NDT - Auto RT - Hois

Recommended Practice for the in-service
inspection of wall loss in pipes by computed

radiography

HOIS(09)RP1 Issue 1

A Report prepared for HOIS

By

S F Burch, ESR Technology
steve.burch@esrtechnology.com

December 2009

HOIS(09)RP1 Issue 1
I
HOIS(09)RP1 Issue 1

Authorisation Sheet

Report Title: Recommended Practice for the in-service inspection of
wall loss in pipes by computed radiography
Customer Reference: HOIS
Project Reference: D7888219
Report Number: HOIS(09)RP1
Issue: Issue 1
Distribution List: Open publication

Author: S F Burch

12.12.09
Checked: B A Stow

15.12.09

Address for correspondence

Dr S F Burch
ESR Technology Ltd
16 North Central 127
Milton Park
Abingdon
Oxfordshire
OX14 4SA
UK

Phone: +44(0)1235 213402
Email: steve.burch@esrtechnology.com

COPYRIGHT ESR Technology Ltd

This report is the Copyright of ESR Technology Ltd and has been prepared by ESR Technology Ltd for the HOIS
JIP. ESR Technology Ltd accepts no liability whatsoever to any third party for any loss or damage arising from
any interpretation or use of the information contained in this report, or reliance on any views expressed therein.

II
HOIS(09)RP1 Issue 1
III

HOIS(09)RP1 Issue 1
IV

Executive Summary

This recommended practice gives guidance for the in-service inspection of pipes for wall loss
type flaws (e.g. corrosion pitting, generalised corrosion and erosion) using computed (digital)
radiography (CR). Weld inspection for typical welding process induced flaws is not covered,
but weld inspection is included for corrosion/erosion type flaws (e.g. weld root
erosion/corrosion).

The pipes may be insulated or not, and can be assessed where loss of material due, for
example, to corrosion or erosion is suspected either internally or externally.

This document covers the following inspection techniques:

1. Double-wall single image (DWSI) radiography for the inspection of discrete wall loss
flaws by their effects on image grey level.

2. Double-wall double-image (DWDI) radiography for the inspection of discrete wall loss

3. Tangential inspection techniques for detection and through-wall sizing of wall loss,
including with the source on the pipe centre line, and offset from it by the pipe radius.
This technique is sometimes referred to as profile radiography, but this term is not
used in this document.

Note that DWDI is often combined with tangential radiography with the source on the pipe
centre line.

The recommendations cover the main radiation sources used for in-service inspection, i.e.
Iridium 192, Selenium 75, and for some specialised thick wall applications, Cobalt 60. The
use of portable X-ray sources is also included.

Two different qualities of radiography are considered in this document:

A standard quality of computed radiography for wall loss inspection. This has less
demanding quality requirements than those defined for weld inspection, since in general wall
loss flaws are easier to detect radiographically than welding flaws which can include cracks.

A higher quality of computed radiography for wall loss inspection is also included in this
document. This is for CR inspections requiring higher quality (e.g. inspection of small pitting
flaws).

Recommendations are given for these two different quality classes on source to detector
distances, axial and circumferential coverage, CR image quality measures and indicators,
and exposure times. In addition to these parameters, for tangential radiography, different
methods for dimensional measurements are covered.

HOIS(09)RP1 Issue 1
V

HOIS(09)RP1 Issue 1
VI

Contents

1 INTRODUCTION ..........................................................................................10
2 REFERENCES .............................................................................................11
3 DEFINITIONS...............................................................................................12
4 PERSONNEL QUALIFICATIONS................................................................13
5 GENERAL ....................................................................................................13
5.1 Protection against ionising radiation........................................................... 13
5.2 Size and Strength Of Sources.................................................................... 13
5.3 Source Containers and Collimation............................................................ 13
5.4 In-Situ Inspection of Plant........................................................................... 14
5.5 Identification of Radiographs...................................................................... 14
6 OVERVIEW OF INSPECTION TECHNIQUES .............................................15
6.1 Double wall single image (DWSI) inspection technique ............................. 15
6.2 Double wall double image (DWDI) inspection technique............................ 15
6.3 Tangential inspection techniques ............................................................... 17
6.3.1 Source on pipe centre line ....................................................... 17
6.3.2 Offset source position tangential radiography.......................... 18
6.4 Both Tangential and Double wall techniques combined............................. 19
7 RADIATION SOURCES ...............................................................................20
7.1 Type of source............................................................................................ 20
7.2 Source selection......................................................................................... 21
7.2.1 Double wall techniques............................................................ 21
7.2.2 Tangential technique................................................................ 22
8 SOURCE TO DETECTOR DISTANCES (SDD) ...........................................27
8.1 Double wall single image (DWSI) inspection.............................................. 27
8.2 Double wall double image (DWDI) inspection ............................................ 28
8.3 Tangential Inspection ................................................................................. 31
8.3.1 Source on pipe centre line ....................................................... 31
8.3.2 Combined tangential/DWDI image radiography....................... 33
8.3.3 Source offset from pipe centre line .......................................... 33
HOIS(09)RP1 Issue 1

VII
9 CIRCUMFERENTIAL AND AXIAL COVERAGE AND OVERLAP...............35
9.1 Circumferential coverage............................................................................ 35
9.1.1 DWSI........................................................................................ 35
9.1.2 DWDI ....................................................................................... 36
9.1.3 Penetrated Thickness measurement ....................................... 37
9.2 Axial coverage............................................................................................ 37
9.3 Overlap of images ...................................................................................... 38
10 CR IMAGE QUALITY INDICATORS............................................................39
10.1 Background ................................................................................................ 39
10.2 Image Quality Measures for Double Wall Techniques (DWDI & DWSI)..... 39
10.2.1 Target Grey level range ........................................................... 39
10.2.2 Signal to Noise Ratio (SNR) .................................................... 40
10.2.3 BSR Measurement................................................................... 41
10.2.4 Wire IQIs.................................................................................. 45
10.2.5 Image quality for tangential techniques ................................... 48
11 SCREENS/FILTERS.....................................................................................50
12 EXPOSURE TIME ........................................................................................51
12.1 DWSI and DWDI......................................................................................... 51
12.2 Tangential................................................................................................... 52
13 PENETRATED THICKNESS MEASUREMENTS.........................................53
13.1 Introduction................................................................................................. 53
13.2 Principle of method..................................................................................... 53
13.3 Effects of scattered radiation...................................................................... 53
13.4 Calibration using step wedge ..................................................................... 53
13.5 Key Points .................................................................................................. 54
13.6 Limitations .................................................................................................. 54
14 TANGENTIAL RADIOGRAPHY...................................................................56
14.1 Recommended SDD................................................................................... 56
14.2 Source location relative to pipe centre line................................................. 56
14.3 Dimensional calibration .............................................................................. 57
14.3.1 Measurement of distances....................................................... 58
14.3.2 Dimensional comparator .......................................................... 61
14.4 Use of lead strips to avoid burn-off............................................................. 62
14.5 Wall thickness measurement...................................................................... 62
HOIS(09)RP1 Issue 1

VIII
14.5.1 Interactive on-screen measurements....................................... 62
14.5.2 Grey-level profile analysis methods ......................................... 63
15 SCANNER PARAMETERS, IMAGE RECORDING AND PROCESSING....65
15.1 Scanner parameters................................................................................... 65
15.2 Image recording and storage...................................................................... 65
15.3 Image processing ....................................................................................... 66
15.4 Monitor viewing conditions ......................................................................... 66
16 ACKNOWLEDGMENTS...............................................................................67
17 REFERENCES .............................................................................................67

HOIS(09)RP1 Issue 1
IX

HOIS(09)RP1 Issue 1
10
1 Introduction
The scope of this recommended practice covers the in-service inspection of pipes for wall
loss type flaws (e.g. corrosion pitting, generalised corrosion and erosion) using computed
(digital) radiography (CR). Weld inspection for typical welding process induced flaws is not
covered, but weld inspection is included for corrosion/erosion type flaws (e.g. weld root
erosion/corrosion).

The pipes may be insulated or not, and can be assessed where loss of material due, for
example, to corrosion or erosion is suspected either internally or externally.

This document covers the following inspection techniques:

1. Double-wall single image (DWSI) radiography for the inspection of discrete wall loss

2. Double-wall double-image (DWDI) radiography for the inspection of discrete wall loss

3. Tangential inspection techniques for detection and through-wall sizing of wall loss,
including (a) with the source on the pipe centre line, and (b) offset from it by the pipe
radius. This latter technique is sometimes referred to as profile radiography, but this
term is not used in this document.

Note that DWDI is often combined with tangential radiography with the source on the pipe
centre line.

The recommendations cover the main radiation sources used for in-service inspection, i.e.
Iridium 192, Selenium 75, and for some specialised thick wall applications, Cobalt 60. The
use of portable X-ray sources is also included.

Two different qualities of radiography are considered in this document.

A standard quality of computed radiography for wall loss inspection. This has less
demanding quality requirements than those defined for weld inspection, in EN 1435 for
example, since, in general, wall loss flaws are easier to detect than typical welding flaws. For
tangential radiography, standard quality is sufficient when the wall loss is approximately
uniform, not isolated pitting.

A higher quality of computed radiography for wall loss inspection is also specified in this
document. This is for CR inspections requiring higher quality (e.g. inspection of small pitting
flaws). For tangential radiography, higher quality is recommended when there is a
requirement to size pitting flaws.
HOIS(09)RP1 Issue 1
11

2 References
- EN 444, Non-destructive testing - General principles for radiographic examination of
metallic materials using X-rays and gamma-rays.
- EN 473, Non-destructive testing - Qualification and certification of NDT personnel -
General principles
- EN 584-1, Non-destructive testing - Industrial radiographic film - Part 1: Classification
of film systems for industrial radiography
- EN 462-1 to EN 462-5, Non-destructive testing Parts 1 to 5: Image quality of
radiographs.
- EN 1435, Non-destructive examination of welds Radiographic examination of
welded joints
- EN 14784-1, Non-destructive testing - Industrial computed radiography with phosphor
imaging plates - Part 1: Classification of systems
- EN 14784-2, Non-destructive testing - Industrial computed radiography with storage
phosphor imaging plates - Part 2: General principles for testing of metallic materials
using X-rays and gamma rays
- E 1647-98a, ASTM Standard Practice for Determining Contrast Sensitivity in
Radioscopy.
- ISO 3999-1, Radiation protection Apparatus for industrial gamma radiography
- ISO 11699-1, Non-destructive testing - Industrial radiographic film - Part 1:
Classification of film systems for industrial radiography.
HOIS(09)RP1 Issue 1
12

3 Definitions
Terms used in this document are as follows:

b The distance between the source side of the pipe and the detector.
b The distance between the centre of the pipe and the detector.
CR Computed radiography
d Source size for calculation of geometric unsharpness
DWDI Double wall double image radiographic technique
DWSI Double wall single image radiographic technique
E
w
The exposure time in sec for penetrated thickness w
E
0
The exposure needed to achieve the required SNR_N value for zero
penetrated thickness in units of Ci.min @ 500mm.
f Distance from source to relevant position in component (depends on
technique)
ID Pipe inside diameter
OD Pipe outside diameter
R Distance from detector to pipe axis (centre line)
r Radius of pipe OD (=OD/2)
S Source strength
SDD Source to detector (originally film) distance
SPD Source to pipe axis distance (i.e. source to centre line of pipe)
T Exposure time
Ug Geometric unsharpness at detector
Ug Geometric unsharpness in plane of interest within component
w Penetrated thickness
w
t
Total steel equivalent penetrated thickness, including an allowance for any
product in the pipe
w
max
Maximum penetrated thickness for a pipe which occurs for a tangent to the ID
surface
WT Actual wall thickness
WT Measured wall thickness
Effective material attenuation coefficient

HOIS(09)RP1 Issue 1
13

4 Personnel Qualifications
Personnel qualifications should be in accordance with EN 473 or ISO 9712. Training in
computed radiography is also recommended, if available.

5 General
5.1 Protection against ionising radiation
WARNING Exposure of any part of the human body to X-rays or gamma-rays can be highly
injurious to health. Wherever X-ray equipment or radioactive sources are in use, appropriate
legal requirements must be applied.

Local or national or international safety precautions when using ionizing radiation shall be
strictly applied.

5.2 Size and Strength Of Sources
The strengths (activities) of isotope sources used for CR need to comply with local
regulations.

Radiography contractors should state the maximum strength isotopes within their Local
Rules as required by IRR 1999.

For in-service inspection applications, a typical size of Ir 192 source is a 2 x 1mm disc (20Ci,
740Gbq), whereas for a similar strength Se 75 source, the size is often 2.5 mm (near
spherical).

The effective source size for geometric sharpness calculations should be used to calculate
the required source to detector distance, as given in Section 8.

5.3 Source Containers and Collimation
The source containers should conform to the requirements for source containers given by
ISO3999-1:2000 or BS5650:1978 ISO 3999-1977 and any applicable national standards.

Conventional projection equipment can be used, provided the requirements of the current
radiation safety regulations are complied with. For these systems, a large radiation controlled
area is normally needed, which often requires out of hours working, or even shutdown of
plant.

Systems which keep the source within a single container or single container/collimator
assembly are now available that allow a much smaller size of controlled area (of order 1 m -
5 m). These systems reduce radiation doses to operators, and the small size of the
controlled area generally means that plant operation does not need to be interrupted when
site radiography is underway.

HOIS(09)RP1 Issue 1
14
The ratings of a source container must be checked for compliance with the type and strength
of the isotope to be used.

The selection of an appropriate source container and deployment system depends on the
balance of the above factors for each individual site and inspection application, together with
economic considerations.

Careful collimation of the sources is recommended to minimise unwanted radiation, and to
reduce the effects of scatter on the radiograph.

5.4 In-Situ Inspection of Plant
Use of gamma-ray radiography equipment for in-situ inspection of plant involves significant
safety issues associated with the use of ionising radiation. The appropriate mandatory safety
regulations appropriate to the plant must be adhered to (IRR 1999 in the UK). These include
the construction and maintenance of radiation controlled areas, by means of appropriate
barriers.

Pre-planning of the inspection work to be carried out on a plant is required, to include both a
risk assessment and a practical assessment of how the source container and shielding will
be placed (IRR 1999).

5.5 Identification of Radiographs
Symbols shall be affixed to each section of the object being radiographed. The images of
these symbols shall appear on the CR image outside the region of interest where possible,
and shall ensure unequivocal identification of the section.

HOIS(09)RP1 Issue 1

6 Overview of inspection techniques
6.1 Double wall single image (DWSI) inspection technique
In the double wall single image inspection technique, illustrated in Figure 6.1, the
radiography is usually carried out with the source in contact with or close to the pipe wall.
The detector is placed adjacent to the opposite pipe wall and wrapped around the pipe OD.

X or gamma-ray source
Pipe
Detector
Corrosion
Image of wall loss on
detector

Figure 6.1 Double wall single image (DWSI) inspection technique for the inspection
of wall loss in one wall of the pipe by image grey level variations.

External or internal wall loss pitting type flaws in the pipe wall adjacent to the detector are
detected by the changes in image grey level they produce. Wall loss flaws in the opposite
pipe wall nearer the source are either outside the radiation beam, or highly blurred and not
generally detectable. Large scale loss of wall, due for example to erosion, which produces a
near uniform loss of wall will not be readily detectable with this technique.

The amount of wall loss affects the image grey level, and computer analysis of the CR image
grey levels can be used to estimate the change in penetrated thickness, provided certain
procedures are followed (see Section 13).

6.2 Double wall double image (DWDI) inspection technique
In the double wall double image (DWDI) inspection technique, illustrated in Figure 6.2, the
radiography is usually carried out with the source directly in-line with the centre of the pipe.
For the double-wall double image method, a sufficiently large SDD is used to ensure that
external or internal wall loss can be detected in the portion of the pipe wall nearer the source,
as well as that closer to the detector.

15
HOIS(09)RP1 Issue 1
Pipe
Detector
Corrosion pits
Images of wall loss on
detector

Figure 6.2 Double wall double image DWDI inspection technique for detection of
wall loss in both walls of the pipe by image grey level variations.

This method decreases in effectiveness towards the edges of the pipe, since the sensitivity
will be reduced as the amount of metal penetrated by the radiation beam increases. See
Section 9 for more information on circumferential coverage.

With this method, the size of any pitting type wall loss flaws in the circumferential and axial
directions can be measured directly from the CR image, provided methods are used for
calibration of distances see Section 14.3.

On a single DWDI image, it is generally not possible to determine which side of the pipe the
flaws are located on, and flaws from both sides are superimposed on the same image.

As with DWSI, large scale loss of wall, due for example to erosion, which produces a near
uniform loss of wall will not be readily detectable with this technique.

The amount of wall loss affects the image grey level, and computer analysis of the CR image
grey levels can be used to estimate the change in penetrated thickness, provided certain
procedures are followed (see Section 13).

16
HOIS(09)RP1 Issue 1
6.3 Tangential inspection techniques
6.3.1 Source on pipe centre line
The tangential inspection technique is used for the inspection of the portion of the pipe wall
running tangentially to the radiation beam, as illustrated in Figure 6.3 for a source position on
the pipe centre line. The CR image then shows a direct image of the pipe wall.

Detector
Extended area of
corrosion
Image of reduced
thickness pipe wall
Image of normal thickness
pipe wall

Figure 6.3 Principle of tangential radiography with the source on the pipe centre-
line

The through-wall extent of extended areas of either internal or external wall loss can be
measured directly from the CR images, provided appropriate calibration techniques are used
to allow for the enlargement (blow-up) of the CR image see Section 14.3 for further
details.

The tangential method inspects only a small extent of the circumference of the pipe for a
single source/detector position, and so full coverage can only be achieved by use of a
number of CR images taken circumferentially around the pipe.

The tangential method is often used to measure directly any wall loss identified by the DWDI
method given in Section 6.2, by rotating the source/detector axis to align the wall loss with
the tangential position.

17
HOIS(09)RP1 Issue 1
The tangential radiography method is recommended for the detection and through-wall sizing
of extended areas of wall loss. For small isolated wall loss pits care should be taken to avoid
underestimation of the maximum loss of wall caused by incorrect geometric alignment (the
pit needs to be as close as possible to the tangent position).

Small isolated pits can also be difficult to detect and size using the tangential technique,
particularly if the maximum penetrated thickness at the tangent position is close to the limits
recommended for the pipe and radiation source (see Section 7.2.2 for further information).
For this reason, use of the higher quality standard for tangential radiography is
recommended for applications involving sizing of pitting flaws.

6.3.2 Offset source position tangential radiography
The tangential inspection technique with an offset source position is illustrated in Figure 6.4.
The offset is usually equal to the mean pipe radius, so that the centre of the X-ray beam
passes through the tangential position on the pipe wall and is then incident at an angle
approximately perpendicular to the detector.

This technique is generally used for the inspection of insulated pipes, or larger diameter non
insulated pipes.

Insulation
Extended area of
corrosion
Detector
Image of reduced
thickness pipe wall

Figure 6.4 Principle of tangential radiography with the source offset from the pipe
centre-line

18
HOIS(09)RP1 Issue 1
As with the tangential technique on the pipe centre line, with the offset source technique the
through-wall extent of extended areas of either internal or external wall loss can be
measured directly from the CR images, provided appropriate calibration techniques are used
to allow for the enlargement (blow-up) of the CR image see Section 14.3 for further
details.

The tangential (offset) method is often used to measure directly any wall loss identified by
the DWDI method given in Section 6.2.

6.4 Both Tangential and Double wall techniques combined
For relatively thin-walled, small diameter pipes, a single CR image has sufficient dynamic
range and size to show the presence of wall loss by both the tangential and double wall
double image (DWDI) techniques, as illustrated in Figure 6.5.

Detector

Figure 6.5 Radiography combining both the tangential and double wall double
image (DWDI) techniques in a single radiograph

For this combined type of radiography, as shown in Figure 6.5, the tangential method can be
applied to both sides of the pipe on the same CR image, and the region in between will show
loss of wall by increased transmitted radiation intensity, which cause increased CR image
grey levels.

The DWDI technique can be used to identify the area of greatest apparent wall loss, and the
source/detector can then be rotated to find the remaining ligament for this area using the
tangential technique.

19
HOIS(09)RP1 Issue 1
20
7 Radiation Sources
7.1 Type of source
The majority of in-situ on-site radiography is carried out using gamma-ray emitting isotope
sources, although portable, light-weight X-ray sources can also be used in cases where the
need for electrical power and high voltages do not cause significant safety issues.

Iridium 192 is a commonly used isotope source for inspection of medium steel thicknesses.
The gamma-ray spectrum is complex, containing at least 24 spectral lines: those with relative
intensities of at least 30% are at 296, 310, 320 and 470 keV. The half-life is 74 days. Iridium
192 is the most commonly used isotope source for in service inspection in the oil and gas
industry. Source strengths are available up to 200 Curies or more with physical sizes ranging
from about 1x1 mm to about 4x4 mm. A typical source size used for pipe inspection is 2 x
1 mm.

The isotope source Selenium 75 has been developed for industrial radiography more
recently than Ir 192. It has a lower energy gamma-ray spectrum than Ir192 with main peaks
at 137keV and 265keV and a longer half life (120 days). Source strengths are available
between about 2 and 80 Curies with physical sizes ranging from 1x1 mm to 3x3 mm. Due to
the lower gamma-ray energies emitted by Selenium 75 compared with Iridium 192,
Selenium 75 can give radiographs with higher contrast on components with moderate steel
thickness. However, Selenium 75 is less commonly used for tangential radiography, due to
its reduced penetrating power, compared with Iridium 192.

Ytterbium 169 gives several different energies between 63 keV and 307 keV, with a half life
of 32 days. Its mean effective energy is lower than both Selenium 75 and Iridium 192, and is
occasionally used for the in-service inspection of thin-walled components (penetrated steel
thicknesses between 1 and 15 mm, for Test Class A film radiography, see EN 444 and EN
1435).

Cobalt 60 is a high energy source (photon energies of 1.17 and 1.33 MeV), with an extended
half-life of 5.3 years. This source is sometimes used for the in-service inspection of thick
walled components, for which Iridium 192 has insufficient penetration (i.e. steel penetrated
thicknesses greater than c. 85 mm). Because of its greater penetrating power, there are
substantial additional safety requirements for in-service inspection using Cobalt 60,
compared with Iridium 192.

The highest energy sources used successfully for in-service CR are Betatrons, which are
high-energy portable X-ray sources, based on circular beta particle (electron) accelerators.
Models are available with output energies of typically 2, 6 and even 7.5 MeV. As with
Cobalt 60, there are substantial additional safety requirements for in-service inspection using
these sources, but unlike isotope sources, the radiation beam can be switched off by removal
of the supply voltage.

HOIS(09)RP1 Issue 1

7.2 Source selection
7.2.1 Double wall techniques
For the double wall techniques (DWSI & DWDI), the penetrated thickness for pipe inspection
will be equal to twice the pipe wall thickness, on the centre line of the pipe, and will increase
with distance away from the centre line. This increase in penetrated thickness will initially be
gradual, and then more rapid, as the edge of the pipe is approached, as illustrated for DWDI
in Figure 7.1 below.

0
10
20
30
40
50
60
70
-80 -60 -40 -20 0 20 40 60 8
Distance from pipe centre (mm)
P
e
n
e
t
r
a
t
e
d

t
h
i
c
k
n
e
s
s

(
m
m
)
0

Figure 7.1 Penetrated thickness across a typical pipe (OD 150mm, WT 7.1mm),
assuming a very distant source.

Generally, the penetrated thickness at the pipe centre (2WT) can be taken as the basis for
source selection.

Recommended lower and upper limits are given in using Table 7.1 for the different isotope
sources, for the two different wall loss inspection classes (standard and higher quality).
21
HOIS(09)RP1 Issue 1

Table 7.1 Source selection for the double wall inspection techniques (DWDI & DWSI)

Standard quality wall loss inspection
class
Higher quality wall loss inspection class Radiation
Source
Penetrated
thickness, w
(mm)
Pipe WT

(mm)
Penetrated
thickness, w
(mm)
Pipe WT

(mm)
Yb 169 1 s w s 15 0.5 s WT s 7.5 1 s w s 15 0.5 s WT s 7.5
Se 75 5 s w s 55 2.5 s WT s 27 10 s w s 40 5 s WT s 20
Ir 192 7 s w s 85 3.5 s WT s 42 20 s w s 85 10 s WT s 42
Co 60 40 s w s 200 20 s WT s 100 40 s w s 200 20 s WT s 100

For very high penetrated thicknesses, Betatron sources have proved effective with CR
plates, although as with Co 60, there are additional safety requirements for in-service
inspection.

For the higher quality wall loss inspection class, for X-rays up to 500 kV, Figure 1 of EN
14784-2 (identical to Figure 20 of EN1435 :1997) should be used.

For product filled pipes, the additional radiation attenuation caused by the product should be
allowed for in selection of sources. For a fully product filled pipe, the penetrated thickness, w,
in Table 7.1 should be increased by approximately ID/9 for water, as measured by [1]. For
oil, the factor is likely to be larger (estimated as 11 on the basis of relative densities of water
and oil with 0.8 gm/cm
3
) but no measured values are available.

It is also important to note that the presence of product may increase the scattered radiation
levels in the CR images. Additional thickness of detector screens may then be required to
produce acceptable quality images (see Section 11 for further details).

7.2.2 Tangential technique
For a pipe with wall thickness WT and outside diameter OD, the maximum penetrated
thickness, w
max
, through the pipe wall occurs for a line forming a tangent with the inner
diameter. This maximum path is shown in Figure 7.2 and is given by

( ) WT - OD WT 2 w
max
= (7.1)

22
HOIS(09)RP1 Issue 1

Detector
w
max

Source

Figure 7.2 Maximum penetrated thickness, w
max
, for the tangential technique

Note that this applies to any line drawn through the pipe, forming a tangent to the inner
surface of the pipe. Thus w
max
is independent of the source position.

Values for the maximum penetrated thickness, w
max
, through pipes of various diameters and
schedules are given in Figure 7.3 and Table 7.2, for ease of reference. Note that these
penetrated thickness values are generally much larger than twice the wall thickness of the
pipe.

Also shown on Figure 7.3 are the approximate maximum penetrated thicknesses for Se75, Ir
192 and Co 60, recommended for CR inspection using the standard quality tangential
technique [2].

These values are given in Table 7.2, which also gives recommended reduced values for
higher quality tangential inspection (for sizing of pitting flaws, which are more difficult to
detect and size than generalised wall loss).

23
HOIS(09)RP1 Issue 1
Tangential path lengths
0
20
40
60
80
100
120
140
160
180
200
0 2 4 6 8 10
Pipe nominal bore (inch)
T
a
n
g
e
n
t
i
a
l

p
a
t
h

(
m
m
)
12
Schedule 40
Schedule 80
Schedule 160
Limit for Se 75
Limit for Ir 192
Limit for Co 60
Se 75
Ir 192
Co 60

Figure 7.3 Maximum (tangential) path lengths through the walls of pipe of different
diameter. The maximum recommended penetrated thicknesses for
different isotope sources are also shown.

Table 7.2 Maximum tangential paths in steel for different isotope sources

Isotope Source Maximum tangential path
(mm)

Standard quality
(for generalised wall loss)
Higher quality
(for pitting flaws)
Se 75 c. 55 c. 40
Ir 192 c. 85 c. 60
Co 60 c. 140 c. 100

Table 7.3 shows which pipes can be inspected using the standard and higher quality
tangential techniques with the different isotope sources.
24
HOIS(09)RP1 Issue 1
25

Table 7.3 Maximum paths through different schedule pipes of various diameters, together
with applicable isotope sources for tangential CR.

Isotope applicability
2

Nominal
Bore

(inches)
Outside
diameter,
OD

(mm)

Schedule Nominal
wall
thickness,
WT
1

(mm)
Nominal
maximum
Tangential
path
1

(mm)
Se 75 Ir 192 Co 60
1 33.4 40 3.4 20.2
80 4.5 22.8
160 6.4 26.3
XXS 9.1 29.7
1.5 48.3 40 3.7 25.7
80 5.1 29.7
160 7.1 34.2
XXS 10.2 39.4
2 60.3 40 3.9 29.7
80 5.5 34.7
160 8.7 42.5
XXS 11.1 46.7
3 88.9 40 5.5 42.8
80 7.6 49.7
160 11.1 58.8 X
XXS 15.2 66.9 X
4 114.3 40 6.0 51.0
80 8.6 60.3 X
120 11.1 67.7 X
160 13.5 73.8 X
XXS 17.1 81.5 X
5 141.3 40 6.6 59.6 X
80 9.5 70.8 X
160 15.9 89.3 X X
XXS 19.0 96.4 X X
6 168.3 40 7.1 67.7 X
80 11.0 83.2 X
120 14.3 93.9 X X
160 18.3 104.8 X X
XXS 21.9 113.2 X X
8 219.1 40 8.2 83.2 X
80 12.7 102.5 X X
120 18.3 121.2 X X
XXS 22.6 133.3 X X
160 23.0 134.3 X X
10 273.0 40 9.3 99.0 X X
80 15.1 124.8 X X
120 21.4 146.8 X X
XXS 25.4 158.6 X X X
160 28.6 167.2 X X X
12 323.8 40 10.3 113.6 X X
80 17.5 146.4 X X X
120, XXS 25.4 174.1 X X X
160 33.3 196.7 X X X

1
Note that if corrosion/erosion is present, the wall thickness and maximum penetrated
thickness can be substantially reduced, allowing application beyond the limits implied by the
nominal wall thickness shown in this table.

HOIS(09)RP1 Issue 1
26
2
key to isotope applicability:

Both standard and high quality
Only standard quality
Neither

HOIS(09)RP1 Issue 1

8 Source to detector distances (SDD)
This section contains recommended minimum source to detector (SDD) distances for the
different techniques. If two techniques are combined (e.g. DWDI and tangential) then the
larger of the two recommended distances should be used.

The terminology follows that used in EN1435 (see Section 3).

8.1 Double wall single image (DWSI) inspection
In the DWSI technique for in-service inspection, the source is conventionally positioned close
to, but outside one wall of the pipe, and the detector is wrapped around the opposite pipe
wall, as shown in Figure 8.1.

The distances involved for SDD determination for the DWSI technique are shown in Figure
8.1. The object plane is the source side of the pipe wall nearer the detector.

SDD
b
Source,
size d
Detector
f

Figure 8.1 Distances for DWSI

The recommended SDD given in EN1435 for basic class A inspection is

3 / 2
mm
b
5 . 7
d
f
|
.
|
\
|
> (8.1)

This can be re-arranged to give:

3 / 2
mm
b
7.5d b SDD |
.
|
\
|
+ > (8.2)

Formula 8.2 can be used for both the standard and higher qualities of wall loss inspection.

27
HOIS(09)RP1 Issue 1
For larger diameter pipes, formula (8.2) allows the source to be placed adjacent to the pipe
wall opposite the detector (as is conventional practice for DWSI).

This may not be true for pipes with diameters of 4" or less, or those covered by insulation, so
that the detector could not be placed in close contact with the pipe wall. The SDD should
then be calculated according to formula (8.2) above, and the source positioned accordingly.

For pipes with diameters of less than about 3" to 4", the DWDI technique (see below) may be
preferable in some cases as greater axial coverage can be achieved in a single exposure.

8.2 Double wall double image (DWDI) inspection
For the DWDI technique, the SDD is increased compared with DWSI inspection, allowing
inspection of both pipe walls, as illustrated in Figure 8.2, but with a significant increase in
exposure time. The detector is then usually flat, and not wrapped around the pipe wall. In this
case, the relevant object plane is the external surface of the pipe closest to the source.

SDD
b
Source, size
d
Detector
f

Figure 8.2 Distances for DWDI

In general, a large source to detector distance (SDD) will minimise the unsharpness in the
radiograph caused by the size of radiation source (known as geometric unsharpness).
However, large source to detector distances can lead to very long exposure times, and
increased shielding difficulties. Thus trade-offs must be made, whilst ensuring acceptable
image quality.

There is no universally accepted method for the choice of the source to detector distance
which will provide a satisfactory radiographic technique [3] and existing standards from
various countries differ widely.

For wall loss inspection, it is recommended that a geometric unsharpness criterion is used to
provide a basis for setting source to detector distances, following the HOIS Recommended
Practice on radiographic inspection of small bore piping [4]. In [4], a specified unsharpness
value in the plane of the object was used as opposed to the unsharpness in the detector
plane. The unsharpness in the plane of the object was considered to be more physically
meaningful than the unsharpness in the detector plane, as it is the unsharpness in the object
plane that principally affects the minimum discernable detail size in the component under
inspection.

28
HOIS(09)RP1 Issue 1
29
In [4], it was shown that the source to detector distance, SDD, needed to achieve a specified
geometric unsharpness in the plane of the object (Ug') is given by:

SDD = (d . b) / Ug' (8.3)

Where d is the effective source size for geometric unsharpness calculations and b is the
distance between the detector and the source side of the external diameter of the pipe (OD).

For small bore connector inspection, [4] recommended a value of Ug' of 0.3mm, and it is
recommended that this value is used for the higher quality of wall loss inspection as defined
in the present document.

Thus, for higher quality wall loss inspection, the following equation should be used for SDD:

SDD
min
= (d . b)/0.3 (8.4)

For the standard quality wall loss inspection, a higher unsharpness value of 0.6 mm is
acceptable, i.e.

SDD
min
= (d . b)/0.6 (8.5)

Where b is the distance between the source side of the pipe and the detector.

The recommended SDDs derived using equations (8.4) and (8.5) are shown in Figure 8.3,
for an assumed source size of 2.3 mm. Tabulated values are given in Table 8.1.

These distances will be increased for larger source sizes, and reduced for smaller source
sizes, as given by equations (8.4) and (8.5).

HOIS(09)RP1 Issue 1
0
200
400
600
800
1000
1200
1400
1600
1800
2000
0 20 40 60 80 100 120 140 160 180 200 220
Distance from Detector to Source side of pipe OD, b (mm)
S
D
D

(
m
m
)
Standard Quality
Higher Quality
Approximate pipe nominal bore for detector in contact with pipe
1 2 4 3 5 6 8

Figure 8.3 Recommended Source to Detector distances (SDD) for DWDI, for an
assumed effective source size of 2.3mm.

Table 8.1 Recommended source to detector distances (SDD) for DWDI, for an
assumed effective source size of 2.3mm and for the detector in contact with the pipe
wall.

SDD
mm b

mm
Pipe Nominal Bore
(if detector in contact with
pipe wall)
in
Standard
Quality

Higher
Quality

33.4 1 128 256
60.3 2 231 462
88.9 3 341 682
114.3 4 438 876
141.3 5 542 1083
168.3 6 645 1290
219.1 8 840 1680

It should be noted that the SDD values given in equations (8.4) and (8.5) are constant factors
of the pipe diameter, for detectors in contact with the pipe wall. For a 2.3 mm source size, the
SDD for the standard quality is 3.8b, or 3.8 times the pipe diameter, and for the higher quality
the SDD is 7.7b, or 7.7 times the pipe diameter.

30
HOIS(09)RP1 Issue 1
8.3 Tangential Inspection
8.3.1 Source on pipe centre line
For tangential radiography with the source on the pipe centre line, the dimensions and
distances are shown in Figure 8.4.

SDD
b'
Source, size
d
Film/Detector
SPD

Figure 8.4 Dimensions and distances for tangential radiography (source on pipe
centre line)

For tangential radiography, there are two main factors which affect the accuracy of wall
thickness measurements and hence the recommended SDDs.

- Dimensions measured in the CR image are progressively distorted away from the
source axis due to the finite source to detector distance. This has differing effects on
the accuracy of the various methods used for calibration of dimensions in the CR
images.

- Geometric unsharpness.

Consideration of the accuracy of dimensional measurements is given below. It is likely that
this will lead to a greater value for SDD than those based on geometric unsharpness
considerations (see equations (8.7) and (8.8) below. If this is the case, the larger of the two
values for SDD should always be used, where practicable.

Figure 8.4 shows calculations of the errors on the measured wall thickness as a function of
the ratio (source to pipe centre, SPD)/OD, for four different dimensional calibration
techniques, as described further in Section 14. Note that these calculations apply to
conditions in which the pipe may not be in contact with the detector, due for example to the
presence of pipe insulation, as well as cases in which the pipe OD is in contact with the
detector.

The centre line magnification calculations are based on the approximate equation (14.1). The
ball bearing comparator calculations are based on the equations (14.3) and (14.4). The fan-
31
HOIS(09)RP1 Issue 1
beam geometry calculations assume that the radiation beam diverges from a point source in
a thin, two-dimensional fan beam, as shown in Figure 14.3, and that the actual wall thickness
WT is related to the measured value WT by equation (14.2).

In reality, for all usual forms of radiography, the radiation beam diverges as a three-
dimensional cone from the source. However, if the line joining the source to the centre of the
detector is perpendicular to the detector plane, then the fan-beam geometry applies to any
line drawn through the centre point of the CR image.

For the purposes of the calculations shown in Figure 8.5, the 3-D cone beam complication is
ignored, and it is assumed that the fan-beam geometry calculation method gives no errors
in the derived wall thicknesses (provided the distances are measured accurately). For the
other calibration methods, the wall thickness measurement accuracy decreases with
decreasing SDD.

0
1
2
3
4
5
6
7
8
9
10
0 1 2 3 4 5 6 7 8 9
SPD / OD
E
r
r
o
r

i
n

W
T

m
e
a
s
u
r
e
m
e
n
t

(
%
)
10
Centre line Magnification
Pipe Diameter measurement
Ball bearing comparator
Fan-beam geometry calculations

Figure 8.5 Calculations of effects of different dimensional calibration techniques on
accuracy of wall thickness measurements using the tangential
technique.

From Figure 8.5, it can be seen that the accuracy of all the calibration techniques decreases
rapidly for SPD /OD less than about 3.5 (with the exception of the exact fan beam geometry
calculation method).

It is therefore recommended that for tangential radiography with the source on the
pipe centre line, the minimum source to pipe centre distance, SPD should be at least
3.5 times the pipe OD.

Using the distances given in Figure 8.4, this corresponds to:

SDD = 3.5 OD + b (8.6)

Where b is the distance from the pipe centre to the film/detector.

32
HOIS(09)RP1 Issue 1
33
It should also be verified that the SDD is not smaller than the values given below. These are
based on geometric unsharpness considerations similar to those used for DWDI (see Section
8.2), but adapted for the different plane of interest on the pipe (tangent position instead of the
source side of the pipe):

Standard quality:

SDD
min
= (d . b)/0.6 (8.7)

Higher quality:

SDD
min
= (d . b)/0.3 (8.8)

Where b is the distance from the pipe centre to the film/detector.

8.3.2 Combined tangential/DWDI image radiography
For smaller diameter pipes (< 4" - 6" diameter) , the double-wall double image (DWDI)
technique is often combined with the tangential technique with the source on the pipe centre
line.

The SDDs for this combined technique should be those for DWDI, i.e. as given in equations
(8.4) and (8.5), unless the value given by equation (8.6) is larger, in which case that should
be used where practicable.

8.3.3 Source offset from pipe centre line
For tangential radiography with the source offset from the pipe centre line, the effects of the
different dimensional calibration techniques on measured wall thickness accuracy are much
less significant than with the source on the pipe centre line.

It is therefore appropriate to use recommended SDDs based solely on an unsharpness
criterion. In this case, the plane of the object of interest is the pipe tangent position, as
illustrated in Figure 8.6 below.

HOIS(09)RP1 Issue 1

SDD
b
Source, size
d
Detector
SPD

Figure 8.6 Distances for offset source tangential radiography

The recommended SDD is then given by equations (8.7) and (8.8) for the higher and
standard qualities of inspection.

These recommended SDD values for offset tangential radiography are plotted in Figure 8.7.

0
200
400
600
800
1000
1200
0 20 40 60 80 100 120 140
Distance from detector to centre of pipe, b (mm)
S
D
D

(
m
m
)
Standard Quality
Higher Quality
Approximate pipe nominal bore for detector in contact with pipe (in)
1 2 4 3 5 6
10 8

Figure 8.7 Recommended Source to Detector distances (SDD) for offset tangential
radiography, for an assumed effective source size of 2.3mm.

34
HOIS(09)RP1 Issue 1

9 Circumferential and axial coverage and overlap
9.1 Circumferential coverage
When using the DWDI and DWSI techniques, then full circumferential coverage of a pipe is
achieved by taking a number of different exposures around the pipe circumference.

9.1.1 DWSI
For DWSI, the approach used in EN 1435 : 1997, is for the number of circumferential
exposures to be calculated on the basis of a maximum permissible increase in penetrated
thickness due to inclined penetration at the edges of the diagnostic area. This increase in
penetrated thickness is a function of:

- SDD
- Wall thickness, WT
- Pipe OD

Following EN1435 class A for DWSI, with the standard and higher quality classes of CR wall
loss inspection, it is recommended that for wall loss CR, the maximum permissible increase
in penetrated thickness should be 20%. The minimum number of exposures is then given by
Figure A.4 of EN 1435 :1997 (page 17).

For the source positioned outside the pipe, Figure 9.1 shows the number of exposures
needed, as a function of two dimensionless variables WT/OD and OD/SDD.

Minimum number of exposures DWSI
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0.5
0.55
0.6
0.65
0.7
0.75
0.8
0.85
0.9
0.95
1
1.05
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 0.22 0.24 0.26 0.28 0.3
WT/OD
O
D
/
S
D
D
> 3
> 4
> 5
> 6
> 7
> 8
4
5
7
6
8
9
3

Figure 9.1 Minimum number of DWSI exposures circumferentially around a pipe as
a function of the ratios WT/OD and OD/SDD. This graph shows the same
information as Figure A.4 from EN 1435: 1997, for source positions
outside the pipe.
35
HOIS(09)RP1 Issue 1

If the detector is offset from the pipe wall due for example to the presence of insulation, then
Figure 9.1 is not applicable, and the values given in Figure 9.2 should be used instead. In
Figure 9.2 note that the vertical axis is the pipe OD divided by the distance from the source
to the pipe axis (SPD).

To obtain the circumferential angular difference (in degrees) between exposures, the
following formula should be used for DWSI:

Angular difference = 360 / (Number of exposures)

Minimum number of exposures DWSI
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.9
2
2.1
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 0.22 0.24 0.26 0.28 0.3
WT/OD
O
D
/
S
P
D
> 3
> 4
> 5
> 6
> 7
> 8
4
5
7
6
8
9
3

Figure 9.2 Minimum number of DWSI exposures circumferentially around a pipe, as
a function of the ratios WT/OD and OD/SPD, where SPD is the distance
from the source to the pipe axis (centre). This figure should be used
instead of Figure 9.1 if the detector is not in contact with the pipe wall
(e.g. due to the presence of insulation).

For DWDI (see Section 9.1.2), recent trials [5] have shown that the maximum permissible
increase in penetrated thickness can be greater than 20%, without significant loss of
sensitivity for pitting type flaws in pipes.

Thus for DWSI, it may also be possible to reduce the number of exposures below the values
shown in Figure 9.2. However further experimental trials would be needed to investigate this.
For DWSI large variations in source to detector distance occur across typical images, which
result in significant changes in background image brightness. This effect should also be
considered if new criteria are to be established for DWSI circumferential coverage.

9.1.2 DWDI
For DWDI, the detectability of flaws at the tangential position represents the most demanding
position for circumferential coverage. Recent trial results [5] showed that the detectability of
36
HOIS(09)RP1 Issue 1
37
pitting flaws at or very close to the tangent position was generally reduced compared with
that obtained when the flaws were on the pipe centre line, especially with pipes having larger
maximum penetrated thickness values, w
max
. However rotation of the source/detector so that
the flaws were 15-20 from the tangent position then gave detectability similar to that
achieved with the flaw on the pipe centre line.

In general, at least two exposures are recommended for DWDI, separated in circumferential
angle by at least 30-40 (for best results the angle would be 90).

However, for those pipes having maximum tangential paths, w
max
, less than about 50% of the
maximum recommended values for the standard tangential image quality (see Table 7.2),
adequate sensitivity may be achieved using a single exposure.

9.1.3 Penetrated Thickness measurement
The number of circumferential exposures derived from Figures 9.1 9.3 can be used for
general DWDI and DWSI techniques.

However, if quantitative analysis of the image grey level information is being used to estimate
penetrated thickness (Section 12), it is important to ensure that the feature being measured
on the CR image is as close as possible to the centre of CR image. This is likely to require
additional circumferential exposures to align the flaw of interest more closely with the source
axis.

9.2 Axial coverage
The maximum axial coverage for a single CR image can be determined in a similar manner
to that for DWSI circumferential coverage (see Section 9.1.1). For axial distances, the
geometry is then simpler, as illustrated in Figure 9.4. For a 20% increase in penetrated
thickness at the edge of the area to be inspected, then it can be shown the total extent, L
d
, of
this area on the detector is

L
d
= 1.32 SDD

The corresponding axial coverage on the source side of the pipe, L
p
, is then:

L
p
= 1.32 f

where f is the distance between the source and the source side of the pipe. Lp should be
used when determining the interval along the pipe of exposures.

Note that for DWDI, typical CR image plates are likely to be significantly less than L
p
, so this
is unlikely to be a constraint found in practice, and the axial interval between exposures will
be determined by the dimension of the CR image plate.

However, for DWSI, especially on smaller diameter pipes, the value for L
p
needs to be
calculated and used when determining the axial interval between exposures.

As for circumferential coverage, if quantitative analysis of the image grey level information is
being used to measure penetrated thickness (Section 13), it is important to ensure that the
feature being measured on the CR image is as close as possible to the centre of CR image.

HOIS(09)RP1 Issue 1

SDD
Source, size
d
Detector
f
L
d

L
p

Figure 9.4 Axial cross section for DWDI inspection (also applicable to DWSI),
showing the maximum permissible axial length, L
d
, of the evaluated area
for a single source position, on the detector and along the pipe, L
p
.

9.3 Overlap of images
The separate CR images shall overlap sufficiently to ensure that no portion of the component
remains un-examined. Unless otherwise specified, the minimum overlap shall be 25 mm
axially either side of the diagnostic area, measured on the source side.

38
HOIS(09)RP1 Issue 1
39

10 CR image quality indicators
10.1 Background
Experience has shown that some form of objective check on image quality is important for in-
service computed radiography. Operators should be aware that very short exposure times
with computed radiography can give CR images which may appear to the eye, subjectively at
least, to have adequate quality, when displayed on a computer monitor. However,
quantitative analysis can show that such images may have low signal to noise ratios, and
hence poor sensitivity for wall loss detection by image grey level variations.

It is strongly recommended that some form of image quality indicator, or objective
check on image quality is used on all CR exposures, as described below.

10.2 Image Quality Measures for Double Wall Techniques (DWDI &
DWSI)
10.2.1 Target Grey level range
For the standard quality class of wall loss inspection CR, a minimum check on CR image
quality is to ensure that a specified (target) grey level range is achieved in the area of interest
of the CR image. The grey level range achieved for a particular radiation exposure also
depends on the IP type and the scanner gain (or sensitivity) used, so it is also necessary to
specify the IP type and all user selectable scanner parameters when setting a target grey
level range (pixel size, gain/sensitivity etc).

In setting a target grey level range for a particular CR scanner, IP type and scanner gain,
measurements of image normalised signal to noise ratio (see Section 10.2.2 below) should
first be made for a range of exposures. The results should be plotted as a function of grey
level, as illustrated in Figure 10.1. This allows the grey level corresponding to a normalised
signal to noise ratio of 50 to be determined, for a specified scanner/IP combination and the
specified scanner user settings including gain/sensitivity and pixel size.

In the example shown in Figure 10.1, for low gain, SNR_N > 50 can be achieved for grey
levels >3700, whereas for medium gain grey levels must be >24000. These target grey level
ranges are specific to the scanner, the IP and all the scanner settings and the CR analysis
software. If changes are made to any of these variables, a repeat calibration is required to
establish the target grey level for which SNR_N = 50.

HOIS(09)RP1 Issue 1
0
10
20
30
40
50
60
70
80
0 5000 10000 15000 20000 25000 30000 35000 40000 45000 50000
Grey Level
S
N
R
_
N
CR50P
Low gain
CR50P
Medium gain

Figure 10.1 Measurements of normalised signal to noise ratio as a function of local
image grey level, for two different gain/sensitivity settings with Ir 192
and a specific CR scanner/IP and pixel size.

For the higher quality class of wall loss inspection CR, it is required that one of the more
rigorous measures of image quality described below is also applied to ensure the required
image quality has been achieved.

10.2.2 Signal to Noise Ratio (SNR)
For the standard and higher quality classes, it is recommended that CR image quality is
assessed by measuring the normalised signal to noise ratio (SNR_N) in the area of interest,
using appropriate software if available.

Note that it is important when making SNR measurements for the images to have grey levels
directly proportional to radiation intensity (i.e. linear response). For non-linear response CR
scanners it is necessary to select an appropriate look-up table (LUT), to achieve a linear
relation between radiation intensity and CR image grey level. Different CR scanners can
have different characteristics (e.g. logarithmic or square-root, as well as linear), and the
correct LUT must be selected for the scanner used. If the CR images are not linearised
correctly, the values measured for SNR_N can be misleadingly high. For typical CR images,
SNR_N values will typically be in the range ~ 50 200. Values in excess of this may be due
to use of an incorrect LUT, and should be checked carefully.

When making this measurement, it is important to ensure the analysis area does not include
any significant variations in grey level due to changes in penetrated thickness. Thus the SNR
area should not include, for example, component edges, areas of wall loss, or any other
image areas where there are significant changes in penetrated thickness.

The size of the SNR area should be 20 pixels horizontally x at least 55 pixels vertically to
provide a dataset of at least 1100 values, as required by EN 14784-1.

40
HOIS(09)RP1 Issue 1
41
In all cases, SNR_N measurements should be made at several locations (minimum of four)
within the area of interest, and a mean value derived.

Note that for some components it may not be possible to find a homogeneous area of the
image suitable for SNR_N measurements. An example would be a pipe containing
generalised irregular corrosion over the whole area of the image. In such cases, the target
grey level method described in the Section above should be used instead.

The SNR measured on a CR image needs to be normalised using a factor which depends on
the basic spatial resolution (BSR) of the CR system (see EN 14784-1, equations 2 and 3).

The formula to be applied to calculate the normalised signal to noise ratio is:

SNR_N = SNR
meas
(88.6/BSR) (10.1)

Where

SNR
norm
is the normalised signal to noise ratio
SNR
meas
is the signal to noise ratio measured on the CR image
BSR is the basic spatial resolution of the CR system in microns (depends on the
scanner pixel size and the model of CR plate)
88.6 is the length of the side of a square having the same area as a circle with
diameter 100 microns (see EN14784-1, p11 for explanation)

For double wall techniques (DWSI and DWDI), and the standard quality class, the
normalised signal-to-noise ratio (SNR_N) as calculated from equation (10.1) in the pipe
centre should be at least 50.

For double wall techniques (DWSI and DWDI), and the higher quality class, the
normalised signal-to-noise ratio (SNR_N) as calculated from equation (10.1) in the pipe
centre should be at least 80.

10.2.3 BSR Measurement
The basic spatial resolution (BSR) can be measured according to EN 14784-1, which
involves use of orthogonal duplex wire IQIs. An improved BSR measurement technique for
isotope sources, to minimise the effects of noise, is described in [5]. This method involves
the analysis of the responses obtained with a Duplex wire IQI, as shown in Figure 10.2.

HOIS(09)RP1 Issue 1

Figure 10.2 CR image of the Duplex wire IQI, rotated so that the wires are as
accurately vertical as possible.

The modulation of the different wire pairs can then be plotted as a function of wire diameter.
The variation of modulation with wire diameter is linear on a log log plot, which allows
ready derivation of the wire diameter for which the modulation is 0.2 (or 20%), as illustrated
in Figure 10.3.

Derivation of BSR based on measured wire pair modulations
0.1
1
0.1 1
Duplex wire diameter (mm)
M
o
d
u
l
a
t
i
o
n
BSR

Figure 10.3 Measured Duplex wire pair modulation plotted as function of the wire
diameter. The BSR is derived from the fitted line for a modulation of 0.2.
42
HOIS(09)RP1 Issue 1
43

The advantages of this more quantitative method for BSR determination include improved
repeatability, better noise tolerance, less subjectivity, and determination of values to a
precision of better than the spacing between available wire pair diameters.

In some cases, the measured BSR values vary with direction in the CR image. In this case, it
is important to ensure consistency between the directions of the SNR_N and BSR
measurements. For the default SNR measurement direction of horizontal within a CR image
(along the fast scan direction), the wires themselves within the duplex wire IQI should be
aligned orthogonal to this direction (i.e. vertical in the displayed image). This is achieved
when the long axis of the duplex wire IQI assembly is horizontal in the image, i.e. aligned
with the SNR measurement direction.

Measured values for BSR for some current CR systems are given in Table 10.1, taken from
[5]. Note that the values depend on the radiation source as well as the imaging plate, pixel
size and scanner model.

If a value for the BSR for the screen/scanner combination in use is not available, a
conservative estimate of 200 microns can be used, provided the screen resolution is known
to be similar to those screens given in Table 10.1, and the scanner pixel size does not
exceed 100 microns.

HOIS(09)RP1 Issue 1
44

Table 10.1 Basic spatial resolution (BSR) of some CR scanners and imaging plates,
measured using Duplex wire IQI

CR System

Basic spatial resolution,
BSR*
(microns)

Scanner IP Pixel
size
(m)

Radiation
source
According
to
EN14784-1
More
quantitative
method
[see ref 5]
HD CR 35
NDT
White, pre April
2008
100 Iridium 192 160
HD CR 35
NDT
White, pre April
2008
100 Selenium 75 160
HD CR 35
NDT
White, post
April 2008
100 Iridium 192 145 163
HD CR 35
NDT
White, post
April 2008
50 Iridium 192 115 133
HD CR 35
NDT
White, post
April 2008
100 Selenium 192 145
HD CR 35
NDT
White, post
April 2008
50 Selenium 192 100
HD CR 35
NDT
Blue 50 Iridium 192 100
HD CR 43
NDT
White, post
April 2008
100 Iridium 192 145
HD CR 43
NDT
Blue 50 Iridium 192 90
Fuji system
DynamiX
ST-VI 100 Iridium 192 160 176
Fuji system
DynamiX
ST-VI 100 Selenium 75 160
GE CR50P IPC2 100 Iridium 192 200 236
GE CR50P IPS 50 Iridium 192 130
GE CR50XP IPC2 100 Iridium 192 200
GE CR50XP IPS 50 Iridium 192 145 165
GE CR50XP IPS 50 Selenium 75 130
GE CR100 IPC2 100 Iridium 192 200 230
GE CR100 IPC2 100 Selenium 75 200
GE CR100 IPS 100 Iridium 192 160
GE CRXTower IPC2 100 Iridium 192 160
GE CRXTower IPC2 100 Selenium 75 160
GE CRXTower IPS 50 Iridium 192 130
GE CRxFlex IPC2 100 Iridium 192 160 222
GE CRxFlex IPS 100 Iridium 192 145
GE CRxFlex IPS 50 Iridium 192 100 100

* The effective spatial resolution or unsharpness of the CR plate/scanner is
approximately twice the BSR [see equation (4) of EN14784-1].
HOIS(09)RP1 Issue 1
10.2.4 Wire IQIs
Use of physical IQIs (wire or step/hole) is recommended for CR inspection for wall loss to
provide a quantifiable measure of image quality. In general, the minimum IQI wire diameter
visible on a CR image will be limited by the total unsharpness of the image, which is made up
from contributions from the screen unsharpness and geometric unsharpness. The CR image
signal to noise ratio and radiographic contrast will also affect the minimum visible IQI wire
diameter.

For DWDI, the IQIs should be placed on the source side of the pipe, where possible. If the
pipe is insulated, the IQI will need to be placed on the detector side. For DWSI, the IQIs need
to be placed on the detector side (between pipe wall and detector).

For both DWDI and DWSI, the IQIs should be positioned close to the centre of the resulting
CR image. If the IQIs are close to the edges of the images, a smaller number of wires may
be detected than for centrally placed IQIs.

For DWDI inspection using Ir 192, source-side measured and target values for IQI wire
numbers (as defined in EN 462-1) are shown in Figure 10.4. These measurements cover
pipes with wall thicknesses in the range c. 3 mm to 22 mm. Separate values are shown for
the standard and higher image quality classes (SNR_N 50 and 80 on the pipe centre line,
respectively). For product filled pipes, the total equivalent steel penetrated thickness should
be calculated using:

w
t
= w
s
+ ID/ f (10.2)

where w
s
is the steel penetrated thickness, ID is the pipe internal diameter and f is a factor
representing the lower attenuation of the product compared with steel. Estimates for f are ~9
for water and ~11 for oil (no measured value is available for oil).

DWDI - Ir 192
5
6
7
8
9
10
11
12
13
14
15
0 10 20 30 40 50 6
Equivalent steel total penetrated thickness (mm)
I
Q
I

v
a
l
u
e

W
0
DWDI Bergen Oct 08 IPC2/CR100
DWDI IPC2/CR50P NDT Services
DWDI Bergen Feb 08 IPC2/CRX Tower
DWDI Bergen May 09 ST6/Fuji
DWDI Water filled Bergen Feb 08 IPC2/CRX Tower
DWDI Water filled Bergen Oct 08 IPC2/CR100
DWDI Water Filled Bergen May 09 ST6/Fuji & HD CR 35/White
DWDI Ir 192 Target values
DWDI HIGHER Bergen Feb 08 IPS/CRX Tower
DWDI HIGH Bergen Oct 08 IPS/CR50XP
DWDI HIGHER Bergen May 09 HD CR 35 White IP
DWDI Water filled HIGHER Bergen Feb 08 IPS/CRX Tower
DWDI Water filled HIGH Bergen Oct 08 IPS/CR50XP
DWDI Ir 192 High Target values
EN1435 Class A (DWDI)

Figure 10.4 Measurements of smallest IQI wires visible on DWDI Ir192 CR images, as
a function of equivalent steel total penetrated thickness (on pipe centre
line). The values from EN1435 for film radiography weld inspection are
shown for comparison purposes. The IQI wires were on the source side
of the pipe.

45
HOIS(09)RP1 Issue 1
Corresponding measured IQI values and target values for DWSI inspection using Iridium 192
are given in Figure 10.5.

DWSI - Ir 192
5
6
7
8
9
10
11
12
13
14
15
16
0 10 20 30 40 50 6
I
Q
I

v
a
l
u
e

W
0
DWSI Bergen Oct 08 IPC2/CR100
DWSI IPC2/CR50P NDT Services
DWSI Bergen Feb 08 IPC2/CRX Tower
DWSI Bergen May 09 ST6/Fuji
DWSI Water filled Bergen Feb 08 IPC2/CRX Tower
DWSI Water Filled Bergen May 09 HD CR 35/White
DWSI Ir 192 Target values
DWSI Water filled Bergen Oct 08 IPC2/CR100
DWSI HIGHER Bergen Feb 08 IPS/CRX Tower
DWSI HIGHER Bergen Oct 08 IPS/CR50XP
DWSI HIGHER Bergen May 09 HD CR 35/White IP
DWSI Water filled HIGHER Bergen Feb 08 IPS/CRX Tower
DWSI Water filled HIGHER Bergen Oct 08 IPS/CR50XP
DWSI HIGHER Water Filled Bergen May 09 HD CR 35/White IP
DWSI Ir 192 High Target values
EN1435 Class A (DWSI)

Figure 10.5 Measurements of smallest IQI wires visible on DWSI Ir192 CR images, as
line). The values from EN1435 for film radiography weld inspection are
shown for comparison purposes.

With Selenium 75, increased radiographic contrast is generally obtainable compared with
Iridium 192, within the range of applicability of the source. This results in higher IQI values
with Selenium 75 for images having the same SNR_N values. The available measurements
for Selenium 75 DWDI and DWSI inspections are given in Figures 10.6 and 10.7
respectively.
DWDI : Se 75
6
7
8
9
10
11
12
13
14
15
0 5 10 15 20 25 30 35 40
I
Q
I

v
a
l
u
e

W
DWDI SELENIUM Bergen Feb 08 IPC2/CRX Tower
DWDI Bergen Oct 08 IPC2/CR100
DWDI Bergen May 09 HD CR 35/White
DWDI Water Filled Bergen Oct 08 IPC2/CR100
DWDI Water Filled Bergen May 09 HD CR 35/White
DWDI Ir 192 Target values
DWDI HIGHER Bergen Oct 08 IPS/CR50XP
DWDI HIGHER Bergen May 09 HD CR 35/White
DWDI Ir 192 High Target values
EN1435 Class A (DWDI)

Figure 10.6 Measurements of smallest IQI wires visible on DWDI Se 75 CR images,
as a function of equivalent steel total penetrated thickness (on pipe
centre line). The IQI wires were on the source side of the pipe.

46
HOIS(09)RP1 Issue 1
DWSI : Se 75
6
7
8
9
10
11
12
13
14
15
0 10 20 30 40 50 60
I
Q
I

v
a
l
u
e

W
DWSI SELENIUM Bergen Feb 08 IPC2/CRX Tower
DWSI Bergen Oct 08 IPC2/CR100
DWSI Water Filled Bergen Oct 08 IPC2/CR100
DWSI Bergen May 09 HD CR 35/White
DWSI Water Filled Bergen May 09 HD CR 35/White
DWSI Se 75 Target values
DWSI SELENIUM HIGHER Bergen Feb 08 IPS/CRX Tower
DWSI HIGHER Bergen Oct 08 IPS/CR50XP
DWSI HIGHER Bergen May 09 HD CR 35/White IP
DWSI Higher Water Filled Bergen Oct 08 IPC2/CR100
DWSI HIGHER Water Filled Bergen May 09 HD CR 35/White IP
DWSI Se75 High Target values
EN1435 Class A (DWSI)

Figure 10.7 Measurements of smallest IQI wires visible on DWSI Se 75 CR images, as
line).

The tables below summarise the target IQI values derived from the experimental CR images
collected during the HOIS trials, as given in Figures 10.4 to 10.7 above. Note that the
thickness t is the equivalent steel total penetrated thickness on the pipe centre line, w
t
, as
given by equation 10.2.

DWDI Iridium 192 (source side IQIs)

Standard Image quality Higher Image quality
Thickness
range (mm)
IQI value, W Thickness
range (mm)
IQI value, W
5 s w
t
< 15 9 5 s w
t
< 8 12
15 s w
t
< 25 8 8 s w
t
< 12 11
25 s w
t
< 40 7 12 s w
t
< 15 10
40 s w
t
< 60 6 15 s w
t
< 20 9

DWSI Iridium 192 (detector side IQIs)

Thickness
range (mm)
range (mm)
IQI value, W
5 s w
t
< 10 11 15 s w
t
< 20 10
10 s w
t
< 15 10 20 s w
t
< 35 9
15 s w
t
< 25 9 35 s w
t
< 60 8
25 s w
t
< 30 8
30 s w
t
< 60 7

47
HOIS(09)RP1 Issue 1
48

DWDI Selenium 75 (source side IQIs)

Thickness
range (mm)
range (mm)
IQI value, W
5 s w
t
< 12 11 5 s w
t
< 10 12
12 s w
t
< 18 10 10 s w
t
< 20 11
18 s w
t
< 25 9
25 s w
t
< 35 8

DWSI Selenium 75 (detector side IQIs)

Thickness
range (mm)
range (mm)
IQI value, W
5 s w
t
< 15 11 5 s w
t
< 25 11
15 s w
t
< 25 10 25 s w
t
< 40 10
25 s w
t
< 30 9 40 s w
t
< 45 9
60 s w
t
< 60 8

10.2.5 Image quality for tangential techniques
For tangential radiography, conventional wire or step/hole IQIs are not directly applicable,
because they cannot be positioned near to the tangential pipe position, and the rapid
changes in penetrated thickness in this part of a radiographic image makes it impossible to
assess IQI visibilities in any meaningful way.

However, very noisy CR images will give lower wall thickness measurement accuracies than
less noisy CR images. In addition, the unsharpness of the CR image will influence the WT
measurement accuracy.

Hence, some form of quality control for tangential CR radiography is considered necessary,
as follows.

- Ensure that the SDDs recommended in Section 8.3 are used.

- The exposure time should be adjusted so that the un-impeded radiation beam outside
the pipe wall does not exceed 80-90% of the CR imaging systems saturation value
(see Sections 12 and 14.4 for further information on exposure times and burn-off
effects).

- The resulting exposure time depends on the gain setting of the CR system. It is
important that a sufficiently low system gain is used to allow the SNR_N values given
below to be achieved.

- If the available software provides this function, the normalised signal to noise ratio
(SNR_N) should be measured with appropriate image analysis software. The average
SNR_N values obtained in the free beam outside the pipe should be at least 70 for
the standard quality and 110 for the higher quality classes of wall loss inspection.
HOIS(09)RP1 Issue 1
49

- If the pipe centre line is available for measurement of SNR_N, then the DWDI SNR_N
values of 50 and 80 for standard and higher qualities respectively can be used as an
alternative to the above free beam SNR_N values.

Note that in all cases when measuring SNR_N, it is important that the image is in a form
having the image grey levels directly proportional to radiation intensity, otherwise the values
can be misleadingly high.

HOIS(09)RP1 Issue 1
50

11 Screens/filters
The filmless plates used in CR are more sensitive than conventional film to scattered
radiation, which generally has a lower average photon energy than that of the primary
radiation beam. It is therefore important to reduce scattered radiation by means of lead
screens.

The recommended values given in EN 14784-2 are reproduced for convenience in
Table 11.1. These screen thickness values are generally greater than those recommended
for film radiography and are thicker than standard.

Thinner screens with standard thickness can be used provided the specified image quality
values are achieved (see Section 10).

Table 11.1 Recommended metal screens for computed radiography taken from
Table 2 of EN 14784-2. Thinner (standard) screens can be used provided the specified
image quality values are achieved (see Section 10).

Type and minimum thickness in mm of metal
screens
Radiation Source Penetrated WT, w
mm
Front Back
w < 5 Pb 0.1 Pb 0.1 Yb 169, Tm 170
w > 5 Pb 0.1 Pb 0.1
w < 50 Pb 0.3 Pb 0.3 Ir 192, Se 75
w > 50 Pb 0.4 Pb 0.4
w < 100 Fe 0.5/Pb 1.5 Fe 0.5/Pb 1.0 Co 60
w > 100 Fe 0.5/Pb 2.0 Fe 0.5/Pb 1.0

For the corresponding values for X-ray sources, see EN 14784-2, Table 2.

For tangential radiography, the values given in Table 11.1 should be regarded as minimum
values, and thicker screens should be used, where possible, to reduce burn-off effects at the
outer edges of the pipes, and lower the overall image contrast.

For pipes containing product, increased levels of scatter may be obtained, depending on the
pipe wall thickness, and the penetrated thickness of product. Increased levels of scatter are
likely to be associated with greater penetrated thicknesses of product, and reduced pipe wall
thicknesses.

Front and back screens with increased thickness (e.g. up to 1.5 mm thick front screens) may
then be beneficial in some cases. Note however that use of thicker lead screens will then
require increased exposure times to achieve acceptable SNR_N values, as defined in
Section 10.

For high strength sources, significantly thicker back screens can be needed to compensate
for increased scatter from the plant environment in the vicinity of the detector.

HOIS(09)RP1 Issue 1
12 Exposure time
CR plates have a wide tolerance or latitude to variations in exposure times. However,
experience has shown that using very short exposure times will result in poor image quality,
and hence poor sensitivity to wall loss with the DWDI and DWSI techniques. Correspondingly
lower wall thickness measurement accuracies can be anticipated with the tangential method
for short exposure time images.

To avoid these issues, the exposures times for computed radiography should be sufficient to
give CR images with the required quality, as described in Section 10.

In all cases, for isotope sources, it is recommended that the time taken to move the source
out of the container into the exposure position and back again should be less than 10% of
the exposure time, to avoid effects connected with movement unsharpness.

12.1 DWSI and DWDI
To calculate exposure times, the following approximate formula may be used:

( )
S
SDD
2
tot 0
4 -
w
w exp E 2.4x10 E + = (12.1)

Where

E
w
is the exposure time in sec for penetrated thickness w.
E
0
is the exposure needed to achieve the required SNR_N value for zero penetrated
thickness in units of Ci.min @ 500mm.
S is the source strength in Ci.
SDD is the source to detector distance in mm.
is the measured effective material attenuation coefficient (0.04 /mm for Ir 192 and
0.08 /mm for Se 75).
w
tot
is the total steel equivalent penetrated thickness, including any product in the pipe.

Note that if the exposure, E
0
, is expressed in units of gBq. sec @ 1000mm, then the constant
in equation (12.1) is 2.7 x 10
-8
.

For pipes fully filled by product, the steel penetrated thickness (2 WT) on the pipe centre line
should be increased by approximately ID/9 for water and c. ID/11 for oil (density of 0.8
gm/cm
3
) to give the total steel equivalent penetrated thickness, w
tot
.

The exposure times required to obtained the recommended SNR_N values on the pipe
centre line depend on the imaging plate and scanner.

For the standard image quality class (SNR_N 50 on the pipe centre line), exposures, E
0
in
the range ~3 - 15 Ci.Min @ 500mm have been measured in practical trials [5], depending on
the CR scanner and imaging plate being used.

For the higher image quality class (SNR_N 80 on the pipe centre line), exposures, E
0
in the
range ~10 - 30 Ci.Min @ 500mm have been measured in practical trials [5], depending on
the CR scanner and imaging plate being used.

51
HOIS(09)RP1 Issue 1
52
Significantly longer exposures may be needed if high-resolution imaging plates intended for
weld inspection are used for in-service examination (E
0
~ 30 Ci.Min @ 500mm for standard
image quality and E
0
~ 70 Ci.Min @ 500mm for higher image quality).

12.2 Tangential
For tangential radiography, the times given above for DWDI can generally be used as a
guide.

It is also important to ensure that the scanner gain/sensitivity setting is adjusted so that the
unimpeded radiation beam outside the pipe wall is not saturated, which will cause burn-off
and errors in wall thickness measurement.
HOIS(09)RP1 Issue 1
53

13 Penetrated Thickness measurements
13.1 Introduction
For the DWDI and DWSI techniques, computer analysis of the CR image grey levels can, in
certain conditions, be used to estimate wall thickness changes. It is important to note
however that this a relatively new method which has not yet been fully validated.

This method should therefore be used with caution, and validated in advance for the
component under inspection using test components with closely similar wall thickness and
pipe OD. These test components should contain areas of wall loss of known through-wall
extent, to allow validation of the penetrated thickness analysis method, described below.

It should be emphasised that this is a relative method for measurement of wall loss, and
unlike the tangential method (Section 14), the penetrated thickness method does not provide
direct measurements of remaining wall thickness.

13.2 Principle of method
The software used for this purpose generally assumes that the detected radiation intensity is
related to penetrated thickness by:

I(w) = I
0
exp(- w) (13.1)

Where

I(w) is the intensity for penetrated thickness w
I
0
is the unimpeded beam intensity
is the effective linear attenuation coefficient of the material

13.3 Effects of scattered radiation
In practice, it is important to appreciate that a number of effects, in particular the presence of
scattered radiation, complicates the analysis. Thus equation 13.1 is often only approximate,
and generally only valid for relatively small percentage changes in wall thickness.

The presence of significant levels of scatter on CR images also makes nominal attenuation
values unreliable. For example, a theoretical value (with no scatter) for the linear attenuation
coefficient, , for Ir 192 and steel is 0.084 /mm. However, measured values can vary widely
between at least 0.02 /mm to 0.05 /mm, depending on scatter levels.

13.4 Calibration using step wedge
For this method, it is important to calibrate using a step wedge placed on the component
under inspection, as close to the area of interest as possible. For the DWDI method, the step
wedge can be positioned on the source side of the pipe. For DWSI, the step wedge needs to
be positioned between the pipe wall and detector, which leads to certain practical issues (see
Section 13.6).

HOIS(09)RP1 Issue 1
54
For both DWSI and DWDI methods, the penetrated thickness increases away from the pipe
centre-line, so it is important that the step wedge is positioned so that the penetrated
thickness for the step wedge is similar to that for the wall loss being measured.

The effective attenuation coefficient should be measured from at least the first and second
steps on the step wedge. If the attenuation values derived from the different steps are in
good mutual agreement, an average value can be used, and this provides some confidence
in the accuracy of the subsequent measurements of wall loss. However, if the attenuation
coefficients derived from the different steps are not in good mutual agreement, the value
derived from the first step (smallest increase in penetrated thickness) should be used when
estimating wall loss. Such differences can also indicate the presence of significant levels of
scatter on the image, which lead to a loss of accuracy in the wall loss measurements,
especially for larger percentage wall losses.

13.5 Key Points
The key points for this technique are:

- CR image grey levels must be linearised using the correct look-up table for the CR
scanner, if a non linear amplifier is built into the CR scanner.
- A step wedge must be used to measure the effective attenuation coefficient, and
located such that the local penetrated thickness is close to that for the wall loss being
measured.
- The reference and measurement areas should be as large as possible given the size
of features in the image to measure. (Small areas are more affected by noise).
- The reference and measurement areas should be as close together as possible,
given the limitations of the image content.

With this technique, initial results suggest that accurate (1mm) results can only obtained for
relatively small wall losses of c. 30 50%.

If the actual total penetrated thickness is not known accurately, differences from an assumed
value can be used to measure small percentage wall losses with reasonable accuracy.

Use of substantially thicker lead screens may preferentially absorb the lower energy
scattered radiation and hence increase accuracy in some cases.

13.6 Limitations
If this method is used to estimate wall loss, it is important to appreciate its limitations and to
note that it is a relatively new method which has not been fully validated.

Practical issues can include:

1. For the DWDI method, if wall loss occurs on both pipe walls (those near to and far from
the source) the corresponding radiation intensity changes can be superimposed on the
CR image, leading to an overestimation of the wall loss, if it is assumed to be present in a
single pipe wall only.

2. For DWSI it can be difficult to place a step wedge between the pipe wall and detector,
without distorting the detector and increasing the object to detector distance significantly
(which can lead to a need to increase the SDD see Section 8.1).

HOIS(09)RP1 Issue 1
55
3. For DWSI, and some DWDI images, the CR images can contain significant variations in
background grey level, due the rapidly changing SDD and penetrated pipe wall thickness
values across the image. These background variations introduce uncertainties into the
measurement of an appropriate reference grey level for the area of wall loss, especially if
a single measurement area on only one side of the indication is used.

In the presence of significant variations in background grey level, analysis methods
based on extrapolation of grey-scale profiles extracted from the CR image are
recommended to estimate the background image grey level underneath the area of wall
loss.

HOIS(09)RP1 Issue 1

14 Tangential radiography
This Section contains important recommendations regarding tangential radiography.

14.1 Recommended SDD
The recommended source to detector distances (SDD) for tangential radiography have
already been covered in Section 8.3.

14.2 Source location relative to pipe centre line
For small diameter insulated or non insulated pipes, the radiation source for tangential
radiography is generally positioned on the pipe central axis, as illustrated in Figure 14.1. This
allows both walls, on either side of the pipe, to be inspected on a single CR image.

CR plate

Figure 14.1 Tangential radiography with the radiation source on the pipe centre line.

For larger diameter pipes (both insulated and non insulated), the radiation source should be
offset from the pipe centre line, to be instead in-line with the tangential position on the pipe,
as shown in Figure 14.2.

In all cases, it is important to check that the tangential path for the pipe is within the source
penetration limits given in Table 6.2.

56
HOIS(09)RP1 Issue 1
CR plate

Figure 14.2 Tangential radiography with the radiation source offset from the pipe
centre line by half the pipe diameter, so it is in-line with the tangential
position on the pipe.

There are no specific recommendations concerning the maximum pipe diameter which
should be inspected using the radiation source on the pipe axis. This will in practice depend
on the size of the CR plate available, as well as the method used for dimensional calibration
(see Section 14.3 below).

14.3 Dimensional calibration
For tangential radiography, when making dimensional on-screen measurements of wall
thickness, it is important to calibrate the distances involved in the radiography, to allow for
the image enlargement or blow-up. The geometric magnification effect for tangential
radiography is shown in Figure 14.3.

57
HOIS(09)RP1 Issue 1
Detector
PDD
r
WT
u
SPD
x
SDD

Figure 14.3 Geometric magnification for tangential radiography showing the
measured wall thickness WT.

Two methods can be used for dimensional calibration, to derive the actual wall thickness WT
from the measured value WT as follows.

14.3.1 Measurement of distances
This method involves direct physical measurement of the key distances involved in the
radiography.

Note that reliable and accurate physical measurements of distances may be difficult to
achieve in a plant environment, and the measurements may not be available at the time of
the analysis. If this is considered to be the case, the alternative comparator method (Section
14.3.2) is recommended.

For calibration by the distances method, two of the following distances need to be measured
accurately (to within a few percent):

1. Source to detector distance SDD
2. Distance from source to pipe centre line, SPD.
3. Distance from detector to pipe centre line, PDD.

In addition, the following distance must be recorded:

4. Lateral offset (if any) of source from pipe centre line, x.

This method also assumes that the CR system software already contains correct information
on the scanner resolution (pixel size) used when acquiring the CR image.

58
HOIS(09)RP1 Issue 1
For offset tangential radiography (with x ~ r) , the true wall thickness WT at the tangential
pipe position as shown in Figure 14.2 can be calculated from the projected (measured) wall
thickness WT on the CR plate using the approximate equation (14.1):

' WT ) 1 ( WT =
SDD
PDD
(14.1)

For tangential radiography on the pipe centre line (x =0), equation (14.1) is an approximation
which becomes less accurate as the angle, u shown in Figure 14.3 increases. The exact
calculation then requires application of the more involved equation (14.2) [6].

|
|
.
|
\
|
+
|
|
.
|
\
|
=
SDD
WT'
r
r
1
SDD
WT'
r
r
r WT
2 2
2 2
SPD
SPD
SPD
(14.2)

Where r is half the pipe outside diameter (=OD/2).

Provided all the distances involved can be measured sufficiently accurately, and that the CR
image pixel size is known reliably, then software which applies equation (14.2) to derive the
actual wall thickness WT from the measured value WT is capable of high wall thickness
measurement accuracy, even for values of SDD/OD < 3.

Note however, that equation (14.2) is only valid for fan-beam geometries, and will be only
approximate for three-dimensional cone-beam geometries, i.e. for sections through the
pipe, either above or below the source position in the CR image, for which the angle | shown
in Figure 14.4 is non zero (see also Section 8.3.1).

59
HOIS(09)RP1 Issue 1
Detector
R
|
f
Pipe walls
SDD

Figure 14.4 Section along the pipe axis showing the three-dimensional cone beam
effect (for non zero values of the angle |).
60
HOIS(09)RP1 Issue 1

14.3.2 Dimensional comparator
An alternative method for dimensional calibration is the use of ball bearing or similar
dimensional comparator. This is an effectively radiation opaque object (usually spherical)
with a known diameter, which is placed close to the pipe, and in the same plane as the
tangent position on the pipe wall, as illustrated in Figure 14.5.

Detector
Comparator,
diameter c
c'

Figure 14.5 Tangential radiography showing use of comparator for dimensional
calibration. The comparator should be placed as close to the pipe wall as
possible, without overlapping it.

On-screen measurements of the imaged size of the comparator, using the CR system
software, then allow the pipe wall thickness measurement to be calibrated in mm, hence
allowing for the radiographic magnification or blow-up. The actual wall thickness WT is then
given by:

c'
c
WT' WT = (14.3)

Note that if the comparator cannot be placed adjacent to the pipe tangent position, due for
example to the presence of external insulation, it is recommended that the source is offset
from the pipe centre-line to be aligned with the pipe wall as shown in Figure 14.6.

As an alternative to the use of a dimensional comparator, if the outside diameter of the pipe
is known reliably, then the measured pipe diameter on the CR image can be used for
dimensional calibration. This method does not then require an additional calibration object,
and can have advantages in terms of accuracy provided the pipe outside diameter is known
accurately.

61
HOIS(09)RP1 Issue 1

Detector
Comparator

Figure 14.6 Tangential radiography showing use of offset source position with
comparator for dimensional calibration, for insulated pipes, where the
comparator should be placed as close to the outside of the insulation as
possible.

14.4 Use of lead strips to avoid burn-off
Some CR procedures specify the use of lead strips close to, or coincident with edge of the
pipe to avoid image burn-off effects at the pipe OD.

However, the use of these strips is not recommended for computed radiography, as they
would adversely affect the accuracy of the preferred wall-thickness measurement methods,
which involve analysis of the image values (profiles) extracted along lines orthogonal to the
pipe wall, see Section 14.5 below.

Burn-off should instead be minimised by ensuring the exposure time is adjusted so that the
unimpeded beam beyond the pipe wall is not greater than 90% of the dynamic range of the
CR plate (not too close to saturation).

In addition, the use of lead screens, thicker than the minimum values given in Table 10.1,
can be useful in avoiding burn-off, by reduction of scattered radiation.

14.5 Wall thickness measurement
14.5.1 Interactive on-screen measurements
Many CR systems contain software options which allow on-screen interactive dimensional
measurements using a cursor overlaid on the CR images. The user then judges by eye the
locations in the image of the inner and outer edges of the pipe wall.

This method can however be prone to error, especially if the penetrated thickness of the pipe
wall at the tangent position (w
max
) is approaching the maximum possible (see Section 7.2),
given the radiation source in use. This makes the exact positions of the outer and especially
62
HOIS(09)RP1 Issue 1
63
inner walls difficult to determine by eye. Furthermore, the apparent locations can be affected
significantly by the contrast and brightness settings in use on the CR image at the time.
Thus, the WT measured using this technique can change appreciably as the contrast and
brightness settings are varied on the same image.

If the on-screen measurement method is used, it should, if at all possible, first be checked for
accuracy using the current contrast and brightness settings of the displayed image, by
application to a section of the pipe with known wall thickness (e.g. known to be uncorroded
or not eroded).

For pipes having maximum tangential paths which approach the maximum permissible for
the source in use (see Table 6.2), the interactive on-screen measurement method is best
applied to CR images having logarithmic response functions to radiation exposure/dose,
which reduces the overall dynamic range of the image. This improves the visibility of the
position of internal pipe wall (ID).

Logarithmic CR images can be obtained from logarithmic-response CR scanners (e.g. the
CR100) or by application of an appropriate logarithmic look-up table (LUT) to CR images
resulting from scanners with linear response functions.

In addition, use of high-frequency spatial filtering (sharpening) is recommended since this
improves the accuracy of this measurement method, by emphasizing the positions of the
edges of the pipe wall in the CR images, and reducing any dependence on the contrast and
brightness settings on the image.

It is however recommended that for improved accuracy interactive on-screen measurements
are made in combination with grey-level profile analysis methods as described in the
following section.

14.5.2 Grey-level profile analysis methods
Many CR systems have software which allows the user to mark lines on the CR image
orthogonal to the pipe wall axis. The software extracts a grey-level profile along this line,
which is then generally presented on-screen, superimposed on the image. Measurements of
wall thickness can be obtained by either interactive or automated analysis of these grey-level
profiles.

Automated routines
Automated analysis routines can increase the reliability of the measured wall thickness
values, unless the maximum tangential penetrated thickness (w
max
) is approaching the
maximum possible, given the radiation source in use (see Section 7.2). In addition, other
factors such as the presence of external scale, corrosion products or irregular
internal/external corrosion may affect the accuracy of these automated routines.

In these cases, the automated routines are subject to uncertainties, and the operator should
check the consistency of the derived values with the density profile.

Interactive methods
In the absence of automated routines for wall thickness analysis, the operator should use the
available interactive facilities for analysis of the image profiles. Accuracy is likely to be
improved, especially for pipes having larger w
max
values, if the CR images have a logarithmic
response and are high-pass filtered.

HOIS(09)RP1 Issue 1
Figure 14.7 shows an example of interactive measurement of wall thickness, using cursors
on a grey level profile across the pipe wall, after applying a logarithmic look-up table to the
CR image, and high-pass filtering to enhance details. The position of the outer diameter
corresponds to a clear peak in the profile, and the location of the inner diameter is given by
the minimum and pronounced change in gradient of the profile.

This method, combined with a visual assessment of the image, is recommended.

Figure 14.7 Example of interactive wall thickness measurement using cursors
superimposed on a grey level profile taken across the pipe wall.

It should be noted that the accuracy of all measurement methods decrease as the tangential
penetrated thickness, w
max
, approaches the maximum value recommended for the isotope in
use (see Section 7.2), since the location of the inner wall becomes increasingly difficult to
determine with any reliability due to lack of contrast and increased noise.

64
HOIS(09)RP1 Issue 1
65

15 Scanner parameters, image recording and
processing
15.1 Scanner parameters
For the standard image quality, the scanner pixel size should be 100 microns or less.

For the higher image quality, the scanner pixel size should be 50 microns or less.

For a given radiographic exposure, it is important to note that increasing the scanner gain or
sensitivity increases the grey levels, but has negligible effect on the image quality, as
measured by the normalised signal noise ratio (SNR_N). To increase SNR_N, the exposure
must be increased, not the scanner gain.

For scanners with linear responses between radiation dose and grey level, use of low
gain/sensitivity reduces the risk of image saturation. For higher scanner gains, image
saturation may occur, especially in the free beam areas, for relatively short exposures, which
do not give sufficiently high image SNR_N values to meet the quality criteria given in Section
10.

15.2 Image recording and storage
The CR images from the scanner should be stored in a file format which supports a minimum
of 12-bits/pixel.

The images should be stored at full resolution and full dynamic range, as delivered by the
detector system. Only image processing required for detector calibration should be applied to
the images prior to storage of these raw data.

Any data compression techniques used in the storage of these files should be loss-less.

The following information should be recorded as header information, attached to each CR
image, to be used for reporting purposes.

(a) Name of company carrying out the inspection
(b) Test report number
(c) Component under test
(d) Date of test
(e) Material tested
(f) Nominal pipe diameter and wall thickness
(g) Details of any lagging/insulation, including thickness
(h) Radiographic technique and class
(i) Radiation source: type, size of focal spot & equipment used
(j) CR plate used (IP), screens and filters
(k) X-ray tube voltage & current or gamma ray source strength
(l) Exposure time
(m) Source to detector distance (SDD)
(n) Component to detector distance and/or source to component distance
(o) Type and location of IQIs used
(p) Any deviations from this procedure
HOIS(09)RP1 Issue 1
66
(q) Name and certification of responsible person
(r) Average pixel values (grey levels) in the area of interest
(s) Measured SNR, if software permits derivation

15.3 Image processing
All measures of image quality (mean image grey level in area of interest, signal to noise ratio
if available, IQI wire visibility etc) must be made on images which have not been subject to
any image processing routines which use spatial filtering.

Spatial filtering is defined as a process in which the value of the current image pixel is altered
by an amount which depends on other (usually adjacent) image pixel values. Examples of
spatial filtering or spatial image processing include noise smoothing/filtering, edge
sharpening, edge contrast, un-sharp masking etc. Spatial filtering such as edge sharpening
may however be beneficial for the interactive measurement of wall thickness using the
tangential method (see Section 14.5).

Image processing which changes contrast and brightness of the displayed CR image
(sometimes referred to as level control) is however permissible when evaluating image
quality, and strongly recommended for image evaluation at the monitor.

It is recommended that the results of any image processing are saved as a separate file, and
that the raw data derived from the CR scanner should not be altered by any subsequent
processing.

15.4 Monitor viewing conditions
The computed radiographs shall be examined in a darkened room using a monitor, or film
hardcopy, as specified in Section 7.10 of EN 14784-2.
HOIS(09)RP1 Issue 1
67

16 Acknowledgments
Several members of the HOIS JIP provided useful comments on earlier drafts of this
document.

17 References
1. Burch S F (2009). Results from HOIS CR Trials in Bergen October 2008. HOIS(08)R8
Issue 1. January 2009

2. Development of protocols for corrosion and deposit evaluation in large diameter pipes by
radiography IAEA report, Vienna, 2008.

3. Halmshaw, R (1995) Industrial Radiology Theory and Practice Second Edition, Chapman
& Hall, p125

4. Burch, S F and Collett N J, (2005) Recommended Practice for the rapid inspection of
small bore connectors using radiography, HSE Research Report, RR 294.

5. Burch S F (2009). Results from HOIS CR Trials in Bergen May 2009. HOIS Report
HOIS(09)R5 Issue 1. August 2009

6. Willems P, Vaessen B, Hueck W and U Ewert (1999) Applicability of computer
radiography for corrosion and wall thickness measurements. Insight, Vol 41, No 10, 635-
637.

HOIS(09)RP1 Issue 1

NDT - Auto RT - Hois

Hochgeladen von

Dokumentinformationen

Originaltitel

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

NDT - Auto RT - Hois

Hochgeladen von

Copyright:

Verfügbare Formate

Recommended Practice for the in-service

inspection of wall loss in pipes by computed

Das könnte Ihnen auch gefallen