Unexpected corrosion was being experienced in steel piping handling source-well brine used for secondary oil recovery. On-site analysis for dissolved H2S and CO2 gave results which predicted much lower rates than were observed on coupons and on linear polarization probes mounted in the flowlines. Cathodic polarization curves made on electrodes in the flowlines suggested a mild oxygen influence. Serial dilution bacteria bottle results suggested an additional MIC (microbiologically induced corrosion) influence.
Three components were therefore involved in this corrosion, dissolved acid gases, small oxygen concentration, and microbially influenced corrosion. A corrosion inhibitor was developed in the laboratory which controlled corrosion due to the first two factors. The ratio of CO2 to H2S, the influence of small amounts of added oxygen, and pre-corrosion were factors considered in this development.
Micro-biocide as well as corrosion inhibitor was applied to the field systems; serial dilution bottles indicated good kill. Coupons and linear polarization probes confirmed that corrosion rates became acceptably low.
The most common method of secondary recovery from oil producing formations is waterflooding.1,2 Many papers have dealt with the corrosion properties of this water, which may be from the petroleum bearing formation, from surface sources, from water supply wells, or mixtures of any of the three. Sometimes the source of corrosion is dissolved naturally occurring acid gases, sometimes the source is oxygen3, sometimes bacteria4, sometimes a combination of these factors. This paper deals with solely supply well water where all three corrosion factors were significant.
The source wells were each producing about 45,000 bbl/day (7150 m3/day) from two distinct zones by natural flow. The first indications of corrosion were the observation of black deposits inside the surface pipelines and corrosion coupon rates of 72 to 95 mpy (1.8 to 2.4 mm/yr). This situation had not been expected due to the low acid gas concentrations in this water; approximately 40 ppm CO2, 180 ppm HCO¯3, and less than 1 ppm H2S. Very little oxygen could be detected, probably because of the small amount of sulfide present 5. This information would predict a corrosion rate of only about 2.5 mpy (0.06mm/yr). 6
The deposit on the coupons was found to be 15-25% FeS, 50-70% FeCO3, 15-25%FexOy and 0-15% unidentified. Subsequent to the coupon results, ultrasonic thickness showed steel pipewall losses at some locations which indicated corrosion rates in excess of 100 mpy (2.5 mm/yr). Linear polarization probes were installed and indicated rates between 10 and 100+ mpy (0.25 and 2.8 mm/yr). The first corrosion related leak in the source water system occurred in steel 350 mils (8.9 mm )thick exposed for 1.5 years; this meant the corrosion rate had to average about 235 mpy (6.0 mm/yr) for this period. Examination of downhole tubing from a well also revealed metal loss and iron sulfide deposition. Serial dilution bacteria bottles showed the scale to contain 104 SRB (sulfate reducing bacteria) colonies per ml. Sulfide concentration increased from the wellhead to the plant; this observation is also consistent with SRB activity.
EXPERIMENTAL
The morphology of the corroded coupons suggested that H2S/O2 corrosion was involved.7 Therefore, electrochemical confirmation in the field was desirable. Electrodes on the field-mounted LPR (linear polarization resistance) probes were potentiodynamically polarized cathodically. This technique has been shown to give mechanistic information about corrosion.6,8 Cathodic polarization curves were made on mild steel electrodes exposed for sev
