It’s the question every water presentation eventually gets, usually from the sharpest person in the room: we built pipelines that move crude oil from Alberta to the Gulf of Mexico. The Mississippi floods on schedule. Phoenix does not have enough water. Why hasn’t anyone connected these facts with a pipe?
It’s a completely fair question, and the answer isn’t “nobody thought of it.” Plenty of people thought of it. Some of them had budgets. The answer is that water fails the pipeline test in four separate ways — volume, value, physics, and politics — and any one of the four would be fatal on its own.
First, the honest concession: America absolutely pipes water. Colorado pushes roughly half a million acre-feet a year through tunnels bored under the Continental Divide to keep the Front Range drinking†. California’s State Water Project moves water 700 miles from the wet north to the dry south. The Central Arizona Project lifts Colorado River water nearly 2,900 vertical feet across the desert to Phoenix and Tucson†. The engineering is proven. The question is why it stops at the regional scale — why nobody runs the pipe from the Mississippi to Lake Mead.
Start with the number that breaks the whole analogy. The United States burns through about 20 million barrels of petroleum a day — roughly 840 million gallons. It withdraws around 322 billion gallons of water a day†. Not consumption, to be fair — power plants return most of what they borrow — but even trimmed to true consumptive use, water demand runs hundreds of times larger than oil demand. Same pumps, same steel, nearly 400 times the cargo.
Here’s what that means in hardware. The Keystone XL pipeline — the one that dominated national politics for a decade — was designed to carry about 830,000 barrels a day. Run it full-tilt for a year and it moves roughly 39,000 acre-feet†. The Central Arizona Project canal delivers about 1.5 million acre-feet a year. So one modest regional canal does the work of roughly 38 Keystone XLs. And the CAP is a straw in the drink: to move the Colorado River’s entire modern annual flow — the 12–13 MAF that seven states and Mexico fight over — you’d need on the order of 300-plus Keystone XLs running around the clock†.
Nobody builds three hundred of anything.
Now look at what’s inside the pipe.
| Crude oil | Raw western water | |
|---|---|---|
| Rough value per gallon | ~$1.70† | fractions of a cent† |
| Weight per gallon | ~7.2 lbs | 8.34 lbs |
| What the U.S. moves daily | ~840M gallons | ~322B gallons withdrawn† |
An acre-foot of agricultural water in much of the West trades for somewhere between $50 and a few hundred dollars† — call it hundredths of a penny per gallon. Crude runs a few thousand times that. And water is the heavier cargo. A water pipeline is the worst freight contract imaginable: pay more per pound to haul something worth a thousand times less, in quantities four hundred times larger. Oil pipelines pencil because the cargo pays its own way. Water can’t cover its own shipping.
Water is only cheap while it’s standing still. The moment you lift it, it becomes an electricity bill with a river attached.
The California State Water Project’s Edmonston Pumping Plant heaves water 1,926 feet over the Tehachapi Mountains in a single lift — the highest of its kind in the world† — and the project as a whole is routinely cited as the largest single electricity user in California†. The Central Arizona Project’s pumping load was, for decades, the main reason an entire coal-fired power plant existed†. Those are the regional projects, the success stories. A Mississippi-to-Southwest line would face the Rockies: a mile of vertical against 8.34 pounds per gallon, times billions of gallons, forever. Every serious study of even the modest versions — a Missouri River aqueduct to western Kansas, for instance — lands in the tens of billions to build with pumping bills that never, ever stop†.
Suppose you solved all of that. You still have to ask someone for their river, and the answer is already written down.
The Great Lakes hold roughly a fifth of the world’s fresh surface water, and in 2008 the eight surrounding states — with Ontario and Québec alongside — signed the Great Lakes Compact, which effectively bans large diversions out of the basin. It is one of the few things Michigan and Ohio agree on. Wet regions have watched the Colorado River’s century of litigation and drawn the obvious lesson: once your water is in someone else’s pipe, it has a constituency, and that constituency votes.
The cautionary tale here is a real one. In the 1960s, an engineering firm proposed NAWAPA — the North American Water and Power Alliance — a continental replumbing that would have dammed British Columbia’s Rocky Mountain Trench into a 500-mile reservoir and sent Alaskan and Canadian rivers south to the American West†. It was studied seriously, priced in the hundreds of billions of 1960s dollars, and died of every cause listed above at once. British Columbia eventually did flood a stretch of that trench — with the W.A.C. Bennett Dam, for its own power, keeping every drop. Canada answered the continental-pipeline question fifty years ago, and the answer was no.
And still the idea won’t die: as recently as 2021, Arizona’s legislature formally asked Congress to study piping Mississippi floodwater west†. Every drought resurrects it. Every cost estimate reburies it.
Here’s the twist ending. America does move water from wet regions to dry ones, continuously, at continental scale — it just doesn’t move it as water.
It moves it as alfalfa, almonds, lettuce, and beef. Economists call it virtual water: the thousands of gallons embedded in every crop and carcass†. A truckload of hay is, hydrologically speaking, a small water shipment that loaded itself, and shipping the crop is radically cheaper than shipping the water that grew it. The transcontinental water pipeline exists. It’s the interstate highway system, and it runs on diesel instead of pumps. There’s an old saying in the West that water flows uphill toward money — and it turns out the cheapest way uphill is in a bale.
So the question isn’t really “why don’t we pipe water like oil.” It’s the Ledger’s question in a new costume: there are no solutions, only tradeoffs. Moving water costs energy, ecosystems, treaties, and votes — and against every version of that bill, the alternatives win. Use less. Price it honestly. Grow the thirsty crops nearer the rain. Every compact in the Headwaters Ledger is what a hundred years of that math looks like, written down and signed.