Investigating Continuously Updated History Matching Using Smart Proxy (Surrogate Reservoir Model)

Abstract

Geothermal energy is a globally distributed sustainable energy with the advantages of a stable base load energy production with a high capacity factor and zero SO x , CO, and particulates emissions. It can provide a potential solution to the depletion of fossil fuels and air pollution problems. The geothermal district heating and cooling system is one of the most common applications of geothermal energy, and consists of geothermal wells to provide hot water from a fractured geothermal reservoir, a surface energy distribution system for hot water transmission, and heating/cooling facilities to provide water and space heating as well as air conditioning for residential and commercial buildings. To gain wider recognition for the geothermal district heating and cooling (GDHC) system, the potential to develop such a system was evaluated in the western United States, and in the state of West Virginia. The geothermal resources were categorized into identified hydrothermal resources, undiscovered hydrothermal resources, near hydrothermal enhanced geothermal system (EGS), and deep EGS. Reservoir characteristics of the first three categories were estimated individually, and their thermal potential calculated. A cost model for such a system was developed for technical performance and economic analysis at each geothermally active location. A supply curve for the system was then developed, establishing the quantity and the cost of potential geothermal energy which can be used for the GDHC system. A West Virginia University (WVU) case study was performed to compare the competiveness of a geothermal energy system to the current steam based system. An Aspen Plus model was created to simulate the year-round campus heating and cooling scenario. Five cases of varying water flow rates and temperatures were simulated to find the lowest levelized cost of heat (LCOH) for the WVU case study. The model was then used to derive a levelized cost of heat as a function of the population density at a constant geothermal gradient. By use of such functions in West Virginia at a census tract level, the most promising census tracts in WV for the development of geothermal district heating and cooling systems were mapped. This study is unique in that its purpose was to utilize supply analyses for the GDHC systems and determine an appropriate economic assessment of the viability and sustainability of the systems. It was found that the market energy demand, production temperature, and project lifetime have negative effects on the levelized cost, while the drilling cost, discount rate, and capital cost have positive effects on the levelized cost by sensitivity analysis. Moreover, increasing the energy demand is the most effective way to decrease the levelized cost. The derived levelized cost function shows that for EGS based systems, the population density has a strong negative effect on the LCOH at any geothermal gradient, while the gradient only has a negative effect on the LCOH at a low population density.

A data-driven smart proxy model for a comprehensive reservoir simulation

“…However, there were still challenges with pressure maintenance on the eastern and western margins of the reservoir. To overcome these challenges and after studies showing that the SACROC [66]. After two years of WAG injections, oil production peaked at 211,000 barrels per day.…”

“…Due to the heterogeneity of the reservoir properties, performing history matching for the SACROC Unit is challenging. The area of study of the SACROC Unit has been history matched by He [66] as part of her graduate study research project. In her work, all reservoir properties were considered trustworthy, except for porosity, permeability, and relative permeability [66].…”

“…Both reservoir properties were decreased by 50% and increased up to 400%. However, only one relative permeability curve was used due to the limit of information availability [66].…”

“…Based on different geological scenarios, 80 simulation runs have been developed for history matching. He [66] was able to match water production, oil production, and average reservoir pressure.…”