We need to think harder about solar and wind energy

The path of least resistance doesn't always lead where you want to go

Over the past several months I’ve become increasingly skeptical of the ability of variable renewable energy — solar and wind — to adequately decarbonize the electric grid without substantial assistance or planning.

That doesn’t mean that I think that climate change isn’t a problem. On the contrary, it’s the worst and most frightening problem in the world. But specifically because I think climate change is the worst and most frightening problem in the world, I’m very worried about scenarios where the basic pitfalls of political economy and human psychology undermine our ability to decarbonize. And I think the current zeal for solar and wind might be the product of some of those pitfalls.

First, a caveat: I am not saying that there is no role for solar and wind energy in decarbonization. Clearly there is; many electric grids already heavily rely on these sources.

What concerns me, instead, is the idea that solar and wind energy are so clearly superior to other forms of electricity generation that they will, almost inevitably, through operation of markets, gradually replace carbon-emitting power generation, until the grid is nearly fully decarbonized. This process might be sped along or otherwise encouraged with subsidies. But in the political world, there is a growing sense that the advantages of variable renewable energy are so clear and self-evident that decarbonization won’t require any kind of top-down or central planning.

In short, there is a noticeable shift towards the view that government and utilities need not do hard things, like:

• directly planning the switchover of carbon-intensive plants to zero-carbon plants, or

• creating long-term plans to build an alternative grid that produces little or no carbon, or

• penalizing carbon emissions.

Instead, we simply need to wait for the invisible hand of the market work its magic. Perhaps we do some tinkering far down the road, to clean up the final rump of fossil-fuel electricity generation. But in the meantime, we’re on a glide path to a carbon-free future.

There is a growing private industry promoting this view, and behind it, a whole cluster of political commentary. Oddly, much of that commentary is coming from left-leaning circles, who would be otherwise skeptical of market-driven miracles. But the green aura around solar and wind is so powerful that many writers have broadly abandoned their skepticism over industry promises and thrown in with a utopian vision of an electricity market transformed by innovative new technologies — many of which remain mostly hypothetical. That’s in part because, without those new technologies, it remains impossible to decarbonize with solar and wind.

In my view, we’ve reached a point in the decarbonization process where two paths have begun to diverge in front of us. One is the path of leveraging state capacity, in which we use tools of government to plan and implement decarbonization as a massive societal project — like arming for WWII. The other is the path of market neoliberalism, in which we rely on diffuse market actors, prompted by economic incentives, to complete that decarbonization project.

And increasingly, it seems as if we are being pushed on to the second path by the idea of solar and wind power as almost magically abundant sources of electricity. Compared to more traditional power plants, solar and wind are cheap and easy to install. They do not require years of centralized planning by monumental quasi-public utilities. They can be added smoothly, easily, quickly, to undercut existing power production (especially with the help of subsidies), all driven by operators’ urge to make a quick buck.

Bluntly, the second path is easier, politically. It requires no sweeping policy changes, no controversial government action, little legislation of any kind. The problem is, the second path doesn’t seem to reach all the way to the goal of full decarbonization — or, I worry, even get us close. And while the first path is undeniably harder, the longer we ignore it, the harder it becomes. The further you go down the wrong road, the worse off you are, when you finally figure out you’re going the wrong way.

INTERMITTENCY IS COMPLICATED

Underneath this political conundrum is a technical conundrum. Here is where I’m required to say that I am by no means an expert in the engineering questions. But the basic technical problem with solar and wind does not require a doctorate to understand. In fact, it’s relatively simple.

So what’s the technical problem? Well, it’s the obvious thing: at nighttime it gets dark. And sometimes the wind doesn’t blow.

That’s an oversimplification. The basic challenge of solar and wind energy is intermittency. But intermittency happens different ways, all of which are important. Solar and wind power generation vary three different ways. They vary based on time of day, they vary from season to season, and then they vary randomly a lot based on local weather conditions, which, of course, are not very predictable.

Demand for electricity varies the same three ways: time of day, between seasons, and sometimes randomly (again, to a large extent based on local weather conditions).

Demand for electricity varies in another way, too: it usually rises over time. And it’s probably going to keep rising. The more we decarbonize the more quickly it will rise, as the energy demand for things like transportation and heating is moved onto the grid. More problematically, as the demand for electricity rises, the pattern of that demand will change, in ways that are sometimes hard to predict. For instance, here in Minnesota, winter demand peaks will get much higher as gas heating is replaced with electrified heating systems.

The ironclad restriction on decarbonization is this: the electrical grid has to be built in order to meet demand on the grid’s worst day. This is the criterion which any reliable grid must meet, or it stops being a reliable grid, and starts becoming an unstable patchwork beset with rolling blackouts, the kind of thing that drives countries into political collapse.

Matching capacity and demand is a very significant challenge with any energy source. But with highly variable sources the challenge gets much, much steeper. With variable sources you need have to match shifting demand (which can rise tremendously, based on things like time of year and weather) with capacity that also shifts (and can ebb tremendously, also based on things like time of year and weather). So not only are there obvious mismatches (e.g., you make less electricity in darker and cloudier winter, but you also need more electricity, to heat homes), but there are also an enormous number of highly unpredictable conditions that can create those mismatches. What happens if you have an extremely hot day with little wind? Suddenly you lose a lot of your generating capacity while demand is spiking. What if it’s unseasonably cold and also very cloudy? Same thing.

The upshot here is that for every condition of extreme demand, you need a somewhat different grid. You need a solar-driven grid for the windless days, and a wind-driven grid for the short winter days. Virtually every source of generation on your grid will have times where it’s useless. What’s worse, many of those times will be correlated: seasonally, or because weather conditions are the same across a wide area, or just because it’s night everywhere nearby at the same time. Worse still, weather conditions also often correlate with high demand. As a result variable renewable sources have a risk of literally failing when they’re needed the most. For instance, consider how a winter storm will both spike demand and also knock out a lot of solar power capacity.

Traditional sources of electricity generation are not 100% reliable, of course. Every power plant has downtime. But that downtime is usually planned, and limited; there is not constantly shifting variation in capacity, and this downtime is not usually caused by the same conditions that are driving electricity demand.

THERE IS NO CLEAR SOLUTION TO INTERMITTENCY

So that’s the basic technical puzzle. (There are other technical challenges too, but intermittency is the really thorny one, inherent to solar and wind power.) And of course everyone is aware of this puzzle and many solutions have been proposed. As I see it, there are four broad solutions, all of which have problems:

• Overbuilding. Basically, one answer is just to flood the world with so many wind turbines and solar panels that we can constantly power everything all the time, even if only a tiny fraction of the plants are actually producing electricity. There’s usually a little bit of wind or sun available somewhere, after all. This could work, in the sense that it’s conceptually possible (especially in a theoretical model) to set the number of turbines and panels so high that the grid can produce enormous multiples of overall demand, making electricity abundant even in tough conditions. But it doesn’t strike me as practical. These are space-intensive facilities and not every potential site is equally useful. Over time, the cheapest sunny and windy sites will be taken. And then the next cheapest. And then the next cheapest. At some point you’re going to putting turbines and panels in places that are straightforwardly bad, because they’re far away, hard to access, have bad weather conditions, are being used for other things, etc. Costs will skyrocket. And as demand keeps going up, you’ll still need to keep adding generation — again, many multiples of what is actually ever used at once. Does this sound realistic, to you? Or do you suspect as a practical matter, this may not come to fruition?

• Storage. If you can store energy when it’s available (sunny days, windy nights), and then use it during lulls in generation, then there’s a lot less of a problem here. But the amount of storage needed is massive: in order for storage to address the intermittency of a grid mostly powered by solar and wind, you’d need to be able to run most of the grid from storage for extended periods (e.g., an entire day). Simply put, we do not currently have the technical capacity to store a grid’s worth of energy for that long and it’s hard to find anyone who thinks that degree of storage is right around the corner. Storage also requires even more solar and wind plants to be built, because you need to actually fill that storage. You’re basically building a rainy day fund full of energy, and so you have to make more energy when it’s available. Finally, while storing energy in the morning and discharging it in the early evening is within the technical capabilities of existing batteries, we do not have batteries that can store large amounts of energy in the summer and discharge it in the winter. If you want to charge those winter batteries, you need to milk even more energy from the unfavorable winter weather conditions, meaning, yes, even more wind turbines and solar panels. And here you start to run into the same issues as with overbuilding, above.

• Transmission. The more you can move electricity around, the easier it becomes to patch holes caused by solar and wind. Sure, it’s night on the east coast, but it’s still early evening on the west coast. So you take some power from the solar panels there and shoot it across the nation to New York. Likewise, it’s windy offshore but calmer in the Great Plains. So you generate the power in the Atlantic and transmit it a thousand miles inland. Building massive, nation-spanning transmission lines would let you play these kinds of games. But once again, there are huge problems. First, shifting around power like this helps solve some of the problem, but even with lots of transmission, there are times when electricity gets tight. It’s often night across the whole country at once. Sometimes you get unlucky and there’s just not much wind anywhere. So you’re back to needing storage or redundancy. Moreover, actually building these transmission lines is a monumental undertaking all on its own. It sounds easy — and maybe it is, compared to papering the earth in solar panels — but these are hard projects to pull off. Right now, even the most obvious transmission projects, like lines across New England to pull energy from hydro plants in Quebec, are getting consistently caught in the teeth of NIMBY and environmental opposition.

• Just burn gas. Ah, now here’s an easy, workable solution to renewable energy intermittency. Natural gas is an easy, abundant power source that comes online very fast. It produces a ton of electricity on demand. When solar and wind sag, you just ramp up gas turbines and plug the hole. No muss, no fuss. You don’t have to build huge amounts of infrastructure. The facilities already exist and it’s cheap to build more. The only problem is that it causes carbon emissions, but hey — if you’re a plant operator, the cost of those is almost entirely externalized. It’s the path of least resistance. And as solar and wind become more common, this has been, overwhelmingly, the solution to their shortcomings.

There are other proposals beyond just these four: smart grids, “demand response” incentive programs, hydrogen fuels, storing energy in people’s electric cars overnight, etc. And of course all of these solutions can be combined in various proportions. Every renewable energy advocate has their favorite mix of overbuilding, storage, transmission, gas, and magic tech. Sometimes the numbers do kinda-sorta pencil out. Sometimes it’s not even close. (I saw a proposal the other day that relied on storing thirty days of energy on the grid, which is really akin to saying we should simply solve the problem by inventing cold fusion.)

But I have three broad points to make about all of this.

First, I don’t like staking the future of the planet on technology that doesn’t exist at a large scale yet. And many solutions to intermittency do require some new technology. It seems to me that we are approaching the problem of decarbonization backwards: doing the easy part of adding turbines and solar panels, and waving away the really hard part with vague promises of new tech or total transformations of the grid. Anytime someone says “Don’t worry, we’re just going to do the hard part last,” I get very suspicious. Are you really? Just because you are currently picking the low-hanging fruit doesn’t mean you will (or are even able to) pick the higher-hanging fruit.

Second, although there are many, many proposed solutions to this problem of renewable intermittency, tellingly, few people can agree what the obvious solution actually is. Look, if solving intermittency on a renewable-heavy grid was clearly workable, with a straightforward answer, wouldn’t there be some degree of consensus? But no such consensus exists. This leads me to suspect that many people are hearing what they want to hear: that we can simply keep adding wind turbines and solar panels, a process that’s already well underway, and the problem of decarbonization will solve itself. That’s a compelling idea, a huge relief if true — but why can no one agree what the final, decarbonized power grid will look like?

Third, the margins here are just very tight. Although it’s hard to know for sure, it’s at least possible that there’s some specific, complex balance of storage, overbuilding, storage, and new transmission, that produces enough power to meet demand under almost all conditions. But realistically, how plausible is it that we find that balance through the blind groping of free-market mechanisms? Can we really get there with a minimum of central planning, installing new infrastructure only when it’s profitable to do so?

We are, it seems, trying to shoot a very narrow gap — and we are trying to do it blindfolded. Solving complex infrastructural puzzles requires planning from above; these mechanisms aren’t likely to just fall into place on their own. Unguided markets, functioning on their own, will happily lead us towards catastrophic outcomes, as long as the short-term incentives are correct. That’s literally how we got in this jam in the first place.

It seems to me that if you are planning to fix a very complicated system and the penalty of failure is existential, then you want to have big, generous margins of error. You want to assume unexpected things will happen and those things will make the problem harder. But instead, the renewable energy vision is the opposite: it only works if things are easier than we can see now. Innovation will make storage cheaper and more useful, the cost of new generation will fall forever, smarter grids with better transmission (not yet in evidence) will make it easier to transfer power.

Now, some or all of those optimistic predictions might come to pass. But counting on it seems very dangerous indeed. Let’s plan for the worst-case scenario, not the best-case.

CARBON-FREE RENEWABLE GRIDS REMAIN SPECULATIVE

One final observation, and to me this one is all-important: it’s easy to imagine new futures, it’s a lot harder to build them. Because of this, and given the absolute importance of achieving decarbonization, I put a lot of stock in real-world examples. The cold, hard truth is this: right now, it doesn’t seem like any big grid in the world is close to decarbonizing with solar and wind.

A number of grids are leaning heavily on variable renewables, like California and Texas. California in particular produces massive amounts of solar energy during the daytime. But in both places, when those sources fall short, the problem hasn’t been solved with overbuilding or storage or transmission. It’s being solved the cheap, easy way, the path of least resistance: natural gas. As a result total carbon emissions haven’t fallen all that much. The same easy road that led to the installation of lots of cheap solar panels and wind turbines has also led to the easiest solution for their shortcomings, which is fossil fuels.

But there are a few big modern grids that have almost fully decarbonized. One is Quebec. The other is France. And neither of those grids really use solar or wind. Quebec runs on hydro power. And France runs on nuclear.

The practical, real-world lesson for me here is that decarbonization is possible. There is concrete proof of that. But there is no concrete proof that decarbonization with mostly solar and wind power is possible. The only proven pathways are nuclear and hydro power.

We obviously have the technical capability to build hydro and nuclear power (though good hydro sites are often taken, and dams are hardly environmentally friendly). We know this is possible. And what it requires, more than advances in technology, is advances in planning capacity. We need governments and utilities with the power and incentive to plan these large-scale plants and bring them online, preferably as cheaply as possible.

Right now, we’re going the other direction. The easy direction. The install-a-solar-farm-in-a-field-and-don’t-ask-what’s-keeping-the-lights-on-after-dark-(because-it’s-gas) direction. We are hoping deregulated energy markets will simply shake themselves free of carbon. We soothe ourselves with charts showing the price of solar energy falling and falling. Behind that number is a world of complex obstacles, but we tell ourselves that there must be a technological path through them, even if we can’t say exactly what it is. We build, in our heads but not in the real world, brilliant futuristic grids designed entirely around the need to circumvent those obstacles. We see emissions going up and temperature records falling, and settle our nerves with science-fiction imagery of cities running entirely on solar panels. As long as we can keep that dream alive, we don’t have to confront the reality that no one’s come close to bringing it to fruition. And behind it all, keeping the whole thing afloat, we’re burning natural gas.

I don’t have a perfect solution to climate change, any more than anyone else. But I do know what got us into this mess: denial. Climate change is the product of our collective inability to acknowledge a hard problem and our collective desire to believe that the pessimists are wrong. And I worry that, even if we’ve finally (mostly) accepted the need to fight climate change, we have not defeated our capacity for denial. Humanity followed the path of least resistance into this crisis; why would anyone assume we’re not following it still?

A note:

I know the ideas in this post might make some people angry, including a lot of people I like. People associate skepticism of solar and wind with right-wing politics and climate denial. Many on the right detest these power sources simply because they are not worried about global warming.

All I can say is that my concerns are rooted in the exact opposite instinct. I am very worried about global warming, more than any other problem humanity faces. Because I am worried, I want to find solutions that I am comfortable with. I also worry about how intoxicating optimism and triumphalism can be. I don’t have an ideological attachment to any particular power source, I just want to find the quickest and safest path to full decarbonization, and I can’t quite silence that nagging voice in the back of my head that asks “If solar and wind are the path, why hasn’t anyone done it? When we can expect it?” My skepticism is sincere and good-faith, and sincere skepticism shouldn’t be a threat, since you can prove it wrong. Prove me wrong!