> they buy power at about $0.02/kWh but then they sell it to the residential consumer for up to $0.32.

This is the most important factor in the whole discussion.

The electricity retailers, which consist of a call centre and a two billing systems, accounts payable and accounts receivable, buy electricity for virtually nothing, $1.00 buys them 50kWh, that's about double what my houses uses per day with two adults. edit to add: per day

I'm guessing the electricity retailers aren't making a lot of money after expenses, otherwise investors would flood the market.

How would a reduction in wholesale electricity price translate to a reduction in residential retail price given the wholesale price is already almost at zero.

Wholesale price is definitely not “zero”. It fluctuates wildly, and if a retailer is exposed to the real time market at the flow date, they can expect to lose hundreds of thousands of dollars. RTM prices can easily spike from $50/MW to $8000/MW in an instant if a generator trips. So, retailers buy hedges to cover this risk, at price premiums that reflect the inherent exposure insurance.

I didn't know this, thanks for clarifying. Further investigation for my local area:

Here's a link to the Australian wholesale spot price and peak price https://www.aemo.com.au/Electricity/National-Electricity-Mar...

Where I live, Tasmania, the average for 2019 is $147.91 / MWh and the peak is $163.

That works out to Au$0.163. I pay AU$0.26 / kWh, plus a daily supply charge of, if memory serves me correctly, $2.30 - this is to cover network operating costs.

That's actually a fairly reasonable mark up on the retails side of things.

That makes the other commenter's $0.02 / kWh seem fairly misleading. Maybe for some places some of the time.

The "peak" price they're talking about on that page is the average spot price over the peak electricity consumption time (7AM to 10PM weekdays).

It's not the highest spot price seen in the market. This time of year the east coast electricity market seldom spikes, but if you go to the historical data page you can see that in January this year the 30 minute spot price in Tasmania spiked above $2000/MWh four times (22-Jan 17:00, 23-Jan 12:30, 23-Jan 13:30 and 30-Jan 16:00).

There are many things you can defer, even at home. Mostly A/C related, but stuff like a self-cleaning oven or a clothes dryer doesn't need to run in the instant, necessarily. The dryer might be able to wait a few hours once the clothes are a bit dry (due to mold and such), as it needs to be ready when you get to it in the morning/afternoon. Until then, it only needs to prevent the clothes from water damage.

The oven can maybe just wait a day or a week even to clean itself. It usually works fine even if partially dirty, and just causes some smell due to thermally decomposing food contamination (cheese drips from pizza, etc.).

I don't need to defer anything, as I'm on a flat rate. I don't like peak metering, as it would forces me to micromanage things in a way that doesn't work for me.

I don't have a clothes dryer, they use too much energy. I just hand my clothes on an indoor clothes airer and blow a fan on them. I usually leave the heat pump set to 16 degrees C while I'm out, the combination of mild air temp and the fan dries most of my clothes over night anyway.

Hadn't even heard of a self cleaning over. I don't bake a lot, so there's that.

It seems odd that you worry about the energy a dryer uses, yet set your thermostat to 16 degrees when you're not even in.

Depends on whether they live somewhere hot or somewhere cold.

Youre Right! Parent is apparently from Tasmania, which I lumped together with the rest of Australia as "damn hot", while that apparently isn't the case at all.

Melbourne, Adelaide, Perth. Everywhere on the south coast of Australia experiences winters that necessitate heating, though perhaps not left on constantly.

That seems very odd to me. Maybe Australian regulations or trade situation is different and makes the math different, but in my area (SoCal), TCO for my rooftop solar installation was $0.08/KWh using a Net Present Value calculation, and utility solar is more than 2x cheaper than that. It’s very odd to hear wholesale prices above 10c when rooftop is so cheap.

Not really addressing your point, but a sidenote on the topic of Australian regulations -- Due to some subsidies/incentive schemes which came through in some states about 10 years ago, some residential Australian rooftop solar installations are selling any excess into the grid at a rate of 0.44/KWh and will be until (from memory) 2024.

Here the connection charge is less than $0.50 a day and electric is $0.095 per kWh.

(those are the retail prices for a small US town that isn't particularly close to any power plant)

Companies also need to pay tax and other per-kWh costs on top of the spot price: are you taking that into account?

But what’s happening more and more now is that batteries are arbitraging those rate spike, and ultimately those spikes will disappear entirely.

Utility scale batteries are not yet a proven and deployed technology. Not to say there aren’t successes - Aliso Canyon and the AEMO installation in Australia have both been very well received by their respective system operators. But there’s a very long way to go before batteries will exist as a viable, general alternative to natural gas peaker plants.

I did not say that they are an alternative to natural gas peaker plants. I said that in markets where the spot price can spike to $8,000 per MWh, grid scale batteries are being deployed in order to serve those load spikes more economically.

There is about ~1GWh of grid scale battery storage already deployed in the US alone. 150MW was deployed in 2019 Q1 - which represents 232% growth YoY . ~5 GWh is projected to be installed annually by 2024. [1] I think that this qualifies as proven and deployed.

[1] - https://www.woodmac.com/research/products/power-and-renewabl...

I agree, proven and delayed, when and where it makes sense to do so.

It remains to be seen if this will become a common deployment generally speaking, though with more grid scale wind and solar being built it does seem likely, in my opinion.

Minimum US national load is ~400GWh and peak is ~675GWh, [1] so there’s a tremendous amount of storage which could be used to smooth out the curve.

Perfectly smoothing the curve would require something like 2TWh (charging at 150GW for 12 hours and then discharging at 150GW for 12 hours), which would be total overkill, but every little bit can help stabilize the grid—and spot prices—that much more.

The fact that 10 years from now we could have 100GWh of storage, and adding maybe 50GWh per year is pretty awesome. Global Li-Ion battery production is forecast to be ~1TWh by then.

Between electric cars, home storage, and grid storage, chemical battery production seems like it’s turning into a trillion dollar market. Eventually we’ll add airplanes to that list!

[1] - https://www.eia.gov/todayinenergy/detail.php?id=27192

You're exactly right. Thanks for adding the the knowledge pool today!

I think companies like Cal Edison do a lot more than billing and sales. They have to build out and maintain the transmission infrastructure. Which, considering they charge 32 cents per kwh marked up from 2 cents, it seems the infrastructure upkeep is a lot more expensive than the actual power generation.

Most research suggests that net metering is underpaying the households that deliver it. This seems very counteri tuitive to people, but remember you're paying an average cost. Gas peakers can cost thousands of times that when load is high. If you use AC then solar and high demand are correlated. At the extremes that means building a whole new powerstation that might only be used for literally hours per year.

The utilities often don't want to buy this power - if they felt that they were underpaying, they'd want to buy it. Could you cite the "most research" you speak of? In fact I've read the exact opposite - that solar rooftop is economically possible only because the utility is often forced to buy it back at retail rates.

Utilities often get recompensated based on a percentage of what they spend.

If you assume they are rationally going to maximise their profit potential then saving money isn't really in their interest even before you get into externalities that they impose on others.

Another example of this is utilities moving coal and gas plants that are no longer economic into these kind of compensation deals to they get a guaranteed profit based on what the cost to run. This is why it's estimated that closing all coal plants in the US would save 10s of billions dollars just in lowered electricity costs, even before factoring in pollution and carbon.

For info on the research "value of solar" is the general term. It varies by geography and location (e.g whether demand is growing or falling, what the other power they displace is coming from etc) but it's generally pretty positive for solar. The number is even bigger when you include things that would save customers money rather than the utility, and as regulated industries they should probably be forced to consider those costs.

But what you said and what I said aren't strictly incompatible. You could get a prisoner's dilemma type situation in which it only makes sense for you to do something if someone else is forced to do something too, otherwise they would defect and gain even more. Doesn't mean it's not economically beneficial for both.

I have no real interest in whether solar makes economic sense for a household or for a utility. It clearly is at the society level and we should be organising ourselves so that we maximise the benefit, not throwing our hands up and saying "well, if it would involve changing a minor regulation on an already heavily regulated industry, then I guess I'm going to have to choose the more expensive option instead"

It’s overpaying if anything. Why should the utility be forced to pay retail RTM rates for power that they could have bought wholesale a month before if they actually needed it? Utility ratecase logic often has many ratepayer-unfriendly motivations but this line of reasoning is legitimate, imo.

We're talking about an average net-metering cost of about US$0.12/kWh, right? That's 3.3¢/MJ in SI units, US$120/MWh in the units used in most real-time electrical markets. LMPs do sometimes get as high as US$120/MWh and even higher — I've seen US$160 at times — but never anywhere close to US$12000/MWh, as your "cost thousands of times that" comment suggests. (US$40 is more typical, and sometimes prices go negative, so you can get paid to burn energy.)

RTM prices in ERCOT can easily spike to the price cap of $9000/MWh during bad days: https://www.spglobal.com/platts/en/market-insights/latest-ne...

Interesting! So even if the demand would justify paying $12000/MWh, ERCOT just permits brownouts or blackouts instead of paying that much? (I can't load the article you linked.)

Realistically a price higher than that wouldn't really have any effect on incentivizing generators any more, since there is a limit to how fast they can ramp up, how much spinning reserve capacity they have etc. So, it mostly just serves to protect the market from falling off the rails. The grid operation itself is actually largely disconnected from the market - the ISO primarily calls the shots with scheduling regardless of what the market is doing.

Because the powerwall allows you to "buy" power at the opportunity cost of $0.02/kWh vs paying up to $0.32/kWh ? If you end up generating a lot of excess solar using a lot of grid power at peak times, it could pay for itself.

But of course, you need to do the calculation for your situation.

I don't think powerwall has that functionality. At least original powerwalls only provided backup power. From a C&I perspective this is one of the sources of revenue for batteries - however there are many different ways for electricity charges to be passed on to customers. It isn't only kWh.

"I don't think powerwall has that functionality"

From wikipedia

"When originally announced in 2015, two models of Powerwall were planned: 10 kWh capacity for backup applications and 7 kWh capacity for daily cycle applications"

https://en.m.wikipedia.org/wiki/Tesla_Powerwall

I take it the daily cycle applications include that functionality.