Cleve Hill Solar Park was granted development consent in May 2020 by the Secretary of State (that would be Alok Sharma, who later became the President of COP26). May 2020 was lockdown city, and most of us probably didn’t notice, having apocalypse on our minds. Most of us probably haven’t noticed yet.

Personally I had noticed Cleve Hill, but my reaction to it was a shrug. To me, there was no way on this fair planet that Cleve Hill would be permitted. That meant I could just ignore it and wait for it to go away, not that any intervention on my part would have made the slightest difference.

What is, or will be, Cleve Hill? It’s an array, or a set of arrays, of photovoltaic panels – 884,388 of them is an indicative figure, each capable of a maximum of 395 watts. Yes, when it’s sunny in June. Although maybe not even then because of the particular design – see below. If you multiply the two numbers together you get about 350 MW. Each module is 2 square metres, so they cover 1.7 square kilometres. But the whole thing covers 3.6 square kilometres altogether.

OK, it seems like a pretty dumb idea to build such a monstrosity in a world where an electricity generator has to compete against other electricity generators for business in a fair battle. But that’s not the world we live in, and it’s not a reason that would make Jit shrug and assume it would be knocked back.

That shrug was because of its location. It’s near Faversham in Kent, or some of it is at least. But that’s not the problem. The problem for the developers was that the site was also right next to The Swale SPA and Ramsar site, which is also partly an SSSI, if you like that sort of thing (see featured image). SPAs and Ramsars are “international” sites, along with SACs, and they are the highest designation of protection afforded to biodiversity in the UK. SPA = Special Protection Area, and is designated for birds, SACs are at an equivalent level but for other species, and Ramsars are wetland areas named for the Iranian town where the Ramsar convention was signed. [In practice the three designations often overlap.] Developers often have to consider the effect that their plans will have on international sites at several kilometres distance (for things like increasing visitor numbers from a housing development, nutrient enrichment etc). When it comes to disturbing birds, the distance is lower, but it is still greater than zero. The distance to the SPA was zero as the featured image makes clear. Therefore, the development was bound to affect the integrity of the international site. Therefore, the Secretary of State was bound to refuse the development. Jit could shrug and concentrate on the unfolding viral apocalypse.

Except that isn’t what happened. Alok Sharma gave the nod. If I had thought about it for a minute, it might have occurred to me that being next to the SPA did not necessarily mean that the development would disturb birds on it. Because there is a ruddy great embankment keeping the sea out. There won’t be added visitor pressure. And after construction, there isn’t likely to be any issue with pollution (unless the battery goes up in flames).

Still, there is a strong case that the fields the solar panels will be built on are “supporting habitat” for the SPA, even if they are outside it. Birds that like mudflats have to go somewhere when the tide comes in, right?

There was also the possibility of collisions. As regular readers know, I often bang on about the long losing streak birds have in their battles with wind turbines. They also have a long losing streak in their battles with solar panels. Such battles, we might guess, are exceptionally rare. But when carpeting 360 hectares of land right next to a region with the highest possible protection for birds, it should at least warrant a discussion. Right?

I had always assumed that there was a chance – albeit a small chance – of water birds mistaking a solar farm for water. A solar farm might resemble water because it reflects the sky. It doesn’t have a diffuse reflection like a brick wall. It has a specular reflection like a mirror. Birds can be unwise enough to see a mirrored surface as another bit of sky (mirrored skyscrapers) or as water (solar farms, potentially).

Cleve Hill, as a “Nationally Significant Infrastructure Project” – a generator of over 50 MW, if you’ve been paying close attention – had a lot of hoops to go through. On the Planning Inspectorate’s website, there are 986 documents related to the development, of which 221 documents relate to the Environmental Statement.

One of these is the Ornithology chapter. How does Cleve Hill deal with the issue of collision risk? As part of the planning process, they put out a PEIR – Preliminary Environmental Information Report for public comment. Of all the public and private bodies who responded, including Natural England, the RSPB and the Kent Wildlife Trust, it was left to the Swale Green Party to raise the issue of collisions.

The solar panels themselves may present a challenge to birds who may see them as a “watery surface.”

Swale Green Party, responding to Cleve Hill’s PEIR

This is awkwardly put, but the relevant point is made. Here is the developer’s response:

Section 0 of this chapter assesses the potential effect of collision to birds; Natural England guidance states that there is no scientific evidence of collision risk associated with solar PV arrays.

[There is no Section 0. I think they meant 9.5.2.7.]

In Section 9.5.2.7, Paragraph 174 says:

Natural England has published a review of the impacts of solar farms on birds, bats and general ecology⁸ . The review concluded that there is no scientific evidence of collision risk associated with solar PV arrays and the risk of collision with solar panels is likely to be very low but not impossible, although there could be risk associated with overhead power lines.

[The footnote reference is missing. Is this important, or not? No-one cares. Find it yourself.]

Well, that was rather interesting. Natural England guidance sweeps collision risk off the table? Before continuing, let me just highlight the use of the word “scientific” in a “scientific” document. It is an obvious redundancy. A cynic might conclude that it is designed to produce a rhetorical effect. And note that there is no risk of collision with solar PV arrays, but (bizarrely) a “very low but not impossible” risk of collisions with solar panels, which are what solar arrays are made of.

What does – once we have tracked it down – the evidence review say? First of all, it relies heavily on a single paper for its conclusions on collision, one by DeVault et al (2014). This assesses the use of PV arrays at airports. The review notes (my bold):

In terms of collision risk, DeVault et al (2014) observed no obvious evidence for bird casualty caused by solar panels, despite conducting 515 bird surveys at solar PV sites.

While this statement is true, it omits a little bit of nuance. DeVault et al were interested in “implications for aviation safety”. Airports are installing solar arrays because they have lots of flat land, mostly grass, that can’t be used for anything else. They want to know whether birds are attracted by the PV installations. According to DeVault et al, birds are not attracted to PV installations near airports (nor are they repelled). But they weren’t looking for carcasses, they were surveying flying birds at range. I don’t think this survey tells us anything about what will happen at Cleve Hill.

In any case, in point ix of its Executive Summary, the NE evidence review says:

Some scientific and grey literature data, based upon carcass searches around solar PV developments suggests that bird collision risk from solar panels is very low.

I think there is a leap between the Ornithology Chapter’s (accurate quote) “bird collision risk from solar panels is very low” and its (apparently distorting the review’s description of DeVault’s finding as if it was the review’s settled conclusion) “no scientific evidence of collision risk associated with solar PV arrays.”

So in the NE review, the phrase is “no obvious evidence” of collision risk in a particular study near airports. In the Cleve Hill Ornithology Chapter, this becomes “no scientific evidence” at all.

The Natural England review sees little danger from direct collisions. It’s a pity that the Natural England review did not bother to include Kagan et al 2014.

Kagan et al is well known for describing the in-flight immolation of birds at the concentrating solar plant Ivanpah. But it also reported on deaths at a solar PV plant, imaginatively named Desert Sunlight. Kagan et al walked up and down the rows of solar panels and picked up the dead birds. Then, when possible, they autopsied them. The image below shows the top part of one of Kagan et al’s tables, which shows the ex-water birds they found on the ground. Some of the birds had died from flying into the panels – 8 of 19 birds dying of impact trauma were water birds. Others had been killed by predators:

Predation is likely linked to panel-related impact trauma and stranding. Water birds were heavily over-represented in predation fatalities at Desert Sunlight. Of the 15 birds that died due to predation, 14 make their primary habitat on water (coots, grebes, a cormorant, and an avocet).

Kagan et al, 2014

That study was observational: no conclusions could be made on death rates. It reported the factual observation that solar farms kill birds – and it ought to have been considered in the NE review, and in turn by the Cleve Hill developer’s ecologists. Since 2014 there have been further studies, mostly showing a small but measurable problem:

Penniman & Duffy presented their own best practice guidelines for solar farms on Hawaii in 2021, in which they stated:

There is a clear body of evidence from the mainland USA and internationally, that birds can confuse solar arrays with water sources due to the “lake effect” and attempt to land on them, dying in the process or being injured and/or subsequently depredated.

Penn & Duffy, 2021

(The lake effect being the hypothesis coined by Kagan et al to explain why water birds would fly into solar panel arrays.)

Kosciuch et al have published two recent studies (2020, 2021), showing a relatively minor effect:

The idea of “lake effect” in which birds perceive a PV USSE facility as a waterbody (or the facility creates a lake effect) and are attracted is likely a nuanced process as a PV solar facility is unlikely to provide a signal of a lake to all aquatic habitat birds at all times. The results from our study suggest that some species of aquatic habitat birds could be attracted to PV USSE facilities, and if attraction occurs, it is likely context dependent. The most compelling evidence for attraction is the mortality of water-obligate species (e.g., loons) found at PV USSE facilities in desert environments that lack water, as these species perish on dry land.

Kosciuch et al, 2021

Last, we found that annual fatality rates never exceeded 2.99 fatalities/MW/year (1.03 fatalities/hectare/year) in the SMD BCR, were highest in the CC BCR where the rate was 9.26 fatalities/MW/year (5.17 fatalities/hectare/year), and that fatality rates did not correlate with nameplate capacity.

Kosciuch et al, 2020

9 fatalities per MW per year? Hell, Cleve Hill is only 350 MW. That’s only 3000 birds a year. Hardly worth even talking about.

Now, I’m not suggesting that 9 birds per MW per year is a realistic figure. It’s likely to be much lower. One reason why Desert Sunlight might have culled so many birds is that – as its name suggests – it’s in the desert. Thus we can imagine a migrant bird flying by, seeing the gleam of water, and thinking, “Thank God, some water at last,” and then finding out the hard way that what it thought was water was not water at all. Cleve Hill, being next to some actual water, might therefore be far less attractive. But I do think it would have been worth at least considering.

Cleve Hill is though likely to look more like a lake than your typical solar array. The ones we are mostly familiar with face south, at an angle to take advantage of the sun’s zenith angle. Something like these near Tuxford on the A1:

The kind we are talking about at Cleve Hill aren’t like that. According to the developer, “In the candidate design, tables will be located continuously from north to south without substantial gaps between them”. In other words, the panels will not be in banks facing south with large areas of grass between them. They will be quite flat and butted up together to more or less cover the ground completely. No sort of life can exist beneath them, except maybe troglodytes. They’ll be like these at Cestas in France:

As things stood, the potential for bird collisions was dismissed by the developer, and was not mentioned in the HRA (Habitats Regulations Assessment: a test by the appropriate authority of whether a development will have any effect on the integrity of an international site). The HRA mentioned disturbance, dust, and the loss of habitat for brent geese, golden plovers, lapwings and marsh harriers (if you are particularly interested to read this, go to infrastructure.planninginspectorate.gov.uk, search for Cleve Hill, pick documents, and then search for “habitats” – it’s the first hit. I know, you really can’t wait.)

Final point. Whatever risk there is, mitigation is straightforward. Kagan et al recommend retrofitting visual markers to break up the reflection from the solar panels at 28cm intervals, noting that at 10cm intervals impacts with windows – and by analogy, solar panels – are eliminated. The way to stop wigeons flying into solar panels is the same as the way to stop pigeons flying into skyscrapers. All the skyscraper architect has to do is to divide that vast mirrored surface into 10 cm squares. My guess is that instead of sullying their grand design like that, they’re brave enough to let the pigeon take the risk. One imagines that a solar developer would not be too keen either. (The Cleve Hill modules are about 2 m by 1 m, so you’d have to split that into quite a large number of squares, depending how wide the “visual markers” would be.)

Cleve Hill was successful in bidding for a CfD in Allocation Round 4, offering 112 MW at £45.99:

However, the developer thought that Cleve Hill was not a very warm and fuzzy name, and has now decided to call it “Project Fortress.”

This is perhaps a more apt name, because none of its neighbours seem to want it, afraid that the attached battery storage system might go up in flames one day. The Faversham Eye has more despairing commentary, if you’re not miserable enough already.

Note to diary: stop calling these things solar “farms”. They don’t farm sunlight, they collect it.

Conclusion

The risk of bird collisions with Cleve Hill’s Project Fortress’s solar arrays is probably low, but it isn’t zero. It should have been considered seriously for its effect on the integrity of the SPA rather than dismissed out of hand.

References

You can find pdfs of all these via Google Scholar.

DeVault, T. L., Seamans, T. W., Schmidt, J. A., Belant, J. L., Blackwell, B. F., Mooers, N., Tyson, L. A. and Van Pelt, L. (2014) ‘Bird use of solar photovoltaic installations at US airports: implications for aviation safety.’ Landscape and Urban Planning. Elsevier, 122 pp. 122–128.

Kagan, R. A., Viner, T. C., Trail, P. W., & Espinoza, E. O. (2014). Avian mortality at solar energy facilities in southern California: a preliminary analysis. National Fish and Wildlife Forensics Laboratory, 28, 1-28.

Kosciuch, K., Riser-Espinoza, D., Gerringer, M., & Erickson, W. (2020). A summary of bird mortality at photovoltaic utility scale solar facilities in the Southwestern US. PloS one, 15(4), e0232034.

Kosciuch, K., Riser-Espinoza, D., Moqtaderi, C., & Erickson, W. (2021). Aquatic Habitat Bird Occurrences at Photovoltaic Solar Energy Development in Southern California, USA. Diversity, 13(11), 524.

Penniman, J. F., & Duffy, D. C. (2021). Best Management Practices to Protect Endangered and Native Birds at Solar Installations in Hawaii.