Want to reduce the energy used by buildings? Make cities denser.

For many years, urbanists have been singing the praises of urban density. Done well, density can increase economic output, increase per capita productivity, increase disposable income (by reducing heating, cooling, and transportation costs), and improve physical and mental health.

Compact, walkable cities — or walkable areas within cities — are delightful. That’s why living in them is so expensive almost everywhere they exist — there’s too little supply to meet the demand. There are all sorts of reasons to build more of them.

One of those reasons, or at least one of the claims urbanists often make on behalf of density, is that it reduces energy use and therefore greenhouse gas emissions — that it’s a key tool in the fight against climate change. (I’ve made that claim myself.)

The logic is simple: As people live closer together, the average size of their private living space shrinks, their heating and cooling costs decline, and they have less need for individual vehicles.

So just how much difference could urban density make in terms of energy use at a global scale?

Energy use in buildings is going to rise, but how much depends on choices we make today

In a recent paper in the Proceedings of the National Academy of Sciences, a team of nine researchers attempts to grapple with that question.

The researchers focus on the energy used by buildings for heating and cooling as a kind of rough proxy for urban energy consumption, and model various urban development scenarios (one top-down, one bottom-up, if you’re into scenario talk) out to 2050.

(One important caveat: Density, unlike building-level energy efficiency tech, also brings a reduction in transportation energy use, which was not measured in this study.)

The bad news is that, in every scenario, global energy use in buildings increases through 2050 — that’s what rising global population and falling global poverty will get you.

What’s more, average urban density declines across all regions in all scenarios, at least through 2050. The trend toward “dispersion” (i.e., suburban sprawl) is strong everywhere, especially as older, pedestrian-oriented cities are replaced or supplemented by new car-oriented cities.

The good news is that the range of possible future densities is very wide, especially in the world’s rapidly urbanizing areas. In China, for example, “urban population density in 2050 is forecasted to range from 10 to 250 persons per hectare.” That’s a huge difference! What this means is that the choices people make today in those areas will have enormous effects in the future.

The topline conclusion is that “energy use for heating and cooling by the middle of the century will be between 45 and 59 exajoules per year (corresponding to an increase of 7–40% since 2010).” No need to try to get your head around what an exajoule is; just know that it matters a great deal whether building energy use rises 7 or 40 percent.

“Most of this variability,” they write, “is due to the uncertainty in future urban densities of rapidly growing cities in Asia and particularly China.” As in so many other areas, the world’s fate hinges on how China chooses to grow.

Density vs. efficiency

One of the questions the research attempted to answer is the relative contribution to building energy use of density in urban form versus building-level energy efficiency technologies.

“Globally,” they concluded, “our top-down analysis shows that urban density is about as effective as efficiency improvements for energy savings in building heating and cooling.”

That’s a pretty big deal, given how much hype efficiency gets from climate hawks relative to density.

Here it is more precisely:

So density gets you slightly more reduction in building energy use than efficiency does, if you do it right.

If you do both, you get the best result:

Overall, if different regions around the world commit to a compact urban development and invest in efficiency, a cumulative 300 exajoules could be saved through 2050.

Here’s a graph showing the rise in annual building energy use (A) and cumulative building energy use (B) for different scenarios, in different regions. It’s pretty dense — just look at the bottom right boxes, “global.”

Where efficiency matters most, and when

A couple of the more fine-grained conclusions are also worth mentioning.

One, as you’d expect, the relative contribution of density versus efficiency varies from place to place depending on local conditions. Broadly speaking, in areas that are rapidly urbanizing — building new stock — density is more important. In areas that are already urbanized, efficiency is more important. That distinction roughly lines up as developing versus developed countries, though China fits both criteria and both tools are equally important there.

Two, there’s an important finding about the right timing for policies supporting energy efficiency, which is relevant to the already heavily urbanized US.

Today’s most common and affordable energy retrofit options, they say, can reduce building energy use by 20 to 40 percent. But state-of-the-art “deep” retrofits can reduce it by 70 to 90 percent.

If policymakers require retrofits today, the researchers warn, they may “prematurely lock in the existing built environments for a long time to subpar retrofit options.” From a long-term perspective, it would be better to nurture deep-retrofit technologies to “technological and price maturity,” and get them saturated in the market, before policymakers start requiring implementation. Retrofit decisions last a long time, so lock-in is always an issue.

Anyhow, the main point I want to drive home from this study is that, among its many other benefits, well-crafted density is a crucially important tool in reducing the per capita energy use of people living in a rapidly urbanizing world. Urbanists and climate hawks should be pulling in the same direction.