Carrier announced a pilot project with an HVAC that includes a battery. Next week is Climate Week NYC; perhaps they will provide more details then. An appliance like this can adjust its consumption based on a price signal, like a Copper Stove.
There were many comments from stakeholders and intervenors after the PD; I skimmed through them but I have not had time to really look into them. I fully expected the Commission NOT to take the topic at its earliest opportunity (the 28th) but they did.
I will try to go through the PD to form an opinion on it. I remain positive about Dynamic Rates but the details matter.
Nice post - this the captures the PG&E Hourly Flex Pilot well - better in fact than the PG&E explainer website!
As you know, I very much agree with "a simple and easy to understand tariff may be better than more complex tariffs designed for a weaker grid."
Indeed, I believe this is the core reason why many people get solar even in a NEM 3 era: they want to not worry about erratic bills, and are willing to pay a premium to do so.
I predict we'll see an even more stable grid (ex wildfire prone area shutoffs) because of the deployment of batteries along the entire path: co-located with generation, at substations, in homes behind the meter, and increasing within appliances (i.e. new stoves). This is excellent as
1) it will continue to reduce spiky prices (would love to see an official analysis)
2) it will increase utilization of the fixed transmission/distribution network, leading to lower volumetric costs (i'd like to see more math on this)
3) behind-the-meter batteries in CAISO will be able to export (nascent now vs ERCOT's more mature solution)
Nothing as good as a good technology/manufacturing innovation (batteries!) to make complex economic engineering (marginal consumer pricing!) less relevant.
Thanks Garen. I appreciate our conversations that are helping us understand this topic.
Now that I've finished this post I want to go back to the Proposed Decision for R.22-07-005, which covers this topic. I had done a quick skim when it came a month ago; time to do a deeper analysis.
As far as I understand it the dynamic rates in the Flex Pilot do not include T&D and are only for generation so the actual rate the customer pays will vary much less dynamically.
This rate includes a dynamic hourly price for Generation and Distribution. Hourly Flex Pricing charges will also include Transmission costs and other costs specific to your current rate plan.
I feel like if this flex is only for generation, the actual rate won't vary that much since Transmission costs end up being the bulk (~2/3) of the rate. That's just based on my own analysis of my bills.
@Eduardo Pelegri-Llopart, it sounds like you are on this Flex Pilot rate plan now. What's your experience been like with this?
I am not in the plan because when it started it didn't allow NEM customers but a couple of friends are, and I believe I understand enough of it to try to answer your question.
I tried to separate Dynamic Rates from Dynamic Tariffs, and in this post I only focused on the Rates. With some caveats, the Rates make sense to me, but I still have issues with the Hourly Flex Tariff in how it is using the "Subscription" machinery. I am planning to tackle that area in a future post.
Going through your points...
Q1 - Do the "rates" only include Generation?
A1 - The pilot rates include CLD (Customer Load Distribution), MEG (Marginal Energy Cost) and MGCC (Marginal Generation Capacity Cost), so it is more than generation - see https://pelegri.substack.com/i/171847466/the-rate-formula
Q2 - The actual rate won't vary much
A2 - That is how I see it. The subscription machinery means the effective rate does not change much.
From my POV, the subscription is trying to give me an insurance over price variations in the Grid but the cost of that insurance is very high, and the CAISO grid in 2027 will be much more stable than the ERCOT grid was in 2021. I don't like the machinery.
Plus, I don't understand how the subscription works over years. In year 0, I have "the old plan". In year 1, with a battery, I will adjust my imports as I want it. Then, in year 2, I will see no savings from year 1, because my imports will be the same as in year 1.
Plus, try to explain how this work to anybody... its really hard.
Bottom line, I like dynamic rates; I don't like the tariff used in the pilot. I want to write a post on this, also pulling in some other tariffs out there.
Interesting article! I noticed in your dynamic price graph there’s also <negative values>... where they presumably could pay you to use more energy (and charge you to export).
It also seems like the pilot is based on the difference in your own usage from the same average time interval usage in the bill period 1 year ago (per their FAQ) - that is, they pay you to export <additional amounts> at the dynamic rate, or you pay for the <increased usage> at the dynamic rate. This seems way too complex, and less relevant to immediate grid pricing signals.
Also interesting - the dynamic pricing at GridStatus.io seems to account for generation pricing (with Energy) and transmission pricing (with Congestion and Loss), but not the cost of the last-mile “distribution” on the local circuit, which are downstream of the local priced nodes. With Distribution making up almost 1/2 of the current charge, ($0.21/kWh out of $0.44/kWh avg in PGE south bay), it’ll be interesting to see how much of this gets absorbed into the fixed-monthly base cost. (My numbers from https://www.ampmyhome.com/solar/lazard-solar-and-your-energy-costs/)
* GridStatus does not consider distribution; they only track prices at the p-node because (a) the GridStatus customers are energy developers that want to participate in the market, and (b) I have no idea how to compute that :)
* The HourFlex rate does include distribution congestion. That is only visible if you look at the detail spreadsheet, which requires calling into the API. I just checked at the Sample spreadsheet and I see some very high numbers, back in July 2024 - like $1.3!
I've added a TODO to my list: I'll use the API to extract the full data for my circuit. I want to understand how that distribution congestion is working.
* So, on the Base Services Charge... I have no idea how the volumetric rates will end up playing up in March 2026. The number used for Base Services Charge was a fairly arbitrary number. The argument at the time was "add some fixed cost to reduce everybody's rates", nothing more detailed than that. IIRC $24.15 came because SMUD has something similar.
* Exports when the rate is negative means the household will have to pay -- Precisely! That is one of the beauties of a Dynamic Rate with symmetric export/import. We will stop debating whether rooftop solar are free-loaders. If they export when the grid needs it, they get compensated. If they export when the grid does not need it, they are penalized. As it should be.
* The CalFUSE Subscription tariff -- Yeah, that tariff is not my favorite. That is why I am separating "rates" from "tariffs". There are many possible Dynamic Tariffs based on these Dynamic Rates. I'm delaying the discussion of tariffs to a separate post.
Carrier announced a pilot project with an HVAC that includes a battery. Next week is Climate Week NYC; perhaps they will provide more details then. An appliance like this can adjust its consumption based on a price signal, like a Copper Stove.
https://www.corporate.carrier.com/news/news-articles/202509_carrier-powering-future-energy.html
The Aug 28th CPUC voting meeting actually included the Proposed Decision (PD) on Guidance for Utility Dynamic Hourly Rates. PR is at https://docs.cpuc.ca.gov/PublishedDocs/Published/G000/M578/K060/578060595.PDF
There were many comments from stakeholders and intervenors after the PD; I skimmed through them but I have not had time to really look into them. I fully expected the Commission NOT to take the topic at its earliest opportunity (the 28th) but they did.
I will try to go through the PD to form an opinion on it. I remain positive about Dynamic Rates but the details matter.
Also see this article on Edward Cazalet original work on Dynamic Rates - https://pv-magazine-usa.com/2025/08/27/dynamic-pricing-is-superior-to-virtual-power-plants-says-dynamic-pricing-pioneer/
Nice post - this the captures the PG&E Hourly Flex Pilot well - better in fact than the PG&E explainer website!
As you know, I very much agree with "a simple and easy to understand tariff may be better than more complex tariffs designed for a weaker grid."
Indeed, I believe this is the core reason why many people get solar even in a NEM 3 era: they want to not worry about erratic bills, and are willing to pay a premium to do so.
I predict we'll see an even more stable grid (ex wildfire prone area shutoffs) because of the deployment of batteries along the entire path: co-located with generation, at substations, in homes behind the meter, and increasing within appliances (i.e. new stoves). This is excellent as
1) it will continue to reduce spiky prices (would love to see an official analysis)
2) it will increase utilization of the fixed transmission/distribution network, leading to lower volumetric costs (i'd like to see more math on this)
3) behind-the-meter batteries in CAISO will be able to export (nascent now vs ERCOT's more mature solution)
Nothing as good as a good technology/manufacturing innovation (batteries!) to make complex economic engineering (marginal consumer pricing!) less relevant.
Thanks Garen. I appreciate our conversations that are helping us understand this topic.
Now that I've finished this post I want to go back to the Proposed Decision for R.22-07-005, which covers this topic. I had done a quick skim when it came a month ago; time to do a deeper analysis.
As far as I understand it the dynamic rates in the Flex Pilot do not include T&D and are only for generation so the actual rate the customer pays will vary much less dynamically.
From the PG&E FAQ (https://www.pge.com/en/account/rate-plans/hourly-flex-pricing.html#accordion-cf73526c80-item-23185953eb):
This rate includes a dynamic hourly price for Generation and Distribution. Hourly Flex Pricing charges will also include Transmission costs and other costs specific to your current rate plan.
I feel like if this flex is only for generation, the actual rate won't vary that much since Transmission costs end up being the bulk (~2/3) of the rate. That's just based on my own analysis of my bills.
@Eduardo Pelegri-Llopart, it sounds like you are on this Flex Pilot rate plan now. What's your experience been like with this?
I am not in the plan because when it started it didn't allow NEM customers but a couple of friends are, and I believe I understand enough of it to try to answer your question.
I tried to separate Dynamic Rates from Dynamic Tariffs, and in this post I only focused on the Rates. With some caveats, the Rates make sense to me, but I still have issues with the Hourly Flex Tariff in how it is using the "Subscription" machinery. I am planning to tackle that area in a future post.
Going through your points...
Q1 - Do the "rates" only include Generation?
A1 - The pilot rates include CLD (Customer Load Distribution), MEG (Marginal Energy Cost) and MGCC (Marginal Generation Capacity Cost), so it is more than generation - see https://pelegri.substack.com/i/171847466/the-rate-formula
Q2 - The actual rate won't vary much
A2 - That is how I see it. The subscription machinery means the effective rate does not change much.
From my POV, the subscription is trying to give me an insurance over price variations in the Grid but the cost of that insurance is very high, and the CAISO grid in 2027 will be much more stable than the ERCOT grid was in 2021. I don't like the machinery.
Plus, I don't understand how the subscription works over years. In year 0, I have "the old plan". In year 1, with a battery, I will adjust my imports as I want it. Then, in year 2, I will see no savings from year 1, because my imports will be the same as in year 1.
Plus, try to explain how this work to anybody... its really hard.
Bottom line, I like dynamic rates; I don't like the tariff used in the pilot. I want to write a post on this, also pulling in some other tariffs out there.
Interesting article! I noticed in your dynamic price graph there’s also <negative values>... where they presumably could pay you to use more energy (and charge you to export).
It also seems like the pilot is based on the difference in your own usage from the same average time interval usage in the bill period 1 year ago (per their FAQ) - that is, they pay you to export <additional amounts> at the dynamic rate, or you pay for the <increased usage> at the dynamic rate. This seems way too complex, and less relevant to immediate grid pricing signals.
Also interesting - the dynamic pricing at GridStatus.io seems to account for generation pricing (with Energy) and transmission pricing (with Congestion and Loss), but not the cost of the last-mile “distribution” on the local circuit, which are downstream of the local priced nodes. With Distribution making up almost 1/2 of the current charge, ($0.21/kWh out of $0.44/kWh avg in PGE south bay), it’ll be interesting to see how much of this gets absorbed into the fixed-monthly base cost. (My numbers from https://www.ampmyhome.com/solar/lazard-solar-and-your-energy-costs/)
(Almost) right on all counts...
* GridStatus does not consider distribution; they only track prices at the p-node because (a) the GridStatus customers are energy developers that want to participate in the market, and (b) I have no idea how to compute that :)
* The HourFlex rate does include distribution congestion. That is only visible if you look at the detail spreadsheet, which requires calling into the API. I just checked at the Sample spreadsheet and I see some very high numbers, back in July 2024 - like $1.3!
I've added a TODO to my list: I'll use the API to extract the full data for my circuit. I want to understand how that distribution congestion is working.
* So, on the Base Services Charge... I have no idea how the volumetric rates will end up playing up in March 2026. The number used for Base Services Charge was a fairly arbitrary number. The argument at the time was "add some fixed cost to reduce everybody's rates", nothing more detailed than that. IIRC $24.15 came because SMUD has something similar.
* Exports when the rate is negative means the household will have to pay -- Precisely! That is one of the beauties of a Dynamic Rate with symmetric export/import. We will stop debating whether rooftop solar are free-loaders. If they export when the grid needs it, they get compensated. If they export when the grid does not need it, they are penalized. As it should be.
* The CalFUSE Subscription tariff -- Yeah, that tariff is not my favorite. That is why I am separating "rates" from "tariffs". There are many possible Dynamic Tariffs based on these Dynamic Rates. I'm delaying the discussion of tariffs to a separate post.