Skip to content

Factual record. No finding of causation or responsibility is made here.

Understanding the roadside · Evidence explainer · 5 September 2026

Loose soil banks: how rain can turn exposed earth into a road hazard

A roadside bank is not separate from the road when water can carry its material onto the carriageway. The important questions are what can move, where it can travel, and what stops it.

This explainer sets out the physical processes behind the concern about the bank on the approach to Killardry Bridge. It is not a reconstruction of Bronagh’s collision or a substitute for a site-specific engineering assessment. The separate weather analysis examines the actual station records around 24 April 2025.

Exposed earth bank and stones beside the road, with cones visible, in the family’s later photograph
Family photograph, 1 May 2025: exposed earth and stones along the road margin. This is seven days after the collision, not a photograph of the surface that night. The visible cones cannot establish what signs or protection existed on 24 April. Tap to inspect the full photograph.

1. “Loose”, “exposed” and “clay” describe different things

Exposed means the surface is uncovered. Loose or disturbed describes its condition. Clay describes a particle-size component of the soil. A photograph can show a bare bank and scattered material; its colour does not establish the proportions of clay, silt and sand. That requires appropriate examination or testing. The USDA soil guide treats texture and aggregate stability as separate properties. [1]

Compaction can reduce pore space and infiltration. Conversely, loosening soil can initially improve infiltration while disrupting its structure. It is therefore not sound to say either “compacted means safe” or “loose means all the rain must run off”. Soil condition, surface sealing, cover and existing wetness all matter. [1]

Clay is not simply material that always washes away more easily: cohesion can make clay-rich soil resist detachment. Once fine particles are detached, transport is a separate question. A useful investigation identifies the actual material rather than assuming that every brown deposit has the same properties. [2]

2. Rain can detach material; flowing water can transport it

SourceExposed, erodible material
MovementRaindrop impact and surface flow
DestinationWhere the water actually carries it

Raindrops strike the surface and can break soil aggregates or dislodge particles. Water moving over that surface can transport detached material and can itself erode soil. Where flow gathers into small channels, erosion can become concentrated rather than evenly spread across the bank. These are established processes, not a theory invented for this case. [2]

But detachment is not delivery. Material loosened on a bank may remain at its foot, be intercepted, enter a drain or reach the road. The amount and size of material moved depend on the material and the flow. Visible stones do not prove that rain transported those stones; placement during works or other mechanisms must also be considered. [2]

3. A rain total is only part of the explanation

Runoff can occur when water arrives faster than it infiltrates, or when the ground has little remaining capacity to absorb water. Rainfall intensity and duration, prior saturation, vegetation and slope therefore matter together. A wet preceding period may be relevant, but a nearby gauge does not measure the moisture content of a particular bank. [3]

A daily total cannot reveal the strongest five-minute burst. Nor does an hourly total tell us exactly when within that hour rain fell. The same accumulated depth can be delivered in different ways. That is why the accompanying weather page preserves station distances, observation periods and time zones instead of treating one number as a complete account of the road.

4. Follow the water all the way to its destination

The investigation needs to connect three things: a sediment source, a route for its movement and the place it is deposited. Teagasc’s Irish catchment research distinguishes reducing sediment at source from interrupting its pathway. It also explains that landscape features can intercept sediment, while artificial drainage can concentrate flows and contribute to erosion. This is catchment research, not a finding about this carriageway. [4]

Applied here, the questions are concrete: does the bank fall towards the road or away from it? Where are the low points? Can runoff cross the edge? Does a channel collect the water, bypass it, or become a route for sediment? An outlet somewhere along a stretch does not, by itself, demonstrate that the whole stretch drains effectively. These are questions for levels, flow paths and inspection evidence—not a judgement from one camera angle.

Mud on a road is a recognised safety hazard. The Health and Safety Authority’s farm-machinery guidance warns that road contamination can cause motorists to lose control, and addresses prevention, warning and removal. Its example concerns agricultural road use; it does not establish who deposited material at Killardry, what was present that night, or the grip available to Bronagh’s tyres. [5]

5. Protect the source, manage the route, inspect the result

Depending on the site, erosion controls can include keeping disturbance limited, establishing vegetation, temporary surface cover, managing concentrated runoff and intercepting sediment. US Forest Service road guidance gives priority to treatment at the source. These are options to be designed for the circumstances, not a claim that one particular mat, fence or wall was mandatory here. [6]

Maintenance matters as much as installation. Sediment can reduce drainage capacity. Clearing a ditch may restore flow but also expose fresh erodible material. An inspection should therefore consider both the water route and what has been disturbed, particularly during or soon after storms. The appropriate response cannot be reduced to “dig a drain and leave it”. [6]

There are also two different safety questions: stopping water and sediment reaching the traffic surface, and protecting road users from an exposed edge or excavation. A sediment-control measure is not automatically a vehicle-restraint barrier. The roadworks safety section examines Irish signing, lighting, guarding and inspection requirements separately.

6. What needs to be established at this bank

  • Condition and date: what was exposed or disturbed before the collision, over what length, and by which operation?
  • Water route: surveyed bank and road levels, channel dimensions, inlets, outlets and any intervening protection.
  • Material and timing: contemporaneous surface observations, photographs, sediment evidence, works diaries and cleaning records.
  • Performance: what inspections tested the arrangement in rain and darkness, and what defects or corrective actions were recorded?

The physical concern deserves a proper answer. Rainfall erosion is real; road contamination is a recognised hazard. Establishing that this particular mechanism operated at this particular place and time requires the connecting evidence. Keeping that distinction clear makes the questions about the bank more precise, not less important.

Sources and scope

  1. USDA Soil Quality Test Kit Guide, July 2001: printed pp55–56, 69 and 77. Agricultural soil properties and tests, not a forensic examination of this bank.
  2. USDA NRCS, Rangeland Processes: Hydrology and Soil Erosion, Handbook 646, September 2021: §§646.06, 646.08, 646.23–24. General physical processes; no local erosion rate calculated.
  3. USGS, Infiltration and the Water Cycle, “Factors affecting infiltration”.
  4. Teagasc, Water quality: the role of sediment, Daire Ó hUallacháin, 11 May 2020. Irish catchment research, not a road-friction study.
  5. HSA, Guidance on the Safe Use of Tractors and Machinery on Farms, p38, §11.6. General road-contamination hazard.
  6. US Forest Service, PSW-GTR-272, September 2021, chapter 4: pp64, 81–82, 108 and 126–127. US road-maintenance and erosion guidance; not an Irish statutory standard.

This is an evidence explainer from the Bronagh English record, not an independent expert report. The family photograph is used to show the later roadside condition, with its date disclosed. No conclusion about collision causation or legal liability is made.

Cahir–Cashel road statistics Cahir–Cashel Deadly place to drive ↗ × Cahir–Cashel road statistics Road statistics · sourced comparison Cahir–Cashel: 3.03× the national road-death rate 1 January 2018–31 December 2023 (six complete calendar years) . In the full Tipperary–Cahir–Cashel Municipal District, including its rural areas, the derived annualized rate was 8.61 people killed per 100,000 residents , compared with 2.84 nationally . This is an area comparison relative to population, not a personal likelihood. This page uses the full municipal district, not the populations of Cahir and Cashel towns. The linked road-deaths map and record has broader coverage and a different purpose. The headline figures 23 people killed in 21 fatal-collision records in the district 8.61 annualized people killed per 100,000 residents 3.03× the State rate on this population basis 36.5% of County Tipperary’s deaths and 26.5% of its population; the district rate was 37.6% above the county rate. District results The district’s fixed Census 2022 population is 44,532 . The 23 people comprised 11 drivers, 5 passengers, 5 pedestrians, 2 motorcyclists . They were not all drivers. Annual district deaths were 6, 4, 1, 3, 3, 6 from 2018 to 2023. Scroll horizontally to read this table. Five Tipperary municipal districts District 2022 population Fatal-collision records People killed People killed per 100,000 per year Clonmel 25,003 10 13 8.67 Tipperary–Cahir–Cashel 44,532 21 23 8.61 Carrick-on-Suir 20,191 7 7 5.78 Nenagh 40,879 11 12 4.89 Thurles 37,290 8 8 3.58 Scroll horizontally to read this table. Annual people killed Year Tipperary–Cahir–Cashel County Tipperary State 2018 6 10 134 2019 4 13 140 2020 1 7 141 2021 3 10 132 2022 3 7 152 2023 6 16 179 County comparison National baseline: the Republic of Ireland recorded 878 people killed across a fixed Census 2022 population of 5,149,139 , an annualized rate of 2.84 per 100,000 residents over these six years. Tipperary recorded 63 people killed , a rate of 6.25 , ranking #2 of 26 counties behind Monaghan (6.38). Tipperary’s rate was 2.20 times the national rate. Scroll horizontally to read this table. All 26 counties, ranked by annualized rate; the CSO county aggregate does not supply mapped source-record counts Rank County 2022 population People killed Fatal-collision records People killed per 100,000 per year 1 Monaghan 65,288 25 — 6.38 2 Tipperary 167,895 63 — 6.25 3 Donegal 167,084 53 — 5.29 4 Roscommon 70,259 22 — 5.22 5 Louth 139,703 41 — 4.89 6 Cavan 81,704 23 — 4.69 7 Mayo 137,970 38 — 4.59 8 Offaly 83,150 22 — 4.41 9 Kerry 156,458 41 — 4.37 10 Kilkenny 104,160 27 — 4.32 11 Leitrim 35,199 9 — 4.26 12 Longford 46,751 11 — 3.92 13 Westmeath 96,221 22 — 3.81 14 Clare 127,938 29 — 3.78 15 Meath 220,826 49 — 3.70 16 Sligo 70,198 15 — 3.56 17 Limerick 209,536 44 — 3.50 18 Wexford 163,919 34 — 3.46 19 Galway 277,737 49 — 2.94 20 Carlow 61,968 9 — 2.42 21 Cork 584,156 83 — 2.37 22 Laois 91,877 12 — 2.18 23 Waterford 127,363 15 — 1.96 24 Wicklow 155,851 18 — 1.92 25 Kildare 247,774 25 — 1.68 26 Dublin 1,458,154 99 — 1.13 A higher district comparator Boyle Municipal District, County Roscommon, recorded 15 people killed in 14 source records from a population of 24,421: 10.24 per 100,000 , about 19% higher than Tipperary–Cahir–Cashel. This is not a claim that Boyle is Ireland’s worst district; 11 districts were screened in this targeted comparison. Clonmel’s 8.67 rate is marginally higher than Cahir–Cashel’s 8.61; the largest raw count is not the highest district rate. Public council minutes: 2024–2026 These later public records are separate from the 2018–2023 statistical comparison. They record requests, responses and planned work—not findings of causation or proof that work was completed. 2025-05-26 James Swords: recorded response to a safety request The minutes record Councillor John O’Heney asking for safety measures following a recent tragic incident. James Swords is listed as District Engineer present; the response is attributed to that role. Limit: The minute does not name the bridge or specify a parapet repair. It does not say that Gardaí were to design or direct a bridge repair. there is an ongoing Gardai investigation into the incident and if there are any recommendations from the report it will be implemented. Excerpt from the minutes, not a verbatim speech transcript. Official minutes — attendance and District Engineer, PDF p. 1 — Officials Present; James Swords, District Engineer Official minutes — District Engineer’s report, PDF p. 5 — Item 3.4, District Engineer’s Report Official minutes — recorded response, PDF p. 6 — Item 3.4, District Engineer’s Report 2024-10-14 Máirín McGrath seconded the default-speed-limit motion Motion 2861 was proposed by Cllr Niall Dennehy and seconded by Cllr Máirín McGrath, a Cahir LEA member. The full County Council recorded unanimous agreement. The motion sought selective application of the new default limits, rather than universal application across the county. Limit: This was not a motion introducing a blanket 80-to-60 reduction. The December clarification below records the legal-advice referral. Official October minutes — Motion 2861, PDF p. 12 — Item 8.2, Notice of Motion Ref. 2861 Official October minutes — unanimous outcome, PDF p. 15 — Notice of Motion Ref. 2861 outcome Official Council AGM minutes — elected-area roles, PDF p. 2 — Item 2.1, elected members by electoral area 2024-12-09 Council clarification and legal advice The minutes record Chief Executive Sinéad Carr acknowledging support for Motion 2861, while explaining that legal advice was being sought on whether a plenary motion could supersede national legislation. The October minutes were to be amended to reflect that clarification. Limit: This does not establish a road-specific exception or a suspension of national rules. Official December minutes — clarification, PDF p. 2 — Item 3.1, Matters Arising on October minutes 2025-01-13 80-to-60 implementation and weather-delayed signs The minutes record 7 February 2025 as the commencement date for the rural-local-road default reduction from 80 to 60 km/h. They also record severe weather delaying sign-installation work and distinguish a later, wider Phase 2 review. Limit: This is a dated implementation note, not proof of the limit or weather at any individual crash site. Official January minutes — Speed Limit Review, PDF p. 11 — Item 7.1, Roads, Transportation, Health & Safety 2026-05-25 477 district bridges; 14 applications; three funded The District Engineer reported €1 million in county bridge funding, 477 bridges in the district, applications for 14 bridges requiring work and funding for three. Limit: The minute does not establish whether Killardry was among the 14; funding is not proof of completed work. Official May 2026 minutes — bridge funding, PDF p. 6 — Item 3.6, District Engineer’s Report 2026-05-25 Temporary measures considered at Rehill Bridge Following Cllr Máirín McGrath’s request for urgent rehabilitation, the District Engineer reported that Rehill tender documents were nearly complete and temporary measures were being considered to reopen the bridge, with completion hoped for by the end of that week. Limit: This records a plan and expectation, not verified completion. Official May 2026 minutes — Rehill Bridge, PDF p. 6 — Item 3.6, District Engineer’s Report The earlier bridge inspection The family’s published summary of Council inspection record TY-L3151-B-010 reports “Significant Damage” on 21 January 2025 and records downstream capping as “partly removed”. That inspection predates both the 24 April 2025 collision and the 26 May 2025 meeting response. Limit: This is a summary of a Council record held on file, not the original report online. The records checked do not establish whether or when James Swords personally received or read the inspection report. Published family summary of the Council inspection record When was the inspection circulated to the district engineering team, and what action was recorded? Further accountability research Read the full sourced record on the primary statistics page . Representatives Engineering Payments Road budgets Crash coverage Weather Surfaces and works Exposure comparisons Method and limits People killed ÷ fixed 2022 population ÷ 6 years × 100,000. Populations are fixed Census 2022 references. Counts concern collision locations, not resident addresses; the result is not a rate per journey or vehicle-kilometre. Current 2024 reference boundaries are applied retrospectively. The series is provisional and subject to revision. The figures do not establish why crashes happened or measure road-maintenance spending. Return home · Open the broader road-deaths map and record Sources RSA SourceData — Map of Collisions source data — Fatal-collision source records, 2018–2023; mapped locations assigned to current reference districts. CSO ROA15 — Persons Killed and Injured — State and county people-killed totals, 2018–2023; provisional/revisable series. CSO ROA12 — All road users — Independent national cross-check of people-killed totals, 2018–2023. CSO Census 2022 Table 8.1 background notes — State and county 2022 population background. CSO FY108 — official statutory LEA population CSV — 2022 population basis and statutory LEA/district mapping. Tipperary County Council Adopted Budget 2024, Table 7 (PDF) — Five municipal-district populations labelled as based on Census 2022; physical page 20. Tailte Éireann 2024 statutory boundary dataset service — Checked 2024 Tailte Éireann statutory municipal-district boundary dataset service; current reference boundaries applied retrospectively. Irish Statute Book S.I. 631/2018 — Optional Boyle municipal-district/LEA mapping context. Download the five-district CSV · Download the 26-county CSV · Download the public JSON snapshot Factual record. No finding of causation or responsibility is made here.

Contact the family

If you have information about this road, this bridge, or the records on this site, you can write to Bronagh's family here.

Sent only to the family. Not shared, not stored by any third party.

Explore the record