No abstract

We have used a sintered glass bead core to simulate the spaces and surfaces of reservoir rock in studies of the bacterial plugging phenomenon that affects waterflood oil recovery operations. The passage of pure or mixed natural populations of bacteria through this solid matrix was initially seen to promote the formation of adherent bacterial microcolonies on available surfaces. Bacteria within these microcolonies produced huge amounts of exopolysaccharides and coalesced to form a confluent plugging bioffim that eventually caused a >99% decrease in core permeability. Aerobic bacteria developed a plugging biofilm on the inlet face of the core, facultative anaerobes plugged throughout the core, and dead bacteria did not effectively plug the narrow (33-,um) spaces of this solid matrix because they neither adhered extensively to surfaces nor produced the extensive exopolysaccharides characteristic of living cells. The presence of particles in the water used in these experiments rapidly decreased the core permeability because they became trapped in the developing biofilm and accelerated the plugging of pore spaces. Once established, cells within the bacterial biofilm could be killed by treatment with a biocide (isothiazalone), but their essentially inert carbohydrate biofilm matrix persisted and continued to plug the pore spaces, whereas treatment with 5% sodium hypochlorite killed the bacteria, dissolved the exopolysaccharide biofilm matrix, and restored permeability to these plugged glass bead cores. * Corresponding author. were both replicable and approximately equal to those of an "open" sandstone. Thus, especially in our early studies in which pure cultures of aquatic bacteria were used, any changes in permeability that occurred in the model cores could be attributed fully to microbiological factors.

Multiple displacement experiments in the same pore-throat pair show that the amount of trapped, residual oil is afunction of pore-throat geometry and wettability and is not affected much by differences of viscosity or interfacial tension within the limits studied. Displacement efficiency was least where the displacing phase was strongly wetting (contact angles < 30 degrees) and greatest for conditions of intermediate wettability (contact angles -90 degrees) in systems with large pore-to-throat size ratios.In extending these conclusions to multi-pore media with branching networks of pores and throats, it must be understood that trapping by capillary instability (snap-off) is only one of several mechanisms of trapping which may occur. However, in strongly water-wet rocks with large aspect ratios, it may be the most important mechanism of trapping. Semi-rigidfilmsformed by a contaminant at the interface of an aqueous solution with decane inhibited snap-off and could favour higher displacement efficiency by maintaining continui'-ty of decane during imbibition. Introduction Trapping of oil and gas on a microscopic scale in a petroleum reservoir rock is affected by the geometric and topologic prop-erties of the pores, by the properties of the fluids and by prop-erties related to fluid-rock interaction such as wettability.Trapping also is affected by gravity, capillary and viscous N.C. Wardlaw N.C. Wardlaw is a graduate of Man-chester University (Hon. B.Sc. geo-logy, 1957) and Glasgow University (Ph.D. geology 1961) and is currently Professor and Head

Void spaces in porous media can be considered as three-dimensional networks consisting of bulges (pores) connected by constrictions (throats). Computer simulations of drainage-imbibition processes show that the critical end points of wetting-phase and nonwetting-phase saturation, in drainage and imbibition respectively, and the form of simulated relative permeability curves all were significantly different for uncorrelated and correlated pore-throat models. Since these models were identical except for the arrangement of throats in relation to pores, the degree of pore-throat size correlation appears to be an important property influencing flow and fluid displacement. Examples of uncorrelated and correlated pore-throat structures in rocks are presented and it is shown that this property, although difficult to quantify by direct observation, can be evaluated from capillary pressure curves.

Residual oil was obtained by waterflooding, under water-wet conditions, in unconsolidated packs of glass spheres at low capillary numbers. Trapped oil blobs were mobilized in a sequence of decreasing size with subsequent increases in capillary number. The larger blobs, which extend over many adjacent pores, fission repeatedly upon mobilization and the smaller daughter blobs are restranded. There is negligible oil production until the blob population has been reduced to single pore blobs (singlets).Thus, the critical capillary number for mobilization of significant amounts of oil in unconsolidated bead packs is not related to the original blob size population, but rather to the conditions necessary for mobilization of singlets.On a rCcup6rC de l'huile residuelle par dCplacement avec de I'eau, dans des conditions humides, dans des empilements non tassCs de spheres de verre pour des nombres capillaires petits. Des bulles d'huile piegees ont Ct C mises en mouvement dans I'ordre des dimensions dCcroissantes, par des accroissements successifs du nombre capillaire. Les bulles les plus grosses, qui s'ttendent sur un grand nombre de pores adjacents, subissent une fission rtp6tCe durant leur dtplacement et les plus petites bulles engendrkes sont rCassoci6es. On observe une production negligeable d'huile jusqu'i ce que la population des bulles ait Ct C rCduite i des bulles dans un pore unique (singulets).Par consCquent, le nombre capillaire critique pour le dkplacement de quantitks importantes d'huile dans des empilements de sphtres non consolid& n'est pas relic i la population originale des dimensions des bulles, mais plutBt aux conditions requises pour le dCplacement des singulets.uring waterflooding of oil reservoir rocks, as the water