Musings on COP, Part 2: How do you solve a problem like Dunkelflaute?

By Wendel Hortop, Senior Strategy Analyst at Open Energi

In the first of my two-part blog from COP26 (see here!), I wrote about how some of the hype around hydrogen may just be starting to get ahead of itself. Even if we get to the low prices for green hydrogen some forecast, the volumes required are enormous and scaling up will take a long time. So, in the mean-time, how do we balance a net-zero carbon power system with high renewable penetration, and keep it operating during the dreaded ‘dunkelflaute’?

Just what is a ‘Dunkelflaute’?

Dunkelflaute (literally, dark doldrums) essentially means those periods in winter where we have very low wind. Basically, when we have no wind or solar power, what do we do? These periods tend to be driven by low-pressure systems which have shown an increasing tendency to stall over northern Europe (such as in January of this year), bringing cold weather with low levels of wind for a pro-longed period of time (perhaps up to two weeks) – a worst-case scenario for a grid heavily reliant on wind generation. Right now we have a large gas-fleet to step in (although even that has struggled recently), but in the net-zero grid of 2035 this isn’t going to be an option.

So, just how do you solve a problem like Dunkelflaute?

It’s probably not lithium

When we think about these weather events, the major concern is the duration. We are seeing huge investment in the UK in large-scale, lithium-ion storage – in the next few years we will see the first 100MW and 200MW units commissioned – however pretty much all of this is 1-2 hours in duration. These assets are currently cycling once, twice or perhaps even three times a day, and anything that reduces this is going to heavily impact the business case for investment. Essentially, they are intraday storage vs. the interday storage we need during Dunkelflaute.

We are seeing serious cost reductions in lithium-ion technology, driven by the scale up of battery electric vehicles with many projections seeing storage of 4 or even 6 hours becoming feasible soon. However even in this case we would need a ridiculous amount of battery storage to meet this longer-duration requirement (in the 100s of GW) – and quite simply this is not going to happen from an economic standpoint, let alone the enormous environmental cost of such infrastructure. Lithium-ion has its own part to play, that it is very good at doing, but other solutions are needed for this longer-term flexibility.

What about long-duration storage?

So, if the current trend of lithium-ion is unlikely to solve this problem, are there any other energy storage options that could do? Perhaps some of the long-duration technologies you may well have seen mentioned in the news: iron-air batteries (such as those by Form Energy), liquid-air (such as Highview Power), or even a return to trusty pumped storage. The recent launch of the Long Duration Energy Storage (LDES) Council is an attempt to raise the profile of some of these solutions, as well as the actual need for a solution.

 Firstly, picking up on the above, we need to define what we really mean by long-duration (as do governments, in BEIS’ recent consultation on the topic they define long-duration as anything longer than just 4-hours). There’s a big difference between something that can provide 4 hours of energy, and something that can provide 1 week’s worth, and really it is the latter we need to start seeing if storage is going to have a role to play in managing this sort of event. Basically, we need multi-day storage, or even multi-week, not multi-hour.

Unfortunately, it’s not easy to make storage of this duration work financially – the economics get harder to stack up for every hour we extend the duration. Firstly, any extension on the duration causes the average arbitrage spread achievable to start to reduce (the very highest prices tend to last an hour or two a day at most). This is also before considering the possible lower efficiencies of longer-duration systems. Increasing renewable capacity can help here as we should see price volatility increase with much more times of zero or even negative pricing, however this same effect will persist to some degree.

Secondly, we start to be able to cycle through this additional energy capacity less and less. 4-hour storage can cycle twice a day, 12-hour storage just once a day, week-long storage once every two weeks. Therefore, combined with the available spreads reducing, this additional duration has to come at a much lower cost. Whilst we could see lithium-ion possibly reaching the oft-quoted magic $100/kWh mark soon, for these longer-duration technologies we need $10/kWh or perhaps even as low as $1/kWh for them to ever really work.

Long-duration storage is a super interesting topic as there are so many different technologies out there and is probably worthy of a much focused piece, but if you are interested in finding out more this episode of The Interchange podcast with the CEO of Form Energy is a very good place to start.

What else is there?

So if we might not see long-duration storage on the scale needed, at least anytime soon, what else is there can fill the gap? Firsly, increased interconnection should help, but many in northern Europe will see reliance on wind-power and are likely to experience exactly the same low-wind conditions at similar times as the UK. Perhaps some of the more moonshot-like projects like an interconnector cable to Morocco?

Or perhaps, as some in the industry claim, clean hydrogen can reach the cost levels and volumes required to support the power system sooner than many think? A fleet of hydrogen powered peaking plant could definitely solve the problem, but this is relying upon many different parts of the required infrastructure scaling up to the volumes required, and all at a cost that is affordable, and there are many reasons to think this is unlikely to happen.

That leaves us with carbon capture and storage (CCS) gas plants. This feels like the mostly likely option seeing as we have such a large gas fleet today, but given the high costs of fitting CCS equipment, plus the spotty record of CCS projects to date these are not guaranteed to deliver net-zero carbon power, and will also likely see lower and lower running hours over time as renewables and storage increase in capacity.

All-in-all, it doesn’t seem like there’s a perfect Dunkelflaute solution out there.

So, is this it?

Is that it then, is the dreaded Dunkelflaute going to get us?

As tends to be the case, we probably will (and need to) see some combination of “all of the above” coming through, especially in the 2030s. It seems unlikely any single technology will be the clear winner in terms of cost and scalability by the end of this decade, and we really need to see the MWs start to add up soon. However, the costs of many of these solutions are currently prohibitively high and are probably not going to start falling significantly until we start to see much more capacity being installed.

The key driver here then is policy and market design. How do we incentivise these technologies to come through before the market is necessarily there to support them, or their costs are low enough to be worthwhile investing in? Do we need a capacity market that goes beyond thinking in hours? Or something else? The fact that BEIS is actively looking into this is a good thing, but it’s a very hard problem to solve.

Finally, just is this a global problem?

A final interesting point that I noticed from a comment by Lord Adair Turner (ex- chairman of the Climate Change Committee) at a great panel one morning. A lot of the world sees a very different power problem to those of us in northern Europe, with many countries closer to the equator seeing much more consistent solar production on a daily basis throughout the year. For these regions, there is a very real possibility of just shifting this energy across the rest of the day when it is required, and therefore we really could see multi-hour storage being the solution for much of the world – perhaps even lithium-ion. What this does unfortunately mean is that we won’t see necessarily see some of the same cost-reductions for these longer-duration technologies as we have seen for solar, wind and battery storage, as we likely just won’t see the same need for them on a global scale. Another difficulty to add to all of the others!