You note that TOU has "the intention of modifying the ratepayer’s behavior and shifting their energy usage." But I think we should also consider that TOU is just price gouging during hours of the day when people cannot easily shift load. Cooking dinner and doing household chores. Why do I say this?
The California Energy Code code for buildings uses a weighted time series for each hour of the year for each of 16 state regions, which gives the Long-term System Cost of electricity on that hour, which is based on a 30 year look ahead. (Formerly called Time Dependent Valuation.) The values are highest on hours when CAISO is feeling pain. CAISO does not feel abnormal pain from 4-9pm and I can find no empirical reason why rates are higher during these hours. The peak CAISO hours tend to be 7-midnight (after sunset, while air conditioning is still running), if we had to pick a five hour block, but even five hours is an arbitrary number, and this is getting pushed even later into the night by grid battery deployment.
4-9pm does not align with carbon pollution intensity either. It's either an anachronism or a cooking tax.
There was a time around 2017 when CAISO was pleading with IOUs to set aggressive TOU rates including "super off peak" to handle the duck curve, but IOUs never did it, and instead grid scale batteries solved the duck curve. TOU price differentials have been decreasing as rates overall rise.
I'd love if you could explain why 4-9pm is still a thing. Thanks again for your work.
Thanks for the comment about the Long-Term System Cost. Can you share a pointer to that?
In general, I think we are running the system "as it was", not "as it is". It may have worked OK for the days pre Renewables, Batteries, Electrification, and Data Center explosions, but it can't work today.
re: IOUs and Super Off Peak -- I think the reason PG&E has not done it is because their Billing System is a real mess. Check my latest post on the topic. As a (retired) software engineer, the situation there is appalling. https://pelegri.substack.com/p/that-pg-and-e-meter-to-cash-system
4-9pm is actually a false signal to the ratepayer because electricity from 4-7pm is clean and abundant (in CAISO). It would be a good time to run appliances. However, based on the price signal, someone might delay laundry or dishwashing until 9pm, when electricity is actually congested and dirty. Are you able to compare wholesale pricing with retail pricing during these evening hours to understand why there is such a mismatch between carbon intensity, wholesale pricing, and retail pricing?
And yes, I agree, 4-9 is a false signal. We are fighting a battle, with PG&E, with information asymmetry. They know everything, we don't.
We have solar panels. Our account is under NEM 2.0, so we get 1:1 export credits but NBT/NEM 3.0 compensates at minimal values, often below the LMP of my local node. A question I'd like to know is how much of my exports are going to loads in my circuit. That would address the arguments that IOUs are making about why even NBT is too much. The IOUs know the answer to that question. Maybe the CCAs could know it too. It is very hard for us to know.
CAISO - OASIS has the real-time market pricing and I've looked at it in the past. Now that I understand the data a bit better, in particular the way we / PCE purchases at the NP-15 prices, we could do an analysis. GridStatus ingests the data from OASIS, CAISO and others, and they provide it through graphs and API access. I only have their free level subscription, but I love what they do.
One area I have not yet started trying to understand is congestion. The LMP for my local node should take into consideration congestion at the transmission level but local, distribution-level, congestion is harder. PG&E publishes some data, mostly intended to assess whether it is possible to add DERs to a distribution circuit, but I don't know whether I can make any good use of that info.
See the first two files at this link: https://www.energy.ca.gov/files/2025-energy-code-date-and-hourly-factors. This is for the forthcoming Code taking effect 1/1/26.
Here is the Accounting Methodology Report, which explains LSC (formerly called TDV): https://efiling.energy.ca.gov/GetDocument.aspx?tn=255318-1
I've enjoyed your writing, thank you.
You note that TOU has "the intention of modifying the ratepayer’s behavior and shifting their energy usage." But I think we should also consider that TOU is just price gouging during hours of the day when people cannot easily shift load. Cooking dinner and doing household chores. Why do I say this?
The California Energy Code code for buildings uses a weighted time series for each hour of the year for each of 16 state regions, which gives the Long-term System Cost of electricity on that hour, which is based on a 30 year look ahead. (Formerly called Time Dependent Valuation.) The values are highest on hours when CAISO is feeling pain. CAISO does not feel abnormal pain from 4-9pm and I can find no empirical reason why rates are higher during these hours. The peak CAISO hours tend to be 7-midnight (after sunset, while air conditioning is still running), if we had to pick a five hour block, but even five hours is an arbitrary number, and this is getting pushed even later into the night by grid battery deployment.
4-9pm does not align with carbon pollution intensity either. It's either an anachronism or a cooking tax.
There was a time around 2017 when CAISO was pleading with IOUs to set aggressive TOU rates including "super off peak" to handle the duck curve, but IOUs never did it, and instead grid scale batteries solved the duck curve. TOU price differentials have been decreasing as rates overall rise.
I'd love if you could explain why 4-9pm is still a thing. Thanks again for your work.
Thanks for the comment about the Long-Term System Cost. Can you share a pointer to that?
In general, I think we are running the system "as it was", not "as it is". It may have worked OK for the days pre Renewables, Batteries, Electrification, and Data Center explosions, but it can't work today.
re: IOUs and Super Off Peak -- I think the reason PG&E has not done it is because their Billing System is a real mess. Check my latest post on the topic. As a (retired) software engineer, the situation there is appalling. https://pelegri.substack.com/p/that-pg-and-e-meter-to-cash-system
4-9pm is actually a false signal to the ratepayer because electricity from 4-7pm is clean and abundant (in CAISO). It would be a good time to run appliances. However, based on the price signal, someone might delay laundry or dishwashing until 9pm, when electricity is actually congested and dirty. Are you able to compare wholesale pricing with retail pricing during these evening hours to understand why there is such a mismatch between carbon intensity, wholesale pricing, and retail pricing?
And yes, I agree, 4-9 is a false signal. We are fighting a battle, with PG&E, with information asymmetry. They know everything, we don't.
We have solar panels. Our account is under NEM 2.0, so we get 1:1 export credits but NBT/NEM 3.0 compensates at minimal values, often below the LMP of my local node. A question I'd like to know is how much of my exports are going to loads in my circuit. That would address the arguments that IOUs are making about why even NBT is too much. The IOUs know the answer to that question. Maybe the CCAs could know it too. It is very hard for us to know.
CAISO - OASIS has the real-time market pricing and I've looked at it in the past. Now that I understand the data a bit better, in particular the way we / PCE purchases at the NP-15 prices, we could do an analysis. GridStatus ingests the data from OASIS, CAISO and others, and they provide it through graphs and API access. I only have their free level subscription, but I love what they do.
One area I have not yet started trying to understand is congestion. The LMP for my local node should take into consideration congestion at the transmission level but local, distribution-level, congestion is harder. PG&E publishes some data, mostly intended to assess whether it is possible to add DERs to a distribution circuit, but I don't know whether I can make any good use of that info.