Abstract

The performance of most Abu Dhabi fields to date has not been dramatically impacted by waterflood characteristics. However, a stage has now been reached in field life when current and future reservoir management decisions depend more critically on the prediction and understanding of waterflood performance.

A large number of water / oil relative permeability datasets have been generated on the reservoir studied, by a number of different laboratories. Upon review, a wide scatter was observed in the historical data, both within individual reservoirs and between equivalent horizons, which could not be systematically related to variations in lithology, depth or wettability. To reduce this uncertainty a review of the data was performed to identify the controls on relative permeability behaviour.

A screening process was developed to identify core data validity based on a mechanistic understanding of the impact of different core acquisition procedures, core preparation and laboratory methods. Statistical analysis of the data sought to establish systematic correlations indicating dependency on physical rock properties.

The statistical analysis established that no single property controlled waterflood behaviour. A combination of lithology, sample zone, permeability and Swi appeared to be the major control on water - oil relative permeability. Separating the data by lithotype, sample zone, permeability and Swi, significantly reduced the scatter in the experimental data.

Because of the laboratory procedures adopted by the different laboratories, these historical datasets are unlikely to represent waterflood behaviour in the field. However, the observed trends in the historical data are compared and contrasted with data from recent high quality experimental studies.

Introduction

There is a rich source of historical water / oil relative permeability data available for this giant carbonate reservoir. However, wide scatter has been observed in the historical data, both within individual reservoirs and between equivalent horizons, that sould not be systematically related to variations in lithology, depth or wettability. This lack of consistency / correlatability was considered to be impeding process optimization (and thus effective reservoir management).

The aim of this review was to:

  1. determine data quality and likely representativeness.

  2. identify regional correlations and trends that will enhance consistency of application of these data for reservoir evaluation and performance prediction.

The two main components of the study were:

Mechanistic Screening Of Data.

A screening process was developed to identify core data validity based on a mechanistic understanding of the impacts of different core acquisition, core handling, core preparation and laboratory methods on resulting data.

Statistical Analysis Of Data.

Statistical analysis of the whole data set (unscreened) sought to establish systematic correlations indicating dependency on lithology, height above oil water contact or depth, wettability, porosity, permeability and initial water saturation etc. The ultimate target from this analysis was 'average' rock curves for given lithology, depths, wettability.

This paper presents the results of this review for one particular reservoir within this giant carbonate field.

Mechanistic Screening Of Data.

A screening process was developed to identify core data validity based on a mechanistic understanding of the impacts of different core acquisition, core handling, core preparation and laboratory methods on resulting data.

Statistical Analysis Of Data.

Statistical analysis of the whole data set (unscreened) sought to establish systematic correlations indicating dependency on lithology, height above oil water contact or depth, wettability, porosity, permeability and initial water saturation etc. The ultimate target from this analysis was 'average' rock curves for given lithology, depths, wettability.

This paper presents the results of this review for one particular reservoir within this giant carbonate field.

This content is only available via PDF.
You can access this article if you purchase or spend a download.