What is the EROEI of Wind Power?

Exploring the minefield of Energy Return on Energy Invested calculations

Introduction

While researching several of my articles, I have come across wildly different claims about the energy return on energy invested (EROEI) of wind power. Perhaps the definitive paper on the EROEI of different energy sources is this paper by Weissbach, which arrives at an answer of 16 for unbuffered onshore wind. This calculation is based on an old design of wind turbine and some criticise the result as being too low because it has not taken account of technological progress.

On the other hand, some wind turbine manufacturers, like Vestas, claim numbers as high as 40 for some of their devices. Moreover, Carbon Brief has published an article claiming EROEI for wind power as high as 44:1.

It is difficult to understand such large differences in results. It does not seem credible that the EROEI for wind could improve by nearly a factor of 3 in little more than a decade. This article delves into the detail of how the calculations are made and explains the different results.

How is EROEI calculated?

As a reminder, EROEI is energy output over the life if the turbine divided by lifetime energy input. This is illustrated by Figure 1, that I grabbed from ResearchGate, which I understand originates from Prieto and Hall.

Figure 1 - Energy System and EROEI calculations after Prieto and Hall

This shows the cumulative energy costs over the life of an asset in red and the useful energy to the consumer in blue. The EROEI is the ratio of the blue column to the red column on the right-hand side of the image.

The expected output of a wind turbine is relatively easy to calculate. Take the rated power, multiply by the expected load factor and the anticipated life of the turbine. As we shall see below, the expected load factor is a bone of contention, but the basic calculation is straightforward.

Calculating the energy inputs is much more difficult. The general approach used is to take the bill of materials for a wind turbine and evaluate the material content, for example, copper, steel, resin and plastic for each turbine. Databases exist that contain the energy cost per unit of each material, so the total embedded energy in the turbine can be calculated. They also consider the energy costs of the large concrete foundations and the cables used to connect the wind farm to the grid. The energy costs of operating the wind farm over its life are also calculated.

Impact of Recycling

One controversial aspect of calculating the embedded energy is how recycling of materials at the end of life is treated. One approach is called the “avoided impacts approach”. This method assumes that a high percentage of the materials can be recycled at the end of life. The embedded energy in those metals is subtracted from the input energy. This has the effect of reducing the effective energy input, so the EROEI rises.  The trouble with this approach is twofold. First, not all the metals that can be recycled will be recycled. Second, what you get back at the end of life is not the original energy used to make the turbine but a pile of scrap metal that needs additional energy to remake it into something useful.

Another approach to end-of-life recycling is the “recycled content approach”. This method looks at the fraction of recycled material that is used to make new copper for instance and reduces the input energy per unit accordingly in the first calculation of embedded energy. This does reduce energy input but by a much smaller amount than the avoided impacts method. The recycled content approach is favoured by Weissbach, and the approach I think best represents the real world.

Weissbach Onshore Wind EROEI Results

Weissbach examined what is now a relatively old model of turbine, a 1.5MW Enercon E-66. He used the “recycled content approach” to recycling and assumed 2,000 hours of operation per year, which is effectively a 22.8% load factor. This resulted in an EROEI of 16 for onshore use. The detailed calculations behind Weissbach’s work are publicly available here.

Vestas Onshore Wind EROEI Results

Vestas has published a Lifecycle Analysis (LCA) for the onshore use of many of its turbines. An example is the V136 4.2MW turbine. For their headline results, they use the “avoided impacts approach” to recycling and effectively assume a load factor of 43% for onshore use by assuming 15,825MWh of generation each year from the 4.2MW turbine. This gives an EROEI of 40.

However, later in the report, they do show the impact of adopting the recycled content approach to end-of-life recycling. This has the effect of pushing up the input energy and reduces the EROEI to 28.8.  However, the load factor is still unrealistically high, pushing up the output energy. Over the past five years, the UK onshore fleet has averaged a load factor of 26.4%. Applying this load factor reduces the EROEI to 17.7, which is not too far away from the Weissbach result. This shows some technological improvement for these newer turbines compared to the older machine analysed by Weissbach.

What About Offshore Wind EROEI

Vestas has not yet published lifecycle analyses for its largest turbines which are most likely to be used offshore. Nor has it published a study specifically for offshore use. However, we can make an estimate of the EROEI of offshore wind.

There are competing forces that alter the EROEI of offshore wind compared to onshore. Acting in favour of higher EROEI is higher load factors. Over the past five years the average load factor for the offshore wind fleet has been 41%. This is obviously higher than onshore wind and on its own would result in more electricity being generated over the life of the wind farm and thus push up EROEI.

However, two other factors act in the opposite direction. The first is the “square-cube law”. The power output of a wind turbine increases with the square of the blade length. However, the weight of the turbine increases with the cube of the blade length. The weight and material content of the turbine is a reasonable proxy for the embedded input energy in the turbine. So, as the turbine size increases, we might expect the EROEI to decline because the energy output rises more slowly than the material content. Applying this rule to the 4.2MW turbine and scaling up to a modern 15MW turbine would give an EROEI of 21.2 for the avoided impacts method and 15.3 for the recycled content method, both using a 43% load factor. The latter figure falls to 14.5 using a 41% load factor.

The second factor that tends to decrease EROEI is that the wind farms are further away from the electricity grid. This requires longer cables to connect them to shore, longer transport routes to get the turbines to the operating location and larger foundations to cope with the increased stresses of operating offshore. The longer cables contain more embedded energy and also lead to higher transmission losses.

The sensitivity analysis in the Vestas LCA report (Section 7.2) for their 4.2MW turbine gives a few clues to the impact of the longer transport distances and longer connection cables. However, the long-distance analysis is for transporting the turbine to Australia. The longer connection cables actually reduce the mineral requirements because they use the “avoided impacts” approach that assumes the cable is recycled at end of life and therefore the expended energy is credited back, reducing the energy input. This is clearly an absurd result. For the purposes of this analysis, I have calculated the sensitivity of offshore wind to a 20% increase in input energy driven by longer transport distance, stronger foundations and longer connection cables. This gives a result of 12.1 using the more realistic 41% load factor.

I think it’s reasonable to estimate that EROEI of large turbines operating offshore is somewhere in the range of 12.1-14.5.

Conclusions

Pulling it all together, the different results are summarised in Figure 2 below.

Figure 2 - EROEI of Wind Power with Varying Assumptions on Load Factor and Recycling Method

it is clear that EROEI figures can be easily manipulated by choosing unrealistic load factors and adopting dubious methods for calculating the impact of recycling. I don’t think it is reasonable to credit back the embedded energy in a bunch of scrap metal and effectively claim the energy was not used at all.

It is reasonable to conclude that a realistic EROEI for modern onshore wind power is 17-18 and for offshore wind, 12-14.5.

---

If you have enjoyed this article, please share it with your family, friends and colleagues and sign up to receive more content.