Background
San Diego, CA, is a warm, sunny, idyllic region covering 1,440 square miles along the Pacific Coast. However, its climatic conditions present challenges for the water resource planning agencies. San Diego’s Mediterranean climatic regime has significant variability, both from year to year and from season to season.
Up to the 1940s, the region relied heavily on reservoirs, to capture rainfall from the occasional wet periods to meet demand during the longer dry periods. In the 1940s, growth exceeded the locally sustainable 200,000 people, and the war effort required greater water reliability. As a solution, the Navy decided to go outside the region for water.
Conveyance capacity for imported water became the resource-planning element of choice until greater demands throughout the region caused imported water supplies to begin showing vulnerability in the 1990s. The Water Authority, which inherited management of the Navy’s aqueduct system in 1944, recognized that the region must begin tapping local water sources and decided to integrate facility planning with comprehensive supply planning. This resulted in the 2002 Regional Water Facilities Master Plan. The Master Plan took into account the distribution of possible hydrological conditions, the variability of demand based on weather, and the uncertainty associated with project implementation. The process relied on expert knowledge and complex computer modeling, which provided the Water Authority a better understanding of the interdependencies between water supply, demand, and regional hydrologies, including extreme hydrologic events.
The proposed dam
Supply sources meeting San Diego’s demand require storage to smooth out the long-term variability of runoff in the southwestern US. Most peaks in demand are managed with existing storage reservoirs. Carryover storage allows the San Diego region to weather a low-occurrence, high-volume supply shortage event of about a 30% imported water shortfall. The Water Authority’s master planning process identified a mix of conveyance and storage, a shift from previous water resource planning and management, and reliance on imported water. The importance of the carryover storage project grew, as awareness of other water resources issues–such as climate change and pumping restriction in the bay delta–increased.
The Master Planning Process and Modeling
The Water Authority embarked on the 2002 Regional Water Facilities Master Plan to develop potential solutions to the water infrastructure needs of San Diego County through 2030. The challenges included variation in demand projections, project implementation uncertainty, and supply uncertainty in three source water basins: the Colorado River, the State Water Project, and regional watersheds. The three basic infrastructure concepts tested included a completely new large pipeline conveyance from the east; a new large pipeline conveyance from the north, using MWD’s existing conveyance; and a desalination option. These concepts all shared many components, but were tested with and without some components, such as carryover storage. The Water Authority used uncertainty testing to reveal the degree of confidence in the groups of projects to accomplish the goal of water supply reliability.
Uncertainty testing for the Water Authority included Monte Carlo simulations that included the following parameters:
- Forecast demand varying based on the influence of weather
- Imported supply conditions
- Local hydrological scenarios based on historical patterns
- Local groundwater and recycled water implementation
- Dynamic scenarios of regional conveyance and storage projects
The Water Authority endeavored to innovate its planning process in the region by integrating all these elements into a comprehensive model to analyze historical trends, as well as daily peaking and its effect on the distribution system.
The process started with the fundamental elements of data collection and data management. First, developing a robust demand forecast meant working with the regional planning agency San Diego Association of Governments (SANDAG) for the spatial and temporal distribution of demand growth. The SANDAG Forecast projected a population growth of 1.0 million people to 3.7 million in 2030. The Water Authority found that growth is not equally distributed throughout its service area. Therefore, SANDAG’s insight into how it is distributed was a critical element. Using the foundation of a vivid population forecast, the development of a demand forecast could begin. The Water Authority contracted with Planning and Management Consultants Ltd. (PMCL) to develop a point and probabilistic water demand forecast. Base weather and conservation were addressed in the demand projections for each member agency.
Water supplies are comprised of two components: imported and local supplies. Imported supplies include those directly from Metropolitan and the Imperial Irrigation District (IID) transfer. Metropolitan supplies were given a comprehensive characterization of availability. IID deliveries were assumed to follow the current ramp-up plan, which starts in 2003 at 10,000 acre-feet, increasing to 200,000 acre-feet by 2021. As far as imported water supplies, the Water Authority completed an extensive analysis on the data contained in Metropolitan’s Integrated Resource Plan.
Local supplies include a history of highly variable runoff. The model took into account reservoir operating parameters such as: monthly historical evaporation rates, 100 years of historical precipitation, historical stream flow correlations to precipitation patterns, and member agency reservoir operating plans. Obtaining operating plans meant workshops with member agencies to put institutional knowledge into a narrative provided to model developers.
Other local supplies include recycled water and groundwater (including brackish water desalination). Recycled water and groundwater availability reduce demand for each member agency. Both recycled and groundwater supply represented by probabilistic curves that develop a shape to the likelihood that projects would be implemented. The “planned” level of production from each source was obtained by surveying the member agencies. The planned level represents the amount of recycled or groundwater production that has the highest probability of occurring.
The weather variable influences both supply and demand. To support the modeling, we gathered the daily maximum and minimum temperature and rainfall data from the National Weather Service. We then merged the data from various storage types and used statistical methods to filter errors from the very large database. Statistical methods were also used to fill in data gaps. Within the region, weather can vary; requiring the modelers to develop spatial independence through estimated coefficients for the different weather sub-zones of coastal region, and inland north and south.
To size regional infrastructure, it is necessary to shape daily demand peaking. A & N Technical Services worked with the data to develop the correlation of weather on historical demand trends and the daily and weekly demand shaping during the peak season, while recognizing that growth follows a continuous path, both as it grows and wanes. The modeling software used to integrate all this data was Confluence, developed by Gary Fiske & Associates. The representative model simulated the regional system of aqueducts, treatment facilities, demand, local projects, and surface storage reservoirs with enough resolution to compare reliability relationship between the scenarios evaluated by the mastering plan team. The scenarios were developed through a long process of interviewing experts in the regional system and evaluating the numerous studies developed for the Water Authority. While the list of options was not exhaustive, it held all possible projects that could actually be completed within all known technical and environment constraints.
Modeling completed for the master plan evaluated the various parameters noted above to determine delivery shortfall associated with facility implementation. The model evaluated the dynamic scenarios, including projected demand growth through 2030, 100 years of historical hydrology, available water conveyance projected through the planning horizon, storage facilities, and facilities to treat stored water projected through the planning horizon. The desire was to determine the frequency and duration of such delivery shortfalls to establish the best mix of facilities to maximize regional water delivery reliability.
The worst single year event evaluated was the coincidental occurrence of low hydrology on the Colorado River, the State Water Project (SWP), and San Diego Region such as weather year 1977. The storage use was approximately 47,000 acre-feet above normal reservoir withdrawals. This volume of water was utilized quickly during peak season of June through October. This was a low occurrence, but high volume event, representing 30% imported water shortfall. A new large conveyance facility from Diamond Valley reservoir could be an alternative to expanded reservoir storage, but would be a large facility not used except in a low hydrology event. The use of ESP facilities was determined to be the most cost-effective alternative.
The model evaluated the storage need for a three-year dry cycle, because a goal of the Urban Water Management Plan was to determine shortage in a three-year event. The worst three-year dry cycle evaluated averaged over 33,000 acre-feet per year above normal reservoir withdrawals or for facility planning 99,000 acre-feet of storage. This storage requirement based on modeling indicated that during a three-year dry cycle there is no means to refill storage since supplies source are already stressed. The water usage pattern matched the worst single-year event
The San Diego region has a relatively high occurrence of dry cycles beyond three years. Model duration curves showed possible delivery shortfalls due to dry conditions beyond three years, but required further assumptions on supply conditions beyond the master planning process. The master plan model, without the above mentioned supply assumptions, showed a drop off in unmet demands beyond the three- to four-year period. Further evaluation outside the master planning process indicated a need for another 100,000 acre-feet, if preferential rights were invoked and some local project did not proceed.
Since the master planning process, reduced pumping scenarios from the SWP related to delta fish protection have occurred. Pumping restrictions and dry-year scenarios increase variability to long-range supply forecasts creating the need for greater reliance on local storage to attenuate the effect of water supply variability. The concern over continued dry hydrology coupled with regulatory restrictions on SWP pumping and the need to plan for multiple dry-year periods have placed even greater value and importance on surface water storage. Climate change and the potential for reduced snow pack in the Sierra Nevada’s heightened our concern, but modeling for this scenario has not been done at this time.
The Road to Implementation
Once the need for 100,000 acre-feet of carryover storage was identified, project implementation did not come without challenges. There were three main hurdles to overcome: 1) selection of alternatives for 100,000 AF of carryover storage, 2) state and federal environmental law compliance and permitting, and 3) owner agency coordination.
Alternatives Analysis. The first hurdle was completion of a comprehensive alternatives analysis for siting 100,000 acre-feet of carryover storage. This process was critical for two reasons. First, it set up the alternatives to be analyzed in the joint state/federal environmental document (Environmental Impact Report/Environmental Impact Statement [EIR/EIS]). And secondly, the outcome of the process would help justify the “least environmentally practicable alternative” required for the stringent US Army Corps of Engineers Section 404 of the Clean Water Act permitting process. Under federal law, each project alternative must be analyzed at the same level of detail and compared, to determine the least environmentally damaging practicable alternative. This required the collection of equal amounts of baseline information for each project alternative.
Because of the level of detail required for analysis, determining which alternatives to carry forward in the EIR/EIS is critically important. The process used to identify, evaluate, and compare alternative project sites becomes the basis for presenting the results in the environmental document. The alternatives analysis was a collaborative effort between the Water Authority, City of San Diego, engineering consultant, and environmental consultant. Screening of alternatives for locating carryover storage included these steps:
- Form a technical panel to guide the screening process
- Identify a range of carryover storage options, the “building blocks” to formulate alternatives
- Compile a long list of possible carryover storage alternatives
- Perform coarse screening to identify a short list of alternatives to carry forward
- Perform fine screening of the short list of alternatives
- Identify alternatives for detailed analysis and comparison in the state/federal environmental document
A technical panel consisting of a core group of 20 individuals with diverse backgrounds and experience performed the screening of alternatives. The panel conducted several-day-long workshops to carry out the alternatives screening process. A detailed description of the overall process is covered in a report entitled: Carryover Storage and San Vicente Dam Raise Project–Screening of Alternatives (GEI Consultants Inc., 2007), summarized below.
Alternatives must be limited to ones that meet the project objectives, are ostensibly feasible, and would avoid or substantially lessen at least one of the significant environmental effects of the project. The process resulted in the identification of 28 potential alternatives. Through the process of coarse screening, this list was reduced to 11 for the fine screening phase that included the development of a quantitative decision model to objectively rank alternatives, involving the following steps:
- Step 1–Form a subgroup of the Technical Panel, referred to as the screening panel, to guide the decision model development
- Step 2–Develop a hierarchy of planning values, and reach panel consensus on the hierarchy
- Step 3–Develop appropriate measurement systems for evaluation criteria
- Step 4–Develop “preference relationships” for each criterion, and reach panel consensus on the criteria measures and preference relationships to use in the model
- Step 5–Reach panel consensus on weighting factors for use in the model by a participatory polling process
Fine screening reduced the number of alternatives from 11 to a reasonable number of alternatives for further investigation in the EIR/EIS. Based on the results of fine screening, the Water Authority selected the following three alternatives to be included in the EIR/EIS: 1) Alternative 1–Additional 100,000 acre-feet expansion of San Vicente Dam and Reservoir (the Proposed Action), 2) Alternative 2–New dam and 100,000 acre-feet reservoir in Moose Canyon, and 3) Alternative 9–Additional 50,000 acre-feet of expansion of San Vicente Dam and Reservoir + a new dam and 50,000 acre-feet reservoir in Moosa Canyon.
Environmental Compliance and Permitting. Having identified the Proposed Action and the alternatives, the next hurdle was to navigate through both state and federal environmental reporting and permitting laws within a very tight timeframe. The California Environmental Quality Act (CEQA) governs environmental compliance in California. The federal National Environmental Policy Act (NEPA) is required when a project must undergo a federal action, such as obtaining a federal permit.
The two laws have similar reporting requirements, allowing the Water Authority to prepare a combined document with the US Army Corps of Engineers for both CEQA and NEPA. This not only streamlined the process for reporting by cutting down on the number and volume of documents, but also was also beneficial from a scheduling standpoint. CEQA and NEPA contain prescribed timelines for legal noticing, public and agency review times, and closeout of the process. By conducting the CEQA and NEPA processes together, the Water Authority and the Army Corps completed the process in 17-months time. By comparison, completing CEQA and NEPA in sequence can take up to three years.
After completion of CEQA/NEPA, permitting for the project could begin. The approved project was Alternative 1–Additional 100,000 acre-feet expansion of San Vicente Dam and Reservoir. The Carryover Storage and San Vicente Dam Raise project required permits from several state and federal agencies, including the Army Corps, United States Fish and Wildlife Service, California Department of Fish and Game, California Regional Water Quality Control Board, and the California State Historic Preservation Office.
The Water Authority sought the most time-effective way to manage the various agencies’ permitting processes, as project scheduling allotted a very tight timeframe. It proactively met with the permitting agencies early on to explain the project, the potential environmental impacts that would require permits, and mitigate approach for the impacts. Several coordination meetings were held throughout the CEQA/NEPA process to keep all parties informed of any project changes.
A final solution to the time management problem was the development of agreements with the various permitting agencies to expedite the permit review process time. By gaining the assignment and commitment of staff resources from the agencies up front, both parties understood the critical nature of the shortened time to review permit applications and issue permits. In the end, all project permits were issued within eight months of EIR/EIS certification. Without this solution, the permitting process can take up to one year to 18 months from certification of the EIR.
Owner Agency Coordination. The City of San Diego owns and operates San Vicente Dam. The last hurdle that the Water Authority encountered was the added coordination of working on an existing facility owned and operated by another agency that must remain operational during construction. To meet this challenge and foster a collaborative working relationship, the city and Water Authority memorialized each party’s obligations in an agreement and clarifying document of principles of understanding (POU). This established protocols for cooperation between the City and Water Authority during implementation of the dam raise project, such as development of reservoir operating plans during the active and post construction periods. They also created a water quality model and response plan for various water quality scenarios and supply issues that the Water Authority may face during the construction drawdown period. Finally, the POU addressed preventative measures to be included in the project design to control quagga mussels infestation.
Conclusion
The Water Authority surmounted various hurdles to increase water reliability for three million residents. This included creating a master plan to prioritize projects, gaining environmental approval and permits to implement the projects, and executing agreements with Water Authority partners to clarify expectations during implementation.
Water supply reliability in the San Diego region and the supply sources to meet demand rely on storage to smooth out the long-term variability of runoff in the southwestern US. Most peaks in demand are managed with existing storage reservoirs. Through modeling, a better understanding of the interdependencies between water supply, demand, and regional hydrologies, including extreme hydrologic events, was obtained. The Water Authority’s master planning process identified a need for approximately 100,000 acre-feet of carryover storage to maintain a secure and reliable supply through varied conditions and the planning horizon of 2030. The carryover storage facility ultimately approved was the addition of 100,000 acre-feet to San Vicente Reservoir (above the reservoir expansion for the Emergency Storage Project). By identifying viable alternatives to analyze in the EIR/EIS, and by finding ways to expedite the environmental and permitting processes, project implementation was able to stay on schedule. The next major step will commence Spring 2009, the start of the four-year construction period to raise the dam. Reaching this point makes the decision “to store,” rather than not “to store,” a more proximate, tangible goal.
