We never turn down work — if the customer wants it repaired, we repair it well. But there's a point where equipment is too old or too run-down for repairs to serve the customer, and spotting that point and saying it out loud is part of the job. This is how to recognize it, how to run the numbers, how to present it honestly, and when a failure should trigger a whole-system review instead of a like-for-like swap.
Cumulative repair spend (each dot = a service call) Replacement cost
Aging equipment fails bigger and more often — the curve bends upward. Our job is to see the crossover coming and say so before the customer pays for a new system without getting one.
Video overview
10:07
Narrated overview of this module. The article below has the full procedure and the numbers.
01
The posture: it's service, not sales
The doctrineWe never turn down work — and we never let a customer unknowingly pour money into equipment that's past saving. Both halves of that sentence matter.
Recommending replacement at the right moment is one of the most valuable things a tech can do for a customer. Recommending it at the wrong moment — too early, too pushy, or without evidence — burns the exact trust that makes the recommendation worth anything. The rules of the road:
Evidence, not vibes
The recommendation stands on the service history, the readings, and the photos we've been logging on every PM — not on "this thing looks rough." A customer shown their own repair timeline usually reaches the conclusion themselves.
The customer decides — always
We present options and honest expected outcomes. If they choose to repair a dying machine, we do that repair well — no sandbagging, no "told you so." Our job is that they decide with open eyes.
Recommend early, not at the funeral
The best replacement conversation happens at a planned PM visit months before the failure — when the customer can budget, schedule, and avoid downtime. Waiting until the emergency call makes every option worse and the conversation feel like a stickup.
Document every recommendation
Recommendation made, options quoted, customer's choice — on the Salesforce work order, every time. When the machine fails eight months later, the record shows we flagged it, protects the relationship, and makes the next conversation easy.
02
The signals that equipment is past repairing
No single signal decides it. Two or three together are the pattern — and most of them show up in the PM trend long before the machine dies:
The repair-cost ratio
When a single repair runs past ~50% of replacement cost, the conversation is mandatory. And weight it by age: 40% on a 3-year-old pump is a repair; 40% on a 14-year-old pump is a down payment on the next failure.
Failures compressing together
The hero chart's curve in real life: calls that used to come yearly now come quarterly, and each one is bigger. Frequency trend beats any single estimate — pull the site's service history and count.
Parts are getting hard to find
Discontinued control valves, obsolete circuit boards, "eBay-only" parts. When a repair depends on scavenged components, every future failure means extended downtime the customer hasn't priced in.
The failure is structural, not component
Tank liners delaminating, vessels rusting from inside, frames corroded, manifolds brittle. Component failures are repairable; structure failures are the machine itself dying — a new pump bolted to a rotting skid is money buried.
It can't do today's job anymore
The business grew, the water changed, or the duty doubled — and the system is undersized or mistreating for what it now faces. Repair restores yesterday's capability; the customer needs today's.
Safety, code, or efficiency gap
Unsafe wiring, non-compliant backflow, or a machine burning double the power and water of a modern equivalent. Sometimes replacement pays for itself in operating cost — that math belongs in the quote.
Typical service lives — the age context
Age alone never condemns a machine, but it sets expectations. Rough working ranges — actual life swings widely with water quality, duty cycle, and how well the PM was kept up:
EquipmentTypical lifeWhat usually ends it
RO membranes
3–7 yrsfouling/scale history; chlorine exposure ends them instantly
Chemical metering pumps
5–10 yrsdiaphragms & valves are consumables; body/drive wear ends it
10–15 yrsbearings & seals repairable; windings/casing wear is the end
Softener control valvesresin similar, water-dependent
10–15 yrspistons/seals repairable until the body wears or parts dry up
FRP media vessels
15–20 yrsliner failure, fitting cracks — structural, not repairable
Control panels / PLCs
15–20 yrsthe boards outlive their manufacturers — obsolescence, not wear
Use the table as context, not a verdict. A 12-year-old softener valve with clean history and available parts is fine. A 12-year-old valve on its third rebuild with discontinued internals is a different conversation — same age, different answer.
03
Run the numbers before you open your mouth
The conversation goes better when it starts with arithmetic the customer can check. Three numbers, all from things we already know:
Repair-or-replace advisor
Field guidance to shape the recommendation — not a substitute for a quote from the office.
This repair vs. replacement
29 %
Spent + this repair, 24 mo
49 % of new
Service life used
80 %
Call frequency
1.5 / yr
—
Heuristics: repair >50% of new, or 24-month spend approaching new, or >90% life used with rising calls → replacement discussion is owed to the customer.
Don't forget the costs that aren't on our invoice: the customer's downtime, hauled water, lost production, and emergency after-hours premiums. On a commercial account, two days of downtime can outweigh the entire repair-vs-replace price gap — say that number out loud when you present options.
04
How to present it: the three-option quote
The doctrineWe don't say "you should replace this." We say "here are your options, here's what each one buys you, and here's what we'd do." Then we do whichever one they pick — well.
The standard shape of the conversation, whether it's at a PM visit or a breakdown:
Option 1 · Fix today's failure
The minimum to restore service now. Priced honestly, with an honest caveat: this repairs the broken part, not the machine's age — here's what's likely next and roughly when, based on the history.
Buys: uptime now · lowest cost today
Option 2 · Repair + refresh
Fix the failure and the known next-to-fail items in one mobilization — the worn seals, the tired pump, the aging media. Costs more today, cuts the return-trip risk substantially.
Buys: pushes the crossover out 1–3 yrs
Option 3 · Replace
New equipment — like-for-like or re-engineered (next section). Present with the crossover math: what they've spent, what this repair costs, what new costs, and what modern efficiency saves per month.
Buys: reset the clock · warranty · right-sized
Then the two habits that make this work long-term: quote all applicable options even when you're sure which they'll take — being shown the cheap option builds more trust than being steered past it. And log the recommendation and their decision in Salesforce before you leave. That record is tomorrow's easy conversation: "when we were out in March we flagged this — want to revisit Option 3?"
05
Replace with what they have — or review the whole system?
Once replacement is on the table, there's a second decision: swap the same thing back in, or step back and ask whether the system is still right? The failure that triggered the conversation is often the symptom of a design that no longer fits.
Like-for-like swap — when the design still fits
One component failed; the rest is healthytrend data on everything else looks clean
Water hasn't changedsame source, and recent tests match the original design basis
Demand hasn't changedsame flows, same duty, same use
The design is recent and was righta young system losing one part to bad luck, not old age
Fast, cheapest, minimal downtime — the right call when it's true
Whole-system review — when any of these are true
Multiple components at or near end-of-lifereplacing one leaves it surrounded by the next failures
The water changednew well, source switch, or the raw water drifted from the design basis
The demand changedbusiness grew, occupancy doubled, new process load
Chronic problems the repairs never curedpersistent scaling, fouling, or quality misses = design mismatch, not bad luck
Technology or efficiency leap availablemeaningfully better recovery, energy, or controls since it was built
The failure is the opportunity to fix the design, not just the part
The trap to name for the customer: a like-for-like swap into a mismatched system replaces the part that was killed by the mismatch — and the new one dies the same death. If the chronic-problem history says design, quoting a straight swap without saying so isn't doing them a favor.
How we run a whole-system review
It's a mini re-engineering, and it leans on everything else in this library. The tech's job is the field data; the office turns it into a design:
Fresh water analysis — never design on the old one. Pull a current sample of the raw water and get the certified report. The original design basis may be years stale, and the water decides everything downstream.Same analyte list as the antiscalant article — that report feeds the projection.
Measure today's real demand. Actual peak and average flows, daily volume, pressure requirements — measured or from meter/logger history, not the customer's guess and not the old nameplate.
Audit every stage against the water and the demand. Walk the treatment train asking two questions per stage: is this barrier still the right barrier for this water, and is it sized for this demand? Condition-score each piece (good / aging / end-of-life) with photos.
Capture the site record. Nameplates, layout photos, pipe sizes, electrical service capacity, drain capacity, footprint constraints — everything the office needs to design without a second site visit.
Hand off for engineering. Field packet goes to the GM/engineering: new analysis + measured demand + condition scores. They run the projection and sizing (Proton for RO, softener/carbon sizing for the rest) and build the proposal.
Present it as a system, phased if needed. Many customers can't replace everything at once — a good review supports a phased plan: what must go now, what can wait a year, in what order, so each phase builds toward the final design instead of fighting it.
The one thing to remember
Never turn down the repair — and never let the customer buy it blind. Show the history, quote the options, and let the crossover math speak.
Axis Water Technologies · 855-443-2947 · info@axiswater.com · axiswater.com · 3462 W Loop 289, Lubbock, TX 79407Axis technician training library · content as of October 1, 2026
