Pore-scale modeling of wettability effects and their influence on oil recovery

The wettability of a crude oil/brine/rock system influences both the form o f petrophysical parameters (e.g., P-c and k(rw)/k(ro)) and the structure an d distribution of remaining oil after secondary recovery. This latter issue is of central importance for improved oil recovery since it represents the "target" oil for any IOR process. In the present study, we have developed a three-dimensional network model to derive capillary pressure curves from nonuniformly wetted (mixed and fractionally wet) systems. The model initial ly considers primary drainage and the aging process leading to wettability alterations. This is then followed by simulations of spontaneous water imbi bition, forced water drive, spontaneous oil imbibition and forced oil drive -i.e., we consider a complete flooding sequence characteristic of wettabili ty experiments. The model takes into account many pore level flow phenomena such as film how along wetting phase clusters, trapping of wetting and non wetting phases by snapoff and bypassing. We also consider realistic variati ons in advancing and receding contact angles. There is a discussion of the effects of additional parameters such as the fraction of oil-wet pores, mea n coordination number and pore size distribution upon fractionally and mixe d wet capillary pressure curves. Moreover, we calculate Amott oil and water indices using the simulated curves. Results indicate that oil recovery via water imbibition in weakly water-wet cores can often exceed that obtained from strongly water-wet samples. Such an effect has been observed experimen tally in the past. The basic physics governing this enhancement in spontane ous water imbibition can be explained using the concept of a capillarity su rface. Based on these theoretical calculations, we propose a general "regim e based" theory of wettability classification and analysis. We classify a r ange of experimentally observed and apparently inconsistent waterflood reco very trends into various regimes, depending upon the structure of the under lying oil- and water-wet pore clusters and the distribution of contact angl es. Using this approach, numerous published experimental Amott indices and waterflood data from a variety of core/crude oil/brine systems are analyzed .