Coal's Future Still Dark
Three undervalued utilities are positioned to benefit from the shift to natural gas and renewables.
Cheap natural gas, sluggish power demand, growing renewable energy, and emissions regulations will all continue to crowd out coal as U.S. utilities use more natural gas and renewables for electricity generation through 2025. This year, we have seen more U.S. utilities incorporate renewables and natural gas-fired generation into their long-term resource planning as coal fleets dwindle. The economics of offshore wind energy continue to improve, and two major projects, one off the coast of Massachusetts and another off Rhode Island, were recently selected by state officials to move forward. Connecticut and New Jersey recently approved legislation providing financial support for nuclear plants. However, more than offsetting these positives are three money-losing merchant nuclear plants that recently announced retirement dates early next decade.
We continue to believe that integrated utilities with supportive state regulatory frameworks should benefit as they retire coal plants and replace these generating assets with natural gas and renewables. Dominion Energy (D), Duke Energy (DUK), and Southern Co. (SO) are investing billions in narrow- and wide-moat projects that should result in strong earnings and dividend growth.
Although President Donald Trump has been supportive of the coal industry, we believe coal-fired electric generation can’t overcome the four demand killers and will continue its downward spiral.
Demand Killer 1: Cheap Natural Gas
Coal’s most lethal killers are cheap gas and the development of highly efficient combined-cycle gas turbine power plants. The relatively low initial capital cost of a CCGT plant has resulted in a large fleet of new competitors for baseload coal-fired power generation. As recently as 2005, coal plants produced 50% of electricity in the United States. Beginning in 2007, technologically improved and widespread fracking practices liberated significant quantities of natural gas, and this ample supply drove down prices. By 2016, natural gas overtook coal and generated almost 34% of electricity in the U.S. versus 30% from coal. In 2017, gas-fired generation declined to about 32% of total generation due in large part to a recovery from unusually low gas prices in 2016, but coal remained at approximately 30%.
Utilities opt to dispatch from various sources in their portfolios, depending in large part on fuel prices. Typically, nuclear will provide baseload generation because of its low variable cost and challenges in cycling on and off, whereas gas can be cycled on and off more easily to handle peak demand but may also be used to meet baseload power demand when economics favor it over coal. This has increasingly been the case in the past several years. Wind and solar are intermittent but have zero fuel cost and play a growing role in meeting the daily load in many regions, reducing usage of coal and natural gas for baseload power.
Natural gas prices were modestly higher in 2017 versus 2016, but the decline in gas-fired generation was due mostly to good hydro conditions in the Northwest U.S. and the continued growth of wind energy. The bad news for coal is that natural gas production continues to increase, following a decadelong trend.
Worse yet for coal-fired generation, gas can be substantially cheaper than Henry Hub spot prices in some regions, making coal even less competitive. For example, prices at the eastern Pennsylvania Tetco M3 trading hub were significantly lower than Henry Hub for most of the past four years; as recently as September 2017, gas in this region was $1.60 cheaper than Henry Hub’s $3.02, and even the cheapest coal cannot compete against gas this cheap. Our energy team projects midcycle $2.90/million Btu ($3.00/thousand cubic feet) Henry Hub natural gas and continued negative basis in the Marcellus region, which is in line with the current forwards.
In another strike against coal, gas plants have become more efficient, able to produce greater electricity for a given volume of gas. This means coal will have a harder time competing, even if gas and coal prices hold at today’s levels. A greater mix of more efficient CCGTs versus simple cycle in the fleet improves the heat rate. The heat rate is the amount of energy used by a power plant to generate 1 kilowatt-hour of electricity. Gas turbine design by suppliers such as General Electric and Siemens continues to advance efficiency, further driving down heat rates.
Coal generators’ higher average heat rate--reduced efficiency--is due in part to running emission-control equipment. This is partially offset by more-efficient modern coal units built in recent years as less-efficient, older coal plants retired. This is relevant because the heat rate is a direct input into the economics of operating a plant, thus driving dispatch decisions that determine what plant will be called on to supply electricity.
We forecast natural gas generation will increase to 39% of total U.S. electricity generation in 2025 versus less than 32% in 2017. Our estimate is based on a Henry Hub midcycle natural gas price of $3.00/mcf. The Energy Information Administration’s 2025 natural gas price bear case is $2.97/mmBtu in 2017 dollars, its reference case is $4.07/mmBtu, and its natural gas price bull case is $6.48/mmBtu. The bear case is in line with our midcycle price of $3.00/mcf ($2.89/mmBtu). EIA’s huge range reflects the significant uncertainty of future natural gas prices. EIA’s more bearish view of natural gas prices places our 2025 gas generation forecast of 1,737 terawatt-hours (39% of total generation) near the top end of EIA’s wide range of 992-1,850 TWh (33.2%-42.2% of total generation).
Demand Killer 2: Sluggish Power Demand
The demand for coal-fired generation is also pressured by slowing electricity demand growth. Power demand has decoupled from GDP because of shifts away from power-hungry manufacturing and continuing energy efficiency gains (for example, in lighting, refrigeration, windows, HVAC, and home insulation). We expect power demand to grow at about 1.25% per year versus our assumption of U.S. real GDP growth of 2.2% annually through 2030 based on the International Monetary Fund’s real GDP per capita forecast and state population forecasts.
We assume distributed generation (mostly rooftop solar) represents 2% of our network demand, taking 15 basis points off our utility generation demand. Thus, power required from conventional utility-scale sources such as coal will grow only 1.1% annually.
We forecast that electricity demand will grow more slowly than in any 15-year period in at least the last 70 years. Our 1.25% national electricity demand growth forecast through 2030--1.1% after incorporating distributed generation--is the highest among key third-party forecasts for U.S. electricity demand growth. However, we note that EIA increased its electricity demand growth for 2017-25 to 0.8% from 0.6% in its 2017 forecast.
In general, the key differentiator is our less bearish view of incremental energy efficiency and sector mix shift away from industrial demand. Our differentiated view is that incremental energy efficiency will become constrained by investment cost, consumer inattentiveness, and political will. However, even our relatively bullish view of power demand is no rising tide to float all power generation sources, including coal.
Demand Killer 3: Strong Policy Support for Renewable Energy
Wind and solar energy are considered intermittent resources, sometimes referred to as variable renewable energy. In other words, the wind doesn’t always blow and the sun doesn’t always shine. However, when operating, they displace a portion of the output of competing coal plants because the variable operating costs of wind and solar are low, or sometimes, due to tax policy, even negative. Thus, public policies supporting renewables are demand killers for coal-fired power plants.
The growth of wind has been dramatic. In 2005, wind was less than 0.5% of U.S. net electricity generation. In 2017, generation from wind had increased over 14-fold and represented 6.3% of net U.S. generation. The growth in wind has been driven by federal tax incentives, technology improvements, state renewable portfolio standards, and private-sector sustainability goals. In addition, states have supported transmission expansion to regions that can support wind.
The production tax credit that provides tax incentives for every kWh of electricity generated by a wind turbine has been a significant driver of wind energy growth. Originally enacted in 1992, the PTC provides a credit of $0.015 per kWh in 1993 dollars. The credit is adjusted for inflation annually. The Consolidated Appropriations Act of 2016 reduced the PTC by 20% for projects commencing construction in 2017, 40% in 2018, and 60% in 2019 before being completely eliminated. In 2017, the PTC with inflation and the 20% step-down was equal to $0.019 per kWh.
Also benefiting wind energy has been state renewable portfolio standards. However, illustrating the importance of the PTC and transmission is the fact that Texas surpassed its RPS in 2009, producing 5% of its electric generation from wind, and just kept going without any increase in RPS. In 2017, Texas produced 14.8% of its electric generation from wind and leads the nation in wind power, producing over 26% of the total electricity from wind in the U.S.
Texas spent $7 billion building the Competitive Renewable Energy Zones transmission project connecting wind-rich West Texas with population centers in Houston and Dallas. Looking at all the projects moving forward to benefit from the PTC before it ends in 2019, we expect wind generation to surpass coal generation in Texas by 2023.
Other states with good wind resources and open space for wind farms include Iowa and Kansas. Iowa set an RPS goal in 1983 that it has far exceeded and never updated, and in 2015 Kansas updated its 2020 RPS goal, which it also has already exceeded. These three states provided over 40% of the wind energy in the U.S. in 2017, and RPS in all three states really had just a small impact on wind farm growth. In our opinion, the main driver of wind energy growth in these three states has been the improving economics of wind turbines and the benefit of the PTC, not RPS.
Although the PTC is being phased out, the IRS is allowing projects placed in service four years after the start of construction to claim the credit. Thus, wind projects will benefit from the PTC through 2023. In addition, technology improvements driving down the cost of wind energy and private-sector sustainability goals will probably result in growth through the next decade.
Offshore Wind Energy a Likely Growth Driver
The U.S. has been a world leader in wind energy and, although trailing China in installed capacity, produces more electricity from wind than any other country. Higher capacity factors in the U.S. are due in large part to better wind resources (windy, open areas in the Central U.S.).
However, it is a different story offshore. The U.S. has lagged other countries because of its cheap energy from coal, gas, and nuclear and the abundance of good onshore wind resources and open land. Only one U.S. project is in operation: the 30 MW Block Island Wind Farm off New Shoreham, Rhode Island.
Several states are promoting offshore development, the U.S. Interior Department has received solid interest in offshore lease sales on the East Coast in the past several years, and technology improvements have driven down the cost of offshore wind in Europe.
In May, Massachusetts and Rhode Island gave preliminary approval to projects that could provide as much as 1,200 MW of offshore wind energy. Both projects have European partners, likely providing experience and technology. We believe there is a decent chance the power purchase agreements will be approved by regulators, as Massachusetts legislation requires utilities to contract for 1,600 MW of offshore wind with 15- to 20-year agreements and Rhode Island’s aggressive RPS requires 38.5% of generation from renewables by 2035. Other states with mandates for offshore wind energy include Maryland, New York, and New Jersey.
Another important sign that offshore wind might be ready to accelerate in the U.S. is experienced European wind energy owner/operators putting up hard cash for offshore leases. As background for its July request for feedback, the Bureau of Ocean Energy Management indicated that it had conducted seven competitive lease sales for almost 1.4 million acres in Atlantic federal waters. BOEM has awarded 13 leases with winning bids totaling over $68 million for approximately 17 GW of offshore wind capacity.
As with onshore wind, technology improvements have driven down the cost of offshore wind. With limited experience in the U.S., we look to Europe to observe price trends. Auctions held in Germany, Netherlands, and Denmark in the past couple of years for a total capacity of almost 4,000 MW scheduled to achieve commercial operation before 2025 indicate a dramatic reduction in prices.
Similar to European companies investing in U.S. onshore wind, offshore wind should allow the U.S. to close the gap with Europe. EIA expects the U.S. to overtake Europe by 2020 in total wind energy.
Solar Has Also Had Dramatic Growth
Although starting from a smaller base, representing only a fraction of net electric generation a decade ago, utility-scale solar represented 1.3% of total U.S. net generation in 2017. However, this does not include rooftop solar, which is not separately metered and must be estimated as a reduction in retail demand. The U.S. Energy Department estimates that roughly 1% of total U.S. demand was displaced by rooftop solar installations last year. These installations have significantly reduced or flattened demand growth for utilities in Southern California, Arizona, and Hawaii.
California has been the leader in utility-scale solar. In 2017, over 11% of the net generation in California was from utility-scale solar, representing approximately 44% of the total U.S. generation from utility-scale solar. California’s RPS has been a strong influence on renewable growth, including reduced demand due to rooftop solar installations. California generated over 26% of its net generation from renewables in 2017 (excluding rooftop solar and hydro). Adding approximately 20% from hydro sources, it appears California is on track to exceed its RPS goal (includes hydro) of 50% by 2030. Legislation has been proposed to increase the 2030 RPS to 60% and 100% by 2045.
Although California remains the leader in utility-scale solar, generating over 13,000 megawatt-hours more power from utility-scale solar in 2017 versus 2014 (a 133% increase), other states are gaining ground as technology has improved and costs have dropped. As one might expect, Nevada and Texas have added utility-scale solar on a percentage base faster than California over the past three years. However, strong increases are also being observed in North Carolina, Massachusetts, New Jersey, Florida, Colorado, and New Mexico.
We are also modestly more bullish on utility-scale solar after the investment tax credit was retained in the 2017 Tax Cuts and Jobs Act. The ITC is 30% for projects under construction by 2019 year-end and then declines to a permanent level of 10% in 2022. The significant drop in costs, state-level RPS, and continued corporate demand should result in strong utility-scale solar growth through 2025, especially in regions with poor wind resources.
Note that only utility-scale solar is included in our assumption of renewables. Dramatic increases in rooftop solar have been seen in Arizona and Hawaii, but those result in reduced electricity demand and are included in that coal demand killer forecast.
Bumping Up Our Expectation for Renewable Power Generation
For the entire U.S., we project renewables, excluding hydroelectric, to grow 70% from 2017 to 2025, boosting renewable generation to 659 TWh, representing almost 15% of the U.S. generation mix from 9.6% in 2017. Our previous estimate was 13.3% of the total, but we now believe offshore wind and utility-scale solar will experience stronger growth. Corporate demand is stronger than we thought, and utilities in areas with poor wind resources are now looking at solar to add renewables to their long-term generation mix.
EIA reduced its renewables reference case forecast in the 2018 AEO to approximately 16.1% of total generation in 2025 versus its 2017 forecast of about 17.4%. (Since EIA projects total renewable generation including hydroelectric, we assume hydroelectric generation remains fixed at 275 TWh to calculate EIA’s estimate.) The reduction in EIA’s forecast appears to be driven by the reduction in its natural gas price forecast, which resulted in significantly more generation from this fuel. We note that EIA’s natural gas price bear case, in line with our midcycle price, results in renewable generation representing 15% of total generation, in line with our estimate. Much of this renewable generation will replace coal-fired generation (some replaces older gas-fired plants). Thus, the unrelenting increase in renewables is a demand killer for coal.
Demand Killer 4: Environmental Regulations
As margins decreased because of pressure from the first three demand killers, coal plant owners have been required to make investments to comply with environmental regulations. These regulations had required implementation dates over the past decade and into the next one. Although the most expensive regulations for coal plants addressed air emissions, regulations related to water use, wastewater treatment, and solid waste management could also be material and push a coal plant to early retirement. Costs related to air emission regulations varied due to the age and design of the plant, type of coal used, and site restrictions. These environmental investments are not insignificant for a coal plant owner. The Environmental Protection Agency estimated that a typical 700 MW coal-fired unit could incur these incremental capital costs for the following pollution control installations: scrubber to reduce sulfur dioxide, $287 million-$351 million; selective catalytic reduction to reduce nitrogen oxides, $116 million-$137 million; baghouse for particulates, $97 million-$114 million.
The Mercury and Air Toxics Standards rule was potentially the most expensive for coal plant owners and had deadlines in 2015 and 2016. However, in planning to implement MATS, coal plant owners were faced with historically low natural gas prices and potential additional regulations to address carbon dioxide and other greenhouse gas emissions. This resulted in a surge of coal plant retirements. In 2015, the U.S. Supreme Court stayed MATS because of EPA’s failure to consider the cost of compliance. However, most power generators had already complied with MATS by installing scrubbers or retiring older coal plants. We don’t think there will be material additional coal plant retirements due to MATS, but we do expect additional retirements due to continued economic stress from natural gas and renewables.
The Clean Power Plan put in place state-by-state emissions goals for carbon dioxide that would have put further stress on coal-fired generation. The U.S. Supreme Court issued a stay of the CPP in February 2016. In December 2017, then-EPA Administrator Scott Pruitt indicated that his agency would issue a replacement rule rather than eliminating it outright. The replacement rule was expected in mid-2018 and would only regulate emissions “inside the fence” of existing power plants. The more far-reaching CPP would have directed states to reduce emissions through energy efficiency, purchasing renewables, and operating gas plants at higher utilization over coal plants. We do not expect any policy changes with the departure of Pruitt.
We do not believe the CPP replacement will have a significant impact, positive or negative, on additional plant retirements. States and utilities are acting on their own to reduce carbon emissions. We think the move to abandon the CPP could even embolden states to strengthen renewable energy standards, similar to moves made by politicians and corporations following the Trump administration’s announcement to withdraw from the Paris climate agreement.
Environmental regulations increase the cost of operating a coal-fired power plant and result in new coal plants not being competitive with other methods of electricity generation. An analysis of the levelized cost of electricity is a useful tool to measure the competitiveness of different generating technologies.
The key inputs in calculating LCOE include capital costs, fuel costs, operating and maintenance costs, financing costs, and the expected utilization rate for each type of plant. The importance of these factors varies by technology. Wind and solar have no fuel costs and relatively low variable O&M. Thus, for these renewable generating sources, capital costs and utilization rates have the most significant impact on LCOE. CCGT plants have relatively low capital costs but high fuel and moderately high variable O&M driving LCOE. Nuclear and coal have high capital costs and variable O&M but low fuel costs. Tax incentives for wind, solar, and nuclear also lower the LCOE for these technologies.
The estimated average LCOE for different technologies in different regions of the country is calculated each year as part of EIA’s Annual Energy Outlook. One might disagree with some of EIA’s inputs, but the variances are not significant enough to change the conclusion. Coal is not competitive, and we think this is unlikely to change without a breakthrough in coal plant technology or a dramatic increase in natural gas prices. We believe both are unlikely in the foreseeable future, and currently there are no new coal plants under construction or in advanced development.
We project coal plants will provide about 23% of U.S. power in 2025 versus 30% in 2017 and 50% in 2005. This is a negative 2.1% compound annual growth rate from 2018 to 2025. To project coal demand, we start with total demand, subtract projections for all other sources, and use coal generation as the remainder. We estimate that demand will grow at a 1.1% compound annual rate through 2025. From this we subtract other sources by projecting capacity from plants proposed, pending, and under construction, less retirements, to model incremental generation from natural gas, nuclear, oil, and hydroelectric. We also subtract renewable generation, which we expect to grow to almost 21% of total generation (including hydro) by 2025. Our 2025 coal generation forecast of 1,013 TWh (23% of total generation) is between EIA’s reference case (27.1%) and its natural gas price bear case (20.7%).
Three Positioned to Benefit Trade at a Discount to FVE
In addition to our more bearish forecast of natural gas prices driving more gas generation and less coal and renewables, we also consider the long-range plans of vertically integrated regulated utilities to retire coal plants in our forecast of coal usage for electric generation. Until recently, the coal plants that retired were older, smaller, and less efficient. Recent retirements have included newer, larger, and more efficient plants owned by merchant generators in the Northeast and Midwest. These plants couldn’t compete in a deregulated market with newer CCGTs and renewables. Regulated utilities are protected from low market prices but are now responding to public and regulator pressure to reduce the overall carbon intensity and other pollutants emitted from their power plant fleets.
They are doing this by adding renewables and natural gas-fired plants and retiring coal plants. These companies work with regulators in developing integrated resource plans. These long-range plans provide a view of what resources will provide electricity over the next 10-20 years. Three of the largest vertically integrated regulated utilities in the country--Dominion Energy, Duke Energy, and Southern--have integrated resource plans that are examples of the slow but steady transition away from coal. The long-range plans of these three utilities are similar to the integrated resource plans of many other regulated or municipal integrated utilities across the country.
Dominion, Duke, and Southern all have supportive state regulatory frameworks that allow them to benefit from retiring coal plants and replacing the capacity with natural gas and renewables generation. Transmission and distribution will also benefit due in part to sluggish power demand, as these utilities focus on grid modernization instead of environmental upgrades for existing coal plants. The significant investment opportunities should result in growing rate base, earnings, and dividends for these utilities and other regulated integrated utilities.
Charles Fishman does not own (actual or beneficial) shares in any of the securities mentioned above. Find out about Morningstar’s editorial policies.