Study of the Lower Cretaceous sands in the Van den Bosch Geothermal energy concession

Goal of this project was to get a better understanding of the producing formations in the two doublets in the Van den Bosch Geothermal energy concession in Bleiswijk. It can be concluded based on the GR logs in VDB-04, the lithologs, End Of Well reports, a Biostrat log in VDB-04 and the seismic interpretation that the first doublet (VDB-01 & VDB-02) produces from a combination of the Rijswijk Member and another aggradational stacked sand formation. The second doublet (VDB-03 & VDB-04) produces from the Delft Sandstone. It is expected that producer and injector in each doublet have good communication and interference between the doublets is absent because the reservoirs are separated by the Rodenrijs Shale and production pressures and rates are constant for years. A regional well log correlation based on Maximum Flooding Surfaces is carried out to put the interpretation of the reservoir sands into a regional context with the surrounding oil fields. Using seismic interpretation the thickness and lateral extension of the Rijswijk and Delft reservoirs is determined. With these interpretations a static model is built of the reservoirs of the two doublets. In the second part of the project, field data from the past four years of production is analysed. The data consists of production rates, pump pressures and a well test in VDB-01. Using a Monte Carlo simulation ranges of the average permeability are calculated. The ranges of the average permeability are 200-500 mD in the Rijswijk and 300-600 in the Delft Sandstone. The main uncertainties in these estimated ranges are the Flowing Bottom Hole Pressure (FBHP) calculations, the skin and the thickness of the formation. In a history match exercise, production is simulated in a Dynamic model in Eclipse 100. The tuning parameter is the average horizontal permeability. In addition the influence on the simulation results of the vertical permeability, the skin and the fluvial shape factors is investigated. Resulting ranges of the average horizontal permeability are similar to those calculated with the radial inflow Monte Carlo simulation: 100-300 mD for the Rijswijk and 300-400 mD for the Delft Sandstone. Simulations show that the vertical permeability has a large influence on the reservoir performance and is therefore an important uncertainty. An initial estimation of the Rijswijk kvertical is 70% of Khorizontal and 22/386*khorizontal in the Delft Sandstone. This is based on earlier studies. Especially in the Rijswijk situation, the estimate may be too large. If the vertical permeability decreases, the horizontal permeability should increase to compensate in the history match. In that case the average permeability range approaches the results from the radial inflow equation analysis more. The well test was plagued by wellbore effects and a too short build-up period. Therefore no reliable parameters could be derived and compared to the radial inflow analysis and history match results. However the analysis showed that a well performed well test could give insight in skin, average permeability and flow boundaries.Petroleum EngineeringGeoscience & EngineeringCivil Engineering and Geoscience