Floodpath

by Jon Wilkman

  Coroner Nance denied receiving evidence that St. Francis Dam concrete crumbled, but that didn’t stop Cushing’s enthusiasm for sabotage. Like Rieber, the Texas explosives expert claimed it wouldn’t take much to bring down the St. Francis Dam. Releasing a landslide would do the trick. If explosives were set underwater, the only sound would be a barely audible “grunt,” he said.20 Cushing admitted there wasn’t much credible evidence yet, but he argued against rushing to a verdict without a more thorough investigation. A grateful Mulholland later sent Cushing a personal letter expressing thanks for his testimony and included a DWP check for $612.70 to cover expenses.21

  Annoyed by this interruption in their straightforward path to judgment, D.A.s Keyes and Dennison struck back, enlisting members of the Mayberry fact-finding committee to lead the questioning, an unusual prosecutorial tactic. Geologist Allan Sedgwick was unleashed on Frank Rieber’s facts and interpretations. “We have apparently two schools,” Sedgwick declared drily. “One is reasoning from the theory to take facts, and one trying to take from the facts a theory.”22 He pushed the geophysicist about the accuracy of the Stevens Gauge and the positions of the downstream fragments. R.R. Proctor, the DWP field engineer on the St. Francis Dam, and surveyor W.W. Hemborg were called to defend the precision of triangulation measurements that showed the dam had moved, and by how much. During the Inquest, Proctor was considered an especially useful DWP witness for both sides, called to testify nine times.

  More than a dispassionate search for the facts, Keyes’s counterattack was designed to preserve the credibility of the Governor’s Report and the results of the Mead and Mayberry investigations. If these prestigious inquiries were wrong about which abutment was to blame, let alone the possibility of sabotage, what else was in doubt?

  Unfortunately for William Mulholland and the DWP, unexpected witnesses and explosive allegations were not enough to undercut mounting criticism of the choice of location and construction methods used to build the St. Francis Dam. Even worse, shortly after the Inquest got under way, the Chief was forced to fight a two-dam war.

  Among the first questions Inquest investigators posed to DWP engineers was “Who designed the St. Francis Dam?” Everyone pointed to Mulholland as the man with the final word in all matters, including site selection, design, schedules, and construction methods. Others admitted to surveying the location and supervising the work, but no one was willing to accept credit as the ill-fated dam’s designer. As an example, when Coroner Nance asked DWP Construction General Superintendent Stanley Dunham what plan he followed, Dunham answered, “The plan handed me.”23

  The mysterious paternity of the St. Francis Dam was clarified when BWWS Assistant Office Engineer Edgar A. Bayley, a veteran of the Owens River Aqueduct project, testified on March 26. Bayley ducked direct responsibility, telling the jury that he was in Colorado when work on the San Francisquito project began. However, he acknowledged being involved in the design of the Mulholland Dam, constructed in the hills overlooking Hollywood. Bayley went on to reveal that the St. Francis Dam, completed later, was not conceived independently but based on specifications for the concrete barrier in Hollywood. DWP engineers simply adapted existing plans to fit the site the Chief had chosen between Powerhouses 1 and 2.24 This was confirmed by testimony from DWP office engineer W.W. Hurlbut, a trusted Mulholland lieutenant who supervised the design transfer process.25 Apparently the St. Francis and Mulholland Dams were twins separated at birth. Soon scrutiny would turn to the Chief’s surviving concrete barrier. Was a second disaster waiting to happen?

  On April 10, Los Angeles County Coroner Frank Nance declared the Inquest investigation complete. He gathered the jury and gave them instructions. Nance said their job was to determine if the death of Julia Rising—along with the others in Los Angeles County who had died because of the St. Francis flood—was the result of homicide, accident, or other causes. As part of their conclusions, the jury was asked to decide if the St. Francis Dam “was properly located, erected and maintained with due caution and circumspection, and to make any recommendations that could help prevent such a failure again.”26

  In Assistant District Attorney Dennison’s closing comments, he was unexpectedly lenient toward William Mulholland. “It would be monstrous to place a man on trial for the crime of manslaughter or murder, who merely made an error of judgment,” he told the jurors. “I don’t want you to send me into Court with him or anybody else to prosecute him for crime [sic], because it would only come to disaster … if he used all the care and prudence he was capable of using and found afterward that he had made a mistake there would be no criminal negligence.”27

  Two days later, on April 12, the Coroner’s jury returned a verdict. The failure of the St. Francis Dam, they concluded, was the result “the very poor quality of the underlying rock structure … and the design of the dam was not suited to inferior foundation conditions.” William Mulholland and the Bureau of Water Works and Supply had committed a tragic error of judgment when they built a dam on such a foundation, but the verdict continued: “We, the jury, find no evidence of criminal act or intent on the part of the Board of Water Works and Supply of the City of Los Angeles, or any engineer or employee in the construction or operation of the St. Francis Dam, and we recommend that there be no criminal prosecution of any of the above by the District Attorney.”28 Bill Mulholland wouldn’t face an indictment and the possibility of joining the Watterson brothers in San Quentin.

  There was no criminal indictment, but the Inquest jury didn’t absolve William Mulholland from the way he ran the Bureau of Water Works and Supply: “A sound policy of public safety and business and engineering judgment demands that the construction and operation of a great dam should never been left to the sole judgment of one man, no matter how eminent, without check by independent expert authority.”29

  Although the Inquest jurors, like the Governor’s, Mead, and Keyes investigators, considered the basic gravity design acceptable and the quality of the concrete satisfactory, they itemized a number of safety features that should have been included or were inadequately applied, such as effective anchoring into the canyon abutments and insufficient grouting to limit leakage.

  Although Mulholland had installed drainage pipes in the center portion of the dam’s foundation, all studies of the failure concluded that they should have been drilled across the entire width of the structure and up either abutment. Inspection galleries, a feature found in the interior of dams like Elephant Butte that allowed engineers to directly observe drainage systems, were also deemed not essential, but they would have provided access to valuable safety and maintenance information.

  The Inquest jurors and other major investigative panels considered predetermined contraction joints “standard practice” to handle stresses that occur as concrete cools, far more effective than the Chief’s “let them occur where they will and patch with oakum” approach.

  In the end, the Governor’s Commission saw little hope for the survival of the St. Francis Dam constructed where it was: “it is improbable that any or all of these devices would have been adequately effective, though they would have ameliorated the conditions and postponed the final failure.”30

  What about the dynamite theory? The Inquest verdict acknowledged the possibility but dismissed it for lack of evidence. The same was true with an earthquake explanation. The jurors acknowledged that landslides played a role in bringing down the dam, but concluded the saturated foundation on the west side initiated the collapse.

  As for Frank Rieber’s evidence supporting an initial east-side failure, the Inquest jurors refused to engage the argument: “The exact sequence of these events is of great engineering interest but has little bearing on the question of the basic cause and responsibility. A susceptibility to landslide was one of the defects of the site that should have been foreseen.”31

  In a final recommendation, and a direct blow to Mulholland’s independent power, the Coroner’s jury followed the lead established by the Gove
rnor’s Commission and called for changes in the 1917 law exempting dams built by large municipalities like San Francisco and Los Angeles from State oversight.

  During his testimony at the Inquest hearing, the Chief resolutely denied seeing anything wrong with the St. Francis Dam. Had he told the truth? Did the percolation tests show the red conglomerate was impermeable, and was the leak he investigated the day of the disaster really running clear? In fact, it’s possible that the percolation holes actually held water as Mulholland said, because unknown to him, soil deep in the hole could have collapsed, filling the bottom and holding the water up.

  As for the seepage coming from the leak that worried dam keeper Tony Harnischfeger, David C. Henny, a Portland, Oregon, consulting engineer, published a report that noted water seeping from the crack was probably flowing through the earthquake fault seam on the dam’s west abutment. Since the seam contained clear gypsum, he speculated that the water wouldn’t darken as it passed through.32 If true, it was a possible reason why Mulholland and Van Norman didn’t see a muddy-looking leak on March 12.

  While the Coroner’s Inquest received extensive coverage in the press, full transcripts of the hearings were not readily available to the public and over time the few copies became extremely rare. As a result, the California Governor’s Commission Report was the primary source used by those who wished to study and understand the St. Francis failure. State Engineer Hyatt’s office reported an unexpectedly high demand for reprints of what was commonly considered the final word about the how and why of the St. Francis Dam disaster.

  In fact, it wasn’t.

  12.

  Hasty Conclusions and High Dams

  In addition to the California Governor’s Report and the Los Angeles County Coroner’s Inquest verdict, the results of other St. Francis Dam investigations appeared in construction and engineering journals, but few outside the profession read them, and the findings were not commonly reported in the popular press. Four days before the Inquest verdict, engineers Carl E. Grunsky and his son Eugene took a fresh look at the evidence. They were joined by Stanford Emeritus Geology Professor Bailey Willis. Only two years younger than Bill Mulholland, Willis was the first formally trained American engineering geologist.1

  On April 6 and 7, Willis and the Grunskys visited the dam site. After surveying the scene and evaluating the evidence, the team agreed with previous conclusions concerning the absence of the latest safety measures in Mulholland’s design. They also agreed that the concrete quality appeared to be adequate, but they questioned how it was placed.

  During his 1925 visit on behalf of the Santa Clara River Protective Association while the dam was still under construction, Grunsky noted that the layers of freshly poured concrete on the crest were uneven, “presenting the appearance of small hummocks.”2 This is the way Mulholland preferred to work. Using the down-to-earth language he was known for, he compared the technique to “packing figs.” The Chief believed this made each five-foot deposit adhere better to the one below. But concrete generates heat as it cools, and Grunsky noted that Mulholland’s pouring method could result in differing rates of density and drying, affecting the strength and solidity of the completed dam. The San Francisco engineer couldn’t say if this had any influence on the collapse.

  Grunsky was especially interested in exploring the role played by uplift forces. Like Frank Rieber, he was convinced the failure started with the schist on the east abutment, not the red conglomerate on the west. “The dam apparently failed at both ends at or very nearly at the same time,” Grunsky wrote in a report. “This is an almost unexpected occurrence, indicating a condition at the time of the failure that could not be accounted for by a mere yielding of the foundation material.”3 In other words, the red conglomerate may have been saturated, but it wasn’t the sole or even initial culprit in the collapse.

  As early as 1925, cracks, both vertical and at 45-degree angles, were apparent on both sides of the dam’s downstream face. Some considered them the result of contraction as the concrete cooled and settled, but Grunsky suggested that shortly after the reservoir began to be filled, beyond expected contraction, the entire structure was under considerable stress and moving. The cracks were evidence of that. He speculated that the winter rains of 1927–28, along with water rising in the reservoir, started to seep between the layers of east-abutment schist, making them slippery like a slick deck of cards. “The uplifting forces of the swelling red sandstone on the west, and the horizontal and up-lifting pressure of the schist at the east, lifted the dam, [and] broke it from its foundation,” he concluded.

  An especially intriguing piece of evidence seemed to confirm that the dam did more than lift and slide. Investigators found a carpenter’s ladder jammed into a crack in the upstream heel of the Tombstone. How did it get there? Grunsky theorized that as the dam tilted forward, the crack opened. When the structure rocked back, the concrete clamped shut, crushing the three-inch-thick ladder. “Grunsky’s ladder,” as it was dubbed, provided dramatic evidence that as uplift pressures increased, powerful forces were at work. As its support system gave way, like a monster in agony, the St. Francis Dam not only slipped and twisted, it rocked on its foundation.

  C.E and Eugene Grunsky’s geologist colleague Bailey Willis agreed with previous investigators that “The rocks at the St. Francis dam site are too weak to support a dam of the concrete arch design.”4 However, going beyond that simple conclusion, Willis introduced an entirely new perspective to understanding the failure.

  Landslides on the east abutment, unleashed on the night of March 12, were obvious to any visitor to the Tombstone. Not as evident were signs of older, far larger land movements slumping along the east side of San Francisquito Canyon. Willis reported that the remains of these landslides dated from thousands if not millions of years ago, not just minutes before the collapse. “They are so large that they are easily mistaken for firm spurs in the mountains,” he wrote, “but once a slide, always a slide.”5

  The upstream crack that apparently opened and crushed “Grunsky’s Ladder” (Author’s collection)

  The Stanford geologist explained that the terrain of the eastern hillside had been formed by repeated landslides, like waves of geological surf. Ironically, the narrow opening in San Francisquito Canyon, which seemed ideal for a dam, was probably created by these slippages, known as paleo-landslides. Willis speculated that the latest of these ancient earth movements initiated the collapse of the St. Francis Dam.

  The hazardous terrain of San Francisquito Canyon was not unique. On March 12, 1928, the same day as the failure of the St. Francis Dam, an article in the Santa Paula Chronicle reported an unexpected catastrophe in the town of Santos, Brazil. Hundreds were reported killed when the slopes of Mt. Serrat collapsed, unleashing a crushing landslide. The newspaper noted that people had lived in Santos since 1543, unaware of the geological menace looming above.6

  The foreign minister from the U.S. offered condolences for the unforeseen “Act of God” in Santos, and his Brazilian counterpart expressed sympathy for the victims of the St. Francis Dam disaster.7 But unlike in the Santos tragedy, Bailey Willis saw the hand of man at work in the slipping hillsides in San Francisquito Canyon: “The old slide had ceased moving, and presumably would not have renewed its activity if it had not been disturbed by the excavation of the [east] end, in search of bedrock on which to rest the dam, and if its base had not become saturated when the reservoir was filled.”8 In short, it wasn’t necessary to accept the widely doubted dynamite theory to acknowledge the credibility of an east-side collapse scenario.

  Like every visitor to the site of the St. Francis Dam before Bailey Willis, William Mulholland was unaware of paleo hazards in the San Francisquito Canyon hillsides, but he knew the importance of designing a dam that sat securely on a safe site. He considered the St. Francis Dam to be a massive structure, depending primarily on gravity to keep it in place. As he told the Inquest jury, it wasn’t a true arched dam. The Chief had included a curve fo
r “an added factor of safety.” Normally, an arch is a good thing, spreading hydrostatic pressure against the abutments. In this case, however, Willis theorized that the arch made conditions worse when the east abutment failed. As the landslide gradually pushed against the east side of the dam, it added stress to the structure, applying pressure that bent and tightened the curve of the arch. “Moving forward, even though fractions of inches only,” he wrote, “the passive abutment had become the source of a force acting to push the dam away, to increase its curvature, and to throw the concrete into tension in a manner not anticipated.”9 It was like an archer’s bow, bent to a breaking point.

  From the beginning, Mulholland believed a major earth movement was a factor in the St. Francis Dam collapse. Privately he may have suspected sabotage, but even if it wasn’t a dynamite attack, the Chief and DWP officials were convinced a full explanation of the failure remained incomplete. Others agreed.

  Charles H. Lee was a San Francisco consulting engineer who had worked on the Owens River Aqueduct and was often employed by the Los Angeles Department of Water and Power as an expert witness in court cases. On April 5, as the Coroner’s Inquest approached a conclusion, Lee sent a telegram to DWP public relations officer Don Kinsey, offering to write a two-thousand-word article for an independent journal. Kinsey replied that he thought it “would be very helpful for you to contribute same.”

  On April 17, after the Coroner’s verdict was released, the San Francisco engineer wrote to DWP Special Counsel, W.B. Mathews, reporting that private utility interests were using the Governor’s and Mead reports to spread “incomplete and incorrect interpretation[s] of data furnished, such as [the] water stage [Stevens Gauge] record.” Lee volunteered to write a DWP-published pamphlet with a “collection of facts and information pertaining to the failure which is known to the department and has not yet been made public.” No record exists of Mathews’s reply, but the DWP wasn’t in the habit of sharing proprietary information unless it worked to the Department’s benefit. Apparently the pamphlet was never written.