NOTE: This post does not reflect my latest thinking on water storage. Please see the new summary.
We can build more than 2 terawatt-hours of pumped hydroelectric storage, significantly cheaper than lithium-ion batteries, using a novel combination of well-established technologies (no breakthroughs needed), with low environmental impact, on unoccupied land with no economic value, where there is little or no water or vegetation, reasonably close to each of the major cities of the American Southwest.
The hydroelectric machinery—pumps, turbines, motor/generators—is off-the-shelf commercial product. Only the water storage is different. Instead of two monolithic reservoirs, water is stored in thousands of large, collapsible purpose-built tanks made of fiber-reinforced polymer materials: in short, bags. These bags, each storing on the order of a million gallons of water, rest on individual flattened spots on the terrain. The piping system that connects them makes it possible for the bags to be sited at different elevations within either the upper or the lower storage area, so earthworks to create a large, level reservoir area are not needed (nor are dams). I call this approach Encapsulated Pumped Storage.
Because the water is fully enclosed, there are no evaporative losses, no seepage, and no interaction with aquatic ecosystems. The site is filled with water as it comes online and uses much the same water throughout its life, which should be many decades.
Many of the possible sites have an elevation difference on the order of 1,000 meters. This is two to three times higher than most existing pumped storage plants, which proportionately reduces the amount of water that must be stored per kWh of stored energy. Operation at 1 km head has been done commercially before, but suitable sites were rare. This new form of water storage is compatible with a much greater variety of topographies.
Because the water essentially sits on top of the terrain, not sunk into it in reservoirs, construction times will be reduced very significantly. Also, a plant can come online quickly, starting small and expanding incrementally.
If the real-world scenario I modeled is representative, decarbonizing the grid using only wind and solar power is feasible up to about 60% carbon reduction, but by approximately 80%, costs begin to rise exponentially. Encapsulated pumped storage can provide large-scale, long-duration storage sufficient to reach 100% grid carbon elimination throughout the Western states, at an acceptable cost.
Because encapsulated pumped storage consists of very familiar, well-understood technologies which are integrated in a new way, given a reasonable R & D investment, it should be possible to complete meaningful prototypes within a year, and start deploying it at scale in 3 to 5 years’ time.
I’ve now written 18 posts on this topic, and I’ve barely scratched the surface. But a proposal shouldn’t go on forever, and it seems like time for a shift to a more collaborative way of working, so I’m going to put writing on the back burner, and spend more time talking to people. Some of the things I hope to do in partnership with others are:
1. Finding out, from experts in various fields, what building EPS systems would really cost;
2. Accurate computer modeling of the hydrodynamic behavior of a whole system; and of course:
3. Building some prototypes.