| Abstract |
The Silangkitang Geothermal Field (SIL) is located in the North Sumatra province, Indonesia. The field is operated by Sarulla Operation Limited (SOL), a consortium of Medco Energi Internasional Tbk., Itochu Corporation, Kyushu Electric Power Co. Inc., Ormat Technologies, Inc., and INPEX Corporation. The field was brought onto full production in March 18, 2017 at a capacity of 110MW. As part of the field development plan, a 3D Silangkitang (SIL) earth model has been continuously updated by integrating multidisciplinary (geology, geophysics, geochemistry and reservoir engineering) data and interpretations to produce an accurate geothermal conceptual model. Most resource estimates are undertaken using computer-based block modelling methods utilizing 3D block models based on a geological earth model of the resource. For these block models to be valid, it is important that they are constrained by the controlling elements of the resource. These include major structural boundaries, lithological changes within rock units, hydrothermal alteration type, compaction effect with depth, and proposed fluid pathways. Extreme heterogeneity in volcanic lithology is one of the challenges of modelling geothermal reservoir. Multiple sources of volcanic products are commonly found in the one geothermal prospect. Furthermore, in the reservoir zone, lithology can be difficult to recognize since mud circulation is normally lost and there are no rock cuttings at the surface. As it is not straightforward to accurately distribute reservoir and rock properties between wells due to the limited spatial distribution or in other word there is great uncertainty of data, the geologist then undertakes statistical and variogram analysis to give a measure of the spatial distribution of rock and resource properties for the field. An example of where this application becomes powerful in geothermal development is for distributing key reservoir parameters for flow simulation across different formations or ‘rock types’ when moving from the conceptual model to the flow simulation model. This is often achieved by giving different geological formations or rock types a bulk value for parameters such as porosity and permeability as examples. Normally this is based on analogues and/or multiple models run with different values. This paper highlights a workflow where geothermal conceptual model understanding and statistical methodology are used to populate rock properties from downhole measurements for flow simulation and its impact on the history matching process. It also gives feedback on how this process might inform concepts within the conceptual model. |