Public utilities in CA don't own any solar or wind stations. They are required to purchase the power if serving load or are required to control voltage & flows across their system to accommodate the generation but they aren't allowed to own it. IPP's (independent power producers) own the stations & sell the generation into the market either through power purchase agreements or through the CAISO market in day ahead or real time.
Filling in the California Duck Curve
- Thread starter Snagglefoot
- Start date
So that's why California has some of the most expensive electricity in the country and has a power grid on the edge of failure every year.
It's one of many reasons. Having a deregulated market could work great. In CA's case, we are only deregulated in name & it's become a political game for the state to mandate unreasonable policy that doesn't coincide with the utilities long term planning processes & forecasts. Hell, NERC even issued a statement in a risk profile report having government policy as the top risk to grid reliability.
https://www.nerc.com/comm/RISC/Rela...es_Report_2023_Board_Approved_Aug_17_2023.pdf
Short Utility Dive article on the above report:
https://www.utilitydive.com/news/nerc-assessment-new-risk-grid-reliability-energy-policy/691590/
Yep. Messed up the TWH. Not used to dealing with TWH. I see now the graph was for a month of production.
The California duck curve is about 10 years old now and they have added about 2 GW of solar (and now include wind in that) per year averaged over those years. The latest addition added 5 GW which needed a home which was in the form of the battery farms. The latest 5 GW was spread across about 6 hours or 30 GWh. So at $400,000,000 per GWh requires a 12 billion investment to make that solar and wind investment usable to California. The recently completed Site C dam cost $11 B USD and can deliver 1.3 x 24 = 31.2 GWhr each 24 hours. The value of 30 GWh at say $0.30 per kWh is $0.30 x 1000 x 1000 x 30 = $9 million per day and one years worth is 9 x 365 = $3.3 B per year. The 12 billion investment would be paid off in 12/3.3= 3.6 years.
Rinse and repeat. Of course the cost of the panels and transmission line upgrades is not included but California has no intention of stopping solar development just for the reason they need battery farms to be able to use the power that is coming on each year. We will see what happens in the next year and decide if they are serious about it.
I sure hope I "carried the knots" correctly. Where is Jethro?
Last edited:
OVERKILL
$100 Site Donor 2021
Cali wind has been pretty static. Annual output since 2017:
2017: 19TWh
2018: 16.5TWh
2019: 16.5TWh
2020: 16.4TWh
2021: 20.1TWh
2022: 21.6TWh
2023: 20.9TWh
Installed wind capacity is 6.29GW, so I'm not sure where you are getting that they are adding 2GW per year average?
Now, solar is absolutely increasing, though the elimination of the generous NEM 2.0 subsidies put a big damper on that.
Surplus solar that the grid can't use (which I assume is the 5GW you are referencing) being shuffled into batteries at incredible cost pretty much illustrates my point, and you are still burning massive amounts of gas
As it stands, that 30GWh represents 1/16th of Cali daily consumption in the low demand fall.No, Site C will produce at less than 50% capacity factor, something like 48%. So its 1.25GW will actually generate around 600MW on average, so it will produce around 5.26TWh/year. It's also a 100-year asset however, owned by the province with an extremely long-term payback schedule, rolled into the provincial utility's debt, so the rate impact of the project will be negligible.
This is more complicated than that.
One thing we have to consider is how the batteries are charged. You are looking at it like a monolith, but the reality is that these are individual battery projects all with their own CAPEX (and as I said, that $400K/MWh is low) and some may be colocated with solar, while others aren't. So, ones that are colocated, they may be charged by solar that's behind the inverter, so no real cost other than the CAPEX of the whole project (no round-trip charge cost). Others however, are being charged from the grid, so you can't use the discharge price plotted against CAPEX to determine the payback schedule, because the battery has to buy power on the grid, like almost 1GW of it was doing this AM at 1AM, generated by gas. And so then the round trip cost is the cost of the power, plus losses, plus CAPEX, O&M...etc. That can be as high as $0.48/kWh IIRC. When Diablo Canyon is delivering 2,260MW for $0.036/kWh, that's not looking like much of a deal. But then, they've been paying full retail to rooftop solar for years (NEM 1 and 2) so the whole rate situation is already a disaster.
Batteries are also never fully cycled.
As you and I already discussed, the actual discharge right now is ~20GWh, from almost 10GW of installed capacity, or roughly 2 hours at nameplate and this was all procured at considerably more than $400,000/MWh, so there's also the legacy cost of existing projects (which, if I had to guess, is probably around $16 billion).
A $400M say 250MW/1,000MWh (4hr) standalone battery project presumably charged via solar can charge/discharge once a day if being cycled close to capacity (you can cycle it twice if you can charge it overnight using something else, but obviously not solar, so colocated batteries would only be able to do one full discharge unless they are buying power from the grid). If it wants to discharge twice, it can only discharge less than half of capacity during each of the two ramp periods (morning/evening). This is likely the reason we only see battery output in Cali hit about 50% of nameplate, most of them are probably restricting output so that they have capacity to discharge again during the morning ramp (hence only ~1GW charging with gas).
Either way, you have say 800MWh of usable capacity if you aren't trying to max cycle the batteries (which you won't be for longevity). So, that's 292,000MWh/year. If we wanted to pay off the battery in 10 years, we'd need to generate $40 million in revenue every year, above and beyond O&M, so that's a base cost of $137/MWh, which is a lot better than the $300/MWh you used, but still very expensive, more expensive than Vogtle, and Vogtle is generating 2,274MW; 54.6GWh/day for 60+ years while a battery generates nothing, you still have to factor in the cost of generating the electricity to charge the battery into the equation.
Yep, it's certainly an interesting exercise to watch. And their electricity prices are insane, but it's a potential lesson I guess. What that lesson is, I don't think we'll know for a few more years.
I've predicted that California will be suing it's neighbors for their power and charging an average price of over 40c per kWh before the end of the decade. Not seeing anything that convinces me that my prediction won't come true.
Last edited:
This has the same problem with electric car: battery depreciation cost. We are in the era where PV panels are cheap and battery / grid capacity is expensive. What NEM 1.0 and 2.0 were intending to do got solved by the Chinese long ago and now the problem is how do you pay for all the solar without shifting the cost to the grid's other users.
I do think the demand side should be a better solution than the storage or supply side. You can have low cost EV charging at certain hours, you can have better tax write off for ice based AC in residential scale, you can provide tax incentive for grid timed electric laundry dryer, but battery to store solar electric then power AC later seems, wasteful to me.
Solar industry here is almost wiped out after NEM3.0. Nobody would have expected SunRun going bankrupt a few years ago. If they are going to waste money on battery for grid storage might as well provide cheap charging at work between 10am-2pm.
I do think the demand side should be a better solution than the storage or supply side. You can have low cost EV charging at certain hours, you can have better tax write off for ice based AC in residential scale, you can provide tax incentive for grid timed electric laundry dryer, but battery to store solar electric then power AC later seems, wasteful to me.
Solar industry here is almost wiped out after NEM3.0. Nobody would have expected SunRun going bankrupt a few years ago. If they are going to waste money on battery for grid storage might as well provide cheap charging at work between 10am-2pm.
Please open the box above to catch my reply on the duck curve. Had editing issues.
Regarding site C’s operating characteristics, In believe the 6 generators are capable of putting out 1.1 GW ( I accidentally typed 1.3 ). I typed in a simple piece of math multiplying that by 24 hours and 365 days for GWh, but yes, that can’t be maintained all year as the reservoir fill rate from the Peace River will dictate that. However, I would think it would not have a problem delivering 1.1 GW for a 6 hour period once per day, just like the California solar based duck curve delivery once per day. I was alluding that the dam’s cost was not far off from the cost of a battery that could deliver 5 GW for a few hours. I guess there are other reasons why they wouldn't want to open up all 6 water turbines at the same time but it would be cool to see.
For sure there are all sorts of extra details but it’s fun to throw out a few numbers to figure out what California is up to.
Once thing for sure, California is not interested in going down the rabbit hole of 15 year builds on nuclear reactors. No politician would take that on. Undeniably they are investing on a very expensive solution to further develop the use of solar energy but they have some of the best solar insolation, much like the enviable position of Ontario having a number of old nuclear reactors that they can refurbish.
Last edited:
OVERKILL
$100 Site Donor 2021
Yes, most of it has been solar, Cali doesn't have much more wind than Ontario
Yes, that's generally how reservoir hydro is operated, there's a constant level of output with some "peaking" capability utilized when needed. So you'd never have all the turbines being bypassed, but they'd also not be running full-bore all the time, the output would vary depending on demand. The Site C reservoir is very small compared to others in BC, but I'm sure whatever its operating profile is, that this has been well modelled by BC Hydro.
The difference between Site C and a battery is that Site C generates power, while a battery does not. That's an important difference. The other is that Site C is designed to last 100 years, while we don't really know how long grid-scale BESS's are going to last, but it's not 100 years (at least not with current chemistries) and may not be even 20 years based on how many have caught fire.
No, not at this time they don't seem to be. Extending Diablo Canyon seems to be about "it" at this stage. Like Australia, they do have excellent solar resources, quite unlike more northernly latitudes, but there are still challenges. Also like Australia, many are watching with keen interest
Here is something released by Oil Price.com. I take it that much of the solar is still being developed in California but the article used the entire USA to show the percentage gain. I guess a large local gain, spread across the entire USA looks large, but here it is. Enjoy.
Similar threads
- Locked