Sei sulla pagina 1di 13

Ultrasonic Thickness Testing for Compliance With AST Fuel Tank Inspection Requirements

John V. Cignatta, PhD, PE Datanet Engineering, Inc. 11416 Reisterstown Road Owings Mills, MD 21117

ABSTRACT There is much confusion about non-destructive testing as it relates to aboveground storage tank (AST) inspections. Frequently, facility managers are unaware of what types of services are even being employed in the conduct of mandatory tank testing inspections for their sites. This paper shall focus on accomplishing ultrasonic thickness (UT) testing for API 653 and STI SP-001 AST inspections. Examples of state-of-the-art UT testing protocols shall be explained along with the advantages and weaknesses of each. Enhanced graphical presentations of numerical data shall also be shown to illustrate ways to learn faster what the UT data streams are saying about the condition of a give fuel tank and its fitness for continued service.

KEY WORDS Ultrasonic Thickness, A-Scan, B-Scan, C-Scan, API 653, STI SP-001

INTRODUCTION Fuel storage is a field governed by a myriad of regulations, codes and standards. These encompass everything from the certifications of the tanks to many of their component systems (e.g. emergency venting devices by UL or other Nationally Recognized Testing

Laboratory). Unfortunately, everything can wear out and one cannot assume that anything was installed correctly per applicable codes and standards in the first place. There is a very well defined requirement to have ASTs inspected and certified for continued fitness for service under Federal Regulation 40 CFR 112.1 This regulation has gone through a series of revisions over the past decade. There is some understandable confusion over these amendments. For example, a comprehensive set of changes and updates were released in December 2008. They subsequently were placed on hold when the new administration took office in January of 2009. Despite confusion over these changes, the regulation is in place and affects the installation, operation and maintenance of the ASTs at all sites where more than 1,320 gallons of combined storage exists. In Section 112.8.c and 112.12.c, there is very definitive guidance on the inspection and testing of regulated ASTs. The EPA did not create a prescriptive standard (as was done in the case of the regulation of underground storage tanks). Rather, the specific methodology and frequency of inspections were defined as the use of current industry tank inspection standards. The exact wording of this assignment is as follows: Industry standards are technical guidelines created by experts in a particular industry for use throughout that industry. These guidelines assist in establishing common levels of safety and common practices for manufacture, maintenance, and repair. Created by standards-setting organizations using a consensus process, the standards establish a minimum accepted industry practice.2 There are two industry standards which are referenced in the Preamble of the Amendments to this regulation. They are American Petroleum Institute (API) Standard 653, Tank Inspection, Repair, Alteration and Reconstruction3 and Steel Tank Institute (STI) SP-001, Standard for the Inspection of Aboveground Storage Tanks.4 These standards both mandate the use of non-destructive testing protocols such as Ultrasonic Thickness testing or UT to assess the amount and location of thinning of the tanks shell walls, floor and roof. Criteria are then compared against the amount and severity of the metal loss to assess if the tank can continue to be used in service for fuel storage, if repairs are needed and when the next type of inspection will be required. It is a mistake that some have made after a quick reading of these standards that a grace period is available to defer the expense of such testing until a tank exceeds a certain category of testing intervals. Rather, the Federal AST regulations (See 40 CFR 112) are explicit that a baseline on the condition of a particular regulated tank is assumed to exist. It is from that baseline of information that the schedule and type of subsequent inspections are drawn.

Referring to the EPA guidance to their regional inspectors, the following explanation is provided about this issue: When a container has no prior inspection history or baseline information, the implementation of the baseline inspection program is important in order to assess the containers suitability for continued service. Both API 653 and STI SP-001 include details on how to assess a containers suitability for continued service. In some cases, where baseline information is not known, the testing program may include two data collection periods to establish a baseline of shell thickness and corrosion rate in order to develop the next inspection interval (or regular schedule), or an alternative inspection schedule established by the PE in accordance with good engineering practice.5 Thus, all regulated tanks must have a starting point on their overall condition (i.e. a baseline of shell thickness) before an inspection interval can be established. It is the process of obtaining the thickness of the tanks walls, floor and ceiling which will be the focus of this paper. TECHNOLOGY Measuring wall loss of a tank is routinely accomplished around the world by technicians employing UT technology. There are many manufacturers of the necessary equipment and there are also different accreditations for the personnel accomplishing this type of nondestructive testing. By the use of a short pulse of sound such as sonar, the time delay of the returning echo is captured to very precisely measure the thickness of the structure. However, the key to accurately measuring the remaining wall thickness of the steel wall of a tank lies in the technician. To begin, a review of certifications of UT personnel will be presented. This will be followed by the types of UT equipment in use for typical tank evaluations. Lastly, a presentation of the various methods to view UT data will be offered. 1. UT Personnel The purpose of accomplishing UT on in-service and out-of-service fuel storage tanks is to find problems and fix them before a release of fuel occurs. Thus, the technician accomplishing

the work bears a weighty responsibility to accomplish the work properly and not miss a corroding area that must be repaired. Referring back to the referenced API and STI Standards, guidance is provided on the qualifications of these personnel. API 653 calls out in Appendix F that the technician must be either American Society of Nondestructive Testing (ASNT) Level II or III under Standard SNTTC-1A Personnel Qualification and Certification in Nondestructive Testing.6 New personnel with only a Level I certification may accomplish testing of tanks as long as they are under the direct supervision of a Level II/III engineer. STI SP-001 also references the same requirements as API 653. However, an exception is made for spot checks of a tank under certain circumstances by non-accredited personnel.5 One major problem with the required certifications is how a given engineer received their accreditation. To explain, there are two major paths for certification in UTindividual and corporate. Individual certification is typically obtained directly through ASNT. However, it is estimated that the vast majority of UT technicians have corporate certifications. There are also other industry certification programs under ANSI/ASNT Standard CP-189, ASNT Standard for Qualification and Certification of Nondestructive Testing Personnel and CP106:2008, Nondestructive Testing - Qualification and Certification of Personnel. The entire individual certification process of minimum education, examination and experience will not be presented here. The problem that is going to be mentioned is that some corporate certifications are a sham by skipping these requirements for their personnel certifications. A new hire at a certain company might only be appointed to be a corporate Level I, II or III UT technician without education, experience or examination. Thus, tank owners must be concerned with the qualifications of the personnel accomplishing UT testing on their tanks. While many advances in technology have simplified the process of conducting UT testing of tanks and other structures, confidence in the results obtained is totally a function of the qualifications of the technician. The advances in both built-in capabilities and simplicity of operation of current equipment still do not eliminate the need for education, examination and experience. Without these three critical items, the technician is incapable of properly interpreting the information derived in the course of scanning the wall of a tank. To explain, there are two example scans depicted in the photo, Figure 1. The left image shows the information Figure1UTScansfromDifferentSurfaces
4

from a UT scanner where smooth non-corroded steel is tested. In the lower right portion of the data screen is the number 0.223. This is the thickness of the steel floor of a tank. As will be explained subsequently, the more important information is contained in the wiggly line above the numbers. The wavy line is the amplitude scan or A-scan depiction. It is the primary focus of the UT technician to assess whether or not the numeric information is providing a valid reading of the thickness of the steel tank wall. In the left A-scan, the wiggly line of acoustic energy resembles a bouncing ball with ever attenuating amplitude as it gradually stops bouncing. The right image is the result of scanning a nearby section of corroded steel on the floor of the tank. The instrument is providing a steady reading of 0.086 as the thickness of the steel. An untrained technician despite properly calibrating the machine might write down that the floor thickness at this location is less than one tenth of an inch thick (i.e. over 60% loss of metal). Qualified personnel would quickly dismiss the numeric information due to noise that is depicted by the A-scan at this location. It follows that technicians with limited training and experience unproven by examination are not capable of knowing when they are getting the correct thickness readings. The purpose of the UT scanning of a fuel tank is not to get data and fill out reports. Rather, it is to find problems. This is the reason that a facility manager must ascertain the background of the UT technician being hired to evaluate a fuel tank. Thus, the problem lies in that UT data collection is integrally connected with UT data interpretation. If the individual operating the UT equipment cannot interpret the acoustic amplitude waveform depicted for a given scan, then no credence should be given to any numeric thickness data that they collect from the instrument. The minimum items which should be submitted for approval of a tank inspection company include the following for the personnel accomplishing the work on your tank: API 653 or STI-SP-001 certification, And ASNT-TC-1A UT Level II or III individual certification, Or ASNT-TC-1A UT Level II or III corporate certification and include: o Name of the Level III UT engineer that certified the corporate program if the Level II engineer has corporate certification o Name of the Level III UT engineer that provided the corporate training program o Name of the Level III UT engineer that provided the corporate UT examinations

2. UT Equipment

The basic UT gear needed to assess wall thickness losses on a fuel tank is a unit employing longitudinal wave technology with a dual beam transducer and A-scan imaging. A basic set up with transducers and meter can be obtained new for several thousand dollars. Figure 2 is a depiction of the typical UT unit used for A-scan readings. Again, an A-scan reading is a single point reading measuring the thickness of the steel tank wall. The transducer is the combination transmitter and receiver that sends out and listens for the acoustic pulse traversing through the steel. Figure2TypicalAScanUTInstrument For corroded steel, a dual beam transducer must be employed. For new steel, a cheaper (i.e. <$1,000) and more accurate single beam transducer can be used. Single beam technology employs a piezoelectric crystal to convert a pulse of electricity into sound and send it downward into the steel. After a brief wait, the same crystal is used to convert the returning echo of sound into electricity that is amplified and evaluated by the UT meter. The amount of time between the release of the pulse of sound and the first peak of returning echo is divided by two and then multiplied by the acoustic velocity of the material being evaluated. Thus, a UT meter is basically a gong, an acoustic receiver and a time clock. The evolution and advancement of UT technology has progressed to the point that it is the work horse for tank evaluations all over the world. Figure 3 depicts the two piezo-electric crystals and their configuration in a typical dual beam transducer. Note that Figure3CrossSectionofDualBeamUTTransducer there is an angular path from where the transmitting element projects the pulse

of sound downward towards the steel surface. The main reason for the angular path is to maximize the recovery of acoustic energy at the surface of the receiving element. To keep the wave of energy from re-entering the transducer at the site of the transmitter, there is an acoustic barrier that splits the unit in half. The barrier keeps the energy emanating off the transmitter from directly affecting the receiver. Therefore, the receiver is only hearing the acoustic energy that is echoing back from the far side of the steel wall of the tank. Figure 4 provides a cross section of the dual element transducer and the AScan depiction of the acoustic Figure4UTScanProvidingAccurateMetalThicknessData amplitude data received by the UT instrument. As shown previously, the UT gage will depict a numeric value for the thickness of the steel being scanned at this location. However, it is the A-Scan which tells the operator if the numeric information is valid. Note that the A-Scan depiction imitates a ball bouncing across the floor. The signal is gradually attenuating. The small triangle beneath the X-axis is also indicating that the basis for the thickness numeric information is the first large peak received after the initial gong. Thus, the numeric information is correct on the thickness of the steel of the tank wall at this one point. The Figure 5 depiction illustrates when more energy is received from the second echo than the first. The operator must note that the small triangle is beneath the X-Axis is now below the second peak. If the thickness data is recorded at this point, then it will be actually double the thickness of the steel.

Figure5ErroneousThicknessReading(Doubling)

To further explain, an uncorroded (i.e. smooth) surface will cause minimal attenuation of the sound energy as the echo returns back up to the receiving element. If the beam was pointed

directly downward as with a single element transducer, a rough surface would be expected to scatter the energy in all directions. With dual beam transducers, the angular path assures that the maximum amount of the energy will end up on the side of the receiver. One of the biggest problems that must be overcome to obtain accurate thickness data from a steel tank wall is acoustically coupling the transducer to the surface of the steel. Any gaps or irregularities between the tip of the transducer and the surface of the steel will greatly compromise the amount of acoustic energy that can be imparted downward into the steel. With excessive loss of the sound pulse, the echoes will be too faint for the transducer to hear. A typical tool to minimize this surface scatter of energy is the use of a couplant between the transducer and the steel. While air is a good attenuator of sound energy, water and other viscous fluids are excellent conductors of sound. This principle is the reason that a layer of a couplant is employed between the transducer and the steel to minimize energy loss during the measurement of the structures thickness. Despite the improvements in energy conveyance with the use of a couplant, not all surface irregularities can be overcome by its use on a fuel tank. All rust, loose paint and scale must be removed from the surface before coupling the transducer onto the substrate. Figure 6 is a real world condition of many tanks. The blistered and delaminated paint due to the ongoing corrosion has created thousands of voids which will severely attenuate the pulse and make a thickness reading practically impossible. To scan anywhere in the area where the paint is popping off the tank, the surface must be polished down typically with a non-woven abrasive paint removal wheel. Figure6CorrodedandIrregularSurfacesonTanks

It should be appreciated that a single point thickness reading does not provide a very good depiction of the integrity of a fuel tank. Figure 7 is an elephant. Consider the man blind from birth reaching out and describing an elephant after touching it in just one spot. He will have in his mind vastly different concepts about an elephant if he touches first the tusk, the ear or the end of the tail. From his very limited data gathering exercise (i.e. one touch), his concept of an elephant is far from complete. To provide the blind man with a better picture of the animal, he is permitted to walk along its side and touch it continuously four feet Figure7DifferentSurfaceConditionson from the floor. This line of information will then ComplexObjects give the man a far better (though still limited) concept on what is an elephant. If the blind man is then permitted to walk all around the elephant and touch him on all sides, it will complete his picture of the animal. There is a strong analogy of the blind mans limited information on the elephant with evaluating fuel tanks. An A-scan is again the thickness of steel at a single point. This is comparable to the man being limited to a single touch of the giant beast. When the blind man walked down the side of the elephant, he was actually obtaining many points of information. In the same fashion, a string of UT readings down the side of the tank provides a two dimensional crosssectional depiction of the tank along that line. This technique is called a B-scan. Lastly, many lines of cross-section of a tank can be spaced across the tank to generate a topographical map of the steel. This is referred to as a C-scan. Referring back to the industry standards for tank inspections, there is not a definitive guidance on which type (i.e. A-, B- or C-scans) of UT measurements must be provided. But, enough readings must be taken to provide a defensible record of the conditions of a tank on a given inspection day. Another analogy for taking enough readings is the game Battleship. In the early stages of the game, it is unusual to find any of the opponents ships across the big ocean unless you are

Figure8ManuallyScanningSideof FuelTank

lucky. With liabilities from missing problems on tanks, inspection personnel do not rely upon luck. Rather, thousands upon thousands of readings are taken to increase the likelihood of finding a problem. The Figure 8 photo is an engineer rappelling down the side of a fuel tank. In his right hand is the transducer which is obtaining a continuous B-scan of the tank. By accomplishing a series of these vertical cross sections of the tank, a good picture emerges on the shell condition. Any bath tub ring effect from poor water control corroding along a line anywhere around the tank would be quickly detected. At one site, it was found that a mistake had been made when the tank was erected at the site. The thickest sheet stock of the steel should be the lowest course. The thinnest steel should be at the highest course. On this tank, the steel courses had been mixed up. This resulted in a de-rating of the tank due to the weakness of the thinner steel being on a lower course of the shell. To facilitate obtaining B-scan information on fuel tanks, there are now robotic carts that will convey the transducer wherever the operator guides it with their simple joy stick controller. Figure 9 is a unit with magnetic traction wheels that are so strong that the unit can hang inverted from a ceiling while scanning the steel. There is a considerable step up in expense when advancing from A-scan technology to that of B- and C-scans. This is one of the reasons why there are such great disparities in bids obtained for tank inspection services. While one firm plans on taking a few A-scans, another details in their proposal that B-scans will be accomplished along the lower sixty degrees of a horizontal cylindrical tank on a one foot spacing. The difference in these two proposals will be thousands of dollars. A tank owner might be pound foolish in thinking that there is considerable savings by using the low bid contractor. However, good decisions require good data. Further, a very incomplete picture of a tanks integrity can be the basis for disastrous decisions. Figure9MagneticWheeledRoboticUTCart

3. UT Data Presentation

10

The UT data from an evaluation of a given tank must be preserved in the records for the facility. As mentioned previously, it is needed to show both the baseline of the tank and changes over time. The first problem of the mountain of data from a tank evaluation is organization. Figure 10 is one method of mapping out a horizontal cylinder tank. All UT data is organized in rows and columns that are defined by distances along the tanks length. In some cases, complete circumferential B-scans are made. Typical fuel tanks have the majority of their corrosive deterioration on their interiors limited to the lower several feet Figure10MappingTankstoOrganizeUTData of their storage area. Some tanks like heated residual fuel oil tanks can have serious crown corrosion to their ceiling areas due to moisture condensation above submerged heaters. Regardless of the sampling strategy, there must be a defensible engineering argument for its selection and use. Further, there must be an established way to go from the data files back to a location on the tank. Figure 11 is a graphical depiction of a Bscan for a vertical cylindrical tank. The darker blue colors reflect the thinner steel at the upper courses of the shell. The lighter green colors represent the thicker steel of the lower courses of the shell. The black lines are areas missed due to welds, rust or Figure11GraphicalDepictionofUTThickness DataDowntheSideofaVerticalTank excessively thick paint. The numbers across the bottom reflect inches for the measurements starting at the roof of the tank. The overall file designation is the tank number, the inspection date and the compass position around the tanks circumference.

11

Figure12CScanProvidesaTopographicalMapofTankShell When many B-scans are accomplished in a close spacing across a tank, then a topographical map of the cross-section of the tank steel can be derived such as in Figure 12 above. Each band of horizontal data is comprised of A-scans taken every 0.1 around half of a horizontal tanks circumference. Thus, this depiction is a map of the left side of the tank. The yellow bands are areas where brackets were welded to the tanks unseen interior wall. The dark blue patches are areas where further testing is needed. Figure 13 is a simple Excel spreadsheet presentation of C-scan data. By the use of both colors and numeric information, conclusions can quickly be made as to the severity of corrosion

Figure13ExcelSpreadsheetPresentationofUTDatafromTankShell
12

detected and its significance for compliance with industry tank inspection standards.

CONCLUSIONS It is no longer acceptable to allow interior corrosion problems on a fuel storage tank to progress to the point that a release occurs. To control such eventualities, Federal tank regulations mandate ongoing inspections and testing be conducted per the requirements of industry standards. One critical part of these inspections is wall loss measurements. There are a variety of methods to accomplish this task. However, the qualifications of the UT technician along with the number of UT scans and their graphical/tabular presentation are all critical for finding and correcting problems before yet another avoidable fuel release occurs. Tank owners must be cognizant of the latitude exercised by some inspection companies in accomplishing the requisite UT testing portion of a tank inspection. By specifying minimum inspector qualifications and training, corporate errors & omissions insurance with pollution coverage, types of scanning and depiction of results, the tank owner will obtain the best return from the expense of these mandatory inspections.

REFERENCES 1. 2. ENVIRONMENTAL PROTECTION AGENCY, 40 CFR Part 112, Oil Pollution Prevention; Non-Transportation Related Onshore and Offshore Facilities ENVIRONMENTAL PROTECTION AGENCY, 40 CFR Part 112, Oil Pollution Prevention; Non-Transportation Related Onshore and Offshore Facilities, Amendments Dated November 21, 2008, page 109 AMERICAN PETROLEUM INSTITUTE, API Standard 653 Tank Inspection, Repair, Alteration, and Reconstruction, Third Edition, Includes Addendum 1 (2003), Addendum 2 (2005), Addendum 3 (2008) STEEL TANK INSTITUTE, STI-SP-001, Standard for the Inspection of Aboveground Storage Tanks, Issued July 2006, 4th Edition ENVIRONMENTAL PROTECTION AGENCY, SPCC Guidance for Regional Inspectors, Version 1.0, November 28, 2005, Chapter 7, Page 7-17 AMERICAN SOCIETY for NONDESTRUCTIVE TESTING, Standard SNT-TC-1A Personnel Qualification and Certification in Nondestructive Testing, Fifth Edition, 2006

3.

4. 5. 6.

13

Potrebbero piacerti anche