The valve on Maple Street had been serviced eight months ago. I pulled it Tuesday morning because a resident reported low pressure. The packing was binding—not seized, just wearing faster than the two-year replacement cycle assumed.
I repacked it, worked it through its range until it moved clean. Same supplier, same specs we'd used for years. But I'd been seeing this—valves wearing faster, pressure spikes during heavy rain, sections failing ahead of schedule. Not catastrophic. Just consistent acceleration.
The foreman asked if I wanted to flag it for the replacement list. I said no. The valve would hold now. We'd check it in six months, adjust if we needed to. You found what needed attention, you responded, you kept the schedule flexible enough to catch things before they failed completely.
By afternoon I was on Jefferson Street, where a relined section from three years ago had developed a circle break. The crew was cutting out the failed section, splicing in new pipe. The liner had been rated for twenty years. It made it three.
"Thought this one was good," the foreman said.
I figured we all did. But I'd been tracking the relined sections, and they weren't holding projections. Twenty-year extensions becoming fifteen, then ten, then less. The pattern was clear enough.
What wasn't clear was what to do about it.
You could flag every section for early replacement, but that meant pulling out infrastructure that was still mostly functional, spending money you didn't have, disrupting service for repairs that might not be necessary yet. Or you could wait for failures, respond when they happened, adjust the schedule based on what actually broke rather than what might break.
I'd seen both approaches fail. In the nineties, the utility tried to replace everything at once—bond measure, five-year plan, wholesale infrastructure overhaul. They got through maybe thirty percent before the money ran out and the political will collapsed. The rest of the system kept aging while resources went to the replacement zones. When the bond work stopped, the deferred maintenance backlog was worse than before they started.
The adaptive approach worked because it distributed the load. Small corrections across the whole system instead of big fixes in some areas and neglect in others. You couldn't replace everything, so you maintained everything, upgraded what you could, replaced what was done. The schedule flexed to meet conditions rather than conditions meeting the schedule.
I drove past the Oakwood section on my way back—new ductile iron main from last year, rated for a hundred years. The old pipe they'd pulled out was from 1952, had failed at seventy years instead of projected hundred. Not dramatically. Just steady deterioration, increasing break frequency, the system showing you it was done.
The new pipe would probably fail sooner than a hundred years too. But probably later than seventy. You worked with the time you had.
That night I drafted the quarterly report for the maintenance supervisor. Valve replacement cycle holding at eighteen months average instead of twenty-four. Relined sections averaging twelve years instead of twenty. New installations projected to ninety years instead of hundred.
The numbers showed compression, but they also showed response. We were catching things earlier, adjusting faster, maintaining closer. The schedule wasn't matching original projections, but it was matching actual conditions.
I added a section recommending increased maintenance capacity—two more crews, expanded monitoring, faster response times. Not emergency replacement protocols. Not wholesale overhaul. Just more of what was working: distributed attention, flexible scheduling, small corrections before big failures.
The system kept running because we stayed close to it. You couldn't predict exactly when things would fail, but you could watch for the signs, respond when they showed up, adjust the timeline to match what the infrastructure was telling you. The compression meant more work, tighter cycles, faster response. But the approach was sound.
I figured that was the work. Managing the aging. Catching failures before they cascaded. The schedule gave you a framework for that. You worked through the system methodically, you responded to what you found, you kept adjusting.
The valve on Maple Street would need service again in a year, maybe less. The Jefferson Street replacement would hold for now. I'd check both next month, then three months, then six. The work was careful. The system kept running.
Some sections would fail sooner than projected. You'd adjust for that. The schedule would compress further. You'd adjust for that too.
The work continued as long as the infrastructure needed it, which was always.