Repair or Replace? The Decision Line for a Food Plant Floor

Resin floor repair is a decision, not a purchase. Three questions settle it: how much of the area is affected, whether the failure sits at the surface or at the bond line, and what the next five years cost on each route.

Most plants arrive at the question after a partial repair has already failed, which is why the first job is usually to work out what the last repair missed rather than to price a new floor.

This article sets out the three questions, the limits of each repair route, and the arithmetic that makes a capital plan decision defensible.

The three questions that decide it

The first question is extent. A defect covering a few square metres is a patch repair; one covering a third of the room is a different conversation about removal and reinstatement.

The second is depth. A surface failure can often be ground back and recoated, while a bond-line failure means the system has separated from the substrate and will continue to do so under traffic.

The third is money over time. A lifecycle cost comparison that includes downtime, disruption and the next repair usually changes the answer that the first two questions suggested.

Sequencing a resin floor repair matters as much as the answers. Extent first, because a room-wide failure makes a patch irrelevant; depth second, because it decides which route is technically possible; cost last, because it decides which route is affordable.

Answer them in that order. Extent tells you whether a resin floor repair is even the right frame; depth tells you which route is technically possible; cost tells you which route the business should take.

Table 1: The three questions and what each one eliminates.

Question 1: how much of the area is affected

Measure it, do not estimate it. A tap test on a grid gives a percentage of hollow area in an hour, and that percentage is the number the decision turns on.

As a working rule, under about 10 percent is a repair, between 10 and 35 percent is an overlay system question, and above that the delamination extent is general enough that removal is usually cheaper than chasing it.

A delamination extent is rarely uniform. It spreads from a drain, a joint or a doorway, and the boundary between hollow and sound is the line that decides the scope. Marking that line on the plan is the most useful hour spent on the project.

Extent also has a shape. A failure that follows a slab joint or a drain line is a detail problem with a defined boundary; a failure scattered across a traffic lane is a system problem.

Extent and depth also interact. A shallow defect over a wide area is an overlay question; a deep defect over a small area is a patch. Reading them separately produces the wrong route in both directions.

Where a resin floor refurbishment has been done before, map both the new and the old defects on the same drawing. A repair that has failed twice in the same place is telling you the cause was never addressed.

Question 2: is the failure at the surface or at the bond line

Cut a sample. A core or a cut-back at the edge of the defect shows whether the resin has separated from the concrete or whether the wear course has simply worn away.

A when to recoat decision also depends on how much wear course is left. Where the aggregate is still visible across most of the area, a recoat restores years of life; where it is polished through, the recoat is cosmetic.

Surface failure is a when to recoat question: the film is thin, the aggregate is polished, or the topcoat has chalked. The substrate is sound and the repair is a recoat over prepared surface.

There is a third case that sits between the two: a thin film over a weak, dusty substrate that has never bonded properly. That is a when to recoat question in appearance and a removal in practice, and the core sample is what separates them.

Bond failure is not. Water, contamination or movement has separated the system from the slab, and any repair that does not reach the cause will lift again within a season.

The remaining service life estimate is the number that keeps a resin floor repair honest. It is a judgement, and it should be written down with the reasoning, so the next person can disagree with the reasoning rather than the conclusion.

The distinction also sets the remaining service life. A surface failure with 60 percent of the wear course left may have years of life after a recoat; a bond failure has none, whatever the surface looks like.

Figure 1: A tap test grid turns scattered tapping into a counted area, and the count decides the route.

Question 3: what does the next five years cost either way

A lifecycle cost comparison has four lines:

1. The work.
2. The downtime.
3. The disruption to production.
4. The expected repeat spend.

The last line is the one that is usually left out and the one that decides the answer.

Take a repair route that costs a third of a replacement and lasts two years. Over five years it is bought twice, with two shutdowns, and the total is usually close to the replacement it deferred.

A patch repair that is repeated three times in five years is not cheaper than a replacement, even when each visit looks small on its own. The pattern only becomes visible when the repairs are logged in one place.

Take a replacement that costs three times a patch. If it lasts fifteen years instead of three, the annual figure is lower, and the capital plan argument becomes a simple one.

Downtime is the line most often underestimated in a repair-and-replace comparison. Two shutdowns of three days each cost more than a single shutdown of five, and the schedule rarely allows the tidier option once the floor is open.

Put both on one page with the downtime in hours, and state the assumption behind each figure, because that is what the next review will challenge.

An option has value only if it is real. Waiting a year is worth something when the floor is stable and the production plan is uncertain. It is worth nothing when a defect is spreading under traffic and the scope grows with every month. Two costs sit outside the comparison and belong in it anyway: the risk of a food safety finding while the floor waits, and the disruption of an unplanned shutdown if it fails mid-run. A finance committee can compare two numbers per year; it cannot compare a rate per square metre with a shutdown schedule.

Repair routes and their limits

A patch repair works when the cause is in one bay and has already been dealt with. Sound substrate under the lifted edge is the test: if the edge will not lift cleanly, the decision has already moved to replacement.

A recoat works where the wear course is thin but bonded. It needs surface preparation and profile restoration before handover, and it fails quickly if the substrate moisture was the original problem.

A resin floor refurbishment that involves grinding the whole room and rebuilding the system is a removal in all but name, and it should be priced and phased as one.

The limits are worth stating plainly: no repair route fixes a damp slab, none fixes movement at a joint, and none lasts on a contaminated substrate. Those are causes, not defects.

Table 2: Repair routes, what they fix and where they stop.

Overlay versus full removal

An overlay system is the middle route: grind the existing floor back, prime, and build a new system on top, in epoxy or in cementitious urethane. It saves the removal and the disposal, and it keeps the floor level if the build-up is thin.

It only works if the existing bond is sound. An overlay on a floor with a general delamination extent transfers the old failure to the new system, and the second failure is more expensive because the new floor goes with it.

Removal is the honest answer when the substrate needs work: a damp slab, a failing screed, contamination that cannot be removed. In those cases the overlay is a delay, not a solution.

An overlay system also depends on the primer. A floor that has been previously coated may need a different primer from a bare slab, and the bond between old and new systems is the weak point in the assembly.

Level is the other constraint. An overlay raises the floor at doors, drains and equipment bases, and a 3 mm raise can be enough to break a door clearance or a drain connection.

Figure 2: An overlay is a new system on an old bond. It saves the removal, not the preparation.

Timing it against the capital plan

A capital plan runs on a cycle, and a floor that fails between cycles is funded from maintenance, which means a patch rather than a replacement. That is a legitimate choice if it is made deliberately.

A resin floor refurbishment funded from maintenance is a different decision from a replacement funded from capital, and the two should not be confused in the same approval. Naming the route makes the budget line honest.

Make it deliberately by recording the remaining service life in the asset register. A floor with two years left is a line in next year’s plan; a floor with none is an emergency in this year’s budget.

Where the replacement can wait, a resin floor refurbishment that buys three years is good value, provided it is scoped as a holding measure and not as a fix. The distinction matters when the same area fails again.

Where it cannot wait, the plan should be phasing a replacement around production rather than around the financial year. A floor replaced in the shutdown window costs less than one replaced in an emergency.

When the three questions disagree, area wins. A wide repair that returns in two years is a replacement deferred, and the second visit costs more than the first one saved.

A decision table by symptom

The table below is the version we use on site. It assumes the cause has been identified; if it has not, the first resin floor repair step is diagnosis rather than pricing.

Read the symptom, check the extent, then take the route. Where two routes are possible, choose the one that addresses the cause, even if it is the more expensive of the two.

Table 3: Symptom, extent and the route that usually follows.

A resin floor repair decision on a floor that is still serviceable is worth revisiting once a year, because lifecycle cost changes when energy, downtime or product mix changes. The route that made sense at 40 percent extent may not at 15.

Whichever route you take, record it with a date and a photograph in the same file as the condition log. The next decision is easier when the last one is written down. Keep the photographs with the dates as well: a decision recorded twice, with evidence and a review date, reads as management rather than as neglect when an auditor asks about the floor.

Frequently asked questions

How do I know whether to repair or replace a resin floor?

Under about 10 percent affected area with the cause removed, a patch repair holds; above about 35 percent the next five years cost more on the repair route than on the replacement.

Can a delaminated floor be patched?

Locally, if the cause has been removed and the boundary is defined. A patch over a cause that extends beyond it will lift again, usually within a season.

What is the difference between a recoat and an overlay?

A recoat restores the wear course on a sound system. An overlay builds a new system on a prepared but retained floor, and needs a sound bond to work.

How long does a resin floor repair last?

A patch on a known, removed cause can last as long as the surrounding floor. A patch on an unaddressed cause typically fails within one to two years.

Does an overlay raise the floor level?

Yes, by the thickness of the new system. At 3 mm that can be enough to affect door clearances, drain connections and equipment bases, so check the details before choosing it.

When should a floor go into the capital plan?

When the remaining service life falls below the lead time for a replacement, which is usually two years. Recording that date in the asset register is what turns a failure into a planned project.

Is a patch ever the right long-term answer?

In a low-care area with a local cause, yes. In a high-care area with a spreading failure, it buys time rather than solving the problem, and it should be scoped and recorded that way. A patch that is not recorded becomes an invisible repair in the next audit.

Why food plants work with Kaida Paint

Kaida Paint manufactures epoxy, polyurethane, cementitious urethane and MMA systems for food and beverage plants, and supplies them as finished systems or as OEM and private-label production for distributors. Each batch ships with a certificate of analysis and a batch record, and TDS and SDS documents are issued in the buyer’s language. If a resin floor repair is being considered, our technical team will survey the area, test the bond and the moisture, and price both the repair and the replacement so the comparison is made on the same evidence.

Send photographs, the plan and any previous repair records, and the reply is a scope with the test results that support the route recommended.