Failure of winch brake on air diving bell system

One of their vessels was undertaking air diving work using a wet diving bell. When the wet bell, which was deployed at deck level over the side of the DSV, was awaiting the diver to enter, it went into free-fall dropping 6m in air and a further 15m in water on to the clump weight. The wet bell was unmanned, nobody was injured although the diver was within seconds of entering it.

The subsequent investigation determined that the wet bell free fall was a direct result of a failure of the secondary band brake on the winch. The fault was clearly visible as the band brake adjustable actuating rod had snapped, rendering the brake totally ineffective. The winch control position is remote from the winch, hence the operator was not aware of the impending failure prior to it happening.

The primary brake on the winch, the hydraulically driven motor system, had supported the wet bell for approximately 5 or 6 minutes from the time the winch had been operated and normally the secondary mechanical brake would have been operational during this decay period. Since the secondary system had failed, when the hydraulic pressure decayed beyond the point that the motor would be held, the weight of the bell caused the winch to freewheel allowing the wet bell to free-fall from deck level into the sea.

The secondary brake is designed to be fail-safe in that it is applied by spring pressure and released by hydraulic pressure when the winch is operated. A hydraulic ram pushes against the spring to release the brake. The ram’s linear action is applied to a lever to increase the braking force and therefore the lever moves through an arc. To permit the slight rotation of the lever, the band brake adjustable actuating rod is connected to the lever via a hinge pin/turnbuckle type arrangement. It was found that this hinge pin had become seized due to lack of lubrication and therefore had not been rotating as the brake was applied and released. This resulted in the brake band actuating rod being subjected to a bending moment as well as the tensile force for which it was only designed. This regular bending, small though it may have been, was sufficient to eventually cause the 24 mm diameter threaded adjustable rod to snap due to fatigue.

The root cause of the secondary brake failure was attributed to lack of maintenance. Although some of the winch band brake mechanism was being lubricated on a periodic basis, two grease nipples to lubricate the turnbuckle type hinge pin were both difficult to see, rather inaccessible and consequently had not been greased.

The primary brake, the hydraulic motor, ‘failed’ after a period of 5 or 6 minutes from the time when the winch was last operated. This failure was due to the pressure decay in the hydraulic system as all hydraulic motors leak internally. The hydraulic system should have been configured to include an accumulator and/or valve(s) to allow hydraulic oil, at pressure, to remain behind the motor and thus only permit the winch to creep as oil pressure decays, rather than freewheel out of control.