Pumped Hydro Accelerating Into Grid Storage Future – CleanTechnica

Sign up for daily news updates from CleanTechnica on email. Or follow us on Google News!

---

There’s been a lot of great pumped hydro news from around the world in the past month, so a bit of a round up is in order. As a reminder, pumped hydro is the gravity storage solution that actually works, unlike concrete blocks, elevators and hillside rail systems.

We’ve been building and operating pumped hydro since 1907, most of the legacy facilities were built to give nuclear and coal plants something to do at night and to bridge their poor flexibility during the day, and it’s by far the biggest form of energy storage on the grid.

Three years ago I built a grid storage analysis based on abstracted business factors, assessing all the major forms of storage and projected the winners out through 2060. The winners were pumped hydro, redox flow batteries and cell-based batteries, with the also rans like hydrogen and compressed air storage making up about 100 GW of mixed nuts storage. Like all of my projections, it’s wrong, but I think it’s less wrong than most. There are big error bars on this, but it’s based on solid fundamentals.

The specific type of pumped hydro I’m bullish on is closed loop, off river pumped hydro storage with high head heights. Researcher Matt Stocks and team at the Australia National University ran a GIS assessment looking for locations off of protected lands, near transmission with room for twinned top and bottom reservoirs with at least 400 meters of vertical head height and within a few kilometers of one another.

Even with those restrictions and lack of sufficient data on large portions of the world like most of Russia, they found that there was resource for 616,000 sites with sufficient energy and power capacity to be 100 times the requirements for all grid storage for the world, and 200 times in North America, blessed as it is with lots of steep land on both coasts.

This model means that pumped hydro reuses water over and over again with little loss, doesn’t impact existing rivers and the head height means that the reservoirs are small. Per Stocks, a gigaliter of water with 500 meter head height would provide a gigawatt hour of energy storage with two small reservoirs only a kilometer on a side if square, more ponds than new lakes.

Time passed and my lithium ion perspective became a multi-chemistry cell-based battery perspective and I’m assessing whether it will be a bigger part of the solution than I’d projected. The pros of massive manufacturability and Wright’s Law argue for it, but the competition for electrification of all ground transportation, all inland shipping, two-thirds of short sea shipping and a remarkable amount of aviation counterbalance that.

Certainly the past couple of years have seen a massive surge of mostly lithium-ion battery grid storage going live globally, but little pumped hydro seeing commissioning. The 20 GWh Swiss facility went live in 2022 while Snowy 2.0 in Australia went from disaster to disaster, with the most recent news being that their tunnel boring machine only made it 150 meters before getting stuck for a year. For context, the tunnel boring machine is 143 meters long. Yeah, every time anyone mentions Snowy 2.0 I like to say it’s not an example of pumped hydro, it’s an example of a terribly planned and managed megaproject.

But just before the end of the year came the news that Norwegian energy giant Statkraft has acquired the 450 MW Loch Ness-sited Red John pumped hydro development from Scotland’s Intelligent Land Investments. I’ve been talking with ILI and its founder and CEO Mark Wilson for years, including dinner with him when I was in Glasgow last year debating maritime decarbonization for a European ship owner operator, so this was delightful news.

Red John is one of three pumped hydro facilities ILI has developed to being shovel-ready with all transmission connections, designs, engineering and approvals completed. The three all share a top reservoir on steep hills with lochs acting as the bottom reservoir, but otherwise follow the closed loop, off river model. Between them, they have 2.5 GW of power capacity and 60 GWh of energy capacity.

Just after that news crossed my screen, something I’d also talked with Wilson about years ago crossed my desk. The UK government finally adopted a cap and floor fiscal model for pumped hydro. This provides the fiscal certainty for development of the billion dollar infrastructure at least in that country, and there’s a lot of resource potential in Scotland to sop it up.

Finally, this crossed my screen this week. When I had originally done the grid storage assessment three years ago, China had 35 of 40 sites globally under construction representing 85% of the energy capacity at 51 GWh.

As of 2023, they had finished 19 GW of power capacity and have doubled down on the technology. They have a further 89 GW of power capacity under construction and another 276 GW planned.

The average pumped hydro facility is long duration storage, with 12 to 24 hours of storage. Hong Kong’s Guangdong facility, for example, has 2.4 GW of power capacity and 25 GWh of energy capacity. That ratio isn’t unusual, as the 2.5 GW / 60 GWh energy to power ratio, a full 24 hours of energy delivery, for the ILI facilities shows.

For cell-based battery grid storage, the power to energy ratio is usually closer to 1 to 2, so a 100 MW power capacity facility might have 200 MWh of energy, which it can deliver economically up to four hours later. With new chemistries and generally plummeting battery costs, there are a couple of cases where eight hour storage with batteries has become cost effective.

The value of cell-based battery storage, in other words, isn’t storing massive amounts of energy, but storing smaller amounts and delivering it very rapidly. The electrochemistry of cell-based batteries is instant on, unlike the pumping of fluids in redox flow solutions or the opening of sluice gates and spooling up turbines in pumped hydro facilities. Each technology has strengths and weaknesses which complement the other.

For context, while pumped hydro has long tailed risks, with tunneling being at exactly the half-way point on Professor Bent Flyvbjerg’s categorization of over 16,000 megaprojects ranked by running over budget, it’s still much lower risk than building nuclear plants.

As I recently pointed out, China only managed to connect 1.2 GW of nuclear generation capacity to the grid in all of 2023, as its nuclear program continues to face strong and apparently self-created headwinds. Its pumped hydro program has been much more successful in recent years, with 19 GW of power capacity and likely 190 to 240 GWh of energy capacity. All, of course, linked to demand centers with HVDC.

I’m waiting to update my 2021 projection of grid storage for a year or two more so that more data is in. One thing I didn’t factor into my projection at the time was that the developed world appears to have lost the ability to execute on infrastructure megaprojects. That will certainly be a factor that will be added. The rapid decrease in cell-based batteries and new chemistries like sodium which promise even lower costs and a removal of mineral constraints will also play a factor.

But it’s safe to say that global deployment of cell-based batteries exceeded my projection over the past three years. I’m still comfortable that redox flow batteries won’t take off in a major way until around 2030 as those technologies mature, and their power to energy ratio will be increasingly valuable with time.

All in all, however, it was a great beginning to the year for one of my go-to grid storage picks. And I’ll tease with the likelihood that more will be forthcoming. Among other things, I’m going to spend some time with Wilson next week recording a couple of episodes of my podcast Redefining Energy – Tech, talking about the eight year journey of Red John to date, the massive cell-based battery successes ILI has been delivering in parallel and what’s next for the firm. And maybe I’ll be involved with a pumped hydro initiative personally as well.