The water-energy nexus is not what you expect
Earlier this summer, researchers at UC Davis confirmed what a lot of us already know — that saving water saves energy. The analysis from the UC Davis Center for Water-Energy Efficiency found that California’s mandatory 25 percent reduction in urban water use, adopted in May 2015 due to the ongoing severe drought, resulted in significant energy and greenhouse gas savings.
From June 2015 to February, the electricity saved by reducing urban water use is estimated to have been nearly 922 gigawatt-hours. Because electricity production oftentimes relies on fossil fuels such as coal and natural gas, this energy savings also significantly reduced greenhouse gas emissions — similar in scale to taking almost 50,000 cars off the road.
Saving water saves energy because of the large amount of energy needed to extract, transport, treat and distribute water to our homes and businesses. Still more energy is needed to collect and treat the wastewater that then comes from our sinks, showers, toilets, clothes washers and other sources. This energy use is referred to as embedded energy.
In California, the embedded energy in water can be quite large, especially for regions such as Southern California, which rely heavily on imported water supplies from the Sacramento-San Joaquin Bay Delta and the Colorado River. To divert water from one river basin to another, major facilities pump water over long distances and steep terrain.
In fact, roughly half of the energy embedded in water in California comes from the state's major long-distance conveyance systems — the State Water Project (SWP) and the Colorado River Aqueduct (CRA). The former conveys water that falls as rain or snow in the northern part of the state, and the latter conveys water from the Colorado River.
As my colleague Ed Osann explained last fall, NRDC has been involved for several years in a proceeding at the California Public Utilities Commission (PDF) that focused on estimating how much embedded energy can be reasonably saved by investing in end-use water-savings measures.
While energy and water utilities are moving forward with use of the calculator tools approved by the commission, there is a distinct possibility that the embedded energy savings attributed to end-use water efficiency measures significantly could be exaggerated unless the limitations of the calculator tools are more clearly recognized.
For example, while the calculator accurately assesses the benefits of water efficiency in reducing the energy used to treat and distribute water to the customer, it likely overestimates the benefits of water efficiency in reducing the energy used to move water long distances.
The energy used by most local utilities that treat and distribute water to retail customers (energy used by your local drinking water treatment plant) certainly can be reduced through water conservation and efficiency. But while the movement of treated drinking water is generally in sync with the demands of water users, the conveyance (pumping over long distances) of untreated water from its natural source to a carry-over storage facility is primarily driven by source water availability and may not be influenced by reductions in end-use water demand for many years or even decades.
The SWP and CRA systems purposefully store untreated water to buffer against large fluctuations in water supply due to year-to-year changes in the amount of water available from rain or snowpack. Because of the significant embedded energy related to conveying water over long distances to these carry-over storage facilities, there is the potential for the calculator tools to overstate the energy savings attributable to end-use water conservation measures.
This is particularly true if other state agencies (Air Resources Board, State Water Resources Control Board, Department of Water Resources) use the calculator tools without imposing the same restrictions that the commission has put in place, such as only allowing embedded energy savings from investor-owned utilities (IOUs) to be claimed.
The energy used by SWP and CRA operations primarily comes from public power agencies so by only considering IOU energy, the embedded energy due to water conveyance is rightfully excluded as a potential source of energy savings for end-use water conservation efforts. Notably, UC Davis researchers follow the commission’s approach by only considering embedded energy from IOUs in their analysis.
To more fully illustrate the embedded energy and long-distance water conveyance conundrum, we analyzed the energy consumed by the SWP and the CRA since the beginning of the drought in 2010. It’s important to note that drought conditions in California have affected the water available to the SWP whereas water for the CRA comes from the Colorado River Basin, from which California has continued to withdraw its full allotment under the applicable interstate compacts.
Due to the severity of California’s drought, the California State Water Resources Control Board (SWRCB) imposed mandatory water conservation targets on urban water suppliers in May 2015. If end-use water conservation had an immediate or relatively near-term impact on long-distance water conveyance, we would expect to see a fairly distinct inverse relationship between water conservation rates and SWP and CRA energy consumption.
In other words, we would expect that as more water is conserved, SWP and CRA energy consumption would also decrease because less water is conveyed.
However, statewide water conservation rates and SWP energy consumption appear to both increase over the period where the datasets overlap. As water use trended down, power consumption continued to trend up. Thus, retail water conservation appears to have had no real effect on long-distance water conveyance during this 23-month period. This is further supported by the fact that monthly CRA energy consumption stays relatively constant.
On the other hand, there appears to be a very clear relationship between SWP energy consumption and April 1 snowpack in the Sierra Nevada. Snowpack is an important source of water for California. In a typical year, snowpack stores about 15 million acre-feet of water, or about one-third of the water used by California’s cities and farms each year.
The trends in April 1 snowpack and annual SWP energy consumption have been in sync since the beginning of the current drought. As snowpack has dwindled, SWP energy consumption also has decreased due to the increasingly scarce water supplies available for conveyance to wholesale customers, such as the Metropolitan Water District with major carryover storage facilities.
Once again, CRA energy consumption remains relatively constant and is not affected by Sierra Nevada snowpack levels (or the California drought) because water withdrawals from the Lower Colorado River have not yet been curtailed due to drought.
Our analysis of SWP and CRA energy use during the drought clearly indicates that the state’s major long-distance conveyance systems more closely follow changes in source water availability than changes in end-use water demand on a year-to-year basis. Declining source water availability had the effect of both reducing power consumption by the SWP and inducing the governor to call for first voluntary and then mandatory water conservation measures. Water consumption by end users then declined sharply after power consumption for conveyance already had dropped due to limited source water supplies.
This evidence supports our long-standing position that the commission must resolve the potential for the calculator tools to overstate the embedded energy saved by end-use water efficiency measures in future updates to the tools. Until those changes are made, state agencies must exercise caution when using the tools approved by the commission to determine which energy-savings projects receive public funding.
Without acknowledging and resolving this fundamental flaw, energy efficiency funds (cap-and-trade auction proceeds, utility customer programs) could be diverted to end-use water efficiency measures when other energy efficiency measures lead to greater and more cost-effective energy savings and greenhouse gas reductions.
While end-use water conservation may not have a near-term impact on long-distance water conveyance operations, some might speculate that sustained reductions in end-use water demand eventually might result in declines in the volume of water conveyed (and therefore, the energy consumed) by the SWP and CRA systems in the long run, even after accounting for the replenishment of carryover storage.
However, such a scenario would require that SWP contractors forego water allocations that they are being charged for regardless of delivery volume, and — just as unlikely — that the state of California decline to make full use of its interstate compact entitlement to Colorado River water even in years when sufficient supplies are available. Unfortunately, existing legal and contractual frameworks are unlikely to facilitate such an outcome without laborious overhaul involving issues and stakeholders far afield from energy savings.
More likely, as source water availability becomes further impacted by climate change, reduced source water withdrawals may become unavoidable — and end-use water efficiency will emerge as an essential coping strategy, rather than a driver for reducing conveyance volumes.
As the UC Davis analysis shows, there’s no doubt that saving water saves energy, but we must be careful not to overstate the energy savings. There are more than enough reasons to save water in California, now and in the future, without claiming more energy and greenhouse gas reductions than such water savings are likely to achieve.