UNDERGROUND COAL GASIFICATION HISTORY, ENVIRONMENTAL ISSUES, AND THE PROPOSED PROJECT AT BELUGA, ALASKA

Transcription

1 UNDERGROUND COAL GASIFICATION HISTORY, ENVIRONMENTAL ISSUES, AND THE PROPOSED PROJECT AT BELUGA, ALASKA Kendra L. Zamzow, Ph.D. Center for Science in Public Participation March

2 Table of Contents CIRI's proposed project... 4 Natural Gas and Coal Gas... 5 How UCG works... 7 Getting coal to burn... 7 Operational parameters... 9 Deep and Thick Temperature Water Faults and fractures Historical Perspective Regional experiences Former Soviet Union United States Europe Other countries Key test sites Hoe Creek Centralia, WA Rocky Mountain I El Tremedal Chinchilla Environmental Impacts Structural Integrity of Host Rock Formation of contaminants Migration of contaminants CO Life cycle greenhouse gas emissions Carbon capture Carbon sequestration Summary Bibliography Appendix A: UCG reactions and Syngas Reactions Appendix B: Natural Gas Processing Appendix C: UCG sites worldwide Appendix D: Water Analysis at Contaminated UCG Sites

3 Figures Figure 1. Location of proposed UCG project, Beluga, AK. 4 Figure 2. CRIP process of gasification. 8 Figure 3. Depth and thickness of coal seams by global region..9 Figure 4. Faulted and dipping seams 11 Figure 5. UCG at Chinchilla, Australia...15 Figure 6. Comparison of life cycle greenhouse gas emissions by fuel type. 22 Tables Table 1. Composition of Natural Gas, Syngas, and UCG gas...5 3

4 In November 2009, Cook Inlet Region, Inc. (CIRI), an Alaskan Native regional corporation, filed for exploration permits to examine the potential to fuel a 100 MW power plant using "underground coal gasification" (UCG) technology. This paper explores the history of UCG technology globally, including environmental impacts during historical trials, how those impacts might be mitigated, and risk of environmental impacts at the Beluga, Alaska project. CIRI's proposed project Information on CIRI's proposed project is limited since the project is only at the conception and permitting states. The property is located in a remote location on the west side of Cook Inlet, just north of the Beluga River, on CIRI lands (Figure 1). The site is 5-10 miles from the current 385 MW Chugach Electric Plant, located at Beluga, which utilizes natural gas from nearby offshore platforms to provide electricity. The CIRI project proposes a 100 MW combined cycle power plant run on the syngas that would be the product of 'gasifying' coal in the ground. 1 Although no maps have been produced to indicate where the power plant would be located, it would almost certainly be adjacent to the targeted coal fields. Figure 1. Location of proposed UCG project, Beluga, AK. 1 CIRI presentation to the Alaska State House Resources, Senate Resources, and Senate Energy committees, October 9,

5 Natural Gas and Coal Gas Natural gas, as extracted from production wells, is mostly methane (CH 4 ) with a high heating value (>1000 BTU/ft 3 ). The composition depends on how the natural gas was formed. If it is biogenic (produced by living organisms) it is nearly pure methane. Thermogenic natural gas (produced by the breakdown of organic matter) is methane with contaminants (small hydrocarbons, water, sulfur, and CO 2 ) and must be processed before it can be used (Table 1). 2 The hydrocarbons can be separated and sold while the other constituents are corrosive. 3,4 Cook Inlet gas is primarily biogenic, and North Slope natural gas is thermogenic. 5 At the Beluga Power Plant, seven gas turbines and one steam turbine together produce 385 MW of electricity, the primary source of electricity to Anchorage. In the combustion process, methane is burned and water, CO 2, and oxidized sulfur and nitrogen products are the primary waste products. "Syngas", or synthetic gas, is the term that refers to a carbon monoxide-hydrogen gas mixture. It can be made from coal, natural gas, or biomass. "UCG gas" is used in this paper to refer to gas produced from burning coal underground, although some literature also refers to this as "syngas". Both are primarily carbon monoxide (CO) and hydrogen gas (H 2 ), but the processes of making them are different (Appendix A). To make coal-derived syngas above ground, coal is put under heat (>700 o C) and pressure to make carbon monoxide and hydrogen gas. Hydrogen or the building blocks for chemical products like methanol are the products. To make electricity, the CO and H 2 are reacted with steam to form CO 2 and more hydrogen. Hydrogen is combusted to produce power. Water, CO 2, and oxidized sulfur and nitrogen products are the waste products. Syngas plants are relatively common, with over 150 in operation. 6 The UCG process burns coal under heat ( o C) and pressure with steam while the coal is still underground. UCG gas as it comes out of the product well is carbon dioxide (CO 2 ) and hydrogen gas (H 2 ), with more methane and less CO than syngas, and lower in sulfur, tars, mercury, and other metals which are left underground in the residual ash after the burn. The actual proportions of each component will vary depending on how the burn is operated: Thickness of coal seam. Thin coal seams (< 2 m) produce a gas that is mostly CO 2, with little H 2, CO, and methane. This is a low quality gas, and may not be economic. Thicker Clayton, G. 1980; Goldsmith and Szymoniak 2009; 6 Simbeck, 2002, in Burton et al

6 seams don't change the amount of CO 2, CO or methane, but produce a lot more H 2. This is useful for generating hydrogen or a CO/H 2 syngas. Depth of coal seam. Burning deep seams where the pressure is greater creates more methane and less CO and H 2, and therefore a higher quality gas. Temperature. High temperature reactions produce more methane,, but temperatures that are too high reduce the quality of the gas Water. Changing the amount of water changes the amount of methane and hydrogen. Oxygen. Injecting air to start the burn instead of oxygen increases the nitrogen content of the gas, lowering the gas quality. The product gas will also contain sulfur, nitrogen, and volatile trace metal species. Like thermogenic natural gas, it needs to be 'cleaned' to remove hydrogen sulfide and other contaminants (Appendix B). UCG produces either hydrogen or a methane-hydrogen mixture for combustion after clean-up. If a methane-hydrogen mixture is combusted, CO 2 will be the primary air emission. If hydrogen is combusted, water will be the primary emission, but CO 2 will be produced as part of the process of making hydrogen. Hydrogen may also have other industrial uses, or may be used for hydrogen fuel cells and other parts of the emerging hydrogen economy. Similarly, chemical reactions can begin with UCG gas to make methane, methanol, fertilizer, and other products. Table 1. Composition of Natural Gas, Syngas, and UCG gas. Natural gas and UCG gas composition is as it occurs at the well-head, not after cleanup. UCG gas composition will vary depending on the purpose the gas is to be used for. and the conditions under which it is generated. The table is a compilation from the following sources: Shafirovich and Varma 2009, Goldsmith and Szymoniak 2009, Friedmann 2007, Bakker 2004, Claypool 1980, and the following websites: Natural Gas Syngas UCG gas Methane (CH 4 ) 70-90% 1-2% 5-14% Hydrogen (H 2 ) 24%-30% 25-40% Carbon dioxide (CO 2 ) 0-12% 4%-15% 25-40% Carbon monoxide (CO) 35-65% 5-20% Hydrogen Sulfide (H 2 S) 0-5% 1% 2-8% Nitrogen (N 2 ) 0-5% 1%? 6

7 Water (H 2 O) saturated 15-25% 33% Producing electricity by burning these different gases will generate different environmental footprints. For instance, syngas requires mining and transporting coal; natural gas production requires deep, often offshore wells; while UCG gas may require deep wells but has very little above ground disturbance. Similarly each has a different potential to pollute water or air, and different greenhouse gas types and amounts. How UCG works The basic concept is to drill two wells into a deep coal seam and burn out the coal between them. Compressed air or an oxygen/steam mixture is injected through one well, the coal burns and releases gases, and the gases come out of the ground at the second well, called the production well. The burn creates a cavity or "combustion chamber", and the process works best with deep seams of low-quality coal, exactly the material that is difficult to traditionally mine economically. Water flowing into the cavity is not pumped out, but is used as part of the burn reaction. Because of the cost of transporting the gas, power plants would likely be sited adjacent to the coal field. Pilot projects have been conducted since the 1940's, and particularly in the 1970's and 1980's in the US and the 1990's in Europe and China. Trials focused on how to get the coal to burn, how to control the burn, maximizing efficiency, and minimizing environmental contamination. Getting coal to burn The first difficulty in the early trials was getting the coal to burn. Although there are numerous instances of uncontrollable underground coal fires around the world, the fires are dependent on oxygen reaching the coal. Deep coal seams several hundred feet underground are saturated with water and isolated from oxygen, and attempts to simply set the seam on fire fizzled. In researching science journals, federal documents, and company literature this author was unable to find any instances of uncontrolled coal fires during UCG. Eventually two techniques proved to be successful. Inject air at the injection well. The ignition source is placed at the production well to "draw" the fire towards the high oxygen area in a process called "reverse combustion", burning a path through the seam. Drill a simple vertical well as the production well. The injection well begins vertical then bends to become a horizontal tunnel through the coal seam towards 7

8 the production well. A "controlled retractable injection point" (CRIP) is a point where coiled tubing burns through the horizontal tunnel borehole casing; oxygen and steam are forced through the point to ignite the coal. This is initially placed near the juncture of the injection and production wells. The coal burns for a while, forming a cavity as hot gases move up and outward, but it eventually fizzles. When the burn is done, the point is retracted and ignition is started again at a point in the horizontal tunnel closer to the injection well. In this way the burning coal front proceeds in a controlled manner (Figure 2). The CRIP technique allows for several production wells for each injection well, reducing the overall footprint. A proprietary technology developed by Ergo Energy was successfully used in the 30-month long Chinchilla Australia project. However, the specifics of the technology are not known. 8

9 Inject oxygen and steam Gas Overburde n Coa l New injection point, new burn burned out injection well Old burned out cavity with ash/rubble pyrolyis products line heat and hot gases water influx Figure 2. CRIP process of gasification. The movable injection point begins the burn near the production well. When the first burn expires, a second burn is initiated closer to the injection well. This procedure continues until the seam is burned out. Adapted from Burton et al In general, the CRIP process and the Ergo technology have worked better than hydraulic fracturing and reverse combustion in controlling the UCG process. 7 Operational parameters Constraints such as depth, thickness, dipping of the coal seam and temperature, pressure, oxygen, and water in the burn cavity all play a role in gas quality, economics, and potential environmental impacts. 7 Hydraulic fracturing is a technique of fracturing the coal seam between the two wells to encourage gas flow; this did not work, as gases spread out and did not flow consistently in the desired direction. 9

10 Deep and Thick Trials have determined that UCG should be conducted on deep, thick coal beds. This provides a better quality gas and reduces the risk of groundwater pollution or surface subsidence. The pressure in the burn cavity affects how well the reaction will proceed, the composition of product gas, and how the surrounding rock will be altered. Burns conducted at depth (200 m to over 1000 m) where the pressure is greater produce better quality gas with more methane. This has been proven in the field: using oxygen injection, gas produced at 4 bars of pressure contained 5% methane (and 21% CO and 38% hydrogen) while gas produced at 5.3 bars contained 13% methane (and 13% CO and 25% hydrogen. 8 The remainder of gases in both cases was carbon dioxide and hydrogen sulfide. The difference affected the heat content of the gas, with 8.7 and 10.9 MJ/m 3 produced respectively. Deep seams also cause the economics of air versus oxygen injection to change. Injecting air Figure 3. Depth and thickness of coal seams by global region. The US tends to have thick seams of coal near the ground surface, while Europe tends to have thin, deep seams of coal. Thick seams produce better quality gas, but deep seams are more likely to isolate contaminants and reduce the risk of subsidence. From Burton et al requires compressing nitrogen and injecting oxygen requires an oxygen separation plant both are expensive. Oxygen injection becomes economically favorable for deeper coal seams as the cost of making oxygen becomes less expensive than additional compressors for injecting air. 9 Thick seams allow gas production with fewer wells. Seams that are too thin (less than 2 m) allow heat to escape into surrounding rock too easily, and the resulting product gas is of poor quality. 10 Some UCG development companies suggest seams should be at least 10 m thick, 11 while others believe seams as thin as 0.5 m can be used; this is likely biased by the depth of available coal (Figure 3). 8 Kreinin in Shafirovich et al 2008; Shafirovich and Varma Boysen et al 1998; Burton et al Bowen and Irwin Shafirovich et al

11 Temperature Temperatures that are too low allow tars to form and can cause the pathway between wells to plug, but if too high reduce the efficiency of the gasification process and the heat value of the final gas. Water Although coal contains water and combustion cavities are expected to have water flux into them, too much water will reduce the methane content in the gas, reducing the heating value. For this reason, if UCG is being used to provide electricity it is preferable to use coal with low moisture with no overlying aquifer within 25 times the height of the seam. 12 This also reduces the risk of groundwater contamination, particularly groundwater that might reach surface water or enter drinking water wells. Faults and fractures Coal seams may dip up and down, or a fault may cause the seam to be discontinuous, suddenly stopping at one depth and starting again at another. To adjust for fault discontinuities, drilling equipment has the capacity to contain "eyes" that "see" the geologic structure ahead of the drill bit; the drill can then be adjusted as necessary (Figure 4). To use the technology on steeply dipping coal seams, combustion occurs at the deep end and the coal above "gravity feeds" down into the fire, with production gases working their way upward (Figure 4) and coal tar flows down away from the burn. Testing on dipping seams has been conducted at Rawlins, WY and in Russia (Juschno-Abinsk). Figure 4. Faulted and dipping seams. (Left) Drilling faulted seams. (Below) Gasification of steeply dipping seams. 12 Sury et al 2004, in Shafirovich et al

12 Historical Perspective This section lays out some the general history of underground coal gasification in different countries, and describes some key sites in detail. A list of all gasification sites is available in Appendix C. Regional experiences Former Soviet Union The technology for burning coal into gas while it was still in the ground began in the 1930's in the USSR, and by the 1950's the USSR was producing about 300 MW of electricity with this gas. 13 About 200 pilots were conducted in the USSR and, after 1991, in Russia, Ukraine, and Uzbekistan. One station (Yuzhno-Abinsk, Kuznetsk Basin, Russia) produced gas for 14 boilers from , finally closing as equipment failed during the post-soviet era. Another plant built in the 1950's in Uzbekistan is still operating. Most closed as cheap natural gas came on-line. During the trials in the former Soviet Union, it was learned that injecting oxygen rather than air produced gas with higher heat value, that transporting the gas any distance is often uneconomical, and that high temperatures ( o C) cause rocks to swell. 14 Environmental modeling was also developed. United States In the US, initial tests began in Alabama in the 1940's and 1950's. Testing was revitalized during the years of high oil prices, with 31 tests conducted , mostly by the Department of Energy (DOE). They were short-term projects, with a total of only 50,000 tons of coal gasified. It was during this period that the CRIP technology was developed by Lawrence Livermore National Labs (LLNL). Most trials attempted to answer specific questions about managing the burn, shutdown and startup, gas consistency and quality, and groundwater impacts. New projects are scheduled to begin in Wyoming: in July 2007, British Petroleum (BP), LLNL, 13 Shafirovich et al 2008; Shafirovich and Varma Den'gina et al 1994, in Shafirovich et al

13 and a UCG developer signed an agreement for a pilot in the Powder River Basin of Wyoming that would incorporate carbon capture and sequestration (CCS) with UCG. 15 Europe In the European Union, a series of experiments was conducted , primarily in Belgium, France, and Spain. These trials tested methods to get coal to ignite, move gas to the production well, and determined that coal burning could be conducted in deep coal seams. Repeated testing of reverse combustion and hydraulic fracturing to direct gas to the production wells failed. Not until directional drilling and oxygen injection were attempted at the El Tremedal site in Spain were researches able to link wells productively. All were short-term tests primarily designed to learn more about the technology. No electricity is currently generated from UCG in Europe, although more trials are planned. A consortium of countries led by Poland has started a pilot to test the feasibility of using UCG as a cornerstone of developing a hydrogen economy and the feasibility of integrating it with geothermal heat exchange and CCS. 16 The UK is examining the feasibility of conducting UCG in a coal seam that lies under the Firth of Forth in Scotland; the gas would be used in conjunction with fuel cells to make electricity. This would be the first UCG project beneath ocean water. Other countries In Australia, one of the most successful pilots was conducted. The Chinchilla project ran from and demonstrated that UCG could be controlled, including shutdown and restart. During the 4-year period, 35,000 tons of coal from a seam 140 m deep was gasified with no environmental problems. Today two major pilot projects are in development, and several other smaller projects. Pilots include a planned 400 MW combined cycle gas turbine (CCGT) power plant and a 100-day pilot to test a module-based system to produce 20 MW per module, 17 with the goal of developing a commercial CCGT plant. In October 2008, a coal gas-to-liquids fuel production facility was started. 18 Canada has not had any historical UCG projects, but two projects are moving forward in Alberta to use UCG for power, fuel, and hydrogen and sequester CO 2 using enhanced oil recovery (EOR). 19 The steam from UCG may be used in tar sands oil recovery. 20 The proponents of these projects, Laurus Energy, may become a partner in the CIRI Beluga project Shafirovich and Varma Rogut Shafirovich et al Friedmann et al ibid 20 Maev 2008; Shafirovich and Varma Bluemink

14 A test project in New Zealand in 1994 only lasted 13 days. The information available says the area was "tectonically active with coal deposits faulted and folded, providing a geologic challenge" but does not explicitly detail the issues, except to say that they did not achieve good gasification. 22 South Africa started a pilot UCG project in January 2007 with the intent to use it for both power and coal gas-to-liquid fuel. The small amount of UCG gas would be used to cofire turbines at a large natural gas facility. The co-firing was successful , but it is unclear if a scale-up occurred after that. China has had 16 UCG trials since the 1980's. One project currently operating in XinWen uses six UCG reactors to provide gas for cooking and heating, while one in Shanxi uses the gas to produce ammonia and hydrogen. A $100 million pilot commercial project has started in Inner Mongolia next to a coal mine. Other plants are used to produce fertilizer. More trials are planned, including feasibility of UCG for hydrogen and methanol production. 23 Key test sites 24 Hoe Creek The Hoe Creek site in Wyoming was operated from The coal seam was 10 m thick, lying m below ground, with a shallow layer (5 m thick) of siltstone and clay separating it from an upper coal seam; overburden above that was primarily silt, sand, and sandstone. The sand and coal seams were the primary aquifers. Three experiments (Hoe Creek I, II, and III) were conducted and heavily monitored to examine the burn process, gas composition, cavity formation, and geotechnical data. The primary research at this time was in getting gas to move to the production well. At Hoe Creek I, explosives were used to fracture the coal bed and air was injected for 11 days. The test was not very successful, with about 7% of the gas lost to the overburden. At Hoe Creek II, three separate trials of 2-43 days used reverse combustion with air or oxygen. Water entering the burn cavity lowered gas quality, so the pressure was increased to keep water out. However, this forced much of the gas out of the cavity away from the production well, and about 20% of the gas was lost. Hoe Creek III used directional drilling and reverse combustion over a 47 day test. Unfortunately the burn at the lower coal seam target moved into the upper coal seam, a mere 10 m above, and again nearly 20% of the gas was lost. Eventually subsidence occurred at both Hoe Creek II and III. Twelve monitoring wells were sampled before, during, and for up to two years after gasification. 25 Groundwater contamination occurred within seven days of the start of gasification. The pressure used in the cavity to try to push water out also pushed out soluble 22 Shafirovich and Varma ibid 24 These compilations are derived from Burton et al 2008, except where noted 25 This section from Campbell et al

15 volatile organics (such as phenols) and other contaminants (like cyanide) into the aquifer above. The problem was exacerbated by surface subsidence, which occurred due to the shallow depth of the seam and the lack of structural integrity in overlying rock. Toxic organics in residual ash dissolved in inflowing water and moved into all three aquifers. Due to the extensive monitoring well system and groundwater analysis, the contamination was picked up and monitored. Analysis was done for 250 different organic and inorganic compounds, and 70 were detected (Appendix D). Testing up to two years after the burns found all contaminants were within 30 m of the burn zone, and concentrations decreased very rapidly with distance; many probably sorbed to overburden and residual coal layers. However, by 1993, the DOE found that contaminants remained in an aquifer 55 m below the surface, and had migrated off of the original BLM-owned property the testing was conducted on. Contaminants included phenol and benzene (known carcinogens) and other organics known to cause kidney and nerve damage; all were small, highly soluble molecules that do not sorb well to soils. In 1998, DOE installed 64 air-sparging wells to remediate the site, and another 50 were installed in A variety of remediation technologies were in use as of Much of what we know now came out of this test, and later tests based on these findings. This was the first successful use of oxygen/steam injection and a movable injection point. What was learned from the subsidence and groundwater contamination became part of the basis for site-based risk assessment by today's standards, the site would have been considered as having high environmental risk due to the shallow depth of the coal and proximity to aquifers. Centralia, WA Between 1981 and 1982, the CRIP system was further tested at Centralia, WA for 4 and for 30 day burns. Different oxygen/steam ratios as well as a propane-silane (SiO 4 ) combination were used to ignite the burns, and drilling configurations and slants were tried to examine changes in syngas quality. The variations did not change gas quality much. This trial was the first real test of the CRIP system, and also tested whether models could predict how cavities would grow. Cavity shape and size models were validated by quarrying out the actual burn cavities. Quarrying also allowed researchers to examine the products left in the cavity dried coal, char, and ash. No subsidence was predicted, and none was observed. Rocky Mountain I The Rocky Mountain I test in Wyoming November 1987-February 1988 was considered the most successful US test to date. The project focused on siting the project to prevent groundwater contamination. Significant effort went into pre- and post-burn water, temperature, and mineral analysis to determine how the burn changed the underground make-up of the rock and water chemistry. The coal seam was 10 m thick and 130 m below ground. The successful directional drilling and CRIP processes tested in Centralia were used continuously for several 15

16 months. 26 Negative pressure was used to ensure that water flowed into, not out of, the burn cavity, and water that filled the cavity post-burn was pumped to the surface and treated to remove underground contamination from dissolution of ash and pyrolysis products, both to ensure that no contaminated water remained underground, and also to cool the cavity quickly to reduce steam, which can crack the rock above and induce fractures, and reduce transfer of hot gases to surrounding rock. No environmental contamination was found by the 19 groundwater monitoring wells. Research indicated that the heat in rocks surrounding the burn cavity does not dissipate quickly, and rocks can still be 4-12 o C hotter than normal two years after a burn. Similarly, groundwater temperatures did not always rise until several months after gasification ended. 27 The rise in temperature in wells was entered into models to calculate the temperature along production lines. Within 1 m of the well, rocks could be o C, nearly as high as temperatures in the burn zone. This is potentially high enough to cause the rock around the gas lines to change and affect the cement-rock seals and could lead to gas leaks. Temperatures decreased rapidly with distance from the line: as modeled they would have been 100 o C four meters away and within 16 m they were only 4.5 o C higher than background. Although the testing was successful and the operators intended to go into commercial production of ammonia, the Rocky Mountain UCG site was shut down when cheap oil became available. El Tremedal The El Tremedal site was a joint project of Spain, the UK, and Belgium located in Spain and operated Directional drilling and oxygen injection were used. The tests were conducted to determine if gasification could be done on deep seams (550 m) while maintaining negative cavity pressure to prevent groundwater contamination. A methane explosion damaged the injection well and stopped the project, but no environmental contamination was detected by the several monitoring wells. 28 Chinchilla 29 The Chinchilla project emerged from testing in the 1980's at the University of Newcastle, Australia. It was conducted over 30 months from December April 2003 using the proprietary Ergo technology and consisted of 9 injection/production wells surrounded by 19 monitoring wells (Figure 5). The coal was 140 m deep and 10 m thick. The test was conducted under low temperatures (300 o C) and reverse combustion with air/water injection 26 Clean Air Task Force Gosnald Friedmann et al Information from Shafirovich et al 2008 and from Burton et al

17 (rather than the CRIP technology) was successfully used between vertical wells. Up to 675 tons of coal per day was gasified, with 75% total energy recovery. No groundwater or surface water contamination was detected, nor was there any subsidence. A gas-to-liquids plant was constructed in 2008 at the site, with the intent of using UCG product gas. What made this project important was that it validated the concept of keeping the cavity at a pressure less than the surrounding rock to allow groundwater to flow into the cavity and keep volatiles from being pushed out; essentially a successful scale-up of the testing done at Rocky Mountain. Figure 5. UCG at Chinchilla, Australia. From Hattingh, L Underground Coal Gasification. Sasol. 20GASIFICATION%20%20-%20Lian%20Hattingh.pdf Environmental Impacts The primary concerns are the potential for uncontrollable fire, sinkholes (subsidence), groundwater contamination, and air emissions, including increased greenhouse gases. Essentially, the risks can be broken down into: will contaminants dissolve, how much CO 2 can be captured and sequestered without leakage, and will any contaminants reach anything important? Although a literature review has not revealed any instances of uncontrolled fires, most projects have been conducted for only a short period of time and little information is available regarding the New Zealand pilot in a tectonically 17

18 complicated area. Current recommendations are that there should be no major faulting within 45 m of the proposed gasifier. 30 The potential for new faults to develop and provide a route for air to reach the coal seam will need to be assessed in Beluga. Subsidence and groundwater contamination have been issues in past projects where shallow coal seams were burned; the recommendation now is to use coal seams greater than 200 m deep with an impermeable, structurally sound layer above the seam and no potable aquifers nearby or within 25 times the height of the coal seam. CIRI proposes to use a seam 198 m deep. If an impermeable overburden layer is present it also helps prevents product gas from flowing into the surrounding rock, improving the quantity of gas retrieved. However, it should be noted that a structurally sound layer does not eliminate the risk of subsidence. Any rock overlying a burned out cavity could develop fractures. About half the mercury, arsenic, sulfur, tars, and particulates produced from burning coal remain underground. While this reduces air emissions, it is a potential concern for groundwater contamination. Groundwater can become contaminated with volatile, soluble organics like benzene and phenols. 31 Site-specific geologic and hydrologic assessment will need to assess whether the aquifers in the area are fresh or saltwater, the potential for connection between the coal seam and aquifers, and the potential for the aquifers to reach surface water. A connection to surface or tidal water is a serious risk, in that benzene at levels safe for humans can cause genetic damage in salmon exposed to it consistently. 32 High temperatures in production wells could cause well casings to crack and release hot gas; 33 if the well passes through an aquifer this could be a route for contamination. To gasify coal above ground, the coal must be mined, transport, and put under great pressure and heat before it can fuel turbines. The environmental impacts include all the impacts of mining (water contamination, methane release, potential subsidence for underground mining, human health impacts for miners) as well as air pollution from combustion (CO 2, mercury, sulfur and nitrogen oxides). By gasifying the coal below ground, many of the mining impacts are eliminated, and groundwater and air pollution become the primary risks. 30 Surey et al 2004, in Burton et al Campbell et al Carls et al Gosnald

19 Structural Integrity of Host Rock Coal will be surrounded by "host rock". In the nearby Chuitna coal fields near the Chuit River (less than 100 m deep), the host rock is primarily permeable sandstone saturated in water. If the same geologic forces that created this set of conditions also created the coal and host rock at the Beluga coal fields (200 m deep), they could also be overlain by a permeable sandstone aquifer. 34 This would increase the risk of a UCG burn contaminating an aquifer, and would be important unless the aquifer were saline. Tectonic activity can create faults and fractures that allow UCG gas to escape, allow water to move in unexpected directions, and provide a route for contaminant transport. 35 Not only do any current faults need to be assessed, but the potential for high temperature activity to cause stresses and fractures and provide new pathways, collapse of the burn cavity, or subsidence needs to be assessed. Subsidence occurs when coal is removed, leaving a void under the surface. Subsidence does not always occur; it was minimal in pilot tests in Centralia, WA and Chinchilla, Australia. However, these were pilot projects, and it is not known what would happen in a commercial situation where large quantities of coal are removed. Given the remote location, the primary risk is the potential to create pathways for contaminants rather than direct risk to habitation. Formation of contaminants The high temperatures of the burn cause volatile hydrocarbons and some trace metals to become gases and carbon in coal and carbonate rocks to release carbon dioxide. These generally partition into either the production gas or end up in the residual ash that stays in the cavity after the burn is complete. If there is a route to an aquifer, highly soluble off-gassed volatiles like phenol can cause persistent water contamination, as can material in ash. Organic compounds such as tars, benzene, toluene, phenols, and polycyclic aromatic hydrocarbons (PAH's) will be created as heat dries and burns coal. The rocks themselves will change: carbonate rocks will release calcium and CO 2 ; mafic rocks will release iron and magnesium, and so forth. Metals from rocks will volatilize and move out with production gas, remain in residual ash, and may move into pore spaces of surrounding rock. A lining of burn products can be generated around the burn cavity. There are two periods to consider: during the burn and after the burn. During the burn, contaminants are most likely to volatilize and move out with product gases. After the burn, contaminants are most likely to become soluble in water and migrate out of the burn cavity as normal hydrologic flow re-establishes. 34 Burton et al 2006 Section Creedy and Garner 2004, in Burton et al 2006; Gregg 1977 in Burton et al

20 Migration of contaminants Very high temperatures (greater than 1000 o C) cause rocks to crack and burn and solid metals become gases; also water becomes less dense and less viscous so it moves more easily allowing easier transport of contaminants. High temperatures in the production well, carrying the gas product, may be high enough to crack the well casing and allow gases to escape gases that can contain metals and toxic organics. The production of contaminants and their movement is entirely different from any other industry, and prediction needs to rely strongly on results from pilot tests. Whether contaminants become a risk depends on whether they are able to reach water being used by aquatic life or people. Just as steam and oxygen, temperature and pressure affect the quality of the UCG product gas, they also affect what happens to the unintended byproducts. Burns will be operated at very high temperatures in order to shift the reaction to produce methane, and the higher the temperature the less byproduct. However, higher temperatures also increase the solubility of organics, allowing them to move further in water. High temperatures can thermally drive water up through the burn cavity roofing, cause cracks or collapse of the burn cavity that allow water to migrate out, and cause organics to become soluble in water. 36 Deep UCG projects will need to be run at higher pressures to keep the burn going, risking outflow of water from the cavity, but are more likely to be far from potable aquifers. High pressure and the buoyant gas forces can combine to overcome the pressure surrounding the cavity, resulting in vaporized material moving out of the cavity and condensing in the outer rock. If the burn is advancing in that direction, the process may repeat. 37 As material is pushed away from the hot cavity, it condenses, absorbs, adsorbs, or in other ways reacts to precipitate away from the cavity. Organics and ammonia may sorb to coal or surrounding clay. This material may be encountered as groundwater re-establishes its natural flow post-burn. After the burn, the normal hydrologic flow will fill the underground chamber and dissolve the ash left behind. When it encounters the precipitated or sorbed material outside the burn cavity, different reactions may occur. Some material may dissolve; some will be detoxified if the groundwater is high in oxygen for instance, ammonia will become the non-toxic nitrate and some may be broken down by aerobic bacteria. The migration of contaminants may be irrelevant if the coal was capped by an impermeable layer or no potable aquifer is at risk. However, some of the contaminants are toxic to fish, if they are able to reach fish-bearing waters: ammonia, high concentrations of calcium or other cations, high concentrations of total dissolved solids (TDS commonly mostly sulfate), and low but persistent concentrations of 36 Under room temperature conditions, many organics are not soluble in water, which is why oil forms a sheen on water instead of dissolving. 37 Burton et al

21 PAH's. One method of mitigation to prevent harm to drinking water or aquatic life is to pump and treat water as it enters the burn chamber until all toxic compounds are below safe levels, as was done at Rocky Mountain I. One author has suggested that UCG sites should be at least 1.6 km from rivers and lakes, and 0.8 km from major faults to prevent groundwater contamination conditions that may be difficult to meet at Beluga. 38 CO 2 Carbon dioxide is the defining pollutant of our age endangering entire populations of people, plants, and animals through its role in global warming and ocean acidification. Models developed by international consensus through the IPCC are proving to have underestimated the rise in global temperatures. Feedbacks such as reduced ice cover at the poles (less reflection of sunlight, more absorption), release of methane from warming Arctic tundra, 39 and positive biological feedback mechanisms such as vast stretches of dying trees in the Pacific Northwest (due to increases in beetle kills because winter temperatures no longer kill the beetles) and no longer removing CO 2 may account for the unexpectedly rapid temperature increase. Ocean ph is dropping, Arctic ice is melting, and permafrost is thawing at rates much faster than predicted, and there has been increased drought in Australia, the US, Africa, and the Middle East; increased flooding; eroding beaches in Hawaii and villages in Alaska; and more. The measured physical observations indicating the fast rate of global warming, the human face of it, and the likely fiscal impacts on individual gas emitters make it an imperative to consider greenhouse gas emissions in any large scale project. Life cycle greenhouse gas emissions Carbon dioxide will be produced from the UCG process as the raw gas exits the production well and also when methane is combusted in the power plant if a methane/hydrogen mixture is used. All carbon products become CO 2 during combustion if the product gas entering the power plant contains CO 2, CO, and methane, all of these will exit the stack as CO 2. If CO 2 and CO are removed during a "cleanup" or carbon capture process, only the methane will be converted to CO 2 in the stack. Although no studies could be found that analyzed the life cycle greenhouse gas emissions of syngas made through the UCG process, analysis has been done comparing coal, syngas, natural gas, and liquefied natural gas both with and without mitigation technologies (Figure 6). 40 The study notes that natural gas is one of the largest sources of greenhouse gas emissions in the 38 Bowen, BH. A review and future of UCG. Powerpoint pdf 39 Methane is a greenhouse gas more than 20 times as potent as CO 2 40 Jaramillo et al

22 US when processing, transmission, and combustion are included, producing about 800 lbs of CO 2 -equivalents per MWh, or an estimated 250 lbs if CCS could be incorporated. But this is less than traditional pulverized coal plants, which produce about 1800 lbs of CO 2 -equivalents per MWh, or an estimated 400 lbs if CCS could be utilized. UCG product gas is likely to be similar to natural gas in the combustion, processing, and transmission components, although it will require extra release of CO 2 for air compression or making oxygen; it will be significantly lower than traditional above-ground gasification CO 2 releases in that no coal mining, processing, or transportation are required, nor is energy required for the gasification process as in above-ground facilities. While UCG combined with carbon capture is likely to produce much lower greenhouse gas emissions than a traditional natural gas plant, it is not a zero-emissions technology. In 2001, the Beluga plant supplied 300,000 MWh of electricity. 41 If UCG with CCS fueled a similar amount of electricity, it would generate at least 375,000 tons of CO 2 -equivalent annually if the estimates of about 250 lbs of CO 2 per MWh are correct. 41 ISER

23 Figure 6. Comparison of Life Cycle Greenhouse Gas Emissions by Fuel Type. Although UCG product gas was not analyzed, it is likely similar to natural gas in production, processing, and storage. It will likely produce slightly more CO2 than natural gas due to energy requirements for compressed air or oxygen generation, but could produce less sulfur and nitrogen if some of those components remain underground. UCG is not like syngas, in that it does not require coal mining or transport, nor does it require power for gasification. All but the PC scenario assume a "combined cycle" turbine system would be used, in which gas drives the first turbine and energy from that is used to make steam to drive a second turbine. The natural gas plant at Beluga is NGCC, and CIRI envisions a combined cycle plant using UCG gas. PC=pulverized coal IGCC=integrated gasification combined cycle NGCC = natural gas combined cycle LNGCC = liquid natural gas combined cycle SNGCC = syngas combined cycle. 23

24 Carbon capture CO 2 is commonly captured from natural gas wells and re-injected into oil fields. This is because ideally methane is all that goes into the gas turbines at power plants; CO 2 is corrosive in pipelines and lowers the heat value of the natural gas. Due to these economic factors, removing CO 2 from natural gas is a long-tested technology. Amine scrubbers, membranes, and "pressure swing absorption" are the main processes used CIRI has discussed capturing CO 2 from UCG gas and re-injecting it into natural gas fields, however there has not been any interest on the part of the natural gas produces in Cook Inlet. Removing CO 2 from UCG gas has never actually been done. There are three groups of methods for capturing CO 2, and all could theoretically be adapted to fit a UCG operation. Pre-combustion removal injecting oxygen and steam to start the burn and adjust the water to converts CO to CO 2, so the product gas contains mostly CO 2 and H 2 ; the dense stream of CO 2 can be separated by amine scrubbers, membranes, or other technologies before reaching the combustion facility. Post-combustion removal removal of CO 2 from the combustion stack by amine scrubbers or other technology Oxy-fuel method gasification and combustion can both be carried out using oxygen instead of air; the stack gas will be primarily CO 2 and steam and CO 2 is recovered by allowing the steam to condense. While CIRI has discussed capturing CO 2 from the product well, it is unclear if they intend to also capture the CO 2 produced from the proposed power plant, or any additional facilities that may be considered in the future, such as a liquid natural gas plant. While CO 2 capture from production wells at natural gas fields is commonly done, capture from power plant stacks (post-combustion CO 2 capture) is a concept in its infancy. CO 2 sequestration has just started at the Mountaineer 1300 MW power plant in West Virginia, but they are only capturing 1.5% of the emissions during this initial test period. 42 Not only is CCS technology just developing and likely to take decades to mature, it is also likely to be expensive. Estimates by the International Panel on Climate Change are that capturing CO 2 from natural gas power plants will increase the capital cost by %. Due to the energy needs of the compressors and capture equipment, up to 30% more natural gas may be needed to produce the same amount of electricity as without the capture equipment, although the Mountaineer plant is testing a new capture technology that requires less power. The cost of 42 Biello

25 electricity is estimated to be 35-70% higher for a natural gas combined cycle plant, such as is used at Beluga, if CO 2 capture is installed. 43 Pre-combustion technology itself (Selexol) costs about $25/ton CO 2 captured. 44 Carbon sequestration The locations where CO 2 is removed from natural gas are in the southern US, where pipelines transport the CO 2 to declining oil fields. Geologic sequestration has been discussed and theorized, but rarely implemented. The only commercial-sized long term sequestration of CO 2 outside of enhanced oil recovery is at the Sleipner, Norway natural gas production platform, where CO 2 has been injected into saline aquifers 1000 m beneath the ocean floor since This project has been driven by Norway's high carbon tax, $55/ton CO 2 in 1991 (the equivalent of over $100,000 per day for Sleipner). Drilling the injection well and installing a compressor added $100 million to the project; adding scrubbers to remove CO 2 and monitoring equipment were additional expenses. While Sleipner has been successful, not all projects have gone smoothly. Norway's Snohvit natural gas platform has had significant technological problems with storing CO 2. The CO 2 freezes at temperatures required to make LNG, blocking the LNG transport pipe. The plant was shut down twice in 2008 and again in Pilot projects that injected CO 2 into rock formations to make solid carbonate rocks instead caused carbonic acid to form, and the acid dissolved the rock cavity intended to contain it. The process stopped when neutralizing rock was encountered. 46 Injecting CO 2 into basalt rock to make mineral carbonates failed when the rock swelled and plugged the underground pore spaces. 47 CO 2 captured at a power plant in Wisconsin (as a demonstration project) did not store the CO 2 because the geology under the plant was not favorable. 48 Currently there has not been enough test-drilling to determine if the geology at the Beluga coal fields would support sequestration. CIRI has suggested injecting the CO 2 into declining natural gas or oil fields, but currently no producers have showed interest in the idea. A report from the National Energy Technology Lab suggests that sequestration can only be done as 43 Thambimuthu 2005 in Burton et al Burton et al Hurst 2008 and 46 Kharaka et al Sturmer et al

26 EOR or as saline aquifer injections in the Beluga area, and that both are likely to be costprohibitive. 49 CIRI has discussed geologic sequestration of CO 2, 50 although they have not shown how this would be economically feasible. It is likely that storing the CO 2 in depleted underground burn chambers will be considered; it is possible the capacity will be available, and injecting CO 2 into residual coal causes swelling that would plug fractures and migrating CO 2 would tend to adsorb to coal and not move far. However, at Beluga the burn cavities are only expected to be 200 m below the surface, and CO 2 storage should be at least m below the surface to maintain CO 2 in a dense supercritical state. Nevertheless, the CO 2 is still likely to be less dense than water, and will be "buoyed" up to the top of a caprock layer, making it important for the caprock to remain impermeable in perpetuity. 51 This may be particularly important in a seismically active area such as Beluga. The heat and steam may cause the rock around the cavity to be quite different than pre-burn, potentially initiating cracks, fractures, and section collapses. Volatile organics (benzene, etc) left behind in the cavity dissolve easily in CO 2 and will be carried upwards by CO 2 if the rock above the cavity is permeable. CO 2 forms carbonic acid as it dissolves in water and may form sulfuric acid on contact with coal and ash. These acids lower the ph of groundwater and potentially allow metals in surrounding rock to dissolve and migrate in a plume along the groundwater pathway. The act of injecting CO 2 will also create changes in temperature, pressure, ph, rock-water chemistry, and gas-water chemistry. If injected too quickly after a burn, the CO 2 could boil, increasing the pressure in the cavity. If injected with too much pressure, the water that has filled the cavity and dissolved volatile organics and ash material could be flushed out or fractures could be created. CO 2 that dissolves decreases water ph, and CO 2 that does not dissolve can push up on the cavity, putting pressure on it. Should CO 2 migrate up and out of the geologic storage location, it is likely to kill plants and ground-dwelling animals at the discharge location. Slow, non-catastrophic natural leaks of 49 Chaney and van Bibber 2006 Chapter 2 50 CIRI's Coal Development plans presentation to the Alaska Bar Association Environmental Law/Natural Resource Law Section Nov Keith et al

27 CO 2 continue to kill forests in the Sierra Mountains in California, and very large discharges from natural sources have in the past asphyxiated plants, humans, and animals. 52 This means that it is not feasible to safely remove product gas then use the same wells to pump CO 2 back down into the burned out coal seams at the proposed Beluga project. At the very least, injection wells will need to be drilled much further down, and the geology will need to be favorable both for safe UCG reactions at the proposed 200 m coal seam level and the 800+ m CO 2 storage level. The seismic analysis during the feasibility period of the project will be critical to determine whether there is a risk of air entering the coal seam during the burn, and further analysis post-burn may be required to determine the risk of CO 2 leaks from deep storage locations if earthquakes open new faults. In the feasibility studies for the UCG project, the true feasibility and costs of carbon capture from both the product well gas and the power plant need to be presented, along with the feasibility, costs, and risks of geologic storage. Summary The CIRI Beluga UCG project proposes to take components of two emerging technologies and join them together. This will require scrutiny of both components. The UCG component has been conducted successfully in pilot scale tests around the world; the one longterm plant in existence (Angren, Uzbekistan) does not have environmental information readily available. The operators will need to satisfy both the requirements of producing high quality gas and the requirements of maintaining environmental integrity. Given the proximity of the proposed project to the Beluga River, Cook Inlet tidelands, and the Castle Mountain Fault, it is particularly important to examine the hydrogeologic and geophysical details to ensure Geologic conditions that preventing subsidence o At least 200 m below ground o Structural integrity of host rock o Geophysical modeling of temperature/pressure stresses on fractures Siting to prevent contaminant migration o Impermeable caprock o a distance at least 25 times the depth of the coal seam between the seam and aquifer o a minimum of 1.6 km from rivers and lakes o a minimum of 0.8 km from major faults 52 Wilson et al

28 o seams should be thick and widely separated to prevent burn-through between seams In addition to the conditions that must be satisfied for coal gasification, conditions also must allow for carbon capture and sequestration. No UCG projects currently capture and sequester carbon. Separating CO 2 and transporting it to an appropriate declining oil field will require extra financing and negotiations with Cook Inlet oil and gas companies. If the CO 2 is to be injected back into the coal fields, injection wells at least 800 m deep far deeper than the 200 m deep target coal seam will need to be drilled and the geologic conditions at that depth will need to be sufficient to entrain the CO 2 for thousands of years. 28

29 Bibliography Bakker, W High temperature corrosion in gasifiers. Mat Res Vol. 7 (1). Biello, D Burying climate change: efforts begin to sequester carbon dioxide from power plants. Scientific American. September 22. Bluemink, E CIRI plans coal-to-gas electrical power plant. Anchorage Daily News. Anchorage, AK October 9. Bowen, BH and MW Irwin UCG CCTR Basic Facts File #12 (powerpoint). Indiana Center for Coal Technology Research. [Online] Boysen, JE, MT Canfield, JR Covell, and CR Schmit "Detailed evaluation of process and envrionmental data from the Rocky Mountain I underground coal gasification field test: final report." Gas Research Institute and US Department of Energy. Chicago, IL and Morgantown, WV. Burton, E, J Friedmann, and R Upadhye "Best practices in underground coal gasification (draft)." Lawrence Livermore National Laboratory. Livermore, CA. Campbell, JH, FT Wang, SW Mead, and JF Busby Groundwater quality near an underground coal gasification experiment J Hydrology 44: Carls, MG, L Holland, M Larsen, TK Collier, NL Scholz, and JP Incardona Fish embryos are damaged by dissolved PAHs, not oil particles. Aquatic Toxicology 88 (2): Chaney, R and L Van Bibber Beluga coal gasification study. Phase I final report for subtask DOE/NETL-2006/1248. July. 20Study9_15_06.pdf Clayton, G Biogenic and thermogenic origins of natural gas in Cook Inlet Basin, Alaska. AAPG Vol 64 DOI: /2F91944F-16CE-11D C1865D. Clean Air Task Force Coal without carbon: an investment plan for federal action. Boston, MA. Creedy, DP and K Garner Clean energy from underground coal gasification in China. DTI Cleaner Coal Technology Transfer Program Report No. COAL R250, DTI/Pub URN 03/1611. Friedmann, SJ, R Upadhye, and F-M Kong Prospects for underground coal gasification in a carbonconstrained world. Energy Procedia 1: Goldsmith, S and N Szymoniak Energy Analysis: propane from the North Slope - could it reduce energy costs in the Interior? Institute of Social and Economic Research for Alaska Natural Gas Development Authority. Anchorage, AK. Gosnald, WD Postgasification thermal regime of the Rocky Mountain I underground coal gasification test site. Gas Research Institute and US Department of Energy. Washington, DC. 29

30 Gregg, DW Ground subsidence resulting from underground coal gasification. Lawrence Livermore National Laboratories. UCRL Livermore, CA. Hurst, S Snohvit CO2 storage underway. Petroleum News Vol 13 (21). May 25. Institute of Social and Economic Research Alaska electric power statistics University of Alaska Anchorage for Alaska Energy Authority. Anchorage, AK. Jaramillo, P, WM Griffin, and HS Matthews Comparative life-cycle air emissions of coal, domestic natural gas, LNG and SNG for electricity generation. Environ Sci and Technol 41: Keith, DW, JA Giardina, MG Morgan, and EJ Wilson Regulating the underground injection of CO2. Environ Sci and Technol, pp. 499A-507A. Kharaka, YK, DR Cole, SD Hovorka, WD Gunter, KG Knauss, and BM Freifield Gas-water interactions in Frio Formation following CO 2 injection: implications for the storage of greenhouse gases in sedimentary basins. Geology 34 (7): Maev, S Development of a UCG based project in Canada. Twenty-fifth annual international Pittsburgh coal conference. Paper Pittsburgh, PA. Rogut, J Hydrogen Oriented Underground Coal Gasification.Twenty-Fifth Annual International Pittsburgh Coal Conference. Paper Pittsburgh, PA Shafirovich, E and A Varma UCG: a brief review of current status. Ind Eng Chem Res Vol 48: Shafirovich, E, M Mastalarz, J Rupp and A Varma Potential for UCG in Indiana: Phase I report to the Indiana Center for Coal Technology Research. Purdue University, Indiana. Simbeck, D Carbon separation and capture from energy systems: the forms and costs of separation and capture. Complements to Kyoto: technologies for controlling CO 2 emissions. National Academy of Engineering. Washington, DC. Sturmer, DM, DD LaPointe, JG Price, and RH Hess Assessment of the potential for carbon dioxide sequestration by reactions with rocks in Nevada. Nevada Bureau of Mines and Geology Report 52. University of Nevada, Reno. Thambimuthu, K et al IPCC special report on carbon dioxide capture and storage, Chapter 3. International Panel on Climate Change, Walter, K Fire in the Hole. Sci Tech Rev, pp Wilson, EJ, TL Johnson, and DW Keith Regulating the ultimate sink: managing the risks of geologic CO 2 storage. Env Sci Technol 37:

31 Appendix A: UCG reactions and Syngas Reactions The table below lists the primary reactions found in producing syngas (SNG) or UCG product gas (UCG). The most important reaction for both is Step 1, the actual transformation of coal into gases. Other reactions either provide the heat to drive the desired reaction (burning coal) or are reactions to produce a desired product (methane, etc). Adapted from Burton et al 2006, Table 4-1. Step Reaction Name Chemical reaction Chemical equation UCG SNG Notes 1 Gasification reaction (Water- Gas Shift reaction) 2 Shift conversion 3 Methanation Hydrogenating gasification Partial oxidation (incomplete combustion of coal) Oxidation (complete combustion of coal) Carbon + water hydrogen and carbon monoxide Carbon monoxide + water hydrogen and carbon dioxide Carbon monoxide and hydrogen methane and water Carbon + hydrogen methane Carbon + oxygen carbon monoxide Carbon + oxygen carbon dioxide C + H 2 O H 2 + CO x x CO + H 2 O H 2 + CO 2 x x CO + 3 H 2 CH 4 + H 2 O C + 2H 2 CH 4 C + ½ O 2 CO side reaction side reaction C + O 2 CO 2 x x x main reaction; makes hydrogen for combustion; requires heat from steps 5,6 react CO to make more hydrogen increase methane content of gas; to make hydrogen from natural gas, reverse the reactions increase methane content of gas releases heat to drive step 1 releases heat to drive step 1 7 Boudouard reaction Carbon + carbon dioxide carbon monoxide C + CO 2 2CO side reaction requires heat, provides CO for steps 2,3 31

32 Appendix B: Natural Gas Processing from 32