March 21, 2015
Suppose US West Gets A Mega-Drought
If a mega-drought lasting decades comes to pass in the American and Canadian West what should we do about it?
According to Cook, the current likelihood of a megadrought, a drought lasting more than three decades, is 12 percent. If greenhouse gas emissions stop increasing in the mid-21st century, Cook and his colleagues project the likelihood of megadrought to reach more than 60 percent.
However, if greenhouse gas emissions continue to increase along current trajectories throughout the 21st century, there is an 80 percent likelihood of a decades-long megadrought in the Southwest and Central Plains between the years 2050 and 2099.
Could we build a a large fleet of nuclear power plants to desalinate water and pump it several hundred miles inland? Picture massive amounts of water pumped to the head waters of the
Missouri river and rivers that flow into it such as the Milk, Yellowstone, North Platte, and South Platt? Then also pump massive quantities to the head waters of the Columbia, Spokane, Colorado, and Arkansas? Also to the inland valleys of California?
What is at stake: The plains bread basket farm fields. The viability of many states as places to live with all the infrastructure, homes, office buildings, and factories that have been built up in them. One has to compare all that sunken cost to the cost of bringing in enough water to supply enough for agriculture, industry, and residential living.
This could be the future of the American West:
So how to do a rough back-of-the envelope calculation of costs?Starting with the headwaters of the Missouri to bring, say, 20,000 cubic feet per second (cfp) of desal water to western Montana from the Washington State Pacific coast. What's the energy cost first of all? There is the energy cost of desalination. Then there is the energy cost of pumping the water uphill and
over 600 miles to around Three Forks Montana. Possibly the water would get pumped to multiple locations in Montana for local uses and to feed different tributaries.
Anyone know how to do rough calculations on this?
Randall Parker, 2015 March 21 05:28 PM
The answer to this problem is large diam low pressure pipes layed on the seafloor which could deliver copious freshwater from river estuaries in North California or Oregon down to South California and thence up dry river courses. A 10m diam pipe sending water at walking speed would pump over 2 billion cubic meters a year. One could imagine vast arid areas being irrigated this way at much lower cost than desalination.
The answer to this problem is large diam low pressure pipes layed on the seafloor which could deliver copious freshwater from river estuaries in North California or Oregon down to South California and thence up dry river courses. A 10m diam pipe sending water at walking speed would pump over 2 billion cubic meters a year. One could imagine vast arid areas being irrigated this way at much lower cost than desalination. "
But we still need to make some calculations to determine how much energy is needed to maintain a low pressure in the region of the large diameter pipe at the bottom of the ocean. This is because even if we just install that low pressure pipe at the bottom of the ocean, once the external pressure on the filter fills that pipe with salt-free water, the relative pressure will gradually decline and become equalized, which means that in order for the process to continue, there must be a powerful pump at the bottom of the ocean to send ALL the pure water to the surface level, to restore the relative low pressure inside the lowest parts of the pipe at the bottom of the ocean.
To be exact, referring to the Wikipedia article about the classical reverse osmosis method for separating pure water from sea water, we first note that the pressure differential between the interior and exterior of the pipe at the bottom of the sea must be 600-1200 psi:
Armed with the knowledge that the relative pressure exerted on the exterior of the pipe must be approximately 1,000 psi, we must then calculate the required depth at which the pipe must be installed at the bottom of the sea. The following website actually gives a chart to calculate the depth as a function of the pressure, and it reveals that the required depth must be approximately 700 meters:
And now we must calculate the amount of energy that is required to pump ALL the purified water back to the surface in such a way that the pressure at the bottom is zero. You see, it is not enough to wait until the pressure at the bottom of the sea to cause the pipe to get filled with pure water, because this just won't happen: in fact, only a few meters of water will be filled at the bottom of the pipe until the process stops because this filling process actually reduces the pressure differential by raising the pressure inside the bottom portion of the pipe! This means that ALL the water at the bottom of the pipe must be pumped all the way to the surface of the sea, or else the required relative pressure at the interior of the pipe near the bottom of the sea will not remain zero.
Now we can also calculate the amount of energy that is needed to pump one liter of water to the surface.
This energy is E = m*g*h, where m = mass of 1 liter of water = 1 kg, g = 9.8 m*s^(-2), and h = 700 meters.
Thus the necessary energy to pump back 1 liter of water to the surface is approximately 1*10*700 =7,000 Joules.
And this is just for one liter of water.
Thus it appears that the amount of energy that is needed, is the same as what a traditional reverse osmosis plant that uses at the surface of the sea on the shore of the continent. This means that putting the pipe at the bottom of the ocean will not make a difference, the needed energy is the same for reverse osmosis.
That being said, apparently the new Dutch technology that is mentioned in the quoted article above, is capable of using a lot less energy than the usual reverse osmosis factories. So there is plenty of hope, but it won't be as free as putting the pipes at the bottom of the sea. But if the new thorium reactors are developed, this will make a big difference since the energy will be far cheaper than coal and it will be practically inexhaustible and remarkably clean and safe.
You are on the west coast, so it may be hard to appreciate, but fresh water is absolutely, gluttonously abundant east of the Mississipi and in the Great Lakes region. It is now mud season across New England and in the Great Lakes area as it is every year.
The Great Lakes hold utterly monumental amounts of fresh water. The amount of fresh water pouring into the Gulf of Mexico from the mighty Mississipi every minute is impossible to imagine.
Gordon: Actually I must admit that the large diameter pipes on the seafloor would work, but the depth must be significantly more than the minimum reverse osmosis pressure level, so that this excess pressure overcomes the additional cost of raising the purified level to the surface of the sea.
If the required reverse osmosis depth is approximately 700 meters (let's round this to 1000 meters to make it more reliable), then it is not enough to lower the pipe to 1,000 meters, the pipe must be lowered to at least 2,000 meters. so that the rising water level towards the top does not cause the pressure differential at the bottom to drop below the reverse osmosis level, or else reverse osmosis will actually stop. Probably 3,000 meters of depth might be required to make it workable.
For many countries this might be a difficult project, but the Pacific Ocean does have enough depth, and I am sure some investors would take the risk and invest in this long term project. Once it is built, the maintenance should be affordable.
Here are some figures to chew on:
Lake Mead, a big supplier of the west, has 4 trillion gallons in it right now (half full).
The Great Lakes hold 6 quadrillion gallons.
The St. Lawrence River flows at around 150 billion gallons a day *out* of the Great Lakes. In other words the Great Lakes leakage could, in a month, totally restock the west.
The Mississippi empties around 200 billion gallons a day.
But I think the entire problem could be solved by just one easier waterworks project. If the Columbia River (100 billion gallons a day) were linked in to the California aqueduct system, you could solve this with a works project no bigger than the original California aqueduct system (and the rest of the west would see relief since California would stop draining the Colorado River).
On the other hand, it would probably be easier overall to run an aqueduct all the way from the Mississippi River than overcome SWPL political resistance in Portlandia.
In 2005, with a US population of 295.5 million, public and domestic water use was 48030 million gallons per day (United States Geological Service website). This excludes irrigation and industrial uses and amounts to 163 gallons per day per person. A reverse osmosis desalination plant currently being built in the San Diego area will supply water at about $2100 per acre foot, or about $6.50 per 1000 gallons. The cost of desalination to supply cities water then is just a little over $1 per day per person. To this must be added the costs of any pipeline system to get water from the coast to inland destinations. Keeping cities alive in a drought looks viable, but there is no water for crop growing other than what can be reclaimed from sewage water.
Actually it seems that what I said is wrong again: it won't work because even if the pipe is lowered all the way to the bottom of the ocean to levels below 7,000 meters, if the purified water were to fill the pipe all the way to the surface, then the pressure differential at the bottom of the pipe would be zero, and reverse osmosis would stop. So this means that the cost of pumping the water to the top is comparable to using a reverse osmosis factory at the surface.
Whoever made the odds calculations has no background in climate science. Hotter environment is associated with wetter environments. The Holocene climatic optimum was warmer than it is now and had a wet North Africa. If the world continues to warm the odds of this draught drop. If it cools then the odds go up.
For you a thought experiment; what if we used a fleet of nuclear plants to send very large amounts of steam/water vapor into the atmosphere off the west coast? Where would that precipitate out? How much would that warm the planet?
Dan, you are correct about the Columbia River. This idea has been floated for 40+ years, and there is NO chance of it ever happening, solely due to politics. That was true decades ago, even more so now. I fear that the same could be true of projects to import water to CA from the East instead of the north. However, that aside, it would make much more sense to build giant pipelines and pumping stations than to engage in massive desalination. (UNLESS a more energy efficient desalination process were developed, and there is some hope of that.) They've had giant pumping projects in operation in CA for half a century, and it is just a matter of scaling up. They certainly have no trouble doing it with petrol lines, and this is easier than that. Going over the mountains is the energy-hungry part of it, but what goes up also goes down, and at least part of the energy can be recouped by running the water through generators on the downhill run. Realistically, this is mostly about CA and AZ. AZ can be supplied by going south of the mountains, CA can be supplied by running lines to the crest of the Sierras. It's all gravity and existing water systems from there.
The eastern half of the US is doing just fine in the water department and since they (mostly) don't rely on stockpiling in reservoirs, droughts only last until the next wet spell. A large interconnected pipeline system could allow us to shift water diversion around according to which rivers were running high at any particular time, avoiding any burden on areas that were unusually dry. You might think it would be a pretty popular idea, given the potential job creation!
To put these numbers into perspective, 200 billion gallons a day (a la the Mississippi river each day) amounts to 500,000 acre feet. So in one day, it could supply a season's worth of water to +/-150,000 acres in the hot Central Valley. Let's say you took just 2% of the MS river flow; it could supply that in 50 days. All year long, that's over a million acres that could be fully irrigated (during winter, it would be stored in existing reservoirs and used in summer). Add the Great Lakes outflow to that, at the diversion same rate, and you approach 2 million irrigated acres. To put that in perspective, Fresno County alone has almost 2 million acres of farmed land, so it's not like it can water the whole West Coast. But still, that would water all the grapes and almonds there, two of the biggest cash crops in the state.
The rub is, as always, the cost. You're looking at $2000 per acre foot, most likely, if you amortize the cost of the construction over 10 years and assume similar cost to petrol pipelines. That's comparable to desalinization's cost. Amortize over 20 years, and it's $1000 an ac/ft. High, but similar to the rates paid now during the drought in CA.
Desalination and pumping seems like a rough way to go. It seems to me that the same energy would be better used to drive condensing plants wherever water is needed. Pound for pound, cooling water vapor is more efficient than heating water. And there's an enormous amount of water vapor in the air, even in the desert.
"Actually I must admit that the large diameter pipes on the seafloor would work, but the depth must be significantly more than the minimum reverse osmosis pressure level, so that this excess pressure overcomes the additional cost of raising the purified level to the surface of the sea....Actually it seems that what I said is wrong again: it won't work because even if the pipe is lowered all the way to the bottom of the ocean to levels below 7,000 meters, if the purified water were to fill the pipe all the way to the surface, then the pressure differential at the bottom of the pipe would be zero, and reverse osmosis would stop. So this means that the cost of pumping the water to the top is comparable to using a reverse osmosis factory at the surface. "
But wouldn't that only be true if you insist on using one continuous pipe from seafloor to shore? What if you dropped the desalination piping at 2000 meters, made it a self-contained system, then just tapped it at intervals into a separate pipe system which would then pump the fresh water to the surface? At worst, the desalination would stop during the tap, but restart when the tap was sealed.
A fair bit of what I do is organize financing and associated hedging for projects with both physical and capital markets risks.
There are reasonable statistics for average cost per cubic meter of water, both traditional and newer membrane methods. But all-in must include the new investment, operating, maintenance and depreciation cost. Transport would find comparables in things like existing oil pipelines, at least for order of magnitude.
Glad to share if that's of interest.
For a different solution that occurred to me, take a look at a global precipitation map (
map ). The heavy rains (13 mm/day) fall about 1,600 miles south of California. If you had an inflatable collection area 15 miles on a side in the 13 mm/day tropical rainfall band, it would collect enough water to provide 170 gallons/day/household for California, and without robbing any from land or the outflow of a river.
You just have to deliver it. To do that, a giant tanker that can carry 700,000 tons of water would supply a million households with 170 gallons each, so California would need 11 such tankers arriving per day. If they averaged 15 knots they’d have a 9 day round-trip time, so a fleet of one hundred tankers should do it.
The question is whether that would cost more than an equivalent desalination capacity.
"Actually it seems that what I said is wrong again: it won't work because even if the pipe is lowered all the way to the bottom of the ocean to levels below 7,000 meters, if the purified water were to fill the pipe all the way to the surface, then the pressure differential at the bottom of the pipe would be zero, and reverse osmosis would stop."
Not so, Wolf-dog: Fresh water isn't as dense as salt water. The pressure difference for that sort of reverse osmosis system to work isn't supplied by depth alone, but by the different pressure at a given depth between fresh and salt water.
It does require that the pipes be rigid, though, and not hoses, because they have to sustain the pressure difference.
I attended a conference on nanotechnology back in the 90's, where this type of system was discussed as a potential early use of nanotechnology. As it relies on very thin low resistance osmotic membranes.
I think that something like that would be very useful for the rest of us who are laymen and don't have that kind of knowledge.
AFAIK, water from the Great Lakes cannot be diverted because of a treaty or compact with Canada.
Please share. I am very interested to see.
My assumption was that diversion from Canada or Great Lakes would be blocked by politics. Hence the need for the West to get the water within the West/Pacific.
Also, mouth of Mississippi: Way less if the Missouri River isn't getting rain run-off and ditto the Arkansas. Picture a massive dust bowl drought several times bigger in area than the 1930s and far longer lasting.
$2000 per acre-foot: Any idea what that does to the cost per bushel of wheat or corn?
I'm concerned most of all about the plains states agriculture: Dakotas, Iowa, Nebraska, Kansas, etc. They would be hit hard by a huge Western drought.
Capturing the rains that fall on the ocean is a pretty radical idea. Practical? One would need ships to deliver it. That sounds more expensive than pipelines.
The idea of capturing fresh water just as it flows into the ocean is a very interesting idea that I've never heard before. It would eliminate desal cost. So then how much cost/energy to pump the water from the mouth of the Columbia back up hill to Montana?
An alternative: Send more water down the Ohio and then pump it up hill to the higher plains states.
Don't Believe Them,
Warming will cause changes in wind patterns and temperature gradients (water comes out of cooling air) and therefore shifts in where the rain comes down. Sure, more water will evaporate from the oceans. So more water will be in the atmosphere to come down. But it will come down in different places and at different times.
Also, just as higher temperatures cause more ocean evaporation they also cause more evaporation of moisture from soil.
I am not worrying about the city dwellers in coastal California. I am worrying about the plains states. That's where most of our calories come from. Hence my emphasis on feeding the Mighty Missouri.
Why not do what dozens of pre-Columbian civilisations have done previously? Depopulate the West until a steady state [proportional to rainfall] is achieved.
Commenters seem to assume that the solution to the problem is to maintain the supply of water in more expensive ways. But how much water people use isn't a fact written on stone, it's a function of what water costs. I predict that at a price well below the cost of at least some of the solutions being suggested, farmers would shift to less water intensive technologies, water intensive industry would move to wetter places, and total quantity of water consumed would drop by a factor of several.
That's a huge land area with huge infrastructure to abandon. Wherever people moved to would become a lot more densely populated too.
In event of a decades long massive drought over most of the West I certainly expect people would move, industries would move, agriculture would change and move. Agriculture can't switch to less water intensive crops is there is no water at all though. The plains states have the bulk of our grain crops because they are such a large, sparsely populated, but wet enough area. Where in the United States could wheat, corn, and soy farmers move to?
Even if huge water projects were undertaken they'd only replace a portion of what ceased to fall from the sky. So huge adjustments would be necessary in any case. But could massive scale engineering projects reduce the size of human adjustment in an affordable way? I still do not know.
On the west coast there are many locations ripe for the Atmospheric Vortex Engine (AVE). These could easily be driven by nuke plants, or even solar power, and would put enormous amounts of water high up in the atmosphere.
Water high in the atmosphere is a positive climate forcing. High clouds trap IR and, as they're cold, radiate poorly to space.
Yes, that's true. I'm not sure what the magnitude of the effects might be, but the rain might be pretty valuable. It's an area ripe for experimentation. A little data here would certainly clarify the picture.
High clouds don't generate rain. Low ones do.
If water gets injected into the air off Pacific Coast where prevailing winds blow inland toward tall tall mountain rangs my expectation is the water will come down on the mountain sides. So in that case I'm not expecting much climate forcing.
Getting water into the atmosphere to the east of the Rocky Mountains to come down to the east seems harder. But heavier rains falling on the western slopes of the Rockies would supply a lot of states and at least put water closer to the east side.
I think David Friedman is mostly right: water is badly underpriced, so it's incredibly mis-allocated. Household water only accounts for about 4% of consumption: 80% is farming, and 80% of the remainder is...watering the lawn!!
So, we can cut out the lawns without losing much. Cities can recycle household water (Las Vegas does it!), and it's only 4% anyway.
Agricultural water is mostly free to farmers, governed by archaic water rights systems. So,we have rice in the desert, almonds, etc.
Just correctly pricing water will fix 90% of water problems.
Now, what do we do with a megadrought over all of the US? If we want to consider massive scale engineering projects, I'd look first at enclosed greenhouse type farming projects. If they really worked, they'd fix the problem once and for all, by eliminating the dependence on the vagaries of weather and climate.
If water gets injected into the air off Pacific Coast where prevailing winds blow inland toward tall tall mountain rangs my expectation is the water will come down on the mountain sides.
Interesting notion. A seawater-pumping wind turbine which only has to get it up to the hub (after which spraying is done by centrifugal pressure head) could be very simple and possibly cheap; that would let you raise the offshore RH and drop the air temperature somewhat. What sort of sea-surface temperatures, air temperatures and absolute/relative humidity do you have to work with, especially inland?
Creating a low, dense mist of seawater in the wake of an offshore wind farm would be a negative climate forcing. It might also be a hazard to navigation.
water is badly underpriced, so it's incredibly mis-allocated.
If farming is such a small fraction of California's economy, it's surprising that cities haven't bought out farms for their water rights yet.
A commenter on a more recent post about water said that farmers in CA don't generally have the right to sell their water - it's use it or lose it.