AN ANALYSIS OF THE PRODUCTION PERFORMANCE OF THE WINDALIA SAND RESERVOIR OF THE BARROW ISLAND OIL FIELD

Abstract

The Windalia Sand is a high porosity, low permeability oil reservoir. Currently 454 wells penetrate the unit for production or water injection operations, and are drilled on a north-south, east-west 16 ha (40 ac.) spacing. Early production performance data indicated a trend of water break-through into wells located east and west of water injection wells in an inverted nine-spot pattern. This early trend has continued and the east- west break-through has become more widespread with time. It was recognised that it could be possible to improve the performance of the waterflood if the factors causing the phenomenon were able to be identified. A detailed geological review of well data was initiated to investigate causes and possible controls of the phenomenon and to determine if oil recovery could be improved. This work was augmented by an engineering study of production data. Subsequently, a computer model was developed to investigate the simulated effects of changes to well patterns on the field's production performance.

The geological review determined that the reservoir contains significant local and transitional irregularities (or inhomogeneities). The mapping of a number of reservoir parameters has shown there are genetic patterns or trends and these are postulated as being at least partial controls of preferential direction of fluid movement.

Previously the reservoir had been regarded as being a more uniform "layer-cake" sand. Well completion practices and timing together with production and injection methods are thought to have accentuated the latent genetic controls. Imposed pressure parting has been postulated, on engineering premises, as a control of fluid movement. The modelling study used the notion of anisotropic permeability in attempting to history-match production performances.

Because of the reservoir size and anisotropy it was impractical to model the entire field. Selected type areas within the reservoir were studied. Good history-matching of various well types (based on location within a pattern) was possible. Predictions of production performance can be made for various simulated pattern changes allowing feasibility studies to be made of possible conversion programs.

East-west producing wells are being converted to injectors as they water out. This program has converted part of the reservoir to a line-drive injection configuration and improved performance in these areas is evident.