Grid-scale battery fire in upstate New York

[goingplacesanddoingstuff]

Every single energy storage mechanism has some risk. It’s storing energy, after all.

Some are safer than others. I wonder if these batteries used safer Li polymer or LiFePO4 technology. I doubt it.

Flow batteries have been mentioned. To my knowledge there have been no successful commercial flow battery companies. For whatever reason, it’s just too hard to make a profit on them as of yet.

Pumped hydro only works where you can pump water into a giant reservoir with greater head than the start. That doesn’t work in many places and carries all the same risks as dams. New York blessed with the geography to already be one of the most hydro-heavy states, but it’s not an infinitely scalable solution. There’s only so much land you can flood.

Another option is inertial flywheel batteries, where a magnetically levitated heavy disk spins at insane rpm. You can imagine the risks there. Yet another could be chemical storage, using ammonia. Again, you can imagine the risks there too.

No free lunch in physics.

 

[MVAR]

Every single energy storage mechanism has some risk. It’s storing energy, after all.

Some are safer than others. I wonder if these batteries used safer Li polymer or LiFePO4 technology. I doubt it.

Flow batteries have been mentioned. To my knowledge there have been no successful commercial flow battery companies. For whatever reason, it’s just too hard to make a profit on them as of yet.

Pumped hydro only works where you can pump water into a giant reservoir with greater head than the start. That doesn’t work in many places and carries all the same risks as dams. New York blessed with the geography to already be one of the most hydro-heavy states, but it’s not an infinitely scalable solution. There’s only so much land you can flood.

Another option is inertial flywheel batteries, where a magnetically levitated heavy disk spins at insane rpm. You can imagine the risks there. Yet another could be chemical storage, using ammonia. Again, you can imagine the risks there too.

No free lunch in physics.

The talk about stored energy is why there is AGC (automatic generation control). The power plant operators bid into the regulation market once their facility is certified. If awarded, there will be a regulation award on top of their base load award. The AGC computer will control Hz via the most economical dispatch at a facility with enough regulation available & needed ramp rate.

Hydroelectric generators are preferred as they offer, usually, the fastest ramp rates as well as the largest D-Curve. Combustion & steam turbines are next while PV & Wind are not used to regulate. I’m also not aware of any nuclear units in service for regulation. Only have older Gen 2 units in my area.

My username is being changed. Looks like a company started using the username I’ve used since the mid 90’s.

 

[OVERKILL]

Every single energy storage mechanism has some risk. It’s storing energy, after all.

Some are safer than others. I wonder if these batteries used safer Li polymer or LiFePO4 technology. I doubt it.

Flow batteries have been mentioned. To my knowledge there have been no successful commercial flow battery companies. For whatever reason, it’s just too hard to make a profit on them as of yet.

Pumped hydro only works where you can pump water into a giant reservoir with greater head than the start. That doesn’t work in many places and carries all the same risks as dams. New York blessed with the geography to already be one of the most hydro-heavy states, but it’s not an infinitely scalable solution. There’s only so much land you can flood.

Another option is inertial flywheel batteries, where a magnetically levitated heavy disk spins at insane rpm. You can imagine the risks there. Yet another could be chemical storage, using ammonia. Again, you can imagine the risks there too.

No free lunch in physics.

Yes, PHES is geographically constrained, but currently the only storage mechanism that can offer true long-term storage. As you mention, New York has abundant hydro, PHES could work there. The combined size of all of these projects is far lower than even the Adam Beck PGS, which in turn, is much smaller than Racoon Mountain.

In terms of flow batteries, Spencer Energy is the first one I know of:

The Project - Yadlamalka Energy

yadlamalkaenergy.com

Which uses a Vanadium flow battery. It's a paltry 2MW/8MWh, but it's a demo project.

It was completed in June:

Habitat Energy to optimise the largest solar & storage project of its kind in the Southern Hemisphere - Habitat Energy

The Spencer Energy project combines a 6MWp solar farm with a 2MW/8MWh flow battery, the largest of its kind in the Southern Hemisphere.

www.habitat.energy

At $22M though, it was rather expensive. But, depending on how long it can last, this may be a much better solution than lithium.

 

[OVERKILL]

The talk about stored energy is why there is AGC (automatic generation control). The power plant operators bid into the regulation market once their facility is certified. If awarded, there will be a regulation award on top of their base load award. The AGC computer will control Hz via the most economical dispatch at a facility with enough regulation available & needed ramp rate.

Hydroelectric generators are preferred as they offer, usually, the fastest ramp rates as well as the largest D-Curve. Combustion & steam turbines are next while PV & Wind are not used to regulate. I’m also not aware of any nuclear units in service for regulation. Only have older Gen 2 units in my area.

My username is being changed. Looks like a company started using the username I’ve used since the mid 90’s.

Our nukes all do frequency control as part of standard operation here in Ontario, FWIW. But, ours can also island and be used for black start, which is apparently not permitted in the US, based on my discussions with operators there. I have an old article that goes into the various operating modes the plants can run in, including load following by varying reactor power and steam bypass, but this mode isn't used anymore, though we can, and do, if required, use steam bypass to reduce output (curtail), but this is done in blocks.

 

[MVAR]

Yes, PHES is geographically constrained, but currently the only storage mechanism that can offer true long-term storage. As you mention, New York has abundant hydro, PHES could work there. The combined size of all of these projects is far lower than even the Adam Beck PGS, which in turn, is much smaller than Racoon Mountain.

In terms of flow batteries, Spencer Energy is the first one I know of:

The Project - Yadlamalka Energy

yadlamalkaenergy.com

Which uses a Vanadium flow battery. It's a paltry 2MW/8MWh, but it's a demo project.

It was completed in June:

Habitat Energy to optimise the largest solar & storage project of its kind in the Southern Hemisphere - Habitat Energy

The Spencer Energy project combines a 6MWp solar farm with a 2MW/8MWh flow battery, the largest of its kind in the Southern Hemisphere.

www.habitat.energy

At $22M though, it was rather expensive. But, depending on how long it can last, this may be a much better solution than lithium.

Another positive for turbines is droop control. When there is a large disturbance to the grid & Hz takes a hit in either the up or down direction, the governor droop control will respond nearly instantly in the opposite direction to help arrest Hz decline or increase. The physical mass of the spinning generators are electromechanical coupled to the grid. Their increase or decrease in speed is used to maintain Hz in these situations of trouble.

This feature is not available with PV & wind generation. There’s no mass associated with battery storage either. NERC has released several IBR (Inverter Based Resource) recommendations & papers the last few years.

 

[MVAR]

Our nukes all do frequency control as part of standard operation here in Ontario, FWIW. But, ours can also island and be used for black start, which is apparently not permitted in the US, based on my discussions with operators there. I have an old article that goes into the various operating modes the plants can run in, including load following by varying reactor power and steam bypass, but this mode isn't used anymore, though we can, and do, if required, use steam bypass to reduce output (curtail), but this is done in blocks.

View attachment 169387

I’ll have to read that when back at a desktop.

Nuke units down here are baseloaded only. No regulation (the only test I previously read had very slow ramp rates that just barely reached 3mw/min). In fact, top priority during blackout restoration is to supply 230kV offsite power to any nuclear stations.

 

[OVERKILL]

Another positive for turbines is droop control. When there is a large disturbance to the grid & Hz takes a hit in either the up or down direction, the governor droop control will respond nearly instantly in the opposite direction to help arrest Hz decline or increase. The physical mass of the spinning generators are electromechanical coupled to the grid. Their increase or decrease in speed is used to maintain Hz in these situations of trouble.

This feature is not available with PV & wind generation. There’s no mass associated with battery storage either. NERC has released several IBR (Inverter Based Resource) recommendations & papers the last few years.

Yes, the inertia big turbines have inherently contributes to grid stability by absorbing variations. Batteries provide "synthetic" inertia, but are obviously limited, I remember when the VRE manure spreader Renew Economy was claiming that the Hornsdale battery was able to rapidly step-in and prevent grid collapse when a unit, several times the size of the battery, tripped.

This is the graph they used:

Look at the Y axis...

Here's the corrected graph:

Apparently, many years ago, they did investigate Diablo Canyon doing load following. Units in France regularly do it and the units in Germany, before they stupidly shuttered them, did it as well. However, the economics are the best in pure baseload, so they didn't take it any further beyond confirming that the plant could, if the operator wanted to.

I find it interesting that US plants can't island. We used Bruce B and Darlington to bring the grid back up in Ontario after the 2003 blackout.

Bruce, being a good distance away from our biggest load centre, experienced a lot of interruptions during its early days and the folks there became extremely good at riding them through, which included islanding units when necessary if the transmission capacity wasn't available.

 

[MVAR]

Yes, the inertia big turbines have inherently contributes to grid stability by absorbing variations. Batteries provide "synthetic" inertia, but are obviously limited, I remember when the VRE manure spreader Renew Economy was claiming that the Hornsdale battery was able to rapidly step-in and prevent grid collapse when a unit, several times the size of the battery, tripped.

This is the graph they used:
View attachment 169391

Look at the Y axis...

Here's the corrected graph:
View attachment 169392

Apparently, many years ago, they did investigate Diablo Canyon doing load following. Units in France regularly do it and the units in Germany, before they stupidly shuttered them, did it as well. However, the economics are the best in pure baseload, so they didn't take it any further beyond confirming that the plant could, if the operator wanted to.

I find it interesting that US plants can't island. We used Bruce B and Darlington to bring the grid back up in Ontario after the 2003 blackout.

Bruce, being a good distance away from our biggest load centre, experienced a lot of interruptions during its early days and the folks there became extremely good at riding them through, which included islanding units when necessary if the transmission capacity wasn't available.

DCPP was in my old control area as a transmission operator never knew about the test for AGC with one of the units.

Isochronous mode is selected for the unit with the largest mw range and fastest ramping ability in the stated island. Isoch is a zero droop setting so the unit selected will quickly react for Hz control. Due to the fast response requirements, if available, hydro units are usually selected if available. I’ve seen geothermal units selected as isoch units when needed though.

 

[OVERKILL]

DCPP was in my old control area as a transmission operator never knew about the test for AGC with one of the units.

Isochronous mode is selected for the unit with the largest mw range and fastest ramping ability in the stated island. Isoch is a zero droop setting so the unit selected will quickly react for Hz control. Due to the fast response requirements, if available, hydro units are usually selected if available. I’ve seen geothermal units selected as isoch units when needed though.

Yeah, most of our load following is done by hydro units, hydro provides ~25% of our total electricity, nuclear ~60%. The rest is gas plus a smattering of wind/solar/biomass.

 

[fdcg27]

Events of this sort should bring pause to those considering a home powerwall.

 

[OVERKILL]

DCPP was in my old control area as a transmission operator never knew about the test for AGC with one of the units.

Heather, who runs Mothers for Nuclear recently shared a post about it on twitter, I wasn't aware they had look at it until that point either. A good friend of mine and co-advocate was a grid operator for Ontario Hydro, then later our IESO here in Ontario. I've had many a good chat with him and he has been any invaluable asset for me in learning about grid operations stuff. After we toured Bruce, he and I sat down with @Rand and chatted for several hours about some of the history of the transmission and control stuff in relation to Bruce specifically, as well as how load rejection was used and the islanding.

 

[goingplacesanddoingstuff]

Events of this sort should bring pause to those considering a home powerwall.

As long as it's maintained, is it really that much different from storing gasoline in the garage, or having natural gas lines?

 

[repairman54]

As long as it's maintained, is it really that much different from storing gasoline in the garage, or having natural gas lines?

Storing gas in an attached garage is a big no. Natural gas lines are safe if not leaking, same with propane. Vapors mixing with oxygen are the issue with fuels and combustion.
Powerwall is just another big lithium battery inside your house, no different than an EV in your attached garage. Internal shorts in batteries are not ''maintainable''.

 

[OVERKILL]

As long as it's maintained, is it really that much different from storing gasoline in the garage, or having natural gas lines?

I'd argue it is, as lithium ion batteries can develop dendrites through use that can then cause sudden failure of the cell. This can lead to sudden, and unpredictable, events that result in thermal runaway and fire. A jug of gasoline will not spontaneously ignite and burn, it requires being exposed to air and an ignition source.

How many people in New York State do you think have gasoline jugs in their garage? Now, how many of those have caught fire in the last three months? Then compare that to the number of grid batteries in the state, and the number that have caught fire in the same period. This is an alarming trend, and seems to be happening at a much higher rate than with other devices that are also using lithium batteries, likely due to the sheer volume of cells in these systems.

 

[goingplacesanddoingstuff]

I'd argue it is, as lithium ion batteries can develop dendrites through use that can then cause sudden failure of the cell. This can lead to sudden, and unpredictable, events that result in thermal runaway and fire. A jug of gasoline will not spontaneously ignite and burn, it requires being exposed to air and an ignition source.

How many people in New York State do you think have gasoline jugs in their garage? Now, how many of those have caught fire in the last three months? Then compare that to the number of grid batteries in the state, and the number that have caught fire in the same period. This is an alarming trend, and seems to be happening at a much higher rate than with other devices that are also using lithium batteries, likely due to the sheer volume of cells in these systems.

I said well maintained, mind you. Don’t ignore that. There are ways to engineer around degradation.

When gasoline automobiles were new, there was tremendous concern over the volatility of petrol. Accidents were more common than today. People freaked out. The technology and best practices then evolved.

This happens for each new energy technology. We already have safer Li battery tech, such as the aforementioned Li-po and LiFePO4. These are being used in more applications and incident rates are lower.

 

[OVERKILL]

I said well maintained, mind you. Don’t ignore that. There are ways to engineer around degradation.

You can't "maintain" away dendrites though unfortunately, they grow inside the cells. Yes, proper maintenance should result in the packs lasting longer, and being safer, but the packs being discussed here were all big name brands with claims as to their resilience and safety.

When gasoline automobiles were new, there was tremendous concern over the volatility of petrol. Accidents were more common than today. People freaked out. The technology and best practices then evolved.

OK, but that took what, 30-40 years? Things were pretty well sorted by the 1960's.

This happens for each new energy technology. We already have safer Li battery tech, such as the aforementioned Li-po and LiFePO4. These are being used in more applications and incident rates are lower.

These are all recent projects

And of course Lithium ion isn't new, it's pushing 40 years old now.

I agree, we have safer cell tech available, which is why it is concerning that we continue to build installations that catch fire, such as the recent Aussie project.

 

[goingplacesanddoingstuff]

You can't "maintain" away dendrites though unfortunately, they grow inside the cells. Yes, proper maintenance should result in the packs lasting longer, and being safer, but the packs being discussed here were all big name brands with claims as to their resilience and safety.

At the macro scale, there are there technologies that can separate, monitor, regulate overcharge, and deactivate cells before they become problematic. Proper maintenance must also include identification and modular replacement of cells at correct intervals.

At the micro scale, there are also technologies which reduce dendrite formation on the SEI, such as increasing polymer yield strength in Li-po as well as potential alternatives in electrolyte composition, reducing ethylene carbonate, changes to an anode structure, using SICPEs, and more. And when it comes to LiFePO4, the fire danger is much lower in the first place.

This is a hot topic in active research with progress being made every year. My point is just that I wouldn’t write it off yet.

These are all recent projects

And of course Lithium ion isn't new, it's pushing 40 years old now.

“Traditional old” LiCoO2 is still in heavy use today. There is a lag time before the latest tech gets deployed.

 

[Cujet]

I have to say it. There is only one way forward, and it does not include big fans or 93 million square miles of black panels.

 

[Cujet]

Fire and smoke map, with an Air Quality Index rating. I find it quite helpful. I've also discovered that the AQI is accurate. When it is in the 100 range the smoke obscures the horizon and I get a sore throat, even if I can't smell the smoke. We've had tons of smoke in NY this summer. Mostly due to the burning nuclear plants, er, ah, forests, in Canada (a little joke)

Fire and Smoke Map

fire.airnow.gov

 

[OVERKILL]

At the macro scale, there are there technologies that can separate, monitor, regulate overcharge, and deactivate cells before they become problematic. Proper maintenance must also include identification and modular replacement of cells at correct intervals.

At the micro scale, there are also technologies which reduce dendrite formation on the SEI, such as increasing polymer yield strength in Li-po as well as potential alternatives in electrolyte composition, reducing ethylene carbonate, changes to an anode structure, using SICPEs, and more.

I appreciate the technical points, but at the end of the day, we are still experiencing a staggering number of fires despite this, three in the last three months in New York alone. We can't just hand-wave that away because we have systems in place that *should* prevent them. Again, as I noted, these are all from big names that many would recognize, that is rightfully concerning.

And when it comes to LiFePO4, the fire danger is much lower in the first place.

Yes, I appreciate that the lithium iron phosphate chemistries are less likely to catch fire (this is what my dear friend has for his off-grid setup), but there IS still a risk. I was under the impression that Tesla was switching to this with their Megapacks, but it's unclear as to when that transition actually took(?) place, and both Victoria and Moss Landing have had fires, the latter being an extremely recent project, commissioned the middle of last year.

This is a hot topic in active research with progress being made every year. My point is just that I wouldn’t write it off yet.

I'm not writing it off, far from it, I think batteries are quite practical in cars, if we can get some of this stuff sorted. And, the rate of seemingly spontaneous self-ignition seems much lower in the the transportation segment than it does the grid storage one. But, that doesn't mean we shouldn't be concerned about, or discussing these sorts of events.

I do think there are better options for grid storage however.

“Traditional old” LiCoO2 is still in heavy use today. There is a lag time before the latest tech gets deployed.

Yes, so I think it would be beneficial if we had more details on what the types of cells were in these systems that have caught fire. That's often omitted, and while Tesla advertised the Megapack as being Lithium Ion, it seems to have potentially already transitioned to LFP as early as 2021, so sometimes the media is outdated or inaccurate.