Autolib in Paris Shows How Electric Car Sharing May Not Be Green

By · October 02, 2013

One Autolib car in Paris

One Autolib car in Paris

A lot of people are excited about car sharing in Europe. It's part of the "sharing economy," in which individuals belong to an organization that provides access to services (such as mobility), without having to own things (like a car). In other words: access versus ownership. But one question needs to be asked: Is it green?

It sounds like a simple question with an obvious answer: Is there less energy consumption with an EV sharing service than with individuals driving their own gas cars? After a close look at the Parisian Autolib service, the answer is not so obvious.

An Autolib rental kiosk

An Autolib rental kiosk in Paris.

People only see the cars on the street, but they're the tip of the iceberg. There are Autolib stations (parking and charging only) and Autolib kiosks that offer subscription facilities. There are also facilities with people providing information. All those are connected 24/7 with the Autolib HQ, where there is a call center and many powerful computers to constantly monitor the fleet of nearly 2,000 cars.

In addition, the service uses diesel trucks—they can be seen everyday in Paris—taking cars away for maintenance. All in all, this is a large organization, with more than 500 employees. Everybody knows that electric cars are very efficient, with have very low energy consumption, but in assessing the service's virtues, it's important to account for the energy consumption in every part of the operation—down to the smallest detail, and not only the cars.

Autolib cars waiting for drivers

Autolib cars waiting for drivers.

Even the cars might not be so green. The Bolloré Bluecar uses lithium metal polymer cells, made by Bolloré itself. These cells have been around for nearly 10 years. Many European car manufacturers had a look at them, but none bought any. They have a problem in the sense that they are "warm cells," which do not work at ambient temperature. They need to be warmed up, and kept warm at all times. It's not much of a concern for car sharing use, where cars are only driven for a short period of time, before quickly going back to a plug. But it would matter for a private automobile.

The battery of a parked but unplugged Bolloré Bluecar goes from fully charged to totally discharged in less than 72 hours! To remain at operating temperature, the battery draws current from its plug at all times, and with 2,000 cars plugged, this consumption is significant. How significant? Enough to be worth hiding, as I asked Bolloré three times about the subject, and they replied three times that they don't communicate on this matter.

A conservative estimate would put the electrical consumption of the entire service at more than five times the electrical consumption of the cars when driving, and that alone is enough to question the greenness of Autolib car sharing.

In European cities though, the concept remains valuable because of high population density. Parking is so scarce in Paris, that the idea of reducing the number of cars makes sense. But it's awful to learn that it may end up rising energy consumption. At least, this calls for precise tools to assess, measure and control car sharing services. It's especially important considering most car sharing services receive some kind of public support. The Bolloré group is about to float 10 percent of "Blue Solutions" unit, which groups its various battery activities other than Autolib, on NYSE Euronext. Investors in EV stocks should not expect with Bolloré the same success they've enjoyed with Tesla shares.


· · 4 years ago

Very interesting report, Laurent. As you know, I've been intrigued with the Blue Car (more so the automobile itself as opposed to the car sharing business model) for some time now.

Boy, I'd really like to know more about these lithium metal polymer batteries! With the exception of the very high self discharge characteristic (might be something that can be solved, as Sanyo did with Eneloop consumer device rechargeable NiMH cells,) I'm intrigued about lithium cells that actually LIKE the heat. Sounds like a battery chemistry that would work very nicely down here in the desert. Meanwhile, Bollare could swap out the LMP cells in Europe with something more suiting the indigenous climate.

· · 4 years ago

Why the random and incorrect attack on car sharing? Having employees doesn't make a service not green, those people were going to exist anyway. I'm glad you investigated Bollore's "warm battery" issue, I wasn't aware of that. And and of course it makes sense to add up everything else. But obviously at the end you have to compare the car share social cost (in which each car supports something like 20 users, at least here in Philadelphia), to the equivalent social cost if there were no car share (i.e., many electric vehicles). I think it's obvious that the environmental benefit of reducing the number of cars in the program by a large factor trumps the impact of a few servers and kiosks, and while I don't know the details of this battery warming scheme, it sounds it's unlikely to raise the impact so high. Bad analysis Laurent!

· · 4 years ago

the batterie is great for middel east, saud arabia ...or arizona.

· · 4 years ago

Articles which deal with numbers should source them, as evaluation is otherwise impossible.

I have dug out the operating temperature of the battery:

It is 60-80C.

The claimed, and unsourced, discharge in 72 hours would amount to an energy drain of 416 watts per hour for the 30kwh battery pack.

I suppose I could do calculations on the battery mass to work out whether that is a reasonable claim, but it is Laurent's thesis, not mine.

In any case, that amounts to a drain of 10kwh per day.
That is enough to take you in an electric car 30 miles or so.
These are high use shared cars, so that is a pretty minimal estimate for daily mileage.

In fact, as Laurent says:
'They need to be warmed up, and kept warm at all times. It's not much of a concern for car sharing use, where cars are only driven for a short period of time, before quickly going back to a plug. But it would matter for a private automobile.'

So it is 'not much of a concern', as they aren't for private use anyway.

So we are talking at MOST of a doubling of electricity use due to constant charging.

Somehow, from this, Laurent by means unshown arrives at the claim that:

'A conservative estimate would put the electrical consumption of the entire service at more than five times the electrical consumption of the cars when driving'

So I await with great interest the demonstration of this, although it is Laurent's usual practise to simply disappear when his claims are challenged.

If one wishes to talk about total energy costs, and look at the manufacturing energy for the battery and so on, then the energy embodied in cars not produced for at least some of the folk who would otherwise own one, were it not for car-sharing, has to be accounted for.

I think in view of this the likely result of full energy accounts are obvious to most of us.

· · 4 years ago

Bollore batteries
cathode is LiFePO4
but note well, even though the blue car has a 30kWh battery pack, its power is augmented by capacitors.

· · 4 years ago

Actually, thinking about it a bit more, the claim that:
''A conservative estimate would put the electrical consumption of the entire service at more than five times the electrical consumption of the cars when driving''

Is materially impossible, unless the cars spend most of their time standing there unused.

It is just possible that the total ENERGY use, including things like producing the battery, is 5 times the amount of energy used in the car running around, but total electricity use, no way.

· · 4 years ago

Yes five times. There's more than the self-discharging batteries. There are offices, large servers (computers), diesel trucks... More than 400 people work for Autolib. My estimate is low, others see it as even higher.

· · 4 years ago

So, where are your figures and sources?
You make large claims, with no back-up whatsoever.

I have provided a detailed breakdown, you have provided nothing.

No wonder Bollore ignore you.

· · 4 years ago

I should add that a substantial amount of the downtime when the car is draining power and not being used will be overnight.
France gets ~75% of its energy from nuclear, which can't all be used at night times.
The marginal cost of this power is truly tiny, as the costs of nuclear are in building the plant, and you might as well carry on running it as the cost of fuel per kilowatt hour is minuscule.

So a lot of the energy used for heating the pack, likely at least half as for perhaps 6 hours almost all the fleet will simply be charging, not running, will be at minimal cost with almost no impact on CO2 emissions or anything else.

· · 4 years ago


You don't need to make things personal. There are more and more critics regarding Autolib in France, and I'm not the only one asking embarrassing questions Bolloré doesn't want to answer. I have data and pictures to prove my point, there are just outside the scope of this article.

· · 4 years ago

You are the one giving embarrassing non-answers when asked to back up your claims.

If I can't back something up, particularly highly damaging claims, I retract.

You just say that you could back them up if you felt like it but can't be bothered.

That is what you call personal.
It is also weak.

New to EVs? Start here

  1. Seven Things To Know About Buying a Plug-In Car
    A few simple tips before you visit the dealership.
  2. Incentives for Plug-in Hybrids and Electric Cars
    Take advantage of credits and rebates to reduce EV costs.
  3. Buying Your First Home EV Charger
    You'll want a home charger. Here's how to buy the right one.