A Free-Market Energy Blog

Texas Power Outages: A Preliminary Analysis (Cold snap brings failure–isolated ERCOT an issue)

By Michael Giberson -- February 4, 2011

[Editor note: Dr. Giberson is an instructor and research associate at the Center for Energy Commerce at Texas Tech University’s Rawls College of Business. He blogs on energy economics and other topics at Knowledge Problem.]

On Wednesday morning, The Electric Reliability Council of Texas (ERCOT), operating the power grid for much of the state, called upon local distribution companies to cut power to blocks of consumers on a rotating basis.

The rolling outages produced hardship for millions, and even isolated instances of severe harm. Consumers and policymakers are dissecting what went wrong and what should be done about it. The following is a preliminary analysis based on public data and news reports. A subsequent post will present more details once more complete information becomes available.

In brief, extreme cold weather pushed power demand to very high winter levels. At the same time, fifty of the state’s power plants were offline due to the effects of the cold, and several others were undergoing planned maintenance. The combination of very high demand and reduced supply left the ERCOT grid perilously short of reserves. Rolling consumer outages were employed to protect the system from failing completely.

Some wondered whether wind power was at fault, but wind contributed about seven percent of ERCOT’s power during the emergency – about the same as this time last year.

No power system is immune to hazards. But policy decisions that increase the likelihood of hazards or multiply the resulting damages ought to be given careful reconsideration. In this case, the choice by Texas policymakers to keep ERCOT isolated from surrounding power systems prevented power companies within ERCOT from accessing excess power capacity elsewhere in the state and in neighboring states. Other policy issues also are raised by the emergency, but few solutions are likely to be as cost-effective and technically simple to implement as linking ERCOT to its neighbors.

Cold Snap

ERCOT reported that severe weather led to the loss of 50 generation units amounting to 7,000 MW of capacity on Wednesday morning. From news accounts it looks like a few large coal plants failed after water pipes burst. Some natural gas generators found insufficient fuel supplies due to heavy demand for natural gas. Other natural gas generators found their access to fuel curtailed by state rules that give priorities to other customer classes when supplies run short. In addition, a larger than usual amount of generation was off-line for scheduled maintenance – one estimate put this quantity at about 12,000 MW.

Demand for power was sharply higher on Wednesday morning compared to earlier in the week, reaching over 53,000 MW between 9 AM and 10 AM. The rolling outages eliminated about 3,000 MW of demand during that period, so the true demand for power was nearer 56,000 MW. By comparison, the same hour on Monday saw demand of just 33,500 MW. ERCOT has seen demand at this level in the winter before –last winter the system handled demand of 57,000 MW without incident. The high demand was only a problem because so much generation was offline.

Rolling power outages are a way to limit power demand during emergencies in an attempt to prevent an uncontrolled cascading blackout. While the rolling outages were controlled, they still impose heavy costs on consumers. Hospitals and other priorities locations are protected from rolling outages, but schools are not. Several San Antonio-area schools losing power resorted to busing students to school buildings that continued to have power. Traffic accidents in Austin were attributed to traffic signals being out due to the rolling outages.

ERCOT’s Electrical Isolation

Texas has pursued a policy of isolation for the ERCOT power grid so as to keep the state’s largest utilities subject primarily to state, rather than federal, regulation. Two minor links connect ERCOT and utilities in Oklahoma, but they are of little commercial significance. A small interconnection with Mexico was activated to send power into Texas for a few hours, but cold conditions in Mexico required it to suspend the assistance.

The policy of isolation is questioned from time to time, but remains popular with the industry and many state policymakers. While the policy has important benefits, the costs are particularly visible at times of system stress.

In the Southeastern corner of the state, Beaumont was not experiencing outages. The local electric utility, Entergy Texas Inc., is not connected to the ERCOT power grid. If Entergy Texas had excess power capacity on Wednesday morning, they could have sold it east into Louisiana or elsewhere as far as Florida or even Maine. However, even thought the utility borders against ERCOT near Houston, no power could flow to help out the rest of the state. Nearby CenterPoint Energy had to blackout an average of about 330,000 customers at a time during the emergency.

Amarillo’s Xcel Energy reported operations were running smoothly despite temperatures falling below zero overnight in the region. If the utility had excess power, however, none of it would have been able to reach ERCOT. Like Entergy Texas, Xcel and other utilities in the Panhandle and South Plains are connected into the Eastern Interconnection, which stretches to the Atlantic coast in the east and to Canada in the north. (On Thursday Xcel called upon consumers in the Panhandle to conserve power and natural gas, as heavy demand for gas was temporarily making the fuel harder to obtain.)

El Paso Electric Co. in the western tip of Texas is not connected to the ERCOT grid, but it also implemented rolling outages Wednesday morning after two of its power plants suffered partial shut downs due to the cold. In the case of El Paso, connected by power lines running throughout the western United States, while it worked to bring the generators back online it could seek out supplies from neighboring states of Arizona and New Mexico, or from as far away as Washington or British Columbia.

Other Single-State Power Systems

Two other regional power grids are contained wholly within a single state – the New York ISO and the California ISO. The California ISO relies on imports for about a quarter of its annual energy consumption. The New York ISO similarly imports and exports high quantities of power. Only Texas pursues a policy of isolation.

The inter-system trade in power surely lowers the overall cost of electricity for consumers in New York and California. And, despite some high profile exceptions like the August 2003 blackout that spread from Ohio to New York, these interconnections tend to improve the reliability of power systems, too. More relevant for the current discussion, when emergency conditions arise, neighboring power systems can cooperate to help solve the problem.

How Did Wind Power Do?

A few rumors bounced around the radio waves and Internet forums on Wednesday linking the rolling blackouts to ERCOT’s wind capacity, one rumor even claiming that wind power had dropped to zero. The rumors were false. News reports indicate that some wind turbines were out of service due to the cold, but the problems appeared not to be widespread. ERCOT spokesperson Dottie Roark said that wind power plants from between 3,500 to 4,000 MW of power during the worst parts of the emergency, about normal for this time of year.

Wind power may have had an indirect effect. The significant investment in wind power capacity may have discouraged some added investment in natural gas or coal powered plants. But given conditions Wednesday mornings, a few additional new thermal plants may not have made much difference. Some existing natural gas generating plants saw their access to fuel curtailed by rules giving higher priorities to other customer categories when supplies become short, other plants were confronted by low pressure in gas pipelines. Additional natural gas plants may have just added to the number of plants without access to fuel. A few of the new coal plants built in recent years were among the plants that were forced out of service yesterday by the cold, key contributors to the problem.

The system needed all of the power it could get. Had more thermal plants been built, at least some of them would have been in service and helpful. Outages would have been moderated a little. Wind generated power was used and useful, but couldn’t be dialed up to produce more during a time of need. Wind power was neither the cause of the problem, nor of any special value in reaching a solution.

Infrastructure Interdependencies a Problem

Emergency actions by ERCOT prevented the generation outages from causing the entire system from failing. ERCOT’s emergency operations seemed to work okay, given the difficult situation. The primary problem on Wednesday was a lack of generator preparation for the extreme cold and the hazards that the weather brought with it. Given those problems, ERCOT probably did as well as it could.

Potential policy problems mostly lay elsewhere. In some cases there were interdependences between the power system and other infrastructure systems that magnified the costs of the rolling consumer outages. For example, some of the controlled outages idled natural gas pipeline compressor stations, reducing pipeline pressure and hampering the ability of natural gas generation plants to get fuel they needed. Other power plants found their fuel supplies curtailed under natural gas priority rules that were last updated in the early 1970s. While the linkages between the electric power system and the natural gas pipeline system can’t be severed, actions can be taken to make each system a little more robust to problems with the other.

Another example, the power outage left many traffic signals out, so a power system problem added to an already difficult roadway congestion problem. Reports from Austin attributed at least one traffic accident to the loss of power to a traffic signal. Battery backup systems are widely available for traffic signals, and the City of Austin was already planning to begin installing the systems later this year. Other cities should take note.

Perhaps most or all cell phone towers have battery backup power, helping to assure continued lines of communication when power goes out. But one cable company served by El Paso Electric reported intermittent loss of service after its battery backup system was drained from repeated loss of power from the grid. While the loss of mid-day movies and soap operas may not be a serious public policy concern, it isn’t too hard to imagine conditions under which timely dissemination of information about health issues could be critical. Companies in the communications business should consider whether further steps are necessary to make their communication systems robust to failures in supporting infrastructure systems.

ERCOT Interconnecting?

It is entirely likely that, had power companies in ERCOT been linked more substantially to other utilities in the state and utilities in neighboring states, Wednesday’s rolling blackouts could have been completely averted. This conclusion is obviously not enough of an argument by itself to justify reforming the state’s policy of isolating ERCOT. But it may be sufficient to rekindle discussions about the costs and benefits of ERCOT’s electrical isolation.

Connections from ERCOT into to the southeast corner of the state would be valuable, in case of another emergency. It seems a shame for excess power capacity in various corners of the state to be unavailable at times of stress.

But possibly it is the case that Texas can, as the saying goes, have its cake and eat it too. The Tres Amigas project has proposed building a high-tech transmission link that would simultaneously link up the Eastern, Western, and ERCOT interconnections. At full capacity, the project would be capable of supplying up to 5,000 MW of power to the ERCOT grid – more than sufficient to cover Wednesday’s shortfall (assuming sufficient in-state transmission to carry the power).

State regulators and many power industry players in the state are reluctant to support the project, citing a desire to protect the current regulatory status quo. But the Federal Energy Regulatory Commission too has acted to protect the status quo in the past, and has signaled a willingness to continue to protect it should ERCOT link up to Tres Amigas. Texas policymakers should explore the opportunities available.


Accidents happen, and no power system will be resistant to all challenges. It would be too expensive to build a power system that would never fail. Yet, when failures come, we ought to do our best to learn from them.

No doubt coal and gas-fueled generators across the state are reexamining their readiness for extreme coal weather. I suspect we could survive another severe storm as early as next week if one comes about. By this time next year, ERCOT and the industry will be well prepared to weather another storm like we’ve had this week. Our problems can make us better, as the late Julian Simon emphasized.

At the same time, we have to consider the ability to respond to the next surprise. Not another storm like we’ve had this year, but something new that Mother Nature will surely toss our way. Linking up with neighboring power systems would give ERCOT additional resources to draw upon during an emergency.

When policymakers in Austin next reconsider ERCOT’s current electrical isolation, the Tres Amigas plan ought to get the hearing it deserves.


Related Readings

ERCOT, “Grid situation improving but conservation still needed,” February 2, 2011.

Jared Fleisher, “ERCOT’s Jurisdictional Status: A Legal History and Contemporary Appraisal,” Texas Journal of Oil, Gas, and Energy Law, Vol. 3, Issue 1 (2008).

Kate Galbraith, “The Rolling Chain of Events Behind Texas Blackouts,” The Texas Tribune, February 3, 2011.

Kate Galbraith, “Head of Texas Grid Discusses Blackouts,” The Texas Tribune, February 3, 2011.

Julian Simon, The Ultimate Resource 2. Princeton University Press, 1996.


  1. Cold snap brings rolling power outages to Texas; is ERCOT policy of isolation at fault? « Knowledge Problem  

    […] [Note: This item was originally posted at MasterResource as: "Texas Power Outages: A Preliminary Analysis (Cold snap brings failure–isolated ERCOT an issue)"] […]


  2. JavalinaTex  

    Thanks for the information. I am not 100% sure you can lay the outages on anyone else other than the generators. Not even the Texas ERCOT “de-reg”. We had outages in the early 1990’s when we had vertically integrated power. Then, the newly operational South Texas Nuclear Project (STNP) – a huge cause of rate shock – had to be taken off line because they hadn’t prepared it for freezing temps. What amazes me is that the new coal plants aren’t ready. I had grown up in the north, and while there was a coal pile freeze crisis in the early 1980’s that was for -30 temps. Hopefully, they will be ready for the next cold front.

    That said, what concerns me the most, and is already happening, is many gas gathering facilities and pipeline compressor stations are moving to electric power for their primary mover (compressor drivers) due to air emissions reasons. Further, you can bet older stations have let their self generated power go for running fin fan coolers and oil and water pumps. In fact, these may even be more vulnerable as they tend to be on a local grid and not on the high voltage high rate segments.

    If GHG emissions go through and midstream companies are faced with getting forced into a PSD permit on CO2 (when critia pollutants wouldn’t be a problem with a gas driver) they will go with electric drives to avoid PSD and let the power companies deal with the GHG issues.

    There is relatively little (no) thought state or nationally on this issue with the gas pipeline system. I asked Chairman Wellinghoff about this issue at a public forum. He didn’t seem to have even thought about the issue and thought perhaps the EPA might need to consider it. I was rather amazed as he is an environmental lawyer and the “cascading interdependency” was a long time concern of both the extreme environmental movement and right wingers who distrust large organizations.

    Regardless of what regulatory and organizational structures we have in the energy industry, the energy infrastructure on the oil and gas side must have a much higher energy priority or exemption from curtailments. Also, perhaps, we need to start ensuring that large scale customers like hospitals and schools pay for the premium service, they infact receive. I know as homeowners, we definitely pay a big premium for our “no notice” services.


  3. John T A Miller  

    excellent explanation, thanks.

    FYI, my understanding is that 5 GW is just the initial phase of Tres Amigas and that it could eventually expand to 30 GW.



  4. Nobody At All  

    Pleasant to see Master Resource highlight the 7,000 MW of dependable offline fossil fuel generation.


  5. Greg Rehmke  

    Very helpful analysis. The concluding comment that next year EROT and generators will be ready sounds good, but a lot can happen in a year both positive and negative. Companies can work to address problems with pipes to reduce the likelihood of outages. But also various state and federal legislation and regulations can be passed that alter incentives, add uncertainty, and limit flexibility.

    After the black-outs though, every commercial and private electricity user knows that the electricity grid cannot be relied upon during cold weather. They must balance the reality of what they just experienced with promises that government regulators and monopoly providers will try to prevent it from happening again.

    I would expect thousands of firms and families to research and invest in backup generators, as many did in preparation for Y2K.

    Any Texas company that uses computers will have lost a lot of money with rolling blackouts unless they had already invested in individual battery power and backup generators.

    Do state regulation allow electricity providers to significantly raise prices during deep cold or high heat periods as a way of reducing demand? Such excess revenue could have been used to compensate firms with backup generators to drop off the grid and generate their own power for a few hours or longer.

    Could Texas power generation be further deregulated, allowing hundreds or thousands more firms with cogeneration capacity to offer electricity to homes and firms near their facilities?


  6. Pierre de Rochemont  

    I think this article misses the main point that EPA regulators are preventing Americans from adequately supplying ourselves with sufficient energy resources. In the name of “sustainability” Austin Energy has shut down its coal-fired power plants and “replaced” them with power generation systems that simply are not economically sustainable.

    Regardless of the administration, the Department of Energy, created under the President Carter to resolve the “oil crisis”, has made it more difficult to consider, let alone achieve, U.S. energy self-sufficiency.

    In the mid-1990’s, the Clinton Administration declared our nation’s (and the world’s) largest clean coal reserve to be a “national park”, providing James Riady’s Lippo Group, (a global mercantilist and Clinton campaign contributor), a monopoly on clean coal.

    Similarly, in 2008, then candidate Barack Obama, promised to bankrupt clean coal producers by levying excessive taxes and regulations upon the industry. (http://www.infowars.com/obama-agenda-to-bankrupt-power-plants-triggers-blackouts/).

    Likewise, then Kansas’ governor, Kathleen Sebelius, (now HHS Secretary), killed two of that state’s coal-fired plants, frustrating new energy production, to protect us from “global warming”. (http://www.nytimes.com/2008/03/22/us/22kansas.html?ex=1363924800&en=08316361230b6a8b&ei=5124&partner=permalink&exprod=permalink)

    And, if that is not enough, just yesterday Al Gore declared the record cold we are experiencing is caused by “global warming”. This declaration stands in stark contrast to reputable, though seldom covered, climatologists predicting significant global cooling at precisely this time based on their observations of solar magnetic cycles.

    Skeptics can validate the link between solar magnetic cycles and terrestrial temperature by reading “Science Has Spoken: Global Warming Is a Myth”, by Arthur B. Robinson and Zachary W. Robinson, with the Oregon Institute of Science and Medicine, published as an OpEd essay in the Wall Street Journal, Thursday, December 4, 1997, and comparing their scientific data with historical through contemporaneous data found at: http://hockeyschtick.blogspot.com/2010/02/solar-update-ap-index-of-geomagnetic.html.

    Is our federal government serving the public good, or laying siege to the nation?

    I found it frightening that the only solution proposed herein is to have Texas, this nation’s largest energy producing state, risk being subject to those very federal regulations.

    Should we submit our public wellbeing to political forces that have either sold out our public interest to global mercantilists and seemingly embraced United Nation’s plans for “global de-industrialization” proposed under “Agenda 21”, or should we act in our own better interest?

    (Note: Agenda 21 further advocates mass depopulation in the name of saving the environment–the fact that flora needs carbon dioxide to thrive and humans need thriving flora to breathe seems to be lost on these global technocrats.)

    Clearly, Texas can do much better than that!! A little more common sense, reality-based planning with a true commitment to our public welfare and the complete avoidance of Malthusian federal regulation will quickly resolve these problems.


  7. Paul Lindsey  

    JavalinaTex – I’m interested in the info about the SNTP having problems with low temps. This would be a very big deal for a nuc plant, because it would impact the ability to perform core decay heat removal after a shutdown/scram. Please post any appropriate links. Thank you.


  8. Inter-area Oscillations in Power Systems: A Nonlinear and Nonstationary Perspective (Power Electronics and Power Systems) | BUKU PDF  

    […] “Texas Power Outages: A Preliminary Analysis (Cold snap brings failure – isolated ERCOT … (masterresource.org) […]


  9. Robust Control in Power Systems (Power Electronics and Power Systems) | BUKU PDF  

    […] “Texas Power Outages: A Preliminary Analysis (Cold snap brings failure – isolated ERCOT … (masterresource.org) […]


  10. JavalinaTex  

    Paul Lindsey… You do understand the event I was talking about happened almost 20 years ago if not longer?

    I would guess the issue was long since resolved given the 90% plus online performance. As I recall it was a precautionary shut down but it was pretty amazing they weren’t prepared for the cold weather (but there was a lot else the utility wasn’t prepared for in that snap). There was never an issue regarding cooling the shut down reactor.

    I would have to go back and try to pull up paid links in from the Chronicle or some other Texas media. But these are often like searching for a needle in a haystack.


  11. Paul Lindsey  

    I understand fully that this was 20 years ago. I’ve been unable to find any reference to cold temperature problems at STNP, even on the anti-nuclear power websites, who I would expect to highlight any problem, even a precautionary shutdown. Unfortunately, googling anything with the word “freezing” in the criteria is generating a huge list of current events. I just thought that you had the reference handy.


  12. JavalinaTex  

    Paul, here are the links. The freeze situation and rolling blackouts were in December 1989. A lot of water under the bridge since then.

    This article covers the STNP issue in depth.


    here is another one that goes into the larger system wide issues that HL&P had. The new Limestone Unit Lignite plant also had failures as did most of all of HL&P’s plants. But given STNP and Limestone both had just been commissioned and had failures in the cold snap, would seem to indicate that a new plant – at least in Texas – will have some issues.

    I think it is important to understand, despite the headlines of the articles, that STNP being down only contributed and exacerbated the 1989 rolling blackouts.


  13. George C. Loehr  

    This article is a very good illustration of why economists should never, ever be permitted to make policy decisions in technical areas. The author doesn’t seem to understand the physics of electric power systems. He has no appreciation of what “isolation” means in an electrical sense, nor that “isolated” synchronous interconnections can exchange power over asynchronous HVDC ties. He seems to have no knowledge that there are HVDC interconnections between ERCOT and the Eastern Interconnection, or that power can be (and is) exchanged between them. Further, he doesn’t understand that the proposed Tres Amigas is nothing more than a complex of several HVDC ties! I suggest that he (or anyone wanting to understand this better) take a look at my article on smaller synchronous interconnections and HVDC ties at http://www.elucem.com/pdfs/takemygrid.pdf.


  14. Trey  

    Nice breakdown of what happened.

    Our power was out for 5+ hours Wed. morning. I had to return home to unfreeze an outside faucet (successfully) and plug up some vent holes under our pier and beam house. (Others in the neighborhood were reporting broken pipes.) Fortunately nothing busted.

    Interconnecting ERCOT’s grid to the outside sounds like a good idea. When I mentioned this to one person I know who works at the Capitol, they scoffed and said that this would never happen. Who knows, the legislature is in session. Now’s the time to get the message across, while it’s still in their memory. If it could be spun that this would save the state billions, that might help. 😉

    A few questions: Texas has ~ 100,000 MW power capacity (http://ftp.eia.doe.gov/state/state_energy_profiles.cfm?sid=TX). We needed 55,000 MW. What happened to the other 45,000 MW?? OK, so a few 500 MW plants were completely down due to frozen pipes, some others were partially down (low gas pressure). Some were down for scheduled maintence. Were too many down for scheduled maintence for this time of year (winter)? How do they decide when? Seems like if this had been better managed we could have had excess capacity, but there’s a lot I don’t know. I’m just trying to get the numbers to add up and help/sources would be appreciated.

    Regarding wind: I suppose 3500-4000 MW is good for wind. It has done a lot worse (see http://www.robertbryce.com/node/374 ). But nameplate capacity for Texas is >9000 MW. I guess wind met expectations (30% average over the year would be considered nominal), but those are pretty LOW expectations. If wind has excelled, it could have met the demand.


  15. Steve Koch  

    Whatever we do, it is imperative to keep it out of the hands of the feds. There should be an independent audit to see what went wrong. Preferably by some EEs who know what they are doing. With all due respect to the author, an economist is not well trained to understand electrical distribution technology.


  16. Can Tomatoes Take Any More Global Warming?  

    […] winter is strongly correlated with cold and flu. Winter can also cause or contribute to power outages, travel disruptions and delays, traffic accidents, and injuries from slipping on […]


  17. As politicians play blame game over power outages, local consortium is already implementing solutions  

    […] from America’s two other main “grids”, the Eastern and Western Interconnections (here is his […]


  18. Committee of Chief Risk Officers seeking wholesale power “passport” for ISOs & RTOs | The Energy Fix  

    […] Council of Texas, or ERCOT — had access to such a  passport, it might have averted temporary power blackouts during 2011. That’s when high demand and dormant wind farms left the Lone Star State without adequate […]


  19. Can Texas go 100% renewable? | Energy Matters  

    […] Combined wind and solar generation show considerable day-to-day fluctuations but no strong seasonal pattern. Demand, however, is significantly higher in the summer, with peak demand usually occurring in August (because the data are daily averages hourly fluctuations in demand, which can reach 40MW in the summer months, are not visible). An exception to this trend was the demand spike in February 2011, which coincided with the descent of an Arctic air mass into Texas that caught ERCOT with its pants down . […]


Leave a Reply