5 ms·

The alternative is large grid scale batteries, like Tesla’s ultra packs or whatever, right? Has anyone studied what the impact of large scale battery installati

The alternative is large grid scale batteries, like Tesla’s ultra packs or whatever, right? Has anyone studied what the impact of large scale battery installations would be?

pfdietz 2y ago
Batteries, yes, or peakers run on clean e-fuels. The latter has lower round trip efficiency but potentially lower capex.
- danans 2y ago
  > The latter has lower round trip efficiency but potentially lower capex Lower capex, but massive opex since clean e-fuels are incredibly expensive and AFAICT not on a fast path to getting cheaper, unlike batteries.
  - pfdietz 2y ago
    Ah, but if run at very low capacity factor the cost of fuel is not very important. Batteries handle most of the stored energy throughput, but the e-fuel backup provides most of the storage capacity. Think of it as like the relationship between cache memory and slower RAM: the combination performs better than either alone.
    - danans 2y ago
      > Think of it as like the relationship between cache memory and slower RAM: the combination performs better than either alone. Seems to me like upcoming long term storage battery chemistries chemistries might be more up to the job analogous to RAM than e-fuels. E-fuels seem more useful for heavy transport and flights than for feid power.
      - pfdietz 2y ago
        The capex for underground hydrogen storage capacity can be very low (like $1/kWh of storage capacity; keep in mind this is not including per-power costs). This is a very challenging price point for batteries to hit. The key point here is that, like flow batteries, e-fuels can separate per-energy and per-power costs. Ordinary batteries can't do that, at least to the same extent. Things like iron-air batteries might be good for an intermediate timescale, like ~1 week. They would be intermediate on the efficiency-capex tradeoff. They might not be good for rare event backup (the dreaded Dunkelflaute) or seasonal leveling, but could be a good match for normal wind variability, which tends to be on a somewhat longer timescale than solar. BTW, electrolysers from China have been on a downward price curve, similar to that seen in other renewable technologies. Low cost electrolysers (at perhaps mediocre efficiency) are the pacing technology for green hydrogen from intermittent renewables. One would like to be able to use them when renewables are providing excess power, not have to run them 24/7 to amortize high capex (as the old nuclear-powered hydrogen economy idea had it.)
ZeroGravitas 2y ago
Yes, Texas and California have actually done it so the results are there to be studied: https://blog.gridstatus.io/ercot-hot-streak-cold-winter/ https://blog.gridstatus.io/ercot-hot-streak-cold-winter/ > ERCOT set new records in load, net load, solar, prices, and battery discharge within a few days, let's explore the data. https://blog.gridstatus.io/batteries-ercot-ancillary-services-2024/ https://blog.gridstatus.io/batteries-ercot-ancillary-service... > For most of their tenure on the ERCOT grid, batteries have largely made their bones supplying ancillary services as opposed to energy, but that is beginning to change. https://blog.gridstatus.io/caiso-batteries-apr-2024/ https://blog.gridstatus.io/caiso-batteries-apr-2024/ > Batteries have taken a huge leap forward in CAISO this spring, shifting from a noteworthy trend into a major force impacting operations of the grid https://blog.gridstatus.io/caiso-beats-the-heat/#batteries-eureka-moment https://blog.gridstatus.io/caiso-beats-the-heat/#batteries-e... > Since the last challenging stretch in CAISO, battery storage has expanded dramatically. Last week, batteries discharge peaked from ~5 to ~7 GWs on each day. In 2022, the peak day was around half of that 7GW value, while most days throughout the week peaked well under 3 GW.

The alternative is large grid scale batteries, like Tesla’s ultra packs or whatever, right? Has anyone studied what the impact of large scale battery installations would be?

pfdietz 2y ago

Batteries, yes, or peakers run on clean e-fuels. The latter has lower round trip efficiency but potentially lower capex.

danans 2y ago
> The latter has lower round trip efficiency but potentially lower capex Lower capex, but massive opex since clean e-fuels are incredibly expensive and AFAICT not on a fast path to getting cheaper, unlike batteries.
- pfdietz 2y ago
  Ah, but if run at very low capacity factor the cost of fuel is not very important. Batteries handle most of the stored energy throughput, but the e-fuel backup provides most of the storage capacity. Think of it as like the relationship between cache memory and slower RAM: the combination performs better than either alone.
  - danans 2y ago
    > Think of it as like the relationship between cache memory and slower RAM: the combination performs better than either alone. Seems to me like upcoming long term storage battery chemistries chemistries might be more up to the job analogous to RAM than e-fuels. E-fuels seem more useful for heavy transport and flights than for feid power.
    - pfdietz 2y ago
      The capex for underground hydrogen storage capacity can be very low (like $1/kWh of storage capacity; keep in mind this is not including per-power costs). This is a very challenging price point for batteries to hit. The key point here is that, like flow batteries, e-fuels can separate per-energy and per-power costs. Ordinary batteries can't do that, at least to the same extent. Things like iron-air batteries might be good for an intermediate timescale, like ~1 week. They would be intermediate on the efficiency-capex tradeoff. They might not be good for rare event backup (the dreaded Dunkelflaute) or seasonal leveling, but could be a good match for normal wind variability, which tends to be on a somewhat longer timescale than solar. BTW, electrolysers from China have been on a downward price curve, similar to that seen in other renewable technologies. Low cost electrolysers (at perhaps mediocre efficiency) are the pacing technology for green hydrogen from intermittent renewables. One would like to be able to use them when renewables are providing excess power, not have to run them 24/7 to amortize high capex (as the old nuclear-powered hydrogen economy idea had it.)

ZeroGravitas 2y ago

Yes, Texas and California have actually done it so the results are there to be studied: https://blog.gridstatus.io/ercot-hot-streak-cold-winter/ https://blog.gridstatus.io/ercot-hot-streak-cold-winter/ > ERCOT set new records in load, net load, solar, prices, and battery discharge within a few days, let's explore the data. https://blog.gridstatus.io/batteries-ercot-ancillary-services-2024/ https://blog.gridstatus.io/batteries-ercot-ancillary-service... > For most of their tenure on the ERCOT grid, batteries have largely made their bones supplying ancillary services as opposed to energy, but that is beginning to change. https://blog.gridstatus.io/caiso-batteries-apr-2024/ https://blog.gridstatus.io/caiso-batteries-apr-2024/ > Batteries have taken a huge leap forward in CAISO this spring, shifting from a noteworthy trend into a major force impacting operations of the grid https://blog.gridstatus.io/caiso-beats-the-heat/#batteries-eureka-moment https://blog.gridstatus.io/caiso-beats-the-heat/#batteries-e... > Since the last challenging stretch in CAISO, battery storage has expanded dramatically. Last week, batteries discharge peaked from ~5 to ~7 GWs on each day. In 2022, the peak day was around half of that 7GW value, while most days throughout the week peaked well under 3 GW.