How muddling facts with beliefs devalues a scientific study

Today we take a critical look at the study „Electricity Generation in LCA of Electric Vehicles: A Review“ from Benedetta Marmiroli, Maarten Messagie, Giovanni Dotelli and Joeri Van Mierlo.
Starting with rock-solid analyses, the paper of August 2018 [i] initially arouses great hopes. The methodological approach of 44 (!) scientific publications is meticulously examined.  The authors check closely whether the objectives of the investigations are stated, the type of LCA is named (Attributional Life Cycle Analysis or Consequential LCA), whether current data are used and on which electricity mix the analyses are based. Due to unclear goals and widely differing assumptions they note large deviations:

The paper sets out the reasons why practically every claim about BEV’s greenhouse gas emissions can be supported by scientific studies, and describes under which conditions life cycle analyses are meaningful:

According to Weidema, a consequential approach is a “system modelling approach in which activities in a product system are linked so that activities are included in the product system to the extent that they are expected to change as a consequence of a change in demand for the functional unit.“
A CLCA is basically concerned with identifying the cause and effect relationship between possible decisions and their environmental impacts. The cause and effect relationships are based on models of equilibrium between supply and demand, borrowed from neoclassical economics. In practice, this consist in the identification of the potential suppliers/technologies that will be affected by a change in demand (marginal suppliers/technologies).

Most of the work initially seems to confirm that if you want to know who generates the electricity for BEVs you have to look at the marginal current, and therefore choose the path of a CLCA:

However, in the LCA of EV, it emerges that the difference between ALCA and CLCA is often identified with the difference between average and marginal electricity mixes, and the terms are sometimes used as synonyms.

This has a strong impact on the results of the studies:

What is common to every study applying marginal mixes is the identification of the EVs as the marginal consumers, both in present and in future energy system. The result is that the effects of using present or future energy systems convey similar results, because technologies on the margin tend to remain the same also in projected scenarios. Therefore, EVs do not benefit from the general decarbonization in the future … Short term marginal mixes are generally higher than the average.

Of course this is the case, and it can’t be any other way. As long as fossil-fueled power plants supply the marginal electricity, electric cars will not benefit from the increase in the green electricity quota. Only changes in the composition of the fossil-fueled power plant park are relevant in this respect.

This means, however, that BEVs will not be able to contribute to the reduction of greenhouse gas emissions for decades. The opposite is the case: BEVs increase greenhouse gas emissions. So if the authors were to draw the only sensible, logical conclusions from their own findings, they would have to advise against BEVs.

This causes difficulties for them, as can be seen from the following remarks:

… we think this is not the best option to inform policy makers on wide-ranging policies such as the paradigm shift in the transportation sector …
Similarly, considering EVs as the short-run marginal consumers in projected scenarios is not coherent when the goal is to inform policy makers on the introduction of EVs in the transportation system.

Whether a methodology is correct, i.e. appropriate to the context to be examined, does not depend in any way on whether the outcome can be used to legitimize policy measures. These authors just dislike the results of their own work. Because they are determined to support the introduction of BEV, they are also prepared to put the cart before the horse: If analyses conclude that electric cars will increase greenhouse gas emissions for decades to come, they conclude that the methodology is wrong and recommend changing it in order to align the result with their personal convictions. The goal is not “to inform policy makers on the introduction of EVs in the transport system”, but to encourage them.

But how can they reverse the previous results to the contrary? Well, they’ve got some ideas.

First of all they try to justify „doubts about the methodology“ by pointing out that “literature distinguishes between short term and long term effects of a change.” The key point is summarized in these sentences:

As stated by Soimakallio et al. [78], if ‘new consumption’ is adequately anticipated before it occurs, there is no unambiguous reason to assign short term marginal production to this particular consumption.
if the utilities add more base capacity (beyond that projected by the U.S. Energy Information Administration [EIA]) as a result of the growth of PHEVs, then the correct CO2 emission allocation should be based on that new base capacity.

The flaw is almost obvious at first glance; it consists in the assumption of a causal relationship that does not exist in reality. The implicit assumption is that higher demand for chargeable electricity triggers an expansion of electricity generation capacity from renewable energies beyond the level that would be achieved without BEVs.

In truth, this is pure fantasy. Almost all countries have serious difficulties in achieving their expansion targets. For this reason introducing more electricity consumers such as electric cars merely postpones the energy turnaround.

In addition, as long as fossil power plants supply the marginal electricity, every expansion of renewable energy power plant capacity naturally increases the green electricity quota. But it  cannot change anything at all about the fact that the electric car is an additional electricity consumer. The sad truth is that charging current prevents the replacement of more fossil electricity with electricity from renewable sources. This will be so for decades to come. Doubts about the short-term marginal electricity mix cannot be justified in this crude fashion.

Nevertheless, the authors are determined to stay on the wrong track:

However, it can have applications when it comes to optimiying recharging time, since EVs are loads more flexible than others. … The studies discouraging EVs are works focusing on the short term introduction of EVs (both in present and future scenarios), where the energy system is not able to adjust to the increased demand of electricity.

Actually, charge management is necessary to solve a problem that BEVs themselves are exacerbating due to their own additional electricity demand. By trying to synchronize load peaks with electricity production peaks, the construction of new power plants to supply BEVs is to be avoided. However, no electricity can be generated in this way. The charging current remains marginal current.

After all, the authors did not even hesitate to add a remark that makes absolutely no sense even after rereading it a few times:

The adoption of average emission factors in the case of EVs could be justified by the low rate penetration of EVs, while the use of marginal could be explained by the high variation in load.

In fact the market share of EVs is completely irrelevant in this respect. How can it be seriously questioned that additional charging current of additional electric cars is additional energy demand which has to be delivered from marginal suppliers – unless the aim of the work is to write a study that glosses over the climate balance of electric cars?

The Forschungsstelle für Energiewirtschaft (FFE) correctly summarized all essential points in one paragraph: [ii]

Using an average approach will overestimate the renewable electricity in the displacement respectively consumption mix. Meanwhile, the marginal approach will more accurately estimate the environmental impact of the power grid due to additional generation and consumption, signaling more adequately to both policymakers and consumers the environmental impacts of appliances generating or using or this electricity.

The authors‘ criticism of the marginal mix is weakly founded and contradicts their own previous findings. It seems to be merely a bow to the expectations of politicians who want to see their questionable decisions legitimized.

Kai Ruhsert, June 2020

[i] https://www.researchgate.net/publication/327066754_Electricity_Generation_in_LCA_of_Electric_Vehicles_A_Review

[ii] https://www.ffe.de/attachments/article/797/EU%20Displacement%20Mix.pdf

%d Bloggern gefällt das: