The Emerald Mile: The Epic Story of the Fastest Ride in History Through the Heart of the Grand Canyon (44 page)

The surge of water from a collapsing dam is called a dam-break wave, and it can move with astonishing speed while generating unimaginable havoc. Even in modern times, it’s hard to beat the raw and visceral impact of such a wave. Proof of that came in May of 1889, in a mountainous valley just east of Pittsburgh, Pennsylvania, where a group of wealthy steel-industry magnates had established a private hunting and fishing retreat around a dam and reservoir above a thriving community known as Johnstown. Following a series of ferocious rainstorms, the seventy-two-foot-high earthen dam—a structure only thirty-five feet higher than Sadd el-Kafara—abruptly gave way, sending some twenty million tons of water
cascading through the Conemaugh Valley toward Johnstown. The main wave traveled at more than fifty miles per hour and scooped up everything in its path—trees, buildings, horses, cows, even a locomotive that was racing down the tracks as fast as it could go, its whistle tied down by the engineer in an effort to warn the town.

Minutes later, the roiling mountain of water plowed into the Cambria Iron
Works, where it picked up a load of railroad cars and barbed wire. When that mass of debris finally piled up against a massive stone bridge in the center of Johnstown, the wreckage caught fire. Many residents who had not drowned in the initial surge found themselves trapped inside an inferno and burned to death. The fire raged for three days, and the tangled wreckage eventually had to be blown to pieces with dynamite. Even for a generation inured to suffering by the carnage of the Civil War, Johnstown was appalling. The death toll, which included ninety-nine entire families, came close to the number of lives that would later be lost in the bombing of the World Trade Center. It was the largest loss of life in a civilian disaster in US history up to that point.

If Johnstown wasn’t enough of a cautionary tale, other reminders have since occurred. In March of 1928, the St. Francis Dam, a 195-foot-high concrete structure about forty miles northwest of Los Angeles, gave way in the middle of the night, releasing a reservoir that held thirty-eight thousand acre-feet of water. The dam-break wave, which was at least 125 feet high, traveled fifty-four miles before discharging into the Pacific. The bodies of some victims were wedged into deep underground caverns, while others were carried into the ocean and washed ashore as far south as the Mexican border.
The exact number of victims is still unknown—remains continued to be discovered throughout the 1950s—but is estimated at more than six hundred.

In the decades after that disaster, civil engineering made tremendous strides, but flawed design, shoddy construction, and inept oversight continued to take a toll, even as recently as 1976. On the morning of June 5 of that year, a stream of water began erupting on the face of a large earthen dam on the Teton River in southeastern Idaho. A pair of bulldozers failed to plug the leak, and in less than two hours the embankment had washed out, both dozers were engulfed, and their drivers had to be pulled to safety with ropes. Forty-two minutes later, a third of the dam wall disintegrated, and the entire reservoir gushed through the breach.
By eight o’clock that night, the Teton reservoir was empty and eleven people were dead, along with thirteen thousand head of cattle.

Johnstown. St. Francis. Teton. Those events were stamped into the mind of every engineer, technician, and manager within the Bureau of Reclamation. And although the team confronting the crisis at Glen Canyon in the summer of 1983 were far too busy to worry about drawing historical analogies, every one of them would readily have conceded that those incidents raised a rather terrifying specter, given that the stakes on the Colorado River were so much greater.

Glen was almost ten times higher than the dam whose collapse caused the Johnstown flood, nearly four times higher than St. Francis, and well over twice the height of Teton. But far more sobering was the sheer amount of water in Lake Powell. The reservoir’s myriad tentacles and arms, its countless alcoves
and bays, extended along a shoreline that covered nearly two thousand miles, and the nine billion gallons of water within that sandstone labyrinth was
ninety-four times bigger than the volume of Johnstown, St. Francis, and Teton
combined
. What would happen if a catastrophic flood surged out of the headwaters of the upper Colorado River basin, wiped out the Glen Canyon Dam, and sent all of that water racing through the Grand Canyon?

As it happens, the Bureau of Reclamation has studied that question, and the details make for a sobering read.

The initial dam-break wave, traveling about twenty miles an hour, would begin sweeping through Lee’s Ferry sometime between thirty-six and forty-two minutes after the dam failed. Within six hours, the ferry itself would be under 520 feet of water, the height of the Standard Oil Building on the southern tip of Manhattan. Meanwhile, the peak stage of the dam-break wave would continue moving downstream, in the process taking out virtually every piece of infrastructure within the canyon.

Four miles below Lee’s Ferry, the steel girders of Navajo Bridge, the only highway crossing between Lake Mead and Lee’s Ferry, which sits 460 feet above the river, would be swatted aside like a spiderweb. Phantom Ranch, along with the pumping station that supplies all of the freshwater on the South Rim, would be inundated and destroyed. So too would every mooring spot, camp site, and Anasazi ruin along the river.

It would take almost a full twenty-four hours for the peak stage of the flood to reach Lava Falls, and ninety-eight miles beyond, the water would begin spreading across the surface of Lake Mead. Meanwhile, the effects of the surge within the canyon would be severe and lethal.

“Anyone still on the river,” notes the report,
“would have to climb the equivalent of a 40-story building, at a minimum, to have any hope of surviving.”

G
len’s
chief designer was an engineer named Louis Puls, who had been born in the nineteenth century and whose wardrobe and habits—he dressed in jodhpurs and enjoyed smoking a corncob pipe—seemed lifted from that distant era. One of the finest dam designers the bureau had ever produced, Puls was legendary for being meticulous and for refusing to take chances. Even those who hated his dam or were convinced that its sandstone abutments offered a poor anchorage readily conceded that the structure itself was exceedingly well made.

When he began his assignment, one of Puls’s most important decisions was what type of dam Glen would be. He had two choices. The first was a gravity structure—essentially an enormous pyramid that was wide at the bottom, narrow
at the top, and would sit at a right angle across the river, directing the force of its reservoir downward into the dam’s foundations. Squat and ponderous, a gravity dam’s stability derives from brute mass: its sheer weight prevents the wall of water behind the dam from shoving it downstream. The alternative was an arch-type dam, a dam far more elegant and potentially even stronger (and which, as the name suggests, was invented by taking the kind of arch that the Romans had conceived and flipping it on its side). Instead of relying on its own bulk, an arch dam transfers the pressure of the water across its convex face and concentrates those forces in the abutments that anchor the structure to the walls of the canyon. Quite literally, the dam commandeers the rock walls to support its own aim. The harder the water pushes against the walls of the canyon, the harder the walls push back and the more resistant the dam is to failure.

Arch dams can be far thinner than gravity dams, but they require exceptionally dense and stable rock because the stresses on the abutments can be tremendous. Inside the canyon where Glen was going to be built, Puls knew that those forces would exceed the tolerance of the brittle Navajo sandstone. For this reason, as the writer Russell Martin has explained in
A Story That Stands Like a Dam
, his definitive chronicle of Glen’s planning and construction, Puls ultimately opted for a hybrid design: a gravity dam that had a double-arched curvature.

When viewed from the side in cross-section, Glen was shaped like an enormous traffic cone, three hundred feet thick at its base and twenty-five feet at its crest. From above, however, the structure looked more like the rind on a slice of watermelon—an arc that bent more than a quarter mile from keyway to keyway. Puls’s design ensured that the weight of the dam would relieve some of the stress on the sandstone walls while still retaining the strength of an arch.
As a result, Glen was spectacularly strong.

Knowing this, almost none of the foremen and supervisors who were in charge during the spillway crisis feared that the dam itself was in any danger of suffering the kind of catastrophic breach witnessed at Johnstown, St. Francis, or Teton. However,
a number of the Denver engineers did harbor concerns about what might unfold if things started to spin out of control. What vexed them most deeply was the integrity of the spillway tunnels and whether the cavitation damage might, in a worst-case scenario, impair the dam’s ability to hold back the lake. This concern had its roots in one of two nightmare scenarios that were possible.

The first—overtopping of the dam—was rather straightforward. If the damage worsened to the point where the tunnels collapsed or were blocked by debris, the penstocks and the river outlets would be unable to cope with the runoff and the lake would rise uncontrollably toward the crest of the dam.
Eleven feet above the tops of the spillway gates, the surface of the lake would sluice over the parapets and begin sliding down the concave face of the dam. As the lake rose higher, the thickness of that waterfall would increase, and so too would the force with which a portion of it would be deflected squarely against the rear of the power plant. As hours turned into days, the plant’s concrete superstructure would start to give way. When the walls and the concrete support pillars were gone, the eight generators would be exposed to the full fury of the cascade and eventually be shorn from their mountings, along with the turbines and the ninety-ton transformers, all of which would topple into the current and form a tangled wreckage at the bottom of the river.

The damage would be enormous and the repairs would be horrendously expensive. But the overtopping of the dam was largely a self-limiting disaster. Sooner or later, the runoff would peak, the lake would subside, and the destruction would cease. This scenario thus paled in comparison with the
real
nightmare, one whose mechanism hinged on the most vulnerable part of the dam—the elbow joint, where Phil Burgi had surmised that the worst of the cavitation damage was unfolding.

W
hen Louis Puls and his team had opted to save some money by marrying the spillway system to the diversion tunnels, they traded a fiscal problem for a structural one. While they wanted to leave the lower portion of the diversion tunnels open to accept water from the spillways, they had to find a way of blocking off the upstream sections. Failing to do so would have created two drain holes at the bottom of the reservoir, which is why they sealed each of the upstream sections of the diversion tunnels with a massive plug of concrete.
Each plug was 150 feet long and notched into the tunnel wall in such a way that no amount of water pressure could dislodge the concrete from the surrounding rock. However, as the Denver engineers now recognized, the position of those plugs was vaguely unsettling. Just in front was the elbow joint that was sustaining the bulk of the cavitation damage. And directly behind the plugs lay the bottom of Lake Powell.

This raised a disturbing possibility. The sandstone abutments were riddled with fissures, and if the cavitation damage eroded the sandstone
around
the concrete plugs enough to establish a connection with the water at the bottom of the reservoir, the pressure on that leak would be enormous, causing the leak to expand like cancer and, ultimately, the abutment to fail.

During the initial phase of the spillway crisis,
the Denver engineers had weighed this possibility before dismissing it as absurd. The length of those concrete plugs and the manner in which they had been notched into the rock made
it highly unlikely they could ever be dislodged. Moreover, the forces inside the spillway tunnels were almost certainly being directed
downstream
and away from the concrete, rendering the scenario even less likely. But a document in the bureau’s own archives suggests that this remained a concern in the minds of the Denver engineers.