Ireland has spent twenty years learning what happens when wind farms are built on wet ground. New research warns that a changing climate will make that ground more dangerous, and the planning rules have yet to catch up.
I. The hill
On the afternoon of Thursday, 16 October 2003, a hillside in the Slieve Aughty mountains of south Galway began to move. Contractors were excavating near turbine 68 of a wind farm above the village of Derrybrien when the bog beneath them let go. The developer put the figure at 450,000 cubic metres of peat. The planning inspector who examined the file seventeen years later recorded about 450,000 tonnes disturbed across some 25 hectares, of which roughly 250,000 tonnes travelled three kilometres downslope through a State forestry plantation. Some seventy workers were taken off the site. The mass stopped just short of an unoccupied farmhouse.
Late in the month heavy rain came, and the peat moved again. It reached the Owendalulleegh River and travelled on towards Lough Cutra, the lake from which the town of Gort draws its water. Inland Fisheries Ireland later counted about 50,000 dead fish along eighteen kilometres of river: brown trout, brook lamprey, eel, gudgeon, stone loach.
The inspector's account of the cause is worth reading slowly, because every element of it will recur. The peat was deep. The slope broke convexly above a watercourse. The bog had been damaged years earlier by forestry and its drains, some of them buried and undetected. And excavated peat had been stockpiled downslope of the turbine base, on a steeper gradient, where its weight bore on ground that could not carry it. There had been warnings. Earlier that same month, the report notes, there were small failures at three other turbines, under side-cast material.
There is a detail in the planning register that tends to be forgotten. The wind farm was consented in three phases, seventy-one turbines in all. Residents told an Oireachtas committee that December that permission had run out on 11 October, five days before the slide. Galway County Council's own register shows that applications to extend the first two phases were lodged on 1 October 2003, a fortnight before the hill moved, and granted on 24 November, five weeks after. In the interval a mountainside had come down, a river had been smothered and a town's water supply was under watch. The extensions issued anyway.
Seventy turbines were eventually built. The seventy-first, number 68, was omitted from the scheme.
It would be convenient to regard Derrybrien as a freak, an accident of an earlier and more careless era. A body of research, capped this year by two papers in Irish Geography, makes that position difficult to hold.
II. The ledger
Niamh Cullen and Mary Bourke are geographers at Trinity College Dublin, and their first paper is, on its face, an act of bookkeeping. Until now the island kept two landslide inventories, one in each jurisdiction, each compiled by a different agency. Cullen and Bourke merged them, added their own records, and attached weather conditions to 86 events.
The combined ledger holds 3,098 landslides. Of these, 134 had never been reported. Peat is the material that fails most often, in roughly 45 per cent of cases. And about 95 per cent of the peat failures are associated with wet conditions.
That last figure deserves a second reading. Peat in Ireland does not, as a rule, fall down because it is steep. It fails because of water: how much arrives, how fast, and what state the bog was in when it came. A planning system that asks an applicant how steep the slope is, and how deep the peat, is asking a question about geometry. The inventory says the answer lies in hydrology.
The authors do not say this about planning. Their papers are about landslides and climate, and they make no recommendation to any planning authority. The inference is drawn here, and it should be judged as such.
III. Three bogs in one year
The second paper asks what a changing climate will do to those numbers. It is a position review, not a field study, and it rests on three blanket-bog failures from a single year. Three is a small sample. But 2020 has since become the most closely studied year in the history of Irish peat, and the other studies point the same way.
| Site | When | Setting | What the studies found |
| Shass Mountain, Leitrim | 28 June 2020 | Deep peat beside forestry; drains diverting water onto the bog | A dry spring that cracked the surface, then intense rain |
| Meenbog, Donegal | 12 November 2020 | Afforested blanket bog; wind farm under construction | Weeks of wetness, no single storm; road construction implicated |
| Mount Eagle, Kerry | 15 November 2020 | Forestry plantation on blanket bog; an earlier failure alongside | Cumulative rain on ground near saturation |
Cullen and Bourke reconstructed the hourly rainfall. At Shass Mountain, they suggest, a spring about 72 per cent drier than average shrank the peat and opened tension cracks, and the cracks carried the June rain straight down into the bog. At Meenbog there was no violent storm at all. The hill had simply been wet for a long time. Three slides, then, and three different routes to failure. There is no one rainfall threshold that a site engineer can watch for.
Alan Dykes of Kingston University, working from drone footage during the pandemic, reached a compatible view. Two of the failures were most likely triggered by rainfall. The third, at the wind farm, appeared to follow the loss of support from an earlier peat flow that was itself probably caused by building an access road. Forestry operations, he concluded, probably contributed to two of the three.
Geological Survey Ireland and the National Parks and Wildlife Service then commissioned a full study of all three sites. Its causal report, finished in late 2024, found recurring ingredients: deep peat, a break in slope, forestry drains concentrating water, and at two of the sites the dry spell of spring 2020.
A fourth group, including Bourke, looked from orbit. Using satellite radar, they measured how far the surface of each bog rose and fell as it took on and lost water, a motion sometimes called bog breathing. The ground that later failed had been breathing hardest. The signal was visible up to two years before the slides.
The forecast compounds all of this. By mid-century Ireland is projected to have wetter autumns and winters, drier summers, longer dry spells and heavier downpours. That is the Shass Mountain sequence and the Meenbog sequence, both, more often. Cullen and Bourke conclude that landslide frequency is likely to rise.
One more finding belongs here. On the national landslide susceptibility map, Shass Mountain and Meenbog were rated low to moderately low.
IV. What the engineers already knew
None of this was new in kind. It was new only in volume.
In 2004 the people of Derrybrien commissioned their own inquiry from Richard Lindsay and Olivia Bragg at the University of East London. Their report observed that the slide had come, unusually, during a dry spell after an exceptionally dry summer, and that rain reactivated it twelve days later. It described how conifer plantations crack blanket peat along the plough furrows until, after twenty or thirty years, the bog is a set of loosely connected blocks. It described how roads floated on peat cut across drainage lines, ponding water on one side and drying the other. And it said of the environmental statements that they presented a best case.
In 2011 Michael Long, Paul Jennings and Roselyn Carroll catalogued at least thirteen Irish peat slide events between 2006 and 2010. The events came in clusters that coincided with intense rainfall, and human activity such as road construction contributed. The paper added a caution that planners might have underlined. Stability calculations of the standard kind, it said, should be treated carefully, because they do not capture what happens at the sliding surface, and in particular the lubricating role of water.
The Scottish Government's guidance on peat landslide hazard, revised in 2017 and used by Irish consultants for want of an Irish equivalent, asks for a demonstrable understanding of site hydrology and expects layouts to avoid ground of medium or high risk. And researchers at Aberdeen calculated in 2014 that, as the grid decarbonises, wind farms on undegraded peat will by 2040 mostly fail to deliver a net carbon saving at all.
So the mechanisms were described in 2004, the pattern was counted in 2011, the method was codified in 2017. What Cullen and Bourke add is the ledger and the weather forecast.
V. Luxembourg
Derrybrien went to the Court of Justice twice. In July 2008, in Case C-215/06, the Court held that Ireland had allowed the wind farm to be consented without a proper environmental impact assessment, and that its system of retention permission let developers regularise such projects after the fact. Ireland amended the law and invented substitute consent. It did not, for another eleven years, assess the wind farm.
So in November 2019, in Case C-261/18, the Court fined the State a lump sum of €5 million and €15,000 a day until it complied. The final bill was reported at about €17 million. The ESB's subsidiary applied for substitute consent in August 2020.
The inspector recommended a grant. A retrospective stability assessment in the file had rated the risk across the site, as it stood before construction, as possible to likely, and had identified 38 zones where a slide was very possible or likely, the ground around turbine 68 among them. The consultants added that such results would not be abnormal for a wind farm on upland blanket bog. It is a remarkable sentence to find in a planning file. The inspector was satisfied that, after 2003, the works had been managed and controlled.
On 4 February 2022 the board refused. Its order does not dispute that the later construction was better run. It says that this is beside the point. The effects of the slide were, in the board's words, "clear, profound and unacceptable". They were permanent. No subsequent care could bring the development within the exceptional circumstances the statute requires, and the board held itself precluded from granting. The ESB switched the turbines off.
It is worth being exact about what Derrybrien decided. The Court of Justice did not rule that wind farms may not be built on peat. It ruled that the assessment must come first and cannot be supplied later by way of apology. The board's refusal added a harder corollary: some damage, once done, cannot be made lawful at all.
VI. A site where none was anticipated
Meenbog is the sequel, and the more unsettling case, because at Meenbog the assessment was done.
The site is afforested upland blanket bog between Barnesmore Gap and Ballybofey, draining for the most part towards the Finn and Foyle river systems and across the border into Tyrone. A forty-nine-turbine scheme had been refused there. In 2017 the developer returned with nineteen turbines and a peat stability assessment prepared on the Scottish method. The assessment computed a factor of safety at each turbine: 1.3 or below would signal risk, and the lowest on the site was 1.67. The risk of instability was rated zero to negligible everywhere.
Not everyone was persuaded. The inspector's report (ABP-300460-17) records objections from the Loughs Agency, from the Northern Ireland environmental authorities, and from a local group that had retained its own peat expert to contest the calculations. Several submissions mentioned Derrybrien by name. The inspector noted that there was no recent history of slides in the area, found the assessment robust, and wrote that she was satisfied the works "would not give rise to peat instability or slippage" provided the mitigation was stringently applied. The board granted permission on 25 June 2018. Its environmental assessments concluded that peat slides were not anticipated.
Construction began in November 2019. What happened next is set out in two judgments, Donegal County Council v Planree Ltd in the High Court ([2024] IEHC 193) and the Court of Appeal ([2024] IECA 300). Within a year the developer's own consultants had logged eight instances of peat failure. One group of slides in June 2020 involved about 40,000 cubic metres. The consultants attributed these to excavated peat being cast onto in-situ peat that could not bear it. According to the Court of Appeal, it appears none of the earlier incidents was reported to the council.
On 12 November 2020 a twenty-metre section of a floating road, under construction towards turbine 7, failed beneath its fresh load, and the failure spread. About 86,000 cubic metres of peat slid. Some 66,000 left the site by the Shruhangarve stream, entered the Mourne Beg River, and went on to protected sites on both sides of the border. The companies pleaded guilty in the District Court in June 2022 to pollution offences.
By the time of the High Court trial, three things were agreed. The works had caused, or at least contributed to, the slide. Those works conformed with the permission. And there was no evidence of negligence. The developer's own 2017 peat management plan, the judge noted, had named the side-casting of excavated material as one of the main stability risks on the site.
Mr Justice Holland drew the inference carefully. Assuming the works were conducted properly, he said, the facts showed peat slides occurring repeatedly despite due care, on a site where the assessments had anticipated none. He described the 2018 assessment as sanguine. Ms Justice Butler, for the Court of Appeal, put it more plainly still: the original assessments had not identified the weak peat, and any future assessment would have to reckon with that.
Fairness requires the other half of the record. The developer remediated the slide under the supervision of the Environmental Protection Agency and the council, and both accepted the result. A fresh stability assessment in 2021 concluded that the wind farm could be finished safely. The injunction that halted the project in 2024 rested on twenty-five unrelated deviations from the permission, which rendered the whole development unauthorised. A first substitute consent application was withdrawn in February 2025. A second was lodged on 3 July 2026.
Still, the essential fact is the agreed one. Derrybrien failed without an adequate assessment. Meenbog failed with one, approved, on the best available method, with the mechanism named in advance in the developer's own documents. The same mechanism, seventeen years apart: loose peat set down on a slope that could not hold it.
VII. The rulebook
One might expect that, after Derrybrien, the State rewrote its rules. It did not. The Wind Energy Development Guidelines that govern these applications date from 2006. Revised drafts appeared in 2013, 2017 and 2019. Successive climate action plans promised a final text by 2023, then by 2024. In 2019 the Supreme Court, in Balz v An Bord Pleanála, quashed a wind farm permission because the board had declined even to consider a submission that the guidelines were out of date. In March 2026 a minister of state said the department was working towards a draft National Planning Statement and that the 2006 guidelines remained in force. The Government's stated aim is to finalise it before the end of 2026.
There is a national landslide susceptibility map, and applicants cite it routinely. But Geological Survey Ireland itself cautions that susceptibility is not hazard and not risk, and the map reassured at both Shass Mountain and Meenbog. The Survey's own consultants, in a technical note completed in December 2024, listed what the country lacks: a national dataset of peat depth, LiDAR coverage of its peatlands, enough rain gauges at critical sites, any body of research on climate change and peat stability, and Irish best-practice guidance for assessing peat landslide hazard. The note recommends that stability be assessed upslope and downslope of a development and not merely beneath it.
Meanwhile the map is being redrawn from Brussels. The revised Renewable Energy Directive requires Ireland to map its territory for renewables and to designate acceleration areas, where projects are not expected to have significant environmental effects and move faster. The Environmental Protection Agency, in its submission on that exercise, observed that counties apply different criteria, so that suitability changes at a county boundary. The constraints it lists include biodiversity, landscape, water, soil and flood risk. Peat stability does not appear.
The legal standard, by contrast, is exacting. Where a project may affect a European site, the assessment must contain complete, precise and definitive findings, with no gaps, capable of removing reasonable scientific doubt. That is the test of the Court of Justice in Sweetman (C-258/11), applied to wind farms by the High Court in Kelly and by the Supreme Court in Connelly. Active blanket bog is a priority habitat, and many upland wind farm sites drain to a protected river. A stability assessment that cannot say how a hillside will behave after a dry spring and a wet autumn has, on the research, a gap in it.
And there is pressure from the other side. In February 2026, in Coolglass Wind Farm v An Bord Pleanála, the Supreme Court held that the planning commission must perform its functions consistently with the State's climate objectives so far as practicable, and must engage with that duty in substance, even where a development plan points the other way. The duty does not compel a grant. But a refusal now has to be reasoned against the 9,000 megawatts of onshore wind the State has promised itself by 2030.
VIII. The case against a line, and where to draw it
The argument for leaving things as they are deserves to be put at its strongest.
Ireland has more than three hundred wind farms and over 5,000 megawatts of onshore wind, which supplied almost a third of the island's electricity in 2024. Many stand on peat. Two have produced catastrophes of national note. The industry's own mapping suggests that less than two per cent of the State remains realistically available for new turbines, and every exclusion pushes development towards lowland communities who object for reasons of their own. Dykes, the geomorphologist, has said that failures of this kind are "unforeseen but not unforeseeable", which is an argument for better investigation, not prohibition. And the Aberdeen carbon finding concerns undegraded peat. Derrybrien and Meenbog were conifer plantations, already drained, where the carbon case for a wind farm is at its best.
That last point, though, turns on itself. Afforested blanket bog is where the carbon arithmetic is kindest and the ground is worst. It is the peat that Lindsay and Bragg described cracking along its furrows, and the setting of all three failures in 2020. The land that looks most expendable is the land most likely to move.
So the line proposed here is a narrow one. It concerns a category of ground, not a technology, and it would work like this.
Decide it on the map, before the application. Ireland is drawing the maps now: the national territory mapping, the regional renewable electricity strategies, the wind strategies in county development plans. Sloping blanket peat that has been drained or afforested, with a break in slope and a watercourse below leading to a protected river or a public water supply, should be designated not normally permissible for turbines, roads, borrow pits or peat storage. It should never be an acceleration area. This is spatial exclusion at plan level, with the strategic environmental assessment that entails. It is a different thing from screening a project out of assessment, and the two should not be confused.
Retire the susceptibility map from a job it was not built for. It rated two of the 2020 sites low. Designation should rest on the landslide database, on measured peat depth, on forestry and drainage history, and on satellite measurement of how the bog surface moves, which flagged all three sites up to two years ahead.
Make the stability assessment a hydrological one. A factor of safety is a snapshot of a dry calculation. The research describes failure as a sequence: drought, cracking and rewetting, or weeks of saturation. An assessment should model a range of water levels and antecedent conditions under mid-century rainfall, extend upslope and downslope of the works, and say plainly what it cannot predict.
Put the controls in the consent, not the conditions. Both disasters began with excavated peat set down in the wrong place. Where spoil goes, how much, on what gradient, and at what rainfall the work stops are matters of design. They belong in the assessed scheme, with thresholds stated, and not in a management plan to be agreed afterwards.
Require every peat movement to be reported. At Meenbog a series of failures preceded the large one. A condition obliging the developer to notify the planning authority of any instability, and to stop until it is assessed, costs almost nothing.
Finish the rulebook. The National Planning Statement on wind energy should carry a chapter on peat, and the State should adopt the Irish guidance on peat landslide hazard that its own geological survey has been told is missing.
None of this would have stopped most of the wind farms now turning. It might have stopped two.
IX. The paperwork
At Derrybrien, the seventy turbines have stood idle since early 2022. Taking them down is itself development on a hill that has already failed once, and so it too requires consent. In July 2025 the ESB's subsidiary applied to An Coimisiún Pleanála. The commission's file gave 28 January 2026 as the date by which the case was due to be decided. In mid-November 2026 it recorded no decision.
Under the plan, the towers and blades would go. The foundations, the hardstands and the access tracks would stay in the bog.
More than two decades after the hill moved, the paperwork is still trying to catch up with it.
Sources
Research
- Cullen and Bourke, An all-island landslide database for Ireland, Irish Geography 58(1)
- Cullen and Bourke, The potential impact of climatic change on landslides in Ireland, Irish Geography 58(2)
- Dykes (2022), Landslide investigations during pandemic restrictions: initial assessments of recent peat landslides in Ireland, Landslides 19(2)
- Islam et al. (2022), Potential use of APSIS-InSAR measures of the range of vertical surface motion to improve hazard assessment of peat landslides, Mires and Peat 28
- Long, Jennings and Carroll (2011), Irish peat slides 2006–2010, Landslides 8(3)
- Lindsay and Bragg (2005), Wind Farms and Blanket Peat: The Bog Slide of 16th October 2003 at Derrybrien, 2nd edition, University of East London
- Smith, Nayak and Smith (2014), Wind farms on undegraded peatlands are unlikely to reduce future carbon emissions, Energy Policy 66
- Fehily Timoney for GSI and NPWS (2024), Assessment of main contributing factors leading to three major peatland failures: project page, causal factors report, data gaps technical note
- Scottish Government (2017), Peat Landslide Hazard and Risk Assessments: Best Practice Guide, 2nd edition
- Geological Survey Ireland, Landslide Susceptibility Mapping
Judgments
- Case C-215/06 Commission v Ireland, Court of Justice, 3 July 2008
- Case C-261/18 Commission v Ireland, Court of Justice, 12 November 2019 (press release)
- Case C-258/11 Sweetman v An Bord Pleanála, Court of Justice, 11 April 2013
- Donegal County Council v Planree Ltd and Mid-Cork Electrical Ltd [2024] IEHC 193, Holland J, 10 April 2024
- Donegal County Council v Planree Ltd and Mid-Cork Electrical Ltd [2024] IECA 300, Butler J, 18 December 2024
- Kelly v An Bord Pleanála [2014] IEHC 400, Finlay Geoghegan J, 25 July 2014
- Connelly v An Bord Pleanála [2018] IESC 31
- Balz and Heubach v An Bord Pleanála [2019] IESC 90, O'Donnell J, 12 December 2019
- Coolglass Wind Farm Ltd v An Bord Pleanála [2026] IESC 5, 4 February 2026 (summary by Mason Hayes & Curran)
Planning files
- An Bord Pleanála, Inspector's Report, Derrybrien wind farm substitute consent, ABP-308019-20, February 2021; Addendum Report, December 2021; Board Order, 4 February 2022
- An Bord Pleanála, Inspector's Report, Meenbog wind farm, ABP-300460-17; Board Order, 25 June 2018
- Galway County Council planning register, extension of duration files 03/5642 and 03/5637
- An Coimisiún Pleanála, case 323014 (Derrybrien decommissioning) and case 324433 (Meenbog substitute consent)
Official and press
- Oireachtas Joint Committee, Derrybrien Landslide Action Group, 11 December 2003
- Dáil written answer on C-215/06 and C-261/18, 9 November 2021; Dáil debate, 27 November 2025
- EPA submission on National Territory Mapping for Renewable Electricity
- RTÉ, 4 November 2003; RTÉ, 18 July 2025; Irish Times on the 2004 consultants' report; Irish Times, The science of bogslides; Irish Farmers Journal, March 2026
- Wind Energy Ireland, statement to the Oireachtas climate committee; Irish Farmers Journal on installed capacity
Sources consulted in November 2026. The planning recommendations in section VIII are the author's own and are not made by the researchers cited.
No comments:
Post a Comment