Up Time Magazine- April/May 2008 - (Page 44) RCFA revisited….. When we analyze component failure we successively delete the non-applicable ones and zero-in on the one where a deviation occurred. Then we make sure we change the component configuration, metallurgy, work process, procedure, in whatever manner needed to avoid repeat failures. All the while, we recall that people make it happen. Figure 6 - Impeller damage caused by heavy-handed dismantling. thus the coupling had “hung up.” There also had been excessive misalignment, the pump discharge line had been anchored 30 feet from the pump discharge nozzle. It pushed and pulled the pump 3/8-of-an-inch relative to the electric motor driver. The piping design was also deficient, and the vibration monitoring was inadequate. With no periodic regreasing or the wrong grease on the coupling, maintenance action was flawed. Whichever segment you pick, operation, maintenance, project-technical – all are “people-affected.” People make it happen; people make all the difference in the world. When equipment is being dismantled, the failed components must be examined for tell-tale signs that explain a failure. This examination will be flawed if the failed part is mistreated, as shown in Figure 7. So, next time, don’t just change parts, but also identify why parts have failed. Next, take steps that avoid repeat failures. Along those lines, we want to acknowledge that many experienced and very competent people work at, for instance, your pump repair shop. I am completely convinced that an individual with 35 years of pump dismantling and re-assembly experience does this better and faster than I could do it. But is it perhaps true that he has replaced the same bearing several times in the last decade? Imagine what additional contribution he would make if you would encourage him to determine exactly why the bearing failed? And what to do so it doesn’t fail again? All he has to learn is that components fail due to only four reasons: “FRETT”— Force, Reactive Environment, Time, and Temperature. Again: that’s RCFA at the grass-roots level. Teach the maintenance person that fact and you will reduce failure frequencies. Again a reminder –“People Make It Happen”. Will you teach him? Will you hand him one of the many articles or books that describe “FRETT”? Will he read the articles or a book? experienced a bearing failure when it was being restarted a few hours later. While it had been sitting there, the various cylinders were not unloaded. Now, the operating and maintenance manuals of every manufacturer I know of state that the cylinders should be unloaded when a machine is stopped. If one doesn’t follow that requirement, the hot bearings are loaded over just a portion of their periphery and no cooling oil flows into the clearance, because now there is no such clearance. The mechanical folks claimed that, for years, they had stopped reciprocating compressors without unloading the cylinders. But I know that by that fateful day, the grouting under the compressor frame had been allowed to degrade. There were now two deviances (Figure 8). Because of this lack of support, the compressor now had a slightly bowed crankshaft and no cooling oil in the bearings. At least one of the bearings was now edge-loaded. These two deviations together caused a multi-million dollar outage event. We won’t spend any time discussing how a good mechanic can determine if too much or not enough cylinder lubricant is consumed, or why it rarely makes economic sense to use the same type of oil from both frame and cylinder lubrication. But people make these choices, and people have sometimes done it wrong – you can tell from repeat failures. Whenever you have a repeat failure, you PISTON ROD PISTON ROD PACKING CONNECTING ROD CRANK CROSSHEAD Figure 7 - Failed parts need to be examined. Even a failed bearing must be treated with due respect. CRACK-END CYLINDER VOLUME HEAD-END CYLINDER VOLUME CROSSHEAD GUIDE Earlier, we had explained that all machines fail because of oversights or errors that fit into one or more of only seven possible root cause failure categories. Now we also learn that components fail only due to four possible reasons: Force, Reactive Environment, Time, or Temperature. Figure 8 - When several deviances combine, machines will fail. Forgetting to unload the gas-containing portions of a compressor cylinder while simultaneously shutting down the lube supply will greatly increase bearing failure risk. Unload Your Compressors Next, a power outage caused a large reciprocating compressor to stop instantly. It have either not discovered the root cause or, if you actually know the root cause of the earlier failure, somebody (a person) has chosen not to do anything about it. It’s really that simple. april/may 2008 44
For optimal viewing of this digital publication, please enable JavaScript and then refresh the page. If you would like to try to load the digital publication without using Flash Player detection, please click here.