Numerical Simulation for Development Scenarios of Nw-sabalan Geothermal Reservoir, Iran

A numerical model of the NW-Sabalan geothermal field was developed. The model covers a total area of 96 km and extends vertically from variable elevation of 2200-3600m a.s.l to 1,000m b.s.l. The model has 2,595 grid blocks in l4 horizontal layers. The boundary conditions and distributions of horizontal and vertical permeabilities were obtained by trialand-error matching of the initial temperature and pressure distributions in the wells. Other hydraulic and thermal properties were obtained from exploration drilling and well testing. Observed and calculated temperature profiles and pressure were well-matched, confirming the validity of the conceptual model and providing the first stage of calibration of the numerical model. This effort yielded reliable estimates of the locations and rates of fluid recharge and discharge in the initial state. Performance prediction was made for required steam for 20MWe for three different reinjection scenarios. Assumed that the productivity of new wells to be drilled lies between that observed in Wells NWS1 and NWS4. The results show that the reinjection is required and the area of reinjection is preferable to be located in injection area B than area C. The field can sustain a power generation of 20MWe with six production and four reinjection wells. One make-up well is needed to be drilled of 5 years after operation start in full reinjection scenarios. Two make-up wells for no-reinjection scenario in years 2 and 5 are needed to maintain the required steam for 20 MWe. INTRODUCTION The NW-Sabalan geothermal project is located in the northwestern flank of the Sabalan trachyandesitic stratovolcano, in the province of Ardebil in the northwest of Iran where significant amount of geoscientific studies have been performed by Iran Ministry of Energy since 1975. The project area is within the Moil Valley that is the dominant topographic feature and a major structural zone on the northwest slope of Mt. Sabalan. Several warm and hot springs of neutral Cl-SO4, acid Cl-SO4 and acid SO4 types are discharging in the valley. The isotopic composition of the spring waters and their seasonal variation in flow rates with little change in temperature or chemistry suggested that a large regional ground water aquifer overlies the potential geothermal reservoir. An MT survey implied a presence of a large low resistivity zone (70 km) in the project area (Bromley et al. 2000). A low resistivity zone (<4Ωm) associated with the thermal features was initially selected for target area for exploratory drilling. Therefore, in order to identify the deep reservoir, a drilling program was adopted. This simulation studies with the objective of natural state simulation and production performance of the resource was undertaken for the primary purpose of predicting and assessing the response of the reservoir to the planned developments by implementing TOUGH2 computer codes with EOS 1 (Pruess, et. al., 1999). The work carried out for the numerical modeling has includes the following: Construction of conceptual model of the field Construction of grid blocks and assigning the rock types to the model Definition of initial and boundary conditions, based upon the conceptual model and measured data Validation of the model by matching available downhole temperature and pressure data from the existing wells with the simulated results. This requires running the models to steady state and comparing the simulated data with the known or interpreted conditions in the system. This is an iterative process that continues until a good match is obtained and requires changing model properties, such as permeability, rock density, specific heat capacity and inflow/outflow conditions. Predicting the reservoir response to the planned production and reinjection scenarios. EXPLORATION DRILLING Three exploration wells were drilled in the study area based on the result of geological, geochemical and geophysical studies. Table 1 shows the specifications of the exploration wells. Table 1 Specification of three exploration wells CONCEPTUAL MODEL Before a simulation model of the given geothermal field can be setup, a conceptual model must be developed. A good understanding of the important aspects of the system leads to develop a conceptual model. The model is usually represented by two or three sketches showing a plan view and vertical sections of the geothermal system. Setting up a conceptual model requires the synthesis of information from a multi-disciplinary team composed of geologists, geophysicists, geochemists, reservoir engineers and project managers. Conceptual model of the study area along AB cross section in Figure 1 from north to south was developed. Three exploration wells were interpolated to the AB line and the intersection of the faults and cross section was defined as location of the faults in the model. The faults were assumed to be vertical or sub-vertical (Sinclair Knight Merz, 1998). In central part of the study area, Dizu formation presents and extends to the depth of about 200 m. This formation consists of terrace deposits mainly conglomerates with sand intercalations. It is well exposed around the Site A (NWS1), Site C (NWS3) and throughout Moil Valley. The conglomerates consist predominantly of rounded clasts of andesite, trachyandesite and subordinate trachydacite. The Valhezir formation reaches to the surface near the Site B (NWS4) and it extends to the depth of about 700 m in study area. This formation consists of andesitic lava, tuft and tuff breccias. Pyroclastics predominate in the upper portions of the unit with lavas beneath. A regional Miocene monzonite batholith was interpreted for deep reservoir on the basis of high gravity that extends from the area of the monzonite’s surface exposure in the west into the project area (Sinclair Knight Merz, 1998). Interpretation of MT resistivity data along AB line Well NWS1 NWS3 NWS4 Elevation (m a.s.l) 263