8

Bulk Water Transfers

Humans have been diverting water since civilization began, since they stopped hunting and gathering and invented farming, enabling them to settle into stable communities. In most of the places early civilizations began, such as Egypt, Mesopotamia, and China, secure food supplies needed irrigated water, since rainfall was either erratic or insufficient. This meant containing water, then moving it to where it was needed. Dams provided containment. Canals, qanāts (which are generally gravity-fed underground channels), aqueducts, and later, pipes, did the diverting. The technology may have been rudimentary, but the thinking of the early hydrological engineers was as sophisticated as it needed to be. The Romans, of course, were the pre-eminent engineers of antiquity, and their network of canals and aqueducts, many of which are still in use, has hardly been surpassed.

We’re bigger now, much more populous, and with access to energy undreamed of in antiquity, and our engineers have had greater dreams than were ever possible before. Which raises the question: We can dream big and build big, but should we?

In modern times, the three best-known large-scale, completed hydrological projects are the California canal and pipeline system, which takes water from the Sacramento and Colorado Rivers to Los Angeles and other drought-prone places; the Israeli National Water Carrier, which extracts water from Lake Kinneret (known outside Israel as the Sea of Galilee); and the late Libyan strongman Gadhafi’s Great Man-Made River, a series of gigantic pipelines that take fossil water from the Saharan aquifers to a new agricultural zone on the coast, a transport now fatally disrupted by the chaotic politics of Libya.

Modern California could hardly exist without bulk water transfers. The State Water Project and its 650-kilometre California Aqueduct is the largest engineering endeavour the world has yet seen, though it may be outdone, and soon, by the Chinese and the Indians. Every year, California shifts fifty-three trillion litres of water southward, capturing it behind twelve hundred dams on every river and stream of consequence before pumping it hundreds of kilometres, lifting it over some mountain ranges and under others, fitting the barren landscape with a massive caul of pipes, ditches, canals, and siphons that irrigates an agriculture that provides more than half of America’s nuts, fruits, and vegetables. Not incidentally, it also waters the lawns and powers the car washes of the Los Angeles metropolis, all this in a region that receives less than forty centimetres of rain a year, when it gets any at all. About half the water Los Angeles uses comes through the aqueduct from the overstressed and over-allocated Colorado; the other half, notoriously, comes from the 375-kilometre Los Angeles Aqueduct, which siphoned water from Owens Lake and destroyed picturesque Mono Lake, turning it into a saline wasteland. Still, if the Owens Valley was destroyed, it was not via rape but through consensual sex, as a hydrologist at University of California, San Diego put it: the local residents were happy enough at the time to sell out to whoever would buy.1

Nor could Israel exist in its present form without its National Water Carrier, which turned fifty years old in 2014. It is nowhere near the size of California’s water transfers, but it shifts around four hundred million cubic metres a year from Lake Kinneret to the cities of the coast and to the Negev, further south, through 134 kilometres of pipes, tunnels, reservoirs, and ditches. Among its consequences have been dropping water levels in the Kinneret itself, almost reaching crisis levels, and the ruination of the Jordan River and the Dead Sea. The decision a few decades ago to use the carrier to supply drinking water to the highly populated coastal cities made it a strategic asset, and Israeli security agencies have become very guarded about its capacity and reach. The construction of a series of desalination plants along the coast, now supplying almost all the country’s drinking water, has lessened the strain somewhat.

Libya’s Great Man-Made River was built to tap into the Saharan fossil aquifers and to take that water from deep in the desert to the coast, where new agricultural zones would be created — despite knowing full well that in thirty years, or fifty, the water would inevitably run out. Stage one, completed before Gadhafi’s overthrow, was a nineteen-hundred-kilometre waterway consisting of pipes large enough to drive a truck through, carrying two million tonnes of water to the coast every day. More than four-fifths of the water would be used to irrigate new farming zones, some of them in the desert itself. The cost was somewhere around $32 billion — but then, in Colonel Gadhafi’s day, oil revenues seemed even more endless than the Saharan water.

The two largest works-still-in-progress are India’s vast array of pipes and canals tying virtually all of the subcontinent’s rivers together into one massive interconnected network, and China’s startling notion to hydrologically re-engineer the entire country, moving massive quantities of water from the more or less sodden south to the increasingly arid north. The most ambitious not-yet-built system is the Red-Dead Canal, carrying sea water from the Gulf of Aqaba to the Dead Sea, and desalinating it along the way.

These six aren’t the only inter-basin water transfers that have been built or contemplated. Plenty of other pipelines thread across most continents, ranging from the functional to the bizarre. There is a new pipe from the Austrian Alps to southern Europe. Turkey’s former president, Halil Turgut Özal, once proposed what he called the Peace Pipeline, which would carry more than two billion cubic metres of water a year from the Ceyhan and Seyhan Rivers and essentially spread it through the Middle East and the Gulf. The cost, before the whole thing was shelved for its political improbability, was going to be around $20 billion. As previously mentioned in Chapter 6, the Congolese dictator Mobutu once proposed a pipeline carrying Congo water through Angola, where it was hardly needed, to Namibia, Botswana, and South Africa, where it was. So far, no one has proposed emptying the Guarani aquifer and piping its water to whoever needs it, mostly because, so far, no one does need it.

In addition, there are many ideas that Peter Gleick has called zombie projects, unlikely and usually grandiose projects that should never be built but stubbornly refuse to die. One of the most outlandish is to shut off San Francisco Bay and turn it into a freshwater reservoir, but there are plenty of others almost as grotesque: the notion of an undersea pipeline from the Amazon to North Africa; damming the Strait of Gibraltar to dry up the Mediterranean and “create” thousands of square kilometres of new land. The list includes turning north-flowing Siberian rivers (the Ob and the Lena) southward to the Volga basin or the Aral Sea; the crazed notion of turning James Bay into a freshwater reservoir and diverting its water southward into the Great Lakes; and the NAWAPA (North America Water and Power Alliance), which takes pride of place in Gleick’s chapter on zombies, taking water from British Columbia to pretty well everywhere, including California, the Mississippi, even the Great Lakes (again!).2

If California’s water transfer system is huge, what China is proposing is more massive yet. This is the $80 billion project to re-engineer the country’s fresh water, taking vast quantities from the lake-filled and often flooded south to the arid north, where much of the population lives and most of the agriculture is sited, and which has been rapidly running out of water. The North-South Water Diversion Project will subsume existing canals, among them the Grand Canal between Beijing and Hangzhou (finished around the year 500 CE). It will also require more than 3,000 kilometres of new canals, tunnels, and conduits, some of them climbing high elevations to cross the Himalayan Plateau, and in many cases having to dive beneath the rivers they must cross. The first, easternmost stage was completed in 2013, pumping fifteen billion cubic metres along 1,160 kilometres of canals, among them the Grand Canal. Most of the water pushed through the new system is highly polluted, even toxic, and must be cleaned before use. The midstream link, delayed by several years, saw its first water reach Beijing in December 2014, through another 1,300 kilometres of canals. The third (most ambitious and much-delayed westernmost) link, is the one supposed to climb to the Himalayan Plateau. Total water movement will be around forty-five billion cubic metres a year — a large number but a mere 7 percent of China’s supply, nowhere near enough to solve North China’s water problems. Perversely, although the capital is stressed for water (and its aging pipelines beginning to rot), water remains ridiculously cheap to consumers at less than a dollar per cubic metre.

Few outside China believe the project makes any sense, and it has many critics within China too. On the other hand, most of the politburo members are engineers, and the former president Hu Jintao, who championed the project, was a water engineer, which might explain China’s predilection for massive projects. Critics point to the ecological senselessness of pushing water to the north to grow food that is then shipped back to the south; they point to the bizarre notion that the Chinese government is planning half a dozen “new cities” of more than a million each, some of them in the Gobi desert; they point to the idiocy of a policy in which water is still cheap in areas where supplies are tight; they point to the ecological hazards of mass water transfers between two utterly different natural ecosystems; and they note that the Yangtze, from which most of the water will come, is polluted to a dangerous degree.3

Simultaneously, the Chinese are building a seawall along the coast. When it is finished, it will cover 60 percent of the total length of the country’s coastline and will be longer than the Great Wall. Their rationale is that the coastal region covers only about 13 percent of the country’s landmass but produces 60 percent of the GDP — and in this ecologically primitive view, all those coastal wetlands are just taking up space. The portions already constructed have caused a dramatic decline in internationally shared biodiversity and associated ecosystem services. Virtually all environmentalists and ecologists think the seawall a bad idea, but it is proceeding nonetheless.4

Not to be outdone in grandiosity, India is building a pipeline from the Tehri dam in the Himalayas to divert water to the upper reaches of the Ganges River and thus supply Delhi with drinking water. This is just a small part of a much larger and more ambitious scheme, which is to build a fifteen-thousand-kilometre network of tunnels and canals to shift 174 billion cubic metres of water a year from areas with surplus water, mostly in the north, to arid regions, mostly in the west and south, such as Karnataka and Tamil Nadu. There will be, altogether, thirty new links between Himalayan-fed rivers and rivers in other basins, and the whole thing will cost, when it is finished, somewhere around $170 billion. The imperative that is driving the project is growth: the need for economic growth to feed the ever-expanding population. The Indian government calculates that it needs to increase its irrigated agriculture from around 100 million hectares to 135 million in a few decades. The National River Linking Project, as it is called, will also help to minimize flooding in the monsoon season and add thirty-four gigawatts of hydro power to the grid.5

Another grand dream, this time in the Middle East, is the Red-Dead Conveyance, and its sometime variant, the Med-Dead Canal. If built as originally conceived, it would be a huge desalination plant that would replenish (or fatally alter) the Dead Sea and provide emission-free electricity to both Jordan and Israel. It was originally proposed by the American engineer Walter Loudermilk in the 1950s and has been received in both the Israeli and Jordanian capitals with enthusiasm that waxed and waned over the decades according to the politics of the time. Finally, in October 2013, Jordan green-lighted the project and two months later signed a water-sharing pact with Israel and the Palestinian Authority, prompted no doubt by a raft of studies suggesting that regional temperatures would sharply increase by century’s end and that rainfall could diminish by 30 percent or more.

But because this is the Holy Land and a place with a deep human history, the provenance of the project goes back a lot further than the modern era — at least, it does if you interpret scripture in a certain way. An American engineer, Randolph Gonce, a soil and water conservation specialist, became involved in plans to replenish the Dead Sea through an odd combination of familiarity with a Tennessee underground pumped-storage facility near Chattanooga called Raccoon Mountain, an interest in Chunnel engineering, news reports on the rapidly decreasing levels of the Dead Sea, and an interest in Holy Land history. He knew, for example, that King Hezekiah had built tunnels in the soft sandstone under Jerusalem to protect the Gihon Spring in the Kidron Valley, the city’s water supply, and he read of the Prophet Ezekiel’s curiously detailed description of a river flowing in a channel from beneath the Temple. Ezekiel measured the channel’s drop at one in one thousand, and then recounted how the Dead Sea would thereafter spring to life. Gonce was taken with the fact that a tunnel bored from the Mediterranean coast to the Dead Sea would plausibly have the same slope as Ezekiel’s one in one thousand, about the same drop as decent drainage for sodden agricultural fields in the modern era.

The Med-Dead tunnel, which Gonce fancifully called the Ezekiel Project, was briefly favoured by the Israelis over the Red Sea counterpart, partly because it would be entirely within state borders and it would be shorter, therefore cheaper. It would, in fact, be gravity-fed and self-sustaining after construction, unlike the Red-Dead variant. But the involvement of Jordan, the interest of international funding agencies like the World Bank, and the enticing prospect of rare political cooperation favoured the longer and more difficult version.6

The plan was to push a channel from the Gulf of Aqaba on the Red Sea to the Dead Sea, following more or less the Jordanian-Israeli border. To get there, it would have to be pumped up a full 200 metres, somewhere around Mount Seir, before beginning a 660-metre descent to the Dead Sea, widely known in the tourist brochures as the “lowest place on earth.” The conveyor would carry just short of a billion cubic metres of water. As it dropped, gravity would generate enough power to desalinate about 40 percent of its flow, the resulting saline brine flowing into the Dead Sea and lifting it back to levels last seen in the 1960s. The idea was that the power generated and the water produced would make the project economically self-sustaining.

Naturally, critics abound. The most vocal opponent is an NGO called Friends of the Earth Middle East, based in Amman. It protests that the impact of mixing Red Sea water into the Dead Sea is, at best, unknown, though some studies have shown that the newly invigorated Dead Sea would be prone to algae blooms and gypsum crystals with uncertain but probably negative consequences.

The deal actually signed at World Bank headquarters in 2014 is rather different and much less ambitious, though it still involves dumping ocean water into the Dead Sea. It involves a conventional desalination plant in Jordan, at Aqaba, producing about forty-eight million cubic metres of fresh water for both Israel and southern Jordan; the resulting brine would then be pumped 150 kilometres to the Dead Sea, again with a view to replenishing its water (or again, contaminating it, depending on your point of view). The agreement has two other political advantages. In return for fresh water from Aqaba, Israel agreed to provide Amman, the Jordanian capital, with somewhere around thirty million cubic metres of fresh water from Lake Kinneret, and the Palestinians would be entitled to buy an allocation at “preferential prices,” a phrase left awkwardly unspecified. Private companies would build the Aqaba plant at their own expense, recouping their investment through sale of the resulting fresh water. The brine pipeline would be financed by donor countries with a World Bank bridge loan. It would be located on the Jordan side of the border, thus lessening the possibility of environmentalist disruptions, as the Israelis cynically admit, Israel being more susceptible to environmentalist pressure than Jordan. Indeed, Friends of the Earth Middle East has already expressed reservations about the whole thing, and the tourist operators on the Dead Sea itself are grumpy about any change whatever.

Sometimes bulk water diversions are not physically possible or would so be so expensive that they simply price themselves out of possibility. A recent Alaskan governor, Walter Hickel, had the idea to take water from Alaska’s Copper River in a pipeline down the coast past parts of western Canada, Washington, and Oregon to southern California. When the estimated cost surged past $30 billion or so, the concept lost steam — desalination, though it was then still expensive, was cheaper than that.

Plenty of beguiling ideas have been put forward to overcome this “not possible” problem. Why not direct glacial melt to the Atacama Desert of Chile? Ship Alaska water in supertankers to San Diego? Use giant plastic bags to tow water? If you look at the literature, it sometimes seems that shifting large quantities of water from water-rich to water-starved places is the perpetual-motion machine of hydrology, attracting all kinds of dreamers, deluded entrepreneurs, and scam artists, my favourite being the Canadian dentist who secured rights to an Iceland glacier before the securities commission caught up with him and fined him a million bucks for fraud. Speaking of water running uselessly into the sea — the notion that icebergs calved off the Greenland glaciers slowly melt fresh water into the sea as they drift southward past Newfoundland has excited the avarice of more than one would-be entrepreneur into envisaging what Modern Farmer magazine sardonically called “the cold rush.”7

Why not lasso these wandering water farms and either chip them into usable chunks or tow them to someplace their meltwater would be useful? A Saudi prince, Mohammed al-Faisal, came up with this notion in the 1970s; it was he who figured he could take all those useless Antarctic icebergs and tow them up to Mecca. The study he sponsored proved conclusively that no iceberg would ever cross the equator — by the time it got past the tropics there wouldn’t be enough ice left to fill a martini glass, never mind a cistern. Even to get icebergs to South Africa, which at least is in the same hemisphere, would be a challenge: a four-month journey under some of the most extreme weather conditions on earth, and to what purpose? South Africa is water stressed, but not critically so. Namibia could use the water, if it could get up the coast another thousand kilometres or so, but who in Namibia could pay for it?

Northern Hemisphere to Northern Hemisphere towing might be feasible, but which places outside the tropics needs water that badly? Simulations have been run, for example, to see if towing icebergs from Newfoundland to the Canary Islands is feasible — but again, to what purpose and for what market? And how to deal with these behemoths when they reach their destination? Icebergs are big. They’d have to be mined and cut up into usable chunks, an effort that would probably cost somewhere around $10 a tonne, far more expensive than desalinating sea water. Moreover, icebergs are damnably unstable, likely to tip at any moment, putting lives at risk. And if you moored an iceberg in San Diego, to take one possible destination, you’d have to purify the water afterward because the berg would be contaminated by coastal pollution. Still, a Newfoundland company is already selling $100,000 “hair nets” that oil companies can use to tow icebergs away from drilling rigs, so the technology is feasible.

A second proposed method of getting estuary water to someplace useful would be to use superannuated oil tankers for transport. Hundreds of single-hulled tankers are being broken up in India’s scrapyards or are simply rotting away in situ, so why not refit them to transport water instead of oil? Single-hulled vessels, banned for oil because of the risk of spills, would not be a problem for water — a breach would simply dump fresh water into saltwater, something that would puzzle a few fish but would otherwise cause no harm. But the cost of cleaning up each ship for water transport proved a major problem, possibly costing as much as $6 million per ship, and it has never been proven to be completely successful, at whatever price. Benzene is not something you want in your drinking water.

Some attempts have been made to use non-oil tankers. A decade or so ago, Turkey struck a deal to take surplus Turkish water to Tel Aviv in tankers and built an extravagant water-fuelling depot at Manavgat to fill them. The trade never happened — Libya’s then strongman, Gadhafi, vigorously objected to anyone supplying Israel, and the resulting outcry killed its already remote chance of success. And in the dire Spanish drought a few years after the turn of the millennium, Barcelona was running so low on water that it hired tankers to bring water in from Marseilles. Only one made the trip. It dumped nineteen million litres of nice fresh water into the city’s distribution system — and it was all used up in less than an hour, proving to everyone’s satisfaction that tankering in water was a fool’s game. (The city was finally rescued by a break in the drought.)

Another idea, seemingly arrived at by a number of people more or less independently at more or less the same time, was huge plastic bags for liquid transport. Sometime in the 1950s — the exact date is unclear — a Cambridge scientist originated the idea and it was picked up by the Dunlop Rubber Company of Bristol that went on to make what it called the Dracone bag, after the Greek word for sea serpent. Dracone bags were tough, long, and narrow, and could be pulled through the water at a good rate of fifteen kilometres an hour, though at that speed they thrashed about a bit (hence the name). Some were actually deployed, mostly for military transport of fuel oil and to help clean oil spills. It wasn’t a big segue to try to do the same thing with fresh water, though at that time there were few accessible places that seemed to need the water that badly. The only real test of the technology to date has been the use of smaller (thirty-five thousand cubic metre) bags that for years have been chugging between Turkey and its arid province of Northern Cyprus. The operator, Nordic Water Supply, is planning to expand its operations to take water from Turkey to the United Arab Emirates.

One of the newer big-bag dreamers is businessman Terry Spragg, who earlier had joined al-Faisal to experiment with towing icebergs. He, along with everyone else who tried it, abandoned that, but water transport remained a passion. Lately, he has been trying to float, if that’s the right term, a company that would employ massive plastic bags to take water from places of plenty to places of none. His Spragg Bags are modular and can be zipped together in a train, and would be towed to their destination by tug. As a test, he wanted to fill a bag or two at Turkey’s Manavgat terminal and tow them down to Gaza, a notion simple enough in theory but surrounded by a thicket of prickly politics. As I write this, no test has been made.

Another outsize personality who has gravitated into the water bag business is Ric Davidge, usually described as “Alaska’s former water czar,” which he was, or as a “water mogul,” which he is trying to be. Once a combat medic in Vietnam, he has formidable political connections, having been an adviser to Ronald Reagan and to former Alaska governor Walter Hickel. After reinventing himself several times, Davidge ended up as president of an Alaska-headquartered consortium called WorldWater, whose partners included the shipping company Nippon Yusen Kaisha and the Abdul Latif Jameel group of Jeddah, Saudi Arabia. The company’s declared aim is to take water from Alaska, or wherever it is plentiful, to wherever it is needed, anywhere in the world. In any case, WorldWater has, as of 2014, yet to ship a single drop.

Of all the doers and dreamers, though, the most plausible idea remains that put forward by former oil-patch engineer James Cran, whose Medusa Corporation is headquartered in Calgary, Alberta. He has been working on the notion since the mid-1980s, when he conceived the idea of transporting water from the Columbia River estuary to San Francisco and Los Angeles. The Columbia, as he points out, is eight times the size of the Colorado, which remains the only out-of-state water supply for California, and even extracting six billion cubic metres a year would comprise only 3.7 percent of the Columbia’s flow, not enough to affect either fishermen or fish. At the time, he took out a patent for his version of the giant bag, but nothing came of it.

Cran’s Medusa bags — named after the jellyfish — would be massive if they ever get built, as much as 500 metres long and 150 wide, with a draft of 22 metres — they’d be as long as six football fields and carry somewhere more than a million cubic metres of water. They would be made of high-tensile polyester fibre; Cran calls them pillow tanks. They would carry fresh water at very slow speeds, somewhere around five and a half kilometres an hour, and because of their deep draft would have to be filled and emptied at offshore buoys. Medusa bags would stay afloat without help because fresh water is less dense than salt. Cran’s analysis is that they would perform best when 40 percent filled, at which point they would be largely impervious to wave action — a saltwater wave could cause an internal freshwater wave, which would then cause another saltwater wave on the other side in the same direction.

Medusa has already deployed a small 4,500-tonne bag in tests off Vancouver and is planning a larger test with a 300,000-cubic-metre bag equipped with strain gauges and other devices for metering performance. The California Department of Water Resources told Cran that the state’s annual ongoing water deficit, currently handled by drawing down groundwater, is around five billion cubic metres per year; he seems confident that Medusa bags could shift some 600,000 cubic metres a year by 2020, with a target of about 6 billion by 2030. More than 600 million cubic metres would require 25.5-million-tonne bags, or multiple trips by smaller numbers — substantial, but far from impossible. The source? Still the Columbia River. Whether he will get the chance remains to be seen.8

All of the major water transfer projects described at the start of this chapter were built without a thought for the ecological consequences — there really weren’t thought to be any ecological consequences. Certainly none of the people who conceived and constructed them, California included, balked in any way at moving water out of one water basin and into another. We needed these massive diversions, so we built them; it is hard, now, to make a real case that we are worse off.

This doesn’t mean we should do more of it. Or does it?

For better or worse, ecologists have only recently taken to examining the matter, and it is still imperfectly understood.

A fundamental principle about water management today, widely accepted by engineers and environmentalists, is that to manage global water supplies properly, we should treat the river basin as the core hydrological unit — river-basin integrity should be the first priority in supply-and-demand management. This recognition is really not political; many river basins cross international frontiers, and while there might be quarrels over allocations, the underlying principle is understood: for long-term health of the system, withdrawals must not endanger the ecological services the river provides to all riparians. This applies not just to surface water. Groundwater is equally affected and should be treated similarly. In fact, this notion of water-basin management has been enshrined in international water law, though not always honoured.

Managed in this way, water retains at least a simulacrum of natural balance. After all, in most water uses, for agriculture or even for municipal water and wastewater supply and treatment, the water is not actually used up, but put to some use and then returned to the hydrological cycle, not always in pristine condition but still there — and treatable. Exporting bulk water destroys this natural balance. So again, it is widely though not universally accepted that moving a billion or so cubic metres from, say, the upper or middle reaches of a river and fluming it across a divide into another, drier, basin has large and mostly negative consequences. Not just for those creatures downstream, now deprived of the water they had been used to, but also for the river and the basin itself. It can even be argued that such export has negative consequences for the receiving basin too, in that it encourages insecure development in places where it shouldn’t happen — Las Vegas, a thirsty city in a desert, is everyone’s favourite example, but Los Angeles and much of the irrigated agriculture of the Central Valley are others.

To come back to the example of the American Great Lakes, eyed enviously by thirsty places in many states and viewed with anxiety by nationalist protectionists across the border in Canada. As we saw in Chapter 3, the Great Lakes do indeed hold a great stock of fresh water, almost a quarter of the global available supply, but little of it is renewable: taking a tiny few percentage points more from the inflow will eventually cause the lakes to shrink, so it is the inflow that counts, not the volume of water stored. This is poorly understood, even by professional water managers who should know better. A pristine example is the “water czar” of Las Vegas, Patricia Mulroy, who to her credit has persuaded (and bullied) the city and its casinos into becoming admirably water thrifty. Still, when she was told that states bordering the Great Lakes (and not just on the Canadian side either) were balking at devising ways to send her the water she wanted, she exploded: “We take gold, we take uranium, we take natural gas from Texas to the rest of the country. We move oil from Alaska to Mexico. But they say, ‘I will not give you one drop of water!’ They’ve got 14 percent of the population of the United States, and 20 percent of the fresh water of the world — and no one can use it but them? ‘I might not need it. But I’m not sharing it!’ When did it become their water anyway? It’s nuts.”9 But it is not nuts. It is prudent and sensible. What is nuts is the city of Vegas itself, but that’s another story.

Considerably less nuts is when water is diverted from the upper or middle reaches of a water basin to lower down in the same basin. Manhattan’s water grid is an instructive example. Manhattan’s water delivery tunnels, begun in 1917, were only (almost) completed in 2013. This is mass water movement in spades. New York City’s nineteen reservoirs and three controlled lakes hold 2.2 trillion litres of water, spreading over eight New York State counties and stretching 150 kilometres up to the Catskill watershed. The whole thing is entirely gravity-fed, and the water can take three months to get to its final reservoir, Hillview in Yonkers, which itself holds only about a day’s supply. From that last reservoir, the water flows into underground aqueducts and tunnels, the last of which, prosaically known as Tunnel #3, will not be finally completed until 2021, though in 2013 the water it carried finally reached Manhattan. Until then, the city’s water supply was vulnerable, either to natural disasters or sabotage. Up to 2013, the five boroughs relied on Tunnels #1 and #2 for their drinking water, put into service in 1917 and 1936 respectively. Tunnel #3, at a construction cost of over five billion, finally provides New York with the critical redundancy it needs, and allows for repairs to be made on the other two — for the first time since they were constructed.

What if you plan to divert water from one basin to another, not from the upper reaches of a watercourse but at the point it empties into the sea? Do the usual strictures still apply? If you take, say, water from Alaska’s Copper River just as it empties into Prince William Sound, does that still cause harm? Or take water from the Columbia River’s estuary to San Diego down the coast, as James Cran is proposing? You can hear many a would-be exporter complaining about all that water running “uselessly” into the sea. The Alaskan town of Sitka, which has been trying unsuccessfully to export water from nearby Blue Lake for decades, is typical, offering all and sundry more than thirty-four billion litres a year and a cost of one cent per. True, Blue Lake is not the estuary exactly, but it is not far off, and Garry White, boss of Sitka’s economic development association, has argued to all comers that removing 8 percent of the watershed flow every year will not harm anything “because much of it is already being lost to the ocean” less than a mile away.10

But is it really “lost” to the ocean?

In some ways, the political and emotional arguments against bulk water exports are stronger than the ecological ones, often stronger than the facts warrant. It is not always easy to separate opinion from fact, to separate a dispassionate discussion of whether more water should be moved from the idea of water grabs by foreign entities taking “our” water without our permission, or under the sanction of a trade treaty like NAFTA. When I first started to write about water more than a decade ago, several companies were preparing to export the water of this or that lake (one in Ontario, up near Sault Ste. Marie, as I remember, another in Newfoundland) and sell it to thirsty customers. But it turns out that they really didn’t have any customers for tankers full of water, and in any case, the economics made no sense: water is heavy — the definition of a metric tonne, after all, is one cubic metre of water — but it is not yet very valuable, and whatever profits were envisaged would be wiped out well before the tanker got anywhere. Nevertheless, the issue caused a minor uproar in the Canadian Parliament, mostly over ill-founded fears that if we allow one company to commodify and sell some of our water, much larger companies with much larger budgets will come and take much larger quantities of our water away to, presumably, somewhere in the United States. Essentially, that’s what the Canadian fear came down to: that “they,” the Americans, would come and take “our” water from us. Of course, if the Americans really wanted to do something like that, they could dip into “their” side of the Great Lakes, since the national frontier runs pretty much through the middle of most of the lakes. What is stopping them is the existence of the International Joint Commission, a bilateral management body always trotted out, justifiably, as one of the most successful water-management institutions anywhere. What is stopping them also is that they don’t actually want to do it: US states that border the lakes are just as adamant about stopping massive extractions as any Canadian provinces are. In any case, under NAFTA, water — except for bottled water — is exempt from reciprocal trade rules, so no one can sue us for not letting them have our water. We’ll come back to this issue in Chapter 12.

The ecological arguments against bulk transport of estuary water are a good deal more elusive than is the notion of abstracting water upstream. One such argument is that altering outflows damages estuary ecosystems and destroys wildlife and fish habitat. This is self-evidently true. But, as I’ve mentioned before, using fish as a reason not to do something that would otherwise benefit a large number of people has run out of steam as an engager of public opinion and now threatens to become merely old-fashioned, as attention is turned to more urgent matters like global warming. A second argument is a little more persuasive: that such transfers permit and even encourage us to live beyond our natural means, make us eternally dependent on energy-intensive transfer industries, and allow us to ignore the virtues of conservation. A third argument is that, if done on a large enough scale in a large enough number of rivers, the ocean and its currents might be affected, with unknowable ecological and climatic consequences. This is the same worry that plagues the project to dam the rapids near the mouth of the Congo — the freshwater plume that travels a hundred kilometres out to sea is beneficial, or so it is supposed, but beneficial to what, and to what degree?

On that, we have only speculation and worry, not data. As with so many other aspects of water management, dire need, and even perceived need, trumps ecological caution pretty much every time.