BMW i3 – Lifetime Carbon Emissions Are Lower Than Conventional Car


BMW Carbon Fiber Production

BMW Carbon Fiber Production

SGL Facility

SGL Facility

Plenty of fuss has been made over whether or not carbon fiber,  EV batteries, etc. are eco-friendly.  The argument in favor of or against certainly won’t be put to rest here.

Here’s what BMW’s Klaus Draeger, head of purchasing has to say on this topic in regards to the BMW i3:

“If you take together the battery, the carbon fiber, the aluminum and what is on the car, you start slightly worse than a conventional car.”

“But by using much less energy during driving, the carbon emissions over the lifetime of an i3 is better than a conventional car…and if you drive the car during the lifespan with renewable energy it is actually much better.”

Focusing only on the carbon fiber for the BMW i3 and i8, Ward’s Auto states:

“…the automaker plans to use all 9,000 tons (8,165 t) of carbon fiber annually set to come from SGL in Moses Lake.”

Draeger adds:

“Our prognosis currently shows that we will need all this we are producing here. But, of course, if there were to be some overcapacity, we could think about (sales to other automakers) as well.”

Check out the Ward’s Auto article at the link below for more in-depth analysis on the carbon fiber used in the BMW i3.

Source: Ward’s Auto

Categories: BMW

Tags: , , , ,

Leave a Reply

21 Comments on "BMW i3 – Lifetime Carbon Emissions Are Lower Than Conventional Car"

newest oldest most voted

Next step – sequestering carbon from emissions into Carbon fiber 😉

Isn’t that happening already or are the carbon atoms already there as part of the initial fibres being converted?

How much CO2 is created to ship the carbon fiber from Moses Lake, WA to Leipzig Germany, then ship the cars back to the US?

Excellent question, and a good example of why doing a full LCA (life cycle analysis) for greenhouse gas emissions or water consumption or energy consumption or whatever can be quite a challenge for a complex product like a motor vehicle that consumes a non-trivial amount of resources post-manufacture.

Hard to say. If there is a half empty ship that is going to embark anyways, just adding on these extra cars may not be a huge difference. The weight of the cars relative to the weight of the ship is sometimes pretty small.

This came up in our Facebook group. There is data on this. Generally, ocean shipping has low CO2 footprint per ton and mile traveled. Look it up. It’s about 1/10 of the impact large catgo trucks have. The contribution of ocean shipping on the i3 lifetime CO2 emissions is insignificant. Trucking is much worse. BMW has recently switched the delivery of newly produced vehicles from Leipzig to Bremerhaven from trucks to fully electrified trains. It’s in BMW’s sustainability report.

Where does the carbon in carbon-fiber actually come from? If they can sequester it from the atmosphere, that would actually be pretty cool! That way the manufacture of the car actually reduces greenhouse gasses!

“After drawing or spinning, the polymer filament yarns are then heated to drive off non-carbon atoms (carbonization), producing the final carbon fiber” from Wikipedia. This suggests to me that the process will produce emmissions, not reduce.

The carbon emissions from the generation of the energy required to make the CF is far, far higher (20:1 or more) than the amount of carbon in the CF.

These calculation does not bear much relevance, because the portion of solar powered EV charging is increasing exponentially. Therefore in five or ten years, very significant amount of EV charging is directly solar powered, e.g. workplace charging.

And if Tesla had started a carbonfiber gigafactory construction today, it would most likely powered by wind and solar only using mostly locally produced electricity.

As with aluminum smelting, these factories need to be located next to hydro power plants. Hydro is much cheaper than solar and wind, is not intermittent, and can run 24/7.

In southern latitudes, solar power is well enough predictable that intermittency does not matter. Production rate can be adjusted on the availability of solar power and the rest can be filled from the grid.

The only problem is that today solar power is too expensive that it cannot compete e.g. with hydroelectric power, but this should change in five to ten years. Actually new hydro projects are not very cheap as they require too long investment period that long term returns are very uncertain, because any technological breakthrough in solar panels or batteries will undermine the capital expenses of hydroelectric power.

Being next to a hydro plant does almost nothing, because all electricity on a grid is shared. Hydro is a limited resource. Take away BMW’s plant, and that hydro power is used elsewhere, making a coal plant reduce output.

BMW’s CO2 footprint per kWh used is the same as that of any other company on the west coast, and definitely not zero.


… Tesla doesn’t have the know how to produce a carbon fiber “gigafactory”.. They could make a factory it would be a “microfactory” cause they’d be cranking CF out one piece at a time by hand at a couple thousand dollars each just like all other manufactuers when they use the stuff. Take the carbon fiber VW XL1 for example.

That’s the significance into BMWs research into mass producing and molding CF.

Although… that last quote “… we could think about (sales to other automakers) as well.” is interesting. He basically hinted/implied that someone has approached BMW about procuring low cost CF. My guess would be Tesla.


I’ve still never really seen a proper apples to apples manufacturing comparison of petrol and EV cars. Every time I read one of the studies in great detail, I see one side or the other being penalized by a step in the process that isn’t accounted for in the other. I just read on that heaped the extra energy required to obtain the lithium on to the EV (reasonable), but didn’t account for any demerit to obtain the platinum or palladium for the petrol catalyst, even though they are mined together in some cases! And show me a study, please, that doesn’t assume carbon emissions scale linearly with pack capacity (in reality, it does not scale linearly). Or they account for a battery replacement at 100,000 miles, but don’t account for the percent of engines that will also fail and require replacement at that point (not all, but not zero). It’s just frustrating to make heads or tails of the claims.

Tesla has promised for analysis about the lifetime carbon footprint now for more than a year. But it is still waiting itself. However, JB Straubel once said that payback period of Tesla S is less than 10 000 miles, depending on grid mix used for charging.

My rough guess is that ecological footprint of Model S batteries is less than one fourth compared to Roadster batteries. This is because the energy density is almost doubled and the amount of cobalt used is significantly reduced. There are also lots of other efficiency gains in manufacturing and gigafactory will provide further efficiency gains not least because it is 100 % solar/wind powered. Also recycling of batteries will provide significant savings in resource consumption.

All these points above are ignored in most of the life cycle calculations, because those who are making the calculations are looking the past, but they fail to see how things are after 10 years in the future.

What is the point of comparing to conventional cars ? We want to see carbon fiber vs if i3 was made just of aluminum/steel.

It doesn’t make sense to make i3’s out of CFRP, because the use-phase energy consumption of a BEV is just not that sensitive to mass changes, so you never get back the extra energy you put into the CF.

It makes even less sense under a scenario in which the energy is very low-carbon, because this makes the use-phase emissions reduction even smaller.

It is worthwhile to compare EVs to conventional cars, though, particularly cars fueled with bio-fuels, if only to understand the role materials and fuels play in a vehicle’s overall impact.

For example, a conventional, mild steel IC car fueled by ethanol made from Brazilian sugar cane could have a total life-cycle CO2 impact that is less than the impact of just manufacturing the i3, because the ethanol has such a low (relative to gasoline) impact, and the CFRP has such a high (relative to steel) impact. Even if the electricity for the i3 is from renewables, you never recover the additional emissions from making the CF.

That assumes that the carbon storage value of the tropical forest that was cleared to make room for the sugar cane production for the ethanol is zero.

I’m guessing that’s not a correct assumption.

Ethanol was pushed in Brazil not out of concern with CO2, but because Brazil was thought to have little oil till the recent offshore finds. It was a strategic energy independence decision, not an environmental one.