Using Production and 4D Seismic Data for History Matching and EOR Optimization on the Norne Field

Abstract

The oil industry faces challenges in the years to come. New oil discoveries are getting smaller, and a major part of the hydrocarbon resources in the world is thought to lie in already producing fields. The oil companies need to get more hydrocarbons out of these fields and increase their recovery factors. There lies huge potential in optimizing well patterns, localizing any pockets of remaining oil, and subsequently drilling to recover it. Traditionally reservoir simulation models have been aiding in these processes, but there are major challenges in correct modeling of the subsurface and simulation of the fields production. In the recent years, 4D (time-lapse) seismic data has been used qualitatively in the industry, to increase the understanding of the subsurface and improve the models of the reservoirs.

This Thesis presents an approach for integrating 4D seismic data in the history matching process alongside the traditional utilization of production data. The proposed approach is applied on an Eclipse 100 black oil model of the E-segment in the Norne Field with success. The discrepancy between the observed production data and the simulated production data is heavily decreased, and simulated OWCs shows similar trends as OWCs interpreted from two 4D datasets. Several potential history matched models which showed good match between observed and simulated production data were disregarded because of the OWCs from the 4D seismic data. Still, there are uncertainties related to the general fluid flow patterns in the reservoir, and a more extensive utilization of 4D seismic data is encouraged, regardless of it being qualitative or quantitative utilization.

Furthermore a manual production optimization process on the simulation model of the Norne field E-segment was carried out. The existing development scheme was revised and several potential well targets were tested. The simulated recovery in the E-segment was increased from 48.8% to 52.2% and the predicted NPV of the E-segment was increased from 676 million US$ to 979 million US$ by adding an additional production well. The simulated recovery was further increased to 52.8% by adding an additional injection well, but the drilling cost and increased water injection costs rules out an additional injection well as a feasible investment.

Low-salinity waterflooding has been simulated with limited success. The mixing of low-salinity and high-salinity water is evident and more oil is mobilized in the simulation model, but the prediction period is too short for the mobilized oil to reach the production wells. Further studies on low-salinity injection on the Norne Field are encouraged.