ERCOT: shrinking reserve margin
Texas is the only organized electricity market in the United States where there is growing demand for electricity. ERCOT’s latest capacity, demand and reserve report projects a 9.3% reserve margin by this summer, which is below the 13.75% target. Some developers see an opportunity for new flexible generation, like gas peakers, that can balance out the large amount of wind farms on the ERCOT grid. Large base-load power plants are being retired because wholesale power prices are so low in ERCOT, due to low natural gas prices, that it is hard to recover fixed costs. However, developers of peaker plants expect electricity shortages this summer that will allow such power plants to make money off price spikes. There could also be development of new “switchable” power plants along the southern border to sell capacity into Mexico while continuing to earn energy payments in ERCOT.
Three close observers of the Texas market talked at the Infocast “projects & money” conference in New Orleans in January about the Texas market. The three are Karl Dahlstrom, a partner in Halyard Energy Ventures, which has 2,000 megawatts of natural gas peaking plants and 100 megawatts of storage facilities under development in Texas, Bob Helton, senior director of regulatory affairs at Dynegy, a large Texas-based independent power company, and Kevin Smith, president of Tenaska Power Services Company, which transacts physical and financial wholesale power and provides congestion management, hedging scheduling, settlement, market interface and other services for 44,000 megawatts of third-party generation. The moderator is Deanne Barrow with Norton Rose Fulbright in Washington.
MS. BARROW: ERCOT is undergoing a number of fundamental changes. Karl Dahlstrom, what are they?
MR. DAHLSTROM: The largest fundamental in ERCOT is consistent load growth. ERCOT has seen about 1.5% load growth per year for the past few years, and it looks like this this will continue for the foreseeable future. In a market of about 70,000 megawatts, that means about 1,000 megawatts a year of additional capacity is needed.
MS. BARROW: Bob Helton, anything to add?
MR. HELTON: The Texas Public Utility Commission has a docket open to look at some very fundamental and large changes to the market. We are an energy-only market, which means that all your revenues have to come from the energy produced or the ancillary services market.
We are revisiting whether to move to a capacity market. We are looking at possibly changing our operating reserve demand curve. We are looking at changes to reliability unit commitment. I think there is a high chance of the reliability unit commitment changes before summer. The odds of the operating reserve demand curve changes are not as high.
Another issue is now to deal with marginal losses, but that is very contentious and down the road, if it happens at all. When it comes to whether that helps or hurts, it depends on what marginal losses you have. I am not sure that it will curtail any renewables that are being built, even though that is the goal of some people pushing it. Another potential change is co-optimization in real time. We will see it eventually, but it is five years down the road.
MS. BARROW: So load growth and potential market reform. Kevin Smith?
MR. SMITH: Coal retirements are another big fundamental change in ERCOT. There have been announcements in the last few months that about 5,000 megawatts of coal is being retired, almost all of it early this year with the rest at the end of the summer. That will have a material effect on capacity reserves. The expectation is that we will see more scarcity pricing this summer than we have in the past.
MR. HELTON: The overall picture is a mixture of retirements, load growth and the renewables that have been built. There are more than 20,000 megawatts of wind farms alone that are non-dispatchable resources. You couple that with the low cost of natural gas and strong supply and it is making it harder for base-load plants to cover their fixed costs. I think the story in ERCOT is a mixture of regulatory, load growth, increased renewables, retirements and low price of natural gas.
MR. SMITH: You said a mouthful.
MS. BARROW: Vistra Energy is retiring, as Kevin said, almost 5,000 megawatts of coal capacity. What does the panel feel is best placed to replace it?
MR. DAHLSTROM: This may sound a little self-serving, since I am a greenfield developer of natural gas peakers in ERCOT, but if you look at the fundamentals we just talked about, we think that the market will require flexible generation with low installed costs and low operating costs to help balance out the renewables. Thus, our view is that the generation to replace coal should be low-cost gas peakers as well as battery storage.
MR. HELTON: I can’t argue with that. Flexibility, flexibility, flexibility is going to be the rule of the game moving forward, especially with the volume of renewables being built. The question is when is the right time to bring gas peakers on line, and that is where I think we get into market design issues with our energy-only market. No one will build unless he can get a rate of return, and there has not been one to date. There has actually been a negative rate of return as is apparent from the retirements and bankruptcies. There are several prominent bankruptcies currently in the Texas market.
I think this summer is going to be when things start to change. We have said that almost every summer, but this year already feels different.
You start to get scarcity pricing once the operating reserve demand curve hits around 2,700 to 3,000 megawatts. When you have 4,000 or 5,000 megawatts of wind on the system over peak, it is nearly impossible to get any kind of scarcity pricing. That is where the problem lies. How to rectify that is one of the things we are going to have to address.
So what happens this summer? If we get through this summer with a 9% reserve margin, which is about where we are going to be when we hit the summer and scarcity is not factored into pricing the way the market expects, then that will be a problem. That is why several of us on the generator side, and some on the load side, are advocating for changes in the operating reserve demand curve to where it will add some additional revenues and change the slope of the curve for the summer.
MS. BARROW: Bob Helton, earlier this month there was actually a record set in ERCOT with an all-time winter peak demand. Did that lead to scarcity pricing? Why or why not?
MR. HELTON: Not really. Scarcity pricing turns on the level of reserves and the way the curve is shaped. When you have high wind, it adds another 4,000 to 5,000 megawatts to the system, and there is no scarcity. That is the problem.
MR. SMITH: The scarcity was limited. There were a few hours of scarcity pricing. In south Texas, you saw several hours of $3,000 prices, and you saw an hour of $1,500 here and $600 there. That kind of scarcity pricing for that duration is not going to cause anyone to build a new power plant. There does not appear to be any interest in base-load generation in ERCOT. So when you ask what kind of generation will replace the retired coal plants, I think wind will continue to be built. I think we will see more wind than anything else.
There is a growing queue for solar. There are about a couple thousand megawatts of solar in the interconnection queue that posted security for the next couple years. There are probably about 13,000 megawatts of solar that are under evaluation. As far as gas-fired generation, until people have confidence that they will see the kind scarcity necessary to get a rate of return, I do not think you will see anything built. Karl Dahlstrom may disagree.
MR. DAHLSTROM: I think there is a clear line of sight to scarcity pricing in ERCOT based on the fundamentals. We had it just yesterday. A weather event triggered it, but this shows how delicate the reserves are. We believe that there is funding to build low-cost gas peakers.
MR. HELTON: The average price last year was somewhere around $26 a megawatt hour. That was the energy price. That is the average price 24/7, and the ancillary service prices were about $1.07. That creates some problems for you. I hope that we do get to scarcity pricing, because we need it. We need the market to be able to function without brownouts and blackouts that can happen with reserve margins at 9%.
MR. SMITH: The fundamentals suggest we should be building, but every time we hit an all-time peak load, we have record-high winds. The knock on wind is supposed to be that it does not blow during peak hours, but when ERCOT has had summer peaks, the wind has been blowing like crazy. We have not seen the kind of scarcity pricing that we would have expected given our load shape.
MS. BARROW: Diving deeper into the effect of wind, let’s bring up the elephant in the room. It is not an elephant, it is a panda. Panda Energy Partners sued ERCOT last year charging that misleading and faulty data in ERCOT’s capacity, demand and reserve reports caused it to invest $2.2 billion in three merchant gas plants that eventually lost money. So talking about wind and renewables, Kevin Smith, how does ERCOT calculate reserve margins given renewables, and is it a rational methodology?
MR. SMITH: Around 2012 through 2014, ERCOT would apply an 8% to 9% capacity factor for wind. Then ERCOT changed to a methodology where it would look at the top 20 hours of peak load seasonally, both summer and winter, look at each wind farm’s output across those peaks and then average that. It would average back to 2009, if the wind farm was operating then, to come up with an average.
I think that is a fairly rational way of applying a capacity factor to wind. It takes into consideration location. The capacity factor for coastal wind is significantly higher across the summer peak than the panhandle west Texas wind, so I think you see around a 50% to 54% summer capacity factor for coastal wind, which is actually quite high.
MR. HELTON: The inland capacity factor in west Texas is about 18%. It took us a long time to get there. I think that is the right way to do it. No forecast is completely accurate.
MS. BARROW: The projected reserve margin for this summer according to ERCOT is 9.3%. ERCOT’s target is 13.75%. What happens if the reserve margin goes even lower? Would there be a potential NERC violation?
MR. HELTON: I prefer to let ERCOT and NERC fight that one out. This summer, we will see whether the commission and the legislature have the intestinal fortitude to let an energy-only market work.
When you get to 9%, you would hope to see scarcity pricing to send the right price signals. There is a standing joke in Texas that we are always just one outage away from a capacity market. Blackouts and brownouts are possible with a 9% reserve margin.
MR. SMITH: To the point about intestinal fortitude, the commission has said publicly that it needs to educate the legislature before the summer about what to expect with an energy-only market.
MR. DAHLSTROM: ERCOT has benefited from an energy-only market for a long time with low power prices. The market was supposed to be one where scarcity price events create an incentive to build more generation. We think there are a lot of Milton Friedman fans in ERCOT that will allow the market incentives to work. The legislature will take some heat from the voters if prices spike. We understand that the PUC is leaning toward letting the market work and continuing with an energy-only market, but you guys are closer to it than I am.
MR. HELTON: All indications point to that.
MR. SMITH: When wind capacity margins were low in 2012, NERC sent ERCOT a letter expressing concern. NERC said it could not order NERC to build capacity or transmission, but NERC is responsible for the reliability of the bulk power system, and it had a responsibility to notify those in charge.
That started a discussion about capacity markets. The customers were not for it. The generators were. Ultimately, the PUC did nothing. The perspective in ERCOT has been that if it is not broken, don’t fix it. When people talk about parts of it being broken, the response is power prices are low and that is what we want.
The customers have enjoyed the benefits of low power prices, and the generators have been challenged by them. It is a political hot potato to move to a capacity market. We probably will not see such a move until something goes seriously wrong. Maybe we start with a capacity obligation on load-serving entities.
We have a bifurcated market where 25% of the load is served by vertically-integrated utilities, municipal utilities and coops, and 75% of the load is competitive under one-to-three-year contracts. The municipal utilities and electric cooperatives have integrated resource plans and plan accordingly. The retail providers have short-term contracts, and it is a challenge for them to go out and make investments in steel.
MS. BARROW: Let’s take this in a different direction. Karl Dahlstrom, Halyard has 100 megawatts of battery storage under development. Can you tell us more about that, and what makes those projects economic?
MR. DAHLSTROM: The projects are not economic today. We believe in flexible generation to help support the intermittent nature of renewables. We believe battery storage will eventually have a strong place in ERCOT. There is no place for it today. We are developing storage to get ahead of the curve and are closely watching the price of energy storage come down until a point, in the next few years, when it will make economic sense. We believe it makes sense to combine storage with the natural gas peakers we plan to build to help balance out VARS to support the grid as well as provide capacity.
MS. BARROW: Does anyone else on the panel have views about the potential for storage in ERCOT?
MR. SMITH: Everybody knows the issue with storage is the economics. It is just a matter of when. The economics will work first in the panhandle and west Texas. In the north zone in 2017, there were about 320 hours of negative prices in the first 11 months of 2017.
MR. HELTON: I have a different view because these guys are thinking about utility-scale storage, and I think the greater opportunity for storage may be behind the customer meter. Batteries will allow a move to virtual transmission. This will also spare the utilities from having to make the investment and then recover it from ratepayers. The ratepayers, or third-party actors, can pay for storage directly.
Let’s do that on the distribution system. You can open up to some degree the distribution system planning and, as a competitor, I can offer up storage units as a solution to transmission constraints. I will sign a PPA with you to fix your problem. Only 20% of the battery is needed for that function. I can take the rest of the battery and bid the storage capacity in the competitive market.
The market is not open for this type of play today. We need to do some design changes to get there. We have to do system mapping down to the distribution system level. We are working currently to do that, but we are talking a few years still before this will be a reality.
MS. BARROW: Let’s talk about the potential for solar in ERCOT. There were 1,800 megawatts of installed solar capacity in ERCOT at the end of 2017, placing Texas seventh among the states. Kevin Smith, how rapidly will the solar market grow?
MR. SMITH: The best opportunity for utility-scale solar is in west Texas. The challenge is transmission. There is lots of cheap land in the west, and there is better irradiance. But the electricity load is in the east, so you need transmission. This makes it a challenge.
Thus, we are seeing smaller-scale solar projects — less than 50 megawatts in size — being built closer to load and being connected at the distribution level. One of the benefits of connecting at the distribution level is it avoids congestion.
MR. HELTON: We are seeing utilities put out requests for proposals for solar of five megawatts in one case and 15 megawatts in another. These are small projects that will connect to distribution lines.
There is an interesting wrinkle in the way we price and pay for transmission in ERCOT. Demand charges are based on your percentage of four coincidental peaks during the summer. If a commercial or industrial customer installs a one-megawatt solar facility with storage, that does not count as generation. It counts as negative load and lowers the four coincidental peaks and, in turn, lowers the demand charges. We tend to see these types of plays mainly in areas served by municipal utilities and electric cooperatives.
MR. DAHLSTROM: I agree. I think the largest challenges are grid congestion and lack of PPAs.
MS. BARROW: How hard is it to get a PPA today in ERCOT for any type of generation?
MR. DAHLSTROM: It depends on who is responsible for the reserve margins. The utilities have benefited from an energy-only market where they had very low prices. They have been signing two-year strips to cover their current peak season and a future peak season and avoiding having to take a long-term view on the market.
They have done well to date with this strategy. It will take one or more significant scarcity price events for utilities to commit to a long-term power purchase agreement and take a view on the future. Until that happens, I do not see a lot of opportunities for PPAs. We will see whether that changes by this coming fall.
MR. HELTON: I agree. There are no long-term PPAs to be had in the current market. The few utilities that feel the need to take a view on future prices would rather own the generating facilities than sign long-term power purchase agreements.
For developers, that means they do better currently to offer to build and transfer projects to utilities than to count on PPAs.
MS. BARROW: How easy or difficult is it to arrange a hedge in ERCOT?
MR. DAHLSTROM: Physical hedges in the form of heat-rate call options and revenue puts are available from financial parties. They are short-term contracts in the five- to 10-year range. Those are readily available. It just comes down to the price. When looking at the price, there are two things to consider: how much will a letter of credit cost to secure your obligations, and how much value are you getting from the hedge?
We believe that it makes sense to secure a heat-rate call option for at least a portion of the output to help attract lower-cost debt.
MR. SMITH: There is a pretty robust market for hedges. Hedges price at a hub. Such a hedge may or may not work for you depending on how far the generating facility is from the hub and what type of resource it is.
MS. BARROW: Several of you have mentioned transmission issues. What has been the effect of CREZ — competitive renewable energy zones — on transmission issues in ERCOT?
MR. SMITH: CREZ has lowered the overall price of wholesale energy because it facilitated the renewables boom in west Texas and the Texas panhandle. It is unclear to what extent it has affected retail prices.
CREZ has also contributed to potential price volatility associated with system deviations. CREZ has helped renewables displace thermal generation in the day-ahead market, which leads to fewer dispatchable resources on line to respond to system deviations, like load forecast error or unit trips, and that leads to more volatility in prices.
MR. HELTON. Before CREZ, everything was bottled up. You used to have negative prices all the time in the west zone. Curtailments due to congestion were common. CREZ reduced, but did not eliminate, both the frequency and amount of negative pricing, but such pricing then spread to the rest of ERCOT.
Now you have a lot more ERCOT-wide negative pricing, albeit not as bad as the $14 to $15 range. Maybe it is $2 to $4.
MR. DAHLSTROM: CREZ has brought low-priced wind to where the load is on the eastern side of the state, and this is putting pressure on base-load plants to cover their fixed costs where they probably did not have such pressure before.
MR. HELTON: The question now becomes how to take care of base-load units in a zero-marginal-cost world. That is where we are headed in ERCOT. We need these units because of low reserve margins, but we have to find a way to make it economic for such plants to remain on the margin. Perhaps the return has to come through ancillary service capacity payments. This is part of the fallout from CREZ.
MS. BARROW: Somewhat related to transmission issues, do you see the potential for any new transmission lines to be built into Mexico?
MR. HELTON: Asynchronously connecting ERCOT to Mexico will not happen. It would probably give the Federal Energy Regulatory Commission jurisdiction over ERCOT. That is considered a sacrilege in Texas. The Baja peninsula is already asynchronously connected to California and Arizona. Connecting Texas to Mexico would give Texas power a path to markets in those two states and give FERC jurisdiction over the Texas market.
While you will not see that, what you may see is switchable units. Mexico has a very good capacity market.
Mexico does not really accept any switchable units in the capacity market currently because Mexico is unsure whether the units will be there when needed, especially given the high potential for emergency events in Texas.
If that can be worked through, and there was a bill in the last legislature to do that, to the point where you can build and offer a switchable unit in the Mexican market, including during an ERCOT emergency, in exchange for a capacity payment, then you will see a lot of new development along the southern border into Mexico with switchable units that can earn a capacity payment in Mexico while still earning energy payments in ERCOT.
MR. SMITH: There are currently three DC-tie lines from Texas into Mexico. All of those are in rate base. I would be surprised to see another one put into rate base.
MR. DAHLSTROM: My understanding is that FERC sent a letter that was a shot across the bow suggesting that a cross-border transmission line would create FERC jurisdiction. I think Texas enjoys having its own grid and retaining the option to secede from the Union.
MR. HELTON: The shot across the bow is leading to a protocol revision that that will allow ERCOT to disconnect the existing DC tie lines preemptively if necessary to prevent Texas from becoming subject to FERC jurisdiction.
MS. BARROW: It sounds like FERC staff and ERCOT staff just need to sit down and have a cup of coffee.
MR. HELTON: There are some conversations on how that might take place and on the potential to get waivers. There are two DC tie lines that go north from Texas into neighboring US states, and they were given exemptions when they were built. The trouble with the ones into Mexico is they don’t have exemptions. There will be discussions about whether they can be given exemptions.