Storage Tank Floor Scanning: MFL vs UT Comparison [2026]

MFL scans a 100-foot tank floor in 4-8 hours and catches pits 2-3mm deep across the plate; manual UT takes 40-80 hours and can miss a pit between grid points entirely.

By Anoop Rayavarapu, ASNT NDT Level III · · Inspection Applications

Storage Tank Floor Scanning: MFL vs UT Comparison 2026

Detecting bottom plate corrosion in atmospheric storage tanks is one of the most critical inspection challenges in the petroleum industry. Two primary technologies compete for this task: Magnetic Flux Leakage (MFL) and Ultrasonic Thickness (UT). This comprehensive guide compares both methods for tank floor integrity assessment.

The Challenge: Storage Tank Bottom Plate Corrosion

Why Tank Floors Fail

Tank bottom plates experience three corrosion attack mechanisms:

  • Underside Corrosion: From soil contact; creates general thinning and pitting (10-50+ mm deep pits common)
  • Product-Side Corrosion: From stored fluid chemistry (high chloride, sulfur, water content accelerate corrosion)
  • Lap Seam Corrosion: Accelerated attack at overlapping welds where moisture is trapped

Corrosion Rate in Tank Floors: 1-15 mils/year depending on soil chemistry, moisture level, and coating condition. A 5-year interval without inspection on high-corrosion floor could mean 50 mils (1.3 mm) loss — potentially catastrophic.

Failure Consequence: Tank bottom leakage = massive environmental liability. Remediation costs escalate rapidly with contamination severity and regulatory requirements.

Current Inspection Challenge

Tank floor inspection traditionally required tank removal from service, complete drainage and cleaning, and internal inspection — a process costing cost varies by specification+ and taking weeks. New technologies now enable in-service or reduced-downtime inspection.

Magnetic Flux Leakage (MFL) Technology

How MFL Works

MFL is an electromagnetic NDT technique specifically designed for ferromagnetic material (carbon steel) inspection:

  1. Magnetization: Strong magnetic field (much stronger than eddy current) magnetizes the steel to magnetic saturation
  2. Leakage Detection: When a defect (pit, crack, corrosion loss) is encountered, magnetic flux "leaks" from the material
  3. Signal Capture: Hall-effect sensors or SQUID sensors detect the leaked magnetic flux
  4. Imaging: Multiple sensors create detailed 2D/3D map showing corrosion geometry and severity

MFL Advantages for Tank Floors

  • Pit Detection Sensitivity: Detects pits 2-3 mm deep on floor; excellent for pitting corrosion identification
  • Speed: Typical tank floor (100 ft diameter): 4-8 hours scanning vs. 40+ hours manual UT
  • Coverage: Single run covers 100% of floor plate (or very close); doesn't miss isolated pits
  • Corrosion Mapping: Provides detailed loss profile showing "worst" areas and general corrosion rate
  • Depth Characterization: MFL can estimate pit depth more reliably than UT in some cases
  • No Coupling Required: No need for water, couplant, or probe contact (fully non-contact electromagnetic method)
  • Minimal Preparation: Paint and scale don't significantly affect results

MFL Limitations

  • Material Limitation: ONLY works on ferromagnetic materials (carbon/low-alloy steel). NOT suitable for stainless steel, aluminum, or austenitic stainless tanks.
  • Coated Tanks: Thick coatings can reduce signal strength; thick insulation may interfere
  • Equipment Cost: MFL systems expensive (a modest cost+ for new equipment). Usually accessed through specialized contractors.
  • Expertise Required: Requires specialized training to interpret MFL signals correctly; certification limited (not ASNT-standard like UT)
  • Data Interpretation: Distinguishing pitting from general corrosion requires experience; automated software helps but human expertise critical
  • Depth Uncertainty: MFL depth estimates are approximate; confirm significant indications with UT

MFL Typical Scan Results

For a 100-foot diameter tank:

  • Scan Duration: 4-8 hours continuous scanning
  • Output: Full-color floor map with corrosion zones highlighted
  • Pit Reporting: Every pit >2 mm detected and mapped; pit depth estimated
  • Cost: cost varies by specification per tank depending on size
  • Report Detail: High-quality visual maps suitable for engineering and repair planning

Ultrasonic Thickness (UT) Method

How UT Works for Tank Floors

Ultrasonic thickness measurement is the conventional method for tank floor inspection:

  1. Transducer Setup: Handheld probe with ultrasonic crystal transmitter/receiver
  2. Coupling: Requires water, oil, or specialized couplant for acoustic coupling to steel
  3. Sound Wave Travel: Ultrasonic pulse travels through steel and reflects off backwall
  4. Thickness Calculation: Time for echo = steel thickness; displayed on instrument
  5. Accuracy: ±0.01 inch (0.25 mm) typical

UT Advantages for Tank Floors

  • ASNT Standard: Widely accepted by inspectors and regulators; ASNT Level II/III credentials standard
  • Equipment Availability: Handheld UT instruments common (a significant capital item); widely available
  • Cost-Effective for Limited Inspection: If inspecting only specific areas, UT is cheaper per-point than MFL full scan
  • Measurement Point Data: Provides exact thickness at each measured point; excellent data for trending over time
  • Material Flexibility: Works on all ferromagnetic AND non-ferromagnetic steels, stainless, aluminum (material-specific velocity)
  • Calibration Control: Straightforward calibration on known steel blocks; easily verified on-site

UT Limitations for Tank Floor Inspection

  • Manual Process: Requires technician to measure point-by-point; typical 40-80 measurements for complete floor
  • Time-Consuming: 40-80 hours for comprehensive floor survey vs. 4-8 hours for MFL
  • Pit Detection Gap: Measures general thickness but may miss localized deep pits between measurement points
  • Labor Intensive: Requires 2-3 technicians working full shifts for days; significant cost in labor
  • Grid Pattern Required: Measurement points must follow logical pattern (typically 5-10 ft grid); labor-intensive layout
  • Coupling Requirement: Water/couplant messy; cleanup required; weather-dependent if outdoor
  • Surface Condition: Heavy scale, paint, rust can reduce signal quality; may require surface prep (grinding, descaling)

Direct Comparison

Factor MFL UT
Material Coverage Carbon/low-alloy steel only All steel types + non-ferrous
Tank Size 100 ft Diameter 4-8 hours scanning 40-80 hours with technicians
Detection Sensitivity (Pitting) 2-3 mm pit depth detection Misses pits between measurement points
General Corrosion Detection Excellent (full coverage map) Excellent (at measurement points)
Depth Measurement Accuracy MFL depth estimates approximate; UT prove-up needed UT exact thickness at point (±0.01")
Equipment Cost a modest cost+ (contractor rental preferred) a modest cost (commonly owned)
Cost per Tank 100 ft Diameter a modest cost a modest cost (labor intensive)
Downtime Required Tank out of service, cleaned and gas-free for floor scanning (same as UT) Full drainage recommended (safer access)
Certification/Training MFL is a method recognised in ASNT SNT-TC-1A; API 653 Annex G covers qualification of floor-scanning examiners and procedures ASNT Level II/III standard
Surface Preparation Minimal (scale/paint acceptable) Ideally clean; scale reduces signal
Trending Capability Yes (year-to-year comparison) Excellent (same points measured)
Report Quality Excellent visual maps; quick interpretation Detailed data tables; requires analysis

Best Practice Recommendation by Scenario

Use MFL When:

  • Tank floor corrosion history is unknown or suspect (safety concern)
  • Need comprehensive 100% coverage to identify all pit locations
  • Tank is large (>80 ft diameter) — cost/time trade-off favors MFL
  • High-consequence failure (product loss would be a modest cost environmental liability)
  • Budget allows (a significant capital item) and tank size justifies cost
  • Downtime is constrained — need fastest inspection method
  • Detailed corrosion mapping required for engineering/repair planning

Use UT When:

Hybrid Approach (Recommended):

  • Phase 1: Run full MFL scan (a significant capital item) for baseline 100% coverage and pit identification
  • Phase 2: Use UT to precisely measure thickness at worst pits identified by MFL (a significant capital item)
  • Phase 3: Plan targeted repair on worst pits; defer others based on remaining life assessment
  • Total Cost: a modest cost (expensive but comprehensive and lowest-risk approach)
  • Benefit: Combines MFL's full coverage with UT's precise thickness measurement; best for risk assessment

Frequently Asked Questions (FAQ)

Q1: How deep can MFL detect pits in tank floors?

A: MFL reliably detects pits 2-3 mm deep and deeper. Shallower pits (1-2 mm) may be detected but less reliably. Compare to UT which detects thickness change but may miss isolated pits between measurement points entirely.

Q2: Can UT detect pits or just general corrosion?

A: UT detects thickness at measurement points. If you measure at a pit location, you'll see the wall thinning. However, if a deep pit exists between measurement points (5-10 ft spacing), you'll miss it entirely. MFL detects all pits >2-3 mm regardless of spacing.

Q3: Is MFL safe for tanks with flammable liquids?

A: Yes. MFL is electromagnetic non-contact; creates no sparks or heat. However, check facility hazardous area (HAZLOC) classification and follow confined space entry procedures. Consult site safety before MFL equipment operation.

Q4: How accurate is MFL pit depth measurement?

A: MFL is a screening technique: its depth estimates are approximate and depend on the equipment, calibration and plate condition, which is why API 653 notes that UT is used to confirm MFL indications. Prove up significant indications with UT before making repair decisions.

Q5: Can UT be used to measure pit depth directly?

A: Yes. If you know the original floor thickness and measure current thickness at a pit with UT, pit depth = Original - Current. This provides exact pit depth. Example: Original 0.375", pit measures 0.220" → pit depth = 0.155" (155 mils).

Q6: What's the cost difference between MFL and UT?

A: MFL a modest cost per large tank. UT a modest cost depending on lab costs. For small tanks (<50 ft), UT comparable. For large tanks (>100 ft), MFL more cost-effective due to time savings.

Q7: Can I use both MFL and UT on the same tank?

A: Absolutely. This is the best practice: MFL for comprehensive pit detection, then UT for precise thickness at worst locations. Redundancy ensures you don't miss critical defects. Total cost varies with capability but highest confidence result.

Q8: How often should tank floors be inspected?

A: API 653 recommends: Good condition tanks every 10-15 years. High-corrosion-risk tanks every 5-10 years. Best practice: Use RBI (Risk-Based Inspection) to determine intervals based on corrosion rate history and consequence of failure.

Q9: What happens if MFL finds deep pits?

A: Follow API 653 repair procedures: (1) Determine pit depth with UT, (2) Assess acceptability against API 653 (minimum bottom thickness at the next inspection per Table 4.4 and the code's pitting and critical-zone rules), (3) Plan repair (local patch or full floor replacement), (4) Conduct repair with full NDE verification (MPI + UT post-repair).

Q10: Is MFL better for stainless steel tanks?

A: No. MFL only works on ferromagnetic materials (carbon/low-alloy steel). For austenitic stainless steel tanks, use UT only. No MFL equivalent exists for non-ferrous materials.

Conclusion

For comprehensive storage tank floor integrity assessment, MFL provides superior pit detection and coverage speed, while UT offers cost-effective targeted measurement and trending capability. Optimal strategy: Use MFL for baseline comprehensive inspection and pit identification, then supplement with UT measurements at critical locations. This hybrid approach minimizes risk while controlling costs.

Ready to plan your tank inspection? Review API 653 certification requirements to understand tank inspection standards, or explore NDT training to develop inspection expertise.

Putting this data on the asset model

Inspection data is far more useful bound to a location on the asset than filed as a report. The Atlantis Digital Twin maps every reading to its CML so corrosion rates trend automatically, and the vendor comparison covers how the major platforms differ on inspection-data depth.

Atlantis NDT Products & Services

Atlantis NDT pairs field expertise with software: NDT inspection management software — Atlantis ERP (certification tracking, work orders, method-specific reporting on every business app you need), a digital twin platform for asset integrity (3D corrosion mapping and inspection-data overlay), and NDT reporting software. Build your team with NDT training & certification (ASNT SNT-TC-1A) and ASNT certification pathways, or bring in ASNT Level III consulting for written practices, procedures and audits — plus independent inspection data review on API 510/570/653-governed assets. Capture as-built reality with 3D laser scanning services. Affordable, accessible, fully customizable — book a free consultation.

Where MFL fits inside an API 653 internal inspection

Under API 653, MFL floor scanning is one examination inside the out-of-service internal inspection. Its purpose is to give the authorized inspector enough bottom thickness and corrosion-rate data to judge whether the floor will stay above the code's minimum remaining thickness until the next internal inspection. API 653 names MFL as a commonly used bottom examination tool, often paired with ultrasonic thickness measurement to confirm and size what MFL finds. The decisions themselves belong to the owner's API 653 inspector and tank engineer.

This section adds the code framework to the method comparison above. References are to API 653, fifth edition (2014) with its addenda. API issued Addendum 4 in July 2025, which revised the text on tank-bottom thickness measurement and its references to API RP 575. Check the current edition and addendum before relying on any clause wording here; the following is a paraphrase, not code text.

In practice, MFL supports three API 653 decisions:

  • Is the bottom fit to return to service? Areas below the minimum thickness the owner needs for the planned interval are repaired, lined or replaced before the tank goes back in service.
  • How long until the next internal inspection? API 653 sets the subsequent internal interval from measured bottom corrosion rates and the minimum remaining thickness calculation, up to code maximums. Alternatively, an owner can use an RBI assessment.
  • What are the corrosion rates? Product-side and soil-side rates are the inputs to the minimum remaining thickness method, and MFL with UT prove-up is the usual way to get them across the whole floor.

The MFL threshold and the minimum remaining thickness method

API 653 gives a method for calculating the minimum remaining thickness (MRT) of the bottom at the end of the planned operating interval. It starts from the lowest remaining thickness after repairs, on both the product side and the soil side, and subtracts the expected corrosion on each side over the interval. The result must still meet the code's minimum bottom thickness at the next inspection. A probabilistic method based on statistical analysis of scan data is also allowed.

MFL changes one input in an important way. Because a scanner does not report every pit shallower than its detection threshold, the code addresses scanned areas specifically. For floor areas examined by MFL that lack effective cathodic protection, the remaining thickness used for the soil-side corrosion term is the lesser of the MFL threshold or the minimum thickness of unrepaired corrosion. API 653 defines the MFL threshold as the minimum remaining thickness to be detected in the areas examined, and expects the tank owner to set it in advance based on the inspection interval they want.

This has practical consequences on a tank job:

  • Set the threshold before scanning, not after. The owner and the inspector should agree on the threshold as part of the inspection plan. The scanning procedure and sensitivity settings then have to be able to detect and report indications at that level.
  • The threshold drives the interval. A less sensitive threshold means the calculation must assume more possible undetected loss, which shortens the interval the floor can support. A more sensitive threshold usually means slower scanning and more prove-up work.
  • Cathodic protection matters. API 653 lets the soil-side rate be taken as zero for areas with effective cathodic protection, and the product-side rate as zero for areas with a coating whose expected life covers the interval. MFL data still matters there to confirm the starting thickness.
  • Repairs reset the calculation. Corroded areas that are repaired are evaluated using the corrosion rate for the repaired area unless the cause was removed, with any patch plate thickness added in.

Minimum bottom thickness at the next inspection

API 653 sets the minimum bottom plate thickness that must remain at the next internal inspection, and the figure depends on the tank's design. For a bottom and foundation with no way to detect and contain a leak, the minimum is 0.10 in. Where the design detects and contains a bottom leak, or the bottom has a reinforced lining over 0.05 in. thick applied to API RP 652, the minimum is 0.05 in. If the projected thickness at the end of the interval falls below these values, or below the value an RBI assessment shows gives acceptable risk, the bottom must be lined, repaired or replaced, or the interval shortened.

The critical zone near the shell is treated separately. Unless a stress analysis is done, the critical zone has its own minimum based on the original bottom plate thickness and the lower shell course minimum, with an absolute floor of 0.10 in. at the next inspection, and averaging is not permitted. Isolated pitting in the critical zone may be judged against the general bottom minimum, but the storage tank engineer decides whether it affects structural integrity. This matters for MFL because the critical zone next to the shell is exactly where floor scanners struggle to reach.

How the scan result feeds the internal inspection interval

API 653 limits the first internal inspection of a new tank, or a tank with a new bottom, to 10 years. Credits for listed safeguards can extend that: a reinforced lining, cathodic protection to API RP 651, a release prevention barrier to API 650 Annex I, extra initial bottom thickness, or certain stainless steel bottoms. The initial interval is capped at 20 years without a release prevention barrier and 30 years with one. An RBI assessment can be used as an alternative.

Subsequent intervals are where MFL data does its main work. The code sets the next interval from the measured bottom corrosion rate and the minimum remaining thickness calculation, with the same 20-year and 30-year maximums, or alternatively through RBI. The code also expects the owner-operator to understand how effective the examination techniques were at detecting and measuring the expected damage. An MFL report that states the threshold, coverage and limitations is what makes that judgement possible. See how API 653 sets tank inspection intervals for external and UT shell intervals as well.

RBI for tanks is an engineering assessment the owner commissions under the code. Atlantis does not perform RBI; we provide the examination data it relies on.

Annex G: qualifying the procedure and the scanning operators

API 653 Annex G gives guidance on qualifying tank bottom examination procedures and the people who perform them. It is informative, so owners may apply it as written or adapt it, and the specific qualification requirements should be agreed between the owner-operator and the authorized inspection agency. API 653 points to Annex G both in its bottom thickness measurement clause and in its examination section for MFL.

In summary, Annex G describes:

  • A written tank bottom examination procedure covering the scanner, any hand scanner and the prove-up method, with essential variables (which cannot change without requalification) and non-essential variables. Plate thickness range, scan overlap, scanning speed and equipment settings are examples of procedure variables.
  • Operator training. The annex recommends at least 40 hours of training for each scanning operator, covering the scanner's NDE principles, equipment limits, calibration and operating variables, plus hands-on scanning under the supervision of a qualified examiner, followed by a written examination.
  • Performance qualification on test plates with designed underside and topside pits and a general-corrosion area, facilitated by the owner or by an independent third party, and recorded on a tank bottom examiner qualification record and a procedure qualification record.
  • Separate qualification for prove-up. Examiners who size indications, usually by UT, are qualified for that task, with expectations for depth-sizing accuracy that depend on whether the bottom is coated.
  • Requalification triggers, such as a gap since the operator last scanned a tank or used the procedure, a procedure change, a move to a new employer with different essential variables, or doubt about the examiner's ability.

Annex G qualification is in addition to the method certification the owner requires for UT prove-up. Owners commonly ask for ASNT SNT-TC-1A certification for the UT examiners. When choosing a floor-scanning contractor, ask for the procedure, the procedure qualification record and the operator's qualification record before mobilisation.

What MFL floor scanning cannot see, and how crews cover the gaps

MFL is a screening method with known blind spots. The main gaps are the zone next to the shell and the critical zone, lap welds and plate edges, areas under internal obstructions, and floors with thick coatings or heavy distortion. It also has limited ability to separate topside from soil-side indications without added sensors or visual confirmation. API 653's position that MFL is commonly used with UT reflects this: MFL finds the indications and UT sizes them.

GapWhy it happensTypical coverage approach
Annular ring and critical zone next to the shellScanner body and wheelbase cannot reach the shell-to-bottom weld; edge settlement curves the plateHand MFL scanner, UT scanning or phased array mapping along the annular plate; visual of the shell-to-bottom weld
Lap welds and plate overlapsGeometry change disturbs the magnetic field; scanner cannot sit flat across the lapScan to the weld from both sides; targeted UT or visual; MT or VT for weld cracking where needed
Under columns, sumps, heating coils, internalsPhysical obstructionHand scanner, manual UT, documented as limited coverage
Thick or degraded coatingsLift-off reduces sensitivity; disbonded coating changes signalProcedure qualified for the coating thickness; coating inspection; prove-up through or after coating removal
Topside versus soil-sideThe basic MFL signal responds to metal loss on either surfaceVisual examination of the product side; surface-discrimination sensors where fitted; UT prove-up
Thick plateSaturation becomes harder as thickness increasesScanner and procedure qualified for the plate thickness range

The report should state uncovered areas, by location and area, and how each was otherwise examined. Without that, an inspector cannot judge examination effectiveness, which the interval calculation depends on.

Pipeline MFL versus tank floor MFL

Many searches for "MFL corrosion inspection pipeline" or "MFL tools" end up on tank-floor pages. The physics is the same: magnetise the steel close to saturation and detect flux leaking from metal loss. The application, regulation and equipment are different.

  • Pipeline in-line inspection (ILI). MFL tools ("smart pigs") travel inside the line, driven by the product, and record metal loss around the full circumference over long distances. This is governed by the pipeline operator's integrity management program under PHMSA rules (49 CFR 192 for gas, 195 for hazardous liquids), with industry standards such as API 1163 for ILI system qualification. Atlantis does not provide in-line inspection tools or ILI services; this is background only.
  • Tank floor MFL. Scanners are pushed or driven across the floor of an emptied, cleaned tank during an API 653 out-of-service internal inspection, followed by UT prove-up. It is governed by the owner's API 653 program and, for oil storage, the EPA SPCC requirement to test or inspect containers for integrity on a regular schedule.
  • Above-ground piping. Short pipe runs can be screened with external MFL or with other methods such as guided wave testing. That falls under the owner's API 570 program, not pipeline ILI rules. See guided wave testing.

For the pipeline side, see MFL pipeline inspection: when to use it versus UT and the API 1163 overview.

Regulatory and Canada notes

In the United States, the EPA's SPCC rule requires owners to test or inspect each aboveground oil container for integrity on a regular schedule and whenever material repairs are made, following industry standards. For field-erected tanks, API 653 is the standard most often used, and floor scanning is how its internal inspection is usually carried out. Some states add their own aboveground storage tank rules, so confirm with the state environmental agency. Facilities under OSHA PSM treat API 653 as recognised practice for covered tanks.

In Canada, aboveground storage tanks are regulated provincially and federally depending on the product and the facility. Federally regulated petroleum tank systems have their own regulations, and provinces add fire code and environmental requirements. API 653 is widely used, but the applicable regulator decides inspection obligations. Confirm with the relevant authority.

How Atlantis supports API 653 floor examinations

Atlantis performs tank floor MFL scanning with UT prove-up, annular-plate and critical-zone examination, corrosion mapping, shell UT thickness surveys and weld examination. We work to written procedures reviewed by our ASNT Level III and use ASNT-certified technicians. We deliver the bottom data, coverage map and stated MFL threshold to the owner's API 653 inspector, who remains inspector of record and sets the interval. We do not perform RBI and do not provide pipeline in-line inspection. See magnetic flux leakage testing, aboveground storage tank inspection and NDE support for API 653 programs. Send tank sizes and outage dates and we will reply with a quote within 24 hours: request an MFL floor scan quote.

More questions about MFL and API 653

Does API 653 require MFL floor scanning?

No specific method is mandated. API 653 requires enough bottom data to evaluate thickness, corrosion rate and integrity and to set the next internal interval. It describes MFL as commonly used, often with UT. The owner and inspector choose the methods.

What is the MFL threshold in API 653?

It is the minimum remaining thickness the scan is set up to detect in the areas examined. The tank owner sets it in advance based on the interval they want. For scanned areas without effective cathodic protection, the minimum remaining thickness calculation uses the lesser of this threshold and the minimum thickness of unrepaired corrosion.

Is Annex G mandatory?

Annex G is informative. Owners may apply it as written or adapt it, and the qualification requirements are agreed between the owner and the authorized inspection agency. Many owners require Annex G-style procedure and operator qualification in their contracts.

Can MFL scan a tank floor while the tank is in service?

Not in the way an API 653 internal inspection requires. Floor scanners work on an emptied, cleaned, gas-free floor. In-service robotic methods exist, but they are a different technology, and whether their data satisfies an internal inspection is a decision for the owner and inspector.

What is the difference between MFL pipeline inspection and tank floor MFL?

Pipeline MFL uses in-line tools travelling inside the pipe under PHMSA integrity management rules. Tank floor MFL uses scanners on the floor of an emptied tank under API 653. Atlantis offers tank floor MFL but not pipeline in-line inspection.

Who sets the next internal inspection date after a floor scan?

The owner's API 653 authorized inspector and storage tank engineer, using the scan data, the corrosion rates and the code's maximum intervals, or an RBI assessment. The NDE contractor supplies the data.

Planning a tank outage? Ask us to scope the bottom examination or request Annex G qualification records for our procedure.