A Risk-based Groundwater Modelling Study for Predicting Thermal Plume Migration from SAGD Well-pads Rudy Maji, Ph.D., Golder Associates Solaleh Khezri, M.Sc., AB Scientific Intern (Golder Associates) Don Haley, M.Sc., Golder Associates Michael De Luca, M.Sc., P.Geol., Brion Energy
Outline Motivation Problem Statement Numerical Model Construction Numerical Model Results Summary and Conclusions April 30, 2015 2
Motivation Why Thermal Plume Migration Matters Elevated temperatures increase the mobilization of chemical constituents that are naturally present in sediments. Since the start of thermal in-situ oil sand production, increased levels of Arsenic was observed in groundwater downgradient of several steam injection wells at the Cold Lake site. April 30, 2015 3
Problem Statement Steam Assisted Gravity Drainage (SAGD) Area of Interest April 30, 2015 4
Problem Statement Conceptual Model of Heat Plumes Near Wellbores Aquitard Groundwater Flow Aquifer Aquitard Aquifer Aquitard April 30, 2015 5
Problem Statement Modelling as a Screening Tool Generic modeling can be used as a screening tool as one input into a risk based assessment of solute migration from multiple SAGD well pads. If the risk at a particular well pad is considered potentially significant, site specific models (deterministic or stochastic) can be developed to help evaluate and quantify the risk of thermally enhanced solute migration. Site specific model could also be used to aid in the design of the Groundwater Monitoring Plan (GWMP), as directed in the DRAFT Guidance for Groundwater Management Plans for In Situ Operations: Assessing Thermally-Mobilized Constituents April 30, 2015 6
Numerical Modelling SAGD Well-Pads in MacKay River Commercial Project Area Streams (SW Receptors) Well-Pad AA Well-Pad AJ Well-Pad AB April 30, 2015 7
Numerical Modelling Borehole Lithology Around Well-Pad AJ Model Layer Formation Horizontal Hydraulic Conductivity (m/s) Thickness (m) Layer 1 Undifferentiated overburden (Clay Silt Till) 1E-7 10 Layer 2 Undifferentiated overburden (Clay Silt Till) 1E-7 10 Layer 3 Undifferentiated overburden (Sand Till) 5E-5 5 Layer 4 Undifferentiated overburden (Clay Silt Till) 1E-7 10 Layer 5 Joli Fou Formation (Shale) 5E-8 10 Layer 6 Grand Rapids 4 Formation (Sandstone) 3E-4* 10 Layer 7 Grand Rapids Formation (Shale) 5E-8 5 Layer 8 Grand Rapids 5 Formation (Sandstone) 1.6E-5* 5 Layer 9 Grand Rapids Formation (Shale) 5E-8 5 Layer 10 Layer 11 Grand Rapids 5 Formation (Sandstone) 1.6E-5* 10 Grand Rapids 5 Formation (Sandstone) 1.6E-5* 10 * Hydraulic conductivity values derived from pumping tests. April 30, 2015 8
Sources Pad AJ, Pad AA and Pad AB Numerical Modelling Sources and Receptors Receptors Surface water streams, Aquifers (Overburden Aquifer, Grand Rapids 4 and 5) April 30, 2015 9
Numerical Model Construction Model Domain Well-Pad AA Well-Pad AJ Wells with Geology Logs SAGD Steam Injection Wells Well-Pad AB Model Domain April 30, 2015 10
Numerical Model Construction Numerical Mesh Well-Pad AA Element Size Around SAGD Wells: <1 m to 9 m Well-Pad AB Well-Pad AJ April 30, 2015 11
Numerical Model Construction 3-Dimensional Numerical Block Model Streams 11 Numerical Layers Each Layer 5 m to 10 m thick 9 Different Hydrostratigraphic Units April 30, 2015 12
Numerical Model Construction Boundary Conditions Y Average Hydraulic Gradient: 0.25% X Constant Temperature of 5 C at Surface and Inflow Nodes April 30, 2015 13
Numerical Model Construction Heat Loss in the Steam Injection Well Versus Depth Steam Temp= 230 C at surface Steam Temp= 226 C at bottom of Grand Rapids 5 Steam Temp= 220 C at top of McMurray April 30, 2015 14
Numerical Model Construction Representative Steam Injector Design April 30, 2015 15
Numerical Model Construction Representation of a SAGD Well Using the BHE Boundary Recovery Injection Borehole Heat Exchanger April 30, 2015 16
Numerical Model Results Calibration to Temperature Change Along the Wellbore Recovery Injection Recovery Injection April 30, 2015 17
Numerical Model Results Modelling Cases Case 1: Steaming of well-pad AJ for 38 years Case 2: Steaming of all three well-pads for 38 years simultaneously and movement of thermal plume after cession of steaming April 30, 2015 18
Numerical Model Results Case 1 (Single Well Pad): Thermal Plumes at Well-Pad AJ 363 m Grand Rapids 4 Depth: 45 m Max Temp: 104 C K H = 3E-4 m/s 380 m Quaternary Aquifer Depth: 20 m Max Temp: 145.7 C K H = 5E-5 m/s 418 m Grand Rapids 5 Depth: 80 m Max Temp: 153 C K H = 1.6E-5 m/s Note: After 38 years of steaming April 30, 2015 19
Numerical Model Results Case 2 (Cumulative Effects): Thermal Plumes in Quaternary Aquifer - All 3 Well-Pads 366 m K H = 5E-5 m/s Depth: 20 m 363 m 367 m Note: After 38 years of steaming April 30, 2015 20
Numerical Model Results Case 2 (Cumulative Effects): Thermal Plumes in Grand Rapids 4 Aquifer 422 m K H = 3E-4 m/s Depth: 45 m 418 m 420 m Note: After 38 years of steaming April 30, 2015 21
Numerical Model Results Case 2 (Cumulative Effects): Thermal Plumes in Grand Rapids 5 Aquifer 386 m K H = 1.6E-5 m/s Depth: 80 m 380 m 382 m Note: After 38 years of steaming April 30, 2015 22
Well Pad AJ Well Pad AB Numerical Model Results Case 2 (Cumulative Effects): Vertical Profiles Through Each Well Pad K=1E-7 m/s m/s K=5E-5 K=5E-5 m/s m/s K=1E-7 K=1E-7 m/s m/s K=5E-8 K=5E-8 m/s m/s K=3E-4 K=3E-4 m/s m/s K=5E-8 K=5E-8 m/s m/s K=1.6E-5 m/s m/s A A A AB A B Well Pad AA CA 380 m A C Note: After 38 years of steaming April 30, 2015 23 A A
Numerical Model Results Case 2 (Cumulative Effects): Length of 50 C Isotherm From Steam Injector Depth (m) Length of 50 C Isotherm (m) Hydraulic Conductivity (m/s) 1 1.3 1E-7 10 4.6 m 1E-7 20 8.3 5E-5 Intermediate Aquifer K 25 8.6 1E-7 35 6.9 5E-8 45 5.2 3E-4 Highest Aquifer K 55 5.9 5E-8 60 7.5 1.6E-5 65 9.9 5E-8 Lowest Aquifer K 70 13.4 1.6E-5 80 20 1.6E-5 90 22 1.6 E-5 Note: After 38 years of steaming April 30, 2015 24
Numerical Model Results Case 2 (Cumulative Effects): Temperature Increase in Underlying Aquifers Well-Pad AA Grand Rapids 4 Well-Pad AA Well-Pad AB Well-Pad AB Grand Rapids 5 Grand Rapids 4 Quaternary Grand Rapids 5 Quaternary Note: Location is immediately downstream of the injector and below the streams April 30, 2015 25
Numerical Model Results Case 2 (Cumulative Effects): Thermal Plume Migration after Cessation of Steaming Time= 0 Max T= 152.5 C Time= 10 years Max T= 34.5 C Time= 25 years Max T= 22 C 380 m 520 m 450 m Time= 50 years Max T= 13.7 C Time= 100 years Max T= 10 C 605 m 600 m 210 m April 30, 2015 26 NOTE: Results are for Well Pad AJ
Summary and Conclusions The distance from Pad AJ to the nearby stream is larger compared to that of Pads AA and AB; hence, the simulated thermal plume from Pad AJ doesn t reach to the nearby streams. The simulation results suggest that thermal plumes originating from the Pads AA and AB would intersect the nearby streams, hence pose a greater risk compared to the Pad AJ if solute migration is thermally enhanced. The 50 0 C temperature plume travels only a few tens of metres; hence, the potential zone of Arsenic mobilization is simulated to be within a few tens of metres, provided the enhanced mobilization effect decreases as temperature drops. April 30, 2015 27
Summary and Conclusions The temperature of the porous medium in the immediate vicinity of the steam injection well is significantly less (approximately 150 0 C) than that of the steam inside the well bore (approximately 230 0 C). The change in temperature in the porous medium around a steam injection well is affected by: The insulating properties of the well bore casing and grout system Formation hydraulic conductivity and thereby groundwater velocity (efficiency of the groundwater to flush heat away from the well bore) April 30, 2015 28