French large scale battery test

  • Active since 1995, Hearth.com is THE place on the internet for free information and advice about wood stoves, pellet stoves and other energy saving equipment.

    We strive to provide opinions, articles, discussions and history related to Hearth Products and in a more general sense, energy issues.

    We promote the EFFICIENT, RESPONSIBLE, CLEAN and SAFE use of all fuels, whether renewable or fossil.
Status
Not open for further replies.

EatenByLimestone

Moderator
Staff member
  • Like
Reactions: Cynnergy
The article indicates that the batteries do work to even out the grid but currently are too expensive, except for some parts of Germany and Denmark in the Euro zone. The tipping point on current economic viability is at electric rates of about 30 euro-cents/kwy, or about $0.35/kwh US. What was also interesting is power management with customers to shift usage. When I talked to our utility about better home power management on a voluntary basis, the utility had no interest.
 
The high nuke % on the French grid will make RE integration harder, due to its lack of throttle-ability. Indeed, high nuke penetration is almost as difficult as high RE penetration, when it comes to load matching, but for the opposite reason. Most pumped storage facilities around the world were built for diurnal load balancing nukes.

Bottom line: it makes sense the French would be on the vanguard for advanced storage technology.
 
The article indicates that the batteries do work to even out the grid but currently are too expensive, except for some parts of Germany and Denmark in the Euro zone. The tipping point on current economic viability is at electric rates of about 30 euro-cents/kwy, or about $0.35/kwh US. What was also interesting is power management with customers to shift usage. When I talked to our utility about better home power management on a voluntary basis, the utility had no interest.

What kind of home power management are you talking about? Using power while an array is producing or something else?
 
The kind of power management talked about in the article linked by the OP:
The Nice Grid pilot has also experimented with "demand response" systems to use discounted tariffs to encourage citizens to use more electricity when sunshine is abundant and less during winter evening demand peaks.

Some 200 households signed up to let ERDF temporarily switch off their heaters or hot water boilers during winter evening peak consumption using the new "Linky" smart meters which France plans to roll out nationwide in coming years.

Another 70 customers signed up to receive an SMS warning the day before an expected sunny day so that they can benefit from midday power prices that are 33 percent lower.

"I switch on my washing machine when the sun shines," Carros resident Lara Muzzarelli told reporters in her red-stone villa.
We offered to do similar kinds of things. Washing and drying clothes, operating the dishwasher, using hot water, even many kinds of cooking and oven uses, and much more can be scheduled to shift demand to even out the grid.
 
I imagine tiered rates are a great motivator for that. I don't know that I'd want a smart meter though.
 
Our power co tried offering incentives to reduce power consumption... and they worked... too well. They said they had to raise rates due to the "unanticipated" drop in revenue.
 
  • Like
Reactions: Woody Stover
The way it was explained to me was that regional governments long ago entered into exclusive franchise contracts with private companies to supply electricity to all parties within that region. Governments named them "Utilities". The governments prevented any competition, gave the companies rights to easements against private property, blanket exemptions from local building codes, but retained the right to regulate the company's profits.

So is it any surprise that when the government set up incentive programs that would be harmful to the Utilities profits, that the government would not let the Utilities recover thru other means? They are longtime bedfellows.

What other industry promotes the slogan: "Don't buy my product, I'll give you money!" !!!
 
The article mentions LiIon... I wonder if the economics would be more favorable if they used old fashion lead acid cells?
 
The article mentions LiIon... I wonder if the economics would be more favorable if they used old fashion lead acid cells?

When I was looking into power outage backups, I did some pricing comparisons for each. The lithium ion batteries cost a lot more up front, but they've reached the point where the manufacturers are now claiming they'll last quite a few more cycles than the lead acid. The result is the life cycle costs of the two seem to be very close - around $0.50/kWh.

However, the utilities are often more interested in the battery power rating (Amps or Watts) than the energy storage rating (kW hours). The way they're currently used is not to store large amounts of energy, but to handle the difference between the constantly varying demand for electricity and the constantly varying (especially with wind and solar) supply.

Historically, they've handled that demand by keeping extra power plants running below capacity, ready to ramp up in seconds. This "spinning reserve" reduces the total efficiency, and those short moments when the ramp up is necessary results in the effective cost of those kWh delivered to meet the change in net demand having a very high effective cost.

The huge current capacity of lithium ion batteries gives them the advantage in trying to displace that excess spinning reserve to meet those short spikes in net demand. So if you need 1 MW worth of ready reserve, you might be able to get that with 0.5 MWh worth of lithium ion batteries, instead of 1 MWh worth of lead acid (very rough ballpark estimate). A few minutes may be enough for a demand spike to end or for an offline natural gas plant to start producing power.

I was surprised to see this recently in a consumer product - portable jump starters. I've got one of the heavy lead acid ones that's been around for years. I just looked up the price on it - $120, for 300 cranking amps. It weighs 18 pounds and has a 17 Amp-hour battery in it.

Lowes now has a a lithium ion unit for $75. It only has a rated capacity of 10.8 Amp-hours, but can supposedly provide 400 cranking amps. I don't think it weighs more than 4 pounds, although it looks like the jumper cables are seriously undersized.
 
Iam, your comments pushed me into a better understanding of the grid power supply/demand and the challenges, real or exaggerated, of solar/wind RE. On a micro scale, yesterday my PV system power moved from 6.7kw to 0.4kw as clouds passed overhead within the 15 minute logging interval of my monitoring system. That move likely was much faster, down and up (maybe several times in the 15 minute interval) and yesterday was a day with rapidly alternating bright, clear sky and large dark clouds. My system was rapidly pushing power into the grid and then my house demand was rapidly taking power from the grid.

So, how does the grid portion in my neighborhood, which I understand to be a 2400V distribution line with linear transformers supplying each house or small group of houses, handle my micro rapid changes in power supply/demand? Theoretically, my PV high power output would raise the grid voltage and my PV low power output would drop the grid voltage. Since my system impact remains "small," even in my neighborhood, those changes in voltage also would be small and of little consequence to any user. But if everyone in my neighborhood had PV, and disregarding the next level up utility distribution line, those changes could be material, enough so as to result in an excessively high or low grid voltage which could adversely affect devices drawing power from the grid.

I have a micro-inverter system, and I understand that if grid voltage gets too "high" or "low," my micros will shutdown, both to protect the micros and I assume the grid itself.

Iam, based on your comments, is this the point where rapid response batteries come into play? ... to quickly absorb excess power and just as quickly supply additional power when needed? BTW, those li-ion starting batteries are pretty amazing.
 
When I was looking into power outage backups, I did some pricing comparisons for each. The lithium ion batteries cost a lot more up front, but they've reached the point where the manufacturers are now claiming they'll last quite a few more cycles than the lead acid. The result is the life cycle costs of the two seem to be very close - around $0.50/kWh.


Interesting, I did not realize that the cost gap had closed so much... although I guess that is not surprising now that I think about it considering how far Li battery prices have fallen.

I was honestly suprised when Tesla came out with the power wall... as traditionally LiIon' biggest benefit has been the high energy density - making it a natural evolution for traction power applications in any vehicle, but not as big as a benefit in static power reserve where the weight of lead batteries is irrelevant and they are so cheap you can just compensate by buying more capacity.

Now here is the part I dont quite follow... I know its easy to stretch LiIon batteries to very long cycle lifes by limiting the discharge depth, and never fully charging (using just the middle from 25%-75%). Ive seen references that using only 50% of capacity this way you can boost cycle life from the typical 300-500 in consumer devices up to a couple thousand (Ive seen 4k referenced as the theoretical ceiling with charge capped to 3.92v/cell or 72%). This is better than lead which might manage 1000 at 50% DoD....

But what about calendar life?.... deep cycle lead optimized for storage like the kind sold for off grid solar can last 5-10 years or more... but LiIon seems to loose capacity with age a lot faster and even treated gentle most of the commercial batteries ive seen struggle to last longer than 2-3 years.
 
Last edited:
Interesting, I did not realize that the cost gap had closed so much... although I guess that is not surprising now that I think about it considering how far Li battery prices have fallen.

I was honestly suprised when Tesla came out with the power wall... as traditionally LiIon' biggest benefit has been the high energy density - making it a natural evolution for traction power applications in any vehicle, but not as big as a benefit in static power reserve where the weight of lead batteries is irrelevant and they are so cheap you can just compensate by buying more capacity.

Now here is the part I dont quite follow... I know its easy to stretch LiIon batteries to very long cycle lifes by limiting the discharge depth, and never fully charging (using just the middle from 25%-75%). Ive seen references that using only 50% of capacity this way you can boost cycle life from the typical 300-500 in consumer devices up to a couple thousand (Ive seen 4k referenced as the theoretical ceiling with charge capped to 3.92v/cell or 72%). This is better than lead which might manage 1000 at 50% DoD....

But what about calendar life?.... deep cycle lead optimized for storage like the kind sold for off grid solar can last 5-10 years or more... but LiIon seems to loose capacity with age a lot faster and even treated gentle most of the commercial batteries ive seen struggle to last longer than 2-3 years.

You are the battery expert Jeremy, but re Li-ion cells I think the cycle life depends a lot on the details of the electrode construction. I think there has been progress since your specs. For currently fielded cells, I hear 1000 full cycles (at not super high rates) with acceptable degradation (<=20% SOC), and more like 2000-3000 cycles when you 'baby' the cells (at both high and low SOC). The 1000 number floor is useful for EVs since you just multiply the electric range by 1000 to get the min battery lifetime in mileage....a 85 mile LEAF with current batteries should get to or exceed 85k miles before the battery starts to act beat. The 200 mi Tesla will prob have a happy battery until you approach 200k miles, etc. Both use most of the SOC range. For the Volt, given the smaller cells, they DO baby the cells, since with a 30-40 mile electric range, they might be asked for a few thousand cycles by 100k miles.

Lots of misunderstanding out there...people think the Volt batteries and Tesla batteries are 'better made' than the LEAF ones....not necessarily....in the Volt they HAD to engineer it to get better cycle number (and could settle for lower range), in the LEAF they needed to max the range (and accept some battery degradation) and in the Tesla, the range is so huge no one can actually run up enough cycles.

As for calendar life, it depends of course on both temperature (hot is exponentially bad) and SOC, with calendar life at middling SOC better than calendar life at high or low SOC. This whole function depends in detail on the chemistry as well. Thus the use of temperature management systems (Tesla and Volt) or special chemistry to be more heat tolerant (the so-called 'lizard' chemistry standard in the 2014 and later LEAFs). Calendar life is supposed to be better than best Lead acid. The 10 year warranty on the Power-wall product (for diurnal storage) suggests that it can handle 3000+ cycles AND 10 years calendar life...its smaller capacity relative to the backup product suggests they are babying the cells SOC, of course.

It is also obvious that EV batteries were/are sold before their calendar and cycle performance were adequately known. I was told not to quick charge (1-2C) my 2013 MY Leaf 'too often', like more than once a week. Now Nissan says thats not a problem at all, so long as I don't overheat the battery.

As for cycle cost....currently at $200/kWh capacity and 1000 cycles....that is $0.20/kWh, a lot better than AGW. Expected to keep dropping with learning curve it could drop below 10 cents/kWh in a few years.
 
Last edited:
But what about calendar life?.... deep cycle lead optimized for storage like the kind sold for off grid solar can last 5-10 years or more... but LiIon seems to loose capacity with age a lot faster and even treated gentle most of the commercial batteries ive seen struggle to last longer than 2-3 years.

I didn't find good data on that, but as woodgeek noted, they seem to have gotten pretty good in recent years, with some offering 10 year warranties.

Keep in mind also that there are several different types of lithium ion batteries which tend to trade off capacity or amp ratings for reliability. A lithium-iron phosphate battery in a car might get 10 years no problem, while a lithium-manganese oxide cell phone battery struggle to get 2 years. However, while there are real life span differences between the various lithium ion chemistries, note also that a big part of that is also the different usage (which again, woodgeek also alluded to). A smart phone battery will probably be almost fully cycled daily - 700+ times in that 2 years, while a Nissan Leaf battery would likely get half cycled 2000+ times in its 10 years.

As far as $200/kWh prices - I've seen that figure tossed around, but it's not clear to me that you can actually buy a lithium ion battery for that price at the moment. Apparently it comes from Tesla saying replacement batteries for the Model S would be $17,000, which for an 85 kWh battery, is $200/kWh, but that doesn't mean it's what it cost them to make batteries now. Elsewhere, I've seen the figure as a cost target for when Tesla's battery factory is up and running which makes sense.

A recent study indicated current prices for the market leaders are more like $300/kWh, and for lower volume users, more like $400-500/kWh. The Powerwall is priced $428/kWh.

Regardless of the battery tech in question, the battery isn't the only cost, as there's also installation and power conversion (for a home, an inverter) to consider. At the utility scale, the pricing gets more hazy because they have better economy of scale, but have to integrate with a much more complex grid. I'm loathe to use this example, but since I have the numbers handy - my own utility just installed a "large" (for this sort of project) grid storage battery system with a capacity of 7400 kWh for $16 million - that's $2160/kWh.

I'm loathe to use the example because that price appears to have been artificially inflated by corruption - It recently came out that the company hired to do the installation is owned by a personal friend of the PUD executive. He was previously hired on to the PUD by that executive into a special position created for him and wrote their specification for a grid storage system, then left the PUD to start the company as a reseller of other companies' batteries and was very shortly afterward awarded a contract without a competitive bid for the spec he wrote. That's not even the entirety of the shady details. Somehow, all that's happened so far, however, is the PUD hired an "independent" lawyer to investigate whether there were any "inappropriate activities" at the PUD, who unsurprisingly released a report than concluded the agency paying her fee was innocent.
 
Iam, based on your comments, is this the point where rapid response batteries come into play? ... to quickly absorb excess power and just as quickly supply additional power when needed? BTW, those li-ion starting batteries are pretty amazing.

Yep. That's a big part of it. A steam plant can take hours to ramp up. A shut down natural gas plant takes several minutes, at a bare minimum. An idling natural gas plant can respond in less than a minute. So if you can gain 15-30 minutes worth of storage, you might be able to switch a standby gas plant from idling to off, saving fuel.

Even without the batteries, some amount of change can be absorbed by letting the voltage vary, which is why part of why the grid worked before batteries started getting installed. 120 VAC power is not exactly 120 volts. Equipment is supposed to be designed to tolerate +/- 10%, and the utility voltage, to have some extra margin, is allowed to vary +/- 5%. That means 114 to 126 volts.

If you consider an ideal resistive load (which the grid isn't, but this isn't a serious engineering discussion), that voltage change results in the power consumed by the load changing by +/- 10%. That gives your gas plant or battery at least a little bit of time to start delivering power. So if you're looking at a region consuming 1 GW of power, it can potentially lose 100 MW of production without immediate problems (with inductive loads and power managed industrial equipment, the real tolerance is significantly lower).

If you and 1% of Minnapolis have 8 kW arrays that get overshadowed by a giant cloud, thinks are are probably ok without backup, as long as the rest of the supply and demand picture doesn't change, but if 10% of the households in Minneapolis have 8 kW arrays and the same thing happened, it's probably going to cause brownouts.
 
If you and 1% of Minneapolis have 8 kW arrays that get overshadowed by a giant cloud, things are probably ok without backup, as long as the rest of the supply and demand picture doesn't change, but if 10% of the households in Minneapolis have 8 kW arrays and the same thing happened, it's probably going to cause brownouts.

Is that much different than the effect if 10% of the grid users cycled their electric dryers and electric ovens (or electric central heat) on or off in the same time window as the cloud passing over?

With my 4.4kW array, I actually alter my load on the grid more by flipping my 5.4kW dryer ON, then stopping it to toss in a dryer sheet, then flipping it ON again at 10PM, than I do by having a cloud pass over my house at 1PM.

Running throughout the day, my grid voltage varies from 238V-248V, and I am attached to the last transformer on this dead end spur of the grid.
 
Is that much different than the effect if 10% of the grid users cycled their electric dryers and electric ovens (or electric central heat) on or off in the same time window as the cloud passing over?

Somewhat. It's overwhelmingly unlikely that everyone will turn their dryer on at the same time. People do turn their heat on at roughly the same time (when they get up in the morning and when they come home in the evening), but the difference there is it's a predictable demand that the power company can start preparing for hours ahead of time.

Clouds covering large expanses of solar or a sudden calm hitting a major wind farm can be just as rapid and less predictable.

That said, there's been a lot of ongoing discussion about this by grid planners, and the general consensus seems to be that this won't become a major problem until we get up around 20% of the power in a given region coming from solar and wind. There's a lot of room to grow before we should need to start think about really large amounts of storage capacity.

Also, I like to point out that some regions will have a lot more flexibility than others - in hot climates, peak solar output usually tracks fairly closely to peak electricity demand. There's a few hours difference, as the air conditioners really get cranked up in the late afternoon, but even by then, solar panels can still be producing a lot of energy, and modest behavior changes, particularly encouraging pre-cooling of buildings with time-of-use billing can go a long ways towards smoothing out that difference even without enough batteries to cover hours worth of demand.

In contrast, living in a cool climate like I do, people start getting up at 6 or 7 AM and usually have the heat set to come on slightly before that, but in the middle of the winter, the sun doesn't come up until almost 8 AM. It will be much more difficult to integrate large amounts of solar in such areas.
 
The high nuke % on the French grid will make RE integration harder, due to its lack of throttle-ability. Indeed, high nuke penetration is almost as difficult as high RE penetration, when it comes to load matching, but for the opposite reason. Most pumped storage facilities around the world were built for diurnal load balancing nukes.

Bottom line: it makes sense the French would be on the vanguard for advanced storage technology.
Take a look at http://www.gridwatch.templar.co.uk/france/ - lots of nuclear power, but they cope with the difficulty in load-following by having a lot of interconnectors and a lot of Hydro. Their demand is also surprisingly flat - presumably the result of metering decisions over the years.
 
  • Like
Reactions: woodgeek
Status
Not open for further replies.