Random radiography

What the fabricator or contractor and the owner need to agree on

By Walter Sperko, P.E.

July 12, 2005

Fabricators and contractors always should check to be sure that the extent of nondestructive examination and the acceptance criteria required are understood clearly by all parties and documented in writing.

Fabricators and contractors always should check to be sure that the extent of nondestructive examination and the acceptance criteria required are understood clearly by all parties and documented in writing.

Piping fabricators and contractors occasionally get contracts that specify "random" radiography. The American Society of Mechanical Engineers (ASME) Code B31.3, "Process Piping," requires 5 percent random radiography of each "lot" of pipe as the default inspection requirement. B31.3 also says that once the radiographs representing that lot are accepted, the entire lot is accepted.

With that background in mind, consider the following horror story.

Attached for your review is a chronology of events related to pipe weld radiography and a sample page from the project piping line list.

After reading the attached, you may find the following explanations to be helpful:

1. The lots of pipe that the owner claims were required to be radiographed may be confusing. This may help: Our shop fabrication subcontractor used what we believe is the B31.3 recommendation for lots of pipe (fluid service combined with percentage radiography). At some point in the project, the owner instructed our field quality control inspector to use the "process service systems" as the designated lots for radiography at the job site. This resulted in approximately 35 to 40 lots. The project pipe was turned over and accepted by the owner using the process service systems as the designated lots for weld radiography after additional radiographs were made on shop welds.
   
   After the project was completed, the owner told us that we should have known from the engineering documents what the lots should be, so we requested copies of the relevant documents. The owner sent a copy of the piping line list and his interpretation that the X-ray column meant that each line on the piping line list was a lot. This would have resulted in approximately 450 lots instead of 35 to 40.
   
   
2. The engineer who developed the piping line list supports our case. Although we believe his comments are correct, his remarks don't carry much weight because he was contracted to us.
   
   
3. You may ask why our quality control inspector at the job site didn't question if the pipe fabricator was using the same designated lots after the owner told him to use the process service systems as the basis for lots at the job site. Our quality control inspector at the job site didn't have responsibility for the fabricator. In addition, the owner was performing inspections at the pipe fabrication shop, so our site quality control inspector logically assumed that the owner's inspector directed the fabrication shop regarding the designated lots that he expected.
   
   
4. At the bottom of the sample page from the piping line list are related notes from the "Legend and General Notes" page. When the owner sent us the piping line list as his evidence, he highlighted the note that reads "Nondestructive Test Requirements Shall Be in Accordance With Owner's Specification." Our people have reviewed the owner's specification and cannot find any requirements beyond ASME B31.3.
   
   
5. During our investigation into this matter after this project was completed, we discovered that we had performed two B31.3 projects for the same owner at the same site within the past five years where, to the best of our knowledge, the designated lots for radiography were the same as the pipe fabricator used on this project.

It appears that there was poor communication before the project started regarding what constituted a designated lot. The owner claims that each line on the piping line list designates a lot, and he insists that additional radiography be performed to satisfy his interpretation—at our cost.

We do not believe that we should have been expected to interpret the piping line list as the owner has interpreted it without additional notes on the "Legend and General Notes" page.

This whole matter seems to boil down to how the piping line list should have been interpreted reasonably. We are asking you for an independent and impartial opinion on this matter.

How bad is this situation? To the casual observer, the only action that needs to be taken is radiography of more welds. But those familiar with the random radiography requirements of B31.3 recognize the contractors' potential exposure to significant financial damage is nothing short of gigantic.

What B31.3 Means to Radiography

Under B31.3 rules, if a radiograph of a weld representing a lot is rejected, two more welds have to be radiographed. If one of those is rejected, two more welds have to be radiographed. If one of those is rejected, the entire lot has to be radiographed. The contractor's exposure, therefore, is the cost of doing 100 percent radiography on piping that is already completely installed.

When random radiography is imposed, agreement on the size of a lot is critical, and there has been so much controversy over it that in 1995 ASME added the following footnote to paragraph 344.1.3 of the B31.3 specification:

"A designated lot is that quantity of piping to be considered in applying the requirements for examination in this Code. The quantity or extent of a designated lot should be established by agreement between the contracting parties before the start of work. More than one kind of designated lot may be established for different kinds of piping work."

The basic definitions in B31.3 use "lot" extensively:

• 100 Percent Examination:Complete examination of all of a specified kind of item in a designated lot of piping.
• Random Examination:Complete examination of a percentage of a specified kind of item in a designated lot of piping.
• Spot Examination:A specified partial examination of each of a specified kind of item in a designated lot of piping; e.g., of part of the length of all shop fabricated welds in a lot of jacketed piping.
• Random Spot Examination:A specified partial examination of a percentage of a specified kind of item in a designated lot of piping.

Clearly, agreeing on the definition of a "lot" is important in determining the extent of examination that is required. For percentage radiography, B31.3 has a second footnote that is important:

"Items not examined in a lot of piping represented by such examination may contain defects which further examination could disclose. Specifically, if all radiographically disclosable weld defects must be eliminated from a lot of piping, 100 percent radiographic examination must be specified."

This footnote says that if the owner wants 100 percent radiographic-quality welds, he needs to specify (and pay for) radiography of those welds. The owner also needs to state this in his specifications to establish prior agreement regarding the extent of radiography required.

Random Radiography and the Owner

So when an owner specifies random radiography, what is he buying? Simple: the fabricator's or contractor's undivided attention. When random radiography is imposed, the following will happen with reputable contractors and welding personnel:

• The contractor or fabricator will pay more attention to the skill of each welder assigned to the work.
• The welder will spend more time preparing ends and precleaning.
• The welder will spend more time on fit-up and alignment. The fabricator may have to add weld metal to the inner and outer surfaces and grind it locally to get internal alignment that will result in an acceptable radiograph.
• The welder will spend more time making and preparing tack welds so that they tie in properly and easily to the rest of the root.
• The welder will spend more time making the root pass. (Sometimes two welders work together.)
• The welder will spend more time cleaning and grinding to get uniformly contoured layers of weld metal between passes.
• The welder will spend more time preparing the cover pass for examination.
• Additional supervision or inspection personnel will be needed to track, monitor, and verify that the welders are doing the job in the optimum manner so that the radiographs pass the first time.

Obviously, random radiography has its desired effect only if the welders and supervisors know it will occur. This is supported by the following note in ASME Section VIII ("Unfired Pressure Vessels") Division 1, UW-52:

"Spot radiographing of a welded joint is recognized as an effective inspection tool. The spot radiography rules are also considered to be an aid to quality control. Spot radiographs made directly after a welder or an operator has completed a unit of a weld prove that the work is or is not being done in accordance with a satisfactory procedure. If the work is unsatisfactory, corrective steps can then be taken to improve the welding in the subsequent units, which unquestionably will improve the weld quality."

If an owner receives more of the welder's and the fabricator's attention when percentage radiography is imposed, will he always get welds that are the equivalent of 100 percent radiographed welds? No. When specifying random radiography, the owner should realize that there is potential for significant discontinuities to exist in welds that were not radiographed.

To maximize the effectiveness of percentage radiography, it must be performed contemporaneously with welding. Percentage radiography performed weeks after the work is done provides no feedback to the welder or supervisor about the quality of the work, making it a waste of time and money. Randomly radiographing welds after all the work is done will contribute nothing to the quality or workmanship of the welds that are not radiographed. Further, it will raise many ugly issues that, after much heated battle, will have little or no effect on the integrity of the piping system. It's simply not possible to inspect the quality into the system after construction is complete; it must be built into the system from the start and throughout the project. Random radiography attempts to achieve this goal while avoiding the cost of performing 100 percent radiography.

The range of nondestructive examination B31.3 imposes depends on the fluid service category that the owner specifies for a given piping system. These categories are:

• Normal—ordinary process piping service
• Category M—Lethal fluid service
• High Pressure—Exceeds ASME B16.5 pressure limits for Class 2500 flanges
• Severe Cyclic—Stress at a joint exceeds 80 percent of the allowable stress range and the number of thermal cycles will exceed 7,000
• Category D—Innocuous fluid, low-pressure, and not very high-temperature service

B31.3 specifies that the owner must evaluate his piping systems and assign fluid service categories as appropriate. If the owner doesn't designate a fluid service category, normal fluid service rules apply.

The extent of examination for fluid service categories is:

• Normal—5 percent random radiography of butt welds
• Category M—20 percent random radiography of butt welds
• High Pressure—100 percent radiography of butt and branch connections
• Severe Cyclic—100 percent radiography of all butt welds and branch connections
• Category D—Visual examination only

Each fluid service category also has its own radiographic acceptance criteria. Some unexpected highlights are:

• Normal—Allows some incomplete penetration
• Category M—Same as normal fluid service
• High Pressure—Tight limits on reinforcement both OD and ID, no undercut permitted (either OD or ID), and a surface roughness limitation of 500 microinches; no incomplete penetration permitted
• Severe Cyclic—Similar to acceptance criteria for high pressure

Fabricators and contractors should pay attention to what radiographers use as acceptance criteria.

During a recent job an owner was upset over a high radiographic reject rate (18 percent) for the contractor erecting piping on a polypropylene plant. Investigation revealed that the radiographer was interpreting the film to the severe cyclic acceptance criteria rather than to normal fluid criteria because that's what "someone" told him to do. Once the miscommunication was cleared up, the reject rate dropped to 2 percent. Thus, it's important for the fabricator or contractor to ensure that the proper radiographic acceptance standards are applied.

Because it specifies different fluid service categories, B31.3 is a difficult section of ASME B31 to deal with relative to nondestructive examination requirements. B31.1, "Power Piping," is not so difficult. Table 136.4 of B31.1 requires 100 percent radiography of butt welds and branch connections that are:

• In service over 750 degrees F regardless of pressure.
• In service over 1,025 pounds per square inch gauge and over 11/8 inches thick and between 350 and 750 degrees F.

All other welds require only visual examination.

B31.1's acceptance criteria are more stringent than those of B31.3 for normal fluid service in that no incomplete penetration is permitted and the size limits on other types of discontinuities are smaller.

The easiest B31 code sections are B31.5, "Refrigeration Piping," and B31.9, "Building Services Piping." Neither has additional examination requirements beyond visual except for B31.5 when the refrigerant is flammable. In that case, 5 percent radiography is required, but unlike B31.3, it has no escalation clause for rejected radiographs.

Specifying the Right Code Section

Being sure that everyone knows which B31 code section applies is critical, since the section followed dictates the extent of nondestructive examination required.

One of the most dangerous scenarios occurs when the specification lists several B31 sections and then leaves it up to the fabricator or contractor to select the section that applies. Don't fall for it. Be sure that the applicable code section is specified in the contract documents. Although the introduction to each B31 section requires the owner to specify which section applies to the work, some owners are reluctant to do so. In these cases, specify the section that you will follow in your quote for the work.

It is acceptable for more than one B31 section to apply to different systems within a facility (for example, the process piping might follow B31.3, the steam plant may follow B31.1, and the office piping may follow B31.9.), but the section that applies to any given system has to be clearly and unambiguously defined in the contract documents. No step in a piping contract is more basic than agreeing on the code that will apply to the work.

Revisiting the story at the beginning of this article, owners and their engineers frequently make unreasonable demands in the engineer's specifications, but the worst situation occurs when, toward the end of a job, the owner insists on radiographing welds that were not required to be radiographed by code or by contract. Or worse—the owner radiographs them himself and gets results that he doesn't like. This violates standard industry practice expressed in ASME B31.1, paragraph 136.1:

"The degree of examination and the acceptance standards beyond the requirements of this Code shall be a matter of prior agreement between the manufacturer, fabricator or erector and the Owner."

Standard industry practice expressed in this paragraph requires the fabricator or contractor to agree to purchaser-imposed examinations and the acceptance standards before the job starts. Similar paragraphs are found in B31.3, B31.5, B31.9, ASME Sections I and VIII, and American Welding Society (AWS) D1.1 through D1.6, so this practice is not unique to piping.

The reason for this practice is that it costs more to make welds that are radiographic-quality than it does for welds that won't be radiographed. It costs more for the same previously noted reasons: It requires better personnel, more effort, and more supervision and generally takes longer. Put simply, it costs more to make welds that can pass radiographic examination than it does to make commercial-quality welds—and the contractor who makes such welds should be compensated for the additional work.

Another reason for prior agreement pertains to piping that has already been erected: The cost to radiograph a system after it's installed is steep compared to the cost of radiographing the same system during erection. Scaffolding must be set up, insulation removed, radiography done in situ, welding equipment located near the needed repairs, postweld heat treatment possibly performed, and nearby equipment protected from damage during repairs. The costs that would not have been incurred during construction can be huge.

Fabricators and contractors always should ensure that the applicable piping code section and the extent of nondestructive examination and the acceptance criteria are understood clearly by all parties and documented in writing before the work starts. Anything less is just asking for trouble.

Walter J. Sperko, P.E., is president of Sperko Engineering Services Inc., 4803 Archwood Drive, Greensboro, NC 27406, 336-674-0600, fax 336-674-0202, sperko@asme.org, www.sperkoengineering.com.

This article is adapted from Walter J. Sperko's conference presented at Pipe Fabricating 2005 April 5-6, 2005, Appleton, Wis., ©2005 by the Fabricators & Manufacturers Association Intl. (FMA).