Methodologies for Magnetic Flux Leakage (MFL) and Above Ground Storage Tank (AST) Inspection
The history of MFE/MFL technology and tank floor inspection
Initially, tanks were visually inspected and only obvious defects were noted. At some point, the responsible parties brought pick hammers into the tank and hit the floor to check the solidity of the floor. Some hammer swingers noted that they could tell the condition of the floor based on the sound coming from the floor when struck by the hammer. This was some sort of pre-acoustic technique which led to the cutting of coupons and the first evidence of a statistical analysis approach to tank floor inspection. Eventually, ultrasonic NDT was applied to this application. Ultrasonic coupons were born and yet tanks continued to leak at an alarming rate. Meanwhile, pipeline inspection companies were utilizing MFE/ MFL technology to perform pipeline integrity inspections. These companies were asked to help resolve the tank floor inspection problem the way they had helped resolve pipeline inspection applications. The tank floor scanner was invented based on the same MFE/ MFL principle technology applied to pipeline scanners. The main point here is that nothing is 100% as regards to this application, but each step is an improvement in moving closer and closer to the ultimate solution.
Manual tank floor scanning with real time display
Manual tank floor scanners with a real time display are scanners that are pushed across the tank floor and do not include a drive motor. The real time display provides the technician with a visual display of a signal response from flux leakage. If the response is only used as a location device, and is not used to implicate pit depth, then the speed the scanner is pushed has a minimal impact on the inspection outcome. With the manual scanner, an alarm is included which will flash if the real time display indicates a signal. The alarm must be reset to stop the flashing alarm. Some companies strongly support manual scanning because this process is the least dependent on tying signal amplitude to pit depth. The argument is that signal strength is more relative to the overall volume loss of the pit than it is to the pit depth. If the unit uses a coil sensor, it would need a fixed speed motor to provide a uniform inspection speed if the signal response is going to be evaluated in terms of some linear response to pit depth.
Stop-On-Defect tank floor scanning without real time display
This type of MFE/ MFL scanner would use a hall sensor and would have a drive motor but no real time display. This technique has been dubbed a “blind single level threshold”. This unit would be calibrated trying to match pit depth to signal amplitude. The unit would use a calibration plate with a known depth of defect. The unit would have a gain switch that could be adjusted to amplify or de-amplify the signal, such that the signal from the calibration plate would trigger the threshold alarm, and thereby shut down the drive motor. The unit would have lights across the width of the scanner to represent the approximate location of the pit in the X or side to side direction. Any light that illuminated would be marked for prove up by ultrasonic inspection. One issue with this approach, is that there are some who believe that since the signal strength from flux leakage is affected by the total volume loss, the stop on defect without real time display could pass over a pit that has the required depth but not the required total volume loss to equal or exceed the threshold volume loss. The defect could, therefore, be missed. If this were to occur, the technician would not know that the anomaly was missed, because there is no real time display to show the pit response that failed to reach the required amplitude threshold. Another issue with this approach, is that general corrosion on part of the tank floor could cause the technician to desensitize the unit as compared to areas where there was isolated pitting. If the volumetric assumption is correct, then it would explain why pits and even holes have been missed with this approach. On the other hand, there are supporters who believe that the lack of a real time display allows the technician to find signals higher than the general noise level, but ignore the signals within the noise level range.
Stop-On-Defect tank floor scanning with real time display
The addition of a real time display to the stop-on-defect approach allows the technician to see a signal even if the signal does not reach the preset threshold. This advantage allows the operator to mark and check any signal displaying on the real time display. Then, the technician can decide if the signal that did not reach the threshold has sufficient pit depth to be included in the anomalies that are of interest. The technician can then adjust the threshold to include this anomaly or the technician can adjust the sensitivity of the unit to cause this anomaly to trigger the threshold. Supporters of stop-on-defect units agree that the main advantage of this approach is that it can be used to discern signals that are just slightly higher than the general noise level. This could be difficult to do with a manual unit where the technician relied on a real time display and had to find all the signals that were just one or two bars above the general noise level.
Mapping tank floor scanning using a Multiple color map presentation
Mapping systems are motor driven and have the added sophistication of computers, software, and some form of encoder that measures wheel travel and tells the software program where the sensor is on the floor at all times. The signal responses are recorded and saved to a file. At some point, the files are transferred to the tank floor drawing. At the end of the inspection, the technician can provide an automated report that provides color images of the tank floor plates with pitting that is relative in volume loss or pit depth to the color on the tank floor map. The map of the tank floor makes it easy for the tank owner to know immediately what type of repair strategy will be needed. Some argue that the map is more representative of the volume loss than of the pit depth. It is the pit depth that determines repair strategy and remaining floor life. However, supporters of the mapping approach argue that the map provides valuable repair information and that if the owner is reasonable as regards to the known limitations of the technology, there is considerable value to this approach.
So, which method is correct?
I believe the application of manual, stop-on-defect, and mapping is absolutely tied to the condition of the tank floor. Tanks that have low density of general corrosion or low density of isolated pitting are most suitable for manual scanning. This approach will find all the pitting and will do so very economically.
If the tank floor has a mid to high density of general corrosion pitting, then the stop-on-defect unit is preferable. The ,stop-on-defect can be set up to stop only when signals come out of the noise and exceed the general noise level. This screening would be difficult to do simply by looking at signals and trying to make the determination on the fly. This method would successfully find the pitting that is visible above the general noise level and this is the best that can be inspected in this situation. It is also the most economical approach for this particular top floor condition.
If the scanner is placed on the tank floor and the screen lights up with multiple levels of color, the technician can become overwhelmed in regards to signal interpretation. This condition favors a mapping tool, where all these signals can be mapped. Some of the signals can be checked with ultrasonic equipment to determine an expectation of pit depth in general regions in comparison to map color. These map colors are more representative of bands of pit depth and not necessarily relative to absolute pit depth. This approach would allow for a fairly rapid inspection and is relative information when planning a comprehensive repair.
The bottom line is that there are applications for all of the approaches. Sometimes the decision of which unit may be made on price or prior experience of tank floor conditions. It is my opinion that the decision of which application should not be made until the technician is in the tank and has performed some scanning to determine density and depth of defects.
Are you ready to move forward?