|
|
| Fig. 1: The Central Valley Project. [4] (Source: Wikimedia Commons) |
The Central Valley is the agricultural region of the state of California which is bordered by the Klamath and Sierra Nevada mountains in the north and east, respectively, and the Pacific Coast Ranges and San Francisco Bay to the west (Fig. 1). Despite its dry and sterile appearance, when water is added, the land becomes extremely suited for agriculture. Because of its agricultural production, water is an essential resource for the Central Valley. However, along with the surrounding areas, the Central Valley experiences a dry season starting in May and a wet season typically starting in November. In addition, California has been plagued by severe drought periods going as far back as 1925 and as recently as 2016. Coupled with the rise of a major urban center in dry Los Angeles, there is a unique and longstanding water problem for the state of California that is challenging to solve to this day.
There are three major water conveyance systems which work to mitigate water demand in dry areas of the state: the Central Valley Project (CVP), the State Water Project (SWP), and the Colorado River Aqueduct (CRA). The CVP was a New Deal project started with the purpose of delivering water from Lake Shasta and Trinity Lake in northern California to the Central Valley for irrigation, controlling flooding, and generating hydroelectric power with pumping-generating plants. It is thus a federal project overseen by the U.S. Bureau of Reclamation (Fig. 1). The SWP is a state project that began in 1960 and is overseen by the California Department of Water Resources. Its function and purpose is basically the same as the CVP, but it has the additional focus of providing water to urban areas like San Francisco and Los Angeles. The CVP and SWP use a large series of aqueducts, canals, pumping plants, dams, and reservoirs to control and transport water.
 |
| Fig. 2: The State Water Project. Note that San Luis Canal becomes the California aqueduct. [4] (Source: Wikimedia Commons) |
The San Luis unit is a joint effort of the CVP and SWP, involving both the state and federal governments. Completed in 1967, it involves the San Luis Reservoir, B.F. Sisk (formerly San Luis) Dam, O'Neill forebay, the San Luis Canal, and two pumping-generating plants (PGPs): O'Neill PGP and Gianelli (formerly San Luis) PGP. [1] The San Luis unit is a major junction of water conveyance in California, and quantifying the energy gain and usage from the PGPs is a source of great confusion in understanding CVP/SWP energy because of its shared nature. The purpose of this report is to estimate how much energy the San Luis unit PGPs are consuming each year.
A study by the California Measurement Advisory Council (CALMAC) in 2007, commissioned by the California Energy Commission (CEC), is the largest effort to assemble and quantify water flow and energy data in the CVP/SWP over a decade (from 1995-2005). [2] The data from this study was primarily used for the analysis in this report. The report notes that moving water from northern California to the rest of the state consumes a huge amount of energy because of elevation rises and the hundreds of miles of transportation required. However, despite the amount of energy used to convey surface water, the amount of electricity used to pump groundwater by farmers in the Central Valley to supplement surface water supplies still exceeds the amount used by the three largest water conveyance systems (CVP, SWP, CRA) combined. [2] This underscores the main takeaway of this report: water, and specifically water in the right locations at the right times, is far more important in California than energy.
Given this context, how exactly do the PGPs in the CVP and SWP contribute to generating hydroelectricity for California?. There are many resources for a more detailed understanding of how pumped storage hydroelectricity works, as well as a report for this class from 2010. [3] The idea is that a plant lies between two water sources at different elevations, and allowing the water to flow from the higher source to the lower source through the plant generates electricity when demand is high during the day. When demand is lower at night, the plant moves the water back up. This is the mechanism through which O'Neill and Gianelli PGPs are implied to generate electricity for California. However, they do not function this way in reality.
 |
| Fig. 3: Net yearly energy use for O'Neill and Gianelli PGPs. [4] (Image Source: J. Pan) |
A closer look at the San Luis inset in Fig. 1 allows us to see how each plant is supposed to function. San Luis reservoir is an off-stream reservoir that stores water during the wet season so it can be released for use in the dry season. Gianelli PGP works between San Luis reservoir and O'Neill forebay. In the wet season, Gianelli pumps water up to San Luis reservoir for storage. When water needs to be released, it flows through Gianelli, generating energy, into O'Neill forebay. Gianelli's main purpose is managing this storage and release mechanism for San Luis reservoir. O'Neill PGP works between O'Neill forebay and the (CVP-owned) Delta-Mendota canal. When water flows from O'Neill forebay into the Delta-Mendota canal to the rest of the CVP system, O'Neill PGP generates electricity. To store water in San Luis reservoir, the plant has to pump water into O'Neill forebay to then be pumped by Gianelli.
O'Neill forebay connects to the San Luis canal, a joint state-federal entity, which is part of the SWP-owned California aqueduct (Fig. 2). Confusion between CVP water and SWP water is a major contributor to CVP's ability to make claims that it is a net generator of power. The main energy usage of the Southern part of the water conveyance system comes from elevation rises in the California aqueduct, which SWP takes responsibility for. In addition, Gianelli PGP cannot and does not operate as a pumped-storage hydropower plant. In order to generate electricity, water must flow from the San Luis reservoir to O'Neill forebay. However, if water were released into O'Neill forebay each day for hydroelectricity generation, it would have to be conveyed down the San Luis canal into the energy-expensive California aqueduct. [4] This opposes the goal of the SWP to minimize conveyance during the day, when electricity is expensive, and maximize it at night. In the data, this can be seen by the fact that high and low flows for pumping/generating directly coincide with the dry and wet seasons, rather than a daily cycle of high/low electricity demand. Because of these constraints, O'Neill and Gianelli do not function as regular pumped storage plants.
 |
| Fig. 4: Extreme minimum net yearly energy use for O'Neill and Gianelli PGPs. [4] (Image source: J. Pan) |
In Appendix C of the CALMAC report, plots are provided showing water flows and energy use/production for pumping/generating in both Gianelli (pp. C-30-C-36) and O'Neill (pp. C-127-C-132) for each month in the years 1995-2005. [4] The data is given in acre-feet (AF) and kiloWatt-hours (kWh), but it has been converted to cubic meters and Joules in this report. Because of difficulty getting the precise data points from the provided plots, I have analyzed the maximum and minimum values per year for pumping/generating at each plant, as well as the average yearly value of energy intensity, which is the average amount of energy used/generated per unit volume of water.
Firstly, the maximum and minimum yearly values of energy use (while pumping water up) and production (while water flows through the generator) can be used to roughly calculate a maximum and minimum net yearly energy consumption for each plant. The net energy use is calculated
One could infer that the minimum energy used and the minimum energy produced would likely coincide with the same year (for example, in a very dry year where there is not much water to perform either task). The same could apply for the maximums. Then, we could estimate bounds on the net energy as (Fig. 3)
Using this measure, as shown in Fig. 3, O'Neill's net energy usage per year is between 7.53 × 1014 to 1.28 × 1015 J. Gianelli's net energy usage per year is between 4.25 × 1014 to 6.69 × 1014 J.
 |
| Fig. 5: Average energy intensities for O'Neill and Gianelli PGPs. [4] (Image source: J. Pan.) |
Another calculation could be of the extreme minimum net energy usage per year, which would give us a better chance of getting a net energy generation rather than a net energy use for each plant. This can be calculated by subtracting the maximum generating year from the minimum pumping year:
Using this measure, as shown in Fig. 4, O'Neill's minimum net energy usage per year is 7.23 × 1014 J. Gianelli's minimum net energy usage per year is 2.74 × 1013 J.
Secondly, I plotted the average energy intensities (energy per volume of water) for pumping, generating, and net (pumping-generating) for each plant (Fig. 5). This tells us that, per cubic meter of water flow in each plant, there is a net energy usage of 7.97 × 104 J/m3 for O'Neill and 3.53 × 105 J/m3 for Gianelli. Note that neither plant has a net generation of energy, which is not surprising, but they consume an enormous amount of energy each year. This supports the idea that their primary task is water conveyance, not electricity generation.
Water is a far more important resource than energy in California because of agriculture in the Central Valley and the rising needs of urban areas for large amounts of consistent, running water. This fact is reflected in the function of the pumping-generating plants in the San Luis unit. While they have a secondary benefit of generating hydroelectric power when water moves through them in the right direction, the primary purpose of O'Neill and Gianelli PGPs is water conveyance. Gianelli is physically and financially constrained from operating as a pumped storage facility because it becomes energetically expensive to move water down the California aqueduct once it has flowed through Gianelli for electricity generation. In addition, O'Neill can only generate power by releasing water into the Delta-Mendota canal, which is only CVP water and does not go to the California aqueduct where it is most needed. Energy data relating to water in California has been historically underestimated and is increasingly difficult to track because of the arbitrary separation of CVP/SWP water despite their joint nature, which is highlighted by this study of the joint-owned San Luis unit. If the data is to be trusted, calculations show that O'Neill and Gianelli PGPs had a net energy usage on at least the order of 1014 and 1013 J per year in the years 1995-2005, respectively. Regardless, California is facing a serious energy and water problem that will only grow with continued drought and climate change. The water conveyance systems are not immune to drought, and getting water to the areas that desperately need it will continue to be a challenge for years to come. It is still unclear how energy will ultimately factor into this growing issue.
© Jessica Pan. The author warrants that the work is the author's own and that Stanford University provided no input other than typesetting and referencing guidelines. The author grants permission to copy, distribute and display this work in unaltered form, with attribution to the author, for noncommercial purposes only. All other rights, including commercial rights, are reserved to the author.
[1] "Central Valley Project: San Luis Unit, West San Joaquin Division", U.S. Bureau of Reclamation, 1984.
[2] "Embedded Energy in Water Studies. Study 2: Water Agency and Function Component Studyand Embedded Energy-Water Load Profiles," California Public Utilities Commission, August 2010.
[3] A. Boysen, "Pumped Storage Hydroelectricity," Physics 240, Stanford University, Fall 2010.
[4] "Embedded Energy in Water Studies. Study 1: Statewide and Regional Water-Energy Relationship," California Public Utilities Commission, August 2010.
