| Abstract |
The 3D Geologic Model is a tool for better understanding the subsurface geology and is an essential component of numerical reservoir modeling. The first Salak 3D Earth Model was built in 2005. Since then, one of the main uses of the 3D Geologic Model has been in planning and designing make-up wells. A major update of the Salak 3D Geologic Model was started in 2012 with the main objective of incorporating results of the wells drilled during the 2008-2009 and 2012-2013 drilling campaigns and about seven years of additional reservoir monitoring. One of the challenges of geothermal reservoir modeling is the description in three dimensions of the geologic characteristics of the complex fractured systems. Volcanic geothermal reservoirs are characterized by extreme heterogeneity in lithology both laterally and vertically. Lava flows and pyroclastics are usually not deposited as widespread as sediments. Even more, in the reservoir section, lithologies can be difficult to identify from cuttings since mud and rock cuttings are often not circulated up to the surface to be sampled and examined. Due to the limited spatial distribution of deep well data, it is not an easy task to accurately estimate the distribution of reservoir and rock properties between wells. Furthermore, these subsurface heterogeneities and uncertainties influence reservoir assessment and production. This paper documents the philosophical approach and workflows that were used to update the Salak 3D Geologic Model. Quite different from approaches taken in past geologic modeling exercises, where the limits of the “commercial” reservoir are defined in each well and then extrapolated beyond well control to build a 3D representation of the reservoir container, the Salak 3D Geologic Modeling Team emphasized the 3D description of the overall geothermal system first, thus providing a means of modeling uncertainty of the commercial limits of the reservoir directly from the 3D geocellular model. The very first step in building the Salak 3D geologic model was to create a well–organized database of core measurements, cuttings description and wireline logs. For the rock-type modeling, lithology types described from the analysis of rock cuttings were grouped on the basis of depositional style and mineral composition to simplify the description of the heterogeneity to a level that can be justified with reasonable confidence. A stratigraphic/structural framework was created based on well correlation and surface geology outcrop description. Rock-type proportions on each stratigraphic unit were assigned based on statistics derived from the database. Another key step in modeling the Salak geothermal system was to build a detailed 3D description of pre-production (equilibrium conditions) reservoir temperatures. Hard data used to constrain the temperature model were measured well temperatures and locations of the surface thermal manifestations. Microseismicity data were used, along with temperature data, to posit the zone of thermal upwelling. Recorded occurrence of high-temperature epidote, the first permeable entry locations in the wells, integrated with resistivity from magnetotelluric survey data and interpretations help to establish the depth of the clay-altered cap, which is a key parameter for defining limits of commerciality. Porosity and permeability were assigned on the basis of formation, rock-type and depth. The next step in the development of the Salak 3D Earth Model is define probabilistic low (P10), mid (P50) and high (P90) reservoir containers based on a number of criteria defining commerciality. Upscaling can then be performed on the gridded geologic model to produce coarser reservoir models suitable for numerical simulation. |