The Economics of the Delta Tunnel
One of the most controversial water issues in California is the proposed Delta Conveyance. The 45-mile-long tunnel will have the capacity to move up to 4 million acre feet per year from the Sacramento River safely under the fragile delta ecosystem, delivering water to southbound aqueducts. That’s not bad. But the reservoir storage necessary to allow the tunnel to operate year-round at capacity, or even at half-capacity, is unlikely to ever be met. The state’s own estimate of the tunnel’s eventual throughput is only 500,000 acre feet per year.
A clarifying way to evaluate the cost effectiveness of water supply infrastructure investments is to examine a simple ratio: the total project construction cost divided by the expected average annual yield. Using that method, at a cost of $20 billion and operating year after year at its full capacity of 4 million acre feet of annual throughput, equates to a cost/yield ratio of $5,000 per acre foot. But this is a theoretical best case estimate. How would it compare to other cost-effective ways to increase California’s water supply?
For example, a water supply investment that could make good economic sense is to raise the Shasta Dam. An 18 foot raise is estimated to cost just under $2 billion to add 630,000 acre feet of storage capacity. But total storage capacity isn’t equivalent to annual yield, which is defined as how much water is stored and released from a reservoir in an average year. For raising the height of the Shasta Dam, a realistic worst case estimate (opponents produce even lower projections) of the increased yield is 100,000 acre feet per year. That equates to a cost/yield ratio of $20,000 per acre foot, less the revenue from additional hydroelectric power, which at 2 AF/megawatt-hour (MWh) and a wholesale price of $40/MWh gives back about $2M per year. That won’t put a dent in recovery of construction costs, but will help defray operating costs. It bears quantifying because it comes up a lot.
A realistic best case estimate (proponents produce higher projections) for Shasta’s increased yield if the dam is raised 18 feet is 200,000 acre feet per year. That assumes the dam fills to capacity about once every three years, something which doesn’t depend solely on rain and snow, but also on how the water is managed and released over the course of each winter. But using this estimate, the cost/yield for raising the Shasta Dam drops to $10,000 per acre foot, which is very good. We should note that Shasta has filled to capacity the last three years in a row.
The Sites Reservoir, to offer another comparison, is expected to yield 500,000 acre feet per year at a construction cost that might actually come in under $5 billion. That’s a cost/yield of $10,000 per acre foot, also very good. Then again, in order to keep the Sites project alive, half its yield is to be reserved for environmental flows.
Returning to the Delta Conveyance, what is a realistic worst case estimate of its cost/yield? Suppose we use the officially projected throughput of 500,000 acre feet per year, and suppose we double the cost, since there probably isn’t a single serious gambler in the world who would bet against that possibility. That would be $40 billion divided by 500,000, resulting in a cost/yield of $80,000 per acre foot.
Suppose the state manages to push 1 million acre feet through the tunnel in the average year, twice the current projection, and manages to keep total construction costs down to $30 billion. That’s a more realistic best case total cost than the official number and is probably still a stretch. Suppose as well that the existing Delta Pumps are not throttled down one bit by litigators and regulators who will argue the tunnel is not meant to supplement the pumps, but to replace them. At a cost of $30 billion and an average net throughput of 1 million acre feet per year, the cost/yield for the tunnel is $30,000 per acre foot. That is probably the absolute best we can hope for, and, just for perspective, is more than the cost/yield for desalination plants.
It’s not fair, however, to judge the tunnel purely on the basis of how much additional water it may move south. The compelling case for the tunnel is that it diversifies the ways in which water can be transported from California’s water rich north to the 5 million acres of irrigated cropland and 25+ million urban water consumers south of the delta. If a superstorm or massive earthquake causes several delta levees to fail, the tunnel will be the only way left to get a lot of water from north to south. California’s delta farmland and parkland used to be moisture-rich peat bogs before they became islands behind levees, and over the past century, the land in the delta dried out and subsided. They now lie at or below sea level, and if they fail, salt water from the San Pablo Bay will pour in. The existing delta pumps would be useless if that happened, since they’d be pumping salt water into the southbound aqueducts.
This is the case for the Delta Conveyance, and it’s hard to argue.
The biggest risk for farmers from Stockton to Bakersfield is that while debate over the tunnel demands legislative and executive attention, while it consumes agency staff time, and, eventually, as it consumes tens of billions of dollars, it crowds out everything else that has to be done. The urgency for everything else cannot be overstated.
The delta tunnel, if it is ever built, even with streamlined approvals, is unlikely to operate before 2040. Between now and then, delta levees need to be reinforced, delta channels need to be dredged, the delta pumps need to operate under more reasonable constraints, new ways to safely take the so-called “big gulp” out of the delta during major storms need to be explored and implemented, fast-charge aquifers need to be identified and developed to efficiently harvest storm runoff, California’s cities need to get the nutrients out of their wastewater discharges if not actually reuse the water, and, gasp, we need to expand our reservoir and desalination capacity at a scale that makes a difference.
Focusing attention on the Delta Conveyance is timely. But so is everything else.