COVER STORY: The cost of cutting coal -- Xcel is backing away from coal power, but who pays?

Today, rumbling conveyor belts feed a stream of coal to units 1 and 2 at Xcel Energy’s Comanche Generating Station, just outside Pueblo. But the endless flow of coal and the roaring boilers could begin to go quiet in the next four years under a plan before Colorado regulators.

But in shuttering the two 1970s-vintage power units 10 years earlier than scheduled, there will be costs and questions: Who pays, how much and how? Xcel has put the cost figure at $193 million, although that number is in dispute.

These aren’t simply local questions. Across the country, utilities are shutting coal-fired power plants, in some cases because they are old and no longer market competitive, and in others because  to continue running them requires expensive pollution control equipment—akin to adding a new catalytic converter to a 1988 Chevy Nova.

Since 2009, 268 U.S. coal-fired units have closed, leaving 262 operating plants, according to the Sierra Club’s Beyond Coal Campaign.

In 2018, another 12.5 gigawatts of coal-fired generation is expected to close, according to energy analyst Bloomberg New Energy Finance. That is equivalent to closing 40 Comanche 1 units.

Many of those plants will close ahead of their projected operating life, before the investment in building them has been amortized, making each a “stranded asset.” In the case of a regulated utility, like Xcel, it has the right to recoup that money from customers.

“Stranded assets are high on the list of problems, especially for regulated utilities,” said Zach Pierce, a spokesman for the Beyond Coal Campaign. “This is a big challenge all over the West.”

It could be a big challenge in Colorado, where Xcel still has an estimated $1.9 billion in coal plants still operating. “We have to find a way to pay,” Pierce said. “There is no simple solution.”

The Comanche story is, however, a bit different since the idea of closing the two plants isn’t based on the immediate need for investment in the units, which went on line in 1973 and 1975, or a lack of market competitiveness.

The closures are part of a plan to radically reposition Minneapolis-based Xcel’s Colorado subsidiary, Public Service Company of Colorado , as a predominantly renewable energy-generating utility.

“We are very committed to decarbonizing when the technology and policy choices make it possible to do so,” said David Eves, Xcel’s top executive in Colorado,  said in announcing the plan last September. “It’s all about the economics.”

The utility’s “Colorado Energy Plan” (CEP) would retire the two Comanche units, with a combined 660 megawatts of generating capacity, and seek bids from developers for 1,000 MW of wind power, up to 700 MW of solar facilities and 700 MW of natural gas-fired generation or energy- storage projects.

The CEP would also involve the construction of a new switching station for a southern Colorado transmission “energy resource zone” to foster development of renewable-energy power resources in rural Colorado.

A third Comanche unit — the $1 billion Comanche 3 unit, which went online in 2010 — would continue to operate; it alone can generate up to 750 MW.

The plan, the company said, could lead to $2.5 billion in clean-energy investments It could result in Xcel producing 55 percent of its electricity from renewable resources, and cut the utility’s carbon emissions 60 percent below 2005 levels by 2026.

In 2016, coal provided 46 percent of Xcel’s electricity, followed by natural gas at 25 percent and wind at 23 percent. Solar and hydropower provided the rest.

The CEP isn’t an act of corporate altruism, as testimony to the Colorado Public Utilities Commission shows. The plan is in response to customer calls for more clean energy, the risk that there could be future regulations on emitting carbon, and the fact that it is likely the two Comanche units would need $190 million in pollution controls to operating operate through 2035.

Xcel’s caveat is that it would not go ahead with the project if it didn’t create savings for customers or at a minimum, not cost them any more than they already pay.

At the outset, the plan looks promising as Xcel issued its request for projects and received 430 proposals for non-coal power development — about eight times as many as in its 2013 call for new projects.

The rates quoted were some of the lowest in the country, including 96 wind projects with a median price of at $18 a megawatt-hour (MWh) and 152 solar photovoltaic projects with a median cost of $29.50 a MWh. The price of 28 stand-alone battery-storage projects averages $11.30 a MWh and 11 wind-plus-storage projects average $21 a MWh.

By way of comparison, Xcel’s estimate for the all-in costs of operating the two Comanche units is about $31 a MWh.

Xcel packaged the CEP plan to the Colorado PUC in a “stipulation” that was supported by 14 other parties, including labor groups, big industrial customers, consumer advocates, environmentalist, the solar energy industry, and independent power producers who sell electricity to Xcel. The PUC staff also signed on.

“This is an opportunity to removed 4.5 million tons of carbon dioxide a year. In 10 years, that is a lot,” said Erin Overturf, an attorney with the environmental policy group Western Resources Advocates, which supports the stipulation.

“It has clear environmental benefits and does that without raising customers’ bills,” she said. “It is a win-win, and it is voluntary on Xcel’s part.”

Xcel had hoped that with such broad backing, the PUC would approve the stipulation giving the utility a green light to proceed. But it didn’t, and here is where the complex question of costs and who pays reared its head.

The first question of who pays is the simplest. Customers will pick up the cost of closing the coal plants and building or buying the new renewable energy. How they pay and how much are more bedeviling questions.

The PUC did not give Xcel’s stipulation a blanket approval. Instead, it asked for more information and figures, in part because of challenges by the Coalition of Ratepayers, a group of businesses and citizens lead by the libertarian Independence Institute. The coalition contends it will be cheaper to keep the two coal plants running.

The coalition questioned some of Xcel’s accounting techniques, arguing they overstated the value of the renewable-energy plan.

“Xcel has the ability to put its thumb on the scale,” said Amy Oliver Cooke, the Independence Institute’s executive vice president. “We went into this being skeptical of a utility saying that is going to make a profit and save customers money. We are just asking is there truth in advertising.”

The PUC asked Xcel to run its numbers stripping away some of the accounting techniques, and even the PUC staff said the annuity method Xcel used as one measure could “skew” the numbers and should be “given little weight.”

Xcel was set to present several scenarios for the adoption of renewable resources, but said it wouldn’t provide a “least-cost” option unless it offered a minimum of $50 million in savings. The commission said it wanted to see a least-cost analysis whatever the savings.

The coalition also called for an annual cost-impact report, which the commission also adopted. “I think the coalition raised a good point,” Commissioner Wendy Moser said, before the commission voted on Xcel’s plan last March.

Xcel had go back to the drawing board on some measures and also had to evaluate the bids it has received for renewable resources. It was supposed to deliver a report to the PUC later this month, but was granted a delay till June by the PUC.

“Things are up in the air in regard to the Colorado Energy Plan,” said Mark Stutz, an Xcel spokesman.

As for the cost of closing the Comanche plants, that was split off into a separate and even more arcane case, or in PUC parlance, “docket.” Here, the accounting and regulatory jungle grows thick. As daunting as it is, a little tweak here or there can redirect millions of dollars from customers’ pockets to corporate coffers, so it is worth venturing into the weeds.

A lot of the complexity stems from the fact that, unlike a standard corporation, a utility that has been granted a monopoly over a service area is regulated by a government utility commission — in this case, the Colorado PUC — with the goal of making sure investments are prudent, rates are reasonable and the company is economically viable.

The process is ponderous. A utility must apply for an OK to build a plant or lines, which requires a docket. A request for a rate increase requires a docket and commission approval. Many dockets take a year or two to complete.

It can take millions of dollars and years to build a new plant, but a utility can’t include that plant in its rate base until it is “used and useful.”

“It is a completely archaic system,” said Leslie Glustrom, who as a private citizen has intervened more than 20 Xcel dockets, including the Comanche case. “It dates back to the beginning of the 20th century.”

Since everything takes so long, a variety of provisional measures that allow a utility to collect some revenue has developed, such as the General Rate Schedule Adjustment (GRSA), a sort of interim rate; the Construction Work in Progress, which allows some construction cost to be recouped before a plant is running; and the regulatory asset, a way for gathering the costs for a new plant or project in one place.

It is in this thicket of accounting gizmos that the battle over the cost of closing the units is being fought. Xcel has proposed setting up a regulatory asset for the closure that would be where the $192.9 million would be banked between 2022 and 2028. It is also proposing that it get a return on the regulatory asset through a GRSA.

To hold customers harmless, Xcel proposes diverting half of the 2 percent charge each customer pays on the monthly bill for solar-energy projects, the Renewable Energy Standard Adjustment, to pay off and decommission the Comanche plants.

Sound complicated? That’s what intervenors in the case objecting to the proposal say.

“This approach is needlessly complex,” Kevin Lucas, director of rate design for the Solar Energy Industries Association, said in PUC testimony.

Lucas and Glustrom have both called for a separate charge on Xcel’s bills for closing the plants. There is already an item charge on the bill for closing six plants under the state’s Clean Air Clean Jobs Act that comes to about $2 on the average monthly bill.

Scott Brockett, Xcel’s director of regulatory administration, said in an email that the company’s approach has been used before in accounting for other units that have been retired early and that using a separate GRSA helps manage the costs so rates do not jump.

Still, Energy Outreach Colorado, which helps low-income households with their energy bills, is also against the approach.

“We are opposed to GRSA because that goes into fixed costs, and one of our goals is to keep fixed costs down,” said Andrew Bennett, Energy Outreach Colorado’s  director of advocacy. “Before you flip on a light switch, (your) electricity bill would be higher.”

The state Office of Consumer Counsel was a signatory to the stipulation, but opposes Xcel’s plan to get — in addition to the funds to amortize and close the Comanche plants — about 7 percent return on the regulatory asset. “This was not part of the agreement,” Cindy Schonhaut, director of the OCC, said.

When a regulatory asset is used to keep track of costs for building a new project, the utility is spending money, and a return on the cost of that capital is appropriate, Schonhaut said. But this asset is being used to retire an existing plant. “There is no steel in the ground,” she said.

Xcel, Brockett said, is seeking a return on “prudently incurred costs,” and that “because the proposal represents a compromise made in order to benefit customers,” the uncollected depreciation for the plants remains in the rate base, even after the plant closes, until they are paid off. Xcel calculates savings customers as a result of closing the plants of $223 million.

Each one of these accounting arguments could translate into big dollars. The proposed mechanism could lead to a “windfall” of $14.1 million, Charles Griffey, the Coalition of Ratepayers’ expert witness, said in testimony.

The tax rate used for carrying costs in the Xcel proposal was 35 percent, even though the federal corporate rate was reduced to 21 percent last year, a group of rural cooperatives that buy wholesale power from Xcel pointed out in testimony, increasing the capital requirement by $8.3 million.

“The $8.3 million referenced is not correct,” Brockett said. “The company will update the tax rate in its 120-day report and rebuttal filing.”

The testimony, the supplemental director testimony, the answer testimony, the motions, and rebuttal filings will continue to pile up, thousands of pages. There are already six months’ worth of paper in the docket and more to come.

“The issue is transparency,” said Cooke. “We are just asking for honesty and transparency in these numbers and then let everyone decided. But the way they’ve packaged it, calling it creative accounting is kind.”