A quote lands on your desk and it is bigger than you hoped. The instinct is to ask whether the repair is worth it. That is the wrong question, and it is why this decision goes wrong so often. Here is the question that actually answers it, and the arithmetic behind it.
A $28,000 repair that buys you 4,000 more hours costs $7.00 an hour. The same repair, if the machine is back in the shop after 900 hours, costs $31.11 an hour. Identical invoice. Completely different decision.
So the only figure that settles this is the one nobody asks for: how many hours will this buy before the next major event? Ask your shop or your dealer directly. A good technician will give you a range, and a range is enough.
Everything below assumes you know roughly what the machine costs you to run per hour today. If you do not, start with the equipment cost per hour calculator and come back.
| Step | What to do |
|---|---|
| 1. Price the hours | Take the repair quote. Ask how many hours it should run before the next major event. Use the low end of the range you are given, not the high end. |
| 2. Amortize it | Repair cost divided by hours bought. That is what the repair adds to your hourly rate. |
| 3. Build the keep-and-repair rate | Remaining depreciation over those hours, plus cost of capital on what the machine is still worth, plus insurance and taxes, plus current operating cost, plus the amortized repair from step 2. |
| 4. Compare to the replacement rate | Total cost per hour of whatever would take its place, at the hours you would actually run it. Lower rate wins, then apply the overrides in the last section. |
Note what is not in this test: the original purchase price. What you paid is gone either way, and letting it drive the decision is the most expensive habit in fleet management.
Hydraulic pump and a final drive. The machine is worth about $60,000 today and would be worth about $35,000 after another 3,000 hours. It currently costs $46 an hour to operate, and it runs 1,200 hours a year. The replacement would be a $250,000 machine with a $75,000 residual over seven years, which the calculator puts at $79.45 an hour.
| Keep and repair, per hour | |
|---|---|
| Remaining depreciation ($60,000 to $35,000 across 3,000 hours) | $8.33 |
| Cost of capital on average value still tied up, at 8% | $3.17 |
| Insurance, taxes and storage at 3% | $1.19 |
| Operating cost today | $46.00 |
| The repair, spread across 3,000 hours | $9.33 |
| Total for the repaired machine | $68.02 |
| Total for the replacement machine | $79.45 |
Repairing wins, by about $11 an hour, or roughly $13,700 a year at 1,200 hours. That is often the counterintuitive part: a tired machine with a big invoice can still be the cheaper way to buy an hour of work, because its depreciation is largely behind it.
But watch how fast that flips. Everything above holds except the hours the repair buys:
| If the repair buys | Repair, per hour | Total, per hour | Versus replacing at $79.45 |
|---|---|---|---|
| 4,000 hours | $7.00 | $63.60 | Repair wins clearly |
| 3,000 hours | $9.33 | $68.02 | Repair wins |
| 2,000 hours | $14.00 | $76.85 | Repair still wins, narrowly |
| About 1,800 hours | $15.56 | $79.45 | Breakeven |
| 1,000 hours | $28.00 | $103.35 | Replacing wins clearly |
One input decides the whole thing. This is why "how many hours will it buy" is the question to press on, and why a repair history that shows how long previous fixes actually lasted is worth more than any rule of thumb.
A certified rebuild sits between the two: a large share of the machine's life reset for a fraction of new. On a cost-per-hour basis it frequently wins, because you are buying a lot of hours for a moderate sum. Run it through exactly the same four steps, with the rebuild cost in place of the repair and a much larger hours figure.
| Rebuilding tends to work when | Rebuilding tends to disappoint when |
|---|---|
| The base machine and structure are sound, and the wear is normal rather than the result of a chronic fault | You are rebuilding around a recurring problem that the rebuild does not address |
| You already stock parts, tooling and training for that model, so the fleet stays simple | The model is being phased out of parts support, which turns future downtime into a waiting game |
| The work is documented well enough that a buyer will pay for it later | The paperwork is thin, in which case the resale market treats it as an old machine regardless |
| The machine still fits the work you actually win | Your job mix has moved and the machine is the wrong size or spec now |
Track cumulative average cost per hour across a machine's life and it makes a shallow U. Early on, spreading the purchase price over more hours drags the average down faster than repairs push it up. Later, repair cost climbs faster than depreciation falls, and the average turns back up. The bottom of that curve is economic life.
Most fleets pass it without noticing, because a machine that still starts and still works does not feel like a problem. The cost shows up quietly, in a rate that has been drifting upward for two years. That is the case for tracking cost per hour continuously rather than pulling it together only when a quote forces the question.
Your own history is the best predictor you have. If the last three major repairs on a machine each bought fewer hours than the one before, that is your curve turning, and it is worth more than any industry average.
| Trigger | Why it wins the argument |
|---|---|
| Parts support is ending | A cheap rate means nothing if the machine sits for three weeks waiting on a component nobody stocks. Price the downtime, not the part. |
| Compliance or emissions | If a machine cannot legally work the jobs you are bidding, its cost per hour is irrelevant. This is increasingly common on public and urban work. |
| Safety | A defect that cannot be properly corrected ends the conversation. No hourly saving justifies it. |
| Downtime risk on critical work | If one machine failing idles a crew or misses a pour, the true cost of its unreliable hours is far above the rate on paper. |
| The resale cliff | Some models hold value to a point and then fall away quickly. Selling just before that is worth more than squeezing out another season. |
Most of the published guidance on this question comes from equipment dealers, and much of it is technically excellent. It is also written by organizations that sell both machines and service, so it is framed by what they can supply. Independent shops earn on the repair. Neither is being dishonest, and both usually know the iron better than you do.
What none of them can see is your side of it: how many hours you actually put on the machine, what it has cost you over the last twelve months, and what the alternative would cost at your utilization. That part only exists in your own records. Take their technical read on how long a repair will last, then do the arithmetic yourself.
When the repair does not buy enough hours to beat the cost per hour of the replacement. Divide the repair cost by the hours it should run before the next major event, add that to the machine's current running cost per hour, and compare against the replacement's total rate. If the repaired machine is more expensive per hour, replacing is cheaper. Parts availability, compliance and safety can override that result.
Frequently, on paper, because a rebuild buys a large number of hours for a fraction of new. It works best when the base machine is sound, when you already support that model, and when the work is documented well enough to protect resale. It disappoints when the underlying problem is chronic, or when parts support for the model is already fading.
The point where cumulative average cost per hour stops falling and starts rising. Before it, spreading the purchase price over more hours outpaces rising repairs. After it, repairs climb faster than depreciation falls. That turning point, rather than a fixed age or hour count, is when replacement usually makes sense.
No, and this is the most common error. What you paid is spent whichever way you decide. Only what the machine is worth now, what it will cost from here, and what the alternative costs should carry any weight.
Ask the shop, and use the low end of whatever range they give you. Then check it against your own history: pull what previous major repairs on that machine actually bought in hours. If each one has bought fewer hours than the last, your economic life curve has already turned.
Torgix keeps owning and operating cost per hour current for every machine you run, from your own labor, parts, fuel and usage records, so repair-or-replace becomes a lookup rather than a research project. Free trial on your real fleet, no credit card required.