The Knowledge That Leaves With the Retiring Technician

Every utility has one. The technician who has been there twenty-five years, who knows which lift station has the relay that was added in 2009 and why, who can get the old programming software to run on the laptop nobody else can log into, who knows that the high level alarm at the west tank is a float that sticks and the one that matters is the transmitter. That person is retiring, and the industry is retiring them by the thousand. The control systems they maintained were built, in ways nobody quite intended, on the assumption that they would always be there.

What is actually lost

The drawings show what is connected. The program shows what the logic does. What neither shows is why: why the lag pump starts a foot higher at station six than anywhere else, why the ferric feed is paced to raw water flow rather than influent, why the dissolved oxygen setpoint in basin two is lower, why the alarm from the generator at the plant is ignored on Tuesdays. Some of those reasons are good ones, decided in a meeting fifteen years ago. Some are workarounds for a problem that was fixed a decade later. Nobody can tell which without the person who was there, and a new engineer who removes the wrong one learns why the hard way.

Why it was never written

It was not negligence. Writing down why a setpoint is where it is takes time, and the utility never needed it written because the technician could be asked. Documentation systems reward the drawing that is required for the project and not the note that explains the exception made afterward. And a great deal of the knowledge is not the kind that fits on a form: it is the sense that the trend looks wrong today, built from thousands of days of looking. The person did not hoard it. The organization never built a place to put it.

A capture program

1. 1

   Start early

   Three years before the expected retirement, not three weeks. Knowledge comes out in the course of work, not in an interview.

2. 2

   Verify the as-builts

   Walk every panel with the technician and the drawings, and mark every difference. The relay that is not on the drawing gets drawn and gets a note that says why it exists.

3. 3

   Annotate the settings

   Export every setpoint, band, delay, and alarm limit from every controller and SCADA, and go through them with the technician, writing the reason beside each one that has a story. The ones with no story are candidates for review.

4. 4

   Record the walk-throughs

   A phone camera and a site visit. The technician explains each site as they would to a new hire: what it does, what fails, where things are. The video goes in the engineering library. It is imperfect and it is far better than nothing.

5. 5

   List the things nobody else knows

   Ask directly: what would break if you were gone tomorrow. Software, passwords, cables, vendor contacts, the box of spares in the back of the truck. Write the list, then fix each item so that it is no longer true.

6. 6

   Pair them

   A second person on every site visit and every change for the last years, so that the judgment transfers by practice. It is the only way the trend sense moves.

7. 7

   Fix the workarounds

   Every workaround captured is a defect. Fix what can be fixed, so that the knowledge needed to operate the plant shrinks.

The permanent fix

Capturing what one person knows is triage. The durable fix is a control system that does not depend on anyone: a control narrative that says what the system does and why, drawings that match the panels and are revised with every change, program backups with the software and licenses to use them, standards that make every site alike so there is less to know, a credential vault instead of a memory, and two people who can do every job. Utilities that have those find that the retirement is an event rather than a crisis. Utilities that do not are one retirement, or one accident, from a control system nobody understands.

The other side

The technician who is leaving usually wants this to go well. They built the system, they are proud of it, and they know better than anyone what will happen if the knowledge goes with them. What they need is time set aside to do the capture instead of being asked to fit it around the work, a person to hand it to, and the sense that the organization values it. A utility that gives them that gets far more than a binder.