By the PICA Corp Engineering Team  |  Updated August 2026  |  Est. reading time: 9 min

Most water utilities aim to replace about 1 percent of their distribution network every year. That is a 100 year replacement cycle. Average life expectancy for US drinking water pipe sits at roughly 78 years and has been trending down. Those two numbers cannot both hold and still produce a functioning system. What gives is reliability: mains stay in the ground past the point anyone planned for, and the utility learns which ones were too far gone when a break wakes the on-call crew at 3am.

The way out is not more money. If you know which pipe is genuinely deteriorated and which has thirty good years left, the 1 percent you can afford to replace goes to the right 1 percent. That requires condition assessment data gathered at the right points in a main’s life, not one panicked assessment after the failures start.

---

The replacement math that does not work

On a mid-sized utility with 800 miles of main, a 1 percent replacement rate means 8 miles a year. At a 78 year average life you need closer to 10 just to stay level, and that assumes the network was installed evenly over time. It never was. Most North American systems have large cohorts installed during post-war expansion, so the need arrives in waves rather than a steady drip.

The funding side offers no relief. The EPA’s 7th Drinking Water Infrastructure Needs Survey put 20 year national needs at $625 billion, two thirds of it in transmission and distribution pipe. No utility is going to rate-increase its way past that.

The lever that is actually available is accuracy. The AWWA Partnership for Safe Water sets an optimization target of no more than 15 main breaks per 100 miles per year. Hitting it while replacing 1 percent of the network annually is possible, but only if the 1 percent is the right 1 percent. That is the argument for building a pipeline condition assessment program rather than an age-based replacement list.

---

What protection looks like at each stage of a water main’s life

Utilities tend to treat inspection as something you do to old pipe. That framing costs money, because the most useful inspection a utility ever runs is often the earliest one.

Years 0 to 25: build the baseline almost nobody builds

New pipe does not need a condition assessment. It needs a recorded starting point. A wall thickness baseline on a representative sample of a new transmission main gives you something no amount of money buys later: known original condition for that pipe in that ground. Twenty years on, a second measurement yields an actual corrosion rate instead of a textbook assumption. Most utilities skip this, and the fallback is inferring original wall thickness from manufacturer specifications, which introduces error in exactly the calculation that matters most. This is also when installation defects surface, and a visual run with CCTV, laser and lidar documents joint condition and liner integrity before anything has aged.

Years 25 to 50: the first honest condition question

Deterioration becomes measurable here but not yet urgent, which makes this the cheapest time to find out about it. Cast iron in aggressive soil can be well into graphitic corrosion by year 30 with no surface symptoms. PCCP from the problem manufacturing era can have wire breaks accumulating quietly. The right action is targeted, not exhaustive: pre-screening on critical runs, with follow-up only where the pre-screen flags something. Utilities that wait for the break rate to climb are gathering condition data through failures, which is the most expensive collection method available.

Years 50 to 75: the decision window

The pipe is now inside the range where replacement is a live question, and this is where inspection earns its budget. A full water main inspection sorts the network into three groups: replace now, rehabilitate, and leave alone. That third group is usually larger than anyone expects, and it is where the savings come from. The mistake at this stage is inspecting only the mains you already suspect, because suspicion is built from break history, and break history is biased toward pipe that has already failed rather than pipe that is about to.

Years 75 and beyond: running on evidence rather than faith

Plenty of North American mains are past 75 years and performing acceptably. Keeping them in service is defensible, but only as a data-backed decision. The inspection interval should tighten here, because corrosion compounds and the margin between current and minimum allowable wall thickness has narrowed. A main cleared at year 60 should not be assumed clear at year 85.

---

Age is a proxy, not a condition

Everything above is organized by age because age is how utilities organize capital plans. Age itself causes nothing. Three factors do the actual work.

The ground the pipe sits in

Soil resistivity, chloride and sulphate content, and moisture cycling drive external corrosion far more than elapsed time. Ductile iron in highly corrosive soil fails at roughly six times the rate of identical pipe in less aggressive ground. Two mains installed the same week can be decades apart in condition because one crosses a former marsh.

Material and manufacturing era

Grey cast iron, ductile iron, steel, PCCP and bar-wrapped pipe deteriorate through different mechanisms on different timelines. Production period matters too. Certain PCCP manufacturing years are documented as higher risk because of hydrogen embrittlement in the prestressing wire.

Operating history

Pressure transients from pump starts and rapid valve closures fatigue joints and fittings, and high velocity sections erode internally. A main that has spent forty years on a pump discharge does not belong in the same replacement tier as one of identical age on a gravity-fed section.

---

Why break history is a bad replacement trigger

Break history is the most common input into North American replacement decisions and one of the weakest. It is a lagging indicator by definition. A main with three breaks in five years tells you something has gone wrong, but says nothing about the main beside it that has not broken yet and is in worse shape. Ranking capital projects by break count funds yesterday’s problem.

The alternative is direct measurement, and no single technology measures everything. PICA’s tiered NDT approach exists because the tiers answer different questions:

• The Navigator multi-sensor acoustic sphere travels through a live, pressurized main and identifies leaks and gas or air pockets. It cannot measure how much wall thickness remains on pipe segments, so it triages rather than diagnoses.
• CCTV, laser and lidar document internal condition, liner status and joint defects on a dewatered line from 6 to 108 inches. They see surfaces, not what is happening inside steel or concrete.
• Near Field Technology (NFT) quantifies broken prestressing wires in PCCP and bar breaks in bar-wrapped pipe, on large-diameter concrete pressure pipe up to 136 inches out of service.
• Remote Field Technology (RFT) maps remaining wall thickness along metallic pipe and, on concrete pressure pipe, detects broken prestressing wires, cylinder wall loss, and loss of pre-load on pipe segments. It reads through cement mortar linings without removing them.
• Handheld EM and UT verify tool findings at excavated sections and spot-check where in-line inspection is impractical.

A utility making a multi-million dollar replacement decision on a single data source is working from an incomplete picture.

Which tier fits also depends on what the pipe will allow. RFT inspects live, pressurized mains from 2 to 36 inches, though flow must be reduced to hold tool speed in the 5 to 20 feet per minute range, so the run is not operationally invisible. On 6 and 8 inch cast iron and ductile iron mains, the HydraSnake launches through a fire hydrant, removing the excavation that normally makes small-diameter inspection uneconomic. Above 36 inches the main has to come out of service, which means planning around a shutdown window years ahead rather than scheduling on demand.

For PCCP and concrete pressure pipe the sequence differs: acoustic pre-screening is usually skipped, because large-diameter concrete mains do not tend to develop the small leaks acoustic tools detect. When one fails, it fails abruptly. The question there is wire and bar break distribution, not leak location, and PICA’s guidance on avoiding PCCP failures covers that sequence in detail.

---

What inspection costs against what replacement costs

Electromagnetic inspection of small and medium metallic distribution mains from 4 to 24 inches runs from several thousand dollars up to roughly $40,000 per mile. Advanced NDT on large-diameter concrete pressure transmission mains runs $100,000 to $200,000 or more per mile, depending on diameter, access complexity, number of tiers, and mobilization distance. Across a full program the honest range is tens of thousands to hundreds of thousands of dollars per mile.

Those are not small numbers, and anyone presenting inspection as cheap is misleading you. The case is comparative. Full replacement of large-diameter transmission main costs millions per mile once excavation, traffic management and surface restoration are counted, and an unplanned failure on the same segment typically costs ten to fifty times what inspecting it would have. The largest return from condition assessment is usually not the failure you avoided. It is the four miles you were about to replace on schedule and did not need to.

PICA delivers preliminary analysis of wall loss defects of 50% or greater within about a week of the run, with final findings following off-site QC typically delivered in 8 to 12 weeks. PICA analysts walk the utility’s engineering team through results segment by segment, so the data arrives in a form the capital plan can use.

---

Frequently asked questions

When should a water main be replaced instead of repaired?

Replace when measured wall loss or structural damage is distributed along the pipe rather than concentrated at one defect. A single corrosion pit or one damaged joint is a repair. Wall loss exceeding 50 percent across multiple pipe segments, or clustered wire and bar breaks in concrete pressure pipe, means the deterioration mechanism is active along the whole run and spot repairs will not keep pace. Inspection data is what separates the two cases. Break history alone cannot, because it only reports where the pipe has already failed.

Does a water main need to be replaced just because it is 50 years old?

No. Installation year predicts nothing reliable about remaining wall thickness. Two mains installed the same year in the same city routinely show completely different condition depending on soil chemistry, bedding, pressure history, and manufacturing quality. Ductile iron in aggressive soil can fail at roughly six times the rate of identical pipe in benign ground. Utilities that replace strictly by install date spend capital on pipe with decades of service left while leaving genuinely compromised segments in the ground.

How do you calculate the remaining useful life of a water main?

Remaining useful life comes from a measured wall thickness baseline compared against a second measurement taken years later, which gives an actual corrosion rate for that specific pipe in that specific ground. Project that rate forward to the minimum wall thickness the operating pressure requires and you have a defensible replacement year. A single inspection gives current condition. Two inspections spaced several years apart give a rate, and the rate is what capital planning actually needs.

What is the first inspection a utility should run on an aging main?

For most distribution mains, pre-screening with the Navigator multi-sensor acoustic sphere is the sensible first step. It travels through the pressurized line while the main stays in service and identifies leaks and gas or air pockets, which narrows a long run down to the segments worth investigating further. It cannot measure how much wall thickness remains on pipe segments, so it is a triage tool rather than a condition assessment. On large-diameter transmission mains the acoustic step is usually skipped, because those pipes tend to burst rather than develop small detectable leaks.

Is it cheaper to inspect a water main or replace it?

Inspection is far cheaper on a per-mile basis. Electromagnetic inspection of small and medium metallic distribution mains from 4 to 24 inches runs from several thousand dollars up to roughly $40,000 per mile. Advanced NDT on large-diameter concrete pressure transmission mains runs $100,000 to $200,000 or more per mile. Full replacement of large-diameter transmission main runs into the millions per mile once excavation, traffic control, and surface restoration are counted. The real value is not the unit cost difference but the avoided spend on pipe that did not need replacing yet.

How many broken wires does it take before a PCCP main is at risk?

Risk rises sharply when breaks cluster rather than scatter. Five or more adjacent broken wires or bars in one location matters far more than the same total spread randomly along the pipe, because adjacent breaks remove the prestress compression holding the concrete core in place over a continuous length. This is why counting alone is insufficient. The inspection has to locate breaks precisely enough to show whether they are clustered, and Near Field and Remote Field electromagnetic tools both report position along with count.

Can inspection extend the service life of a water main past its design life?

It can extend the period a utility is willing to keep the main in service, which amounts to the same thing operationally. Design life is an assumption made before the pipe went in the ground. Measured wall thickness is a fact about the pipe as it exists now. Utilities routinely find mains at 80 or 90 years old with wall loss well inside acceptable limits, and those mains stay in service on the strength of the data rather than being replaced on schedule. The condition assessment does not make the pipe stronger. It replaces guesswork with a number.

---