America can beat China on energy

It's time to leapfrog China in the single most important industrial input

America needs to start beating PRC on basic industrial stuff if we want to remain economically relevant in the 21st century. It will take some time to get better at things like phones and drones, but energy generation is a different story. I've been poking around a bit inside some electricity pricing datasets lately, and this has led me to the conclusion that America can beat China in the most important industrial input of them all: grid electricity.

Energy is a major component of the cost of everything that we produce and consume. Much cheaper electrons would be helpful with pretty much every other policy priority. It would encourage more industrial buildout of all types (which means more jobs), it would make electrification of transportation much less painful, and it would do wonders for household budgets.

We have reason to do this, and unlike areas like phones or drones, it's pretty straightforward to generate way more electricity by scaling systems that we already have. America has a long history of booming domestic energy production in response to crises. This is a crisis, and we should boom in response. If we do this right, we can create a durable American industrial advantage that China can’t easily compete with.

Overview

• How much do electrons cost these days? … Electricity market overview with selected examples from US and PRC, tldr we can beat them

• The location problem … America makes cheap power, just in the wrong places

• The photon problem … Domestic deployment of solar farms and battery storage is held up by stupid culture war stuff, we should stop that

• The neutron problem … Delete NRC, deploy SMRs, maybe state-owned SMRs

• Towards $0.01/kWh … We should do all this to beat China, but after that we should think much bigger

How much do electrons cost these days?

Let's look at industrial grid pricing for Southern China:

The entire world's electronics supply chain depends on the Southern China's Pearl River Delta. In Shenzhen and its environs you can expect to pay a base rate of ~$0.09/kWh for electrons to your manufacturing facility. If you factor in things like provincial and municipal governments giving favored entities pricing perks you probably end up with a somewhat lower real number, but it's still in the high single digit cents per kWh ballpark. This is a high demand region with no locally available hydrocarbons and lots of air conditioning requirements given the tropical clime, so lots of customers are competing over scarce grid capacity and we see significant demand variability over time. We see similar prices (~$0.09/kWh) in Beijing, Shanghai, and Chongqing - some of China's other core urbanized regions.

Compare this to America's economic mecca, California, where companies pay ~$0.24/kWh for industrial grid access. When electrons on one side of an ocean cost roughly three times what electrons on the other side of that ocean cost it's not very hard to predict where the industrial capacity will be installed. Massachusetts comes in at ~$0.18/kWh - the rest of the Northeast (less New York) is within a few cents of that number.

This is bad, but thankfully not all of America is like California or Massachusetts. Texas actually beats pretty much every region in China besides the major coal producing provinces at ~$0.06/kWh. Washington and Oregon come in at ~$0.067/kWh and ~$0.077/kWh respectively, owing to numerous hydroelectric dam projects. Illinois (a place known for its high cost of doing business) comes in at ~$0.095/kWh, creditable to a large nuclear reactor fleet.

New York has some of America's highest residential and commercial electricity prices, but still managed to keep statewide industrial prices below $0.10/kWh in October 2024. Unfortunately this is trending in the wrong direction; October 2023's number was ~$0.065/kWh. There are enormous electricity pricing differences between NYC and the Upstate region. On the day that I'm publishing this piece (January 27th, 2025) we saw negative electricity prices near Buffalo and some of the country's highest spot prices in NYC and its environs.

Generally speaking, American industrial electricity prices rise sharply anywhere near dense concentrations of human capital. This is a deeply pernicious phenomenon. American factories have to eat an enormous increase in operational costs if they want to be nearby talent. If we fix this, we gain an advantage that China will have a hard time duplicating.

The location problem

Donald Trump wants to double American electricity generation, and has declared a national energy emergency in order to expedite actions that further that objective. I think the goal is admirable and the use of emergency powers is justified, but I don't think either the ambitions or the current means of reaching them go nearly far enough.

This buildout is intended to supply power to the rapidly growing artificial intelligence sector. We need vast supercomputing clusters to train AI models, and even larger deployments of inference compute to take advantage of their capabilities. We should absolutely build power plants in the wilderness and spin up supercomputing clusters next to them, but that's only the first of many steps that must be taken to fully exploit America's energy advantage.

Datacenters are great, but you can locate them basically anywhere as megabytes are pretty cheap to ship to endusers compared to, say, aluminum castings. Economic functions that require proximity to a skilled workforce do not benefit much from $0.05/kWh electrons in South Dakota. Even AI inference tasks would return results with lower latency if done close to the enduser.

We need to add to our total electrical generation capacity, but our inability to distribute electrons is the more immediate problem. We can generate cheap electricity at hydroelectric dams in Tennessee or solar farms in Nevada, but our ability to get it to major population centers is severely hampered by decaying transmission infrastructure. Grid equipment is so badly cared for and so severely out of date that it routinely causes forest fires in California and other western states. This is an area that acutely requires government intervention. Utilities that won't replace failing infrastructure should be severely punished by regulators. In some cases (PG&E perhaps?) this might mean a forced sale to more competent operators.

New transmission capacity must be heavily subsidized, particularly when it connects new generators to major population centers. $0.05/kWh power in the San Francisco Bay Area would make it the obvious place to put the electronics manufacturing facilities that we urgently need to substitute for imports from China. The physics of transporting cheap electrons to a skilled workforce are far more tractable than the economics of transplanting a skilled workforce to cheap electrons. In an age of high voltage direct current trunk lines our failure to do this is purely a policy problem.

The photon problem

It is heartening to see a positive vision from policymakers around dramatically scaling up energy production, but the new administration is getting certain specifics very wrong for very dumb reasons.

GOP policymakers are stuck in an Obama-era culture war mindset with respect to solar panel and battery development. This attitude was myopic then and is myopic in the extreme now. GOP temper tantrums around solar and batteries contributed a lot to the definite advantage that China now enjoys in these two categories. Notably, this attitude is not shared by large red states like Texas, which is building solar at a quick clip to augment it’s already sizable wind turbine fleet.

The rest of the country should follow the example that Texas is setting: solar farms deserve at least as favorable regulatory treatment as is afforded to natgas wells. If we unblock projects and start new ones we may even find that solar is a useful adjunct to existing natgas infrastructure.

We need government intervention to recover the ground that we've lost to China over the last two decades in. Solar and battery infrastructure is made riskier by the fact that PRC owns the production stack for affordable solar panels and batteries. We need the government to invest in domestic production of both. The Department of Energy Loan Programs Office is a good vehicle for this, their mandate should be expanded significantly, and Jigar Shah (a fairly nonpartisan Biden appointee who did a great job) should be put back in charge.

It is worth noting that President Trump was elected on a platform that directly opposes economic structures like the one which produced the current dynamic around solar and batteries. America paid for all the basic research into photovoltaics and battery chemistry and then outsourced all of the jobs and all of the profits to a hostile country overseas. This is a bad deal. We cannot rely on China for basic inputs.

The neutron problem

We need more nuclear reactors. If you're reading this blog you're probably part of the choir so I won't preach too much on this point. Nuclear buildout is blocked by NRC, which bans reactor installation in practice by requiring absurd safety standards that are grounded in bureaucratic rent extraction and irrational risk aversion, not physics.

There's a lawsuit working its way thru the courts right now that might give states a bit more leeway with around small modular reactors, but that's not a national fix, and it could take a very long time to play out because there will be a state level fight to follow.

NRC just needs to be deleted and replaced with an agency that actually has a mandate to install reactors, there is not incremental solution here, no amount of armtwisting will actually get the agency out of the way.

I have come to the conclusion that we need a state-backed reactor fleet (either DoD or DoE would work) to provide baseload power to currently expensive coastal markets and pressure utilities to lower prices. This is annoying to admit given my relatively free-market-oriented politics, but I don't see a more direct path to deploying lots of small modular reactors quickly. For a sketch of what that might look like, see this Twitter thread. I'll probably write more on this at a later date.

China is installing 1GW reactors at a steady clip, but they have no ready answer to a rapidly scaled American SMR program should we commit to one. SMRs deployed offshore or on military bases or close to major population centers would go a long way to solving the location problem without many of the land use issues that often block reactor construction. I think Navy S9Gs are the obvious candidate for mass deployment, but there are also commercial options. We could, and probably should, do both, but I think it’s worth explicitly making the case for a fully state-owned option to light fires under asses. If you don’t like the idea of government baseload then it’s on you to come up with a better idea.

Towards $0.01/kWh

China's electricity comes mainly from coal and hydro. They've dammed most of their dammable rivers and they've created far more smog by burning coal than is healthy for their population. Their nuclear program is firmly 20th century in character and already has a way of doing business that will be difficult for them to adjust. The Chinese deserts that are best suited for large solar farms are far away from the coasts that consume most of the energy. They might be able to cut their energy prices in half, but I doubt they’d be able to do it quicker than America can, and I really doubt they’d be able to follow us even further if we were to shoot for an objective like $0.01/kWh.

It's crazy that a relatively energy-poor country operating under all these physical and organizational constraints manages to get cheaper electricity in its major economic centers than we do. Our government had a lot to do with the perverse financial incentive structures and stupid regulations that got us to this point, and only our government can undo it in timeframes that matter in the context of fierce global competition. We have to throw money at the problem at grand scale, but if we do we can expect to cover a lot of ground quickly, and in a way that China can't easily duplicate.

Scaling American energy generation by enough and in the right places might be a >$1t project, but we will reap far more than that sum if we pursue it aggressively, and we may even enable victory in the most consequential geopolitical confrontation of our times in the process. We have the means, now it is a question of will.

Economic prosperity is a function of energy availability to a great extent. We’re currently underperforming the Henry Adams curve by at least a factor of five. We must be competitive with Chinese industry in order to maintain our security and prosperity, but once we’ve accomplished that objective we must stretch our ambitions further. I’ll settle for doubling energy generation for now, but in the long run we should be discussing orders of magnitude, not factors of single digits.

Sources:

PRC industrial electricity prices: https://china-briefing.com/news/chinas-industrial-power-rates-category-electricity-usage-region-classification/

USA industrial electricity prices: https://eia.gov/electricity/monthly/epm_table_grapher.php?t=epmt_5_6_a

Disclosure:

I founded and run an industrial automation company, General Fabrication. Please buy our exceptionally capable and designed-and-assembled-in-USA 3D printer and ping me with manufacturing problems that we could solve with a robot. We would benefit financially from the policies I lay out above, particularly if somebody managed to deliver $0.05/kWh in San Francisco with a nuclear power plant ship. Depending on how you look at it that means that I have skin in the game or a significant narrative bias, caveat lector.