Ricochet is the best place on the internet to discuss the issues of the day, either through commenting on posts or writing your own for our active and dynamic community in a fully moderated environment. In addition, the Ricochet Audio Network offers over 50 original podcasts with new episodes released every day.
I’ve been fascinated by the St. Francis Dam failure since I first found out about it. For those who are unaware of or who’ve forgotten about it, the St Francis Dam failure, which occurred in 1928, was the greatest civil engineering failure in the United States in the 20th century (the Johnstown Flood killed many more people, but it took place in 1889), and except for the San Francisco Earthquake, caused more deaths than any other event in California history. Until recently, however, it was relatively hard to find much information on the topic. There was a book about the disaster by a local retired rancher, Charles Outland, who had been a high school senior in Santa Paula at the time the St Francis flood waters raged through town, which was published in the early 1960’s, but that was about it. Since then a couple more books have been published and an engineering professor who has extensively studied the failure and developed a detailed analysis thereof has written and given talks on the subject so that it’s now possible to flesh out the subject in great detail (I’ll provide links to the books at the end of this article; all other links will be in the text). The most interesting aspect of the story to me, however, is the way in which this event touches on and impacts so many other stories.
Los Angeles Aqueduct
The series of events that led to this tragedy started much earlier – 1900 or so would be as good a place as any to start this tale. It was around this time that the city fathers of Los Angeles and William Mulholland (1855-1935) the Belfast, Ireland-born, self-taught engineer and now superintendent of the newly formed Los Angeles Water Department realized that additional water sources would need to be found and developed if the rapidly growing city was to continue to grow and thrive. At this time, the sole source of water for the city was the Los Angeles River plus groundwater within the LA drainage basin.
Then as now, Los Angeles has a Mediterranean climate with warm, dry summers and mild winters. It averages 15” of precipitation (essentially all in the form of rain) with three quarters of that precipitation occurring in the four months from December through March. By 1905, the city was diverting all the surface flow of the Los Angeles River and had constructed underground galleries under the river to tap the sub-surface river flow in the summer months. In addition, groundwater wells were displaying steady subsidence.
Various potential water sources were investigated and it was eventually determined that the Owens River would be the best source for an adequate quantity of high quality water for the city. The story of the Owens River & Los Angeles Aqueduct is filled with much intrigue and certainly deserves its’s own post. However, I’ll just hit the basic project features here in order that we can concentrate on the story at hand.
The Owens RIver Los Angeles Aqueduct was a major engineering achievement at the time. It became a national story and it made a sort of folk hero out of Mulholland. The project was financed by the city via two bond issues – a $1.5 million bond for the purchase of land passed in 1905 and a $23 million bond for the construction costs passed in June 1907. The city performed all but a very small portion of the work with their own forces – starting construction in October of 1907 and completing the work in November 1913 on time and on budget. The aqueduct would transport the Owens River waters a total of 233 miles from its source at the eastern slope of the Sierra Nevada Mountains. The aqueduct intake at the Owens River is at an elevation of 3814 feet and is approximately 15 miles north of Independence, the county seat of Inyo County. From there the water flows by gravity to its’ terminus at what was then called San Fernando Reservoir No 1 at elevation 1135 feet near the Newhall Pass. The major features of this original aqueduct were:
Unlined Canal . . . 23.726 miles @ 900 cfs capacity
Open Lined Canal . . . 37.054 miles @ 900 cfs capacity
Haiwee Bypass . . . 2,001 miles @ 420 cfs capacity
Covered Conduit . . . 97.642 miles @ 420 cfs capacity
Lined Tunnels . . . 42.903 miles @ 420 cfs capacity
Concrete Flumes . . . 0.165 miles @ 420 cfs capacity
Inverted Siphons . . . 12.052 miles @ 420 cfs capacity
Power Waterways . . . 9.845 miles @ 1,000 cfs capacity
Reservoir . . . 7.87 miles
Note: cfs = cubic feet per second.
St. Francis Dam Planning, Design, and Construction
Although the aqueduct was satisfactorily completed and delivering water to the city, work remained to complete and improve the aqueduct to fully meet the city’s long term needs. Several reservoirs and hydroelectric power plants had been planned all along but had been left out of the original aqueduct in order to keep initial costs down. In the years after 1913, the Los Angeles Bureau of Water Works & Supply (I use the abbreviation BWWS for this from here forward) planned, designed and constructed these facilities. However, the continued dramatic population boom of Los Angeles exacerbated by an extended dry spell in the early 1920’s led Mulholland and the BWWS to seek additional storage facilities. Among these additional facilities would be the St Francis Dam & Reservoir. The chart below illustrates this ongoing population boom for Los Angeles.
1880 . . . 11,183
1890 . . . 50,395
1900 . . . 102,479
1910 . . . 319,198
1920 . . . 576,673
1930 . . . 1,238,048
This population growth would play a role not just in the decision to build the St. Francis Dam, but also figure in the dam’s eventual failure. Although the BWWS had designed and built a number of dams by this time, all but one the previous dams had been embankment dams and none had been anywhere as big as the proposed St Francis Dam. The St. Francis Dam would be the second concrete gravity dam to be designed and built by the BWWS. The first was the Weid Canyon Dam (re-named Mulholland Dam in 1925) in the Hollywood Hills, which had been constructed between August 1923 and December 1924.
In order to improve reliability of the system, Mulholland wanted a large storage facility somewhere near the southern (or downstream) end of the aqueduct. This would provide additional security from a number of potential and real hazards to reliable water deliveries. First, since the aqueduct crosses the San Andreas Fault via the five- mile long Elizabeth Tunnel about 40 miles from the aqueduct terminus at San Fernando Reservoir No. 1, the reservoir needed to be downstream of the fault. Second, siting the reservoir near the downstream end would also allow for regular maintenance and minor repairs of the aqueduct with little or no disruption to water deliveries. Finally, the populace of the Owens Valley had never really forgiven the BWWS and the city of Los Angeles for their underhanded dealings in the purchase of land for the original aqueduct, and many in the valley viewed the city as something of a foreign power stealing a resource that rightfully belonged to them. These tensions were intensifying in the early 1920’s. When negotiations between the city and Owens Valley interests to share in the water from a planned new reservoir fell apart several acts of violence ensued. In the early morning hours of May 21, 1924, an aqueduct spillway gate near Lone Pine, CA was dynamited. Six months later, on November 16, 1924, sixty or so men led by one of the two Watterson brothers (who were the Owens Valley leaders of the resistance to the city of Los Angeles) overwhelmed a BWWS armed guard at the aqueduct control gates near the Alabama Hills (a location that was just starting to become a popular locale for Hollywood westerns), opened the gates to release water into the Owens River and occupied the headwork for the Los Angeles Aqueduct for several days.
Mulholland originally preferred Big Tujunga Canyon for the dam site; however, the local ranchers requested exorbitant fees for their properties (one owner requested $500,000 while the city was offering $12,000) Mulholland sited the dam for this reservoir at a gorge in the San Francisquito Canyon. The dam site would be located between two of the Bureau’s powerplants – about a mile and a half upstream from the Bureau’s San Fernando Power Plant No. 2 (placed in service in 1920) and about 5 miles downstream from San Fernando Power Plant No. 1 (placed in service in March 1917). In 1922 Mulholland had promised the commissioners of the city’s Board of Water Supply that the dam would be sized to store a year’s supply of water for the city of Los Angeles. This called for a reservoir capacity of 30,000 acre-feet. Due to the continuing population growth, it was decided to increase the reservoir capacity to 32,000 acre-feet in July 1924, shortly after construction had commenced, and then increased its’ capacity again to its’ ultimate capacity of 38,168 acre-feet in July 1925. This resulted in an addition to the dam height of 20 feet without increasing the base width of the dam.
Construction on the dam began in April 1924 and the first concrete was placed into this concrete gravity-arch dam in August 1924 and the dam would be completed in May 1926. As with the Los Angeles Aqueduct, construction was performed by the city using their own work forces. There are a number of details involving the planning, design and construction of the dam that merit discussion (especially so in light of its eventual collapse), but I think it best to leave those details to a bit later. Below is a tabulation of data for the completed dam.
Crest of Parapet . . . 1838.06 feet
Crest of Spillway Lip . . . 1835.00 feet
1st Outlet upstream elevation . . . 1799.00 feet
2nd Outlet upstream elevation . . . 1763.00 feet
3rd Outlet upstream elevation . . . 1727.00 feet
4th Outlet upstream elevation . . . 1691.00 feet
5th Outlet upstream elevation . . . 1658.26 feet
Size of each Outlet Pipe . . . 30 inches
Natural Grade . . . 1650.00 feet
Bottom of maximum section . . . 1630.00 feet
Length of main dam measured at centerline of dam crest . . . 700 feet
Length of wing dyke . . . 588 feet
Maximum thickness at base as per plans . . . 176 feet
Indicated maximum thickness from photos . . . 156 feet
Thickness at Dam Crest . . . 16 feet
Radius of arch of dam measured to the upstream face at the crest . . . 500 feet
Volume of Concrete in main dam . . . 130,446 cubic yards
Volume of Concrete in dyke . . . 3,826 cubic yards
Width of steps (offsets) on downstream face of dam . . . 5.5 feet at the base decreasing to 1.4 feet at the top
Spillway . . . 11 panels measuring 20 feet wide by 1.5 feet high clear inside dimensions
Reservoir Area . . . 600 acres
Reservoir Capacity . . . 38,168 acre feet
Distance from dam to head of reservoir . . . 2.8 miles
Capacity of canal that returned reservoir water to the Los Angeles Aqueduct below Powerhouse No. 2 . . . 1056 cfs
Cost . . . $1,250,000
Before getting any further along I should make a point regarding terminology. In water resources engineering, the directions left and right are always referenced based on the observer looking downstream. For the St Francis Dam, the left abutment is also the east abutment and the right abutment is also the west abutment. I wanted to make this clear because I’ll probably use these directions interchangeably from here on out.
A number of myths have developed surrounding the St Francis Dam, some of which are false and some of which require a more nuanced understanding of the events in question. One such myth is the idea that the dam failed upon initial filling. That is simply not so.
The BWWS diverted water from the aqueduct and began filling the reservoir in March 1926 and filled the reservoir to elevation 1780’. The reservoir was then lowered by 20 feet over the summer months this water being used to satisfy BWWS’s customers during the peak season for water demand. Over the 1926-27 winter the water level in the reservoir was steadily raised until it reached elevation 1832 – three feet below the spillway sills. It remained at 1832 for three weeks before the BWWS began withdrawing water again for the 1927 summer season and fluctuated between elevations 1813 and 1819 for the rest of the year. Over the next winter the water was again slowly raised until it reached elevation 1834.75 – three inches below the spillway sills on March 2 1928. It would stay there until disaster struck. After the initial filling the dam experienced some seepage problems. Most notably, seepage along a portion of the right abutment would continue from relatively early on continuously. However, the discharge water was so clear that a small pipe was set up to deliver the seepage to the nearby residence of the damkeeper, Tony Harnischfeger, for his domestic use (muddy discharge would tend to indicate a potentially serious piping problem). In addition, four large vertical tension cracks more or less equally spaced from the outlet pipes located at the center of the dam developed and expanded as the water level was raised during the 1927 season. They leaked profusely and the BWWS dealt with this by grouting and caulking each crack with oakum. During 1928, similar cracking occurred at the wing dyke. Since the BWWS design failed to include any contraction joints for this 130,000 block of concrete, it is likely, this cracking was due to the normal shrinkage associated with cement heat of hydration and played no role in the dam’s ultimate failure.
Around 8 AM on March 12, 1928, dam keeper Tony Harnischfeger called his superiors in Lo Angeles to report that muddy water was leaking from the west abutment. Mulholland and his assistant, Harvey Van Norman, immediately made the trip from downtown Los Angeles to investigate Harnischfeger’s concern. They arrived on site around 10:30 AM and with Mr. Harnischfeger made an inspection trip over the entire dam starting with the new leakage. They found the leak to be running clear – the muddy appearance was due to contact with soil as it exited the dam. They did not detect anything that could portend imminent failure of the dam and they left the dam around 12:30, about 11-1/2 hours before the dam would collapse but first stopped at Power Plant No 2 at which they directed employees at that site to 1) stop flows into the reservoir, and 2) to open three of the six 30” diameter outlet pipes. This would have been too little, too late to prevent the impending dam failure; but, I believe their goal was just to minimize the new leakage they’d just inspected until it could be repaired.
On March 12, 1928 at 11:57-1/2 PM the St Francis Dam would fail and the first victims would be dam keeper Harnischfeger, his girlfriend and his young son. They would be the first of many. The wall of water 140-foot-high as it left the dam trundled 7300 feet down the canyon to Power Plant No.2 where the now 110’ high floodwaters would obliterate the power plant and take the lives of all but three of the 67 BWWS employees and family who lived at the site (Many but not all Power Plant employees lived on site at BWWS-provided cottages). The flood wave would start as a slurry of water and sediment and would continue to accumulate sediment and debris as it wended its way along its 54-mile route to the sea (San Fransiquito Creek to Santa Clara River) destroying everything in its path. The destruction included the Edison Saugus substation, the Ridge Route bridge at Castaic Junction thereby severing the main route between Los Angeles and northern California and an Edison construction camp near the Kemp RR siding killing 84 of the 140 construction workers camped there (it should be noted that this total would’ve been much higher if not for the night watchman’s actions in alerting the sleeping workmen. The night watchman did not survive the flood). It would spread out as it passed through the Santa Clara River Valley eventually to a width of 2 miles as it flowed into the Pacific Ocean about 5 and a half hours after the failure.
The Outland and Wilkman books provide accounts from a number of the survivors. These accounts describe utter chaos, confusion and sheer terror. The noise from the oncoming flood was deafening from miles away. Many survivors described lights heading toward them and many thought the oncoming flood waters was a freight train. One thought emergency vehicles were coming their way.
The disaster also brought out the best in many. As the floodwaters raged and powerlines and substations went out, calls about the approaching disaster came into the towns of the Santa Clara River valley – Fillmore, Saticoy, Santa Paula, et al. The operators in these towns (all of them women) stayed at their board, transferring calls, contacting emergency personnel and relaying messages even though they had no way to know if they would end up in harm’s way. Local police and sheriffs raced up and down the valley in squad cars and motorcycles sirens blaring, stopping and knocking on doors warning of the danger. One motorcycle cop Thornton Edwards barely survived the flood and became known as “the Paul Revere of the St Francis Flood.” These heroes and heroines saved many lives. A steel sculpture “The Warning” was completed in 2003 in Santa Paula in honor to these heroes.
Finally, here’s a tabulation of the flood from the Outland book;
Location . . . Arrival Time . . . Elapsed Time . . . Distance . . . Avg. Velocity
St Francis Dam . . . 11:57-1/2 PM . . . ———— . . . ———– . . . ———–
Powerhouse 2 . . . 12:02-1/2 AM . . . .5 minutes . . . 1.5 miles . . . 18.0 mph
Saugus SubStation . . . 12:40 AM . . . 37.5 min . . . 8.0 miles . . . 12.8 mph
Edison Camp . . . 01:18 AM . . . 38.0 min . . . 7.9 miles . . . 12.5 mph
Barsdale Bridge . . . 02:20 AM . . . 62.0 min . . . 12.7 miles . . . 12.3 mph
Santa Paula Bridge . . . 03:05 AM . . . 45.0 min . . . 8.4 miles . . . 11.2 mph
Saticoy Bridge . . . 04:05 AM . . . 60.0 min . . . 7.4 miles . . . 7.4 mph
Ocean . . . 05:25 AM . . . 80.0 min . . . 7.9 miles . . . 5.9 mph
Fr Dam to Ocean . . . 5 Hours, 27.5 minutes . . . 53.8 miles . . . 9.8 mph
Aftermath: Recovery and Restitution
An exact death total for the disaster has never been established. For months afterward bodies washed ashore along the southern California coastline. Charles Outland, writing in 1977, stated that “…any death total over 450 or under 400 is unrealistic.” Jon Wilkman in his 2016 book considers a death total of 500 or thereabouts to be the most realistic figure.
After some initial wariness, civic leaders from the city of Los Angeles and from the Santa Clara Valley realized it made sense to work together to reimburse flood victims for their damages (for death and injuries as well as property damage) and to repair and replace. An entity called the Joint Restoration Committee was created with seven members representing Los Angeles and seven members representing Ventura County, led by George Eastman president of the Los Angeles Chamber of Commerce and Charles Teague, a prominent banker and rancher in Ventura County. All the costs associated with this effort would be borne by the City of Los Angeles, all payments required that city was not accepting liability for any damage, and there would be a six month statute of limitations during which claims could be made. The City of Los Angeles signed an “at cost” contract with the Associated General Contractors to clean up all flood debris and repair/replace all damaged property. Most observers agreed that this process was a success.
Although detailed accounting records were kept, there is also no agreed figure as to the overall cost of the St Francis Dam catastrophe. Outland provides detailed breakdowns of the costs for death payments, injury payments, repair and replacement costs, and agricultural and land reparations costs. He puts the total cost at around $13,500,000. Wilkman puts the cost at around $30,000,000.
Aftermath: Why Did It Happen?
BWWS officials and others traveled to the dam site on the morning of March 13 to see what was left of the dam and to begin to ascertain what had gone wrong. All that was left of the dam was the center section (which would come to be called “the tombstone”) and the wing dike that extended 600 feet beyond the right abutment. The rest of the dam had broken into numerous pieces of varying sizes (the largest weighing as much as 10,000 tons) and laying in a variety of locations – some near where they had once stood and others up to 3,000 or so feet downstream. Also, readily visible was a large landslide extending well above the height of the dam where the left abutment had once been.
It was also necessary to investigate the dam failure and attempt to determine the cause(s). Several different commissions and entities were put into place for this purpose; but, only two mattered at the time – a commission put into place by California Governor C. C. Young, and the Los Angeles County Coroner’s Inquest.
In the immediate days and weeks after the disaster numerous unfounded rumors as to the cause arose and spread like wildfire. Mullholland started the first one when he was asked shortly after the disaster what happened and speculated “an investigation will prove there was an enormous earth movement preceding the flood”. This theory did not last long as Harry O. Wood, chief of the Seismological Laboratory of the Carnegie Institution of Washington (soon to become part of the California Institute of Technology) noted that there had been no recorded earthquake activity on the night of March 12. Another theory put forward by BWWS partisans was that the dam had been dynamited and they provided as evidence the number of dead fish found downstream of the failed dam. Based on the previous violence on the part of Owens Valley this was a possibility. In any event, these claims necessitated autopsies of the dead fish which found heavily silted gills – the fish had died of asphyxiation.
California Governor C. C Young appointed a six-man Board of Inquiry comprised of four prominent engineers and two prominent geologists to investigate the causes leading to the failure of St Francis Dam. The Board convened on March 19, one week after the failure, made either one or two site visits (a visit on the 22nd and maybe an earlier visit on the 20th sources vary on this) and submitted their 79-page report to Governor Young on March 27th in Los Angeles. Their conclusion was that the foundation material along the right abutment (called the Vazquez Formation at the time and now called the Sespe formation) was unsuitable for a dam foundation due to its’ tendency to slake when exposed to water and the most likely cause of the failure was seepage in this area causing piping and eventual failure. The gradual decline in the dam’s water surface elevation in the 40 minutes before the rapid decline indicated on the Stevens Automatic Water Stage Recorder (here is a photo and description of a similar type recorder) was cited in support of this theory.
The Board was composed of serious, highly-regarded and well-intentioned men, but it seems to this observer their report and the haste in which it was completed does not speak well of the process. The Board came to their conclusion long before all available physical evidence had been obtained (for example, the BWWS was still surveying and attempting to ascertain the disposition of the various concrete blocks and new geological investigations by BWWS were also not yet completed), and without interviewing more than a few witnesses – all of whom were BWWS employees. In addition, the Board relied on others to gather data, perform measurements and did not have complete access to project documents (drawings as well as construction photographs) which the BWWS had turned over to the Los Angeles County Coroner’s Office. In addition, it appears they did not take time to digest and fully comprehend the information they were in possession of. Despite this, the Board’s conclusion would become the accepted account of the St Francis Dam failure.
The Coroner’s Inquest began on March 21 (pdf of transcript). It would include several days testimony from dozens of witnesses – starting with BWWS engineers from Mr. Mulholland and on down the chain of command, including construction laborers, relatives of the deceased and others who had grievances to air – and a field trip to the dam site. It is most notable for two things – two statements made by Mr. Mulholland and a finding/recommendation made. During his testimony, Mr. Mulholland made two statements that stood out. At one point he stated, “the only ones I envy in this thing are those who are dead”. More importantly the record shows that he took responsibility for the failure stating “Don’t blame anyone else, you just fasten it on me. If there was an error in human judgement, I am that human”. The main finding of the Inquest would change the way projects of this type would proceed in the future. It found the following; “A sound policy of public safety and business and engineering judgement demands that the construction and operation of a great dam should never be left to the judgement of one man, no matter how eminent, without check by independent expert authority, for no one is free from error, and checking by independent experts will eliminate the effect of human error and insure safety”.
Recent Forensic Studies
In the 1980’s a Geological Engineering professor became interested in the Saint Francis Dam story and began to study it. This engineer, J David Rogers, began his career as an investigator of the 1976 Teton Dam failure. He probably knows more about the Saint Francis Dam than anyone else. He has spent many hours at the site, mapping the geology and has read and studied all of the engineering documents and literature on the topic. With the aid of modern software, he has modeled and developed a comprehensive theory of the failure mechanism and process. A good discussion of his theory is presented in the 2016 Wilkman book.
Before discussing Professor Rogers theory, I first need to provide a little more information regarding the underlying geology at the dam foundation. As discussed earlier, the right abutment was composed of Vazquez Formation which the Governors Board fingered as the most likely culprit for the failure. However, the geology at the left abutment was every bit as problematic as that of the right. The left abutment geology consists of Pelona Schist overlain by an ancient landslide. This material continues across the valley channel and about a third of the way up the right abutment where it meets the Vazquez Formation at a fault. A geologic cross section of the dam foundation is provided. The first to recognize the potential danger associated with building a dam against this ancient landslide was noted geologist Bailey Willis in an article published in the June 25, 1928 issue of Western Construction News, but to no effect.
Getting back to Prof Rogers – His extensive study and analysis of the dam including the use of modern analytical software such as Discontinuous Deformation Analysis, leads him to conclude that a massive landslide on the east side abutment was the cause for the collapse of the dam. In addition to the east abutment geology, he marshals an avalanche of physical evidence in support of his hypothesis. Southern California Edison (SCE) operated a 70 kV powerline delivering power to Palmdale/Lancaster which ran along the east side to the reservoir. The line had two poles above the east abutment. This line went out at 11:57:30 PM causing a blackout at BPL’s San Fernando Powerplant No. 1. San Fernando Power Plant No. 2 would go offline at 12:02:30 AM when the flood demolished the site. The powerline was severed when the east abutment slide caused the powerline poles to fall away. Additionally two witnesses who travelled along the powerline road above the east abutment several hours prior to the dam failure testified that they had to negotiate dips in the road which had not been there before. Most likely the Governor’s Board was unaware of the powerline failure because the powerline and had been relocated and put back in service by SCE by the time the Board visited the site. Rogers also notes that the massive landslide of 700,000 or so cubic yards mixed in with the escaping reservoir water provides the necessary buoyancy to carry the large chunks of concrete as far downstream as they did. The gradual drawdown of the reservoir water level for 40 minutes before the abrupt failure as indicated by the Stevens recorder is due not to leakage at the west abutment, but to the dams center section tilting downstream.
I don’t know whether Rogers theory of failure is correct. However, it is the only theory to either answer or at least be consistent with the physical evidence encountered.
Design and Construction Deficiencies
The design and construction of St. Francis Dam involved many deficiencies. Some of these deficiencies were critical in its’ eventual denouement, while others played no role. Also, some of the deficiencies should be considered as deficiencies only in hindsight as the theory and practice of dam engineering was rapidly evolving at the time. In any event, I will provide a brief list of some of these deficiencies. I’ll try to list them in something close to the order they might affect the integrity of any dam.
- The dam was unknowingly founded upon an inadequate foundation. The left abutment was placed against a paleo-landslide, while the right abutment consisted of a material that tended to slake when exposed to water. The most likely reason for the poor siting of the dam was that the geological investigation of the dam site during the planning and design phase was perfunctory.
- The designers raised the height of the dam by twenty feet with no corresponding increase in its’ width causing the structure to be statically unstable when the water level in the reservoir got near the maximum water surface.
- Hydraulic uplift was ignored in the design of the dam leading to a lower Factor of Safety than the engineers planned.
- Hydraulic uplift wells were provided only in the main section in the channel, but not under either abutment. The center section with the relief wells is the only section of the main dam to remain standing.
- The dam keyways at the abutments and along the channel were insufficiently deep.
- Concrete production and placement procedures were deficient in several ways. First, this 205-foot high, 130,000 cubic yard water-retaining structure was built with no contraction or expansion joints thus guaranteeing that a number of large cracks would develop over the dam’s service life which would require on-going remedial work on an ad hoc basis. Second, aggregates for the concrete were not washed and, except for screening out oversize material, aggregates were not graded. Third, construction joint preparation was deficient. Laitance was allowed to stay in place between each concrete lift. It should be noted that subsequent testing of the concrete revealed that it had sufficient strength for its purpose and played no role in the dams ultimate failure.
There are more deficiencies, but this post is already too long.
One of the most interesting aspects of the St. Francis Dam failure is how it impacted many other stories and still affects life. Below is a brief recital of these downstream impacts;
- State Inspection & Regulation of Dams – In the wake of St. Francis, the California legislature passed a comprehensive dam safety bill in 1929. Under the act, all dams (except federal dams) that were either 25 feet or greater in height or which impounded 50 or more acre-feet were to be reviewed by the State Engineer and monies were set aside for that purpose. Over time this responsibility morphed into a permanent state Safety of Dams organization.
- Civil Engineer Licensing – The Civil Engineers Registration Act was also signed into law by the Governor of California in 1929. The act set forth requirements (age, residency, education & experience) and responsibilities for future licensed civil engineers. A written test was required from 1930 on. However, the law also included a grandfather clause which over 5,700 individuals applied for and of which a little over 5,000 were accepted by the new Engineer’s Licensing Board. It would take another 25 years before the state would license its next 5,000 engineers. So, the St.Francis Dam failure not only ended Mulholland’s engineering career; but, it would also lead to the end of his type of engineering career path – that of the hard working young man with little formal education but plenty of desire starting with a shovel or wrench in his hand and slowly work his way into and up the ladder of the engineering profession.
- Mulholland Dam Engineering Reviews and Modifications – In 1923-24, the BWWS had constructed the Weid Canyon Dam (renamed Mulholland Dam in 1925) in the Hollywood Hills and had put it into service as a reservoir. In the aftermath of St. Francis it came to light that Mulholland Dam had been the design template for St. Francis. This rightly caused consternation among those living and working below the dam. There followed one study after another of the dam. First, a restriction was placed on the dam limiting it to half capacity. Later, in 1933-34, a large buttress fill of compacted earth was placed against the dam’s downstream face. Both the volume restriction and buttress fill remain in place to this date.
- Boulder Canyon Act & Design of Hoover Dam – The Reclamation Service had issued a report in 1922 for a large dam along the Colorado River to serve as both a flood control facility and as a source of electrical power. The St. Francis Dam failure put the future of what would become Hoover Dam into doubt. The Boulder Canyon plan had been promoted by Los Angeles politicians (and William Mulholland) for years. With St. Francis, politicians from Utah and Arizona who had long felt their state(s) had not come out well in the 1922 Colorado River Compact allocating water rights of the seven states contiguous with the river, saw an opportunity to either kill the plan or get a better deal for their state. What followed was the formation of a board – the Colorado River Board – comprised of five members – two noted engineers, two noted geologists and chaired by a retired Major General, William L Sibert. The Board was charged with reviewing documents and recommending the final site for the proposed dam. The Board recommended the Black Canyon site and also suggested the following; 1) Reduce the contact pressure against the foundation rock from 40 tsf (tons per square foot) to 30 tsf, 2) increase the capacity of the diversion tunnels from 100,000 cfs to 200,000 cfs, 3) increase spillway capacity to greater than 11,000 csf, and 4) increase the volume of active flood storage. The Bureau would incorporate all of these suggestions except for the reduced foundation pressure.
- The BWWS would need a replacement for the St. Francis Dam. They would find the site for the new reservoir at Bouquet Canyon which is about 3 miles east of the St. Francis site. The new reservoir would require two earth embankment dams – a main dam 221 feet high and 2,125,000 cubic yards in volume and a smaller saddle dam (42 feet high and 135,000 cubic yards in volume). During construction of the new reservoir, the BWWS engineer in charge of construction, Ralph Proctor developed a new method for testing the in-place compaction of cohesive soils. He would publish his test procedures in a series of articles in Engineering News Record in August and September 1933. Proctor’s test method used the soil dry density as the standard and allowed for the two variables of soil compaction (compactive effort and moisture content) to be tested separately. The test was rapidly adopted within the engineering industry and is still the industry standard.
Odds and Ends
As soon as the road to the San Francisquito Canyon was re-opened the dam site became a destination for sight-seers. The main attraction was the still-standing Tombstone. And, with its stair-step configuration it was an attractive nuisance. On May 27, 1928, an eighteen- year- old Leroy Parker would fall from the Tombstone to his death when he was startled by a friend who threw a snake at him. His family sued the city and the BWWS decided the Tombstone needed to go. On May 10, 1929, the Tombstone was dynamited in place.
The San Francisquito dam site is still undeveloped so one can still see what is left of the dam and its ruins. Concrete blocks or what is left of them are still visible lying where they ended up on March 12/13 of 1928, the wing dike or what is left of it is still in place and the east abutment is also in plain. The site is 30-40 minutes north of Los Angeles and for anyone considering a trip to the site here are driving directions.
Instead of reading my long post, you can get the gist of the St. Francis Dam story in this Frank Black & the Catholics tune “St. Francis Dam Disaster.”
Links for Books
- Man-Made Disaster: the story of St. Francis Dam by Charles F. Outland
- Floodpath by Jon Wilkman
- Los Angeles Aqueduct Final Construction Report by Los Angeles Board of Public Service Commissioners
- The History of Large Federal Dams: Planning, Design, and Construction by David P. Billington, Donald C. Jackson, Martin V. Melosi