← Back to context

Comment by dgacmu

1 year ago

This was a little buried, so surfacing some #s that seemed interesting to put this in perspective:

- 565 MWh of storage capacity

- 185 MW of instantaneous power delivery capacity

- $219M of financing for the project

Hawaii's residential electricity price is roughly $0.415 per kWh vs a US average of $0.162.

https://ourworldindata.org/battery-price-decline

https://www.energy-storage.news/global-bess-deployments-to-e...

Start where electricity is expensive and/or the revenue you steal from thermal generators (grid support mentioned, synthetic inertia, black start capability, etc) supports the economics, and work your way down as battery costs decline and you force thermal generators to become uneconomic due to compressing their runtimes. Think in systems.

  • Yup, absolutely. Places with high energy costs due to being geographically isolated / without a lot of local energy resources have always struck me as some of the best initial places for solar+battery.

    • I worked on a solar project a number of years back that was one of the first that was actually independently financially sustainable. It was in west Texas in an area that had a highly distributed population and very hot summers. So the existing energy sources were already higher than normal and had the added dimension of spiking demand. Perfect environment for solar to be competitive.

      6 replies →

  • Can you clarify your usage of "thermal" here? Most everything except photovoltaic is thermal.

    In the US, we usually name the heat source -- coal, natural gas, nuclear -- even though these are all thermal in operation. And the word 'thermal' does not show up in any of those when we talk about them.

    The only time the word 'thermal' shows up in US usage is with the 'geo' prefix, and I can't imagine compressing the runtime of a geothermal plant, it's the perfect base-load plant. Are we talking about different things?

  • If only the "systems" we were considering were meant to provide limitless and virtually free electricity (nuclear), which is congruence with the "systems" of reducing poverty.

    • Enough sunlight lands on the Earth every 2 minutes to power humanity for a year [1]. ~500-600GW of solar will be deployed in 2024 globally, and we are accelerating to 1TW deployed annually [2].

      Commerical nuclear fission is unviable at this point [3], even at nimble startups [4] [5], but proponents are free to argue in support of it to anyone who will still listen. Renewables and batteries have reached an escape velocity trajectory [6].

      This global energy system will eliminate energy poverty in our lifetime, and like bankruptcy, it'll happen slowly, and then all of a sudden.

      [1] https://news.ycombinator.com/item?id=37502924

      4 replies →

    • Electricity from nuclear is neither limitless nor free. While we would have been much better off (in terms of global warming) if we had not hobbled nuclear power generation decades ago, at this point it's cheaper and faster to build out solar and wind than nuclear.

      14 replies →

    • The fifties want their nuclear advertising back…

      Nuclear is rather expensive and, with current technology, not „limitless“ in any sense of the word

      3 replies →

    • Nuclear is definitely part of the mix we need, but we can easily do multiple things.

      For one thing, it's neither limitless nor free - the limit is the amount of radioactive ore we mine, and the cost is the cost of setting up a plant, running it, mining the ore, purifying it, transporting it,... The cost of nuclear is actually pretty high. I'm not talking about safety except that the cost factors in both passive and active safety mechanisms. And, they take _forever_ to build and bring to operation.

      On the other hand, the price of solar (even without subsidy) is already cost competitive with _coal_ leave alone nuclear.[1] But it's intermittent, and batteries like the article are expensive.

      So, the question is not either this or that, but what's the right mix...

      [1]: https://upload.wikimedia.org/wikipedia/commons/4/48/Electric...

      1 reply →

  • Don't forget to factor in the thermal generators' owners abandoning their business way before you thought they would, decades before there's a viable replacement for on-demand power to run an advanced industrial economy.

People always forget that batteries also absorb power. Having a lot of renewables means there are energy spikes far exceeding what can be used in that moment. Without batteries, that energy is lost. Having batteries means that energy can be buffered and used later (e.g. in the evening). So they improve the capacity factor of existing installed renewables. Add domestic batteries, EV batteries, etc. to the mix and you also get the potential for demand shaping where you charge those when renewable energy production is spiking and prices are low. And of course even though that is currently not utilized on a large scale, all those EVs could technically provide energy back to the grid as well.

Another point is that batteries like this are not actually intended for long term storage. They are instead about stabilizing the grid and dealing with short term spikes and dips in supply and demand of energy. Unlike a coal or gas plant, a battery can respond in milliseconds and be very cost effective for that. Spinning up coal and gas plants is expensive and slow. And they cost money when they are not running.

And while that single coal plant was able to provide so-called baseload; it would only have been able to do so if it was up and running 24/7/365. And that wouldn't be true. They are very reliable but occasionally coal plants have to be down for maintenance, repairs, etc. and this can take quite some time (weeks/months). Same with nuclear plants. So, relying on that to not happen was never a good plan.

Long term storage is always assumed to be needed to compensate for a lack of this baseload. However, baseload is actually a fuzzy notion until you express it in gwh and gw. Hawaii seems to be in the process of proving this might be a lot less than some people seem to assume. At least I'm not aware of them having any long term storage. They'll probably add more battery and resilience to their grid over time in the form of more wind and solar generation and additional batteries. But if these people modeled this correctly and did their homework, this might actually be fine as is. We'll find over time I guess.

  • Do we currently have enough renewables installed in (eg) the the UK for batteries to increase capacity factor? Is there ever enough renewable production that energy is lost?

    • Yes, but it's at a local level.

      One of the benefits of batteries is that they can be spread around and used to alleviate bottlenecks. Building transmission is very expensive, so this is a good early market for them.

      These are called Non-Tranmissikn Alternatives or Non-Wires Alternatives:

      NTAs are programs and technologies that complement and improve operation of existing transmission systems that individually or in combination defer or eliminate the need for upgrades to the transmission system.

    • Wind farms often get turned off because there is too much Wind, solar is also often throttled. There is a lot of "lost" power that could rather find its way into batteries or H2-electrolyzers.

      1 reply →

Hawaii, a remote island in the middle of the pacific, pays less than the 2024 pg&e prices for the bay area. PG&E are the worst.

So estimating the lifetime of the battery at 5000 cycles and lets say round trip efficiency at 95% we end up with $0.082 / kWh. (EDIT: originally I claimed $0.074 which is wrong) that the battery adds.

So I'm guessing in the long run this will considerably lower the cost of electricity on the island as adding PV capacity is much cheaper than keeping a coal plant running and this battery allows to install much more and use the energy at night. Not sure whether Hawaii has much wind power but it would seem to be rather windy place.

  • Can you explain your logic a bit more? I'm struggling to understand how you calculated the $0.074, and what you are saying it represents.

    Edit: I suspect your calculations just represent depreciation over the batteries lifetime, which is only one of the costs involved.

    • The capacity of the battery is 565 MWh.

      The cycle life of these kinds of batteries is about 5000. Meaning they get about 5000 charge and discharge cycles before their useful life is over. It could be 2000 it could be 10000 and the definition of useful is also dependent on application.

      So in it's lifetime this battery can store 5000 * 565 = 2825000 MWh

      The cost of the system was $219M.

      About 5% of energy is going lost due to inefficiencies.

      $219M / (5000 * 565 * 0.95) = $81.6/MWh = $0.082 / kWh.

      I am sorry for calculating the efficiency incorrectly in the original post.

      This does not take into account the maintenance cost.

      5 replies →

  • Typically at the moment we talk about a price of about $15 a MWH for Wind and $14 for Solar (last year anyway). So around $0.15 p/KWH for the power to charge and discharge it. Assuming the wind/solar is only going for a third of the day that brings the average price up to about $.209 p/kwh when we take into account battery wear cost. That is definitely economically viable in a very large number of places in the world.

    Incidentally the totals work out about the same on a home solar system, my battery is 0.09 p/kwh and the Solar output averages out to about 0.07p/kwh but get paid for export at 0.15 p/kwh.

  • That's close to my guesstimates of about $0.10/kwr. So I tend to believe it.

    The important thing is battery storage is competitive with peaking plants over a period of hours. And lowest cost when it comes to short term supplies on the order of seconds to an hour.

    Also the logistics of containerized batteries is great. You need a place to put them and a grid connection. And nothing more than that.

  • Are you assuming zero cost for the power to charge the battery?

    • No. This is additional on top of energy production. Energy production cost was already in the base price quoted. The energy consumption will be roughly the same unless the price changes dramatically.

      But this allows more PV generation to be put in which is the cheapest way of producing energy.

  • it won’t make any impact on the prices there because it’s a drop in the bucket compared to what they spend importing oil and diesel to burn for the majority of their electricity

  • rountrip is closer to 85% and you have to add the power electronic, also the graph is cell cost of module/pack/gigapack and security systems...

What I don't get is that this is meant to replace a 180MW coal plan, so we are talking about 3h worth of electricity at full load. Not sure how volatile is the weather in Hawaii, but in Europe, when there is no wind, it can last days not hours.

$219,000,000.00 / 185,000 kwh = $1,183.78 per kwh.

Seems kind of on the expensive side, but maybe it's reasonable for this kind of project -- and there might be some big one-time costs like connecting the site to the power grid.

Seems like there's a lot of room to drive costs down though. Some company could plausibly buy the batteries for $100/kwh, sell a completed power station for $200/kwh, and still make a profit.