Batteries can do this, along with their other functions. ISOs should require them if you are interconnecting non-dispatchable generation. I say this is a renewable project developer.
I thinnk that you are spot-on in your analysis, btw. I learned about this in 2002 when I started out as a renewable developer. I was schooled in it by a very kind utility engineer.
California has been working for 15 years to develop standards to make solar inverters play as nicely as possible. Part of the problem is that their first reaction when there is a frequency event is to trip off. When the frequency event is caused by a sudden drop-off of generation (which I believe happened in Iberia), having the solar drop-off, when it is contributing a huge amount of the total generation, could have been truly catastrophic.
Tripping off or reducing output due to over frequency is pretty easy. That is what the plants in Spain should have done - and maybe they did do it.
If the system was importing power during the separation, they would be in under frequency. Getting solar and wind plants to contribute additional power to solve under frequency is much harder. They basically have to always operate at 80% of their capabilities so they can ramp up to 100% during an under frequency event.
Paying for 100% of a plant and then only using 80% of it is expensive.
I have often seen this claim made, but I am not buying it. No matter how good your grid forming inverter is, it still has to depend on PID logic. A PID loop senses error, then reacts to correct it, that takes time. Time is something you don't have during a disturbance. So the inverters can mimic a rotating machine, but they will never do what a rotating machine does.
I too haven't seen the claim verified however it should be possible since an accurate PPL cpuld provide the frequency/phase reference and electronics can respond within nanoseconds to any frequency/phase deviation. Perhaps it's the large instantaneous currents that high voltage triacs simply cannot deal with?
That is the question, to be able to respond in nanoseconds, it takes some pretty sophisticated processing power. Is that kind of power available in a field hardened device? Regardless of the processing speed, there is dwell time in gating the thyristors (or other silicon device). 10 milliseconds per half cycle on a 50Hz system.
Yeah. There are alternatives. Key West uses a clutch based system which allows their backup generators to serve as syncons when their importing power from the mainland,
which is most of the time. But in this case, lack of inertia was not an issue, and there is no fly wheel.
The economics of conversion versus purpose built must not be overwhelming in favor of conversion, or ABB would not be selling independent syncons. It's one thing to convert a generator to a motor for voltage support and reactive power balancing. But the turbine has most of the inertia, and a lot of windage. So I think you need to pull out the turbine
and put in a flywheel if its inertia youre after, but not sure about this.
To get batteries to simulate grid inertia you are going to need some really big and robust inverters and a charged up battery. My superficial reading of the literature is that that technology is not quite there, but once again I could be wrong. One thing is certain. The UK grid operator thinks that syncon/flywheel is most economic. They and the Irish grid have several other syncon/flywheel projects in the works.
There are lots of hidden zombie assets on the power system. It is common to mothball old plants when retiring them because it is cheaper than tearing them down.
With the right price signals and/or policy guidance, many of them could be partially or fully returned to service.
I think we mostly get new syncons instead of syncon conversions because transmission planners are responsible for inertia studies and they don't know what old generators are out there or what condition they are in.
Yeah, you can probably buy most mothballed plants for next to nothing or maybe get paid to take them. They're liabilities.
The hard part is convincing someone to pay you to provide inertia. I think they are mostly being studied in the same way that transmission planners study VAR control mechanisms - the solutions are just added to the transmission investment portfolio, third party solutions are not generally investigated.
Not really: they were designed specifically to provide power pulses to JET for fusioin experiments.
The peak of consumption here is during a 300s JET pulse – where over 300 MW of electrical power is pulled from the grid, and up to 400 MW is supplied from two large flywheels located in Culham, U.K. However, this is only for 30 seconds, every 20-30 minutes.
Although it is nitpicky of me, the proper unit for giga-Watt seconds as I'm sure the author knows is GWs, not GW-s. A more correct alternative to GW-s would be GW⋅s.
I appreciate and like your analysis - $500/kiloWatt on a name-plate 1000MW solar farm costing $1000/kiloWatt but averaging 300MW per day?
Well I've seen a quoted 500MWe Thorium- Sodium-Berilium Fluoride salt MSR built in a Korean shipyard, (like your mammoth double-hulled bulk-carrier); not only completely fitted out with two 250MWe walk-away safe nuclear pots, heat exchangers and 2×250MWe turbines; but also a standby cold start generator; towed barge-like to the Java sea, completely fieted out except for fuel for about $1000/kW.
The only cost to the customer were the site preparation and linking to their local grid.
It was to have the potential 24/7 base load service with a pre-profit levelized cost of less than $30/Megawatt.hour (<$0.03 per unit kiloWatt.hour) but also capable of automatic load-following ability to compensate for Mickey Mouse, short life, un-recyclible "renewable" solar and wind farms.
So one of these 500MWe MSRs would produce more sustainable power at ⅔ the total Capital cost of your quoted example and also have a life 3× longer?..
It is not changing the value of this excellent article dealing with fast transients and reacting to them, but 30% capacity factor solar is an extremely generous possibility. Real solar PVs in Germany or UK show values closer to 15% and that number alone hides the problem that most of said generation happens when less power is needed(warm months).
Such a synchronous condenser can only supply power during the time it coasts down from 50 Hz speed to 49.95 Hz speed, then it cuts out. That's just 0.001 of the rotating mass energy. If a 1 GW turbine is rated to deliver 4 GW-sec of energy during the slowdown, then the total rotating energy is 4000 GW-sec, over 1 GW-hour! Is that right?
I believe one cause of the Chernobyl accident was the experiment to confirm the "rotating inertia" energy of the turbine-generator could power the safety systems in a grid disconnect.
Chernobyl accident or the botched experiment that caused it was slightly different. Grid is lost, the reactor scrammed and emergency cooling and critical functions must be supported before the diesel generators start and take the load. So the coasting of the turbine and generator was meant to support much lower power demand for critical time period(not supplying full output power and frequency of the working reactor to the grid). Similar concept, but different timescales and power levels included. If I understand correctly (would frequency drop matter for this application?)
Thanks for an informative article. It is sorely lacking as engineers are trained to delay publication of technical items until they confirm their findings. However I am comfortable in saying. Many W&S advocates (& journalists) understate the increasing vulnerability of large grids to natural and technical induced instabilities with increasing W&S penetration. They also overstate the "potential" system inertia contribution of unproven technologies to fully substitute for heavy mass of classic generators. Engineers could drive a massive vessel through this rift in technical discourse. Keep up the good work.
Only if it has grid forming inverters. Standard inverters are grid following. My understanding is that getting batteries to ape inertia is still developmental, but I'm not sure about this. There'a difference between smearing out a step change, and rapidly ramping up and down which batteries are good at.
They are common in Australia and Australia only, hardly a resounding recomedation. They exist there because of government regulation. It remains to be seen if NERC will adopt that same enthusiasm in the new IBR standards that will roll out soon.
The AEMC rules exist because they address the exact concern raised in this blog post. The NERC is about a decade behind the leaders in this field such as Australia and China simply due to smaller renewable penetration and more dispatch-able spinning reserve. What they choose to do is essentially irrelevant when it comes to good policy, just like how we don't look Tonga for space technology. On the other hand, Texas is already installing FFR as fast as possible.
See also the European Network of Transmission System Operators for Electricity:
Commission Regulation (EU) 2016/631 (the “RfG” itself)
• Article 21 (2)(b) defines “synthetic inertia” and makes it a mandatory feature of all new power‑park modules (i.e. inverter‑connected generators)
Article 21
Requirements for type C power park modules
1. Type C power park modules shall fulfil the requirements listed in Articles 13, except for Article 13(2)(b) and (6), Article 14, except for Article 14(2), Article 15 and Article 20, except for Article 20(2)(a), unless referred to otherwise in point (v) of paragraph 3(d).
2. Type C power park modules shall fulfil the following additional requirements in relation to frequency stability:
(a) the relevant TSO shall have the right to specify that power park modules be capable of providing synthetic inertia during very fast frequency deviations;
(b) the operating principle of control systems installed to provide synthetic inertia and the associated performance parameters shall be specified by the relevant TSO.
Despite Buystander's advice, Electranet the South Australia grid operator has installed four large syncon/flywheels: two at Davenport, and two at Robertstown. AEMO says the inertia from these devices allows it to lift its limit on asynchronous generators.
I don't know the specs, but from the pictures these syncons are at least as large as the Lister Drive units. Have not been able to find any cost numbers.
The link does not inspire confidence. Just 60 pages of bureaucratic handwaving.
My tentative layman's take is the jury is still out on whether inverters and batteries alone can provide the "inertia" required for a stable 100% W/S grid. Perhaps Australia will do that experiment for us.
If your argument is that you didn't understand it or couldn't be bothered to understand it, I accept that. South Australia has demonstrated islanding multiple times with batteries holding up just fine.
I believe the key difference lies in the amount of energy that can be delivered to the grid over a short period of time during a frequency event. If we consider a scenario where the grid frequency drops to 49 Hz over 10 seconds:
A 100 MW battery delivering full power for that duration would supply only about 0.28 MWh of energy to the grid.
A 500 MW gas turbine with an inertia constant (H) of 2 seconds has 1,000 MJ (≈ 278 kWh) of stored kinetic energy per MW, so a total of 139 MWh of rotational energy. A 1 Hz drop (from 50 to 49 Hz) corresponds to about a 2% reduction in speed, which means it would release roughly 4–5% of its kinetic energy, i.e., ~5.6 MWh, which is a power of over 2GW over those 10 seconds.
For the purpose of arresting the frequency drop, I believe it's this energy availability over the critical initial seconds that matters most.
Your numbers for the gas turbine are out by a factor of 500. A 500MW generator can at most deliver a small multiple of 500MW for 10s, a 500MW battery likewise can deliver a small multiple of 500MW for 10s (due to capacitors, thermal limits etc). However, a battery can continue to firm the grid until it goes flat, rotating mass can only support the grid until it slows to 47.5Hz or so and trips out.
Thanks. It's great to have input from somebody who actually has experience in the area. Everybody should read Kilovar's post. But couldn't a fast enough acting inverter effectively ape inertia?
With respect to your battery buffered wind/solar grid, I think you are still going to need fossil backup. The first question I have for any all wind/solar proposal is what's the longest dunkelflaute your grid can handle? I don't think your post answers that question.
One lesson I take from the Spanish blackout is the value of resilience. Near 100% gas back up capacity is not that expensive, at least in the US. In the REPOWER plan, that backup is provided by local coops. If the whole grid goes to pieces, the coop just disconnects, fires up its generators, and turns the lights back on. This is effectively the system we had in the Florida Keys. It proved invaluable when a storm takes out the SE FL grid. And thanks to the Turkey Point nukes which were built for less than a $1000/kW in today's money, we had some of the cheapest electricity on the country.
I love when I get put on the spot, no seriously I do. It means you're interested I was pretty frank in my write up that battery backed renewables were still only good for supplemental power, not as base load energy. So in that theoretical renewable only backed grid, that's easy, four hours. That is the limit of present battery technology. You can make it last longer by using less, but ... The green folks are talking about eight hour batteries, but they are like carbon storage, hasn't been invented yet
Inverter as inertia, even really fast, not IMHO. Because rotating mass resists change in speed BEFORE it happens. A grid forming inverter by design has to react AFTER the speed starts to change.
this does not make sense, an 8 hour battery is just a 4 hour battery with the inverter limited to 50%. Typically you want a 4 hour battery so you can capture the peak solar generation, as the 4 hours refers to both the discharge rate and the charging rate, and for diurnal cycle the charging rate is the limiting factor.
Critical control of input frequency in big machines like Synchrotrons is achieved with large flywheels that separate the input lower quality from the final input to the "big machine" - in this case an electron accelerator with buzzing around at 90% of the speed light.
This helpful machine is an inertial spinning mass that is impervious to "upstream fluctuations".
Pumped hydro, until the top bucket is empty, can perform the same function on a larger scale.
However, these solutions add to the primary cosy of supply, and not just a little bit.
If ecological accounting is taken into consideration greenfield hydro batteries can be really challenging to justify. And hydro "batteries" backing up 15 to 20% greenfield solar + grid extension is a non-starter in pristine landscapes.
Wind turbines (25-38% capacity factors) are worse and "environmental exemptions" common in Australia make the ecological devastation inevitable (birds, bats, insects, ground dwelling species that burrow......)
Nuclear + fast response gas turbines and custodianship of legacy assets with logical life extensions makes much more sense.
Highly local concentrated solar combined with overnight molten salt generation is immature but worthy of more work.
Interesting idea but you will have losses at every step, and unless you expand your wind/solar capacity, you will shorten the design dunkelflaute, the longest wind/solar lull your grid can handle. It's kind of symbolic of the contortions a nearly all W/S grid forces us into. All this stuff would come for free if we had cheap nuclear. Which reminds me to remind everybody to get the How We Can Make Nuclear Cheap Again book.
I was addressing the "resiliency"/inertial issue. That is not the same as "intermittency"
I agree that nuclear helps on both issues and that secularly nuclear has not received the same levels of support/subsidy that wind and solar have and consequently has more scope for future movement along cost curves.
I saw a grid costs projection that had 6% of the cost of a high VRE grid as synchronous condensers, that might assume batteries have a significant stabilizing capability as well
I don't remember which it was. ChatGPT says AEMO in Australia's ISP report have about 2% of the cost of an 82% renewable grid as condensers. It says higher penetration estimates assume grid forming inverter technology supercedes it at a lower cost
A question about pony motors being used with flywheels and synchronous condensers. I assume they run on gasoline? I assume this is inconsequential? Doesn't gasoline go stale, so if this is run occasionally, maybe gasoline needs to be discarded? Maybe at some point if there is more and more of these systems that need to be deployed with increasing wind and solar deployments, this may not be so inconsequential?
The pony motor is electric. It's not for black starts. It's a variable speed induction motor. It's job is to get the syncon up to grid frequency speed at which point that synchronous machine can be locked into the grid by connectignto the grid when the phase angle is right. If you want to know more, you are going to have to find a real electrical engineer.
Once the syncon is on the grid the pony motor does nothing.
Batteries can do this, along with their other functions. ISOs should require them if you are interconnecting non-dispatchable generation. I say this is a renewable project developer.
I thinnk that you are spot-on in your analysis, btw. I learned about this in 2002 when I started out as a renewable developer. I was schooled in it by a very kind utility engineer.
California has been working for 15 years to develop standards to make solar inverters play as nicely as possible. Part of the problem is that their first reaction when there is a frequency event is to trip off. When the frequency event is caused by a sudden drop-off of generation (which I believe happened in Iberia), having the solar drop-off, when it is contributing a huge amount of the total generation, could have been truly catastrophic.
Tripping off or reducing output due to over frequency is pretty easy. That is what the plants in Spain should have done - and maybe they did do it.
If the system was importing power during the separation, they would be in under frequency. Getting solar and wind plants to contribute additional power to solve under frequency is much harder. They basically have to always operate at 80% of their capabilities so they can ramp up to 100% during an under frequency event.
Paying for 100% of a plant and then only using 80% of it is expensive.
I have often seen this claim made, but I am not buying it. No matter how good your grid forming inverter is, it still has to depend on PID logic. A PID loop senses error, then reacts to correct it, that takes time. Time is something you don't have during a disturbance. So the inverters can mimic a rotating machine, but they will never do what a rotating machine does.
I too haven't seen the claim verified however it should be possible since an accurate PPL cpuld provide the frequency/phase reference and electronics can respond within nanoseconds to any frequency/phase deviation. Perhaps it's the large instantaneous currents that high voltage triacs simply cannot deal with?
That is the question, to be able to respond in nanoseconds, it takes some pretty sophisticated processing power. Is that kind of power available in a field hardened device? Regardless of the processing speed, there is dwell time in gating the thyristors (or other silicon device). 10 milliseconds per half cycle on a 50Hz system.
Existing generators at plants slated for retirement can be converted into syncons for lower cost than new builds.
Case study:
https://www.google.com/url?sa=t&source=web&rct=j&opi=89978449&url=https://www.eaton.com/content/dam/eaton/services/eess/eess-documents/eaton-sync-condenser-conversion-iou-case-study-cs027005en.pdf&ved=2ahUKEwjk7cW9zICNAxUjIjQIHXWGGncQFnoECB8QAQ&usg=AOvVaw2iJ5cG3x3lCXmIsoy1iEqf
Yeah. There are alternatives. Key West uses a clutch based system which allows their backup generators to serve as syncons when their importing power from the mainland,
which is most of the time. But in this case, lack of inertia was not an issue, and there is no fly wheel.
The economics of conversion versus purpose built must not be overwhelming in favor of conversion, or ABB would not be selling independent syncons. It's one thing to convert a generator to a motor for voltage support and reactive power balancing. But the turbine has most of the inertia, and a lot of windage. So I think you need to pull out the turbine
and put in a flywheel if its inertia youre after, but not sure about this.
To get batteries to simulate grid inertia you are going to need some really big and robust inverters and a charged up battery. My superficial reading of the literature is that that technology is not quite there, but once again I could be wrong. One thing is certain. The UK grid operator thinks that syncon/flywheel is most economic. They and the Irish grid have several other syncon/flywheel projects in the works.
I was surprised when the flywheels at JET in Culham were mothballed, they seemed like an oven-ready syncon
Those flywheels still exist, so they may yet be repurposed for grid stability or other research projects.
There are lots of hidden zombie assets on the power system. It is common to mothball old plants when retiring them because it is cheaper than tearing them down.
With the right price signals and/or policy guidance, many of them could be partially or fully returned to service.
I think we mostly get new syncons instead of syncon conversions because transmission planners are responsible for inertia studies and they don't know what old generators are out there or what condition they are in.
Sounds like an opportunity for engineers-cum-private equity analysts to make a big bet
Yeah, you can probably buy most mothballed plants for next to nothing or maybe get paid to take them. They're liabilities.
The hard part is convincing someone to pay you to provide inertia. I think they are mostly being studied in the same way that transmission planners study VAR control mechanisms - the solutions are just added to the transmission investment portfolio, third party solutions are not generally investigated.
Not really: they were designed specifically to provide power pulses to JET for fusioin experiments.
The peak of consumption here is during a 300s JET pulse – where over 300 MW of electrical power is pulled from the grid, and up to 400 MW is supplied from two large flywheels located in Culham, U.K. However, this is only for 30 seconds, every 20-30 minutes.
Sounds perfect
Is it possible to use the inertia in load as it is in generation? ie kinetic energy in industrial processes?
If there are any
Although it is nitpicky of me, the proper unit for giga-Watt seconds as I'm sure the author knows is GWs, not GW-s. A more correct alternative to GW-s would be GW⋅s.
I appreciate and like your analysis - $500/kiloWatt on a name-plate 1000MW solar farm costing $1000/kiloWatt but averaging 300MW per day?
Well I've seen a quoted 500MWe Thorium- Sodium-Berilium Fluoride salt MSR built in a Korean shipyard, (like your mammoth double-hulled bulk-carrier); not only completely fitted out with two 250MWe walk-away safe nuclear pots, heat exchangers and 2×250MWe turbines; but also a standby cold start generator; towed barge-like to the Java sea, completely fieted out except for fuel for about $1000/kW.
The only cost to the customer were the site preparation and linking to their local grid.
It was to have the potential 24/7 base load service with a pre-profit levelized cost of less than $30/Megawatt.hour (<$0.03 per unit kiloWatt.hour) but also capable of automatic load-following ability to compensate for Mickey Mouse, short life, un-recyclible "renewable" solar and wind farms.
So one of these 500MWe MSRs would produce more sustainable power at ⅔ the total Capital cost of your quoted example and also have a life 3× longer?..
Cheers
JC
It is not changing the value of this excellent article dealing with fast transients and reacting to them, but 30% capacity factor solar is an extremely generous possibility. Real solar PVs in Germany or UK show values closer to 15% and that number alone hides the problem that most of said generation happens when less power is needed(warm months).
Such a synchronous condenser can only supply power during the time it coasts down from 50 Hz speed to 49.95 Hz speed, then it cuts out. That's just 0.001 of the rotating mass energy. If a 1 GW turbine is rated to deliver 4 GW-sec of energy during the slowdown, then the total rotating energy is 4000 GW-sec, over 1 GW-hour! Is that right?
I believe one cause of the Chernobyl accident was the experiment to confirm the "rotating inertia" energy of the turbine-generator could power the safety systems in a grid disconnect.
Chernobyl accident or the botched experiment that caused it was slightly different. Grid is lost, the reactor scrammed and emergency cooling and critical functions must be supported before the diesel generators start and take the load. So the coasting of the turbine and generator was meant to support much lower power demand for critical time period(not supplying full output power and frequency of the working reactor to the grid). Similar concept, but different timescales and power levels included. If I understand correctly (would frequency drop matter for this application?)
Thanks for an informative article. It is sorely lacking as engineers are trained to delay publication of technical items until they confirm their findings. However I am comfortable in saying. Many W&S advocates (& journalists) understate the increasing vulnerability of large grids to natural and technical induced instabilities with increasing W&S penetration. They also overstate the "potential" system inertia contribution of unproven technologies to fully substitute for heavy mass of classic generators. Engineers could drive a massive vessel through this rift in technical discourse. Keep up the good work.
On the other hand a typical 100MW BESS at just $30M can provide 1.4 TW.s of inertia.
Only if it has grid forming inverters. Standard inverters are grid following. My understanding is that getting batteries to ape inertia is still developmental, but I'm not sure about this. There'a difference between smearing out a step change, and rapidly ramping up and down which batteries are good at.
Synthetic inertia inverters are fairly commonplace in grid BESS now and provide exactly the inertia you are looking for. for example https://www.aemc.gov.au/rule-changes/clarifying-mandatory-primary-frequency-response-obligations-bidirectional-plant means that all new BESS systems have fast-frequency-response (where the inverter simulates the phasor/rotation of the grid/generator) which provides grid-forming capacity.
They are common in Australia and Australia only, hardly a resounding recomedation. They exist there because of government regulation. It remains to be seen if NERC will adopt that same enthusiasm in the new IBR standards that will roll out soon.
The AEMC rules exist because they address the exact concern raised in this blog post. The NERC is about a decade behind the leaders in this field such as Australia and China simply due to smaller renewable penetration and more dispatch-able spinning reserve. What they choose to do is essentially irrelevant when it comes to good policy, just like how we don't look Tonga for space technology. On the other hand, Texas is already installing FFR as fast as possible.
See also the European Network of Transmission System Operators for Electricity:
Commission Regulation (EU) 2016/631 (the “RfG” itself)
• Article 21 (2)(b) defines “synthetic inertia” and makes it a mandatory feature of all new power‑park modules (i.e. inverter‑connected generators)
Article 21
Requirements for type C power park modules
1. Type C power park modules shall fulfil the requirements listed in Articles 13, except for Article 13(2)(b) and (6), Article 14, except for Article 14(2), Article 15 and Article 20, except for Article 20(2)(a), unless referred to otherwise in point (v) of paragraph 3(d).
2. Type C power park modules shall fulfil the following additional requirements in relation to frequency stability:
(a) the relevant TSO shall have the right to specify that power park modules be capable of providing synthetic inertia during very fast frequency deviations;
(b) the operating principle of control systems installed to provide synthetic inertia and the associated performance parameters shall be specified by the relevant TSO.
Well, I guess we know where you stand, all comments are welcome.
Despite Buystander's advice, Electranet the South Australia grid operator has installed four large syncon/flywheels: two at Davenport, and two at Robertstown. AEMO says the inertia from these devices allows it to lift its limit on asynchronous generators.
I don't know the specs, but from the pictures these syncons are at least as large as the Lister Drive units. Have not been able to find any cost numbers.
Buy,
The link does not inspire confidence. Just 60 pages of bureaucratic handwaving.
My tentative layman's take is the jury is still out on whether inverters and batteries alone can provide the "inertia" required for a stable 100% W/S grid. Perhaps Australia will do that experiment for us.
If your argument is that you didn't understand it or couldn't be bothered to understand it, I accept that. South Australia has demonstrated islanding multiple times with batteries holding up just fine.
I understood the document just fine. There were no technical specs in its. Just legalese and advertorial.
What upsets did SA ride thru? How little synchronous inertia was on line?
I believe the key difference lies in the amount of energy that can be delivered to the grid over a short period of time during a frequency event. If we consider a scenario where the grid frequency drops to 49 Hz over 10 seconds:
A 100 MW battery delivering full power for that duration would supply only about 0.28 MWh of energy to the grid.
A 500 MW gas turbine with an inertia constant (H) of 2 seconds has 1,000 MJ (≈ 278 kWh) of stored kinetic energy per MW, so a total of 139 MWh of rotational energy. A 1 Hz drop (from 50 to 49 Hz) corresponds to about a 2% reduction in speed, which means it would release roughly 4–5% of its kinetic energy, i.e., ~5.6 MWh, which is a power of over 2GW over those 10 seconds.
For the purpose of arresting the frequency drop, I believe it's this energy availability over the critical initial seconds that matters most.
Your numbers for the gas turbine are out by a factor of 500. A 500MW generator can at most deliver a small multiple of 500MW for 10s, a 500MW battery likewise can deliver a small multiple of 500MW for 10s (due to capacitors, thermal limits etc). However, a battery can continue to firm the grid until it goes flat, rotating mass can only support the grid until it slows to 47.5Hz or so and trips out.
I think you're right. I had a large factor out. It's not 2GW, but rather 4MW I think.
I wonder if the US navy could build SMRs for the civilian market?
Lukas,
This might be the only way of making US nuclear even more expensive than it is. See
https://jackdevanney.substack.com/p/a-tale-of-two-ships
Jack take a look at my Substack https://open.substack.com/pub/kilovar1959/p/making-wind-and-solar-work-what-would?utm_source=share&utm_medium=android&r=23kggy
Kilo,
Thanks. It's great to have input from somebody who actually has experience in the area. Everybody should read Kilovar's post. But couldn't a fast enough acting inverter effectively ape inertia?
With respect to your battery buffered wind/solar grid, I think you are still going to need fossil backup. The first question I have for any all wind/solar proposal is what's the longest dunkelflaute your grid can handle? I don't think your post answers that question.
One lesson I take from the Spanish blackout is the value of resilience. Near 100% gas back up capacity is not that expensive, at least in the US. In the REPOWER plan, that backup is provided by local coops. If the whole grid goes to pieces, the coop just disconnects, fires up its generators, and turns the lights back on. This is effectively the system we had in the Florida Keys. It proved invaluable when a storm takes out the SE FL grid. And thanks to the Turkey Point nukes which were built for less than a $1000/kW in today's money, we had some of the cheapest electricity on the country.
Pls let us know what you think about this idea. https://gordianknotbook.com/download/the-repower-plan
I love when I get put on the spot, no seriously I do. It means you're interested I was pretty frank in my write up that battery backed renewables were still only good for supplemental power, not as base load energy. So in that theoretical renewable only backed grid, that's easy, four hours. That is the limit of present battery technology. You can make it last longer by using less, but ... The green folks are talking about eight hour batteries, but they are like carbon storage, hasn't been invented yet
Inverter as inertia, even really fast, not IMHO. Because rotating mass resists change in speed BEFORE it happens. A grid forming inverter by design has to react AFTER the speed starts to change.
this does not make sense, an 8 hour battery is just a 4 hour battery with the inverter limited to 50%. Typically you want a 4 hour battery so you can capture the peak solar generation, as the 4 hours refers to both the discharge rate and the charging rate, and for diurnal cycle the charging rate is the limiting factor.
https://www.energy.gov/energy-storage-grand-challenge/energy-storage-grand-challenge?nrg_redirect=353266
PS, take a look, I have a whole series of stuff to read. 😁
Critical control of input frequency in big machines like Synchrotrons is achieved with large flywheels that separate the input lower quality from the final input to the "big machine" - in this case an electron accelerator with buzzing around at 90% of the speed light.
This helpful machine is an inertial spinning mass that is impervious to "upstream fluctuations".
Pumped hydro, until the top bucket is empty, can perform the same function on a larger scale.
However, these solutions add to the primary cosy of supply, and not just a little bit.
If ecological accounting is taken into consideration greenfield hydro batteries can be really challenging to justify. And hydro "batteries" backing up 15 to 20% greenfield solar + grid extension is a non-starter in pristine landscapes.
Wind turbines (25-38% capacity factors) are worse and "environmental exemptions" common in Australia make the ecological devastation inevitable (birds, bats, insects, ground dwelling species that burrow......)
Nuclear + fast response gas turbines and custodianship of legacy assets with logical life extensions makes much more sense.
Highly local concentrated solar combined with overnight molten salt generation is immature but worthy of more work.
Why not inertial batteries? Route low MC power from solar / wind at peak times into flywheels.
Thomas,
Interesting idea but you will have losses at every step, and unless you expand your wind/solar capacity, you will shorten the design dunkelflaute, the longest wind/solar lull your grid can handle. It's kind of symbolic of the contortions a nearly all W/S grid forces us into. All this stuff would come for free if we had cheap nuclear. Which reminds me to remind everybody to get the How We Can Make Nuclear Cheap Again book.
I was addressing the "resiliency"/inertial issue. That is not the same as "intermittency"
I agree that nuclear helps on both issues and that secularly nuclear has not received the same levels of support/subsidy that wind and solar have and consequently has more scope for future movement along cost curves.
I saw a grid costs projection that had 6% of the cost of a high VRE grid as synchronous condensers, that might assume batteries have a significant stabilizing capability as well
Smope,
Do you have the reference?
I don't remember which it was. ChatGPT says AEMO in Australia's ISP report have about 2% of the cost of an 82% renewable grid as condensers. It says higher penetration estimates assume grid forming inverter technology supercedes it at a lower cost
Nice description of the inertia problem by Raul Bajo Buenestado
https://www.bakerinstitute.org/research/iberian-peninsula-blackout-causes-consequences-and-challenges-ahead
One of the issues he highlights is coordinating multiple grid forming inverters.
A question about pony motors being used with flywheels and synchronous condensers. I assume they run on gasoline? I assume this is inconsequential? Doesn't gasoline go stale, so if this is run occasionally, maybe gasoline needs to be discarded? Maybe at some point if there is more and more of these systems that need to be deployed with increasing wind and solar deployments, this may not be so inconsequential?
Todd,
The pony motor is electric. It's not for black starts. It's a variable speed induction motor. It's job is to get the syncon up to grid frequency speed at which point that synchronous machine can be locked into the grid by connectignto the grid when the phase angle is right. If you want to know more, you are going to have to find a real electrical engineer.
Once the syncon is on the grid the pony motor does nothing.