February 27, 2021 (OPINION): “How economic theories influence energy policy“, Austin American Statesman
“Simply put, more of us need to think about the broader relationship between energy and economic theory” Read more …
Texas House of Representatives Joint Hearing
Joint Hearing: State Affairs and Energy Resources
February 25, 2021
Summary and notes by Carey W King (http://careyking.com/)
|Chair:||Rep. Chris Paddie|
|Vice Chair:||Rep. Ana Hernandez|
|Members:||Rep. Joe Deshotel|
|Rep. Sam Harless|
|Rep. Donna Howard|
|Rep. Todd Hunter|
|Rep. Phil King|
|Rep. Eddie Lucio III|
|Rep. Will Metcalf|
|Rep. Richard Peña Raymond|
|Rep. Matt Shaheen|
|Rep. Shelby Slawson|
|Rep. John T. Smithee|
|Chair:||Rep. Craig Goldman|
|Vice Chair:||Rep. Abel Herrero|
|Members:||Rep. Rafael Anchia|
|Rep. Tom Craddick|
|Rep. Drew Darby|
|Rep. Jake Ellzey|
|Rep. Charlie Geren|
|Rep. Tracy O. King|
|Rep. Ben Leman|
|Rep. Oscar Longoria|
|Rep. Ron Reynolds|
Mr. Morgan stated: “I was a big proponent of this [ERCOT energy only] market, and my faith has been shaken.” We’re risking this economic engine of this country [implied as Texas].
Main problems affecting the blackout:
We produced 35% of total ERCOT generation during February 15-17, and we generally have 18% market share.
Coal Power: They mine coal for Oak Grove and they had issues of their local coal delivery (via local rail). Also, they don’t normally cover their coal piles that got wet or froze up. So the coal became gummy and blocked the process of pulverizing coal to input into boilers.
Natural Gas: Their natural gas fleet was 90% available at NG plants, but they could not get the gas at the pressures that were needed to operate.
Nuclear: They came within 3 minutes of having the Comanche Peak (nuclear power plant) automatically trip off due to low grid frequency.
Fossil fuels vs. renewables: Mr. Morgan brought up how zero marginal costs of renewables stresses the market structure. We’re going to need to investigate a change in market structure. This is a policy level change in decision to figure out how to deal with this “We are at a crossroads”. We need to continue to figure out how to evolve the competitive markets. “There is a policy angle here”. “There is a big debate in this country on the future of fossil fuels”. We need to have revenues for companies to invest in capacity.
The lack of an enforced weatherization standard is not the entire problem that occurred on Feb. 14 and 15, 2021.
When asked why his company does not have reserve gas on site, or if more NG on site would have prevented the ERCOT blackouts the week of Feb. 14 (e.g., with 3 or 5 days of storage)? Mr. Morgan: The system has been so reliable (99.99%) that it doesn’t make economic sense to store more NG. We do have some offsite storage and deliver it through price. More storage would have helped, but hard to say how much. Some of us have dual fuel capability, and we did lean on that a bit (and was limited by diesel trucks coming in).
None of their customers will be exposed to high energy bills.
They have 8 GW of generation capacity and it performed at 80% of its capacity. They produced almost 2X the amount of electricity as they produced in normal winter times at the beginning of the month.
There were 3 main reasons for lower generation from NRG fleet:
There are many factors that impacted their power plant performance, and they are:
Gutierrez said there is now the chance for the Legislature to define “reliability” and “resilience” for the electric grid. The state agencies can find the best way to achieve these goals. “Give industry the guideposts”. There is no standard for weatherization. There is the opportunity to have that conversation.
When asked why his company does not have reserve gas on site, or if more NG on site would have prevented the ERCOT blackouts the week of Feb. 14 (e.g., with 3 or 5 days of storage)? Gutierrez: We need to think about the quantity of gas needed to store many days of NG on site. You’d have to store NG as LNG because the quantity is so large. This is cost prohibitive.
It is important that we look at extreme weather in a different light, given what we have seen. I believe climate change is real, and as a company we [NRG] are targeting to reduce carbon emissions 50% by 2035 and net zero by 2050 in accordance with United Nations targets to 1.5 degree C rise.
He stated it seems ambiguous as to whether ERCOT is a governmental organization, and this is in decision at Texas Supreme Court. He wondered if there is any reason we can’t make ERCOT a government agency under PUC or other agency?
To each person testified, Representative Hunter asked pointed questions about who (persons or entities) were at fault. He asked questions such as: Who’s at fault? I don’t want to hear about “systems”? I want people to know who screwed up? I want details.
Answers to these questions were generally that many parties are responsible: 1) natural gas suppliers, 2) power generators, 3) Texas Public Utility Commission, 4) ERCOT, and 4) transmission and distribution utilities
Representative Hunter emphasized the need for better communication by all parties (power generators, utilities, ERCOT, Public Utility Commission, Governor’s office).
This “feast or famine” 9000 $/MWh creates volatility that doesn’t really incentivize financing.
He emphasized the need to revisit how to incentivize investment in ERCOT outside of the 9000 $/MWh offer cap of the Operating Reserve Demand Curve.
Rep. Anchia expressed that the current ERCOT market design seems to have the perverse outcome that we can’t incentivize new baseload power to be installed unless we have extreme weather that juices up costs.
Rep. Anchia: Does the market still have perverse outcomes and increases volatility that we need to address?
Mr. Morgan responded to Rep. Anchia that much of their equipment (they buy?) is designed to operate at 10 deg. (F?), and they weatherized $10 million was supplemented to try to get us closer to 0 degree capability, and these were more short term changes
Mr. Morgan responded to Rep. Anchia stating that the ERCOT market has have increased volatility relative to years ago due to the 9000 $/MWh offer cap and steepness of the Operating Reserve Demand Curve.
Rep. Anchia stated we need to create a market that avoids too much volatility and scares investors. A recent Wall Street Journal survey suggests that the ERCOT wholesale matket design has caused Texas customers to pay 10s of billions (?) more than other areas.
Asked for ideas in 1 week: Last session I introduced HB 34 on emergency communication system. It didn’t get a hearing to pass. Hopefully now it will. He asked those giving testimony to tell the legislature what can be done to make this better, and report back in 1 week.
On subsidies: During testimony of Mr. Clark and Mr. Hemmeline, Representative Raymond stated that we have to try to move away from politics as last week’s events are so tragic. He realized that people sometimes don’t like discussing tax credits and subsidies etc. But he said we should talk about all the tax credits that oil and gas gets (state and federal), let’s put it all on the table.
I’m interested in the market issues. We sort of know how things work but not all of the details. What about a lower price cap with a capacity market? Do you know what you would really want? What structure meets the needs of reliability, affordability, and lower risk (even it not financial risk, but risk to livelihoods)? How do we restructure? Since NG prices are not capped, this sounds like it doesn’t work.
Morgan: We could consider a capacity market. This does keep some old capacity on the system that only runs 1 week or so per year. But that might be nice in the situation such as last week. There are things we can do on the “all energy” market. We can increase the amount of reserves that are required and then drop the price cap. Let’s remember, ORDC is a construct, but the pure marginal cost system wasn’t working either. We could add a specific product for fuel security. We can’t ignore the idea of re-regulation, but there is a reason we left that idea and I think it has served us well, so I’m not suggesting we go back. I like idea of some reduction in price cap with an increase in reserves, but I’m not advocating yet.
Hill: We had about 2000 MW going offline at about the same time.
We started losing gas supply broadly by late morning or early afternoon Feb. 15th. These were back on Wednesday (Feb. 16th). We had some firm capacity and producers cut us off since they couldn’t get the gas into pipelines. 40% of their deliveries were cut from suppliers on Wed. Feb. 16. There were times when 40% were offline due to loss of gas supply.
Freezing at out power plants: We started in 2011 on best practices with ERCOT and TRE, and we do this every year. We still lost 2 of 12 plants due to freezing conditions.
Hemmeline: Solar output generated more than ERCOT planned (starting Feb. 15?) given expected generation. Several solar sites were set offline, many (200+) inverters had to be manually reset, and one solar photovoltaic plant was shut down due to icing on transmission lines.
Jeffrey Clark: We represent, wind, solar, batteries, investors, and even NG customers. This issue has shaken my confidence in a reliable system, and we have to fix this going forward. We also had a failure of transmission. Freezing fog is what affected some wind turbines. We’re looking at ways to shed ice from wind blades, limit ice accumulation, and other methods to keep wind turbine blades from icing.
Jeffrey Clark (on renewable industry obligations in ERCOT market): It is important that most renewables are funded via Power Purchase Agreements. There is a perception that wind and solar can just show up and produce whenever they want to, and don’t feel repercussions. These generators have been affected by the grid blackout. When do we change any ERCOT rules? What are the mechanisms? You can be assured that wind and solar members of my organization did not benefit from the high prices and events we experienced.
Jeffrey Clark (on wind turbine performance): Mechanically, we’re aware of weather issues, but we are assessing if this is a “new normal” and how do we make the power available when needed. Freezing fog in San Angelo/Sweetwater area had a particularly rough impact. We’re looking into that. Shaking ice off of blades is possible but tricky (e.g., if you shake off ice above the nacelle onto the nacelle). We’re talking with OEMs on the options for de-icing on wind turbine blades.
Cold weather by itself is not an issue in itself. There are heating elements to keep lubricants functioning properly. The typical [wind turbine] machine can go to -15 deg. C (some to -30 deg C). So it is not the cold itself that is the problem.
Question: Is snow generally a problem? Clark: Every region had trouble with power generation, even in SPP and West Virginia. It is the icing of the wind turbine blades that seems to be the major problem here.
Clark was asked: If we are going to have a discussion of how we get more capacity, are we going to have to consider how subsidies for wind and solar affect generation investment? Clark: I think we need to look to all subsidies. Coal has been subsidized since before we were a country.
Nye: We did plan to do rolling outages, and we did this at first late Feb. 14 and early Feb. 15. At 2:02 am Feb 15 we experienced 59.302 Hz and this was a critical event. By 4:43 we know multiple gen was offline. By 8:03 state was shedding 16 GW. Why was 2:02 am important? The importance of under frequency relays are required to have an automatic response to drop load faster than people can respond. They are the last measure to prevent total grid blackout. They are set to trip at 59.3 Hz. There were within 0.002 Hz of getting to this limit. We did have some under frequency relays trip, we don’t know why some did or didn’t. The situation was so fast we were worried about customers being in dark for say 1 month if ERCOT went completely dark.
What about inability to do rolling blackouts? Nye: We’ve got to communicate better. We should have contacted people to let them know that rolling blackouts of 15-30 minutes were not going to happen anymore.
We are required to have 25% of our load on feeders that have special devices called under frequency relays. To get to 25% of load, this is 40% of their feeders that have these relays, and these circuits stay online 100% of the time to act in emergencies and these feeders STAY ON ALL THE TIME because we need that automatic control capability. The distribution of these is to GEOGRAPHICALLY DIVERSE to buffer the system (he can show the maps of these) and they are generally heavier when you have more load. There are more in urban areas since there is more load there. They are distributed based on where the load is NOT rich vs. poor areas. He gives a campfire analogy to say these relays are like
2nd bucket of critical circuits make up 8-9% of his feeders that serve hospitals, critical care facilities, airports, 911 call centers. Now this leaves only 51% of feeders that are left as even possible to rotate or turn off.
My Nye knew that equity would be a question. He called COO Debbie Dennis and Chief Counsel to make sure they were not cutting off people in an inequitable manner. We (Oncor) were as equitable as we could be.
Total premises out of power in Oncor regions:
Other metrics they used, # underfrequency relays, # loadshed feeders, % customers interrupted and all of these are pretty equal metrics. I see no institutional equity.
Speaker Craddick: Are O&G producing wells not on critical infrastructure with you, Oncor, but they are with Centerpoint? Thus you shed load associated with O&G production, so do you know how that impacted the rest of the grid?
Nye: On gas supply issue, after 2011, there was a push that critical gas facilities were on a critical list to provide gas. There is a process to be designated as such, and to your point (Craddick), I don’t know how power plants get gas or who is delivering. The NG producers have to tell me (Oncor) if they are delivering gas to generators, otherwise I don’t know. We had an identified 35 specific facilities that were delivering gas and we never cut those off from electricity. During the event, PUC Chairperson Walker would call with names of folks indicating desire to turn on the power so they could deliver gas, and we did this, and by the end of the event we added 168 more locations to the critical list and we never again turned them off. We need better coordination of NG producers and generators with T&D.
Question from Representative Hunter: Here comes March, another storm is coming, what do you think you can do better?
Nye: Communication for sure. We received more inbound telephone calls in 2 days than we did in all of 2020. Our phone system (with AT&T) crashed and our customers could not get through to us. Make sure we have adequate technology for people to contact us. Make sure we are talking to legislature better. Better communication with O&G suppliers to not cut them off. Work on water districts and know much they can be rotated. Biggest issue with water utilities is they don’t like to be surprised (so tell them the plan). We had about 1.33 M out on their system at one time, but on Wed. when they were allowed to turn on more, all but ~140K were able to come online and this was mostly due to ice storm damage around Waco/Temple/Killeen area. SO this tells me Oncor was generally ready to come online when they could (T&D was generally working). This last ~ 100K customers was actually pretty tough to turn back on. We can use data analytics to more quickly know what parts of our system are broken. We can encourage conservation (as mentioned by Mercado as Centerpoint has done). For our industrial customers, who are mostly at transmission voltage, and they are at backbone of system and we didn’t want to turn on transmission level lines. Over 50% of industrial customers voluntarily came down, so we can use that idea going forward.
Rep. Herndandez: [To Mercado]: We had a visit how you determine who to turn off. There are “buckets” of types of loads.
Mercado: We look at the circuits in thinking of load shed, we don’t look at neighborhoods. The load shed is pre-programmed. Heavier demand for load shed where there are larger loads. We got 59 orders for load shed from ERCOT, and each order changed the buckets of load shed changed. We don’t have the infrastructure in our telecommunications to do more coordinated load shed using smart meters? We have to measure and trigger load shed in milliseconds. The only way we can actually measure load is via SCADA system. We need to figure out a way to get the meter speed of communication to our feeders faster.
Question from Representative Hunter: Here comes March, another storm is coming, what do you think you can do better?
Mercado: Not so much to add. Communications is key. I realized how important it was to be an expert communicating with the public in real time. The problem with my message on Day 1 (with NPR, Houston stations, etc. that I then learned to correct) was I could not tell them “how long” the grid will be down. I only know what was happening in our company, so I was limited in knowledge, and we need more information to make people comfortable. We need a better software approach for better load shedding. We planned for losing 2-3 units, but now we have a situation when 50% of generation can go out, and we need a better solution. We need to be more transparent and dynamic in our load shed while still protecting critical loads and underfrequency relay feeders.
Mercado: Due to turning circuits on and off many times, we had to replace 1000 pad-mounted transformers after this event (due to their cycling broke them)
Magness: Texas suffered. Our goals is to prevent this type of event again. We are like air traffic control for the ERCOT grid (90% of customers, 75% of land mass). We can only stay at 59.4 Hz for 9 minutes before generators automatically trip offline from low frequency relays. Figuring out how to actually do rotating blackouts is critical.
Mr. Magness was repeatedly criticized for ERCOT not having a more effective and direct communication that the grid situation was so severe that rolling blackouts were not going to be able to be in place and that some regions of the grid were going to be in a long-term power outage while some regions were going to stay on.
[note from this note taker: Apologies for not as many notes on Bill Magness testimony]
Walker: By statute ERCOT is accountable to the PUC. PUC can apply penalties per PURA or ERCOT protocols. Can impost administrative penalties, and they max at $25K/day/violation. Largest to date is $2M. She noted that the FERC 2011 report had 26 recommendations but none to PUC [[Note: but PUC would have been involved in overseeing and agreeing to any new mandates]]. Quanta Energy Services ensured that 2011 legislation calling for a report on preparedness, that a report was filed, and it had 9 recommendations for PUC and PUC implemented all of them.
I don’t think PUC has explicit authority to require winterization of generation because we’ve not been given clear authority. The 2011 TX bill did have this requirement but it was removed before passed.
Each year ERCOT is to file a report on weatherization plans, and it filed that report in January 2021 stating compliance of the power plants.
Regarding the PUC order on Feb. 15: They heard that how their models we rerunning, in load shed that you will go to 9000 $/MWh cap. But ERCOT was getting reserves and bringing on customers, and so the reserve was sitting there and sending the signal that not all generation was needed. So we told them to move to 9000 $/MWh. On Feb. 21 we had another emergency meeting, we urged all REPs to delay their invoicing to customers, and not disconnect them, and give the option for a long-term payment plan. This only applies to IOUs, not municipal utilities or co-ops.
Here are some things for PUC and community to work on:
On Wed. Feb. 10, we knew a storm was coming in, but not 100% sure how bad. I had a generator call me (Feb. 10 or 11) and said 3 of their units we going to be curtailed and needed help getting NG. The RRC quickly passed an order to allow electric generation to have high priority getting gas. In the moment, she was discussing coordination in the NG supply chain (well heads, processing plants) and worked with Oncor to get these facilities power turned on. Typically their call center is not open on the holidays or weekends, so she did keep it open on the weekend. We “worked” with the industrial customers telling them they need to have them come offline. That Thursday and Friday she was asking generators and TDUs checking to see that PUC could do for them. What did you know and what did you know … really I started actions on Wednesday (before the storm?)
Representative Lucio criticized the Chairwoman Walker and the Public Utility Commission as largely lacking any communication to the public on the severity of the electricity crises and what they can do to prepare even though he believes the PUC (not ERCOT) had the responsibility for this communication.
[note from this note taker: Apologies for not as many notes on DeAnn Walker testimony]