Why is there almost nothing on the left hand side of the USA? Water scarcity!

We’re missing 300 million Americans. We’re missing  30 global cities west of 100 degrees longitude. We should do something about it!

The western US is a parched opportunity to create millions of acres of prime land for the next billion Americans to live on. Only one ingredient is missing – water.

“Cadillac Desert” (1986) by Marc Reisner correctly pointed out that within the limits of natural precipitation, we’ve expanded habitation in the West close to its maximal extent. Nearly 40 years after he wrote, however, the answer to shrinking flows of the Colorado and ever more demand for living space is not to stage some kind of retreat from land otherwise blessed with climate, solar power potential, mineral and human capital wealth. The answer is to flex our industrial might and finish what the irrigators began a century ago, and bring water in vast quantities to the high desert, to terraform a few select valleys in Nevada, and build a 21st century aesthetic vision.

We’ve already Terraformed California and Florida. 63 million people live in sparkling prosperous modern metropolises that were formerly uninhabitable swamps, within living memory. How did we do this? Large scale infrastructure projects that moved natural resources, principally water, from one place to another. 

March 1938 flood in Los Angeles.

In Los Angeles, we channelized the rivers and lowered the water table, creating 284,000 acres on which we’ve built one of the most dynamic cities on Earth. At the same time, we built the LA aqueduct, the Colorado River aqueduct, and the California aqueduct to bring in millions of acre feet of water every year to serve agriculture, homes, and businesses. 

In Florida, a combination of development, drainage, and air conditioning created one of the most desirable cities on Earth from a previously pestilential swamp. 

As a people, we need to ask ourselves whether we have an aesthetic sensibility. Do we need to curate landscapes? Cultivate life? Create beauty? Do we want to continue our historical pattern of striving to build better towns, cities, and opportunities for the coming generation?

My sense is that the answer to these questions is yes, of course, but that water scarcity in the American West blocks the sort of truly ambitious development we would otherwise like to see. 

We should solve that problem.

The key to solving water scarcity at the scale necessary to terraform Nevada is not to build bigger dams or to resort to rain dances, but to make more water using solar powered desalination technology.

In this post two years ago I explained how cheap solar unlocks usefully cheap desalinated water, with current (2024) state-of-the-art plants producing at just $0.40/m^3 – and radical further improvements are within reach. 

In this post a year ago I showed how a desalination project in the Imperial Valley could not only double southern California’s water availability but also jump start an enormous brine mineral extraction industry, convert the Salton Sea environmental catastrophe into a warmer version of Lake Tahoe, and pay for itself in short order. 

Both of these posts were based on using solar+batteries+reverse osmosis desalination technology (RO) to produce cheap water – with no technical risk and almost no market risk. I firmly believe regulatory risk is a solvable problem, once the participants in the usual zero sum water disputes realize that there’s an easy path to unconditional future abundance. We can radically expand the pie, everyone can have a big slice!

In the previous posts I baselined solar power at $0.50/W and batteries at $100/kWh, together with RO at $2000/kW. Amortized over 10 years, these systems produce water at around $500/acre-foot, or $0.40/m^3. This is a great price for municipal supply and even for the farming of certain high value cash crops. Indeed, the Federal Government now pays $521/af for water use avoided! Matching California’s annual 5 maf extraction from the Colorado would cost $2.5b per year, a tiny fraction of the economic productivity enabled by this water but a significant premium on the price California pays for the first 5 million acre-feet per year! 

We should be able to do better than this. Solar is getting cheaper.

Indeed, solar PV is the first mass produced product where energy is an output rather than an input. We have literally invented a way to produce more and more energy at ever lower prices, which are currently falling at 15-20% per year. Earlier this week, we saw modules changing hands for the record low price of $0.07/W! Solar modules are commodities, inert slabs of glass that you put on the ground and which spit out wealth. If we can maintain similar cost curves on racking and installation systems, we’re within striking distance of $0.10/W for a large scale array.

At the same time, and for the same reason, batteries have dropped in cost too.

I will have to update this chart too, since earlier this week we saw LFP cells changing hands for less than $50/kWh. 

Scaling these advances to the Salton-Imperial project, we can cut the cost of the array from $10b to $2b, the battery storage system from $12b to $6b, while the intervening (and rather modest) 5% RO cost reduction brings that part from $10b to $9.5b, for a total of just $17.5b, compared to $32b a year ago. Feeding forward, that reduces water cost to $280/af or $0.22/m^3, cheap enough that you could almost double your irrigation while still cutting legacy extraction to zero and pocketing the $521/af from the feds. 

At the same time, more than half of the residual cost is just the amortized RO system. If we want to go even lower, we’re going to have to adapt to the solar supply curve. In this post earlier this year, I explained how adapting legacy industrial processes for lower capex and intermittency can capture the upside of future solar PV cost reductions more effectively than relying on solar and batteries. As a rough rule of thumb, an even split of capex between the solar array and the application, collocated in some off-grid distributed plant, minimizes product cost for customers. 

So, for a $200k 1 MW solar plant, a $200k desalination plant that operates at 25% utilization and 3 kWh/m^3 produces 7.3 million cubic meters over 10 years at an amortized cost of just $0.055/m^3 or $68/af. 

Current RO plants cost more like $2000/kW, so they’re both financially and technically unsuited to intermittent operation, which fatigues their membranes. Thermal desalination could achieve radically lower cost, albeit at lower energy efficiencies, so there’s work to be done here designing new, low cost desalination machines that fully exploit the upside of cheap solar PV. 

In particular, a multi-effect distillation apparatus could be produced from injection molded low cost plastics in enormous volumes and at minimum practicable cost. Multi-effect distillation uses a series of heat exchanges to get more distillation bang from the high latent heat of vaporization buck. 

In addition to cheaper desalination, terraforming of the Nevada high desert or similar landscapes will require pumping water uphill. Fortunately, once again solar PV can provide the necessary low cost energy. Pumping one million acre feet of water per year 1000 m uphill requires a 6 GW solar PV array plus the usual low cost, high efficiency water pumps. Pumping water doesn’t involve nasty thermodynamic transitions so is relatively easy to perform. Depending on desalination efficiency, pumping water 1000 m uphill roughly doubles the energy cost, with a lower incremental increase in price due to the relatively low cost of pumps vs desalination equipment. 

Let’s Terraform Nevada

Nevada is a fascinating place. 71 million acres, of which 80% is federally managed land. Only three million people, of which almost all live in Las Vegas, a city which depends on the Colorado River for its water. Reno, at the other end of the state, houses just over half a million people. In contrast, Ohio is roughly one third the size and has nearly four times as many people. Unlike Ohio, Nevada receives so little rain that it lacks rivers with significant perennial flow. Water-induced erosion is so minimal that the geological formation of the basin and range geography has fragmented the drainage into a few dozen endorheic basins. In the battle between mountain formation and water erosion, mountain formation is winning handily. 

This wasn’t always the case. During the last ice age, only 10,000 years ago, Nevada enjoyed a much wetter climate, with numerous lakes, rivers, forests, and even North American megafauna such as mammoths, mastodons, and giant sloths. 

I’m hardly the first to observe that cheap energy can radically transform Nevada, so here I’ll try to be more constructive. How do we do this?

We use cheap solar PV and, ideally, cheap desalination technology to generate millions of acre feet of water on the coast. We move the water inland. We release it onto the landscape to substitute for a lack of natural rainfall. 

Let’s get specific. 

First, we have to get fresh desalinated water to Nevada, a landlocked state. There are a number of ways this could be accomplished, depending on aesthetic preferences. 

By far the simplest and easiest is to swap water desalinated in the Colorado River delta for water otherwise released from Lake Meade, which lies at altitude adjacent to Las Vegas. Five million acre feet per year generated either in Mexico or the US part of the Imperial Valley can feed the All American Canal, the Coachella Canal, and thence the Colorado River Aqueduct. As part of a water deal, Nevada can lease California’s water rights to Colorado River water and divert it on a loop or two through the state before returning flow to the river.

The lower Colorado River has three reservoirs, through which coastal desalinated fresh water could even be pumped uphill to Lake Meade, depending on seasonal demand and compensating for future variability in flows and releases from the Upper Colorado river. 

This is not the only way to get water to Nevada. The next easiest is to restore the Mojave River, a now dry/underground watercourse that runs from the Cajon Pass near San Bernardino north and east towards Death Valley, ultimately terminating at Badwater. In this case, however, flow would be diverted upstream of the National Park, then pumped up the Amargosa River valley across the border into Nevada.

This approach, instead of or in combination with a canal from Lake Meade skirting Las Vegas, brings water into the south west corner of Nevada, still at relatively low altitude. 

The upstream end of the Mojave river can easily be fed by the California Aqueduct, which brings water from the Sacramento Delta south to Los Angeles. This flow would be augmented by significant desalination anywhere on the California coast. It could even be in the LA basin – the land requirement for 20 GW of distributed solar is hardly onerous. For one example, an array adjacent to the I5 between San Clemente and Camp Pendleton – all undeveloped land – would be more than sufficient. 

If the desalination is occurring in the vicinity of the Sacramento delta, it may even be possible to pump water over the Sierras, then let it flow down the Truckee, Carson, or Walker rivers.

In some ways, the upstream instantiation of Nevada’s water supply is an aesthetic choice: Would you rather deal with California regulators or Mexican regulators? Fortunately, both are aligned on the broader political necessity to increase the supply of low cost water in essentially unlimited quantities. The best approach would be to set up a situation where sensible rents cause regulatory competition for deployment, not unlike early railroad siting. Save some money by skipping sealant on canals to ensure everyone along the route gets some free water, and all is well. 

Routing within Nevada

Geographically, Nevada’s valleys run primarily north-south. There is an east-west trending divide in the middle of the state, with better ordered north-flowing drainage feeding the Humboldt river, which in turn brings water to the Carson Sink in the western part of the state, near Reno. Southern drainage is mostly endorheic, though the White River (usually dry) drains south along the eastern edge of the state, ultimately into the Colorado. 

Here’s a lovely map showing the various watersheds in the American West. The hodgepodge of colors in the middle reflect that Nevada’s valleys are mostly unconnected, at least for now. 

The low cost of PV affords considerable flexibility in routing of canals and choice in which valleys are terraformed and which may be left in their inhospitable desert state. It’s quite fun to peruse maps and connect the dots.

One effective approach is to build a canal network that parallels US Route 95. This route has logistic access and relatively benign terrain. It would pump water uphill from Las Vegas, Amargosa Valley, and the Carson Sink to Tonopah, and from there to a flooded playa lake in Big Smoky Valley. 

From here it can be pumped over ranges into the Reese Valley and/or Monitor Valley, both of which ultimately drain north to the Humboldt river, bolstering existing agriculture in these valleys and closing the loop on the western side. The Monitor Valley drainage passes through Diamond Valley, refilling a large lake which overflows into Huntington Creek and thence to the Humboldt river. 

For water return to the Colorado River, water from Diamond Valley Lake can be diverted south via Eureka following the path of US Route 50 through Hiwee Valley, Newark Valley, and Jakes Valley into the White River watershed. In this scheme, the White River would also become a generous perennial river as opposed to a mostly dry valley. 

The complete system, roughly mapped, could look like this. But infinite variations are possible – what is your ideal?

About 500 miles of canals (red) feed just over 1000 miles of natural drainages (black), creating more than 750 square miles of new directly irrigable agricultural land. Up to 1460 square miles of new lakes, depending mostly on how much water we leave in the Carson Sink, and 240,000 acres of prime waterfront real estate. Ultimate water consumption through evaporation and ground water recharge would be up to 3 maf per year, and commercially significant brines in some of the sinks may enable mineral development in addition to agriculture, commerce, and real estate. In all, over a trillion dollars of land value appreciation alone – an enormous windfall in a state with an annual GDP of about $200b and relatively little economic complexity.

Not to disturb desert lovers (myself included) too much – more than 90% of the state, including all existing National Parks, State Parks, and military bases, would remain unaffected by this project. There is plenty of empty land to go around. 

How much would it cost? Baselining on the CAP and the earlier analysis of low cost desalination, I project:

  • $4b for a 20 GW solar desal array
  • $4b for a matched low cost desal plant
  • $6b for canal construction, based on the CAP in Arizona and adjusting for scale.
  • $2b total for pumps and solar arrays to power them

In all, ~$16b to restore a huge swath of the Pleistocene ecosystem, with over $1t of realizable value generation, in addition to ongoing revenue from development. 

There are other possible approaches, but it seems to me that the simplest way to fund this sort of development is to grant options to purchase small pieces of state and federal land at unimproved prices, in much the same way that railway infrastructure was funded in the US 150 years ago, along with the necessary regulatory permissions. 

This model generalizes

Here we’ve looked at Nevada, and previously we’ve looked at the southern edge of California. I think the observation applies generally. Something like a third of Earth’s land surface area is very sparsely inhabited desert, much of it coastal and low lying, unlike Nevada. Arizona, Utah, New Mexico, the eastern Cascades, much of India, Australia, Namibia, Chile, Argentina, the Middle East, Central Asia, and Saharan Africa could be, with the use of less than 1% of its land for solar desalination, converted into an arbitrarily fertile paradise. Israel, Saudi Arabia, and the UAE are well on the way, but the first solar desal megaproject awaits development. It could be you! Go forth!