Unpacking the Texas power outage and how to avoid similar in future. Legal expert Andrew Roman breaks down what happened and how to increase reliability.
For the original version of this posting, see here.
For more analysis by Andrew Roman, check out his blog here:
Why am I blogging about electricity in Texas? Because some of my legal career involved electricity supply and demand issues in regulatory hearings, and in advising the Ontario government on restructuring the Province’s electricity system. All electricity utility systems have similar components, albeit in different quantities and configurations.
The Power Failure in Two Words
I can explain the Texas blackout – without playing the blame game – in two words: reliability and cost. There is a trade-off between them. Everyone wants electricity 24/7/365 when they switch the lights on or to heat or air condition their homes. But the cost increases exponentially with increased reliability. A small percentage increase in reliability – say from 99.90% to 99.99% can be many times more costly than a small percentage increase, and even more so to get to 99.999%. At some point, the people who manage the system and the consumers who pay for it implicitly decide that they have enough reliability and don’t want to pay for more – until it proves to be insufficient. Then everyone gets angry and blames everyone else, or on tribal lines, blames renewables or fossil fuels as the “main cause”.
Laws and Government Policies Affect System Reliability
The power system is a complex series of components, but ultimately, they all operate within a framework of laws. The laws implement government policies, influencing how much reliability, and at what cost, the system provides. These laws often grant some discretion in setting consumer rates and deciding the design of electricity generation and transmission, but that discretion is also subject to political control.
Causes of Power Failures
Most of the widespread power failures of the past have occurred in the northeastern part of the USA, up into Canada, from Ontario to the eastern coast. The cause has been transmission lines, brought down by ice storms and fallen trees. The generators were able to keep generating, but there was no way to get the power to consumers.
The causes of the Texas power failure were unusual: (i) a failure of local generation, and (ii) a lack of interconnection to other power grids that could have wheeled electricity into Texas when local generation failed. Because the Texas power system is designed for peak summer heat it was not prepared for this extraordinary cold weather. Its wind generators froze and stopped working; likewise the gas wells and pipelines, preventing gas from getting to the generators.
Every failure of reliability will cause property damage, health issues and, in cold weather, several deaths. (A lot more people die of cold and heat.) The public focus now is whose fault it was, but as the anger subsides over time, the questions will be who should do what about it.
This blog post will first explain the likely causes of what went wrong, without assigning blame, and then suggest potential ways of increasing reliability in future. To make all of this comprehensible, we first need a description of how electricity systems operate. (You can skip this part if you already know or aren’t interested.)
Local extreme weather conditions in North America and Europe have been mitigated for several decades by creating large interconnected transmission grids, including North-South connections between Canadian provinces and nearby US states. This greatly increases reliability by overcoming local weather, and also permits cost reductions through purchasing electricity from generators with lower costs operating in other jurisdictions. My understanding is that the Texas grid does not have these interconnections (except for a small connection to New Mexico) and is therefore almost entirely dependent upon both local generation and local transmission reliability. This isolation is not an engineering necessity but a policy choice. (I am not qualified to comment on the benefits and costs of this policy, which would require a detailed multidisciplinary study.)
Transmission lines require coordination with generation across the entire grid, and also, to decide, in real time, which generator gets dispatched to put its electricity into the grid. The Texas grid operator for 90% of the state’s homes, ERCOT, doesn’t operate in a legal or political vacuum. It is subject to Texas law and policy.
Different jurisdictions give their grid operators different degrees of control over whose electricity generation goes into the grid hourly, as well as what types of generation is to be added to or subtracted from the grid, at which locations. In some jurisdictions the grid operator is politically independent, for the most part, and in others its operation is mostly determined by politicians, directly or indirectly. To the extent that there may be federal or state policies governing CO2 emissions or other policy objectives not based on the lowest cost bidding, there are government policies influencing the reliability-cost trade-off.
The grid must be operated to keep electricity demand and supply in balance at all times, failing which, the frequency falls below the operating range for thousands of electric motors. That would damage water pumps, elevators and buildings and other essential facilities. The grid is designed to shut down when the frequency falls out of range, to avoid damaging electrical equipment. This is probably what happened in Texas.
Power grids require various proportions of three generic types of generation:
- baseload, which generates all the time (typically, hydro, nuclear, coal or gas);
- intermediate, which is added above the baseload as the system approaches peak (typically, gas, small hydro, coal or biomass (usually wood chips or other plant-based material, solar and wind); and
- peak load, which runs for relatively short times at system peaks.
Thus, electricity is not a single product, but has a different generation mix almost every hour, with different levels of cost and reliability.
Nuclear cannot be ramped up or down, it’s either off or on, so it is only useful for baseload. In states like Texas, where there is not much hydro, the baseload is provided by coal and gas. Wind and solar are intermittent, so they cannot respond quickly to changes in demand. That’s why they cannot be used for peaking supply. Nevertheless, in most jurisdictions, including Texas, wind and solar are given preference over other sources of supply. This preference for intermittency adds to cost while reducing reliability.
With the usual interconnected grids, the generation mix includes more than just local generators. If the grid operator can purchase electricity from another connected jurisdiction at a lower cost it will usually do so, shutting off local generators.
Another important factor is what the grid operator pays generators to supply. When there is a widespread power outage someone needs to ramp up power in the grid, just as your car needs something to start the engine. Similarly, and of greater importance, reliability requires redundancy in both transmission lines and generation. Both starting power and redundancy add costs that raise consumer prices.
In what is called an “energy market,” the grid operator only pays for energy actually put into the grid. (Although there are some favoured exceptions for wind and solar power which may be given take-or-pay contracts, meaning they are paid the same whether or not their generation is dispatched into the grid).
The alternative is a “capacity market,” in which some generators are paid to provide the stand-by capacity to add to the grid if necessary, and are paid more when their capacity is ramped up to put electricity into the grid. Texas has an energy market, not a capacity market, again, as a policy choice. Although critics tend to criticize politicians currently in power, I would expect that these policy choices were made, and modified, over several decades.
The Generation Mix:
The percentage of coal, nuclear, gas, solar and wind generation available for customers is a strong determinant of reliability and cost. Entire books have been written about the optimum mix for different regional climates and resource availability. This short blog post is not the place for that analysis. However, I can state in general, that apart from extraordinary equipment failures such as the gas plants in Texas, wind and solar generation is intermittent. When the wind doesn’t blow enough or the sun doesn’t shine enough, the lights go out without fossil fuel backup. Wind and solar cannot be counted on to generate sufficient reliable electricity to supply more than a small proportion of the demand, especially in grids with a high degree of peakiness. Until there is new and better generation technology invented or developed, fossil fuels and nuclear will have to provide the bulk of the generation. (While hydro is theoretically an alternative, most of the large-scale hydro locations in North America have already been developed, or would require massive power dams that would flood large tracts of land.)
Ways of Increasing Reliability in Texas
1. Creating a Capacity Market
If Texas is willing to pay some generators to maintain reliable backup capacity, as many other North American jurisdictions have done, that can enhance reliability – at a cost. If some of these generators are outside of Texas they would have to be given long-term contracts to justify their capital investment and ongoing operating costs, even if their electricity is only required intermittently. Also, given the failure of all forms of generation to operate in, what is for Texas, extreme cold, the new capacity market entrants will have to engineer their systems for operation at those temperatures. If that generation is to be by natural gas, it will require extensive and costly reconstruction of the existing gas wells and pipelines to avoid freezing, suggesting that the generation will have to be provided by other fuels.
2. Building Transmission Lines for Interconnection
Transmission lines require rights-of-way over land, which may require some expropriation. New lines are often strongly opposed, blockaded or repeatedly sued by local residents and anti-development activists, who can also attack the pipelines. Transmission lines will take several years to obtain construction permits and then to construct.
Interconnection with grids in other states will require their permission and contractual arrangements. Once these transmission lines are built, Texas could contract with particular generators in other states to supply Texas with baseload, intermediate or peaking power, either on demand or on a standby capacity basis. Of course, the other states with which Texas may wish to interconnect are unlikely to have a transmission line already constructed that ends at the Texas state border, just waiting for Texas to interconnect. These other states will also have to build new transmission lines to connect with those being built by Texas. Texas would have to make it financially worthwhile for these other states to make this expenditure.
The fact that so many other North American states and provinces have established multijurisdictional grid connections demonstrates that they can provide both greater reliability and lower electricity prices through broader competition among generators. For example, Ontario’s nuclear generation and Hydro-Québec’s water generation can provide baseload to the interconnected grids they are in.
3. Requiring Generation to Re-engineer for Colder Weather
If this record-breaking cold is treated as a freak event unlikely to be repeated for perhaps another century, Texas will not want to spend the enormous sums necessary to rebuild its entire generation system to be able to operate in these temperatures. (Of course a once-in-a-century event does not mean it will only happen 100 years from now; it could happen again next year.) Rebuilding for such a level of cold would require modifying every type of generation, and even natural gas suppliers that are not generators, to be able to operate at below freezing temperatures. The cost would be massive and take many years to complete, requiring substantially higher electricity rates, for a relatively modest reliability increase.
Several decades ago, Québec endured a terrible ice storm which cut transmission to large parts of the province, requiring costly and time-consuming repairs. This resulted in a lengthy and costly inquiry to consider the costs and benefits of burying transmission lines. It concluded that this would be too expensive for the probable benefits. I would not be surprised to see a similar cost-benefit conclusion in Texas.
4. Individual Actions
Texas homes are not insulated for heat retention as they are in the north. It would not make economic sense to rebuild these homes with insulation designed for colder climates.
It might be useful for homeowners to install propane or diesel generators to generate enough electricity for lighting and refrigeration, and perhaps some modest level of baseboard electric heating during a prolonged power outage. Gas fireplaces or stoves for heating could operate on propane stored in tanks on the property, even if natural gas is unavailable due to frozen gas wells and pipelines. Again, these measures would be costly, but less than the damage from burst water pipes and the discomfort and risk of cold weather “camping” in the living room.
The Insurance Industry Response
Many Texas homes and businesses will have insurance claims from water frozen in pipes, toilet bowls and other fixtures, spoiled food, hotel costs, frostbite and other medical problems, etc. The insurance industry is unlikely to have included the cost of paying these unforeseen claims in their premiums. Many insurers will suffer unanticipated financial losses. It will be interesting to see how many insurance companies will be able to pay the claims.
Looking ahead, Texas insurance premiums are likely to be increased to recover these losses and as protection for the future, or to be listed as exclusions in their policies. Competition between insurance companies may require a reduction in premiums for homes and businesses that have alternative methods of heating and electrifying their premises.
The collapse of generation in Texas last week affected both fossil and non-fossil generation alike. Whether the grid had operated on all wind or all natural gas, the result would have been equally bad without engineering for these unusually cold temperatures. That does not mean the generation mix is currently ideal or provides the optimum trade-off between reliability and cost, including costs imposed by various government policies.
As with every trade-off that produces widespread societal impact, the ultimate praise or blame will go to the politicians. They create the laws and policies that determine the balance between reliability and cost.