In the past, the question of grid/load defection for residential customers was limited to solar PV with battery backup. In the studies I and others conducted, the economics mostly pan out except for some hours or days during the year. Depending on the region of the country, the hours that a combination of solar and battery was capable of providing the energy needs of "typical" homes were 93-97% of the hours of the year. The issue is solar dunkelflaute, to borrow an expression from the wind turbine folks. There are periods of insufficient solar insolation to both provide energy and maintain batteries sufficiently charged to carry homes through nights and other periods of solar deficiency. The authors of the paper being discussed relieve this constraint by permitting fossil-fueled backup generation to relieve that constraint. The authors' use of backup generation doesn't affect the validity of their findings, but prior research only sought renewable alternatives as the basis for grid/load defection. It wouldn't be surprising to find other areas that are also susceptible to residential grid/defection with a solar PV/battery/backup generation combination available to them.
One scenario I've been considering more is the neighborhood/housing development that defects into a solar+storage microgrid with natural gas backup (like what Enchanted Rock provides today to its commercial customers). Better than diesel, not zero carbon but the derivative has the correct sign, more local/community. May have a lot to recommend it.
I think it's preferable to use fuels that need to be transported to the site, fuel oil, propane, etc., rather than relying on the natural gas distribution system. In the long run, the gas distribution system's fixed costs will be spread over fewer and fewer therms, so its unit costs will increase dramatically. In addition, I believe the relative GHG impacts per therm, including losses on the distribution and transportation systems, etc., are smaller, so they represent a more climate-friendly alternative. Microgrids with centralized batteries might solve the backup problem if they have enough battery capacity, possibly siting them at substations, for example, or at some aggregation point in a neighborhood, like school or public building parking lots.
Agree completely on the aggregation point strategy.
Economics is not predictive so don’t put any weight on this, but my sense is that the gas sunsetting will occur over two decades, which is time enough to recoup investments in gas backup where the gas infrastructure already exists (sorry, New England!).
I'm less optimistic than you about gas distribution costs. The average cost per mile for replacement pipes is about $3M/mile, $5M in NE. The typical depreciation period for pipes is 40 years. You'd need to double the depreciation rate and assume that neither new mains are built as a result of expansion nor any additional pipeline constructed for replacement purposes over the twenty years. Natural gas distribution companies will vigorously continue to expand and replace pipes as long as the states don't have a stranded asset plan in place for them. So far as I know, none of them do. I think the lack of such plans is possibly the single most significant political/regulatory roadblock to electrification.
It's the mirror of the electrical grid death spiral. The NG distribution system will become a fixed capacity cost, and be the vast majority of the gas bill. It's the only way they can halt the slide.
So I agree liquid / liquefied fuels make way more sense as backups.
When I visited Mumbai, I saw water trucks. They would come and refill the water tank on the roof of your building because of water grid defection. Similarly, imagine a solar dunkelflaute being resolved by a mobile fossil fueled generator stopping by to refill the storage.
Hi Lynne, totally agree that grid contestability does the the things you mention - but it also increases the uncertainty of repayment for fixed assets deployed in the past (no doubt part of Barclay's concerns). So the cost of defection has to include that buy-out, no?
Looking forward, how should a classic IOU structure the financing of the local (or long distance) infrastructure capex?
Hi Michael, yep, financial implications for the IOU for sure. I have a whole line of thought I wonder about along the lines of “what happens when a long-lived capital-intensive industry confronts innovation involving shorter-lived capital?”. Depreciation mismatch + regulatory “used and useful” requirements = pressures on profitability, dividend, and share value.
I am not a financial economist so I have no depth here, but in terms of the economics I think about instruments for laying off that risk on other parties — securitization (which this industry knows better now thanks to coal retirements), etc. But I think an imaginative IOU (and its regulators) can do better than that by leaning in to the innovation, saying yes, grid architecture is changing and must change, and that means more of our capex will be on standard, modular, shorter-lived assets. In total their capex may not fall because in a more decentralized architecture you need a lot more of that stuff, so their rate base may not fall, but decentralizing their architecture provides a physical hedge by shifting their capex toward shorter-lived investments while also enabling them to compete in a more contestable setting.
Hmmmm … I’ve never thought about it that clearly before. Thanks for the prompt! And if I’m off base do let me know :-).
For new residential communities, Why can't we simply install enough PV and a geothermal network on-site and build in enough batteries using heat pumps to provide enough to operate the community without the electric grid or needed fossil fuels as a backup?
"enough" is the magic word, though. Go read the newspapers from 100 years ago during the winter when the trains couldn't get through with the coal, and the coal supplies started to run low. Same problem, no solution.
Question: Can't a city, state, etc simply make defection unlawful? When we had our solar panels installed we were told (and I believe had to sign as a requirement of installation) they must be linked into the local power grid, any excess energy we produced could not be stored in a local battery, and went to the power company, by law. What's to prevent that approach by the industry? Enforcement would of course make it difficult to implement, especially in rural areas I would think. But in a city/suburb area? They could simply make it illegal in order to protect the public utility/public distribution of energy, couldn't they?
A utility death spiral / major grid defection is fatal to deep decarbonisation both for all the generators (and far more fueled heating) it implies, and for the costs that will land on the users who can't defect (high density housing, industrial, some commercial / institutional, even renters...
I don't agree that it's terrifying. We'd be better off if the dynamics of contestability were allowed to play out, if the threat of grid defection induced regulated utilities to do a better, more timely job of offering their customers products and services that they value so that they don't defect. And, if the costs get low enough/benefits get high enough, then defection is the economically right thing to do.
The big challenge, and where I agree with your comment, is that wires costs are an increasing share of the monthly bill, so partial defection would concentrate those wires costs on the incumbent's remaining customers. Which itself raises another important question: wires innovations to reduce wires system costs, and the utility incentives to adopt them.
That renters can't defect is becoming a rapidly outdated assumption. 2kW of plug and play solar, with a 1.6kWh battery can be bought for less than €1,500. How would the utility even know?
Sadly that’s still illegal in the US. Stories here abound of “hey German renters can buy solar fencing for their apartment balcony why can’t we do that?”.
If you need any amount of grid power, you are going to end out being stuck with the full 'wires' bill, which will likley end out being most of the cost...
You "can't" install solar on your roof in Elmira NY without getting a permit for it. You also "can't" hook them up to a grid-tie inverter without all kinds of permission.
In the past, the question of grid/load defection for residential customers was limited to solar PV with battery backup. In the studies I and others conducted, the economics mostly pan out except for some hours or days during the year. Depending on the region of the country, the hours that a combination of solar and battery was capable of providing the energy needs of "typical" homes were 93-97% of the hours of the year. The issue is solar dunkelflaute, to borrow an expression from the wind turbine folks. There are periods of insufficient solar insolation to both provide energy and maintain batteries sufficiently charged to carry homes through nights and other periods of solar deficiency. The authors of the paper being discussed relieve this constraint by permitting fossil-fueled backup generation to relieve that constraint. The authors' use of backup generation doesn't affect the validity of their findings, but prior research only sought renewable alternatives as the basis for grid/load defection. It wouldn't be surprising to find other areas that are also susceptible to residential grid/defection with a solar PV/battery/backup generation combination available to them.
One scenario I've been considering more is the neighborhood/housing development that defects into a solar+storage microgrid with natural gas backup (like what Enchanted Rock provides today to its commercial customers). Better than diesel, not zero carbon but the derivative has the correct sign, more local/community. May have a lot to recommend it.
Thank you for the like and comment.
I think it's preferable to use fuels that need to be transported to the site, fuel oil, propane, etc., rather than relying on the natural gas distribution system. In the long run, the gas distribution system's fixed costs will be spread over fewer and fewer therms, so its unit costs will increase dramatically. In addition, I believe the relative GHG impacts per therm, including losses on the distribution and transportation systems, etc., are smaller, so they represent a more climate-friendly alternative. Microgrids with centralized batteries might solve the backup problem if they have enough battery capacity, possibly siting them at substations, for example, or at some aggregation point in a neighborhood, like school or public building parking lots.
Agree completely on the aggregation point strategy.
Economics is not predictive so don’t put any weight on this, but my sense is that the gas sunsetting will occur over two decades, which is time enough to recoup investments in gas backup where the gas infrastructure already exists (sorry, New England!).
I'm less optimistic than you about gas distribution costs. The average cost per mile for replacement pipes is about $3M/mile, $5M in NE. The typical depreciation period for pipes is 40 years. You'd need to double the depreciation rate and assume that neither new mains are built as a result of expansion nor any additional pipeline constructed for replacement purposes over the twenty years. Natural gas distribution companies will vigorously continue to expand and replace pipes as long as the states don't have a stranded asset plan in place for them. So far as I know, none of them do. I think the lack of such plans is possibly the single most significant political/regulatory roadblock to electrification.
It's the mirror of the electrical grid death spiral. The NG distribution system will become a fixed capacity cost, and be the vast majority of the gas bill. It's the only way they can halt the slide.
So I agree liquid / liquefied fuels make way more sense as backups.
When I visited Mumbai, I saw water trucks. They would come and refill the water tank on the roof of your building because of water grid defection. Similarly, imagine a solar dunkelflaute being resolved by a mobile fossil fueled generator stopping by to refill the storage.
Hi Lynne, totally agree that grid contestability does the the things you mention - but it also increases the uncertainty of repayment for fixed assets deployed in the past (no doubt part of Barclay's concerns). So the cost of defection has to include that buy-out, no?
Looking forward, how should a classic IOU structure the financing of the local (or long distance) infrastructure capex?
Hi Michael, yep, financial implications for the IOU for sure. I have a whole line of thought I wonder about along the lines of “what happens when a long-lived capital-intensive industry confronts innovation involving shorter-lived capital?”. Depreciation mismatch + regulatory “used and useful” requirements = pressures on profitability, dividend, and share value.
I am not a financial economist so I have no depth here, but in terms of the economics I think about instruments for laying off that risk on other parties — securitization (which this industry knows better now thanks to coal retirements), etc. But I think an imaginative IOU (and its regulators) can do better than that by leaning in to the innovation, saying yes, grid architecture is changing and must change, and that means more of our capex will be on standard, modular, shorter-lived assets. In total their capex may not fall because in a more decentralized architecture you need a lot more of that stuff, so their rate base may not fall, but decentralizing their architecture provides a physical hedge by shifting their capex toward shorter-lived investments while also enabling them to compete in a more contestable setting.
Hmmmm … I’ve never thought about it that clearly before. Thanks for the prompt! And if I’m off base do let me know :-).
For new residential communities, Why can't we simply install enough PV and a geothermal network on-site and build in enough batteries using heat pumps to provide enough to operate the community without the electric grid or needed fossil fuels as a backup?
That's grid defection!
"enough" is the magic word, though. Go read the newspapers from 100 years ago during the winter when the trains couldn't get through with the coal, and the coal supplies started to run low. Same problem, no solution.
Question: Can't a city, state, etc simply make defection unlawful? When we had our solar panels installed we were told (and I believe had to sign as a requirement of installation) they must be linked into the local power grid, any excess energy we produced could not be stored in a local battery, and went to the power company, by law. What's to prevent that approach by the industry? Enforcement would of course make it difficult to implement, especially in rural areas I would think. But in a city/suburb area? They could simply make it illegal in order to protect the public utility/public distribution of energy, couldn't they?
"Just pass a law" always raises my hackles. What will the now-regulated parties do? Will they just buy legislators?
This is terrifying!
A utility death spiral / major grid defection is fatal to deep decarbonisation both for all the generators (and far more fueled heating) it implies, and for the costs that will land on the users who can't defect (high density housing, industrial, some commercial / institutional, even renters...
I don't agree that it's terrifying. We'd be better off if the dynamics of contestability were allowed to play out, if the threat of grid defection induced regulated utilities to do a better, more timely job of offering their customers products and services that they value so that they don't defect. And, if the costs get low enough/benefits get high enough, then defection is the economically right thing to do.
The big challenge, and where I agree with your comment, is that wires costs are an increasing share of the monthly bill, so partial defection would concentrate those wires costs on the incumbent's remaining customers. Which itself raises another important question: wires innovations to reduce wires system costs, and the utility incentives to adopt them.
That renters can't defect is becoming a rapidly outdated assumption. 2kW of plug and play solar, with a 1.6kWh battery can be bought for less than €1,500. How would the utility even know?
Sadly that’s still illegal in the US. Stories here abound of “hey German renters can buy solar fencing for their apartment balcony why can’t we do that?”.
Is that enough to actually defect though?
If you need any amount of grid power, you are going to end out being stuck with the full 'wires' bill, which will likley end out being most of the cost...
Even if you consume no watt-hours, here in NY you need to have a meter connected and pay the connection charge in order to have an occupancy permit.
You "can't" install solar on your roof in Elmira NY without getting a permit for it. You also "can't" hook them up to a grid-tie inverter without all kinds of permission.