Jan

3

1 GW = about 1 nuclear power plant

Five-year US load growth forecast [for power] surges to 128 GW

U.S. electricity demand is forecast to increase 15.8% by 2029, according to a new report from Grid Strategies. Six regions of the country are driving the growth

The report's load growth estimates are based on annual planning reports submitted to the Federal Energy Regulatory Commission by electric balancing authorities and updated with additional data from utilities and planning regions.

Consider this question from another list member: What would happen to the grid if Silicon Valley companies found technologies a decade hence that would provide similar server services with less electric power?

Answer 1a: The utility could face stranded assets, including underutilized power plants, transmission lines, substations, and distribution systems. Remember that most utility investments are 30-year-plus assets, are leveraged, and have their levelized costs, plus margin, firmly embedded in utility bills.

Answer 1b: How would investors hedge their position if they considered building a $1 billion gas turbine in states that deregulated their power plants?

David Lillienfeld responds:

Basically, you're going to see a mismatch between where demand from data centers are and where there's generating capacity. You can build demand a lot faster than you can build supply though, and if you get efficiency on the demand end, you have overpriced supply relative to the ability of the region to pay for the power generated. At some point, someone's going broke.

But here's the curious question–how much economic activity can be attributed to a server building? It's like a parking lot for data–nothing more than that. And if there isn't that much taxable revenues that it's generating, what's the appeal to governments–the risk for the electorate of holding the bag at the end of the day is non-trivial it would seem. So what's the appeal?

Henry Gifford writes:

In other industries power saving strategies are known but not adopted yet, but could be at any time without much warning.

Take cars for example. I heard that in about three years the whole car industry in Europe is going to switch from the now-standard twelve volt electrical system (actually about fifteen volts) to seventy-six volts. One advantage of voltage about five times higher is that electric motors can be about one-fifth the size they are now. This includes the starter motor used to start the engine, the motors used to raise and lower the windows, the heating/cooling system’s fan motor, the engine’s cooling fan motor, the alternator (which is basically a motor wired to work in reverse), and others. As motors are made in large part from Copper and rare earth magnets, smaller motors can save a lot of money. Another advantage is that the wires, usually made from Copper, can be about one-fifth the cross sectional area. Another advantage is that some things typically driven directly off the engine by rubber belts, such as the air conditioning system’s refrigerant compressor and the power steering pump can instead be driven by a small electric motor that can be located anyplace the designer chooses, instead of now having to be located in line with a belt wrapped around part of the engine. Shrinking all these things would make the car lighter, saving fuel. Voltages higher than seventy-six would of course extend these advantages in an ever-diminishing way, but be more capable of going through a person’s skin, thus seventy-six is thought to be the best choice.

The problem is, I heard the three year prediction about twenty years ago, and a few times after that, but not more recently than about ten years ago. So maybe it won’t happen anytime soon, or ever, but the technology and advantages are well known and waiting to be used.

Similar changes could gradually or suddenly drop the power used by data centers.


Comments

Name

Email

Website

Speak your mind

Archives

Resources & Links

Search