Direct insights into the pore-scale mechanism of low-salinity waterflooding in carbonates using a novel calcite microfluidic chip

摘要

One of the key open questions in the area of low or controlled salinity water flooding (LSWF or CSWF) is how the observed oil recovery at macro-scale (e.g. Darcy or core-scale) can the explained and what underlying microscopic mechanisms drive it. Thus far, the micro-model investigation of LSWF has been limited to sandstones, remaining challenging to apply to carbonates. In this paper we aim to i) extend the capability to fabricate a novel calcite micromodel using Iceland spar calcite crystal, ii) investigate the pore-scale mechanisms leading to oil recovery from carbonates.A target crude oil-brine-rock (COBR) system was first selected. To screen potential brines which can produce low-salinity-effect (LSE) and to guide the design of the micromodel experiments, contact angle measurements were carried out using two methods: i) contact angle under fixed, and ii) under dynamic salinity condition. The micromodel displacement experiments were then performed by flooding an oil saturated model with high salinity water followed by low salinity water injection to displace the high salinity water and observe any potential changes to the configuration and saturation of the residual oil. Additionally, the effect of connate water presence on the efficiency of LSE was investigated. To account for the time effects of the low salinity process, the experiments were monitored for an extended time period in order of several days to a month.For the COBR system studied in the micromodel, the results clearly show that when brine salinity is lowered the microscopic sweep efficiency is improved; providing a direct in-situ evidence for wettability alteration to a more water-wetting state. The presence of connate water enhanced the efficiency of LSWF both in terms of speed (time-scale) and quantity of oil recovery. It is postulated that in the presence of connate water an initial water-film around the calcite surface is present which facilitates the diffusive transport of brine ions when low salinity is injected. Thus it is favorable to have an initial water film present; a case for mixed-wettability. We observed that the oil production was non-instantaneous characterized by a prolonged induction time and a slow "layer-by-layer" recovery either from the pore body or throat wall; a process we refer to as "peel-off". Before the oil can be removed from the calcite surface, de-wetting (or de-pinning) patterns were formed which grew and coalesced toward formation of a clearly visible larger pattern. Ultimately, the remaining oil under low salinity was comparatively much less compared to the end of high salinity step.The observed mechanism of the oil recovery and the slow associated time have direct implications for the pore-scale simulation of the process and upscaling to Darcy-scale, and the design of laboratory experiments to avoid false negative results. They would also likely imply lack of a clear oil-bank observation at core scale.