Cost-effectiveness is a crucial criterion for sorting through possible climate actions, but not everything is up for replacement or upgrade today, or even in the next decade.
The California Energy Commission’s recent announcement to require solar on new homes has made waves – waves of enthusiasm, anger, eye-rolling, ambivalence, and gritty determination for climate action, in every possible blend. The economic and technical arguments for and against, as well as the political realities, have received the vigorous discussion they deserve, as the world’s now-fifth-largest economy considers a tricky and controversial tweak to its market for new housing.
There are good reviews of this issue elsewhere, including quite critical ones, but one crucial item seems to have has escaped most observers. Not all potential climate action items are available in a given moment. In other words, among the many changes we might consider for carbon reduction in transportation, energy, buildings, agriculture, and other sectors, many of them pop up only occasionally.
I don’t mean this in a political sense, though of course that’s true as well – “Political will is not fungible” from one cause to another, as David Roberts explains. Instead, I mean this in a technical sense. Not all possible investments are available right now, or even soon. And the moment a house is under construction, certain windows of opportunity are rapidly closing.
As a result, the residential solar mandate is a strategic play to make investments that wouldn’t otherwise show up as options for a long time. Once a house is already built, putting solar on it is a more expensive task. Looking back from 2030, this will look like a darn good idea.
Let’s briefly appreciate how restricted the menu of action items is at any given moment, and then return to California’s mandate, as well as the legitimate reservations with it.
Long-lived capital and key leverage points
There is one fundamental reason that we don’t have every possible action item at our disposal: Most capital is long-lived. This simple fact is true for power plants and cars, dishwashers and industrial equipment.
This is true for energy-using equipment; it’s also true for infrastructure and building systems that don’t use energy directly, but nonetheless shape our energy use. The long-lived highway system requires energy to maintain, but it mainly shapes our pattern of transportation, now the leading source of greenhouse gas emissions. The roof and windows of a home don’t use energy, but they play a significant role in determining the heating, cooling, and lighting load of that home.
- Refrigerator: 8-12 years
- Gas range: 15-23 years
- Car: 10-20 years (or 150,000-250,000 miles)
- Residential roof – concrete tile: 35-50 years
- Residential roof – high-quality asphalt shingle: 30-35 years
- New commercial building: 25-50 years
- Coal-fired power plant: 30-60
- Nuclear power plant: 40+
(There are many sources alleging similar figures; I used this one, this one, and this one, and a few others.)
That’s a big range. For this discussion, the most important aspect of that range is not necessarily obvious: Inside the lifespan of many of those investments, we will want to take a lot more action on climate. When you look out even just 10-15 years, you know there will be a lot of long-lived investments that feel like big liabilities; this rule is a way to keep a lot of houses from giving us the same regret.
The caricature of this situation is notoriously long-lived coal plants, and the legitimate arguments over when and if to retire them (for example in Illinois and Arizona). Less obviously, it’s similar with houses, as more than half of them are fifty years old or older.
As the coal fights show, it’s hard to reconsider – much less do away with – long-lived assets. That’s why it might pay to get ahead of ourselves a bit.
What is this “menu” anyway?
There are many ways to slice and dice climate action. There are multi-sector views, such as McKinsey’s famous abatement cost curves. There are energy-specific descriptions of cost-effectiveness, such as Lazard’s annually-updated Levelized Cost of Energy publication. We’ve seen other analyses specific to a wide range of sectors for many years, including transportation and land use too. Together, these analyses generate a wide range of potential action items for addressing climate change.
McKinsey’s analytical beauty was to force an apples-to-apples comparison of the cost of carbon reduction. It highlighted the vast assortment of cost-effective measures at our disposal, especially in energy efficiency. The last full update of the abatement cost curve (pictured below) came out nearly a decade ago, so it vastly understated the eventual viability of wind and solar, whose costs have fallen by 75% or more in the intervening years.
One might view McKinsey also as a menu in order. By arranging mutually exclusive measures by cost-effectiveness, the implication is clear: start on the left (where things are cheap), and move gradually to the right (where things are expensive). Only a fool would order expensive items first, right? Indeed, that might be not just inefficient but downright counterproductive, as current extravagance will slow the overall effort by squandering scarce funds.
And yet not all of those menu items are truly on the menu today. The menu of actions we might consider, now and in the near future, is much narrower. Just as the table above suggests, we simply don’t replace everything every day. Perhaps this debate, then, is just about timing: being a little bit early vs. showing up way too late.
So, does this make sense right now in California?
This all brings us back to California’s rule. Given the urgency of climate action and the inflexibility of the housing stock, mandatory solar looks like appropriate opportunism: Let’s grab the opportunity for rooftop solar at the best possible moment in a home’s lifetime, even if rooftop solar isn’t the least expensive option.
That’s my main assertion: Looking back from 2030, we will be hugely relieved that California forced some additional solar onto the roofs of a few million homes. We will celebrate that source of generation, and the innovation it spawned.
To be clear, I’m not suggesting that we shift our effort from utility-scale renewables to distributed generation. Large-scale renewables must continue to be the major storyline in generation. And again, I acknowledge the good critical questions that others have asked, notably that mandatory solar on new homes carries risks and inconveniences. The duck curve gets worse, and utilities lose even more grid-supporting revenue.
But these inconveniences and the technical challenges they pose will bring the change we need, and fast. This move surely accelerates the arrival of a future we know we must confront anyway. Add to this rationale that a broad mandate will push competition from “Do I do solar?” to “How can I make solar work best?” and we can count on almost certain reductions in cost – and that’s starting from a point of net costs savings already at the outset.
Should the policy be tweaked a little more to address local circumstances? Probably. Are there a few parts of California where the fundamentals simply aren’t good enough? Maybe – though not many without a lot of sun. (I encourage skeptics to spend a few minutes looking at a geographic variety of capacity factors with NREL’s PVWatts tool to appreciate just how few unsunny places there are in the state. For scale, anything above 1300 or 1400 kWh/kW in a year is a really good capacity factor.) And shouldn’t all of this roll out with more flexibility, such as a clear option for behind-the-meter control by utilities? Again, yes – though again, increased behind-the-meter generation will force the question sooner.
Joshua Skov is a long-time climate and sustainability consultant, and currently faculty in the Center for Sustainable Business Practices at the University of Oregon.
Joshua Skov 