilling in the Devonian carbonates, thus further suggesting fracture permeability and fluid loss potential. GEOMECHANICAL IMPLICATIONS OF CHARACTERIZATION Evaluations of the geological and geomechanical properties of the caprock and underburden seals are necessary as these strongly correlate to the thermal, hydraulic and mechanical properties of these GMUs. A caprock seal for a SAGD project must have adequate strength and yield characteristics to withstand the loads and deformations induced in this GMU during life of the SAGD project. Ideally, the caprock and underburden seals have the following characteristics: Adequate continuity and thickness. Minimal discontinuities. Low vertical transmissibility. Strength and deformation properties which are resistive to shear failure. Limit heat and fluid losses. Stress and temperature loadings developed in the reservoir during SAGD operations typically cause thermal expansion, increased pore pressure and promote rock dilation within the reservoir, which may cause yielding and deformation in the caprock and underburden. Existing sub-horizontal fractures within the caprock are not necessarily be detrimental to a SAGD project and can assist with dissipating the generated stresses. Problems occur when these fractures form connective systems and preferential pathways for stress migration and steam chamber losses (AOS, 2010; McLellan et al., 2000).

12 12 SPE PP Geological and geomechanical characterization of the caprock and underburden always includes some element of risk which can be difficult to quantify. Uncertainties exist between the data available and real world system. Exact representations of the caprock and underburden physical properties are not entirely possible and hence, assumptions about these for a given study area are necessary. Factors contributing to these uncertainties are often caused by information deficiencies and the inherent variability in physical properties. Key uncertainties associated with the core assessment of the caprock and underburden seals at the Pilot Project were typically related to geological heterogeneity and geomechanical features. For the caprock unit, heterogeneity, anisotropy and lamination and bedding planes can impact permeability and pressure build-ups from the SAGD operation. Another key uncertainty is the existence of large fractures that were not identified during the data review, which can form localized regions of increased permeability and reduced rock strength, thereby impeding caprock containment. Evaluations of the Quaternary overburden are also necessary as it can impact caprock integrity. Glacial meltwater channels insized into the caprock can reduce its continuity and thickness, influence local hydrodynamics and form stress relief areas. At the Pilot Project, meltwater channels were not identified; however, these are known to exist at other locations in northeast Alberta (Andrianshek, 2003). Geological events of the Devonian Period have impacted the characteristics of the Clearwater Formation caprock in northeast Alberta. Though these affects were not readily observed in the caprock at the Pilot Project, deformations within the underburden are known to induce caprock fracturing and yielding at other locations (ERCB, 2010b; Hein, 2010). SUMMARY AND CONCLUSIONS The authors recently completed an evaluation of caprock integrity and underburden seal characteristics at a proposed LP- SAGD Pilot Project in northeast Alberta using core photographs and geological logging data collected from the study area. This evaluation is similar to those applied in geotechnical investigations and can be readily applied to other in situ thermal developments of the oil sands industry. Geological and geomechanical characteristics of the caprock and underburden seals were assessed both qualitatively and quantitatively using the data collected from 26 wellbores advanced at and proximal to the Pilot Project. The authors determined that the information available provided adequate coverage to complete a thorough evaluation of the Clearwater Formation caprock and Beaverhill Lake Group underburden at the Pilot Project. Having reasonable data coverage for such investigations is critical for evaluating caprock integrity at SAGD operations. The results of this evaluation demonstrated that the average thickness of the Clearwater Formation caprock at the Pilot Project is 51 m. The caprock unit is mainly comprised of a mudstone and shale unit with interbedded silts and sands. Generalized RQD and fracture assessments determined from core photographs likely overstated in situ conditions, when compared to those determined from geotechnical and FMI logging. The average FSI determined from core photographs of the caprock was about 5/m, notably higher than the corresponding average of 1.2/m determined from a geotechnical borehole log. Even though core photographs overstate in situ fracture spacing, these along with geotechnical and FMI log information demonstrated that the main fracture orientation was sub-horizontal and followed lamination and bedding planes. The induced fractures observed in core photographs of the caprock were attributed to core handling methods. The authors did not identify a pervasive fracture system in the caprock at the Pilot Project using the available information. Such fracture systems could compromise caprock integrity. Evaluation of the underburden seal suggested that dissolution and karsting events had little effect on the characteristics of the caprock at the Pilot Project. Characterization of the geological and geomechanical properties of the caprock are necessary for SAGD developments as these influences its thermal, hydraulic and mechanical properties and its response to the stress and strain loads induced by the in situ thermal operation. Uncertainty always exists between the data evaluated and the real world system; however having adequate data coverage and conducting a thorough investigation of the information available can reduce these uncertainties and inherent risks associated with evaluating caprock integrity.

13 SPE PP 13 ACKNOWLEDGEMENTS The authors of this paper would like to thank Alberta Oilsands Inc. and RPS Energy Canada Ltd. for permission to publish these findings. The information referenced in this paper was obtained primarily from the technical work documented by AOS (2010). The authors would like to thank Ms. Karen Lausen of RPS Energy Canada Ltd. for her assistance with document formatting and reviewing. NOMENCLATURE AOS = Alberta Oilsands Inc. bgs = below ground surface CIWH = Condensation Induced Water Hammer ERCB = Energy Resource Conservation Board FMI = Formation MicroImage FSI = Fracture Spacing Index GMU = Geomechanical Unit MOP = Maximum Operating Pressure RQD = Rock Quality Designation SAGD = Steam-Assisted Gravity-Drainage LP = Low Pressure TCR = Total Core Recovery TOTAL = Total E&P Canada Ltd. WCSB = Western Canadian Sedimentary Basin REFERENCES Alberta Energy Resources Conservation Board (ERCB), 2010a. Directive 078: Regulatory Application Process for Modifications to Commercial In Situ Oil Sands Projects Issued. Directive Approved December 3, Alberta Energy Resources Conservation Board (ERCB), 2010b. Total E&P Canada Ltd. Surface Steam Release of May 18, 2006 Joslyn Creek SAGD Thermal Operation. ERCB Staff Review and Analysis. Document submitted February 11, Alberta Oilsands Inc. (AOS), Project Update for Clearwater West LP-SAGD Pilot Project, December 22, Project update prepared for Alberta Energy Resources Conservation Board (ERCB). Website: Andrianshek, L.D., Quaternary geologic setting of the Athabasca oil sands (in situ) area, Northeast Alberta. EUB/AGS Earth Sciences Report Bachu, S., Underschultz, J.R., Hitchon, B., and Cotterill, D., Regional-scale subsurface hydrogeology in Northeast Alberta. Bulletin No. 61., Alberta Geological Survey, Alberta Research Council. Carlson, M.R., SPE, And now for something completely different: Condensation Induced Water Hammer and Steam-Assisted Gravity Drainage in the Athabasca Oil Sands. 14 th International Topical Meeting on Nuclear Reactor Thermalhydraulics, NURETH- 14. September 25-30, 2011, Toronto, Ontario, Canada. Paper in press. Carlson, M.R., 2010, SPE. What every SAGD engineer should know about condensation induced water hammer. Presentation at the SPE Technical Luncheon, Calgary Section. Website: Collins, P.M., SPE, Geomechanical effects of the SAGD process. SPE Reservoir Evaluations and Engineering Paper pgs Hein, F., Cap Rock Integrity and Geologic Risk Assessment of the Athabasca Oilsands and Development Strategies. Presentation at the Canadian Heavy Oil Association (CHOA) Technical Lunch Series. Website: Kulander, B.R., Dean, S.L., and Ward Hr., B.J., Fractured Core Analysis Interpretation, Logging, and Use of Natural and Induced Fractures in Core. AAPG Methods in Exploration Series, No.8. Tulsa, 88 pgs. McLellan, P., SPE, Read, R., and Gillen, K., Assessing Cap Rock Integrity for Steam-Assisted Gravity-Drainage Projects in Heavy-Oil Reservoirs (GEM Presentation). SPE/PS-CIM International Conference on Horizontal Well Technology. Calgary, Alberta, Canada. November SPE Paper

14 14 SPE PP Palmgren, C.T.S., SPE, and Walker, I., SPE, Alberta Oilsands Inc., Carlson, M., SPE, and Uwiera, M., SPE, RPS Energy Canada Ltd., and Torlak, M., SIEMENS, Reservoir design of a shallow LP-SAGD project for in situ extraction of Athabasca Bitumen. World Heavy Oil Congress, WHOC11-520, Edmonton, Alberta, Canada. Paper in press. Total E&P Canada Ltd. (Total), Summary of Investigations into the Joslyn May 18 th 2006 Steam Release. Issued December 2007.