One passenger died, four people suffered serious injuries and 23 sustained minor injuries when two Transport for Wales Class 158 units collided head-on near Talerddig, Powys, at approximately 19:26 on 21 October 2024, at a closing speed of around 50 km/h. The Rail Accident Investigation Branch's 50,000-word final report identifies failed sanding equipment, an approach speed too high for the available adhesion, and an ETCS design that assumes effective braking regardless of railhead conditions as the factors behind the collision.

Train 1J25, the 17:xx-ish Shrewsbury–Aberystwyth service, should have stopped in Talerddig passing loop to allow 1S71 to pass in the opposite direction. Instead, it overran its limit of movement authority — the ETCS equivalent of a red signal — on exceptionally low adhesion down a steep gradient, and entered the occupied single-line section with emergency brakes applied but no effective deceleration.

Why did the sanders fail?

Both sanding systems on 1J25 failed to dispense sand. RAIB found two defects: a high-resistance contact in the Sander Isolation Switch and a failure of the Low Speed Relay to energise. Either would disable the automatic sander, and the onboard test switch does not exercise either component. The blocked sander hoses found on the leading car after the accident were a consequence, not a cause — no sand sat behind them, proving no discharge had been attempted since the blockage formed.

The driver never used the one-shot emergency sander, which operates only with emergency braking applied. RAIB quoted the driver's account: "It had not occurred to them to use the emergency sanders on the day of the accident. They recalled thinking that the adhesion conditions would improve, as this was their previous experience of braking in low adhesion conditions." Once the driver decided to call the signaller, "using the emergency sanders at this point did not come to mind."

RAIB's simulations showed that a functioning automatic sander, or a timely emergency sander deployment, would have prevented the collision.

What role did ETCS play?

The Cambrian line, where Talerddig sits, carries the UK's first ETCS Level 2 installation without lineside signals, with all commands delivered through the cab DMI. ETCS functioned as designed, demanding full service braking and then emergency braking. But the system assumed braking would be effective and took no account of poor adhesion.

Because ETCS provides continuous supervision, its designers accepted reduced margins. The standard overlap beyond a conventional signal is 180 metres; the overlap at Block Marker MH1078 was 81 metres — less than half. Standards in force when the system was specified assessed braking performance at an undefined "normal adhesion level" and treated "low adhesion", equally undefined, as out of scope. A 2010 design document acknowledged that "very low adhesion levels would require a driver to adapt their driving style".

RAIB describes a "longstanding misalignment between the safety model for the Cambrian ETCS signalling system and the safety model for the class 158 rolling stock" as a possible underlying factor.

What did the operator miss?

TfW driver managers reviewed performance using only the OTMR data recorder; JRU data showing ETCS Service Brake Interventions had to be requested from Network Rail. With access to both, RAIB found SBI events occurred far more often than TfW believed, indicating drivers were following DMI speed profiles rather than adjusting for adhesion — which the Cambrian DMI specification cannot vary for low adhesion in any case. The ETCS driving simulator built at commissioning had fallen into disuse.

The driver's non-emergency radio call led the Machynlleth signaller to treat the situation as routine; an emergency call or ETCS emergency stop would not have prevented the collision but could have reduced its consequences.

RAIB issued nine recommendations covering sanding equipment standards, Class 158 automatic sander redesign, Cambrian overrun risk assessment assumptions, future ETCS specifications, DMI interaction, driver emergency training, railhead treatment research, interior fitting standards, and staff competencies. In February 2026, RSSB published a 55-page report on how professional driving policies must change under ETCS, and further work on adhesion-aware ETCS design is expected to follow.