top of page

High Resolution Mapping of Tank Floor Underside Corrosion using MFL and PAUT Array DLA Technology

Introduction

Aboveground storage tank floors can be inspected for bottom side corrosion using several non-destructive testing techniques (NDT) including acoustic emission testing (AE) [1}, magnetic flux leakage testing (MFL) [2-4], ultrasonic testing (UT/PAUT) [5] and visual inspection (VT). Tank floor inspection is most performed using magnetic flux leakage (MFL) testing due to its screening capability for tank bottom side metal loss. Most MFL inspections require follow-up spot checking with ultrasonic thickness testing (UTT) to quantify accurately the metal loss. While conventional manual UTT is effective, it is a low-resolution sample of the MFL screened area and it may be possible that thinner areas are missed. The article presents encoded phased array using dual element probe technology as a fast and effective high-resolution method for MFL inspection spot checks.


Figure 1: High resolution PAUT scan of tank bottom underside corrosion.


Aboveground Storage Tank Soil Side Corrosion

Corrosion on the tank floor bottom side, or soil side, occurs when the tank floor components, underlying materials and environment facilitate the creation of an electrochemical reaction to occur and initiate underside metal loss [6-7]. The most observed types of tank floor corrosion are generalized and localized (pitting) corrosion. General corrosion is characterized by thousands of microscopic corrosion cells initiating on the soil side tank floor surface resulting in in relatively uniform metal loss. Localized corrosion is characterized by individual corrosion cells that are larger, and metal loss from localized corrosion may be concentrated within relatively small areas with substantial areas of the surface unaffected by corrosion.

Tank Floor Magnetic Flux Leakage Testing

The soil side of steel aboveground storage tanks can be screened for metal loss rapidly using MFL technology. The ferromagnetic tank floor is magnetized using a large permanent magnet that introduces a strong directional magnetic field into the tank floor. The MFL scanner's powerful magnetic bridge introduces a magnetic flux into the material near to saturation level. Any localized pitting or generalized corrosion will result in a magnetic flux leakage at the tank floor top side. An array of Hall and/or inductive sensors are placed between the poles of the magnetic bridge to detect these leakage fields. The strength of the leakage field is a function of volume loss and is not a reliable indication of remaining wall thickness. Even though the amplitude of the signal generated by the sensors should be proportional to the degree of bottom side corrosion the MFL amplitude alone should not be used for accept/reject criteria purposes. It is generally recommended to use MFL tank floor scanning technology as a detection tool. Accurate assessment of metal loss of MFL screened areas should be performed with high resolution ultrasonic thickness testing.



Figure 2: MFL detected tank floor underside corrosion markup. MFL screened area is marked up for follow-up UT with PAUT DLA probe.


High-Resolution Tank Floor Ultrasonic Scanning

High resolution scanning of tank floor soil side corrosion may be obtained by using a PAUT dual element array (DLA) probe with a standard magnetic wheel encoder. Olympus’ 7.50L32 REX1-IHC dual element linear phased array probe can be configured with a magnetic wheel encoder to provide high resolution – near surface resolution. The 1.0 mm pitch allows for small to large virtual probe configurations. The 7.5 MHz PAUT DLA’s 32 elements permit up to 1.25” wide scans compared to 0.25 or 0.50” wide scans with conventional ultrasonic contact or dual element probes. An encoded scan across the area above produced the high-resolution data presented in the Olympus X3 data shown above.

In the PAUT video, the top left data is the top view C-scan. The vertical green axis on the left is the 1.25” wide index axis. The actual virtual probe width is dependent on the user selected element quantity. For example, if the user selects and element quantity of 8, the virtual aperture will be 8 mm and there will be a total of 32 – 8 + 1 = 25 virtual probe apertures. The blue horizontal axis is the scan axis and was approximately 10” in this case.

In the top right is the B-scan across the green index axis, or transducer width. The purple depth axis is on the left side. Note that the thinning shown in the C-scan will not be displayed in this view unless the C-scan data cursors are inside the corrosion area.

The bottom right is the current VPA A-scan. No main bang, dead zone, or initial pulse is present since a DLA probe is used. A strong backwall vertical edge is observed crossing the red detection gate.

Finally, the bottom left data is the traditional B-scan generated along the scan axis. Pitting corrosion is observed from about 2.75 through 6” with severe pitting observed at 3.5”. the tank floor had a nominal thickness of approximately 0.375”. The minimum wall in this area was roughly 0.20” for almost 50% wall loss. Considering that the deep pitting area is only 0.5” x 0.4” in area, it is possible, or even likely, that a smaller 0.50” or 0.25” diameter probes would miss this area during a manual UTT spot check.

The scan axis resolution is defined by the encoder resolution selected. In this case it was set to 0.0