CARBONATE RESERVOIRS….. Simeon Agada, Fuzhen Chen, Sebastian Geiger Institute of Petroleum Engineering, Heriot-Watt University, U.K. Email: [email protected] Numerical simulation of IOR AND EOR processes in a carbonate reservoir outcrop analogue DYNAMIC MODELLING USING 3D HIGH-RESOLUTION DIGITAL OUTCROP MODELS PURPOSE ANALYSING THE IMPACT OF HETEROGENEITY ON DIFFERENT RECOVERY MECHANISMS CONCLUSIONS: 1. Outcrop analogues are used as an economic means to decipher fundamental flow attributes during IOR and EOR in carbonates. This is achieved by relating transient pressure response to visible geological heterogeneity. 2. Secondary recovery (waterflood and gasflood) and tertiary recovery (miscible flooding) simulations show that fluid properties and engineering parameters can be varied to optimise recovery during IOR and EOR. 1 10 100 1000 0.1 1 10 100 Pressure (psi) Time (hr) Zero Trans (P) Zero Trans (dp/dt) 0.25 Trans (P) 0.25 Trans (dp/dt) 1.0 Trans (P) 1.0 Trans (dp/dt) 5.0 Trans (P) 5.0 Trans (dp/dt) 1 10 100 1000 0.1 1 10 100 Pressure (psi) Time (hr) WT1 (Press) WT5 (Press) WT7 (Press) WT1 (dp/dt) WT5 (dp/dt) WT7 (dp/dt) 1 10 100 1000 0.1 1 10 100 Pressure (psi) Time (hr) WT1 (Press) WT2 (Press) WT6 (Press) WT1 (dp/dt) WT2 (dp/dt) WT6 (dp/dt) 1 10 100 1000 0.1 1 10 100 Pressure (psi) Time (hr) Base_WT1 (Press) Base_WT1 (dp/dt) Bioherms_WT1 (Press) Bioherms_WT1 (dp/dt) 1 10 100 1000 0.1 1 10 100 Pressure (psi) Time (hr) Base_WT4 (Press) Base_WT4 (dp/dt) Bioherms_WT4 (Press) Bioherms_WT4 (dp/dt) 1 10 100 1000 0.1 1 10 100 Pressure (psi) Time (hr) Base (Press) Base (dp/dt) Base + Fractures (Press) Base + Fractures (dp/dt) (a) (f) (e) (d) (c) (b) Flow simulation and pressure transient analysis workflow Upscaling Flow Simulation Fracture Modelling Outcrop Analogue Geological Model Numerical Well Testing A systematic process is used to adequately capture geological heterogeneities and represent them in high-resolution geological and flow simulation models. The workflow we employ enables us to identify unique signatures of geological features by numerical well testing and to optimise hydrocarbon recovery in carbonates by secondary (water and gas injection) and tertiary (miscible flooding) methods. (a) Oil recovery factor and (b) oil production rate for secondary and tertiary oil recovery simulations. Impact of (c) petrophysical rock types and (d) hysteresis on ultimate recovery from WAG. f f Upscaling is required for quick turn around time in numerous optimisation simulations. It is achieved by a vertical layer optimisation procedure and validated against the fine scale geological model using streamline simulations (full field) and finite difference simulations (sector). - 350 -0 - 200 - 100 - 300 Dip azimuth (deg) (a) (b) (c) CAD Rhino Representation of Outcrop Window ANSYS Mesh of Outcrop Window after Extrusion to 3D The influence of faults is seen (below) by the slope of the pressure transient at late time and always dominates transients (a & e). The symmetry/asymmetry of a well’s drainage area is discernible from the pressure transient derivative “hump” (b). High permeability oyster bioherms in varied reservoir locations are also identifiable from transients (c & d). The classical textbook transients for a fractured reservoir are not visible possibly due to severe heterogeneities (f). a) Regular 5-spot waterflood (favourable mobility) b) Inverted 5-spot waterflood (favourable mobility) c) Direct line drive d) Staggered line drive e) Regular 5-spot gasflood (favourable mobility) f) Regular 5-spot waterflood (unfavourable mobility) g) Tornado chart showing impact of key engineering parameters (a) (f) (c) (d) (b) (g) Digital outcrop model showing positions of synthetic numerical test wells Digital outcrop model showing high permeability mollusc banks Drawdown pressure transients enhance identification of geological features Oyster Bioherms Base Geological Realisation Model with oyster bioherms Outcrop Analogue The matrix model is coupled with fractures using a Discrete Fracture Network (DFN) model or Discrete Fracture and Matrix (DFM) model that incorporates input from RHINO & ANSYS. DFNs are generated for 0.1, 0.5 & 0.9 fracture intensity to investigate fracture impact on flow dynamics. The DFNs are then combined with dual porosity-dual permeability models in IMEX/GEM for fracture flow simulations. In the numerical well test analysis, synthetic wells are placed at different locations in the outcrop model to test the impact of small- and large scale heterogeneities including mud mounds, oyster bioherms and fractures. The theoretical pressure responses (e.g. linear flow and recharge) that are derived for idealised geological structures can be validated by comparing them to pressure responses obtained for realistic geological structures. This procedure can also be employed to improve the calibration of static and dynamic reservoir models with production. The secondary recovery simulations show the potential of varied fluid and engineering parameters to mitigate flow challenges due to heterogeneity. Final flow simulation model • Intermediate oil-wet reservoir • Single and multiple P c & K r curve • Secondary recovery simulations with IMEX • Miscible Flood simulations with GEM DFN with 0.1 fracture intensity (P32) DFN with 0.5 fracture intensity (P32) DFN with 0.9 fracture intensity (P32) Injector Producer 0.00 Injector Injector-CO2 Producer 0.00 s 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00 Injector Producer (b) (a) S o Water Injection Continuous Gas Injection (Miscible) Water Alternating Gas Injection (Miscible) • Carbonate reservoirs are highly heterogeneous across all length scales, rendering it difficult to predict flow behavior in the subsurface. • The multi-scale heterogeneities result from complex depositional, reactive and diagenetic processes. • An accurate characterization of carbonate reservoir architecture is therefore critical for successful modelling and multiphase fluid flow prediction. • The detailed reservoir architecture is commonly studied using outcrop analogues, but these are rarely used for flow modelling. We use a high-resolution 3D outcrop model of a Jurassic Carbonate ramp in order to perform a series of detailed and systematic flow simulations. The aim is to test the impact of small- and large-scale geological features on reservoir performance and oil recovery during IOR and EOR. We also investigate the effect of dynamic matrix-fracture interaction and phase miscibility on the flow simulations. (a) (b) (c) (d) Research kindly supported by the ExxonMobil (FC) 2 Alliance