Abstract

Steamflood experiments are carried out in a low-pressure scaled physical model to examine the effects of reservoir heterogeneity on oil recovery. Oil reservoirs are often composed of many layers of different permeabilities and porosities. This stratification can have significant impact on the success of a steam injection process. The experiments simulate three different types of reservoirs: homogeneous, favourable stratification (the lower permeability layer is on the top) and unfavourable stratification (the lower permeability layer is on the bottom). Steamflood performance for different experiments utilizing vertical and horizontal well combinations, and a fracture are compared to determine the optimal recovery strategy for each of the three types of reservoirs.

It is found that permeability stratification has a significant impact on steamflood performance with the unfavourable stratification case achieving the lowest oil recovery. Fully communicating fractures and the use of horizontal injector are not effective in minimizing steam override in unfavourably-layered reservoirs.

Introduction

Petroleum reservoirs are deposited over the course of millions of years. As a result of the changes in the depositional environment the reservoirs are often composed of many layers of sand. This stratification of the reservoirs has a significant impact on the sweep efficiency of many displacement processes, particularly in the case of thermal recovery processes such as steamflooding, cyclic steam stimulation and in-situ combustion. In cyclic steam stimulation and steamflooding operations the density contrast between the steam and steam condensate, oil and water leads to the steam override (or gravity segregation) phenomenon whereby the injected steam rises to the top of the reservoir some short distance from the steam injector leaving a large sand body in the lower part of the reservoir unswept and unmobilized. This steam override henomenon can be accelerated or delayed depending on the stratification of the sand body. Specifically, when the stratification is unfavourable (i.e. the upper layers of the sand sequence have higher permeabilities than the lower layers) the vertical upward migration of the injected Steam is enhanced - leading to a reduced sweep efficiency. On the other hand, vertical steam migration is delayed for the case of "favourable stratification" when there is a declining trend of permeability upward. This study investigates experimentally the steamflooding performance in stratified reservoirs. One of the main objectives of the study was to investigate the potentials of using mechanical means such as drilling horizontal wells and/or creating fractures in appropriate sand layer(s) of the formation to improve steamflood recovery.

One of the earliest studies using scaled physical models was that by Pujol and Boberg [1] who found that for highly viscous oils (μ>100 000 mPas) accurate scaling of the capillary pressure was not crucial, mainly because the ratio of capillary to viscous forces was so small. For oils with medium viscosity (μ< 10 000 mPa's) the results of not scaling the capillary pressure generally led to optimistic predictions of oil recovery. The authors also determined that oil recovery depended mainly on the energy injected, and not on steam injection rates. Pursley [2] used the scaling criteria developed by Pujol and Boberg [I] to build several 3-D models of the Cold Lake reservoir.

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