| Title | INVESTIGATING THE EFFECTS OF NON-ISOTHERMAL RESERVOIR CONDITION ON PRESSURE TRANSIENT ANALYSIS OF AN INJECTION/FALL-OFF TEST USING NUMERICAL MODELLING |
|---|---|
| Authors | R.J.A. Guerra, J. OSullivan |
| Year | 2018 |
| Conference | New Zealand Geothermal Workshop |
| Keywords | Pressure transient analysis, non-isothermal effects, injection/fall-off test, numerical model, TOUGH2, PyTOUGH, SAPHIRTM, permeability, skin |
| Abstract | Injection/Fall-off tests are commonly performed in geothermal wells to derive reservoir permeability and a skin factor using pressure transient analysis (PTA). Pressure transient data are commonly interpreted using analytical models which usually assume that fluid flow in the reservoir is isothermal. However, for geothermal reservoirs analysis of pressure transient data is complicated by non-isothermal reservoir conditions. The challenge in pressure transient analysis from non-isothermal reservoirs is determining the effects on the pressure response of the fluid and rock temperature-dependent properties. Using the numerical reservoir simulator TOUGH2, the non-isothermal effects during injection/fall-off tests can be simulated. Different test scenarios were designed to examine the effects of varying reservoir and injectate parameters on the derivative plot and on the analytical PTA derived reservoir permeability (k) and skin factor (S). Six test cases were used, representing different types of geothermal reservoirs, such as homogeneous reservoirs, fissured or fractured reservoirs, and layered reservoirs. Pressure transient data, generated by the numerically modelling the different cases, were analyzed using analytical PTA software SAPHIRTM. Results showed that the non-isothermal effects can influence the SAPHIRTM derived values of k and S. Non-isothermal effects on the derived k are less significant at low temperatures than at high temperatures. Non-isothermal effects on the derived S resemble a positive skin effect. The derived S is more sensitive to change at a high ΔT (difference between reservoir and injectate fluid temperature). All SAPHIRTM derived k and S values generated from different models are higher than the specified model values under non-isothermal condition. Analytical model also incorrectly introduced a reservoir boundary effect to match pressure data from some model cases. |