Assessment of Chemical Disequilibria in Oil Reservoirs: Implications For Production- and Diagenetic- Timescale Geochemical Modelling

J. R. Bunney, K. S. Sorbie, and M. M. Jordan

The accurate prediction of reservoir fluid-rock interaction in oil reservoirs is important in both production engineering and reservoir geology. Produced and/or mixed brine scaling tendencies, formation damage and production treatments all require a knowledge of mineral dissolution and precipitation both close to the wellbore and in the interwell volume. This can be carried forward into the longer timescale over which diagenetic processes occur to predict reservoir quality and to trace migration pathways.

Geochemical packages which are available to model such processes are based on equilibrium thermodynamics. However, during the relatively short timescale of oilfield production, the introduction of large volumes of external brines generally causes the fluid-rock system to be far from equilibrium. In order to apply quantitative geochemical modelling to prediction of production chemistry, the degree of chemical disequilibrium must be assessed.

The Oilfield Scale Research Group at Heriot-Watt University has conducted an extensive series of flooding experiments in which many (~40) reservoir cores have been reconditioned to reservoir conditions and seawater (or produced brine) has been injected over long time periods (up to months) at elevated temperatures. The composition of injection fluid and effluent are monitored throughout the flood and the full mineralogical analysis of the core both before and after these floods has been established. Thus, we are able to quantify the fluid-rock interaction. Both carbonate and aluminosilicate systems have been examined.

Comparing the experimental core flooding data with thermodynamic simulations of those systems illuminates the degree of disequilibria present over production timescales. The results from many core floods in various mineralogical systems have been analysed and will be presented. Certain reactions can be assigned to be rate independent over the timescale in question, by either occurring spontaneously or not at all. This enables the kinetic requirements of any predictive model to be identified and satisfied before any coupled reactive transport simulation is undertaken.

AAPG Search and Discover Article #91019©1996 AAPG Convention and Exhibition 19-22 May 1996, San Diego, California