The Thirtieth Berth: What PS Waverley Hitting Brodick Pier Tells Every Ship's Engineer About Known Failure Modes

Brodick, Isle of Arran, 3 September 2020 — 1640

The afternoon excursion had gone well. PS Waverley — the world's last seagoing paddle steamer — had taken 186 passengers down the Arran coast and was returning to Brodick pier. The vessel had berthed here 29 times before. The approach was familiar. The chief engineer had been at the controls during the berthing the previous week.

At 1640, Waverley entered Brodick Bay making 13 knots. The master briefed the quartermaster: enter the bay, shape a course to the eastern berth at the closed end of the pier, apply a 10° port turn in the final approach. Standard procedure.

At approximately 100 metres from the concrete wall at the closed end of the pier, Waverley was travelling at 7.3 knots. The master ordered slow astern. The chief engineer (C/E), manually controlling the engine from the engine room below, stopped the engine. He moved the controls to the astern position. The engine did not restart.

The high-pressure piston had stopped at dead centre — the full extent of its travel. Steam could not flow through the engine in either direction. The vessel was still moving toward the pier at over 7 knots.

What Dead Centre Means on a Triple-Expansion Steam Engine

Waverley is propelled by an inclined triple-expansion steam reciprocating engine built by Rankin & Blackmore in Greenock in 1947. It drives two feathering paddle wheels via three pistons connected to a single crankshaft. There are no lateral thrusters. The vessel's turning circle is 548 metres — almost three cable lengths for a 73-metre vessel. She can stop in her own length from 16 knots when the paddle wheels are put to full astern. But she cannot do anything without propulsion.

If the high-pressure piston stops at the top of its travel — dead centre — steam cannot act on it in either direction. The engine is locked. The recovery technique is to admit impulse steam to the medium-pressure or low-pressure cylinder to nudge the crankshaft, moving the HP piston off dead centre, and then restart normally.

The C/E was aware this could happen. It had happened once before, in open water, with no danger. He applied the impulse steam technique to the MP cylinder. The crankshaft did not move.

The master ordered half astern. Full astern. Double full astern — an emergency command achieved by ordering full astern twice in rapid succession. None of these could be executed while the HP piston was at dead centre. The engine was locked.

The C/E eventually admitted steam to the HP cylinder in the ahead direction to shift the piston, and then restarted the engine in the astern direction. The paddle wheels began to turn astern. Waverley was decelerating.

At 1646, the bow struck the concrete structure at the closed end of the pier at 2.8 knots. No warning had been broadcast to passengers. Many were standing on the open Promenade Deck and Aft Boat Deck, watching the arrival.

The Impact

When Waverley struck the pier, passengers who were standing were thrown forward. They fell onto the deck, made contact with fixtures and fittings, hit railings. Twenty-one passengers and three crew required treatment. Three people suffered serious back and pelvis injuries. Two were evacuated by helicopter to a mainland hospital. Nine more were taken to the island's hospital. Waverley's bow was buckled inward, with three holes in the bow plating above the waterline.

No-one was killed. The vessel remained afloat. By any measure, the outcome could have been worse.

What the Investigation Found

The Defect That Made It Harder

In the post-accident inspection on 15 December 2020, investigators found that the locking nuts securing the HP piston valve to its rod were loose. This extended the piston valve's travel by 5.4% — advancing the engine's timing and causing excessive steam to enter the HP cylinder ahead of top dead centre. On an inclined engine this provides 'cushioning' of the piston when it approaches top dead centre, which directly counteracts the impulse steam technique for dead centre recovery.

The C/E almost certainly did not know this. The defect meant the standard recovery technique — which would normally work — was operating against a significant mechanical resistance it was not expected to face.

Why the Defect Was There

In 2016 and 2017, most of Waverley's experienced engineering personnel left the company. The outgoing C/E left handover notes — which were then lost during a further turnover of senior personnel. In 2015, the vessel's piston valve rings had been replaced with rings made from the wrong material because the correct specification was unknown; the original rings had to be reinstalled while the correct material was researched. The accumulated engineering knowledge that kept Waverley's complex machinery running correctly had been held in people, not in systems.

By 2020 the vessel had qualified but less vessel-specific engineers. They had not been trained in Waverley's specific dead centre failure mode. The PMS was still being developed. The maintenance records were dispersed, incomplete, and not standardised. The vessel was operating under an Interim Safety Management Certificate.

The loose locking nuts were a product of four years of maintenance knowledge degradation. Nobody had set out to leave them loose. Nobody had a documented procedure that would have identified them as a critical maintenance item requiring verification at each season start.

Why No Procedure Existed

The ISM Code requires companies to identify equipment whose sudden operational failure may result in hazardous situations, and to establish measures promoting reliability and managing those situations. Dead centre events were known and occasional — they had occurred once, in open water, with no consequences. They had never been assessed for their impact in a closed-end pier approach, which removes all recovery options except the propulsion system.