Monitoring of fluid–rock interaction and CO 2 storage through produced fluid sampling at the Weyburn CO 2-injection enhanced oil recovery site, Saskatchewan, Canada

Abstract

The Weyburn Oil Field is a carbonate reservoir in south central Saskatchewan, Canada and is the site of a large CO 2 injection project for purposes of enhanced oil recovery. The Weyburn Field, in the Mississippian Midale Formation, was discovered in 1954 and was under primary production until secondary recovery by water flood began in 1964. The reservoir comprises two units, the Vuggy and the Marly, and primary and secondary recovery are thought to only have significantly depleted the Vuggy zone, leaving the Marly with higher oil saturations. In 2000, PanCanadian Resources (now EnCana), the operator of the field, began tertiary recovery by injection of CO 2 and water, primarily into the Marly. The advent of this project was an opportunity to study the potential for geological storage of CO 2. Using 43 Baseline samples collected in August 2000, before CO 2 injection at Weyburn, and 44 monitoring samples collected in March 2001, changes in the fluid chemistry and isotope composition have been tracked. The initial fluid distribution showed water from discovery through water flood in the Midale Formation with Cl ranging from 25,000 to 60,000 mg/L, from the NW to the SE across the Phase 1A area. By the time of Baseline sampling the produced water had been diluted to Cl of 25,000–50,000 mg/L as a result of the addition of make up water from the low TDS Blairmore Formation, but the pattern of distribution was still present. The Cl distribution is mimicked by the distribution of other dissolved ions and variables, with Ca (1250–1500 mg/L) and NH 3 (aq) increasing from NW to SE, and alkalinity (700–300 mg/L), resistivity, and H 2S (300–100 mg/L) decreasing. Based on chemical and isotopic data, the H 2S is interpreted to result from bacterial SO 4 reduction. After 6 months of injection of CO 2, the general patterns are changed very little, except that the pH has decreased by 0.5 units and alkalinity has increased, with values over 1400 mg/L in the NW, decreasing to 500 mg/L in the SE. Calcium has increased to range from 1250 to 1750 mg/L, but the pattern of NW–SE distribution is altered. Chemical and isotopic data suggest this change in distribution is caused by the dissolution of calcite due to water–rock reactions driven by CO 2. The Baseline samples varied from −22 to −12‰ δ 13C (V-PDB) for CO 2 gas. The injected CO 2 has an isotope ratio of −20‰. The Monitor-1 samples of produced CO 2 ranged from −18 to −13‰, requiring a heavy source of C, most easily attributed to dissolution of carbonate minerals. Field measured pH had increased and alkalinity had decreased by the second monitoring trip (July 2001) to near Baseline values, suggesting continued reaction with reservoir minerals. Addition of CO 2 to water–rock mixtures comprising carbonate minerals causes dissolution of carbonates and production of alkalinity. Geochemical modeling suggests dissolution is taking place, however more detail on water–oil–gas ratios needs to be gathered to obtain more accurate estimates of pH at the formation level. Geological storage of CO 2 relies on the potential that, over the longer term, silicate minerals will buffer the pH, causing any added CO 2 to be precipitated as calcite. Some initial modeling of water–rock reactions suggests that silica sources are available to the water resident in the Midale Formation, and that clay minerals may well be capable of acting as pH buffers, allowing injected CO 2 to be stored as carbonate minerals. Further work is underway to document the mineralogy of the Midale Formation and associated units so as to define more accurately the potential for geological storage.