Media & news

It’s early Sunday morning, and a couple of issues are on my mind. The electric utility industry has several challenges and will certainly face great change related to the enormous amount of electricity data centers are projected to need in the future.

Data center growth is led by huge data consumers and the leaders of the Artificial Intelligence (AI) digital revolution, including Microsoft, Apple, Meta (or Facebook), Amazon, and Google. However, other companies pursuing AI technologies are also exponentially increasing their energy demand.

The U.S. Department of Energy projects that the percentage of data center electric usage as a total of U.S. electricity usage will increase from 4.4% in 2023 to 7%-12% by 2028. McKenzie and Co., a global management and consulting firm, estimates data center electric demand will increase by 100% from 2022 to 2030. Goldman Sachs Research forecasts that data center global demand will increase by 50% from 2023 to 2027, and as much as 165% by 2030. Regardless of the numbers, electric growth to serve data centers between now and 2030 will be completely unprecedented.

PowerSouth Energy Cooperative is an example of the potential impact. We are a large business, but a small electric utility. We have been in business 84 years and have a peak demand of 2,769 megawatts (MWs), which we set this January. The data centers that have contacted PowerSouth about service have requested as many as 1,000 megawatts (MWs) over a few-year period. Simple math would indicate that three data center additions would more than double our peak demand. That growth in such a short period by two or three new customers would stress PowerSouth’s ability to add generation plants, expand the transmission grid, provide natural gas transmission, and protect our distribution members from the economic risk of data center business failure in the future.

Large utilities will also be stressed to add facilities fast enough to keep up with projected growth. Gas turbine manufacturers, transmission and substation equipment vendors, natural gas pipeline companies and others in the electric generation and delivery industry will be equally challenged.

The greatest obstacle will be providing the actual generation and the environmental permitting requirements for the generation. The industry has never added that much generation in such a short period of time.

The expected growth in natural gas generation is interesting in light of a February 28 article in The Wall Street Journal, “The Clean Energy Revolution is Unstoppable,” claiming the clean energy revolution is inevitable, and that we will see global Net Zero global carbon emissions by 2050.

The authors, who are professors at The University of Oxford in Oxford, England state that fossil fuels have held their historical cost, adjusted for inflation, since their adoption a century ago. However, the authors claim, fossil fuel costs will rise in the future because the costs of extraction will continue to increase. I’m not sure if they have noticed the natural gas fracking revolution — it has made natural gas extraction much easier and cheaper, and it has created the potential for more abundant and affordable natural gas development in the future.

They also claim the cost of clean energy solutions have continued to decrease. Since 1990, they claim, the cost of wind power has declined by about 4% a year, solar power by 12% a year, and batteries by 12% a year. They claim the International Energy Agency (IEA) calculates the electricity from solar, with battery storage, is less expensive than new coal plants in India and new natural gas plants in the U.S. They project that by 2050, renewable energy, combined with batteries and existing nuclear and hydroelectric power, will effectively replace fossil fuel electric generation.

The professors believe the revolution will take over because solar, wind, and battery technologies can be mass produced and will benefit from advances and economies of scale in related sectors like semiconductors, consumer electronics, and aerospace industries. Those projections of increased efficiencies are based upon Moore’s Law, which predicts computing power in a circuit will double every two years, and the application of the “S Curve,” which predicts the cost and availability of technology through the different phases of introduction, development and maturity.

If Moore’s Law and the S Curve actually apply to the mass production of renewable energy, they will be correct, and the cost of renewable technologies will certainly decline. However, if the development of solar and wind power elements is actually governed by the laws of physics and thermodynamics, there will be a collision with Moore’s Law and the S Curve. The laws of physics and thermodynamics stand for the proposition that energy is neither created or destroyed – it is merely changed in form. Yes, there is the possibility that solar and wind generation may become more efficient, but no amount of mass production will create any new energy.

Also, the public is starting to strongly resist the idea of having solar and wind generation farms in their communities. I predict the public’s resistance will grow and the exponential growth of data center electricity demands will lead to even stronger public rejection of new wind and solar facilities.

I hope the authors of “The Clean Energy Revolution is Unstoppable” are correct, because the lower cost of electricity would help lift billions of people out of poverty and greatly improve their lives. However, I expect this is just another academic application of an environmental pipe dream, and the industry will have to deal with the coming data center explosion with known resources.

I hope you have a good month.