The entry says that the city of San Diego bought one experiment and ran another, and that the second was read in twenty minutes on the afternoon of January 27, 1916. This piece is about the people who read it: the engineers who went over the wreck of the Lower Otay dam that spring and argued, in the pages of Engineering News and Engineering Record, about what it had been and what it meant. They did not agree on everything. They agreed on the design.

What it was

The clearest description came from Charles Whiting Baker, the editor of Engineering News, who had walked over the dam on October 27, 1915, with the hydraulic engineer Allen Hazen and John Bacon, the city’s assistant manager in charge of water, and who published what he had seen a week after it was gone. The dam had been “fully described at the time of its construction, in Engineering News of March 10, 1898,” Baker wrote, “and has been described also in the late James D. Schuyler’s well-known treatise on ‘Reservoirs.’ In some important respects, however, the actual construction of the dam differs from the published descriptions, notwithstanding the fact that the article in Engineering News was prepared by the engineer in charge of the work.”

It had been built by the Southern California Mountain Water Company and finished in 1897, and “the originator of this design was not an engineer, but a promoter and financier, E. S. Babcock, at that time president” of the company, who “not only originated the plan of building the dam in this manner, but directed how the work should be carried on.” The foundation was masonry, carried thirty-one feet down into a seamy porphyry and up to eight feet above the old water level. Then Babcock changed his mind. The rest of the dam was rock, blasted from the canyon walls just below the site by charges of six tons of powder and dropped into place by cableway, and through the middle of it stood a wall of riveted steel plate, a third of an inch thick in the three bottom courses and a quarter of an inch above, mopped with hot asphalt, wrapped in burlap, boarded, and cased in a thin wall of concrete on each side. The boards were left in. “According to information furnished the writer by those directly connected with the work of construction,” Baker wrote, “slap-dash methods were pursued.” The steel was often run up well ahead of its concrete, and in the sun “the top of the diaphragm expanded and assumed a series of curves, so that it was impossible to do a good job in holding the layers of asphalt, burlap, boards and concrete against the steel. There was of course no way in which to determine whether the riveting and calking of the diaphragm was a water-tight job.” The dam was 130 feet high at its original crest, 565 feet long, and its faces were as steep as dumped rock will lie. “It has always leaked.” By 1913 the city was pumping the leakage back into the reservoir with a fifty-horsepower oil engine.

For its first two years it had no spillway. “Building a dam without a spillway is a good deal like building a boiler without a safety valve,” Baker wrote, “and it is a particularly noteworthy illustration of the ideas prevailing in the construction of the dam that no spillway whatever was made until the dam had been completed for two years.” The one then cut through the rock at the east end was thirty-eight feet wide and eight feet deep. The only other way to get water out was the outlet tunnel to the city, eighty-two feet below the crest, with a blow-off valve at its lower end that could pass about 330 cubic feet a second. The drainage basin above was about a hundred square miles. In nineteen years it had never put the water within fifteen feet of the spillway floor.

Baker set down the engineers’ objection in one sentence, and it is the sentence this journal exists for: “one chief objection to the peculiar design of the dam is the impossibility of determining with accuracy its margin of strength.”

How it went

Roy Silent, from J. B. Lippincott’s office in Los Angeles, examined the site on January 31 and February 1 and wrote the account the Geological Survey adopted. The first storm, January 15 to 20, put 5.60 inches of rain in the gauge at the dam and raised the reservoir from 96.5 feet above the datum to the spillway floor, at 122 feet 8 inches, by the morning of the 21st. It went on rising through the spillway. The second storm put another 3.54 inches in the gauge in three days, and “the rain on Jan. 27 was extremely heavy, and by noon the water had risen so high that Mr. Weuste, in charge at the dam, deemed it advisable to open the outlet gate. This failed to check the rise, and it was realized that the dam would probably be overtopped before evening. Men were accordingly dispatched to warn residents in the valley to move to higher ground.”

“At 4:45 p.m. the water had reached the top of the dam and had seeped through and filled the boxes that were sunk in the top to allow an examination of the steel core. Water began running down the lower face on the east side of the dam at approximately 4:50 p.m. About this time several spouts or small streams of water appeared on the lower face of the dam, in one instance loosening a large boulder which rolled down to the bottom. From this time on, the destruction was very rapid. The lower face of the fill quickly melted away, thus removing the support from the core wall. At 5:05 p.m. the tension was so great that the steel diaphragm tore from the top at the center, and the dam opened outward like a pair of gates. The released water rushed through and filled the cañon to a point approximately 20 ft. below the top of the dam.” The reservoir emptied in two and a half hours. “Practically all the fill was washed completely away. The steel core was deposited in varying-sized sections along the valley, a large part being found at Palm City, 10 mi. below the dam.”

Engineering Record drew the stages of the failure from the eyewitnesses in three sketches, and added two things. “Failure did not commence until the water had overtopped the 5-ft. earth dike surmounting the core wall, but witnesses agree that the depth of the overflow before failure was very slight, probably between 4 and 6 in.” And: “There is no indication that the structure slipped along its foundation or that there were excessive leaks or blow-outs before overtopping occurred.” The steel, when they found it, “was in perfect condition, although it had been in service 20 years.” Numbers of its rivets had been sheared in two.

What they made of it

The Engineering News editorial of February 10 was anxious, above all, that the public should not draw the wrong lesson. “In the literature published by the company that built the dam and later sold it to the City of San Diego the structure was praised as a marvel of engineering skill. So far as the public was aware, the Lower Otay dam was as substantial and permanent a structure as was ever built. In the engineering profession, however, the Lower Otay dam was not held in such high repute. In very few places has its peculiar design been imitated.” The disaster “carries no warning as to lack of safety of dams in general or of the rock-fill dam as a type”; what it showed was that a rock-fill built as this one was “is not materially safer, if overtopped, than an earth dam.” The defects were three: too small a section; no paving on the downstream slope; and no crest that could stand an overflow, because the concrete core wall “stopped some 5 ft. below” the top. All of it, the editorial said, would have cost “a very small percentage” of the dam. “It may be asked why the dam was not built in this way in the first place. The answer is that, when the dam was built, the company which owned it was chiefly anxious to do the work at the least possible cost. The company and its successors have always worried far more over lack of rainfall to fill the dam than over floods that might overtop it.”

Engineering Record, a week later, put the Lower Otay failure beside the dams that had held. The Upper Otay arch, “overtopped by a flood pouring 3 ft. deep over its crest,” was “wholly uninjured.” Sweetwater, arched and overtopped by three and a half feet for nine hours, “stood without serious damage an overpour” that would have carried it away had it been straight. “The Lower Otay dam was a hybrid structure,” built “without adequate engineering supervision” over a ten-year period, with a hundred square miles of catchment and a spillway thirty-eight feet wide whose lip was lower than the top of the core. The lesson it drew was about the Southwest: “flood hazards in the semi-arid Southwest have been underestimated,” and earth and rock-fill dams need spillway capacity above any credible flood, and kept clear.

In June, George S. Binckley, a consulting hydraulic engineer in Los Angeles, wrote the longest of the post-mortems, and the one that read the experiment as an experiment. “The more disastrous a failure is, the more valuable are its lessons. In engineering especially, much is learned from failure, little from success. The construction and continued existence in service of the Lower Otay dam taught nothing. The margin of safety that held it together under a head of water 15 ft. below the spillway level in 1907 is unknown, and can never be known. But the failure of this recklessly conceived and crudely constructed example of lay engineering has a value of its own. So long as it stood up, no matter how fearfully narrow the margin by which it stood, the fact that it did stand was to the layman sufficient and triumphant proof of its safety.” Even engineers had been mild, he said, because to question the dam was to be asked for a mathematical proof of its weakness, and “the nature of the stresses and the elements of resistance in a pile of loosely dumped rock are not susceptible of mathematical determination.” Now that it was gone, “criticism may rest on the safe basis of disastrous fact,” and “the dissection of this unpleasing corpse will be profitable to the engineering profession.”

Binckley’s own reading of the corpse was that the overpour had not washed the dam away; it had drowned it. The water that went over the top filled the voids in the downstream rock faster than it could escape, the fill lost its effective weight as it became submerged, hydrostatic pressure worked inside the mass, and a slope built at the natural angle of repose needed only to lose material at its toe to set everything above it moving. There had been “an interval of twenty minutes between the time the first serious break occurred in the earth dike on top of the dam and the moment of failure of the structure itself, a period quite long enough for the downstream portion of the rock fill to become disastrously gorged.”

Then he turned to Morena, the great rock-fill the city had hired a rainmaker to fill, and which had come within eighteen inches of the same fate because floating wreckage had choked the iron racks in front of its gated spillway. “With the spillway entirely useless, a few hours more of heavy rain would have produced an enormous overpour. It is, the writer believes, extremely unlikely that the Morena dam would have sustained such a condition for more than the short period that would have been required for the gorging of the fill.” His closing paragraph is the profession’s verdict on the whole month. “The destruction of the Lower Otay dam carries one kind of lesson. But the narrow escape of the great Morena rock fill carries another and perhaps more valuable warning. The Otay dam cannot be considered as an example of engineering design or even good construction. But the Morena dam is the work of an engineer of standing and experience, and is therefore more properly subject to engineering criticism where weakness or error in design has been developed. The lesson in both cases is plain. From the Lower Otay we again have tragically emphasized the folly of intrusting engineering works to untrained or reckless hands. And from the narrow escape of the great Morena rock fill few engineers will fail to realize with the shock of sudden remembrance the grim hazards of an obstructed or insufficient spillway.”

The city engineer’s answer

George Cromwell, the city engineer, had been at Morena during the high water and at Lower Otay twenty-two hours after it went, when the inflow was still over 2,000 cubic feet a second. In March he wrote to Engineering News to correct the figures of the men who had written before him. Silent had the spillway passing 2,100 cubic feet a second at the moment of failure; it was 4,700. Silent had the catchment at 139 square miles; it was 98.3. Another critic, Francis Sellew of La Jolla, had the dam 130 feet high with the spillway six feet below the crest; it was 135.5 feet, with the spillway eleven feet down, and its slopes had been measured at one and a quarter to one, not “probably 1 to 1.” Sellew had discussed the dam’s “coefficient against sliding”; Cromwell pointed out that it had not slid. “The dam failed from the top down, it did not slide on its base, fail at foundation or overturn.” Then he conceded the whole case. “It is useless to discuss the merits of the design of this type of dam. I do not think any engineer of standing ever approved the design of this dam. It was a hybrid type beyond question.”

The failure, he said, “was due to overtopping because the spillway was of insufficient capacity to take care of an unprecedented flood resulting from a cloudburst. However, I do not think there was an engineer in the country who would have recommended a larger spillway in view of the runoff records prior to the recent storm.” Every consultant the city had ever called in had doubted “if the Lower Otay reservoir would ever be filled from its own watershed, even to the height of the spillway crest.” He recorded one decision taken in the week between the storms, when the reservoir was already spilling: E. S. Babcock, the dam’s builder, and “several other prominent citizens” had urged that flashboards be put in the spillway to hold more water, and Cromwell and the mayor, Edwin Capps, himself a member of the American Society of Civil Engineers, “refused to allow any such operation of the dam.” The inflow in the last hour before the failure had been 23,500 cubic feet a second, “over seven times as great as any runoff shown by the prior records.” To the complaint that the blow-off valve should have been opened sooner, he answered that at 330 cubic feet a second it would have had to be opened wide ten days before the dam went, before the first storm, which “would have been looked upon as the act of a lunatic at that time by both engineers and laymen.” He would, he said, have been in favour of emptying the reservoir the previous autumn and enlarging the spillway four times, “if we had known about this storm! It seems too bad the Almighty does not see fit to inform us a few weeks ahead of time when He sends such a deluge as He visited upon this section on Jan. 27, 1916.” He hoped Sellew would retract his charge of mismanagement “and acknowledge that the failure of the Lower Otay dam was due to an act of God in sending a deluge that was beyond human possibility to anticipate or cope with.”

The Geological Survey’s engineers, compiling the official record of the flood two years later, used Silent’s narrative, Cromwell’s spillway figure, and the reservoir’s own staff gauge, from which they computed an inflow of 32,700 cubic feet a second in the last forty minutes, and a total of 37,400 with the spillway added. Under the station’s heading they wrote: “Data collected by the city of San Diego were withheld because of pending litigation.” The city rebuilt the dam in concrete and called it Savage Dam.

Sources

• Charles Whiting Baker, “Otay Rock-Fill Dam Failure,” Engineering News, 3 February 1916, with Allen Hazen’s photographs of 27 October 1915 and John L. Bacon’s wired account. archive.org
• “Is the Rock-Fill Dam a Safe Type of Dam?”, editorial, Engineering News, 10 February 1916. archive.org
• Roy A. Silent, “Failure of the Lower Otay Dam,” Engineering News, 17 February 1916. archive.org
• George Cromwell, “San Diego’s City Engineer on the Otay Dam Failure,” letter of 23 March 1916, Engineering News, 13 April 1916. archive.org
• “California Floods Wreck Otay Rock-Fill Dam,” Engineering Record, 12 February 1916; “Safety in Dam Design,” editorial, 19 February 1916; George S. Binckley, “Are Rock-Fill Dams Safely Designed? What Failure of Lower Otay Dam Teaches,” 10 June 1916. archive.org
• H. D. McGlashan and F. C. Ebert, Southern California Floods of January, 1916, US Geological Survey Water-Supply Paper 426, 1918: the Lower Otay station record and the withheld data. archive.org