Cadillac Desert (72 page)

In 1969, the year before the water-injection tests, the Bureau had taken the rather unusual step of performing a test-grouting program, so unsure was it of the conditions at the Teton site. Holes were drilled in the rock, and grout was pumped in under high pressure; then the job was tested to see how well it worked. As far as the Bureau was concerned, it had worked fine. “Once we decided that the cracks in the abutments could be sealed with grout,” Harold Arthur, then head of dam design and construction, told a reporter from the Los Angeles
Times,
“we never reconsidered the suitability of the Teton site, despite the difficulties we experienced later in construction.”

Only one thing had been wrong with the Bureau’s test-grouting program. There was a road leading to the damsite from Sugar City, a few miles to the southwest, but none from the north. All of the test grouting was performed on the south abutment of the dam. None whatsoever was performed on the north side, the right abutment of the dam—the side where three hundred gallons of water per minute injected into holes had simply disappeared, day after day after day.

With the defeat of the environmentalists in court, there was no way to stop the dam. From an appropriation of $1,575,000 in 1971, funding for Teton jumped to over $10 million in 1972 and went even higher for the next four years, reaching an apogee of $15,217,000 for fiscal year 1976, when the $85 million dam was completed. Or, to be more accurate, when the Bureau’s engineers thought it was completed.

In his 1970 memorandum, Clifford Okeson, the Bureau geologist, had said that the largest cracks he could find after extending a miniature televison camera and light down the length of a thirty-five-hundred-foot borehole were about an inch and a half wide. That was a small crack, easily grouted—nothing to worry about. In February of 1974, however, as the Bureau’s main contractors, Morrison-Knudsen of Boise and Peter Kiewit of Omaha, were excavating the huge keyway foundation trench—which would replace the worst of the fractured surface rock with a man-made concrete foundation—they came on the right abutment’s great secret. It was a discovery that five years of boring, injecting, and test grouting had failed to reveal. What they found, Robbie Robison, the Bureau’s project engineer, wrote his superiors in a memo, were “unusually large” fissures in the rock of the right canyon wall.

“Unusually large” was hardly apt. The fissures were gigantic. They were caves. One of them was eleven feet wide and a hundred feet long. Another was nine feet wide, in places, and 190 feet long. One by one, other fissures were discovered. The whole right canyon wall was full of them.

If Robison’s description of what had to be considered an appalling discovery was understated in the extreme—even if the fissures weren’t a safety problem, it was astonishing that they had been missed—his recommendation of a course of action displayed an arresting mental paralysis. “We do not recommend to grout these voids at this time.” Robison wrote Harold Arthur in Denver. “The claims situation [by the contractor] ... makes us hesitant to cause any delays.... Furthermore, grouting of these voids is not critical at this time as they are located outside the dam area and could be grouted at a later date if you should so desire.”

Robbie Robison, barely thirty years old, was on his first big project. It had been a troublesome project from the beginning, racked by delay. Costs were up; schedules were behind. For four years the Teton project had been officially underway, and now, in 1974, there was still nothing to show for it but a huge amount of excavation at the bottom of the canyon and some trailer sheds and a lot full of earthmoving equipment. The two biggest voids alone would eat a trainload of grout. Who knew what others would be found? The important thing, Robison figured, was that they were
beyond
the keyway trench; they were
beyond
the point where the Bureau had arbitrarily decided no further grouting was required; they were, therefore, beyond the limits of reasonable concern. After all, if you wanted to be
really
secure, you could have extended the keyway trench all the way to Ashton, which was twelve miles out from the north abutment of the dam. That was what Robie Robison sarcastically told a reporter, later on. It might not have been a bad idea.

Though the airy caves in the rock were a shocking discovery, no one besides Robison, the contractors, Harold Arthur, and a small circle of Bureau officialdom knew about them. Gil Stamm, the commissioner, was probably never told. The people of Rexburg and Sugar City, the two towns lying directly in the Teton River floodpath, were entirely in the dark, as were the politicians who had so assiduously promoted the dam. Of course, had they known about the voids, it probably wouldn’t have mattered to them anyway. After all, the Bureau of Reclamation had the best engineers in the world.

The dam was finished, more or less, on October 3, 1975, when the flow of the river was interrupted for the first time. Even with the biggest voids left unfilled, the job had taken 503,000 cubic feet of grout—more than twice as much as the Bureau predicted it would have to use. That winter, a series of Pacific storms bashed into the Teton Mountains, depositing a big snowpack. As spring was about to arrive, Robbie Robison had two worries: how he was going to settle with the contractors over the cost of the extra grouting, and how he was going to capture the snow that was about to melt out of the Grand Tetons without violating the Bureau’s time-honored rule about filling reservoirs behind earthfill dams.

The rule is simple: the rate of fill is to be kept at or below one foot a day measured vertically along the reservoir walls. That way, if problems develop with the dam or the abutments, or back along the reservoir itself—where rising water sometimes loosens rock and causes landslides, or causes the bedrock to shift under its weight, producing the same result—they can be dealt with. At a slow rate of fill, such problems are less likely to develop in the first place. It was a sensible rule, and, like most sensible rules, it had already been violated on a number of occasions. Why not dispense with it again, with all that precious water coming down from the Teton Range? On March 3, 1976, Robison wrote Harold Arthur formally requesting permission for a two-foot-per-day filling rate. Ironically, one of the arguments he used in support of his request was that a faster rate of fill would permit the Bureau to observe how effective its grouting program had been. It was, in a way, like arguing for a hundred-mile-per-hour speed limit on the grounds that motorists would spend less time on dangerous highways if they drove twice as fast. But on March 23, Arthur readily acceded to Robison’s request.

Actually, the whole business—formal request, formal permission granted—was a meaningless charade. The main outlet works—the tunnel and appurtenances that would carry water out of the reservoir and into the adjacent canals—were not yet finished. The auxiliary outlet works were, but they were designed to carry a maximum flow of 850 cubic feet per second. Engorged by a snowpack half again as deep as normal, the Teton River was about to peak at several thousand cfs. Without a functioning main outlet works, the reservoir would rise as fast as the Teton River felt like filling it. It was likely to rise a lot faster than two feet per day.

Harold Arthur was unconcerned about such a fast rate of fill because he had ordered a series of observation wells to be drilled around the dam, which would—in theory—inform the Bureau of any developing problems. The water table around a damsite will often show a rise as the reservoir fills, because a certain amount of seepage into outlying terrain is inevitable. If the water table rises precipitously, however, and if wells far from the reservoir are affected—especially wells downstream—it could mean that the reservoir is seeping excessively. The only other possibility is a pressure response, where the adjacent water table rises out of proportion to the actual rate of seepage because of hydrologic pressure, much as the constriction of a hose nozzle turns a placid gurgle into a sixty-foot jet.

From what Arthur had heard from Robbie Robison, the observation wells in the vicinity of the dam were showing what he termed a “predictable buildup”; that was the term he used in his March 23 memo. Obviously, he had not yet seen, or even been told about, a Bureau report written almost exactly at the same time, which disclosed the startling fact that “the rate of travel of the rising water table north of the reservoir is over 1,000 times that calculated for predicted movement of water.” The memo came from Gordon Haskett, the Bureau geologist who had been monitoring the observation wells. To engineers who had spent their lives working with microtolerances, who considered almost any adjective hyperbolic, something as extreme as a thousandfold increase should have leaped off the printed page. Haskell’s report, however, was routinely routed through the Boise regional office, from where it went to Denver, where it reached Harold Arthur’s in-box on April 13, three weeks after he had already consented to the faster rate of filling.

It probably wouldn’t have mattered if it had arrived the day after it was written. After looking Haskett’s memo over, Arthur filed it away. In a way, he cannot be blamed. Having reported the bizarre thousandfold increase in the predicted rise of the groundwater table, Haskell had felt obliged to explain it. It was, he said, too
excessive
to be attributable to seepage. “Therefore,” he concluded, “[it] must be a pressure response.”

Actually, a relatively simple and inexpensive piece of gauging equipment, a piezometer, could probably have told the Bureau whether something drastic was going on or whether the inexplicably rapid rise of the adjacent groundwater table was merely a pressure response. Forty miles across the Rexford Bench, on Willow Creek, the Corps of Engineers had just erected Ririe Dam, and all forty-nine of its observations wells were equipped with piezometers. Their use had been routine practice for years. The closest thing to an official explanation as to why they weren’t used at Teton came from Richard Saliman, the chief of the Bureau’s design division. “We do use them on other dams,” Saliman told a reporter, “but basically, we had such an excellent foundation we didn’t feel it necessary.... With the rock types we had we just didn’t see the need for it.” For his part, Harold Arthur doesn’t think the piezometers would have detected anything “unless one of them happened to be exactly where the leakage was occurring. It would have been a matter of luck.” But even if luck had been on the Bureau’s side, it might not have made a difference. The Bureau didn’t believe in luck—it believed in itself. “Suppose we’d gotten a reading from a piezometer that there was massive seepage from the dam,” Arthur told an interviewer in 1983. “We might not have believed it. We had a perfect record up to then. We might have thought the thing was giving us a wrong reading.”

By mid-May of 1976, the Teton River was a frigid deluge. Square miles of snowfields were melting into it under a hot, high sun and the reservoir was rising much faster than it ought to have been, approaching four feet a day. As the reservoir filled, the emergency outlet works were the only real insurance against catastrophe. If the dam gave evidence that it was going to fail, the outlet works would permit a rapid but controlled drawdown of the reservoir. But the outlet works were still not operational; they were completely sealed off by a huge metal barrier and in the process of being painted. On May 14, Robison was finally concerned enough about the rapid filling to write his superiors. “Request your comments for flood control operations,” he said in a terse memo. It was a pro forma exercise: the Bureau, by then, was completely in the river’s hands.

On the 3rd of June, a Thursday, the first equipment operator arriving at the damsite early in the morning noticed a small leak pouring out of the canyon wall about a third of a mile below the dam. From the canyon rim, three hundred feet above the river, the leak looked like nothing; one could barely hear it bubbling above the quiet rush of what was left of the river flowing out of the auxiliary outlet works. The leak was coming out of the north abutment—the right canyon wall. The water was clear. Five hundred feet closer to the dam was another leak, even smaller, also clear. The next day there was still another. All three leaks were coming out of the right canyon wall.

Robbie Robison stood on the canyon rim watching the leaks for a while. Looking back at him impassively was his masterwork, Teton Dam: an average-size modern dam, but a monument that would have made a pharaoh reel. Although Robison was, as he later put it, “just a cog in a great big wheel,” it was his monument. The reservoir was sitting quietly behind the dam, looking utterly serene. Suddenly set free, it would have a calculable energy release approximating that of a quarter-megaton bomb.

Robison returned to his office in the trailerlike project building. Then, restless, he went outside and watched the leaks again. Finally, he went down into the canyon and crossed the river by boat. The dam loomed above him, 305 feet high. Robison jumped over the rocky bed and climbed up the fifty-degree slope to the first leak and measured it. Sixty gallons a minute, about a seventh of a cubic foot per second. The second leak was flowing at about forty gallons a minute, the third—the one closest to the dam—at about twenty.

Robison went back across the river, climbed to the Bureau’s trailers, and wrote a brief memo to Harold Arthur telling him about the leaks. At the end of the memo, he said, “I’ll keep you advised.”

Off and on during the day, Robison’s men monitored the leaks through binoculars. By nine o’clock in the evening it became too dark to see, and they went home.

Saturday, June 5, dawned pellucid and bright, a warm and somnolent day. The first Morrison-Knudsen man arrived at the Teton site at seven in the morning. In the shadowy postdawn light, the downstream embankment, facing west, was still dark. He looked at it and saw nothing. Sometime around seven-thirty he looked again and saw something. There was a roiling creek of muddy water emerging from the right abutment adjacent to the dam.