Inspection, Assessment and Repair of Coke Drums and

Peripheral Components in Delayed Coking Units

API Technical Report 934J (Complements API Technical Report 934G)

Date

1
 Table of Contents

1 Scope (Relationship to API Technical Report 934G)

2 References
2.1 Terms and Definitions
2.2 Acronyms and Abbreviations
3 Inspection and Monitoring of Coke Drums
3.1 Inspection for Cracks
3.1.1 Visual Testing Examination (VT)
3.1.2 Liquid Penetrant (PT) Inspection
3.1.3 Magnetic Particle (MT) Inspection
3.1.4 Ultrasonic (UT) Inspection
3.1.4.1 Shear Wave Ultrasonic (SWUT) Inspection
3.1.4.2 Time of Flight Diffraction (TOFD) Inspection
3.1.4.3 Phased Array Ultrasonic (PAUT) Inspection
3.1.5 Alternating Current Field Measurement (ACFM) Inspection
3.1.6 Acoustic Emission Testing (AET)
3.1.6.1 Instrumentation
3.1.6.2 Monitoring Coke Drums with AET
3.1.6.3 Data Recording and Analysis
3.1.6.4 Using Results from AET Monitoring
3.2 Inspecting for Bulges
3.2.1 Use of VT to Find Bulges
3.2.2 Use of Straight Line Reference to Measure Bulges
3.2.3 Use of Laser Profiling to Measure Bulges
3.3 Inspecting for Metal Loss and Cladding Damage
3.4 Monitoring for Drum Bowing, Tilting and Lift at Base
3.5 Use of Strain Gages and Shell Temperature Measurements to Monitor Drum Damage
3.6 Frequency of Inspection
4 Damage Assessment
4.1 Thermal-Mechanical Loading of Coke Drums
4.2 Shell Cracking
4.3 Shell Bulging
4.4 Corrosion and Erosion
4.5 Skirt Cracking and Distortion
4.6 Fire Damage
4.7 Fatigue Life
5 Welding Considerations Associated with Repairs
5.1 Filler Metal Selection
5.2 Electrode Baking
5.3 Preheating
5.4 Controlled Deposition Welding (CDW)/Temper Bead Welding (TBW)
5.5 Post Weld Heat Treatment (PWHT)
6 Types of Repairs Associated with Coke Drums
6.1 Weld Repairs Made from Inside Surface
6.2 Weld Repairs Made from Outside Surface
6.3 Weld Repairs Made from Both Sides
6.4 Replacing Large Sections of Cylindrical Shell Plate
6.5 Skirt Repairs, Retrofits and Replacement
6.5.1 Full Skirt Replacement Around Entire Circumference
6.5.2 Skirt Window Replacement
6.5.3 Skirt Attachment Weld Repairs

2
 6.5.4 Skirt Slot Repairs
6.5.5 Retrofit Designs for Skirts
6.5.6 PWHT of Skirt Repairs
6.6 Repairs and Modifications to Unheading Devices
6.6.1 Modification to Cone Section to Accommodate Unheading Device
6.6.2 Trouble Shooting of Gasket Joints
6.7 Repairs to Feed Nozzles
6.8 Inspection Insulation Windows
6.9 Foundation Repairs
6.9.1 Repairs for Drum Tilting
6.9.2 Anchor Bolt Repairs
6.9.2.1 Anchor Bolt Types
6.9.2.2 Anchor Bolt Distress
6.9.2.3 Anchor Bolt Assessment
6.9.2.4 Anchor Bolt Repair Scenarios
6.9.2.5 Non-Bolted Restraints
6.9.3 Grouting Repairs
6.9.4 Support Structure Repairs
6.10.4 Support Structure Cracking
7 Life Extension
7.1 Temporary Repairs of Cracking
7.2 Repairs Designed to Last an Extended Period of Time
7.3 Use of Structural Weld Overlay to Extend Life
7.3.1 Structural Weld Overlay Methodology
7.3.2 Experience with Structural Weld Overlay
8 Bibliography

3
1 Scope

This document includes information and guidance on the practices used by industry practitioners on the
inspection, assessment and repair of coke drums and peripheral components in delayed coking units. The
guidance is general and does not reflect specific details associated with a design offered by li censors of
delayed coking technology, or inspection tools, repair techniques, and/or engineering assessments offered
by contractors. For details associated with the design offered by a licensor or services provided by
contractors, the licensor or contractor should be consulted for guidance and recommendations for their
specific design details, inspection techniques, assessment procedures and repair practices. This
document is a technical report and as such, provides generally used practices in industry and is
not an API Recommended Practice for coke drums in delayed coking units.

This document complements information contained in API Technical Report 934G which also contains
information and guidance on the practices used by industry practitioners on coke drums and peripheral
components. There is some overlap in the information and guidance provided in this document and in API
Technical Report 934G; however, one of these documents will be considered the primar y source of
information or control document for this information. Table 1 shows which Technical Report will serve as
the primary source of information or the control document for each of the general categories of information.

Table 1: Primary Source of Information or Control Document for General Categories of

Information on Coke Drums and Peripheral Components

Information Category Control Document

General Information on Delayed Coking API TR934G
Results on Past Industry Surveys of Coke Drums API TR934G
Description of Damage Mechanisms for Coke Drums API TR934G
Design API TR934G
Materials Selection API TR934G
Fabrication API TR934G
Operating Practices API TR934G
Inspection Associated with Finding and Characterizing
API TR934G
Damage
Inspection Associated with Repairs API TR934J
Damage Assessment API TR934J
Repairs API TR934J
Welding Considerations API TR934J
Life Extension API TR934J

2 Background

2.1 References

The following referenced documents are cited in the application of this document. For dated references,
only the edition cited applies. For undated references, the latest edition of the references document
(including any amendments) applies.

ACI 201.1R-08 “Guide for Making a Condition Survey of Concrete in Service

ACI 305-10 “Guide to Hot Weather Concreting”

ACI 306R-10 “Guide to Cold Weather Concreting”

4
ACI 364.1R-07 “Guide for Evaluation of Concrete Structures prior to Rehabilitation”

ACI 562-13 “Code Requirements for Evaluation, Repair, and Rehabilitation of Concrete Buildings”

ACI 546.3-06 “Guide for the Selection of Materials for the Repair of Concrete”

API 510, Pressure Vessel Inspection Code: In-Service Inspection, Rating, Repair, and Alterations

API RP 582, Welding Guidelines for the Chemical, Oil, and Gas Industries

API RP 934-C - Materials & Fabrication of 1-1/4Cr-1/2Mo Steel Heavy Wall Pressure Vessels for High-
Pressure Hydrogen Service Operating at or Below 825 deg. F (441 deg. C)

API RP 934-E Recommended Practice for Materials and Fabrication of 1-1/4 Cr- ½ Mo Steel Pressure
Vessels for Service above 825 F (440 C)

API Technical Report 934-G, “Design, Fabrication, Operational Effects, Inspection, Assessment, and Repair
of Coke Drums and Peripheral Components in Delayed Coking Units”

API Proceeding of 1958, Volume 38, Weil N. A. and Rapasky F. S., Experience of Vessels of Delayed Coking
Units

API Proceeding of 1980, Pressure and Tanks Developments, Tomas, J. W., API Survey of Coke Drums
Cracking Experience

API Final Report, 1996 API Coke Drum Survey

API Standard 579-1 / ASME FFS -1, Fitness-For-Service

ASME1 Boiler and Pressure Vessel Code, Section II–Materials; Part A–Ferrous Material Specifications; Part
C, Specification for Welding Rods, Electrodes and Filler Metals; Part D–Properties

ASME B16.5, Pipe Flanges and Flanged Fittings

ASME B16.20, Ring-Joint Gaskets and Grooves for Steel Pipe Flanges

ASME B16.47, Large Diameter Flanges

ASME Boiler and Pressure Vessel Code, Section V–Nondestructive Examination

ASME Boiler and Pressure Vessel Code, Section VIII–Rules for Construction of Pressure Vessels, Division 1

ASME Boiler and Pressure Vessel Code, Section VIII–Rules for Construction of Pressure Vessels, Division 2

ASME Boiler and Pressure Vessel Code, Section IX–Welding and Brazing Qualifications

1 ASME International, 3 Park Avenue, New York, New York 10016, www.asme.org.

5
ASME Post Construction Code PCC-2, Repair of Pressure Vessel Equipment and Piping

ASME SA-20, Specification for General Requirements for Steel Plates for Pressure Vessels

ASME SA-182, Specification for Forged or Rolled Alloy-Steel Pipe Flanges, Forged fittings, and Valves and
parts for High-Temperature Service

ASME SA-193, Specification for Alloy-Steel and Stainless Steel Bolting Materials for High Temperature or
High Pressure Service and Other Special Purpose Applications Specification for Forged or Rolled Alloy-
Steel Pipe Flanges, Forged Fittings, and Valves and Parts for High-Temperature Service

ASME SA-194, Specification for Carbon and Alloy Steel Nuts for Bolts for High-Pressure and High-
Temperature Service

ASME SA-263, Standard Specification for Corrosion-Resisting Chromium Steel-Clad Plate, Sheet, and Strip

ASME SA-264, Standard Specification for Stainless Chromium-Nickel Steel-Clad, Sheet and Strip

ASME SA-335, Standard Specification for seamless Ferritic Alloy-Steel Pipe for High-Temperature Service

ASME SA-336, Standard Specification for Alloy Steel Forgings for Pressure and High-Temperature Parts

ASME SA-369, Carbon and Ferritic Alloy Steel Forged and Bored Pipe for High-Temperature Service

ASME SA-387, Standard Specification for Pressure Vessel Plates, Alloy Steel, Chromium-Molybdenum

ASME SA-435, Standard Specification for Straight-Beam Ultrasonic Examination of Steel Plates

ASME SA-516, Specification for pressure vessel Plates, Carbon Steel, for Moderates- and Lower-
Temperature Service

ASME SA-578, Standard Specification for Straight-Beam Ultrasonic Examination of Rolled Steel Plated
for Special Applications

ASNT2 RP SNT-TC-1A, Personnel Qualification and Certification in Nondestructive Testing

ASTM A380, Standard Practice for Cleaning and Descaling Stainless Steel Parts, Equipment, and Systems

ASTM3 G-146, Standard Practice for Evaluation of Disbonding of Bimetallic Stainless Alloy/Steel Plate for
Use in High-Pressure, High-Temperature Refinery Hydrogen Service

2American Society for Nondestructive Testing, Inc., 1711 Arlingate Lane, P.O. Box 28518, Columbus,
Ohio 43228, www.asnt.org.
3 ASTM International,100 Bar Harbor Drive, West Conshohocken, Pennsylvania 19728, www.astm.org.

6
AWS4 A4.2, Standard Procedures for Calibrating Magnetic Instruments to Measure the Delta Ferrite
Content of Austenitic and Duplex Austenitic-Ferritic Stainless Steel Weld Metal

AWS A4.3, Standard Methods for Determination of the Diffusible Hydrogen Content of Martensitic,
Bainitic, and Ferritic Steel Weld Metal Produced by Arc Welding

ICRI Guideline 310.1R-2008 “Guide for Surface Preparation for the Repair of Deteriorated Concrete
Resulting from Reinforcing Steel Corrosion”

ICRI Guideline 320.1R-1996 “Guide for Selecting Application Methods for the Repair of Concrete
Surfaces”

ICRI Guideline 320.2R-2009 “Guide for Selecting and Specifying Materials for Repair of Concrete
Surfaces”

ICRI 330.1-2006 “Guide for the Selection of Strengthening Systems for Concrete Structures”

NBIC, National Board Inspection Code Part 3, Repairs and Alterations

Society for Protective Coatings, SSPC-SP-10/NACE No. 2, Near-White Blast Cleaning

WRC5 Bulletin 342, Stainless Steel Weld Metal: Prediction of Ferrite Content

WRC Bulletin 452, Recommended Practices for Local Heating of Welds in Pressure Vessels

S.S. Manson, Experimental Mechanics, Volume 5 (Number 7), 1965, page 1932
ASME PVP papers. Ma2JIC

2.2 Terms and Definitions

For the purpose of this technical report, the following terms and definitions apply.

ASME Code

ASME Boiler and Pressure Vessel Code, Section II, Parts A though D, Section V, Section VIII, Division 1 and
Division 2, and Section IX, including applicable addenda and Code Cases.

final PWHT

The last post weld heat treatment after fabrication of the vessel and prior to placing the vessel in
service.

4 American Welding Society, 550 N.W LeJeune Road, Miami, Florida 33126, www.aws.org.
5Welding Research Council, 3 Pak Avenue, 27 th floor, New York, New York 10016,
www.forengineers.com.

7
fracture ductility

Fracture ductility is the term used to define the limiting ductility before fracture occurs as a result of low
cycle fatigue as modeled using the Coffin-Manson equation. It is typically defined as follows:

fracture ductility = ln(100/(100-RA))

where RA is reduction in area during a tensile test

hot forming

Mechanical forming of vessel components above the recrystallization temperature which is well above
the final PWHT temperature.

Larson-Miller parameter (LMP)

Formula for evaluating the effect time at temperature has on mechanical properties from heat
treatment of steel. This same formula can be used to evaluate the effect time at temperature has on
the life of stressed equipment operating in the high temperature creep range.

LMP = T  (C + log t )

where

T is the temperature in K (Kelvin);

t is time in hours.

C is a constant normally equal to 20 for ferritic steels

maximum PWHT

Specified heat treatment of test specimens used to simulate all heat treatments performed on a vessel
including austenitizing, tempering, the final PWHT, a PWHT cycle for possible shop repairs, and a
number of extra PWHT’s to account for repairs in the future.

Note: To determine the equivalent time at one temperature (within the PWHT range), the Larson-
Miller parameter may be used. Results to be agreed upon by purchaser and manufacturer.

minimum PWHT

Specified heat treatment of test specimens used to simulate the minimum heat treatments
(austenitizing, tempering and one PHWT cycle).

Note: To determine the equivalent time at one temperature (within the PWHT range), the Larson-
Miller parameter formula may be used; results to be agreed upon by purchaser and manufacturer.

8
Manufacturer

The recipient of a direct or indirect purchase order for coke drums, materials, fabricated components, or
subassemblies used in the construction of coke drums. In this technical report a direct order is one issued
to a manufacturer by a contractor representing the owner/operator or the owner/operator. An indirect
order is one issued to a manufacturer by a vendor (recipient of a direct order).

Owner/Operator

The owner/operator of the Delayed Coker Unit is located where the coke drums are or will be installed.
The owner/operator is represented by a group of people responsible for the reliable operation of the coke
drums in a specific facility or site.

Owner/Operator’s Quality Assurance & Quality Control Authority

The owner/operator’s technical representative is responsible for implementing and coordinating the Quality
Assurance and Quality Control Program for the construction of coke drums

Shop Inspector

An inspector assigned by the owner/operator’s QA&QC authority to supervise all shop inspection during
fabrication of the coke drum and following Quality Assurance and Quality Control Program.

2.3 Acronyms and Abbreviations

For the purposes of this technical report, the following acronyms apply.

CMTR Certified Material test Report

CDW Controlled Deposition Welding

FN Ferrite Number

HAZ Heat-Affected Zone

HBW Hardness Brinell with tungsten (W) carbide indenter

HV Hardness Vickers

LPWHT Local Post Weld Heat Treatment

MDMT Minimum Design Metal Temperature

MT Magnetic Particle Testing

NDE Nondestructive Examination

9
PQR Procedure Qualification Record

PT Dye Penetrant Testing

PWHT Post-Weld Heat Treatment

RT Radiographic Testing

UT Ultrasonic Testing

VT Visual Testing Examination

WPS Welding Procedure Specification

3 Inspection and Monitoring of Coke Drums

Coke drum inspections are performed to find and measure the extent of damage that occurs over time.
The primary damage mechanisms targeted by inspection and monitoring of coke drums include:

• Cracking

• Bulging

• Metal Loss and Cladding Damage

• Drum Bowing and Tilting

Each of these are discussed separately below, along with drum monitoring and inspection frequency.

3.1 Inspection for Cracks

One of the primary damage concerns for a coke drum is cracking. Cracking can occur at circumferential
and longitudinal welds in the shell, within the shell plate away from welds, at or near the skirt to shell
attachment, or in other crack-prone areas in the skirt such as keyholes and nozzle welds. Cracking also
has been reported to occur in the bottom cone of the drum and at shell external appurtenances.

There are several inspection techniques used to detect cracks whether the cracks are initiating at the
internal surface or at the outside surface of the coke drum. Each technique is discussed in the following
sections.

3.1.1 Visual Testing Examination (VT)

One of the primary methods to detect cracks in coke drums is visual testing (VT) examination. Cracks
initiating from the inside surface can be detected from an internal examination which requires drum

10
entry and scaffolding unless the inspection is performed by video cameras, normally attached to the
coke drill stem. Figure 3.1.1 shows a photo from a VT examination of a coke drum internal surface
showing “elephant skin” cracking. More recently the use of high definition digital technology and
digitally transmitted video has greatly improved remote visual inspection of the inside surface of coke
drums. This has made it significantly easier to detect and determine the length of cracks on the inside
surface. However, it is important to note that even these high quality video examinations of the inside
surface should not be considered a definitive approach for crack detection and sizing. The use of other
inspection techniques as described below are needed to provide a complete inspection of a coke drum.

Figure 3.1.1: VT examination of a coke drum surface displays “elephant skin” cracking.

VT examination for cracks initiating on the outside surface typically requires scaffolding and the removal
of insulation. Since it is generally impractical to remove all of the insulation and install scaffolding to
access to the entire outside surface of the drum, VT examination for cracks initiating on the outside
typically is performed locally where it is known that cracking is likely. Some owners have utilized a
follow up VT examination of cracks with specialized crews using rope access when the drum was empty
during the operating cycle.

3.1.2 Liquid Penetrant (PT) Inspection

Liquid penetrant (PT) inspection can be used to find surface breaking cracks and to determine the extent
of a crack along the drum surface. PT typically is used for non-magnetic materials such as Ni alloy weld
clad restoration welds but also can be used for ferritic steel portions of a coke drum, such as Cr-Mo shell
and 12Cr cladding.

Figure 3.1.2a shows cracking found by PT inspection from the outside surface of a coke drum in the
vicinity of the skirt attachment weld. Figure 3.1.2b shows “elephant skin” cracking by PT inspection
from the external surface of the coke drum shell.

11
 Figure 3.1.2a: PT inspection from the outside surface highlighted cracks in the vicinity of the skirt
attachment weld.

Figure 3.1.2b: PT inspection from the outside surface displays an area with “elephant skin”
cracking.

3.1.3 Magnetic Particle (MT) Inspection

Magnetic particle (MT) inspection can be used to find surface or near surface breaking cracks in suspect
areas and to determine the extent of a crack along the drum surface. MT inspection can only be used on

12
a magnetic material so it is not appropriate for nickel alloy restoration welds or 300 series stainless steel
or nickel alloy weld overlay. Figure 3.1.3 shows cracking at an insulation support clip highlighted by
fluorescent wet MT inspection.

Figure 3.1.3: Cracking on the external surface at an insulation clip displayed by fluorescent wet MT inspection.

3.1.4 Ultrasonic (UT) Inspection

There are several ultrasonic examination techniques that can be used to find and size cracks in drums
and skirts as discussed below. For each UT inspection technique, the surface where the transducers are
placed typically needs to be buffed to a smooth surface in order to insure good contact.

3.1.4.1 Shear Wave Ultrasonic (SWUT) Inspection

SWUT inspection typically is performed from the outside surface to detect cracks initiating at the inside
surface and propagating through wall. SWUT inspection from the outside surface on cracks originating
in the nickel based restoration weld normally is not effective when the crack is propagating through the
nickel based restoration weld. This is not an issue once the crack propagates beyond the restoration
weld and into the base metal. At times it can be difficult to detect cracks that form in the cladding due
to disbonding of the cladding away from the underlying base metal of the coke drum.

SWUT inspection frequently is used to find and size cracks in the vicinity of the skirt attachment weld as
illustrated in Figure 3.1.4.1.

13
Figure 3.1.4.1: SWUT can be used from the outside surface to find and size cracks typically associated
with skirt attachment welds. (Shown as types “A”, “B” and “C” cracks)

Experience has shown that it can be difficult to detect cone side cracks (Type B) using SWUT. In
one incident a crack propagated around the full circumference and through wall despite repeated
focused inspections in the past of the joint using SWUT.

3.1.4.2 Time of Flight Diffraction (TOFD) Inspection

TOFD is an ultrasonic technique used to detect diffracted waves from crack tips and size the cracks
from the arrival times of those waves. TOFD has two major advantages compared with the
conventional pulse echo technique. First, it is weakly sensitive to flaw orientation, whereas pulse echo
relies on specular reflection of the waves. Second, the determination of the flaw size relies only on
being able to measure the arrival time of the signals and not, as with pulse echo, on measuring the
signal amplitudes.

The basic arrangement of the twin probe technique for TOFD is illustrated in the Figure 3.1.4.2a. One
probe (transducer Tx) transmits ultrasound and the other (transducer Rx) acts as the receiver. Probes
with beams separated as much as possible are used to increase the beam coverage. Commented [JM1]: Deleted this sentence because it is more
information than one needs to know. Only need to know that one
wants to maximize the distance between sending and receiving
probes in order to increase coverage near surface where probes are
placed.

14
 Lateral
Wave

Back wall Waves Diffracted

Wave

Figure 3.1.4.2a: Typical transducer arrangement for TOFD inspection of welds.

TOFD has an advantage over conventional UT examination techniques because it can detect the
diffracted signals from discontinuities and/or flaws, process the diffracted signals and compare them
to the back wall signal and structural noise to locate and size the discontinuity and/or flaw. It also
allows the creation of an image which shows the discontinuity and/or flaw location in relation to the
wall thickness.

The TOFD technique provides information not only on the diffracted waves but also on the different
propagation modes such as:

• The lateral wave: This is a propagation wave that follows the shortest path between the
transducers which is the first wave arriving at the receiving transducer. These surface waves
detect surface breaking flaws and indicate the location of the flaws in relation to the
transducers.
• The longitudinal and traverse waves: These waves spread inside the material at different
speeds, sweeping the entire volume, detecting internal discontinuities/flaws and producing a
specular reflection from the opposite surface. This second signal, commonly called a back wall
reflection, reflects off of the opposite surface from where the transducers are located. The
region of the image defined by the lateral wave and the back wall reflection represents a full
thickness cut through the thickness.
• The diffracted and reflected waves: These waves, take place due to the interaction of the
ultrasonic beam with the present discontinuities/flaws inside the material. These signals
arrived in an intermediate time between the lateral wave and the back wall reflection.

The detection and sizing of imperfections between TOFD sensors are in relation to the time flight of
diffracted and reflected signals as the ultrasound beam interacts with discontinuities/flaws. Orientation
of the detected discontinuities/flaws does not have effect on TOFD processing of the signals.

All the signals obtained from the propagation waves, shown in Figure 3.1.4.2a on a two-dimensional
image, commonly known as a B-Scan, is obtained in real time, as the 2 transducers move on the
surface of the component. A typical image is shown in Figure 3.1.4.2b. In this figure, the horizontal
axis represents the displacement of the transducers on the inspection surface. The vertical axis
represents the times of arrival of the signals coming from the lateral wave, the reflected and diffracted
waves for the observed discontinuities/flaws in the material, and the back wall reflection.

15
 External Surface

Crack Thickness

Internal surface

Figure 3.1.4.2b: B-scan image obtained by the application of the TOFD UT technique showing
the presence of an embedded flaw.

Shallow cracks connected to the inside or outside surfaces cannot be detected by TOFD due to dead
zones on the lateral and back wall signals. It is normally recommended that other UT techniques such
as PAUT (See Section 3.1.4.3) and SWUT (See Section 3.1.4.1) be used to detect shallow surface
breaking flaws.

Figure 3.1.4.2c illustrates TOFD inspection results from the OD surface on a clad coke drum
containing a deep crack initiating at the ID surface in the cladding.

Figure 3.1.4.2c: Typical output for TOFD inspection of crack initiating in the cladding on the inside
surface and propagating through wall.

16
Figure 3.1.4.2d: TOFD output for deep crack in coke drum shell.

Figure 3.1.4.2e: TOFD Image from crack in nickel repair weld.

17
3.1.4.3 Phased Array Ultrasonic (PAUT) Inspection

The phased array UT (PAUT) technique is a process wherein UT data is generated by constructive phasing
formed by a single PAUT probe containing multiple elements (e.g., 10, 16, 32, 64 elements) controlled by
accurate time delayed pulses to each element. A PAUT probe can sweep the sound through an angular
range (sectorial or S-scans), at a fixed angle (electronic or E-scans), focus the sound beam with lateral or
line scans, or perform raster scans depending on the array and programming of the PAUT instrument. Each
element consists of an individually wired UT probe with appropriate pulsers, multiplexers, and converters.
Each of the PAUT elements are acoustically insulated from each other. Imaging using a PAUT instrument
includes A-Scans, B-Scans, C-Scans and S-scans. As shown in Figure 3.14.3a, the echo from the desired
focal point hits the various transducer elements with a computable time shift.

Beam
Forming

Time
Delay

Focusing
Principle

Figure 3.1.4.3a: Beam forming, time delay and focusing principle involved with a PAUT probe.

There are several characteristics associated with the application of PAUT as follows:

1. PAUT technology has an advantage over conventional single crystal UT technology in detecting and
sizing flaws of variable orientation. Figure 3.1.4.3b shows difference between a single crystal and a
PAUT in detecting crack-like flaws that are randomly oriented in a steel plate.

18
2. Some of the limitations of PAUT include: Hardware is 10 to 20 times more expensive than conventional
UT hardware. PAUT probes are very expensive. Software controlling PAUT instruments typically
involves costly upgrades.
3. Basic training for PAUT technology is not readily available resulting in a lack of qualified technicians
that are needed for a large scale inspection effort typical for a refinery turnaround.
4. Multiple calibrations required for probe and system plus periodic routine checking for functionality
take a large amount of time.
5. Redundancy of flaw data from PAUT technology makes the interpretation/analysis time consuming.
6. PAUT technology is difficult to integrate into existing standards due to the complexity of this
technology; however, it is more commonly used for a vessel inspection than TOFD.

Single Crystal UT PAUT Crystal Arrangement

Figure 3.1.4.3b: PAUT crystal arrangement provides an improved ability to detect crack-like flaws
compared with single crystal UT.

19
Figure 3.1.4.3c shows the results from a PAUT examination of a Type B crack-like flaw found in the area
of the shell to skirt connection in a coke drum.

Figure 3.1.4.3c: PAUT inspection for a Type B crack at a shell to skirt attachment. PAUT transducers
are placed both on the inside and outside surface of the coke drum.

Figure 3.1.4.3d shows the results of a PAUT examination of cracking (identified as a Type A crack) from
the outside surface of a coke drum at the shell to skirt connection. In this case one transducer is used
for PAUT examination which is placed on the outside surface. This is the same cracking shown by PT
inspection in Figure 3.1.2a. In this case, PT inspection is used to find cracking and PAUT inspection is
used to size it.

20
 Figure 3.1.4.3d: PAUT inspection of cracking (Type A) from the outside surface on the coke drum at
the skirt attachment.

3.1.5 Alternating Current Field Measurement (ACFM) Inspection

ACFM is an electromagnetic technique capable of both detecting and sizing (length and depth) surface
breaking cracks. It is a relatively fast inspection tool that requires access to the cracked surface. Surface
preparation requirements are minimal for ACFM. The basis of the technique involves an alternating
current can be induced to flow in a thin skin near the surface of any conductor. By introducing a remote
uniform current into an area of the surface under test, when there are no defects present the electrical
current will be undisturbed. If a crack is present the uniform current is disturbed and the current flows
around the ends and down the faces of the crack.

Figure 3.1.5a presents a plan view of a short-length surface breaking crack where a uniform AC current is
flowing. The field component denoted Bz in Figure 3.1.5a responds to the poles generated as the current
flows around the ends of the crack introducing current rotations in the plane of the component. These
responses are principally at the crack ends and are indicative of crack length. The field component
denoted as Bx responds to the reduction in current surface density as the current flows down the crack
and is indicative of the depth of the defect.

21
 Figure 3.1.5a: ACFM currents flowing around a surface crack.

For longer cracks, it is not practical to locate the ends of the crack but the Bx field change can still be
measured to determine the depth by moving a probe transversely across the crack.

Special probes have been developed which contain a remote field induction system, for introducing the
field into the surface, together with special combined magnetic field sensors that allow accurate
measurement of the of the resulting magnetic field. Variations of these probes can be used longitudinally
along or transversely across the crack. The probes require no electrical contact with the surface and can
therefore be applied without the removal of surface coatings or grime and without the use of a coupling.
In order for ACFM to size crack depths properly, the user has to ensure the cracks do not contain
electrically conductive or magnetic materials such as wet or dry iron sulfide scales which may be
impractical for an internal surface in a coke drum.

One inspection contractor has developed an automated robotic ACFM device that can be launched from
a drill stem onto the drum shell and then the robotic probe crawler can be steered remotely to permit
internal surface inspection and video recording of not only welds but anywhere on the drum clad surface.
Figure 3.1.5b shows the results of ACFM inspection of cracking that was observed during a remote visual
examination of the inside surface of a drum using a video camera. ACFM determined the length and depth
of the crack that initiated on the inside surface of the drum.

22
 Figure 3.1.5b: ACFM tool determined the depth and length of a crack that initiated on the coke drum
internal surface. A crack was found during the internal visual examination of the coke drum using a
remotely controlled video camera.

Since this method can only detect surface breaking cracks, an externally initiated crack that has
propagated through most of the wall will not be detected using ACFM from the internal surface. This
method is not affected by cleanliness of the inspected surface and can detect internally initiated cracks
that are covered with a thin layer of coke residue.

To assist inspectors with crack detection on circumferential welds, a combined system using Eddy Current
Arrays (ECA) and Alternating Current Field Measurement (ACFM) along with cameras to rapidly and
accurately characterize the presence of surface breaking cracks could be deployed. The ECA eddy current
technique is used for its rapid survey capability and high probability of detection to identify indications,
while the ACFM technique is used to provide depth sizing. Neither ECA nor ACFM require surface
preparation and they both can detect and/or size cracks through a thin layer of coke residue as normally
present after coke is dumped from a drum so they are suitable for remote deployment into a coke drum
without the need for vessel blinding or scaffolding.

23
3.1.6 Acoustic Emission Testing (AET)

Acoustic emission testing (AET) has been used to perform in-service inspections of coke drums. Early
applications of AET on coke drums faced several difficulties such as high background noise levels duri ng
portions of the operating cycle (steaming, drilling, filling, etc.), lack of knowledge about the stress fields
causing damage, and a lack of data processing and filtering capabilities of the old AET instrumentation.
Recent applications of AET for coke drums have provided a more reliable technique to detect and locate
active crack-like flaws. These applications provide the owner with valuable data to optimize inspection,
maintenance and operations of coke drums.

The most important advancement of applying AET to assessing the structural integrity of a coke drum is
based upon combining traditional AET procedures for inspecting high temperature components with
modern high temperature strain gauge technology and skin thermocouples. The data gathered during t he
monitoring of consecutive process cycles includes vital information not only regarding the presence of active
cracks, but also a validation (from the simultaneously collected strain and temperature data) that these
signals were driven by thermo-mechanical loads imposed on the drum during operation.

3.1.6.1 Instrumentation

Monitoring of coke drums with acoustic emission involves the use of a relatively large number of
piezoelectric transducers or sensors. Coke drums are large vessels with diameters in the 15’ t o 30’ range,
and heights varying from 45’ to 90’ (tangent-to-tangent). Sensors are typically distributed over the drum’s
cylindrical shell surface to form the traditional “triangular” array to facilitate source location. Sensor proximity
to circumferential welds is crucial since the vast majority of cracking problems associated with drums are
related to fatigue cracks at these welds.

The drum’s bottom cone requires the use of a separate array of sensors. Sensor distribution is often
completed by installing the necessary number of sensors on the top head (typically semi-elliptical), to cover
its geometric elements such as welds, nozzles and support structures. Typical sensor number to cover a
large coke drum can be as high as 70. Self-calibrating AE probes are very useful in these situations
involving a large array of sensors installed on a large vessel, on top of a tall structure.

During a typical operating cycle, coke drums are heated to 900 oF and cooled to 120oF within a few hours.
These cycles are repeated throughout their entire life accumulating thousands of cycles. The driving force
behind crack initiation and propagation involves complex thermo-mechanical stresses imposed on a coke
drum during a typical operating cycle. As a result, it is important to collect the following data simultaneously
during operating cycles of a coke drum to understand the cause of the acoustic emission.

• AET sensors strategically placed on the coke drum as discussed above

• Skin thermocouples to monitor metal temperatures to better define thermal gradients and resulting
thermal loads
• High temperature strain gages to measure resulting thermos-mechanical loads

AE sensor attachment to coke drums is done by means of spring loaded waveguides to withstand high
metal temperatures during a coke drum operating cycle. Care should be taken when designing and
installing waveguides because of the significant movement experienced by a drum during an operating
cycle compared with the surrounding structure at a constant ambient air temperature. Conventional coaxial
cables (RG-58) are used to bring the amplified signal to the AE data acquisition system.

The AE monitoring system can be located at a number of positions, each with advantages and
disadvantages. Locating the AE system on an upper deck, at mid drum height requires less cabling, but it
presents problems with the “outdoor” conditions in a “classified” area. Installing the AE data acquisition
system on the “skirt deck” has the advantage of better access to operators and other instrumentation.
Longer term monitoring of coke drums often requires the use of conduit to carry coaxial cables back to the
control room, or a nearby building.

24
 3.1.6.2 Monitoring Coke Drums with AET

Coke drums can be monitored over a short period of time during a number of cycles typically varying from
3 to 6. The aim of short-term monitoring is to detect and locate “active” crack-like flaws at that moment in
time in the coke drum. This type of short-term monitoring does not reflect an indication whether the detected
flaw will grow with continued operating cycles.

Longer term monitoring of coke drums can last for months. The aim of long-term AE monitoring is to
determine if there are indications that flaws are growing and pose a potential for through wall penetration
and a leak. Often long-term monitoring is used to help make plans for an upcoming shutdown, better
defining inspection and repair needs for the shutdown.

AE data filtering is important when monitoring coke drums. Several types of noise are created during normal
drum operation. Mechanical, process, atmospheric, structural, electromagnetic interference, amongst
others, are the typical background noises present during data collection. Collecting AE data should be
carefully planned, so it takes place when the combined maximum thermal stresses and minimum
background noise are present.

3.1.6.3 Data recording and analysis

Due to the nature of coke drum cycles, heating and cooling of drums can occur at different times of day
and night. Attending to the data acquisition system(s) during the periods of data acquisition requires close
coordination between the AE crew and unit operators. AE data acquisition systems that can be “triggered”
by other means such as a change in the drum operating cycle or conditions are useful when long term
and remote monitoring is desired.

Data files are identified by cycle number. It can contain AE data from a drum be ing quenched and from
another drum being heated. Data from corresponding thermocouples should be incorporated into the
applicable AE channel data. The quantity of recorded AE, skin temperature and stain gage data can be
overwhelming if a relatively large number of cycles are being monitored. Figure 3.1.6.3 shows a graph
containing AE activity (purple line), skin temperature (red line) and stress from a strain gage (blue line)
over time from Location 1 on a coke drum.

Figure 3.1.6.3: Plot developed from AE monitoring performed simultaneously with strain gage
skin temperature measurements

Analysis of recorded AE data should concentrate on searching for the characteristics normally found in
active flaws, on a “per channel basis”. Typical AE data trends from active flaws include increases in data
rates and amplitudes, and cumulative energy when the thermal gradients are most severe during an
operating cycle.

It is not uncommon to detect AE activity from active flaws in one area during a cycle, but no AE activity
from the same area during subsequent cycles. Consecutive cycles on coke drums are seldom similar or

25
comparable, since the application of thermo-mechanical loads during a cycle at a given location can
significantly vary from one cycle to the next.

Once AE data has been considered relevant, a zonal location can be determined for complementary
follow-up inspection(s) using other NDE technologies discussed in this document. More exact location
methods such as planar and linear positions can assist in better defining a location where follow-up
inspection should be performed during a shutdown period.

3.1.6.4 Using the results from AET monitoring

Results from acoustic emission monitoring of coke drums can be used in several ways as follows:

1. Detect and locate crack-like flaws – As discussed earlier, a short term AE monitoring effort can be
used to find flaws which can be further characterized by other NDE techniques.
2. Define inspection and repair efforts for a planned shutdown – Prior to an upcoming planned
shutdown, AE monitoring can help better define where inspection and repairs are needed on the
drum. A more focused inspection and repair effort during a turnaround can reduce cost for a
planned shutdown.
3. Adjustment of the operating cycle – AE monitoring in conjunction with monitoring skin
thermocouples and strain gages can indicate when during an operating cycle it is most likely that
crack propagation is occurring. Adjustments to the operating cycle can be observed by changes
in the AE signal, and skin temperature and strain gage measurements. For example, increasing
the time for steam cooling prior to introducing quench water into a drum should reduce thermo -
mechanical loads introduced during a water quench, and be observed as a reduction in the AE
signal level, temperature gradient and strain gage measurements.
4. AE monitoring during initial hydrotest - An initial AET during hydrotesting after fabrication of a
coke drum can be used to find and locate crack-like flaws that may not be a cause for rejection
based on requirements of the construction code. However, knowing that a crack -like flaw may
exist in a specific area of a coke drum before going into service provides valuable information to
help plan future AE monitoring and other inspections of the drum.

3.2 Inspecting for Bulges

Bulges occur in coke drums as a result of thermal loads generated during the operating cycle. As the
operating temperature cycles during an operating cycle, the metal temperature from one location on
the drum to another varies at points in time. As a result, one section of the drum will expand or shrink
relative to other sections of the drum. Over time this will result in permanent deformation or bulges in
the drum. With the exception of few cases such as plastic collapse in the skirt, in general, bulges do not
affect the structural integrity of a coke drum, but it does provide an indication of high thermal loads and
a higher possibility for cracking in the bulged section of the drum. Additionally, it is important to
monitor the extent of bulging over time. This provides an important indication of how the deformation
is accumulating over time and increasing the likelihood for cracking that could lead to a leak.

Each of the inspection procedures for bulging provides information in assessing damage and how it
accumulates over time. It is an essential part of optimizing coke drum reliability.

3.2.1 Use of VT to Find Bulging

VT is an important first step that can be used to find bulging. Since a coke drum is insulated on the
outside surface, VT typically is limited to internal inspections. It can be performed using a video camera
or with the naked eye using oblique lighting as shown in Figure 3.2.1a. The major drawback of a VI
inspection is that multiple bulges can be time consuming to accurately document.

26
 Figure 3.2.1a: Bulges in a coke drum found by internal VT examination are highlighted using oblique
lighting.

3.2.2 Use of a Straight Line Reference to Measure Bulges

A rigid four to eight-foot-long straight edge that has a ruler along one side, such as a construction
bubble level, can be used to measure bulge vertical height. It can also be used to as a straight-line
reference to measure the bulge radial depth as shown in Figure 3.2.2a. After placing the straight edge
against the drum wall, a six inch ruler or digital caliper can be used determine the distance to the
maximum bulge depth.

27
 Figure 3.2.2a: Use of a straight edge from the ID and OD to measure bulging.

3.2.3 Use of Laser Profiling to Measure Bulges

Today, laser scanning tools frequently are used to provide a 100% dimensional measurement on a coke
drum. Lasers can provide very accurate dimensional measurements and determine if a bulge is developing
at the initial stages of bulge formation. It is common that laser tools are mounted on the drilling stem or
the inlet/outlet nozzles on the drum to provide a complete surface profile. Several owners perform a laser
profile prior to or soon after the initial commissioning of the drum to acquire the baseline measurements.
Having baseline measurements improves the accuracy of future measurements and the ability to detect
the early onset of bulging.

Figure 3.2.3a shows a flat surface representation of laser profile data taken on a coke drum on 2 separate
occasions over a 5-year period. It shows that bulging got progressively worse over this period in the areas
highlighted with a purple circle.

28
 Figure 3.2.3a: Laser profiling measurements show the growth of bulging over a 5-year period.

Figure 3.2.3b shows a vertical section plot and polar plot from laser profile data taken on 5 separate
occasions over a 6-year period.

Figure 3.2.3b: Laser profile data is showed for 5 sets of measurements in a vertical section
comparison and a polar plot comparison.

29
As mentioned earlier, the presence of a bulge provides an indication of whether cracking will eventually
occur on the drum. Figure 3.2.3c shows how laser profiling found a very sharp and deep distortion in
the middle course of two different drums. A through wall crack was reported in the first case while
severe elephant skin damage was registered in the second case.

Through wall crack at a inward bulge

Elephant skin at a outward bulge

Figure 3.2.3c: Laser profiling detected a sharp deep distortion at middle course of two different
drums.

Attention must be paid to the use of data smoothing techniques such as moving-average filters to filter
out measurement noise and instrument vibration. This practice can produce smoother scans and
eliminate or minimize high-frequency components noise that would otherwise appear as local saw-tooth
artifacts, particularly in severity analyses that utilize 2nd derivatives of bulge profile. However, the
undue use of smoothing techniques can result in an artificial flattening of bulges and, consequently, an
unrealistic reduction in their severity. To ensure accurate assessment of bulges, techniques for
removing scan platform position uncertainties and motion must not rely solely on smoothing. These
techniques can be applied during post-processing of data and can include motion frequency analysis and
oversampling.

30
3.3 Inspecting for Metal Loss and Cladding Damage

Metal loss in the form of general corrosion and pitting has been observed on internal coke drum
surfaces. Cladding damage such as disbonding and detachment from the drum wall also has been
observed. Typically, the optimal way to detect metal loss or cladding damage is by an internal visual
inspection using a video camera. This is routinely performed using a video camera attached to the drill
stem as discuss earlier. UT measurements taken from the outside of the drum are not as efficient,
largely because insulation must be removed and scaffolding needs to be installed.

Figure 3.3a shows a collection of photos from a video camera examination of the internal surface of a
coke drum illustrating the various forms of damage that has been found.

Figure 3.3a: Photos from a video camera show internal damage in coke drum

31
3.4 Monitoring for Drum Bowing, Tilting and Lift at the Base

Coke drum tilting is caused by a combination of grout/shim/anchor bolt failure and skirt deformation
due to uneven loading on the base plate resulting from uneven thermal expansion and contraction of
the coke drum during the normal operating cycle. Bulging in the shell, caused by local yielding due to
high thermal gradients will exacerbate the situation by effectively shortening the vertical wall of the
drum. Due to the nature of coke drum operations and the limitations of today’s technology, these
disproportionate temperature effects will continuously impart strains and cause subsequent
deformation of the drum and movement of the baseplate resulting in coke drum tilting.

The extent of coke drum titling can be monitored using different tools as discussed previously in
paragraph 3.2.3. The information collected should be used to track its progression, making sure it does
not go beyond acceptable limits. Figure 3.4a shows a practical example of tilting monitoring in two coke
drums over a period of 5 years. Results of the measurements taken in these drums identified as Drums
A and B, indicate that tilting progression rate was similar in both drums with Drum B exhibiting the
worse titling. Measurements taken until 2013 by the owner/operator, raised some concerns in regard
to drums stability and operability due to drill stem misalignment preventing drilling operation after
quenching the coke with water. As a consequence, an engineering study was performed to establish
acceptable limits. Results of the study indicated that a tilting less than 15 inches would be acceptable in
terms of structural stability, while tilting limits based on operation of the drill was lower at 13.8 inches.
The results of this assessment along with tilting measurements taken on the 2 drums over a 5-year
period are shown in Figure 3.4a.

Figure 3.4a: Monitoring of tilting progression in two coke drums and comparison with site defined
acceptable limits.

32
An evaluation of the measured tilting in Drum A and Drum B indicated that the major contributor was
asymmetrical bulging of the drums. In this analysis, results of a laser scan inspection of the shell section
were used to analyze several sections of the drum, comparing the amount of measured tilting with the
degree of shell shortening. A schematic representation of the results is shown in Figure 3.4b. The
sketch in this figure represents the degree of shell shortening (green line) in one of the drums which was
more severe in the Southeast side, same direction of permanent tilting (blue arrow). This relationship
between shortening and tilting direction was observed in both drums.

Figure 3.4b: Comparison of tilting and shortening in a coke drum. In the right figure the green line
represents the degree of shell shortening and the red arrows represents the degree of tilting. The
direction of permanent tilting is represented by the blue arrow.

Figure 3.4b shows a schematic representation of the results, which are the degree of skirt ovalization
and deformation of the skirt base ring. Over years, uneven bulging of a drum can lead to ovalization.
Measurements taken on the skirt shows that the ovalization was oriented perpendicular to the direction
of tilting as indicated in Figure 3.4b. Lifting of the base ring was also measured and results are
represented in the sketch of Figure 3.4c with the green lines. These results showed that the
deformation of the base ring in both drums was most likely caused by the skirt ovalization.

In the analysis, contribution of other factors such as deterioration of the foundation (grout and
concrete), anchor bolt failures, coke drum piping loads, etc. were also considered, and it was concluded
that they do not significantly contribute to tilting.

In summary, results of the measurements made and analysis conducted indicate that non-uniform
plastic deformation (bulging) in both drums leads to tilting in a preferential direction which leads to
ovalization of the skirt perpendicular to the tilting direction and lifting of the base ring.

33
 Lifting of the base
ring pointed out by
red arrow

Measurements of
lift in the drum
base plates

Figure 3.4c: Measurements taken on 2 coke drums show ovalization and the resulting lift at the
base ring.

34
3.5 Use of Strain Gages and Shell Temperature Measurements to Monitor Drum Damage

High temperature strain gages have been attached to a drum shell and skirt at the attachment weld to
measure strain during the operating cycle. Strain gage measurements taken on a drum and skirt
indicate that the equivalent stress on the shell, shell to skirt junction and bottom cone can exceed yield
during the water quench. Similarly, measurements made on the shell to skirt junction shows stress
levels exceeding yield during rapid heating while filling an empty drum with feed. Strain gages indicate
average strain over the gage length. They do not accurately reflect stress levels that exist at welds and
other areas of stress concentration. Engineering judgment and analysis are required to translate the
measured cyclic stress range appropriately when evaluating stress at welds and other stress
concentrations on a drum.

Placing strain gages near cracks may create inconsistent results because of the effect the crack has on
the compliance of the material immediately surrounding the crack. Statistical histograms of the stress
ranges derived from strain gage measurements may be used to characterize damage at instrumented
locations. Thermocouples have been attached to the drum shell to measure temperature difference on
the shell during the operating cycle and provide an indication of the thermal loads that can exist. Large
differences in temperature, up to 500°F (278°C), have been observed over short distances on the drum
vertically and circumferentially. These indicate that thermal loads can exist and that in some cases may
cause local stresses in the area of a weld, stress concentration or even base metal to exceed the shell
steel’s yield strength. Therefore, multiple cycle testing is needed to more fully characterize the stresses.
Statistical and probabilistic techniques have been used to evaluate the data.

Rapid cooling rates are a strong indicator of trouble. High cooling rates will contribute to non-uniform
cooling and pronounced local hot zone/cold zone issues. In many instances of modern shot coke
operation, local areas on the steel shell are being quenched rapidly compared to other areas on the shell
due to water channeling through the coke bed to the shell.

Experience of some operators has shown that the use of skin thermocouples and/or strain gages can be
useful in identifying operating practices that produce the highest thermal stresses and damage on a
coke drum. Several operators performed continuous temperature and strain gage monitoring on a
bottom cone and found that by modifying the introduction of feed and quench water during the various
times of an operating cycle, both temperature differences and stresses from the strain gage
measurements could be reduced. They were able to develop improved operating procedures that
allowed them to significantly reduce the cycle time and simultaneously reduce the thermal stresses and
resulting damage imposed on the bottom cone and drum shell during each cycle.

3.6 Frequency of Inspection

In general, new drums do not require an initial inspection until after the first 4 to 6 years of service.
However, many operators are using laser profiling or mapping in their new drums to have a baseline of
the shell section before placing drums in service. Sometimes, baseline scans reveal the existence of
initial bulges formed by the fabrication process. After an initial inspection, subsequent inspections will
need to be scheduled based on the level of cracking and bulging that has been observed during the

35
 initial inspection and the severity of operation since the last inspection. Each drum in a coking unit may
need to have a different inspection plan, because experience shows that each drum can display a unique
damage pattern over time.

Table 3.1 provides an example of guidance on use and frequency for each of the inspection techniques
discussed earlier. This table does not reflect any specific inspection requirements. It indicates general
industry experience as reflected in the survey results as discussed in Section 4.2 of API Technical Report
934G. It is important that owners/operators at each location develop a comprehensive inspection plan
tailored for each drum.

The plan should specify the type of inspection which should be performed, and the timing of future
inspections based on the results from previous inspections. Adjustments to the inspection plan should
be made based on the drum operating conditions (for example changes in the cycle time and specific
steps in the operation such as the addition of quench water and addition of hot feed to an empty drum).

Most operators make it a practice to update the inspection plan on a regular basis to ensure it reflects
current information on the drum and anticipated changes to the operation of the drum. Experience has
shown it is prudent to perform an update on the inspection plan well before a planned turnaround so
that onstream inspections before the turnaround can be conducted and the required downtime
maintenance during the turnaround can be better anticipated.

Table 3.1: Listing of typical onstream and downtime inspection techniques for coke drums and
frequencies for use of the inspection technique

On stream
Inspection Off-stream
(non TA) Typical frequency Typical frequency
technique inspection
inspection

Spot skirt attachment weld and

spot skirt keyhole area once a
ID surface, skirt attachment, skirt
year. ID surface approximately 6 Yes
keyhole area, and previously
VI Yes months before TA, during last
identified cracking areas every
de-coke before TA. Once
T/A.
cracking detected, as
appropriate to situation.

Used for follow-up to known or

Used for follow-up to known or
suspected areas identified by
suspected areas identified by
Yes visual or other means to confirm
MT Yes visual or other means to confirm
and determine length of cracks.
and determine length of cracks.
External welded attachments
every T/A

PT No Yes Used for follow-up to known or

suspected areas identified by

36
 visual or other means to confirm
and determine length of cracks,
especially in non-magnetic
materials such as Ni-alloy weld
metal on ID circumferential welds
and overlay on ID.

For crack-depth determination

after cracks are detected by visual
UT or other means.

For crack-depth determination Yes To inspect drums from ID to detect

(Shear wave
Yes after cracks are detected by and size cracks on OD where
UT, TOFD,
visual or other means. external attachments are located
and Phased
Array UT) on bulge peaks without removing
insulation.

Approximately 6 months before

TA, during last de-coke before
Video camera Yes No
TA. Once cracking detected, as
appropriate to situation.

At the skirt and drum surface to At the skirt and drum surface to
ACFM Yes measure crack depth once cracks Yes measure crack depth once cracks
are detected are detected.

To be considered only as a
supplementary global inspection
Acoustic
Yes tool for locating and monitoring No
Emission
active cracks. Not to be used as
sole inspection tool.

If desired, approximately 6
months before TA, during last
de-coke before TA. Once bulging
Yes detected, as-desired and No
LASER appropriate to situation.

37
4 Damage Assessment

4.1 Thermal-Mechanical Loading of Coke Drums

Typically, most coke drums are designed to ASME Boiler and Pressure Vessels Code Section VIII, Division
1. However, assessment of coke drum damage is complex to perform because coke drums are exposed to
significant thermo-mechanical loads that cannot be predicted or completely quantified. In most cases,
highest thermo-mechanical loads occur during the quench portion of an operating cycle when the shell
metal temperature can drop by over 300oF, as water is injected into the drum to cool the hot coke before
it is dumped. Thermal loads are generated as water randomly channels through the coke bed and
quenches localized areas on the shell interior surface. Mechanical loads may possibly be generated from
the resistance of solid coke to the contraction of the cooling shell. Thermal-mechanical loads may be
generated from relatively large temperature gradients across and along the coke drum walls. The
magnitude of these thermo-mechanical loads at any given location cannot be predicted because quench
water randomly reaches different areas on the drum shell during each cycle. Since water channeling takes
place in the shell below the coke fill level, the lower half of the drum is most affected. These random
thermo-mechanical loads are significantly more influential on causing damage than design loads such as
internal pressure. Skin thermocouples placed on the drum surface can provide significant insight on the
actual thermal distributions experienced by coke drums and may help with the assessment of damage
through characterization of loading [1].

In addition to the thermo-mechanical loads generated during quench, there can be significant thermal
loads generated during the fill part of the cycle. During injection of hot feed into a drum, thermal loads
are generated at the fill line as it moves up vertically on the drum. This load can be particularly significant
in the cone if hot feed impinges at high speed against one side of it. In addition, thermal transient loads
can generate high stresses at the skirt to shell weld due to the differential temperature between the skirt
and drum. While the skirt does not directly experience the heating and quenching occurring inside the
drum, the delay or lag in heat transfer to the skirt can create a significant thermal stress at the welded
joint between the skirt and the drum.

In addition, depending on the combination of inlet nozzle configuration, coke morphology and quench
procedure, bulging and deformation may occur preferentially on one side of the drum, resulting in a
“banana” effect. This effect, and its consequences on the stability of the drum, are discussed in Reference
[1].

The main effect of drum tilting and bowing in the shape of a banana is that it adversely affects an
operator’s ability to cut the coke during the coke removal portion of the operating cycle. Assessment of
drum stability due to tilting or bowing is extensively addressed in Reference [2]. Challenges associated
with performing damage assessments on coke drums are discussed further below. Such assessments
require specialized technical expertise and experience with coke drums.

38
4.2 Shell Cracking

Cracks form in coke drums due to a variety of reasons, but in general are due to thermal fatigue. Cracking
with the typically observed “elephant skin” appearance or multiple crazed cracking appearance has been
observed originating from both the ID and OD surfaces of coke drum shells. Inside surface tensile stresses
will be created when a cooler liquid contacts a warmer shell surface, whereas outside surface tensile
stresses will be created for the reverse case (warmer liquid contacting a cooler inside surface). The
magnitude of the temperature difference between the fluid and metal will directly determine the
magnitude of these stresses. While the location of tensile stress often correlates with cracking, crack
formation can also be strongly influenced by weld quality, local stress concentrations, and strength
mismatch. Even coke type and quench details can have a strong influence on where cracks form in
practice. For example, both hot and cold spots can form during a given quench based on the variable path
of quench water.

Cracks at bulges are primarily initiated by excessive cumulative strain (local failure) due to repeated high
thermo-mechanical loads. The existence of bulges creates additional stresses that can elevate strain levels
and secondary stresses due to bending moments created by the deflection of the bulged shell. These are
directly additive to thermal and coke resistance stresses resulting in complex stress fields. Regardless of
location, cracks propagate due to fatigue from cyclic stresses. Even in areas away from bulges or in older
un-bulged vessels with gentle operating cycles, fatigue cracking will eventually occur once enough cycles
are reached.

Crack assessments for coke drums are challenging in that mechanical properties of aged drum materials
such as fracture toughness and crack propagation model constants are difficult to determine. This
uncertainty is over and above the loading uncertainty discussed in Section 4.1. In the absence of
mechanical and metallurgical testing, the API 579-1/ ASME FFS-1 Standard provides general material
properties guidance that may be used to produce conservative results.

4.3 Shell Bulging

There are several bulging types in coke drums most of which occur as a result of high thermo-mechanical
loads generated during water quenching. Occasionally, bulges may also be present from fabrication. A
detailed discussion of bulging types and contributing load mechanisms is found in Reference [1]. It is
generally believed that bulging and bulging-induced cracking is a function of stress, low cycle fatigue
strength and ductility of the material.

The assessment methods of Part 8 of Fitness-for-Service Standard API 579-1/ASME FFS-1 are not
commonly used to evaluate the fitness-for-service of coke drum bulges because of the loading
uncertainties described in Section 4.1. In addition, the process of simulating the development and growth
of numerous interconnected bulges in a drum, as required for a Level 3 assessment of thermo-
mechanically induced shell distortions, is challenging. Because of these difficulties, over the last few
decades, industry has utilized approximate methods to assess the severity of bulges in the following order:

39
1. Monitoring of bulging magnitude

The monitoring of bulging magnitude is conducted by trending the maximum radial growth of the
shell as measured manually or by internal laser scanners. This process, which provides a qualitative
description of relative deterioration in drums, cannot be used for assessment of specific bulges. The
1996 API survey showed no correlation between cracking and depth of bulges [4] .

2. Calculation of stresses or strains using finite element modeling

The calculation of stresses or strains using finite element modeling has been utilized using one of
the following approaches:

a) An FEA model is developed using the actual measured bulge geometry. A stress analysis is
performed using internal pressure as the only load. Thermal loads are not included in the analysis.
Stress fields in and around observed bulges under pressure loading are examined and rated. See
Ref. 3.
b) An FEA model is used to create the observed bulge by plastically deforming the shell which
determines plastic strain levels corresponding to the observed bulge. Skin thermocouple data is
applied to calculate thermal loads, and in addition to internal pressure, determines the
operating stresses in the bulged region. Fatigue life is estimated using appropriate low cycle
fatigue data. In lieu of actual field thermocouple data, thermal analysis can be substituted for
an approximate solution.

3. Geometric methods

The shape of bulges has been used as an indication of severity:

a) Geometric pattern recognition methods use known bulging shapes associated with bulge cracks
to assess the severity of bulges. The application of this assessment approach is described in
Reference [5].
b) Dimensional measurements such as depth-height ratio have been used to screen for severity.
c) Sharpness of a bulge is a characterization used to screen for damage in coke drums. This approach
utilizes the second derivative of the vertical profile of curvature of the bulge which is directly
related to strain (and therefore stress). The computed sharpness value is scaled to fit within a
range of 1 to 10 covering a no bulge condition (1) to extremely tight bulges (10). This
categorization considers various factors including but not limited to drum metallurgy, diameter,
wall thickness and the location of the bulge (mid-course vs circumferential weld). The values are
further grouped into five categories of increasing likelihood of surface damage with bulging
located on circumferential welds showing higher levels of damage at lower sharpness values when
compared to mid-course deformations. See reference 30.

4. Calculation of plastic strains

40
 Based on the observation that bulge-induced cracking is initiated by plastic strain, the calculation of
plastic strains from distorted geometry has been used to quantify bulging severity and identify
bulges that are most likely to develop cracks. The strain-based methodology identifies and ranks
areas that are most susceptible to local failure using strain limits provided by API 579-1/ASME FFS-1
Standard that were calibrated using a database of known internal and external failures. The strain
measure at any point on the drum is the ratio of calculated plastic strain to the calibrated strain limit
in a percentage form. Ranking of severity between a design threshold and strain limit is defined
using a severity system that specifies failure initiation on the inside and outside surfaces of the wall.
This measure is also used to determine needed frequency of laser scanning. A description of this
method, its correlation with observed bulging-induced cracks, and a comparison with the stress
concentration method are provided in Reference [7]. Note that while the strain limits provided by
API 579-1/ASME FFS-1 are related to (cumulative) ductility exhaustion and not the assumed primary
damage mechanism of fatigue, the limit does incorporate triaxial stress state effects on ductility.

4.4 Corrosion and Erosion

Some coke drums experience general and/or localized wall loss from corrosion that can lead to
significant metal loss on the 12Cr cladding and underlying base metal. Metal loss in the bottom cone
section of a coke drum has been attributed to solid particle erosion. Such erosion is observed most
frequently in coke drums that produce “shot” coke. It is believed that this erosion is caused by the coke
rapidly exiting the drum during coke removal, resulting in coke particles abrading against the bottom
cone surface.

To verify compliance with design requirements for minimum required wall thickness, general and
localized metal loss above the maximum coke fill level (that has no bulging) can be assessed by Parts 4
and 5 of API 579/ ASME FFS Standard, respectively. For the rest of the drum, however, the load
definition difficulties discussed above make this type of damage difficult to assess.

Some coke drum owners have installed Alloy 625 cladding and/or weld overlay to minimize both
corrosion and erosion in coke drums. Alloy 625 has reportedly improved corrosion and erosion
resistance over 12Cr which is the standard cladding alloy used on coke drums.

4.5 Skirt Cracking and Distortion

As discussed above, welded skirts typically experience cracks in the vessel-to-skirt attachments. Cracks
on the skirt side of the attachment often grow through-wall and eventually around the entire
circumference of coke drums. Cracks on the shell and cone sides are also possible and can have more
serious consequences due to possible loss of containment. Skirts that are designed with slots to
minimize cracking at attachment welds tend to develop cracks at the top keyholes of slots.

In addition to weld attachment cracks, skirts can develop plastic distortions that may be related to upset
loading conditions, installation damage, and/ or design and fabrication issues. While skirt distortions are
uncommon, it is one of the most dangerous types of damage in coke drums because it can potentially

41
lead to instability and catastrophic collapse. It is also a problem that can be exacerbated by the severity
of the cyclic operating condition.

The presence of any tilting or leaning of the drum may have a significant adverse impact on stability,
especially in slotted skirts. As discussed in API 934G, it is important that the drum and skirt-to-shell
attachment be fabricated to tight tolerances in order to minimize eccentric loads.

Rigorous assessment of skirt distortion starts with an accurate description of the distorted shape. Both
manual and laser-based measurements have been used to provide this data which is utilized for building
a numerical model, per Level 3 procedures of Part 8 of the API 579-1/ ASME FFS-1 Standard. However,
since vessel supports are not typically designed per the ASME BPV Code, an examination of structural
safety margins must be conducted to ensure compliance with applicable design standards such as
Reference [8].

4.6 Fire Damage

Coke drums can be exposed to heat from an internal or external fire. Metallurgical damage can be
assessed using Part 11 of API 579-1/ASME FFS-1. Shell and skirt distortions may be evaluated as
described in the other parts of Section 4.

4.7 Fatigue Life

Improvements in fatigue life may be achieved using design and fabrication enhancements described in
API 934G Sections 5 and 6 as well as this document.

Fatigue life is not easily determined mainly because of load uncertainty, as described in Section 4.1.
Methods of fatigue analysis are discussed in more detail in Part 14 of API 579-1/ASME FFS-1 which may
be used to assess fatigue performance at some locations such as the skirt junction. The Structural Stress
method and other analysis techniques have been utilized to evaluate fatigue life. Examples of shell
fatigue assessment studies are found in References [1], [9] and [10]. A study that compares various
techniques for calculating fatigue life at the skirt attachment is found in Reference [11] fatigue life is
typically determined from the standard fatigue curves in API 579-1/ASME FFS-1, or by using more
specific data such as that found in Reference [12].4.7 Tilting and Bowing

5 Welding Considerations Associated with Repairs

This section provides the general welding practices that have been used in industry when performing
repairs to coke drums. These practices are commonly employed in making repairs as more specifically
described in Section 6.

5.1 Filler Metal Selection

The base material for coke drums varies from carbon steel to Cr-Mo grades, which are typically clad on
the ID with a 12Cr steel for resistance to sulfidation. The two main criteria in selection of a welding filler
metal are to match the mechanical properties and chemistry of the base metal. Table 5.1 provides
typical filler metal for each base metal per the intended field welding process. Once the base metal

42
chemistry and mechanical properties are determined, the filler metal selection is narrowed to a limited
number of choices. One vital choice pertains to the selection of welding consumables for the P-4
materials. For repairs where the drum will be post weld heat treated (PWHT’d), the straight carbon
grade consumables are normally used because they will have matching strength properties as the base
metal in the PWHT’d condition. However, if the repair is not PWHT’d, then the low carbon or “L” grade
consumable is normally used in order to minimize the hardness of the weld metal and provide a closer
match of the strength of the existing PWHT’d welds and base metal. Performance characteristics of each
filler metal will vary from manufacturer to manufacturer and as a result a fabricator may have a
preference to use one type of electrode over another.

Table 5.1
Typical Filler Metal Choices for Coke Drum Repairs(1)

Base SMAW(3) GTAW/GMAW FCAW(2)(3)

Material

P-1 E7016, E7018 E(R)70S-2 E71T-1, -5,

Materials or -9 or -12
(CS) E7018-1

P-3 E7016-A1 or E(R)70S-A1 E71T5-A1

Materials E7018-A1
(C-Mo)

P-4 With PWHT With PWHT With PWHT

Materials E8016-B2 or E(R)80S-B2 E81T1(or 5)-B2
(1 or 1 ¼ Cr) E8018-B2 Without PWHT(4) Without
Without E(R)70S-B2L PWHT(4)
PWHT(4) E81T1(or 5)-
E7018-B2L B2L

P-5A With PWHT With PWHT With PWHT

Materials E8016-B3 or E(R)80S-B3 E81T1(or 5)-B3
(2¼Cr) E8018-B3 Without PWHT(5) Normally
Without E(R)80S-B3L requires
PWHT(5) PWHT(5)
E8018-B3L

Notes
1. Table lists filler metals designated with US Customary Units. Filler
metals designated with metric units have not been shown, but can be
used.
2. FCAW electrodes should be used with an Argon/CO2 mixed gas and
therefore should be designated with the supplemental prefix “M”.
3. SMAW and FCAW electrodes should be specified with the
supplemental suffix “H4” which designates the diffusible hydrogen level

43
 of the deposited weld metal. The “H4” designation means that the weld
deposit will have an average of 4 mL (H2)/100g of metal.
4. For controlled deposition welding applications, the low carbon grade
welding consumables should be used because the resulting weld
deposits will be lower in hardness and easier to temper by subsequent
weld passes.
5. Some companies perform weld repairs on 2.25Cr without PWHT
using a temper bead procedure while others require PWHT because of
the high hardness levels produced.

If the inside surface is clad, base metal repairs are often followed by an in-situ back cladding using a
nickel based filler metal depending on the depth and extent of cracking. ENiCrFe-2 (INCO-WELD A) is a
common coated SMAW electrode used for clad restoration welds, while ERNiCr-3 ( INCONEL 82) filler
wire can be used with the GTAW or GMAW process. FCAW consumables similar in composition to
ERNiCr-3 , such as ENiCr3T0-4, are also now available. Alloy 625 welding consumables (ENiCrMo-3 for a
SMAW coated electrode and ERNiCrMo-3 for GTAW and GMAW wire) also are frequently used for
restoring cladding on repair welds. However, some owners do not use Alloy 625 welding consumables
for this application because it is very strong and over-matches the mechanical strength of the base
metal and 12Cr cladding.

Nickel based filler metals have a thermal expansion just slightly greater than that for carbon steel and
are therefore better suited than stainless steel electrodes like E309L. The difference in thermal
expansion is less than 5% between carbon steel and these nickel alloys. For this reason, nickel based
filler metals are a good choice for restoration of the ID cladding after performing through thickness
repairs.

ENiCrFe-3 (INCONEL 182) also has been used for cladding restoration of coke drum repairs. ENiCrFe-2
and ERNiCr-3 are considered the best welding consumables to restore the cladding at repair welds
because they best match the strength and thermal expansion properties of the base steels (CS, C-Mo or
Cr-Mo) and 12 Cr cladding. ENiCrFe-3 typically is not recommended because of its tendency to embrittle
at elevated temperatures and its poor sulfidation resistance above 750°F (400°C). The Alloy 625 welding
consumables have greater sulfidation resistance above 750°F (400°C) than the other nickel alloys due to
its higher chromium content, and would therefore seem to be the ideal choice. (Coke drum wall
temperature ranges in service from 820°F (440°C), when exposed to coke drum vapor, to 730°F (390°C)
when covered by coke.) However, the strength of an Alloy 625 weld deposit highly mismatches the
strength of the low alloy steel base metals and 12 Cr cladding, making it less than ideal for cyclic service.

In some cases, owners have not restored the back cladding weld after a repair to a weld seam using one
of the ferritic welds listed in Table 5.1. In all reported cases, metal loss due to sulfidation of the ferritic
repair weld and the adjacent base metal has been insignificant. It is believed that coke fills the space
formed by the missing back cladding weld and provides protection against sulfidation.

44
In cases where a temporary repair is performed from the outside surface during the short outage that
occurs as part of a normal operating cycle, a nickel based welding consumable is frequently used.
Typical practices employed for this temporary repair are discussed in Section.6.2.

Industry Sponsored Research Effort on Welding Consumables for Repairs and Back Cladding
Restoration

At the time of the development of this report, there is a research effort underway at the Ohio State
University and University of Tennessee to evaluate the performance of welding consumables used in the
repair of welds and restoration of back cladding in coke drums. This effort will include fatigue testing of
the various consumable/base metal combinations with and without back cladding to simulate the
thermo-mechanical loads experienced by coke drums in service. Some of the considerations being
evaluated in this program include:

a) Weld composition and effect on properties (matching or dissimilar composition)

i. Intermix zone chemistry/properties for dissimilar welds
ii. Physical properties such as thermal coefficient of expansion and thermal
conductivity
iii. Mechanical strength in fatigue in relation to base metal
b) Weld process
c) PWHT versus Controlled Deposition Welding (CDW)/Temper Bead Welding (TBW)

One of the key developments being explored by this effort is a matching 12Cr welding consumable for
cladding restoration. A 12Cr consumable has been available and used in Japan to restore 12Cr cladding
in vessels. A 12Cr consumable for this application could result in an improved performance. Many
cracks in coke drums initiate on the ID surface at the interface between the nickel based restoration
weld and the 12Cr cladding. The use of a 12Cr welding consumable for the restoration weld should
eliminate the mismatch in properties between the dissimilar nickel weld and 12Cr cladding.

5.2 Electrode Baking

Both SMAW and FCAW electrodes contain coatings that can absorb moisture when removed from
protective packaging. If the coatings contain moisture it can lead to welding defects like delayed
hydrogen cracking. Consequently, it may be necessary to bake out coated electrodes to drive off any
moisture that may have been absorbed after removing the electrode from protective packaging.
Specific requirements for the baking and storage of SMAW and FCAW electrodes prior to welding vary
depending upon the electrode being used. The manufacturer’s recommendations for proper preparation
and use of electrodes should always be checked before use. Some examples of baking and storage
requirements for commonly used SMAW electrodes are shown below. Note that these steps do not
over-ride the manufacturers’ recommendations.

Nickel based coated electrodes

• Electrodes should be baked at either a) 600°F (315°C) for one hour or b) 500°F (260°C) for two
hours in a furnace that permits moisture to escape.

• Store electrodes at 250°F (120°C) in a portable electrode heater or drying oven near the work
site.

45
 • Electrodes that have been re-baked twice or exposed to precipitation should be discarded.

Carbon Steel (E7018/E7018-A1)

• Store electrodes at 250°F (120°C) in a portable electrode heater or drying over near the work
site.

• If electrode has been exposed to atmosphere for an extended period of time, place in 250°F
(120°C) oven and slowly increase temperature to 600°F (315°C); bake at 600°F (315°C) for 1 hour
in an oven that permits moisture to escape.

• Electrodes that have already been re-baked once or exposed to -high relative humidity should
be discarded.

E7018-B2L-H4/E8018-B2-H4

• Store electrodes at 250°F (120°C) in a portable electrode heater or drying oven near the work
site.

• Electrodes do not require a high-temperature bake unless they have been exposed for more
than 4 hours.

• Exposed electrodes can be reclaimed by baking at 700°F (370°C) for one hour in a furnace that
permits moisture to escape.

• Electrodes that have been exposed to precipitation should be discarded.

5.3 Preheating

Prior to performing thermal cutting, arc-gouging or welding, the base metal should be warmed to a
suitable temperature. Preheat makes the base metal more crack resistant, drives off moisture and
allows for more uniform cooling rates. Preheating temperatures are specified, as shown in Table 5.3,
depending on the coke drum base metal and whether or not the repair will be PWHT’d.

46
 Table 5.3
Recommended Preheat Temperatures for Coke Drum Repairs

Preheat Temperature

With PWHT Without PWHT(1)

Base Material

P-1 Materials 200°F (95°C) 300°F (150°C)

(CS)

P-3 Materials 200°F (95°C) 300°F (150°C)

(C-Mo)

P-4 Materials 250°F (120°C) 300°F (150°C)

(1 or 1 ¼ Cr)

P-5A Materials 300°F (150°C) 400°F (200°C)2

(2 ¼ Cr)

Notes

1. The minimum preheat temperature should not be less

than that used in the procedure qualification test.
2. Some owners do not use CDW on P-5A steels and
require PWHT

Preheat Control

Preheating is performed using a torch assembly or electrical resistance heating elements. It should not
be performed by a hand-held torch. To monitor the preheat temperature, thermocouples or other
temperature indicating instruments such as temperature indicating crayons (i.e. Tempilstik) should be
used. It is imperative that the temperature measurements be taken away from the heat source.
Therefore, it is recommended to control the preheat temperature by measuring metal temperature on
the opposite side from which the heat is being applied to ensure the heat has been applied uniformly
through the thickness of the base metal. Preheat temperatures are normally maintained at least 75 mm
on both sides of the weld.

47
5.4 Controlled Deposition Welding (CDW)/Temper Bead Welding (TBW)

The most common weld repair method used on coke drums fabricated from Cr-Mo steels involves the
use of controlled deposition welding (CDW) or temper bead welding (TBW) techniques to avoid post
weld heat treatment (PWHT). API 510 defines CDW as any welding technique used to obtain controlled
grain refinement and tempering of the underlying heat-affected zone in the base metal. ASME Section
IX defines TBW as a weld bead placed at a specific location in or at the surface of a weld for the purpose
of affecting the metallurgical properties of the heat affected zone (HAZ) or previously deposited weld
metal. The bead may be above, flush with, or below the surrounding base metal surface. If above the
base metal surface, the beads may cover all or only part of the weld deposit and may or may not be
removed following welding. There are several variations of these techniques including the half-bead
technique, the consistent layer technique, the alternate temper bead technique, and the ambient
temperature technique. In general, all of these techniques can be utilized with the SMAW, GTAW
(machine only), GMAW, or FCAW welding processes with strict control of weld bead placement, bead
sequence, and heat input control for at least 2 layers minimum.

Weld repair by CDW/TBW techniques are an attractive option for in-service repair of Ferritic pressure
vessel steels. The use of these techniques has permitted repairs to coke drums to be performed very
quickly and within an operating cycle, thus avoiding the need for a shutdown which would be required if
a PWHT was performed after the repair. These techniques are specifically developed to refine the
coarse grained HAZ in the parent metal and subsequently deposited ferritic weld metal, thus controlling
hardness levels. Dissimilar metal welds such as nickel based filler metals can also be used with the
techniques. Filler metals to be used can be dependent upon the base metal of the coke drum and are
highlighted in Table 5.1.

It should be noted that CDW/TBW techniques will not reduce weld residual stresses. Any residual stress
produced by a CDW/TBW repair is subsequently reduced by “shakedown” which occurs when high cyclic
thermal loads are imposed on the drum during service. Therefore CDW/TBW repairs do not effect
cracking or bulging tendencies of coke drums.

Welding Procedure Qualification

The first step in evaluating CDW/TBW techniques for repairs of a coke drum should be a review of the
qualified welding procedure that will be used by the welding contractor to make the repair. Rules for
qualification of a CDW welding procedure are outlined in API 510 section 8.1.1.4.3 “Preheat or
Controlled-deposition Welding (CDW) Methods as Alternatives to PWHT”. Rules for qualification of a
TBW welding procedure are outlined in the NBIC Part 3 “Repairs and Alterations” section 2.5.3
“Alternative Welding Methods Without Post Weld Heat Treatment”. Additionally, both the API CDW
and NBIC TBW rules invoke the requirements of ASME Section IX “Qualification Standard for Welding,
Brazing, and Fusing Procedures; Welders; Brazers; and Welding, Brazing, and Fusing Operators” section
QW-290 “Temper Bead Welding”.

It should be noted there are slight differences between the requirements for CDW and TBW
qualification. The CDW rules require that the test material for the welding procedure qualification be of
the same material specification (including specification type, grade, class, and condition of heat
treatment) as the original material specification for the repair whereas the TBW rules just require the
same P-number and Group number when there are impact test requirements. The CDW rules do not
allow for repairs on 2.25C-1Mo (P5a) steels whereas the TBW rules do allow such repairs (see

48
Alternative Welding Methods #2,4,&5). The CDW rules allow for full thickness (full penetration) repairs
for P4 and P5a steels while TBW rules only allow partial thickness (partial penetration) repairs.

Typically, there is not the necessary time during a turnaround to perform the necessary qualification for
a CDW/TBW welding procedure. This can be an issue for emergency repairs, when the decision to use
CDW/TBW techniques is made. Because these methods are commonly performed for emergency
repairs today, it is typically possible to find a service provider with approved procedures. It is common
that users who need to perform weld repairs using CDW/TBW techniques have procedures in place that
are applicable to the owner’s specific coke drum vessels. These procedures can be used during both
planned maintenance shutdowns and unplanned emergency shutdowns to avoid unexpected delays.

CDW/TBW Parameters

CDW/TBW essentially involves controlling the first several passes of a repair weld to insure a fine grain
HAZ and weld deposit is formed. The remaining fill passes can be performed using a normal bead
sequence. The welding parameters for a typical CDW/TBW procedure using a SMAW process and a
GMAW process are shown in Appendix A..

5.5 Post Weld Heat Treatment (PWHT)

PWHT is performed after welding is completed to reduce welding residual stresses and temper hard
bainitic or martensitic phases that may form during the welding process. Typically, PWHT is not
performed on carbon steel for repair welds unless dictated by Code maximum thickness requirements,
normally at thicknesses of 38 mm (1½”) and above. C-½Mo also possesses limited hardenability and
does not require PWHT for repair welds less than 5/8-inch (16 mm) thick. 1Cr-½Mo, 1¼Cr-½Mo and
2¼Cr-1Mo have progressively increasing hardenability compared to either carbon steel and C-½Mo and
typically require PWHT in order to control hardness of repair welds. As discussed in the Section 5.4,
CDW is frequently used for repair welds on 1Cr-½Mo and 1¼Cr-½Mo as well as 2¼Cr-1Mo steels as an
alternative to performing PWHT to maintain acceptable hardness levels and mechanical properties.

PWHT of repair welds is almost always performed locally. Local PWHT (LPWHT) should be performed in
accordance with Welding Research Council WRC Bulletin 452 “Recommended Practices For Local
Heating of Welds in Pressure Vessels” and NB-23. WRC Bulletin 452 provides guidelines for LPWHT in
terms of soak bands, heated band and gradient bands that are necessary to avoid unacceptable thermal
gradients that can result in high residual stresses. LPWHT typically takes place in a circumferential band
in cylindrical portions of a drum or a spot on spherical heads on the drum.

All thermocouples to monitor and control a LPWHT should be placed on the opposite side of the shell
from the heating elements. Temperature gradients need to be minimized by employing heated and
gradient bands as outlined in subsections (1) through (3) below. Figure 5.5 illustrates typical PWHT
heating and insulating details consistent with the requirements included in WRC Bulletin 452.

(1) Soak Band. The soak band width that is exposed to the full PWHT temperature T1 needs to
extend for a distance of at least 2t beyond each edge of the weld, where t is the nominal base metal
thickness at the weld.

(2) Heated Band. The temperature decay along the longitudinal axis of the vessel should be
controlled at a distance equal to 2 Rt from the edge of the soak band, where R is the internal radius
of the vessel shell and t is the nominal base metal thickness at the weld. The temperature T2 at this
point should nominally be one-half of the actual PWHT temperature T1 maintained at the weld. The

49
tolerance used for temperature T2 shall be +100°F (63°C) / -0°F (0°C). Additional heating elements may
be required in this area to ensure that the target temperature is achieved and maintained.

(3) Gradient Band. Thermal insulation should be applied to both the internal and external surfaces
of the vessel in the area of all heating elements, to facilitate heat conservation and to control the
temperature gradient along the shell. Insulation needs to extend for a distance of at least 2 Rt
beyond the edge of the heated band.

b. Local PWHT bands should be located a sufficient distance away from nozzle and
manhole attachments to ensure they do not influence the smooth temperature gradient down
the shell. When this is not feasible, the band widths need be increased as necessary to fully
encompass the nozzle or manhole, and the attachment should be completely insulated and
heated during the PWHT operation.

c. Proposals to use heating band and insulation configurations different from Figure 5.5
should be supported by an elastic-plastic stress analysis to show that the residual stress in the
vessel after LPWHT and hydrostatic test does not exceed 50% of the base material specified
minimum yield strength. Refer to Welding Research Council Bulletin No. 452, Recommended
Practices for Local Heating of Welds in Pressure Vessels for details.

d. Spot (bull’s-eye) PWHT should not be permitted on cylindrical shells. Spot PWHT may
be performed on the spherical portion of heads only when approved by the owner. Proposals
need to be supported by sufficient analysis, as outlined in subparagraph c above. See NB-23.

Figure 5.5: The dimensions for the soak band, heated bands and thermal gradient bands for PWHT of
a drum as provided in the guidelines in WRC Bulletin 452. This drawing only shows one side of the
heated bands and the thermal gradient bands.

50
Gas firing or blowing of hot gas from a burner for PWHT is a low cost option when extensive weld repairs
are made over a large area of a coke drum; however, the use of gas firing or blowing of hot gas from a
burner to achieve a PWHT should be carried out in a carefully controlled manner. The biggest concern
with internal firing for PWHT is stratification of the heated gas. If the gas flow is left uncontrolled hotter
gases will rise vertically causing hot spots on the drum shell. This can cause high thermal gradients and
excessive residual stresses in the drum after the PWHT. In a properly designed gas firing PWHT,
insulation, baffles and mechanical controls are used inside the coke drum to control the gas flow to
ensure even heating.

Changes to material properties such as tensile strength and Charpy impact toughness occur during
PWHT. Both base material and weld material are qualified for a specified PWHT time and temperature
so that mechanical properties are maintained. Typically, the weld procedure qualification will include
both a minimum and maximum PWHT combination of time at temperature. The minimum PWHT
condition will reflect a minimum temperature for a minimum amount of time, while a maximum PWHT
condition will reflect a maximum amount of time at the maximum temperature not only for one PWHT
cycle during fabrication but also a PWHT cycle after a repair during initial fabrication and several
additional repairs during the life of the drum. Typically, the vessel fabricator will determine the amount
of PWHT time at temperature required for material qualification. Larson-Miller parameter calculations
should be used to account for time at temperature for all temperatures above 900°F (480°C) during a
PWHT cycle. Material property tests are conducted for the minimum and maximum PWHT conditions.

The coke drum must be analyzed for the PWHT conditions to ensure the drum will not buckle, deform or
induce damaging residual stresses during PWHT or LPWHT. Typically these analyses are performed
using a thermo-mechanical finite element analysis (FEA) consistent with guidance for a Level 3
assessment in API 579-1/ASME FFS-1. The analysis needs to use the material properties of the steel at
PWHT temperatures. Selecting the correct material properties for the analysis of structural integrity
during PWHT is critical. The analysis needs to include the effects of time dependent creep and high
thermo-mechanical stresses created by temperature gradients and geometric discontinuities. This type
of analysis should also identify locations for support spiders which are placed on the inside of the drum.
Additionally, the number, size and angle for support saddles on the outside of the drum can be
determined from this analysis.

A PWHT will heat a coke drum to temperatures that will cause coke to burn. It is important that all coke
deposits are cleaned from the coke drum before a PWHT is performed.

6 Types of Repairs Associated with Coke Drums

This section provides information on commonly employed practices to perform repairs to coke. Section 7 includes
discussion on how long a repair using these practices is expected to last.

6.1 Weld Repairs Made from Inside Surface

When cracks originate from the inside surface and penetrate no more than 50% of the thickness, a
repair normally is made from the inside surface. A repair from the inside surface, involves the following
steps.

1. As a first step, coke deposits should be removed from the area where there is cracking. This
should include an area at least 6-inches (150 mm) around the cracking.
2. The crack size should be determined using UT.

51
 3. Any internal cladding (normally a 12 Cr steel) should be removed by grinding. The extent of
cladding removal should include the crack and the intended weld beveling needed to prepare
the excavation for welding plus an additional 3/8" (10 mm), to prevent contamination of the
repair weld by the cladding. Figure 6.1.1 shows the extent of cladding removal needed for a
weld repair.

Figure 6.1.1: Removal of the cladding as required for

weld repairs made from inside surface.
Groove profile is shown to
illustrate how much cladding
needs to be removed.

4. The area where the cladding has been removed should be tested using an acidic copper sulfate
(CuSO4) solution per guidance in ASTM A-380 to ensure the cladding was completely removed.
5. Cracks should be removed by grinding or arc-gouging. If arc-gouging is used, the area should be
preheated to 200°F minimum (95°C) if the drum is fabricated from carbon steel or C-½Mo, or
preheated to 250°F minimum (120°C) if the drum is fabricated from Cr-Mo steel. The width of
the excavation at the surface should be 2.5 times wider than the depth.
6. The groove should be filled with weld metal. If a CDW/TBW process is used to fill the groove the
guidance in Section 5.4 should be followed. If a CDW/TBW process is not used to fill the groove
a normal welding sequence with PWHT should be followed as incorporated in a qualified weld
procedure meeting the requirements in the ASME Code Section IX.
7. After the weld repair is completed, the weld reinforcement is ground off to be flush with the
inside surface of the vessel as shown in Figure
6.1.2.

Figure 6.1.2: After welding the weld reinforcement is

ground flush with the inside side surface of the base
metal.

52
 8. If a repair is made to the seam welds of the drum, then volumetric examination such as RT or UT
is required before clad restoration.

9. The removed cladding at the repair site needs to be restored. Typically, the 12 Cr cladding is
restored using a SMAW process with an INCO WELD A electrode. The restoration typically is
performed by depositing a butter layer on the beveled edge of the cladding, followed by 2 layers
of INCO WELD A over the entire repair area. Figure 6.1.3 illustrates the restoration weld.

Figure 6.1.3: The cladding restoration weld involves

depositing a butter layer on the cladding bevel,
followed by depositing 2 layers of weld metal over the
entire area where the cladding has been removed.

10. The reinforcement from the weld overlay should be ground flush with the surface of the
cladding as shown in Figure 6.1.4.

Figure 6.1.4: The cladding restoration weld is ground

flush with the cladding surface.

11. After grinding the restoration weld flush with the

cladding, the entire repaired area is inspected using PT.
12. If PWHT is specified, it should be performed following the guidance provided in Section 5.5.

53
6.2 Weld Repairs Made from Outside Surface

When cracks originate from the outside surface and penetrate no closer than 1/8-inch (3 mm) to the
internal 12Cr cladding, a weld repair can be made from the outside surface. A repair from the outside
surface does not have the same complications as a repair from the inside surface because there is no
cladding on the outside surface. However, it is essential that the repair weld does not penetrate into
the 12Cr cladding causing the repair weld to pick up chromium from the cladding. This could result in a
very hard repair weld deposit with poor mechanical properties. Repair welds from the outside surface
not penetrating the 12Cr cladding are made with a welding consumable with a composition matching
the coke drum base metal. The repair weld can be made using a CDW/TBW process as discussed in
Section 5.4 or using a normal welding sequence followed by PWHT as required. In each case, it is
important that a qualified welding procedure is provided by the welding contractor for review by the
owner during early planning stages for the repairs.

Frequently, it is desirable to make repairs from the outside surface of a coke drum when cracks initiate
on the inside or outside surfaces and penetrate through the entire wall or almost through the entire
wall. Since the internal 12Cr cladding cannot be removed when repairing from the outside surface, it is
necessary to use a consumable suitable for welding a coke drum base metal (carbon steel, C-½Mo or Cr-
Mo) to 12Cr steel. As with a cladding restoration weld, a nickel based welding consumable is used to
perform a full thickness weld repair from the outside surface. Typically, ENiCrFe-2 electrode is used for
a SMAW process, ERNiCr-3 wire is used for GMAW or GTAW processes and ENiCr3TO-4 electrode is used
for a FCAW process. Table 6.2 provides typical welding parameters when using a nickel based welding Commented [JEM2]: As noted in Section 5.1 on welding
consumable to repair cracking from the outside surface of a coke drum. These welding parameters are consumables, an Alloy 625 welding consumable is generally not
used because it is very strong and causes a large strength mismatch
intended to provide a fine grain HAZ in the base metal and promote tempering of the base metal HAZ as with base metal. AZZ has included the use of Alloy 625
is done using a CDW/TBW process as discussed in Section 5.4. consumables for repairs from the outside surface and for cladding
restoration. Need to resolve this issue.

Table 6.2: Typical Welding Parameters Using a Nickel Based Consumable

Welding Process SMAW GMAW or GTAW FCAW

Typical Electrode ENiCrFe-2 (Inco Weld A) ERNiCr-3 (Inconel 82 wire) ENiCr3TO-4
3/32” 1/8” .045” 3/32” 1/8” .045”
Electrode Diameter
(2.5mm) (3.2mm) (1.2mm) (2.5mm) (3.2mm) (1.2mm)
Current 50 – 75A 70 – 95A 50 – 80A 70 – 120A 90 – 150A 130 – 175A
Voltage NA 18 – 23V 22 – 26V
Heat Input 1.5 kJ/mm max. 1.5 kJ/mm max. 1.5 kJ/mm
max.
Preheat As appropriate for drum base metal – See Table 5.3
Interpass
300°F (150°C) max.
Temperature
Bead Type Stringer
500°F for 2 hours for Cr-Mo steels only, CS and C-½Mo do not require post
Post Heat
heat

A weld repair on a coke drum of a through wall or almost through wall crack is performed from the outside
in order to minimize the downtime required for the repair, since vessel entry is not required. However, it is
important to note that a full thickness repair made with a nickel based welding consumable should be
considered a temporary repair. The narrow high hardness zone that exists along the fusion line of the

54
nickel weld deposit and coke drum base metal or cladding is susceptible to cracking. See Section 7
which discusses the expected life for repairs including temporary repairs made with a nickel based
consumable. Most owners replace the temporary nickel based repair welds during the next planned
turnaround with a repair weld procedure using a welding consumable that matches the base metal as
discussed in Sections 6.1 and 6.3.

6.3 Weld Repairs Made from Both Sides

When cracking originating from either the inside or outside surface of the coke drum is deeper than 50%
of the thickness, it normally is necessary to make a full thickness repair weld from both sides with a
double bevel weld geometry. Figure 6.3.1 shows a typical double bevel weld geometry for a repair weld
made from both sides in a coke drum.
Bevel opening
2.5 times the
depth

Figure 6.3.1: Typical geometry for a weld repair

made from both sides of coke drum wall.

Cladding ground
back from weld
bevel

Welding from the inside surface should follow guidance from Section 6.2 while welding from the outside
surface should follow guidance from Section 6.3. One addition to this guidance is that after welding
from one side is performed, it is normally necessary to back gouge the root pass before making the weld
from the opposite site. After back gouging, the surface should be inspected for cracks using MT or PT.
Additionally, in some cases when the gap at the double bevel is large, it may be necessary to install a
back ring before making the initial pass. It is important that the ring has a matching chemistry with the
base metal. Again, after making the weld from one side with the backing ring, it is necessary to back
gouge the backing ring with root pass before making the weld from the opposite side.

6.4 Replacing Large Sections of Cylindrical Shell Plates

There have been many occasions in coke drums that large sections of cylindrical plate were required to
be replaced because of severe bulging, banana effects on the coke drum and even extensive cracking,
not only at seam weld locations, but also in other locations within the plate itself. Plate replacement can
be consist of replacement of entire ring sections, partial ring sections either in one course or multiple
adjacent courses and even vertical strakes if the drum was constructed using this technology. In either
case, it is essential to use only contractors who have successful previous experience and the requisite
equipment to carry out the type of work specific to the coke drum to be repaired.

Before starting any replacement work, whether it be replacing large sections of drum shell plates or
making smaller replacements or repairs, it is useful to make a determination as to why individual
sections of shell plate require replacement or repair at one or multiple locations so that possibly some
action may be taken to mitigate or reduce the detrimental action that causes the damage to the shell
plates. Also, it is useful to carry out an evaluation of the repair related to the anticipated remaining

55
drum life and make a determination of how the repair may affect the remaining drum life or whether
plans need to be made to initiate drum replacement.

One of the first steps that need to be carried out is to determine by calculations the width of a plate
section that can be removed safely, giving consideration to prevailing wind and potential seismic
conditions, so that the remaining shell will still keep safely supporting itself. The entire replacement
process must be very carefully planned. In this way, entire 360º shell courses may be replaced section by
section, otherwise the entire coke drum section above the one to be removed also needs to be
removed. In this case, space and crane permitting, entire ring(s) may be replaced with the upper section
of the coke drum then finally replaced onto the new shell course(s). In some cases, the structure around
the drum has also been used to provide additional support to the upper section of the drum during the
modifications. Sections 10..3 and 10.4 in API 934-G also provide valuable additional information on
minor and major shell replacements in coke drums.

Once the appropriate plate width has been determined for safe removal, a new section (or multiple
sections) to be replaced must be first be cut to size and then rolled or pressed to the correct diameter.
An internal laser scan can provide valuable information regarding the actual diameter that the
replacement plate needs to be manufactured ensure a good fit. When the replacement plate(s) are
available, the drum cut locations are carefully marked and the drum is cut and the plate sections are
removed one by one. The drum cut lines are then beveled and examined by the prescribed NDE
techniques as required.

Depending on the extent of the bulging of the drum, the new section to be installed may not necessarily
be totally welded out as the new plate may not necessarily mate with the existing drum until several
other sections have also been removed, but the reinstallation should strive to be such that provides
support to the remaining drum section above. The junction of longitudinal and circumferential seams
requires extra care in order that there is no mismatch of adjacent plates as such locations of mismatch
can be sources where future potential cracks can take place. The completion of the welding of all of the
seams may in some cases not take place until all the plates that make up the replacement in a specific
location have been installed. Properly engineered and installed temporary supports strategically placed
to support larger spans of removed plate have also been used on many plate installation projects.

6.5 Skirt Repairs, Retrofits and Replacement

There can be numerous different options for repairing and replacing coke drum skirts depending on the
damage that has occurred that may require either full or partial skirt replacement.

These options include but are not necessarily limited to the following:

• Section 6.5.1: Full vertical length of skirt replacement around the entire skirt circumference
starting from drum attachment weld and extending to a point above the anchor bolt chair cap
ring.

• Section 6.5.2: A skirt window replacement-

o Starting at the drum attachment weld location and extending to an undamaged section
of the skirt at a lower elevation

56
 o Starting at a location below the drum attachment weld and extending to a lower
location or to the base plate.

• Section 6.5.3: Skirt attachment weld repairs

• Section 6.5.4: Skirt slot repairs

• Section 6.5.5: Skirt Retrofits

6.5.1 Full Skirt Replacement Around Entire Circumference

It is suggested that the lower cut line of the skirt be located above the projection of the anchor bolts to
facilitate ease of installation unless the skirt damage is so great that the base plate and anchor bolt chair
ring and anchor bolt gussets also require replacement. In case that skirt weld PWHT is required, the
lower cut line elevation should allow for as much temperature gradient drop as possible between the
gradient control band edge and the base plate.

Note: The same installation procedure below may be used for full vertical length skirt replacement
partially around the circumference

As previously indicated, one of the first steps that needs to be carried out is to calculate the width of a
skirt section that can be removed safely so that the remaining vessel will be safely supported. In many
cases, the replacement process can include replacing two diametrically opposite skirt panels at the same
time, however, this must be confirmed by calculations. The entire replacement process must be very
carefully planned and by doing so, the entire 360º skirt may be replaced section by section. The need
for the determination of the amount of skirt that can be safely removed will be applicable for each type
of skirt removal.

The replacement process will include making the vertical skirt panel cut on each side of the panel that
has been determined to be able to be safely removed, cutting the skirt horizontally at the determined
cut line location above the anchor bolt chair cap and removing or cutting at the upper support weld as
the case may be. One vertical seam of a pair of adjacent panels is welded after completion of the top
and bottom welds after the second panel is correctly fitted into its respective place. In this way the last
panel is ultimately fitted and welded to the first panel when the installation takes panel by panel
circumferentially around the drum.

The engineer responsible for designing the skirt replacement will provide the details of how and where
to make the cuts at the skirt to shell support location as well as at other locations based upon the
damage that exists at the skirt to shell attachment location and in the rest of the skirt. The engineer will
also ensure that hot box removal and replacement details are included as well as the requisite insulation
and fireproofing removal details and scope are clearly defined. The engineer will also ensure that
appropriately fabricated skirt replacement panels closely follow key aspects of the original design, such
as keyholes, are correctly fabricated to correct dimensions and tolerances and are at site prior to
starting any skirt replacement work

The upper skirt to shell attachment detail may be one of a number of typical different configurations
listed below in Figure 6.5.1:

• Typical old style skirt to shell fillet weld (see Figure 6.6.1, a). Modified fillet weld with internal
crotch radius (see Figure 6.5. 1, b).

57
 • Forged skirt to shell attachment (see Figure 6.5.1, c).

• Externally wrapped skirt straight lap joint over shell (see Figure 6.5.1, d).

• Externally wrapped skirt with wavy lap joint over shell (see Figure 6.5.1, e).

Each style of skirt attachment indicated above may also have keyholes as indicated in Figure 6.5.2.
Following the skirt panel removal, the existing skirt upper attachment weld may need restoration to
some extent and this is usually accomplished by applying the requisite preheat and by manual welding
and subsequent grinding, although an automatic welding processes can also be utilized if the amount of
welding warrants doing so. Prior to the reinstallation of panels, the cut locations where re-welding will
take place should be beveled for welding and all edges and weld restoration areas should be magnetic
particle examined prior to fit up and welding. In some installations, the decision may be made to remove
the entire skirt attachment weld because of extensive damage in welds such as those depicted in Figure
6.5.1, a. b, d & e. However, in such cases, the weld removal and subsequent replacement will be time
consuming and will require additional manpower and equipment resources and careful planning

The new prefabricated panels are then fitted and installed taking care for proper alignment and welded
out top and bottom. The junction of longitudinal and circumferential seams requires extra care in order
that there will be no mismatch of adjacent plates nor mismatch of the top section of the skirt plate with
the weld that it must abut to. The completion of the welding of all of the seams may in some cases not
take place until all the plates that make up the replacement in a specific location have been installed.
Once replacement plates have been properly fitted they can be welded out per the applicable welding
procedure with the application of the requisite preheat. NDE before and after final PWHT (if applicable)
should include full radiography (on radiographable welds), UT testing (on welds where meaningful UT
can be performed) and magnetic particle examination of welds.

58
 Figure 6.5.1: Typical skirt attachment details

6.5.2 Skirt Window Replacement

The basic procedure for a skirt window replacement will be very similar to the procedure provided
above for full skirt replacement except the window panels will generally be smaller and there will not be
as many as with a full skirt replacement. Again, the junction of longitudinal and circumferential seams
requires extra care to ensure that there will be no mismatch of adjacent plates.

59
6.5.3 Skirt Attachment Weld Repairs

As discussed above, welded skirts typically experience cracks in the vessel-to-skirt attachments (Ref. Fig.
6.5.3a). Cracks on the skirt side of the attachment often grow through-wall and eventually around the
entire circumference of coke drums. Cracks on the shell and cone sides are also possible and can have
more serious consequences due to possible loss of containment. Skirts that are designed with slots to
minimize cracking at attachment welds tend to develop cracks at the top keyholes of slots (further
discussed in 6.5.4).

Figure 6.5.3a Most common locations for cracking of a welded vessel-to-skirt attachment.

A common repair sequence for repairing cracks originating at the upper end of the attachment weld and
propagating into the coke drum shell is as follows (see Fig. 6.5.3b):
1. Excavate area to completely remove cracks & previous repairs
2. PT or MT exam the excavation to ensure the crack has been completely removed.
3. Utilize a CDW/TBW welding technique to re-fill the repair excavation as discussed in Section 5.4.
4. Re-contour the repair surface profile to minimize stress concentrations

60
 Figure 6.5.3b Skirt attachment repair sequence

Note that sometimes this transition region is extended further onto the drum in order to further
minimize stress concentrations.

Repairs to the inner crotch area within the “hot-box” is more complicated and usually involves removing
windows within the skirt to access this area to make repairs, then re-welding the windows back in place.
Sometimes this type of repair is done in concurrence with a full skirt replacement or skirt section
replacement. In extreme cases, the complete skirt attachment weld has been completely removed and
rebuilt with weld metal buildup[13].

6.5.4 Skirt Slot Repairs

Installations which have slots with or without keyholes will usually at some time develop extensive
cracks predominantly radiating out from the upper end (see Figure 6.5.2a&b) towards the skirt
attachment weld. Cracks emanating from slots and/or keyholes should be routinely monitored and
repaired before they encroach onto the skirt to shell attachment weld. Weld repair will include grinding
out the crack and re-welding the ground area. Sometimes keyholes are added to existing slot
terminations, or existing keyholes are enlarged. Enlarging of existing keyholes to a diameter that
consumes nascent radial cracks is a good first phase repair for this type of cracking.

Note that sometimes this transition region is extended further onto the drum in order to further
minimize stress concentrations.

Repairs to the inner crotch area within the “hot-box” is more complicated and usually involves removing
windows within the skirt to access this area to make repairs, then re-welding the windows back in place.
Sometimes this type of repair is done in concurrence with a full skirt replacement or skirt section
replacement.

61
 Figure 6.5.4a: Typical Coke Drum Skirt Keyhole System

Figure 6.5.4b Cracking initiated at the slot keyhole and propagating through the skirt attachment
weld.

62
6.5.5 Retrofit Designs for Skirts

In addition to above conventional designs for skirts on coke drums, several unconventional skirt
designs have been developed to minimize stresses and the likelihood of cracking at the skirt
attachment. Several unconventional designs are non-welded and are patented. An example of a
non-welded design that has been implemented as a retrofit is the tongue-in-groove design shown in
Figure. 6.5.5a and described in Reference [14].

Figure 6.5.5a: A Tongue-and-Groove Skirt Design

Damaged skirts can be retrofitted to non-welded skirt designs to minimize the likelihood of future
cracks. The first such retrofit of an operating set of coke drums was conducted in 2011 by converting a
scallop-welded skirt to a bracketed sliding skirt, as shown in Fig. 6.5.5b and described in Reference [15].

63
 Figure 6.5.5b: A Welded Skirt Retrofitted to a Bracketed Sliding Skirt

6.5.6 PWHT of Skirt Repairs

Care must be taken during post weld heat treatment (when required) to make sure that the soak band,
heated band and gradient control band widths are maintained as recommended in WRC Bulletin 452
and discussed in Section 5.5. PWHT should consist of full circumferential bands where possible. As
indicated above, when skirt PWHT is required, the lower weld line elevation should allow room for as
much temperature gradient drop as possible from the gradient control band edge to the base plate.
Full band local post weld heat treatment is recommended for Cr-Mo welds and welds in which Code
requirements mandate PWHT. It is to be noted that some owners have repaired keyhole Cr-Mo welds
without PWHT.

6.6 Repairs and Modifications To Unheading Devices

Unheading devices are very specialized pieces of equipment and repairs to these should be carried out
only by the device manufacturer service reps who have the knowledge and the parts to make the repairs
and who can usually be at site very soon after a call is made.

64
6.6.1 Modification to Cone Section to Accommodate Unheading Device

When retrofitting an unheading device to coke drums, modification typically is required to the cone
section of the drum. Generally, the angle of the cone section has to be modified to fit the unheading valve
and this will require a cut in the cone section of the drum at a certain elevation to permit the attachment
of the new cone section with unheading device. In the past a bolted flanged connection had been used
for the attachment, but this flange connection has a tendency to leak even with hot bolting used in the
field. For better reliability a welded connection should be used with minimal change to the cone angle to
reduce the structural discontinuity at the new weld. The new weld should be PWHT’d.

6.6.2 Trouble Shooting of Gasket Joints

The gasket joints that have given problems over the years in coke drums were the gaskets at new flanged
inlet sections above unheading valves that originally bolted to coke drum bottom flanges. These new
bolted sections were later replaced with inlet sections welded to the coke drums. Additionally, the
unheading valve attachment gaskets at the drum bottom flanges also are prone to leak after a number of
cycles. In each case the unheading valve manufacturer has determined specific bolt tensioning procedures
as well as bolt tensioning frequency following initial startup. The unheading valve manufacturer also
recommends a specific type and brand of gasket which helps to alleviate the leakage problem. Owners
are recommended to contact their specific unheading valve manufacturer for specific recommended
tensioning procedures and tensioning frequencies and also for recommended information on gasket types
and manufacturers.

6.7 Repairs to Feed Nozzles

There are basically two different types of feed nozzles. 1) Ones in the bottom cover of a coke drum, and
2) Ones in coke drums which have unheading valves where the feed nozzle(s) have been moved and
attached to the drum vertical section just above the unheading valve and below the bottom of the cone.
Figure 6.7aa indicates a typical nozzle penetration through a bottom cover which receives some thermal
protection from the hot feed and cold quench water with an internal layer of refractory. Figure 6.7b
indicates an upsweep inlet nozzle penetration into a short vertical section of a drum below its cone. The
inlet nozzles above the unheading valve can also be single nozzles or diametrically opposed dual nozzles
located either on the cone or on the vertical shell section below the cone as indicated in Figure 6.7b.
These inlet nozzles can also be radial, or have an upsweep as indicated in Figure 6.7b, or can even
project at right angles from the cone.

Feed nozzles can be subject to fatigue cracking after a number of years of operation because of the
severe cyclic thermal loading they undergo during the operation and switches from the drum steam
heating cycle, the vapor heating cycle, the hot feed inlet cycle, the steam quench cycle, the water
quench cycle and finally drum drain cycle. During such cycles, the feed nozzles also undergo mechanical
loadings from the expansion and contraction of the feed piping and these combined thermal, pressure
and displacement loadings are resisted by the nozzle attachment to the bottom cover or in the case of
unheading valves to the vessel cones. Where possible, integrally reinforced forgings with radiographable
weld lips with a generous attachment corner radius are preferred to decrease local stress
concentrations and increase fatigue life.

In each case, the replacement of a nozzle should only be undertaken after calculations using heat up and
cooling data, pressure information and nozzle loading information can verify that a replacement nozzle

65
design is suitable for the estimated remaining drum life. For feed nozzles located on cones, the nozzle
neck should preferably be an integrally reinforced forging with radiographable lips with a generous neck
to shell radius to reduce stress concentrations and provide maximum fatigue life. For feed nozzles that
penetrate centrally through a thick flat bottom cover a weldable expanding fitting with a central core is
recommended as illustrated in Figure 6.7a. Such a design will help reduce the thermal stresses
generated at the interface of the flat cover and the feed nozzle.

Various operators have concluded that both single and double side entry nozzle flanges are prone to
leakage and hence they should be adequately designed for the thermal conditions. Frequent re-
tensioning of the flange bolts normally is required.

Figure 6.7a Typical Insulated feed entry into a drum bottom cover

66
 Figure 6.7b Typical Insulated feed entry into the side of the drum bottom cover

6.8 Inspection Insulation Windows

Inspection insulation windows (removable insulation covers) are used for routine monitoring, inspection
and assessment of "critical areas/welds" of a coke drum to avoid unplanned, unsafe and costly
shutdowns caused by unanticipated leaks. Inspection windows are used to complement 360 degree
inspection lanes typically used during turnaround inspection of welds. See Figure 6.8a, b and c for details
for inspection insulation windows.

The number of inspection windows around each circumference is determined by the inspection team,
but as a minimum, it typically is 8 but can be 12 to 16 depending on number of tee intersections of
circumferential to longitudinal welds. Inspection insulation window size typically is 250mm by 250mm
(nominally 10-inch by 10-inch) to keep them between circumferential bands which are at a 300mm
(nominally 12-inch) spacing.

The critical welds and areas to monitor and justifying inspection insulation windows are typically as
follows:

- Skirt Junction area including inner crotch using weld build-up (WBU) design. A forged Y-ring design
normally would not justify the need for an inspection insulation window since the attachment welds are
removed from the high stress area at the inner crotch and the base metal stress concentration may have
greater fatigue life then the remote welds.

- Upper portion of keyhole in a slotted skirt

- Circumferential closure seam weld(s) between coke drum subassemblies subjected to local PWHT

67
 Figure 6.8a: Side view of inspection Insulation window at skirt junction

Figure 6.8b: Typical detail for inspection insulation windows at shell closure weld.

68
 Figure 6.8c: Typical inspection insulation window

6.9 Foundation Repairs

6.9.1 Repairs for Drum Tilting

As indicated in Section 3.4, monitoring of drum tilting or “out of plum” is essential to make sure its
progression does not go beyond acceptable stability and operability limits. In those cases where tilting
is approaching the limits and progression rate indicates that corrective action is needed, the owner/user
needs to explore alternatives. A repair technique that has been used with success is the repositioning
and stabilization of the drum by reinforcing the base ring. This type of procedure is executed by
contractors with recognized experience and is always supported by an engineering assessment prior to
the execution.

The following example is used to illustrate some of the steps involved in a tilting correction project. It is
important to keep in mind that every case is different and the steps used in this specific case might not
apply to other projects. There can be other options to correct tilting, and depending on the

69
configuration of the specific drum or set of drums and the nature of damage that cause the tilting, the
solution may require either full or partial replacement of the drum(s).

In this specific example, tilting was cause by an internal fire in the shell section of the drum. The coke
fire resulted in damage to the drum cylindrical section, leaving a bulged area on the drum’s north side.
Figure 6.9.1a shows an internal view and external view of the drum obtained by laser mapping after the
fire event. The bulge caused the base ring to draw upward on the north side and the top head to draw
downward to the north. The deformation at the top of the drum affected the penthouse restricting
independent movement of equipment in the top portion of the drum, such as RV discharge headers,
steam lines, walkways above the penthouse, and the drill stem assembly rail supports.

The best course of action was to reposition, or “tilt”, the drum to the south to reposition and correct all
equipment affected in the top section. A specialized structural contractor was hired to design and
execute the repositioning. Flat jacks were used to raise the north-side base ring approximately 1-inch.
After an engineering study, the contractor selected the total number of jacks to be used. Base ring
modifications were required to structurally support the lift. This amount of lift of the base ring was
expected to result in approximately 4-inches of southward movement at the top of the drum. Anchor
bolts that were affected by the tilting were also replaced. As shown in Figure 6.9.1b, base ring
modifications included the addition of gusset plates and a base ring extension. Concrete from the table
top was removed to allow insertion of flat jacks under the modified base ring, and temporary shims
were installed to maintain drum stability after concrete removal. In the excavated area, leveling pads
were installed for the flat jacks as shown in Figure 6.9.1c. The flat jacks were set on the leveling pads
and shim plates were installed to fill the gap between the jack and the modified base ring. The jacks
were inflated with water to “tilt” the drum. Once the drum was in the desired position, the water in the
jacks was displaced with grout. The jacks were kept permanently in place and the excavated area and
the area between the base ring and the table top were filled with grout as shown in Figure 6.9.1d. The
drums remained out of service until the grout completed the initial cure. Figures 6.9.1e through 6.9.1h
show some of the steps during the repair procedure.

Bulge

Laser Scan – Internal View External view of drum; made from scan data.

Figure 6.9.1a: Internal view and external view of the drum cylindrical section generated with laser
mappings.

70
Figure 6.9.1b: Base Ring modifications and concrete excavation.

Figure 6.9.1c: Setting the Jacks and Lifting the Drum.

71
 Figure 6.9.1d: Grouting process after lifting.

Figure 6.9.1e: Modifications of base ring including new gusset plates, and positioning and installation
of multiple flack jacks

72
Figure 6.9.1f: Grouted to match table top surface.

Figure 6.9.1g: Final grout stage.

73
 Figure 6.9.1h: Completed installation.

6.9.2 Anchor Bolt Repairs

Deterioration to existing anchor bolts that fasten delayed coke drum vessels to
reinforced concrete table top structures can be caused by a number of conditions
ranging from standard metal corrosion to drum vessel movement and unless the failure
mechanism is well understood, the Owner can be placed in a position of frequently
repairing-the-repair. Once the failure mechanism is identified, the anchor bolt repair or
replacement can provide a significant service-life extension to these important fastening
devices. References [16] and [17] provide useful information on anchor bolt design.
References [18] and [19] provide specific information on anchor bolts in coke drums.

6.9.2.1 Anchor Bolt Types

Anchor Bolts used to fasten coke drum vessels to reinforced concrete table top
structures are typically metallic and cast into the concrete at the time of original
construction. Typical anchor bolt details include the following:

• L hook anchor bolt type

• J hook anchor bolt type
• Threaded rod with welded base plate
• Threaded rod with double nuts and plate washer

74
 Less frequently, a straight rod (threaded both ends) is placed through the entire
concrete slab thickness via a pipe conduit and has nuts and washers on both the top
side and bottom sides of the reinforced concrete slab to facilitate tightening.

6.9.2.2 Anchor Bolt Distress

Much of the working length of the anchorage is hidden from view since it is embedment
within the reinforced concrete slab. Unfortunately, the condition at the “business end”
of the anchor bolt is unknown until a failure occurs such as:

• Bolt tensile failure

• Pullout (concrete tensile) failure
• Lateral bursting (blowout) failure
• Localized bearing failure
• Concrete splitting failure

These failure modes can be initiated by environmental/service conditions associated

with the delayed coking process. High operating service temperatures coupled with a
moist/wet process exposure create corrosive conditions. Additionally, vibratory
movement associated with the process can move coke drum vessels (i.e., drum walking
or drum twist) from their original set position and bend/deform fixed anchor bolts.

6.9.2.3 Anchor Bolt Assessment

In order to understand the condition of anchor bolts, an investigative program

employing both NDE and semi-destructive testing is required. Additionally, a
representative stock sample (i.e., similar length, diameter and metallurgy) of the type of
Anchor Bolt being evaluated should be used for NDE equipment calibration.

On-site investigation should begin with a 3 lb. maul strike to each anchor bolt to
determine “relative soundness” and whether the anchor bolt is broken or corroded in
near surface regions. Once it has been determined which anchor bolts require further
investigation, these bolt top surfaces should be ground flat using an abrasive grinder to
prepare the surface for NDE.

a. Anchor bolt NDE typically involves ultrasonic testing. A comparison of ultrasonic

testing results on the service exposed anchor bolts with the ultrasonic testing
results on a new stock anchor bolt of the same size, metallurgy and
configuration should highlight potential damage in the form of cracking and/or
metal loss in the service exposed anchor bolt. Commented [JM3]: Description of how ultrasonic testing works
b. Anchor bolt semi-destructive testing is used to validate results generated by NDE is not necessary.

techniques and generally includes shallow localized excavations at selected

anchor bolt locations to visually determine the actual condition of the anchor
bolt. Typically, excavations only extend down a distance of approximately 6-
inches as most deterioration occurs within this region based on experience.

75
6.9.2.4 Anchor Bolt Repair Options

Several repair options exist for anchor bolts including the following:

a. Mechanical coupling and stud extension

b. Welded stud extensions
c. Relocation of anchor Bolts
d. Full depth anchor bolt extraction and reinstallation

Each of the anchor bolt repair options listed above have pro’s and con’s with repair
options a and b requiring significant excavation adjacent to the coke drum skirt base
ring. As a result, these repair options are normally performed off-line during an outage. Commented [JM4]: Pretty sure that galvanic corrosion comes
into play in this situation
Repair option c above requires relocating anchor bolts to another location within the
region supported by the existing embedded reinforcing steel bars. Relocating new
anchor bolts require cutting of some of the embedded reinforcing steel bars in the slab.
This should be done only after a a structural engineer evaluates the overall effect of
reducing the loading capacity of the structural member (i.e., slab) by the loss of
reinforcing steel bar support. Additionally, new anchor bolt chair assemblies will need to
be welded to the skirt wall to accommodate the new relocated anchor bolts. Due to the
amount of work required on skirt wall faces, this repair option is typically performed off-
line during an outages.

Repair option d listed above employs a core rig equipped with a diamond-tipped, water-
cooled extendable core bit capable of cutting concrete and embedded steel. Since the
anchor bolts are being replaced in kind at the existing location, it is important to have
bolts of the same dimension on hand. This repair procedure has the advantage of not
requiring any cutting of the embedded reinforcing steel bars as noted for repair
procedure c.. Depending on the location of the bolts, insulation removal may be
necessary for core-rig access. Once the existing anchor bolt is core-cut, the core is
removed from the hole, the hole cleaned of slurry/debris and subsequently dried. A new
mechanical expansion anchor bolt is placed into the prepared hole and engaged
according to manufacturer’s recommendations. After engagement, the remaining void
around the new anchor bolt is filled with grout and allowed to cure. It is very important
that the new anchor bolt extension be designed by the anchor bolt manufacturer and
the design be reviewed by a structural engineer. Unlike the other repair options, repair
option d can be performed while the delayed coking unit is on-line.

6.9.2.5 Non-Bolted Restraints

Recent studies have demonstrated that significant anchor bolt stresses can result from
uneven thermal expansion of coke drums under non-uniform thermal gradients, Ref. 4.

76
 To minimize the likelihood of bolt failures, a new patented restraint system that utilizes
non-bolted restraints has been developed and implemented. The new restraint system
consists of multiple anchor blocks placed around the drum that initially do not contact
the base plate. Anchor blocks only engage the base plate after base plate displacement
exceeds designed axial, circumferential, or radial displacement gaps between the base
plate and the anchor blocks. In the first reported implementation of this system, shown
in Fig. 6.9.2.5, anchor blocks allow the bottom of the skirt to displace in reaction to non-
uniform transient thermal gradients in the drum with minimal restraint, as described in
Reference [20].

Figure 6.9.2.5 – Photo of Non-Bolted Anchor Blocks

6.9.3 Grouting Repairs

Repairs to existing grout beds supporting delayed coke drum vessels can be required
due to damage by several different causes. It is very important before a repair is
performed that the root cause for the observed damage is fully understood and
mitigated.

Surface Preparation

Typically, when a grout bed has sufficiently deteriorated, the entire grout bed is
replaced after the underlying concrete substrate is cleaned of grout residue and other
surface contaminates. Concrete substrates should be free of laitance, curing

77
compounds, dirt and debris, as well as hydrocarbon residue all of which can prevent
cementitious grout from bonding to concrete.

a. Demolition of cementitious grout typically involves the use of hand

tools and small pneumatic chipping guns (15 lb. class or less) or electric
hammer drills fitted with chisel bits. Care should be taken when
excavating around the skirt base ring to avoid damaging the base ring.
Resultant excavated base surfaces should be evaluated for soundness.
If “soft conditions” or obvious deterioration is encountered, concrete
substrate repairs should be made prior to new grout bed placement in
accordance with Section 6.9.4 of this document.
b. Review existing coke drum skirt ring base plate shim(s), for corrosion.
Corroded shims should be replaced.
c. Inspect undersides of the skirt ring base plate for the presence of
corrosion. Corrosion scale should be removed mechanically by wire
wheel, needle scaling or abrasive grit blasting. Resultant cleaned steel
surfaces should then be coated with a corrosion-resistant primer prior
to grout bed placement.
d. Prepare resultant sound concrete surfaces mechanically (i.e., chipping,
abrasive grit blasting, needle scaling, etc.) to ensure a roughened
surface profile.

Grout Formwork

Formwork for grout containment will be required for all grout placement methods
except for dry packing. Forms should be designed to withstand all loads imposed on the
forms during poring and curing. For grouting around the coke drum vessel skirt base ring
plate, formwork is placed along interior table top penetrations as well as along exterior
base plate perimeters.

a. Formwork should be coated with bond-inhibiting materials (i.e., form

oil, wax, etc.) to allow easy form removal.
b. Joints and formwork base regions should be adequately caulked,
foamed, or taped to provide an effective seal from leakage of grout
materials during placement and subsequent hardening.
c. To facilitate edge detailing, formwork should be designed to allow easy
removal subsequent to early age hardening.

Embedments

Various items may be embedded into the grout bed and require preparation in the form
of corrosion protection or covering/coating to prevent grout bond development with

78
the embedment. Embedment items can include anchor bolts, shims and movement
assemblies (sole plates/pads).

a. Bond breakers can include wax, plastic sheeting, duct tape and plastic
tubing.
b. Corrosion resistant coatings can include zinc-rich primers appropriate
for the service conditions.

Grout Material

Grouting used below the coke drum vessel skirt base ring plate are generally
cementitious, non-shrinking and without metallic aggregate. Resinous polymeric epoxy-
based grout is not recommended due to high elevated service temperatures potentially
resulting in creep deformation under sustained loads.

a. Grout consistencies can range from fluid to plastic-flowable to dry-

pack. The most common grout placement consistency ranges from
Fluid to Plastic-Flowable. Dry-Pack consistencies typically are used in
emergency situations for temporary on-line repairs where a low-water
content, stiff grout consistency is required because of exposures to
high temperatures during installation and curing.
b. Mixing of cementitious grouts should be in strict accordance with grout
manufacturer’s directions. However, all cementitious grouts should be
mixed in a mortar mixer (a stationary drum with rotating blades and
not in a concrete mixer (a rotating drum with attached fins). Hand
mixing of grout materials should only be performed when a very small
amount of grout is needed. Inadequate mixing of grout can lead to Commented [JM5]: What is meant by blending? Would mixing
lower strengths, lumps, segregation and reduced flowability. be a better word? Does this mean blending of one or more
grouting products?

Grout Installation

Grout can be installed using several methods as follows:

a. Grout pump - hand, electric, pneumatic or hydraulically actuated

b. Head box - for gravity/plunger placement
c. Bird’s mouth - inclined chute placement methods
d. Bag placement – for very fluid grout placements which employs the use
of a heavy-duty pastry/frosting bag with integral nozzle

During cementitious grout placement, with even more fluid-consistency grouting

products, consolidation techniques typically are required that remove air bubbles that
may form along grout contact surfaces and adversely affect grout properties. These
consolidation techniques include:

79
 I. Pencil vibrator
II. Mallet tapping
III. Rod “puddling/plunging”
IV. “Banding” with flat steel bands

Finishing shoulders occurs when the grout stiffens sufficiently to the point where it will
hold its shape when scored with the point of a towel. At that point, the forms can be
removed and the shoulders cut-back at a 45o angle from the bottom edge of the
baseplate to the foundation. The grout shoulder should be finished using a trowel with
care taken to maintain the elevation slightly below the base of the skirt base plate.

Curing

Cementitious grout should be cured as directed by the grout material manufacturer.

Typically, this involves wet curing immediately after hardening. To prevent rapid surface
drying, all exposed grout surfaces should be covered with wet burlap, blankets or rags
that are maintained in a continuously “wet” condition. The additional use of plastic
sheeting and membrane-forming curing compounds should be allowed; however, wet
curing will provide optimal grout strength and long-term durability. Minimum wet
curing duration normally is 3-days unless noted otherwise by the grout material
manufacturer.

Special Considerations for Curing

Special “means and methods” may be required based on site-specific environmental

conditions at the time of grout installation. American Concrete Institute (ACI) provides
curing guidance to achieve a long-lasting and durable grout bed. Hot and cold weather
can be complications that can be dealt with in a systematic manner when following ACI
guidelines stipulated below:

a. Hot Weather precautions/curing practices should follow ACI 305-10

“Guide to Hot Weather Concreting”
b. Cold Weather precautions/curing practices should follow ACI 306R-10
“Guide to Cold Weather Concreting”

Additional Information

ACI Committee 351 has developed two pertinent documents, ACI 351.1R-12 “Report on
Grouting Between Foundation and Bases for Support of Equipment and Machinery” [21]
and ACI 351.4-14 “Specifications for Installation of Cementitious Grouting Between
Foundations and Equipment Bases” [22]. Both are excellent resources that provide an
overview of current practices for grouting that supports equipment.

80
6.9.4 Support Structure Repairs

Repairs to existing reinforced concrete structures and concrete fireproofing covering

steel structures can be required due to the effects of deterioration caused by a harsh
and aggressive environment and service conditions that exist in delayed coking units.
The effects associated with exposure to high temperature and wet conditions, in
addition to poor original construction practices can result in damage after years of
service. Damage is most commonly related to corrosion of steel rebar in reinforced
concrete structures and steel structures covered with concrete fireproofing.
Additionally, upgrades to the unit (i.e., new slide valves, un-heading assemblies, etc.)
can add significant loads to a structure resulting in damage. Repairs can be
accomplished by following the Concrete Repair Manual [23].

Structure Assessment

The condition of existing reinforced concrete structures can be determined by

implementing methodologies developed by ACI in their Guidelines ACI 562-13 “Code
Requirements for Evaluation, Repair, and Rehabilitation of Concrete Buildings”, ACI
364.1R-07 “Guide for Evaluation of Concrete Structures prior to Rehabilitation” and ACI
201.1R-08 “Guide for Making a Condition Survey of Concrete in Service”.

Repair Surface Preparation

Once the degree, and depth and extent, of deterioration has been identified, surface
preparation prior to a repair can follow ICRI Guideline 310.1R-2008 “Guide for Surface
Preparation for the Repair of Deteriorated Concrete Resulting from Reinforcing Steel
Corrosion”:

Repair Material Selection

Repair materials should be selected so that installed material functions compositely

under load. Repair material selection should follow ACI 546.3-06 “Guide for the
Selection of Materials for the Repair of Concrete” and ICRI Guideline 320.2R-2009
“Guide for Selecting and Specifying Materials for Repair of Concrete Surfaces”.

Repair Material Installation

As each repair is unique, the selection of the application process needs to be repair
specific. The concrete repair installation method should be performed in accordance
with ICRI Guideline 320.1R-1996 “Guide for Selecting Application Methods for the
Repair of Concrete Surfaces”.

Structural Strengthening

81
 Opportunities arise when unit upgrades are planned that involve larger, heavier drums,
new de-heading valves, etc., and the existing support structure needs an enhancement.
Selecting an appropriate strengthening strategy should follow guidelines set up in ICRI
330.1-2006 “Guide for the Selection of Strengthening Systems for Concrete Structures”.

Special Considerations

Special “means and methods” may be required based on site-specific hot or cold
weather conditions that may exist at the time of concrete installation. ACI provides
guidance to achieve a long-lasting and durable concrete repair during weather
conditions that can make concrete repairs difficult.

a. Hot Weather precautions/curing practices should follow ACI 305-10

“Guide to Hot Weather Concreting”
b. Cold Weather precautions/curing practices should follow ACI 306R-10
“Guide to Cold Weather Concreting”

7 Life Extension

Once damage is observed in a coke drum and a leak occurs or it is determined either by a damage
assessment or experience that a repair is needed, it is commonly necessary to determine how long the
repair will last. Section 6 contains commonly used practices to perform a repair to a coke drum.
Frequently, a crack or other damage is observed while a coke drum is empty, but the unit is still
operating. In many cases a temporary repair is called for that needs to be performed quickly and only
needs to last until the next planned turnaround when a more permanent repair can be made. This
section of the document discusses industry experience with how long different types of repairs are
expected to last.

7.1 Temporary Repairs of Cracks

Section 6 of this document discusses various forms of weld repairs that are performed in coke drums.
Typically, temporary weld repairs on cracks in coke drums described in Section 6 involve:
• the use of a nickel based filler metal for a full thickness weld repair made from the outside
surface
• the use of filler metal that matches the base metal for a repair from the outside surface through
part of the thickness that does not affect the internal cladding
• the use of filler metal the matches the base metal for a repair from the inside surface without
restoring the cladding.
For each of these temporary weld repair procedures, it is expected that a more permanent weld repair
as defined in Section 6 will be employed at the next planned turnaround. Based on experience, it is
expected that, on average, the temporary full thickness weld repair procedure using a nickel based filler
metal as outlined in Section 6 will last 2 to 4 years before re-cracking and a possible leak. Repairs made
from the outside surface through part of the wall and not affecting the internal cladding and repairs

82
made from the inside surface using a filler metal that matches the base metal are expected to last
longer. Of course this life estimate can be much greater or even less depending largely on how the coke
drum is operated. Additional preheating before introducing hot feed and/or additional steam cooling
before introducing quench water will tend to reduce cycle loads and increase the period before cracking
and leaks.

7.2 Repairs Designed to Last an Extended Period of Time

Section 6 provides guidance for various types of weld repairs on coke drums. The guidance focuses on
providing a repair weld best suited to resist thermo-mechanical fatigue. Typically, this is achieved by
promoting the following properties for the repair weld.

1. A fine grain size in both the weld deposit and the heat affected zone on a repair weld will have
good fracture ductility which is required to resist thermo-mechanical fatigue. Typically, a fine
grain size is achieved by using a weld procedure that incorporates a low heat input by low metal
deposition rates. This is an important feature of the control deposition welding (temper bead)
procedures detailed in Appendix A.
2. A second important feature of a repair weld is to ensure that it has a hardness level that
matches existing base metal and welds in order to avoid a strength mismatch. This insures that
imposed thermal loads are not concentrated in lower strength areas adjacent to a high hardness
(strength) repair. Several steps can be taken to ensure that a strength mismatch is minimized.
1) A controlled deposition (temper bead) weld procedure can be used that maximizes interpass
tempering. 2) A lower carbon level welding consumable can be used to minimize hardness
levels in the repair weld deposit. 3) The repair weld can be PWHT’d to temper the weld and
heat affected zone.
3. A repair weld should be ground flush with the surrounding base metal in order to avoid any
mechanical reinforcement provided by a “crown” on the weld. This is commonly performed on
all welds made during the fabrication of a new coke drum.

It is noted that residual welding stresses are not mentioned as a contributing factor for thermo-
mechanical fatigue. As noted in Section 5.4, it is expected that residual welding stresses will be greatly
reduced by mechanical shakedown that occurs once a significant thermo-mechanical load is imposed on
a repair weld.

It is difficult to determine with any precision how long a repair weld will last in a coke drum. Even repair
welds with a fine grain size, minimal strength mismatch and ground flush with the base metal will crack
in a short period of time if exposed to severe thermo-mechanical fatigue conditions. Experience has
shown that changes in an operating practice such as increasing preheat before introducing hot feed or
increasing steam cooling before introducing quench water generally will result in greater life extension
than any improvement in repair weld properties.

83
7.3 Use of Structural Weld Overlay to Extend Life

Over the past three decades, automated structural weld overlays have been installed on several coke
drums worldwide. The great majority of these applications have been in service for over five years and a
significant number has been in service for 10-15 years. The method is used as a life extension option
because it could be faster and less disruptive to implement than section replacement. Although the
method does not indefinitely postpone the eventual end of life of the vessel, some owners use the
technique to manage drum life so that the eventual replacement can be executed during an opportune
time for the owner. Even though the outcome is highly dependent on the rigor of design and quality of
implementation, the structural weld overlay repair method has been found by some owners to be
logistically and financially effective in extending the life of coke drums ([24],[28]).

Because of its success and increased use, over the last ten years there has been an increase in the
number of vendors providing this service and the results have been mixed. The success of structural
overlay installations is highly dependent on several key factors:

• The availability of current vessel distortion profiles (mapping) to permit analysis and design.

• A quality engineering assessment and design of the overlay deposit that includes historical as
well as future operating plans.

• A quality installation with well-designed welding procedures and depth and surface profile
quality.

• If the extent and nature of damage found during the turnaround that is different than the
damage used to design the overlay deposit, it is necessary to make field changes to the overlay
deposit. It is essential that the overlay starts and ends in the proper location of the damage
area. Starting or terminating an overlay deposit in the wrong location could significantly affect
its effectiveness.

7.3.1 Structural Weld Overlay Methodology

The structural weld overlay is an engineered repair method consisting of the following steps:

Pre-Turnaround

• ID mapping of the vessel geometry is performed to capture the existing bulging profile of the
vessel. See Section 3.2.3. This step is typically performed well in advance of an upcoming
turnaround to allow time for analysis and planning for the overlay deposit.

• Operating parameters and history for the coker are included in the evaluation.

• An engineering analysis is performed to identify areas of primary interest for inspection and to
identify areas where there is a high probability for damage to occur during the time period of
interest to the owner. High probability areas are typically ranked to show where immediate
action should be taken and areas of concern for future turnarounds. (See Section 4)

84
 • A structural overlay is designed to mitigate bulge severity and to extend the time to crack
initiation in areas where bulging and cracking are most likely..

• Vessel design information is reviewed to ensure that proper weld procedures consistent with
Section 5 and 6 are being used.

During Turnaround

• Base material inspection is performed to identify existing cracks. Existing cracks are repaired and
inspected as needed prior to the installation of the structural overlay.

• The weld overlay area is marked out and cladding is removed by arc gouging for overlays to be
applied on the inside diameter.

• Surface inspection is performed in the overlay area and pre-heat blankets are installed.

• Overlay is deposited using a CDW procedure as discussed in Section 5.4. Overlay deposits on
internal surfaces typically are installed using ERNiCrMo-3 (Inconel 625) and overlay deposits on
the outside surface typically are made with a matching filler wire.

• Post overlay surface grinding to taper edges is performed to reduce stress risers in the deposit.

• Post weld bake-out is performed in accordance with the procedure in use.

• Final visual and surface examinations are performed.

Several laboratory tests have been performed to determine the properties of as-welded overlay
deposits as reported in References [25], [26] and [27].

85
Figure 7.3.1a Inspection and crack excavation performed prior to weld overlay repairs.

86
 Figure 7.3.1b Through-wall crack repaired.

Figure 7.3.1c Finished Structural Overlay Deposit.

87

Temperbead Base Metal Repair Deposit

7.3.2 Experience with Structural Weld Overlay

Performance of structural weld overlay is heavily influenced by quality of the design, quality of
implementation, and severity of operating conditions. Structural weld overlay has been applied on
bulges on several coke drums based on bulge criteria calculations as described in Section 4.2.

The following observations were provided by owner/operators that have utilized weld overlay repairs of
bulged areas on coke drums.

Operator 1:

Eight coke drums were commissioned in 1999. The first application of the Alloy 625 weld overlay was
performed in 2010, followed by a second application in 2015. In total 6 of the 8 drums contained Alloy
625 weld overlay in bulged areas on the drums, while the remaining 2 drums did not contain any weld
overlay. It appears the weld overlay did extend the life of the 6 drums; however, it is not possible to
quantify how effective the application of weld overlay is in extending life of these 6 drums. The
following trends were observed in the 6 drums that contained Inconel 625 weld overlay:

• The first crack in the weld overlay surface was observed 3 to 4 years after it was applied. After 5
years of service, cracking in the weld overlay became more widespread and significantly more
pronounced.

• Bulging at the overlaid locations initially appeared to be reduced but over time bulging
increased with subsequent operation and cycling.

• Inspections performed after 4 years of service after application of the overlay found a number
of cracks on the overlay surface ranging in length from 75 to 1000 mm with some of the cracks
penetrating through wall.

• Inspection of the overlay from the OD surface was ineffective, while inspection from the ID was
challenging over the entire overlay surface and peripheral areas.

• Cracks were observed on both the overlaid shell course and at circumferential seams which
typically contained the bulge peak. Cracking occurred in the overlay at termination points, even
at termination points with a 10:1 taper.

• Initially, it was thought that surface finish was a primary factor for a successful overlay
application; however, it was noted that cracking also was observed at locations with a smooth
ground surface.

• Hardness measurements were made on the Alloy 625 overlay. The as-welded deposit had a
Vicker hardness of Hv 250 while the weld overlay deposit after extended high temperature
service had a Vickers hardness reading between Hv 350 and Hv 450.

• Currently, operator plans to replace bulged drums with no plans to use weld overlay to extend
life.

88
Operator 2:

A second operator applied Alloy 625 weld overlay in bulged areas on several drums. Their experience
showed the following trends:

• Immediately after the application of the weld overlay, it appears crack initiation and
propagation was slowed down. However, after 4 to 5 years of service, crack initiation and
propagation appeared to accelerate significantly.

• Cracking occurred in both the weld overlay and areas immediately adjacent to the weld overlay.
It appears reinforcement created by the weld overlay increases cracking in areas adjacent to the
overlay.

• The shortest time between installation of the weld overlay and through wall cracking at or near
the weld overlay was 5 years.

• In severely bulged areas, it was difficult to control bead placement and surface finish/contours
resulting in locations of stress concentration prone to crack initiation.

• It appears there was a corrosion fatigue cracking mechanism occurring at the interface between
the 12 Cr cladding and Inconel 625 overlay.

Operator 3:

A third operator has used Alloy 625 weld overlay in bulged areas on several drums since the early
2000’s. Their experience has been in multiple drums built with a variety of metallurgies, including
carbon steel, 1Cr-1/2Mo, 1-1/4Cr-1/2Mo, and 2-1/4Cr-Mo base metal showed the following trends:

• All weld overlay in coke drums for cladding restoration and bulge reinforcement are done
following a methodology developed after years of experience. For each repair, the complete
review of the drum history is the first step of the repair process, and includes a detailed analysis
of inspection files, laser scans, and a detailed bulge assessment. Weld Overlay procedures and
parameters are selected based on base metal type, bulge dimensions (weld overlay dimensions),
and drum age, and they are discussed with the selected contractors (this operator only uses
selected specialty welding contractors with long term coke drum repair experience). Procedures
include detailed inspections before complete removal of the internal cladding, inspection after
cladding removal and before application, inspection after weld overlay application, and after
PWHT (when applied)

• All weld overlay reinforcement projects have included a previous engineering analysis with the
intent of optimizing the coking cycles to reduce damage accumulation rate.

• Weld overlay has been completed with success in areas severely bulged, requiring a good
preparation of the repaired area and the application of multiple layers of Alloy 625. Bulging rate
at the repaired area decreased and through wall cracking at or near the weld have not been

89
 observed. The experience of this operator has shown that bulging moved to areas away from
the repair.

• Inspections at major turnarounds of the weld overlay reinforced areas includes visual inspection
of the weld overlay area, PT examination and ultrasonic testing at the termination of the
reinforcement area (taper). Online inspection of the taper from the external surface has been
done only in the situation where previous cracking was an issue in the weld overlay area or
when the weld overlay includes shell welds. Cracking has not been found after years in service
since weld overlay repairs have been following the current procedures that includes QA/QC.

• Hardness is controlled during and after weld overlay application. The operator has never seen
an increase of hardness in the repaired area after several years of service (up to 11 years in
currently installed repaired drums). In drums that have been replaced with new equipment due
to aging, issues with in-service increase of hardness was never observed.

• Corrosion fatigue cracking was found in one of the old weld overlay repairs (no reinforcement
area, but a weld crack repair) because there was an area where the baking material at the
interface between of the 12 Cr cladding and Alloy 625 overlay was left exposed to the service
without protection. The cracking was found during a turnaround and it was not through wall.
The issue was resolved by eliminating the cracking and applying the Inco 625 overlay but
extending it in the top of the original cladding and providing an adequate taper.

• In one particular case, weld overlay was applied in a set of 4 drums to extend their service after
finding unusual internal corrosion in the top section of the drums (refer to API 934G); thickness
readings were found to be below 50% of the original minimum thickness; at the beginning and
due to time constraint, the overly was applied to restore thickness on the external surface and
using backing material matching electrode; after more than a year in service and in order to
prevent severe thinning due to internal surface corrosion, internal surface weld overlay was
applied using Alloy 625. The set of drums operated successfully for several years until drums
were replaced due to aging (more than 25 years in service).

• One of the most recent applications was done in 2016 in a set of 4 drums that have been in
service for almost 45 years. After some changes done in this set of drums in the late 2000’s, at
least 7 through wall crack in period of 4 years starting in 2012. After some operational
adjustments and the major weld overlay reinforcement project in 2016, there have not been
through wall crack events. The set of drums are in continued monitoring that includes laser
scans to track bulge performance. These drums were set to be replaced in the next 2-3 years at
the time weld overlay was applied.

• The oldest weld overlay application has lasted 11 years without cracking or severe bulging of the
reinforced area. A combination of weld overlay bulged reinforcement and can replacement was
necessary in this set of 4-drums only after 9 years of service due to very short coking cycle,
uncontrolled quench, some fabrication issues. This major repair was accompanied with cycle
optimization. After 11 years of the repair and adjustment, a slight increase in bulging continues

90
 but away from weld overlay; there has not been through wall crack events in these drums since
the original weld overlay reinforcement repair.

• Currently, this operator plans to continue using weld overlay for crack repairs, thickness
restoration, and bulged reinforcement of coke drums. The operator considers this a long-term
temporary repair that is useful in the late stages of coke drum life and allows preparing strategic
replacement of their installed based.

8 Bibliography

The bibliography contains a listing of all of the references imbedded in this document.

[1] Seijas, A., Bedoya, J., Stoller, A., Perez, O., Marcano, L. (2017) “Assessment, Mitigation,
Management and Extension of Coke Drum Life Through Equipment Health Monitoring Systems and
On-Line Inspection”, Proceedings of the 2017 ASME Pressure Vessels and Piping Division
Conference, PVP 2017-65903.
[2] Penso, J., Bedoya, J., Bouse, S., Ramamoorthy, S. (2017) “Assessment of the Critical Tilting Angle of
a Coke Drum Vessel Subject to Seismic Loading”, Proceedings of the 2017 ASME Pressure Vessels
and Piping Division Conference, ASME, PVP 2017-65870.
[3] Samman M. “Bulging Patterns of Coke Drums.” Proceedings of the 2016 ASME Pressure Vessel and
Piping Conference, ASME, PVP2016-63812.
[4] 1996 API Coke Drum Survey, 2003, American Petroleum Institute, Washington, DC.
[5] Samman, M. and Du Plessis, P. (2007) “The Bulging Intensity Factor (BIF) - A technique for
assessing the bulging severity of coke drums.” Proc. 2007 NPRA Reliability & Maintenance Conf.,
RMC-07-100, National Petrochemical & Refiners Association, Houston, TX.
[6] Egler A., Daryl R., Darren L., Stephen P., Rick C., Jason D., 2018 “Correlating Coke Drum Profiles
with Observed Surface Damage”. Pressure Vessels & Piping Division Conference, ASME, PVP2018-
84766, Prague, Czech Republic.
[7] Samman, M., Tinoco, E. B., and Marangone, F. C. (2014) “Comparison of stress and strain analysis
techniques for assessment of bulges in coke drums.” Proc. Pressure Vessel and Piping Conference,
ASME, PVP2014- 28139.
[8] Steel Construction Manual, American Institute of Steel Construction, Chicago.
[9] Ramos AJ, Rios C, Vargas JAR. Fatigue life prediction of delayed coke drums. Vision Technology
1999;6:93–100.
[10] Chen, J. and Xia, Z. (2014) “A fatigue life prediction method for coke drum base, weld, and HAZ
materials from tensile properties” Elsevier Materials and Design 63, 575–583.
[11] Penso, J. A. and Hazime, R. (2010) “Comparison of Thermo-Mechanical Fatigue Life Assessment
Methods for Coke Drums” Proc. Pressure Vessel and Piping Conference, ASME, PVP2010-25810.
[12] NRIM Fatigue Data Sheet No. 28, “Data Sheets on Elevated-Temperature, Low-Cycle Fatigue
Properties of SCMV 3 (1.25Cr-0.5Mo) Steel Plates for Pressure Vessels,” National Research Institute
for Metals, Tokyo, Japan. 1981.
[13] Kay, A., Lester, P., and Barborak, D., 2017, “Weld Repair of C, Cr-Mo Cokedrums (& Pressure
Vessels) without PWHT”, Proc. of the ASME 2017 Pressure Vessel & Piping Division, ASME,
PVP2017-65161.
[14] Samman, M. (2017) “New Tools for Addressing Coke Drum Failures”, API Inspection Summit,
Galveston.
[15] Schmidt, M. and Samman, M. (2012). “Repair and Retrofit of a Coke Drum Skirt Attachment Weld”,
Coking.com Meeting, Fort McMurray.
[16] American Concrete Institute Standard and Report Building Code Requirements for Structural
Concrete (ACI 318-14) Chapter 17, Anchoring to Concrete
[17] Wind Loads and Anchor Bolt Design for Petrochemical Facilities, ASCE, Copyright 1997, USA,
Library of Congress Catalog Card No. 97-20890, ISBN 0-7844-0262-0
[18] Kline, Thomas Anchor Bolt Repair in Petrochemical Environments Hydrocarbon Engineering
Magazine, September 2013

91
[19] Samman M. (2016) "Anchor Bolt Failures in Coke Drums." Proc. Pressure Vessel and Piping
Conference, ASME, PVP2016-63813.
[20] Samman, M. and Kaye, A. (2018) "A Non-Bolted Restraint For Coke Drums", Proc. Pressure Vessel
and Piping Conference, ASME, PVP2018- 84734
[21] ACI 351.1R-12 “Report on Grouting Between Foundation and Bases for Support of Equipment and
Machinery”
[22] ACI 351.4-14 “Specifications for Installation of Cementitious Grouting Between Foundations and
Equipment Bases”
[23] Concrete Repair Manual, American Concrete Institute (ACI) and the International Concrete Repair
Institute (ICRI).
[24] Du Plessis, P. and Samman M. (2017) “A Successful Strategy for Managing the Mechanical
Integrity of Coke Drums.” ASME, PVP2017-65066.
[25] Tinoco, E. B., Marangone, F. C., and Bedoya, J. (2014) “Fatigue Testing of Coke Drum Plates
Reinforced with Weld Overlay”, Coking.com, Rio de Janeiro, Brazil.
[26] Samman, M., Williams, E., and Gemmel, G. (2015) “Structural Vessel Repairs Using Automated
Weld Overlays”, IPEIA Meeting, Banff.
[27] Williams, E. and Samman, M. (2013) “Automated Weld Overlay Repairs of Large Damaged
Equipment”, API Inspection Summit, Galveston.
[28] Samman, M. and Ajmera G. (2018) “Performance of Bulges after Weld-Overlay Repairs”,
Pressure Vessels & Piping Conference, ASME, PVP2018- 84736.

92
Appendix A: Controlled Deposition Welding (CDW)/Temper
Bead Welding (TBW) Using a SMAW Process and a GMAW
Process

93
 Appendix A: Controlled Deposition Welding (CDW)/Temper Bead Welding (TBW) Using a SMAW
Process and a GMAW Process

A.1 CDW/TBW Using a SMAW Welding Process

The first 3 passes of a typical CDW/TBW procedure should use the welding parameters as illustrated in
Table A.1. These welding parameters are listed as a typical example of the parameters listed in a
CDW/TBW procedure for a SMAW process. Variations on these parameters can be used to achieve
similar results.

Table A.1: Typical Welding Parameters for the First 3-Passes of a CDW/TBW procedure using a
SMAW process

1st LAYER 2nd LAYER 3rd LAYER

WELDING PROCESS SMAW SMAW SMAW

ELECTRODES TYPE As appropriate for base metal. See Table 5.1

3/32" 1/8" 1/8"

ELECTRODES DIAMETER
(2.5 mm) (3.2 mm) (3.2 mm)

CURRENTS 80 - 85 Amps. 120 - 125 Amps. 125 - 130 Amps.

VOLTAGE 20 - 21 Volts 20 - 22 Volts. 20 - 22 Volts.

9.5 - 10.0 kJ/in 17.0 - 17.5 kJ/in 19.5 kJ/in

HEAT INPUT
(3.7 – 3.9 kJ/cm) (6.7 – 6.9 kJ/cm) (7.7 kJ/cm)

PREHEAT As appropriate for base metal. See Table 5.3

INTERPASS TEMPERATURE 550°F (290°C) maximum

10.5" (26.5 cm) while burning 12" (30.0 cm)

RUN-OUT LENGTH
of electrode. (For all position welding)

BEAD TYPE Stringer beads - electrode tip aimed at toe of

94
 prior pass to achieve 50% bead overlap

550°F (290°C) for two hours, prior to allowing the

POST HEAT
weld to cool down below the preheat temperature.

Details of Performing CDW/TBW Using SMAW Process

1. The excavation for a CDW/TBW repair should be about 2.5 times wider at the surface than it is
deep as illustrated in Figure A.1

2.5 d

d
Figure A.1: The excavation for a CDW/TBW repair
should have an opening that is 2.5 times the
depth.

2. Prior to performing any welding the excavated area should be 100% inspected using MT
examination.
3. The initial layer of weld metal should butter the entire excavation area and extend over onto the
unexcavated area (See Figure A.2). 3/32" (2.5 mm) diameter electrodes should be used for the
first layer.

Figure A.2: Application of first layer using CDW/TBW.

4. A dry powder MT inspection should be performed while the weldment is at preheat

temperature to insure no cracking has occurred when applying the first butter layer of weld
metal. Grinding may be required to remove surface irregularities which can mask indications. If
cracking does occur, it is advisable to increase the preheat temperature and consult a welding
specialist.

95
5. The second and third layers should be deposited using 1/8" (3.2 mm) diameter electrodes
covering the entire first layer (See Figure A.3). Note that the outer beads of these 2 layers
should not be any closer than 1/8" (3.2 mm) to the base metal.

Note: The use of an ammeter (amps) or a multimeter (amps/volts meter) may be

required to verify and monitor that the specified ranges for current and voltage are
followed. Documentation of these measurements may be required by the site.

Figure A.3: Application of second and

third layer using CDW/TBW

6. The remaining groove should be filled using normal bead sequencing as shown in Figure A.4.
After the 3 layers of weld metal have been deposited, the remaining fill passes can be deposited
using conventional SMAW welding parameters.

Figure A.4: The remaining fill passes are

made using normal bead sequencing.

7. The completed weld should be MT inspected using dry powder while the weld is still at preheat
temperature.
8. The preheat temperature should be increased to approximately 550°F (290°C) and hold it for a
minimum of two hours to allow any hydrogen remaining in the weldment to diffuse out.
9. The completed weld should be cooled slowly to ambient temperature by keeping or installing
insulation around the repair area.
10. The weld crown should be ground flush with the surrounding metal surface to remove any weld
reinforcement. The grind marks should be oriented perpendicular to the long axis of the weld,
so as not to inhibit fatigue crack initiation. Figure A.5 shows the completed welding after
grinding it flush with the surrounding metal surface.

96
 Figure A.5: Completed weld with crown ground off

11. Visual, MT and UTSW inspections should be performed immediately after grinding to detect
welding flaws. This same inspection combination should be repeated 24 to 48 hours after
reaching ambient temperature to detect delayed hydrogen cracking.

As defined in the provided repair specification, hardness testing should be performed on the completed
weld after the crown is removed. Typically, hardness readings are taken across the base metal, HAZ and
weld deposit using the Vickers method with a 10 kg load. Acceptable maximum hardness levels
established in repair specifications vary from 250 to 300 Vickers or equivalent. Some repair
specifications do not set an acceptable maximum hardness level, but require that hardness
measurements be taken for information.

A.2 CDW/TBW Using a GMAW Welding Process

The first 3 passes of a typical CDW/TBW procedure should use the welding parameters as illustrated in
Table A.2. These welding parameters are listed as a typical example of the parameters listed in a CDW
procedure for a machine GMAW/TBW process. Variations on these parameters can be used to achieve
similar results.

Table A.1: Typical Welding Parameters for the First 3-Passes of a CDW/TBW procedure using a
GMAW process

:
1st LAYER 2nd LAYER 3rd LAYER

WELDING PROCESS Machine GMAW

ELECTRODE TYPE As appropriate for base metal. See Table 5.1

ELECTRODE 0.045 in.

DIAMETER (1.2 mm)

CURRENT 185 - 220 Amps 190 - 240 Amps 190 - 240 Amps

VOLTAGE 23 - 26 Volts 20 - 26 Volts 22 - 27 Volts

25 in./min. 25 in./min. 25 in./min.

TRAVEL SPEED
(10.6 mm/sec.) (10.6 mm/sec.) (10.6 mm/sec.)

97
 230 – 295 in./min. 260 – 330 in./min. 270 – 320 in./min.
WIRE FEED SPEED
(97.4 – 124.9 mm/sec.) (110.1 – 139.7 mm/sec.) (114.3 – 135.5 mm/sec.)

12.4 kJ/in. max. 13 kJ/in. max.1 14.3 kJ/in. max.2

HEAT INPUT
(0.49 kJ/mm max) (0.51 kJ/mm max.) (0.68 kJ/mm max)

PREHEAT As appropriate for base metal. See Table 5.3

INTERPASS
450°F (232°C) maximum
TEMPERATURE

Stringer beads - 50% bead overlap. Subsequent weld layers shall be no

BEAD TYPE further than 1/8 in. from the edges of the previous layer including starts and
stops.

450°F (232°C) for two hours, prior to allowing the

POST HEAT
weld to cool down below the preheat temperature.
1. Per QW-409.29 of ASME BPVC.IX-2015, An increase or decrease in the ratio of heat input between the first
tempering bead layer and the weld beads deposited against the base metal shall not exceed 20% for P‐No. 1 and P‐
No. 3 metals and 10% for all other P‐Number metals.
2. Per QW-409.29 of ASME BPVC.IX-2015, An increase or decrease in the ratio of heat input between the second
tempering bead layer and the first tempering bead layer of more than 20% for P‐No. 1 and P‐No. 3 metals and 10%
for all other P‐Number metals.

The details for performing a CDW/TBW procedure using a machine GMAW process will follow similar
steps as performed for a CDW/TBW procedure using a SMAW process as illustrated in steps 1 through
12.

98