Thoughts for January 9, 2022
Good morning/afternoon/evening everyone. This week’s topics include urban density, fusion, and electrification.
Urban Density
There was an article by Judge Glock in the Breakthrough Journal a few months ago, but I didn’t see it until now. Titled “Sprawl Is Good: The Environmental Case for Suburbia”, the article delivers exactly what is promised. The article makes a good case and I think is a valuable read, though it has a few weaknesses.
Glock builds from a solid theoretical foundation on agglomeration effects in cities. He opens with the value of agglomeration economies—the reason why we have cities in the first place—and then he goes into the negative agglomeration effects of cities: air pollution, traffic congestion, crime, and urban heat islands (he doesn’t talk about noise and light pollution, but these are also critical issues). The distribution of sizes of cities in an economy is roughly determined by the equilibrium between positive and negative agglomeration effects; if the former is why we have cities at all, the latter is why we don’t all live in one giant city.
It is critical to note that positive agglomeration economies are themselves not a function of density per se, but of transportation connectivity. This paper demonstrates that cities with higher highway density see more productivity gains than can be explained simply by the highways allowing a greater city population. This point may seem obvious, but it is frequently overlooked.
Glock argues that there are environmental costs to density, particularly exposure to air pollution. This, as far as I can tell, is true if interpreted properly. A person in a less dense environment will generally create more air pollution, mainly through more driving, but be exposed to less because there are fewer neighbors in a given radius. Which of these two metrics is more relevant is a value judgment to some degree. A person who takes a biocentric or ecocentric view of nature is more like to be disposed to the pro-density case than someone who takes a strictly anthropocentric view.
Glock also argues that, while there are environmental costs to sprawl, those costs are minor and worth being paid. He cites a report that a doubling of density is associated with a 5-12% decline in vehicle miles traveled. This is roughly in line with other figures, though larger reductions are possible if density is accompanied by mixed use development and active and mass transit. In any case, when it comes to pollution and greenhouse gas reduction, we will get far more mileage from higher CAFE standards and electric cars than we will from density, pun intended.
Land use is, by definition, an irresolvable cost of sprawl, though it is partially mitigated by the fact that low density development supports more biodiversity per acre than high density development. Glock observes that even under the most extreme plausible projection, no more than 5% of the land in the United States will be urbanized. If we are concerned about land use, intensifying agriculture and developing alternatives to meat, particularly beef, will save far more land than dense cities.
One area where I think Glock misses the mark is in comparing the non-transportation relative costs of dense and sprawled development. Under a scenario where urban density doubles over the next century—which is feasible technically, maybe economically, and probably not politically—most people will still live in detached single family homes, and of those who don’t, most will live in rowhouses, du-, tri-, and quad-plexes, and 1-2 story walk-up apartments. Most of this density increase would be met by reducing lot sizes and unit size minimums, and only a small amount through apartment midrises and highrises. Under this scenario, we would save on emissions embodied in materials, and we would also save on heating and cooling through smaller houses and shared walls. Studies in Ontario and Australia confirm this.
Glock observes that dense development has a place. Dense urban cores are a vital part of the economy, and many people prefer them. I probably would too if political leadership had credible responses to high cost of living and crime.
Ultimately, I think that density is the wrong question for urban planners to focus on. Density, or lack thereof, is the outcome of good planning, not the objective.
Fusion
Last month, the National Academy of Sciences, Engineering, and Medicine published a report on private development of fusion and how the Department of Energy could support those efforts.
In their motivation, the NAS focuses on fusion as one of several tools for mitigating climate change, and they also discuss that black start capability in the event of a disaster is appealing. They don’t approach fusion from an energy abundance standpoint, considering the many valuable things we could do with a low-cost, unlimited energy source, which I think is an oversight. Fusion would offer baseload power, which is attractive on a possible grid dominated by wind and solar.
In their discussion of regulatory challenges, the report emphasizes the value of building fusion plants where there is already grid infrastructure available. They also discuss the need to reform NEPA, anticipating how it might be used against fusion projects.
There are many important developments that fusion would need to achieve and that private ventures have not yet achieved. They include
Net energy gain (Q>1), followed by net electrical energy gain.
Ash (the helium produced from deuterium-tritium reactions) removal.
Tritium breeding.
HTS (high temperature superconductor) magnets. YCBO (Yttrium barium copper oxide) appears to be the most promising candidate.
Advances in materials, particularly for the blanket for energy removal, tritium breeding, and neutron shielding.
Cost-effective solutions for plasma heating and actuators.
The report emphases that D-T fusion with a tokamak remains the leading route, as also shown in this 2019 review.
Tritium breeding is critical. A single 500 MW(thermal) D-T fusion plant—a modest-sized power plant—would deplete the world’s tritium stockpile in a year. The report also recommends further development into advanced fission reactors that would breed more start-up tritium.
The two other major reactions considered are deuterium—helium-3 and proton—boron-11. There are plenty of protons and boron-11 for the latter, but for the former, helium-3 availability would be a problem. The report says that lunar mining of helium-3 “may” be necessary for this route, but they don’t assess the possibility of breeding tritium and harvesting He-3, the decay product of tritium. I have a very hard time believing that lunar mining of helium-3 will be feasible in the foreseeable future. Current U. S. inventories would be sufficient to run a 400 MW(thermal) plant. The main challenge of the D-He3 and p-B11 routes, especially the latter, is that the plasma is less reactive and thus would have to be run at a higher temperature, further complicated the materials science problems. The advantage is that they produce fewer neutrons. Ash removal would remain a problem.
The report describes a pilot plant, to be constructed in the 2035-40 time frame, with targets of 50 MW(electric) production and an overnight construction cost of $5-6 billion. This would be way too expensive for mass power generation, but the pilot plant is meant to help drive down costs by kicking off learning curves and economies of scale.
The report touches on continued U. S. involvement in ITER and recommends that this involvement continues. The previous Burning Plasma report, as well as Clery’s review cited above, helped convince me that this is a good idea. Though the report sees commercialization of fusion occurring in parallel, rather than in series, with ITER.
I am still enthusiastic about fusion and want to see continued effort. And while I am grateful for the recent wave of investment and enthusiasm, it really must be emphasized that commercial fusion is not close, and there are still fundamental technical issues that need to be resolved. I fear that recent excitement in the press could breed false expectations, followed by more cynicism when those expectations are inevitably dashed.
Electrification
Although there is much more I would like to do, it is time to put the electrification analysis aside and move on to other projects. Here are my last observations.
Over time, the United States has gotten a greater portion of its energy in the form of electricity, and there is considerable variance in how energy much the different states get as electricity. But how much of this is simply due to a changing economy? Maybe the United States is exporting heavy non-electrified industry, and maybe the most electrified states are simply those that don’t have much industry.
This isn’t really the case. Using an adaptation of structural decomposition analysis, I found that about 30% of the electrification in the U.S. from 1960 to 2019 is due to sectoral shift, and the rest is due to electrification within sectors. Similarly, about 34% of the variance of electrification between states is due to different relative amounts of the sectors, and the rest is due to the individual sectors being more electrified. Of course, there are only four broad sectors, and maybe if we looked at energy consumption more granularly, such as through the Manufacturing Energy Consumption Survey, we would find a greater role for sectoral shift.
The one factor that seems to make a real difference is the price of electricity. This should be obvious: if electricity is cheaper relative to other forms of energy, then there is a financial case to use more of it. And this result is confirmed in the data.
A write-up of these results can be seen here, and the code is here.