Tank floors fail more often than any other component because they sit in the wettest, dirtiest part of the tank. Water settles on top of them, sludge accumulates against them, and on the underside they sit against a foundation that traps moisture. When an internal API 653 inspection finds floor problems, the question is rarely whether to repair, but how. Patch plates, partial re-bottom, or full re-bottom each carry very different costs, downtime, and engineering requirements.
What follows breaks down the three primary floor repair options, when each one is appropriate, what API 653 requires, and how the inspection findings that drive the decision are typically captured.
Three repair options at a glance
Floor repair under API 653 falls into three broad categories. Each has its own typical scope, cost range, and decision criteria.
| Option | Typical Scope | When to Use | Relative Cost |
|---|---|---|---|
| Patch plates | Localized welded plates over isolated damaged areas | Pitting or thinning confined to small, well-defined areas | Lowest |
| Partial re-bottom | Replacement of a defined section of floor plate | Damage exceeds patch limits but is concentrated in one region | Medium |
| Full re-bottom | New floor over the existing one or full floor replacement | Widespread damage, underside corrosion, or end of original floor service life | Highest |
How floor problems get discovered
Floor damage almost always shows up during an internal API 653 inspection because the tank has to be empty for any meaningful floor evaluation. The standard inspection toolkit covers four primary methods:
- Visual inspection. The starting point, but limited to topside conditions and surface defects.
- Magnetic flux leakage (MFL) scanning. Standard method for evaluating floor plate corrosion across the entire surface, capable of detecting both topside and underside indications.
- Ultrasonic thickness testing (UT). Used to characterize specific MFL indications and measure remaining wall thickness at points of concern.
- Vacuum box testing. Verifies weld integrity on suspect areas, particularly shell-to-bottom welds.
Findings typically fall into recognizable categories: isolated pitting, generalized thinning, weld defects, settlement-induced cracking, or underside corrosion that is invisible until MFL flags it. The category drives the repair decision more than the absolute amount of corrosion does.
Patch plate repairs
Patch plate repairs are the most common floor repair under API 653. They work well when damage is localized, well-defined, and doesn't approach the limits API 653 sets on patch geometry.
Lap-welded vs butt-welded patches
API 653 Section 9 recognizes two patch configurations:
- Lap-welded patches sit on top of the existing floor plate and are fillet-welded around the perimeter. Easier to install, but the lap creates a crevice that can trap product and accelerate corrosion. Best for tanks in mild service or for short-term repairs.
- Butt-welded patches replace a cut-out section of floor plate with new plate welded flush. More work to install, often requiring a backing strip and weld qualification, but eliminates the crevice and produces a longer-lasting repair.
Where patches are appropriate (and where they aren't)
Patches make sense when damage is confined and the rest of the floor has good remaining life. They become less defensible when:
- The number of patches exceeds what API 653 permits per square foot of floor area
- Damage is widespread enough that patching becomes more expensive than partial re-bottom
- Underside corrosion is suspected, since patches address topside damage but do nothing for what's happening beneath the floor
- The floor is approaching end of original service life, in which case repair cost is better invested in re-bottom
The minimum thickness rules in API 653 also matter. If remaining floor thickness in the patched area or surrounding plate is below the calculated minimum, the inspector will require additional repair beyond the patch.
Partial re-bottom
Partial re-bottom involves cutting out and replacing a defined section of floor plate while leaving the rest of the floor in service. Useful when damage is too extensive for patches but concentrated in a region rather than spread across the floor.
When partial re-bottom makes sense
The classic case is a tank where one section of floor (often near the shell where water-bottom corrosion concentrates) has thinned beyond patch limits, while the rest of the floor still has years of remaining life. Replacing the affected sector preserves the investment in the rest of the floor.
Partial re-bottom requires careful planning around the annular plate, the weld between the new and existing plate, and any shell-to-bottom weld implications. The engineering review that API 653 requires for floor repairs becomes more involved as the scope grows.
The economic catch
Partial re-bottom often loses to full re-bottom on pure economics. Once the tank is empty, cleaned, and accessible, the marginal cost of replacing the entire floor versus a section is smaller than people assume. The labor of mobilization, gas-freeing, scaffolding, and demobilization is largely fixed regardless of repair scope. Many partial re-bottom projects end up being upgraded to full re-bottom mid-project once the contractor sees the actual condition. This is worth modeling honestly during planning rather than discovering on site.
Full re-bottom
Full re-bottom is the most expensive and most disruptive floor repair, but in many cases also the most economical when total life-cycle cost is calculated. Two main approaches exist.
New floor over existing
The simpler approach: a new steel floor is installed over the existing one with a sand or concrete fill between. Faster and cheaper than full removal because the existing floor stays in place. The downside is loss of secondary containment if the existing floor still functions as a barrier.
API 650 Appendix I addresses the design considerations when installing a new floor over an existing one, including release prevention barriers and leak detection systems.
Full floor removal and replacement
The more involved approach: the existing floor is cut out and removed, the foundation is inspected and prepared, and a new floor is installed on the prepared subgrade. Significantly more labor and waste handling, but produces the cleanest result and provides an opportunity to install a release prevention barrier with leak detection from the ground up.
Release prevention barriers and leak detection
Modern re-bottom projects almost always include a release prevention barrier (RPB) beneath the new floor. The RPB is a synthetic liner or membrane that prevents floor leaks from reaching the soil and groundwater. Leak detection systems route any liquid that does penetrate the floor to monitoring points, providing early warning long before contamination becomes a regulatory problem.
Federal and state environmental regulations increasingly require RPBs and leak detection on re-bottom projects, particularly for tanks holding regulated substances. Even when not strictly required, the cost of installing them during a re-bottom is small compared to the future cost of a release.
A re-bottom is the right time to add an RPB and leak detection. The tank is empty, the foundation is exposed, and the labor for installation is at its lowest. Skipping these systems to save short-term cost is a common regret on tanks that later see a release.
The role of NDT in repair planning
The repair decision rests on inspection data. MFL scanning establishes the floor's condition map: where corrosion is, how deep it is, and whether it's topside or underside. UT confirms specific indications and measures remaining thickness at points where MFL flags concern. Visual inspection captures everything else, particularly weld defects and settlement-induced damage.
An experienced inspection provider does more than report findings. They translate findings into repair categories, flag the boundary cases where the patch-vs-rebottom decision is genuinely close, and provide the engineering basis the repair contractor needs to scope work properly. Tanks with thin or unreliable inspection records often get conservative repair recommendations because the inspector cannot confidently bound the damage.
Documentation that drives better repair decisions
Repair planning works best when the inspection produces:
- Floor maps showing MFL indications with severity classifications
- UT thickness readings at indication locations and across surrounding areas
- Photo documentation of visual findings, including weld defects and topside conditions
- Settlement survey data if differential settlement is suspected
- Calculated minimum required thickness and corrosion rate analysis
Tanks inspected with thorough documentation get more targeted repair scopes; tanks with sparse documentation get larger, more conservative repair scopes by default.
Coatings and linings as a complement to repair
Floor repair is often paired with installation of an internal lining or recoating of the existing one. A new coating or lining over a freshly repaired floor protects the steel from the conditions that caused the original damage and can substantially extend the floor's service life until the next repair cycle. The economics work because the labor of cleaning, prep, and access is already invested in the repair project.
Lining selection depends on tank service: water service requires NSF/ANSI 61 compliant systems, petroleum service uses different chemistry, and chemical service demands case-by-case selection. The right lining system can change the next repair interval from 10 years to 20 or more.
Working with an inspection and repair provider
The cleanest floor repair projects are the ones where inspection, repair planning, and any related coating or lining work are coordinated by a provider who handles the full scope. Multiple contractors, multiple mobilizations, and multiple sets of documentation introduce coordination friction that adds days and dollars to every project.
NDT Tanknicians performs API 653 inspections, repair planning, and coatings and lining installation as coordinated services. To discuss an upcoming inspection, plan for repair work after a recent inspection, or evaluate options on a tank approaching end of service life, contact us.
Recent Comments