E82：瓦特问题？

Transcript: Rob Rowe (0:00) Welcome to our Research@Citi podcast. I'm Rob Rowe, the current actingGlobalHead ofResearch. And with me on the podcast today is TonyYuen. Two years ago, we thought that the power grid was under enormous strain because of risingdemand andsupply, both on the data-center side and renewablesside. How has that situationchanged since then? Now we have even more demand due to data centers. We have even more [See our related note:Must C: Overcoming Gridlock 2.0] TonyYuen(0:34) So the power grid certainly is really important because our lives really right now revolve aroundelectricity. But then the power-supply side tend to come from places that are really far away. And But at the same time, power demand—as you talked about, the data-center side hasbeen risingvery quickly. Over the last couple of years, the estimates for data-center power demand have just But then thereis also thebroadernarrativeand megatrend going forward aswell, which also needthe gridto expand. So it's not just within the U.S., but also globally, because this issue is actuallyexacerbated by what happened in the conflict in the Middle East, because some of the countriesright now are rethinking the whole notion of energy security. And these countries are reallythinkingabout energy sourcesfor power generation that want to be domestic. And also you can see the So the IEA just published a report earlier this year, basically summed it up really nicely. It says thatroughly 2,500 gigawatts of renewables, storage, and large demand projects, such as data centers,are stalled because of the grid problem. So to give asense of scale, a city with a million peoplegenerally uses one to two gigawatts of power. We're talking about 2,500 gigawatts of renewablestorage and large demand projects like data centers are stalled. So what that means is that we are Rob Rowe (2:35) And beyond thosepower-supply challenges, what are other constraints do you think you would TonyYuen(2:41) So I think what happens is that right now, given that expanding/improving the grid proved to berather difficult because ofregulatory and labor constraints and others,that's why, on the one hand,there's a lot of demand for electrical equipment just to improve the grid to the extent possible, but But then if you just step back, why is it so difficult to expand a grid? Well, if you look at theregulatory framework, it is basically necessary because you want to take care of the stakeholders,the people who are living there, and then environmental issues—you certainly don't want the And then also the other side is that, for example, in the U.S., the labor issue is really big. So the BLS,Bureau of Labor Statistics, Department of Labor,essentially it's saying that there are shortagesoflabor, professional trade labor by roughly about 220,000 workers in critical industries likeconstruction, HVAC, electrical, and many of them basically you require to build data centers, but But then these skilled tradespeople,they don't suddenly come out and can call upon them,andyou cannot train themand then get them ready within a month or what-not.And basically BLSexpect that to be a shortage well intothe 2030s.And so then not only is the labor being anobstacle But thenthere’ssomething interesting about the kind of demand that data centers draw from interms of electricity. Because it's not just that the demand for electricity is big. It's just that they'rehighly volatile. So let's say a training modediscul suddenly needs to ramp up. This actually It's almost like the demand side of renewables. Because if youthink about wind generation, solargeneration, if there's no wind at the cloud, then basically generations just come off. And so whatthat means, the grid also needs to accommodate that not just volume aspect, but also volatility of Rob Rowe (5:22) That's interesting. And so given the rapid growth of data centers, and I guess we would say demand,what's the current landscape for critical power equipment? And what's your outlook for its TonyYuen(5:35) Sothere's actually definitely significant opportunities. Ifyou look attheU.S. data-center electrical-equipment market alone, it's projected to grow from roughly about $20 billion in 2025 to roughlyabout $65 billion by 2030. And then to get a more broad scope, the global transformer market willlikely expand from roughly about $60 billion in 2024 to $93 billion by 2030, somewhere aroundthere. This surge in demandfor these electrical-equipment,by fuel,by data centers, utilities, So the grid is not just power lines—in fact,there's a lot of overlapin terms of what the data centerswant in terms of power equipment as well as the renewablesside.So definitely there'sa lot ofdemand for large power transformers,a lot of generated step-by transformers,high-voltage switchgear,medium-voltage switch gear,and so that's why recently our colleague talked about solid- And that's why the market is getting quite big. But also because of such