| dc.description.abstract | The purpose of this research is to characterize the Norne aquifer and testing its performance in the reservoir in terms of additional oil recovery factor with aquifer model included in reservoir simulation model by considering the use of Carter-Tracy model, Fetkovich model and van-Everdingen-Hurst (VEH) model.
The methodology applied involves four stages: Aquifer characterization, model ranking aquifer modelling and economic analysis.
The characteristics of the Norne aquifer considered are aquifer strength and aquifer properties such as aquifer size, aquifer permeability, aquifer porosity and water encroachment angle. The aquifer strength is determined by using diagnostic plots such as a production decline curve of oil rate versus time in semi-logarithm scale and drive indices plot. Estimation of aquifer properties is achieved by using a non-linear regression method in material balance software (MBAL). This approach involves history matching of the average reservoir pressure with computed pressure data by using production data, fluid injection rates and PVT properties. The computed pressure data from the tank model are history matched by regressing the most uncertain parameters in aquifer such as aquifer size, permeability, porosity, and water encroachment angle until the computed pressure matches with historical average reservoir pressure.
In model ranking, three models are investigated. These are van-Everdingen-Hurst (VEH) model, Carter-Tracy model and Fetkovich model based on the standard deviation after regression analysis.
In aquifer modelling, the Carter-Tracy model in Eclipse software is applied by using the estimated Norne aquifer properties. To introduce the Carter-Tracy aquifer in reservoir simulation model, three keywords are added in Eclipse data file. These are AQUDIMS for specifying dimensions for aquifer, AQUCT which specifies property data for Carter-Tracy aquifer and AQUANCON which specifies connection data for analytical aquifer.
In economic analysis, total additional revenue is computed by using two main parameters, additional volume of oil produced with aquifer model included in reservoir simulation model and an average WTI crude oil price for historical period from 1997 to 2006.
The findings of this research show that, the van Everdingen and Hurst (VEH) model is the best model to describe the Norne aquifer with standard deviation of 8.38919, followed by Carter-Tracy model with standard deviation of 19.57410 and Fetkovich is the least model with standard deviation of 43.67590.
Further, results show the water drive strength in the Norne field is strong with aquifer size approximately three times the size of the reservoir (as the reservoir radius is 9618 feet), the aquifer permeability is 2495 millidarcy, aquifer porosity is 20.04 percent, aquifer thickness is 561.12 feet and water influx angle is 51.9133 degree.
Furthermore, the behaviour of the cumulative water influx into the Norne reservoir is observed to correlate with the average reservoir pressure trend.
In addition, by adding an aquifer model in full field Norne reservoir simulation model of 2004, the oil recovery factor is observed to increase by 1.0-1.6 percent at the end of historical and prediction periods respectively. This means the corresponding total additional revenues between 435.0-696.1 million USD are generated when the aquifer model is included during the historical and prediction simulation periods respectively.
Therefore, including the aquifer in Norne reservoir simulation model improves the Norne field reservoir description for better management of the reservoir. | |
