[slapshot]

An interesting problem is highlighted on that graph: solar seems to be peaking around mid-day, but energy demand peaks at 5 - 7 pm when solar is pretty weak. The curves showing pre- and post-solar have almost identical peaks; solar hasn't yet successfully reduced peak load. (The California energy system posts a similar graph, with a similar shape: http://www.caiso.com/Pages/TodaysOutlook.aspx , made worse by the fact that the California wind farms apparently perform best in the morning and worst in the afternoon.)

The difficulty is that, unless storage gets dramatically better, we'll still need a huge conventional electricity base. And it will still burn a huge amount of fuel: most generators are not "instant-on" -- it can take weeks to spin up a nuclear plant, days to spin up a coal plant, and hours [1] to spin up a natural gas plant. Of course, there are some savings, but there is huge waste if you have to keep the whole conventional energy infrastructure spinning during the day just to fuel that 6 pm peak load.

Advice to entrepreneurs: finding a cost-effective way to store solar energy for 4 hours will be worth more than another 2% increase in efficiency. And inventing instant-on conventional-fuel plants will also make a huge difference in GHG emissions.

[1] - Page 8 of http://www.euec.com/getattachment/euecjournal/Paper_3.pdf.as... gives times between 1.2 hours for a warm start to 6 hours for a cold start.

[reitzensteinm]

We don't need any breakthroughs. Electric cars could handle the entire peak load with today's technology, the right incentives, and capital flowing in the right place.

The peak load for South Australia was 1.8 gigawatts. There are 1,275,041 motor vehicles registered in SA.

If you could get all of them turned into electric cars plugged into the grid, it would be a 1.4 kW power drain per vehicle, which is less than the power drain of an ordinary kettle here (240V, 10A, ~2.4 kW).

The cheapest Tesla Model S has a 40kW hour battery meaning it could sustain that power drain for over 28 hours (a total of 51 GWh). It's got a 10kW charger standard, which is more than enough.

So if everyone had the cheapest Model S, kept it plugged in with the ability to send power back into the grid, and didn't mind that their car was sometimes down to 20% charge, you could power the entire grid on solar alone.

That's obviously a set of unrealistic assumptions, but it does indicate that it is feasible to solve the problem like this.

A more realistic set of assumptions:

  * 1/10 cars are turned into a Model S base model equivalent
  * Through financial incentives, the owners are willing to keep them plugged in 
    during the day and using up to 20% of the battery to push back into the grid.
  * Cars need to shave 0.2 gigawatts off of peak load for four hours to bring 
    it in line with the rest of the day

That's 800 MWh of power to satisfy peak demand, at 200 MW.

We have 127,504 vehicles, can can use up to 8kWh and 20kW each, giving:

1 GWh of capacity and 2.5 GW maximum power draw.

So it's just enough. Alternatively, 5% allowing 40% usage works, etc. The power draw is insignificant.

Owners can be heavily compensated for pushing energy back into the grid. I won't run the numbers here for length and time reasons, but knocking out peak power usage is incredibly profitable. You're literally decomissioning a large percentage of power plants. It would be feasible to make all the energy your car uses free; likely the lithium ion battery packs, too.

The question is; would 5% of the driving population buy a $50k car if they no longer had to pay for fuel or battery packs? Financially that would probably put it closer to a BMW. I think it's feasible. What about in 10 years when the cars are $20-30k? Undeniably. In 20 years when you only need 2% of the population to be in the scheme due to battery increases and the cars cost $20k? No brainer.

The system would require a smarter grid (so you can plug your car in at work and have it all taken care of), 10 years of Moore's Law for batteries, etc.

But I think the numbers check out, and it means we could go crazy with solar (the explosion in solar must continue to charge these vehicles during the day). What's needed is the will to make it happen.

[brc]

>So if everyone had the cheapest Model S, kept it plugged in with the ability to send power back into the grid, and didn't mind that their car was sometimes down to 20% charge, you could power the entire grid on solar alone.

The OP said cost effective, not pie-in-the-sky.

Even if you took out the battery packs and sold them separately, the cost would exceed 10 years worth of electricity supply.

>Owners can be heavily compensated for pushing energy back into the grid.

You're asking people who cannot afford the expensive technology to subsidise those who can. This is the exact reverse logic of most progressive taxation regimes.

>The question is; would 5% of the driving population buy a $50k car if they no longer had to pay for fuel or battery packs? Financially that would probably put it closer to a BMW.

So you want to subsidise the purchase of expensive cars to the point where it is financially a good deal. This magic money no doubt comes from other taxpayers.

And for what end? Just so you can have some type of boutique distributed power generation system?

>What's needed is the will to make it happen.

No, what's needed it pots of other peoples money.

I've spent the best part of 5 years trying to hose down the jetsons fantasies of people pushing ridiculous schemes like this as not only unworkable, but inequitable for forcing up a basic cost input of life - energy - for effectively vanity purposes of a small subset of the population. I usually cop a pile of flamebait and downvotes each time, but I do so because there seems to be a mass delusion going on, and this has become one of those things you can't say.

[caf]

So you want to subsidise the purchase of expensive cars to the point where it is financially a good deal. This magic money no doubt comes from other taxpayers.

The proposal isn't a subsidy. It's a straight market payment for an economic benefit - the difference in cost between off-peak and peak electricity is real, and this means that storing off-peak electricity and releasing it at peak is an activity that's worth money.

I'm not sure I agree that the economics work out exactly as described, though - if it would be worth paying for everyone's lithium car battery packs to do this, then it would be even more profitable to build a giant lithium storage battery in a central location.

[reitzensteinm]

Well, the benefit is that it combines the capital costs of owning a vehicle and storing energy. Many people will be purchasing a 160+ mile range car, but day to day use only 20-30 miles of that - an unused capacity large enough to power an average house for over 24 hours.

This is in contrast to dedicated lithium ion batteries, which just don't work out economically. Otherwise you're right - the utilities would do it themselves at scale.

[caf]

I can see how that's true if the car owner's continue to pay the capital costs of the battery - the part I'm taking issue with is where you say that it would likely that the car owners would have the capital cost of their battery packs paid for by taking part.

[reitzensteinm]

My back of the envelope calculations elsewhere in the thread were $1571/yr savings per car (10% overall reduction in grid cost); the Model S battery replacement insurance is $12k per car, which is a ~7.5 year payback (by which time, they claim the pack will be at 70% effectiveness).

Elsewhere someone corrected me that the peak grid usage in a year is actually significantly higher - the figures I was looking at were average daily figures - which may make grid savings quite a bit higher than 10%.

I don't think it necessarily is a clear financial win now; I'm just saying, if you squint just right it kind of makes sense now, and with the march of battery technology and PV, not to mention the inevitable adoption of EVs whether or not this scheme exists, it's going to look better and better from here on in.

[reitzensteinm]

First off, I specifically said that the first lot of calculations was based on unreasonable assumptions, to gauge the size of the problem.

Attacking them because they're "pie in the sky" frankly sounds like you're deliberately misinterpreting me, or at least you didn't bother reading the post.

Secondly, the scheme I'm proposing does not require any government subsidies. The money comes from selling the power back to the grid.

With my proposed figures - 5% EV ownership (63,752 cars), 40% battery usage sold back to the grid, you could reduce peak power usage by 15%. South Australia spent ~$990 million on electric power last year (not exact figures; based on 619k households from the census with an average spend of $1600).

Assuming 15% reduction in peak power is a 10% reduction in cost (which is reasonable, because capital costs and fixed running costs dominate electric spending expenses), that yields $99 million a year in savings.

The cost of 63,752 Model S cars is $3.1 billion. However, you have to look at the marginal cost of purchasing a Model S, vs whatever car the high end of the market is buying anyway, AND subtract the spending on fuel.

We're saving $1571 per car per year. Tesla's prepaid replacement program is $12,000 (over ten years), which eats a big chunk of it, however we're also charging with cheap peak solar (the abundance of which is causing this issue to begin with), plus maintenance is cheaper.

BMW 5 series MSRPs are from $38k-68k. Taking the lower end of that, we're paying a $12k premium for a car with significantly lower maintenance costs and little to no fuel costs.

Could you get 5% of cars on the road taking that deal? I don't know. Today? Maybe, maybe not. But I do know that as the price of peak solar and battery technology falls, it will go from being a 'maybe' to a 'no brainer' over the next ten or fifteen years.

In any case, I may be wrong in my analysis. I'm merely presenting an argument for the way I see it. If you disagree on a specific point, I'd be happy to talk more about it, but nothing I wrote deserved your response.

[MikeCapone]

Power plants aren't free, people pay for the service they provide. If you can replace them with less expensive batteries (especially since people will be ready to buy electric cars for their main benefit, which is transportation -- getting paid to lease you battery is a side benefit) to store cheap solar power, that's a good deal, not a subsidy.

[brc]

>If you can replace them with less expensive batteries

Agreed. And when that is the case, then this behaviour will appear at large and be taken up enthusiastically.

At the moment this manifestly isn't the case (stored solar by batter cheaper than large-scale power), so any calls for it to happen is usually a disguised call to force it to happen by producing regulatory benefits to a small subset by driving up the cost for others.

I have zero problem with people electing to spend their own money on solar + battery storage. I find the tech interesting. I am simply determined to point out whenever I see this type of call that usually this means taxing poorer people to pay for richer peoples desires.

[MikeCapone]

Maybe I misread, but I think the OP pointed out that it could be done with today's technology, not that it could necessarily be done cost-effectively today. But batteries are improving and their costs are coming down, so it's certainly possible to predict that it'll be possible at some point in the not-too-distant future. When that happens, people will be compensated because it makes economic sense, not because they are subsidized into an un-economic activity (though that could happen too, but I don't think that's what was argued here).

[prawn]

FWIW, I forwarded the article and these comments to my friend who works in the energy sector in South Australia. His comment:

"The comments talk about electric cars effectively being used as batteries to help supply the grid during peak times which is a concept I haven’t heard of. Electric cars being plugged in overnight to increase the night-time demand would certainly help as you would significantly increase the base-load demand, thereby improving the load factor of the network. This then distributes the fixed infrastructure costs over a higher number of kWh lowering the per/unit cost of energy."

[rdl]

I think cycle counts for expensive low mass batteries are going to be a lot more of an issue than you think; it would be cheaper to put thermal, flywheel, or gravitational batteries as retrofits into substations and transformers instead, I think, or at least to put them at fixed locations like businesses which draw a lot of power. Even better, it's a free backup power supply for when the grid is down, too.

We agree on some form of distributed storage and maybe generation, though, combined with a smarter grid. It's all current or 5 years out technology.

[reitzensteinm]

To clarify, I am very much not in favor of standalone lithium ion battery packs; I'm just talking about harnessing the inevitable unused capacity of electric car batteries to smooth out the grid. The revenue from which could incentivize faster adoption of EVs, since the marginal cost of getting an EV is falling. If you had to pay for them as dedicated energy storage, it would be horrifically expensive.

As you say, many other technologies are far more suited to this kind of use.

You may be right about the cycle count; EVs are designed to be daily drivers, but charging and discharging 40-50% of a battery daily is likely more punishing than 10-20%. Though even if that is the case, it would just push my analysis 10 years into the future, when cheaper batteries and PV will make up the difference.

[uvdiv]

The peak load for South Australia was 1.8 gigawatts.

No, that's the average load at a particular time of day. (Averaged over all days over two years). The peak is about twice that:

South Australia experienced a mild summer with only a few days exceeding 40°C. A relatively short heat wave occurred in late January 2011.

The maximum demand for the year was 3,433 MW, and occurred 4:30 PM (Australian Eastern Standard Time) Monday 31 January 2011 (at a temperature of 42.9°C). A higher maximum might have been expected if the same conditions had occurred later in the week, after an extended hot weather period.

http://www.aemo.com.au/~/media/Files/Other/planning/0400-003...

[reitzensteinm]

Oops, my mistake - thanks for pointing that out!

That said, it actually makes my case significantly stronger. With 10% EVs and 20kW chargers, you can put 2.5 GW into the grid at peak, which is over 2/3 of the instantaneous maximum demand for the year.

I'd love to see figures on how much you could save providing a grid with 1.8 GW of base load vs having to engineer it to be able to handle nearly twice that, once a year.

You'd have to go to 15-20% EVs to support that, but if you're shutting down half your traditional power stations and investing heavily in solar, it may actually make sense with today's prices. Then again, lithium prices would skyrocket going to those kinds of extremes...

[SoftwareMaven]

Until fast-charge batteries are a reality (such as the carbon-laced Li-ion discussed previously here), people very much will mind their car being at 20% charge. That means they don't get to drive until the next day, except 5-7pm is key commuting time, when they need to drive now.

[reitzensteinm]

My proposal was not that people would sacrifice 80% of their battery. The first half was exploring the limits of the idea. I was suggesting 20-40%.

[nl]

Advice to entrepreneurs: finding a cost-effective way to store solar energy for 4 hours will be worth more than another 2% increase in efficiency. And inventing instant-on conventional-fuel plants will also make a huge difference in GHG emissions.

This is a solved problem. It's called pump storage[1], and is in use world wide. The US has some existing pump storage power plants (eg [2]), but typically they are used to smooth out demand on non-renewable generators (Coal, Nuclear etc).

[1] http://en.wikipedia.org/wiki/Pumped-storage_hydroelectricity

[2] http://en.wikipedia.org/wiki/Bath_County_Pumped_Storage_Stat...

[gvb]

[2a] http://en.wikipedia.org/wiki/Ludington_Pumped_Storage_Power_...

[brc]

You still need the conventional energy base. You cannot have a grid reliant on solar, because sometimes the sun doesn't shine for days on end.

Solar can top up the grid, and correlates well with air-conditioning driven demand on hot days. But it doesn't coincide with peak load - especially in winter - and will always be a diffuse power source.

You would not only need a cost effective storage solution (and none is in the foreseeable solution) you also need a lot of available space. Even if the panels were virtually free - there still remains large problems with solar power as a baseload energy source. If you solved the storage problem - cheaply - you'd still need 4 - 5 times the collecting capacity to get close to being able to reliably produce power over even a 48 hour period.

It's a great boutique energy solution, particularly for remote usage. But large scale electricity generation will remain the dominant energy source well into the foreseeable future.

[MikeCapone]

One potential solution is grid-scale liquid metal batteries:

http://lmbcorporation.com/

http://www.ted.com/talks/donald_sadoway_the_missing_link_to_...

[Andys]

I propose a new type of grid with electronic signaling of spot prices.

For heavy duty appliances (eg. air conditioning, water heating) the consumer pre-programs it with a desired average cost and a maximum price per kWh.

Then as demand and supply rises and falls on the grid, the spot price changes, and the devices receive signals and turn themselves on and off as needed. (There would be a certain minimum of fixed-price energy per month to make sure financially strained families never have to go without any)

[comicjk]

Or just install twice the solar panels, if they're getting cheap enough. Or put them 4 time zones East (they will have to float in some cases) and send the electricity West.

[grecy]

Better yet, go crazy with daylight savings, moving everything by 5 hours.

The sun will be high in the sky at 5-7pm, generating power when we need it most.

[nl]

Or put them 4 time zones East (they will have to float in some cases) and send the electricity West.

This is a good idea, except electricity transmission costs are significant over long distances like that.

[pippy]

Capturing energy should only come till after solar PV's generates surplus electricity. If solar panels continue to become cheaper and more efficient it's likely no revolutionary capturing technology needs to be developed. There are plenty of simple, cost effective energy capturing technologies currently in use today (see: http://en.wikipedia.org/wiki/TV_pickup)

[Daniel_Newby]

> The difficulty is that, unless storage gets dramatically better, ...

Refrigeration! Put all the air conditioners and refrigerators on timers. Run them hard during peak solar hours. Then turn them off during the evening peak. If that would make the buildings too cool, make ice and melt it later to chill the air. (Some event venues already do this to handle air conditioning ten thousand people during midday.)