Comment by cptcobalt | Hacker News Reader

Comment by cptcobalt | Hacker News Reader

I'm not OP, but the timing context of this post is important:

Today, California has issued a FlexAlert, and Powerwalls in PG&E and SCE will start discharging at 6pm to help reduce load on the grid. It's not a daily action, but it's emergency support when the grid needs it the most.

And the time window when it happens is important: Powerwalls are contributing to to reduce load when the grid needs it the most. It's not meant to support the entire grid, just directly attack the problem at the time where every Watt and Watt-hour counts.

You don't need to own Powerwall to benefit from this, just like how you don't own the peaker plants that typically spark up at these times. You benefit because there's a big chance that there won't be rolling blackouts in California this afternoon—distributed assets like this can be dispatched.

Dispatch Info: https://twitter.com/AlexGuichet/status/1559748677626777600

A VPP tracker in the Tesla community: https://www.lastbulb.com/virtual-power-plant

Some technical details of how Tesla VPPs work, as organized distributed systems: https://www.youtube.com/watch?v=ggdYts4muu0

CAISO Grid Outlook: https://www.caiso.com/TodaysOutlook/Pages/default.aspx

Has anyone done the math to determine whether the added wear-and-tear on the battery is negated by the payout?

I wouldn’t want to go from a 7-10 year battery replacement cycle to a 3-5 year battery replacement cycle out of pocket just because “I’m Doin’ My Part for the Grid(TM)”.

At $2/kWh you’re making as much as $16-$20 per cycle so it is almost certainly worth it to the battery owner to use your battery for this vs. load shifting, etc.

$2/kWh is insane though. My power rate is $0.12/kWh.

$2/kWh is still 6x California’s most expensive power.

> $2/kWh is still 6x California’s most expensive power.

You only need that power ("operating reserve") for extremely short times in a day, the rest of the time the power plant stays idle. Mostly these are hydro pump dams or gas peakers (as they can start in well under 60 seconds), batteries are a relatively new thing - and all three of them are pretty expensive to build, so it is obvious that their power generation comes at a high enough price that the plants still stay profitable.

Yesterday the peak price for electricity in Europe was 4€/kWh (that was in Estonia, was one hour period, due to some production issues at multiple power plants).[1]

[1] https://news.err.ee/1608686533/estonia-s-electricity-price-t...

And FWIW: I pay 0.42029€/kWh to my provider in the Netherlands while getting back 0.115€/kWh back for every kWh delivered back to the grid (after I generated and used most of it in my home)

Used to be 0.23ct or something.

I don't want to talk about or look at the prices for a cubic meter of gas...

Here in Estonia the spot market pricing works the same both ways. So I set my system to feed at 7kW during the hour and made about 28EUR for helping the grid. That is twice the $2/kWh this Tesla peak power provides but it's literally the only time price has gone this high.

You're confusing peak pricing with average pricing. Peak price is about 50-250x as much as average price.

During the trial its $2/kwh. A powerwall costs ~$10.5k for a single unit with a capacity of 13.5KWh. You'd need to do 5MWh to pay for the powerwall, and that would be ~370 cycles.

The pack should be good for over 1k cycles. This is definitely a beneficial thing for the owner at these prices.

For reference:

https://www.tesla.com/sites/default/files/pdfs/powerwall/pow...

If it's being used for solar-piped home backup it's covered for unlimited cycles, and for any other usage they rate it for 37.8 MWh of lifetime AC output.

Now, I don't see anything talking about the Powerwall's warranty within this article explaining Tesla VPP, so whether or not this might be considered when they determine warranty eligibility is anyone's guess.

I would think that just charging $2/kWh during peak time would be enough to bring down peak load without using any batteries.

Well, yeah, probably. You'll also make a lot of people unhappy

$2/kWh is very expensive. I guess by paying the powerwall users you can limit how much you spend with them. Not sure with how much local granularity can you turn them on, but it definitely helps

How can they pay powerwall users $2 if they're not actually charging $2?

Presumably most of the time they are charging end users more than they are paying energy providers.

By amortizing that across off-peak pricing?

By not having to build more power plants or not putting the battery equivalent of all the power walls.

Also, I imagine this is to get the power wall users 'hooked'. Then over next few years that $2 will turn into 20 cents.

This isn’t actually unique to powerwall users - this is a program for energy providers to supply emergency power. There’s a minimum amount of power needed to be eligible and a powerwall is too small. Tesla realized that they have the unique ability to manage a fleet of small batteries to act as a single large one to qualify.

Will the price go down in the future? Maybe. Is this a gimmick to get people hooked? Probably not, assuming Tesla isn’t paying end users more than they make to join in.

And hope that won't realize they're destroying their battery without enough compensation?

You're buying the bare minimum to stabilize the grid and you're buying it "delivered" already

Only if people reacted to it reliably and predictably, which generally at the moment they can't.

So they're being paid not just for not using this power at that moment, but being part of a system of "demand response" that lets the grid plan ahead and modulate demand.

They also do this with people's AC and various large industrial energy users.

But, time-of-use rates are a currently underutilized idea, and with greater computerisation, you can get more and more value from this on short-timescales.

You want to start with the low hanging fruit first, and EV chargers are that easy win as they are being installed now.

> But, time-of-use rates are a currently underutilized idea

Is it not normal in the US to pay less for electricity at night and on weekends?

That's very common in the places (Netherlands and Switzerland) where I've lived, for as long as I can remember, and seems much more low-tech than using consumer batteries to supply back to the grid and paying them a custom rate.

It's still 'controversial' and little used in the US:

https://www.utilitydive.com/news/an-emerging-push-for-time-o...

> A new idea is to vary the price of kWhs, making the price higher for residential customers during peak demand, and lower at other times, allowing for more precisely pricing kWhs at what they cost.

> With today's flat per-kWh rates, "customers who use the grid more during peak demand are underpaying for the costs of meeting that demand," Brattle Group Principal Ahmad Faruqui told Utility Dive. "Customers who use the same kWhs during off-peak times pay the same bill and are subsidizing the others. TOU rates can redress that inequity."

The quotes drive home why it is "controversial"; because generation and transmission companies see it as a potential way to improve revenue, not as a way to improve the grid.

I will assent to these powerful features when there is a strong legal guarantee that it will be used to improve the world, not to juice up someone's q2 numbers by charging me more.

Remember that these are the same companies that by and large lobby against being able to sell solar power back to the grid.

> lobby against being able to sell solar power back to the grid.

Worse, they lobby against residential solar as a whole. I could install enough solar panels & batteries to cover 100% of my needs through the year, but unless I disconnect from the grid completely, I'd still pay about half of my current electric bill, without using any electricity from the grid. The utility company with local monopoly is specifically making sure as little residential electric is installed as possible.

In San Diego, we get to choose whether we want to opt into the time-of-use pricing, or whether you want to pay based on the traditional flat-rate tiered pricing (where your monthly consumption is compared with your neighbors to determine your price per kilowatt).

It sounds fancy and pro-consumer, but in reality, we have some of the worst rates in the nation. Jon Oliver even did a recent bit about how ridiculous it is. (Search for "Jon Oliver SDG&E".)

How long will the trial prices last? $2/kwh is insanely high; usually the wholesale electricity prices hover closer to $0.02-$0.04/kwh, although the recent gas price increases have disrupted that for now. The war in the Ukraine is not going to last forever though, and if gas prices remain this high we might see an accelerated transition to non-fossil based electricity generation.

At $0.20/kwh, so prices 10x above "normal", you'd need 3.7k cycles to pay for the powerwall. That would be well above the lifespan of the pack, so at lower prices the profitability of the scheme (to the owner of the pack) falls off rapidly. I suppose it all depends on how long you estimate the trial pricing will last. TFA mentions in the fine print that this project "shall continue until at least December 31, 2023, unless expressly extended by PG&E or terminated sooner." Take from that what you will.

Peak pricing is much, much higher than the prices you quoted and in the ballpark of the trial. E.g. before the 2021 Texas outage the offer price for additional power was the legal cap, $9 / KwH at the time.

Peak pricing is indeed higher, but also much rarer. Prices of $2/kWh are reached extremely rarely, so the VPP in this scheme might only be operating once or twice a year. The calculation in the GPP about payback time for the powerwall assumed you could get a lot of cycles in, but at these elevated prices that seems unlikely to happen. Note btw that the $2 is flat, so even if prices reach $9/kWh you'd still only get $2/kWh back.

I notice that the article does not include any expectations about how often the "events" would happen btw, even though the utilities would surely have enough data to at least make an estimate.

The payback calculation was a direct reply to the concern that VPP use will wear out the battery too fast. The only reason it was done was to show that if VPP would trigger often enough to meaningfully harm the pack, it would also pay you more than enough to replace it.

In practice, VPP will be rare, and thus the payouts will be low. But so will the harm caused to you by participating, assuming you already own a suitable system.

Also the 1k cycles is until the pack capacity goes down to 80% of the original. You could still use the pack well past that unless tesla somehow limits it or complete short of some of the elements in which case you should be able to replace the element.

TBH, I think it goes well past 1k even for the warranty. Someone else pointed out that its closer to 2,800. I wasn't sure, so I picked a conservative number.

Although it was a much more extreme event, no fewer than 246 people died because of the Texas outage in 2021. It’s not just the economy that can suffer if there is a grid outage.

https://en.wikipedia.org/wiki/2021_Texas_power_crisis

That 246 includes all deaths directly or indirectly attributed to the storm. Not specifically the power outages.

"In addition to hypothermia, DSHS attributed the storm-related deaths to "exacerbation of pre-existing illness" (10%), motor vehicle accidents (9%), carbon monoxide poisoning (8%), fires (4%) and falls (4%)."

Seeing as the storm does not produce carbon monoxide itself, I think "storm related" here includes "people dying because the storm knocked out their power"

not having power also doesn't produce carbon monoxide

Portable combustion heaters do. People were using those because they didn't have gas or power for heat, and that's usually how people end up dead from CO. I don't think there's an honest way to spin that into not being the storm and grid failure's fault.

The flip side is that in an event like that, the power is worth far more than $20 to the owner of the PowerWall.

I lived through that outage, and I lost more than $20 in food that spoiled in the fridge/freezer (didn't have a good storage container to put it outside where the raccoons/cats couldn't get it). Some people lost dramatically more than that when their pipes froze and burst.

Central heating also doesn't work without power. They draw less than 600 watts. In that situation, would you rather have $20 or enough power to have your heating run for 24 hours? And that's actual runtime, you could probably keep your house warm-ish for 2 or 3 days by turning it on and off.

Personally, I'd only be interested if joining something like VPP came with a substantial subsidy on the price of the PowerWall or a monthly payment that could offset a significant portion of the cost. Or it had enough storage that I actually had extra power after running essentials (it does not currently).

These stationary battery should not require anywhere close to a 7-10 year cycle. The chemitry should be above 80% for much longer and te battery should stay usable for decades.

> out of pocket just because “I’m Doin’ My Part for the Grid(TM)”.

Exact mentality why we'll never ever fix our ways of life

The grid in California is run by a problematic for-profit company that has a history of starting deadly wildfires.

People should generally be willing to make sacrifices to help the greater good, I’m sympathetic to people who don’t feel obligated to do their part.

For Australia, it was not worth it, except to get a few thousand off the purchase price (then quit VPP as soon as humanly possible).

My experience was their algorithm is terrible and it gets into a charge+discharge cycle for no reason for hours at a time, costing $0.40 to charge and earning $0.10 to discharge, plus wear, for no benefit to the grid.

I used to turn off my battery a lot.

Important to have some perspective on scale, though. All powerwalls acting together can source 125MW, max. Yesterday, grid-scale batteries in California discharged a peak 2751MW. The powerwall tricks are cool, but I’m not sure it solves a problem the ISO actually faces.

We've come so far in so little time and people like to keep puting progress down.

Regulated market participants are adding the equivalent capacity every month. That's where the progress is coming from, and where you should expect most future progress to occur. Long-tail stuff just isn't efficient enough to really make a dent.

But the number of power walls is not fixed, and going from 125MW to 12,500MW is only going to happen there are proper incentives from the grid.

And as it is, utilities in the CAISO are generally too fossilized and change averse to view behind the meter aggregation as something that they can benefit from, and they will likely instead fight it. Smarter utilities would see the opportunity and approach regulators with rules changes that profit everybody. Instead, we will need legislators to impost such schemes on regulators and utilities, as there is zero innovation among leadership of utilities, and regulators have zero capability for this too.

(Perhaps I exaggerate, SMUD is tiny but they may have the vision for at least part of this. Vermont's Green Mountain Power has the leadership that we need in California, though.)

Regulated market participants went from zero to 3000MW/12000MWh of capacity in only 2 years. That's 100x faster than powerwall installations. I don't see how you can conclude from these facts that CAISO or the CPUC are stuck in the mud.

Central installations are perfect for the current incentives: they use the same amount of T&D, for which utilities can always charge their profit margin, as long as they can convince regulators that the expenditure is justified.

Behind the meter storage is a fundamentally different beast, because it has the potential to drastically reduce T&D costs, which are the vast majority of bills (at least mine under PG&E).

Christopher Clack has some great economic modeling of the grid at a very fine level, showing that deploying a bunch of storage and solar at the grid edge right now will greatly reduce costs and enable far more utility scale installs in the future, because the local storage and solar reduces the need for T&D peaks.

This sort of cost optimization is not incentivized but anyone in the current setup, and is impossible to achieve without direct regulation to change incentives for utilities.

I don't disagree with any particulars of your analysis except one omission: powerwalls compete with electric cars for relatively scarce high-tech batteries. It is globally optimal for the time being for those batteries to all be used to electrify transportation, while utility power can be enhanced with large, fixed installations using whatever kind of battery they can get. Lead-acid or anything else, it doesn't really matter. But nobody is selling turn-key residential power packs using anything but lithium-ion batteries.

I thought a lot of these were recycled from vehicles when their energy density doesn’t meet the needs of a vehicle anymore but make sense in a house since the house doesn’t move?

Note this same argument is used against EVs. Saying the batteries should be used for hybrids instead.

In both cases, the fundamental mistake is to think of these as costs, when they are investments with rapid paybacks and so the obvious answer (and the reality) is massive expansion of battery manufacturing capacity.

The hybrid thing is true though. With the same number of batteries you can cut carbon emissions from cars by either 50% with hybrids or 2% with electrics. It makes perfect sense as the thing to do now with the capacity we actually have.

No it doesn't.

The thing to do now, is increase our battery capacity.

Not all parts of the energy grid are as “clean” as a gas cars’ ICE unfortunately. You’re better off driving an ICE in those regions than charging an EV. Globally we’re better off creating a sustainable carbon-free grid in regions with poor grid generation first.

In California, where this applies, the grid is very clean, so yes the dirties/most polluting part of the system is probably the cars.

Seems like having an available 5% bonus on peak (!) usage is a huge deal, actually. Buffering shortfalls have an exponential frequency curve. I wouldn't be surprised if 5% extra generation would have been enough to cover, what, 60-90% of all brownout/failure conditions on the grid over the past few decades.

Also, too, it's a brand new technology just being rolled out, and building more capacity at the limits of factory bandwidth. You're acting like the number of powerwall users are all we're ever going have, which seems silly.

> Seems like having an available 5% bonus on peak (!) usage is a huge deal, actually.

CalISO showed a peak demand today of ~42GW. 125MW is more than nothing but nowhere close to 5% of peak usage.

And solar was providing about 3GW less than it was about a month ago. Lower solar output but spiking fall temperatures. It's sort of a "coffin corner" for the state, currently.

I think they meant that if grid batteries are a good thing (which they are), then distributed batteries equal to 5% of their peak contribution, isn't nothing. (125MW/2751MW)

The baseline is not zero GW. If the grid planned for 35 to 40GW of demand for today and then sees 42GW load, the 0.125GW are over 5% of the excess load.

Serving baseline from batteries is unrealistic and will be for quite some time. Smoothing peaks or helping with unexpected high (or low!) load however is slowly becoming viable.

I'm surprised it is so low. Our pack discharges at at least 10kw (max averaged by hour over the last two weeks; didn't look up specs). That means we're over 1/12,500th of the state wide pilot's footprint.

Apparently, (as of 2021) Tesla has only sold about 200,000 powerwalls globally, so I guess that sort of checks out.

Annoyingly, our batteries are set to minimize our bill, but didn't get the memo about the special rates today (they are not powerwalls). They stopped selling back to the grid just before 6pm, and are almost half full as I write this.

More annoyingly, every morning they aggressively charge themselves during the precommute CO2-intensive demand spike, since power is cheap then (due to a policy setting, they normally only charge from solar)

I wish I could forbid them from charging before 10am or that PG&E would fix their rates.

Many of these devices do let you set 'fake' time of use details if you want to further restrict them. Not ideal, but a quick way to make your energy use greener if you consult one of the APIs that tells you when your local grid is greenest (which probably roughly, but not quite matches their time of use rate structure since they have to simplify, possibly averaging over a year when it various seasonally)

The Tesla stuff is just a name-brand subset of a larger demand response program in California though:

https://www.caiso.com/participate/Pages/Load/Default.aspx

I think the demand response stuff adds up to about 2GW, so the same as the grid batteries at the moment, and both are set to grow at rapid rates.

Reply on news.ycombinator.com