Electricity providers hard pressed to keep up with growing tech-heavy demand
Q. After two decades of practically no growth in U.S. power consumption, grid planners are now expecting annual growth over the next decade of about 1.3 percent. What changed?
I think if you were to have asked that same question last year, most people would have said, “No, there's still no growth.” So, this is new. What's interesting is that when you go back to the ’80s and the ’90s, we saw a lot of power growth.
We saw 3 percent annual growth in the ’80s, 2 percent in the ’90s. If you think about households, we were buying lots of different appliances like fridges. In Texas, you know, we sometimes have two or three fridges. We were also buying TVs.
But then moving into the last 20 years, it's just been very slow. This is due to two main trends: efficiency and offshoring. I think everyone's seen all these energy-efficiency stickers that are on everything that they buy. Everything we were buying became more efficient. The other big thing was a lot of offshoring of manufacturing, whether it was steel, aluminum or other products. We offshored a lot of heavy, energy-intensive manufacturing.
More recently, the surprise has been AI [artificial intelligence]; there's a need for more electricity to power AI. There's also a big push for onshoring of manufacturing. And so, that's essentially what's changed.
Q. What about data centers?
People talk about data centers, and they tie them into AI. Data centers are not new; they have been used since the advent of the internet to process all the data that we use for the internet, whether it's for storage or whether it's just to go to a website. But these centers also run AI applications.
Q. So, is it the combination of AI and the reliance on data centers that's going to contribute to the demand for more power in the future?
That's correct. You can literally see every day or two a news release about a new data center being built. They're being built all over the U.S., all over the world, and they do consume power. Now, we don't know exactly how much; there are no reliable statistics. The number that comes out is somewhere between 2 percent and 4 percent of U.S. power.
Q. You mentioned nearshoring and industrial policy bringing some plants back to the U.S., for example, computer chip manufacturing. Obviously, those are not built yet, but when they're up and running, is that going to be a significant factor in the growth of future power consumption?
I think so. Manufacturing generally is energy intensive. The specific one I'm following is chip manufacturing, because chip manufacturing is a 24/7 process that uses a lot of power. Battery manufacturing also uses a lot of power. Tracking the amount of projects that are out there, there's about $500 billion in announced projects, although there will likely be less in the final number.
I think that [many of] these projects will be built over the next 10 years. So, you should see a slow and steady increase in power consumption.
Q. You mentioned batteries. Are those batteries for electric vehicles? What kind of batteries are those?
They're for electric vehicles, but they're also for grid storage. Previously, there was just no way to store power. You produce it and you consume it. Batteries have been a way to store that power but for only a limited amount of time—four to eight hours. In the future, as we use more wind and solar, we're going to need more battery storage.
When I'm trying to divide up [power] demand between the two—electric-vehicle manufacturing and batteries—I lean a little bit more toward electric vehicles [consuming most power].
Q: Why has Texas power demand grown in the past 20 years even before the recent national growth spurt?
I think if you were to ask an expert about power demand for the past two decades, they’d say, “It's not going anywhere in the U.S.” And then they would say, “Well, except for Texas.” In Texas, we've had a lot of net migration. Population increases do contribute to power demand, but it's not the primary driver.
We've also had some specific manufacturing projects in Texas—new chemical plants, new LNG [liquefied natural gas] export facilities. In the oil fields, there's a push to electrify, essentially moving from using generators to actually pulling power from the grid.
And in Texas, we've had a lot of crypto mining. Part of the reason to do crypto mining in Texas is our deregulated market. Power prices are cheaper than in other parts of the U.S. We also saw the price of electric power in Texas decline in the past decade, before rising in recent years.
That's different than in the U.S., where we've seen increasing power prices. That's because in Texas, as natural gas prices have fallen, we've seen lower power prices. So, all of these different factors have led to an increase in power consumption in Texas over the past decade. I'd put that number at about 2 percent annual growth.
Q: A lot of our growth in terms of power capacity has come from renewables—solar and wind energy. Are renewables still going to be the answer? What's the power supply going to look like for the grid?
When we're looking at wind and solar, they make up about 15 percent of the U.S grid. In some locations, such as Texas, it's around 30 percent. And we will see more wind and solar. Anybody that's looking to build a data center is also looking to sign a power purchase agreement for wind and solar power.
I think the big question going forward is that with wind and solar, we use the word intermittent—their availability depends on whether the sun is shining or the wind is blowing. As we retire a lot of older dispatchable power sources—coal-powered plants, for example—the big question is will we have enough power on a very hot or cold day?
And so, the big thing everyone is looking at is what will be the advances in battery storage, because battery storage is very important for wind and solar. My expectations are we're going to see more wind and solar. But we really need to watch and make sure that we have the battery storage out there, that we have the dispatchable energy generation out there.
Q. These redundancies that you talk about—dispatchable power—is that going to raise the price of electricity?
We've seen power prices move up a bit in the U.S. and Texas, but the price you pay as a consumer isn't exactly the price that's in these tradable [electricity] markets. When you get a bill, you'll see that there is a price for power. And then there's a fee for transmission and distribution—the cost of actually delivering electricity from the power plant to your home. And there are also taxes.
You should expect that we're going to have to spend more on transmission and distribution. Our infrastructure was built 40 to 60 years ago. It has to be replaced. There's a lot of capital spending that goes into that. So, you should see your bills continue to increase.
We've seen power prices increase for retail consumers. I expect it [price of electricity] to slowly increase, but nothing too crazy.
Q. We know that reliability has become an issue with the dependency on renewables. Would connecting the Texas power grid to other states’ grids improve reliability and maybe reduce some of the vulnerability we experienced, such as during the 2021 winter storm blackout?
When we think about the 2021 freeze, there has been a big debate. With an interconnected grid, you have the ability to import power from other parts of the U.S. However, that depends on if they are having the same issues we are having in Texas, as was the case during that winter storm.
On the other side, there is a benefit of having our own grid in that we can make our own decisions. As long as we're making the right decisions, then we benefit. The other big thing with the interconnected grid is that if there is an issue in one part of the grid, it could cascade to other parts of the grid.
Q. What are the benefits and costs of investing in the power grid, and who should pay?
Over the next 10 years, 20 years, 30 years, it seems very clear that we're going to see more power consumption. The question is how much? I don't think anybody has a number out there. I've seen estimates for 1 percent all the way up to 5 percent growth per year.
I think a reasonable number is generally 1 to 2 percent. I think the big question, of course, is who pays for all of these investments?
When business investments are made, there are economic benefits. These businesses provide jobs; they provide capital spending. Now, each one is different. Some of these businesses, like a data center, will have a lot of capital spending but not as many jobs.
But if you move to chip manufacturing, if you look at battery production, there are a lot of jobs, a lot of spending, a lot of bringing in people from out of state. There are a lot of different changes that must occur. People are trying to weigh the benefits of that versus the costs.
