The First Battery in the Country Capable of Storing Wind Energy?

Michael Giberson

From the Star-Tribune in Minnesota, “Xcel looks to harness wind energy for use even when there’s no wind “:

Next spring Xcel Energy Inc., the state of Minnesota and a Virginia-based technology firm will test the first battery in the country capable of storing wind energy.

Well, that’s a bit wrong. Any battery is capable of storing wind energy. I can store wind energy on my laptop battery if I happened to be plugged into an outlet at a time when local wind power projects are producing. The problem isn’t storing electrical energy per se, but doing it economically.

When it is fully charged, the massive sodium-sulfur battery — which weighs about 80 tons — can store 7.2 megawatt-hours of electricity. That’s enough to power 500 homes for about seven hours. It will cost more than $5.4 million to buy and install the battery and analyze its performance. …

Xcel, which invested $3.6 million in the project, expects the battery “to become very important to both us and our customers,” [Xcel Chairman and CEO Dick ] Kelly said.

The article doesn’t say where the other $1.8 million is coming from, but the subtitle notes, “the project … also includes the state and a tech firm,” so maybe it is partly state tax dollars and partly entrepreneurial investment.

13 thoughts on “The First Battery in the Country Capable of Storing Wind Energy?”

  1. Hi Jeff! How are you?

    7.2 MWh, $5.4m => $0.00133333333/kWh.

    I wonder what the battery’s life will be: how many times do you get to do a full 7.2 MWh recharge?

    I defer to Mike in doing a more accurate analysis of how cost-effective this project is, though.

  2. 7.2 MWh… Hmmm, not bad for the battery, but what about the actual mean production of the wind power generator. Will the battery be mostly half-full or even sometimes more than full? How many half-load cycles can the battery survive? Are there any memory-effects that could limit the usability.

    I believe the company may well addresse those problems, but I am highly sceptical… Also, 7.2 MWh (7.2 MJ) are not that much all things considered.

  3. A battery that can store 7.2 mWh and discharge it over 7 hours is a one mW battery, more or less. The cost is therefor $5,400/kW. Here are some other articles about NaS batteries: USA Today NYTimes. The USA articles gives a 15 year life for the batteries.

    By way of comparison, Nuclear plants are now expected to cost about $6,300/kW. Future Pundit. They are also expected to deliver power about 90% of the hours in a year.

    We have yet to see any cost figures for giga-watt scale wind or solar plants, that include the storage and transmission mechanism that will be necessary. I expect that when they come-in they will be a lot higher than nuclear or fossil fuel.

  4. A battery that can store 7.2 mWh and discharge it over 7 hours is a one mW battery, more or less. The cost is therefor $5,400/kW. Here are some other articles about NaS batteries: USA Today NYTimes. The USA articles gives a 15 year life for the batteries.

    By way of comparison, Nuclear plants are now expected to cost about $6,300/kW. Future Pundit. They are also expected to deliver power about 90% of the hours in a year.

    We have yet to see any cost figures for giga-watt scale wind or solar plants, that include the storage and transmission mechanism that will be necessary. I expect that when they come-in they will be a lot higher than nuclear or fossil fuel.

  5. Lynne! Check those numbers! 7.2/5.4 is 1.3, and that’s watt-hours per dollar. What did you do? 😉

    You can’t go from an asset cost to a per-kWh cost unless you assume the battery’s life, the number of cycles, etc. But that isn’t the real economic analysis in any case. It’s a matter of the *value* of the energy(minus losses) as shaped with storage minus the *value* of the energy as-generated, present valued over the life of the asset and compared with the cost of the asset. You’d need an hourly locational price profile, storage cycling characteristics, and a probabilistic hourly generation profile to do it right. But it could be ball-parked with optimistic assumptions just to find out the best it could be.

    Remember that 7.2 MWh at $100/MWh is only $720 worth of energy. If the wind energy as-generated were worth $50/MWh, and if we made this trade-off daily, then you’d gain $360 per day for an investment of $5.4M. In a year we’d earn $131K, which is a return of around 2.5% per year. That’s way optimistic, assuming a full cycle and a $50 differential every day of the year. Let’s expand the differential by $25, that increases the return by 50%, about… 3.8%/year? It doesn’t seem like a great investment at this cost level, but this is a prototype, isn’t it?

    Somebody check my decimals…

  6. Assume that the battery goes through one complete cycle per day, e.g. it is charged by nuclear in the night and releases it in the day. Also, assume $0.10 per kWh.

    Income per day = 0.10 * 7200kW = $720 per day income
    Income per year = $263k

    Assuming 5% interest and infinite lifetime, that gives a value of $5.26 million.

    Ofc, there are other costs (including things like the actual nuclear/wind turbines), and it isn’t likely to have to have infinite lifetime. However, at least the price is on the right order of magnitude.

    Also, the cost price seems to include research costs, which would mean that the marginal cost is lower.

  7. Assume that the battery goes through one complete cycle per day, e.g. it is charged by nuclear in the night and releases it in the day. Also, assume $0.10 per kWh.

    Income per day = 0.10 * 7200kW = $720 per day income
    Income per year = $263k

    Assuming 5% interest and infinite lifetime, that gives a value of $5.26 million.

    Ofc, there are other costs (including things like the actual nuclear/wind turbines), and it isn’t likely to have to have infinite lifetime. However, at least the price is on the right order of magnitude.

    Also, the cost price seems to include research costs, which would mean that the marginal cost is lower.

  8. Thanks D.O.U.G.! I always get tripped up by the intersection of the engineering math and the econ math. At some point when someone explains how to work with MW and how to work with MWh to me, I hope it will finally stick.

  9. Thanks D.O.U.G.! I always get tripped up by the intersection of the engineering math and the econ math. At some point when someone explains how to work with MW and how to work with MWh to me, I hope it will finally stick.

  10. Note that while Sodium-Sulfur batteries have relatively efficient charge/discharge cycles (~89–92%), and are expected to have relatively long lifetimes (> 1500 charge/discharge cycles), they are suitable only for fixed-mount installations, since both the sodium and sulfur electrode materials are molten — these batteries operate at 300–350 C (570–660 F).

    Note also that sodium-sulfur batteries will incur a constant daily “overhead cost” because power must be constantly consumed to keep these batteries at their fairly high operating temperature. (Allow these batteries to cool off, and they become useless.) This constant daily “overhead cost” to keep these batteries hot will need to be factored into any economic analysis.

    Finally, note that 80 tonnes of molten sodium and sulfur at 300–350 C represents a significant fire and toxicity hazard, since either substance will spontaneously burst into violent flame and emit toxic smoke if they are accidentally exposed to air by a cracked battery housing — so the “Not In My Backyard” factor is likely to come into play after one of these installations inevitably catches fire…

  11. Note that while Sodium-Sulfur batteries have relatively efficient charge/discharge cycles (~89–92%), and are expected to have relatively long lifetimes (> 1500 charge/discharge cycles), they are suitable only for fixed-mount installations, since both the sodium and sulfur electrode materials are molten — these batteries operate at 300–350 C (570–660 F).

    Note also that sodium-sulfur batteries will incur a constant daily “overhead cost” because power must be constantly consumed to keep these batteries at their fairly high operating temperature. (Allow these batteries to cool off, and they become useless.) This constant daily “overhead cost” to keep these batteries hot will need to be factored into any economic analysis.

    Finally, note that 80 tonnes of molten sodium and sulfur at 300–350 C represents a significant fire and toxicity hazard, since either substance will spontaneously burst into violent flame and emit toxic smoke if they are accidentally exposed to air by a cracked battery housing — so the “Not In My Backyard” factor is likely to come into play after one of these installations inevitably catches fire…

Comments are closed.