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.
6 Effects on the Physical Environment Extraction of oil and gas from subsurface deposits in- volves deliberate alterations of the surface and subsurface physical environments, which in turn affect organisms liv- ing on the North Slope. This chapter focuses on the effects of industrial activities on the physical environment. Most of the effects on organisms that are caused by changes in the physical environment are dealt with in chap- ters 7,8, and 9, but some of those effects are mentioned here as well. For example, small areas of vegetation have been contaminated by spills of oil, other petroleum products, and saltwater, and road dust; vegetation has also been damaged by bulldozers, offroad vehicles, and ice roads; and it has been destroyed where it underlies gravel pads and roads, or where it has been removed to make way for gravel mines (Forbes et al. 2001, Jorgenson and Joyce 1994, McKendrick 2000a, Walker 1996~. Alterations in vegetation can affect other organisms on the North Slope. Physical disturbances can affect fish migrations, the movements of caribou, and in the marine environment migration and distribution of animals, especially bowhead whales and fish. Oil-field ac- tivities can affect the number and distribution of predators, which can in turn affect the number and distribution of birds and some mammals. PERMAFROST The climate of Alaska's North Slope is much colder than that of other U.S. oil fields. As a consequence, the ground is permanently frozen to great depths (about 200-650 m, or 660-2,130 ft) (Lachenbruch et al. 1982a,b). In this perma- frost zone natural processes are significantly different from those in unfrozen sediments, and this imposes a wide range of unique constraints on the development of industrial infra- structure and on the preservation of functioning ecosystems (Lachenbruch 2001~. A shallow surface layer, the active layer, thaws in sum- mer to become an extensive wetland even though the climate 64 is arid. The meltwater cannot filter downward through the underlying impervious permafrost. The active layer contains and sustains the living tundra vegetation mat, which in turn sustains the region's diverse populations of land animals and influences landform processes like runoff, erosion, and soil Stowage. Below the active layer, the permafrost generally contains a substantial fraction of ice. The integrity of the surface and any buildings, roads, or pipelines placed on it depends on the presence of that ice for support. If not prop- erly designed and maintained, these structures can destroy themselves by upsetting the natural heat balance and thaw- ing their foundations. Thus, changes in the thermal condi- tion of the surface in permafrost terrain can have widespread effects that accumulate throughout the physical, biotic, and human systems. Many of the physical effects of oil and gas development enumerated in Chapter 4 (from gravel roads, heated struc- tures, offroad traffic, oil spills, gravel mining, oil wells, and pipelines) can trigger environmentally significant changes in the physical behavior of the permafrost and the active layer. The ultimate environmental effects of such develop- ment cannot generally be anticipated without knowledge of the intermediate functioning of permafrost. For exam- ple, road dust can cause deep pools on the tundra surface by collecting solar radiation and thawing the underlying permafrost. The general pattern of the committee's efforts has been to identify effects of the industrial activities described in Chapter 4 on the environment (described in Chapter 3~. For permafrost, however, it is important also to consider recipro- cal effects of the natural environment on development. In fact, the permafrost environment influences the de- sign of the industrial infrastructure in ways that significantly influence the effects of the infrastructure on the environ- ment. Structures must generally be designed to avoid thaw- ing their own foundations. Dealing with this condition es- tablishes the engineering configuration, or architecture, of
EFFECTS ON THE PHYSICAL ENVIRONMENT North Slope oil development, which consists of a conspicu- ous network of 954 km (596 mi) of roadways elevated on thick gravel berms, and 725 km (450 mi) of pipeline clusters elevated on pilings (Table 4-2~. They join more than 2,354 ha (5,817 acres) of thick gravel work pads above which are large heated buildings elevated on pilings and, more recently, closely spaced oil wells that are cooled by extensive refrig- eration equipment to preserve the superficial permafrost rem- nant that supports them. Some effects of this unique assem- blage of thick gravel and associated permafrost-resistant infrastructure are discussed in chapters 7 (plants), 8 (ani- mals), and 9 (people). To understand the importance of permafrost in the evo- lution of the oil and gas infrastructure and the effects of physical development on the North Slope environment, it is useful to examine the controlling thermal processes. In this section we discuss natural and artificial processes relevant to oil and gas development. Most are mentioned again in later chapters in connection with specific effects. Active Layer The active layer that thaws each summer lies between the top of permafrost and the ground surface. It controls the influ- ence of permafrost on surface processes and the influence of human activity on permafrost. The permanently frozen base of the active layer is impermeable to water and impenetrable to roots. Consequently, the active layer is the growth medium for surface plant communities; the reservoir for their water and nutrient supply; the locus of most terrestrial hydrologic activity; and a boundary layer through which heat, moisture, and gases are transferred between permafrost and atmosphere. The active layer protects the permafrost from summer warmth. Its thickness is a measure of the ground's ability to transmit heat. The active layer varies from as little as 20 cm (8 in.) in some areas of peat or poorly drained sphagnum moss to more than 2 m (80 in.) at some well-drained inland gravel sites. The thickness varies with the active layer's properties as well as with the locally variable thermal properties of the permafrost it protects (Lachenbruch 1959~. Disturbance to the surface, whether anthropogenic or natural, can affect the thickness and mechanical nature of the active layer and ultimately the composition of its plant communities. Such disturbances might include disruption of peat and living vegetation, changes in radiation properties, or alterations in the abundance of water. Disturbances can be initiated by off-road vehicular traffic, removal of vegetation, addition of gravel, paving, oil spills, and saltwater spills, deposition of airborne dust, or the modification of surface drainage (Chapter 7~. Ice-Wecige Polygons and Thermokarst Typically, permafrost is cemented by ice that, in the upper few meters at least, occupies more space than the 65 water-filled pores that would remain after thawing. Melting in such ice-rich or thaw-unstable permafrost results in thaw settlement and disruption of the surface and whatever is on it. The ice in permafrost, or ground ice, can be mainly in the pore space or it can be in large segregated masses. Ice con- tent varies greatly from one North Slope location to another and so consequently does the potential for thaw settlement. Ice content increases sharply from Prudhoe Bay to Alpine on the Colville delta (Hazen 1999) and between many loca- tions on the Arctic Coastal Plain and in the Brooks Range foothills. The most troublesome disturbances are those for which a thickening of the active layer is not self-arresting; that is the thawed increment from permafrost flows off as a slurry rather than remaining in place to augment the active layer' s insulation. This unstable process, called thermokarst, can lead to deepening pits and trenches, retreating scarps, and mud flows (Lawson 1982~. It occurs commonly on the North Slope where the active layer is disturbed over ice-wedge polygons widespread patterns of troughs shaped like giant mud cracks omnipresent on the coastal plain and much of the foothills (Figure 6-1~. Below each trough is a wedge of almost pure ice a meter (3.3 ft) or more wide and several meters deep (Figures 6-2 and 6-3~. The wedges form over centuries in recurring contraction cracks, opening during the winter and filling with ice each spring by downward perco- lation and refreezing of melted snow (Lachenbruch 1962~. A roadway, heated building, or pipeline that thaws the under- lying permafrost will soon be unsupported in a free span across the melting ice wedges. This thermokarst condition is illustrated by early roads on the North Slope that consisted of gravel laid 1 m (3.3 ft) thick on the tundra (Figure 6-4~. Because the natural active layer in North Slope gravels is about 2 m (6.5 ft), the roads soon became impassable, even to foot travel. Compression by the gravel destroyed the insu- lating capacity of the organic mat, and the too-thin gravel disappeared into a network of deepening trenches left by thawing ice wedges (Ferrians et al. 1969, Lachenbruch 1966~. Heatecl Builclings On the North Slope the typical result of a heated build- ing constructed on the ground surface is a surface patch the shape of the foundation maintained at a temperature 25 °C (45 °F) or so above the ambient mean temperature of ap- proximately -10 °C (14 °F). The shape of the thawed basin that grows in the permafrost under a building depends on these two temperatures, the width of the building, and its age, and it is easily estimated (Lachenbruch 1957b). Sub- stantial thawing can develop quickly, leading to destructive thaw settlement, a fact well known to Alaskan cabin dwell- ers. This process leads to a conspicuous constraint on North Slope development heated structures, even very large ones, generally must be elevated above the surface to let the cool
66 CUMULATIVE EFFECTS OF ALASKA NORTH SLOPE OIL AND GAS FIGURE 6-1 Ice-wedge polygons on the Arctic Coastal Plain; troughs are underlain by ice wedges. Peripheral ridges represent material displaced from permafrost by ice-wedge growth. Photograph by George Gryc. FIGURE 6-2 Intersection of three small ice wedges exposed in an undercut bank of Elson Lagoon, near Barrow. Photograph by Gordon Greene.
EFFECTS ON THE PHYSICAL ENVIRONMENT FIGURE 6-3 Upper portion of an ice wedge exposed in a riverbank on the Colville delta near Nuiqsut, Arctic Coastal Plain. Photograph by H. J. Walker. FIGURE 6-4 Roadway destroyed by thawing ice edges near Umiat. Gravel fill was not thick enough to replace insulating effect of the organic mat, which it destroyed. Photograph by Gordon Greene. SOURCE: Lachenbruch 1966. 67
68 air circulate below. The load that can be carried by their pilings depends on the temperature of the enclosing perma- frost. If necessary, building design might be revised to ac- commodate a warming climate by refrigeration with thermo- siphons (e.g., Kinney et al.1983) like those used to maintain the mechanical integrity of closely spaced oil wells. Where heated buildings must be placed on the surface it is generally necessary to insulate and refrigerate their foundations; insu- lation alone only delays thawing. Although heated buildings generally are not built on grade, local warm surface patches from snow drifting against structures, or the above-freezing temperatures of well houses can be significant. If those dis- turbances are superimposed on others they must be consid- ered in thermal design to avoid destructive thermokarst (see the section on "Well Pads" below). Moclifiecl Lakes and Gravel Mines Winters on the North Slope are cold enough to freeze lakes to a depth of about 1.8 m (6 ft). Most lakes and ponds on the North Slope are shallower than that, so they freeze solid to the bottom, and are part of the active layer. How- ever, lakes that are deeper cross an environmental thresh- old the bottom remains unfrozen, and a permanently thawed basin, or talik, grows downward into the permafrost under the lake bed. An inverted dimple of unfrozen ground grows upward from below the base of permafrost. If the lake area is large with respect to the natural depth of permafrost, and if the lake is old enough (thousands of years), the basin and dimple can join to form a thawed hourglass shape through the permafrost. Typically, the mean lake-bottom temperature is 1-2 °C (34-36 °F), or about 10 °C (18 °F) warmer than its surroundings (Brewer 1958b). In this sense deep lakes behave like heated buildings, and the same pre- dictive theory for the thawed basin applies. When such a lake is drained, refreezing of the sediments in the thawed basin beneath it often creates a pingo, a mound formed by ice expansion like the bump formed in the middle of an ice cube as it freezes. If a shallow lake is deepened for a winter water supply, as is done occasionally, for example at Kuparuk, it will gen- erally continue to deepen on its own from thaw settlement as its thawed basin grows. Once the lake no longer freezes to the bottom it works differently in the ecosystem. Gravel mines that become deep lakes behave similarly. Gravel Roacis To prevent thermokarst, the gravel placed under roads must be thicker than its depth of summer thaw to ensure that the subgrade remains frozen as the road crosses a variety of thaw-unstable permafrost materials (Lachenbruch 1959~. On the Arctic Coastal Plain this requires that roads be placed on gravel berms up to 2 m (6.5 ft) above the tundra surface. Like the elevated pipelines, this network of elevated roads CUMULATIVE EFFECTS OF ALASKA NORTH SLOPE OIL AND GAS has a conspicuous visual impact on the landscape. Addition- ally, the continuous berms intercept natural drainage, creat- ing ponds that collect solar radiation, thicken the active layer, and initiate thermokarst (Walker 1996~. Road dust also can perturb the thermal balance of the active layer and underly- ing permafrost. The effects of these processes on vegetation are described in Chapter 7. Pipeline Burial Permafrost poses severe obstacles to pipeline burial, the preferred mode of construction in nonpolar environments. The principal difficulty is that subsurface heat from the trans- mission of warm fluids thaws the surrounding permafrost, causing differential settlement, which strains the pipe. This consequence of permafrost leads to one of the most con- spicuous impacts of oil and gas development on the North Slope a network of elevated pipelines. They can impede free overland travel by subsistence hunters (newer ones are higher to permit passage), and they constitute an imposing visual alteration of the landscape (see Chapters 8 and 9~. The Northstar pipeline is buried in the seabed in a shal- low trench that extends to an artificial island 10 km (6.2 mi) offshore. The pipeline has a planned operating temperature of 29 °C (85 °F); warm oil started to flow through it late in 2001. Special problems are posed by its burial in ice-bearing permafrost in the seabed within 3 km (1.9 mi) of shore (Intec Engineering 1998) heat from the oil will eventually thaw the permafrost and strain the pipe. Although it has been care- fully designed, such an offshore-buried pipeline is without precedent in the Alaskan Arctic; its performance will be instructive. Well Pads and Annular Thawing The layer of gravel used for roads is also generally ad- equate to protect large work pads from seasonal thaw settle- ment. However, different thermal designs may be necessary if there are additional sources of heat, such as the heated foundations of well houses, snow drifts that insulate pads from winter cold, or subsurface sources like heat in a "thawed chimney" the annular region thawed through per- mafrost around a warm production well (see "Effects of Fluid Withdrawals. The chimneys have become much more critical with the recent emphasis on decreasing the footprint using directional drilling from closely spaced wells on small well pads. In the process of drilling a well or extracting oil, natural gas, or formation waters, fluid circulating through a borehole trans- fers heat advectively from warm formations at depth to the colder ones near the surface. The drilling targets are gener- ally at depths where the formation temperatures are 40- 90 °C (104-194 °F). Heat from the warm fluid is conducted radially through the borehole wall (the well casing), quickly thawing annular zones, or chimneys in the surrounding per-
EFFECTS ON THE PHYSICAL ENVIRONMENT mafrost (Lachenbruch et al. 1982a). The radius of the chim- ney is calculated by using properties of the permafrost and borehole specifications. For typical Prudhoe Bay production wells, Perkins and colleagues (1975) estimated a thawed chimney radius of about 2-6 m (6-20 ft) after a year or two and about 6-11 m (20-35 ft) after a decade of production. The smaller values apply near the surface where the perma- frost is colder. During the early Prudhoe Bay development, wells were drilled about 50 m (160 ft) apart (BP 1998a). Even after decades of production, thawed chimneys were relatively unconnected to one another, and most of the permafrost remained intact and able to resist damaging settlement (sometimes aided by insulating the well casings). In the cur- rent design, which is used for pads at the Alpine oil fields, wells at 43 °C (109 °F) are spaced only 3 m (10 ft) apart (Hazen 1999~. Accommodating such a concentration of heat in permafrost requires sophisticated design with extensive refrigeration by passive heat pipes (or thermo-siphons) and insulation. Hazen (1999) calculated that without refrigera- tion the thaw chimneys would coalesce at all depths, and all of the permafrost about 850 ft (280 m) thick under the row of wells would thaw causing the natural surface, gravel pad, and well houses to settle nonuniformly from 6 to 19 ft (1.8 to 5.8 m). With refrigeration to a depth of 50 ft (15 m) and insulated conductor pipe to 80 ft (24 m), Hazen (1999) estimated that the top 40 ft (12 m) of permafrost would re- main intact to form a supporting arch that will deform slowly and smoothly, with only 1 to 2 ft (30 to 60 cm) of thaw settlement at the surface. SUBSURFACE ENVIRONMENT: POSSIBLE EFFECTS OF THE WITHDRAWAL AND INJECTION OF FLUIDS AND OTHER MATERIALS The subsurface physical environment in an oil field con- sists of the layers of sediment and rock and the fluids that naturally fill their fractures and pore space. Potential effects in this environment generally relate to its possible invasion by gas or oil along unintended flow paths through failed oil- well casing and cement seals or through artificially fractured rock. Hydrocarbons gas and oil have the potential to de- grade two principal receptors: tundra surface habitats and subsurface water sources. On the North Slope oil fields the consequences are complicated considerably by the presence of permafrost and to some extent by waste-disposal prac- tices. Problems related to the thawing of permafrost are mainly controlled by fluid withdrawal, the subject of the next section. However, in the North Slope oil fields, fluids are injected into the wells in volumes comparable to those that are withdrawn; they too can have environmental conse- quences for the tundra surface and for subsurface water sources. Injection serves two purposes: It enhances produc- tion by restoring lost pressure in a waning reservoir, and it is used to dispose of drilling mud and other wastes by placing 69 them in previously undisturbed porous rock strata. The first procedure has a long history of worldwide use; the second is relatively new and used most intensively on the North Slope. The first requires neither new flow paths nor injection pres- sures above natural ambient values the second requires both. They are treated separately below. Fluicl Withdrawal and Its Effects The fluids produced by a well are oil, gas, and forma- tion water. In a typical reservoir, the shallowest portion can be filled with gas resting on oil that, in turn, rests on saline- to-brackish, rarely fresh, formation waters. As oil or gas is extracted, formation water moves up into the portion of the reservoir previously filled with oil and gas. Those fluids are at temperatures and pressures that are largely controlled by the natural thermal gradient, the pres- sure gradient, and the depth of burial. The North Slope reser- voirs are "normally pressured"; the rate of pressure increase and the pressure at any given depth are close to the hydro- static gradient 0.427 kg/cm2 per m (0.445 psi/ft). The ini- tial reservoir pressure at Prudhoe Bay was 309 kg/cm2 (4,390 psi) at 2,682 m (8,800 ft). In the shallower Kuparuk River field, the original reservoir pressure was 229 kg/cm2 (3,250 psi) at 1,890 m (6,200 ft). The pressure gradients were 0.48 kg/cm2/m (0.50 psi/ft) and 0.50 kg/cm2/m (0.52 psi/ft), re- spectively, at Prudhoe and Kuparuk. The thermal gradient determines the temperature of the fluids in the reservoir. It affects their viscosity, corrosiveness, and tendency to thaw permafrost. The thermal gradient in the Prudhoe Bay area is variable its average is about 5.6 °C per 100 m (30.7 °F per 1,000 ft). The original tempera- ture of the oil at the Prudhoe Bay field was 97 °C (207 °F) at 2,682 m (8,800 ft); the temperature of the Kuparuk oil was 71 °C (160 °F) at 1,890 m (6,200 ft) (AOGCC 1998~. The withdrawal of subterranean fluids from oil fields in the Arctic causes thawing of the permafrost in the neighbor- hood of the wells. Ground that had been frozen solid loses its rigidity. This results in potential environmental and struc- tural disruption. Thawed chimneys create three potential problems that are specific to oil production through permafrost: · Annular path to the surface. The thawed chimney is more permeable to infiltration by fluids than the same area is before disturbance. To the extent that permafrost might be expected to form a barrier to uncontrolled borehole fluids, the relatively permeable thawed chimney provides a pos- sible path outside the casing for broaching to the tundra sur- face (ARCO/BPIExxon 1997~. Such a fluid path to the sur- face could affect natural plant communities, although no such effects have been observed. · Stress on the well casing. The formation of the thawed chimney leads to two sources of stress on the well casing (Wooley/ARCO/BP 1996~. The first is caused by
70 ~7 ~7 thaw settlement when the annular region loses strength and settles against the casing adding axial (i.e., vertical) drag forces anc' rem al pressures to the pipe. The second is caused by increased radial pressure as ice reforms during freezeback of the chimney after drilling or production ceases. Alter- ations in casing design and cementing procedures seem to have solved problems of casing failure caused by external forces in the permafrost chimney (D. Andrews, Consulting Petroleum Engineer, personal communication, 2001; Perkins et al. 1975~. A fluid with a freezing point below ambient temperature is commonly used inside the casing to prevent the internal forces caused by refreezing. · Surface subsidence. When ice-rich permafrost thaws, its volume generally decreases, leading to subsid- ence of the overlying ground and damage to structures on it. Producing oil wells on the North Slope thaw a chimney through the entire 300-600 m (1,000-2,000 ft) permafrost column beneath the pad that supports those structures. Un- derstanding and controlling the surface effects is impor- tant. If the chimney were only a pinhole through perma- frost, the surface material would not sink far into it before it would be supported by shearing stresses from the chim- ney walls. But if a chimney were wider than its height (thickness of permafrost) the walls could not support the slumping mass, and surface subsidence could be extreme. In the older development areas where wells were drilled 36.5 m (120 ft) apart (BP 1998a), thawed chimneys did not coalesce and most of the permafrost remained intact and able to resist damaging settlement, especially if casings were insulated. However, wells are now being drilled so close together that, without additional mitigating measures, their thawed chimneys would coalesce to cause destructive differential settlement of the pads (see "Well Pads and Annular Thawing". Injection for Enhancecl Recovery Several kinds of wells are used to inject fluids either for enhanced recovery or for disposal of wastes. Those wells must meet specific design requirements to isolate the injected fluids from the surface and to place them in specific hori- zons (Wondzell 2000~. The return of some produced fluids and the introduction of others to the producing formations is designed to improve the recovery of oil from the reservoir, either through main- taining pressure or through increasing the fluidity of the oil. To achieve this, produced fluids, such as formation water and gas, can be re-injected into the original oil-producing reservoirs. Other fluids, such as treated seawater and natural gas or CO2 from other sources, also can be introduced to the reservoir. Enhanced recovery requires that the fluid be beneficial for increasing the ultimate recovery of oil, that it be injected at pressures that will not propagate fractures through the confining zones that protect fresh waters, and that it be CUMULATIVE EFFECTS OF ALASKA NORTH SLOPE OIL AND GAS chemically compatible (for example, not to cause precipita- tion) with the formation water. Formation water and treated seawater are used princi- pally to maintain pressure. The water is placed in the reser- voir by a series of injection wells to "push" the oil toward the producing wells. Natural gas and miscible injectants also are used to maintain pressure, but they have the added ability to increase the fluidity (decrease the viscosity) of the oil and strip it from sand grains (DOE 1999~. The fluids generally are injected at depths below the oil-water contact surface to allow them to sweep through the entire oil column with maxi- mum effect on any remaining mobile oil. Potential environmental consequences would be the risk of escape of fluids to the surface either through fracturing of the overlying stratigraphic section (a highly improbable event) and the permafrost or around the casing through fail- ure of the cement job or the casing itself. This could result in a spill on the tundra. Injection for Waste Disposal Environmentally sound disposal of oil-field wastes has long been a problem. A relatively recent innovation, the down-hole injection of fluid wastes and slurries, has been used for the disposal of large volumes of waste on the North Slope. These drilling by-products and other wastes are now injected into otherwise undisturbed, confined geological for- mations. Down-hole injection eliminated the use of reserve pits for surface storage of drilling waste (Gilders and Cronin 2000), and although it is superior in most respects to older methods, it is not without potential environmental effects. Class I and 11 Wastes Because of the presence of gas and oil, other volatile organic compounds, and metals, Class II (exempt) wastes present a considerably higher risk to the environment than do Class I (nonhazardous wastes), should there be a spill on the surface (API 1987,BP1992~. Such an event might result from a failure in the cement of the well, a pipe collapse, or through a nearby, poorly plugged or monitored, abandoned, or shut-in well. The actual down-hole effects of these fluids are unimportant if they are not injected into an underground source of drinking water (USDW). Grincl and Inject The grind-and-inject process is used for Class II waste disposal of substances associated with reserve pits and drill- ing mud and cuttings from drilling wells. The process in- volves mining materials from the reserve pits, transporting them to a central grinding facility, grinding the solids finely enough to facilitate injection, and injecting them as a slurry into disposal zones (BP and ARCO 1993~. The reserve pit solids contain a wide range of metals and some hydrocar-
EFFECTS ON THE PHYSICAL ENVIRONMENT loons, some of which present a potential hazard. (See Appendix D.) Drilling mud and cuttings from active drilling wells also are ground and injected. More than 42 million barrels (bbl,7 trillion L, 1.8 trillion gal) have been disposed of this way. This method of waste handling has greatly reduced the pos- sibility of environmental damage caused by reserve pit ma- terials. The hazards associated with subsurface injection of these materials are discussed below. Annular Injection Annular disposal is the process of pumping drilling mud and cuttings from drilling operations down the annulus formed when another casing is cemented inside the surface casing. Annular disposal requires porous intervals below the confining zone at the bottom of the surface casing and above the probable top-of-cement depth of the production casing. Only the drilling muds and cutting materials from the drill- ing operations compatible with disposal horizons can be in- jected. The ground and slurried materials are combined with water, if necessary, and pumped down the annulus and in- jected out of the bottom into the disposal formation. The result is a controlled fracturing of the disposal formation that creates more storage space and pushes the particulate mate- rial out into the porous formation. The water tends to sepa- rate from the cuttings and penetrates more deeply into the unit. Alaska Oil and Gas Conservation Commission regula- tions limit the disposal volume to 35,000 bbl (5.7 million L, 1.5 million gal) per well to encourage the construction and use of disposal wells on drill pads and to reduce the possibil- ity of the disposal stream eroding the production casing in the wellhead (Wondzell 2000~. The amount is equivalent to the production of disposable materials from three wells. At least 158 wells have been used for annular disposal of drilling muds and cuttings. A total of 3 million bbl (477 million L, 126 million gal) has been injected at depths of 820-1,340 m (2,690-4,400 ft). The potential hazards are dis- cussed below. Possible Effects of Injection for Waste Disposal Pressure Fracturing and Broaches to the Surface A potential concern is that the pressure required to in- ject wastes into a selected horizon might be enough to frac- ture the confining overburden stratum and allow waste to escape toward the surface. In the enhanced-recovery pro- cess this does not seem to be a serious problem because the injection augments the falling reservoir pressure and it takes place at pressures below the original ambient reser- voir value. By contrast, waste injection is done in previ- ously undisturbed subsurface environments, and it requires pressures above ambient values. Where the waste includes 71 ground rock cuttings, the target reservoir must be fractured to receive the slurry. This requires injection pressures that exceed the ambient by the fracture strength of reservoir rock and poses a substantially greater potential for fracture of the confining overburden than is the case for enhanced recovery. Risk to confinement is much greater if the pres- sure fracture is vertical (not horizontal), an outcome pre- dictable from a knowledge of the formation's stress state (Abou-Sayed et al. 1989~. Pressure profiles for several Prudhoe Bay wells con- firm that the porous disposal formations now have pres- sures of 7-18 kg/cm2 (100-250 psi) above the original hy- drostatic values of about 105-280 kg/cm2 (1,500-4,000 psi). The tests also confirm the integrity of the confining zones. The pressure gradient returns to normal as little as 30 m (100 ft) above the zone of injection. Thus, the escape of injected or in situ fluids through 30 m (100 ft) or so of overburden does not occur, at least at those sites. There is no documented instance of fracturing of the permafrost caused by injection of fluids below the permafrost. How- ever, the annular thawed chimney provides a potential per- meable path to the surface for fluids that might reach the base of the permafrost. Regardless of the steps taken to prevent accidents, there have been some failures of the system, and broaches to the surface have occurred. At least 20 wells in the Prudhoe Bay field (Appendix F) have experienced fluid escape to the sur- face around the surface casing and conductor, a result of the annular injection of drilling mud. All of those fluids have been contained on the pads, largely in the well cellars. None has reached the tundra. Eight of those broaches were the result of a poor ce- menting package. Starting in 1992, operators established more rigid criteria for the selection of wells for annular in- jection. Since Jan. 1, 1992, of the 158 wells with annular injection of wastes, only three (DS 9-47, DS 15-45, and B- 33) have experienced broaching of fluids to the surface dur- ing injection down the outer annulus. Seven of the other broaches were related to casing leaks, and all were drilled without Arnco XT-200 hardbanding, a special treatment that reduces the chance of leaks. Since the implementation of the hardbanding policy there have been no new casing leaks. Three of the remaining wells had wellhead leaks, one of which was the result of a parted casing and one of which was the result of subsidence. The most severe broach occurred at the dedicated grind- and-inject well DS 4-19 in March 1997. A large volume of fluid appeared in several well cellars and even between wells on the DS 4 pad. A cumulative volume of 18,000 bbl (2.9 million L, 756,000 gal) of fresh and diluted seawater came to the surface (ARCO, unpublished material, 1997~. No gas or oil was found, and the solids on the surface contained no evidence of being injected solids. New procedures and guide- lines were adopted subsequently for injection of solid-waste slurries (ARCO, unpublished material, 1997~.
72 Permafrost Thawing from Upflow of Warm Injected Waste A widely discussed thermal study of vertical movement of injected waste was conducted from 1972 to the present on Pad 3 at Prudhoe Bay. Waste was injected into a permeable horizon at the 610 m (2,000 ft) depth below 560 m (1,850 ft) of permafrost and an underlying, impermeable "arresting interval" (e.g., Perkins et al. 1975, Weingarten et al. 2000~. Temperature measurements showed warming in the bottom 45 m (150 ft) of permafrost caused by the 23 °C (65 °F) waste injected into the horizon below. The issue was whether the obvious warming of the bottom of the permafrost by the waste was caused by heat conduction from contained (sta- tionary) waste, or by upward movement of mobile escaped waste. With a highly idealized thermal model, Weingarten and colleagues (2000) concluded that the waste was con- tained, consistent with the finding (Abou-Sayed et al. 1989) that the fracture created by the injection of waste was hori- zontal in this instance. Grouncl Water Degraclation by Waste Injection In recent years, drilling wastes, which previously were stored in environmentally undesirable surface pits, have been injected into subsurface aquifers for permanent disposal. Although much of the water in aquifers below the imperme- able permafrost is too saline to meet standards for a legally protected USDW, some is not. Because the sub-permafrost hydrologic system is poorly understood and incompletely sampled, the possibility of contaminating a potential water resource by waste injection should be considered. The North Slope of Alaska is largely classified as wet- lands underlain by permafrost, which separates the surface- water system active layer, lakes, streams from the rela- tively isolated and little understood groundwater system of sub-permafrost formations (Sloan 1987, Williams 1970~. A1- though water appears plentiful on the surface, the North Slope has an arid climate, and if a significant supply of fresh water exists in deep aquifers it could be a valuable resource in the future. Extensive federal and state regulations are designed to identify potential USDWs and exclude them from waste- disposal programs. The federal Safe Drinking Water Act de- fines USDW as groundwater that contains less than 10,000 ppm (parts per million) total dissolved solids (TDS). In aqui- fers where TDS exceeds 10,000 ppm there is no conflict re- garding waste disposal. Disposal in USDW is allowed only under special conditions (e.g., CFR144.3) the act essen- tially prohibits disposal in water with less than 3,000 ppm TDS (e.g., 40 CFR 146.4, 20 Aac 25.080(e)~1~. The TDS values needed to apply these regulations are generally in- ferred from well logs and are poorly known in detail. Rela- tively few direct chemical analyses are available from the North Slope. Much but not all of the sub-permafrost groundwater in production areas is known to have salinities in excess of the CUMULATIVE EFFECTS OF ALASKA NORTH SLOPE OIL AND GAS 10,000 ppm limit. A map (Fink 1983) of West Sak Sand- stone sub-permafrost water salinities shows that they range from >50,000 ppm TDS in the northeast corner of the Prudhoe Bay unit to <5,000 ppm TDS in the southwest cor- ner of the Kuparuk River unit. Although those concentra- tions are generally based on estimates from down-hole resis- tivity logs, the low values in the Kuparuk River unit were confirmed by extensive chemical sampling showing TDS generally <3,000 ppm (Fink 1983~. Similar low values have been reported from widespread sub-permafrost chemical sampling elsewhere on the North Slope (Collett et al. 1988, Table II-9~. Such results suggest that it might not be uncom- mon for sub-permafrost groundwater to meet the regulatory definition of USDW, but the extent is unknown and data on known occurrences have not been systematically compiled and published (see Chapter 10~. Finclings · Enhanced recovery procedures have not damaged the reservoir or other subsurface formations because of relatively low injection pressures and compatible chemistry. · Production of fluids has not caused significant local or regional subsidence but the thermal effects of warm fluids promote thawing of the permafrost (creating thaw bulbs and chimneys) and could provide potential pathways for the es- cape of fluids to the surface around boreholes. · There have been approximately 20 instances of broaching to the surface, primarily resulting from poor ce- ment packages that cause leaks and from wellhead leaks. Engineering solutions have been effective to date. · Groundwater resources and the effects of waste in- jection on them are inadequately examined and considered. Injection of drilling wastes into porous horizons has elimi- nated the problem of surface waste storage but has raised problems of possible groundwater contamination. . Recommendation It should be confirmed that existing subsurface waste disposal practices are not degrading a groundwater resource intended for legal protection. Rigorously measured total dis- solved solids profiles should be routinely acquired, com- piled, and used to identify patterns of freshwater distribution as a tool for planning and for evaluation and conservation of the groundwater if appropriate. ESCAPE OF INJECTED WASTE FLUIDS IN THE MARINE ENVIRONMENT One concern regarding the effects of industrial activi- ties on the marine environment is that injected waste fluids might travel laterally through a disposal zone to intersect the ocean floor. This is highly unlikely, however, because nei-
EFFECTS ON THE PHYSICAL ENVIRONMENT ther the producing reservoirs nor the disposal intervals inter- sect the ocean floor. Those units are buried hundreds to thou- sands of meters under younger rock and sediment. If such an occurrence were possible, one would expect numerous ac- tive large oil and gas seeps offshore from the major oil fields. They do not exist because the injection horizons are often as deeply buried as a number of the producing intervals, or as at Alpine, Kuparuk, and several of the satellite fields, the injec- tion zones are deeper than production ones. Surface Ballooning Surface ballooning, or rebound, is a phenomenon that can be caused by injection of fluids. The possibility of this occurring on the North Slope has generated some concern. Surface rebound is known to have occurred in areas where a reservoir is relatively shallow compared with its lateral ex- tent, there is a significant reduction in reservoir pressure, and the reservoir is relatively unconsolidated. Studies of the Cretaceous disposal intervals on the North Slope indicate that surface rebound is insignificant. AIR QUALITY Air quality on the North Slope has been affected by in- dustrial activities there and elsewhere. The most important potential accumulation of effects is likely to be a reduction in visibility and an increase in direct human exposures to pollutants caused by synergistic interactions between locally generated and globally transported contaminants. Ecologi- cal degradation also could result from deposition of dust and pollutants on terrestrial and aquatic ecosystems. Air quality on the North Slope meets state and national standards. Ambient concentrations of measured pollutants are often near detection limits at monitoring stations. How- ever, although local air quality does not appear to have been seriously degraded by emissions from oil and gas production facilities (AOGA 2001), emissions from local facilities re- sult in observable haze, increased atmospheric turbidity, and decreased visibility (AOGA 2001~. The most conspicuous air quality problems on the North Slope are the widespread arctic haze, which occurs at higher elevations, and locally produced smog. Research confirms that arctic haze is a common phenomenon in po- lar climates and that it is the result of distant rather than local emissions. Fugitive emissions from industrialized ar- eas in the temperate zone are transported long distances. There has been no research to determine how local and regional air masses and their contained contaminants inter- act. The lack of pre-development baseline data further ham- pers assessment of the effects of local or distant pollution on North Slope air quality. If additional fields are developed, air emissions will increase. If more energy is required to maintain production in declining fields as waterflood or gas-lift injection are 73 used to enhance oil recovery, air emissions could increase as well. Finclings · The only areawide monitoring program on the North Slope has been for priority pollutants as defined by the Clean Water Act, from 1986 through 2002, at a limited number of sites. No large-scale, long-term monitoring sys- tem has been established to provide a quantitative baseline of spatial or temporal trends in air quality on the North Slope. The lack of adequate information limits the accu- racy and precision of assessments of both past and future accumulation of effects. · The quantity of air contaminants reaching the North Slope from distant sources is unknown. · Little is known about the nature or extent of interac- tions between locally produced and globally transported air contaminants on the North Slope. Recommenclation Research and monitoring should be implemented to dis- tinguish between locally derived emissions and those that ar- rive by long-range transport, to determine how they interact, and to monitor potential human exposure to air contaminants. FRESHWATER ENVIRONMENT Industrial activities on the North Slope have to some degree affected the chemistry, flow patterns, and drainage patterns of the area's fresh waters. Effects could accumulate as a result of withdrawal or redistribution of water for con- struction of ice roads and pads, gravel mining in rivers, and blockage of drainage caused by gravel roads. Deposition of air contaminants also could alter water chemistry. Industrial activities to date have been concentrated in areas where lakes are common and there are abundant supplies of gravel. Those conditions do not characterize the foothills of the Brooks Range, many parts of the National Petroleum Reserve- Alaska, or the Arctic National Wildlife Refuge. Therefore, the effects of future development on fresh waters in those regions could differ from those of the past. Water Chemistry During summer, inland lakes tend to have low concentra- tions of dissolved ions, but lakes near the coast that receive nearshore brackish waters have elevated concentrations of dis- solved ions. As the ice grows in winter, electrolytes are ex- cluded from the ice matrix and ion concentrations increase. By late winter or early spring, at maximum ice thickness, ion concentrations in unfrozen water can be more than four times greater than those observed during summer. Evidence sug- gests that no significant changes in seasonal patterns or con-
74 centrations of chemicals in lakes and streams have resulted from industrial activities on the North Slope. Flow Patterns Much of the gravel used for construction of roads and pads has been obtained from deposits within the floodplains of rivers. Concerns arising from this practice prompted the U.S. Fish and Wildlife Service to study the effects of flood- plain gravel mining on physical and biological processes (Woodward-Clyde Consultants 1980~. The study identified numerous examples of habitat modifications, including in- creased braiding and spreading of flows. The study also set forth guidelines on how to mine gravel to reduce flood- plain effects (Joyce et al. 1980~. As a result, gravel mining largely has been restricted to deep mining in upland pits, some of which are flooded on abandonment to create aquatic habitat. Drainage Patterns Much of the gravel used for roads and pads has been de- posited in wetlands. During spring break-up there are substan- tial sheet-flows across the wetlands of the Arctic Coastal Plain into lakes and streams. When long stretches of gravel road interrupt flows, the difference in water surface elevation from one side of the road to the other can produce high flow rates in the cross-road drainage structures. An opposite effect can oc- cur in mid- to late summer when stream flow is low. Finclings · Gravel mining in rivers during the early years of de- velopment substantially altered flow patterns and distribu- tion of unfrozen water in winter, but recent restriction of gravel mining to upland pits has reduced those effects. · Gravel roads and pads have often interrupted both sheet flow and stream flows. Proper construction and place- ment of culverts can greatly reduce, but not eliminate, those effects. · Development in areas where surface water is less abundant could result in effects on freshwater that differ from those in the freshwater-rich Prudhoe Bay region. MARINE ENVIRONMENT Offshore activity in the Beaufort Sea has been limited. Activities that affect the quality of marine waters and flow patterns have included construction of gravel islands and causeways and discharges of materials. Only a few small spills have occurred in marine waters to date, but mechani- cal recovery the method allowed by current regulations- is not efficient and only removes a small fraction of the spilled oil, especially in broken ice. Concerns about con- CUMULATIVE EFFECTS OF ALASKA NORTH SLOPE OIL AND GAS lamination of marine waters center primarily on the poten- tial effects on marine organisms. Those effects are discussed in Chapter 8. There have been three permitted types of discharges to the Beaufort Sea over the life of the oil fields. First, indi- vidual facilities have discharges permitted under U.S. Envi- ronmental Protection Agency (EPA) NPDES (National Pol- lution Discharge Elimination System) program. Second, small or localized discharges have been permitted under the North Slope General NPDES Permit (under EPA). Third, exploratory drilling discharges were permitted under the Arctic General (or Beaufort General) NPDES Permit under either coastal effluent guidelines or offshore effluent guide- lines (Wilson 2001b). Permitted NPDES discharges include effluents from seawater-treatment plants, desalination plants, sanitary- waste-processing units, deck drainage sumps (from offshore production facilities, such as Northstar), temporary construc- tion dewatering, and occasional tests of fire suppression with water. These discharges are permitted for a specific facility, and there are monitoring and reporting requirements. Four facilities currently have NPDES permits: Northstar, the Prudhoe Bay seawater treatment plant (STP), Kuparuk STP, and Endicott. The largest discharges under this program are ocean water and peat detritus from the two STP operations (Wilson 2001b). A North Slope General Permit was issued by EPA for small operations other than those covered under individual NPDES permits. Permitted discharges include small volumes of water pumped from gravel mine sites and wastes from temporary camps. Industry must apply to EPA for coverage under the general permit. Discharges are small, localized, and infrequent. Exploratory drilling discharges are covered under the EPA Beaufort Sea General Permit and include disposal of drill cuttings and fluids from well-drilling operations. Which effluent guidelines are in effect depends on whether the well is drilled near to the shore or offshore. The coastal effluent guidelines in effect since the mid-199Os prohibit discharge of muds and cuttings. Offshore guidelines still allow dis- charges of muds and cuttings. Monitoring is frequently required as a condition of dis- charge permits to ensure that discharges do not exceed water quality standards, are not toxic to marine organisms, do not degrade water quality, and do not pose a threat to human health. Most of the records of the monitoring programs are retained by EPA, the principal permitting agency, and are not readily available. Records also are kept by individual operators, but those records are not summarized, and few annual reports have been produced. NPDES Monitoring Until recently, NPDES stipulations have called for en- vironmental monitoring for all operations covered by the
EFFECTS ON THE PHYSICAL ENVIRONMENT permits. The studies have included monitoring the waste stream as well as the receiving water. Topics include efflu- ent mixing and dispersion, and the effects of effluents on fish, benthic organisms, sediments, and water quality. Per- mits generally required that water, sediment, or biological samples be obtained seasonally from within and outside of the outfall mixing zone. Required monitoring of fish and benthic communities has been discontinued in recent years because effects were found to be minor or not measurable (Wilson 2001a). Water-quality monitoring has included measurement of several variables in the receiving water, including currents, salinity, temperature, pH, dissolved oxygen, total suspended solids, total residual chlorine, and chlorine reaction prod- ucts. Sediment studies have measured grain size distribu- tion, total volatile solids, and concentrations of organohalide compounds. Biological monitoring has included collection of fish and benthic organisms and toxicity studies that use commercially available test organisms (Robilliard et al. 1988, Wilson 2001b). 75 Results of monitoring water quality, sediments, and spe- cies for the Kuparuk STP outfall were summarized by Mont- gomery Watson (1994~. The results of winter and summer measurements of temperature, salinity, dissolved oxygen, pH, total suspended solids, and total residual chlorine showed values that were within permitted ranges both within and outside of the mixing zone. Sediment monitoring at the Kuparuk STP outfall revealed no adverse effects of the dis- charge on sediment grain size. Some variability was noted in silt, clay, and other grain sizes at some sampling stations. Total volatile solids showed a pattern of increasing concen- trations from west to east across the study area. This was attributed to variations in natural peat detritus across the sampling-station array. Fincling Physical effects of discharges and spills have been small and infrequent and have not accumulated. The effects of causeways are discussed in Chapter 8.