| Abstract |
The single porosity numerical reservoir model of the San Jacinto geothermal system, Nicaragua has been created and developed over the past 3-4 years. Recent reservoir tracer studies have shown a connection between an injection well and the main production area. In order to assess the significance of this link in terms of reservoir management, a quick and effective numerical model update was required. When considering the impact of injection returns in a reservoir, the associated thermal impact of injection returns is of critical importance. The matching of temperature declines by injection returns in a single porosity reservoir can be limited due to the relatively large volumes of the model blocks and heat diffusion from reservoir block surfaces. Travel times and return volumes are often very difficult to match. Dual porosity modelling introduces the ability to model fracture flows, and as such provides a mechanism to achieve closer matching of injection returns. It is widely accepted that the conversion of single porosity numerical models to dual porosity models is a difficult and highly time consuming process due to the increased computational requirements, the introduction of additional matching parameters such as fracture density and spacing, and the inherent difficulty in acquiring field measurements for these parameters. In order to achieve the desired level of calibration in an appropriate timeframe, we considered an alternative approach to matching the reservoir tracer injection returns. Given that the connection between the injection well and production area had been interpreted to be structural with linked feed zones, a distinct fractured zone or channel between the injection well and production area was selected to represent the fault structure. Within this fault structure, the volume of each block was reduced and assigned a high porosity. Anisotropic permeability calibration was also performed to attain a good match to injection returns. This alternative approach allowed an accurate match of injection returns in multiple wells and a close match for the travel times overall. Production well enthalpy and pressure responses were also matched during the tracer calibration. A number of injection scenarios were investigated to assess the impact of the injection returns. The results highlighted that there was sufficient heating of the injection returns between injection and production areas at the current production and injection rates, and that the injection returns should not be a concern at present. Long term limits for the main injection well linked to the reservoir were suggested in order to minimise impact, but there was significant flexibility for short term increases in injection load if necessary. The quick reservoir model calibration, high level of accuracy in the matching of field data and sensitivity of the reservoir to injection have provided valuable information for the management of the San Jacinto reservoir in a highly cost effective manner. It has provided confidence that, while field data which initially suggested a potential issue to production, the current reservoir production and injection strategy remains appropriate and sustainable. |