1/26 I’ve seen a lot of confusion over what’s going on in Texas regarding the blackouts we’re having. Unlike normal winter storm blackouts, the vast majority of what’s causing power outages is a lack of CAPACITY rather than damaged wires. The ability to generate power has

2/26 basically collapsed (only about 35% of our normal maximum capacity is online), so they have to turn off power to blocks of houses to keep power running in any form at all. The PLAN was to have short (15-45 minute) outages that moved from neighborhood to neighborhood. That

3/26 never happened.
The current situation in many places is that the lines companies (e.g., Oncor) have been unable to cycle the rolling blackouts the way they have planned to due to what I've seen described as "frozen lines," which I interpret to mean ice on the substation

4/26 disconnect blades. But even before that happened, rolling blackouts weren't going as planned: the criteria for taking grid segments up and down were taking far longer to satisfy than anticipated, which is why it wasn't just 15-45 minutes off, as originally expected.

5/26 Instead, some places (even where the switches were working) were seeing more like 8 hours off, 30 minutes on. I must imagine that they're working to clear the ice, but I doubt they'll be able to run the rapid rolling blackouts that they can in the summer until the lines

6/26 are above freezing. So that's the rolling blackout situation.
But even if that were working as planned, the big issue is capacity. If you take all the power generation that we could possibly have online in the middle of summer, it's around 125 GW. Currently (as of 6 PM),

7/26 Texas's grid has just about 47 GW of power online. Yesterday at noon, this was 50 GW. The midnight before that, it was 68 GW. What's been happening is that the cold weather has been incapacitating various generation plants. There's some frozen wind (and this is getting a

8/26 lot of press, although in practice, extra windy conditions on the coast means that wind is currently overperforming projections rather than underperforming -- it's also not very significant, as total wind projections are about 4 GW, and we're currently producing about 5

9/26 GW of wind). Solar is producing as expected, but that's only 2 GW, and it's only for about 11 hours. There are two major issues that I know of that are hampering production in a way that can't be fixed: natural gas wells and pipelines are shutting down due to freezing

10/26 conditions (mid-day yesterday, about 30 GW of natural gas production had been knocked offline, and that number is probably higher today); and cooling reservoirs for standby plants had frozen, making it impossible to bring them online, which is impacting natural gas, coal,

11/26 and nuclear plants. So that's why the capacity numbers are low and keep dropping.
Capacity has been kind of wandering around between 45GW and 50GW today. If I had to guess based on previous days and the colder temperatures, demand would probably be around 70 GW if

12/26 everyone were still attached, so I'd expect that a little under half of Texas is disconnected from the grid right now. Understanding what's happening here requires a step back to understand how electrical generation capacity is managed in Texas. Since deregulation in

13/26 1999, Texas has relied on market forces to manage its grid capacity, and it looks a lot like other commodities trading (although it's necessarily a lot more real-time). Wholesale energy providers (e.g., Entex) bid on a real-time spot market to sell energy to retail

14/26 service providers (e.g., TXU), who buy at those spot rates to satisfy the demand of their retail customers. (Neither company owns the wires, which are operated by regulated lines companies like Oncor.) Retail providers set their own rates with their customers, which are

15/26 usually fixed rates. Wholesale prices generally float somewhere around 2 to 5 cents per kWh, and retail providers operate on the difference between that price and the 8 to 20 cents per kWh or so they charge their customers.
When demand gets higher, it drives the spot

16/26 price of generation up, making it financially attractive for wholesale producers to bring new supplies online. These are generally allowed to go as high as necessary, although there is a statutory cap of $9.00/kWh. (An interesting side note: because most power can't be

17/26 brought up and down instantaneously, you can end up with enough excess production that spot prices drop below zero -- meaning generators are paying to put power on the grid, and retail providers are being paid to take it. This happens with some frequency in the middle of

18/26 the night during the summer.)
What happened overnight is that the rolling blackouts eased the demand, and spot prices dropped from around $5.00/kWh to $1.20/kWh (with some regional fluctuations). So there's still capacity that could be brought online, but because it would

19/26 be only for a day or two, and because it costs so much to take a plant online and then offline again, $1.20/kWh isn't enough to justify providing that capacity. This morning, ERCOT ruled that the cost structure had been artificially perturbed by the intentional

20/26 disconnections, and no longer reflected actual demand. The decision is that, as long as disconnections are in effect, the spot price market must exchange all power at the statutory maximum of $9.00/kWh. The theory is that this will make it financially viable for this idle

21/26 capacity to come online for the relatively short period of time that it is needed. This *might* be working, as we've seen a slight uptick, as I mentioned above, but it isn't yet keeping pace with ERCOT's projections. (It's also not clear to me whether the $9.00/kWh has

22/26 taken effect yet, since the spot prices I'm seeing on the ERCOT website are still floating).
So that might work. One of the side effects of this situation -- and it was true even before ERCOT made this decision -- is that you have retail providers who have been having to

23/26 buy energy at $2-$5 per kWh and sell it to their customers at $0.08-$0.20/kWh for a couple of days now. They're going to take a hit of about $1,000 to $2,000 per customer, and a lot of them probably don't have enough funds to deal with such a shortfall. I suspect many of

24/26 them will go into bankruptcy this month. In fact, one forward-looking company was paying customers $100 if they left and chose a different provider before the 14th. (The key point being: if you knew enough about how the system worked, you could see this disaster coming).

25/26 In the best case, we'll see higher energy rates all around for years, if not forever, as companies make up this shortfall and hedge for the next such event. If things go really badly, we might see a complete collapse of the deregulated market, and turmoil as we figure out

26/26 how to replace it.

PS 1/2: I'm seeing a lot of quote tweets of this thread knocking the free-market aspects of a deregulated market. And there might be some valid points here! But I have homework for those folks: if you're going to use this thread to highlight a failure of electricity deregulation,

PS 2/2: at least game out in your quote-tweet thread how this situation would have looked in a traditional regulated utility. Watching ERCOT's graphs, I think we need to assume that all capacity that can be online is online now.