Paul Krugman's The Rise of Renewable Energy is a fascinating read covering the sharp rise in the contribution of renewables to the world electricity generation mix. His first observation is that in solar, battery storage, and wind power, we saw a huge reduction in cost, to the point now that the cost per kWh is lower than fossil fuels and coal. Worth mentioning for the schadenfreude that the naysayers have been proven completely wrong.
A second observation is that these three technologies are only tangentially related in engineering terms; he attributes the declines in cost as related to Wright's Law, which describes a relationship between output and falling costs. This is the learning curve for a manufacturer, where over time they get better at making things. Think of a company reducing defects, finding ways to shave costs, iterating on designs, flexing their buying power to drive down costs, and workers simply getting better at their jobs, among other things driving this relationship.
Krugman makes a good case then for the "infant industry" argument that Alexander Hamilton made a couple centuries ago. Essentially, a government should provide some protection from overseas competition while a domestic industry finds its footing, so as to not be overwhelmed by cheap imports before it has a chance to scale. In this light, blocking Chinese electric vehicles from the US market and driving tariffs on solar panels and batteries makes some long term policy sense. Certainly the Chinese have used these tactics, as well as joint ventures and a number of shall we say aggressive knowledge transfer techniques, to benefit the industries they view as critical to their economy.
I am certainly not versed in policy so wanted to take a detour to talk about a few other related renewable topics.
Local burdens block adoption of extractive energy production
Just as we see our current administration oppose wind power based upon the whims of the current office holder, creating new chaos in that industry, solar and battery storage hugely benefit from the tangled regulatory environment. It is very hard to build a new coal, natural gas, or nuclear plant. Even in US states that have energy policy frameworks that tilt towards the extraction industries, people don't want to live next to a coal-burning power plant.
It is not merely zoning regulations, vocal neighborhoods, and pesky environmental lawyers causing the problems for new plants with old technology. Utilities in this country are a patchwork of thousands of independent entities with a huge variety of technical abilities and financial resources. Putting in a power plant is not as simply as clearing a field and pouring concrete, as there are a host of grid connections and upgrades that need to take place, bonds to be floated, state regulatory agencies to approve rate hikes to pay for the bonds, and then specialty equipment to be sourced and installed. Nuclear energy faces even more dramatic issues with building, despite huge recent technological gains.
Long term, AI drives demand for renewables
Data centers are creating huge increases in energy demand. UC Berkeley researchers cite a 1000% increase from 2014 to 2028 of data center power requirements - going from 58TWh to 580TWh. By contrast, EV power demand in 2023 was 7.6TWh.
Any solar panel salesman can tell you that the biggest reason for a homeowner to get solar panels, even with the tax credits phased out, is as an inflationary hedge. The cost to finance your home energy system per month, if it equals your current average energy spend, will save you money over the life of the loan, as energy costs are anticipated to be driven up year after year by data centers. Your payment should stay flat.
This goes doubly so for businesses where the scale of savings is so much greater and the owner does not need to get on a ladder to swap out a light fixture.
One might expect a Wright's Law for AI data center implementation - certainly there is a huge opportunity for savings if anyone can make the chips that run AI more efficient, or innovate around algorithms to require few processing steps for equivalent results. But so far, while we have seen LLMs like Deep Seek get great results on far smaller processor footprints, it is mostly an arms race where electricity use is increasing geometrically with data center implementation.
Barring a great reduction then in data center energy usage, or huge offsetting gains in efficiency for other energy consumers desperate to cut their energy bills, power bills are going up. And the only way for a business to control this cost in the short term (without drastic measures such as keeping the lights off while people work) is to install renewables.
In fact, a data center from Aligned is going in with massive battery storage capacity to help control costs. Drawing from batteries during peak usage and filling the batteries back up when demand is lower seems to be penciling out for new data centers. So not only will the increased price of energy drive renewables adoption in business, but the data centers themselves will find themselves adopting renewables to cut their main recurring cost.
Impact on Electric Vehicles
I am worried that the increased price of electricity from data centers is going to create even more FUD (Fear, Uncertainty, and Doubt) from the usual anti-EV crowd. Certainly the operating cost reductions will remain a great argument for electrifying transportation when you factor in maintenance and flexibility. Sadly I expect some lobbyist or paid think tank to ignore these cost advantages and instead come out with a chart showing EVs aren't cost effective any longer due to this surge in pricing while gasoline prices, subsidized to the gills, remain at fairly low levels in the US.
The good news is that with vehicle-to-anywhere (V2X) technology, vehicles seen as batteries on wheels can be part of the solution. For one, vehicle batteries will drive scale improvements on batteries used elsewhere as companions to solar, perhaps counteracting the upward pressure from data centers. Innovations in batteries are transforming their energy density and reducing the capacity losses over time. Solid state batteries from companies like QuantumScape, Factorial, and Solid Power alongside traditional vehicle manufacturers like Toyota can drive battery costs down further, which drives down the cost of EVs, making the financial case for switching to EVs stronger even if the fueling costs start to look less enticing than they were.
Innovative companies could create incentives for their fleets and maybe employees to spare some electricity during peak demand, reducing their grid demand when energy is most expensive. Certainly implementing solar, batteries, wind, alternative energy generation sources, and electric fleets will long term mean that a company can generate all their own fuel. It becomes a matter of seeing the pricing on renewables and storage drop below the point where buying from the grid, and assuming all of the inflationary risk, makes sense versus financing a microgrid.
Personal note
If you are really pinching pennies (like me), you can make increasing the energy efficiency of your home a delightful new hobby. Replacing old appliances, swapping in LED lights, running lights on timers, installing more efficient HVAC and insulation can drive your usage down 10-20%. In my area in SW PA, you aren't allowed to install solar that is more than 110% of your average demand, so installing solar first, when in an inefficient state, and then reducing your energy usage via these efficiency upgrades means you can cap out at as much as 130% of your demand. Moving to electric water heaters and heat pumps can also drastically cut your winter gas bills. Anecdotally, I am seeing negative summer power bills for my home in cloudy Pittsburgh, more than offsetting the greatly reduced winter energy bills.