Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Appendix G Saline Spills INTRODUCTION Saline water associated with North Slope oil production comes from water produced with the oil or from seawater used for enhanced oil recovery. The produced water is clas- sified as wastewater and injected into Class I and II waste- water wells. Drilling fluids and cuttings generated by drill- ing and associated wastes derived from processing facilities are also injected into these wells (Maxim and Niebo 2001a). Seawater has been used in relatively large volumes since 1984 when the Prudhoe Bay waterflood project began. This is a field-wide enhanced oil recovery system that includes facilities to extract and treat water from the Beaufort Sea and then inject it into injection wells. The injected water main- tains pressure within the oil reservoir and flushes oil toward recovery wells (Maxim and Niebo 2001a). When this project began it was estimated to enable the recovery of an addi- tional billion barrels from the Prudhoe Bay oil fields (ARCO Alaska 1984~. Seawater is also used other purposes, such as testing pipelines for leaks. Produced water is considered saline, though salinity is highly variable, depending on the field and the amount of seawater injected into the oil-bearing strata (Maxim and Niebo 2001a). Spills of produced water may occur at the wellhead, along pipelines, and at central processing facilities. They may also come from leaking tanks or, in the past, leaking reserve pits. Reserve pits have been phased out in recent years, and they are being dewatered and restored. A recent progress report on the ADEC reserve pit closure program states that as of mid-January 2002,184 of 329 reserve pits on the North Slope (56%) have been closed, restored, and approved by ADEC (J. Peterson, ADEC, unpublished material, 2002~. Judd Peterson, ADEC Reserve Pit Coordinator, stated that plans to rehabilitate remaining reserve pits were due in his office on January 28, 2002. He estimates that completion of 228 remaining pit restorations through closure and ADEC ap- proval will take 6 to 8 years. Mr. Peterson also stated that ADEC requires water sampling adjacent to all remaining reserve pits. According to these data, no substances are be- ing leached from these pits that exceed state water quality standards. Some reserve pits contain diesel-based drilling fluids and produce a visible sheen when muds are disturbed. ADEC requires that these be excavated, dewatered, and backfilled on an accelerated timetable even if the diesel can- not be detected in samples. Accelerated restoration is also required for pits subject to erosion and possible contamina- tion of Beaufort Sea waters. Four such sites are currently being restored, and a fifth site is scheduled for restoration in 2003 (J. Peterson, ADEC, personal communication, 2002~. Seawater spills from the enhanced oil recovery process can occur at the seawater extraction plant, the seawater treat- ment plant, holding tanks, along pipelines, and at seawater injection wells. Less common sources include fire control systems, compressors, pig launchers, and meltwater. Causes of saline water spills along the pipeline include leaking valves, pump failures, leaking pipes, leaking tanks and drums, transfer hoses, o-ring and seal failures, leaking vehicles, and human error. In the past, leaking reserve pits were also a cause of spills. SPILL DATA Maxim and Niebo (2001a) examined water spill data from an unpublished portion of the TAPS oil spill database developed by IT Corporation. This spill database contains information on spills of crude oil, refined petroleum prod- ucts, water, and other substances from 1977 to 1999. The database covers exploration and production activities on the North Slope and the entire Trans-Alaska Pipeline. The crude oil and products spills data were presented in TAPS Owners (2001) environmental report. There was some difficulty dis- tinguishing the saline water spills from freshwater spills.
APPENDIX G Maxim and Niebo (2001a) compiled spills listed as wa- ter, produced water, seawater, wash water, meltwater, gelled water (seawater mixed with chemical enhancer to thicken it used in enhanced oil recovery), and chemical mixtures. Any spill record that referred to seawater, produced water, or gelled water was considered to be a saline spill for pur- poses of the analysis. There was no way to separate out the low salinity from the higher salinity (seawater) spills. Some spills did not contain enough information to identify the material spilled, and those were termed unclassified spills. There were 17 unclassified spills between 1977 and 1985. These were excluded from the analysis. Together, they ac- counted for 0.9% of the total spill volume. In addition, spill records for that period were less complete, and reporting appeared to be less rigorous than it has been subsequently. Therefore, the detailed analysis covers only the period from 1986 through 1999. Three spills during this period were water mixed with crude oil. They were considered in the oil spill section; only the water portion was considered in the analysis of saline water spill data. Over the period 1986-1999, there were 929 seawater spills associated with North Slope exploration and produc- tion and the North Slope portion of TAPS. Total amount spilled was 40,849 bbl (1,715,658 gallons). This averages out to 66 spills per year over the period and an annual spill volume of 2,918 bbl (122,556 gallons). (See Table G-1.) Analyses of the TAPS oil spill data have normalized spills to the amount of oil transported (Maxim and Niebo 2001b). This is appropriate for oil spills in establishing time trends, but may not be the best choice when normalizing TABLE G-1 Number and Volume of Saline, Freshwater, And Unclassified Water Spills on the North Slope 229 saline water spill data. Comparing water spilled with the amount of crude oil produced suggests only how well water is being handled in relation to the amount of oil handled, and it may mask inefficiencies in the ANS water handling sys- tem. A more useful analysis may be comparing the amount of water handled on the North Slope with the amount of water spilled. Seawater used in the enhanced oil recovery process ac- counted for the vast majority of water used on the North Slope during the period. Annual data on the amount of pro- duced water and water used for enhanced oil recovery is available from the Alaska Oil and Gas Conservation Com- mission (AOGCC) (McMains, personal communication, 2001~. These data were used to calculate the volumetric spill rate (VSR), measured in barrels of water spilled per million barrels of wastewater handled (bbls/million bbls). Table G-2 lists the annual water handled, the annual brine spill vol- umes, and the calculated VSR. Figure G-1 (3) plots the VSR based on volume of water handled (solid line) as well as the volume of oil transported through TAPS (dotted line). The lines match until 1990 when the volume of water handled increased while the amount of oil transported began to decrease. Over the period from 1986 through 1999, the average VSR for saline spills based on water handled was 3.3 bbl/ million bbl of water handled. If based on the volume of oil transported, the VSR is 5.4 bbl/million bbl transported (Maxim and Niebo 2001b). As is the case with oil spills, there is substantial annual variability in the VSR for saline water spills from 0.25 to 17.85 bbl/million bbl. "Bad" years are the result of a rela- tively few large spills and "good" years result from the lack of large spills, not spill numbers. The years 1997 and 1991 TABLE G-2 Alaska North Slope E&P Saline Water Spill Rates (1986-1999) Unclassified Saline Water Freshwater Water Spill Volume of Annual Spill Rate Volume Water Handled (bbls spilled/ Year no. vol. (bbl) no. vol. (bbl) no. vol. (bbl) Year (bbls) (bbls) million bbls handled) 1986 18 955 8 26,923 16 160 1986 955 588,243,485 1.623 1987 20 177 13 19,758 20 17 1987 177 689,765,315 0.257 1988 52 1,098 15 55 39 45 1988 1,098 726,675,694 1.511 1989 104 3,336 24 231 41 122 1989 3,336 801,407,354 4.163 1990 139 772 36 117 24 17 1990 772 845,450,781 0.914 1991 132 9,295 36 227 25 168 1991 9,295 894,098,366 10.395 1992 80 505 37 227 38 16 1992 505 983,579,753 0.514 1993 73 575 35 52 11 7 1993 575 1,038,007,615 0.554 1994 63 1,728 44 95 12 3 1994 1,728 997,105,134 1.733 1995 63 1,057 21 216 14 52 1995 1,057 1,001,078,993 1.055 1996 56 652 16 32 8 56 1996 652 1,000,648,796 0.651 1997 52 18,407 21 60 17 71 1997 18,407 1,031,291,327 17.849 1998 41 1,910 39 144 8 16 1998 1,910 1,004,600,076 1.901 1999 36 383 26 19 25 58 1999 383 671,552,213 0.571 Totals 929 40,850 371 48,156 298 808 Total 90,290 12,273,505,602 3.30 SOURCE: Modified from Niebo 2001a. SOURCE: Modified from Maxim and Niebo 2001a.
230 100- `~, ~ 10 cd 0 = c Q As U) ~ ° lis \ g Bbls/MM bbls ANS water handled - - - - - Bbls/MM bbls TAPS throughput art, \=_r , ~ \ \ \ \ 0.1~ 1 1 1 1 1 1 1986 1988 1990 1992 1994 1996 1998 Year FIGURE G-1 Saline water VSR on the North Slope, 1986-1999. SOURCE: Reprinted with authors' permission from Maxim and Niebo 2001a. have the highest spill rates (VSRs). In 1997, 18,040 bbl (757,680 gal) of freshwater and diluted seawater came to the surface around several wells. A large spill occurred in 1991 when a valve failed and 8,500 bbl (357,000 gal) of produced water were spilled at Central Processing Facility 2. The 20 largest saline water spills from North Slope op- erations during the 1986-1999 period are listed in Table G- 3. These range in volume from 210 to 18,040 bbl (8,820 to 757,680 gal). Combined, they account for 85% of the total saline water spill volume. APPENDIX G The volume of reported saline water spills range from 0.0024 bbl (approximately 1.6 cups) to 18,040 bbl (757,680 gal). As with oil spills, small spills are frequent, large spills are rare, and the total volume is dominated by the few large spills. A Lorenz diagram (Figure G-2) provides a useful de- piction of these spills. The fraction of spill volume is plotted against the fraction of spills. If all spills were of equal size, the plot would be a straight line. There is substantial curva- ture in the plot, and the computed Lorenz coefficient is 0.97. It is clear that the relatively few large spills account for the most of the spill volume. In fact, 50% of North Slope saline water spills were less than 0.95 bbls (40 gal), and the small- est 90% of those spills accounted for approximately 3.9% of the total volume; the smallest 95% accounts for approxi- mately 8.2% of the total volume spilled (Maxim and Niebo 2001a). EFFECTS OF SALINE WATER SPILLS A study that anticipated potential spills associated with the Prudhoe Bay Waterflood project was sponsored by the Army Corps of Engineering Cold Regions Research and Engineering Lab (CRREL) (Simmons et al. 1983~. The pur- pose was to evaluate the sensitivities of different tundra plant communities to seawater spills. Eight sites representing the range of vegetation types along the pipeline route were treated with single, saturating applications of seawater dur- ing the summer of 1980. Each site was examined prior to the experimental spills, monitored closely for 28 days, and vis- TABLE G-3 Twenty Largest Saline Water Spills from North Slope Operations Volume (bbls) Description 1 4 5 6 7 9 10 11 12 13 14 15 16 17 18 19 20 17 Mar 97 01 Jun 91 16 Dec 89 10 Jan 98 29 Sep 86 28 Jul 89 31 Oct 94 22 Dec 89 25 Jun 92 07 Mar 94 14 Apr 94 15 Mar 95 07 Nov 95 08 Feb 88 30 Mar 88 17 May 88 06 Oct 88 01 Nov 96 12 Dec 91 07 Jan 96 18,040 8,500 1,500 1,500 500 Freshwater and diluted seawater surfaced around nine wells at Drillsite 4. Valve failed and leaked produced water at Flowstation 2. Pipeline weld failed, leaking seawater from a seawater injection line along Oliktock Road. Pipeline leak spilled produced water. Fiberglass bypass line on heat exchanger failed, spilling seawater to secondary containment at Seawater Treatment Plant. 500 Flowstation 2 actuator bonnet failed spilling produced water. 385 Crack in pipeline lY-lR spilled mixture of crude oil and produced water. Only the volume of spilled water is given for this spill. 355 350 320 310 300 300 287 281 250 250 231 230 210 Pipeline leaked seawater to drill pad. Corrosion suspected. Water tank overflowed seawater to a sump when valve leaked at CPF-3. Seawater spilled to drill pad at Prudhoe. Seawater valve in pig launcher module leaked into pigging pit and overflowed. Produced water was released to a drill pad and reserve pit when equipment failure caused pressure change in . . plpelme. Seawater spilled onto drill pad during equipment malfunction. Seawater injection line bled water from a pipe rack on a drill pad. Corrosion caused a produced water leak in a pipeline. Solonoid on seawater line at seawater treatment plant failed. Produced water injection line at drillsite failed due to corrosion. Leak in seawater line at seawater injection plant. Rod failed on seawater pump at central processing plant. Produced water spilled from pig launcher after an ice plug melted out of a partially open valve. SOURCE: Modified from Maxim and Niebo 2001a.
APPENDIX G 1.0 - . . 0.8- 0.6- 0.4 - 0.2 - 0.0 - 0.0 0.2 Actual distribution of saline spills ---- - Curve with all spills the same size 1 0.4 0.6 Fraction of Spills 0.8 1.0 FIGURE G-2 Lorenz diagram of North Slope saline water spills. SOURCE: Reprinted with authors' permission from Maxim and Niebo 2001a. ited less frequently over the following year. Symptoms of physiological stress were observed 8 days after the experi- mental spill. Within 12 days, 17 taxa of vascular plants de- veloped physiological stress attributable to the treatment, ranging from slight chlorosis to total browning and desicca- tion of all the plants foliage. The impact of seawater treat- ment was most severe in the mesic and dry sites. The wet sites were less severely affected. Within a month of treat- ment, 30 of 37 taxa of shrubs and fortes in the experimental plots developed definite symptoms of stress, while the 14 graminoid taxa did not exhibit adverse effects. Live vascular plant cover was reduced by 89% and 91% in the two dry sites and by 54%, 74%, and 83% in the three moist sites. Mosses were unaffected in all but one of the experimental sites. Two species of foliose lichens showed deterioration, while other lichen species were not affected. The absorption and retention of salts by soils is inversely related to soil moisture. In the wet sites, conductivities reached prespill lev- els in approximately 30 days. Salts were retained in soils at the dry sites, concentrating at or near the seasonal thaw line. Soil enzyme and microfloral activity was reduced for up to one year after treatment (Simmons et al. 1983~. In December 1982 approximately 400 bbl (16,800 gal) of concentrated sodium chloride leaked from a damaged stor- age tank at a drillsite in the Kuparuk oil field. The spilled brine spread onto tundra adjacent to the drill pad, covering an area of approximately 0.3 ha (0.7 acre) before freezing (Baker 1985~. Soils and vegetation were studied for two years. By the 1983 growing season all plants within 40 to 60 m (130 to 200 ft) of the center of the spill were dead, and up to 30 m (100 ft) beyond the dead vegetation, many plants showed signs of physiological stress. In July 1983, the size of the affected area was approximately 1.6 ha (4 acres). In the fall of 1983 a road was constructed across the western side of the site, altering the local drainage and forming an impoundment in the spring of 1984. By July 1984, the size 231 of the affected area had increased to approximately 4.5 ha (11 acres). Vegetation was dead within 90 to 140 m (300 to 460 ft) of the spill center, and signs of stress were found in plants 40 to 60 m (130 to 200 ft) beyond the dead vegetation (Baker 1985~. The study was continued and expanded to map and quantify vegetation types, examine the effects of the spill on thaw depths, and document salinity levels in soil and water bodies in the vicinity. Unaffected tundra near the spill zone was used as a reference area. The area affected by the spill was divided into high- and moderate-impact zones. Forbs and shrubs were most se- verely affected, graminoids and cryptogams less severely. Some recovery had occurred by 1984; the size of the high- impact zone was decreasing. There was vigorous growth of the sedge species Eriphorum angustifolium in the moderate- impact zone (Jorgenson et al. 1987~. In June 2000, approximately 1,200 to 1,500 gallons (28.6 to 35.7 bbls) of low-salinity seawater (Electrical Con- ductivity = 5,020 ,umhos/cm; salinity = 3.5 ppt) was leaked from the Alpine oil pipeline during hydrostatic testing. The tundra at this site is moist tussok and wet sedge. The soil active layer was only partially thawed at the time of the spill. The moist tundra was thawed deeper than the wet sedge meadow. The spill site was studied by J. D. McKendrick (2000a). As in the previous study, the moist (tussok) com- munity was affected to a greater degree than the wet (sedge) community. Elevated salt levels in surface water bodies were found as far as 58 ft (18 m) from the leak. The area of veg- etation damaged by the leak was 6.25 ft2 (0.1 m2~. In this area salt levels in soil were elevated. Even though plants lost their leaves, they survived and grew normally during the growing season. McKendrick (2000a) attributes this to the low salinity of the water spilled. In addition, tundra commu- nities close to the coast may be exposed to seawater during storm surges. McKendrick (1997) has examined many saline water spill sites, including those treated with fertilizers or flushed with freshwater or calcium nitrate. He noted much variation in recovery based on such things as grazing pressure. Those sites treated by flushing with freshwater or calcium nitrate recovered faster than non-flushed sites. The calcium nitrate flush was no more effective than freshwater alone. Recovery of flushed sites may still take several years; non-flushed sites may take decades, depending on conditions. The effects of saline water spills are related to salinity. As expected, low-salinity water has less severe initial im- pacts and more rapid recovery than higher-salinity water. Recovery from higher-salinity spills may take several years. Effects of saline water spills can be reduced by flushing with freshwater (Walker 1996~. Joyce (M. Joyce, indepen- dent consultant, personal communication, 6/7/2001) reports that the standard response countermeasure in summer is flushing with warm freshwater. In winter, snow berms are constructed to contain the spill and the frozen material is picked up with scrapers.