Blog post

Texas: lessons learned from a Wild West electricity market

Energy Blog, 26 February 2021

The unexpected winter storm caused millions to be without power in Texas.

Initial noise and commotion were around blaming renewables (as usual and expected). Going forward, will this be enough pressure for politicians to ensure functional infrastructure in Texas? Only they can tell.

Jérôme Guillet, Managing Director of Green Giraffe, discusses the lessons learned.

In Texas it is actually possible, as an individual consumer, to pay your electricity at a price linked to the spot market:

Astronomical bills face customers who opted for floating-rate contracts tied to wholesale prices in the state’s freewheeling electric market. In Burleson, a suburb of Fort Worth, Valerie Williams has been charged more than $6,000 by her electricity retailer Griddy to power her 1,400 sq ft home over the past few days.” (Source: FT – probably behind paywall)

We can discuss the wisdom of doing so as an individual, when you have limited control over your demand, but the episode vividly demonstrates some of the mechanisms that are required to manage a grid using only short-term prices, and what they mean for the actual cost of electricity.

As we all know, electricity is hard to store and demand and supply must match (or be very close) at all times. Other than for a few large industrial users, demand is largely inflexible (people just switch lights or appliances on) so the balancing must be done by the supply side. In most circumstances, that means using the “merit order” where plants bidding the cheapest prices to provide power are called first, until the most expensive plant needed to satisfy demand is reached, which sets the price for the market at that time. As long as you have sufficient generation capacity available, this works relatively well, but one should already note that the last generator available is likely to be very expensive – by design. Its owners will know that it is only likely to be required very rarely, and thus it will set a very high price – high enough that during those very rare episodes when it actually produces it can make enough money to cover being available at all times. That explains why periods of intense demand (in Texas, usually in the summer) can lead to price spikes – that’s the way markets are designed to work.

This time we had an unusual event, not because demand was very high, but because supply was suddenly much lower than expected. Multiple plants did not function due to freezing – either freezing of their own installations (including some wind turbines, but most of the down capacity was gas-fired and included nuclear tranches), or freezing of the gas network supposed to be providing their fuel. With supply abruptly constrained, the only way to balance the system was to bring online all available capacity (thus ensuring a price spike) and then forcing demand reduction. That can be voluntary (for instance an industrial user having access to fixed price electricity or its own power supplies stopping its normal activities and instead selling back the power at a huge premium) or involuntary – in the form of more or less controlled blackouts. With a lot of capacity down for a long time, Texas has both – very high prices for an unusually long time and rolling blackouts.

This demonstrated that there really is limited price sensitivity on the demand side, and that it is hard to ensure enough spare capacity for rare events on the basis of spot market prices only. Most Texas power plants did not have the “winterization” equipment that would have allowed them to cope with the icy conditions because it was expensive and not seen as profitable to make such an investment. This suggests that there is a role for the market regulator to force generators to have additional equipment and capacity beyond what is expected to be profitable. The Texan grid operator ERCOT, where the “R” stands for “reliability”, did not live up to its name. It is worth noting here that ERCOT is cut off from the rest of the US grid in order to avoid federal regulations on the sector which are more constraining – that also prevented any imports of power from other parts of the USA…

The second lesson is to note that the yearly electricity bill of those hapless consumers who chose the market indexation has been multiplied by something like 3-4 with just a few days of consumption at the peak prices (2-3% of consumption at prices >100x times higher than “usual”). This made the average price of power over the year – or even over several years – substantially higher than the expected rate and the available fixed rate plans. That brings us to the difficulty of comparing prices between power generators when one has actual fixed costs and fixed prices (like renewables or nukes) whereas others have most of their cost of generation coming from the purchase of fuel (gas-fired plants, coal plants and fuels/diesel units). The price of electricity from the latter is an estimate, based on scenarios for the price of the fuel. Most price forecasts for these fuels look like gentle slopes – very stable prices, maybe trending up or down depending on expected overall macro-economic conditions and power market trends (power demand, plants being retired and built, etc). As we know, and have seen repeatedly in the past, oil and gas prices look nothing like that, with massive peaks and troughs and a lot of volatility. But the fully certain prices of renewable generators are compared to future gas prices (that are nothing but wild-ass guesses) that generally are on the low side. And whenever gas prices are low (which of course happens), renewables are presented as uncompetitive – even if all it takes is a few days of peak prices to completely change the picture in terms of medium-term average prices!

Renewables, in that context, should be seen as an insurance product – you get the certainty of stable prices, and you avoid the peak prices that could ruin you. This alone would justify a permanent premium over volatile spot prices. But in fact, we now have renewables that are cheap enough to be competitive even against “normal” spot prices – before externalities like carbon emissions and pollution are taken into account, and before the burden of peak prices are added back to the real prices of gas-fired power.

P.S. despite a few well-publicized frozen turbines, the wind and solar sectors performed more or less as expected during the crisis and were definitely not a material cause of the supply collapse.