Gas to Coal Competition in the U.S. Power Sector
Steven Macmillan, Alexander Antonyuk, Hannah Schwind
© OECD/IEA 2012
The views expressed in this paper do not necessarily reflect the views or policy of the International Energy Agency (IEA) Secretariat or of its individual member countries. The paper does not constitute advice on any specific issue or situation. The IEA makes no representation or warranty, express or implied, in respect of the paper’s content (including its completeness or accuracy) and shall not be responsible for any use of, or reliance on, the paper. Comments are welcome, directed to
[email protected].
© OECD/IEA, 2013
INTERNATIONAL ENERGY AGENCY The International Energy Agency (IEA), an autonomous agency, was established in November 1974. Its primary mandate was – and is – two-fold: to promote energy security amongst its member countries through collective response to physical disruptions in oil supply, and provide authoritative research and analysis on ways to ensure reliable, affordable and clean energy for its 28 member countries and beyond. The IEA carries out a comprehensive programme of energy co-operation among its member countries, each of which is obliged to hold oil stocks equivalent to 90 days of its net imports. The Agency’s aims include the following objectives: n Secure member countries’ access to reliable and ample supplies of all forms of energy; in particular, through maintaining effective emergency response capabilities in case of oil supply disruptions. n Promote sustainable energy policies that spur economic growth and environmental protection in a global context – particularly in terms of reducing greenhouse-gas emissions that contribute to climate change. n Improve transparency of international markets through collection and analysis of energy data. n Support global collaboration on energy technology to secure future energy supplies and mitigate their environmental impact, including through improved energy efficiency and development and deployment of low-carbon technologies. n Find solutions to global energy challenges through engagement and dialogue with non-member countries, industry, international organisations and other stakeholders.
IEA member countries: Australia Austria Belgium Canada Czech Republic Denmark Finland France Germany Greece Hungary Ireland Italy Japan Korea (Republic of) Luxembourg Netherlands New Zealand Norway Poland Portugal Slovak Republic © OECD/IEA, 2013 Spain International Energy Agency Sweden 9 rue de la Fédération Switzerland 75739 Paris Cedex 15, France Turkey www.iea.org United Kingdom United States Please note that this publication
is subject to specific restrictions that limit its use and distribution. The terms and conditions are available online at http://www.iea.org/termsandconditionsuseandcopyright/
The European Commission also participates in the work of the IEA.
© OECD/IEA 2013
Coal to Gas Competition in the U.S. Power Sector
Table of Contents Executive Summary ................................................................................................................... 3 Introduction .............................................................................................................................. 4 Background and Recent Trends ................................................................................................. 6 Page | 1 A look back at 1990‐2011: The US dash for gas and its limits ................................................... 6 The evolution of gas‐coal competition during the past two decades ............................... 6 Assessing current potential switchable gas capacity ......................................................... 8 Characterising the portfolio of switchable capacity .......................................................... 8 Factors affecting utilisation of switchable capacity .................................................................. 9 Fuel prices ........................................................................................................................ 10 Coal contracts .................................................................................................................. 11 Variability in plant level efficiency ................................................................................... 11 Technology factors ........................................................................................................... 13 Role of regulation and policy ........................................................................................... 14 Environmental regulation ................................................................................................ 19 Estimates of switching: observed and projected ...................................................................... 20 Observed switching ................................................................................................................. 20 Projected switching ................................................................................................................. 21 Conclusions on projection for switching by 2017 .................................................................... 23 Appendices ............................................................................................................................. 25 Appendix A ............................................................................................................................... 25 Appendix B ....................................................................................................................... 27 References .............................................................................................................................. 30 List of Figures Figure 1 • Evolution of the US wellhead gas price over 1980‐2012 ............................................... 6 Figure 2a • Coal and gas power generation, 1990‐2011 ................................................................. 7 Figure 2b • Coal and gas shares in power generation, 1990‐2011 .................................................. 8 Figure 3 • Switching potential in 2011 ........................................................................................... 9 Figure 4 • United States Henry Hub prices, 2009‐12 ................................................................... 10 Figure 5 • Average length of US non‐lignite coal contracts for power generation, by state ....... 12 Figure 6 • Range of thermal efficiency in US coal and CCGT plants, 2011 ................................... 12 Figure 7 • Gas switch price based on coal price, for different gas efficiencies and fixed coal efficiency ...................................................................................................................... 13 Figure 8 • Deregulation in the US power sector and retail electricity prices, 2011 ..................... 15 Figure 9 • Share of coal and gas generated from the regulated and unregulated sector, 2011 .. 16 Figure 10 • CCGT capacity, GW, 2011 ............................................................................................. 25 Figure 11 • Coal capacity, GW, 2011 ............................................................................................... 25 Figure 12 • Gas costs for power, December 2011 .......................................................................... 26 Figure 13 • Coal costs for power, December 2011 ......................................................................... 26 List of Tables Table 1 • Individual Case Studies .................................................................................................... 27
Coal to Gas Competition in the U.S. Power Sector
© OECD/IEA 2013
Acknowledgements The authors would like to acknowledge the members of the Gas, Coal and Power Division (GCP) of the IEA, in particular Carlos Fernández Alvarez, László Varró and Anne‐Sophie Corbeau, who all provided considerable guidance and feedback to support the completion of this paper. Input was Page | 2 also provided by Warner Ten Kate and Johannes Trüby, both secondees to the IEA during the course of 2012. Steven Macmillan was a secondee to the International Energy Agency with the support of Origin Energy (Australia). Hannah Schwind worked as an intern in GCP Division. Alexander Antonyuk left the IEA in early 2013. Thanks must also go to Associates of the IEA’s Coal Industry Advisory Board (CIAB), who provided valuable input. The Energy Information Administration in the United States assisted in the provision of data and clarification of data points.
© OECD/IEA 2013
Coal to Gas Competition in the U.S. Power Sector
Executive Summary This paper analyses the impact of the shale gas revolution in North America on the American power sector and contributes to analysis of the economic implications of environmental policies. Lower variable production costs triggered by the US shale gas revolution have enhanced the competitiveness of natural gas‐fired power plants during 2012, especially with respect to coal. Page | 3 However, in light of considerable regional diversity across the US power sector, low fuel prices are only one of many elements that determine the role of gas in the power system. This paper provides an overview of factors that influence this switch and identifies a sample of US states that have a meaningful fuel‐switching potential in terms of unused combined‐cycle gas capacity. From a sample of 18 states (accounting for around 75% of unused combined cycle gas output and around 46.5% of net generation in the United States in 2011), estimates are given for switching observed from late 2011, and a projection of switching to occur by 2017. It is assumed that natural gas prices will increase to USD 4.7/million British thermal units (MBtu) by 2017. Similarly, assumptions are made about other factors, such as the retirement of coal‐fired plants, construction of new combined cycle plants and changes in US environmental regulation affecting coal‐fired plants.
Coal to Gas Competition in the U.S. Power Sector
© OECD/IEA 2013
Introduction The United States is by far the largest consumer of natural gas in the world (690 billion cubic meters (bcm) in 2011, or 21% of the world’s gas consumption (IEA, 2012a)) as well as the second largest coal user (697 million tonnes of coal equivalent [Mtce] or 13% of global coal consumption) Page | 4 (IEA, 2012b), albeit standing well behind China. The United States power sector could be considered as a country on its own: its gas use, almost a third of total US gas consumption, amounted to around 230 bcm in 2011, which is equivalent to the combined liquefied natural gas (LNG) exports of the Middle East and Asia Oceania regions. The use of thermal coal by US power generators was 95% of US coal demand, larger than the global seaborne thermal coal trade. Hence the US power sector forms a significant part of both the global gas and coal markets, and any change affecting it in a positive or negative way can have profound implications for the world. Such changes are already under way. The shale gas revolution that resulted in the dramatic and unforeseen increase of US gas production by over 100 bcm over 2007‐11 resulted not only in a quasi‐independence of the United States from LNG imports – and therefore on global gas markets – but also in a significant increase of gas use in the power sector. Over the past three years, as US natural gas prices remained low (on average at around USD 4/MBtuover 2009‐11, and below USD 3/MBtu for the first nine months of 2012), gas‐fired plants have been slowly but surely eroding the position of coal‐fired plants as the first energy source for power generation. In 2007, net coal‐fired electricity generation was 2.25 times that of net gas‐fired generation; in 2011 the ratio dropped to only 1.70. The year 2012, with its remarkably low gas prices, continued to close the bridge between the two fuels, resulting in a 1.16 ratio for the first seven months This shift from coal to gas for power generation, together with increased LNG production in the United States has ramifications for energy markets worldwide. On global gas markets, LNG supplies once earmarked for the United States were available to be redirected to Japan after the Fukushima accident. Additionally, the United States now plans to become an LNG exporter by 2015, deterring investors previously targeting the North American market for LNG exports. Moreover, record low natural gas prices in the United States give a competitive edge to the US industry – notably petrochemicals and fertilisers – while other regions, such as Japan and Korea, face record gas prices almost six times higher. On global coal markets, the lack of US domestic demand for coal in the power sector forced US thermal coal producers to look for other markets, resulting in a marked increase in US coal exports, most of which found their way into Europe. As a result, coal is winning market share against gas in the European power sector, where gas is simply no longer competitive. The irony is that the United States, which did not sign the Kyoto protocol, sees greenhouse gas (GHG) emissions reduced through cheap gas, while Europe, which was the first to put in place an Emission Trading Scheme, may actually see its GHG emissions increase in the power sector through gas to coal switching. These changes, as significant as they have been already, raise two questions that this report investigates: whether the penetration of natural gas in the US power sector might have been more significant except for limitations on switching in particular US regions; and whether this trend is likely to continue over coming years. The first part of this report analyses the historical developments of gas and coal in the power generation sector. In particular, it examines in depth the historic background to coal to gas fuel switching in the United States, with a special focus on the range of factors that have influenced or limited the rate of switching. These factors include relative fuel prices, the location of excess gas‐fired capacity in comparison to existing coal‐fired plants, technological factors, the duration
© OECD/IEA 2013
Coal to Gas Competition in the U.S. Power Sector
of coal contracts between producers and power generators, the role of regulation (including new regulations covering emissions for coal‐fired plants) and transmission constraints in both power and gas markets. This analysis shows that, despite a large over‐capacity of gas‐fired plants at the country level, the limitations mentioned above currently restrict switching to a theoretical ceiling of 613 terawatt hours (TWh) in a year (which is equivalent to 13% of total US power generation in 2011). Page | 5 The second part provides an estimate of switching that occurred in the 12 month period beginning in October 2011 (when Henry Hub prices fell below USD 4/MBtu for a sustained period), and a projection for switching to occur by 2017. These estimates are based on an analysis of the situation in 18 American states that account for approximately 75% of the unused combined‐cycle gas turbine (CCGT) capacity in the United States.1 The conditions for switching vary between states. The key factors examined are relative fuel prices, the length of coal contracts, the existing unused CCGT potential, an estimate of coal plant retirements due to the legislative requirements on air quality, and an estimate of the expansion in the CCGT fleet occurring by 2017. The meaning of the term “fuel switching” varies depending on the time horizon chosen. In the short term, where no changes in the power plant fleet are possible, fuel‐switching designates instantaneous adjustments of the merit‐order as a response to temporary changes in variable production costs. For example, coal‐to‐gas fuel‐switching within the existing generation capacity signifies that some gas‐fired power plants have cheaper variable production costs than coal, thus triggering changes in the merit order. This “short‐run” switching can occur either within days or over periods of many months, and is of interest in terms of the resilience of the electricity systems, as well as for its impact on commodity markets. In the long run, the power plant fleet itself can change due to additions of capacity, retirements, fuel conversions and retrofits. Those factors then have long‐term effects on the power mix. Both “short‐run” and “long‐run” switching are examined in this paper, the former over months (rather than days) and the latter by analysing the impact of projected price changes, additions and retirements out to 2017.
1
The total is comprised of 14 states, in addition to four states (California, Maine, Oregon and Washington) where there is spare CCGT capacity but no coal output to switch. The 14 states are Alabama, Arkansas, Arizona, Florida, Georgia, Louisiana, Michigan, New Jersey, Nevada, New York, Oklahoma, Pennsylvania, Texas and Virginia.
Coal to Gas Competition in the U.S. Power Sector
© OECD/IEA 2013
Background and Recent Trends A look back at 1990‐2011: The US dash for gas and its limits Natural gas‐fired plants have provided around 60% of additional US power generation over the past two decades (1990‐2011). While coal‐fired plants also met additional power demand in the 1990s, their contribution to incremental US power output remained flat over most of the last decade and has been dropping significantly since 2009. The current situation as of late 2012 was therefore the result of a slow evolution which took place over 20 years and resulted in a significant overbuild of gas‐fired capacity. The geography of the US power, coal and gas sectors means that only part of this surplus gas‐fired capacity can replace coal‐fired plants, even when pricing signals would dictate switching. The following analyses in depth factors which need to be taken into account while looking at future switching from coal to natural gas
The evolution of gas‐coal competition during the past two decades Liberalisation of the US power sector accelerated in the 1990s. By this time, the United States had already developed a well‐functioning gas market and Combined Cycle Gas Turbine (CCGT) technology was technologically mature. It is also worth mentioning that in the 1970s and 1980s, US energy policy restricted the use of gas in power generation due to limitations in the supply of gas (IEA, 2012c). As gas was then considered a scarce resource, its share in the power generation sector was kept artificially low during that time. Figure 1 • Evolution of the US wellhead gas price over 1980‐2012 USD/mbtu 9.00 8.00 7.00 6.00 5.00 4.00 3.00 2.00 1.00
2011
2009
2007
2005
2003
2001
1999
1997
1995
1993
1991
1989
1987
1985
1983
0.00
1981
Page | 6
Source: EIA.
During the 1990s and into the 2000s, CCGT was considered an investment of choice by new entrants in the power market. A true boom in gas capacity construction occurred then, with 184 GW of gas‐fired plants built between 1990 and 2010. Indeed, a record of 36 GW of CCGTs added at the peak of the boom in 2002 (or 57 GW if one includes open‐cycle gas‐fired plants). Some observers argue that too much capacity was built, which led to its underutilisation later in the 2000s. Although low gas prices below USD 2/MBtu in the 1990s are often referred to as one
Coal to Gas Competition in the U.S. Power Sector
© OECD/IEA 2013
of the main reasons for the surge in gas generation investments, coal prices also dropped during this period (from USD 33.93 per tonne in 1988 to USD 26.37 in 1999).2 As a consequence, there is no direct correlation between low gas prices and high gas‐fired capacity additions. Levelised cost studies sometimes show coal as more competitive than gas‐fired plants (Hogue, 2012). However, CCGT plants offer many advantages, including high efficiency, lower CO2 emissions, relatively quick and cheap construction, modularity and less local resistance to the Page | 7 siting of new plants than for coal and nuclear plants. Moreover, when the distinctive economic and financial characteristics of CCGTs are taken into account, they reveal their critical advantages for new entrants in liberalised markets. Indeed, a high degree of correlation between gas and electricity prices makes CCGTs “self‐hedged” (Roques, 2007). Finally, CCGT investments take place with a lower capital expenditure than coal or nuclear plants (IEA, 2010). Despite the significant gas price fluctuations and the dramatic but short‐lived surge in gas capacity investments that have occurred over the past 20 years, the share of gas in thermal generation as well as gas‐fired generation increased in a very gradual and steady manner over the same period. Figures 2a and 2b illustrate this gradual progression of coal and gas output and their respective shares in thermal generation. In absolute volumes, gross generation from gas‐ fired plants reached about 1 046 TWh in 2011, which is three times higher than in 1990, and twice more than in the late‐1990s, illustrating the dash for gas that occurred over this period. Meanwhile, coal‐fired generation increased during the 1990s, then hovered at around the same level of 2 000 TWh over the last decade until 2008. Since 2009, it has been globally declining to reach 1 773 TWh in 2011. It appears that during most of these past 20 years, additional gas‐fired generated electricity (+660 TWh) has actually partly filled the gap created by incremental power demand (+1 100 TWh), rather than displacing coal.3 Real competition between coal‐ and gas‐fired plants started in the past four years, prompted by low gas prices. This took place in a context of stagnating power demand. This competition, however, did not occur everywhere in the United States, but was mostly concentrated in the eastern part of the country. Figure 2a • Coal and gas power generation, 1990‐2011 TWh 2500
2000
1500
1000
500
0 1981
1984
1987
1990
1993
1996 Coal
1999
2002
2005
2008
Gas
Source: EIA.
2 3
Bituminous coal, EIA data Additional electricity from all renewable sources added 190 TWh and from nuclear power plants 210 TWh.
2011
Coal to Gas Competition in the U.S. Power Sector
© OECD/IEA 2013
Figure 2b • Coal and gas shares in power generation, 1990‐2011 70%
60%
Page | 8
50%
40%
30%
20%
10%
0% 1981
1984
1987
1990
1993
1996 Coal
1999
2002
2005
2008
2011
Gas
Source: EIA.
Assessing current potential switchable gas capacity As of 2011, capacity utilisation of CCGT plants in the United States at an aggregate national level was approximately 46.4%, compared with 62% for non‐lignite coal. There is considerable variation among states, however, with utilisation rates for gas‐fired plants ranging from below 10% in Nebraska and Iowa to over 80% in Connecticut and Alaska.4 Given the large raw switching potential suggested by the low utilisation of US gas‐fired plants, one could wonder why more switching has not already occurred. The following in‐depth look at the US energy market elucidates which factors have hampered − and could continue to deter − a more significant switch from coal to gas.
Characterising the portfolio of switchable capacity When considering the scope for fuel switching in electricity generation, only a sub‐set of coal‐ and gas‐fired plants is most likely to be substitutable, also taking into consideration the fuel type and combustion method. The United States had 415 GW of generating capacity from gas in 2011, compared with around 315 GW of coal‐fired capacity. Running 415 GW of capacity at 100% theoretically equates to 3 637 TWh of output, shown as a maximum output in Figure 3. Of the 415 GW of gas capacity, as much as 198 GW are open‐cycle plants, which, due to their lower efficiency, are not as competitive as CCGTs and are therefore unlikely to be substituted for coal even at prices in the range of USD 2.50‐4/MBtu. Excluding open‐cycle plants results in a reduction of 1 734 TWh in output. In practice, 85% capacity is a more accurate maximum running level, since plants require maintenance. This results in a reduction in potential output of 285 TWh. 4
The United States Energy Information Administration (EIA) is in the process of refining its measures of capacity utilisation, with a view to reporting consistent monthly figures on capacity utilisation by fuel and state.
Coal to Gas Competition in the U.S. Power Sector
© OECD/IEA 2013
From this, it is then necessary to subtract existing CCGT output of 833 TWh, since the corresponding capacity was already in use and is therefore unavailable for further switching. Lastly, two states in particular represent limited switching potential. Arizona has relatively more expensive gas, longer average term coal contracts and less efficient CCGTs, making switching less economical in that state. Meanwhile, California stands out as a state with a largely underutilised gas‐fired capacity; however, there is no coal capacity to displace. Page | 9 Subtracting California and Arizona nets a further 171 TWh from the remaining total (assuming an 85% potential output). This leaves therefore a maximum switching potential of 613 TWh (or around 17%) of gas‐fired generation to compete with coal as of 2011. This figure represents a ceiling of possible switching rather than an actual switchable amount. Additionally, it is important to look at coal‐fired capacity. The existing 315 GW of coal‐fired capacity is fuelled by lignite, bituminous and sub‐bituminous coal. Out of 16 GW of lignite‐fired capacity in the United States, the majority is concentrated in the states of Texas and North Dakota, with some capacity in Louisiana, Mississippi and Montana. As lignite is a very low‐cost fuel source, it is generally consumed close to the mine; consequently, it is unlikely that CCGT capacity could compete with lignite, even at gas prices in the range of USD 2.50‐4/MBtu. In North Dakota there is little CCGT to switch, whereas in Texas there is sufficient non‐lignite coal capacity to switch that the state can be included in the 613 TWh. Finally, growth in overall demand for electricity is falling in the United States, which may reduce the scope for switching. In this context it is also worth noting the year 2012 was exceptional in many respects: a mild winter, a hot summer, and frequent outages of nuclear power plants. The demand of gas‐fired generation in 2012 was therefore driven, inter alia, by abnormal conditions. Figure 3 • Switching potential in 2011 Twh
4000 3500 3000
1734.117647 2500 2000
3637
285.4323529
1500
833
1000
171
500
613.45 0 Total gas
Remove OCGT
15% limitation
CCGT output 2011 CA,AZ (at 85%) Potential extra gas
Source: EIA data (EIA‐860) and IEA calculations
Factors affecting utilisation of switchable capacity Many factors can further affect potential coal‐to‐gas switching and actually explain why switching has not occurred in a more significant manner. These include:
Coal to Gas Competition in the U.S. Power Sector
© OECD/IEA 2013
the relative fuel prices at state level, the variability in plant efficiency, the length of contracts between coal producers and power producers, technical factors, and Page | 10
other specificities of the US power market. These factors are examined in detail in the following sections. They will provide the background for the analysis of observed switching at state level in the year beginning October 2011, as well as an estimate of switching occurring by 2017.
Fuel prices Fuel prices are the primary determinant of dispatch and switching activity between coal‐ and gas‐ fired plants. First, gas prices at the main US hubs have fluctuated considerably since January 2009, making it hard for power generators to predict how they would evolve. Figure 4 shows that that gas prices remained consistently below USD 4/MBtu for more than three months since the economic downturn in 2009 in the latter half of 2011. While looking at the US gas market, one tends to consider only the Henry Hub (HH) price. Actually, there are many different regional hub prices, the HH usually acting as a reference for international comparisons or long‐term forecasts. While regional gas prices would tend to follow HH price movements, there is also great variability between these prices, resulting in widely different gas prices at the state level, and even among plants. The spread between the different hub prices is seasonal. The greatest spread, more than USD 1/MBtu, occurs in winter months, with New England and New York experiencing significantly higher gas prices than the rest of the country due to lack of gas storage and transportation congestion. Florida also has more expensive gas prices (see Figure 13 in Appendix A). At the level of regional prices (and based on available data), substitution seems most likely to occur on the Eastern seaboard, where relative prices were most favourable. Figure 4 • United States Henry Hub prices, 2009‐12
Source: EIA.
© OECD/IEA 2013
Coal to Gas Competition in the U.S. Power Sector
Coal contracts The terms and conditions of coal contracts influence the choice of fuel for electricity generation in the United States, as a larger proportion of coal than gas is bought on a contracted long‐term basis. For example, 93% of the coal consumed for electricity generation in the United States in 2011 was purchased via long‐term contracts of more than one year, (rather than via spot Page | 11 purchases), against only 44% of gas (EIA, 2012). While the exact conditions attached to these contracts are not made public, anecdotal evidence suggests that many have firm take‐or‐pay clauses, with the result that power producers have frequently committed to consuming a given level of coal output for several years into the future. Of the 15 states with the longest average remaining contract terms, four have CCGT capacity above 5 GW: Arizona, Pennsylvania, Mississippi and Oklahoma. The average contract terms by state are shown in Figure 5, with these four states highlighted.5
Variability in plant level efficiency The relative value of natural gas and coal to electric generators cannot be compared solely on a thermal unit (Btu) basis, since the thermal conversion into electricity (kilowatt‐hours generated per Btu) varies by facility. US non‐lignite coal plants range in age from one to 88 years, with an average age of 38 years.6 Due to depreciation and changes in technology, the efficiency of these plants varies considerably, with most falling in the range of 22 to 35%.7 Likewise, while the CCGT fleet is much younger (with an average age of 12.5 years), there is still considerable variance in efficiency, with the bulk of the fleet falling in a range between 40% and 50%, since the first generation of CCGT plants had considerably lower efficiencies.8 Because efficiency of any one given plant depends on a combination of factors, such as fuel quality, load factor, cooling temperature, etc., efficiencies of different plants are difficult to be reported on standard, comparable bases. There is no single, definitive source on the distribution of efficiencies in the two sets of generators, and available data can deliver differing results. A study completed by the California Energy Commission in August 2011 (Nyberg, 2011) examined the growing efficiency of CCGT plants in California in the period 2000‐10. Analysing data of state regulatory agencies, the study found the average efficiency of new CCGT plants in California to be around 48% in 2010. Aggregate calculations from EIA data (EIA, 2012; EIA, 2013) support this finding; however, EIA data show considerably less efficiency for CCGT plants in Texas, with an average below 44%. Texas and California have respectively the largest and third‐largest CCGT fleets in the United States. At a high level, the range of efficiencies can be estimated, based on a calculation of the net output compared with thermal input. This calculation has been done at plant level in Figure 6, on a high heating value basis and using net generation. While Figure 6 represents a high level estimate of the distribution of efficiencies of coal and gas‐fired plants, it is nonetheless broadly indicative of the variation that exists.
5
In the case of Alabama, one coal contract recorded as expiring in 2099 skews the average by just over two years. The basis for enforcing a contractual term extending in excess of 80 years is questionable, and so this contract has been omitted from the data in Figure 5. 6 Age weighted for MW capacity, EIA data. 7 All efficiencies are on a gross calorific basis, based on net generation. 8 Age weighted for MW capacity, EIA data.
Coal to Gas Competition in the U.S. Power Sector
© OECD/IEA 2013
Figure 5 • Average length of US non‐lignite coal contracts for power generation, by state years 10 9
Page | 12 8 7 6 5 4 3 2
WY
AZ
MT
PA
NM
KS
OH
MS
WV
UT
OK
TX
CO
IN
MO
NJ
OR
IL
KY
NC
NV
FL
AR
AL
MN
IA
GA
SC
WI
MI
MD
VA
NE
ME
LA
TN
WA
SD
NY
HI
NH
DE
0
CA
1
Note: Volume‐weighted average, based on remaining terms on coal contracts from deliveries in 2011. Source: EIA (EIA‐923); IEA calculations.
Figure 6 • Range of thermal efficiency in US coal and CCGT plants, 2011 Capacity: MW 70000
60000
CCGT
COAL
50000
40000
30000
20000
10000
0 20
30
40 Efficiency: percent
Source: EIA (EIA‐923); IEA calculations.
50
Coal to Gas Competition in the U.S. Power Sector
© OECD/IEA 2013
Some uncertainty regarding the absolute level of efficiency notwithstanding, it appears that CCGT efficiencies vary significantly in the US power sector. As a purely theoretical but indicative exercise, Figure 7 shows the “switching gas price”, depending on the price of coal and the thermal efficiency of the CCGT plant. As can be seen from the graph, a change of efficiency from 52% to 44% can require a gas price up to USD 0.60/MBtu lower (for a given coal price of USD 65/t). Interestingly, the differential increases in absolute terms as the coal price increases, Page | 13 which is explained by the fact that a switching price is proportionate to a gas plants’ efficiency, and thus at higher coal (and electricity) prices, gains (and switching) from higher efficiency are larger in absolute terms, while fixed in percentage terms. Figure 7 • Gas switch price based on coal price, for different gas efficiencies and fixed coal efficiency gas 44% eff
gas 48% eff
gas 52% eff
Gas price, USD/MBtu
5
4
3
2 50
55
60
65
70
75
80
Coal price, USD/t6000
Note: Coal thermal efficiency fixed at 39% , $/t6000 represents $ per tonne of coal with a 6000 kilocalorie /kilogramme of net calorific value . Source: IEA calculations.
Technology factors Limitations of coal and CCGT technologies are also likely to play a role in the choice between the two fuels. The increasing peakiness of US power load in recent years places certain restrictions on the dispatch between coal‐ and gas‐fired plants. Where a power producer might otherwise switch from coal to gas for baseload power and utilise coal to follow variable demand, technology may limit this, for the following reasons: Operating range and minimum output, Start‐up rates, and Ramp rates. The geographical location of coal plants may also be a factor in some instances, as discussed later under the section on transmission constraints.
Operating range and minimum output The existing coal and CCGT fleets have different optimal capacity factors and the relationship between utilisation and efficiency differs. The existing US coal fleet may be less well‐suited to
Coal to Gas Competition in the U.S. Power Sector
© OECD/IEA 2013
meeting variable demand because its optimal utilisation falls in a smaller range. The level of optimal utilisation for a given plant depends on a range of factors, notably its age and its design. New coal and CCGT plants have relatively similar ranges and can generally operate optimally at between 70% and 90%, and sub‐optimally at between 40% to 70%, with moderate losses in efficiency. For older plants, the loss of efficiency at lower levels of utilisation will tend to be Page | 14 higher (IEA, 2010; NERC, 2010; Platts, 2003). As coal plants are on average 25 years older than CCGT plants (IEA calculations based on EIA data [EIA‐860]), the efficiency losses associated with sub‐optimal load factors are greater on average.
Start‐up rates The start‐up rates for coal‐fired boilers and the steam component of CCGT plants are in the range 8‐48 hours, whereas the gas turbine components of CCGT plants have start‐up rates below one hour (AEMO, 2010). This allows CCGTs to respond to rapid changes in power demand, albeit at the efficiency level of an open cycle plant in the early stages (efficiency decreases by around 10‐ 20%).
Ramp rates The ramp rates of US coal‐fired generators depend largely on their vintage. The range for plants of the 1960 vintage (the average age for a US coal plant being 38 years (IEA calculations based on EIA data) is around three megawatts per minute (MW/min) for a 500 MW unit. This compares with average CCGT ramp rates of around 15‐25 MW/min, which is roughly similar to the ramp rates for coal plants built since 2000. As is the case with operating range, above, while ramp rates are similar for coal and CCGT plants of similar vintage, the average age of coal plants is much higher than that of CCGT plants, making coal plants more costly for meeting variable load. For these reasons, CCGT plants are better placed to respond to variable demand, making coal plants more expensive to run intermittently. This increases the likelihood that coal plants will be run at higher capacity factors than CCGT, other factors being equal.
Role of regulation and policy Besides the considerations regarding switchable gas‐fired capacity and technical, pricing and contractual limitations on this, there are some specificities of the US power sector to be considered. In some regulated states (in the Southeast), due to already relatively lower electricity prices, there might be less pressure to reduce prices further by switching to cheaper fuels. Although there might be some discontent about higher end‐user prices in the liberalised northeastern states, economic and market design factors play a key role. These states sometimes also have higher fuel prices.
Price regulation Unlike the gas sector, liberalisation of the US power sector reform is at different stages, which affects directly the way power prices are formed. Electricity prices are a key factor when considering electricity market reform. Figure 8 highlights the extent of power sector reform in each state, alongside average electricity prices. Many states, notably in the centre and southeast of the United States, have not deregulated power prices, while deregulation has been suspended in some additional states in the Southwest. States with active deregulation are mostly concentrated in the North east but also include Texas. However, conclusions cannot be extracted
© OECD/IEA 2013
Coal to Gas Competition in the U.S. Power Sector
from the map, as differences in fuel prices between states often play a bigger role to determine electricity prices than the matter of being regulated/deregulated. In the context of analysing fuel switching, the regulatory process in regulated states where the price is lower is expected to put less pressure on power producers to reduce prices further by minimising fuel costs. In the liberalised north eastern states, economic and market design factors Page | 15 play a key role, despite some potential discontent about higher end‐user prices. In competitive markets, such as PJM, market design and structure are potentially significant factors in coal‐to‐gas competition. Capacity payment mechanisms exist in most liberalised US markets and they constitute a significant share of gas plants’ revenues. For example, in 2010, CCGT plants in the PJM market received around 30% of their net revenues from capacity payments (Potomac Economics, 2011). Based purely on the microeconomic theory of profit maximisation, this fixed stream of revenue should not affect decisions to run gas installations. However, the reduced risk of making a loss on gas installations thanks to capacity payments might affect the way market actors make decisions, as utility functions depend on attitudes to risk, especially when the same owner also has coal‐fired plants. As shown in the second chapter of this paper, some 12 TWh of fuel switching was estimated to have occurred in Florida (from a base of coal generation of 56.4 TWh)9 and Florida is a regulated state, suggesting that public utility regulation does not necessarily dampen economic incentives. In reality, the impact of regulation is likely to be ambiguous and specific to local conditions. Figure 8 • Deregulation in the US power sector and retail electricity prices, 2011
Source: IEA research
Some power producers, whose revenues are regulated, face weaker incentives to depart from existing practice in response to changes in relative gas and coal prices. In 2011, 75% of US coal‐ fired generation was in the regulated sector, against 36.5% of CCGT (Figure 9). In many cases, the power‐producing entity has the flexibility to increase charges automatically in response to changes in fuel costs, but should not profit from the fuel component. A regulated entity, which passes through the cost of fuel directly, is likely to have more discretion about the timing for 9
In the year beginning October 2011, compared with the prior 12 months, see the second chapter.
Coal to Gas Competition in the U.S. Power Sector
© OECD/IEA 2013
switching from coal to gas than a competitive energy producer, since it cannot theoretically benefit from fuel cost changes nor does it face any negative exposure to these, as they are simply passed through in higher charges. Even though fuel pass through clauses should theoretically mean that the fuel choice is neutral in terms of profits, in practice, there are likely to be more or less significant temporary or permanent cash flow impacts, depending on the precise pass‐ Page | 16 through arrangement in place. A number of studies have discussed the implications of fuel pass through clauses for the efficiency of regulated energy providers in the United States (Joskow, 1974; Brown et al., 1991; Graves et al., 2006). Theoretically, the need to protect a firm from risks associated with rapid change in fuel costs could be considered to involve some efficiency losses in the use of fuel by that firm (Graves et al., 2006).10 Knittel (2002) examined the operation of firms in regulated American states in the period 1981 to 1996, focusing on a handful of states where regulators modified fuel clauses. Modified fuel clauses require the regulated firm to absorb a portion of the risk associated with positive or negative changes in fuel costs, whereas under a standard fuel pass through clause, the end customer bears all risk associated with fuel costs. Knittel found evidence that providing incentives to regulated firms to keep fuel costs low had increased efficiency in the use of fuel, suggesting that standard fuel pass through clauses do introduce a level of inefficiency in the use of fuel among regulated firms.11 Figure 9 • Share of coal and gas generated from the regulated and unregulated sector, 2011
Terawatt hours 2000 1500 1000 500 0 Coal
Gas Regulated
Unregulated
Unknown
Source: EIA (EIA‐923); IEA calculations
While fuel pass‐through charges may delay a response to a change in relative fuel prices, most regulated utilities will eventually be required to prove to the regulator that their fuel and resource decisions have been prudent, as part of a rate case. Therefore, they will seek to move to the lower cost fuel in the medium term. In the interim, the impact of regulation is ambiguous, since the clauses essentially appear to provide a utility with extra discretion, either to defer or bring forward its decision to switch fuels, without having a negative impact on earnings. Such a decision could then be influenced by additional factors, such as: whether the entity has long‐term coal contracts; 10
Graves et al. (2006) note that there may also be inefficiency in a regulatory process that requires a rate case each time an entity needs to adjust its fuel costs. 11 Knittel examined efficiency in the use of a given fuel, rather than the choice between two types of fuel; however, the implications appear pertinent to a situation where a firm must choose between power generated from fuels of different costs.
© OECD/IEA 2013
Coal to Gas Competition in the U.S. Power Sector
whether the entity has sufficient coal‐ and gas‐fired capacity to switch from one to the other or must contract for gas‐fired electricity. Some utilities may prefer generating from their own power assets rather than purchasing electricity on the wholesale market, since relying on internal generation may be seen as lower risk. Where they do not have sufficient gas‐fired capacity, they may seek to put off fuel switching until they have their own gas‐fired plant; whether the decision to move away from coal is supported in the relevant jurisdiction. Moving Page | 17 to gas in states where this is supported by policy at state level may facilitate regulatory approvals of future investment; conversely, in states where the coal industry plays a central role in economic activity, utilities may seek to delay fuel switching to support policy objectives of state governments; and some level of generalised inertia or path dependence that leads firms to defer change. Consequently, the impact of fuel pass‐through clauses on switching from coal to gas is ambiguous results, since these clauses may lead to faster or slower switching than would otherwise be the case under purely competitive forces. A general conclusion is that the fuel decisions of regulated utilities will be less closely correlated with changes in relative fuel prices than in states where competitive price pressures more strongly influence fuel choice, but that this could lead to faster or slower switching than would otherwise be the case in a state where the power sector was liberalised.
Electricity transmission The United States electricity system is very fragmented due to administrative, infrastructural and natural factors, despite regional coverage of some of the US power markets. For coal and gas plants to compete, they not only need to be connected to the same electricity system, but also to belong to the same administrative unit (market or regulated state). Maps of geographic location of coal and CCGT capacity show that capacity is very unevenly distributed. See Appendix A, Figures 11‐13. California stands out as a state where a large underutilised gas capacity has no coal capacity to displace, and therefore we have subtracted it from the switchable potential, as per above. In states where conditions are otherwise favourable to fuel switching, the electricity grid may limit switching. Specifically, there is very little trade between the three main interconnections in the coterminous American states (Western, Eastern and ERCOT), and within these interconnections, there is limited long‐distance transmission capacity. This means that for fuel switching to occur, a CCGT plant needs to be not too distant from a load served by a coal‐fired plant. Additionally, even where transmission is theoretically available to transport a load, congestion in the transmission system may limit the dispatch of generating units and therefore the coal to gas switch. Finally, in states where coal‐fired generation serves primary as a base‐load energy source, the geographic distribution of plants is relevant to the task of maintaining grid stability. In those states, it may be difficult to switch to base‐load power fueled by CCGT if the geographic distribution of CCGT plants is not comparable. The US Congress Research Service (Kaplan, 2010) conducted a high‐level analysis to identify all major coal plants with one or more existing CCGT plants within a ten‐mile radius, on the assumption that these CCGT plants would be best placed to displace coal within the constraints of the local transmission network. The hypothetical surplus generation for each CCGT within the ten‐mile radius was calculated and assumed to displace generation from the coal plant. At a high level (and examining only this variable), the analysis suggested that existing CCGTs near coal‐fired plants could account for up to 30% of the displaceable coal‐fired generation. Greater displacement of coal by existing CCGTs would depend on more distant CCGTs that might face network constraints. Kaplan (2010) concludes that this analysis “emphasises the importance that
Coal to Gas Competition in the U.S. Power Sector
© OECD/IEA 2013
the configuration and capacity of the transmission system will likely play in determining the actual potential for displacing coal with power from existing CCGT plants.” However, it is impossible to accurately gauge the implications of the layout of the two sets of plants for switching without a fully integrated model of the entire fleet that can estimate the impact of changes in the merit order. Page | 18
Gas transmission In addition to the constraints of the electricity transmission network, the physical capacity of the national gas transmission network also affects switching potential. While the regional pipeline system is designed to deliver the necessary gas to individual plants to allow each to run at full capacity, increasing the utilisation of the entire CCGT fleet to 85% implies a significant increase of natural gas production and distribution at the national level (Beach et al., 2012). Beach et al. estimate that a 7% increase in pipeline capacity (around 4.6 tcf or 130 bcm) would be required compared to 2008 levels, which falls well within the 22.6 tcf aggregate expansion of pipeline system capacity targeted for completion by early 2012. Even if additions to the gas network may seem to have been adequate to allow for a significant expansion in gas‐fired generation, complications arise in relation to markets for wholesale gas and gas haulage. The National Petroleum Council (2011) finds that most power generators – and particularly those selling into unbundled wholesale electric markets – choose less expensive, interruptible transportation gas pipeline capacity rather than firm contractual capacity. During peak winter demand conditions, pipeline customers with firm contractual rights use their full contractual entitlements, meaning that pipelines frequently do not have additional capacity for interruptible transportation customers. The NPC cites an instance in January 2004 in New England where 6 GW of gas‐fired generation was unavailable to run at peak times because the operators had chosen to rely on interruptible transportation, and the winter conditions resulted in all parties with firm commitments using these to deliver natural gas to residential and commercial customers. In a 2006 report on gas‐electricity interdependency, the North American Energy Standards Board (NAESB) identified six areas where standardisation and harmonisation would facilitate better integration of US gas and electricity markets. Given the lack of uniformity in approaches in electricity markets, realising these initiatives is challenging. However, this may become more pressing if is to account for a larger share of the US power generation mix. The preference among gas‐fired generators for interruptible gas haulage contracts stems in part from a mismatch between US electricity and power markets. Access to gas transmission networks is harmonised across states, through standards developed by the NAESB which were designed to improve transparency and efficient scheduling. These standardised approaches stand in contrast to the practices in electricity markets, which are not standardised even within one inter‐ connection. For instance, important differences between gas and power wholesale markets exist in relation to the definition of day and intraday schedules. These differences are exacerbated by the fact that the gas haulage market is far less liquid, reacting more slowly to rapid changes in energy demand. NPC (2011) notes that “as a consequence of these inconsistent timelines, the owner of a gas‐fired generator must either buy gas without knowing if its power will be scheduled, or submit a power bid before knowing if the gas can be purchased and scheduled. The cost of covering the risk created by the inconsistency in timelines must be reflected in generators’ power offers”.12 This has an impact on the relative competitiveness of gas‐fired generation. 12
Fuel switching away from natural gas to diesel is a further option in these circumstances, but this has no impact on overall substitution of natural gas for coal.
© OECD/IEA 2013
Coal to Gas Competition in the U.S. Power Sector
Environmental regulation The future retirement of coal‐fired plants as a result of more stringent environmental regulations may open up opportunities for fuel switching where it might not otherwise be economically viable. Permitting and licensing of new coal‐fired plants has become more challenging in the United States, partly as a result of local opposition to the construction of new plants. Page | 19 Consequently, as older, less viable coal plants are decommissioned, this may create opportunities to switch to CCGTs or build fresh CCGT capacity. Of the 299 MW of non‐lignite coal‐fired capacity as of 2010, around 110 GW did not have emission control equipment (“scrubbers”) or firm plans to install this equipment.13 Among this capacity, in the 55 GW of older and less efficient plants, the necessary investments to meet increasingly rigorous emissions control requirements are relatively less likely to justify. Around 36 MW of this capacity is concentrated in mid‐western and southern states Illinois, Indiana, Michigan, Wisconsin, Alabama, Mississippi, Tennessee and Kentucky. An estimate of the impact of air quality regulations on coal retirements is provided in detail in the next chapter.
13
Based on an analysis prepared by ICF International for the Interstate Natural Gas Association of America
Coal to Gas Competition in the U.S. Power Sector
© OECD/IEA 2013
Estimates of switching: observed and projected As highlighted in the first part of this report, beyond the theoretical current potential of 613 TWh of annual output that could have switched in 2012, there are many other factors to be taken into account to analyse correctly how, where and whether coal‐to‐gas switching can occur. This part Page | 20 provides estimates of coal‐to‐gas switching that occurred in the year beginning October 2011, and further switching that will occur out to 2017. The analysis focuses on the situation in 18 American states that account for approximately 75% of the unused CCGT potential in the United States (selected on that basis).14 The 18 states accounted for 1 011 TWh of net generation output in 2011 in the United States, or 46.5%. The states account for 78% of the installed CCGT capacity in the United States but only 36% of installed non‐lignite coal generating capacity. The estimates of switching simplify a number of factors, including some outlined in the first chapter of this study, focussing primarily on elements that can be readily measured. A further simplification is that the analysis examines each state as a producer and consumer but not an importer of energy, since the EIA data is available at state level rather than at the level of interconnection.
Observed switching An estimate of observed switching has been made by comparing CCGT and coal output in the year commencing October 2011 with output in the preceding 12 months. The year beginning October 2011 is chosen because the last quarter of 2011 was the first time since the economic downturn that gas prices for power producers at HH remained consistently below USD 4/MBtu for a sustained period (longer than 3 months), USD 4/MBtu being a level below which switching has previously been observed to occur. The estimate of observed switching is considered in this manner: Coal and CCGT‐fired output from October 2010 to September 2011 is compared with output from October 2011 to September 2012, the first being a year of higher gas prices relative to coal, and the second a period of lower gas prices relative to coal. Where both CCGT output increased and coal output decreased year‐on‐year, the smaller of the two changes is assumed to represent switching. This is against a backdrop of low or no growth in overall electricity demand. A likely scope for switching is gauged, by considering:
the gas price at which capacity will switch, in the absence of other constraints (in the manner presented in Figure 7), as well as actual gas prices in the period under analysis;
the primary measurable non‐price constraints, being the volume‐weighted average of coal contract terms; the efficiency of CCGT and coal plant relative to national averages; the potential for increasing CCGT output based on existing plants, and the potential for coal retirements; in order to determine whether overall conditions will be favourable to switching. The estimate of observed switching is considered in light of whether conditions appear favourable. In other words, in a state where the gas switching price was met for an extended period during the 12 months from October 2011, with short coal contract terms and considerable surplus CCGT potential, significant switching should already have been evident in the year 14
Fourteen states in addition to four states where there is spare CCGT capacity but no coal output to switch (California, Maine, Oregon and Washington).
© OECD/IEA 2013
Coal to Gas Competition in the U.S. Power Sector
beginning October 2011. In a state where price and other conditions are unfavourable, limited switching should have been evident. This comparison was consistent in all states examined, with the exception of Louisiana and Michigan, where price and other measurable conditions analysed were favourable to switching for much of the 12‐month period beginning October 2011, yet only between 15‐20% of coal output switched to gas in those states in that year. In these instances, it appears likely that other factors less amenable to measurement are at play, such as the role of Page | 21 utility revenue regulation in Louisiana, for example. The estimate of observed switching across the 18 states between the higher gas price period (beginning October 2010) and the higher gas price period (beginning October 2011) was 122.5 TWh, or 19.92% of the coal output in the higher gas price period. The analysis at state level across 14 states appears in the table in Appendix B. Eighteen states represent 75% of the surplus CCGT potential output in the United States (four states not shown in the table have surplus CCGT potential but no material coal output to switch: California, Maine, Oregon, and Washington). This represents a reduction in carbon dioxide emissions of approximately 61.5 million tonnes.15
Projected switching The estimate of projected switching provided here builds on the estimate of observed switching, but with an assumption that the gas price moves to USD 4.7/MBtu at HH by 2017, while the coal price remains largely stable except for some marginal increase in the eastern states. It is also assumed that the gas price for power producers in each of the states remains at a premium or discount to HH that is consistent with long‐term averages. In addition, coal plant retirements and CCGT expansions are considered in detail, as outlined below.
Coal plant retirements The Mercury and Air Toxics Standards (MATS) and the Cross‐State Pollution Rule (CSAPR) are federal measures that aim at reducing polluting substances from power plants. Both rules were widely expected to have the greatest impact on the electric sector between now and 2015 (DOE, 2011). However, in August 2012, the US Court of Appeals for the District of Columbia Circuit ruled that the CSAPR rule was void. With many operators having already announced their intentions to retire some of their coal‐fired plants prior to the decision on the CSAPR, it is complicated to quantify the discrete impact of MATS on coal capacity retirements. Generally speaking, however, it is reasonable to assume that short‐term retirements will be a little lower than expected before the court’s decision. This study analysed the impact of MATS and CSAPR on coal capacity prior to the court’s decision, according to a method outlined below. The estimated impact of these two regulations was such that in each state analysed, excess coal capacity remained ‐ to a degree that it was unlikely that projected retirements would trigger changes in merit‐order switching. Hence, with only MATS coming into effect, retirements should be fewer than projected, reducing the likelihood of a resulting change in merit‐order. In light of the above, the results drawn from the estimation approach developed prior to the court’s decision are still considered valid in the context of the overall findings, even though retired capacity is likely to turn out somewhat lower than the numbers given here. 15
This calculation is based on national average emissions per unit of generation and includes carbon dioxide only, not other pollutants.
Coal to Gas Competition in the U.S. Power Sector
© OECD/IEA 2013
Approach to estimating coal retirements Previous studies attempting to project plant closures due to EPA regulations have utilised different estimation strategies. This study combines multiple plant criteria, as well as operator announcements and EPA predictions, in order to come up with estimates that do not only rely solely upon theoretical predictions or announced retirements. Reflecting the range of current Page | 22 estimation strategies, the chosen methodology provides a realistic quantification of coal capacity that will retire due to EPA regulations. More precisely, retirements were estimated by identifying plants in the following categories: The EPA has published the results from its integrated planning model (IPM) showing a projection of the impact at plant level of the MATS and CSAPR regulations. Plants projected to be “withdrawn as uneconomic” are here projected to retire. According to a comprehensive study at plant level conducted by PJM, coal units older than 40 years and below 400 MW were shown to have the lowest net energy revenues per MW of capacity and to cost the most per MW to retrofit; they are consequently considered to be the least likely to warrant retrofitting. As such, all plants older than 40 years and smaller than 400 MW that do not have NOx or SOx controls are projected to be decommissioned. Data for this category is provided through the EPA NEEDS file. Many operators have already announced plant closures, either in response to EPA regulations or citing economic conditions. Where there have been public announcements of retirements in addition to the units listed under the first two bullet points, the plants are also assumed to be closed. Data is drawn from EIA questionnaires. Based on the above method, the projected retirements in the 14 states examined was 12.8 GW out of 118 gigawatts of coal capacity in the 14 states. In all 14 states, even with levels of retirements as projected by the above method, there is still sufficient excess coal capacity after 2015 to maintain current output from coal generators. As such, the retirements seem unlikely to affect switching in those states. (EIA 2011 data projects planned retirements of 27.8 GW of coal capacity across the nation as a whole in the period 2012‐17,[EIA, 2011] but around half of this falls outside the states that are most relevant to switching.)
Expansion in CCGT capacity Overall demand for energy in the United States is projected to grow only 1.87% over the coming five years (EIA, 2013). In particular, power demand is expected to grow from 3 729 billion TWh in 2012 to 3 798 billion TWh in 2017. Given the low utilisation rates of CCGT plants in many states, there is only limited investment in CCGT expansion planned in the period, around 10 GW in the period from 2013 onwards, an increase of around 4.6% on the existing fleet, concentrated in nine states only (EIA, 2011). Over the coming years, some shale plays such as Marcellus and Eagle Ford are expected to provide significant additional supply. Unlike Eagle Ford, which is located near other existing shale gas plays in Texas which have so far provided the bulk of additional new shale gas supply, Marcellus is located in the Northeast of the United States. A specific analysis of the possible effects of increased shale gas availability in states covering the Marcellus play has been conducted, considering that such a significant fresh supply could change the economics of investment, prompting fresh investment in CCGT capacity and a shift in the merit order. However, in the three states likely to benefit foremost from the Marcellus expansion (New York, Pennsylvania and Virginia), there is already significant spare gas capacity, such that significant fresh investment in CCGT output as a result of increases in local supply of gas appears unlikely in
© OECD/IEA 2013
Coal to Gas Competition in the U.S. Power Sector
the medium term. While the proximity of new CCGT capacity to growing gas production is an important consideration, the United States has in recent years built significant intraregional capacity, which has considerably reduced the spread between regional gas prices. EIA data shows that the plans for expansions in CCGT capacity from 2013 are concentrated in Texas (2.6 GW), Florida (2.4 GW) and California (1.3 GW) (EIA, 2011). In the case of Florida, this will represent an almost 50% increase in CCGT potential output, from 42 TWh to 61 TWh. Page | 23 However, the impact of this will still depend on favourable gas prices relative to coal.
Conclusions on projection for switching by 2017 Based on the above assumptions and consideration of the measurable non‐price constraints, it is estimated that the bulk of the 122.5 TWh that switched in the 14 states in the twelve months beginning with October 2011 will switch back to coal as the HH price approaches USD 4.7/MBtu in 2017, leaving shares of CCGT and coal output similar to those observed in early 2011. In this context it is also worth noting that the year 2012 was exceptional in many respects, with a mild winter, a hot summer, and frequent outages of nuclear power plants. The substantial increase of gas‐fired generation in 2012 was therefore driven, among other things, by abnormal conditions. Inputs to the analysis at state level are shown in Appendix B.
Coal to Gas Competition in the U.S. Power Sector
© OECD/IEA 2013
Acronyms, abbreviations and units of measure Acronyms and abbreviations Page | 24 CSAPR CCGT EIA EPA ERCOT IPM NPC MATS NAESB NEEDS NOx NPC PJM SOx
Cross State Air Pollution Rule combined cycle gas turbine Energy Information Administration Environmental Protection Agency Electric Reliability Council of Texas Integrated Planning Model National Petroleum Council Mercury and Air Toxics Standard North American Energy Standards Board National Electric Energy Data System nitrogen oxides National Petroleum Council Pennsylvania‐New Jersey‐Maryland Interconnection sulfur oxides
Units of measure GW MBtu MW
Gigawatt million British thermal units Megawatt
© OECD/IEA 2013
Coal to Gas Competition in the U.S. Power Sector
Appendices Appendix A Page | 25
Figure 10 • CCGT capacity, GW, 2011
Source: EIA (EIA‐860)
Figure 11 • Coal capacity, GW, 2011
Note: Source: EIA (EIA‐860)
Coal to Gas Competition in the U.S. Power Sector
© OECD/IEA 2013
Figure 12 • Gas costs for power, December 2011
Page | 26
Note: White filling means data unavailable. Gas prices are annual prices reported by EIA. Source: EIA
Figure 13 • Coal costs for power, December 2011
Note: White filling means data unavailable. Coal prices are from November 2012, the most recent data available, on the basis that coal prices have remained largely stable across the United States. Source: EIA
Florida
Arkansas
Arizona
Alabama
State
Page | 27 Georgia
Coal output 10/20109/2011 (TWh)
Fall in coal output, 10/20119/2012 compared with prior year (TWh)
11.7
Estimate of switching observed in year beginning 10/2011, based on fall in coal output/increased CCGT output (TWh)
13.3
11.8
0.4
2.08
14.2
23.0
11.8
0.4
2.8
59.3
43.5
29.3
56.4
65.9
Coal price to power producers ** (USD/MBtu)
Estimated gas price at which switching is predicted to occur (USD/MBtu)
Average premium/ discount of gas price to HH price (times) (price of gas to power producers)
5.45
Gas price projected to 2017, at HH USD 4.7/MBtu (USD/MTbu)
Prices favourable to switching for much of 2011/12, but switching likely to have been limited by longer coal contracts. Unfavourable prices are binding constraint in 2017.
Comments, other factors***
-14.2
Projected switching to 2017 (TWh) (relative to year beginning Oct 2011)
5.12
Prices unfavourable to switching in 2011/12 and projected to remain so to 2017.
-11.8
-0.4
-2.8
6.82
Prices favourable to switching briefly in 2011/12. Unfavourable prices are binding constraint in 2017.
-13.3
Prices unfavourable to switching in 2011/12 and projected to remain so to 2017.
5.41
4.70
1.16
1.15
1.45
1.09
1.00
3.75
4.15
4.45
3.08
2.67
2.95
2.05
2.42
3.47
3.25
Prices favourable to switching for all of 2011/12. Extent of switching may have been constrained by utility revenue regulation or other factors. Unfavourable prices are binding constraint in 2017.
Coal to Gas Competition in the U.S. Power Sector © OECD/IEA 2013
Appendix B
Table 1 • Individual Case Studies*
State
Coal output Oct 2010Sept 2011 (TWh)
Fall in coal output, Oct 2011Sept 2012 compared with prior year (TWh)
Estimate of switching observed in year beginning Oct 2011, based on fall in coal output/ increased CCGT output (TWh)
3.11
Coal price to power producers ** (USD/MBtu)
3.90
Estimated gas price at which switching is predicted to occur (USD/MBtu)
1.13
Average premium/discount of gas price to HH price (times) (price of gas to power producers)
5.31
Gas price projected to 2017, at HH USD 4.7/MBtu (USD/MTbu)
Comments, other factors***
-3.7
Projected switching to 2017 (TWh) (relative to year beginning Oct 2011)
-8.1
3.5
Prices favourable for most of 2011/12, but other factors appear to have limited switching. Unfavourable prices are binding constraint in 2017.
-1.2
3.7
4.47
Prices favourable to switching in 2011/12. Unfavourable prices are binding constraint in 2017.
-2.9
20.3
5.12
Prices favourable to switching for part of 2011/12.
-6.6
Louisiana
1.17
6.01
Unfavourable prices are binding constraint in 2017.
Prices favourable to switching for all of 2011/12. Extent of switching may have been constrained by utility revenue regulation or other factors. Unfavourable prices are binding constraint in 2017.
3.08
1.28
5.73
0.95
2.45
5.00
`1.22
4.75
3.75
1.2
4.08
3.69
1.01
2.92
1.2
2.9
3.15
2.52
8.1
5.4
3.1
6.6
1.99
8.1
Nevada
4.9
6.9
3.8
60.8
New Jersey
11.8
3.8
Michigan
New York
33.9
-3.8
Oklahoma
Prices favourable to switching only briefly in 2011/12. Long coal contracts also likely to have limited switching. Unfavourable prices are binding constraint in 2017.
Page | 28
© OECD/IEA 2013
Coal to Gas Competition in the U.S. Power Sector
State
94.7
Coal output Oct 2010Sept 2011 (TWh)
106.8
21.8
Fall in coal output, Oct 2011Sept 2012 compared with prior year (TWh)
9.3
Estimate of switching observed in year beginning Oct 2011, based on fall in coal output/ increased CCGT output (TWh)
7.3
20.4
15.5
20.4
7.6
Coal price to power producers ** (USD/MBtu)
2.51
1.85
3.62
Estimated gas price at which switching is predicted to occur (USD/MBtu)
Average premium/ discount of gas price to HH price (times) (price of gas to power producers)
5.41
Gas price projected to 2017, at HH USD 4.7/MBtu (USD/MTbu)
5.66
4.70
1.15
1.20
1.00
3.23
2.35
4.62
Observed switching in year beginning October 2011 (TWh) Projected switching to 2017 compared with year beginning October 2011 (TWh)
Virginia
Texas
Pennsylvania
Page | 29 Notes: *To be read in conjunction with explanation in the second chapter. **Coal data drawn from months when data is available, however prices have remained stable over the period analysed ***Where other factors refer to 2011/12 this is the year beginning October 2011, as per analysis outlined in the second chapter. Source: EIA, IEA calculations and projections, as per assumptions outlined in Part II
Comments, other factors***
Price favourable to switching for around half of 2011/12. Long coal contracts likely to have limit3ed switching. Unfavourable prices are binding constraint in 2017.
Price favourable to switching for less than half 2011/12. Long coal contracts also likely to have limited observed switching. Unfavourable prices are binding constraint in 2017.
Prices fourable to switching for most of 2011/12. Unfavourable prices are binding constraint in 2017.
Projected switching to 2017 (TWh) (relative to year beginning Oct 2011)
-9.3
-20.4
-7.6
122 -122
Coal to Gas Competition in the U.S. Power Sector © OECD/IEA 2013
Coal to Gas Competition in the U.S. Power Sector
© OECD/IEA 2013
References Australian Energy Market Operator, (2010), “Introduction to Australia’s Energy Market”, accessible at: www.aemo.com.au/.../0000‐0262.pdf. Page | 30
Beach, F.C., M. Shivers Gonzalez, J. C. Butler, M. E. Webber (2012), “An Analysis of the Potential for Expanded Use of Natural Gas in Electrical Power Generation, Transportation, and Direct Use: Texas as a Case Study”, University of Austin Texas, web paper accessible at: www.webberenergygroup.com/natgas. Brown, L, M. Einhorn, and I.Vogelsang (1991), “Toward Improved and Practical Incentive Regulation”, Journal of Regulatory Economics, vol.3, December, pp. 323–338. Comnes, G.A., S. Stoft, N. Greene and L. J. Hill (1995), “Performance‐Based Ratemaking for, Electric Utilities, Review of Plans and Analysis of Economic and Resource‐Planning Issues, Volume I, Lawrence Berkeley National Laboratory, University of California, Berkeley, California. Department of Energy, U.S. (DOE) (2011) Resource Adequacy Report (2011), accessible at: http://energy.gov/sites/prod/files/2011%20Air%20Quality%20Regulations%20Report_120111 .pdf. Energy Information Administration, United States (EIA) (2011), Annual Electric Generator Report Form EIA 860 Data, accessible at: www.eia.gov/electricity/data/eia860/index.html. EIA (2012), Annual Electric Generator Report Form EIA 923 Data, accessible at: www.eia.gov/electricity/data/eia923/. EIA (2013), U.S. EIA Annual Energy Outlook (2013), accessible at: www.eia.gov/forecasts/aeo/er/index.cfm. Graves, F., P. Hanser and G. Basheda. (2006), Electric Utility Automatic Adjustment Clauses: Benefits and Design Considerations, prepared for the Edison Electric Institute by the Brattle Group, Washington D,C. Hogue, M. (2012) “A Review of the Costs of Nuclear Power Generation”, Bureau of Economic and Business Research (BEBR), David Eccles School of Business, University of Utah, accessible at: www.bebr.utah.edu/Documents/studies/Nuclear_Report_Final_Web_7Mar2012.pdf. IEA (2010) Energy Technology Systems Analysis Programme, Gas‐Fired Power, IEA ETSAP Technology Brief EO2, April, accessible at: www.iea‐etsap.org. IEA (2012a), Medium‐Term Gas Market Report, June, IEA/OECD, Paris, available at www.iea.org. IEA (2012b), Medium‐Term Coal Market Report, December, IEA/OECD, Paris, available at www.iea.org. IEA (2012c), Pricing and Regulation, China’s Challenges and IEA Experience, September, IEA/OECD, Paris. Joskow, P. (1974) “Inflation and Environmental Concern: Structural Change in the Process of Public Utility Price Regulation”, Journal of Law and Economics, 17, October, pp.291‐327. Kaplan, S.M. (2010) “Displacing Coal with Generation from Existing Natural Gas‐Fired Power Plants”, United States Congress Research Service. Knittel, C. (2002) “Alternative Regulatory Methods and Firm Efficiency: Stochastic Frontier Evidence from the U.S. Electricity Industry”, The Review of Economics and Statistics, Massachusetts Institute of Technology, Massachusetts, August, 84(3): 530–540.
© OECD/IEA 2013
Coal to Gas Competition in the U.S. Power Sector
Massachusetts Institute of Technology (2011) “The Future of Natural Gas”, an MIT Interdisciplinary Study, chapter 4. National Petroleum Council (NPC) (2011) “Power Generation Natural Gas Demand”, Working paper #3‐1, September. North American Energy Standards Board (NAESB) (2006) “Gas and Electric Interdependency Final Report to the Federal Energy Regulatory Commission”, prepared by ICF Resources, LLC. North American Electric Reliability Programme (NERC) (2010) “Flexibility Requirements and Metrics for Variable Generation”, Princeton, August. Nyberg, M. (2011) “Thermal Efficiency of Gas‐Fired Generation in California”, Staff Paper, California Energy Commission, California. Platt, K. (2009) “The Implications of Lower Natural Gas Prices for the Electric Generation Mix in the Southeast”, Energy Information Administration Short‐Term Energy Outlook Supplement, May. Platts Research, (2003) “Coal‐Wind Integration Strange Bedfellows May Provide a New Supply Option”, December. Potomac Economics (2011) “2010 State of the Market Report for the ERCOT Wholesale Electricity Markets”, Austin, Texas. Roques, F. (2007) “Technology Choices for New Entrants in Liberalised Markets: The Value of Operating Flexibility and Contractual Arrangements”, Utilities Policy, vol. 16, Issue 4, December 2008, pp. 245‐53.
Page | 31
Online bookshop I e la n t er ed n a 9 ru tiona l Energy Agency •
Buy IEA publications online:
www.iea.org/books PDF versions available at 20% discount
r dé Fé
A number of books printed before January 2012 are now available free of charge in pdf format on our website
at io n•
75 73 9 Pa ris Ce dex 15, France
Tel: +33 (0)1 40 57 66 90 E-mail:
[email protected]
This document and any map included herein are without prejudice to the status of or sovereignty over any territory, to the delimitation of international frontiers and boundaries and to the name of any territory, city or area. IEA Publications 9, rue de la Fédération, 75739 Paris cedex 15 Printed in France by IEA, April 2013
