Repair or replace: building a capital plan from your break logs

The decision this guide makes

You hand the board a list of mains to replace. Someone asks why this one and not that one, and why this year. If the honest answer is "it keeps breaking and I know my system," the ask stalls, and the pipe breaks again on its own schedule.

This guide turns the break records you already keep into a number for every pipe, a ranked list, and a repair-or-replace call you can defend line by line. It is the difference between "our system is old, trust us" and a case a board can vote yes on.

Start with the data you already have

You are not starting from zero. You have system maps or as-builts with pipe size, material, and install year. You have work orders and logbooks of past repairs. You have monthly reports, and you have the crew's memory of the streets that always give trouble.

Pull it into one place. A spreadsheet is plenty. Going forward, capture the same fields on every break: who worked it, what failed, when, where (which segment), why it likely broke, and how you fixed it, including the cost. Standardize the words, so "rusted through" and "old pipe" do not end up as two different causes.

Set aside the breaks that aren't the pipe's fault

A backhoe through your main is not a failing pipe. It is a bad day. Before you judge a pipe's condition, pull out the third-party hits: the backhoe through the main, the car into the hydrant. Keep them for tracking cost and response, but do not let them skew the reliability numbers. What is left reflects the pipe itself: its age, its material, its corrosion.

Put every pipe on the same footing

Raw break counts lie. Two breaks on 200 feet is a worse pipe than two breaks on two miles, but the count reads them as equal. Normalize by length and time:

breaks per mile-year = (breaks ÷ years) ÷ length in miles

Work a real one. A 1,200-foot main is about 0.23 miles. Three breaks over three years is one break a year. Divide by 0.23 and you get about 4.4 breaks per mile-year. Mains nationally run well under one break per mile per year, so 4.4 is a pipe in real trouble, not a hunch about a bad street.

Find the outliers with the 2x rule

Average the rate for similar pipe, the same era and material, then flag any segment running at twice that average or worse. If your 1970s ductile iron averages 0.5 breaks per mile-year, a segment at 1.0 or above is failing twice as often as its peers. That is your short list to investigate or replace. The rule is blunt on purpose: it points at the pipe, not at whoever complained loudest.

Run the repair-or-replace break-even

Turn both choices into a yearly cost and compare them.

Annual repair cost is the expected breaks per year times the cost per repair. Annualized replacement cost is the full replacement cost divided by the new pipe's expected life.

Work it: a $100,000 replacement that lasts 50 years costs $2,000 a year to own. If each repair runs $5,000 and the main breaks about once a year, you are spending $5,000 a year to keep a pipe a new one would cover for $2,000. Replace it. If it breaks once every five years, closer to $1,000 a year, repairs still win for now. A rule of thumb rides on top of the math: three breaks in five years on the same segment makes it a candidate.

Cost is not the only input

Risk is likelihood times consequence. The break rate is the likelihood. Consequence is who sits downstream. A main feeding a hospital, a school, or the only loop into a neighborhood earns replacement on fewer breaks than a dead-end serving ten homes. A critical line with two breaks can outrank a quiet line with five. Write down which lines are critical before you rank, so that judgment is on the record and not made live in the meeting.

Build the list and the stages it moves through

Rank by the numbers, break ties with criticality, and you have the start of a capital plan. Manage it in stages so nothing falls off the radar and nothing jumps the line:

• Watchlist: aging pipe with a break or two, worth keeping an eye on.
• Candidate: hit a written trigger, so it is time to plan replacement.
• Planned: in an upcoming budget or project.
• Funded: money secured, design underway.

Set the triggers before you rank. Common ones: a rate at or above twice the cohort average; three breaks in 36 months, or two on a critical main; or three-year repair costs reaching half of what replacement would cost per foot. Rules written in advance are what let you answer "why isn't my street first" without it turning personal.

Keep it live and keep the receipts

Use a rolling 36-month window, so a bad recent year shows up instead of hiding inside a ten-year average. Update a segment the day it breaks; it may climb overnight. Pull a pipe off the list when you replace it, then watch the system break rate to see the payoff.

Archive each year's list. When someone asks in 2028 why a main got replaced, the 2026 and 2027 rankings are your answer. A one-page top-ten, with a plain reason on each line, is the report that earns a board's trust: high break rate, or serves the school, or half its value already spent on patches.

What the work buys you

The same numbers that rank your pipes are the evidence funders ask for. State revolving-fund loans and infrastructure programs want documented need, not age and good faith. A 50-year-old main that broke 6 times in 3 years and cost $18,000 to patch is a fundable project the moment you can show it on a page. The log you already keep is the start of that case.

Run the numbers on a single main, weighing the running cost of repeated repairs against the cost to replace it, and take the one-page result to your next budget meeting.

Next on the ladder: talking to your board in 60 seconds.