Arguments about which technology is the most energy and carbon efficient over their entire lifetime are good ones to have. This is where the conversation should be focusing, not rehashing questions that are not currently scientifically controversial. But the debate about life cycle efficiency is complex, and often gets abused or misunderstood. We face these questions from biofuels to solar, wind energy, and all-electric vehicles.

With regard to electric vehicles, for example, it is not enough that they do not emit carbon from a tailpipe. We have to consider the energy used and carbon released during the entire manufacturing process, including sourcing the raw material. We also need to consider the source of energy used to charge the vehicle, and what happens to the battery at the end of its lifetime. This is a difficult assessment to make, and every study that attempts to do so must make a number of assumptions which affect the outcome.

The result is many studies with a range of outcomes based on different techniques and assumptions used. This is a common situation in science, and what is typically done is to look at the full range of study outcomes, which should follow somewhat of a bell curve of results, and see where the peak is. It does not make sense to rely upon individual studies that are out by either tail – these are literally outliers. So bottom line – what do these studies show? They indicate that over the entire lifetime of an electric vehicle, under most driving conditions, they produce less carbon than an average gas vehicle, and hover around the efficiency of a gas-electric hybrid. The greatest individual determining variable is the source of the electricity (“fuel cycle” in the chart).

Let’s break this down a bit. Overall electric vehicles are more fuel efficient than gas cars. It’s hard to make a direct comparison because the “fuel” they use is electricity, but if we gauge this by the total energy used to go a mile or kilometer, electric cars generally win. And of course an all electric car emits no direct carbon. The cost of manufacturing the car itself is about the same, but the electric cars have the extra cost (carbon and energy) of manufacturing the battery. In the chart above we are looking at the Nissan Leaf, which is consider at the low end of electric vehicle with a relatively small battery. The chart looks similar if we compare a Tesla Model 3, but the light blue contribution is about double, as these are longer range cars.

So that is one variable – how big is the battery? The bigger the battery, the farther you have to drive it to make up for the extra manufacturing cost. The chart above assumes a 150,000 kilometer lifespan, which seems low to me. The major limiting factor for electric cars is the battery life, which has been getting better over time. Conservatively current Tesla batteries have about a 500,000 mile lifetime, but Tesla claims their latest batteries should best a million miles. This matters because most of the carbon cost of electric vehicles is upfront, and their net carbon benefit gets better the longer you drive them.

But the biggest variable, the yellow part of the bars above, is the fuel cycle – how is the electricity made? As you can see, in Norway there is essentially no carbon from the fuel cycle, because they are all nuclear and hydroelectric. France also does well with its large nuclear infrastructure. Germany, who tried to get rid of its nuclear power and ended up having to build coal-fired plants, does the worst.

This raises an important point – electric vehicles by themselves, even in the worst-case scenario, are better than average gasoline fueled vehicles, and about the same or a little worse than hybrids. But in the best case scenario they are much better. As a strategy for reducing global carbon emissions, however, they need to be just one part of an integrated whole. We do need to understand the contribution of each piece, but we also need to see how those pieces interact. Electric vehicles will become more and more carbon efficient as the energy infrastructure gets more carbon efficient. The more countries go the way of France and Norway, the greater the benefits of EVs will be.

This, of course, brings me back to an argument I have been making for a while – we need nuclear energy as part of our decarbonization strategy. Those countries who have embraced nuclear are simply doing better than those who have rejected it. Further, as use of EVs increases this will magnify the effect. Widespread adoption of EVs will dramatically increase electricity demand. If we meet that demand by burning more coal, there will still be a net advantage to EVs but it will be much smaller than if we meet that demand with carbon neutral sources.

Also, there is good news in all of this – electric vehicles and battery technology are still on the steep part of the progression curve. As battery tech improves, the carbon advantage of EVs will also increase. We also have to consider what happens to those batteries at the end of the vehicle’s lifespan. One great possibility is to use them for grid storage. This will extend their useful lifespan, and further improve the overall carbon efficiency of the system.

The overall point that I want to emphasize is to beware of anyone who is cherry picking individual studies or a limited analysis. We need to take a comprehensive look at all the evidence across the entire life cycle of any technology, and how that technology interacts with other variables, to develop a thorough analysis and make it part of a coherent strategy.