design, but not reliability or availability. It won’t blow up, but it will fail at unknown intervals. Based on the design, you don’t really know what percentage of the time the machine will need to be shut down for maintenance.”
Continuing, Joe pushed another document toward the Plant Manager saying, “This is the original bid comparison. This machine was the least expensive of all the alternatives. I am familiar with two of the other more expensive alternatives. They were selected and installed at two of our other plants. Both are experiencing much higher reliability and significantly lower maintenance costs.”
The Operations Manager commented, “If we selected the most expensive choice for every component, we would never get any new plants.”
Joe responded, “The fifteen year lifecycle cost for this choice will end up costing more than twice as much as the closest alternative. And that is without considering the value of lost production. Again, a comprehensive lifecycle cost comparison was never made during the design. In other words, the cheapest choice up front is the most expensive choice over the long haul.”
By this time, the participants in the meeting other than Joe were giving each other nervous looks and were squirming in their seats.
Joe withdrew another document from the file and pushed it toward the Plant Manager, saying, “This is the record of alignment measurements completed during construction. What the records suggest is that there was an unusually high piping load on the inlet nozzle when the compressor was placed in service. The inlet piping is 24-inch diameter and the area it passes through is quite congested. Apparently re-routing the piping was viewed as too expensive. Therefore, the machine has had to deal with high nozzle stress for its entire life.”
“But that doesn’t directly cause failures,” responded the Assistant Plant Manager.
“Well, stress translates into strain, and strain translates into displacement, and displacement between stationary and rotating components results in more wear and early wear out,” explained Joe. “In other words, it is a ‘defect’ in the system.”
Pulling yet another small stack of papers from the manila folder, Joe described their content. “These reports cover a series of events that resulted in emergency shutdowns of this machine. You can see that most of them were situations when the feed drum was allowed to exceed high level. It appears that in several situations the machine ingested at least some liquid.”
The Operations Manager took the sheets from the Plant Manager’s desk, saying “Now you are trying to blame the operators. I can assure everyone here that my operators do a good job and this was not their fault.”
“I am not trying to blame anyone,” said Joe. “I am just trying to describe some of the things that affect the reliability of this machine. It really doesn’t matter to this machine if it was slugged with liquid due to operator oversight or a malfunctioning suction drum level instrument.”
About this time, several of the people in the meeting began to glaze over. Apparently the meeting was not going as they had expected. They wished they were elsewhere.
Joe again reached into his file and pulled out two documents. He was sensing that his air time was running out and, if he wanted to make a point, he would need to do so quickly. “This first document is a record of predictive and preventive maintenance,” started Joe. “it shows that roughly fifty percent of the time the recommended PM is not being done.”
Joe flipped to the second document, saying, “This is a record of the work that was recommended for the last turnaround and the work that was actually completed. Our analysis showed that several components were at the end of their life and several other components would not survive another run. You can see that the decision was made to defer the overhaul from the turnaround. Here you see that when the bearings failed, the other components with limited life were not changed because of the desire to get the machine back as quickly as possible. The decision to defer the overhaul from the turnaround caused the first shutdown. In turn, the decision to make only a partial repair caused the second.”
By this time, the Plant Manager had seen enough. He said, “I will have to take responsibility for those decisions. I made those choices.”
Again sensing that his audience was running out of patience, Joe pulled out the last two documents from the folder. The first was a document from a recent project and the second was a copy of a recent budget detail.
Pointing to the project document, Joe began by saying, “The first of these documents describes some communications that occurred during the recent capacity expansion project. You may or may not be aware that the corporate project management process does not include any design-for-reliability or reliability analysis steps. As a result, we in the plant reliability department performed a comparison of the pre-project reliability and compared it to the reliability of the proposed post-project configuration. Our calculations showed that the proposed post-project configuration would be less reliable. This is the result of the redundant electrical feeder to this machine being used to supply a new load. We made a recommendation to the project manager that the redundancy be maintained, but was told that was beyond the project scope.”
Once again, the Assistant Plant Manager interjected, “That was analyzed and viewed as an acceptable risk.”
Joe responded, “One problem is that risk is really a measure of hidden cost. In some cases, the costs appear later; in other cases, the costs appear sooner. In this case, we have already experienced a failure associated with this choice. Although the failure was charged to this machine, it was really a failure of the electrical system that supplies the motor — and the decision making process”
Joe continued, referring to the final document, “This final document is a copy of the budget detail for last year. I have highlighted a line item that proposes replacement of several outdated controls on the machine. Also some of the instrument wiring shows deterioration and should be replaced as a part of our plant renewal initiative. As you can see, the line item was struck from the budget and will need to be proposed for some later time.”
“Is that all” said the Plant Manager, “or is there anything else?”
“That’s about it,” answered Joe.
“Well, I can see how some of these things might have an effect, but I guess I don’t completely follow what you are saying,” said the Plant Manager.
Sensing that several of the participants were becoming defensive, Joe started slowly to explain, “At the beginning of the meeting, you said that this machine was not meeting expectations. For reliability, realistic expectations should be based on an assessment of risk. Risk is a measure of the likelihood that an undesired event will occur. In this case, the undesired event is a shutdown of this machine. Each choice and action during the life of the machine will affect the risk of failure. In some cases, choices will improve reliability, some choices will maintain the same level of risk, and some choices will deteriorate reliability and increase the risk of failure. Each of the choices I have mentioned today tended to increase the risk of failure and reduce reliability. Thinking in terms of ‘what we have a right to expect,’ we should think of reducing our expectations or investing in efforts that will enhance reliability.”
“Well, Joe, I think I speak for everyone here when I say that we appreciate your efforts in assembling the information you shared with us today. I am sure you have other things you need to do and we have taken enough of your time,” said the Plant Manager, looking around the room. “If everyone else would hang around a few minutes, Joe, you can get back
