CHEVRON Pressure Vessel - in Service Inspection

700 In-Service Inspection

AbstractThis section discusses in-service inspection of pressure vessels, including code and jurisdiction requirements of the ASME Code, API Pressure Vessel Inspection Code, and the National Board Inspection Code. It also compares the functions of an autho-rized inspector with those of a technician certified by the American Society of Nondestructive Testing (ASNT).

Various forms of deterioration are discussed, including internal and external corro-sion, thermal aging, fatigue, stress corrosion (environmental) cracking, internal erosion, hydrogen attack and blistering, and creep and stress-rupture.

The section also analyzes in-service inspection data for each form of deterioration. The fitness-for-service analysis discusses the information needed for the analysis, the procedure for carrying out the analysis, and required inspections. Among the nondestructive examination (NDE) techniques discussed in some detail in this section are: visual examination (VT), dye-penetrant examination (PT), magnetic particle examination (MT), ultrasonic examination (UT), radiographic examination (RT), and acoustic emission testing (AE). Inspection checklists are provided for distillation columns, pressure vessels, reactors, and spheres.

Note that in-shop inspection of new vessels is discussed in Section 670.

Contents Page

710 Introduction 700-4

711 Design for Inspection

720 Code and Jurisdiction Requirements 700-5

721 ASME Code Requirements

722 Jurisdiction Requirements

723 API 510, Pressure Vessel Inspection Code

724 National Board Inspection Code

725 Authorized Inspector vs. ASNT Certified NDE Technician

730 Forms of Deterioration 700-7

731 Internal Corrosion

Chevron Corporation 700-1 March 1990

700 In-Service Inspection Pressure Vessel Manual

732 External Corrosion

733 Thermal Aging

734 Fatigue

735 Stress Corrosion (Environmental) Cracking

736 Internal Erosion

737 Hydrogen Attack

738 Hydrogen Blistering

739 Creep and Stress-Rupture

740 Analysis of In-Service Inspection Data 700-17

741 Internal Corrosion

742 External Corrosion

743 Thermal Aging

744 Fatigue

745 Stress Corrosion (Environmental) Cracking

746 Internal Erosion

747 Hydrogen Attack

748 Hydrogen Blisters

749 Creep and Stress-Rupture

750 Fitness-for-Service Analysis 700-22

751 Introduction

752 Background

753 Information Required for Fitness-for-Service Analysis

754 Procedure for Fitness-for-Service Analysis

755 Required Inspection at the Next Scheduled Shutdown

760 Nondestructive Examination (NDE) Techniques 700-35

761 Visual Examination (VT)

762 Dye-Penetrant Examination (PT)

763 Magnetic Particle Examination (MT)

764 Ultrasonic Examination (UT)

765 Radiographic Examination (RT)

766 Acoustic Emission Testing (AE)

770 Inspection Checklists 700-74

771 Inspection of Distillation Columns

March 1990 700-2 Chevron Corporation

Pressure Vessel Manual 700 In-Service Inspection

772 Inspection of Pressure Vessels

773 Inspection of Reactors

774 Inspection of Spheres



710 IntroductionIn-service inspection refers to inspection after new equipment is placed in service, although these inspections are usually made during scheduled shutdowns. An in-service inspection program will significantly improve reliability and safety of opera-tion, and increase productivity and profitability by preventing unscheduled shut-downs for emergency repairs.

Pressure vessels can deteriorate during normal operation and periods of process upset due to:

1. Internal and external corrosion.

2. Thermal aging of material.

3. Mechanical and thermal fatigue.

4. Stress corrosion (environmental) cracking.

5. Internal erosion.

6. Hydrogen attack.

7. Hydrogen blistering.

8. Creep and stress-rupture.

Materials for the construction of pressure vessels are selected to minimize deteriora-tion during service (see Section 500), but their actual performance is not always exactly as predicted. Consequently, a well planned in-service inspection program is necessary to assure reliable and safe operation.

All of the above forms of deterioration will not necessarily occur with each vessel. The inspection program for a vessel should be established to detect only those forms of deterioration that may occur during normal operation or process upsets. The specific forms of deterioration that the in-service inspection program covers should be determined by experienced pressure vessel and materials engineers, with regard to the process environment and service conditions.

Nondestructive examination (NDE) techniques can detect most forms of deteriora-tion, and these techniques can be carried out at an early stage to permit continued operation until repair or replacement can be scheduled. Therefore, in-service inspec-tion will prevent most equipment failures that cause unscheduled shutdowns for repairs and interrupt production.

711 Design for InspectionThe requirements for future in-service inspections should be considered when new pressure vessels are designed and constructed. In particular, weld joints should be designed with configurations and geometries that permit nondestructive examinations for deterioration.



nd

hen for

isual d , the

hem y of

or hese

y

Selection of the most appropriate design details for weld joints requires a thorough evaluation of the various forms of deterioration that may occur during service, and the nondestructive examination procedures that can be used to detect the deteriora-tion. Furthermore, inspection during fabrication of the vessel may have to be more comprehensive than required by the ASME Code, to provide baseline data for the in-service inspections. It is often very difficult to distinguish pre-existing fabrica-tion flaws from those that may have subsequently developed during service, and requires a highly skilled NDE examiner. Although nondestructive examination procedures are better now at identifying various flaw types, lack of appropriate baseline data can complicate in-service inspection.

Designing for in-service inspection can exceed the requirements for assuring the integrity of a new pressure vessel as constructed, but the modification of design details and additional inspections during fabrication can frequently be obtained with minimal additional cost. Nevertheless, additional initial expense for vessels that may be susceptible to some form of deterioration during service may be justified by the ability to properly assure through in-service inspection their integrity and reli-ability after a long service history.

720 Code and Jurisdiction Requirements

721 ASME Code RequirementsThe ASME Code, Section VIII, applies directly only to the “design, fabrication, ainspection during construction of pressure vessels.” This statement is normally interpreted to mean that the direct applicability of the ASME Code terminates wthe authorized inspector authorizes application of the ASME Code Stamp to thenameplate on the vessel. It is the intent of the ASME Code to provide “a margindeterioration in service so as to give a reasonably long life,” and openings for vexamination or a manway for entry are required to permit in-service inspection.However, the Foreword to the ASME Code makes it clear that it “deals with thecare and inspection of pressure vessels only to the extent of providing suggesterules of good practice as an aid to owners and their inspectors.” In other wordsASME Code does not directly govern the in-service inspection requirements for vessels.

722 Jurisdiction RequirementsMost jurisdictions require the owner/operators of pressure vessels to maintain tin safe operating condition, but they do not usually regulate either the frequencin-service inspection nor the types of inspection required to keep them safe. However, many jurisdictions refer to either the National Board Inspection Codethe API Pressure Vessel Inspection Code (API 510), and accept conformity to tdocuments as a satisfactory in-service inspection program.

The Company prefers to use API 510 whenever possible. API 510 is specificalloriented to requirements for the safe operation of vessels in the hydrocarbon



of ner/ oth

due ed

ernal 510 s

ally ata um

in-first -e

ina-aph ed not

re een rce

n e

e al

processing industry and provides somewhat greater flexibility based upon the owner/operator’s actual operating experience and the expertise of his inspectors.

723 API 510, Pressure Vessel Inspection CodeAPI's “Pressure Vessel Inspection Code (API 510)” and its “Guide for InspectionRefinery Equipment, Chapter VI, Pressure Vessels,” give useful guidance to owoperators for meeting their obligation to maintain safe and reliable equipment. Bdocuments are included in this manual. Note that Chapter VI is out of print andto be replaced by API RP 572 in 1990. However, a copy of Chapter VI is includfor your reference.

API 510 is more specific about in-service inspection of a pressure vessel for intcorrosion than for any other form of deterioration that may occur. The maximuminterval between internal inspections of pressure vessels recommended in API is one-half of the remaining life related to corrosion, with a maximum of 10 year(API 510, Paragraph 4.3).

API 510 contains separate guidelines for pressure vessels used in oil and gas producing operations.

The corrosion allowance specified as a design condition for new vessels is usudetermined to give an expected life of 20 or 30 years, based on corrosion-rate dobtained from laboratory tests or based on operating experience with vessels insimilar service. It is permissible to adjust the corrosion rate based on the maximcorrosion rate actually exhibited by the vessel, to establish subsequent internalservice inspection intervals. Actual corrosion rate is the rate established at the in-service inspection. Large vessels with two or more zones exhibiting differentcorrosion rates can have different inspection intervals established for each individual zone. It is permissible to determine the depth of corrosion to satisfy thesrequirements while the vessel is in operation, providing the nondestructive examtion (NDE) procedure used can give sufficiently accurate data (API 510, Paragr4.1). Vessels that are known to have a corrosion rate of less 0.001 inch/year nebe inspected internally.

The other forms of deterioration that can occur during the operation of a pressuvessel are mentioned in API 510 (API 510, Paragraph 3.5), but it is not specificabout the in-service inspections that should be performed nor the intervals betwinspections. The experience of the owner/operator inspectors is the primary soufor information about potential forms of deterioration that can occur in a vessel under its operating conditions. Visual examination (VT) is the primary inspectioprocedure. Depending on the form of deterioration, visual examination should bsupplemented by other types of nondestructive examinations.

724 National Board Inspection CodeThe “National Board Inspection Code” is issued by the National Board of Boilerand Pressure Vessel Inspectors, which is composed of Chief Inspectors from thjurisdictions in the US and Canada that have adopted the ASME Code as a leg



requirement for design and construction of boilers and pressure vessels. The NB Code recognizes the existence of API 510 and does not claim to supersede it. However, some jurisdictions mandate compliance with the NB Code in preference to API 510.

The NB Code does not differ greatly from API 510 with regard to the technical requirements for the in-service inspection of pressure vessels that can affect safety of operation, although it is less specific about forms of deterioration other than corrosion that can occur in vessels used for hydrocarbon processing.

One significant difference is that the NB Code requires authorized inspectors employed by owner/operators to obtain a commission from the National Board and that the governing jurisdiction grant approval for owner/operator in-service inspec-tion (NB Code, Glossary of Terms), whereas API 510 simply permits the owner/operator to designate authorized inspectors who have appropriate qualifica-tions and experience (API 510, Paragraph, 1.2.5). The NB Code also formalizes recordkeeping (NB Code U-104), but this does not significantly exceed normal good practice.

725 Authorized Inspector vs. ASNT Certified NDE TechnicianA distinction should be made between the functions of Authorized Inspectors and ASNT (American Society of Nondestructive Testing) Certified NDE Technicians with regard to the in-service inspection of pressure vessels. An Authorized Inspector is required to have a knowledge of the ASME Code and NB Code (or API 510) rules and requirements, and must determine if the condition of a vessel is satisfac-tory for continued service with respect to these Codes. The ASNT Certified NDE Technician, on the other hand, must demonstrate proficiency with nondestructive examination procedures that are used to determine the extent of deterioration that has occurred during service. The Authorized Inspector may not, therefore, have the required skill to satisfactorily perform nondestructive examinations, but the ASNT Certified Technician cannot pass on the suitability of a vessel for continued service. These two functions are frequently performed by the same individual in an owner/operator organization, but this need not necessarily be the case.

730 Forms of DeteriorationThe major forms of deterioration that can impair the integrity and reliability of pres-sure vessels are discussed below, primarily with regard to:

1. The process environments and service conditions that can cause the specific form of deterioration.

2. The physical characteristics of the specific form of deterioration that can be detected by nondestructive examination.

A detailed discussion of the mechanisms responsible for the various forms of deteri-oration is beyond the scope of this manual, but much of this information can be found in the Corrosion Prevention and Metallurgy Manual. Materials are discussed in Section 500.



s of -

0°F. t it

in-es-

at nsed oo nts to -

oint re

731 Internal CorrosionInternal corrosion is the most frequently encountered form of deterioration that can affect the integrity and reliability of a pressure vessel. Fortunately, it is usually also the easiest form of deterioration to detect during an in-service inspection, providing the correct locations inside the vessel are examined and the proper nondestructive examination procedures are used.

Many process environments are corrosive to the materials used for the construction of pressure vessels. For example, process environments that contain sulfur, chlo-rine, and organic or inorganic acids can be highly corrosive to carbon steels and low-alloy steels. The corrosion attack can be a general wastage of the material, or it can be highly localized. Deposits on internal surfaces or held in crevices can trap corrosive compounds in contact with the vessel shell and may cause severe local-ized corrosion.

Visual examination (VT) is usually adequate for detecting internal corrosion of pres-sure vessels. To detect some types of localized corrosion, the inspectors must look at locations that may be obscured by internal components or covered by deposits. Removal of the corrosion scale is necessary to determine the depth of localized corrosion and pitting. Ultrasonic examination (UT) using a longitudinal wave proce-dure from the O.D. surface of the vessel can also be used to detect internal corrosion and to determine the remaining thickness. It is helpful to mark locations of localized corrosion or pitting on the O.D. surface to guide the ultrasonic examination. Ultra-sonic imaging techniques are available to “map” (discussed in Section 764) arealocalized corrosion or pitting and are especially useful for monitoring the progression of corrosion attack from one in-service inspection to the next.

Corrosion by SulfurProcess environments that contain sulfur compounds, such as H2S, can be very highly corrosive to carbon steels and low-alloy steels at temperatures above 50This type of corrosion generally occurs as a general wastage of the material, bucan be more severe at locations of high velocity or impingement. Austenitic staless steel cladding or weld overlays may be required to protect the shell of a prsure vessel from this type of corrosion.

Sulfur compounds can also cause corrosion of carbon steel and low-alloy steeltemperatures below the dew point of water, when they are dissolved in a condeaqueous phase to form an inorganic acid. This type of corrosion is usually not tsevere and can be adequately handled with a corrosion allowance or adjustmethe operating conditions to prevent the condensation of water. Process environments that contain ammonia in addition to H2S can be significantly more corrosive.

Corrosion by ChlorineCorrosion by chloride compounds occurs only at temperatures below the dew pof water. This type of corrosion usually occurs below the liquid level, unless thevapor space is below the dew point. It can be very severe and highly localized.Carbon steel surfaces can become severely pitted, and grooves can occur whe



of

over-n a

. d for

vice

d-n can

tmo-

st

y)

ng

t essel and

continuously moving liquid streams contact a vessel, such as at overflow weirs and reflux return nozzles.

Corrosion by AcidsMany crude oils contain napthenic acid that can cause severe corrosion resembling pitting. Other inorganic acids (including sulfuric and hydrofluoric) are used in various processes in refineries, and they can cause corrosion under certain process conditions.

Polythionic acid can form when water condenses on a vessel surface covered with a sulfide scale. This acid can cause intergranular corrosion and cracking of austenitic stainless steel vessels or of the cladding on carbon steel and low-alloy steel vessels, especially in heat affected zones that are sensitized by welding. This phenomenon is commonly referred to as “knife-line-attack.” Stabilized grades of stainless steel(i.e., Type 321 and Type 347) can be used to minimize susceptibility to this typeattack.

Corrosion Protection Claddings and Weld OverlaysSome pressure vessels are provided with corrosion-resistant claddings or weldlays to protect the vessel shell in severely corrosive process environments, whereasonable corrosion allowance would not be sufficient for high corrosion ratesHigh-alloy materials, such as austenitic stainless steel or Monel, are usually usethese linings. Deterioration of the protective cladding or weld overlay during sercould expose the vessel shell to severe corrosion attack.

Visual examination (VT) is usually sufficient to determine the integrity of the clading or weld overlay on a pressure vessel shell. Dye-penetrant (PT) examinatiosupplement VT to determine if cracks are developing in the lining, and hammertesting can indicate if the lining has separated from the vessel shell.

732 External CorrosionExternal corrosion of a pressure vessel is always attributable to some form of aspheric corrosion. It is highly dependent on the natural atmospheric conditions prevailing at the geographic location. The natural atmosphere at humid sea coalocations can be expected to be more corrosive than at dry inland locations. However, the chemical emissions at the plant site (or from other nearby industrcan considerably increase the corrosivity of the atmosphere.

Locations where rain water can accumulate on pressure vessels for relatively loperiods are especially prone to external attack. Crevices or pockets created bysupports, rings, and other external attachments are typical examples.

Externally insulated vessels should be provided with weatherproofing to prevenrain water from seeping into the insulation where it can be trapped against the vshell. Soluble compounds in the insulation can be leached out of the insulation contribute to the external corrosion, if the weatherproofing is not maintained in good condition.



elop cur eep, al 4

xi-th of

be tly to

have ese T) onic rmine t yet

rely

Vessels that operate at temperatures above the dew point of water are less vulner-able to external corrosion, but the possibility should not be ignored. Insulated vessels that operate at low temperatures between the dew point and freezing point of water may be especially vulnerable to external corrosion.

Visual examination (VT) is the most appropriate procedure for detecting external corrosion. However, it is necessary to remove insulation and weatherproofing, at least at selected locations where the external corrosion is most likely to occur. External corrosion of the vessel supports, including anchor bolts, should not be overlooked.

733 Thermal AgingDeterioration of a pressure vessel during normal operation by thermal aging of the material is not a serious concern for most pressure vessels. Some types of thermal aging that can cause deterioration of vessels in process plants are:

• Graphitization• Temper embrittlement• Creep embrittlement• Sensitization of austenitic stainless steels

Unfortunately, NDE cannot detect these types of thermal aging until cracks devas a consequence of the deterioration. Other forms of deterioration that can ocduring the long-term operation of a vessel at elevated temperatures, such as crhydrogen attack, and thermal fatigue are not included in this definition of thermaging. These causes of failure in pressure vessels are discussed in Sections 73through 739.

GraphitizationCarbon steel vessels that have been in operation at temperatures above appromately 800°F are subject to graphitization. Graphitization can reduce the strengcarbon steel plate below the minimum required by the ASME Code for the maximum allowable design stress, but this effect by itself is usually not a very serious concern. The heat affected zones of welds in the carbon steel plate cansusceptible to a much more damaging type of graphitization that can lead direccracking.

Graphitization can be detected by nondestructive examination only after cracksdeveloped. Visual examination (VT) of the I.D. and O.D. surfaces can reveal thcracks before they cause failure, but magnetic particle (MT) or dye-penetrant (Pexaminations could reveal the cracks at an earlier stage of development. Ultrasexamination (UT) using shear wave techniques can be used to detect and detethe depth of surface cracks, and to detect and size internal cracks that have nopropagated to the surface.

Temper EmbrittlementLow-alloy Cr - Mo steels are susceptible to temper embrittlement after long timeoperation at temperatures above approximately 700°F, which can result in seve



c-s-

ech-ed by

e

vice

h tress to

e to

ent, ) or

the to lu-ntra-

F can ack

st less isk

to the ially

ults in

c

reduced CV-impact toughness at temperatures up to 250°F or higher. Brittle frature is prevented by restricting startup and shutdown procedures to limit the presure to 20% of the design pressure at temperatures below 250°F.

Temper embrittlement cannot be detected by any nondestructive examination tnique. The severity of temper embrittlement that has occurred can be determinremoving “boat” samples from plates and weldments for CV-impact testing. However, this method would necessitate repair of the vessel and should only bperformed after consultation with pressure vessel and materials engineers.

Creep EmbrittlementLow-alloy 1¼ Cr - ½ Mo steel is susceptible to creep embrittlement when in serat temperatures above approximately 850°F. Weld heat affected zones usually exhibit the greatest degree of embrittlement. The material becomes highly notcsensitive, and stress-rupture cracks can develop at locations of relatively high sconcentration. The toes of nozzle welds and fillet welds for attachments appearbe the most likely locations for these cracks to occur.

In-service inspection can detect cracks attributable to creep embrittlement in timpermit safely scheduling repairs without unnecessarily interrupting production. Visual examination (VT) of the O.D. and I.D. surfaces of a pressure vessel is usually adequate for detecting and sizing cracks attributable to creep embrittlembecause these cracks most often originate at the surface. Magnetic particle (MTdye-penetrant (PT) examinations would provide greater sensitivity for detectingcracks. Ultrasonic examination (UT) using shear wave procedures can be useddetermine the depth of cracks originating at the surface and would also be invaable for detecting and sizing cracks that may have originated at a stress concetion associated with an internal flaw.

Sensitization of Austenitic Stainless SteelAustenitic stainless steels in service at temperatures above approximately 800°become sensitized to intergranular corrosion. This is similar to the knife-line-attreferred to in Section 731, but it is not directly related to welding. Intergranular corrosion attributable to sensitization is not considered a serious problem in moprocess environments. However, the use of stabilized grades of austenitic stainsteel (i.e., Types 321 and Type 347) for service at temperatures above 750°F inprocess environments that contain sulfur compounds significantly reduces the rof intergranular corrosion.

Dye-penetrant examination (PT) of austenitic stainless steel surfaces exposed process environment will usually reveal intergranular corrosion and is an especuseful testing tool for cladding and weld overlays. Ultrasonic examination (UT) cannot detect intergranular cracks in cladding or weld overlays and can be verydifficult to use for solid stainless steel vessels. The fine, multiple, and branchedmorphology of the intergranular cracks makes them poor reflectors of ultrasonicpulses. Furthermore, the coarse grain structure of austenitic stainless steels resconsiderable “noise” that additionally complicates interpretation of the ultrasonireflections. Specialized ultrasonic examination (UT) procedures using very low



res-

to be

ntra-

.

d by ed at

frequency ultrasonic transducers can sometimes be successfully applied to austen-itic stainless steel vessel components.

734 Fatigue

Mechanical FatigueMechanical fatigue, caused by cyclic stresses, is likely to occur when the vessel has experienced at least 400 stress (pressure) fluctuations that exceed 15% of the maximum allowable design stress for the material of construction (or design pres-sure for the vessel).

Cracks attributable to mechanical fatigue usually originate on the surface of a pres-sure vessel at locations of relatively high stress concentration. Nozzles and the toes of fillet welds are likely places for fatigue cracks to occur. Pad-reinforced nozzles and partial penetration welds are more susceptible to fatigue than are integrally rein-forced nozzles and full penetration welds. Fatigue cracks can also originate at the stress concentration associated with internal fabrication flaws.

Magnetic particle examination (MT) or dye-penetrant examination (PT) can be used to detect mechanical fatigue cracks that originate on the surface subjected to cyclic stresses. Both the O.D. and I.D. surfaces should be examined at locations of rela-tively high stress concentration. Ultrasonic examination (UT) using shear wave procedures can be used to determine the depth of fatigue cracks that originate at the surface. A properly planned in-service inspection program will allow the surface fatigue cracks to be detected before they propagate to a size that jeopardizes the integrity and reliability of the vessel. Such a program also enables scheduling of repairs to avoid any unnecessary interruption to production. UT can also be used to detect and size the more insidious fatigue cracks that originate at internal flaws, or at surfaces that are not accessible for MT or PT examination.

Thermal FatigueThermal fatigue is attributable to cyclic stresses similar to mechanical fatigue but differs in that the stresses are developed by temperature gradients (i.e., thermal stresses). Temperature gradients that exceed 50°F through the thickness of a psure vessel's shell, or that span a distance less than 2.5(Rt)1/2 (where R = radius of vessel component and t = thickness of shell) along the surface, are consideredsufficient to cause thermal fatigue.

Thermal fatigue cracks tend to occur at locations where the thermal gradient isgreatest, and, therefore, they may not develop at locations of high stress concetion, where mechanical fatigue cracks are most likely to occur. Thermal fatigue cracks almost always originate at the surface, and closely spaced cracks can develop. Because nozzle reinforcement pads and partial penetration welds canmagnify temperature gradients, they may be especially prone to thermal fatigue

The locations most susceptible to thermal fatigue are surfaces alternately wettecooler liquids and dried by hotter vapors, and nozzles where fluids are introducfluctuating rates at either higher or lower temperatures than the bulk contents.



Thermal fatigue cracks in pressure vessels occasionally can be detected by visual examination (VT), but the greater sensitivity of magnetic particle examination (MT) or dye-penetrant examination (PT) is superior. Ultrasonic examination (UT) using a shear wave technique can be used to determine the depth of thermal fatigue cracks, or to detect and size those that might originate at a surface that is not accessible for MT or PT examination.

Corrosion FatigueCorrosion fatigue is the nucleation of fatigue cracks at the stress concentrations associated with corrosion pits and/or the acceleration of fatigue crack growth by the simultaneous occurrence of corrosion in the crack. It is related to mechanical fatigue in that a cyclic stress is necessary, but corrosion fatigue failures can occur more rapidly due to the interaction with a corrosive environment. Residual stresses can also play a very significant role in the nucleation of corrosion fatigue cracks and the acceleration of their growth rates.

Deaerator vessels can be highly susceptible to corrosion fatigue related to oxygen contamination of the boiler feedwater, especially at low pH levels. The cracking has been observed predominantly in welds and heat affected zones of vessels that have not received a postweld heat treatment, due to the relatively high residual stresses at these locations.

Corrosion fatigue cracks in deaerator vessels can be detected by magnetic particle examination (MT) of the I.D. surface of the vessel. The surface should be lightly ground to assure adequate sensitivity for crack detection. Ultrasonic examination (UT) from the O.D. surface using shear wave techniques can be used to determine the depth of the cracks detected, and thus evaluate if the vessel can be economically repaired or must be replaced.

735 Stress Corrosion (Environmental) CrackingStress corrosion cracking can be a very serious form of deterioration that can result in the catastrophic failure of a pressure vessel. It is highly dependent on the vessel material, as well as on process environment and operating conditions. Proper mate-rials selection and fabrication procedures effectively prevent stress corrosion cracking, but the possibility of its occurrence in severe process environments should not be ignored.

The major types of stress corrosion cracking are:

1. H2S stress cracking of carbon steel and low-alloy Cr - Mo steels.

2. Chloride stress corrosion cracking of austenitic stainless steels.

3. Ammonia stress corrosion cracking of carbon steels.

4. Caustic embrittlement of carbon steel.

Refer to the Corrosion Prevention and Metallurgy Manual for detailed discussions of these phenomena and procedures for preventing their occurrence.



An in-service inspection program should include nondestructive examination proce-dures that can detect stress corrosion cracks, regardless of the preventive measures taken. Stress corrosion cracking can be caused by process environments that do not normally contain, but are occasionally contaminated by, the chemical compounds that cause the phenomenon.

Stress corrosion cracks most often originate at the surface exposed to the process environment. Welds and weld heat affected zones can be especially susceptible. However, some types of stress corrosion cracks caused by process environments containing H2S can originate below the surface.

Visual examination (VT) is not reliable for detecting stress corrosion cracks in a pressure vessel, because the cracks tend to be very fine. However, magnetic particle examination (MT) or dye-penetrant examination (PT) are usually reliable. Frequently, MT or PT examinations will reveal a great number of very closely spaced cracks. Many of these cracks can be superficial and not have a significant effect on the integrity and reliability of the vessel unless they grow to a larger size. Ultrasonic examination (UT) using shear wave procedures can be used to determine the depth of the cracks to evaluate if they jeopardize the vessel. However, the typical fine, closely spaced, and branched morphology of stress corrosion cracks makes them very difficult to reliably detect and accurately size by UT examination. Specialized UT techniques are usually required to obtain satisfactory results. UT must be used to detect any stress corrosion cracks that originate below the surface, and it may be used to detect those that originate at a surface that is not accessible for MT or PT examination.

736 Internal ErosionInternal erosion can occur when high velocity process streams come in contact with the vessel shell. Entrainment of solid particles in the fluid stream and direct impingement of the stream on the shell can greatly increase the severity of erosion. Concurrent corrosion can also greatly increase the severity of erosion, due to the continuous removal of corrosion scales.

Wear plates are frequently provided to protect pressure vessel shells, especially at locations where high velocity process streams enter and impinge on the shell. The wear plates are usually fillet welded to the vessel shell and can be made from a high-alloy material that has greater erosion resistance than the shell.

Visual examination (VT) is adequate to detect internal erosion of a pressure vessel shell. Ultrasonic examination (UT) using a longitudinal wave procedure can be used to determine the remaining shell thickness in eroded areas, or to monitor erosion from the outside surface.

737 Hydrogen AttackHydrogen attack can occur in carbon steel and low-alloy Cr - Mo steels at elevated temperatures in process environments that contain a relatively high partial pressure of hydrogen (see Section 500). Hydrogen attack can be very serious, both reducing



in oten-

d d.

a ly be

ore d to at

a-n of

ts that n.

e s cro-aga-he

the

. A ll e d in to

the strength of the material and simultaneously causing microfissures to develop that eventually grow until failure occurs. Weld heat affected zones can be especially susceptible, and the attack can be more severe at locations of relatively high stress, such as nozzles.

Materials selected for construction of a vessel should not be susceptible to hydrogen attack at the operating temperature and pressure (see Section 500). Safe limits for carbon and low-alloy steels are usually satisfactorily predicted by the Nelson Curves in API 941, with the exception of C-½ Mo low-alloy steel, which is not recom-mended for preventing hydrogen attack. (API 941 is included in Volume I of theCorrosion Prevention and Metallurgy Manual). Nevertheless, in-service inspectionprograms should be established to assure that hydrogen attack does not occurvessels that operate at high temperatures and high pressures, because of the ptially serious consequences of the failure of high pressure equipment.

Highly specialized ultrasonic examination (UT) techniques have been developethat can detect hydrogen attack before the microfissures grow to a size that cancause failure, but the applicability to weld heat affected zones is presently limiteRemoving “boat” samples for metallurgical study can reveal hydrogen attack atvery early stage, but doing so will necessitate repair of the vessel and should onperformed after consultation with pressure vessel and materials engineers. Beffailure occurs, ultrasonic examination using shear wave procedures can be usedetect cracks that develop from the growth of microfissures, including cracks thmay appear in weld heat affected zones. However, microfissures can progress through the entire wall thickness of a heat affected zone before cracks can be detected by conventional ultrasonic shear wave techniques. Focused beam ultrsonic shear wave techniques are being developed that may enable the detectiomicrofissures in heat affected zones.

738 Hydrogen BlisteringHydrogen blistering of pressure vessel shells can occur in process environmencause hydrogen to diffuse through and “charge” the shell material with hydrogeProcess streams that contain acids or a relatively high partial pressure of hydrogen at high temperatures are most likely to cause blistering. Some of thhydrogen diffusing through the vessel shell is “trapped” at nonmetallic inclusion(or stringers) in the shell material. The hydrogen pressure builds up in these miscopic traps, causing them to propagate and link together to form blisters. Proption of the microscopic traps and growth of the blisters follow the major axis of tstringers, which lies in the plane of the rolled plate. Cracks can also develop, however, in the blisters that propagate towards the surface of the vessel, whenhydrogen pressure inside the blister becomes high enough to cause swelling.

Visual examination (VT) can be used to detect blisters in pressure vessel shellsflashlight beam directed along the surface of the shell can aid in observing smablisters. Ultrasonic examination (UT) using a longitudinal wave technique can bused to determine the remaining sound shell thickness if cracks have developethe blisters. Care should be exercised when making an ultrasonic examination



ck er-

t the n

sel.

signed and t in

weld ture

no

fore ver, e

he ere

sel n to

ct and d can ir or

distinguish between possible laminations and inclusions at midwall of the plate, and actual blisters.

739 Creep and Stress-RuptureOnly a small number of pressure vessels in process plants operate at temperatures high enough for creep to occur during normal operation. Creep is the continuous plastic deformation of a material under a constant stress and at high temperatures (approximately 700°F for carbon steels, 850°F for low-alloy Cr - Mo steels, and900°F for austenitic stainless steels) and ultimately results in stress-rupture cranucleation. It is a time-dependent process that implies a finite life for vessels opating in the creep range (i.e., at temperatures high enough for creep to occur amaximum allowable design stress). The ASME Code maximum allowable desigstress for vessels with design temperatures in the creep range is based on the 100,000-hour stress-rupture life of the material used for construction of the vesNormal experience shows, however, that vessels give satisfactory service for considerably longer times.

FCC reactors and Rheniformer reactors are examples of vessels that can be defor operation in the creep range. Some other vessels (including “cold wall” FCCRheniformer reactors) are refractory-lined to keep the vessel shell temperaturebelow the creep range. However, deterioration of the refractory lining can resul“hot spots” on the vessel shell, which can subject those overheated locations tocreep and the formation of stress-rupture cracks.

Stress-rupture cracks in pressure vessels operating in the creep range usually develop first at locations of relatively high stress, such as nozzles. Weldments (metals and heat affected zones) also tend to have somewhat shorter stress-ruplife than plates and forgings. Stress-rupture cracks generally originate near the surface of the vessel shell, but it is not unusual for them to initiate internally withdetectable indication on the surface until the latter stages of crack propagation.Internal initiation of stress-rupture cracks is more common for relatively thick (1-inch) vessel shell components.

Creep must cause some plastic deformation of the shell of a pressure vessel bestress-rupture cracks develop and failure occurs. The plastic deformation, howecan be very difficult to detect before the stress-rupture cracks cause failure andshould not be relied on to give sufficient warning of approaching failure. Becausthe plastic deformation can be highly localized in regions of high stress where tstress-rupture cracks develop, it may not cause any noticeable distortion of thevessel. Observable “bulging” of overheated areas may be evident, however, whhot spots created by the deterioration of the refractory lining cause creep.

Stress-rupture cracks can usually be observed on the surface of a pressure vesbefore failure occurs, but visual examination (VT) should not always be relied ogive an adequate indication of deterioration.

Ultrasonic examination (UT) using shear wave procedures can be used to detesize stress-rupture cracks. Stress-rupture cracks propagate relatively slowly anusually be detected at an early stage in their development, in time to allow repa



k-

uces

oded y the ile

ld be n -

e

replacement of deteriorated components of the vessel without unnecessarily inter-rupting production. UT examinations should concentrate on areas of relatively high stress, such as nozzle welds, head-to-shell welds, and longitudinal seam welds. Any hot spots that develop should receive intensive examination.

740 Analysis of In-Service Inspection DataThe information obtained from the inspection of a pressure vessel must be inter-preted and evaluated to determine if the vessel has sufficient integrity for continued service, or if repairs or replacement are required. The judgment of experienced pres-sure vessel or materials engineers may be required to properly determine the dispo-sition of a vessel after in-service inspection.

A very important concept to understand is that detected flaws do not necessarily have to meet the ASME Code acceptance standards for the vessel to be suitable for continued service without repair. The ASME Code acceptance criteria have been established to assure owner/users that vessels receiving the Code symbol have been manufactured to a high level of quality that can be obtained with good workman-ship. These acceptance criteria have no direct relationship to the requirements for safe operation. This is because some flaws occurring in service, though larger than those acceptable to ASME Code at time of vessel manufacture, are too small to cause vessel failure at its design temperature and pressure. Because of the various forms of deterioration that may occur during the operation of a vessel, in-service inspection can detect flaws that the ASME Code would not accept, but this does not always mean that the vessel has to be repaired or replaced.

741 Internal CorrosionCorrosion can reduce the thickness of the shell of a pressure vessel below the minimum thickness required by the ASME Code. The corrosion allowance speci-fied for the vessel when it was manufactured may not always be adequate for the actual corrosion rates occurring during operation, or the vessel may have been retained in service for longer than its initially anticipated life. API 510, “PressureVessel Inspection Code” gives methods for evaluating the reduction in shell thicness by corrosion.

General WastageGeneral wastage of a pressure vessel shell over a relatively large area, that redthe thickness of the shell below the minimum thickness required by the ASME Code, can result in failure by rupture. The primary membrane stress in the corrarea will be increased above the maximum allowable design stress permitted bASME Code, and failure will occur when the increased stress exceeds the tensstrength of the material.

The weld-joint efficiency factor used for the design of the pressure vessel shouused when calculating the minimum thickness required for the vessel shell wheweld joints are within the corroded area. In addition, the minimum required thickness calculations should include wind and earthquake loadings according to th



he ode on he

rate s-ded

ave a ick-n

e

lied of the

the

ick-not nt

-f the

t any y

ure m ving tress

same criteria used for the design of the vessel (see Section 400). It may be possible to exclude wind and earthquake loads from the minimum thickness requirement, depending on the location and geometry of the corrosion, but this should be discussed with a pressure vessel or structural engineer.

API 510 permits “averaging” the remaining wall thickness in corroded areas. If taverage thickness is less than the minimum thickness required by the ASME Cor can be expected to be reduced below this thickness by the prevailing corrosirate before the next scheduled inspection, either the shell must be restored to tminimum required thickness plus a corrosion allowance by weld build-up (see Section 800), or the vessel must be replaced. An acceptable alternative is to rethe vessel (Section 800). Rerating reduces the maximum allowable working presure so the maximum allowable design stress will not be exceeded in the corroareas with reduced wall thickness.

Localized CorrosionAreas of localized corrosion or pitting in pressure vessels can be permitted to hremaining wall thickness that is less than the ASME Code minimum required thness, within certain strictly defined limits. Local primary membrane stresses caexceed the maximum allowable design stress for the material of construction without significant risk of rupture, because of the “reinforcement” provided by thsurrounding material.

The weld joint efficiency factor used for the design of a vessel need not be appwhen calculating the minimum required thickness of the shell for the evaluationlocalized corrosion or pits that are farther from the edge of any weld than twice thickness of the shell. Weld joint efficiency factors are employed by the ASME Code to compensate for flaws that may exist in welds that do not receive full radiographic (RT) examination. However, flaws in the weld metal do not reducestrength of the plate and forging materials used for construction of the vessel.

Wind and earthquake loadings need not be included in the minimum required thness calculations, because highly localized reductions in shell thickness would be expected to significantly affect the structural integrity of the vessel in the eveof high winds or earthquakes.

API 510 limits the total area of localized corrosion or pits within any 8-inch diameter circle to 7 square-inches maximum. Furthermore, the remaining thickness ovessel shell may not be less than one-half of the minimum required thickness apoint, and the sum of pit dimensions along any straight line within the circle manot exceed 2 inches.

Appendix 4 of ASME Code, Section VIII, Division 2, can also be used (for Divi-sion 1 or Division 2 vessels) to evaluate localized corrosion or pitting in a pressvessel. Local primary membranes are permitted to reach 1.5 times the maximuallowable design of the material used for construction of the vessel. The area haa local primary membrane stress that exceeds the maximum allowable design smay not exceed (Rt)1/2 (where R = the radius of the vessel shell, and t = the shell thickness); and individual areas of local high stress may not becloser to one another than 2.5(Rt)1/2. These criteria can provide for the acceptance



t” t

in-r

g and gi-

d as f the ion

anner and essel

ot any ra- the

of somewhat larger areas of localized corrosion than would be permitted by API 510, especially for larger diameter vessels with relatively thick shells.

Areas of local corrosion or pitting that do not meet the requirements of API 510 or Appendix 4 of the ASME Code, Section VIII, Division 2, must be repaired by weld buildup, or by removal of the corroded area and insertion of a butt-patch (see Section 800).

Note that highly localized corrosion and the growth of pits can progress at consider-ably faster rates than predicted by the corrosion rates for general wastage. There-fore, returning vessels that have exhibited localized corrosion or pitting to service may involve risk of failure unless measures are taken to stop the attack (see the Corrosion Prevention and Metallurgy Manual).

Claddings and Weld OverlaysVisual indications that deteriorating claddings or weld overlays are resulting in corrosion of the pressure vessel shell must be further investigated. Hammer testing can provide a useful indication of the integrity of the lining. A “solid” or “resonansound indicates that the lining has not separated from the vessel shell, and thacracks have not developed. On the other hand, a “hollow” or “tinny” sound indi-cates that the liner has disbonded or that cracks have developed. Bulging or wrkling of the lining may also occur. However, indication of deterioration of the linedoes not necessarily indicate that the vessel shell is corroding. Corrosion of thevessel shell can be determined by removal of a section of the deteriorated lininvisually examining (VT) the shell, or by ultrasonic examination (UT) using a lontudinal wave procedure from the O.D. surface of the vessel to determine the remaining wall thickness.

Corrosion of a shell resulting from deterioration of the lining should be evaluatediscussed above. Deterioration of the lining that has not resulted in corrosion oshell need not be repaired, unless continued deterioration may to lead to corrosof the shell before the next scheduled inspection, or the integrity of an internal attachment may be compromised.

742 External CorrosionExternal corrosion of a pressure vessel shell should be evaluated in the same mas evaluating internal corrosion. Severe corrosion of pressure vessel supports anchor bolts may require evaluation by a structural engineer, especially if the vcan be subjected to high wind or earthquake loads.

743 Thermal AgingDeterioration of a pressure vessel due to thermal aging of the shell material is nusually repairable, because the thermal aging processes are not reversible for practical considerations. However, the manifestations of thermal aging (usuallycracking) can usually be repaired as an interim remedy to permit continued opetion until the vessel can be replaced. Alternatively, it may be possible to analyze



acceptability of the crack for continued operation (see Section 750) until the replacement vessel is obtained.

744 FatigueMechanical and thermal fatigue cracks can be repaired by first removing by grinding, followed by weld build-up of the ground area to restore it to the minimum required thickness (see Section 800). The ground area should receive a magnetic particle examination (MT) or dye-penetrant examination (MT) to make certain that all of the fatigue crack has been removed. Furthermore, the remaining shell thick-ness in the ground area can be determined by ultrasonic examination (UT) from the opposite surface using a longitudinal wave procedure. Weld build-up of the ground area may not be necessary if, similar to local corrosion or pitting, the local primary membrane stress due to the reduced shell thickness meets the criteria in Appendix 4 of ASME Code, Section VIII, Division 2 (see Section 741).

Alternatively, the depth of the fatigue can be determined by ultrasonic examination (UT) using a shear wave procedure, and a fitness-for-service analysis can be made (see Section 750) to determine if the crack jeopardizes the integrity and reliability of the vessel. This may allow postponing repair until a shutdown can be scheduled or until replacement materials and parts can be obtained.

Whenever mechanical or thermal fatigue cracks are detected in a pressure vessel, the cause of the fatigue must be determined and eliminated by changes in design or operation. Unless this is done, it is likely that fatigue cracks will reappear in the same locations after the repairs have been made. Regardless of the actions taken to prevent recurrence of the fatigue cracks, all locations where they were detected should be re-examined during the next scheduled inspection. Postweld heat treat-ment can prevent the recurrence of corrosion fatigue cracks in deaerator vessels.

745 Stress Corrosion (Environmental) CrackingStress-corrosion cracks in a pressure vessel can be repaired in a manner similar to fatigue cracks (see Section 744). However, the repairs to the vessel can be very extensive, because large numbers of cracks can occur in susceptible vessels.

The depth of stress-corrosion cracks should be determined by ultrasonic examina-tion (UT) using shear wave procedures before repairs begin. The cracks may be superficial and possibly dormant, in which case immediate repairs may not be necessary if the vessel satisfies fitness-for-service criteria (see Section 750). Deter-mining the crack depth by UT examination may be very difficult due to the form and structure of the crack (see Section 736). Very specialized shear wave proce-dures may be required. If the depth of the cracks cannot be determined by UT exam-ination, grinding to the minimum depth for their complete removal at selected locations can be substituted. Use of this method could also be considered to verify the accuracy of UT examination depth measurements.

It is essential to reexamine during the next scheduled inspection any stress-corrosion cracks that are allowed to remain in a pressure vessel. A pressure vessel or



materials engineer should be consulted to determine the maximum interval advis-able before the next examination.

746 Internal ErosionInternal erosion of a pressure vessel shell can be evaluated in the same manner as localized corrosion (see Section 741). Wear plates should be installed on a pressure vessel shell wherever internal erosion is detected, or existing ones should be repaired or replaced with a more protective design.

747 Hydrogen AttackIt is not acceptable to repair cracks attributable to the terminal stage of hydrogen attack (see Section 738) by local grinding and weld build-up. Cracks are likely to develop rapidly adjacent to the repair, in the material that had sustained a lesser degree of hydrogen attack. Furthermore, fitness-for-service analysis (see Section 750) is not applicable to cracks attributable to hydrogen attack. Therefore, the component of the vessel exhibiting the cracks (or the entire vessel) must be replaced before the vessel returns to service.

It may be possible to retain in service for a limited duration a vessel that has exhib-ited the initial stages of hydrogen attack detected by specialized ultrasonic examina-tion (UT) or metallurgical study. However, this must be discussed with a pressure vessel or materials engineer before making a decision. If the decision is to return the vessel to service, it is necessary to schedule re-examinations at recommended inter-vals to monitor the progression of the hydrogen attack. It must be recognized that the vessel cannot be retained in service indefinitely, because the attack will continue to progress. Therefore, it is advisable to schedule replacement at an early opportu-nity that will not unnecessarily interrupt production.

All vessels, or components of vessels, that are replaced due to hydrogen attack should be redesigned with a material that is more resistant to hydrogen attack.

748 Hydrogen BlistersHydrogen blisters generally do not significantly compromise the integrity or reli-ability of a pressure vessel. The separation causing the blister to appear is parallel to primary membrane stress, and therefore, does not reduce the load bearing strength of the shell.

Drilling a small vent hole into the blister will relieve the hydrogen pressure that builds up inside the blister (see Section 738) and prevent it from swelling enough that cracks develop and propagate towards the surface. Only if cracks have already propagated to the surface will it be necessary to evaluate the blister further. Cracked blisters will effectively reduce the thickness of the vessel shell. The remaining sound shell thickness can be determined by ultrasonic examination (UT) using a longitudinal wave procedure from the uncracked side of the shell. The remaining shell thickness can then be evaluated (see Section 750) for repair, replacement, or rerating for a lower pressure.



749 Creep and Stress-RuptureStress-rupture cracks can be repaired by grinding for removal and then restoring the ground area to the minimum required thickness with weld build-up, similar to that for fatigue cracks (see Section 744). However, this method may not always serve as a permanent repair, because the shell material adjacent to the repair might have undergone sufficient creep for additional stress-rupture cracks to develop after a relatively short period of additional operation. Therefore, it may be more realistic to plan for replacement of the components of the vessel exhibiting the cracks. It is unlikely that all components of a vessel will begin to develop stress-rupture cracks at the same time, and, therefore, it will not usually be necessary to replace the entire vessel. Consult a specialist for guidance.

Stress-rupture cracks tend to propagate relatively slowly, and if the stress-rupture cracks in a pressure vessel are relatively small when they are first detected, it may be acceptable to return the vessel to service until the replacement materials and components are obtained. Fitness-for-service analysis (see Section 750) can be used to help evaluate if the vessel has sufficient integrity and reliability for limited continued service.

750 Fitness-for-Service Analysis

751 IntroductionMany of the forms of deterioration that occur in pressure vessels are characterized by the formation of cracks, at least in their terminal stages before failure actually occurs. Analytical methods are available to evaluate the effect of these cracks on the integrity and reliability of the vessel if it is returned to service. These methods are commonly referred to as fitness-for-service analyses, because they can predict the effect that flaws will have on the performance of a vessel under actual service conditions. Employing fitness-for-service analyses can have significant benefit by:

1. Avoiding the expense of unnecessary repairs.

2. Allowing the vessel to be returned to service until repairs or replacement can be scheduled without interrupting production.

3. Planning in-service inspection programs that will prevent the failure of vessels during operation resulting in unscheduled shutdowns.

Guidelines for performing a fitness-for-service analysis of a pressure vessel are given in this section.

The inspection of pressure vessels during a shutdown occasionally reveals the pres-ence of flaws that exceed the acceptance limits of ASME Code Section VIII and the Company specification that may have been used for the original design and construction. The flaws most commonly detected are corrosion and indications of cracks. Detection of flaws (including cracks) that exceed the ASME Code accep-tance limits does not necessarily indicate that a pressure vessel is unfit for continued operation within its original design conditions. Fitness-for-service



cog-ity of of a ely,

e are re in nal-ME to as

re-cal ld-

.

n in el.

e e cks vice, re lure

analyses can determine if pressure vessels with flaws developed during operation can be returned to service without significant risk of failure.

An experienced pressure vessel engineer should be consulted for making a fitness-for-service analysis. This engineer should have the following qualifications:

1. Appropriate experience concerning the design, operation, inspection, and repair of pressure vessels.

2. Familiarity with the effects that the process environment can have on the behavior of the materials of construction.

3. Knowledge of fracture mechanics.

752 Background

CorrosionGeneral or localized corrosion can reduce the wall thickness below the minimum required by the Code design rules and equations for the design pressure and temper-ature, but it usually does not present a serious maintenance problem. Both API 510, “Pressure Vessel Inspection Code,” and the National Board Inspection Code renize that areas of localized corrosion do not necessarily compromise the integra pressure vessel, and they provide procedures for evaluating the acceptabilitylocal area with a thickness below the minimum required by the Code. AlternativAppendix 4 of ASME Code, Section VIII, Division 2, is generally accepted as a means to evaluate the effects of local corrosion. Stresses at the design pressurallowed to reach 1.5 times the allowable design stress for the design temperatua local area (defined in Appendix 4). These are, in effect, “fitness-for-service” ayses that permit the continued use of pressure vessels that would not meet ASCode requirements for new construction, although they are not usually referredsuch.

Local areas too deeply corroded to meet the reduced minimum thickness requiment, or generally corroded areas that are too large to meet the criteria for a lodeviation from Code thickness requirements, are routinely repaired by weld buiup or replacement of the component.

Indications of CracksIndications of cracks are usually of considerably greater concern than corrosionThey predominantly occur at weld joints and may propagate during operation. However, cracks that exceed the acceptance limits for radiographic examinatiothe ASME Code do not necessarily compromise the integrity of a pressure vess

The acceptance limits for linear indications (including cracks) in the ASME Codare “workmanship” criteria used to establish quality standards that should not bexceeded with good fabrication shop practices. It should not be inferred that craexceeding these limits, detected after a pressure vessel has been placed in seralways jeopardize reliable operation and require immediate repair. Flaws that alarger than the acceptance limits in the ASME Code will not always result in fai



d e l size

the or rate ack ed

mper-era-

e sent a

or ure

ed

than ient ciated

during startup, nor will they necessarily propagate so rapidly at the operating pres-sure and temperature that they cause failure before the next planned shutdown.

Fracture MechanicsTechnology has been developed, and is still evolving, that can predict if a crack will propagate as a fast fracture at the applied stress resulting from the internal oper-ating pressure, or if it will grow in a fatigue mode when subjected to repeated pres-sure or temperature cycles. This technology is based on the principles of fracture mechanics that relate crack propagation to the fracture toughness of the material, crack size, and applied stress. The major concepts involved are the following:

1. Fracture will occur when a critical stress intensity is exceeded at a crack tip. This critical stress intensity depends on the toughness of the material but is independent of the crack size and the applied stress.

2. Slow growth of a crack can occur below the critical stress intensity by a fatigue mode under a cyclic stress. The rate of crack growth by fatigue depends on the cyclic change in stress intensity at the flaw tip, not on crack size or applied stress.

3. The stress intensity at the tip of a crack increases with the size of the crack at a constant nominal applied stress but is not dependent on the toughness or fatigue properties of the material.

Fitness-for-Service Analysis of Cracks or Other FlawsThe concepts of fracture mechanics can be applied to determine if a pressure vessel containing a crack larger than permitted by the ASME Code for new construction is “fit-for-service” at its intended operating conditions. A critical crack size, requirefor failure of the vessel, can be determined for the material of construction at thvessel's operating pressure (i.e., applied stress). Cracks smaller than this criticawill not cause a pressure vessel to fail when it is returned to service. However, crack could grow by fatigue during operation if the vessel is subject to pressuretemperature fluctuations. Under these circumstances, the fatigue crack growth for the stress intensity at the crack tip must be calculated, to determine if the crwill grow to a critical size to cause failure of the vessel prior to the next schedulshutdown when another inspection can be made.

The toughness of carbon and low-alloy pressure vessel steel depends on the teature. Therefore, separate analyses may have to be made for the ambient tempture startup and elevated temperature operating conditions. Crack sizes that aracceptable for operation of a pressure vessel at elevated temperatures may prerisk of failure during startup at ambient temperature. In this situation, it may be necessary to restrict the pressure during startup until the vessel is warmed up, increase the minimum pressurizing temperature (MPT). The restriction on presswould also apply during a shutdown, in that the pressure may have to be reducbefore the vessel is allowed to cool to ambient temperature.

Hydrotest of a pressure vessel could present a much more serious risk of failurestartup at ambient temperature. Although the toughness of the material at ambtemperature is the same for both hydrotest and startup, the applied stress asso



c-ce by the or

cited or.

rack

roce- more y the ting a res- of f crit-

t

hould f the of

with the hydrotest pressure is appreciably higher than that developed at the oper-ating pressure. Therefore, a crack size that is entirely acceptable for both startup and operation could cause a pressure vessel to fail during hydrotest. In other words, a hydrotest is not proof that a fitness-for-service analysis has been performed satisfactorily and should not be conducted with a pressure vessel that contains a crack, unless a separate analysis has been made for the hydrotest conditions.

Neither API 510 nor the National Board Inspection Code provides for the fitness-for-service analysis of a pressure vessel containing a crack, but they do not explic-itly prohibit making such an analysis. ASME Code, Section XI, “Inservice Inspetion of Nuclear Power Plant Components,” does provide for the fitness-for-servianalysis of ASME Code, Section III, pressure vessels, and has been accepted Nuclear Regulatory Commission (NRC). The methodology given in Section XI fmaking an analysis has been adapted for ASME Code, Section VIII, pressure vessels in this section. Therefore, the acceptance of Section XI by NRC can befor precedence if it is necessary to obtain the approval of an authorized inspect

753 Information Required for Fitness-for-Service AnalysisIt is essential to have the following information to perform properly a fitness-for-service analysis:

1. Accurate sizing data for cracks (or other flaws).

2. Toughness and fatigue data for the material of construction at ambient andoperating temperatures.

3. Applied stress developed by the operating pressure at the location of the cor other flaw.

Cracks or other flaws can be accurately sized by ultrasonic testing (UT), but theproper procedures must be used by experienced technicians. Shear wave UT pdures are generally used for sizing, and time-based techniques usually produceaccurate results than amplitude-based techniques. It may be necessary to verifaccuracy of the procedures and the performance of the technicians by construc“mock-up” with implanted defects that simulates the size and geometry of the psure vessel component containing the crack. Ultrasonic imaging techniques areavailable that can provide a permanent visual record of the size and orientationthe crack, and they should be used to support the fitness-for-service analysis oical equipment.

The toughness of the material of construction can be estimated from CV-impacdata, but more accurate toughness data obtained directly from KIC, COD, or J-integral tests should be used whenever they are available. The fatigue data sshow the crack growth rate as a function of the cyclic stress intensity at the tip ogrowing crack. The toughness and fatigue data should also reflect the behaviorthe material in the process environment. Especially important are environmentswhere hydrogen charging can promote crack growth (including stress-corrosioncracking), and high operating temperature where creep can occur.



The applied stress developed by the operating pressure at the location of a crack or flaw must be accurately determined. The primary membrane and bending stresses used for design are usually sufficient for pressure vessel components with simple geometries. However, it may be necessary to perform a finite-element stress anal-ysis to determine discontinuity and secondary, i.e., thermal, stresses for locations that have a complex geometry.

754 Procedure for Fitness-for-Service Analysis

ScopeA fitness-for-service analysis can be made for any pressure vessel containing one or more flaws or cracks regardless of size, thickness, shape, or design conditions. Unavailability of appropriate data on materials toughness and crack propagation is likely to place the greatest limitation on the use of fitness-for-service analyses, espe-cially if the process environment has a significant effect upon the behavior of the material or if the operating temperature is high enough for creep to occur. These limitations should be gradually eliminated as the work of the API, MPC, and PVRC (referred to above) task groups proceeds.

Outline of ProcedureAn outline for making a fitness-for-service analysis is given below. Reference should be made to the ASME Code, Section XI, Appendix A, for details.

A. Determine Flaw Size and Orientation

1. Determine actual flaw size, configuration, and orientation using ultra-sonic examination procedures.

2. Resolve the flaw into a simple shape by completely circumscribing with an ellipse or circle.

a. Circumscribe multiple flaws with the same ellipse or circle if the distance separating them is less than the dimension of the largest flaw in the same direction as the separation is measured.

3. Project circumscribed flaw onto a plane perpendicular to the direction of the maximum principal stress.

a. Flaws closer to a surface than 0.4 times their maximum dimen-sion in a direction perpendicular to the surface are classified as surface flaws.

b. Flaws further from a surface than 0.4 times their maximum dimension in a direction perpendicular to the surface are classi-fied as internal flaws.



B. Determine Applied Stress at Location of Flaw

Applied stress at the location of the flaw should be resolved into membrane bending components. All forms of loading (i.e., internal pressure, disconti-nuity, and thermal) should be considered as follows:

1. Calculate maximum principal membrane stress at the operating pres-sure for cylindrical and spherical shapes using ASME Code design equations.

2. Calculate maximum principal membrane and bending stresses at nozzle opening using WRC Bulletins 107 and 297.

a. Consideration must be given to both internal pressure and signifi-cant external piping loads. Piping stress analysis may be neces-sary to determine piping loads.

3. Calculate principal membrane and bending stresses at internal or external attachments and vessel supports using WRC Bulletin 107.

4. Calculate discontinuity stresses at head-to-shell joints and transitions resulting from internal pressure using ASME Code, Section VIII, Division 2, Appendix 4, and combine with principal membrane and bending stresses.

5. Calculate thermal stresses at locations with significant temperature gradients, using ASME Code, Section VIII, Division 4, and combine with principal membrane and bending stresses.

6. Use finite element stress analysis to calculate all membrane and bending stresses when other methods will not provide satisfactory results.

C. Determine Stress Intensity at Flaw

1. Using the principal membrane and bending stresses calculated above, calculate the stress intensity at the flaw (KI) using the following equation:

(Eq. 700-1)

where:σm,

σb = membrane and bending stresses, psi, in accordance with B above.

a = minor half-diameter, in., of embedded flaw; flaw depth for surface flaw

Q = flaw shape parameter as determined from Figure 700-1 using (m + b)/ys and the flaw geometry

KI σmMm π( )0.5 aQ----

0.5σbMb π( )0.5 a

Q----

0.5+=



Mm = correction factor for membrane stress (see Figure 700-2 for subsurface flaws; Figure 700-3 for surface flaws)

Mb = correction factor for bending stress (see Figure 700-4 for subsur-face flaws; Figure 700-5 for surface flaws)

D. Determine Material Toughness and Crack Propagation Properties

1. Obtain appropriate KIC, COD, or J-integral data at ambient and oper-ating temperature by one of the following:

a. Testing of samples of the actual material used for construction. (This would provide the most accurate data but is rarely obtainable.)

Fig. 700-1 Shape Factors for Flaw Model



b. Published data for the “generic” material of construction. (The lower bound trend curve should be used.)

Fig. 700-2 Membrane Stress Correction Factor for Subsurface Flaws



Fig. 700-3 Membrane Stress Correction Factor for Surface Flaws



Fig. 700-4 Bending Stress Correction Factor for Subsurface Flaws



Fig. 700-5 Bending Stress Correction Factor for Surface Flaws



hey e.)

e s is

en-

. r

k of

nd tes be

ate-

re

-

a

c. Estimate from CV-impact data for the actual or “generic” mate-rial of construction. (Correlations have been made that can be used to predict fracture toughness from CV-impact data. However, there is considerable scatter in the correlations and tshould be used only if fracture toughness data are not availabl

The effect of the process environment, especially hydrogen charging upon fracture toughness, should be considered. Somdata are available to indicate the degree to which the toughnesreduced by hydrogen, and a more complete data base is beingobtained by industry task groups.

2. If the flaw is subjected to internal pressure or temperature fluctua-tions, obtain crack growth rate data (i.e., growth/cycle vs. stress intsity range/cycle) by one of the following:

a. Testing of samples of the actual material used for construction(This is not a common test performed with the material used foconstruction and cannot usually be obtained.)

b. Published data for the “generic” material of construction. (The lower bound trend curve should be used.)

The process environment can have a significant effect on cracgrowth rate and should be taken into consideration. The growthsurface cracks can be accelerated by corrosion fatigue, and hydrogen charging can accelerate the growth of both surface ainternal cracks. Little test data are available for crack growth rain simulated process environments, and estimates may have tomade based upon knowledge of the general behavior of the mrial in the environment.

E. Determine Acceptability of Flaw

1. When the flaw is not subjected to significant pressure or temperatucycles (greater than 20% of design pressure or 50°F), compare thestress intensity at the flaw (KI per C above) with the critical stress intensity for fracture initiation (KIC per D above) at ambient and operating temperatures.

a. The flaw is acceptable for operation if KI is less than KIC at the operating temperature. (It is recommended that the flaw size determined in this manner should be reduced by ½ to provide safety factor of 2 for acceptable crack size.)



The flaw is unacceptable for continued operation if KI exceeds KIC at the operating temperature. The flaw must be repaired, or the operating pressure downrated until KI is less than KIC.

b. The flaw is acceptable for startup and shutdown at full operating pressure if KI is less than KIC at ambient temperature (i.e., atmo-spheric temperature during startup or shutdown).

The flaw is unacceptable for startup and shutdown at full oper-ating pressure if KI exceeds KIC at ambient temperature. Pressure during startup and shutdown will have to be limited such that KI is lower than KIC, or the temperature will have to be increased before full operating pressure is attained.

2. If the flaw is subjected to significant pressure or temperature cycles, determine the extent of crack growth by fatigue until the next sched-uled shutdown.

a. Estimate the number of pressure or temperature cycles until the next scheduled shutdown. (It is recommended to double the number of anticipated cycles to obtain a safety factor of 2.)

b. Determine the extent of crack growth for the number of antici-pated cycles for the stress intensity at the flaw (KI per C) using the fatigue crack growth properties (per D). (It is recommended to double the crack growth/cycle to obtain a safety factor of 2 for fatigue crack growth rate.)

c. The flaw is acceptable for continued operation if it will not grow to a size such that KI exceeds KIC prior to the next scheduled shutdown. (It is not anticipated that the crack will grow to a size that will require repair at the next scheduled shutdown because of the dual safety factors recommended for anticipated cycles and crack growth rate. Therefore, this should be considered an indefi-nite postponement of repair.)

The flaw is unacceptable for continued operation if it will grow to a size such that KI exceeds KIC prior to the next scheduled shut-down. The crack should be repaired, or the operating pressure downrated such that the crack will not grow by fatigue to a size where KI exceeds KIC.



755 Required Inspection at the Next Scheduled ShutdownFlaws that are allowed to remain in a pressure vessel during continued operation based on a fitness-for-service analysis should be reinspected by ultrasonic examina-tion during the next scheduled shutdown.

1. If the flaw has grown, the fitness-for-service analysis should be repeated for the enlarged flaw size.

2. If no crack growth is detected, the need for subsequent reinspection should be evaluated by an experienced pressure vessel engineer.

760 Nondestructive Examination (NDE) TechniquesThe NDE techniques that are most useful for the in-service examination of pressure vessels are:

1. Visual Examination (VT), Section 761

2. Dye Penetrant Examination (PT), Section 762

3. Magnetic Particle Examination (MT), Section 763

4. Ultrasonic Examination (UT), Section 764

5. Radiographic Examination (RT), Section 765

6. Acoustic Emission Testing (AE), Section 766

Each of these NDE techniques employs a different phenomenon to interact with flaws in metal objects in a manner that can be observed with appropriate instru-ments for detection of the flaws. Therefore, each technique can detect some forms of deterioration in a pressure vessel during operation but is unsuitable for detecting other forms. Selecting the proper NDE procedures requires understanding of the attributes of each with regard to the forms of deterioration that may have occurred during operation.

Note that eddy current examinations have been used with mixed results in the past. This technique is being developed and may be valuable in the future.

761 Visual Examination (VT)VT should be employed to some extent for the in-service inspection of every pres-sure vessel, and, in this sense, it is the only universal NDE technique. Direct obser-vation identifies many forms of deterioration (such as corrosion, erosion, hydrogen blistering, and the occurrence of large surface cracks), and also can highlight loca-tions where other forms of deterioration may have occurred that will require exami-nation with other NDE techniques. Therefore, VT is usually the first NDE performed after a vessel is opened for an inspection, and it will give a useful impression of the general condition of the vessel. However, as important as VT is, it alone cannot verify the integrity and reliability of a vessel, because some forms of



deterioration cannot be seen (such as fine fatigue or stress-corrosion cracks, occur-rence and growth of internal cracks, and hydrogen attack).

It is usually necessary to clean pressure vessels internally to facilitate inspection. Steam or chemical cleaning is usually sufficient to remove hydrocarbon liquid films and sludge deposits that might remain inside after shutdown. If corrosion is severe and a relatively thick scale adheres to the vessel shell, hydroblasting or abrasive blasting may be necessary to evaluate the severity of the deterioration. Nozzles and shell surfaces obscured by internals should not be overlooked.

Inspectors frequently use hammers, picks, and scrapers to aid VT for determining the thickness of corrosion scales and the soundness of the shell remaining under the scale. Depth gages can be used to measure the depth of corrosion pits, and calipers are useful for measuring the inside diameter of nozzles to determine corrosion loss. However, whenever it appears that corrosion could have thinned the shell to less than the minimum required thickness, it is advisable to determine directly the remaining thickness with UT.

VT can be relied on for detecting hydrogen blisters and is very useful for appraising the integrity of corrosion protection claddings. However, UT should be used to determine the remaining shell thickness when blisters or deteriorated claddings are visually detected.

762 Dye-Penetrant Examination (PT)PT is used to detect flaws that are open to the surface of a workpiece. It is a rela-tively simple NDE procedure that requires a minimal investment. It is used prima-rily to detect cracks that have developed at, or have propagated to, the surface of the vessel shell.

Article 6 of ASME Code, Section V, gives the minimum requirements for a PT procedure for pressure vessels, and refers to ASTM Standard SE165 for additional details that should be considered when establishing a PT procedure.

Physical PrinciplesFigure 700-6 illustrates the physical principles of PT. A liquid (referred to as a pene-trant) is applied to the surface of the part being examined, and is drawn into cavities in the object by capillary action. The surface tension of the liquid draws it from the wet surface into the dry cavities. A dye (commonly red) is added to the liquid to make it clearly visible.

The surface is then wiped clean of the liquid, which causes a small quantity of the liquid that was drawn into the cavity to reemerge on the surface. Reemergence of the liquid onto the surface at this point is due to reverse capillary action, resulting from the now dry surface and wet cavity. Sufficient liquid may reemerge to be visible when the surface is wiped clean, but a developer is usually applied to the surface that greatly enhances the ability to detect the cavity.

The developer, a very fine powder, serves both to draw liquid penetrant out of the cavity, and to form an opaque layer on the surface that masks the potential



ge- be el-

the

nd e

mita-

to

t to

ace nd

confusing background of the workpiece surface. The developer is applied to the surface after it is wiped clean of the liquid penetrant. The powder forms a “sponlike” layer that draws more liquid out of the cavity by capillary action, which canthought of as blotting, to make it more visible. Furthermore, the color of the devoper (usually white) is chosen to contrast with the dye in the liquid to further enhance the visibility of the liquid reemerging from the cavity.

Dye-Penetrant Examination SystemsThe dye-penetrant examination system that is the most useful is referred to as solvent-removable system. This system employs an oil-base penetrant that is removed from the surface of the workpiece with a solvent. Both the penetrant asolvent can be sprayed onto the surface of the workpiece with aerosol cans. Thsurface of the workpiece must be accessible, but there is essentially no other lition with regard to size, location, and orientation of the workpiece.

Surface PreparationSurface preparation is very important. The surface must be both clean and dry enable the liquid penetrant to wet the surface completely, and to be drawn into flaws. This condition is generally achieved by washing the surface with a solvenremove oil, grease, sludge deposits, etc.

Light grinding to remove surface irregularities can be beneficial. Innocuous surfblemishes can make it difficult to distinguish between these harmless artifacts aflaws that have developed during service. Heavy grinding should be avoided,

Fig. 700-6 Physical Principles of Dye-Penetrant Examination (PT)



because surface deformation can cover over flaws and make it difficult for the pene-trant to enter.

Corrosion scales can block entry of the penetrant into the flaw and must be removed. Wire brushing or light abrading may be adequate for relatively thin scales, but abrasive blasting may be required for heavy scales. Corrosion scales can also fill the cavity and greatly reduce the quantity of liquid penetrant that can be drawn into the opening, which can make surface cracks very difficult to detect with PT.

Dye-PenetrantsThe liquid used for a dye-penetrant must both wet the surface of the workpiece and have a high surface tension to cause a sufficient quantity of the liquid to be drawn into the cavity. Viscosity is not a direct measure of the ability of a liquid to enter a cavity, but liquids with a high viscosity flow too slowly to be certain that they will fill cavities in a reasonable time to make them useful penetrants.

The oil-base liquids that are used as the penetrant for solvent-removable dye-pene-trant systems can be applied to a surface of the workpiece having any orientation, by spraying from an aerosol can. Sufficient dwell time must be allowed for the pene-trant to be drawn into surface flaws in the workpiece. ASME Code, Section V, requires a minimum dwell time of 10 minutes. The sensitivity for detecting very fine surface cracks can be increased by increasing the dwell time. Doubling the dwell time to a minimum of 20 minutes is considered to be a high sensitivity PT procedure, and should be used for tight cracks that can be difficult to detect, such as stress-corrosion cracks in austenitic stainless steel.

It is best to first wipe the excess penetrant from the surface with a dry cloth, followed by a second wiping with a clean cloth dampened with the solvent. The surface can also be flushed by direct application of the solvent to remove the pene-trant, but there is danger of inadvertently removing the penetrant from flaws before they are revealed by application of the developer. ASME Code, Section V, prohibits removal of the excess penetrant by flushing with solvent because of this risk. The excess penetrant must be completely removed from the surface before the devel-oper is applied.

DevelopersThe developer should be applied to the surface of the workpiece immediately after the excess liquid penetrant is removed from the surface. The developers used for solvent-removable dye-penetrant systems are generally suspensions of the devel-oper powder in a quick drying solvent, that can be sprayed onto the surface of a workpiece from an aerosol can. Quick drying of the solvent allows an essentially uniform layer of developer powder to be obtained on the surface without any running, even on vertical surfaces.

The developer should be applied in a layer thick enough to completely mask the surface, but if the layer of developer is too thick, the liquid penetrant drawn out of a flaw may not reach the top of the layer and the indication of a flaw can be obscured. The minimum development time required by ASME Code, Section V, is 7 minutes.



The dye-penetrant drawn out of a flaw by the developer should stain the developer layer a deep red. A light pink color may indicate that excessive cleaning inadvert-ently removed penetrant from the flaws, and that smaller flaws may not have been revealed. Therefore, when pale indications are obtained, the examination should be repeated with less vigorous cleaning.

Applications and LimitationsPT can be used to detect cracks that have occurred due to stress-corrosion (environ-mental) cracking, creep, mechanical or thermal fatigue, and hydrogen attack. However, it must be realized that PT will only detect flaws on the surface of the vessel shell. Surface cracks are often visually apparent upon close scrutiny after the PT has been performed, but this does not imply that the cracks would have been detected by VT alone if they were not previously located by PT.

PT does not give any indication of the depth of a surface crack, which is the primary characteristic of a crack that can affect the integrity and reliability of a vessel. The intensity of the dye stain revealed by the developer is an indication of the quantity of liquid penetrant that reemerges, but a deep crack that is tightly closed may exhibit a less intense stain than a shallow crack with a relatively wide opening. Corrosion scales that fill a deep crack would also reduce the intensity of the dye stain. In addi-tion, the spread of the dye stain in the developer can exaggerate the apparent surface length of a crack, especially for relatively short cracks. Therefore, it is usually necessary to determine the size of cracks detected by PT by other NDE procedures, such as ultrasonic examination (UT), to evaluate their significance, unless the PT indications are removed by grinding.

763 Magnetic Particle Examination (MT)MT can be used to detect surface and near surface flaws in ferromagnetic materials. MT cannot be used for austenitic stainless steel pressure vessels, because these materials cannot be magnetized.

MT’s applicability to inspection of pressure vessels is similar to PT, mainly in that it is most useful for the detection of surface flaws, but MT is generally considered to give greater sensitivity for the detection of fine cracks. It is somewhat more diffi-cult to apply than PT, and it requires considerably more investment in equipment.

Article 7 of ASME Code, Section V, gives the minimum requirements for an MT procedure for pressure vessels, and refers to ASTM Standard SE709 for additional details that should be considered when establishing an MT procedure.

Physical PrinciplesA bar of ferromagnetic material can be magnetized by placing it in contact with the north and south poles of a horseshoe magnet, as depicted in Figure 700-7a. North and south poles are created in the magnetized bar opposite to the poles of the horseshoe magnet, and magnetic lines of force flow through the bar from the south pole to the north pole in a straight line, as shown in Figure 700-7b.



the nti-lines e

nes of e may

A discontinuity (flaw) in the material will disrupt the magnetic lines of force flowing through the material, as shown in Figure 700-7c. New north and south poles are created at the discontinuity, and some of the magnetic lines of force “leak” from surface of the material to bridge the gap resulting from the discontinuity. A disconuity that is open to the surface results in a relatively large leakage of magnetic of force, whereas appreciably less leakage is caused by a discontinuity below thsurface. Deep discontinuities cause greater leakage than shallow ones, but the leakage produced decreases as the gap that must be bridged by the magnetic liforce widens. Discontinuities that are perpendicular to the magnetic lines of forccause the greatest leakage, whereas discontinuities parallel to the lines of forcenot produce any leakage.

Fig. 700-7 Physical Principles of Magnetic Particle Examination (MT)



ll

lines el-

rk-c

sitive at re s,

Magnetic particles applied to the surface of the material will be attracted by the magnetic lines of force leaking from the surface at a discontinuity and will adhere to the surface at this location, revealing the presence of the discontinuity.

Ferromagnetic materials can also be magnetized by the passage of an electric current. In these materials, an electric current creates magnetic lines of force that are at right angles to the flow of the current, as illustrated in Figure 700-8.

Since detection of a surface flaw does not depend on a liquid being drawn into the flaw, MT can detect flaws that a liquid penetrant cannot enter because they are too tight or are filled with corrosion scale.

Magnetic Particle Examination SystemsThe MT systems used for in-service inspection of pressure vessels employ an elec-tric current to magnetize the area of the vessel shell being examined. Either an elec-tromagnetic “yoke” or electric “prod” contacts can be used to magnetize the shewith an electric current, as illustrated in Figure 700-9.

The electromagnetic yoke functions as a horseshoe magnet, with the magneticof force created by an electric coil in the handle. Magnetic lines of force are devoped in the workpiece when the poles of the yoke functioning as a horseshoe magnet are brought in contact. The magnetic lines of force flow through the wopiece from one pole of the yoke to the other. The yoke does not pass an electricurrent through the workpiece to create magnetic lines of force.

The electric prod contacts do not function as electromagnets. They serve as poand negative electrodes that introduce an electric current into the workpiece, thflows between the contact points of the prods. Circular magnetic lines of force acreated in the workpiece that are concentric around the point of contact of eachprod. Prods offer more flexibility for MT of components with complex geometriesuch as nozzle welds, and are more frequently used than yokes.

Fig. 700-8 Creation of Magnetic Lines of Force by an Electric Current



It is very important to note that the direction of the magnetic lines produced by an electromagnetic yoke is different from that produced by electric prod contacts, as can be seen in Figure 700-7. Flaws are most detectable when they are oriented normal to the magnetic lines of force, because this orientation results in the greatest leakage of magnetic lines of force. Therefore, the yoke, or prods must be positioned correctly with respect to the orientation of the flaws to obtain the greatest sensi-tivity for detection. The greatest sensitivity with a yoke is obtained by positioning its poles normal to the orientation of the flaws, whereas positioning prods parallel to

Fig. 700-9 Electromagnetic Yoke and Electric Prod Contacts for Magnetizing Shell of a Pressure Vessel



e lec-o-n rrent

ent.

the orientation of the flaws provides the greatest sensitivity. If the flaws appear to have a random orientation, the same location should be examined twice with the positioning of the yoke or prods rotated 90 degrees. One example where two exami-nations should be made by rotating the yoke or prods is for the detection of both longitudinal and transverse cracks in a weldment.

Electric CurrentThe electric current for magnetizing the workpiece, with either an electromagnetic yoke or electric prod contacts, must be flowing to produce strong magnetic lines of force when the magnetic particles are applied to the surface. Some residual magne-tism is usually retained by the workpiece after the current is turned off, but it is too weak to properly attract the magnetic particles to a flaw.

When electric prod contacts are used, the current should be turned on after the prods are brought in contact with the workpiece, and turned off before the prods are removed from the workpiece to avoid arcing. Arcing can cause small heat affected zones to develop in the workpiece, similar to welding, which could make the vessel susceptible to failure by stress-corrosion cracking in some process environments (see Section 736). Arcing is not a concern when an electromagnetic yoke is used, because the yoke functions as an electromagnet and does not make electrical contact with the workpiece.

Either AC or DC electric current can be used for magnetizing the workpiece with both electromagnetic yokes and electric prod contacts, but the magnetic lines of force created by each type of current have different characteristics. The magnetic lines of force created by a DC current penetrate further below the surface of the workpiece than those created by an AC current. Therefore, DC currents should be used when subsurface flaws must be detected. However, AC currents provide some-what greater sensitivity for detecting flaws that are open to the surface, and are preferred when only surface flaws must be detected.

The electric current must create magnetic lines of force in the workpiece of suffi-cient strength for the magnetic particles applied to the surface to be attracted to the magnetic leakage occurring at flaws. When electrical prod contacts are used, SE706 recommends a magnetizing current of from 90 to 110 amperes/inch of prod spacing for workpiece thicknesses less than ¾ inch, and 110 to 125 amperes/inch of prod spacing for workpiece thicknesses of ¾ inch and thicker. It is undesirable for thprod spacing to exceed 8 inches. The criteria for the electric current when an etromagnetic yoke is used are different, because the yoke functions as an electrmagnet. The magnetizing strength of a yoke is determined by its lifting power. Aelectromagnetic yoke must be able to lift a 10-pound steel plate when an AC cuis used, or a 40-pound steel plate when a DC current is used.

Mobile electrical power sources are available that provide both AC and DC currThese power sources can be transported to the pressure vessel for in-service inspection.



Magnetic ParticlesThe magnetic particles that are applied to the surface of a workpiece when it is magnetized are of two types, dry and wet; and they are classified according to how they are carried to the workpiece. Dry particles are carried to the surface of the workpiece by air, whereas wet particles are carried by a liquid. Both dry and wet particles must have a high magnetic permeability so that they can be attracted by relatively low levels of magnetic leakage at flaws to assure a high sensitivity for detection. Furthermore, the particles must have suitable shape and size for adequate mobility on the surface of the workpiece to move to the locations of magnetic leakage at flaws.

Dry particles should be applied to the surface of the workpiece with low velocity air as a uniform cloud with a minimum of turbulence. This will allow the particles to be attracted by the magnetic leakage at flaws while they are suspended in air and their movement is not unduly influenced by air currents and gravity. Dry particles are especially useful for rough surfaces that inhibit the flow of a liquid, and they are somewhat more sensitive than wet particles for detecting subsurface flaws. Another advantage of dry particles is that they can be applied to surfaces that are too warm for liquids, and therefore can permit some pressure vessels to be examined during operation. The dry particles are usually colored yellow, red, or black to contrast with the workpiece surface. Dry particles are also available with fluorescent coatings that give them a very high visibility in ultraviolet (black) light.

Wet particles are suspended in light oil or water, and the slurries are generally applied to the surface of the workpiece by spraying. The concentration of suspended particles in the liquid must be high enough to give observable indications of fine flaws, but too high a concentration can result in a confusing background that obscures the indications of flaws. Frequent agitation is necessary to maintain a uniform concentration of particles in the slurry. The best way to determine proper concentration of particles in the slurry is to perform a test with a specimen that contains known discontinuities.

Wet particles are more sensitive than dry particles for detecting fine surface cracks. They generally have a fluorescent coating that makes them highly visible in ultravi-olet (black) light, and they will generally adhere to the surface of the workpiece after the liquid has evaporated.

Applications and Limitations of MTMT is preferred to PT for detecting surface cracks because of its greater sensitivity, especially when wet particles are used. Very fine cracks can be detected by MT that escape detection by PT. This makes MT especially useful for vessels that have been exposed to process environments that contain H2S under conditions that can cause stress-corrosion cracking.

Although MT can detect subsurface cracks that are not too far below the surface, do not depend on MT alone to detect internal cracks (see Section 730).

Other methods more suitable for detecting internal flaws, such as UT and RT, should be employed to supplement MT. This is especially true for relatively thick (approximately 1.5-inch thick and greater) vessel shell components. It is more likely



lid

aces quid, ves,

olid-r

mally

rti-

e

that cracks may originate internally in thick shells, and it will normally take longer for them to propagate to the surface where they can be detected by MT.

Similar to PT, MT does not give a reliable indication of the depth of a surface crack, which is the primary characteristic that affects the integrity and reliability of a pres-sure vessel. Therefore, the depth of cracks detected by MT should be determined by other NDE procedures (such as UT) to evaluate their significance, unless the MT indications are removed by grinding.

MT procedures can vary widely in reliability. They should be qualified for an in-service inspection of a pressure vessel by demonstrating their performance with sample flaws representative of the actual flaws that may have occurred in the vessel.

764 Ultrasonic Examination (UT)UT is a very versatile NDE method. A variety of techniques have been developed for using UT to detect most forms of deterioration that can occur during service. Both surface and internal flaws in a vessel shell can be detected from either the O.D. or I.D. surface, and internal corrosion loss can be determined without access to the inside of the vessel.

Of all NDE methods UT is the most able to determine flaw size with sufficient accuracy to evaluate the integrity and reliability of a vessel, and to make fitness-for-service analyses. However, the accuracy of the sizing data obtained is highly depen-dent on the procedure used, and the qualifications and experience of the technician performing the examination.

Article 5 of ASME Code, Section V, gives the minimum requirements for a UT procedure for the in-service examination of a pressure vessel. However, these minimum requirements may not be adequate for the reliable detection and accurate sizing of some types of flaws that can occur during service. Specialized UT techniques are discussed below that can greatly improve detection and sizing capabilities.

Physical PrinciplesA high-frequency sonic (referred to as “ultrasonic”) wave is introduced into a somaterial and is reflected by interfaces in the material. The reflected waves are analyzed to detect the existence of a flaw and determine its location. The interfthat reflect the ultrasonic waves can be between the solid material and a gas, lior other solids. Solid-gas interfaces tend to completely reflect the ultrasonic wawhereas solid-liquid and solid-solid interfaces usually only partially reflect the ultrasonic waves. Flaws in metal pressure vessel shells normally act as either sgas or solid-solid interfaces, and can, therefore, be either very good or very pooreflectors. The surfaces of the vessel shell are solid-gas interfaces that are norvery good reflectors.

Ultrasonic waves are mechanical waves that are created by displacement of pacles (atoms or molecules) in an elastic medium from their equilibrium positions.The displaced particles are restored to their equilibrium positions by interatomicforces acting between the particles. However, the interatomic forces also induc



displacement of adjacent particles that results in propagation of the wave through the medium. Two types of ultrasonic wave can be created, depending on the orienta-tion of particle displacement relative to the direction of wave propagation. Longitu-dinal waves are developed when the particles are displaced parallel to the direction of propagation, whereas transverse (or shear) waves result when particles are displaced perpendicular to the direction of propagation, as illustrated in Figure 700-10. Shear waves will generally propagate only in solids, because the inter-atomic forces in liquids and gases are too weak for perpendicular displacements to induce displacement of adjacent particles.

A longitudinal ultrasonic wave propagating through one medium in a direction that is perpendicular to the interface with a second medium (i.e., angle of incidence = 0 degrees from vertical), will be both reflected back into the first medium and transmitted into the second medium without change in direction as a longitudinal wave. However, if the angle of incidence is not 0 degrees, the wave will be reflected at an angle equal to the angle of incidence. The portion of the wave transmitted into the second medium will be refracted at the interface, and will split into both longitu-

Fig. 700-10 Types of Ultrasonic Waves



curs f the e ound as a er de

dinal and shear waves that propagate into the second medium in directions that differ from the angle of incidence, as shown in Figure 700-11.

A longitudinal wave that is not perpendicular to the interface is refracted when it propagates into the second medium, because of the difference in the velocity of sound between the two media. Splitting of the incident longitudinal beam into refracted longitudinal and shear waves is known as “mode conversion,” and ocbecause the displacement of particles at the interface parallel to the direction oincident longitudinal wave is not parallel to the refracted longitudinal wave. Thesine of the angles of refraction for the refracted longitudinal and refracted (modconverted) shear waves are proportional to the difference in their velocities of sbetween the two media forming the interface. The refracted longitudinal wave hgreater angle of refraction than the mode converted shear wave, due to its highvelocity. Although the angles of refraction for both refracted longitudinal and moconverted shear waves vary with the angle of incidence, the angle between thelongitudinal and shear waves will always be the same.

Fig. 700-11 Reflection and Refraction of Ultrasonic Wave at Interface Between Two Materials



ow) ves is

n d

to ut

mall s

for d be

ed to ). The for lse is

Ultrasonic InstrumentsMost UT systems used for the in-service inspection of a pressure vessel employ a “pulse-echo” method of operation, in which the same transducer (described belthat creates the ultrasonic wave (pulse) is also used to detect any ultrasonic wa(echos) reflected by flaws. A typical ultrasonic instrument for the pulse-echo UTillustrated in Figure 700-12. “Pitch-catch” or “dual-element” transducers have recently greatly improved the reliability of detecting and accurately sizing certaitypes of planar flaws (or cracks) that can occur during service, such as tight anbranched stress-corrosion cracks.

The ultrasonic instrument generates high frequency electrical impulses for acti-vating an ultrasonic transducer and has a control knob for selecting the input frequency to the transducer. Frequencies normally used for UT range from 0.5 10 MHz. High frequencies provide greater sensitivity for detecting small flaws, bpenetration is reduced due to greater attenuation attributable to scattering by sirregularities in the material. Metals with coarse grained microstructures, such asome austenitic stainless steel weld metals, can present an especially difficult problem for obtaining adequate penetration while retaining sufficient sensitivity detecting small flaws. Grain sizes that are 1/10 of the ultrasonic wave length anlarger tend to have a very serious effect upon attenuation. Low frequencies canused to increase penetration, but sensitivity for detecting small flaws will be sacrificed.

Controls are also provided for setting the duration of ultrasonic pulses transmittthe workpiece and the interval between successive pulses (pulse repetition ratesettings used depend primarily on the thickness of the workpiece. It is desirableall echos of one ultrasonic pulse to fade out in the workpiece before the next puemitted. “Phantom” reflections can be obtained if the pulse rate is set too high.

Fig. 700-12 Ultrasonic Instrument for Pulse-Echo UT



e 3. to t for

hout e

at an ent ave

as a , rs at to rface.

s

l

ve air

An amplifier is incorporated into the instrument to amplify the ultrasonic echos received by the transducer, and a gain (or sensitivity) adjustment is provided to opti-mize the signal-to-noise ratio.

The ultrasonic echos received by the transducer are visually displayed on an oscillo-scope. Both the amplitudes of the echos, and the times at which they are received (normally calibrated to indicate distance) are displayed.

Ultrasonic TransducersUltrasonic waves for UT are created in a metal workpiece with a piezoelectric trans-ducer that is coupled to the workpiece. The piezoelectric transducer converts high frequency electrical impulses into correspondingly high frequency mechanical vibrations that induce ultrasonic waves in the workpiece. For detection of reflected waves, the piezoelectric transducer operates opposite to the manner in which it creates ultrasonic waves, by converting the reflected stress (ultrasonic) waves to electrical impulses. The transducers are moved over the surface of the workpiece, usually manually, but automated scanning devices can be used.

UT with a longitudinal ultrasonic wave that is introduced into the workpiece perpendicular to the surface of the workpiece is referred to as “longitudinal wavUT.” A typical transducer for longitudinal wave UT is illustrated in Figure 700-1Movement of the transducer over the surface of the workpiece would subject it abrasion, and, therefore, a wear-resistant faceplate is normally provided. Excepthis relatively thin faceplate, the piezoelectric element in the transducer directlyinduces the ultrasonic wave into the workpiece perpendicular to the surface witpassing through another medium. Therefore, there is no mode conversion of thultrasonic wave at the surface of the workpiece.

UT with an ultrasonic wave introduced into the workpiece at an angle is usuallyreferred to as “shear wave UT.” A typical transducer for shear wave UT is illus-trated in Figure 700-14. The piezoelectric element in the transducer is oriented angle to the surface of the work piece by interposing a wedge between the elemand the workpiece. The piezoelectric element induces a longitudinal ultrasonic win the wedge, which is commonly manufactured from lucite. The lucite wedge hsignificantly different velocity of sound than the metal workpiece, and, thereforemode conversion of the longitudinal wave propagating through the wedge occuthe interface with the workpiece. The angle of the wedge is most often selectedobtain a shear wave propagating through the workpiece at 45 degrees to the suHowever, shear waves at angles of 60 degrees and 70 degrees are also used, depending on orientation of the flaws to increase the intensity of reflected wave(echoes) to maximize the sensitivity for detecting small flaws. It is advisable to perform shear wave UT at two or more angles when the orientation of flaws is uncertain or random.

CouplantsAttenuation of ultrasonic waves propagating through air is very high, and normasurface roughness (especially of the workpiece) will always create an air gap between an ultrasonic transducer and the workpiece. It is not practical to improthe surface of the workpiece to the extent necessary to effectively eliminate the



m

with ting

by or if

etect face gitu-lar to the

that grity rpen- will will

gap. Therefore, a “couplant” that will allow the ultrasonic waves to propagate frothe transducer to the workpiece with significantly less attenuation is used.

Satisfactory couplants are usually viscous liquids or greases that fill the surfaceirregularities of the transducer and workpiece, and thereby eliminate the air gapmoderate pressure on the transducer. Attenuation of ultrasonic waves propagathrough the viscous liquids or greases will not be as high as through air, but thelayer of the couplant must still be maintained relatively thin by pressure on the transducer, to obtain good transmission. Very rough workpieces may have to beground to obtain a sufficiently thin layer of couplant. Some surface preparation light grinding is frequently required when corrosion has roughened the surface scales have formed.

Longitudinal Wave UTLongitudinal wave UT is used primarily to determine the remaining thickness ofcorroded shell components, including the depth of pits. It can also be used to dinternal flaws that have developed during service and that have a reflecting suressentially parallel to the surface of the workpiece (i.e., perpendicular to the londinal ultrasonic wave), such as hydrogen blisters. Surfaces that are perpendicuthe longitudinal ultrasonic wave will usually reflect echos directly back towards transducer with sufficient amplitude to be detected by the transducer.

Longitudinal wave UT is not an especially useful technique for detecting crackshave developed during service. Cracks that have a significant effect on the inteand reliability of a pressure vessel propagate in a direction that is essentially pedicular to the surface of the component. The reflecting surfaces of these cracksbe essentially parallel to the longitudinal ultrasonic waves, and, therefore, they not normally reflect echos towards the transducer with sufficient amplitude to bereliably detected.

Fig. 700-13 Typical Transducer for Longitudinal Wave UT Fig. 700-14 Typical Transducer for Shear Wave UT



e

n be ent d in

on of rom on

en

CalibrationCalibration of longitudinal wave UT consists of developing a “distance amplitudcurve (DAC)” for the instrument and transducer, using test blocks manufacturedfrom a material similar to the workpiece (i.e., with the same velocity of sound). Calibration for determining the remaining thickness of a corroded component caaccomplished by placing the transducer sequentially on test blocks (or on differlocations on a single “step block”) with different known thicknesses, as illustrateFigure 700-15.

The amplitude of the reflection (echo) received by the transducer from the backsurface decreases as the thickness of the test block increases, due to attenuatithe ultrasonic wave propagating through the block. An initial pulse is obtained fthe front surface of the test block, which is positioned on the zero thickness linethe oscilloscope, by adjusting the “time delay” (or “sweep”) control on the ultra-sonic instrument. The back reflections from the other known thicknesses are th

Fig. 700-15 Calibration of Longitudinal Wave UT for Thickness Measurement



shifted as necessary to coincide with the proper thickness lines on the oscilloscope display, by adjusting the material velocity control on the ultrasonic instrument.

Remaining ThicknessFigure 700-16 illustrates the use of longitudinal UT for determining the remaining thickness of a corroded shell. If the transducer is placed on a location of the shell component that is not corroded, as depicted by the position of Transducer A in Figure 700-16, a back reflection will be observed at a distance equal to the original thickness of the shell. On the other hand, if the transducer is placed on a location where internal corrosion has occurred, as depicted by the position of Transducer B in Figure 700-16, a back reflection will be observed that is less than the original thickness. The back reflection from a corroded surface of a vessel shell component can have an amplitude below the DAC, because irregularities in the corroded surface can scatter some of the reflected longitudinal wave.

Fig. 700-16 Determination of Remaining Thickness of Corroded Pressure Vessel Shell with Longitudinal Wave UT Calibrated in Figure 700-15



The distance of the back reflection on the oscilloscope display is the remaining thickness of the shell component, and the size of the corroded area can be deter-mined by moving the transducer along the surface until the back reflection returns to the original thickness.

Hydrogen BlistersThe use of longitudinal wave UT for detecting, locating, and determining the size of hydrogen blisters is illustrated in Figure 700-17. The blisters are internal flaws that have a reflecting surface at a depth from the front surface that is less than the distance to the back surface (thickness) of the shell component. Therefore, the blis-ters will cause reflected peaks to appear in the oscilloscope display at a distance less than the thickness of the shell.

The peaks attributable to the blisters will usually be lower than the DAC, because the surface of the blisters will not normally be as good a reflector as the back surface (similar to the effect of a corroded surface). Furthermore, very thin blisters may have sufficient contact between their opposite surfaces to permit propagation of

Fig. 700-17 Detecting Hydrogen Blisters with Longitudinal Wave UT



in shift from

d ding

ace hole cillo-

by

some of the ultrasonic wave through the blister. This situation is depicted by the position of Transducer B, in which case a small reflection from the back surface will appear in the oscilloscope display. However, no portion of the ultrasonic wave can propagate through blisters with more widely separated opposite surfaces, and no reflection from the back surface will be observed, as depicted by the position of Transducer C.

The distance of the peaks on the oscilloscope display resulting from the blisters indi-cates their depths below the surface, and the size of each blister can be estimated by moving the transducer along the surface until the reflection attributable to the blister disappears.

Shear Wave UTShear wave UT is used primarily to detect and determine the size of cracks that have developed during service. Forms of deterioration that can result in cracking include mechanical and thermal fatigue, creep, stress-corrosion, and hydrogen attack, among others (see Section 730). Although shear wave UT can provide very good data for the evaluation of the integrity and reliability of a vessel, quality of the data is highly dependent on the skill and expertise of the technician performing the examination. Only highly qualified technicians should be used who have demon-strated their ability to correctly detect and size the types of flaws that might have developed in the vessel.

CalibrationShear wave UT is calibrated using a test block manufactured from a material similar to the workpiece (i.e., with the same velocity of sound), that has side-drilled holes and a notch on the back surface as shown in Figure 700-18. The test block should have a thickness within 1 inch of the thickness of the workpiece. The side-drilled holes should be 3/16-inch diameter for test blocks up to 4 inches thick, and increase 1/16-inch in diameter for each 2-inch increase in thickness of the test block greater than 4 inches.

The transducer is first moved along the surface until the reflection from the ¼-thick-ness drill hole attains maximum peak amplitude at the position of Transducer AFigure 700-18, and the delay control on the ultrasonic instrument is adjusted tothe peak to a distance on the oscilloscope display equal to the depth of the holethe surface.

Next, the transducer is moved along the surface until the reflection from the ¾-thickness drill hole attains maximum peak amplitude at the position of Trans-ducer C, and the material velocity control on the ultrasonic instrument is adjusteuntil the peak coincides with the distance on the oscilloscope display corresponto the depth of this hole from the surface. Moving the transducer along the surfuntil a reflection of maximum peak amplitude is obtained from the ½-thickness at the position of Transducer B should produce this peak at a distance in the osscope display corresponding to the depth of this hole below the surface.

Then, the transducer is moved until a reflection of maximum peak amplitude is obtained from the ¾-thickness hole when the ultrasonic shear wave is reflected



the back surface of the test block at the position of Transducer E. This peak should appear in the oscilloscope display at a distance corresponding to 5/4 the thickness of the test block. The maximum peak amplitudes determined in this manner will estab-lish a distance amplitude curve (DAC) for shear wave UT. The sensitivity control on the ultrasonic instrument should be adjusted until the amplitude of the highest peak is at least 80% of the oscilloscope screen height.

A very strong reflection is normally obtained from the notch on the back surface with an ultrasonic shear wave at 45 degrees, which can exceed the DAC. However, ultrasonic shear waves at 60 degrees and 70 degrees will usually produce reflec-tions from the notch that are below the DAC.

Crack DetectionShear wave UT is very useful for detecting cracks that have developed during service. Figure 700-19 illustrates how shear wave UT, calibrated according to Figure 700-18, can be used to detect a crack in the heat affected zone of a weld joint. This crack has started at the back surface of the workpiece (I.D. of the vessel), and has propagated towards the front (O.D.) surface.

Fig. 700-18 Calibration of Shear Wave UT



Shear wave UT is performed by moving the transducer along the O.D. surface of the vessel towards the weld joint. No reflection will be observed in the oscilloscope display with the position of Transducer A in Figure 700-19, because the ultrasonic shear wave is reflected by the I.D. surface away from the transducer.

A relatively strong reflection from the base of the crack will be observed in the oscilloscope display for the position of Transducer C, because the base of the crack at the I.D. surface of the vessel acts like a notch in the back surface of a test block. The amplitude of this peak can exceed DAC, and it will appear at a distance in the oscilloscope display corresponding to the thickness of the vessel shell at this location.

Fig. 700-19 Shear Wave UT Calibrated According to Figure 700-18 for Detecting a Crack in Heat Affected Zone of Weld Joint



is ck

nt

ret ive er ssel s, to

will ave s-h indi-he le

ssel ucer

ope then at l not ks,

are a-eaks they

ed. e ure xper-SNT

When the transducer is moved closer to the weld, as depicted by the position of Transducer D, the amplitude of the peak will usually decrease significantly, and it will appear in the oscilloscope display at a distance corresponding to a depth less than the thickness of the vessel shell. The crack does not always provide a good surface for reflecting the ultrasonic shear wave back to the transducer for detection, because it is usually not oriented perpendicular to the wave. However, cracks normally have a rough (or faceted) texture that will reflect at least a small portion of the ultrasonic wave. The amplitude of this reflection can be very low and difficult to distinguish from the “noise” that results from the relatively high gain setting thatrequired to obtain satisfactory peak heights from the drilled holes in the test bloduring calibration. ASME Code, Section V, Article 4, now requires recording all reflections with amplitudes of 20% DAC and greater as flaws. This is a significaincrease in the sensitivity of crack detection from the previously used 50% DACrecording level.

Considerable skill and experience is required for a technician to properly interpreflections with amplitudes this low as an indication of a crack. Cracks that givevery low amplitude reflections with ultrasonic shear waves at 45 degrees can gstronger reflections at 60 degrees or 70 degrees. Ultrasonic waves at the greatangles will be more nearly perpendicular to a crack propagating through the veshell. Therefore, it is advisable to perform shear wave UT at two or more anglemaximize the probability of detecting the cracks.

When the transducer is moved still closer to the weld, the ultrasonic shear waveeventually pass above the tip of the crack, as depicted for Transducer E. The wwill once again be reflected by the I.D. surface of the vessel away from the tranducer, and the peak will disappear. Disappearance of the peak can give a rougcation of the depth of a crack, but it should never be used by itself to evaluate tintegrity and reliability of a vessel. Other shear wave UT techniques are availabthat provide much more accurate crack depth and size data.

It is possible for the ultrasonic shear wave reflected by the I.D. surface of the veto be reflected subsequently by the crack, as depicted for the position of TransdB. The peak for this type of reflection from the crack will appear in the oscilloscdisplay at a distance that is greater than the thickness of the vessel shell. It willbe necessary for the technician to use geometry to determine the actual depth which the wave is reflected by the crack. The actual I.D. surface of a vessel wilalways be a good enough reflector to produce reflections of this type from cracbecause this surface can be roughened considerably by corrosion.

Very fine cracks, such as those resulting from stress corrosion, and cracks thatfilled with a corrosion scale may permit a portion of the ultrasonic wave to propgate through. This wave propagation will further reduce the amplitudes of the pin the oscilloscope display attributable to these types of cracks, and, therefore, can be difficult to detect. Proper calibration of shear wave UT with test blocks containing side drilled holes will not guarantee that cracks will always be detectA special test block containing cracks of the type that may have developed in thpressure vessel should be used to demonstrate that the shear wave UT procedbeing used is capable of detecting the cracks, and that the technician has the etise to properly interpret the data. This concept is gradually being adopted by A



ll the that re

pth h-

ain

e in - that

r to . ld

the

d ve is nt of

er is

and ASME. Contact CRTC Engineering analysis for information and assistance.

Using shear wave UT to detect cracks in nozzle welds or other vessel components with complex geometries can be more difficult than for the previous example. The technician performing the examination will have to know the actual geometry and dimensions of the component, to determine the angles for the ultrasonic shear waves that are required to probe the locations where cracks are most likely to develop. In addition, more complex component geometries can have surfaces that reflect the waves back to the transducer. The technician will have to be able to distinguish these reflections from the surfaces of the vessel component from reflections coming from cracks.

Crack SizingThe ability to determine the depth of a crack through the thickness of a vessel shell is a very important attribute of shear wave UT. However, the accuracy of the depth measurements made with UT can vary considerably, depending on the technique used and the skill of the technician.

UT crack sizing techniques are classified as either “amplitude based” or “time based.” Amplitude based techniques were the first to be developed, and are stimost frequently used. They are relatively simple to use, but it has been learnedthey do not always give accurate results. Time based techniques have been morecently developed and are capable of providing significantly more accurate demeasurement. However, they are more difficult to use than amplitude based tecniques, and, therefore, appreciably more skill and expertise is necessary to obtaccurate results.

Amplitude Based SizingFigure 700-20 illustrates how the depth of a crack is determined using amplitudbased sizing techniques. The crack was detected by shear wave UT as shown Figure 700-19. The occurrence of a relatively high amplitude peak in the oscilloscope display at a distance corresponding to the thickness of the shell confirmsthe crack has initiated at the I.D. surface.

The peak will normally decrease in amplitude as the transducer is moved closethe weld, as shown for the position of Transducer B in Figure 700-20, when the“corner” reflection attributable to the base of the crack at the I.D. surface is lostThe amplitude of the peak reflected from the crack can be quite low, and it shoube increased to approximately 80 percent of screen height (12 dB) by adjustingsensitivity control on the ultrasonic instrument.

Amplitude of the peak will remain relatively constant as the transducer is movecloser to the weld. Only a portion of the ultrasonic wave will be reflected by thecrack as the transducer is moved further towards the weld, and some of the wapasses over the tip of the crack. It is generally assumed that the tip of the cracklocated by a decrease in peak amplitude to 50% of the maximum after adjustmethe sensitivity control (i.e., 6 dB drop), which is shown for the position of Trans-ducer C. The depth of the crack is determined by the distance that the transduc



to arge

moved along the surface from the indication of the base of the crack (position A) to the indication of the tip of the crack (position C).

Amplitude base sizing is reasonably satisfactory for crack depths that exceed the diameter of the transducer. However, other characteristics of the crack (such as changes in orientation, roughness, and width) can affect the amplitude of the peak, which can lead to errors in sizing.

Amplitude based sizing is generally not satisfactory for cracks shallower than the diameter of the transducer. This is the situation with most vessel inspections. A 1/8-inch deep crack may not significantly affect the integrity and reliability of a vessel, whereas a ¾-inch crack would require repair before the vessel is returnedservice. Amplitude based sizing tends to oversize small cracks and undersize l

Fig. 700-20 Amplitude Based Sizing Using Shear Wave UT



tip n of

r for nt on

ones, and, therefore, may not accurately discriminate between the shallow superfi-cial crack and the deeper one that could jeopardize safe operation.

Time Based SizingTime based sizing generally provides greater accuracy for determining depth, espe-cially when the crack is relatively small with respect to the diameter of the trans-ducer. Therefore, time based sizing should be used whenever a fitness-for-service analysis is made to evaluate the integrity and reliability of a vessel.

Several variations of time based sizing have been developed, but only the general concept is discussed here. Different vendors may use different techniques, but they are all significantly more complicated to apply than amplitude based sizing. All time based sizing techniques are highly dependent on the skill and expertise of the tech-nicians performing the examination, and only the most qualified vendors who can appropriately demonstrate the accuracy of their procedures should be used.

Figure 700-21 illustrates one shear wave UT technique that can be used to deter-mine the depth of a crack. The ultrasonic wave propagates through the material as a wave front having a width at the surface equal to the diameter of the transducer. The width of the wave front spreads somewhat as it propagates, but this can be mini-mized by the use of high frequencies and large diameter transducers. Two peaks, referred to as a doublet, can be observed in the oscilloscope display for the position of the transducer.

A peak with a relatively high amplitude (Peak R) will be developed by the corner reflection of the ultrasonic wave from the base of the crack at the I.D. surface. This reflected peak will normally have maximum amplitude at the distance in the oscillo-scope display corresponding to the thickness of the vessel shell.

The portion of the ultrasonic wave that passes over the tip of the crack is diffracted, which results in the formation of a second peak (Peak D) with an appreciably lower amplitude (referred to as a tip-diffracted satellite pulse). The amplitude of the tip-diffracted peak can be maximized for better observation by moving the transducer, which moves the reflected peak from the distance in the oscilloscope display that corresponds to the thickness of the shell and decreases its amplitude. Nevertheless, the sensitivity control on the ultrasonic instrument usually has to be adjusted quite high to be able to see the tip-diffracted pulse, and, therefore, it can be difficult to distinguish from the background noise. One aid to distinguishing the tip-diffracted peak from the noise is that the tip-diffracted peak will move in unison with the reflected peak (i.e., as a doublet) across the oscilloscope display as the transducer is moved.

The separation between the reflected and diffracted peaks is constant whenever they are observed together, regardless of the position of the transducer, and is the result of the difference in the “time-of-flight” of the ultrasonic wave from the base and of the crack. Therefore, the distance in the oscilloscope display of the separatiothe peaks indicates the depth of the crack (∆× in Figure 700-21).

Other techniques for time base sizing can be used, and, in fact, may be superiosome types of cracks. The actual technique that is used is based to a large exte



the vendor’s knowledge and experience. The vendor should be required to demon-strate the accuracy of the technique by examining a test block containing cracks similar to those in the pressure vessel that is being evaluated. Consult Company Inspection Groups, or a specialist for recommendations.

UT ImagingUT imaging can be used to obtain two- and three-dimensional pictures of cracks, or other types of flaws, in a pressure vessel component from an automated ultrasonic examination. An ultrasonic transducer (either longitudinal or shear wave) is moved over the vessel’s surface with a scanning fixture. The X-Y coordinates of the trans-ducer locations on the vessel and the UT signals obtained by the transducer are recorded on a magnetic disk as the transducer is moved over the vessel’s surface. A computer subsequently constructs an image of the crack (or other type of flaw) based on the UT signal amplitude or distance data correlated with the location of the transducer.

Both plan view and cross-sectional UT images can be obtained for permanent records to very accurately determine the location, orientation, and size of a crack.

Fig. 700-21 Time Based Sizing Using Shear Wave UT



f a the ss rred

s. ving law for

teg-, rk

of ni-l

only

cu-chni-ng i-ld be vide ure

alifi-

ed ently,

g

These images are especially useful to a fitness-for-service analysis, that is critically dependent upon an accurate knowledge of the crack’s dimensions with respect to the maximum principal membrane stress (see Section 750).

UT imaging can also be used to “map” internal corrosion from the I.D. surface ovessel, using a longitudinal wave UT transducer. The image constructed by thecomputer will depict the I.D. surface contour of the vessel, which represents theinternal corrosion loss. This can be most useful when it is desired to determine size of locally corroded areas with a remaining wall thickness that is possibly lethan the minimum required, to evaluate if repairs are necessary when it is prefenot to open the vessel for an internal inspection.

It should be noted that UT imaging does not improve upon the accuracy of the UT data itself, and, therefore, it is not a substitute for using the optimum UT proce-dures for flaw detection and sizing discussed above for manual UT examinationHowever, the UT image can be very beneficial for analyzing the UT data and gia permanent record of flaw location, orientation, and size. Characteristics of a fcan become readily apparent in a UT image. The image can be very important evaluating the significance of a flaw that could be missed with a manual UT examination.

Applications and LimitationsUT is a very efficient NDE method. A large amount of data for evaluating the inrity and reliability of a vessel can be obtained in a relatively short period of timewithout requiring extensive preparation of the vessel or interfering with other woin the area.

Longitudinal wave UT is applicable for determining the remaining wall thicknessa corroded pressure vessel under almost any circumstances, and most UT techcians have the skill and experience to perform these examinations. Longitudinawave UT will also detect and locate hydrogen blisters, or similar internal flaws. Shear wave UT is very useful for detecting cracks and provides essentially the method for determining the size (depth) of cracks with sufficient accuracy for making a fitness-for-service analysis to evaluate the integrity and reliability of avessel.

The greatest limitation on the use of UT for in-service inspections is that the acracy of the data obtained is highly dependent on the skill and expertise of the tecians performing the examinations. This is especially true for detecting and sizicracks by shear wave UT. Qualification of procedures and certification of techncians is not sufficient to guarantee acceptable results. Only those vendors shouused that can demonstrate that they have the required skill and expertise to proaccurate data. Samples of various types of cracks that have developed in pressvessels during service should be saved for future use as “standards” for the qucation of UT procedures and technicians for in-service inspections.

Another limitation on the use of UT for in-service inspection is that it was not usduring the construction of most of the vessels that are now operating. Consequmany indications of flaws detected by UT are very difficult to classify as either innocuous fabrication flaws or more serious indications of deterioration occurrin



during service. UT procedures are available that can usually distinguish between fabrication flaws and those that have developed during service, but they require highly skilled UT examiners to properly apply. The increasing use of UT during the construction of vessels, as provided for in PVM-MS-4750 and PVM-MS-4749, included in this manual, may eventually alleviate this difficulty.

765 Radiographic Examination (RT)RT is very useful for detecting both surface and internal flaws, and it is the primary NDE procedure required by the ASME Code to verify the quality of welds during construction. However, the associated radiation hazard makes it difficult to use for inspection of pressure vessels during shutdowns, when other personnel are working on the same vessel or close to it. The investment in equipment (radiation sources and darkroom facilities for processing film) can also be quite high, but this can be offset by the use of qualified contractors.

Article 2 of ASME Code, Section V, gives the minimum requirements for an RT procedure for pressure vessels.

Physical PrinciplesX-rays and gamma rays penetrate steel, but the intensity of the incident radiation will be attenuated as it passes through the material. The degree of attenuation depends on the thickness and density of the material.

Flaws can have the effect of reducing the thickness of material through which the radiation must pass by interposing cavities or impurities of lower density in the workpiece. Therefore, there is less attenuation of radiation passing through the flaw than through the surrounding material, as illustrated in Figure 700-22.

Fig. 700-22 Attenuation of Radiation by Workpiece Containing a Flaw



Photographic film placed opposite the source of radiation will be exposed by the radiation that has passed through the workpiece (transmitted radiation) and, conse-quently, the flaw will appear as a dark image on the developed negative (referred to as a radiograph). It is important to recognize that the radiation passing through the workpiece does not directly interact with the flaw. The flaw is detectable only because it alters the thickness of material through which the radiation passes. The image of the flaw on the radiograph is actually the silhouette of the three-dimen-sional flaw projected onto the two-dimensional surface of the film, as illustrated in Figure 700-23.

Flaws will not always significantly reduce the thickness of material through which the radiation must pass. The flaw in Figure 700-22b is identical to the flaw in Figure 700-22a, except that it is rotated 90 degrees. With this orientation, the flaw will not significantly reduce the thickness of material. Therefore, the attenuation of radiation passing through this flaw will be essentially the same as that for the radia-tion passing through the surrounding material, and there will be no indication of the flaw.

Fig. 700-23 Projection of Three-Dimensional Flaw in Workpiece onto Two-Dimensional Surface of Film



of e -

rue tive

t five

oper

Radiation SourcesBoth x-rays and gamma rays can be used as the incident radiation. Energy and intensity are the most important characteristics of the incident radiation. The energy of the incident radiation determines its ability to penetrate the workpiece. Higher energy radiation is required to penetrate thicker workpieces but it reduces the scatter of radiation passing through the workpiece. The intensity of the radiation reaching the photographic film after it has passed through the workpiece controls the length of time required to properly expose the film. Higher intensity radiation is required for thicker workpieces to obtain reasonable exposure times, because attenuation of the radiation passing through the workpiece increases with material thickness.

X-rays are usually produced by x-ray tubes, which are electronic devices that convert electrical energy into x-rays. The voltage of the x-ray tube determines the energy of the x-rays produced, and the current controls the intensity of the radia-tion. Consequently, higher voltages and currents are required for thicker work-pieces, which necessitate larger x-ray tubes and electrical power sources. The bulk of the equipment required to produce high energy-high intensity radiation with x-ray tubes severely restricts portability. Therefore, RT systems suitable for pres-sure vessels are generally not satisfactory for shell thicknesses greater than 2 inches.

Linear accelerators (LINAC) can also be used to produce x-rays. They are much more practical than x-ray tubes for producing the high energy-high intensity radia-tion required for thick workpieces, and some systems are sufficiently portable. However, the equipment is still quite cumbersome to handle.

Gamma-rays are created by the radioactive decay of unstable isotopes of naturally occurring or artificially produced elements. Cobalt-60 and iridium-192 are the two isotopes most commonly used for RT. The radiation has a relatively high energy for penetrating thick workpieces. However, the intensity of the radiation is generally lower than that produced by x-ray tubes or LINACs, and it diminishes with time as the radioactive isotope decays. Therefore, the time required to properly expose the film can be quite long. Nevertheless, the equipment required for RT with gamma rays from a radioactive isotope is much less cumbersome than that required to produce x-rays, and it is therefore more suitable.

A “point” source of radiation would provide the sharpest images of flaws on thephotographic film exposed to the radiation passing through the workpiece. However, actual sources of radiation used for RT are provided with an apertureapproximately 1/10 to 1/4 inch to obtain sufficient intensity of radiation to exposthe film in a reasonable length of time. Consequently, some “geometric unsharpness” results in the image of the flaw, because the source of radiation is not a tpoint source. The unsharpness is reduced by large source-to-film distances relato the thickness of the workpiece. The source-to-film distance should be at leastimes the thickness of the workpiece to give satisfactory clarity.

It is extremely important to recognize that the radiation sources used for RT have much higher energies and intensities than those used for medical x-rays and therefore can very severely damage animal tissues. Severe disability or deathcan result from exposure to the radiation. Therefore, it is essential to provide pr



s the in by ible for

ia-peed atis-

vice rain of an-

sually rate a ensi-nd d lead

shielding of the radiation source and to prevent unauthorized access to the area when RT is being performed. Guidance for shielding and restriction of access should be obtained from knowledgeable safety and health specialists before performing the RT.

Photographic FilmThe radiation which passes through the workpiece is recorded by photographic film. The radiographs are usually interpreted visually with the aid of a high intensity light source (light box), but optical densitometers or image analyzers are occasionally used.

Two primary characteristics of the film can affect the sensitivity of RT for detecting flaws: gradient and grain. Gradient is the difference in optical density of the nega-tive resulting from exposure by different intensities of radiation. A film with high gradient exhibits relatively large differences in density when exposed by radiation of relatively small differences in intensity. In other words, a high gradient results in a relatively high contrast in the negative, which makes small differences in the intensity of radiation passing through the workpiece visible. Therefore, films that have a high gradient will provide the greatest sensitivity for detecting small flaws that cause only a small attenuation of the radiation.

Grain results directly from photo-sensitized crystals in the film. The darkened crys-tals impart a visually apparent “graininess” to the transparent negative that limitdetail that can be resolved by viewing the negative, regardless of the differencedensity of adjacent areas of the film. Therefore, very fine flaws can be obscuredthe grain of the film can obscure very fine flaws, so the flaws and may not be visin a radiograph, although they significantly reduce the effective thickness of theworkpiece. Fine grain films usually also have a high gradient, and are preferredRT to obtain the greatest sensitivity for the detection of flaws.

Another characteristic of film that can affect the sensitivity of RT for detecting flaws is speed, which is a measure of the sensitivity of the film to the radiation passing through the workpiece. A high speed film requires less exposure to radtion to produce the same optical density in the developed negative than a low sfilm. Therefore, less exposure time is required for high speed films to produce sfactory radiographs than with low speed films for the same intensity of incident radiation. The reduction of exposure time can be significant, especially for rela-tively thick workpieces, when low intensity radiation sources are used for in-serinspection. However, high speed films provide less gradient and have coarser gthan low speed films. It is usually inadvisable to sacrifice the greater sensitivity fine grain films with a high gradient for detecting flaws in order to obtain the advtage of shorter exposure times.

The film is loaded into flexible cassettes in a darkroom, and the cassettes are uheld against the surface of the workpiece with magnets. The cassettes incorpo“radiographic screen” that improves the image recorded on the film, by both intfying the exposure of the film by the radiation passing through the workpiece, aby filtering out scattered radiation to reduce fogging. Lead foil is commonly usefor the radiation screen. Radiation penetrating the workpiece interacts with the



the hile ugh

ifi-

ce -xpo-1.8 nsity

asses er, nnot -y of

ork-ate-

in

atoms in the foil, which causes them to “fluoresce,” intensifying the exposure offilm. The lead foil will also absorb most of the low energy scattered radiation, whaving little effect upon the intensity of the higher energy radiation passing throthe workpiece. In this manner, fogging of the film by the scattered radiation is greatly reduced, but exposure of the film by the transmitted radiation is not signcantly affected.

ExposureThe exposure of the photographic film by radiation passing through the workpieis determined by the intensity of the radiation multiplied by the time of the exposure. The optical density of the developed negative increases with increasing esure. ASME Code, Section V, requires a radiograph to have a density between and 4.0 for proper visual interpretation. The exposure must be adjusted if the deof the radiograph is not between these limits, by changing either the time of theexposure or the intensity of the incident radiation.

Flaws that reduce the thickness of the workpiece through which the radiation pwill increase the exposure, and, therefore, the density of the radiograph. Howevthe density of the flaw image compared to that of the surrounding workpiece cabe relied upon to give an accurate indication of the depth of the flaw. The threedimensional shape and orientation of the flaw can significantly affect the densitthe image and are not always revealed by the two-dimensional silhouette in theradiograph.

Radiograph QualityThe acceptability of a radiograph for the detection of flaws is determined with adevice, referred to as a “penetrameter,” that is placed upon the surface of the wpiece when the exposure is made. Penetrameters are relatively thin pieces of mrial with radiation attenuation characteristics similar to the workpiece that contaholes with diameters of 1, 2, and 4 times the thickness, as illustrated in Figure 700-24.

Fig. 700-24 Penetrameter for Evaluating Quality of Radiograph



The quality level required for a radiograph is designated by a two-part expression X-YT, where X is the maximum thickness permitted for the penetrameter as a percentage of the thickness of the workpiece, Y is the diameter of the hole as a multiple of the thickness, and T is the thickness. A quality level of 2 to 2T is adequate for most applications of RT for the in-service inspection of pressure vessels, and is consistent with the requirements of ASME Code, Section V.

A radiograph is considered to be acceptable for the quality specified if the entire outline of the penetrameter is visible, the density of the penetrameter is within the required range of 1.8 to 4.0, and the hole is discernable. However, the acceptability of the radiograph for flaw detection is limited to areas that have densities within 15% under and 30% over the density of the hole in the penetrameter. These limits may not encompass the entire range of densities in a radiograph, and interpretation of areas with densities outside these limits is of questionable validity. It is important to recognize that the workpiece must have an essentially uniform thickness for the background density of the radiograph to be within these limits.

Note that the penetrameter functions only to determine that a radiograph has accept-able quality. The penetrameter does not serve as a calibration standard. Therefore, it should not be used to estimate flaw sizes, and should not be used to establish accep-tance limits for flaws based upon relative densities in the radiograph.

Applications and LimitationsIt is sometimes thought that RT is the best NDE method that can be used for detecting flaws, because it is mandated by the ASME Code for certifying the quality of newly constructed pressure vessels. This is not true. RT has attained its preemi-nence in the ASME Code by virtue of the evolution of weld joint efficiencies and permitted design details around the types of fabrication defects that can be readily detected by RT.

The reliance of the ASME Code upon RT should not be construed to imply that it is the optimum NDE method to use for the in-service inspection of vessels. In fact, most of the flaws that can develop as a consequence of the deterioration of a vessel during service can be better detected by other NDE methods.

One circumstance where RT can be used to considerable advantage is when a direct comparison is desired between the present condition of a vessel and its condition when new, and other NDE methods were not used during construction to provide baseline data.

A significant limitation upon the use of RT for in-service inspection is that it will only detect cracks that are essentially parallel to the direction of the incident radia-tion, and have a sufficient width to be visible in the radiograph as limited by the grain of the film. The detectability of cracks diminishes greatly as they deviate further from this orientation. Taking several radiographs with different angles of incident radiation can overcome this shortcoming, but this would considerably increase the time and cost of the inspection.

RT does not give a reliable indication of the depth of a flaw through the shell of a pressure vessel. Therefore, the depth of a crack detected by RT must be measured



by another NDE method, such as UT, to evaluate the effect of the crack upon the integrity and reliability of the vessel.

RT is also severely limited for the in-service inspection of nozzle openings and welds, which tend to be locations of relatively high stress where deterioration during service is likely to occur. Nozzles are usually fabricated from plate and forging or pipe materials with different thicknesses. Satisfactory radiographs of nozzle open-ings can rarely be obtained, because the workpiece must have an essentially uniform thickness for the variation in density of the radiograph to be within an interpretable range.

RT is the most time-consuming and expensive of all NDE procedures. The set-up time for the equipment is usually much greater than the time required for the expo-sure, and this must be followed by development of the film in a darkroom and inter-pretation of the resulting radiograph. Several man-hours can be required for each exposure. Additional time and cost penalties are incurred indirectly by restricting and delaying other work in the area due to the serious radiation hazard associated with RT. Therefore, the suitability of other NDE methods for detecting the forms of deterioration that might have occurred during service should be investigated before RT is employed.

Despite all of the disadvantages associated with the use of RT for the in-service inspection of pressure vessels, there can be no dispute that a radiograph can provide very valuable data concerning the integrity and reliability of a vessel. The radio-graphs provide permanent records that can be compared to the results of future inspections or reinterpreted in the light of new information concerning the deteriora-tion that can occur during service.

766 Acoustic Emission Testing (AE)AE was developed to assess the integrity of pressure-containing equipment in the nuclear power industry. Acoustic sensors are attached to the O.D. surface of the equipment to detect and locate flaws in the material during a pressure test. AE can be a very useful NDE method to use for the in-service inspection of pressure vessels, especially because one test will cover all components of a vessel and weld joints in the shell.

Physical PrinciplesSonic impulses are emitted by a flaw in a ductile material subjected to a gradually increasing applied stress, due to the release of strain energy in the intensified stress field associated with the flaw. The release of strain energy creates stress waves that propagate through the material at the speed of sound. These stress waves behave like sonic impulses, and are, therefore, commonly referred to as acoustic emissions. However, there may be no significant release of strain energy in a brittle material to cause acoustic emissions prior to the occurrence of fracture.

Acoustic emissions attributable to flaws in metals are usually inaudible, but they can be thought of as being similar to the familiar sounds emitted by a piece of wood as it is gradually bent to the breaking point. The sounds not only reveal that the wood is



por-

fied oc-laws

ener-le to enever

sting be

r-ing

mploy is-ors

cracking, but an individual hearing the sound can also sense the direction that it is coming from to help locate the piece of wood. It is useful to draw another analogy for brittle materials, such as a glass rod that is bent to the breaking point. No warning sounds will be heard as the glass rod is bent before it breaks.

The nominal applied stress in a component of a pressure vessel manufactured from a ductile metal can be well below the yield strength of the material, but the local stress concentrations developed at flaws can be high enough to cause highly local-ized plastic deformation or enlargement of the flaw by microcrack propagation. Both plastic deformation and crack propagation produce acoustic emissions that can be detected by acoustic sensors to disclose the existence of the flaw.

A single sensor can usually detect acoustic emissions originating anywhere in a relatively large vessel, because attenuation of the sonic impulses in a metal shell is very low. However, several sensors are usually employed to obtain confirming indi-cations, and, more importantly, to locate the origin of the emissions in the shell by triangulation.

Test PressureAcoustic emission tests of a pressure vessel are usually conducted by filling the vessel with water, and slowly increasing the hydrostatic pressure until it exceeds the maximum pressure that the vessel has experienced during recent operation.

Early development of acoustic emission testing showed that significant acoustic emissions are not observed until the recent maximum operating pressure has been exceeded. This phenomenon is known as the “Kaiser Effect,” and it is a very imtant fundamental of acoustic emission testing. Acoustic emissions are directly related to localized plastic deformation or microcrack propagation at the intensistress fields associated with flaws in the material. Plastic deformation and micrrack propagation at flaws will not normally recur unless the stress fields at the fare increased above the previous maximum levels.

The minimum acceptable hydrostatic pressure for an acoustic emission test is gally 10% above the recent maximum operating pressure. However, it is desirabhave the test pressure reach 20% above the maximum operating pressure, whpossible, to be certain that all significant flaws have been detected. It is not normally necessary to attain the ASME Code hydrotest pressure, but if hydroteof the vessel is required for any other reason, the acoustic emission test shouldconducted at the same time at this higher pressure.

An acoustic emission test of a pressure vessel must not be conducted when the ambient temperature is below the minimum design temperature (minimum pressurizing temperature). The material is susceptible to brittle fracture at tempeatures below this limit, and the acoustic emissions may not give sufficient warnto stop the test in time to prevent failure of the vessel.

Acoustic Sensors and InstrumentationMost acoustic sensors used for acoustic emission testing of pressure vessels epiezoelectric ceramic elements that convert the stress waves (i.e., acoustic emsions) propagating through the vessel shell into electrical impulses. These sens



es” -ws. r n r is

m-h

o ydro-

ata. f the

h is

yzed . dor

or-able

ally t

ignal ayed

and re

ca-

are usually attached to the vessel shell with magnets or an epoxy glue. Several sensors are positioned on the shell in a geometric pattern that is based on the size and configuration of the vessel, to be certain that significant emissions from any major component of the vessel will be detected by at least two of the sensors. Multiple sensors are also required for the location of the acoustic emission sources by triangulation.

The electrical outputs of the acoustic sensors are sent separately to a preamplifier that contains a band pass filter to cut off frequencies below 50 kHz. Acoustic emis-sions with lower frequencies are predominantly mechanical and hydraulic “nois(such as movement of insulation, vessel supports, internals, and flange connections), and they could seriously confuse the interpretation of emissions from flaSome emissions from these mechanical and hydraulic sources can have highefrequencies that coincide with those that originate at flaws, which must be takeinto consideration when analyzing the data. The filtered signal from each sensosubsequently sent to a main amplifier, and then on to various multichannel recording and monitoring instruments.

Recording of the acoustic emission data for permanent retention is usually accoplished with a magnetic disk. The amplitude of every emission detected by eacacoustic sensor is recorded on a time base. Audio and visual monitors can be provided for “real-time” observation of the test results. X-Y recorders are used tplot the acoustic emission data as a function of some test parameter, such as hstatic test pressure. These plots can be made during the test, but they are morecommonly made immediately after the test for a more detailed analysis of the dX-Y recorders can also be used to plot on a “map” of the vessel the locations oflaws revealed by the acoustic emissions.

Pressure transducers are used to determine the hydrostatic test pressure, whicalso recorded on the magnetic disk on a time base.

Analysis of DataThe data obtained from an acoustic emission test of a pressure vessel are analto determine the existence of flaws and their locations in the shell of the vesselAccurate analysis of the data is highly dependent upon the expertise of the venperforming the test. Only those vendors that can demonstrate that they have acquired appropriate knowledge and experience should be used. It is vitally imptant for the vendor to be able to distinguish between acoustic emissions attributto flaws, and other “noises” detected by the acoustic sensors during the test.

The acoustic emission signals detected by each sensor during a test are genercharacterized for analysis by number of counts above a threshold amplitude thadetermines the sensitivity of the test, amplitude, duration, energy (area under senvelope), and rise time as illustrated in Figure 700-25. The data are then displon CRT monitors as illustrated in Figure 700-26.

Plots of cumulative count, and count rate vs. time are shown in Figure 700-26ab respectively. A plot of cumulative count rate vs. test pressure is shown in Figu700-26c. An abrupt increase in counts vs. either time or test pressure is an indi



r is ing

en the

tion of the existence of a significant flaw. Innocuous flaws are generally character-ized by a more gradual increase in counts.

A plot of the number of counts vs. amplitude is shown in Figure 700-26d, which can be useful for distinguishing between various plastic deformation and flaw propagation mechanisms, or for separating emissions associated with flaws from background noise. Figure 700-26e is a plot of the cumulative number of counts of equal or greater amplitude, which is more useful for evaluating the severity of a flaw.

Figure 700-26f is an acoustic emission source location display, which is basically a map of the vessel with the computer location of each emission source. This map can be used to focus subsequent NDE, such as MT and UT, that are usually desirable to confirm the existence of flaws in the vessel and to determine their sizes for fitness-for-service evaluations (see Section 750).

The location of flaws is determined by analyzing the arrival time of the same acoustic emission from a flaw at different acoustic sensors attached to the shell of the vessel. The arrival time of an acoustic emission at a sensor is dependent on the distance of the flaw from the sensor, and the speed of sound in the material. There-fore, the differences in arrival time of the same emission at different sensors can be used to locate the flaw by triangulation. A computer is used to perform this func-tion using the raw time-of-arrival data recorded during the test, and to drive an X-Y recorder to plot the location of each flaw on a “map” of the vessel. The computeprogrammed to calculate the distances between the flaw and the sensors followthe curvature of the vessel shell, and not as straight lines through the air betweflaw and the sensors.

Fig. 700-25 Characteristics of Acoustic Emission Signal



The above displays of acoustic emission data can be obtained during the test to monitor the results, and can subsequently be recreated from the recorded data for more detailed analyses.

Applications and LimitationsAE can be used for the in-service inspection of pressure vessels to detect cracks that have resulted from any form of deterioration during service that can cause cracking. It can be a very sensitive NDE method for detecting cracks, but the reliability of the test results is highly dependent upon the qualifications of the vendor conducting the test. It makes no difference if the cracks are internal or surface, very fine or filled with corrosion scale, or are in vessel components with complex geometries. However, the deterioration of the vessel must reach the stage where cracks have developed for AE to detect the deterioration. It will not detect the initial stages of creep or hydrogen attack (see Sections 738 and 739) that could lead to the develop-ment of cracks shortly after a vessel is returned to service.

The greatest benefit of AE is that a single test will detect and locate cracks that have developed in any component or weld seam in a pressure vessel. Full coverage in-service inspection of a vessel cannot be obtained with any other NDE method in as short a time. However, it must be understood that AE will detect and locate cracks

Fig. 700-26 Typical Displays of Acoustic Emission Data



s-

sels, ac-ed or

is of

e parts

y

only. It provides no information that can be directly used for a fitness-for-service analysis, or to estimate the remaining life of the vessel. Other NDE methods, such as UT or RT, must be used to determine the size and orientation of cracks detected by AE. Nevertheless, AE is a valuable NDE method because it pinpoints the loca-tions where these other examinations have to be made, and, therefore, can signifi-cantly reduce the total time required for in-service inspection.

The greatest disadvantage of AE is that the pressure vessel must be prepared for a hydrostatic pressure test. This should not be thought of as a deterrent to using AE, particularly if full coverage in-service inspection is desired and the results of the test can be further evaluated with other NDE methods at specific locations.

Alternative procedures have been used to develop a stress sufficiently high to cause acoustic emissions at flaws (such as thermal stresses during cool-down or increasing the operating pressure 10% to 20% during service), but they may not always provide reliable AE test data.

770 Inspection ChecklistsThe following inspection procedure is reproduced from the CUSA Manufacturing Inspection Manual. It may be used as is, or modified to prepare local documents. Note that an API inspection guide is also included in this manual. API 510, “Presure Vessel Inspection Code,” is also included.

GeneralThe general procedure given in the following inspection guide on columns, vesand spheres applies to pressure vessels of all types of construction and serviceconditions. The exterior and interior inspections pertain to all vessels, drums, frtionating columns, packed columns, blow cases, water washes, etc., whether linunlined. Details of inspection will vary for the type of vessel to be inspected.

The inspector should be familiar with the history of the column or vessel that heabout to inspect through a study of the records. He should be aware of the typecorrosion or defects that have occurred in the past and their exact location. He should know the materials of construction as well as the original thickness of thpressure-containing parts, and know the minimum allowable thickness of those as given in the Safety Instruction Sheets or the Manufacturer's Data Sheet.

771 Inspection of Distillation Columns

External Inspection Prior to Entry1. Support and Foundations

Distillation columns are usually supported by a cylindrical skirt which usuallis fireproofed inside and outside with concrete, brick, or other insulating material.



Inspect the fireproofing material carefully for evidence of cracking, spalling, or defective seal at the top edge. Water which penetrates the insulation can cause severe corrosion of the support. Remove small sections of the insulation, if necessary, to permit inspection.

Check the anchor bolts and nuts for corrosion or evidence of failure, and observe the condition of the concrete around the bolts. Cracking of the concrete around the bolts usually indicates corrosion is taking place.

Inspect all foundation support rings, brackets, or lugs for corrosion or indica-tion of distortion or settling. If settling is noted, the equipment should be checked for plumb.

2. Ladders, Platforms, Handrails and Davits

External structures attached to the column for the purpose of servicing the column and providing entry should be inspected with the column. Check ladder supports and clips. Look for missing bolts, broken or weakened handrails, loose toe-boards or defective floor plates. Davits used for lifting materials should be sturdy and not deformed from misuse or overloading. Davits must be load tested. Angle supports where welded to columns are frequently found corroded under the insulation where water collects. Remove the necessary insulation to permit inspection for this condition.

3. Insulation

Note the general condition and effectiveness of the insulation on the shell and heads. Openings around manways, nozzles, and brackets should be sealed to prevent the entry of water. All insulation ties or retainer strips should be intact and tight. Evidence of weathering or cracking of the weathercoat should be noted, and repairs made if necessary.

4. External Corrosion

Inspect the external surfaces of the column where exposed or where corrosion is indicated. Note particularly welds, brackets, nozzles, breaks in the insula-tion, or where water has been allowed to run on the shell, support legs, or skirt. Check the weep holes in the reinforcing pads for evidence of leakage. Gage the areas where corrosion is found and establish the severity, extent, and rate of corrosion.

5. Lines, Instrument Leads and Conduit

Visually inspect and hammer-test all small piping, including instrument piping, level gage connections, vent and drain fittings, and sample connections. Note the condition of all conduit attached to the column or platforms for lighting, instrument, and thermocouple leads. Look for exposed wires, missing cover plates, broken fittings, and improper support.

6. Nozzles

Inspect the gasket faces on all flanges that might be opened, including the manway nozzles and covers. An adequate inspection cannot be made if the



gasket is still in place. Remove it or have it removed. Hammer-test all small nozzles. Inspect the manway cover hinge pins for binding. A frozen pin can prevent tight closure of the cover.

Internal Inspection1. Safety

Entry of a column can be made only after the approved, signed entry tags are in place on the column. Safety rules in effect at the vessel must be observed.

Inspectors must not work alone in vessels but always in pairs to provide maximum safety. Do not attempt an internal inspection until the vessel has been adequately cleaned to permit a satisfactory inspection. If you cannot see it, you cannot inspect it.

2. Shell and Heads

Inspect the shell and heads for defects. Look particularly behind downcomer plates, on the shell opposite nozzles at points of impingement, under nozzles for rundown attack, at the liquid level on the trays, and at shell welds. Corro-sion can appear in many forms, some of which are not readily apparent. A smooth general loss can look like the original surface. Observe for pitting in isolated areas. Measure the depth of the corrosion, if possible, with a depth micrometer and ultrasonically gage the remaining thickness of the shell or heads where corrosion is severe. Measure all fixed gage points using a depth micrometer, or ultrasonically gage at established locations.

3. Column Internals

Inspect all bubble cap trays for out-of-level, and for leaks and holes that could affect the liquid seal on the trays. Check the weirs at the edge of the trays to be sure they will maintain the proper level. Note the condition of the internal tray manways and the gasket surfaces on tray and cover.

Examine all tray support members for mechanical defects or corrosion. If corrosion is a problem, measurements should be obtained to establish corro-sion rates.

Bubble caps, chimneys, bolts and holding members should be tight and in posi-tion. If they are loose enough to rattle, the tray will leak excessively. Note the condition of the caps. Check for corrosion.

Visually inspect and hammer-test all internal piping and spargers. Check spray holes in reflux headers to see that they are not plugged. Inspect the tray and shell carefully in the reflux area as corrosion is frequently severe in this area. Inspect the reboiler baffle for leaks or holes, and ensure that the baffle manway cover fits tightly and is properly gasketed. Check that the vortex strainer is securely in place.

4. Fouling



Note carefully during the complete inspection the degree of fouling in the column. Are the bubble caps plugged? Are the trays covered with scale? If so, to what depth? Are the drains or weep holes plugged? Is the bottom head clean? Internal fouling of this kind can affect the efficiency of the column and make good separation of the cuts difficult.

5. Linings

a. Strip Lining

Special corrosion resistant linings may be installed in columns or vessels where corrosion rates are excessively high. These linings are usually of light gage strips of alloy material welded to the shell. Where lining has been installed it should be carefully inspected for cracks or corrosion. Where lining failures have occurred, representative sections should be removed for gaging of the shell.

Carefully check the condition of all shell and head lining for full protec-tion. If there is any evidence of leakage or corrosion deposits between the lining and vessel wall, request removal of a section of the lining. Recom-mend abrasive blasting the exposed vessel wall and inspect for corrosion. Record the extent of corrosion observed. The lining must be maintained in such condition as to prevent any circulation of stock between it and the vessel wall.

Inspect lined nozzles for bulges, collapse of the liner seams, and for torn or cracked weld seams.

Whenever lining is removed, inspect the metal surface of the column, particularly at the welds, connections, riveted seams, and vacuum bracing for corrosion.

In case of new lining installation or lining repairs, abrasive blast and inspect all column welds and joints before lining, and any which are exposed during repair. Inspect all new installations of lining for workman-ship and condition of welds before returning the column to service.

b. Cladding

Some columns may be protected from severe corrosion by the use of alloy-clad materials. Cladding is a thin alloy sheet, factory bonded integrally to the carbon steel plate.

In those corrosive services where cladding is used, the cladding must be carefully inspected, since penetration to the carbon steel usually results in vessel failure in a relatively short time. Such penetration is sometimes evidenced by rust stains on the cladding.

The remaining thickness of nonmagnetic claddings can be measured nondestructively by the use of the coating gage or the cladding gage. Clad-ding thickness can also be measured with a depth micrometer by grinding through the carbon steel. The interface between the carbon steel and the



cladding material is found by using copper sulfate to plate copper on the steel. This requires welding up the test area, and is not recommended if measurements can be obtained with the coating gage.

c. Weld Overlay

Check weld overlay areas for signs of corrosion, cracking or leaks. Use copper sulfate solution to check the integrity of the alloy overlay if corro-sion/erosion is indicated.

6. Hydrostatic Testing

After repairs have been made to the shell or heads of a column, it may be decided to hydrostatically test the vessel for strength or tightness. Engineering instructions define the procedure for such tests to provide safety and to prevent damage to the equipment.

Be sure that the test pump is equipped with a suitable pressure gage and a safety valve that is properly set and tagged. All air must be bled from the column and the pressure raised at the pump or the column top to the pressure stipulated by the engineer. All pressure-containing parts, particularly at welds, nozzles, reinforcing pad weep holes, as well as at the repaired area are then carefully inspected for drips or leaks or visible signs of weakness.

At the conclusion of the test, the inspector should check to see that the vessel is properly vented to prevent a vacuum and the collapse of the vessel as the water is emptied.

A field hydrostatic test is seldom applied to a whole column due to the great weight of the water, which the foundation might not be designed to withstand. For nozzle replacement or repairs it is sometimes advisable to test the area affected only. This is usually done by applying a welding pipe cap over the inside opening of the nozzle, then pressuring up the nozzle-to-shell weld. Make sure that the pipe cap is removed after the test is completed.

772 Inspection of Pressure VesselsThe methods and techniques of inspection described for distillation columns apply generally to all pressure vessels and should be used where applicable. Other consid-erations will apply as unique designs or materials of construction might dictate.

SafetyBefore entering vessels with agitators, mixers, or other mechanically driven equip-ment, the wiring must be disconnected and the motors tagged out. Properly signed and dated entry tags must be installed on each vessel to be entered.

Inspection1. External Corrosion

Horizontal drums resting on concrete support saddles frequently corrode exter-nally in the area of the saddle. Inspect the water seal at the edge of the concrete.



If evidence of corrosion exists, it might be advisable to lift the vessel off the support or cut away the concrete for more complete inspection. Look for corro-sion on the external surface of vessels where salt water drips on them from overhead coolers or condensers.

2. Packing

Rock, pall rings, or Raschig rings, etc., used to provide large surface area and good mixing in some vessels should be inspected for excessive breakage and fouling. Fouling or plugging of the bed can cause undesirable channeling of the flow through the vessel. Corrosion is usually more severe in a vessel in the area of the packing.

3. Demister Pads

Demisters fabricated from wire mesh pads are installed in vessels to prevent a carryover of liquids. Inspect carefully to see that the supports, ties, and retainers are securely fastened and that they cannot be bypassed. Severe fouling or plug-ging of a demister pad can cause it to be blown out of position.

4. Impingement Plates

Impingement or wear plates in vessels should be inspected for corrosion or erosion, and the attachment welds or bolting checked to prevent loss of the plate. Inspect the shell of the vessel carefully for corrosion adjacent to the impingement plate to be sure that the plate is large enough to cover all of the affected areas.

5. Nonmetallic Lining Materials

Vessels in highly corrosive service may be internally protected by many mate-rials. Among those commonly used are glass, various grades of fiberglass, rubber, plastic and numerous paint-like products.

Glass linings are highly effective to protect against unusually severe corrosion, but are subject to damage by impact or localized temperature changes. Inspec-tion usually consists of a careful search for cracks or chipped areas. Cracks, pinholes, or gas pockets in the glass can result in a leak in the vessel.

Extreme care must be taken during the entry and inspection of glass lined vessels to protect the glass from impact damage. In addition, the vessel shell and nozzles must be protected from external blows or heating. No welding, flame cutting, or hammer-testing is permitted, and the vessel should be so sten-ciled. Temperature limitations may be imposed for steaming, washing, and placing the vessel in service to avoid thermal expansion strains.

All other types of protective coatings should be carefully inspected to be sure that the bond to the shell is tight and for surface deterioration. Fiberglass, plastic, and rubber linings sometimes pull loose from the shell, permitting a corrosive attack on the vessel. Loose pieces of lining material may plug off the outlet of the vessel and cause a shutdown. It is advisable to take a few thick-



ness readings on the coatings of all lined vessels while performing an internal inspection. Most of these coatings will deteriorate at elevated temperatures.

773 Inspection of Reactors

Internal Inspection1. Inspection Ports

Obtain measurements at corrosion reference points on reactor shell. Inspect for corrosion or scale at these locations. Loose scale on the bottom head is prob-ably a combination of scale and catalyst. This scale can cause corrosion and must be removed.

2. Basket Tray

Inspect for basket screen deterioration and plugging. Report percentages of plugging. Plugged screens can be cleaned by hydroblasting. Thin or corroded screens should be replaced.

3. Quench Flex Hoses

Look for broken bonding or severely distorted hoses. If hoses are distorted and bonding is bunched up, replacement will probably be required. Hoses may require hydrotesting to determine soundness. Hydrotest up to 250 psig.

4. Catalyst Support Screens

Inspect for corroded or holed-through areas often caused by hot spots, where the catalyst can migrate through the screen. Do not overlook internal manway cover screen. Inspect screens for plugging, and report the percentage of plug-ging. In most cases plugged screens can be cleaned by hydroblasting. Inspect space cloth where screens have corroded through; any thinning might necessi-tate replacement. Loss of aluminized coating on screens will reduce screen life to approximately 1 year, and screens so affected should be replaced. Obtain a screen sample for a bend and brittleness check.

5. Tray Support Beams

Visually inspect for any distortion or cracked welds. Inspect to determine if each tray (bed) is properly supported. Slight bowing upward was noted on one reactor which was caused from reverse flow through the reactor.

6. Tray Sections

Inspect for cleanliness and condition.

7. Thermowell Bundles

Inspect for external iron sulfide scaling and corrosion. Obtain O.D. measure-ments. Several thermowells have required replacement because of severe external loss.



Pressure test with helium or nitrogen. If pressure drops, leaks can be located with a Delcon ultrasonic translator (noise detector).

8. Quench Lines

Note if they are in position; failure can be expected at threaded connections below support tiers. This has occurred at Richmond.

9. Catalyst Dump Pipe (in sections)

Inspection to be made after removal from the reactor. Look for corrosion, wear, and distortion. Any leak can cause bypassing around catalyst beds, and may require unloading the reactor to repair.

10. Inner Bottle

Inspect wall for scale, pitting and cracks. Any corrosion or excessive scaling should be reported.

11. Ladders

Inspect for condition.

External Inspection1. Nozzles

Visually inspect gasket faces if opened, ORJ gaskets and bolting material. On vessels with pressure sealed gaskets check for distortion, corrosion, and cleanliness.

2. Shell

a. Obtain UT thickness measurements at representative locations on shell and heads.

b. For Isomax reactors, ultrasonically shear wave top and bottom heads, head-to-nozzle welds, head-to-shell welds, nozzles, shell-to-support welds. This inspection is to be performed by qualified personnel using the established procedure. All findings should be recorded on appropriate drawings for comparison with future inspections. This inspection will be performed at scheduled frequencies as necessary.

3. Structure

Inspect for corrosion and spalled concrete. Inspect condition of platforms, stairs, and guard rails. Deteriorated areas should be cleaned and painted.

774 Inspection of SpheresIn general, the practices that apply to pressure vessels also apply to spheres although the techniques may vary. The tall legs necessary to support spheres should be care-fully inspected to determine the condition of the fireproofing and the rain seals.



Remove cracked or spalled fireproofing and inspect the exposed steel. Earthquake or sway bracing must be tight and in good condition.

Inspect the cooling deluge rings on the upper shell, if any, for deposits of rust and sediment. Severe external corrosion can occur at these rings if the drain openings are not kept clean.

The internal surfaces can be inspected by filling the sphere with water and making the inspection from an inflatable raft. Life jackets must be worn while making this inspection. The sphere is usually filled to a point near the top and then inspected and gaged at various levels as the water is emptied.

Voice communication is difficult in a sphere due to reverberation and echoes. Do not take the portable ultrasonic instruments into the sphere on a raft. A large horse-shoe magnet is available for use in the spheres to assist in holding the raft to the shell during gaging operations. All tools and equipment should be tied to the raft to prevent loss. Another method to gage the shell is the use of a magnetic ultrasonic crawler.
