If there’s one takeaway from the massive Power Gen International conference, it's this quote from the first day by CAISO CEO Steve Berberich:
“Capacity will no longer be the coin of the realm. Capability will be the coin of the realm.”
It’s no secret that the growing penetration of renewable generation is changing the way the power grid works. While utilities have traditionally relied on steady baseload fossil fuel plants, intermittent renewables – whose generation fluctuates over time, often within minutes – require the grid to act fast to balance supply and demand.
Energy storage could theoretically fix the problem, but it remains limited in scale and too expensive in the short term to be an immediate solution. To cope with the challenge, utilities and grid operators have instead been calling on their traditional power plants — especially gas-fired ones — to respond faster to market signals. In practice, that means plants are called upon to ramp up and down more often, delivering more or less power depending on what the market needs.
The problem is that most of the existing power plants on the system today weren’t originally designed to cycle up and down with the frequency demanded by grids with increasing amounts of variable generation. Power plants are being asked to perform functions that may be outside of the historical norm — and that's causing generators, grid operators, and regulators to rethink how power plants will need to operate in the future.
The rise of plant cycling
The new demands of grids with renewables, said James Schetter, president of the energy research firm Renewable Impacts, are beginning to increase costs for many cycling plants.
“Some say that renewables are starting to stress the operations of conventional power plants because they are the force behind increased cycling,” Schetter said.
While those cycling costs in many areas of the country remain low, he said, market trends and state renewables targets mean that the costs of increased cycling are likely to increase in the future.
“Cycling cost impacts seem low now, but as renewables penetrations increase, they may increase as well,” he said. “Those that are familiar with the CAISO duck curve, you can imagine what that curve will look like when we hit 50% renewables.”
Schetter has a novel idea to correct for cycling costs: making renewable energy developers pay for the increased cycling they cause. But while he and his team at Renewable Impacts seem to have devised a comprehensive methodology for measuring the causes of increased cycling, not everyone at the conference agreed that it was the responsibility of renewables companies to pay for it.
In crafting their new report, Schetter and his team analyzed 100 different power plants in the California ISO (CAISO) footprint, producing more than 300 cycling causality reports outlining the sources of the increased ramping. The firm analyzed 12 different drivers of cycling, including renewables, system load, other thermal plants, power imports and more.
“There is no need to guess at what causes what,” Schetter said. “There is a demonstrated way to determine the causality … of power plant cycling.”
The firm used publicly available EPA data to measure the movement of baseload plants — referred to as reactors in the report — and data from CAISO to measure the movement of the cycling drivers. The researchers found that while renewables aren’t responsible for most of the cycling that plants must do on the grid today, they are a fast-growing piece of the pie.
In 2012, Schetter said, solar energy was responsible for 2.6% of the cycling of the 100 traditional generation plants examined in the CAISO footprint. Wind energy drove 7.7% of the cycling, and other renewables, such as small hydro and biomass, caused 1.7%. The balance was made up of the demands of power imports, thermal and nuclear plants (31%), and the demands of CAISO load (57%).
But the numbers themselves weren't as important as how Renewable Impacts saw them change over time. When they went back to examine cycling causality in 2013, they found the drivers had significantly changed from even just the year before.
While solar had been responsible for 2.6% of the cycling a year before, the firm found it was responsible for 7.6% of it in 2013. Wind’s causality grew as well, although less so than solar, leaving it responsible for 7.9% of the cycling in CAISO in that year.
“Solar was up a whopping 187% in a year,” Schetter told the audience, “and wind was up 3%.”
Examining data from 2011 to 2013, the Schetter said a few trends began to emerge.
“Wind energy penetration vs. wind causality was found to have a relationship of 1:1,” he said. “In other words, a 1% increase in wind penetration resulted in a 1% increase in wind causality.”
But for solar, “the ratio was 1:3,” he said. “Solar was found to create three time the cycling causality of wind."
Using these measures, Schetter said utilities and regulators can assign costs of cycling to each of their drivers. The key question remains whether renewables should pay for the increased flexibility of fossil plants that their presence on the grid demands.
“My answer is yes,” Schetter said. “Renewables should pay, up to their determined causality level.”
The debate over renewables payments
While no one at the panel discussion challenged Renewable Impacts' methodology for assigning cycling costs, a number of audience members and panelists critiqued Schetter’s proposal to make renewables pay for cycling.
A lively discussion between panelists and audience was kicked off when Utility Dive asked Schetter why renewables should pay for increased plant flexibility, instead of expecting traditional generators to foot the bill for upgrades demanded by more variable generation on the grid.
Schetter said it is largely because of the fast-growing role of renewables in pushing more and more cycling.
“The initial development of these renewables has been very small, so the impact of that has really been in the noise of the system, but now that we see penetrations coming along to a level where they’re starting to have this impact that can now be identified,” he said. “With states like New York and California going to as much as 50% [renewables], the effect of greater penetration of renewables will cause this type of cycling activity, and the policy question becomes who should pay for that?”
Other panelists were not as enamored with Schetter’s logic.
“Just because we can say that renewables are contributing to the increased need for cycling, what doesn’t follow is that renewables are responsible for paying for that,” said Sky Stanfield, special counsel to the law firm Shute, Mihaly & Weinberger, who works on behalf of the Interstate Renewable Energy Council (IREC).
“The fact that something takes away part of your market or changes the way you need to upgrade doesn’t mean that external force is responsible for paying you for the change — why are renewables responsible immediately for those changes?”
“Because they’re intermittent,” someone from the audience cut in. To that, Schetter added: “Because they’re the driver.”
“You can say they’re the cause,” Stanfield responded, "but that doesn’t mean they’re the ones that need to pay you for your change. Maybe that’s just an inherent part of the nature of your resource is that you have to respond.”
Schetter said that his proposal is not that different from how regulators and utilities set electric rates.
“My understanding of setting electric rates in the industry is that you first start from a cost assessment, and then from there it can be modified based on policy directives or issues,” he said. “This is essentially the same step. This would be analogous to maybe a cost of service study so you can do an effort without a form of bias and say there’s a cause and effect going on here and we can now identify this and assess it.”
In a nutshell, his approach would show regulators “what it would look like if you would like to allocate costs from a production bucket to a resource bucket.”
That argument did not placate former Colorado utility regulator Ron Lehr, who is now working as a lawyer for America’s Power Plan.
“The cost of service is a rough tool to assure end-use customers that don’t have other choices, that their bills have some relation to the cost to serve them,” he said. “But the cost of service notion applied to competitive entry in the generation market simply doesn’t wash in my mind. Regulation has never been about protecting the utilities’ generators from competitive entry. That is not what regulation is for.”
The responsibility should not be on the new market entrants to subsidize plants that can’t keep up with the changing grid, he said.
“In my view, those plants are obsolete that can’t flex to meet society’s requirements,” he said, noting that Colorado will soon surpass 30% renewables “simply because of economics.”
“If the old plants can’t keep up and do what they need to do, they’re obsolete,” he said. “If they’re obsolete, they need to leave the market.”
There are other ways of compensating fossil plants for their increased cycling without burdening the emerging renewables industry, Lehr and Stanfield agreed.
Costs could be spread across ratepayers or different participants in the markets, Stanfield said, and grid operators can create market products for grid flexibility that allow plants to earn more for the grid services they provide.
“There’s a lot that can be done to make the plants more flexible at very low cost,” Lehr said. “There are many other strategies that we want to see happen before we start taking money from wind and solar to take care of old, obsolete fossil plants.”
The debate over cycling costs underscored the key trend at the generation conference Power Gen International: Capability — and not simply capacity — is increasingly what matters for generators today.
Looking ahead to when 'capacity will no longer be the coin of the realm'
For generators, the message from Berberich about capability and capacity is that it’s no longer enough to simply provide power on the 21st Century power grid. Modern power plants, especially those providing readily dispatchable power, will need to be flexible enough to respond to momentary changes in supply and demand.
At present, fast-acting gas turbines are predominantly the cheapest option to offer many of the services that the grid demands, such as quick-start and cycling.
“Today we’re designing gas turbines that in 10 minutes get to simple cycle operation and in 30 minutes get to combined cycle operation,” Joe Mastrangelo, CEO of GE Power Generation Products, said in his keynote on Monday. “So, how quickly can you bring a gas turbine online and how can you provide tools to allow it to sync to the grid faster so you can balance the intermittency that occurs with renewables [is the key question].”
But while gas remains the most attractive technology for these services from a cost standpoint, Mastrangelo cautioned the audience to keep an eye on energy storage. It largely remains more expensive than gas turbines on a per kWh basis, but batteries can ramp much faster than even the most efficient gas plants, given their lack of many moving parts, and their price is coming down quickly.
“Battery storage – there’s a technology that we all need to think about that’s going to challenge the dynamics … about how gas remains a competitive fuel source in the marketplace.”
To stay competitive with batteries in the decades to come, Mastragelo said the gas generation industry needs to increase its efficiency. While generating power from gas turbines today is 35% more economical today than it was five years ago, most of that is due to historically low gas prices. If prices were equal to what they were then, the GE exec said, “it would only be 7% more cost effective.”
“We as an industry can’t rely on Henry Hub gas staying in the $2 range,” Mastrangelo said. “What we need to be able to do is to continue to drive that 63% combined cycle efficiency up toward 65%.”
“That’s what’s going to continue to keep gas power generation as a competitive fuel source into the future.”
Correction: The original version of this article identified Sky Stanfield as an attorney at Keyes, Fox & Wiedman LLP. That is incorrect. Stanfield is Special Counsel to Shute, Mihaly & Weinberger.