What works better in troubleshooting plant operations — a structured and systematic approach or a more-random exploration of possibilities? Unfortunately, there’s no universal answer. One of the best troubleshooters I know uses an extremely systematic approach. It involves working in a consistent fashion through a very comprehensive set of questions and logic charts. In contrast, the single best plant troubleshooter I’ve met takes an apparently near-random approach of wandering around the plant until something catches his eye. The two vastly different approaches can produce excellent results. So, what can we learn from this?
To understand how different approaches to troubleshooting work, we must examine the types of problems and how to solve them. Let me offer a half dozen observations based on my experience.
First, most problems have common and recurring sources. Operating problems are like safety problems. For every 100 near-misses, a larger safety event occurs. The same is true with operating problems. For every 100 common problems, a single unusual one comes up. Using checklists and conventional logic diagrams to reach a conclusion works well with common problems. Indeed, empirical evidence clearly indicates that checklists play an important role in solving such problems.
The reality is that most problems that arise in the plant have been seen before. Well-documented checklists will identify these problems. Working through a checklist easily can solve 90% or more of plant problems. Many managers would call troubleshooting that achieves a 90–99% solution rate a great success.
Second, new hires out of school often are dropped into plant support roles and immediately given the responsibility for troubleshooting. Such inexperienced personnel need checklists. Someone without a lot of knowledge and time in the plant isn’t seasoned enough to randomly walk around a unit or talk to operators to glean something “interesting” that can identify a problem. Checklists are critical.
Training courses, books and articles are full of structured approaches. The reason is simple: a structured approach does solve most problems. However, it doesn’t solve all of them.
Third, context is important. Outside troubleshooters only get summoned when plant staff can’t solve a problem. They rarely are called in to deal with a straightforward problem. Plant personnel usually can find the cause of a simple problem.
Fourth, to extend the previous point, simple problems shouldn’t require outside troubleshooters. If an outside troubleshooter encounters a simple problem, it’s always a concern because it indicates that something prevented plant staff from solving the problem. The most common reasons simple problems don’t get solved are due to management or internal political issues. Political issues may include:
• A pet project or operation is implicated as a source of the difficulty.
• Someone with influence will lose power, money or status (either directly or indirectly) if the problem is solved.
• The belief that something is (or isn’t) the problem has reached near-mythic proportions, preventing staff from seeing the answer.
At this point, the troubleshooter faces a very difficult management task rather than a strictly technical one.
Fifth, structured approaches can’t easily identify unique problems. As already noted, flowcharts, checklists and other aids are extremely valuable in troubleshooting common problems. However, by their very nature, they aren’t set up to deal with rare and special problems. Based on observation, the more-random approach of looking around to see what’s unusual seems to work better with unique problems.
Nevertheless, most problems are common problems. That’s why a structured approach works so well. However, if you get to the end of the structured flowchart and still haven’t come up with a solution, switch tot he more-random approach.
Sixth, what expert troubleshooters have in common is a great depth of technical knowledge in many areas. What makes the less-organized approach work is both depth and breadth of experience. That’s why letting a new hire wander around will be unproductive. Decades of experience are required. Even very experienced troubleshooters keep a checklist of common problems in their head while they ramble around. But they also rely on pattern recognition to identify things that aren’t “normal” that demand further investigation. That experience, resulting from effort, time and willingness to learn, is what defines a great troubleshooter.