In This Issue
Summer Bridge Issue on Engineering for Disaster Resilience
July 1, 2019 Volume 49 Issue 2
The articles in this issue present examples of engineering innovation to develop resilient infrastructure.

Improving the Resilience of Southern California Water Supply Aqueduct Systems to Regional Earthquake Threats

Monday, July 1, 2019

Author: Craig Davis and John Shamma

Southern California is home to over 19 million residents who rely on three independently owned and operated aqueduct systems that provide on average 50 percent of the region’s annual water needs. The three aqueducts import water from outside the region and cross the San Andreas Fault. The potential loss of all three aqueducts in a single major San Andreas earthquake event poses a serious threat to Southern California and the United States economy.

In response to this threat, a multiagency Seismic Resilience Water -Supply Task Force was created to examine assumptions regarding earthquake-induced damage, review likely repair durations for the aqueduct systems, and develop a coordinated approach to restore water supplies to the region if an earthquake event compromises one or more of the aqueducts.

The Southern California Aqueduct Systems

The three aqueduct systems that import water to Southern California are the California Aqueduct as part of the State Water Project (SWP), -Colorado River Aqueduct (CRA), and Los Angeles Aqueduct (LAA) (figure 1). The independently owned and operated systems work in conjunction to supplement local water resources and meet the demands of the region. The overlapping of the supply operational areas provides an opportunity for the responsible agencies to coordinate their efforts should any of the aqueducts become damaged from a seismic (or other) event.

The California Department of Water Resources (DWR) operates the SWP, the nation’s largest state-built water development and conveyance system. Placed into service in the early 1970s, the SWP provides water for 25 million residents, farms, and businesses statewide, including 19 million people in the Southern California urban area (figure 1) via the California Aqueduct’s East and West Branches. The water is purchased and distributed by various state water contractors; the Metropolitan Water District of Southern California (MWD) is one of these contractors.

MWD is the nation’s largest wholesale water agency and owns/operates the CRA, which was constructed in the 1930s to import water from the Colorado River to Southern -California. MWD also owns and operates five water treatment plants and over 800 miles of pipeline to distribute water imported from the CRA or the SWP California Aqueduct East and West Branches to member agencies. Additionally, an interconnection between the LAA and MWD system can be used to supplement the portion of MWD’s service area that is normally supplied by the SWP West Branch with water from the LAA.

The Los Angeles Department of Water and Power (LADWP, an MWD member agency) is the nation’s largest municipal water agency and the owner/operator of the LAA system, which was constructed in the early 1900s and expanded in the late 1960s to import water from the Owens -River in the Eastern Sierra Nevada. The LADWP and the area supplied by the LAA can also be supplied by SWP and CRA water.

The San Andreas Fault

As seen in figure 1, the three aqueducts cross the San Andreas Fault (SAF), which runs nearly the entire length of California.[1] It is primarily a strike-slip fault moving with right-lateral motion; however, there is a compressional portion in the vicinity of San Gorgonio Pass (about 90 miles east of Los Angeles) where the fault would be expected to have significant uplift (-Weldon et al. 2016), which could be especially disruptive to the flow of water through aqueducts.

 Figure 1

The SAF has produced earthquakes of moment magnitude (Mw) 7.8–7.9 (in 1857 and 1906). Major earthquakes of Mw 7.25 or larger on the southern SAF occur every 100–150 years (Scharer et al. 2017). It has been over 150 years since the last earthquake ruptured across locations of the SWP and LAA and over 300 years since the last rupture crossed the CRA location (Philibosian et al. 2011; Scharer et al. 2010).

Assessment of the Threat: The ShakeOut Scenario

In 2008 the US Geological Survey, along with many partners, completed a study of a plausible Mw 7.8 earthquake on the southern SAF (Jones et al. 2008) for use in the Great Southern California ShakeOut exercise (hereafter, the ShakeOut Scenario). Figure 2 shows the region, ruptured portion of the SAF (black), shaking intensity for this scenario, and how it affects the broad Southern California urban area (figure 1). The hypothesized Mw 7.8 earthquake has an epicenter at the Salton Sea and ruptures along a 300 km (190 mile) distance north to Lake Hughes (Graves et al. 2008; Hudnut et al. 2008).

Figure 2

The ShakeOut Scenario was used to enhance understanding of the effects of a Mw 7.8 event on critical infrastructure and identify the physical, social, and economic consequences. It provides a realistic illustration of possible fault displacement and shaking throughout the region and useful insight into water supply impacts.

A major SAF earthquake could have severe impacts on aqueduct operations, including the potential to completely disrupt all imported water supplies to Southern California until repairs can be made.

Challenges to Repair and Service Restoration

According to the ShakeOut Scenario, the three aqueducts would be significantly damaged by fault movements, ground shaking, liquefaction, and permanent ground movements (differential settlement, lateral spreading, slope and embankment deformations). Afterslip and aftershocks would continue to compromise the aqueducts and slow repair efforts (Davis 2009, 2010; Davis and O’Rourke 2011). Supply and distribution to hundreds of water systems, large and small, would be disrupted, affecting customers and hampering the ability to fight fires.

The duration of disruptions would be a function of damage level and repair time. Restoration times would be affected by resource conflicts (labor, equipment, materials, supplies), transportation difficulties after the earthquake, and additional impairment from aftershocks and afterslip. In addition, fault displacements, shaking, and other ground failures would disrupt -other critical lifelines such as highways, railroads, power transmission, oil and natural gas, and fiber optic communication lines, impeding aqueduct repair efforts and severely impacting social and economic activity.

Independent studies based on the ShakeOut -Scenario noted that repairs to all three aqueducts would likely require more than 6 months. This timeframe is longer than previously anticipated, as described in the following section, revealing new challenges for regional response and recovery efforts (Davis 2009, 2010; Davis and O’Rourke 2011). The study authors recom-mended “a supply agency coordination team consisting of -LADWP, MWD, and the DWR to coordinate aqueduct post-earthquake response and recovery and pre-earthquake mitigation efforts” to improve the regional seismic resilience of the imported water supply systems (Davis and O’Rourke 2011, p. 474).

Economic Impacts

Total losses from the ShakeOut Scenario are estimated at $213.3 billion, with approximately 25 percent due to impacts on water systems (regional supply and local distribution): “business interruption from water disruption is greater than 50% of total business interruption losses because of long duration of outage (4–6 months in heavily impacted areas)” (Jones et al. 2008, p. 279). This gross loss in output from water disruption alone is enough to drive the region into a recession. The results assume that the LAA, SWP, and CRA are fully restored to service within 6 months. But as noted above, this assumption may be optimistic. Water disruption -longer than 6 months could lead beyond a recession to a regional economic catastrophe (Jones et al. 2008; LADWP 2014).

Countermeasures to the SAF Threat

Historical Measures

The aqueduct systems were generally designed and constructed with knowledge of the SAF hazard and precautions were taken to help mitigate impacts from a fault rupture. For example:

  • The City of Los Angeles developed large water supply reservoirs south of the SAF (Lund and Davis 2005).
  • The California Aqueduct West Branch crosses the SAF at Quail Lake, a natural body of water that may allow aqueduct water to flow safely across a rupture at that location.
  • The SWP is equipped with gate structures at strategic locations along its alignment to allow water to be distributed upstream of the gate while repairs from shaking and fault rupture are made downstream.
  • MWD designed the CRA with an additional 2.3 m (7.5 ft) of head to account for potential gradient -changes resulting from fault movement. It also used inverted siphons or shallow conduits for crossing identified active traces of the fault to make damaged areas more accessible for repairs, and segmented conduit sections rather than typical monolithic construction to reduce potential damage from displacements (Hinds 1938).
  • MWD established emergency storage requirements in 1991, as described below.

Regional Emergency Water Storage

MWD’s emergency water storage facilities, located on the coastal side of the fault, can supply the region after a seismic event on the SAF. The emergency storage requirements are based on the potential for a major SAF earthquake to damage the CRA, SWP, and LAA.

MWD coordinates with its member agencies to set criteria for emergency storage, assuming that damage from a major SAF earthquake could render the aqueducts out of service for 6 months (MWD 1991). The objective is to ensure water supply and delivery to all member agencies during the outage. This objective allows MWD to provide 75 percent of member agencies’ retail demand under normal hydrologic conditions, supplementing local water production to avoid severe water shortages while the aqueducts are out of service.

A major accomplishment to meet local emergency water storage objectives was the completion of -Diamond Valley Lake in 2000, which nearly doubled the available emergency supplies in Southern California. In addition, MWD may draw from water-bank agreements during an emergency, if necessary and available.

Ongoing Collaborative Efforts

Resilience by Design

Under Mayor Eric Garcetti, a strategic plan for the City of Los Angeles, Resilience by Design (Mayoral Seismic Safety Task Force 2014; Jones and Aho 2019, this issue), was developed, with specific recommendations for improving the seismic resilience of the Los Angeles water system. One key recommendation is to “create a Seismic Resilience Water Supply Task Force . . . with the LADWP, MWD, and DWR, in an effort to create a collaborative and regional approach to protecting the resilience of our water supply.” This recommendation was informed by the ShakeOut Scenario, previously mentioned studies of the regional water supply, and LADWP (2014). As part of the Mayoral Seismic Safety Task Force, the LADWP developed an internal water system resilience program (LADWP 2014, which made the same recommendation as Davis and O’Rourke 2011).

Seismic Resilience Water Supply Task Force

In August 2015 a multiagency task force was formed to collaborate on postearthquake response and recovery, undertake studies, and implement mitigation measures to improve the seismic resilience of imported water supplies to Southern California using a regional approach (Clark et al. 2018; Davis et al. 2017). This unprecedented collaboration, which required overcoming institutional constraints and differences in organizational culture to achieve unity of purpose and commitment, has great potential for significantly improving resilience and reducing risk on a regional level.

 Figure 3

The task force, whose organizational structure consists of a Management Oversight Committee, Seismic Resilience Team, and working groups, is composed of managers and staff from the planning, engineering, and operations groups of each agency: DWR, LADWP, and MWD (figure 3). Recognizing that damage cannot be prevented from a major SAF event, the agencies agreed that “The Task Force will collaboratively develop strategies to mitigate risks and prepare for an event on the SAF with the goal of restoring imported water supplies to Southern California rapidly” (Clark et al. 2018).[2]
 The task force goals are to

  • establish a framework for the agencies to coordinate response and recovery efforts;
  • establish a common understanding of indi-vidual agency aqueduct seismic vulnerability assessments, -projected damage scenarios, and planning -assumptions;
  • revisit historical assumptions regarding potential aqueduct outages due to seismic events; and
  • discuss ideas for improving the seismic resiliency of Southern California’s imported water supplies through multiagency cooperation.

The task force will determine steps to develop a coordinated response to emergencies. It already maintains an emergency contact list for agencies to communicate in the event of a major SAF earthquake, and is completing a memorandum of understanding on how they will work together following a serious earthquake. A tabletop exercise was conducted, and a multiagency recovery plan is under development.

Aqueduct Workshop

With an initial focus on aqueduct assessments and mitigation and emergency preparedness and response, the first major task force activity was a workshop to establish a common understanding of seismic vulnerabilities on each aqueduct, revisit historical planning assumptions, and identify action items to increase seismic resilience. The March 2016 workshop focused on (a) potential damage to Southern California’s imported water aqueducts from a major seismic event on the southern SAF, (b) potential outage durations for restoration of partial and full aqueduct capacities, (c) identification of regional priorities, and (d) mitigation options (Clark et al. 2017). The format allowed the candid exchange of information and ideas among all participants.

Agency participants considered preparation for and response to the ShakeOut Scenario from a regional perspective and were asked, “If all aqueducts were owned and operated by a single agency, what steps should be taken now to mitigate potential damage and what would the priority of repairs be to most rapidly restore imported water deliveries to the region?” From their responses the task force concluded that recovery times would exceed historic planning assumptions, restoration of partial aqueduct flows could take at least 2 months, and restoration of full aqueduct capacities could exceed 6 months.

Another finding from the workshop is the importance of dependencies of the three aqueduct systems on other lifeline systems for continued operations and the ability to rapidly recover. The most important systems are transportation, communication, and electric power.

The workshop helped develop a consensus on potential outage durations and specific actions to -better prepare for large -seismic events. The region will be best served if the three agencies

  • move forward with recently identified mitigation projects on the CRA and LAA;
  • prioritize additional known vulnerabilities on the CRA, LAA, and SWP;
  • execute an agreement to allow for a coordinated response to emergency events; and
  • develop a joint emergency recovery plan to share resources and prioritize repairs when responding to emergency events.

LADWP’s Resilience Expert Panel noted the significance of the nation’s largest municipal utility -(LADWP), largest water wholesaler (MWD), and largest state-owned water agency (DWR) joining together to address a major hazard for the first time and encouraged the task force to continue working together long into the future. Common issues can be studied more efficiently and Southern California will be better prepared for seismic events if the task force continues its efforts to facilitate coordinated vulnerability assessments, evaluate mitigation options, and develop agreements for coordinated emergency responses to major seismic events.

Soon after the workshop, the agencies developed a 5-year plan to accomplish the identified goals and additional collaborative actions, such as sharing approaches to facility vulnerability investigations and conducting multiagency emergency exercises, including emergency communications. Efforts were also initiated to work with the LADWP power system and Southern -California Edison to address aqueduct system dependencies on the electric power systems.

Investigation of Aqueduct Enhancements

The LADWP is developing a seismic enhancement -project where the LAA crosses the SAF through the 8 km (5-mile) long Elizabeth Tunnel. It consists of reinforcing the tunnel at vulnerable locations and installing two 61 cm (24-inch) diameter high-density -polyethylene pipes for approximately 244 m (800 feet) across the SAF zone (figure 4). The objective is to increase the probability of supplying water across the fault zone for events that do not fully rupture the 2.7 m (9-foot) wide horseshoe-shaped concrete-lined tunnel.

 Figure 4

The LADWP is also undertaking other seismic improvements along the LAA and a detailed investigation to characterize the SAF at the Elizabeth Tunnel crossing. These efforts will inform the planning to engineer a solution for the potential maximum fault offset (Lindvall et al. 2018).

MWD is continuing to evaluate and improve the CRA seismic resilience. Recent CRA seismic vulnerability assessments identified the Whitewater Tunnel No. 2 as the most likely component of the system to sustain major damage from a strong earthquake on the southern SAF (MWD 2016; Weldon et al. 2014, 2016). The tunnel’s alignment closely parallels the SAF and crosses a splay of the fault. An event like the ShakeOut Scenario could damage the tunnel at the fault crossing and shallow sections near the tunnel portals.

Given the specialized nature of tunnel repairs, a workshop convened industry experts in tunnel engineering and construction to identify repair options and estimate repair durations. Participants concluded that, with preevent planning, design, and upgrades, repairs to -Whitewater Tunnel No. 2 could be completed within 6 months. These preevent steps include (1) design and construction to strengthen vulnerable tunnel sections near the two portals, (2) design and construction of a new access structure at the west portal, (3) design of a new tunnel section in advance of an earthquake to bypass collapsed and/or blocked portions of the existing tunnel and enable construction to proceed quickly after an event, (4) stockpiling of steel sets needed for construction, and (5) prequalification of tunnel repair contractors. MWD has initiated preliminary design of the tunnel upgrades.

Potential interconnections between the aqueducts are being investigated as one way of increasing system resilience, as discussed during an aqueduct workshop in March 2018. Initial planning efforts are being under-taken to connect the LAA with the California Aqueduct East Branch near the Elizabeth Tunnel north portal. This project increases opportunities both to import water by transferring between the two systems after different types of SAF events and to accelerate seismic improvements along the LAA between the SAF and Los Angeles.


A major earthquake on the San Andreas Fault could disrupt all imported water supplies to Southern California, resulting in serious social and economic impacts to the area’s 19 million residents and businesses. Through the Seismic Resilience Water Supply Task Force, the DWR, MWD, and LADWP are working together to minimize these impacts. This unprecedented collaboration will benefit the region in many ways, such as the following:

  • comprehensive assessments, evaluations, and studies for the region (for individual aqueducts and how they collectively operate as a regional system);
  • clearer understanding of interdependencies both within each system and between the water supply systems and other lifeline systems such as electrical power, transportation, and communication;
  • identification of goals and specific recommendations for addressing vulnerabilities; and
  • coordination of critical areas for the region’s ability to continue operating and recover after an event, including emergency storage needs, resource sharing, and overall emergency response activities.

The result will be a more resilient and sustainable Southern California that will be better protected from major seismic events.


Clark DK, LeCocq P, Davis C, Rennie D, Grahek M, Rojas M. 2017. Seismic Resilience Water Supply Task Force March 30, 2016, Aqueduct Workshop and Five-Year Action Plan. Report No. 1536. Los Angeles: Metropolitan Water District of Southern California.

Clark D, Davis C, Rennie D. 2018. Multi-Agency Seismic Resilience Water Supply Task Force. 11th National Conference on Earthquake Engineering, Earthquake Engineering Research Institute, Jun 25–29, Los Angeles.

Davis CA. 2009. Scenario response and restoration of Los Angeles water system to a magnitude 7.8 San Andreas Fault earthquake. Proceedings, 7th US Conference on Lifeline Earthquake Engineering, ASCE, Jun 28–Jul 1, Oakland.

Davis CA. 2010. Los Angeles water supply impacts from a M7.8 San Andreas Fault earthquake scenario. Journal of Water Supply: Research and Technology 59:408–417.

Davis CA, O’Rourke TD. 2011. ShakeOut Scenario: Water system impacts from a M7.8 San Andreas earthquake. Earthquake Spectra 27(2):459–476.

Davis CA, Clark D, Rennie D. 2017. Multi-agency collaboration takes aim at water delivery. Source 31(4):21.

Graves R, Aagaard BT, Hudnut KW, Star LM, Stewart JP, -Jordan TH. 2008. Broadband simulations for Mw 7.8 southern San Andreas earthquakes: Ground motion sensitivity to rupture speed. Geophysical Research Letters 35(22).

Hinds J. 1938. Major problems of aqueduct location. Engineering News Record 121:646–652.

Hudnut KW, Aagaard B, Graves R, Jordan TH. 2008. Constructing the scenario event: The earthquake source. In: The ShakeOut Scenario, eds Jones LM, Bernknopf R, Cox D, Goltz J, Hudnut K, Mileti D, Perry S, Ponti D, Porter K, Reichle M, and 4 others. US Geological Survey Open File Report 2008-1150 and California Geological Survey Preliminary Report 25. Reston VA: USGS.

Jones LM, Aho M. 2019. Resilience by Design. The Bridge 49(2):52–59.

Jones LM, Bernknopf R, Cox D, Goltz J, Hudnut K, Mileti D, Perry S, Ponti D, Porter K, Reichle M, and 4 others. 2008. The ShakeOut Scenario. USGS Open File Report 2008-1150 and California Geological Survey Preliminary Report 25. Reston VA: USGS.

LADWP [Los Angeles Department of Water and Power]. 2014. Water System Seismic Resilience and Sustainability Program: Summary Report.

Lindvall S, Kerwin S, Heron C, Davis C, Tyson J, Chestnut J, Mass K. 2018. Characterizing the Los Angeles Aqueduct crossing of the San Andreas Fault for improved earthquake resilience. 11th National Conference on Earthquake Engineering, Earthquake Engineering Research Institute, Jun 25–29, Los Angeles.

Lund LV, Davis C. 2005. Multihazard mitigation Los Angeles water system: A historical perspective. In: Infrastructure Risk Management Processes: Natural, Accidental, and Deliberate Hazards, eds Taylor C, VanMarcke E. ASCE Council on Disaster Risk Management Monograph No. 1. Reston VA: ASCE.

Mayoral Seismic Safety Task Force. 2014. Resilience by Design. Los Angeles: Office of the Mayor.

MWD [Metropolitan Water District of Southern California]. 1991. October 1991 Final Environmental Impact Report for the Eastside Reservoir. Los Angeles.

MWD. 2016. 2015 Urban Water Management Plan. Los Angeles.

Philibosian B, Fumal T, Weldon R. 2011. San Andreas Fault earthquake chronology and Lake Cahuilla history at Coachella, California. Bulletin of the Seismological -Society of America 101(1):13–38.

Scharer KM, Biasi GP, Weldon RJ, Fumal TE. 2010. Quasi-periodic recurrence of large earthquakes on the southern San Andreas fault. Geology 38(6):555–558.

Scharer K, Weldon R II, Biasi G, Streig A, Fumal T. 2017. Ground-rupturing earthquakes on the northern Big Bend of the San Andreas Fault, California, 800 AD to present. Journal of Geophysical Research: Solid Earth 122:2193–2218.

Weldon R, Streig A, Hammond W, Yule D. 2014. Colorado River Aqueduct San Gorgonio Pass seismic event vulnerability study, San Bernardino County, California, San Andreas Fault-Southern Branch Surface Deformation Modeling. MWD Report No. 1484. Los Angeles: Metropolitan Water District of Southern California.

Weldon R, de Lamare G, Yule D, Hammond WC, Streig A, Sarmiento A, Freeman ST, Shamma J, Beikae M, Rodriguez A. 2016. San Andreas Fault–South Branch surface deformation modeling and risk to the Colorado River Aqueduct. In: Applied Geology in California, eds Anderson R, Ferriz H. Association of Environmental and Engineering -Geologists (AEG) Special Publication 26. Lexington KY: AEG.

[1]  The fault is about 1,200 km (750 miles) long and runs from just west of Eureka in the north to the Salton Sea (about 150 km, or 90 miles, east of San Diego) in the south, marking the tectonic boundary between the American and Pacific plates.

[2]  Other seismic threats will be addressed as warranted.

About the Author:Craig Davis is recently retired manager of the Water System Resilience Program and seismic manager for the Los Angeles Department of Water and Power. John Shamma is unit manager of engineering’s facility planning with the Metropolitan Water District of Southern California.