Impacts of Uncertainty in Energy Project Costs |
|
From the late 1970s through 2002, steel, cement, and concrete prices followed a general downward trend. Since then, however, iron and steel prices have increased by 8 percent in 2003, 10 percent in 2004, and 31 percent in 2005. Although iron and steel prices declined in 2006, early data for 2007 show another increase. Cement and concrete prices, as well as the composite cost index for all construction commodities, have shown similar trends but with smaller increases in 2004 and 2005 (Figure 9). Recent increases in the costs of basic commodities and increases in capital costs for energy equipment and facilities could have significant effects on future energy supplies and consumption. Higher capital costs could change both the competition among fuels and technologies and the marginal costs of new energy supplies. In the electric power sector, for example, capital costs are generally lower for generating plants that use fossil fuels than for plants that use nuclear or renewable fuels. If capital costs increased on a proportional basis for plants of all types, then capital-intensive nuclear and renewable power plants would become even less competitive with fossil-fired plants when new capacity is planned. In addition, over the long term, higher capital costs would lead to higher energy prices, which in turn could slow the growth of energy consumption. The AEO2008 version of NEMS includes updated assumptions about the costs of new power plants, the costs of drilling and pipeline construction in the oil and natural gas industry, refinery costs, and capital costs in the LNG supply chain. In the reference case, energy project costs are assumed to level off over the long term. To examine the effects of different assumptions about future costs, high and low energy project cost cases were developed, assuming higher and lower costs than in the reference case. Power Plant Construction In the electric power industry, cost estimates for individual construction projects to be completed over the next decade have increased by 50 percent or more in recent years [53]. Increased costs have been reported for power plants of all types, including coal, nuclear, natural gas, and wind. The Handy-Whitman index for electric utility construction (which is used as a proxy for all electric power industry projects) provides an average cost index for six regions in the United States, starting from 1973. A simple average of the regions is used in Figure 10 to show the national trend for power plant construction relative to the cost index for construction materials. The two indexes diverge in the early 2000s, when power plant construction costs began to show a flat to slightly increasing trend, while general construction costs continued to decline. With the sharpest increases in electric utility construction costs occurring over the past 3 years, the electric utility construction cost index for 2007 is 17 percent higher than its low point in 2000. Oil and Natural Gas Industry Exploration and Production According to the American Petroleum Institute’s Joint Association Survey of Drilling Costs (JAS), the average real cost of drilling an onshore well almost doubled in 2004 and increased by another 10 percent in 2005. The increases are attributable in part to the increased drilling activity brought on by higher prices for crude oil and natural gas; however, there is a great deal of uncertainty as to whether the recent escalation in drilling costs represents a fundamental shift in the drilling services industry or is a temporary aberration that will be corrected in the near term. Natural Gas Pipelines Historical trends in pipeline construction costs are more difficult to identify, because the cost data are not readily available; however, average real capital costs for lower 48 pipeline construction appear to have increased by some 70 percent over the past 3 years. Anecdotal evidence suggests that new estimates for the cost of constructing an Alaska pipeline are 50 percent higher than the estimates published in May 2002, and estimates for a Mackenzie Delta pipeline also are higher than the preliminary estimates from 2003. LNG Facilities Construction cost estimates for new natural gas liquefaction facilities scheduled to come on line between 2008 and 2011 increased by 50 percent in 2006 relative to those reported a year earlier for the same period. Some of the increase may be due to strong growth in demand for LNG liquefaction capacity. This cost pressure will not persist as markets adjust and additional projects are announced and completed; however, a portion of the increase is due to increased material costs, shortage of experienced workers, and construction bottlenecks that are likely to persist or take longer to resolve. The costs for regasification facilities and receiving terminals have also increased sharply—by more than 50 percent— over the past few years. Based on contracts signed between 2000 and 2006, LNG shipping costs have also risen by more than 7 percent over the past few years. Petroleum Refineries and Ethanol Plants The Nelson-Farrar refinery construction cost indexes, which track overall costs for refinery construction, show a 30-percent increase from 2003 to 2005 in real dollar terms. Similarly, the Chemical Engineering Plant Cost Index (CEPCI) shows a significant increase in ethanol plant construction costs over recent years. Because there has not been a significant increase in U.S. refining construction activity over the past few years, cost increases in the petroleum refining sector largely reflect higher prices for the various commodities used in the refining industry (steel, nickel, cobalt, etc.) rather than significant increases in demand for refinery services and equipment. Case Descriptions Reference Case The AEO2008 reference case includes updated information on the current costs of construction and investment in the energy industry, based on recent data and estimates that show higher costs than were assumed for AEO2007. In most of the AEO2008 cases, the higher cost levels are assumed to continue throughout the projections. For the electric power sector, initial costs for all technologies are 15 percent higher than those in AEO2007 and continue to be higher throughout the projection, although overnight costs fall over time as a result of technology learning. For the oil and natural gas industry, regional drilling costs are calculated annually from econometrically derived equations, which are based on historical data from the American Petroleum Institute’s JAS, and estimates of the number of wells being drilled and the average depth of each well. The cost increases seen after 2003 are represented by an explicit multiplier that captures the combined impacts of various cost factors other than drilling activity and well depth. In the reference case, the cost escalation factor is applied and held constant over the projection, but its effect is partially offset by an annual technology improvement factor that reflects learning and increased efficiency. Pipeline construction costs are based on average construction cost data filed between 1992 and 2008, and they are assumed to remain constant through 2030. The reference case also assumes that the recent, higher estimates for an Alaska pipeline and a pipeline from the Mackenzie Delta remain constant through 2030. Construction costs for new natural gas liquefaction facilities were increased by 50 percent in AEO2008 to match the 2006 cost estimate for facilities scheduled for completion between 2008 and 2011. The construction costs are assumed to remain constant at that level through 2015, then decline to only 15 percent above their pre-2006 levels in 2018 as the market adjusts, after which the costs are assumed to remain constant at the 2018 level through 2030. LNG shipping costs and construction costs for regasification facilities are assumed to be 15 percent and 7 percent higher, respectively, than their 2006 level throughout the AEO2008 projection. Construction costs for refineries and for ethanol production plants are assumed to remain constant at 2006 levels through 2030, based on the Nelson-Farr index and CEPCI, respectively. High Energy Project Cost Case The high energy project cost case assumes that the cost of construction will continue to rise. For electricity generation plants, the base capital cost for all technologies rises at a rate of 2.5 percent per year— similar to the average increase over the past 3 years— through 2030, offset in part by learning effects. For the oil and natural gas industry, the escalation factor for drilling costs is assumed to increase to twice its original value by 2010 and remain constant thereafter. It is offset in part by an annual technology improvement factor. Pipeline construction costs are assumed to start at the reference case level but grow to about 25 percent above the reference case level in 2030. LNG liquefaction costs match the reference case increase through 2008 and add an additional 20 percent thereafter. Construction costs for LNG regasification facilities are 15 percent above the reference case level in 2008 and then held constant through 2030. LNG shipping costs are increased to 7 percent above the reference case level in 2008 and then held constant through 2030. For the refining sector, construction costs are increased above the reference case level by a factor equal to the percentage difference between the 2004 and 2006 Nelson-Farrar index values and held constant. Construction costs for corn and cellulosic ethanol plants are treated similarly, using the CEPCI. Low Energy Project Cost Case The low energy project cost case generally assumes that the cost of construction will decline to the levels of 5 to 10 years ago. For the electricity sector, the 15-percent capital cost escalation factor included in the reference case is phased out over 10 years, so that overnight construction costs for all generating technologies are 15 percent lower than those in the reference case by 2017. For the oil and natural gas industry, the drilling cost escalation factor applied in the reference case is phased out by 2010. Pipeline construction costs start at the reference case level but decline gradually to about 25 percent below the reference case level in 2030. For LNG liquefaction facilities, construction costs are reduced gradually from those in the reference case, returning to 2006 levels by 2015 and remaining constant thereafter. Similarly, construction costs for LNG regasification facilities and costs for LNG shipping costs decline gradually from reference case levels, return to 2006 levels by 2018, and remain constant thereafter. Refinery construction costs are assumed to return to 2004 levels by 2008 and then remain constant through 2030. Results Electricity: Capacity Additions and Generation The projected mix of generating capacity types added in the electric power sector from 2006 to 2030 does not vary significantly among the reference, high energy project cost, and low energy project cost cases, because increases or decreases in construction costs have similar impacts on new builds for all technology types on a percentage basis. For example, coal-fired technologies provide about 40 percent of all new capacity additions in each of the three cases. More capital-intensive technologies, including nuclear and renewables, are affected somewhat more, however, than those with lower capital costs, including natural-gas- and coal-fired plants. In the high energy project cost case, coal-fired capacity additions are reduced by 13 gigawatts from the reference case level, but with higher costs leading to higher electricity prices and lower demand, less new generating capacity is needed overall. As a result, the coal share of new builds remains almost the same as in the reference case. The technology most affected is nuclear power: no new nuclear capacity is built before 2030 in the high energy project cost case (Figure 11). Renewable capacity additions are 17 percent lower than in the reference case, but total generation from renewable plants is about the same in order to meet the requirements of State and regional RPS programs. The increase in renewable generation comes primarily from biomass co-firing at existing coal plants. Because they are the least expensive to build, natural gas capacity additions increase in the high energy project cost case relative to the reference case, meeting 43 percent of new capacity needs. As a result, natural- gas-fired generation in 2030 is 22 percent higher than in the reference case. Average electricity prices in 2030 are 9 percent higher in the high energy project cost case than in the reference case. In the low energy project cost case, more capacity of all types except natural gas is added over the projection period. The largest increase is in nuclear capacity additions, which are 10 gigawatts higher than in the reference case. Because capital costs make up a smaller share of total costs for natural-gas-fired capacity additions than for other technologies, they are slightly less economical in the low energy project cost case and about 3 gigawatts lower than in the reference case. The fuel shares of total generation in 2030 are similar in the low energy project cost case and the reference case, with a small decrease in the natural gas share (to 13 percent, compared with 14 percent in the reference case). The nuclear share of total generation increases from 18 percent in the reference case to 19 percent in the low energy project cost case. Electricity prices in 2030 are 4 percent lower in the low energy project cost case than in the reference case. Natural Gas: Supply, Consumption, and Prices Natural gas supply volumes are determined primarily by consumption levels, particularly for electric power generation. Capital costs play a role in determining the relative shares of total supply derived from conventional, unconventional, LNG imports, and other supply categories. Total domestic natural gas production in 2030 differs by 1.6 trillion cubic feet between the low and high energy project cost cases (Figure 12). Lower 48 onshore production differs by 1.1 trillion cubic feet between the two cases, with conventional and unconventional production accounting for 0.6 and 0.5 trillion cubic feet of the total difference. Production from Alaska and offshore production differ by 0.4 and 0.2 trillion cubic feet, respectively, between the low and high energy project cost cases. In 2030, total net natural gas imports are 3.1 trillion cubic feet in the high energy project cost case and 3.4 trillion cubic feet in the low energy project cost case. LNG imports account for more than 80 percent of total net natural gas imports in all the cases, and the capital costs for LNG facilities are by far the largest component of LNG supply costs. Net LNG imports are 2.5 trillion cubic feet in 2030 in the high energy project cost case, compared with 2.8 trillion cubic feet in the low energy project cost case. The picture for net pipeline imports of natural gas from Canada and Mexico is more complex. In the reference case, because recent cost estimates indicate that a Mackenzie Delta pipeline would not be economical to build [54], net pipeline imports total only 0.3 trillion cubic feet in 2030. In the low energy project cost case, a Mackenzie pipeline would begin operation in 2014, providing about 420 billion cubic feet per year through 2030; as a result, net pipeline imports to the United States total 0.5 trillion cubic feet in 2030. In the high energy project cost case, with higher U.S. prices for natural gas inducing more production and exports from Canada, net U.S. pipeline imports total 0.6 trillion cubic feet in 2030. Differences in total natural gas consumption in the energy project cost cases are determined primarily by the different amounts used for electricity generation. Because coal, nuclear, and renewables are more competitive with natural gas in the low energy project cost case and capture a larger share of new capacity additions, natural gas consumption in the electric power sector in 2030 is 0.4 trillion cubic feet lower than the reference case projection of 5.0 trillion cubic feet (Figure 13). As a result of the lower level of natural gas use for electricity generation in the low energy project cost case, total domestic natural gas consumption and prices in 2030 are lower than in the reference case: consumption by 0.3 trillion cubic feet (from 22.7 trillion cubic feet in the reference case) and wellhead gas prices by $0.33 (2006 dollars) per thousand cubic feet (from $6.63 in the reference case) (Figure 14). In the high energy project cost case, new natural-gasfired electricity generation capacity is considerably less expensive than competing technologies, and the natural gas share of capacity additions increases, resulting in higher total consumption and prices for natural gas than in the reference case. The increase in consumption for electricity generation leads to higher total domestic consumption (by 1.1 trillion cubic feet) and higher price levels (by $0.49 per thousand cubic feet) for natural gas than in the reference case. Because of the higher prices, natural gas consumption in the residential, commercial, and industrial sectors in 2030 is lower than projected in the reference case. Petroleum Liquids Supply A large part of the domestic oil resource base has been produced, and new oil reservoir discoveries are expected to be smaller, more remote (offshore deepwater, for example), and more costly to exploit. With a few exceptions—namely, deepwater Gulf of Mexico and offshore Alaska—the remaining domestic petroleum basins have been significantly depleted. Consequently, EOR using miscible CO2 is the primary extraction technique expected to keep onshore oil production at a relatively high level through 2030. The assumptions in the low and high energy project cost cases were applied only to the domestic resource. Depletion of domestic oil resources constrains the high and low energy project cost assumptions from having a significant impact on domestic oil production. The low and high energy project cost cases would show larger impacts if the assumptions were applied to world liquid supplies. A slow, continuous decline in oil production is projected for the onshore United States, even with the relatively high oil prices [55]. Future domestic onshore oil production is dominated by large oil fields that were discovered decades ago, and EOR only extends their productive life. For example, although the Prudhoe Bay Field started production in 1976, the largest share of Alaska’s oil production still comes from Prudhoe Bay. Although large oil fields on Alaska’s North Slope came into production more recently [56], the long-term trend is for Alaska’s oil production to decline as the Prudhoe Bay Field declines. The AEO2008 reference case and low and high energy project cost cases include constant or declining U.S. oil production, as smaller and smaller new fields come into production while the larger existing fields continue to be depleted [57]. In the low energy project cost case, total domestic oil production in 2030 is 18,000 barrels per day higher than projected in the reference case. In the high energy project cost case, higher drilling costs reduce both the rates of return on oil production and the cash flow of oil producers, and as a result total domestic production in 2030 is about 300,000 barrels per day lower than in the reference case. Because EOR is highly capital-intensive, most of the variation in domestic oil production across the three cases reflects differences in EOR production. In the reference case, CO2 EOR production in 2030 totals 1.31 million barrels per day, as compared with 1.33 million barrels per day in the low energy project cost case and 980,000 barrels per day in the high energy project cost case. For deepwater production in the Gulf of Mexico, the reference case projects an increase from about 970,000 barrels per day in 2006 to 2.0 million barrels per day from 2013 through 2019, followed by a decline to 1.6 million barrels per day in 2030. The projections in the low energy project cost case are nearly the same, because the constraints on deepwater development are not prices and costs but long development lead times and limited infrastructure. In the high energy project cost case, the capital intensity of deepwater development constrains oil production in the Gulf in the earlier years, with a peak production level of 1.9 million barrels per day from 2013 through 2019. As oil prices increase later in the projection period, however, small deepwater fields that were uneconomical in earlier years begin to be developed. In 2030, deepwater production in the Gulf is about 30,000 barrels per day higher in the high energy project cost case than projected in the reference case [58]. Both CTL and BTL production are also capitalintensive and vary significantly on a percentage basis across the three cases. Combined production from CTL and BTL facilities is about 620,000 barrels per day in 2030 in the low energy project cost case, compared with 510,000 barrels per day in the high energy project cost case. The only other petroleum supply category significantly affected in the energy project cost cases is natural gas liquids (NGL). In the high energy project cost case, which projects considerably more natural gas production than the low case, NGL production is also higher, at 1.6 million barrels per day, compared with 1.5 million barrels per day in the low case. As a result, the difference in combined CTL and BTL production between two cases is almost completely offset by the difference in NGL production. Crude oil prices are not projected to vary significantly across the three cases. The reference case projects a price of $70.45 per barrel for low-sulfur light crude oil in 2030 (2006 dollars), compared with $70.33 per barrel in the low energy project cost case and $70.65 per barrel in the high energy project cost case. Accordingly, total domestic consumption of petroleum liquids does not vary by much, at 22.7 million barrels per day in the high energy project cost case and 22.8 million barrels per day in the low energy project cost case. Imports of crude oil and liquid fuels make up the difference between the projections for liquids production and consumption in each case, varying from 55.5 percent of total U.S. supply in 2030 in the high energy project cost case to 54.0 percent in the low energy project cost case. As noted above, the impacts would be more significant if the assumptions in the low and high energy project cost cases were applied to global markets. Contact: Laura Martin/Philip Budzik |
