Using WAG Laboratory Data to Derisk Water Injection at the Gas-Oil Contact on a Giant Offshore Field

Abstract

The current development strategy of this field is focused on the oil rim, it involves downdip water injection and gas injection in the gas cap to maintain the pressure. A co-development of the oil rim with the large gas cap is being tested at pilot scale and considered at large scale, it requires to create a water fence or water barrier by injection water in updip position, near the gas cap, to isolate if possible oil rim and gas cap, and avoid oil movement into the gas cap. This should enable production of gas and condensate from the gas cap, and still continue simultaneous oil production of the oil rim. This co-development program requires drilling of numerous inner ring water injectors, the well count and placement need to be optimized with help of 3D simulation, which requires relevant flow functions (like KRs). This paper is focusing on the laboratory program followed to quantify these 3-phase flow functions, which happened to be very close to the SCAL program used to constrain WAG operations. Interesting findings related to these measures are presented and discussed. It was found that the near-miscible gas flood clearly outperformed waterflood, and the beneficial gas-oil interaction could be quantified. Three-phase flow (water and gas injection successively) clearly outperformed 2-phase flow, and a succession of several short slugs was also better than long slugs. Trapping of gas by water was also rather low, except if several cycles were followed. Water behavior was also quite surprising, as its mobility in presence of gas was measured higher than expected during imbibition, which implies to increase the water injection rates and water treatment facilities in order to maintain the water barrier in the project, but will also have implications on the water cut and eruptivity of oil producers in the oil rim. When water injection ceases, as in WAG operations when switching to gas, or for intermittent water injection, water mobility was found to be very low for the drainage phase, and water trapping was clearly documented despite rate bumps in the gas injection. This could be interpreted by hysteresis affecting also the water phase, and can be explained logically by water being the least wetting phase in an oil-wet near-miscible environment. This mechanism may be of importance, as it may stabilize the updip water fence if water injection rates are reduced.