Documente Academic
Documente Profesional
Documente Cultură
119
In this issue:
CED Annual Meeting – Page 7
A tribute to Fred Palmer – Page 8
CEOCOR – Page 12
Technical Article – Page 15
CorrosionManagement | May/June 2014
E-mail: Contact@Rustrol.com
Central Fax: 905-333-4313
www.Rustrol.com
2
CORROSION MANAGEMENT
A JOURNAL OF THE INSTITUTE OF CORROSION May/June 2014 No.119
CONTENTS
Published on behalf of the Institute of Corrosion
Square One Advertising and Design Limited
Institute News
84 Queen Street, Sheffield S1 2DW,
United Kingdom. The President Writes 4
Publisher and Managing Editor
Debbie Hardwick
London Branch News 5-6
Tel: 0114 273 0132
Fax: 0114 270 0422 CED Annual Meeting 7-8
Email: debbie@squareone.co.uk
Technical Editor A tribute to Fred Palmer 9-10
Prof. Robert Akid CEng, FIMMM, FICorr
Email: robert.akid@manchester.ac.uk Aberdeen Branch Meeting March 2014 11
Design
Square One Advertising & Design Aberdeen Branch Meeting April 2014 12
www.squareone.co.uk
CEOCOR 13-15
Advertising Manager
Jonathan Phillips
Tel: 0114 273 0132
Fax: 0114 272 1713
Email: jonathan@squareone.co.uk
Technical Article Company News
Editorial copy date for July/August’14 The problem with meeting paint specifications 16-22
issue is: 11th July 2014
Subscriptions
UK £70.00
Europe £80.00
Outside Europe £90.00 airmail Company News
£80.00 surface mail
Enquiries and subscriptions to the Institute of Winn & Coales (Denso) Ltd 23
Corrosion at the address below
The Institute of Corrosion Mobility Oil & Gas Ltd 24
President
Trevor Osborne
Alfred Bagnalls & Sons Ltd 24
Immediate Past President
Dr. Bob Crundwell
Vice President Sustaining Members 25-32
Mr. John Fletcher
Hon. Secretary
Dr. Jane Lomas
The Newton Building, St George’s Avenue
ICATS Registered Companies 33-35
Northampton, NN2 6JB
Tel: 01604 893883 Fax: 01604 893878
Email: admin@icorr.org Diary and Branch Contacts 36
Website: www.icorr.org
All rights reserved Reproduction without written permission from the
Institute of Corrosion is prohibited. Views expressed in editorial text or
advertising copy are the opinions of the contributors/advertisers and
are not those of the Institute or the Publisher.
ISSN: 13 55 52 43
www.icorr.org
3
CorrosionManagement | May/June 2014
INSTITUTENEWS INSTITUTENEWS
4
INSTITUTENEWS INSTITUTENEWS
5
CorrosionManagement | May/June 2014
INSTITUTENEWS INSTITUTENEWS
7
CorrosionManagement | May/June 2014
INSTITUTENEWS INSTITUTENEWS
SUSTAINING MEMBERSHIP The annual rate from 1st July 2014 is £360.00 (plus VAT)
“GOLD” SUSTAINING
The annual rate from 1st July 2014 is £695.00 (plus VAT)
MEMBERSHIP
8
INSTITUTENEWS INSTITUTENEWS
9
CorrosionManagement | May/June 2014
TECHNICALARTICLE
INSTITUTENEWS INSTITUTENEWS
answers at that event, with such coating Fred’s sons – Council that the quiz subjects were very
luminaries as Derek Bayliss, Roger Hudson and Stephen, Geoffrey diverse and ICorr would not do well because
David Storey. – Well done Fred. and Michael they thought we were all just academics! I
Although Fred had come back out of knew Fred could not resist the IMF challenge
retirement at the age of 73, it was a most and being a winner he would not go along
memorable event that all the delegates who with a team that he had any doubt could only
attended will remember for his sharp and come second.
humorous technical responses to a diverse We arranged for three key members, Stewart
range of questions on all coating subjects. Crow an SPC retired coating inspector who had
Regrettably that 2003 event was the last time informed me that he had once come second
that Fred was able to attend any major, public on the BBC radio, ‘Brain of Britain’ and in the
ICorr event and although I was in contact with final he only lost by 2 points, (so he became
him about the 50th ICorr anniversary, which Richard, John and DHD the second cleverest man in the UK) Fred also
was held at the Thames Barrier in 2009, he was arranged for two members of the BP Sunbury
unable, and too ill to attend. staff to join, one was Graham Clayton, who
Fred was not only very sharp technically, had been runner up on TV’s Mastermind and
throughout his career, but an extremely had only lost again by 2 points in the final on
engaging personality, he was also a dedicated his specialist subject, although he had got full
winner. When the CCEJV with NACE was in its marks on his general knowledge questions,
early days, Fred organised a golf tournament they were all paid up ICorr members.
between ICorr and NACE and this ran annually With an ICorr team like that Fred knew that
for four years, the first three years were held in we would out class any team from IMF and
the UK in conjunction with the UK events and although it was the biggest ever IMF quiz night
after ICorr had won these first three, the NACE turnout with 17 four people teams present,
members said that their best golfers were John, Richard and Bill the ICorr team won the trophy and when it
not present in the UK and ICorr should go to was presented we offered to buy all of the IMF
America and play them on their own courses, contestants a drink from the ICorr winners.
where they were certain they would win. We However, not surprisingly only three of them,
therefore took up the challenge and Fred out of the 70 plus people attending, accepted
organised for a strong ICorr team to visit New our drinks offer and we have never been
Orleans to play the Americans at golf on their invited to defend our winner’s trophy, which
own ground, needless to say Fred’s ICorr team we still hold!!. - Well done Fred and his team.
won the day, then the joint venture broke up So yes, in the corrosion and coatings world our
and we have never played them at golf since, Fred was a true Icon and he was also a real
but at least we got to keep the trophy!! - Well winner. Fred you will really be missed, not only
done Fred and his team. by your family, but also by the many members
ICorr learned that IMF held an annual quiz
A second challenge that I remember well, night, which was extremely competitive with who shared your company in both technical
was from the Institute of Metal Finishing between 12 and 15 teams of four people all and social events, over many, many, years,
where as many members will remember, registering and attending and fighting for the Thank you, “Our Fred”.
the administration of ICorr was run from annual trophy, Fred was very keen to put in David H Deacon, a past colleague and
the IMF, Birmingham office. At one meeting an ICorr team, despite being told by the IMF friend of, “Our Fred”.
TO ADVERTISE IN
CORROSION MANAGEMENT
please contact
Jonathan Phillips or Debbie Hardwick at:
Square One, Tel: +44 (0)114 273 0132
Email: enquiries@squareone.co.uk
10
INSTITUTENEWS INSTITUTENEWS
11
CorrosionManagement | May/June 2014
INSTITUTENEWS INSTITUTENEWS
13
CorrosionManagement | May/June 2014
INSTITUTENEWS INSTITUTENEWS
Technical support economical situation covers the reduction uncertain to select the right technical
Support for implementation and using the of damages, the optimisation of reservoir parameters and the right materials for the
guidelines and recommendations for practical management and the reduction of supply risks inner surface of the drinking water reservoirs.
works. Service and qualification for using the for infrastructures and private properties. In The main objective of this group is the
actual technologies and applications. Qualified the field of water supply, the level of water preparation of a best practice guide for the
service and consulting for water plants, losses, the number of damages and the energy planning and rehabilitation of reservoirs with
building installations, selection of materials, management are performance indicators for respect to a general framework of technical
fittings and technologies for installation - and the quality of the network systems. The base and hygienic aspects
rehabilitation works. are network data and the condition data of - like the influence of the water quality on the
pipes together with local, technical, hygienic reservoir surface
Commission 1: Internal corrosion of pipe and economical influences are parameters for
materials and facilities maintenance and rehabilitation planning and - corrosion of steel in concrete
Management for construction and must be include in the corporate management. - effects of electric current on the concrete
rehabilitation strategies concerning water- The goal is the preparation of a decision characteristics
and gas supply systems support guideline for planning and organize
- analyze the processes of the damage
the operation-, inspection- and maintenance
The pipes and fittings of supply systems mechanism
activities and also the rehabilitation rate of
for water and gas have a limited service life - recommendataion for the operating and
transport - and supply pipes with the right
time and must be renewed (replacement by reservoir cleaning
materials and technologies under efficient and
constructions works or rehabilitation of pipes)
long term conditions. Strategies for reasonable and successful
at the right time. The network management
for the improvement of the technical and - definition of the water- and energy- balance use of chemical disinfectants in drinking
water systemsprevention of damages of
- evaluation and analyse the condition of the
construction products
pipe system
Chemical disinfection of drinking water with
- investigation of the local influence factors
chlorine, chlorine hypochlorite and chlorine
for pipelines
dioxide as a commonly used method and
- investigation of hydraulic and hygienic state-of-the-art against the biological load.
factors of the pipes Efficiency of disinfection depends not only on
- selection of priorities for rehabilitation works the type of disinfection agent and the dosage
concentration but also on the water quality
- selection of rehabilitation techniques for (pH, turbidity etc.), on structural and hygienic
pipelines conditions of the transport- and distribution
- recommendation of a efficient inspection-,
maintenance -, rehabilitation - and
investment strategy
Influencing of water quality by water
reservoirs
According to the European Standard EN 1508
there are different possible construction
systems for drinking water reservoirs in order
to obtain water hardness and construction
design. Operating, designing, construction
and expert companies are in many cases systems, the domestic installations and
on the operation conditions (temperature
and flow regime etc.). Disinfectants might
have unfavourable effects on the water
quality (formation of harmful THM, change
of odour and taste) and on the degradation
of installation materials or on localized
corrosion, which can result in serious damages
of metallic and non metallic pipes and
construction products. From that reasons the
use of disinfectants should be minimized and
all processes should be optimized what should
include water treatment and disinfection
processes, construction and rehabilitation
of water supply systems, improvement of
operation conditions and selection of suitable
installation materials.The main objective of this
working group is the collection of information
and data based on the available knowledge the
14
TECHNICALARTICLE TECHNICALARTICLE
15
CorrosionManagement | May/June 2014
TECHNICALARTICLE TECHNICALARTICLE
Clearly this is not the case for all yards Paint Scheme: Multi-coat system to a Nominal
Introduction and all owners, but there does seem to be Dry Film Thickness (NDFT) of 320 µm with up
Over recent years there have been interesting inconsistencies that the industry should work to two stripe coats.
developments in the way coatings and linings to overcome. In most cases this is interpreted as 2 x 160 µm
are specified that have unwittingly resulted In addition to these changes the advent of the and one stripe coat.
in a situation that can make it impossible to International Maritime Organisation (IMO)
meet a paint specification as currently written. Requirement: Minimum DFT: defined by the
Performance Standard for Protective Coatings 90:10 rule
The increase in demand for offshore structures (IMO PSPC) has seen an increased focus on
and the move by many shipyards into the protective coating for all areas of a vessel but Maximum DFT: in accordance with
offshore market has seen owner paint specifically for ballast tanks. In particular the manufacturer’s recommendations.
inspectors moving into shipyards, which has PSPC introduced the concept of a minimum In general, there is also wording in the PSPC
often resulted in much tighter subjective and Dry Film Thickness based on the 90:10 rule. that indicates that the coating is to be
objective requirements being placed on the This paper will focus on an example of the applied in accordance with manufacturer’s
yard during the build process. problems currently being faced in meeting recommendations.
Firstly, there has always been a problem specified Dry Film Thickness (DFT), which This may seem fairly straight forward, as may
in terms of meeting a paint specification is deemed the best understood and most a specification for the underwater hull such as:
because of the subjective nature of some of objective element of application. The paper
The scheme specified:
the inspection assessments such as a visual will show that even this most basic aspect
assessment of surface cleanliness, rust and mill of the paint specification is neither well 2 x Epoxy anti-corrosive 250 µm
scale removal, dust, and profile height when understood nor well specified. 1x Modified epoxy 100 µm
using a surface comparator. If the coating process cannot achieve the 3 x Self-polishing anti-fouling 390 µm
It is well known that the surface appearance specified DFT then to some extent the coating
system performance through-life will be Total scheme DFT 740 µm
of a steel substrate can change markedly
depending on the nature of the abrasive media compromised. How could any such reduction The fact that reality can vary dramatically
used, thereby making an assessment of surface in performance manifest itself? In many ways from the specification as can be seen in Figure
cleanliness particularly subjective. as exemplified by an increased chance of 1 below for this example of an underwater hull
corrosion for a low DFT and increased chance coating system specification:
Secondly, up to 2008, the shipbuilding boom
of cracking for a high DFT or an increase in
and the strong market for shipping has seen
time to service or solvent entrapment.
a considerable demand for ships and the
attractive charter rates had encouraged owners
to accept ships as quickly as possible. Many of The DFT
the Chinese yards and some of the second tier One of the key elements of any paint
South Korean yards were established in this specification is the Dry Film Thickness that is
boom time and therefore developed systems required to be achieved for individual coats
and processes in line with what owners were that make up the whole system and for the
then willing to accept in terms of vessels that system as a whole. The figures introduced into
were perhaps of a lesser standard than may the specification are usually taken from the
have been accepted if there was not such a Technical Data Sheet (TDS) for the product
boom. being used. The TDS normally provides a value
Now that market conditions have worsened, or a range of values, e.g. 125µm or 125-150µm
Figure 1. DFT and scheme
owners are more circumspect and have and usually refers to a single coat. Thus if a
become more cautious about what is and scheme is 2 x 125µm, then the Specified DFT The scheme as applied
what is not acceptable. As a result standards (SDFT) is 250µm. 2x Epoxy anti-corrosive 209 µm
have tightened. This has posed difficulties to
yards that grew up in the boom time (including 1x Modified epoxy 317 µm
many of the larger and new Chinese yards).
The Specification 3 x Self-polishing anti-fouling 213 µm
These yards have developed their systems First of all let us consider a typical specification
Total scheme DFT 739 µm
based on owner’s expectations during the as required by the IMO PSPC for Water ballast
boom period and are now struggling to meet tanks for the majority of areas (rather than If the scheme applied was assessed based on
the higher expectations. In the meantime repairs and erection joints): two inspections:
Owners are being frustrated by what is now Surface preparation: To Swedish Standard Sa • Surface cleanliness and
perceived as poor quality performance that 2.5 (Note: This is a requirement for mill scale • Final DFT
although acceptable in the boom time, is not and rust removal and does not cover the
so acceptable during the recession. specification of profile height.) Then the scheme would likely be accepted (if
16
TECHNICALARTICLE TECHNICALARTICLE
17
CorrosionManagement | May/June 2014
TECHNICALARTICLE TECHNICALARTICLE
18
TECHNICALARTICLE TECHNICALARTICLE
Let us consider then what this means for a • 66.6% of all values lie within the range process can only be improved by a change in
specification of 2 x 160 µm ±1σσ process technology. Thus improved process
It has first to be assumed that the 160 µm is • 95.4% of all values lie within the range capability is critical as many assignable causes
a nominal DFT; perhaps it should be the mean ±2σσ as possible must be identified and managed.
or the mode. • 99.75% of all values lie within the range The assessment of process stability/
The mean would require that the average ±3σσ performance is often made using control
reading taken in a given sample (for sample charts. There are a variety of control charts, the
The mode can be below or above the mean simplest of which is shown in Figure 5 below:
sizes see: SSPC PA2, ISO19840, IMO PSPC, [Ref depending on the readings taken and would
2, 3 and 4]) would be given by the arithmetic generate a “skewed” distribution with the
mean, while the mode is the value that occurs majority of readings being at the mode not
the most often of the set of readings taken. the mean (see Figure 3).
For example take the following set of numbers:
1,2,3,3,3,3,5,5,6,7,10,10
Sample size n = 12: Mean = 4.83: Mode = Figure 5. Simple control chart
3.00
The key elements of the control chart are the
The difference between mean and mode can setting of upper and lower tolerance limits
also be shown diagrammatically as shown in and a mean value (this would normally be the
Figures 3 and 4. specified or target value).
An example is as shown in Figure 5. Here
The Mean the specification limits or tolerances are set
Figure 4. Schematic illustrating both the Mode outside the capability limits. If this is the
and Mean case, the process can be said to be capable of
undertaking the work to tolerance. If however
the specifications limits are inside the upper
Process stability and and lower capability limits, then the process
control would not be capable of carrying out all work
to within the required tolerance. Obviously
Accuracy in all processes is critical in the further inside the capability limits the
shipbuilding. It has been stated that “The specification limits can be set, the greater
successful application of accuracy control the probability of being able to meet the
Figure 3. The mean for a normal distribution techniques to shipbuilding is quite fundamental specification. Just because the capability limits
In the normal distribution, the mean is the to achieving high levels of productivity [Ref 5] are within the specified tolerance, this does
value above which you would expect to Perfection is not often possible, especially not automatically mean that no readings will
find 50% of your DFT readings and below when it comes to applying coatings on board be obtained beyond the specification limits,
which you would expect to find 50% of your ships. This is as a result of many factors that just that there should be a low probability of
readings. DFT gauge software assumes you create an inherent variability in the process, this happening.
have a normal distribution when they provide making it difficult to control [Ref 6].
a statistical summary that often includes: Often however there is a lack of appreciation
The variability of any process comprises two of the process capability that results in the
• Average DFT elements: specification limits being set so as they fall
• Maximum • Assignable causes outside the capability limits.
• Minimum • Random variations It is important to note that it is normal in most
• Standard deviation processes for the tolerance limits to be set
Examples of these for coating application work around a mean value e.g. a typical value may
• Range may be: be 150 µ ± 50 µm.
The average is then simply the mean of the Assignable causes
readings taken, the maximum is the highest
reading taken, and the minimum is the lowest
• Use of the wrong or a worn spray gun tip Impact of minimum and
reading taken, while the range is the difference • Using the wrong pressure maximum values
between the maximum and the minimum. • The wrong stand-off distance The practical impact of the minimum and
The standard deviation (σ) is a measure • The addition of a cosmetic coat to the scheme maximum values needs to be carefully
of the spread of the curve. A low standard understood. To demonstrate this, a number
Random variation
deviation would indicate a tight process that of samples of coating inspections have been
can perform accurately and a low value for the • The size of the atomised particles taken to get an understanding of what is
range (the difference between the maximum • The workers physical capabilities generally practically achievable and these are
and the minimum values). A large standard presented in the following sections.
deviation would indicate a poorer control of • Wind gust and temperature changes
Control of DFT is dependent on many factors,
the application resulting in a larger range. • Available air pressure
such as worker skill, equipment, access
For a normal distribution the following While assignable causes can be addressed considerations and the complexity of the
approximate values are used: and managed, the inherent variability of the structure to be coated.
19
CorrosionManagement | May/June 2014
TECHNICALARTICLE TECHNICALARTICLE
Performance is likely best on a flat surface [Ref under application is more likely to occur for a ratio to the mean is considerably higher as a
8] while more complex surfaces will tend to number of reasons, such as: result.
increase the range. The size of the area to be • Weather/wind conditions This would imply that design complexity has
coated will also have an influence. a much greater influence on the variability of
• Worker skill
Data presented in Ref 1, indicated that, for the coating process rather than the number of
small areas, the achieved DFT readings on • Equipment capability/maintenance coats.
a nominal DFT of 85 µm for a single coat of • Roughness of the surface to be coated Furthermore, this would also imply that to
inorganic zinc silicate would give a range of maximise the probability of a good coating
This would indicate a process that is not
240 µm with a minimum of about 20 µm and application, both the design complexity and
well controlled which could easily result in
maximum of about 260 µm. the number of coats should be minimised.
excessive over application of coatings, which
On larger areas, work carried out on the can penalise a yard in a number of ways: Notwithstanding, the simplification of design
underwater hull of US Navy carriers gave the would offer the greater benefits.
• Increased cost of paint and thinners/cleaners
following results: In practice, the problem is aggravated further
• Increased application time
USS Nimitz: because not all the coating work in one
• Increased curing/drying time location will be carried out by the same team
Specified scheme DFT 680 µm
• Increased emissions and the skill/ability and equipment, as well as
Mean: 720 µm local conditions may vary to a considerable
• Increased waste
Standard deviation: 220 µm degree.
• Delay to build schedule
Range approximately 1000 µm Of course the process capability is also likely
• Increase in utilisation of facilities to change for different ship sizes, with smaller
Ratio of standard deviation
In more complex areas, research by the authors vessels providing more complex/tighter
to the mean 0.44
has shown these results for Cargo holds gave structures. The authors suggest that perhaps
Process capability to 3 σ σ 60 – 1380 µm the Compensated Gross Tonnage Coefficients
a mean of: 0.19 for the ratio of standard
USS Lincoln deviation to the mean. [Ref 8] could be utilised to help establish the
complexity of different ship types and sizes.
Specified scheme DFT 680 µm Specified DFT: 250 µm
Mean 850 µm Average DFT: 649 µm
Standard deviation 250 µm Standard Deviation: 133 µm
Impact on a coating
Range approximately 1000 µm Process capability to 3 σ σ 250 - 1048 µm
scheme
Let us consider a specification of 2 x 160 µm
Ratio of standard deviation Thus while the process for cargo holds does as required by the IMO PSPC and as shown on
to the mean 0.29 show a greater variability (higher standard most paint supplier data sheets. In this case
Process capability to 3 σ σ 100 – 1600 µm deviation) than that for the outside shell, the the value on the TDS is the “nominal value”
ratio of mean to standard deviation is about which the authors have interpreted as a target
Thus assessment of the US Navy figures cannot the same (0.18 to 0.19).
be made without knowing the specification for value or the mean/average.
the system that was applied (e.g. how many The reason for this is relatively simple. The Maximum DFT
coats of paint). Also these coating projects outer hull scheme typically comprises 4 or
more coats of paints as compared to 2 coats Good practice from paint company guidelines
were in a refit scenario rather than a new build
in the cargo hold. The variability in the DFT of would mean that the maximum DFT applied
scenario.
each coat is additive, thus the more coats of should be x2 the specified DFT for each coat
Figures from work carried out by the authors paint applied the greater the variability. and for the total scheme. These would give a
for the outer hull for a new build commercial maximum scheme of 2 x 320 µm.
ship for the same ratio gave a mean of 0.18 Thus the more steps in a process (i.e. the more
coats of paint in the scheme) the greater the Minimum DFT
for the ratio of standard deviation to the mean
(also known as the coefficient of variation): variability that should be expected, irrespective While applying the 90:10 rule or the 80:20 rule
of the complexity of the surface to be coated. would give minimum values of:
Specified DFT: 610 µm
Results for ballast tanks, which are also 90:10 rule - 2 x 144 µm or 288 µm total:
Average DFT: 990 µm generally two coat schemes and are more
80:20 rule – 2 x 128 µm or a 256 µm total.
Standard Deviation: 170 µm complex areas, should therefore offer a better
comparison to the cargo holds. The standard deviation for water ballast tank
Process capability to 3 σσ 480 – 1500 µm
(WBT) application has been calculated at 162
Thus even on the relatively uncomplicated area These give a mean of 0.26 for the ratio of
µm. Thus if the minimum acceptable value is
of the underwater hull there is a considerable standard deviation to the mean
288 µm as per the IMO PSPC, then 3 standard
range of the achieved the required DFT with Specified DFT: 320 µm deviations would provide a mean of 774 µm
the ratio of standard deviation to the mean Average DFT: 602 µm [ given by: 288 + 3(162) µm] and the maximum
ranging from a relatively good 0.11 to a value that could be expected would be 1260
relatively poor 0.44 for the US Navy. In simple Standard Deviation: 162 µm µm [given by 774 + 3(162) µm].
terms, the closer the standard deviation value Process capability to 3 σ σ 116 – 1088 µm The mean that is likely to be achieved will itself
is to the mean value (the higher the ratio)
Considering ballast tanks despite having only exceed the recommended guideline of most
the greater the spread of the curve. Hence,
two coats of paint (as is the case for cargo paint suppliers which is set at x2 the specified
as the process is not well controlled, over or
holds) both the standard deviation and the DFT (640 µm in this case) and also surpasses
20
TECHNICALARTICLE TECHNICALARTICLE
the x3 value in ISO 12904 [Ref 9]. Mode 564.5 µm how the data from a DFT gauge is presented.
To achieve the required specification: Breakdown of readings in bands of microns. However all is not lost. A little more statistics
• Minimum 288 µm Given that the recommended practice can come to our aid. Basically it requires a
review of how we collect DFT readings.
• Maximum 640 µm 200- 400- 600- 800-
<200 If instead of taking individual readings, the
Then the standard deviation would have to be 400 600 800 1000 requirements of SSPC PA 2 are considered,
58.7 µm or about 36% of that being achieved then there is a requirement to take 3 readings
in the field based on the authors research. 0 36 286 158 67
for a spot measurement. This “grouping” of
The problem of excessive DFT is compounded 1000- 1200- 1400- 1600- readings will tend to result in the data being
even further. If during inspection areas of low >1800 forced to form a normal distribution (this
1200 1400 1600 1800
DFT are identified e.g. an area of 250 µm occurs as a result of the Central Limit Theory –
thickness is identified. This would then be 15 4 0 0 0 CLT) and is why the SSPC PA2 method results
marked for touch up. If this is done by airless in a requirement for fewer DFT readings. If you
spray, it will not be brought up to 288 µm or do not group the readings in that way you
320 µm but likely by an additional 160 µm to (maximum of 2 times the DFT specified) require more readings to invoke the CLT to
410 µm. If the coating is applied by brush an would give a maximum of 640 µm then 193 generate a normal distribution. This is the basis
additional 80 µm could be added. Therefore (34%) of the readings taken exceeded the of Control Chart Theory [Ref 10].
any application of “touch up” coats to achieve maximum, while very few readings were below Therefore if in doubt, the more readings that
the minimum DFT is likely to increase the the minimum, this despite the mean and the can be collected the better the overall picture
mean DFT and push the scheme further out of mode being below the 640 µm maximum. you will obtain of the coating of a particular
the recommended guidelines provided. Thus the actual distribution of the DFT readings area, however, where time is limited, a smaller
will be greater than the expected specified set of data if collected correctly can give you a
Practical distribution values. In particular the actual distribution reasonable overview.
will tend to be skewed towards higher DFT
In practice however, the applied DFT data does
not result in a normal distribution but rather
values and this is aggravated by the use of a
minimum DFT rule.
Conclusions
a skewed distribution as indicated in Figure The variability of the coating process, the
4. An actual set of data from a ballast tank is In a nutshell, as soon as a minimum rule is
number of coats of paint, the complexity of
presented in Figure 6 below. introduced then the mean DFT achieved will
the surface and the use of a minimum DFT
result in considerably higher than the specified
This ballast tank coating was specified rule, result in a mean DFT far greater than that
DFT. This combined with a space with complex
according to IMO PSPC and thus should have specified. The shift in the mean can be close
geometry results in the mean DFT being close
a nominal DFT of 320 µm. Analysis of the data to or exceed the x2 DFT maximum value that
to or greater than the x2 DFT maximum cited
gave the following results: paint companies generally recommend as
in paint company guidelines.
good practice.
Total number of readings 566
Clearly, it is easier for a shipyard to apply more
Minimum DFT: 272 µm DFT Gauge readings paint to make up for low DFT than to remove
Maximum DFT 1326 µm DFT gauges are set up to assume that the paint in the event of excessive DFT. While is
Range 1100 µm readings being collected are represented by a important to achieve a certain minimum
normal distribution and provide the statistical DFT for the coating to perform, there is a real
Mean 611 µm results associated with such a distribution. danger that the use of minimum DFT rules will
Standard deviation not relevant as However, as demonstrated, the DFT readings lead to higher than expected DFT readings and
this is not a normal for a ship tend to fall into a skewed distribution, this can also lead to performance drop off or
distribution which potentially would raise concern about even failure of the coatings.
The problem for the shipyard is that the
application of this extra paint not only
increases the man-hours and cost of coating
the area but also extends over-coating and
drying times as well as increasing VOC
emissions.
For the owner the problem faced is that
the DFT provided may be in excess of that
recommended by the paint supplier as good
practice and the impact (if any) of the excessive
DFT on the performance of the coating may
not be well understood.
The reality is therefore that unless current
coating application techniques are improved,
then the range of readings that will be
obtained in practice for any given specification
will depend on several factors such as;
Figure 6. Example of a skewed distribution in ballast tank DFT readings
21
CorrosionManagement | May/June 2014
TECHNICALARTICLE TECHNICALARTICLE
• The number of coats It is recommended that the TDS should simply The range to assure the required in-service
• The structural design complexity contain a maximum and minimum value for performance (assuming drying/cure times and
DFT rather than some individual ambiguous over-coat intervals etc. are adhered to. The
• The skill of the applicator value. This would leave each paint supplier problem with ships in particular is if 1 or 2%
• The condition of the equipment used. to determine the DFT range over which their of the readings are accepted to be outside the
products will provide the claimed performance. range provided the total area equating to such
The TDS’s provided by paint suppliers need
Of course this would add some complications, a small percentage can still be quite large. For
to be very specific as to the DFT value that
in that drying times, cure times and other data example 5% out of specification (equating to
is being quoted. It is likely to be preferable
that may be affected by DFT (such as time to ±2σ for a normal distribution) of 30,000 m2
to simply quote a range from the minimum
service) will need to reflect the range that is of cargo hold would mean that 1,500m2 would
to the maximum acceptable for each coat,
provided. be outside the range. This is not an insignificant
rather than some vague value that could be
Those developing coating specifications area.
interpreted as a minimum, a mean or some
other measure such as nominal. should also consider the range of DFT as In the final analysis it is clear, that variability
more important than a specific DFT value should be minimised and technical
It is the author’s contention that paint suppliers
(nominal, mean or otherwise) the range would documentation needs to better reflect the
would be prudent to test their products
then reflect any minimum/maximum values in service practicalities of process variability
performance at expected DFT’s that may be
recommended by the paint supplier. and capability so that applicators and end
achieved in the field and provide data on the
The challenge will be to achieve a range that is users understand the limitations that may be
TDS for the elevated thickness expected.
achievable by the application process. imposed on them and how if at all coating
Thus the IMO PSPC specification may be performance may be affected by the real
better written as a range for example; 288 µm life conditions, rather than the laboratory
– 640 µm, this would imply a mean of about
464µm.
Other sources of conditions which are used to generate the
information for the TDS.
Even with a range consideration must be given
variability
It would be nice to be able to conclude that the
to the structural complexity of some parts of
problem stops there. However there are other
References
a ship. The range must be achievable with the
sources of variability that also have an impact Francis RA; Thickness of Marine Coatings:
current process capabilities. It is suggested
on results and even these can be significant Measurement, standards and problems; RINA
that the maximum value would need to be at
when measuring the DFT achieved. These can Conference on Marine Coatings, London April
least x3 the nominal value of 320 µm, thus
be summarised as follows: 2013
giving a range of 288 – 960 µm. The latter
would imply a mean of 624µm (assuming a • Sampling regime: How many samples are SSPC PA2: Procedure for Determining
normal distribution). taken? Were they are taken; does everyone Conformance to Dry Coating Thickness
follow the same sampling regime? If not there Requirements. www.sspc.org.
Of course, this is only used as an example. It
is known to the authors that for ballast tank can be significant differences in the results as ISO19840 : Paints and varnishes - Corrosion
coating in particular, the maximum limits set shown in Figure 2. protection of steel structures by protective paint
in South Korean yards are 2000 µm. While this • Accuracy of DFT gauge/repeatability of systems - Measurement of, and acceptance
seems high, it is clearly an attempt to push readings: This can also result in a different criteria for, the thickness of dry films on rough
the upper specification limit well beyond the result even for people using the same sampling surfaces. www.iso.org
capability limit of the application process to regime, an observation highlighted by Ault IMO PSPC
ensure that there is never any need for re-work [Ref 11].
in the form of removing excessively thick paint. Dr R Vaughan, Productivity in shipbuilding. Trans
• Operator capability: The skill of the operator NECIES Volume 100 1983-84
The introduction and use of minimum value is critical and can also affect the readings taken.
rules in a specification (such as 90:10 or C Jefferson, R Kattan. Accuracy Control Training
• Location of readings and substrate: Are Manuals, British Shipbuilders 1987
80:20), will tend to excessively increase the
the readings taken across a representative
mean DFT excessively values. The outcome Coating of ships the Design Challenge;
sample of the structure (as per IMO PSPC
will tend to move the whole distribution to Broderick, Kattan, Wright RINA Conference
requirements). Is the substrate condition
the right. The process variables tend to also 2010
known to enable allowances for the profile to
lead to a non-normal distribution and result OECD Council working party on shipbuilding,
be made?
in a skewed distribution. The most significant Compensated Gross Ton (CGT) system
of those variables is the touch up process that • It is a well-known adage that the more you
takes the lower DFT values and moves them inspect the more you will find wrong and so in ISO 12904
significantly to the right by the addition of part the inspection process needs to be well Statistical Process Control – Grant and
more paint through brush or spray application). defined to assure that it is not punitive, while Leavenworth, McGraw Hill.
achieving the required assurance.
Coating inspection data reproducibility, J Peter
• Given a range of DFT’s to achieve, the
Implications for Paint obvious question will be: what if some few
Ault, Megarust 2008.
22
COMPANYNEWS COMPANYNEWS
Email: info@mobilityoilandgas.com
Corrosion Control in the Oil and Gas Industry
London, 24-27 June
Aberdeen, 19-22 August
Houston, 16-19 September Web: www.mobilityoilandgas.com
24
SUSTAININGMEMBERS SUSTAININGMEMBERS
CATHELCO
CATHODIC PROTECTION Marine House, Dunston Road,
CONSULTANCY SERVICES Chesterfield S41 8NY
Tel: +44 (0) 1246 457900 Fax: +44 (0) 1246 457901
BEASY Email: sales@cathelco.com www.cathelco.com
Ashurst Lodge, Ashurst,
Southampton, Hants, SO40 7AA
Tel. 02380 293223 Fax: 02380 292853 CORROCELL LIMITED
e: t.froome@beasy.com www.beasy.com 17 South Meade, Maghull, Liverpool L31 8EG
Tel: 0151 249 8461 Fax: 087087 92585
Email: dp@corrocell.co.uk
CORROSION CONTROL
3 Ivy Court, Acton Trussell, Staffordshire ST17 0SN
Tel: 01785 711560 Fax: 01785 711561
CORROSION CONTROL INCORPORATED
Email: brianwyatt@controlcorrosion.co.uk 494 Fairplay Street, Rutledge,
www.controlcorrosion.co.uk
Georgia 30663, USA
Corrosion Control
Company AB
Box 72, 268 03 Billeberga,
Sweden
Your specialist when it comes to Cathodic Protection
Systems for Power Plants, Tank Farms, Pipelines,
Concrete Structures, Harbours, Jetties and Refineries.
Tel: +46 418 411 900 Fax: +46 418 411 935
Email: info@3ccc.se Website: www.3ccc.se
25
CorrosionManagement | May/June 2014
SUSTAININGMEMBERS SUSTAININGMEMBERS
26
SUSTAININGMEMBERS SUSTAININGMEMBERS
27
CorrosionManagement | May/June 2014
SUSTAININGMEMBERS SUSTAININGMEMBERS
NUSTEEL STRUCTURES
Lymane, Hythe, Kent CT21 4LR
Email: simon.slinn@nusteelstructures.com
Website: www.nusteelstructures.com
28
SUSTAININGMEMBERS SUSTAININGMEMBERS
DNV
Cromarty House, 67-72 Regent Quay, Aberdeen AB11 5AR
Tel: 01224 335000 Fax: 01224 593311
Reader Enquiry: CM013 Email: yee.chin.tang@dnv.com Website: www.dnv.com
29
CorrosionManagement | May/June 2014
SUSTAININGMEMBERS SUSTAININGMEMBERS
INTERTEK SPECIFIERS
PRODUCTION & INTEGRITY ASSURANCE (P&IA)
Bainbridge House, 86-90 London Rd, Manchester M1 2PW FORTH ESTUARY TRANSPORT AUTHORITY
Tel: 0161 933 4000 Fax: 0161 933 4001
Website: www.intertek.com/corrosion Forth Estuary Transport Authority
Email: production.assurance@intertek.com Forth Road Bridge
Administration Office,
MOTT MACDONALD South Queensferry, EH30 9SF
Tel: 0131 319 1699 Fax: 0131 319 1903
Materials & Corrosion Engineering
Email: customer.care@forthroadbridge.org
Spring Bank House, 33 Stamford Street
Altrincham, Cheshire WA14 1ES
Tel: 0161 926 4000 Fax: 0161 926 4103 SSE Ltd
Email: paul.lambert@mottmac.com www.mottmac.com
Grampian House, 200 Dunkeld Road, Perth PH1 3GH
Tel: 01738 456000 Fax: 01738 456647
Paint Inspection Limited
Milton House, 7 High Street, Fareham PO16 7AN
Tel: 0845 4638680
Email: ian@paint-inspection.co.uk
www.paint-inspection.co.uk SUPPLIERS COATINGS
Plant Integrity Management Ltd BREWERS PROTECTIVE COATINGS
1st Floor Office, Woodburn House, Reform Rd, Maidenhead, Berkshire SL6 8DA
Woodburn Road, Blackburn AB21 0RX
Tel : 01628 784964 Fax : 01628 672578
Tel: 01224 798870 www.www.pim-ltd.com
E-mail : info@brewersprotectivecoatings.co.uk
Email: maraneda@pim-ltd.com
www.brewersprotectivecoatings.co.uk
30
SUSTAININGMEMBERS SUSTAININGMEMBERS
31
CorrosionManagement | May/June 2014
SUSTAININGMEMBERS SUSTAININGMEMBERS
32
ICATSREGISTEREDCOMPANIES ICATSREGISTEREDCOMPANIES
33
CorrosionManagement | May/June 2014
ICATSREGISTEREDCOMPANIES ICATSREGISTEREDCOMPANIES
International Energy Services Ltd Northern Protective Shutdown Maintenance Services Ltd
94 Awolowo, Ikoyi, Lagos State, Nigeria 16 High Reach, Fairfield Industrial Estate, Bill Quay, Kingsnorth Industrial, Hoo, Rochester,
T: 014615636 Gateshead, Tyne & Wear, NE10 0UR Kent, ME3 9ND
T: 0191 438 5555 T: 01634 256969
Interserve Industrial
Unit 2, Olympic Park, Poole Hall Road Nusteel Structures Solent Protective Coatings Ltd
Ellesmere Port, Cheshire, CH66 1ST Lympne Industrial Estate, Lympne, Hythe, Tredegar Wharf, Marine Parade
T: 0151 3737660 Kent, CT21 4LR Southampton, Hants, SO14 5JF
T: 01303 268112 T: 02380 221480
34
ICATSREGISTEREDCOMPANIES ICATSREGISTEREDCOMPANIES
Coastground Ltd
Morton Peto Road, Gapton Hall Industrial , Matthew James Services
Torishima Service Solutions Europe Ltd
Great Yarmouth, Norfolk, NR31 0LT Unit 4, Shibdon Business, Cowen Road
Sunnyside Works Gartsherrie Road
T: 01493 650455 Blaydon, Newcastle-Upon-Tyne, NE21 5TX
Coatbridge ML5 2DJ
T: 0191 414 5700
T: 0123642390
Corroless Eastern Ltd
Greens Road, Greens Industrial Estate, Dereham, Moore Steel Developments Ltd Transvac Systems Ltd
Norfolk NR20 3TG Station Road, Thorney, Peterborough PE6 0QE Monsal House, 1 Bramble way
T: 01362 691484 T: 01733 270729 Alfreton, Derbyshire, DE55 4RH
T: 01773 831100
Darcy Spillcare Manufacture Optimal Rail Ltd
Brook House, Larkfield Trading Estate, New Hythe Lane, Unit 5, Moorgate Crofts Business Centre
Larkfield, Kent ME20 6GN Alma Road, Rotherham
T: 01622 715100 S60 2DH
T: 01709 331153 TO ADVERTISE HERE
D F Coatings Ltd
Unit 17, Willments Ind. Estate, Hazel Road, Woolston
CONTACT
Paint Inspection Ltd
Southampton SO19 7HS
T: 0238 044 5634
Milton House, 7 High Street, SQUARE ONE
Fareham PO16 7AN
T: 0845 4638680 Tel: +44 (0)114 273 0132
E G Lewis & Company Ltd
Suite 5, 3 Shawcross Industrial Estate,
Ackworth Road, Portsmouth PO3 5JP
Parks Fabrication Ltd Email: enquiries@
Park Farm, Holme-upon-Spalding-Moor,
T: 01792 323288
York, YO43 4AG squareone.co.uk
T: 01430 861628
35
INSTITUTE
CORROSION EVENTS May/June 2014 No. 119
London Branch publish a monthly Newsletter; to be included on the circulation list please contact Sarah Vasey sarah.vasey@akzonobel.com