Near-miss involving winch failure

The winch had been stripped down nine months before and the controls were newly built during the same period. One month prior to the incident, when the system was mobilised, the handling system and winch controls had been set up and tested by a reputable hydraulic company and had been given a clean bill of health. The brake system and all other functions were tested and passed.

Divers had been working at a depth of 270 feet (bottom depth being 460 feet) and had completed their bell run. They returned to the bell and secured for recovery for surface. The bell was cross-hauled at this time. As the bell was brought through the interface it appeared a little sluggish. However, it was retrieved out of the water. When it was about two feet from the bell catcher, the winch started paying out. The hoist lever was still in the ‘up’ position. However, the bell kept descending and the spring-loaded brake remained open. The technician turned the engine off, but the descent continued. The supervisor instructed the divers to blow the internal of the bell deeper, using onboard gas, while doing so himself. The bell ended up with an internal bottom depth of 390 feet.

At a depth of about 315 feet, the cross-haul wire checked the descent of the bell. The winch was restarted and the bell recovered to surface. Just below the surface, the deck crane was attached to the bell. The bell was then recovered, mated on and the divers locked through. All saturation operations ceased in order to determine the cause and rectify the problem. A hydraulic specialist consultant was dispatched to the site.

The investigation noted that the problem had been created over a period of time. One of the shuttle valves had an end cap on one side and oil was slowly seeping past the ball seal to the back of the shuttle valve and was trapped between the ball seal of the shuttle valve and the end cap. Eventually, this caused a build up of pressurised hydraulic oil, resulting in the ball seal being pushed against the opposite side of the shuttle valve, thereby sealing the return line to tank. This created a hydraulic line full of trapped pressure with no means of release and since both brakes were on a common line, it kept them open. Although there is a small vent hole in the counter balance valve, the ‘fail-safe’ spring on the brake cylinder kept the hydraulic line pressurised and acted like an accumulator. This also caused the counter balance valve to remain open putting the winch in neutral.

The root cause was an error in the plumbing of the control system.

Corrective measures were taken to first correct the plumbing, run a bleed line from behind the shuttle valve to the tank and to isolate the shuttle valve and the spring loaded brake. As a precaution, the shuttle valve was changed out.