Sulfide Remediation by Pulsed Injection of Nitrate into a Low

Nov 11, 2009 - (A, D) Sulfide, (B, E) sulfate, and (C, F) nitrite concentrations as a ..... Telang , A. J.; Ebert , S.; Foght , J. M.; Westlake , D. W...
0 downloads 0 Views 4MB Size
Download PDF
Environ. Sci. Technol. 2009 43, 9512–9518

Sulfide Remediation by Pulsed Injection of Nitrate into a Low Temperature Canadian Heavy Oil Reservoir G E R R I T V O O R D O U W , * ,† ALEKSANDR A. GRIGORYAN,† ADEWALE LAMBO,† SHIPING LIN,† HYUNG SOO PARK,† THOMAS R. JACK,† DENNIS COOMBE,‡ BILL CLAY,§ FRANK ZHANG,§ RYAN ERTMOED,| KIRK MINER,| AND JOSEPH J. ARENSDORF⊥ Petroleum Microbiology Research Group, Department of Biological Sciences, University of Calgary, 2500 University Drive NW, Calgary, Alberta, T2N 1N4, Canada, Computer Modelling Group Limited, 150, 3553 31st Street NW, Calgary, Alberta, T2L 2K7, Canada, Enerplus Resources Fund, 3000 333 Seventh Avenue SW, Calgary, Alberta, T2P 2Z1, Canada, Baker Hughes Incorporated, 208 Saskatchewan Drive NE, Redcliff, Alberta, T0J 2P0, Canada, and Baker Hughes Incorporated, 12645 W Airport Boulevard, Sugar Land, Texas 77478

Received July 27, 2009. Revised manuscript received October 12, 2009. Accepted October 28, 2009.

Sulfide formation by oil field sulfate-reducing bacteria (SRB) can be diminished by the injection of nitrate, stimulating the growth of nitrate-reducing bacteria (NRB). We monitored the field-wide injection of nitrate into a low temperature (∼30 °C) oil reservoir in western Canada by determining aqueous concentrations of sulfide, sulfate, nitrate, and nitrite, as well as the activities of NRB in water samples from 3 water plants, 2 injection wells, and 15 production wells over 2 years. The injection water had a low sulfate concentration (∼1 mM). Nitrate (2.4 mM, 150 ppm) was added at the water plants. Its subsequent distribution to the injection wells gave losses of 5-15% in the pipeline system, indicating that most was injected. Continuous nitrate injection lowered the total aqueous sulfide output of the production wells by 70% in the first five weeks, followed by recovery. Batchwise treatment of a limited section of the reservoir with high nitrate eliminated sulfide from one production well with nitrate breakthrough. Subsequent, field-wide treatment with week-long pulses of 14 mM nitrate gave breakthrough at an additional production well. However, this trend was reversed when injection with a constant dose of 2.4 mM (150 ppm) was resumed. The results are explained by assuming growth of SRB near the injection wellbore due to sulfate limitation. Injection of a constant nitrate dose inhibits these SRB initially. However, because of the constant, low * Corresponding author tel: 403-220-6388; e-mail: voordouw@ ucalgary.ca. † University of Calgary. ‡ Computer Modelling Group Limited. § Enerplus Resources Fund. | Baker Hughes Incorporated, Alberta, Canada. ⊥ Baker Hughes Incorporated, Sugar Land, Texas. 9512

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 24, 2009

temperature of the reservoir, SRB eventually grow back in a zone further removed from the injection wellbore. The resulting zonation (NRB closest to and SRB further away from the injection wellbore) can be broken by batch-wise increases in the concentration of injected nitrate, allowing it to re-enter the SRB-dominated zone.

Introduction Much of the world’s oil is produced by injection of water to maintain pressure and push the oil toward the production wells (1). This results in production of an oil-water mixture, which is separated into produced oil and produced water in above-ground facilities. In fields on land, where the availability of fresh water is limited, the reservoir is usually subjected to produced water reinjection (PWRI). In offshore reservoirs, which are injected with seawater, PWRI is rarely practiced. Instead, the produced water is cleaned and discharged. In fields, where oil is produced by PWRI, makeup water is often needed to bridge the shortfall in volume of produced water versus needed injection water. Oil production by water injection often leads to souring, the production of sulfide by resident sulfate-reducing bacteria (SRB) or other sulfidogenic bacteria. SRB use oil organics to reduce the injected sulfate to sulfide. Souring is noted by the appearance of sulfide in produced gas and water in production wells. Production of sulfide can be delayed if reservoir rock contains a significant fraction of iron minerals such as siderite (FeCO3), which bind injected sulfide by conversion to FeS (2, 3). Amendment of injection water with nitrate can diminish formation of sulfide and can remove sulfide already formed. Injection of nitrate boosts the activity of heterotrophic and sulfide-oxidizing nitrate-reducing bacteria (hNRB and soNRB, respectively). The former are thought to outcompete SRB for the use of the same oil organics as electron donor for nitrate and sulfate reduction, respectively, a mechanism referred to as competitive exclusion (4). The latter oxidize sulfide directly to sulfur or sulfate, while reducing nitrate to nitrite and then to nitrogen (5, 6). Reduction to ammonium occurs also (7). The produced nitrite strongly inhibits SRB activity (8, 9). Hence, addition of nitrate to injection water has the potential to lower produced sulfide concentrations. Successful application has been reported for fields with a relatively high down-hole temperature (>60 °C), producing light oil (10). Microbial growth in such fields may be limited to the near injection wellbore region, which is cooled down by the continuous inflow of cold injection water and may have a lower concentration of toxic water-soluble oil hydrocarbons, due to continuous washout (3, 10). Hence, in high temperature reservoirs, sulfide production may originate just from the near-injection wellbore region, which can be relatively easily treated with nitrate. In contrast, microbial growth in fields with a low down-hole temperature producing heavy oil, may not be subject to spatial limitations, making treatment more challenging, as is reported here.

Experimental Section Source of Samples. Samples were obtained from the Enermark field in the Western Canadian Sedimentary Basin near the town of Medicine Hat in South Eastern Alberta (Figure 1 top). The field consists of a shallow, oil-containing, glauconitic sandstone at a depth of ∼1000 m. The in situ temperature is ∼30 °C. The field has been under primary production, meaning that oil was produced without water injection, since the early 1980s. Water injection was started 10.1021/es902211j CCC: $40.75

 2009 American Chemical Society

Published on Web 11/11/2009

FIGURE 1. (Top) Survey of sampling sites in the Enermark field. The field has 3 water injection zones, each with its own water plant. Many horizontal wells are indicated. Sampling sites are water plants 1, 17, and 20 (squares), makeup water sources 21, 22, and 23 (diamonds), injection wells 8 and 14 (circles), and production wells 2-7, 9-13, 15, 16, 18, and 19. Note that only a fraction of the injection and production wells were sampled. (Bottom) Schematic of water flow and nitrate injection in water injection zones 1-WP and 17-WP. Injection water originating from 1-WP served as makeup water at 17-WP, as explained in the text. in 2000 to maintain reservoir pressure. Souring, as determined by increasing sulfide concentrations in the gas phase, started in 2006. In 2007 approximately 2500 m3/day of water were injected with approximately 1000 m3/day of heavy oil with an API gravity of 16 being produced (11). Produced water is separated from oil and mixed with makeup water in three water plants (Figure 1) to obtain injection water, which is distributed to injection wells throughout the field. In 2007 the daily output of injection water from the three water plants was approximately 1900, 600, and 40 m3, respectively. This had increased to 2900, 1200, and 300 m3, respectively, in 2009. Characteristics of produced water, makeup water, and injection water at the main water plant 1-WP are given in Table S1 in the Supporting Information (SI). Its makeup water was the output of the municipal sewage treatment plant and had a sulfate concentration of 4-5 mM (Table S1), causing the injection water to have a sulfate concentration of 0.8-1.2 mM. Injection water from 1-WP was used as the makeup water at 17-WP (Figure 1B), whereas deep formation water resembling the produced water described in Table S1 was

used as the makeup water at 20-WP. As a result the injection water output of 17-WP and 20-WP had a lower sulfate concentration than that of 1-WP. Fifteen sites were sampled initially. This was increased to 23 sites over the course of the study as indicated in Table S2. The breakthrough time for travel of water and dissolved ions from injectors to producers through high permeability zones is estimated at 5-7 weeks from data presented below. Samples were collected in 250mL and in 1-L Nalgene bottles filled to the brim to exclude air. The latter were frozen onsite in dry ice. Samples arrived in the lab within 5 h of collection. The 1-L bottles were transferred to a Coy anaerobic hood with an atmosphere of 5% (v/v) H2, 10% CO2, 85% N2. Aqueous sulfide concentrations in these samples were determined within 24 h. The 250-mL samples were stored frozen at -70 °C. Analytical Determinations. Water (5 mL) was removed from the 1-L samples using a disposable pipet. Following centrifugation (10 min at 14 000 rpm) to remove particulate material, part of the supernatant was used immediately for determination of the sulfide concentration, while the remainder was stored frozen at -70 °C. Aqueous sulfide concentrations were determined colorimetrically with N,Ndimethyl-p-phenylenediamine (12). Sulfate was assayed by a turbidometric method using BaCl2 (13) and/or by highpressure liquid chromatography (HPLC), as described elsewhere (7, 11, 14). Nitrate and nitrite concentrations were determined by HPLC (14), whereas nitrite concentrations were also determined with sulfanilamide/n-(naphthyl)ethylenediamine reagent (15). Determination of Microbial Activities. Although the presence of selected microbial groups is often evaluated by the most probable number (MPN) method, this works poorly when growth rates are slow, when multiple cells are needed to establish a positive reading and/or when cells are clumped or attached. Unfortunately, these conditions apply to the evaluation of anaerobes in oil fields. An alternative is to do a single dilution and measure the time required to score the tube as positive. This time is inversely proportional to the logarithm of the MPN of microorganisms in the original sample (16). The time required for a microbial population to convert substrates to products also represents its activity and we will use this more direct term, rather than a derived MPN, to refer to these data. Microbial activities were determined by adding 2.5 mL of sample to 50 mL of CSB-K medium (7) in a 100-mL stoppered serum bottle with a gas phase of 10% (v/v) CO2, balance nitrogen. The medium contained 20 mM sulfate and 3 mM volatile fatty acids (VFA, a mixture of acetate, propionate, and butyrate) for determination of SRB activity, or 10 mM nitrate and 3 mM VFA or 10 mM nitrate and 4 mM sulfide for determination of hNRB and soNRB activity, respectively. The concentrations of sulfate and sulfide, of nitrate and nitrite, and of nitrite, sulfate, and sulfide were determined to evaluate SRB, hNRB, and soNRB activities, respectively. Microbial activities in arbitrary units were calculated as 100/t1/2 (units/day), where t1/2 (days) was the time needed for 50% of the electron acceptor (sulfate or nitrate) to be reduced under the stated conditions. For measurement of SRB or hNRB activity with toluene as the electron donor, media were supplemented with 4 mM toluene and 0.4 mL of 2,2,4,4,6,8,8-heptamethylnonane, as the inert carrier. Incubations were at room temperature without shaking. Nitrate Injection. Field-wide injection of 45% (w/w) calcium nitrate was started at 1-WP and 17-WP on May 7, 2007 (week 0) and at 20-WP on September 15, 2008 (week 71), as indicated in Table S2. The injection rates were adjusted such that an average concentration of 2.4 mM nitrate (200 ppm of calcium nitrate) was maintained at 1-WP and an average concentration of 1.8 mM nitrate (150 ppm of calcium nitrate) was maintained at 17-WP for weeks 1-63. From weeks VOL. 43, NO. 24, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

9513

FIGURE 2. Nitrate concentrations at two water plants (1-WP and 17-WP) and two injection wells (14-IW and 8-IW) in the Enermark field. Although most data shown were measured, the following were entered based on information provided by field personnel: (all panels) the peaks at 14 mM indicating week-long nitrate pulses and the zero values for weeks 40, 85, 86, and 87; (panel B) all values of 10 mM, indicating nitrate batches injected from weeks 33 to 101. All 11 nitrate pulses are represented in (A)-(D), whereas 22 of 62 injected nitrate batches are represented in (B). For panel (C) all data for weeks 1-32 were given the average value determined for weeks 35-62, because sampling of 17-WP did not start until week 35. 64 to 96 pulses of high (14 mM) and low (2.4 mM) nitrate were alternated weekly. No nitrate was injected in weeks 40 and 84-86. Batch-wise injections of 1 m3 of 45% (w/w) calcium nitrate were done weekly for 1-2 h at injection well 14-IW from weeks 33 to 101. These were diverted to the 17WP water plant from week 101 onward. The nitrate injection regime is indicated in Figure 2 and in Table S2.

Results Microbial Activities in the Enermark Field. We verified that NRB activities, required to successfully remove sulfide with nitrate, were present in the Enermark field. Average hNRB and soNRB activities in samples from production wells 2-, 3-, 4-, 5-, 7-, 9-, 11-, 13-, and 15-PW, collected just prior to the start of nitrate injection (week 0; May 7, 2007), were (51 ( 23) units/day and (26 ( 16) units/day, respectively. These activities increased as nitrate injection progressed. In week 63 these same 9 wells had an average hNRB activity of (135 ( 108) units/day and an average soNRB activity of (74 ( 50) units/day. Toluene, present in Enermark oil at a concentration of 6 mM, is an important electron donor for nitrate reduction (11). Testing the same week 63 samples for hNRB activity with toluene as the electron donor gave an average activity of (18 ( 15) units per day. SRB activity with VFA as electron donor was not determined prior to nitrate injection. In week 63 the average SRB activity for production wells 2-, 3-, 4-, 5-, 9-, 11-, and 15-PW was (6.1 ( 1.2) units per day. Wells 13-PW and 7-PW had 0 units/day, which may be related to the fact that nitrate was being produced at 13-PW and was close to breaking though at 7-PW at that time. No SRB activity with toluene as the electron donor was found for week 63 samples. Field-Wide Distribution of Nitrate. Because the injection water contains residual oil organics (11), some of the nitrate 9514

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 24, 2009

added at the water plants may be reduced by NRB during transport of water to the injection wells, which took approximately one day. Determination of the decrease in electron acceptor (sulfate, nitrate) concentrations of the transported injection water indicated a loss of 7% nitrate and 11% sulfate during water transport from 1-WP to 14-IW and of 14% nitrate and 80% sulfate from 17-WP to 8-IW (Figure S1). Despite the significant sulfate loss, water at 8-IW did not contain significant dissolved sulfide. However, it was blackish indicating the possible presence of iron sulfides. Nitrite was found both in samples from the water plants and from the injection wells. Nitrite did not increase significantly during transport, except at 14-IW in weeks 43-45 (Figure S2). Overall, the data indicated that 86-93% of the nitrate added at the water plants was injected into the subsurface. Effect of Field-Wide Nitrate Injection on Aqueous, Dissolved Sulfide Concentrations at Production Wells. Sulfide is present in the aqueous phase as H2S and HS(assuming near neutral pH) and in the oil and gas phases as H2S only. Suspended metal sulfides, such as FeS, can also be present. However, souring is by definition the increase in free sulfide concentration with time and is most often measured semiquantitatively by recording the H2S concentrations in the gas phase. This was done by field personnel using sulfide gas detector tubes. However, aqueous sulfide concentrations, which are directly impacted by nitrate, are most relevant, and were determined in this study. All samples were centrifuged prior to analysis to remove precipitated sulfides. Data obtained for produced waters from the 1-L bottles were in good agreement with those from the 250-mL bottles frozen on site. This indicated that negligible sulfide loss in the former occurred due to oxidation during transport of samples to the laboratory. Aqueous sulfide concentrations were, therefore, routinely determined for samples removed

FIGURE 3. (A, D) Sulfide, (B, E) sulfate, and (C, F) nitrite concentrations as a function of time for two production wells with nitrate breakthrough. (A-C) Control of sulfide at production well 13-PW due to batch-wise nitrate treatment at injection well 14-IW from weeks 33 to 101 (Figure 2B; Figure S4). (D-F) Control of sulfide at production well 7-PW due to field-wide, week-long nitrate pulses from weeks 64 to 96 (Figure 2). from the 1-L bottles. The changes of sulfide concentration with time, measured in produced waters from 15 production wells for a 2-year period, are shown in Figures 3 and 4, and Figure S3 in the SI. Prior to nitrate injection, the aqueous sulfide concentrations ranged from 0.01 mM in 7-PW to 0.52 mM in 13-PW (week 0 in Figure 3D and A, respectively). As noted also elsewhere (17), a decrease in sulfide concentration of on average 70% was observed in 11 of 11 production wells in the first 5-7 weeks of the 150 ppm and 110 ppm nitrate injection at 1-WP and 17-WP, respectively (Figures 3A, D and 4A; Figure S3). This decrease was also observed at 15-PW (Figure S3: V) even though no nitrate was injected at the nearest water plant 20-WP during this period. Hence, subsurface flows of injection water are not restricted to the suggested water injection zones (Figure 1 top) and nitrate likely reached 15PW through water injected on the east side of the river. Following the 70% decline sulfide levels increased to values close to those prior to nitrate injection in many production wells from weeks 7 to 20 (Figures 3 and 4, Figure S3). Effect of Batch-Wise Nitrate Applications at an Individual Injection Well. Following a drop in weeks 1-7, the aqueous sulfide concentration at 13-PW rebounded to 0.4 mM (Figure 3A). Because the gas phase H2S concentration at 13-PW was also among the highest (not shown), additional treatment through weekly batch-wise nitrate applications at injector 14-IW, which is most likely connected to 13-PW (Figure S4), was started in week 33. The batch-wise nitrate applications caused a zigzag pattern of injected nitrate concentration at 14-IW (Figure 2B). The actual peak nitrate concentrations were considerably higher than represented. The average water injection rate at 14-IW was 150 m3/day (6.2 m3/h). Injection of 1 m3 of 45% Ca(NO3)2 in 1 h, when mixed with 6.2 m3 of injection water would give a peak concentration of 47 000 ppm (760 mM) of nitrate. Batchwise application of such high concentrations at 13-PW led to

elimination of the aqueous sulfide concentration from week 40 onward (Figure 3A). This was associated with an increase in the concentration of sulfate in the produced water to 0.2 mM (Figure 3B), as well as with breakthrough of nitrate (not shown) and of up to 0.6 mM nitrite (Figure 3C). Stoppage of the batch-wise nitrate injections in week 101 gave recovery of sulfide concentrations at 13-PW to 0.12 and 0.23 mM in weeks 108 and 112 (Figure 3A), respectively. This provides further evidence for the effectiveness of batch-wise nitrate injections in reducing sulfide levels at 13-PW. Monitoring of production wells 16-, 18-, and 19-PW, which were also in close vicinity to 14-IW (Figure S4), was started in weeks 35-48 to evaluate whether they were similarly affected by the batch-wise nitrate application at 14-IW. Production well 13-PW was unique in its strong response to the batch-wise treatment. Decreases in sulfide concentration at the other three production wells were less strong (Figure S3: Y, AB, AE) and no nitrate or nitrite breakthrough was observed at these other production wells (results not shown). Effect of Field-Wide Week-Long Nitrate Pulsing on Sulfide Concentrations at Individual Production Wells. As discussed below, the initial decrease in sulfide concentrations followed by recovery indicated emergence of possible microbial zonation in which NRB are located in close vicinity to the injection wells, whereas SRB grow in more distant zones. The successful batch-wise treatment at 13-PW suggested that this zonation might be broken by applying pulses of high nitrate concentration. This strategy was tested fieldwide by injecting high (1000 ppm) or low (200 ppm) calcium nitrate at the water plants in alternate weeks from weeks 64 to 96. At production well 7-PW sulfide concentrations below the detection limit and nitrite breakthrough were observed from week 70 onward, following the start of nitrate pulsing at week 64 (Figure 3D, F). Significant sulfate concentrations were found in water produced by 7-PW when nitrate injection VOL. 43, NO. 24, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

9515

FIGURE 4. (A) Sulfide concentration (Csulfide in mM), (B) water production (m3/day) and (C) sulfide production (mol/day) at production well 2-PW. (D-F) Averages and sums for 11 production wells for which data were collected since May 7, 2007 (Table S2). (D) Production water volume-weighted, average sulfide concentration, (E) total water production, and (F) total sulfide production. was first started and during subsequent nitrate pulsing (Figure 3E). However, similarly strong effects of nitrate pulsing in weeks 64-96 were not observed at other production wells. Following the initial decrease, production well 2-PW showed an increase in sulfide concentration from weeks 7 to 60, a decrease from weeks 60 to 90, and an increase from weeks 90 to 112 (Figure 4A). Production of water and sulfide at 2-PW increased significantly over the 2-year period (Figure 4B, C). Aqueous sulfide production reached 100 mol/day at week 102, approximately half the daily output of 11 production wells (Figure 4F). At production well 3-PW the sulfide concentration was near zero from weeks 7 to 20 and from week 92 onward (Figure S3: A). At production well 12-PW sulfide concentrations were significantly lower than the week 0 value of 0.08 mM from weeks 3 to 20 and weeks 90 to 112 (Figure S3: S). No breakthrough of nitrate or nitrite was observed at 2-, 3-, or 12-PW. Sulfate concentrations, as determined by HPLC, were generally near zero, except during the initial 7 weeks, when sulfide concentrations decreased strongly (not shown). A survey of data obtained for 9 other production wells is provided in Figure S3. Effect of Field-Wide Nitrate Pulsing on Well-Averaged Sulfide. Because both the volume of produced water and the aqueous sulfide concentration differed substantially among individual wells, the average sulfide concentration was calculated as a volume weighted average, as shown in Figure 4D for 11 production wells for which data were collected since May 7, 2007 (Table S2). To prevent the calculated averages from being influenced by missing data from temporarily shut-down wells, these were obtained by in9516

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 24, 2009

terpolation from data obtained before and after the shutdown period. Total produced water (m3/day) and total produced sulfide (mol/day) are also shown as a function of time (Figure 4E, F). The observed patterns are dominated by 2-PW, which contributed nearly half of the produced sulfide toward the end of the 2-year period (Figure 4C, F). A decline in the average sulfide concentration from 0.130 to 0.04 mM was observed in weeks 1-7, when nitrate was first applied (Figure 4D). This was followed by an increase to 0.150 mM from weeks 7 to 60, a decrease from weeks 70 to 90, and an increase to 0.160 mM from weeks 90 to 112. The decrease and subsequent increase in average sulfide concentration from weeks 70 to 112 correlates with the start and stoppage of nitrate pulsing in weeks 64 and 96, respectively. The increase in total sulfide production (Figure 4F) can be credited largely to the increase in injected water volume with time (Figure 4E).

Discussion Field-wide injection of nitrate is particularly suited to diminish the production of sulfide in oil fields down-hole. Aboveground other options, such as the use of H2S scavengers or biocides, are available. However, the active organic compounds used for this purpose generally do not penetrate well into the reservoir, preventing their use down-hole. Instead, nitrate and its derivative nitrite track the injection water well, allowing them to emerge in production wells, if an excess is injected. Field-wide nitrate injection has been proven to work for fields with a high reservoir temperature, producing light oil (10), likely because microbial activity is

FIGURE 5. Proposed microbial zonation due to nitrate injection in the Enermark field. The indicated concentrations are for the aqueous phase. (Top) In the absence of injected nitrate, SRB live close to the injection wellbore due to sulfate limitation. (Bottom) Continuous injection of nitrate gives rise to microbial zonation in which NRB live in close vicinity to and SRB live further away from the injection wellbore. Zonation gives sulfide production similar to that without nitrate injection. The zonation can be broken by sudden increases (pulses, batches) of the injected nitrate concentration, allowing nitrate to re-enter the SRB dominated zone. confined to the near injection wellbore region due to hydrocarbon toxicity and the high temperature elsewhere in the reservoir. The Enermark field produces a heavy oil, which contributes less water-soluble oil organics with associated hydrocarbon toxicity than is contributed by conventional fields producing light oil (11). It also has a downhole temperature of ∼30 °C throughout. Although microbial activity may thus be envisaged throughout the reservoir, it is likely that SRB activity is also limited to the injection wellbore region due to sulfate-limitation. The injection water contains only ∼1 mM sulfate and, prior to nitrate injection, no sulfate is produced. The average aqueous dissolved sulfide concentration of produced water was ∼0.1 mM (Figure 4D). A diagram explaining the results obtained is depicted in Figure 5 (top). Here the reservoir has been divided into blocks (A) to (D) from injection to producing well. Injected water and produced oil move from (A) to (D) as indicated. Oil organics (e.g., VFA or toluene) dissolve into the water phase throughout and are in excess over injected electron acceptor. In the model, SRB quantitatively convert sulfate (1 mM) to sulfide in block (A). The produced sulfide is transferred to blocks (B), (C), and (D), decreasing in concentration due to distribution over oil and gas phases and due to precipitation as iron sulfide (2, 3), causing 0.1 mM sulfide to be produced. When nitrate is injected, NRB form nitrite in the zone where SRB reside giving a strong reduction of sulfide production in the first 5-7 weeks (Figures 3 and 4; Figure S3). This also corresponded with a strong decrease in gas phase H2S (18). The lower concentrations of sulfide in produced waters correlate with increased sulfate concentrations (Figure 3B, E). However, once SRB have grown back deeper in the reservoir, the sulfide output is similar to that prior to nitrate injection (Figure 5, bottom). NRB are believed to prevent SRB action by competitive exclusion (4), in which hNRB preferentially use organic electron donors for nitrate reduction, preventing their use by SRB. However, this mechanism requires that the two groups grow in the same near-injection wellbore region. This may not be the case in low temperature fields, as shown

here. The resulting zonation (Figure 5, bottom) causes most of the injected nitrate to be used to oxidize available oil organics, not to oxidize sulfide or inhibit SRB action, as intended. Hence no benefit of continuous nitrate injection may be achieved once a zoned microbial community has formed, as in Figure 5 (bottom). We have shown here that this zonation can be broken by adding pulses of nitrate, either at individual wells or field-wide. The suddenly increased nitrate concentration can, apparently, not be respired by the NRB population residing near the injection wellbore and nitrate and nitrite penetrate in the SRB-dominated zone, inhibiting sulfide output once again. Nitrate pulsing led to nitrate and nitrite breakthrough in two production wells (Figure 3C, F). Increased concentrations of sulfate were also observed, indicating oxidation of sulfide with nitrate and supporting our conclusion that the injected nitrate pulses and batches were targeting SRB-inhabited zones. In conclusion, it appears that for low temperature fields, injected with water with a low sulfate concentration, the increase in souring can be arrested by nitrate injection (Figure 4D). A pulsing strategy appears to give better results than continuous field-wide injection of the same concentration. The optimal pulsing frequency and peak concentration remain to be determined. Batch-wise treatment with a high concentration for a short time (e.g., 1 h per week of 40,000 ppm of nitrate) during a low level (e.g., 200 ppm) continuous application may be most promising, providing the field-wide injection system can handle such high concentrations. Overall such a strategy would approximately double the quantity of injected nitrate and associated costs. However, our work has shown this strategy to be more promising than, for instance, raising the continuously injected concentration to 400 ppm. Determining which wells are the high sulfide producers (e.g., Figure 4C: 2-PW) and concentrating pulsed treatment on injectors that most likely service these wells may be another promising method for achieving successful sulfide oxidation with nitrate in low temperature fields. VOL. 43, NO. 24, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

9517

Acknowledgments This work was supported by a Strategic Grant from the Natural Sciences and Engineering Research Council (NSERC) of Canada to G.V. with financial contributions from ConocoPhillips Company and Baker Hughes Incorporated, as well as by an NSERC Industrial Research Chair Award to G.V. The latter was also supported by Baker Hughes Incorporated, Commercial Microbiology Limited (Intertek), the Computer Modelling Group Limited, ConocoPhillips Company, YPF SA, Aramco Services, Shell Canada Limited, Suncor Energy Developments Inc., and Yara International ASA, as well as by the Alberta Energy Research Institute. We are indebted to many others who contributed to this project including Sabrina Cornish, Shawna Johnston, Johanna Voordouw, and Rhonda Clark from the University of Calgary, Pat Stadnicki and Doug Irwin from Enerplus, as well as Rob Mather, Dan Johnson, and Mike McWilliams from Baker Hughes Incorporated.

Supporting Information Available Four figures and two tables with information on loss of nitrate and sulfate during pipeline transport, nitrite in injection waters, sulfide concentrations in producing wells as a function of time, field details for injector 14-IW, characteristics of oil field waters, and a survey of sampling sites. This material is available free of charge via the Internet at http://pubs.acs.org.

Literature Cited (1) Planckaert, M. Oil reservoirs and oil production. In Petroleum Microbiology; Ollivier, B., Magot, M., Eds.; ASM Press: Washington, DC, 2005; pp 3-19. (2) Lin, S.; Krause, F.; Voordouw, G. Transformation of iron sulfide to greigite by nitrite produced by oil field bacteria. Appl. Microbiol. Biotechnol. 2009, 83, 369–376. (3) Vance I.; Thrasher, D. R. Reservoir souring: mechanisms and prevention. In Petroleum Microbiology; Ollivier, B., Magot, M., Eds.; ASM Press: Washington, DC, 2005; pp 123-142. (4) Sandbeck, K.; Hitzman, D. O. Biocompetitive exclusion technology: a field system to control reservoir souring and increase production. In Proceedings of the 5th International Conference on Microbial Enhanced Oil Recovery and Related Biotechnology for Solving Environmental Problems; Bryant, R., Sublette, K. L., Eds.; National Technical and Information Services: Springfield, VA, 1995; pp 311-320. (5) Telang, A. J.; Ebert, S.; Foght, J. M.; Westlake, D. W. S.; Jenneman, G. E.; Gevertz, D.; Voordouw, G. The effect of nitrate injection on the microbial community in an oil field as monitored by reverse sample genome probing. Appl. Environ. Microbiol. 1997, 63, 1785–1793.

9518

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 24, 2009

(6) Gevertz, D.; Telang, A. J.; Voordouw, G.; Jenneman, G. E. Isolation and characterization of strains CVO and FWKO B, two novel nitrate-reducing, sulfide-oxidizing bacteria isolated from oil field brine. Appl. Environ. Microbiol. 2000, 66, 2491–2501. (7) Hubert, C.; Voordouw, G. Oil field souring control by nitratereducing Sulfurospirillum spp. that outcompete sulfate-reducing bacteria for organic electron donors. Appl. Environ. Microbiol. 2007, 73, 2644–2652. (8) Kaster, K. M.; Grigoriyan, A.; Jenneman, G.; Voordouw, G. Effect of nitrate and nitrite on two thermophilic, sulfate-reducing enrichments from an oil field in the North Sea. Appl. Microbiol. Biotechnol. 2007, 75, 195–203. (9) Myhr, S.; Lillebo, B. L. P.; Sunde, E.; Beeder, J.; Torsvik, T. Inhibition of microbial H2S production in an oil reservoir model column by nitrate injection. Appl. Microbiol. Biotechnol. 2002, 58, 400–408. (10) Sunde E.; Torsvik, T. Microbial control of hydrogen sulfide production in oil reservoirs. In Petroleum Microbiology; Ollivier, B., Magot, M., Eds.; ASM Press: Washington, DC, 2005; pp 201213. (11) Lambo, A. J.; Noke, K.; Larter, S. R.; Voordouw, G. Competitive, microbially mediated reduction of nitrate with sulfide and aromatic oil components in a low temperature, western Canadian oil reservoir. Environ. Sci. Technol. 2008, 42, 8941– 8946. (12) Tru ¨ per, H. G.; Schlegel, H. G. Sulphur metabolism in Thiorhodeaceae. I. Quantitative measurements on growing cells of Chromatium okenii. Antonie van Leeuwenhoek 1964, 30, 225– 238. (13) Cypionka, H.; Pfennig, N. Growth yield of Desulfotomaculum orientis with hydrogen in chemostat culture. Arch. Microbiol. 1986, 143, 396–399. (14) Grigoryan, A. A.; Cornish, S. L.; Buziak, B.; Lin, S.; Cavallaro, A.; Arensdorf, J. J.; Voordouw, G. Competitive oxidation of volatile fatty acids by sulfate- and nitrate-reducing bacteria from an oil field in Argentina. Appl. Environ. Microbiol. 2008, 74, 4324– 4335. (15) American Public Health Association. Standard Methods for the Examination of Wastewater; American Water Works Association and Water Pollution Control Federation: Washington, DC, 1992; pp 439-440. (16) Arensdorf, J. J.; Cowan, J. Methods for the rapid enumeration of live bacteria in oilfield systems. In Proceedings of the 15th International Oil Field Chemistry Symposium; Geilo, Norway, March 22-25, 2004, paper 6. (17) Grigoryan, A.; Lambo, A.; Lin, S.; Cornish, S. L.; Jack, T. R.; Voordouw, G. Souring remediation by field-wide nitrate injection in an Alberta oil field. J. Can. Petrol. Technol. 2009, 48, 58–61. (18) Arensdorf, J. J.; Miner, K.; Ertmoed, R.; Clay, B.; Stadnicki, P.; Voordouw, G. Mitigation of reservoir souring by nitrate in a produced water re-injection system in Alberta. In Proceedings of the SPE International Symposium on Oilfield Chemistry, Houston, TX, April 20-22, 2009; Paper 121731.

ES902211J