Your best process engineer retires in three years. Where does the knowledge go?

Process industries skew experienced. A large share of the deep operational knowledge sits with people who have been doing the job for decades and are closer to the end of their careers than the start. That is not a criticism, it is the shape of the workforce, and it means a predictable amount of hard-won knowledge is going to leave over the next several years. The only question is whether it leaves with the person or stays with the business.

It is rarely the documented facts. It is three harder things:

• Judgement. Not the facts, which are usually written down somewhere, but the sense of which fact matters when, and how to weigh a trade-off that the manual does not cover.
• Context. Why a process is run a particular way, what was tried before and abandoned, which customer has a quirk that is not in the system. The reasons behind the decisions, not just the decisions.
• The map. Where things are. A thirty-year veteran does not remember every detail, but they know exactly which folder, which colleague, and which old study to reach for. That index is often the most valuable thing of all.

• They ask for a data dump. Sitting someone down to "write everything down" produces a document nobody reads and that captures the easy, explicit knowledge while missing the tacit judgement that was the point.
• They happen too late. The project starts when the retirement date is announced, which is exactly when the person is least available and most checked out. The knowledge walks out on schedule.
• The output is not usable. Even good notes become another unsearchable archive. If nobody can find the answer at the moment they need it, the knowledge has not really been retained.

The common thread is that a documentation project treats knowledge as content to be dumped, rather than as something to be captured in use and made findable. Making the result searchable and answerable is as important as capturing it in the first place.

The version that works is undramatic. Capture happens as part of the normal work, not as a farewell project: decisions and their reasons recorded as they are made, the old archive brought into one searchable place, and the veteran’s answers captured when they are actually asked rather than in an artificial sit-down. The goal is not a perfect record. It is that when the person is on holiday, or gone, the next colleague can find the answer and, just as importantly, the reasoning behind it. That is the same source-of-truth idea that runs through the whole chemical and process industries approach.