APPENDIX 5
Memorandum by Dr John M Leeds
CONTENTS
1. Background Information
2. Introduction: Pipeline Corrosion Control
3. Corrosion Control and BTC Field Joint
Coatings
4. Specific Comments on the Field Joint
Specification
5. Comments on the Warley Parsons Review
6. Cathodic Protection
7. Principles of the DCVG Technique
8. Comments on the J P Kenney DCVG Survey
9. Short-term Activities
10. What to do with the Field Joints already
Buried
1. BACKGROUND
INFORMATION
My name is Dr John Michael Leeds and I am the
Managing Director of several small companies. I am a practising
Corrosion Engineer of more than 40 years' experience, having originally
studied Corrosion under Professor L L Shrier for my PhD in Electrochemistry
(Corrosion is Electrochemical in nature) as a Student of the Faculty
of Science of London University. All my studies since 1962 have
been concerned with problems that exist with all types of protective
coating systems, specialising in the causes, location and assessment
of holes in protective coatings.
I have been a Fellow of the Institute of Corrosion,
a Fellow of the Royal Society of Chemistry, a Fellow of the Institute
of Metal Finishing as well as a member of several other professional
bodies.
I am the author of more that 70 technical papers
on protective coatings, coating inspection techniques etc, and
have lectured at more than 25 International Conferences as well
as running Training Courses in Pipeline Coating Fault Location
Techniques, Cathodic Protection and Pipeline Rehabilitation at
more than 53 separate locations for some of the world's largest
oil and gas companies. I have also worked internationally in 39
different countries. In several countries I was in charge of the
company-wide corrosion control activities with five-year budgets
as high as 75 million dollars. I have also been technically responsible
for what in the 1980s was one of the world's largest pipeline
rehabilitation projects of live large diameter gas pipelines.
I have been requested to technically review
several documents that relate to the BTC pipeline being built
through Azerbaijan, Georgia and Turkey, for which BP is the Consortium
Operator.
In reading through and commenting on these documents
I must make several things plain to prevent any thoughts as to
any conflicts of interest:
(1) I have never done any work for or with
BP.
(2) I did work for Pipeline Induction Heat
Ltd from 1989-94 when I left to set up my own companies. At PIH
I set up and operated a consulting Group that had nothing to do
with the day-to-day field joint activities of PIH. I have done
no work for or with PIH since 14 June 1994. PIH is the BTC field
joint coating Contractor.
(3) I have never done any work with or for
Worley Parsons Energy Services.
(4) I have never been in Azerbaijan, Georgia
or Turkey.
(5) I have never done any work on the BP
BTC pipeline.
(6) I have done work on pipelines in Kazakhstan
for Chevron etc that feed eventually into pipelines operated by
BP.
(7) I do not know any of the BP BTC employees
or subcontract employees of BP that are currently working on the
BTC pipeline project.
(8) I have never done any work with or for
J P Kenny or Ionik Consulting.
(9) I do know Mr D Mortimore who has worked
for BP on the BTC pipeline project. I have known Mr Mortimore
since 1983 in his capacity as a pipeline Coating Expert and Corrosion
Engineer. My dealings with Mr Mortimore, which are often irregular,
have only ever been on a professional level of one Expert Corrosion
Engineer talking to another.
(10) I have met Mr M Gillard only once, quite
recently.
The Documents reviewed below are as follows:
(1) The Specification for Field Joint Coating
No 410088/00/L/MW/SP/015
(2) Desktop Study Final Report, Field Joint
Coating Review. Redacted Version by Worley Parsons Energy Services.
(3) DCVG Pipeline Surveys in Azerbaijan and
Georgia. J P Kenny.
(4) Caspian Connection Published in Frontiers
August 2003 p19
There are more reports of studies made on the
Field Joint problems but these have not been made available to
me. Judging from the comments in the Worley Parsons report which
did look at a bigger selection of reports, the additional studies
offer little to solving the real problems that still exist. Warley
Parsons also fail badly to identify and highlight these real problems
in their report.
2. INTRODUCTION:
PIPELINE CORROSION
CONTROL
For corrosion (which is an electrochemical process)
to take place the following five parameters need to be simultaneously
present:
(1) An anode area where the corrosion actually
takes place.
(2) A cathode area where the counter-electrode
process can occur.
(3) Both anode and cathode areas must be
in the same electrolyte, namely moist soil.
(4) The anode and cathode areas must be electrically
connected together and this is through the steel pipe itself.
(5) There must be a driving force between
anode and cathode for the corrosion to occur. This can be because
of different metallurgical structures, surface films oxygen content
of the soil, salt variations etc.
The process of controlling corrosion is simple
in concept. Remove any one of the above five parameters and corrosion
will not occur. We do this in a number of ways. We use a protective
coating to separate the anode and cathode areas from the electrolyte,
the moist soil. We can change the driving force by applying Cathodic
Protection thus suppressing the corrosion reaction. We could change
the electrolyte, for example bare pipe does not corrode in dry
(very high resistance) soil. We could change the pipe material
and use stainless steel instead of carbon steel.
The normal parameters used to control corrosion
of buried pipelines are those that have been chosen for use in
the BP BTC project:
(1) To apply a protective coating (the premier
corrosion mitigation mechanism) to separate potential anode and
cathode areas from the moist soil.
(2) To apply cathodic protection (the supporting
corrosion mitigation technique) to change the driving force between
anodic and cathodic areas on the pipeline to suppress the corrosion
reaction at any holes in the protective coating as it is impossible
to practically coat a pipeline so that it has no coating faults.
3. CORROSION
CONTROL AND
BTC PIPELINE COATINGS
Reviewing the corrosion mitigation techniques
to be used on the outside of the BTC pipeline in view of their
fundamental activities in the corrosion process the following
parameters are identified as important. These are widely known
and well documented facts for any technical person working in
the coating industry:
1. The pipeline protective coating along
its complete length must be resistant to water uptake. This is
a very important coating property.
2. Where there are holes in the coating the
cathodic protection must be shown to be effective. This point
will be dealt with in detail later.
3. The corrosion mitigation techniques must
be effective all the time for the full lifetime of the pipeline.
This means that the corrosion mitigation techniques for a 40-year
life must be as effective at the end of year 40 as they are at
day one. If not, then there are limited means to stop the pipeline
corrosion, ultimately resulting in a leak with catastrophic effects
on the environment.
4. It is a fact that most pipelines start
off being looked after but soon a change in people, company philosophy,
company profitability, off-load corrosion control and maintenance
to ineffectual third party companies etc, etc ensures that most
pipelines quickly reach the state of being out of sight, out of
mind, until a leak occurs. No such leak should occur if the pipeline
received proper surveying and maintenance. Adequate tools are
available, they require only the engineers who know how to apply
them and interpret the results, with the adequate budgets to ensure
proper corrosion control. Thought should be given to full implementation
of the ECDA (External Corrosion Direct Assessment) concept on
the BP pipeline from its day of construction. Unfortunately the
above means that the symptom of the real problem is being chased.
It is better to install a good coating from day one.
5. It is a fact of life that when a coating
starts to absorb water, initially chemically bound and later as
a separate phase, the Cathodic Protection assists the coating
failure mechanism through a process called electro-osmosis, forcing
water through the coating to the steel coating interface to form
blisters that eventually burst, exposing bare steel. Depending
upon the coating properties when degraded by water absorption,
some blisters can be as big as a hand. A coating that absorbs
water has very different mechanical properties to one that resists
water. Again laboratory tests should always be carried out on
samples that have been aged in water once their water resistance
has been determined so correct preconditioning conditions for
lab testing can be identified.
6. The parent outer coating of polyethylene
would be expected to be resistant to water uptake but not so for
any FBE (Fusion Bonded Epoxy) sub-coat or more important any epoxy
such as SP 2888 RG used as a field joint coating. Hence for the
long term performance of the field joint area knowledge of the
water resistance of SP 2888 RG is critical plus if it does absorb
water how does this affect adhesion to the steel and change the
mechanical properties of the coating particularly resistance to
soil stresses. I have not been able to obtain any independent
assessments of the water resistance or mechanical property variations
of SP 2888 RG with water content. It is expected that typical
of most pipeline coating suppliers, no such information is available
as such measurements are difficult to obtain if done correctly.
Measurement is not easy as it requires the correct electrical
equipment and techniques to be used. Occasionally it is possible
to find some measurements reported but these are often done by
immersing a lab prepared sample in water and observing any weight
gain. This is not the correct technique as it is too crude. The
suppliers of pipeline coatings are often negligent in investigating
this important property of their coatings.
7. The layer of FBE under the LD/HD polyethylene
coatings is usually only a very thin coating and it is common
for no properties particularly DSC values as to the degree of
cure of the FBE coating to be recorded during manufacture. If
applied to a pipeline that is too hot or held at a hot temperature
for to long the FBE forms a foam like structure, see Figure 1,
that has negligible resistance to water absorption and very poor
adhesion to the steel pipe. In such circumstances undermining
of the parent polyethylene coating occurs where shielded corrosion
can take place.
8. A major mistake in the Specification assumes
(as has the Worley Parsons Review) that the field joint area for
coating has a 100mm section either side of the weld of bare steel
then an area of 50 mm of exposed FBE then the cutback Polyethylene.
This is not what happens in practice. In the pipe mill the complete
pipe including the potential weld area are all coated in a continuous
process first with a thin layer of FBE then the LID polyethylene
then HD polyethylene. The two outer layers are then cut back and
stripped from the FBE leaving behind traces of polyethylene imbedded
in the FBE. The FBE must then be stripped back so the weld area
is clean. Invariable the amount of bare FBE left showing is variable
and usually not the nice 100mm as per Specification. Further,
there does exist a significant step from the FBE coated bare steel
to the outside of the HD polyethylene outer coating. This step
should be feathered to give a gradual change in diameter. However,
in reality this is usually not the case, the physical step remains
and any field joint coating method must be expected to adequately
cover this significant step.
9. The chosen field joint material, SP 2888
RG is applied not only on the steel but also as an overlap onto
the HD Polyethylene coating. The use of an overlap is important
in concept to get complete isolation of the steel pipe from the
moist soil. It is claimed that at the overlap adhesion exists
between the SP 2888 RG coating and the polyethylene. This is totally
wrong because as everybody who is in the industry knows, there
is no chemical adhesion between epoxies, (modified or not by urethane)
and HD polyethylene (the chemistry is totally wrong) and any adhesion
that exists is mechanical due to any roughening of the polyethylene
surface. The fact that poor adhesion will exist is recognised
within the Specification for Field Joint Coating 410088/00/L/MW/SP/015
itself by stating in Section 5.6 that adhesion is a pass when
a Ranking of =/>1 is achieved, described as "Peels in
Large Pieces Adhesively from the substrate", the HD polyethylene.
This Ranking, a classical example of under-specification, used
as a pass (acceptable for use) means that coatings with no bond
between the epoxy and the polyethylene are acceptable. This interpretation
in terms of the long term performance of the coating is totally
useless (totally unacceptable) and I am surprised that BP have
allowed this very low poor quality adhesion interpretation to
be used. Basically the Specification is saying that no adhesion
of the epoxy to the polyethylene is acceptable which is a very
strange assessment for the suitability of any field applied coating
and totally adverse to the corrosion control concept of stopping
corrosion by inserting an impermeable coating separating the anode/cathode
areas and the moist soil. No Coating Specification ever written
and used in practice would allow such a poor assessment of the
adhesion between coatings at such an important area to be considered
as acceptable for a 40 year life time pipeline. The Ranking should
be at a minimum of =/>3, the same as for steel as the substrate
and described as "Peels cohesively in large pieces"
which means that the epoxy bond to the polyethylene should be
strong and stronger than the bonds within the epoxy coating itself.
There is nothing that a field joint coating Contractor could do
to change the Ranking to =/>3 as the two coating systems are
incompatible.
10. A distinction is usually made between
an adhesion and cohesion. An adhesive force acts to hold two separate
bodies together (or to stick one body to another) Mechanisms for
adhesion include both mechanical adhesion and specific adhesion.
Mechanical adhesion occurs when the applied epoxy flows into the
texture of a roughened polyethylene substrate. Specific adhesion
includes electrostatic forces, Van der Waals forces and acid-base
interactions that should take place between the epoxy coating
and the substrate. But the polyethylene surface is non-polar with
a very low surface energy so these natural bonding forces do not
come into play. Adhesion failure happens when one layer delaminates
from another because there is no bonding. A cohesive force acts
to hold together the like or unlike atoms, ions, or molecules
of a single body and refers to the strength of the material to
support itself. Cohesion failure is when delamination occurs within
the epoxy coating. The big problem with putting the modified epoxy
onto the polyethylene surface is:
(1) The need to roughen to provide a
mechanical key. This is usually done by abrasive blasting which
is only effective up to a point. It is a very difficult surface
to roughen and in any case the Coating Specification calls for
only sweep blasting which is insufficient to significantly roughen
the surface. The epoxy then has to wet the surface to get intimate
contact but this is difficult on a surface that resists being
wetted.
(2) Flame treatment, a well known pre-treatment
to condition by oxidising the surface and make it more receptive
to coating by allowing wetting to occur and allowing some limited
bonding. The Specification only calls for Flame treatment of pipe
used in directional drilling. Its general use was not specified
and has not been used.
11. There exist well tried and tested under
field application and use, a number of field joint coating techniques
for 3 layer polyethylene coatings. These existing field joint
coating systems set out in the submission by D.Mortimore, except
for sleeves and tape, do not even seem to have been seriously
considered for the BP BTC pipeline and are dismissed on hearsay
which is most strange for such a high profile pipeline. If you
have something which does the job and these other systems have
been extensively applied and has with a working history why change
and in particular to use this very important pipeline as a proving
ground for an experiment with a new coating system.
12. A well known coating problem not so far
mentioned exists when coating a weld joint with a liquid system
too soon after welding. This problem occurs because in welding,
hydrogen is introduced into the weld joint and occurs no matter
what type of welding rod is used. As an example it is documented
that 30 to 100 ml of hydrogen is introduced into the weld area
for every 100gms of weld metal from cellulose electrodes. This
hydrogen occurring as hydrogen atoms slowly egresses out of the
metal forming hydrogen molecules. If coated by a liquid system
too soon after welding the coating develops a series of small
volcanoes formed as bubbles of hydrogen escape and are arranged
around the full pipe circumference in the heat affected zone either
side of the weld joint, see Figure 2. In setting, the thixotropy
of the epoxy limits mobility to back fill the hydrogen volcano
bubble sites. If the coating sets before the hydrogen has chance
to escape then the build up of pressure between the steel and
coating literally blows the coating off the steel surface and
the result is usually about a one cm wide strip of coating in
the heat affected zones either side of the weld joint completely
stripped around the full circumference. The egress of hydrogen
is temperature/time related. At ambient it can take a month to
disperse. At elevated temperatures it will be much shorter in
time but will also depend upon welding conditions. The preheating
to 80 degrees centigrade requirement in the field joint coating
Specification will help in hydrogen removal but is not applied
for long enough time.
13. ln preheating the joint area another
problem can arise due to moisture content in the FBE coating,
50 mm of which should be exposed as per Specification. A properly
applied and cured FBE will contain about 0.5% by weight of water
and if this is vapourised during any heating there is a big volume
expansion. As any chemist will tell you 18 cc of water will vapourise
to 22.4 litres at NTP. The effect is to seriously blister the
FBE coating making it unsuitable as a substrate on which to apply
a liquid epoxy coating system. Normally it takes a minimum of
15 minutes for any vapourised water to egress from the FBE coating.
Factors such as 10 and 12 above because of the time involved can
significantly add to the cost of each field joint coated.
14. A further problem not so far mentioned
is the onset of Stress Corrosion Cracking of the steel pipeline.
SCC, a most insidious form of pipeline failure, occurs in both
oil and gas pipelines though the more spectacular failures are
in gas pipelines, see Figures 3A and 3B. No pipeline steel can
be considered to be free from the possibility of the two types
of SCC failures, low pH and Carbonate/Bicarbonate. Because of
the environments through which the pipeline is buried, there are
sections of pipeline where either type of SCC can occur. For discussion
we consider only Carbonate/Bicarbonate SCC. No metal loss is necessary
for SCC as cracks in the steel, parallel to the axial direction
of the pipe develop and then progressively grow until they reach
the critical crack length for the pipe wall thickness and steel
strength. The pipeline then spontaneously bursts open. For SCC
to develop a number of different conditions have to occur at the
same time. SCC is more prevalent at higher pipeline operating
temperatures, and low pipe to soil potentials (ineffective Cathodic
Protection.), at a coating fault particularly if the steel is
under a disbonded area of coating where the right environment
can develop. Two other factors are necessary, the soil resistivity
must be below 1,500 to 2,000 ohm cm and the pipeline must be operating
with cyclic stresses that exceed 46% of the SMYS (specified minimum
yield strength of the steel). A coating fault is necessary for
all forms of SCC as it does .not occur in steel that is not exposed
to the correct environment for development of the cracks. Hence
the importance of applying the best coating system with no possibility
of disbondment particularly as the pipeline runs through some
very low soil resistivity areas. A disbonded coating will exist
as there is no effective adhesion between polyethylene and epoxy
thus allowing the ingress of moisture etc at the coating overlap
area. Techniques are available to search for SCC. One is the use
of inline inspection pigs but despite supplier claims which are
often exaggerated, they are limited in the detection of SCC due
to fact that the minimum documented crack size they can detect
is about 2.5 cm long which is a big crack. Also there is the fact
that separating inspection "noise" from real crack locations
is difficult when analysing pig data. Inline inspection sensitivity
will also be limited by the build up on the pipe walls and pig
sensors of wax/asphaltenic deposits from the crude. A SCC failure
in the gas pipeline that is to run parallel to the Oil Pipeline
in part of the route would seriously damaged the oil pipeline
with the high prospect of significant environmental contamination.
4. SPECIFIC COMMENTS
ON THE
FIELD JOINT
SPECIFICATION ISSUES
FOR CONSTRUCTION.
(LATEST DATE
ON DOCUMENT
02.10.02) REFERENCE 410088/00/L/MW/SP/015
4.1 General Comments
A. Considering the total Document, it is
not up to BP's usual high standard and is totally unsatisfactory
to be issued by a professional company such as BP to any Contractor
as a Specification to which he must coat field joints. Reasons
for this comment can be seen below. A number of items are under
specified considering normal pipeline constructing practices.
The general document is so loose that BP would have little recourse
to any claim if poor quality field joint work is produced. A Technical
Specification should be very specific leaving little for interpretation
by the Contractor otherwise it becomes a licence for the Contractor
to print money.
B. A major mistake is in specifying only
one coating material, in this case SP 2888 RG (Claimed to be a
modified Urethane Epoxy) a relatively untried pipeline coating
with little in use performance history. BP therefore has taken
full responsibility for the quality of any field joints produced
where the coating does not perform satisfactorily. Further as
only one coating was chosen the Specification should contain all
the Manufacturers Product and Technical Information and Literature
so limits are clearly defined and can be agreed by all concerned.
More surprising is the fact as stated in the technical article
in Frontiers August 2003 by Terry Knott on page 22 BP say,
"As far as we know this is the first time that such a system
has been employed". Proof that this expensive pipeline was
being used as a test bed which from BP's point of view must be
a silly situation to be placed in.
C. Despite claims summarised in the Worley
Parsons Report, 3.3 p7 and 3.7 p10 and p15 in working in 39 countries
on buried pipeline coating problems over the last 25 years I have
never seen the product SP 2888 RG used anywhere, even in USA.
I have never before seen the SP product mentioned amongst a list
of competitive products bid for any field joint or pipeline coating
rehabilitation project.
D. Not every epoxy especially a modified
epoxy is suitable for use as a pipeline coating so the selection
of a specific product must be thoroughly investigated as to its
long-term suitability. Laboratory tests are of very limited value,
in fact it is possible to get any result from some tests, a classical
example being cathodic disbondment testing, yet the Specification
in Section 1.1.2 specifically states "the coating shall have
proven good resistance to Cathodic Disbondment". Obviously
the Specification writers do not understand the limitations of
the inspection techniques specified. Practical factors such as
past application difficulties, experience in use under different
environmental conditions and a number of years of actual in service
history are vital inputs to the coating selection decision making
process. I have seen no such supporting evidence by the use of
SP 2888 RG on other projects internationally.
E. Section 1.5 of the Specification is mostly
complete nonsense. The first part talks about BP approving the
coating material as though a number of different coatings are
available and requesting technical information, but in Section
4.1 it states that the coating material shall be SP 2888 RG. In
specifying a particular product it suggests that BP has in fact
all the relevant technical information required in 1.5.
F. Section 2 is just general Contractual
Requirements and a professional Field Joint Company should be
use to handling this type of Working Activities, Quality Control
and Testing requirements.
G. Section 3, has item 3.1 impractical under
field conditions particularly the field removal of salt. You can
guarantee no salt was found on the pipe as this part of the Specification
is fluid to interpretation. The contractor most probably will
not have tanks of water etc in very cold conditions with the water
freezing onto the pipe.
H. Section 3.2 seems a typical surface preparation
but many terms are qualitative. Surprised they allow copper slag
as an abrasive as it is known that in corroding environments steel
at any unrepaired coating fault can become covered by a film of
copper metal by a replacement reaction. This can give rise to
accelerated corrosion of the steel because of the bimetallic couple.
Such slag should not be used as it is not possible to ensure it
will not contaminate the pipe trench. There are better cleaner
abrasives that are more commonly used that can achieve the desired
profile quicker and would give a better profile on the polyethylene.
I. In 3.2 Flame Treatment of the HDPE Coating
is mentioned and is well known to be one of the treatments that
will improve adhesion to polyethylene. However, its use is only
required on sections where directional drilling is used. In view
of the well known difficulty of getting anything to adhere to
polyethylene it is surprising that Flame Treatment was not generally
used on the basis of applying all means possible to get some kind
of adhesion to the polyethylene.
J. Section 3.5 is too loose. All conditions
under which surface preparation and coating should not be done
need to be carefully and specifically laid out. References to
the Manufacturer for interpretation are unwise. He is not usually
in the field when field joint coating is taking place. This type
of information should have been sourced from the manufacturer
then together with BP experience, written as mandatory in the
specification clause by clause.
K. Section 4. At this stage the surface
of the field joint area should have been preheated to 80 degrees
centigrade and abrasive blasted. The time specified between surface
preparation and coating application is specified as immediate.
The pipe temperature could well be above the 50 degree centigrade
maximum specified in 4.4. The directions given in the specification
are badly thought out. Epoxies can be applied better and cure
more quickly on a warm substrate. However, they can be subject
to slumping if the coating thickness is built up too quickly.
The control of temperature which is a very important parameter
in the state of the pipe surface and in curing the epoxy is throughout
the whole specification not properly and clearly specified. When
temperatures are low the epoxy will not cure and be very friable.
When the pipe is hot the epoxy will run or slump more easily particularly
at higher coating thicknesses, see Figures 4A and 4B. This lack
of temperature control, a very important parameter, is not acceptable
and a major weakness in the original Specification. The curing
of epoxies has a temperature/time relationship. The higher the
temperature the quicker they cure. No routine measurements seem
to be specified as to the degree of cure. Whilst a Shore Hardness
of 80 is specified in prequalification, it is normal practice
to specify routine measurements in production. One commonly used
field technique is a Barcol hardness tester. Typically epoxies
can take up to 48 hours even in temperatures of 30 degrees centigrade
to harden, the Barcol Hardness reading showing an exponential
relationship to maximum cure hardness. Original Manufacture information
should have included construction of a graphical presentation
of hardness with time, done at different temperatures to clearly
identify the curing envelope of the SP 2888 RG epoxy. This type
of data also needs to be available for different variations in
the proportions of Base and Hardener used to make the epoxy to
identify how much variation in the mix can be tolerated.
L. 5.3 Holiday testing. In simple terms
this name peculiarly applied to the pipeline industry involves
identifying any holes in the coating by using a high voltage spark
tester. In this case it is a 4KV instrument. Holiday testing is
usually defined as volts per micron of coating thickness. In Section
4.5 the coating thickness is stated to be 750 to 1,250 microns.
Assuming a thickness midway of 1,000 microns this gives a Specification
Holiday voltage of four volts per micron. This is inadequate.
All the specifications on FBE and other thin film coatings typically
use the voltage of five volts per micron. This would require a
minimum 5KV holiday tester. In reality a properly cured epoxy
has a breakdown voltage of 40 plus volts per micron. Fundamental
studies published 17 years ago in Australia on what voltage to
holiday test epoxies showed that at five volts per micron you
could get a set of holidays in pass one along the pipe. If you
repeat you can get some more. The work showed that even at five
volts per micron you missed holidays. The research showed that
the best voltage to use was 10 to 15 volts per micron. However,
industry for obvious reasons still sticks to five volts per micron.
It is obvious that at an under specification of four volts per
micron many weak areas of coating will be missed and thus hides
basic faults in the coating material. One pleasing thing in the
specification is the comment "A fine wire metallic brush
electrode shall be used . . ." The published research identified
this type of electrode as the best. The typical coiled spring
electrode widely used was poor particularly at low voltages and
with coil spacing above one cm.
M. Sections 4.4 and 4.5 Thickness Tolerance.
The reference in both sections of the Specification to Manufacturers
recommendations is inadequate. The Specification particularly
as it is a single source product, should be a standalone document
with all parameters and limits etc clearly defined. 4.5 states
the dry film thickness should be 750 to 1,250 microns, but how
does the applicator know what thickness coating he has applied.
The usual method is to use a wet film thickness comb. There is
no mention of this technique in the Specification and its application
during the coating process is more important so coating thickness
can be adjusted rather than wait until coating is cured and then
take readings and then try and make corrections by over spraying
the cured coating.
N. Section 4.4 last paragraph. Suggests
that a roller system is the preferred coating method. This is
impractical on large diameter pipelines and whilst spraying is
suggested as an alternative it is not to be used presumably at
steel temperatures above 50 degrees centigrade though the wording
of the text states Operating temperatures above 50 degrees which
implies the normal operating temperature of the pipeline. The
wording is loose in terms of what the industry would normally
understand. The Specification should insist on the use of a plural
component spray system where the two components are mixed in the
spray so work can be continuous and not on a batch process as
would be the case with a roller. Plural spraying is also more
economical on materials and hence more cost effective. Section
4.3 specifically deals with mixing of the two components. This
section is immaterial when using plural spraying.
O. Section 4.6 deals with Repair of faults
in the field joint coating. The end of the section has a paragraph
on coating repair of CP test Cable connections. This has nothing
to do with Field Joint coating and in any case the use of epoxies
to cover the cable etc of such a cable connection is ill advised.
Also placing a test cable so close (7.5 cm) to a circumferential
weld is also ill advised, it is simply too close. There is no
mention of what the repair coating should be. Often it is different
from the parent coating in this case the epoxy SP 2888 RG.
P. 5.5 Impact Resistance. This test is somewhat
impractical in the field and as the Specification states is only
possible on the top of the pipe. During the lifetime of a pipeline
most coating failures occur on the bottom of the pipeline and
that is the area that needs to have the best coating assured.
The Holiday testing that determines rejection of a joint coating
needs to be updated as stated earlier which will cause more joints
to be rejected. I question the value of this test in the field
as it is not what happens in practice unless the construction
contractor is backfilling with rock in which case a rockshield
layer should be used to provide extra protection to the coating.
I am not aware of any rockshield material being used in the construction
of the pipeline.
Q. 5.6 Adhesion. This is dealt with earlier.
The method of evaluating adhesion particularly to the Polyethylene
coating overlap is totally unsatisfactory and should be properly
upgraded to normal industry standards.
R. 5.7 Penetration test. This part of the
specification has in concept three samples being cut out of a
welded pipe for testing with no thought given to the cost involved
in cutting open the weld area. The existing coating must be removed
and in addition a new cut back of the polyethylene and FBE coatings
is required for a new weld joint. The area where the pipe sample
for testing was taken needs to be cut out, remake the pipe ends
including the bevel then somehow pulling the pipe ends back together
(which may be impossible) and re-welding before recoating. This
is a very impractical requirement in the test procedure and would
be a waste of time on the actual pipeline.
S. 5.8 Hardness, already dealt with. There
is nothing in the Specification that identifies how many should
be taken per field joint and what should happen if it does not
meet specification.
T. 5.9 Cathodic Disbondment test. This is
a pre-qualification test that can only be interpreted as to the
input from the Field Joint Coating Contractor. If there is a problem
with the test due to weaknesses in the coating composition provided
correct mix used etc then it is a BP problem as they specify only
one coating material. The Specification calls for the use of salt
in "ionised water" what ever that is. No mention is
made of the importance of separating anode and cathode electrolytes
which from an electrochemical point of view is important, or the
use of phenolphthalein to more accurately delineate the disbondment
area. These latest additions to the technique are ignored.
5. COMMENTS ON
THE WARLEY
PARSONS ENERGY
SERVICES BTC FIELD
JOINT COATING
REVIEW DATED
15 JULY 2004
5.1 General Comments
1. The Warley Parsons report is simply a
regurgitation of selected information from selected documents
and attempts to pull together the results from various studies
made in order to justify the in-trenched decision as to the single
source use of SP 2888 RG epoxy as the field joint coating material.
The report propagates inherent errors and in some cases makes
ill informed or incorrect comments on the documents reviewed and
criticisms by other Consultants. In the WP report p24 item 7.1
states" that the technical issues had been resolved".
This is incorrect as in all the BTC Studies, fundamental inadequacies
in the Field Joint Coating design in particular the total lack
of adhesion between epoxy and polyethylene have not been resolved.
Neither is long term damage likely to be caused by the use of
22mm rock backfill without any rockguard material. Hence CCIC
and SPJV, the constructing contractors are correct in expressing
concern over their construction warranty.
2. It is interesting to note that the documents
identifiable as reviewed by Warley Parsons are all except one
dated well after the Specification for Field Joint Coating No
410088/00/L/MW/SP/015 was originally issued, the specific coating
SP 2888 RG selected and construction had begun with all the qualification
procedures supposedly approved. The fact that so many fundamental
studies were initiated as an after event sounds very much like
having to justify the coating selection made. "Closing the
stable door after the horse had bolted". This is not an acceptable
technical practice particularly as Environmental Data was supplied
as Table 1 in the 3 July Field Joint Coating Specification and
the basic coating performance characteristics should have been
known particularly temperature characteristics after all it does
get cold in Canada where SP 2888 RG is made and supposedly widely
used. It is noted that a modified Specification was issued on
09/02/04, six months after construction started and incorporated
pre and post heating procedures.
3. Several references are made in the WP
report that SP 2888 RG has a history of use as a pipe coating.
There are no references cited and no detailed evidence in any
document anywhere to support this and it must therefore be dismissed
particularly as in the Frontiers August 2003 document it
clearly states that it's the first time the system had been employed.
4. In the opening line of the WP report
two bad errors are propagated:
(A) "Lack of confidence in the field
joint coatings used with three layer systems to date . . .
" No comparison tests with other established field joint
methods were made prior to the epoxy selection. However, a detailed
trial that was organised in 2 July in France was arrogantly cancelled
by BP as they had already made up their mind to dismiss well tried
and tested coating systems and all other systems, instead, to
select only SP 2888 RG. At such an early date in Specification
evolution, when change is easily possible, this seems a most peculiar
way for BP to act.
(B) The desire to use larger (22mm) backfill.
The Avantica tests were only concerned with the effects of impact
during backfill will have and not more correctly the long term
effects of the backfill causing serious coating damage. The use
of 22mm backfill as anybody who has excavated pipelines will tell
you, is silly. BP should use a much finer padding around the pipe
and where this poses significant logistical problems, at a minimum
they should use a rockguard around the pipeline. The use of a
rockguard, a typical industry practice is not mentioned at all
in any document reviewed. Full of oil some sections of pipe will
weigh about one ton per metre of pipe. It is a fact that 22mm
rock will gouge its way through both any Epoxy coating and also
any three layer coating. It has to be realised that all pipelines
move in the ground and this movement will in time cause very significant
gouging around the pipe circumference with the most gouging being
on the bottom of the pipeline. The epoxy that has no adhesion
to the polyethylene as interpreted by the Specification adhesion
tests, will be ripped from the pipeline causing significant damage
to the step area. When inspecting other buried pipelines, many
of the coating failures that we delineate and excavate are caused
by rock damage, some by very soft rocks such as pumice. During
its lifetime, significant coating damage to the BTC pipeline coating
can be expected by the rock backfill, this damage enhanced by
the fact that the polyethylene coating will be softened by the
pipeline operating at warm to hot temperatures and the epoxy having
no adhesion to the polyethylene.
5. Warley Parsons and their Independent
Engineer have totally missed the point on p17/18 re the mistake
of BP nominating only one field joint coating material. The usual
industry practice is for a company like BP to nominate 3 coating
systems that meet requirements so that the coating contractor
can therefore get the best commercial deal. The way the present
single coating is nominated is a fix—no flexibility is available.
A lot of BP money will have been wasted on paying two times market
price for SP 2888 RG compared to other products, for the unnecessary
extensive testing that has had to be done (with major problems
still unresolved) to gather information that should have been
in place before the epoxy was selected, for extensive repair of
field joints where the coating has cracked and for installing
more expensive Big Fink Zap Guard Test Posts, etc making a mockery
of any Cost Reduction Study.
6. No mention is made in the review as to
what type of performance was being achieved in the Turkey section
where a urethane tar field joint coating is being used. This coating
has been applied successfully to thousands of three layer field
joints by PIH in Algeria, and Tunisia. Why has there been no comparison.
No documents have identified the consistency of 100mm of bare
steel at each end of a pipe length, of the 50 mm cut back of polyethylene
leaving only FBE on the steel, how much contamination exists of
the FBE contamination by residual polyethylene and most important
has every step from FBE to outside of the polyethylene been feathered
before any epoxy coating has been applied.
7. On page 15 of the WP report, the BTC
comment on "negative attitude" is stupid by using a
comparison to coal tar and asphalt coatings as poor and FBE and
three layer as superior coatings. This comparison bears no relationship
to the present case and obviously BTC people are unaware of the
massive failures of FBE in West Australia, Oman, Saudi Arabia,
Venezuela etc FBE on its own has proved a disasterous pipeline
coating for many companies with massive failure showing within
two years of laying. Coal Tar /Asphalt coatings have served the
pipe industry very well for more than 70 years and we still see
their excellent performance during surveys. Remember many coal
tar coatings were hand applied with no quality control, no specifications
etc If we applied coal tar under todays more stringent conditions
they would perform superbly. Coal tar coatings are still widely
used as a sub sea pipeline coating.
6. CATHODIC PROTECTION
6.1 Introduction
A Cathodic Protection system is actually a large
electrochemical cell constructed in the soil where the pipeline
is the cathode and a specially constructed groundbed, the anode.
The cell electrolyte is the moist soil and the electrochemical
reactions at both anode and cathode involve the decomposition
of water. At the anode water decomposes to give off oxygen and
leave behind acid conditions. At the pipe surface the decomposition
of water gives off hydrogen, and also together with the consumption
of oxygen dissolved in the water (secondary process) generate
hydroxyl ions, (alkali) that absorbs carbon dioxide from the soil
to form carbonate and bicarbonate ions. A carbonate/bicarbonate
environment naturally generated by the application of CP is one
of the requirements for the development of Stress Corrosion Cracking
of the pipeline. The hydroxide generated by the CP will also attempt
to saponify the epoxy reducing its resistance to water uptake.
Under the influence of the CP water can be driven through the
epoxy to form blisters resulting in epoxy coating failure.
6.2 Comment
1. No where do any studies refer to the
resistance of SP 2888 RG to strong alkali (pH 14) at the predicted
operating temperature which is what the coating can be exposed
to under normal operation. Similar comments apply to the thin
FBE coating used under the epoxy and polyethylene.
2. It is well known that CP of land based
systems do not work properly when the soil resistivity is very
low or very high. It will not work where a trench has been cut
through solid rock. The trench is surrounded by the high resistance
rock. The backfilled trench becomes a lower resistance drainage
trench for the local area, so the pipeline is often continually
moist and corrosion can occur as there is negligible CP available
to prevent corrosion. The only way CP can be made to work is to
install a continuous anode (such as Anodeflex) laid parallel to
the pipeline at pipe depth at the edge of the trench. CP also
does not work in very low resistivity soils (10 ohm cm) where
there is pure salt (bands of salt have been reported along the
pipeline route), as CP will not "throw" very far along
the pipeline. As an effect of this many more CP installations
will be required in order to protect the pipeline. The CP concept
as specified has installations close to pump and valve stations.
These are probably too far apart in both rocky and low resistance
soils so extra power lines will be required for the additional
installations.
3. The criteria for protection, (a potential
value used as the divide between protection and corrosion) is
typically taken as -850mV OFF vs Cu/CuSO4 reference electrode.
This value is not cast in concrete, instead it is now well known
that at ambient temperatures, in some soils, protection, ie zero
corrosion rate, can occur at -600mV but in other soils -1,150
mV is required. What protective potentials do you interpret pipe
to soil potentials to along the full pipeline length is unknown
as no knowledge exists as to the variation in soil corrosivity
is currently limited? The only real criterion at each coating
fault to ensure protection is to confirm that each fault is a
recipient of net current flow to the exposed steel. Techniques
identified within this report identify the technology to use.
4. In assessing the protection, two types
of coating fault problems exist:
(1) A simple hole or crack of the epoxy through
to the steel.
(2) Faults through the polyethylene or at
the polyethylene to FBE step at a field joint where undermining
of the polyethylene via the FBE coating can occur.
5. Various BTC Studies and reports reference
that CP is a "cure all" and can restrict corrosion of
steel at un-repaired coating faults. I am advised that BP personnel
have decided that coating cracks of six inches or less are now
considered acceptable for burial without repair. This is just
not acceptable, making a mockery of any Specifications and Contractual
requirements and demonstrates the cavalier attitude taken by on-site
people. Burying such coating faults so they are out of site out
of mind, faults that could so easily be repaired before pipe laying
is nothing short of disgraceful.
6. The idea that CP can look after all such
coating faults identifies that such comments are made by people
who do not understand the limitations of CP and the interaction
between coating fault location, severity and CP. Just applying
CP is no guarantee that every coating fault between CP system
installations are receiving enough CP current for protection.
A CIPS (Close Interval Potential Survey) survey, see Figure six,
will not provide adequate information about protection as it is
well documented in the technical literature that the CIPS technique
widely used to assess the level of potential and hence protection
does not see detailed information about small faults such as the
cracks.
7. A number of reported studies by others
as to the effective throw of CP into the orifice and crevice under
a disbonded coating have shown that the depth of throw into the
crevice is very dependent on the actual potential at the orifice
itself. Such a crevice arises due to undermining of the polyethylene
at the step. If the CP is weak then there is little chance of
any CP throw and so no safeguard of protection. It is very important
that there is strong adhesion between the field joint coating
material and the polyethylene to minimise this type of problem
yet no studies correctly address the poor adhesion inherent in
applying epoxy to polyethylene. There is a fixation in all reports
about cracking and virtually nothing about the real problem, the
lack of adhesion between epoxy and polyethylene.
7. PRINCIPLE
OF THE
DC VOLTAGE GRADIENT
BURIED PIPELINE
COATING SURVEY
TECHNIQUE
7.1 Introduction
This section is put into the submission so that
readers can understand the technology. We are the only serious
manufacturers and suppliers of DCVG Equipment and Market leaders
in DCVG Surveys in the Northern Hemisphere.
When DC is applied to a pipeline in the same
manner as in cathodic protection, a voltage gradient is established
in the ground due to the passage of current from the anode through
the resistive soil to the bare steel exposed at a coating fault.
The voltage gradient becomes larger and more concentrated the
greater the current flowing and the closer you are to a coating
fault location. In general, the larger the fault, the greater
the current flow and hence bigger the voltage gradient.
The DC voltage gradient method uses a specially
constructed sensitive milli-volt meter, to indicate the potential
difference between two copper/copper sulphate half-cells placed
in the soil in the voltage gradient at ground level. If spaced
two metres apart in a voltage gradient, one half cell will adopt
a more positive potential than the other, which thus enables the
size of the gradient and direction of the current flow causing
the voltage gradient to be established see Figure 5. To make it
easier to interpret and to separate what is being monitored from
other DC sources such as long line cells, tellurics, other CP
systems, etc, in the DC Voltage Gradient Technique, the asymmetrical
DC signal impressed onto the pipeline is switched ON and OFF at
the rate of 0.45 seconds ON, 0.8 seconds OFF. The DC signal can
be impressed on top of existing CP systems or the pipeline CP
Transformer Rectifiers (T/R) can be switched by using a special
interrupter inserted into the negative lead from the Transformer
Rectifier.
In carrying out a survey, the surveyor walks
the pipeline route testing at regular intervals with the probes
in a position of one in front of the other, separated by one to
two metres, parallel and preferably above the pipeline, (though
not essential provided you can pick up the voltage gradient from
faults in the pipeline route). As a fault is approached, the surveyor
will see the milli-volt meter start to respond to the ON/OFF pulsed
current, which is either a coating fault or interference from
another structure. When the fault is passed, the needle deflection
completely reverses and slowly decreases as the surveyor moves
away from the fault. By retracing, the position of the probes
can be found where the needle shows no deflection, ie: a null.
The fault is then sited midway between the two copper/copper sulphate
half-cells.
This procedure is repeated at right angles to
the first set of observations and where the two midway positions
cross is the epicentre of the voltage gradient. This is directly
above the coating fault to within a 15 cm circle. Once located
a series of electrical measurements are made that allow the severity
of the fault and its corrosion status to be determined. The DCVG
technique is capable of locating a coating fault the size of a
small fingernail buried 2 metres deep to within the 15 cm circle.
The DCVG technique is more than capable at delineating
the cracking that has been observed in the epoxy coating on steel
pipe.
8. COMMENTS ON
THE J P KENNY
BTC DCVG SURVEY REPORT,
PROJECT 052530. REFERENCE
2530-01-A-3-002 DATED 07/05/04
I have reviewed this report and offer the following
comments. It has to be realised that as market leaders in DCVG,
it is not uncommon for companies to pass their DCVG data/reports
etc to us for review/interpretation. The DCVG fieldwork presented
in the Kenny report was carried out in April 2004. The scope of
the work identified in the report was very selective as to sections
of the pipeline surveyed. This is not satisfactory. The complete
pipeline should be surveyed. In order to do this the following
test parameters should be applied and reported.
1. The survey should be carried out at least
three to six months after backfilling, during which heavy rain
should have occurred and the soil thoroughly wetted and consolidated.
This ensures electrochemical contact between the soil electrolyte
and any steel exposed at coating faults. No indication is given
of any time lapse between laying the pipeline and surveying although
at some locations bellholes were reported as full of water indicating
heavy rain had occurred.
2. There should be adequate DCVG pulse amplitude
on the pipeline so that coating faults particularly on the bottom
of the pipeline (42 inch) can generate a gradient that can be
seen on the soil surface. Ideally for the type of work done, this
should be between 1,250 and 1,500mV over all sections to be surveyed.
A number of survey pulse amplitudes reported were half this value.
Also, the temporary CP system used could only consistently output
low amperes so technical set up limitations had prevented better
DCVG pulse amplitudes. Ideally the report should present a graph
of DCVG pulse amplitude vs distance to show survey sensitivity
with distance.
3. The type of equipment used, make and
recent calibration certificate should be presented. None of this
type of information was reported. We have seen some surveyor's
equipment having only 45% of the sensitivity of newly calibrated
equipment. Poor sensitivity means the DCVG equipment would not
detect to the desired accuracy and at 45% misses many significant
faults with severity less than 30%. Calibration is very important.
4. The surveyor should be Accredited to
prove they have been properly trained in the technique and have
the ability to understand and carry out the DCVG survey correctly.
No information on surveyor Accreditation was provided.
5. Sacrificial anodes if connected to the
pipeline during a survey limit the ability to identify faults
located in close proximity to the anodes. The report contains
no information on exact anode locations.
6. GPS was used for fault location. Only
crude GPS information was reported and did not include altitude.
Altitude is required to calculate distance. No indication as to
the type and accuracy of GPS equipment used is reported.
7. When carrying out a DCVG survey it is
possible to identify the orientation of the fault around the pipeline
circumference. This is invaluable particularly to identify faults
on the bottom of the pipeline where most faults will occur. No
such information reported.
8. The DCVG technique is capable of identifying
if a coating fault is receiving enough CP current for protection.
This variation of the DCVG technique identifies if sacrificial
anodes are adequate and also CP from a temporary system can adequately
protect coating faults particularly from the proposed ground bed
locations. DCVG provides coating fault specific information and
provides a more accurate assessment than is possible with pipe
to soil potential (CIPS) measurements. None of this type of information
was reported.
The Kenny report only provides the bare minimum
information. Also I recognise some of the wording in the report
as being plagiarised from my past technical papers and reports.
Unfortunately this is not uncommon and comes with being market
leaders in the technology.
The Kenny report references the DCVG procedure
as set out in Contract Document (ref C-04-BTC-59100) Section 3
Method Statement. I have not had the opportunity to view this
document but if the Kenny report is "to the letter"
then this Method Statement is inadequate.
As an overall comment on the DCVG survey report
by Kenny, the equipment and conditions used and the selected areas
plus lack of use of the other DCVG capabilities suggest that the
survey conditions to report the MINIMUM coating problems seem
to have been selected.
9. SHORT TERM
ACTIVITIES
It is very important that the BTC pipeline is
constructed correctly and put into operation as soon as possible
so that the parties financing the project can begin to recover
their big investment. It is therefore necessary to satisfactorily
and urgently sort the technical problems with field joint coating
even to the extent of abandoning the existing coating system,
instead of as at present, go round in circles continually trying
to justify the selection of the single sourcing of an epoxy whilst
still not addressing some fundamental corrosion control concepts.
It must be concluded that SP 2888RG has not performed satisfactorily
in the field and is not adhering to the polyethylene. It never
will. There are too many separate interests involved in this project
and there seems to be to much pre-occupation in trying to cover
major technical errors.
In order to resolve the situation in which BP
finds itself so the pipeline construction can be completed as
soon as possible to the best possible practices the following
actions need to be implemented.
1. The "cupboard needs to be cleaned".
Because of Technical incompetence, bad on site practices and trying
to disguise major faults and having done no work to resolve the
major technical problem of polyethylene adhesion with the field
joint coating system, all BP personnel involved at all levels
in the intimate process of field joint coating specification material
choice and application should be removed and preferably dismissed.
Their input into the success of the BP project is very questionable.
A number of the decisions they have made also strongly suggest
a different agenda to that of BP.
2. Employ some proper experienced Corrosion
Engineers who have proper corrosion engineering training, with
minimum MSc or PhD from a reputed University such as UMIST. Obviously
BP don't have any working at present on this job.
3. Ensure that the Corrosion Engineer has
a dominant final say in all aspects of the corrosion mitigation
techniques used on the pipeline.
4. Urgently review the use of SP 2888 RG
in relation to its on site performance on the pipeline and be
prepared to totally replace SP 2888 RG with a better product or
field joint coating techniques. We need to know the following:
(a) How SPC are going to make the adhesion
on the polyethylene the same as on the steel?
(b) What is effect of strong alkali on the
SP 2888 RG coating?
(c) Why was temperature effects not available
before starting field work?
(d) Why was data relating to variation of
mix and temperature not available?
(e) Why is price for SP 2888 RG so high?
(f) Where is the real data that identifies
SP 2888 RG as being better than existing tried and tested field
joint coating systems as no proper comparisons seem to have been
made?
(g) Why was such a low holiday voltage chosen?
(h) Why are coating cracks currently being
buried?
(i) Can it be proven that the step in the
polyethylene coating has always been feathered as it should be?
(j) What inspection data is available to
say the base layer of FBE is fit for purpose?
5. Take a fresh look at existing field joint
systems. It is important to get unbiased input from PIH and other
field joint coating companies as to what they can offer as alternative
coating solutions. These companies are experts at applying field
joint coatings but are working under duress with their hands tied
behind their backs with the present coating material.
6. From the existing knowledge base within
the industry it should be possible to have field joint problems
resolved. Solutions exist, there has never been the need to "reinvent
the wheel". To calm Government and Contractor concerns initially
a "belts and braces" approach should be adopted. Implement
a revised coating system within four weeks.
7. The proposed CP design and construction/materials
specifications should be independently reviewed and alterations
made on basis of available information.
8. The permanent CP system should be installed
and made operational as soon as possible. The sacrificial anodes
should then be disconnected from the pipeline.
9. The 22mm rock backfill should be seriously
reconsidered. The application of a rockguard material should be
a minimum requirement if crushed rock is to be used.
10. A proper DCVG Survey should be carried
out on the whole pipeline to form the construction baseline survey.
The DCVG survey should have the following:
(a) DCVG Accredited Surveyor and certificates.
(b) Properly recently calibrated DCVG Equipment
and certificates.
(c) Operate DCVG at 1,250 to 1,500 mV pulse
amplitude.
(d) Record GPS sub-metre location of all
faults and right of way features.
(e) Record Test Post to remote earth pulse
amplitudes.
(f) Record using accurate GPS equipment,
see Figure 8, the epicentre location of all coating faults and
right of way features.
(g) Record Fault epicentre to remote earth
potentials.
(h) Observe the DCVG Corrosion characteristics
of all fault locations to determine if all pipeline coating faults
receiving enough CP current for protection.
(i) Determine approximate orientation of
the fault to identify pipe bottom problems. Also note distance
apart of coating faults to confirm if a field joint problem.
(j) In conjunction with the DCVG survey,
run a continuous EM soil resistivity survey see Figure 7, to identify
the most corrosive soil areas along the pipeline route. Measurements
to be taken every four metres.
(k) Based on the analysis of survey results
decisions should be made on the excavation and repair of any buried
pipeline coating faults.
(l) Every seven years carry out an in line
tool metal loss survey after thoroughly cleaning the pipeline
to remove as much as possible of the deposited wax and asphaltenes
which clogs up the sensors reducing their ability to detect metal
loss areas. The use of such tools should be as a check (audit)
that the corrosion mitigation techniques are being operated correctly
by limiting metal loss. It has to be recognised that metal loss
tools detect the symptom (metal loss) of the real problem which
is ineffective external corrosion mitigation technique (coating
and CP) operation and control that allows metal loss to occur.
10. WHAT TO
DO WITH
THE DAMAGED
FIELD JOINTS
ALREADY BURIED
It is my understanding that there are a significant
number of "damaged" field joints already buried. The
DCVG survey was presumably carried out to locate these. Comments
on the DCVG data so far examined indicate lack of attention to
detail in the survey and the survey report.
The failure of the buried coating will show
itself as follows:
1. Cracks and other types of fault (including
insect damage during setting of the epoxy) in the epoxy on the
steel.
2. Damage at the epoxy at the step from
original coating exposing potential crevices. This can be enhanced
by differences in the thermal expansion of the steel, polyethylene
and epoxy.
3. Ripping of the epoxy from the polyethylene
enhancing damage at the step a process enhanced by operating the
pipe at higher temperatures resulting in the pipeline moving in
the soil as it expands. (All pipelines move in the soil as they
accommodate changes in stress levels.)
4. Stone damage to the epoxy and polyethylene
coatings giving rise to bare steel at the faults.
5. Third party interference as the pipeline
ages.
The coating faults can arise from construction
malpractice or from general degradation or interference in the
surrounding soil and overall, will be faults that exist on the
pipeline right now or will gradually increase in number and size
as the system ages. An aging system will probably have more faults
on the bottom of the pipeline from stones such as the crushed
rock backfill, with also some on the sides resulting from banging
the pipe against rocks in the pipe trench wall as the pipe is
lowered into the trench or caused by thermal expansion in operation
rubbing the pipe against rocks in the trench wall.
What do you do with all the faulty joints so
far buried? Do you dig them all up and overcoat or completely
strip and replace with a proper coating? In reality the excavation
of every joint will be prohibitively expensive and cause tremendous
delay to the project. An alternative is to manage the problem
after ensuring that the problem is not being propagated on joints
currently being coated.
Managing involves identifying problem areas,
defining if they are protected then excavating and repairing according
to the following criteria.
1. Those coating faults in acidic or highly
corrosive soils that are big consumers of CP current and are most
likely to become unprotected (net current flow away) and where
metal loss will occur.
2. Those coating faults in corrosive soils
where the net current flow is away from the pipeline (weak CP)
and where metal loss can occur.
3. Those coating faults in mildly corrosive
soils that are protected (net current flow to) and big consumers
of CP current so by repair we can have the current available for
those faults not to be repaired.
4. Not repaired are those coating faults
of small severity that are nett recipients of CP current and are
in low corrosivity soil.
The survey techniques that should be used are
DC Voltage Gradient and EM Soil Resistivity operated by competent
surveyors with calibrated equipment with all data handled in a
competent software program such as "PRIMAS" (Pipeline
Rehabilitation and Integrity Management Analysis System). Because
of the sensitive nature of the pipeline and its known problems
the survey work should be carried out every two years for the
first six years and then reviewed to increase to three year periods
and after 12 years to four year intervals. A proper coating repair
specification needs to be set up and materials selected.
The main thrust of the monitoring and maintenance
is to identify areas where the CP is weak and then to improve
the coating and CP so by measuring net current flow we can be
assured that the pipeline is protected. The one weakness of this
is to be assured of control at any undercut of the polyethylene
so we can be certain that Stress Corrosion Cracking (SCC) does
not develop. Hence at any step crevice area repaired it is important
to conduct magnetic particle examination of the pipe surface looking
for any SCC particularly at type one and two faults, see above.
Based on experience, the ongoing monitoring
and maintenance of this pipeline should not be subcontracted.
Subcontracting this responsibility will be a disaster as we have
seen for some pipelines even in the UK where no equipment was
provided for the engineers to do their job properly. Everything
done on the cheap such as the reluctance of subcontract companies
to employ experienced engineers. Control must be based on BP retaining
the day to day activities although specialist companies will be
required for the surveying and repair work.
There also exists another problem. Initially
there is a lot of noise but after a couple of years complacency
sets in. If the monitoring and maintenance is done correctly there
should be no leaks and no other problems. The general attitude
common within the oil and gas industry develops in that we have
not had a problem so we don't have a problem (even a long term
problem) so we do not need to spend money doing all this surveying,
repairing etc so we can trim our budgets and add to profitability
by not doing all this "unnecessary" work. The pipeline
becomes "out of sight out of mind". As a practicing
Corrosion Engineer experience has taught me that in order to have
budgets to do your job properly you need a periodic leak to remind
management of the value of the work being done. In other words
BP has to set up a well defined, properly financed and implemented
commitment to an agreed monitoring excavation and repair program
for the next 40 years and this program must be staffed by competent,
experienced Corrosion Engineers. This is the price to pay for
not doing the job properly in the first place.
| 
