Repair or replace industrial equipment: a repair vs replace decision method for corroded and worn assets

Repair or replace equipment? Repair when the damage is characterized, the degradation mechanism is arrested or accounted for, an accepted repair method restores the required integrity, and the repaired asset costs less per year of service than a new one. Replace when any of those four conditions fails — or when no repair can be documented as acceptable.

The question usually reaches the shutdown planner late: inspection has found wall loss, erosion or cracking, the outage window is fixed, and someone must choose between a repair crew and a purchase order. The method below asks four questions — can it be repaired, can it be replaced in time, which option costs less over the service still expected, what risk does each leave — so one file serves operations, the inspector and management. A replacement that cannot arrive in time often turns the real decision into "repair now, replace at the next planned outage".

Technical criteria: can the equipment be repaired?

Integrity and extent of damage

Start with what the inspection measured: remaining wall thickness against the minimum required thickness, the extent and shape of the damage (general thinning, local metal loss, pitting, cracks), and the condition of the surrounding base metal. Localized damage on sound metal is the typical repair case. Generalized thinning, cracking of unknown origin or a distorted shell push toward replacement, because a repair would have to rebuild most of the part.

Is the degradation mechanism still active?

This criterion is the most often skipped: restoring thickness without addressing the cause only resets the clock. Identify the mechanism — general corrosion, pitting, erosion-corrosion, cavitation, abrasion, galvanic or microbiological attack, corrosion under insulation — then decide whether it can be arrested (barrier coating, material change, isolation, new operating conditions) or only slowed. If it will continue, the repair must carry a future corrosion or wear allowance, and its life is set by the degradation rate.

Fitness-for-service: API 579-1 levels 1 to 3

For pressure equipment, piping and tanks, the recognized tool is a fitness-for-service assessment to API 579-1/ASME FFS-1. It compares the damaged component, as measured, against acceptance criteria and returns an outcome: continue operating, continue at reduced conditions, monitor at a shorter interval, repair, or replace. It also estimates remaining life.

Failing Level 1 does not condemn a component. A Level 2 or 3 assessment pays when replacement is expensive or slow.

The codes that frame the answer

The in-service codes govern equipment after construction: API 510 for pressure vessels, API 570 for process piping, API 653 for atmospheric storage tanks. Each defines inspection, remaining life from measured corrosion rates and the conditions for repair, and refers to API 579-1 when damage exceeds its simple rules. ASME PCC-2 provides repair methods — welded, mechanical and nonmetallic, including engineered composites — but, as our guide to ASME PCC-2 explains, it supplies the engineering basis, not the acceptance. See also how API 653 treats composite repairs and the API 570 piping guide. In Canada, pressure equipment is regulated provincially: the owner's program, the authorized inspector and the jurisdiction confirm what the codes allow.

Non-pressure equipment — pump casings, chutes — has no equivalent code, but the logic holds: define the minimum section needed, measure what is left, state the acceptance basis.

Operational criteria: lead time, outage window, obsolescence, criticality

When the manufacturer has discontinued the model, "replace" may mean redesigning the installation. The intermediate route is reverse engineering a replacement part without drawings, often while a cold repair keeps the original in service. On criticality, a risk-based ranking (probability of failure against consequence, as in API 580 for inspection planning) shows which assets justify a spare, an FFS Level 2 or an early replacement.

Economic criteria: repair vs replace on equivalent annual cost

Comparing a repair invoice with a purchase price answers the wrong question: the repaired asset and the new one deliver neither the same years of service nor the same running costs. Engineering economics compares the equivalent annual cost (EAC) of each option over its own expected life:

EAC = NPV × r / (1 − (1 + r)^−N)

where NPV is the present value of all the option's costs (installation, outage, inspection, maintenance), r the discount rate set by finance, and N the service life delivered — for the repair, the remaining-life estimate; for the new unit, its economic life. The method and its pitfalls are detailed in our total cost of ownership guide.

Illustrative calculation (hypothetical figures)

The figures below are hypothetical round numbers in arbitrary cost units, chosen only to show the mechanics — not prices, service lives or project results. Assumptions: discount rate 5 %. Replacement: 100 units installed, economic life 20 years, running cost 2 units per year. Repair: 30 units, running cost 4 units per year (closer inspection), service life after repair varied below.

The repair's credible life — what the fitness-for-service assessment and the arrest of the mechanism give you — decides the ranking more than its price. And outage cost belongs to whichever option causes it: charged to the wrong side, it flips the result.

Risks: temporary vs permanent repair, documentation, owner and insurer

Temporary or permanent. The in-service codes distinguish temporary from permanent repairs and expect temporary ones to be recorded, monitored and replaced or re-evaluated, under conditions set by the code and the inspector. An engineered composite repair is designed for a stated lifetime, not "forever". State which kind you are buying, and until when.

Documentation. A repair is defensible only with its file: inspection data, damage mechanism, FFS result where one was done, repair design basis, material data, applicator qualification, preparation and cure records, post-repair inspection. An undocumented repair becomes a finding at the next inspection — on pressure equipment, possibly a regulatory one.

Owner, insurer and jurisdiction. The owner decides; the authorized inspector accepts; for regulated pressure equipment the provincial authority may have to accept the repair too. Insurers of pressure equipment often have their own inspectors — involve them before the repair.

Hot work. Welding on an in-service or hard-to-clean asset carries fire and gas-freeing risk. Where the code allows it, a cold repair removes that risk; our cold bonding vs welding comparison sets out where each method fits.

The repair vs replace decision tree, step by step

Typical cases

Centrifugal pump. Casing and impeller erosion, cavitation damage and corroded wear-ring seats are classic repair cases: the metal is rebuilt with filled epoxy and the flow path coated. A cracked casing, or wall loss that compromises the pressure rating, points to replacement — or a reverse-engineered casing if the model is obsolete.

Heat exchanger. Corroded or eroded tubesheet faces and water boxes are usually repaired and coated in place. A bundle with widespread tube failures is a retubing or replacement decision, driven by lead time and by whether the mechanism has been addressed.

Storage tank. Localized bottom or shell thinning can be repaired under API 653 once characterized; generalized bottom thinning calls for steel work. A lining arrests the mechanism; it does not replace the structural repair.

Piping. A local external defect can take a clamp or an engineered composite wrap; a spool with generalized internal thinning, or corrosion under insulation along its length, is normally replaced.

Chute or hopper. Abrasion wears the liner first. While the structural shell is sound, replace or upgrade the liner — hardfacing, abrasion-resistant plate, ceramic or filled epoxy — not the chute.

FAQ

When is it cheaper to repair than to replace industrial equipment? When the repair's equivalent annual cost over its credible service life is lower than the replacement's over its own life — typically when the damage is localized, the mechanism is arrested and the replacement carries a long lead time or its own outage.

Is a fitness-for-service assessment mandatory before repairing? Not always. For vessels, piping and tanks, the in-service code refers to API 579-1 when damage goes beyond its simple acceptance rules. For non-pressure equipment it is not required, but the same logic — minimum section, measured damage, remaining life — applies.

Can a repair be considered permanent? That depends on the governing code, the repair design and acceptance by the owner and inspector. Engineered repairs are designed for a stated lifetime; the file should state the status and the next inspection date.

What if the original part is obsolete? Repair to keep the equipment running, then reverse engineer the part from the worn original and its mating components, correcting the material if it caused the failure.

Who signs off on the decision? The owner makes it. For regulated pressure equipment, the authorized inspector and, where required, the provincial authority accept the repair.

In practice

Induscoat helps plants in Quebec and Ontario document the repair side of this decision — damage assessment, repair method, surface preparation, application and inspection records — through its maintenance and rehabilitation service and its cold bonding repair work on pumps, exchangers, tanks, piping and chutes. Send the inspection findings, the outage window and the replacement lead time you have been quoted to request a quote. Technical data sheets for the Induscoat range are published on the brand site, induscoat.com.