A pump has failed and there are two quotes on the desk: one to rebuild it, one to replace it. The rebuild is cheaper today. That is almost the only thing in its favour, and it is not always enough. Here is the framework we use to make that call, and the four numbers that usually decide it.
The 50 per cent rule, and why it is only a starting point
The conventional guidance is simple: if the repair costs more than half of a new pump, replace it. It is a reasonable first filter and it catches the obvious cases. But it compares only capital, and capital is the smaller part of what a pump costs you.
Over a fifteen-year life, a typical industrial pump costs roughly 10 per cent in purchase price, 15 per cent in maintenance, and 75 per cent in electricity. Any decision made on purchase price alone is being made on a tenth of the evidence.
Four numbers that decide it
1. Efficiency loss
Measure the pump’s actual delivery against its original curve. A pump that has lost 15 to 25 per cent of its head through worn wear rings and eroded impellers is burning that difference in electricity every hour it runs. A rebuild that restores clearances recovers it; a rebuild that only replaces bearings and seals does not.
Run the arithmetic: on a 15 kW pump running 4 000 hours a year, a 15 per cent efficiency deficit is roughly 9 000 kWh a year. That figure frequently exceeds the entire price difference between repair and replacement within two years.
2. Failure frequency
One failure is an event. Three failures in two years is a pattern, and patterns do not respond to repairs. Repeated failure of the same component almost always means the pump is wrong for the duty rather than worn out – oversized and running far left of its best efficiency point, cavitating, or fighting a system that has changed since it was installed.
In those cases a replacement that is correctly sized solves a problem that no number of rebuilds will touch. Work through the head calculation for the system as it is today, not as it was when the pump was bought.
3. Age and spares
| Pump age | Default position |
|---|---|
| Under 5 years | Repair – and investigate why it failed early |
| 5 – 10 years | Repair if the duty is still right and efficiency can be restored |
| 10 – 15 years | Compare whole-life cost properly; replacement often wins |
| Over 15 years | Replace unless it is a specialist unit with good spares support |
Spares matter more than age. A twelve-year-old pump from a manufacturer with local stock and a service agent in Gauteng is a better repair candidate than a six-year-old import with a ten-week lead time on an impeller.
4. Downtime cost
What does an hour without this pump cost? For a farm reservoir with three days of storage, very little. For a dairy, a lodge in season, a process line or a municipal lift station, a great deal. Where downtime is expensive, the calculus shifts towards replacement with the old unit kept as a rebuilt standby – which gives you both a new pump and a spare for less than the cost of two pumps.
A whole-life comparison
A 22 kW booster pump in a Centurion office park, eleven years old, failed bearings and visibly worn impellers. It runs 5 000 hours a year. Electricity at R 2.40 per kWh.
| Repair | Replace | |
|---|---|---|
| Capital today | R 38 000 | R 96 000 |
| Measured efficiency | 58 per cent after rebuild | 76 per cent with a correctly sized new unit |
| Annual energy | 96 000 kWh | 73 000 kWh |
| Annual energy cost | R 230 400 | R 175 200 |
| Five-year total | R 1 190 000 | R 972 000 |
| Expected remaining life | 4 – 6 years | 15 years plus |
The repair passes the 50 per cent rule comfortably – it is 40 per cent of the replacement cost. It is still the wrong decision, because the rebuild restores the pump to its own worn-in efficiency rather than to a modern, correctly sized duty point. This is the most common way the 50 per cent rule misleads people.
When repair is clearly right
- The pump is well matched to the duty and its efficiency can be restored to near original.
- A single identifiable cause failed – a seal, a bearing, a capacitor – and the rest of the unit is sound.
- It is a large, specialist or imported unit where replacement means months of lead time.
- The pump is under ten years old and the failure was an event, not a pattern.
- Capital is genuinely constrained and the pump runs few hours a year, so energy cost is not the dominant term.
When replacement is clearly right
- Efficiency cannot be restored – a cracked casing, a scored shaft, a pitted volute.
- The pump has failed repeatedly, which means the selection is wrong.
- The duty has changed: a bigger building, a deeper borehole, a new irrigation block.
- Spares are obsolete or the lead time is unacceptable.
- Energy modelling shows a modern pump plus a drive pays back inside three years.
The third option people forget
Replacement does not have to mean buying new. A refurbished pump from a reputable rebuilder, supplied with a warranty, sits between the two and often makes sense for standby duties or seasonal work. Rental is the other option for short-term or emergency needs – it converts an urgent capital decision into an operating cost and gives you time to specify the permanent solution properly. See our equipment rental page.
Whichever way you go, fix the cause
A pump that failed because it was cavitating, dry running, short cycling or undersized on cable will do it again in a new casing. Before the replacement goes in, confirm the suction conditions, the water level, the control settings and the cable size. The symptoms are all in signs your water pump needs maintenance, and if energy is part of the motivation, reducing pump energy consumption lists the measures worth bundling into the same job.
A simple scorecard
When the case is genuinely marginal, score the six factors below. Three or more points in the replace column is usually the answer.
| Factor | Points to repair | Points to replace |
|---|---|---|
| Repair cost versus new | Under 40 per cent | Over 50 per cent |
| Efficiency after repair | Within 5 per cent of original | More than 10 per cent down |
| Failures in the last 24 months | One | Three or more |
| Age | Under 8 years | Over 12 years |
| Spares availability | Local stock | Import, or obsolete |
| Duty still correct | Yes | No – demand has changed |
The row that most often decides it is the last one. A pump that is failing because the building grew, the borehole deepened or another irrigation block was added is not a maintenance problem at all, and repairing it simply buys a repeat of the same failure at a later date.
Not sure which way to go? We will assess the pump, measure its actual duty point and give you a costed comparison of repair, refurbishment and replacement – with rental cover in the meantime if you need it. See our services, contact us, or call +27 (15) 516-4991.