viable US megaprojects




Real interest rates are now negative. What could be done with all this excess capital?

Well, it's not like there's literally nothing worth doing. Lots of companies have proposals for projects that have >10% estimated returns but get rejected. Middle managers are risk-averse, and the inaccuracy of estimates means a large safety margin for rates of return is demanded. In general, bigger projects have less competition and benefit from economy of scale, but the larger institutions required to run them are less efficient. If competence is held constant, big projects tend to be more potentially profitable.

But anyway, I've listed some things below.



What do I mean by "viable megaprojects"?
"viable" - Things that are profitable when externalities are included, preferably with an IRR >10%.
"megaprojects" - Projects with a minimum efficient scale of $1B or more.

I'm also going to limit my list to things that don't require new or experimental technology.



#1 - a new city


The minimum viable scale for a city these days is around 1M people. Let's say you want 108 m2 of floor space, 108 m2 of road, and services like sewers and power lines. Construction costs vary too widely for me to want to make a precise estimate, but that's obviously more than $100B. While Apple could potentially afford that, it's basically something that only a government can do.


You want somewhere:

- with flat land to build on
- with ground that can support fairly tall buildings
- with a reasonably comfortable climate
- relatively close to an existing city
- relatively close to an existing or potential seaport


That's a lot of criteria, but America does have some mostly-unused locations meeting them well enough.

Why would you build a completely new city instead of expanding an existing city?

- existing cities have extensive regulations for construction
- many existing cities have expanded to natural geographic limitations such as mountains or water
- intracity traffic density increases with size and has negative externalities


As Dubai has shown, it's possible to just build a bunch of tall buildings in some dumb location and have it basically work out. Moderately tall buildings are not necessarily more expensive; you can build 20 to 30 storey buildings for about the same cost as 6 storey ones. 100 storey skyscrapers are pretty expensive, but in general, tall buildings in expensive existing US cities are profitable and some people are making lots of money from getting or giving permission to build them. Without that, and with the economy of scale from building lots of similar stuff in the same area with the same regulations and requirements, it's possible to greatly reduce construction costs.


Maybe the US government could force some agencies to move from Washington DC to this new city to get things started. The leadership of them would absolutely hate that.



#2 - a national HVDC grid


Solar and wind power are unreliable. A national power grid would enable more extensive usage of them. Here's an influential paper concluding that:

US power consumers could save an estimated US$47.2 billion annually with a national electrical power system versus a regionally divided one (~1.1ยข/kWh). This amounts to almost three times the cost of the HVDC transmission per year.

Here's further analysis, concluding that $400B of investment in HVDC transmission and $2 trillion in new generation is justified.


Currently, the US has no national power grid, partly because its largest grids are AC and synchronizing AC over very large distances is problematic. HVDC is now the standard choice for long-distance transmission. The transmission losses are lower, and DC doesn't require synchronization.

Note that America now lacks the ability to even produce enough aluminum for a national HVDC grid.



#3 - copper mines


Under President Biden, two large copper mine projects in the USA have been blocked on environmental grounds: Twin Metals in MN and Resolution Copper in AZ.

Now, I'm not actually saying they should be approved. Mining companies don't pay royalties on material from federal lands, and they should. If the profit would go to companies that think they deserve money because they bought another company with an old mining claim or have the political connections to trade land for more-valuable land, well, screw them.

What I am saying is that they're profitable, and at least theoretically could be done without too much environmental harm. What the US government should do is change the mining laws, auction the rights, and require a joint venture with partial government investment and ownership.



#4 - bus-like vehicles


I don't like passenger trains. Train tracks are more expensive to build than roads, because steel wheels require smooth, straight, flat tracks, and because the stress is more concentrated. Being at low altitude and near the ground increases drag, so very high speed trains are less efficient than aircraft. Low friction means slow acceleration and braking in addition to limiting track grade.

On the other hand, overhead electric wires work well, and large passenger vehicles can work well.

What I'm proposing here is large vehicles with pneumatic tires on dedicated roads, powered by overhead electric lines. Perhaps:

- 120 feet long, 12 feet wide, double-decker
- 165 mph typical speed
- 30 wheels
- powered by 40kV DC overhead wire


This would involve development of a new vehicle class, and construction of a significant amount of dedicated road with overhead electric lines. But that's better than trying to build passenger rail in the USA. Probably.



#5 - a mobile offshore base


I already wrote about this here.



#6 - new shipyards


China, Japan, and South Korea have all subsidized shipbuilding heavily. Why would you want to compete with them?

Well, the US requires that ships for for the US Navy be built in US shipyards, and US shipyards are outdated and inefficient.

Here's a timelapse of the USS Gerald Ford carrier in a drydock. Note that the cranes lift relatively small pieces, and that it sits in a drydock for 4 years. More-efficient shipyards typically make slices of the entire ship section and weld them together.

You can make a decent new shipyard for $2B. (That's about what a single destroyer costs the USA; South Korea can build something slightly better for half the cost.) Maybe you want 2 new shipyards, so $4B?

Well, it looks like some members of Congress want to spend $25B on shipyards - going to the same people who can barely build overpriced ships now, of course. It looks like the US Navy is still "developing" a plan (SIOP) that might be 20 years or 10 years. They have no idea how long it will be but it's definitely 21 billion dollars. Well, I say, lose the people who're bad at building ships, and find some new people.



#7 - methanol


World production of methanol is over 150M tons/year. The US makes about 6M tons/year. China makes over 70M tons/year and is rapidly increasing its production. This is a case of China being smart and the USA being dumb.


The USA has cheap natural gas and can make methanol more cheaply than China. On the other hand, the Chinese government wants to make methanol from coal and turn it into olefins and aromatics to secure a domestic supply of polypropylene, polystyrene, phenol, PET, and nylon. They might even want to make gasoline from coal via methanol.


Anyway, US natural gas to methanol to olefins is viable. Methanol to aromatics and some other stuff is viable but there's some new technology involved in doing that efficiently. Producing certain important chemicals from biomass would also be best done through methanol; plants could be mostly converted between using natural gas and biomass.


Plants need to make 1M+ tons of methanol a year for good economics. Methanol is cheap, so you don't want to truck it around, so you want to integrate conversion. Then if you make it into propylene, you probably want to integrate polypropylene production. So, preferably, a single plant would be billions of dollars. $100B in investment in methanol production and conversion seems justified.


"But what if you make too much polypropylene?"

You can't. The potential demand from polypropylene siding, polypropylene pallets, glass fiber reinforced polypropylene stuff, EconCore, and so on is massive. And that's just one use for methanol - like carbon monoxide, it's one of those chemicals that's practically an element. The only problem is coordination of production and usage. Sure, maybe you don't like increasing plastic usage, but even in that case, believe me, you'd rather have polypropylene in your stuff than PVC.


back to index