I was in my first six months as Corporate Manager of Maintenance and Reliability for a large Miner when I got a late Sunday night call from my boss. There had been a haul truck rollover at one of our mines and he asked me to organize a root cause analysis and determine how and why it happened. There was considerable damage to the truck and the operator sustained some broken bones.
It was an early morning drive out to the mine. When I arrived, I contacted the Maintenance Superintendent and we discussed how the investigation would be handled and the Mine Manager met with me and gave his support. We assembled a Root Cause Team made up of maintainers and operators.
What happened? The truck was pulling an uphill ramp, fully loaded when the operator saw black smoke coming out of the engine compartment and believed the engine was on fire. The standard training was to get off the truck after setting the service and park brakes, turning off the engine and hitting the fire extinguisher system button. The operator did all of this in proper order and would have believed that the truck was properly secured. She exited the cab and headed for the ladder when the truck started to move backward. Immediately she ran for the cab and tried to restart the truck, but by that time the truck had picked up speed, hit the wall and flipped up on its side. The operator, with no seat belt fastened, was thrown across the cab and ended up pretty seriously injured. It took several hours to organize and perform a rescue. She ended up in the hospital.
Before I got to the mine that morning the truck had been righted and moved to the shop so the accident scene was really not preserved very well. The members of the team had no training in Root Cause Analysis so I started out by doing a quick introductory course in the way we would move through the accident. What that team did have that was very useful was good knowledge of maintenance practices and an understanding of the operation of the systems onboard the haul truck.
We went down to visit the truck after requesting some disassembly of the brake system. Here is what we found. One turbo charger had failed and come apart, the smoke was from engine oil exiting the bearings and landing on the hot engine block. That was the only damage in the engine compartment. There was no evidence of discharged fire retardant. It was only black smoke and there was no fire.
The fire extinguisher system squib had discharged. The system had two deployment points one in the cab and one at the bottom of the exit ladder. The line to the fire-retardant system was disconnected at the ladder. Service brakes were in good order and had pads that met thickness specifications. There were two park brake pucks, one on each side installed from different manufactures. One was confirmed to be set correctly and the other when activated did not make contact with the brake rotor.
From operator training the emergency shut down process was followed correctly and confirmed. There was no training that identified what to do when a truck moved backward and the operator was already out of the cab. Standard operating procedure was then and is now to stay in the cab when a machine is out of control but of course in the case of a fire it is required to get off the machine as quickly as possible.
What caused the accident when all the procedures were followed?
First, the turbo failure was the main physical root. Had this turbo not failed, and that was common in those days, there would have been no accident. There were a couple of other failures that when combined caused the roll back, injury and equipment damage.
When the operator set the service brakes and turned off the key a timer started. This timer controlled a solenoid operated discharge valve that released the brake accumulator pressure after a few minutes. It was a safety system designed to make the pressurized hydraulics safe to work on when doing brake work. When the accumulator discharged the service brakes released. The truck shifted and the one park brake engaged did not hold the truck. It was never determined if both park brakes would have held in these conditions, a loaded truck on a ramp.
I am not going to go into the whole root cause exercise here but suffice it to say there was plenty of lack of accountability and responsibility to answer to by mine management. The question of why the turbocharger failed was also answered and perhaps that should be the subject of another blog post.
This is not the purpose here as I am not trying to show how well Root Cause Analysis works. It is a powerful tool not used often enough. A few years after I got involved with wireless remote condition monitoring of haul trucks. Remarkably we learned to call impending turbocharger failures and many others before the components came apart by using fairly simple alarms. It was true that we could only catch these close to failure but we successfully stopped this type of failure over large fleets. Perhaps some lives were saved.
The value of the domain knowledge and the live data resulted in a prognostic solution being developed that could see turbo charger failures starting in some cases weeks ahead. As I have said before many times knowing what’s wrong, what to do about it and how long you have to act is the Holy Grail of Asset Management.
So, the story of moving to on condition maintenance went from identifying a failing component by hearing Mayday on the radio to actually predicting time to failure. This took over 10 years. Don’t think this is easy but you can take your data and your knowledge and solve problems using predictive analytics.
We are still learning just as computational power and predictive analytics are rapidly advancing.
Spring officially came to Wyoming yesterday and for once we had no snow. Don’t get suckered in as the spring storms are sometimes the worst.
Best wishes to those of you trying to stop failures. If there is any way we can help don’t hesitate to call.