This is the second of a five-part series exploring oversupply in the power sector and the factors driving a glut of natural gas-fired power plants.

Ohio, for many decades one of the centers of the U.S. coal industry, has been undergoing a transformation triggered by a shift in demand to other energy sources — primarily natural gas. Today, at least six natural gas plants are in some phase of planning or development in the state, and more than $25 billion is being spent to build new power plants — nearly all of which are natural gas — across the Appalachian Basin states of Ohio, Pennsylvania and West Virginia. The eastern half of Ohio also lies atop the Utica and Marcellus formations, two of the richest shale gas deposits on earth.

With little notice or fanfare, coal country has become gas country.

Monday Why a US gas-building spree continues despite electricity glut Tuesday PJM market rules drive an era of oversupply Power plant sales in PJM: A buyer's market Wednesday In Virginia, Dominion faces challenges to its reign Thursday Hailing renewables, NextEra bet big on gas in Florida Friday Eyeing zero-carbon grid, California grapples with gas exit

The entity charged with operating the electric grid for this region is the PJM Interconnection, the largest market for electric power in the United States. More than 65 million customers in the mid-Atlantic and beyond count on PJM's analysts and officials to keep the power on and to steer billions of dollars of utility investment into building new power plants and essential grid infrastructure. In recent years, PJM has been at the vanguard of a shift to competitive wholesale power markets. The idea behind the shift is that markets, not regulators and lawmakers, should determine the right level of investment in the grid. And if the markets are wrong, investors, not electricity consumers, should pay the price.

However, an S&P Global Market Intelligence investigation indicates that as billions of dollars of investment continue to flow into natural gas-fired power plants in the region, consumers may be paying a heavy price for an unnecessary and unwise boom in fossil-fuel generating assets. And an ongoing effort to reform PJM's capacity market, which is designed to ensure that enough power is available to keep the lights on, has been blasted as inadequate and misguided by federal regulators, power generators, renewable energy advocates and environmental groups.

From 2008 to 2017, generation capacity in PJM expanded by more than 15,000 MW despite essentially flat demand growth. Nearly all of that was in the form of new natural gas plants.

As a result, reserve margins — the built-in cushion of excess capacity above peak demand — have soared. PJM's 2018 summer target reserve margin was 16.1%. The actual margin of excess power was more than twice that, at 32.8%, and the anticipated reserve margin for 2021, according to the North American Electric Reliability Corporation, is nearly 40%.

Nevertheless, according to S&P Global Market Intelligence data, more than 29,000 MW of new natural gas plants are planned or under development in the region.

PJM officials point out that the markets have consistently provided reliable power at historically low prices. But critics say that, rather than opening a free market that gets sufficient power to customers at the lowest possible cost, PJM has created a power glut that serves mainly big utilities and power-plant developers, rather than ratepayers.

In a May 2019 filing with the Federal Energy Regulatory Commission, watchdog group Public Citizen said that PJM’s market structure "benefits a very narrow set of financial interests: its members that own nuclear, coal and natural gas power plants." The group charged that PJM "is run less as an independent transmission operator, and more as a price-fixing cartel.”

A hybrid market

At the center of the power glut is the annual auction that PJM calls the "base residual auction," the mechanism that drives the capacity market.

That auction determines, three years in advance, the price per MW-day that generators will receive to make their output available to grid operators. Those payments provide a guaranteed revenue stream for plant owners, whether they actually produce electricity for the grid or not, while ensuring there is adequate capacity to keep the lights on even on days of peak usage. But while generators compete on the price of their capacity, the buy-side of the auction is predetermined — administrative rules set by PJM officials, not the market, determine how much capacity is procured.

"Unlike in other kinds of markets, PJM's capacity market demand curve, known as the Variable Resource Requirement (VRR) Curve, is not determined by customers or their utilities deciding how much they want and at what price," wrote Jennifer Chen, senior counsel for federal energy policy at Duke University's Nicholas Institute for Environmental Policy Solutions, in a June 2018 blog for the Natural Resources Defense Council. "Instead, PJM designs the VRR curve to procure a certain amount of capacity at each price point."

The result has been to provide investors, utilities and developers with incentives to continue building new plants, despite the growing power glut. A November report from Grid Strategies LLC found that capacity markets in PJM, New York ISO and ISO New England have driven excess capacity at a cost of roughly $1.4 billion per year across the three markets. PJM today "has more excess reserves than the Electric Reliability Council Of Texas Inc. has total reserves," Joe Bowring, president of PJM's independent market monitor, Monitoring Analytics, said during S&P Global Platts' Financing U.S. Power Conference in October.

Those price incentives have been strengthened by load forecasts — predictions of peak demand that help determine the amount of power generation that PJM will procure — that, until recently, have been consistently and dramatically higher than actual demand. For years, PJM long-term forecasts saw demand in the region growing well into the 2020s, sending a powerful signal to developers of and investors in power plants. Since 2011, however, the growth rate has turned negative: peak load in PJM has fallen by nearly 1%.

The 10-year forecast is important because investment and construction decisions on new generation capacity, especially large natural gas plants that cost about $1 billion apiece to build, are not made year-to-year. Instead, they are made based on time horizons of 10 to 30 years.

Since 2014, PJM analysts have gradually begun to temper their forecasts. The 10-year annual growth rate forecast fell from 1.3% in 2011 to 0.6% in 2016. In 2017, the forecast fell to 0.2% — essentially flat growth. That reduction resulted from significant modifications to the forecast model. PJM's analysts, for example, finally started to account for the rapid spread of rooftop solar generation in 2016 — at least a decade after that trend began to accelerate.

It became clear, said Tom Falin, PJM's director of resource adequacy planning in an interview with S&P Global Market Intelligence, that "we had a history of over-forecasting."

Still, new capacity additions — specifically natural gas plants — continue to outpace new demand and retirements, and PJM ratepayers continue to be burdened with excess capacity well beyond required levels. Summer peak load in 2018 was approximately 147,000 MW, so at a 32.8% reserve margin, that's 24,549 MW of excess capacity on the system, above the 16.1% target reserve margin — more than the total generation capacity of Georgia.

A PJM spokesperson pointed out that a significant quantity of excess capacity on the system does not actually clear in the capacity auction and thus is not paid for by ratepayers. The margin of cleared reserves was around 21% in the most recent capacity auction, held in 2018; the remaining capacity, up to the 32% margin, does not receive capacity payments and thus does not result in additional costs to ratepayers.

"'Keeping the lights on' to an acceptable standard is a critically important objective," wrote Chen, "but 100-percent-fail-safe reliability is impossible to achieve and prohibitively expensive to attempt."

A flood of plants

There is broad agreement that, as currently constructed, PJM's capacity market is flawed. The "Quarterly State of the Market Report for PJM: January through September," released by PJM's independent market monitor in November 2018, found that the capacity market is "not competitive" and questioned whether PJM’s markets can coexist with efforts to boost the role of renewable and nuclear power resources through state subsidies. The question is how to fix it.

The grid operator has been engaged for more than a year in a process to reform its capacity market. Unfortunately, the result has satisfied no one except power generators, who, under the latest scheme, would be allowed to continue building new capacity unchecked. In June 2018, FERC rejected two proposals from PJM officials for capacity market reform, resulting in a delay of the scheduled auction. That auction has still not been held. A revised option that was approved in April also failed to satisfy many policymakers and ratepayer advocates.

The new plan, said FERC Commissioner Richard Glick in a dissent to the commission’s April approval of changes to PJM’s methodology for calculating demand, will perpetuate the grid operator's "chronic oversupply" of generation capacity.

The troubled effort at market reform helps explain why developers and investors see opportunities in PJM at energy prices well below the levels at which new plants are ostensibly economic, and why, in an era of oversupply, natural gas plants continue to get built.

Read More: Power plant sales in PJM: A buyer's market

In a 2018 paper in The Electricity Journal, Eric Hsieh, the director of grid systems and components at the U.S. Department of Energy, and James Kennedy, formerly at the DOE and now an equity research associate with Guggenheim Partners, found that, despite the oversupply, the construction boom continues for a relatively straightforward reason: it looks like a safe investment opportunity.

Examining the cost and location data for 20 combined-cycle natural gas projects slated to come into service in PJM between 2017 and 2020, Hsieh and Kennedy found that even under a bearish scenario — if electricity prices turn sharply downward — the internal rate of return over 20 years for a plant developer would reach 11.8%. In a high-price scenario, the rate of return climbs to 16%. Thus, even if market conditions deteriorate dramatically, Hsieh and Kennedy found, investors are likely protected.

That conclusion provides more evidence that investors and developers are building plants based on factors that increase the supply without regard to demand for electricity. The absence of downside risk indicates that investors, shielded by cheap fuel and capacity market payments, can make money on these plants even if PJM remains oversupplied well into midcentury.

In Ohio, the boom continues

Nowhere is the natural gas boom in PJM more apparent than in Ohio. Built at a cost of $900 million, the 940-MW Lordstown Generating Station in Trumbull County came online in October 2018. It is one of three massive natural gas plants that have recently come online in the state, and at least two more are under construction, representing nearly $9.3 billion in investment.

The Lordstown plant was originally developed by Clean Energy Future, a Massachusetts-based developer that has been involved in multiple natural gas plants in Ohio. Macquarie Infrastructure Partners III, an investment fund managed by a subsidiary of Macquarie Group Ltd., acquired a majority interest in Lordstown Energy Center and oversaw the completion and commissioning of the plant

Although 2019 has seen some gas plants in financial distress, Macquarie is among the investors and developers who believe that new, state-of-the-art combined-cycle gas plants with favorable capital structures are more or less impervious to future fluctuations in demand and in the prices of the gas those plants consume and of the power they produce.

"We could see substantial increases in gas prices and still be producing — it's not the sort of plant where if that gas price goes wrong it's not going to run," said Aaron Rubin, a managing director with Macquarie Infrastructure and Real Assets, a unit of Macquarie. Even if market fundamentals deteriorate, he said, "We'll still make a decent return."

In PJM, the biggest power market in the U.S., "it's difficult to see a scenario where a modern, efficient gas-fired plant like Lordstown isn't part of the generation mix."

That mix is increasingly dominated by natural gas plants. PJM officials acknowledge that the development of renewable power capacity has stalled across the region. "In our interconnection queues, we used to have a significant amount of renewables in the queue," said Stu Bresler. "That has come down drastically. I’m not sure why."

Outside observers are less puzzled.

"The world is headed for microgrids and distributed generation," said Andrew Thomas, executive in residence at Cleveland State University's Energy Policy Center. "But utilities continue to invest heavily in 1950s-style centralized command-and-control generation and the ubiquitous grid."

Calling the capacity market "the elephant in the room," Illinois Commerce Commissioner John Rosales, speaking at the National Association of Regulatory Utility Commissioners' annual meeting in Orlando, Fla., in November 2018, said that PJM's "inherently flawed and unnecessarily complex" market rules actually prevent the functioning of the free market.

In Ohio, those rules continue to work to bring more gas-fired capacity online.

In August 2018, a judge's decision cleared the way for yet another major gas plant to be built in the region. Scheduled to come online in 2023, the 940-MW Trumbull Energy Center is under development by Clean Energy Future on a site literally next door to the brand-new Lordstown plant.

The new plants will add more than 1,800 MW of added capacity to PJM's already overstuffed portfolio. The combined price tag: more than $1.8 billion.