03 Hull Roughness

HULL ROUGHNESS
The Ship Hull

Fouling Penalty
(2003)
Dr. R L Townsin, Consultant
‘Garfield’, Whitley Road, Benton
Newcastle upon Tyne, NE12 8BP, UK.
The Ship Hull Fouling Penalty
‘Garfield’, Whitley Road, Benton, Newcastle upon Tyne, NE12 8BP, UK.
Abstract
The ship resistance penalties of slime, shell and weed are discussed in turn.
Methods to measure the hard paint roughness of antifoulings are recapitulated. The determination of a
satisfactory roughness parameter from correlations with measured roughness functions is described. This in
turn, allows a relationship between ship added friction and roughness height to be found. This recapitulation
allows a consideration of using the same route for a surface with filamentous fouling.
Consideration is given to low surface energy coatings and their roughness idiosyncrasies.
Finally, the determination of economic penalties is discussed, both for a particular ship and globally.
Keywords: antifouling development, hull roughness, fouling drag, ablative coatings, non-stick antifouling.
1. Introduction
The penalty referred to in the title, is ship speed loss at constant power, or, power increase at constant speed,
or, consequentially, an economic penalty due to increased fuel consumption and scheduling penalties and
other delays. Penalties arise from an unsatisfactory underwater outer bottom condition, which includes
propeller blade surfaces.
Almost all vessels have an antifouling paint coating over the underwater hull. Generally, propeller blade
surfaces are of polished metal eg. manganese bronze, and will have no antifouling provision.
As far as the hull coating is concerned, a number of problems can arise:
• A new antifouled surface may be hydrodynamically rough, usually as a result of poor paint application
management eg. drips, runs, sagging, overspray, and grit inclusion.
• The coating may become rougher in service due to paint system partial failures and mechanical contact
damage.
• The antifouling provision may be inadequate over time, resulting in slime development, and then weed
and shell growth, variously distributed over the hull.
Propeller fouling will not be dealt with here, partly because it is normally a minor part of the ship fouling
penalty, but also because there is a relatively inexpensive remedy viz. underwater blade cleaning, and even
polishing, by specialist divers. It can be noted however, that coatings can survive on propeller blades and
have been found to have some antifouling benefit. As far as propeller blade surface roughness penalties are
concerned, these may be estimated from the nomograms published in Townsin et al. (1985).
The ship hull fouling in the title, then, consists in the three categories, demotically: slime, weed and shell.
Comment will be made on the ship speed and power performance penalties arising from each in turn. After
that, some comments on hydrodynamic research issues of roughness and fouling would be appropriate.
A survey of the economic penalties will complete the review.
With the exception of Papers written exclusively by International Marine Coatings, the Papers presented do not necessarily reflect the
opinion of International Marine Coatings and, whilst every effort has been made to ensure that the information in this publication is
accurate, International Marine Coatings makes no representation or warranty, expressed or implied, as to the accuracy, completeness
or correctness of such information. Please note that the information does not necessarily stand on its own and is not intended to be
relied upon in specific circumstances. To the extent permitted by law, International Marine Coatings accepts no responsibility
whatsoever for any loss, damage or other liability arising from any use of this publication or reliance upon the information which it
contains.
© Akzo Nobel, 2003
11
The deleterious effects of fouling on ship performance have been feared and recorded from the earliest times.
Between 1862 and 1904 there were 18 papers on corrosion and fouling issues read to the Institution of Naval
Architects in London. The two issues were often seen as one problem in those days, as iron was replacing
wood as the shipbuilding material. Copper sheathing had protected wooden hulls from the depredations of
the teredo worm, and, as a by-product, the copper kept fouling at bay. The efficacy of copper as an anti-
foulant, led to attempts to clad iron ships with copper. The British Admiralty had to call upon the
distinguished chemist, Sir Humphrey Davy, to help with the inevitable accelerated corrosion of the iron and
copper combination. An attempted solution at the time was to clad the iron ship hull with wood and fix the
copper sheets to the wood, without contact with the iron: clearly, fouling was a serious performance problem.
The effects of fouling at these times is well illustrated by a quotation from one of the INA papers referred to
earlier, (Lewes 1889). He says “of some protective and antifouling compositions in use by the Navy, it is no
exaggeration to say that, as far as speed is concerned, one half of our fleet would be useless before one year
had elapsed, from the accumulation of rust, weed and shell”.
As well as copper sheathing, many entrepreneurs were patenting and experimenting with various
compositions with which to coat the outer bottoms of steel and iron ships. Indeed, the first record of an
antifouling coating was in a British Patent of William Beale, as far back as 1625. As Volker Bertram (2000)
quotes in his helpful WEGEMT lecture “Until 1865, more than 300 such ‘patent paints’ were registered. All of
them were quite ineffective”. Some exotic brews resulted from these attempts. One such was of fish scales
pounded up with red lead. What transient antifouling benefit there was, resulted more from the red lead than
the fish scales.
The last of the 18 INA papers referred to earlier, was by Holzapfel,(1904). Holzapfel reports on test patches,
keel to light waterline, of about 100 different compositions, which he applied to some 80 boats sailing from
Genoa, principally in the Mediterranean trade. He also tested composition-coated plates immersed in Genoa
harbour. He distinguished two types of antifouling composition current at the time, which he called
‘varnishes’ and ‘greases’. All his experiments were with ‘varnishes’. Holzapfel found that the more successful
of the ‘varnishes’ were those with substantial copper and mercury content; but, he writes, a successful
composition must also be one “capable of being gradually dissolved in seawater” which he called ‘primary
disintegration’. The ‘greases’, which were much softer, and gave rise to greater surface friction and fuel
consumption (we would now say they were ‘rougher’), were supposed to shed the fouling by what he called
‘secondary disintegration’ under fluid frictional shear. Professor Lewes used the word ‘exfoliation’ in his
earlier support of ‘greases’. The industry went down the ‘hard varnishes’ road for the next 70 years and
exfoliating greases were not pursued much further. However, in Holzapfel’s paper we may read the
beginnings of ideas about what we now call ‘ablative coatings’.
1
The ‘hard varnishes’ road led eventually to inter docking periods of 2–2 ⁄2 years, at best. It became evident
that the leach rate of toxins was falling off with time, so that new coatings had an over-rich leach rate,
whereas an older coating had an ineffective leach rate, even though biocide remained. Holzapfel’s ‘primary
disintegration’ was not being achieved.
All this changed with the British Patents of Milne and especially Milne and Hails (1971), which led to the so-
called self-polishing copolymers loaded with tributyltin as the biocide. The chemistry of these new products
provided an ablation rate which ensured a constant rate of leach of biocide. The effective life of these
products depended on the coating thickness in relation to the ablation rate, and, in practice, could ensure a
foul free hull for up to 5 years, which coincides with the Classification Societies requirements for inter docking
survey periods.
With the advent of these TBT SPCs, suddenly, fouling became yesterday’s problem and interest then began
to centre upon the fluid friction penalties due to the roughness of outer bottom coatings. The roughness of
new ship coatings had always been of interest because of contractual speed and power trials at hand-over
of a new ship. The renowned Lucy Ashton trials, in the Gare Loch, in Scotland, were conducted in the late
1940s and early 1950s with this in mind, Denny (1951). But now, with the prospect of foul free hulls, the
through life roughness of antifouled hulls became of interest to ship operators. The Ship Performance Group
at Newcastle University, among others, pursued this work eg. Townsin et al. (1981). The Group’s formulation
for calculating the resistance penalty for a ship with a moderate and measurable roughness was
subsequently adopted by the 19th International Towing Tank Conference, Madrid, ITTC (1990).
Whilst the ablation of these products and the consequent constant biocide leach rate was their prime raison
d’etre, it was also noted that any initial roughness due to application was smoothed out in service. The name
‘self-polishing’ for these products was therefore applied by the marine coatings industry to indicate
smoothing properties, although, whilst the paint itself became smoother, the hull, overall, often became
rougher due to surface damage. The added resistance due to paint surface damage was a problem
recognised by Holzapfel.
12
The success of tributyltin was not to last. Marine biologists, worldwide, were able to assemble evidence of
the effects of TBT on coastal marine life. The poisoning of oyster beds and imposex among dog whelks were
among the effects observed. The passage of vessels with TBT antifouling in coastal waters was seen as the
cause, but the prime cause may have been the flushing of dry docks where the coatings were applied or
removed and where no filters were used. As is well known, tributyltin is now banned with various conditions
and deadlines.
The marine coatings industry, at the present time, does not wish to lose the benefit of an ablative matrix
containing a biocide. The chemistry is being reconstructed to accommodate different biocides, and copper
returns as the major present candidate, supplemented with booster biocides. Marine coatings chemists are
busy trying to improve these so called tin free, self-polishing copolymers to match the effectiveness of what
is being banned. There is a little way yet to go before the confidence of ship operators returns, but another
antifouling candidate is already being applied to the bottoms of ships viz. the low surface energy, or non-
stick, or foul release coating. Perhaps this non-biocidal paint is the way ahead and the prophetic banning of
copper biocides on some coastal stretches of Scandinavia may be noted. It is interesting that work on these
non toxic coatings was underway as TBT SPC was being developed eg. Callow et al. (1985). The success of
the latter partially obscured the promise of the former.
2. Slime
The fluid friction drag of slime is the least well understood of the three demotic categories of fouling referred
to earlier. A slime layer forms quickly, constituting the beginning of fouling growth. It is sometimes the case
that an antifouling does not inhibit a slime film, yet keeps weed and shell at bay. At present, this seems to
be the case for foul release coatings. Since slime is ‘rootless’, it is easily cleaned off, yet it resists removal
even by the relatively high frictional shear rate of faster vessels. The important question is therefore, what
drag penalty arises from the presence of a slime film?
The classic experiments on the Lucy Ashton, referred to earlier, consisted of a refurbished paddle steamer
hull, with the paddles removed, propelled on the calm waters of the Gare Loch, off the river Clyde, Scotland,
by jet engines mounted on the paddle sponsons. The engine thrust in air was accurately measured, as was
the speed on the measured mile. All manner of surface conditions were explored for their drag, including the
effects of hull roughness and of slime. The hull was allowed to foul for 40 days, on a coating of bituminous
aluminium. Only slime was present. Over the speed range of 5 to 15 knots, the frictional resistance increased
by 5%, ie. 0.125% per day. Conn et al. (1953).
Watanabe et al. (1969) reported an 8 to 14% increase in (frictional) resistance due to slime.
Lewthwaite et al. (1985) reported the 600 day development of fouling on the hull of a 23m fleet tender,
operating around the South Coast of England. The hull had a non-polishing antifouling. The local frictional
resistance coefficient was measured by a total and static head wedge probe mounted near to amidships.
After 240 days operation, divers reported a thin slime coverage, too thin to measure but detectable by touch.
At this time, a 25% increase in the local frictional resistance was measured. At 600 days a thick 1mm slime
film had developed, with some shell and extensive weed above the bilge keels. The frictional resistance had
increased locally by 80%.
An interesting set of full scale trials on a frigate was reported by Bohlander, (1991). The frigate had an
organotin and cuprous oxide antifouling coating. After 22 months out of dock, it had no weed but a very small
amount of calcareous fouling. However, it did have a mature slime film. Speed and power trials were
conducted. The ship was then drydocked, the slime was cleaned off and further speed and power trials
undertaken. A measured 8 to 18% decrease in total propulsive power was attributed to the removal of the
slime film, with the highest percentage being near the highest speed (surprisingly).
The above four sets of full scale trials are sufficient to show that slime can have a substantial deleterious
effect upon ship performance and many laboratory tests have corroborated this conclusion eg. Loeb et al.
(1984), in which a 5 to 8% increase in resistance for a flat plane at 40 knots, was extrapolated from disc tests.
Unfortunately, there is no way of predicting, for a particular ship, trade route and coating, how the slime will
develop. Additionally, at present, there is no way of measuring and characterising slime, which can be
correlated with its drag. The ship operator, therefore, has to consider the cost and inconvenience of
underwater cleaning, when no other substantial fouling than slime, is present, yet without being sure of the
consequential performance gain. The operator may attempt ‘a before and after’ log of fuel consumption in
service, but this is fraught with difficulty and inaccuracy in most cases.
Clearly, more studies are required about the prediction, characterisation and drag penalties of slime films; but
before leaving the subject of slime, it would be well just to recall some of its associated mysteries. Fish do
13
not foul but do have slime films. If fish generally do not carry epiphytic growths, is their slime film an
antifoulant or does it have a cell wall shedding mechanism? Or what? More interestingly, for the naval
architect, is the notion that fish slime, when diluted in water, has drag reducing properties, which some fish
seem to call upon when required, for example Rosen et al. (1971).
3. Shell and weed

The drag penalty for a shell infestation of a hull is relatively simple to calculate. The straightforward shell
characteristics of diameter, height and distribution density can be readily measured. The drag of simulated
shell on a flat surface can be measured in a ship model testing basin or a flow channel. A well known example
is the classic pontoon tests of Kempf (1937). He found that the maximum drag increase occurred with shell
fouling covering 75% of the wetted surface. The shell height was 14mm. But even with only 5% of the wetted
area covered, the drag increase was already 2⁄3rds of the maximum. From his empirical data he was able to
calculate the increased resistance of ships due to shell, in a number of cases. For example, a 120m vessel,
with 75% coverage with shell of 4.5mm height would show an increase of 85% in skin friction resistance. If
shell characteristics are known therefore, it is possible to estimate the added drag. The nature of ship service
nowadays means that calcareous fouling is less common than it was in the 1930’s and is supplanted today
by weed fouling.
Estimating the added resistance of a ship covered with weed fouling, say ectocarpus or enteromorpha
typically, is a more intractable problem. The percentage of area covered can be estimated visually, but what
of the weed itself? Wet weight, dry weight or ash weight are used for other purposes, but these are not
promising parameters for correlation with drag. The average height of pliable filaments, attached
perpendicular to a surface and then subjected to a fluid flow, could be used as a characterising parameter,
along with some measure of the density of distribution of the filaments. Early work by Lewkowicz et al. (1986)
at Liverpool University, following upon their studies of roughness drag in pipes, pioneered this approach to
marine fouling drag. Work along these lines is being pursued, notably by Swain (1999) and by Schultz (2000).
With the advent of laser doppler velocimeters for use in flow channels and other fluids apparatus, it is now
possible to study the effects of surface characteristics on the turbulent structure of fluid flow and hence the
resulting added drag. Interestingly, the marine problem has common cause with those interested in the fluid
boundary layer interactions between the wind and plant canopies eg. a cornfield, which is of interest to
meteorologists among others, eg. Raupach et al. (1981). A little more will be said about this topic later, in the
final section on research directions.
The work on filament drag is making good progress but there will be a difficulty in applying results to the
circumstances of a particular ship. One problem would be the measurement of average filament length or
height characteristic over the hull, presumably by diver. Perhaps it will be sufficient to calculate the drag and
fuel penalty for a few typical cases, in order to write the bad news of weed fouling. However, for the ship
operator, it is a matter of deciding between the unscheduled dry docking and re-coating costs on the one
hand, and the continuing extra fuel cost penalty, in service, on the other.
Another possibility is to provide a weighted, numerical fouling index from a visual inspection of a hull just
after in-docking. The weighting would have to account for fouling type, intensity and distribution. The
difference between the in-docking fouling and the re-coated, foul free bottom could then be correlated with
some performance measure eg. the fuel consumptions immediately before and after docking, somewhat
along the lines of Bohlander’s study (1991). The fouling index has to be designed with care, and in service
performance monitoring is prone to error.
4. Research directions
If the nature of a surface, for example, its roughness, can be numerically categorised, and if the added fluid
frictional drag, due to the surface condition, can be measured, then there is a possibility of an approximate
correlation between the surface condition, however measured, and the added drag. Once the correlation and
its limits of applicability are established, it is then possible to measure another surface condition and hence
predict its added drag. This research route has been followed in the case of hard (antifouling) paint surface
roughness, and a correlation has been established, ITTC Madrid (1990), which was referred to earlier.
In the case of ship hull roughness, one obvious candidate parameter to characterise the surface was a
statistical measure devised for the Lucy Ashton trials. The parameter is Rt(50) and is the maximum peak to
the lowest trough in a 50mm sample length, along the rough surface. The roughness height distribution is
random and therefore it is necessary to use an average value of this parameter over a measurement position
(mean hull roughness, MHR), and then MHR is averaged over the hull (Average Hull Roughness, AHR). There
14
is a standard procedure for measurement of a ship hull surface, Townsin et al. (1981). There is also a special
instrument for taking the measurements, the Hull Roughness Analyser. Although this ‘ready made’ parameter
was available, it was necessary to test a variety of alternatives to correlate with added drag, especially
because texture as well as a height measure has to be accounted for.
The correlation involves the velocity distribution u at a distance y from the surface, within the turbulent
boundary layer. A simple model of this distribution is known as the Logarithmic Law of the Wall :-
u/uτ = (1/κ) ln (yuτ /ν) + Bο
where the friction velocity uτ = √(το/ρ)
and το = the wall shear stress
with ρ = fluid density.

κ is the von Karman constant, which appears to be a universal constant in most circumstances. Bο is also
constant for smooth wall flows. This model has been explored by numerous measurements of fluid boundary
layer flows over smooth surfaces.
Boundary layer velocity measurements over rough surfaces do not obey the Logarithmic Law as written
above, but empirical results may be interpreted by adding a further term (- ∆u/uτ ). This latter term is called
the Roughness Function. Authorities agree that this function is a logarithmic expression involving some linear
measure characterising the roughness. However, there are various proposals and some debate about the
precise nature of the expression. In the end, it is important to find a linear measure of roughness, which does
correlate with the roughness function.
There are only about 30 sets of experimental data available concerning fluid flow over rough painted
surfaces, where the surfaces have been subjected to detailed, statistical roughness measurement and where
there are adequate flow measurements to allow the roughness function to be calculated. The correlation
between all these roughness functions, and a number of linear measures, h, in the quantity log (huτ / ν), has
been explored by Townsin et al. (1990).
The best correlation was when h = √(αm0m2), where m0 and m2 are spectral moments of the random
roughness distribution and α is the band width parameter. This may seem complicated, but for the naval
architect, it is interesting to note that the statistical modelling of a random roughness is similar to the spectral
representation of ocean wave heights and is therefore familiar.
As a concession to simplicity, however, it was found that, providing the data was restricted to values of Rt(50)
<230µm, which is the moderately rough ship range, then the roughness function correlated well enough with
h = Rt(50), for all the paint surfaces tested. The way was now open to correlate Rt(50) directly with roughness
added resistance values, as measured by a number of authorities. Finally, a simple formula for the added ship
resistance in terms of average hull roughness, AHR, was the outcome:
1000∆CF = 44 [(AHR/L)1/3 – 10Rn-1/3 ] + 0.125
where the added resistance coefficient ∆CF = ∆R / 0.5ρSV2, the ship length is L, Rn is the ship Reynolds
Number, at speed V and the ship wetted surface is S.
This rather lengthy explanation of the research direction taken to estimate the drag of a surface covered with
a hard, rough, paint coating, gives background for the question, ‘can the same route be taken in the case of
flexible (weed) roughness?’ A notable paper by Schultz (2000), takes us, with academic rigour, in the direction
of an answer. The use of laser doppler flowmeters, in association with flow channels, allows us to measure
velocities in fluid boundary layers with some convenience. Hence, the roughness function over a filamentous
surface may be determined. However, the same problem arises, as with paint roughness, viz. a linear
measure of the fouling height (under flow), must be determined, in order that the measured roughness
function, which includes the height, may correlate with the linear measure itself. An added complication with
a pliable surface coating, is that changes in the flow can, to some extent, change the fouling height, however
it is measured. Also, the induced waviness caused by the flow over a pliable surface has to be taken into
account. The mobility of the strands of weed fouling may interact with the turbulent velocity perturbations
and it is not beyond possibility that there could be a turbulent drag reduction mechanism. Possibly, the
eruption and bursting of stream-wise turbulent vortices are delayed by filament mobility. All these features
lead to the need for a close examination of the flow characteristics between the smooth and fully rough
regimes. The Colebrook – White model, (1937), upon which we have depended for some time, may not be
adequate for flow over a weed fouled surface.
15
Compared with studies of the flow over hard rough surfaces, there is, as yet, a limited amount of published
boundary layer measurements over weed fouling where there are also adequate numerical data describing
the fouling itself. As yet, therefore, no measure of fouling is available to correlate with its drag. The current
position is well summed up by the last sentence in Schultz’ paper: ‘Scatter in the roughness functions of
these surfaces indicates that further research is needed to better correlate the physical measures of the algae
layer with their roughness function’, and, it might be added, ‘and with their drag’.
Concerns attending the gradual abandonment of tributyltin, range from fear of a return to endemic fouling,
to the increased cost of replacement antifouling provision. However, despite these concerns, there are
promising signs. Marine coatings companies are promoting their antifouling products on the basis of their
smoothness and smoothing properties, once again. In that connection, interesting problems arise with the
biocide free, low surface energy coatings.
It seems that the chemistry of foul release products is such that the applied coating cures with a smoother
finish than other compositions, applied in the same way. Of itself, smoothness seems to inhibit fouling
settlement, but it is the drag issue which is of particular interest. It is generally realised that two types of
parameter are required to characterise hard roughness in order to correlate with added drag; one is a linear
or height measure, the other is usually referred to as a texture parameter. A peak-to-valley height measure
and the first two, even, spectral moments, as texture parameters, have already been referred to. There are
other statistical parameters which have been proposed, eg. Musker (1977). Hitherto, paint surface texture
has consistently related to its roughness height, implying that a height measure alone, eg. Rt(50), could
characterise, at least moderate hull roughnesses. The foul release surfaces, however, have a noticeably
different texture, with more spectral energy in the longer wavelengths. This is likely to mean a lower drag for
the same height, Candries et al. (2002). An implication is, of course, that the whole correlation process of
roughness measurement and analysis, of boundary layer studies and of drag measurement, has to be gone
through again for low surface energy coatings.
An interesting experience for the student of these elastomerised surfaces, is to feel them dry, when they drag
like rubber, and then to feel them underwater, when they are fish slippery. A consequence when using the
standard, Hull Roughness Analyser, is that the stylus judders over a dry surface but the drive wheel slips
when the surface is wet and measurements can therefore be unreliable. Additionally, it is necessary for the
instrument to be modified to record the surface trace digitally so that texture and height parameters other
than Rt(50), may be determined. Chuah et al. (1990).
5. In conclusion
In the matter of hull fouling, it is the marine paint chemists and the marine biologists who pave the route
ahead; naval architects count the cost, and the ship operators pay the bills. It is appropriate therefore, to
conclude with some consideration of the economic penalties of fouling.
In a particular ship case, if outer bottom surface condition can be numerically assessed, whether it be
roughness or fouling, and if that measure can be used to predict the drag penalty, then, using discounted
cash flow techniques, an outer bottom maintenance strategy can be developed and alternatives evaluated.
In the case of roughness, this can be achieved, Townsin et al. (1981). In the case of fouling, the difficulties of
measuring and correlating with drag have already been referred to.
Another approach for a particular ship is to monitor the speed to power relationship in service. Increases in
fuel consumption can then be a determinant in the maintenance strategy. However, monitoring performance
at sea is fraught with imprecision. The effects of wind, waves and ocean currents, must be accounted for.
The measurement and relationship between propulsive power and fuel consumption are difficult to assess in
service. An example of the intricacies is in Townsin et al. (1980). Nonetheless, fuel consumption figures on
passage are often the only hard evidence available in disputes between ship operators and the coatings
industry when an antifouling is thought to have failed.
As a closure, it is worth reflecting upon the global savings due to efficient outer bottom surface maintenance
and the use of effective antifouling. Ablative, organotin antifoulings were introduced in 1974. A survey of the
effects on ships as to their roughness and fouling, over a decade, 1976 to 1986, was undertaken at
Newcastle University, Townsin et al. (1986). The hull roughness of 47 ships was measured on 147 occasions
during the decade, Byrne (1980). Only a few of these ships had ablative antifoulings. A further 98
measurements were taken later in the decade, on ships, most of which had TBT SPC coatings. A striking
improvement in in service roughness was noted.
The most striking change discovered, however, was in antifouling performance. Data from 672 ships prior to
the introduction of ablative coatings, showed only 19% entering drydock with zero fouling and 20% were
16
more than three quarters covered. Later in the decade, 183 ships which had TBT SPC antifouling were
surveyed: 91% entered dry dock with zero fouling, despite longer inter docking periods.
Milne (1989) attempted the difficult task of putting a global figure of savings to the shipping industry,
consequent upon the introduction of ablative coatings, based upon the decadal study referred to. Savings
were calculated in four groups: fuel cost savings due to the reduced ship frictional resistance; savings due
to extended inter docking periods; savings due to consequential lower dry dock costs: and indirect savings,
including, for example, savings due to the lower requirement to transport bunkers to refuelling ports. The
annual savings in the four categories in $ millions came to 720, 409, 800 and 1080, respectively, giving a
grand total annual saving, for the world fleet, of about $ 3,000 million.
Anyone working in the antifouling field could feel proud to have assisted in achieving these savings for the
shipping economy. For many, however, a greater source of satisfaction in producing efficient antifoulings, is
the reduction in smoke stack emissions: the greenhouse gas and global warmer, carbon dioxide, CO2; the
acid rain and ozone depleters, nitrogen oxides and sulpher oxides, NOx and SOx. In line with other
estimates, Milne calculated an annual world fleet fuel consumption of 184 million tonnes, at the time of the
decadal study. From that study he took a figure of 2% reduction due to smoother hulls and a further 2%
reduction due to improved antifouling, yielding an annual fuel saving of 7.36 million tonnes. The carbon
content of liquid fuels is quite constant at 85% – 86% and every tonne of fuel used creates 3.1 – 3.2 tonnes
of CO2 Nurmi (2001). The savings in greenhouse gas due to improved antifouling during the decadal study,
was, therefore, about 20 million tonnes per annum.
Perhaps another decadal study should be undertaken in the post-TBT SPC era.
References
Bertram V (2000) Past, present and prospects of antifouling. Proc. 32nd WEGEMT School, on Marine
Coatings. Univ. Plymouth. July. pp 85-97.
Bohlander GS (1991) Biofilm effects on drag: measurements on ships. Conf. Polymers in Marine Environment.
Marine Management (Holdings). 23-24 Oct.1991. Paper 16 pp.1-4.
Byrne D. (1980). The hull roughness of ships in service. MSc Thesis. Univ. Newcastle upon Tyne.
Callow ME, Milne A (1985) Non-biocidal antifouling processes. Conf. on Polymers in a marine environment.
Trans IMarE (C) 97. pp.229-233.
Candries M, Atlar M, Mesbahi E, Pazouki K. (2002). The measurement of drag characteristics of tin free, self
polishing copolymers and foul(ing) release coatings using a rotor apparatus. 11th Int. Congress on Marine
Corrosion and Fouling. San Diego. July.
Chuah KB, Dey SK, Thomas TR, Townsin RL. (1990). A digital hull roughness analyser. Int. Workshop on
Marine Roughness and Drag. RINA, London. 7pp.
Colebrook CF, White CM. (1937). Experiments with fluid motion in roughened pipes. Proc. Roy. Soc. London.
Ser. A 161 pp. 367-381.
Conn JFC, Lackenby H., Walker WP (1953) Resistance experiments on the Lucy Ashton. Trans. INA 95.
pp.350-436.
Denny M (1951) BSRA resistance experiments on the Lucy Ashton. Part I. Full scale measurements. Trans.
INA 93. pp.40-57.
Holzapfel ACA (1904) Ships compositions. Trans. INA 46. pp.252-265.
ITTC (1990). Proc. 19th International Towing Tank Conf. Report Powering Performance Comm. Madrid 16-22
Sept. pp 235-287.
Kempf G (1937) On the effect of roughness on the resistance of ships. Trans. INA. 79. pp.109-119.
Lewes VB (1889) The corrosion and fouling of steel and iron ships. Trans. INA 30. pp.362-389.
Lewkowicz A, Das DK (1986) Turbulent boundary layers on rough surfaces with and without a pliable overlay:
a simulation of marine fouling. Int. Shipbldg. Progress 33. pp.174-185.
Lewthwaite JC, Molland AF, Thomas KW. (1985). An investigation into the variation of ship skin frictional
resistance with fouling. Trans. RINA 127, pp. 268-279.
17
Loeb G, Laster D, Gracik T. (1984) The influence of microbial fouling films on hydrodynamic drag of rotating
disks. In ‘Marine Biodeterioration: an Interdisciplinary Study’. Proc. of Symp. on Marine Biodeterioration.
Uniformed Services University of Health Sciences. 20-23 April 1981.
Milne A, Hails G.(1971). British Patent 1,457,590.
Milne A. (1989). (Private communication). Aquatic biocides: their benefits and environmental risks.
Unpublished lecture to Royal Soc. of Chemistry. Autumn Mtg. Symp. 6. Industrial Div.
Musker AJ. (1977). Turbulent shear flows near irregularly rough surfaces with particular reference to ships’
hulls. PhD Thesis. Liverpool Univ.
Nurmi J. (2001). Achieving the optimum ‘green’ level by minimising exhaust emissions.
The Future of Ship Design – Part 2. Deltamarin/Journal-The Naval Architect . RINA pp.30-34.
Raupach MR, Thom AS. (1981). Turbulence in and above plant canopies. Annual Review Fluid Mech. 13 pp.
97-129.
Rosen MW, Cornford NE. (1971). Fluid friction of fish slimes. Nature 234 pp. 49-51.
Schultz MP, Swain GM (1999) The effect of biofilms on turbulent boundary layers. ASME J. Fluids Eng. 121
Schultz MP (2000) Turbulent boundary layers on surfaces covered with filamentous algae J. Fluids Eng. 122
Townsin RL, Svensen TE. (1980). Monitoring speed and power for fuel economy. Shipboard Energy
Conservation ’80. Symp. SNAME New York. 22-23 Sept. 24 pp.
Townsin RL, Byrne D, Svensen TE, Milne A (1981) Estimating the technical and economic penalties of hull
and propeller roughness. Trans. SNAME 90 New York. pp. 295-318.
Townsin RL, Spencer DS, Mosaad M, Patience G. (1985). Rough propeller penalties. Trans. SNAME 93
pp.165-182.
Townsin RL, Byrne D, Svensen TE, Milne A. (1986). Fuel economy due to improvements in ship hull surface
condition 1976-1986. Int. Shipbuilding Progress. 33 July pp. 127-130.
Townsin RL, Dey SK (1990) The correlation of roughness drag with surface characteristics. Int. Workshop on
Marine Roughness and Drag. RINA London. 14pp.
Watanabe S, Nagamatsu N, Yokoo K, Kawakami Y (1969) The augmentation of frictional resistance due to
slime. Jour. Kansai Soc. of Nav. Arch. No. 31 March 1969. pp.45-51. (In Japanese, see BSRA Translation
No. 3454).
18
Notes
19
Notes
20
Calculating the Cost
of Marine Surface
Roughness on Ship
Performance
(2000)
‘Garfield’, Whitley Road, Benton
Newcastle upon Tyne, NE12 8BP, UK.
Calculating the Cost of Marine Surface
Roughness on Ship Performance
‘Garfield’, Whitley Road, Benton, Newcastle upon Tyne NE12 8BP, UK
Abstract
This chapter provides the information, means and opportunity to calculate the speed and power penalties
resulting from the roughness of an antifouled hull surface. The information concerns the sources and causes
of antifouled hull surface roughness, the definition and measurement of hull roughness, and the relation
between roughness and added drag. The means are formulae for the calculation of added resistance, power
and fuel consumption. Data is provided for worked examples in the workshop.
Keywords : Marine roughness, Added resistance, Ship performance penalties.
1. Sources and causes of antifouled hull surface roughness.

The new ship at outdocking.
The roughness of the final coat of antifouling on a new ship results from the rheological characteristics of the
paint itself and, more importantly, from the circumstances of its application. In laboratory conditions, a paint
might finish with a peak-to-trough measure of 40 µm MHR (q.v. later). The best ever recorded figure for a new
ship roughness was 78 µm AHR (q.v.) and a typical ‘good’ finish today will be, say, 125 µm AHR. ref. (1).
The principal ‘applied’ roughnesses are as follows:
• Orange peel - this texture arises from the limitations of the flow of the paint.
• Runs and sagging - due to an over thick application.
• Waviness - spray fan blowing waves in the wet deposited paint.
• Overspray - solidified droplets of paint arriving at the wet painted surface giving a ‘sandy’ texture, due
to spray gun too far from surface.
• Grit inclusions - due to wind blown dust, blasting grit etc.
• Crazing - cracks and crazes due to paint and environmental circumstances.
A ‘bad’, new finish might be 450 µm AHR, from any of the above causes.
In service roughness.
The hull roughness of a ship in service increases between re-coatings. This is due to failures of preparation,
application and of the paint itself, and, usually more importantly, due to surface damage in service.
The principal in service roughnesses are caused as follows:
• Welds and surface defects at newbuilding, which are not properly cleaned, will give rise to corrosion and
blistering.
• Old corrosion pitting, when overcoated, gives rise to a ‘negative’ roughness.
contains.
© Akzo Nobel, 2003
23
• Paint detachment may be of the new coating or of the overcoated previous coating and there may be a
substrate compatability problem.
• Recoating over inadequately cleaned off fouling residues.
• Surface damage in service arises from many causes eg. anchors and cables, fendering, tug nosing,
groundings, drydock block damage.
• Underwater cleaning in the hands of non-specialist divers using inappropriate equipment can also cause
damage. Later in this School there will be an account of how underwater cleaning can be carried out
correctly. This raises the important issue of paint hardness, which arises, eg., when considering slime
cleaning from the new, low surface energy antifoulants.
2. The definition and measurement of hull roughness.

Following Froude’s work on roughened planes, one of the earliest attempts at hull roughness definition was
by sand grain comparison. The Prandtl/Schlichting/Nikuradse work on flow through sand grained pipes
(1934) ref.(2) was invoked to attempt both a definition of ship hull roughness and its penalty. During the
1940’s there was an interest in the possibility of laminar (low drag) wing sections for aircraft. An aerofoil
roughness gauge was developed at the National Physical Laboratory, consisting of a flexible steel foil with a
longitudinal slot, which could be bent around a wing section. A trolley could be run over the foil with a stylus
passing through the slot. The perpendicular, mechanically amplified movement of the stylus over the rough
wing surface was then recorded against the distance travelled by the trolley. This idea was developed to
measure ship hull roughness, specifically for the well known ‘Lucy Ashton’ trials of the 1950’s ref.(3). The
method for defining hull roughness then, is still in use today, although the instrument has been improved to
become the BMT Hull Roughness Analyser (HRA). It is singular that no other hull roughness definition has
been developed and no other instrument has been successfully put on the market.
The universally adopted parameter is Rt(50), which is the distance in µm, perpendicular to the surface, from
the highest roughness peak to the lowest trough, in a distance of 50mm. The 50mm determines the long
wavelength cut-off, whilst the diameter of the ball at the tip of the stylus determines the short wavelength
cut-off. It was argued, in the ‘Lucy Ashton’ work, that the longer wavelengths, like plate waviness, would not
affect resistance and could not be called roughness. Similarly, short wavelengths, say less than 1mm, would
also have no effect on drag.
Since Rt(50) varies all over a hull, some averaging system is required, which must link up with a procedure
for surveying a hull.
At a measuring station on a hull, one arm’s length pass of an HRA trolley yields about m = 12 values of Rt(50).
A complete hull survey should cover over 100 equally spread measuring stations on both sides and the
bottom.
AHR = k = Average Hull Roughness
= 1/n ∑MHR ......n about 100
MHR = Mean Hull Roughness (of one pass)
= 1/m ∑Rt(50) ....m about 12.
The histogrammatic representation of survey results, both as grouped (MHR) and ungrouped Rt(50) data will
indicate the influence of surface damage, and this is explored at length in ref. (4). An outcome of such
studies, for example, shows that whilst an ablative antifouling paint (eg. SPC), does indeed ‘polish’, or
become smoother in service, the ship becomes rougher due to surface damage. Ref. (4) also includes a
standard procedure for a hull roughness survey and its analysis.
3. Roughness and added drag.

It should be noted that there is an arbitrariness about Rt(50) as an appropriate roughness measure, both as
to its cut off values and the whole question of whether peak-to-trough height is the correct or adequate
measure to correlate with added drag. Perhaps, as well as a roughness height measure, there ought to be a
‘texture’ parameter to describe the shape of the roughness peaks and to define their distribution. These
questions have been extensively researched and discussed, eg. Ref.(5).
The foundation for research is the measured added drag of various rough surfaces, the topographies of
which have been carefully measured, usually yielding a set of ‘roughness’ and ‘texture’ parameters in
24
addition to Rt(50). Among those parameters proposed, have been kurtosis, skewness and the even moments
of the power spectrum of the roughness profile.
The drag measurements that have been undertaken have usually been on rough flat plates, internally
roughened pipes or on rotating cylinders. The correlation of the various parameters, or combination of
parameters, are not usually with the added drag values directly, but with the roughness function.
If the boundary-layer flow velocity u, at a distance y from a smooth surface, is plotted against log y, the curve
is linear between about 0.5 δ and 10 - 15% δ, (the ‘inner law’), where δ is the boundary layer thickness.
It is usual to non-dimensionalise the plotting as u/uo against loge yuo/ν, where uo is the flow velocity at the
edge of the boundary-layer and ν is the kinematic viscosity.
It has been found that these flow parameters for a surface which is rough, behave in the same linear manner.
The smooth and rough linear sections are found to be parallel, with the rough line ∆u/uo below the smooth.
∆u/uo is called the roughness function, and it is this that research has attempted to correlate with various
statistical roughness measures.
When the roughness function is plotted against loge huo/ν, where h is a particular statistical roughness
measure, all the available experiment data is often scattered and no correlation can be found. However, when
h = Rt(50) a correlation can be detected provided that:
• the surface is an undamaged, conventional, antifouling paint and
• the surface is not excessively rough, viz. Rt(50) < 225µm say.
It was on this basis that the 19th ITTC (Madrid 1990) recommended the adoption of a new roughness penalty
formula and a revised power prediction correlation allowance, ref (6).
4. Formulae for the calculation of added resistance.

The following formulae are based upon the work described above. A more complete account may be found
in the references at the end of this chapter.
The work described in ref.(7) involved the use of an integral momentum method for the calculation of the
frictional drag of five ship types - ULCC, products carrier, containership, small tanker, frigate - both smooth
and with homogeneous roughness over a range of severity. Those familiar with these procedures will know
that a skin friction equation is required and, for this purpose, Coles law of the wall / wake, was used, which
has the roughness function as one term. Any deficiencies in the accuracy of the integral method were largely
offset since only the difference, ∆CF, between CFrough and CFsmooth, was required. The correlation between
the roughness function and any roughness measure, however, has to be as reliable as possible.
The results from the calculations showed:
• a linear relationship between ∆CF and (AHR/L)1/3 at a fixed speed for each ship, where L is length between
perpendiculars.
• for a given ship and hull roughness, ∆CF was found to be speed dependent.
Whilst there is no theoretical justification for the 1/3rd power as a lineariser, the idea came from the work of
Telfer, esp.ref.(8), who reanalysed Froude’s sand roughened planes data to give a relationship of the form:
CFR = A (h/L)1/3 + B
Telfer also advocated a smooth ship extrapolater:

CFS = C + D(Rn)-1/3
Subtracting the smooth from the rough yields:

∆CF = A (h/L)1/3 - D(Rn)-1/3 + constant
which is speed dependent.
This form was applied to the detailed calculations for the five ship types referred to above and a satisfactory
fit to the results was given by:
103∆CF = 44[(AHR/L)1/3 - 10(Rn)-1/3] + 0.125 ————equation 1.
This is the formula adopted by the 19th ITTC, ref (6).
25
Consider a change in roughness from AHR = k1 to k2, k2 > k1. If the ship maintains the same speed, the
increase in the resistance coefficient as the ship roughens from k1 to k2 is:
103∆CF = 44[(k2/L)1/3 - (k1/L)1/3] ———————————equation 2.
k and L must be in the same units!
Let the total resistance coefficient be CT = (total resistance) / 0.5rSV2 and CF is the frictional resistance
coefficient, then the fractional added resistance is:
∆R/R = ∆CF/CT = 0.044[(k2/L)1/3 - (k1/L)1/3] / CT ———————equation 3.
5. Formulae for the calculation of added power and fuel consumption.

The resistance increases due to hull surface deterioration and we may consider two possibilities:
• Speed is maintained and therefore power and fuel consumption increase.
This is appropriate to the maintenance of schedules on a liner service e.g. a containership.
• Power is maintained and therefore the speed drops.
If power is constant the fuel consumption is constant except for the effect of reduced r.p.m. This is like a
tanker operation.
6. Power increase ∆P/P at constant speed.

Let η= the open water propeller efficiency.
Let T= propeller thrust.
Let P= shaft power.
P = TV / η ............smooth hull
P + ∆P = (T + ∆T) V / (η+∆η) ............with added roughness resistance
= (TV/η)(1+ ∆T/T)(1+∆η/η)-1
∴1+∆P/P = (1+∆R/R)(1+∆η/η)-1 —————————————-equation 4.
The added resistance at constant speed may be calculated as shown above: it remains to establish a
relationship between (1+∆R/R) and (1+∆η/η)-1. Noting that ∆R/R is almost the same as ∆T/T, the relationship
can be calculated from the propeller design data. For most propellers, the relationship is closely linear in the
relevant range and, of course, the curves intersect at (1.0, 1.0). As a handy guide, the following approximate
relationships hold, in the relevant range, for a ro-ro ship and a tanker, which typify the liner and the bulker:
(1+∆η/η)-1 = 0.17(1+∆R/R) + 0.83 .............for ro-ro ship
= 0.30(1+∆R/R) + 0.70 ............for tanker. ————equations 5.
7. Speed loss at constant power.

Let R ∝ Vn where n may be regarded as constant over a small speed range V2 to V1
then P = kVn+1
Suppose resistance is increased at any speed due to deteriorating
hull condition and suppose that the resulting fractional increase in
power ∆P/P is constant over a small speed range, then the equation
of the rough power curve will be (see figure right) :
P = kVn+1 + ∆P
Rough and smooth power curves.
26
At V2 power for smooth ship = kV2n+1
V1 power for rough ship = kV1n+1 + (∆P)1

Thus, for the same power kV2n+1 = kV1n+1 + (∆P)1
k(V1 + ∆V)n+1 = kV1n+1 + (∆P)1
[(V1 + ∆V)/V1]n+1 = 1 + (∆P)1/P1
(1 + ∆V/V1)n+1 = (∆R)1/P1 + 1
Using the Binomial Theorum, since ∆V/V is small,
1 + (n+1)∆V/V1 ≅ (∆P)1/P1 + 1
∆V/V ≅ ∆P/P (n+1)-1 - - - - - - - - - - - - - - - - - - equation 6.
and ∆P/P is the fractional change in power at constant speed due to added resistance, which can be found
from the preceding analysis.
8. Determination of the speed index n.

The simple relationship P = kVn+1 can only hold with adequate accuracy over a small speed range around the
speed of interest. It follows that n will not only vary from one ship to another but also vary at different speed
positions along a given speed to power curve. Typical values are given in Table I below, which are taken from
ref. (9).
For a given speed to power curve we note that :
P2/P1 = (V2/V1)n+1
and hence n+1 = log P2/P1 (log V2/V1)-1
from which it is simple to estimate n from a given speed to power curve around a particular speed.
Table I. Values of the speed index n for 4 ships at each of 2 speeds.
ship type length speed (knots) speed index n
low speed cargo ship 122 m 10.5 2.19
8.9 2.11
intermediate cargo liner 140 m 16.0 2.29
13.6 2.36
large passenger ship 230 m 24.0 3.00
20.4 2.29
cross channel ferry 100 m 24.5 4.55
20.8 3.62
Whilst the ship types are somewhat dated, Table I illustrates the likely variations in n.
9. Examples.
A containership of length 274.32m has CT = 2.52 x 10-3 at service speed and the hull roughness deteriorates
from 150µm to 500µm AHR over an interdock cycle. Use equation 3.
∆R/R = 44/2.52 [{500/(274.32x106)}1/3 - {150/(274.32x106)1/3 ]
= 44/(2.52x649.7) [7.937-5.313]
= 0.071 or 7%
Where it is impossible to estimate a value for CT a very approximate estimate may be made from
CT = 0.018L-1/3 which, in the above case, gives CT = 2.77x 10-3 changing ∆R/R% to 6.4%.
Using the ro-ro formula in equations 5, as representative of a containership propeller, and inserting 1+∆R/R
= 1.071, we find:
27
(1+∆η/η)-1 = 0.17 x 1.071 + 0.83 = 1.012
Hence: ∆P/P = 1.071 x 1.012 - 1 = 0.084 or 8.4%
To calculate the speed loss at constant power, we have ∆P/P, above, the power increase at constant speed:
it remains to find a value for n. For a containership n ≅ 2.1 (from the speed to power curves in ref.10.), then:
∆V/V = [1/(1+2.1)] x 0.084 = 0.0271 or 2.7% speed loss.
In this case, of constant power, the fuel consumption rate is constant but there is increased consumption
due to extra time at sea, and there will be other financial penalties.
10. Economics issues.

Usually, the ship operator is principally interested in the financial outturn from alternative outer bottom
maintenance strategies. The drydock, maintenance and out of service costs may be predicted for a given
strategy for a period of say 10 years. The increase of fuel costs due to predicted hull surface deterioration
may be calculated as above. An analysis may then be made using discounted cash flow calculations. It is
not appropriate to develop these issues further in this workshop, but a clear development may be found in
ref.(11).
All of the foregoing assumes a continuously effective antifouling and this seemed possible at the introduction
of organotin biocides in an ablative matrix. Today, as TBT coatings are being phased out, a lower standard
of antifouling provision has to be accepted, at least for the next few years. Some fouling in service must be
expected, and, in the case of the ‘non-stick’ coatings, the presence of slime is likely. Maintenance strategies
must therefore include costed items for underwater inspection and cleaning. At present there is no method
for assessing the added drag of a fouled surface, principally because there is no measure of fouling which
can be correlated with its fluid resistance. It is generally understood, however, that fouling penalties are
usually severe and so it is prudent to ensure that they do not occur.
11. References.
1. Townsin RL, Byrne D, Milne A and Svensen TE. (1980), ‘Speed, power and roughness: the economics of
outer bottom maintenance’. Trans. RINA. Vol 122.
2. Prandtl L and Schlichting H. (1934), ‘Das Widerstandsgesetz rauher Platten’. Werft-Reederei-Hafen.
3. Lackenby H. (1962), ‘Resistance of ships, with special reference to skin friction and hull surface condition’.
Proc. I Mech E. Vol 176.
4. Townsin RL, Byrne D, Svensen TE and Milne A. (1981), ‘Estimating the technical and economic penalties
of hull and propeller roughness’. Trans. SNAME Vol 89.
5. ‘Marine roughness and drag’. (1990), RINA Int. Workshop. London.
6. 19th Int. Towing Tank Conference. (1990), Proceedings-Report of the Powering Performance Committee.
7. Townsin RL, Medhurst JS, Hamlin NA and Sedat R. (1984), ‘Progress in calculating the resistance of ships
with homogeneous or distributed roughness’. Trans.NECIES. Centenary Conf. on Marine Propulsion.
8. Telfer EV. (1969), ‘On taking the rough with the smooth’. Trans. IESS.
9. Lackenby H. (1952), ‘On the acceleration of ships’. Trans. IESS. Vol 95.
10. Townsin RL and Svensen TE. (1980), ‘Monitoring speed and power for fuel economy’. Proc. Shipboard
Energy Conservation ‘80 Conf. SNAME, New York.
11. Svensen TE. (1983),‘The economics of hull and propeller maintenance examined in the face of
uncertainty’. Trans. NECIES, Vol 100.
28
Notes
29
Notes
30
Estimating the
Impact of New
Generation
Antifoulings on Ship
Performance:
the presence of
slime (2000)
M Candries, M Atlar
University of Newcastle upon Tyne, UK.
C D Anderson,
International Marine Coatings.
Estimating the Impact of New Generation
Antifoulings on Ship Performance:
the presence of slime
M. Candries, M. Atlar, University of Newcastle upon Tyne, UK
C. D Anderson, International Marine Coatings.
Abstract
Due to the phase out of TBT SPCs imposed by the IMO, new generation antifoulings are set to replace 80%
of the existing antifouling market. Two types of coatings are claimed to offer satisfactory performance over
5 years: Tin free SPCs and Foul Release coatings, which were both commercially introduced in the mid
1990s. This paper gives an overview of the research at the University of Newcastle upon Tyne which
compares the drag, boundary-layer and roughness characteristics of both coatings when newly applied. It
was found that Foul Release coatings offer less drag than Tin free SPC, by an amount which depends on the
quality of application and which has been related to the respective differences in roughness characteristics.
Assessments have shown that Foul Release surfaces are very effective against macrofouling organisms, but
that the surface is covered by slime films when the vessel returns to drydock. A literature review on the effect
of slime films on ship resistance shows that slime films have a significant effect on drag but in turbulent flows
the effect is likely to remain limited because of detachment processes. Further research is underway to
investigate this.
1. Introduction
For years the most widely applied marine antifoulings have been Tributyltin Self-Polishing Co-Polymers (TBT
SPC). They can keep a ship free of fouling for 5 years by means of a steady release of TBT and other co-
biocides (such as cuprous oxide) contained therein. A chemical reaction occurs at the surface-seawater
interface (known as the “leached layer”) and forms a water soluble product that is able to dissolve away,
resulting in the surface “polishing” with time. However, due to environmental side effects related with TBT,
the International Maritime Organisation (IMO) has decided in October 2001 to prohibit the application of TBT
SPCs from 2003 and hence completely phase out their use by 20081.
There are currently two alternatives on the market that can also offer 5 years of satisfactory antifouling
performance. The first alternative, Tin free SPC, uses the same chemical principle but instead of TBT uses
other chemically bound moieties, based on copper, silyl or zinc compounds, to enable the surface to slowly
dissolve in seawater. This processes allows the gradual leaching of copper based biocides, which are
complemented by so called ‘booster biocides’ since copper alone does not have a sufficiently broad
antifouling spectrum. Some of these booster biocides have come under increasing environmental scrutiny,
but others, such as zinc polymers, degrade rapidly in seawater and have therefore much less impact on the
marine environment.2,3 Unlike the cheaper Controlled Depletion Polymers (CDP) antifoulings, the release of
the toxins continues when the ship is stationary and most prone to foul, as illustrated in Figure 1.
Assessments made during drydocking have shown that the antifouling performance of Tin free SPCs is
equivalent to TBT SPCs over a five year period4.
The second alternative, Foul(ing) Release coatings, acts as a physical rather than a chemical defence against
fouling.5 Instead of killing marine organisms that have attached to the hull, they try to prevent the attachment
of the organisms altogether by virtue of their surface properties. Most of the Foul Release coatings currently
on the market are silicone elastomers based on polydimethylsiloxane (PDMS). PDMS has an extremely
flexible backbone, which allows the polymer chain to readily adapt to the lowest surface energy
contains.
© Akzo Nobel, 2003
33
configuration. The surface energy represents the capability of the surface to interact spontaneously with
other materials. Brady and Singer6 found experimentally that the relative adhesion of fouling organisms on a
material is directly proportional to √Εγc, whereby E is the elastic modulus of the material, and γc its surface
energy. This parameter for silicone materials is at least an order of magnitude smaller than for other materials.
Eventually, fouling organisms will attach to the surface, but it has been shown that macrofouling (algal and
animal) organisms attach less strongly on PDMS than on other materials and that the strength of attachment
of macrofouling is inversely proportional to the thickness of the coating.7,8 This explains why these
macrofouling organisms (such as weeds and barnacles) will release from the surface under the influence of
relatively small hydrodynamic shear forces. The speed at which these organisms release has been measured
by Kovach and Swain,9 who towed a plate, which was coated with a Foul Release system and covered by
fouling, and observed that the organisms started to release at speeds above 12 knots. Further tests have
shown that, with the current Foul Release technology, speeds in excess of 15 knots will prevent most types
of fouling from settling on the surface. Foul Release systems are therefore particularly suited for ships which
spend a short time in port and travel at sufficiently high speeds.
When Foul Release coatings were commercially introduced in the mid-1990s and first applied on a high-
speed catamaran ferry to replace a CDP antifouling, the recorded fuel consumption was lower at the same
service speed, implying lower drag characteristics10. A research project was therefore undertaken at the
University of Newcastle-upon-Tyne with the objective of collecting data on the drag, boundary-layer and
roughness characteristics of Foul Release and Tin free SPC coatings, and to compare them systematically11.
A summary of the findings of this research project is presented in Section 2 of the paper.
Dry docking assessments have indicated that a microbial slime layer is present on ship hulls coated with Foul
Release systems. A literature review on the effect of slime on drag is presented in Section 3. The possible
repercussions on ship performance of hulls coated with Foul Release coatings are briefly discussed in
Section 4, which also describes the ongoing research.
2. Drag, boundary-layer and roughness characteristics of Tin free SPC and Foul Release coatings
This section summarises the findings of a research project carried out at the University of Newcastle-upon-
Tyne to systematically compare the drag, boundary-layer and roughness characteristics of Tin free SPC and
Foul Release coatings. The coatings used in this study, which were provided by International Coatings Ltd.,
were a PDMS and a Tin free copper-pigmented acrylic self-polishing co-polymer that contains zinc pyrithione
as booster biocide.
Drag measurements have been carried out in towing tank experiments with two friction planes of different
size, which showed that the Foul Release system exhibits less drag than the Tin free SPC system. The
difference in frictional resistance varied between ca. 2 and 23%, depending on the quality of application12.
Rotor experiments were also carried out to measure the difference in torque between coated and uncoated
cylinders. In addition to both coatings applied by spraying, a Foul Release surface applied by rollering was
included because there were indications that this type of application might affect the drag characteristics.
The measurements indicated an average 3.6% difference in local frictional resistance coefficient between the
sprayed Foul Release and the sprayed Tin free SPC, but the difference between the rollered Foul Release
and the sprayed Tin free SPC was only 2.2%13. Similar rotor experiments carried out with coatings of a
different manufacturer also indicated lower drag for Foul Release compared to Tin free SPC at high speeds14.
The friction of a surface in fluid flow is caused by the viscous effects and turbulence production in the
boundary layer close to the surface. A study of the boundary-layer characteristics of the coatings was
therefore carried out in two different water tunnels using four-beam two-component Laser Doppler
Velocimetry (LDV) and the coatings were applied on 1m long test sections that were fitted in a 2.1m long flat
plate set-up, as shown in Figure 2. The intersection of the laser beams is characterised by an optical
interference fringe pattern in the “probe volume” which allows accurate determination of the velocity in the
streamwise and wall-normal direction. The probe volume diameter was 276µm for the (blue) wall-normal
channel and 291µm for the (green) streamwise channel. The velocity measurements were conducted over
20s or until 4096 validated samples were collected, whichever came first. A traverse mechanism allows the
probe to be positioned to within ±12.5µm and moves the probe away from the wall so that a boundary-layer
velocity profile is measured. Velocity profiles were measured at five different streamwise locations and at five
different free stream velocities.
Figure 3 shows the boundary-layer velocity profiles at 1.607m from the leading edge for a free stream velocity
Ue = 5m/s. The distance from the surface, y+ε, and the streamwise velocity component U have been scaled
by the viscous length scale ν/Uτ and the friction velocity Uτ respectively. Various experiments have indicated
34
that by this inner layer scaling turbulent boundary-layer velocity profiles collapse regardless of the Reynolds
number15. An outer layer wall similarity method and the Reynolds stress method were used to determine the
friction velocity and both methods showed good agreement with each other.11 The measurements showed
that the friction velocity for Foul Release surfaces is significantly lower than for Tin-free SPC surfaces. This
indicates that at the same streamwise Reynolds number the ratio of the inner layer to the outer layer is
smaller for Foul Release surfaces. The inner layer is that part of the boundary layer where major turbulence
(and hence drag) production occurs.
By definition the friction velocity is equal to:
τw cf
Uτ = = Ue (1)
ρ 2
whereby τw is the wall shear stress, ρ is the density of the fluid, cf is the local frictional resistance coefficient
and Ue is the free stream velocity. Consequently, the downward shift in the log law region (where the velocity
increases linearly with the logarithm of the distance from the wall) of the velocity profiles shown in Figure 3
is a direct indication of the difference in local frictional resistance between a rough and the uncoated smooth
reference surface. This parameter ∆U+ is known as the velocity loss or roughness function.
Statistical analysis of the obtained values of the roughness function by means of multiple pairwise
comparison using Tukey’s test indicated that the roughness function for Foul Release surfaces is significantly
lower than for Tin free SPC surfaces at a 95% confidence level. These findings are consistent with the drag
characteristics measured in the water tunnel and rotor experiments, as shown in the overview in Table 1.
In addition to the difference in frictional resistance and the roughness function, Table 1 shows the average
roughness of each of the surfaces. This parameter was measured with the BMT Hull Roughness Analyser,
which is the stylus instrument that is most frequently used in dockyards. It has a cut off length of 50mm and
a sampling interval of 1.25mm and measures the highest peak to lowest valley roughness height over this
cut off length, known as Rt(50). It is clear from the rotor experiments and the large plate towing tank
experiments that this single amplitude parameter does not correlate with the measured drag increase for Foul
Release surfaces.
It is also worthwhile to note a difficulty of the stylus type instrument in measuring “non-stick” surfaces such
as the Foul Release. When it is dry, this elastomerised surface feels like rubber while the same surface
becomes slippery - like fish skin - when it is wet. In using the BMT Roughness Analyser, if the surface is dry
its stylus will judder over the surface, whereas its wheel will slip when the surface is fully wet. The
measurements taken in both conditions can therefore be unreliable, subject to a compromise of these
conditions which require considerable experience and familiarity to use this instrument. Of course one
alternative to help this problem is to use a non-contact type roughness analyser.
A detailed non-contact roughness analysis was carried out with an optical measurement system fitted with
a 3mW laser. Measurements were taken on sample plates coated alongside the surfaces tested in the towing
tank and water tunnel experiments and representative of their surface characteristics, and on slabs, cut from
the cylinders used in the rotor experiments. A moving average ‘boxcar method’ was applied to filter long-
wavelength curvature. The upper bandwidth limit or cut off length was set at 2.5 and 5mm, whereas the lower
bandwidth limit or sampling interval was set at 50µm. Typical roughness profile for each type of coating are
shown in Figure 4.
The detailed roughness analysis revealed that when the profiles are filtered, the amplitude parameters of the
sprayed Foul Release surfaces are in general lower than those of the rollered Foul Release surfaces and the
SPC surfaces. However, the rollered Foul Release surfaces display a roughness height distribution which is
considerably more leptokurtic (i.e. exhibits a larger number of sharp roughness peaks) than the sprayed Foul
Release surfaces. The greater number of high peaks on the rollered Foul Release surfaces is expected to
engender higher drag than sprayed Foul Release surfaces.
The main difference between the Foul Release and the Tin free SPC systems lies in the texture characteristics,
as shown in Figure 5 and Figure 6 for two typical roughness measurements of a Foul Release and a Tin free
SPC coating applied by spraying. Whereas the Tin free SPC surface displays a spikey “closed” texture, the
wavy “open” texture of the Foul Release surface is characterised by a smaller proportion of short-wavelength
roughness. This is particularly evident in texture parameters such as the mean absolute slope and the Fractal
Dimension. There is relatively little data available in the literature of irregularly rough surfaces on the influence
of texture only on drag. Grigson16 has mentioned explicitly that open textures have a beneficial effect on drag.
It is clear that in order to correlate with drag, the roughness of irregularly rough surfaces in general needs to
take both amplitude and texture parameters into account. In his classical work to investigate the turbulent
35
shear flows near such surfaces with particular reference to ships’ hull, Musker17 has proposed a combination
of various statistical parameters to characterise these surfaces, based on the pipe flow experiments.
Musker’s approach has been further explored by Townsin and Dey18, who have proposed a new set of
statistical parameters, with reference to fluid flow over a limited rough painted ship hull surfaces, where the
surfaces have been subjected to detailed, statistical roughness measurement and where there are adequate
flow measurements to allow the roughness function to be calculated.
Based on the experiments presented here, it is thought that the rheology of the paint (which is dependent on
the viscosity and significantly different for Foul Release and Tin free SPC coatings) has a direct effect on its
texture, whereas amplitudes depend significantly on the application quality. Correlation of the texture
parameters with the amplitude parameters, however, shows that the two are inter-related so that bad
application can be expected to have a knock on effect on the texture parameters.
A correlation analysis between the roughness and drag characteristics was carried out and reasonably good
correlation was achieved if a “characteristic roughness measure” is used which takes both the amplitude and
the texture of the surface into account. At present, the procedure adopted by the International Towing Tank
Committee (ITTC) to correlate roughness with drag only accounts for a single roughness amplitude
parameter.19 The procedure hinges on the use of a practical formula for the added ship resistance (roughness
correlation allowance), which was proposed by Townsin and Dey18 in terms of Average Hull Roughness for
moderately rough ship range, Rt(50) <230µm. Unfortunately, this procedure will not work for Foul Release
surfaces, unless a texture parameter is included in the roughness characterisation. Even then, full scale data
should be gathered in order to adjust and validate the prediction of added drag from measured roughness
characteristics. It is also recommended that more roughness profiles will be collected from dry dockings
since this study has only analysed newly applied coatings. This is only useful if a relatively simple
modification of the BMT Hull Roughness Analyser is carried out to record the entire profiles digitally, rather
than only the average extreme amplitude. Within the framework of necessity to modify the Hull Roughness
Analyser to record the surface profiles digitally, the past work of Chuah et al.20 will have even more
importance for analysing Foul Release surfaces while the work of Fitzsimmons and Ellis21 has practical
importance for the ship owners and operators in combining a two parameter roughness analyser with a
useful performance monitoring software.
3. The effect of slime on ship performance

In the previous section, it was shown how newly applied Foul Release surfaces exhibit drag benefits over
surfaces coated Tin free SPC. These differences in drag may seem relatively small but would nevertheless
offer significant fuel savings. Dry docking assessments, however, have indicated that Foul Release surfaces
quickly become fouled with slime films. The effect of this slime fouling on ship performance has not been
thoroughly investigated, but ship operators who have compared fuel consumption of Foul Release
applications with TBT SPC report that little differences can be seen after a period of time4. It would therefore
appear that slime fouling counteracts drag benefits of newly applied Foul Release coatings but does not
increase drag beyond that. This section reviews the literature on the effects of slime fouling on drag and
addresses the repercussions on ship performance of hulls coated with Foul Release surfaces. Within the
same framework, a recent keynote study by Townsin22 is particularly worthwhile to acknowledge in exploring
the ship resistance penalties due to slime, as well as shell and weed fouling.
Fouling starts from the moment the ship is immersed in seawater. The hull rapidly accumulates dissolved
organic matter and molecules such as polysaccharides, proteins and protein fragments23. This conditioning
process is regarded as the first stage of fouling, which begins within seconds, stabilises within hours, and
sets the scene for later fouling stages. When a conditioning film has been formed, bacteria and unicellular
organisms such as diatoms then sense the surface and settle on it, forming a microbial film24. This slime film
involves the secretion of muco-polysaccharides, and generally eases the way for macrofouling settlement
(i.e. weeds, barnacles).25 In this paper “slime film” and “biofilm” are used interchangeably to describe a
collection of micro-organisms and their associated extracellular products attached to a substratum, in
agreement with the definition given by Palmer and White26.
The genesis of fouling almost invariably occurs when the ship is at rest, most commonly in port. Ports differ
considerably in their tendency to cause fouling and it is commonly known that the problem of fouling is more
severe in tropical waters. The nature of the fouling community depends on the species of animal and plant
life present in the water, the salinity and temperature of the water, the degree of illumination of the hull
surface, the season when berthing takes place and the time spent in port. Considerable differences in the
nature and intensity of fouling on a Foul Release surface were measured by Swain et al.27 at seven different
marine sites.
36
While the consequences of macrofouling are well known because it has a catastrophic effect on resistance,
much less attention has been paid to the effect of slime films. McEntee28 was probably the first to mention
this. A separate section was devoted to the effect of slime fouling in the monumental work on antifouling for
the first half of the 20th century, Marine Fouling and its Prevention29. Towing tank experiments carried out by
the US Navy and involving different paints showed that after 1 days’ exposure the increase in resistance was
very small, but that after 10 days’ exposure the resistance was increased more than 10% was measured and
attributed to the effects of the slime film. It was observed however, that a significant part of the slime film
released and that after 30 days’ exposure the biofilm consisted of an upper layer which sloughs off and with
a harder slime layer underneath that does not release. It was therefore estimated that the eventual drag
increase on ships would be within a few percent of the painted hulls in clean condition, and within this
context it is interesting to mention that the formation of slime on the ‘Lucy Ashton’ after a 40 days’ mooring
period resulted in a 3.5% increase in total resistance30.
Very similar conclusions were reached by Watanabe et al.31 who studied the effect of slime on resistance by
an extensive series of rotor experiments involving cylinders and disks and towing tank experiments involving
a 9m long ship model. Compared to the towing tank experiments of Todd32 with painted surfaces, the added
drag of a slime film would be similar to a painted surface of (now very) poor finish. Watanabe et al. predicted
a 9-10% increase in total full scale resistance, which is comparable to the drag increases of 5-8% at 40 knots
predicted by Loeb et al.33 based on rotor experiments with disks. The rotor experiments by Watanabe et al.
also showed, however, that a very large quantity (>90%) of the slime film will be removed at speeds above
8m/s.
To avoid these detachment problems, a few experiments have been carried out with “artificial slime films”.
Lewkowicz and Das34 used nylon tufts to simulate fouling in general and measured the turbulent boundary-
layer characteristics in a wind tunnel. El-Labbad35 applied agar-agar gel of different concentrations on a
friction plane to simulate light and heavy slime films in particular. The towing tests showed that the frictional
resistance coefficient was increased by 4-11% and 13-21% respectively.
Picologlou et al.36 carried out pipe flow experiments with slime layers of thickness varying between 10 and
1000µm and found that the frictional resistance increased with increasing thickness. Thickness is not the only
factor, however, since they also noted that the viscoelastic nature of the biofilm may cause additional energy
dissipation within the turbulent boundary layer. Stoodley et al.37 also carried out pipe flow experiments and
observed that there are significant differences between biofilms grown in laminar and in turbulent flows. The
former were patchy and consisted of cell clusters separated by interstitial voids, whereas the biofilm grown
under turbulent flow were filamentous. The filaments were formed by the colonisation of filamentous
sheathed bacteria with microcolonies of non-filamentous bacteria. Significant increases in resistance were
only found if filaments were formed. It was also observed that the filaments can oscillate rapidly, causing the
additional energy losses reported by Picologlou et al36.
These observations are consistent with evolving perspectives of biofilm structure38. Driven mainly by the
application of confocal scanning laser microscopy, conceptual models of biofilms have evolved over the last
15 years from a homogeneous planar layered model to a heterogeneous model where convective flow is
possible within the biofilm. This is important because drag for a planar biofilm would mainly occur from skin
friction, whereas fluid flow around the structures in a heterogeneous biofilm will also result in form drag. This
form drag in effect causes the additional resistance and the flow is said to operate in the “transitional
regime”. Depending on the degree of coverage by and heterogeneity of the biofilm, the form drag may
dominate so that flow is in the “fully rough regime” and independent of the Reynolds number.
The effect of slime films on the turbulent boundary-layer characteristics was studied by Schultz and Swain39.
Different surfaces were tested in a water tunnel and the boundary-layer velocity profiles were measured with
two component LDV. They found that the frictional resistance coefficient was dependent both on the biofilm
thickness and on its morphology. The increase in the skin friction coefficient for slime films with a mean
thickness of 163µm and 347µm before testing was 3-70% (with an average of 33%) and 8-133% (with an
average of 68%) respectively. By comparison, the average increase for a surface dominated by filamentous
green algae (Enteromorpha spp.) with a mean thickness of 310µm was 187%. The average thicknesses of
the three surfaces after testing were 126µm, 74µm and 344µm, which seems to indicate that quite a
significant amount of the slime films was removed even at these low velocities (up to 3m/s). It was found that
the displacement thickness, the momentum thickness and the shape parameter of the boundary layer were
increased signifcantly by the biofilms. The turbulence intensities were also increased over a large part of the
boundary-layer.
The boundary-layer velocity profiles of surfaces covered by slime films were also measured by Lewthwaite
et al.40, who developed a technique for determining the local skin friction of a ship’s hull under seagoing
conditions by using a small pitot type probe. It was found that as the hull became covered with a dense slime
37
film but remained virtually free of weed and shell growth, the skin friction was increased by about 80% which
resulted in a 15% speed loss. Other full scale measurements are reported by Bohlander41, who carried out
power trials with a US navy frigate. A significant change in required shaft power, ranging between 8 and 18%,
was measured after the removal of a 22-month old mature slime layer and the maximum speed of the test
ship, was increased after cleaning by about 1 knot.
Thus far, the literature review has indicated that slime increases the resistance, but it should be mentioned
that when used as fluid additives in sufficiently large concentrations, some natural slimes are able to reduce
drag significantly42-45. Rosen and Cornford46 suggested that certain predatory fish like the barracuda are able
to reduce their drag by this mechanism during the short periods when the fish must catch its prey or escape
danger. Additives such as polyethyleneoxide are an established method of reducing drag and are used
successfully in pipe lines47,48. Canham et al.49 showed that significant drag reduction is also achievable for
ships, but overall the results were not economical. There is, however, some scope for economical marine
applications by means of pulsed injection50. This as an aside, the general consensus is that slime films on
ships will increase the resistance which can be significant unless a significant part of the slime layer releases
under hydrodynamic action.
Evaluation tests have shown that the adhesion strength of slime films to Foul Release surfaces is much
higher than the adhesion strength of other organisms. Most data is available on the shear adhesion strength
of barnacles because this is recognised as a standard method to test the efficacy of Foul Release materials51.
Swain et al.27 measured the barnacle adhesion strength of Foul Release surfaces at seven different marine
sites and observed large variations due to factors such as the nature of the biofilm, the temperature and the
salinity of the seawater. The pooled data showed that the barnacle adhesion strength was on average around
0.08Mpa for the coating used in the experiments in Section 2 (i.e. Intersleek). Terlizzi et al.52 measured that
the adhesion strength of barnacles to several silicone materials was between 0.1 and 0.2Mpa. By contrast,
the water jet pressure required to remove slime was between 0.4 and 1Mpa.
Kovach and Swain9 measured a barnacle shear adhesion strength of 0.75Mpa and 0.25Mpa on a 1m long
boat-mounted foil coated with two different silicones. Denny53 used strain gauge force transducers to
measure the drag and lift forces and found that the total hydrodynamic force on a solitary sessile marine
organism consisted of almost 90% lift. By taking for the barnacle a lift coefficient CL = 0.45 and a drag
coefficient CD = 0.5 at Re = 1·105, with the shear adhesion strength being three times larger than the tensile
adhesion strength, Kovach and Swain predicted the velocity at which barnacles would release from the two
silicones at Ue = 20 and 12 knots respectively, which was in good accordance with visual observations up
until 12 knots.
This prediction theory only considers a solitary organism and does not take the effect of growing in clusters
and the presence of other types of organisms into account. Boundary-layer theory is neglected and another
problem for predicting the release of fouling organisms from ship hulls, is that the boundary-layer thickness
is dependent on the Reynolds number at which the ship is operating. As mentioned in Section 2, the inner
part of the boundary layer scales with the friction velocity Uτ. Walderhaug54 gives the following approximate
formula for the friction velocity:
Ue
Uτ =
(ln Re)1.2 (2)
whereby Re is the Reynolds number. Using Eq. 1 the wall shear stress can be approximated by:
ρU e
2
τw =
(ln Re)2.4 (3)
White gives an alternative approximate formula for the wall shear stress:
55
τ w = 0.0135ν 1 /7 ρ U e x−1/7
13/7
(4)
whereby ν is the kinematic viscosity of the fluid and x is the streamwise distance from the leading edge.
Using Eq. 1 the friction velocity can thus be approximated by:
U τ = 0.1162U e Re−1/14
(5)
Figure 7 shows the approximate wall shear stress calculated with Eq. 3 and Eq. 4 against the free stream
velocity for a 1m long friction plane, a 90m long high speed ferry and a 250m long tanker in seawater at 15°C
(for which56 ρ = 1025.9kg/m3 and ν = 1.18831·10-6m2/s). The figure indicates that as the length of the vessel
38
increases at a fixed level of the wall shear stress higher free stream velocities are required to maintain
boundary-layer similarity. Schultz et al.57 recently used similar equations to predict the speed of fouling
detachment at full scale from the speed measured at laboratory scale in channel flow. As mentioned, further
complications to this prediction method based on boundary-layer similarity are the fact that different
organisms exhibit varying adhesion strengths and the fact that different organisms resist the flow stresses to
a different extent so that they can release under different modes (i.e. lift or shear) and mechanisms.
Morgenroth and Wilderer58 distinguish four mechanisms of detachment of which erosion (removal of small
groups of cells caused by fluid shear forces) will dominate when a ship is underway. The mechanism of
detachment interacts with the growth processes of the biofilm and partly influences its morphology. van
Loosdrecht et al.59 report that biofilms can adapt to certain shear conditions. Under high shear stresses the
biofilm becomes smoother and thereby less susceptible to shear forces. This interaction between growth and
detachment process will lead to an equilibrium thickness that decreases with shear rate.
4. Discussion and topics for further research

One of the key issues that has come forward from the literature review on slime films is that slime films will
attach more strongly to Foul Release surfaces than other fouling organisms but a significant proportion of the
slime film is expected to release as the service speed of the vessel increases29,31,39. The evolving perspectives
on the structure of heterogeneous biofilms indicate that the interaction between detachment and growth
processes leads to an equilibrium thickness that decreases with shear rate, or increasing service speed37,59.
This seems in agreement with observations of Klijnstra et al.60, who recently carried out experiments with Foul
Release coated disks and observed that at 25 knots the slime layer is strongly reduced in thickness, while at
45 knots the slime layer is not removed completely but very thin.
This may explain why the effect of slime films on the drag of the most effective Foul Release surfaces is
significant but limited to a few percent, as indicated by the disk experiments carried out recently by Haslbeck
et al.61 In other words, Foul Release surfaces rapidly loose their initial drag benefit compared to other
coatings but do not exhibit more drag than Tin free SPCs over the long term.4 The literature review suggests
there is a possibility that at very high speeds where the adhering slime layer will be very thin, small drag
benefits may persist.
Further research is being carried out at the University of Newcastle-upon-Tyne to study this. Flat plates
coated with Foul Release and Tin free SPC schemes will be submerged in seawater for one year and will then
be tested in the Emerson Cavitation Tunnel as described in Section 2. The composition of the fouling
organisms and the speeds at which they release will be verified and the drag, boundary-layer and roughness
characteristics of the Foul Release surface will be compared with the exposed (foul-free) Tin free SPC and
with the data of the coatings when newly applied11. In addition, this would indicate that the added drag would
be less as the service speed of the vessel increases.
Further research is also being carried out to investigate how the drag benefits of newly applied Foul Release
schemes can be exploited. Knowing that there is a likelihood that slime films will release or be very thin at
high speeds, an experimental study investigates the possible benefits of coating propellers. Water tunnel
experiments are underway to compare the efficiency and roughness characteristics of propellers coated with
a Foul Release scheme with uncoated propellers of varying degrees of roughness. In addition, full scale trials
with a coated propeller with an uncoated propeller will compare the power at the shaft.
Acknowledgements
Dr. Candries’ PhD research was funded by EPSRC Research Grant No. 98316366, by International Coatings
Ltd. and by the Engineering Design Centre at the University of Newcastle-upon-Tyne.
References
1. International Maritime Organization (2001), International Convention on Control of Harmful Antifouling
systems on Ships, 25 October 2001, http://www.imo.org.
2. Evans, S.M., Birchenough, A.C. and Brancato, M.S. (2000), The TBT ban: out of the frying pan into the fire?
Marine Pollution Bulletin, Vol. 40, pp. 204-211.
3. Turley, P.A., Fenn, R.J., Ritter, J.C. (2000), Pyrithiones as antifoulants: environmental chemistry and
preliminary risk assessment, Biofouling, Vol. 15, pp. 175-182.
4. Anderson, C.D. and Hunter, J.E. (2000), Whither antifouling paints after TBT? NAV 2000 International
Conference on Ship and Shipping Research, 13th Congress, Paper 3.7. Venice, Italy, 19-22 September 2000.
39
5. Wahl, M. (1989), Marine epibiosis. I: Fouling and antifouling: some basic aspects, Mar. Ecol. Prog. Ser., Vol.
58, pp. 175-189.
6. Brady, R.F. and Singer, I.L. (2000), Mechanical factors favoring release from fouling release coatings.
Biofouling, 15, pp. 73-81.
7. Baier, R.E. (1970), Surface properties influencing biological adhesion. In: Adhesion in Biological Systems,
R. S. Manly (Ed.), Academic Press, pp. 15-48.
8. Meyer, A.E., Baier, R.E. and King, R.W. (1988), Initial fouling of nontoxic coatings in fresh, brackish and
seawater. The Canadian Journal of Chemical Engineering, 66, pp. 55-62.
9. Kovach, B.S. and Swain, G.F. (1998), A boat mounted foil to measure the drag properties of antifouling
coatings applied to static immersion panels. Proceedings of the International Symposium on Seawater Drag
Reduction, pp. 169-173, Newport.
10. Millett, J. and Anderson, C.D. (1997), Fighting fast ferry fouling. Fast ‘97, Conference Papers, Vol. 1, pp.
493-495.
11. Candries, M. (2001), Drag, Boundary-Layer and Roughness Characteristics of Marine Surfaces Coated
with Antifoulings, PhD Thesis, Department of Marine Technology, University of Newcastle-upon-Tyne,
http://www.geocities.com/maxim_candries/.
12. Candries, M., Atlar, M., Guerrero, A. and Anderson, C.D. (2001), Lower frictional resistance characteristics
of Foul Release systems, Proceedings of the Eight International Symposium on the Practical Design of Ships
and Other Floating Structures (PRADS 2001), Volume 1, pp. 517-523, Elsevier, Amsterdam.
13. Candries, M., Atlar, M., Mesbahi, E. and Pazouki, K. (2003), The measurement of drag characteristics of
Tin free Self-Polishing Co-polymers and Foul(ing) Release coatings using a rotor apparatus. 11th International
Congress on Marine Corrosion and Fouling (to be published in a Supplement to Biofouling, February 2003),
San Diego, 21-26 July 2002.
14. Weinell, C.E., Olsen, K., Christoffersen, N., Martin, W., Kiil, S. (2002), Experimental study of drag
resistance on sea water immersed surfaces using a laboratory scale rotary set-up. 11th International
Congress on Marine Corrosion and Fouling, (to be published in a Supplement to Biofouling, February 2003)
San Diego, 21-26 July 2002.
15. Gad-el-Hak, M. and Bandyopadhyay, P.R. (1994), Reynolds number effects in wall bounded turbulent
flows. Applied Mechanics Review, Vol. 49, pp. 307-365.
16. Grigson, C.W.B. (1982), The drag coefficients of a range of ship surfaces II. Trans. R.I.N.A., Vol. 124,
pp. 183-198.
17. Musker, A.J. (1977), Turbulent shear flows near irregularly rough surfaces with particular reference to
ships’ hulls, PhD Thesis, University of Liverpool, Liverpool.
18. Townsin, R.L and Dey, S.K.(1990), The correlation of roughness drag with surface characteristics.
International Workshop on Marine Roughness and Drag, R.I.N.A., London.
19. I.T.T.C. (1990), Report of the Powering Performance Committee, Proceedings of the 19th International
Towing Tank Conference, pp. 235-287. Madrid.
20. Chuah, K.B., Dey, S.K., Thomas, T.R. and Townsin, R.L. (1990), A digital hull roughness analyser.
International Workshop on Marine Roughness and Drag, Paper No. 3. R.I.N.A., London.
21. Fitzsimmons, P.A. and Ellis, J. (1990), A two-parameter roughness analyser and performance monitoring
software. International Workshop on Marine Roughness and Drag, Paper No. 4. R.I.N.A., London.
22.Townsin, R.L. (2003), The ship hull fouling penalty. 11th International Congress on Marine Corrosion and
Fouling (to be published in a Supplement to Biofouling, February 2003), San Diego, 21-26 July 2002.
23. Egan, B. (1987), Marine microbial adhesion and its consequences. In: Microbes in the Sea, M. A. Sleigh
(Ed.), Ellis Horwood Ltd.
24. Fletcher, R.L. and Chamberlain, A.H.L. (1975), Marine fouling algae. In: Microbial Aspects of the
Deterioration of Materials, R. J. Gilbert and D. W. Lovelock (Ed.), pp. 59-81, London, Academic Press Inc.
25. Fischer, E.C., Castelli, V.J., Rodgers, S.D. and Bleile, H.R. (1984), Technology for control of marine
biofouling - a review. In: Marine Biodeterioration: An Interdisciplinary Study, J. D. Gostlow and R. C. Tipper
(Ed.), Maryland, USA, Naval Institute Press.
26. Palmer, R.J. and White, D.C. (1997), Developmental biology of biofilms: implications for treatment and
control. Trends in Microbiology, Vol. 5, pp. 435-440.
40
27. Swain G., Anil, A.C., Baier, R.E., Chia, F.-S., Conte E., Cook, A., Hadfield, M., Haslbeck, E., Holm, E.,
Kavangh, C., Kohrs, D., Kovach, B., Lee, C., Mazzella, L., Meyer A.E., Qian, P.-Y., Sawant, S.S., Schultz, M.,
Sigurdsson, J., Smith, C., Soo, L., Terlizzi, A., Wagh, A., Zimmerman, R. and Zupo, V. (2000), Biofouling and
barnacle adhesion data for fouling release coatings subjected to static immersion at seven marine sites,
Biofouling, Vol. 16, pp. 331-344.
28. McEntee, W. (1915), Variation of frictional resistance of ships with condition of wetted surface. Trans.
S.N.A.M.E., Vol. 24, pp. 37-42.
29. Woods Hole Oceanographic Institution (1952). Marine Fouling and its Prevention. Annapolis, Maryland,
US Naval Institute.
30. Denny, M.E. (1951), BSRA resistance experiments on the ‘Lucy Ashton’: Part 1. Full scale measurements.
Trans. R.I.N.A., Vol. 93, pp. 40-57.
31. Watanabe, S., Nagamatsu, N., Yokoo, K. and Kawakami, Y. (1969), The augmentation in frictional
resistance due to slime (B.S.R.A. Translation No. 3454). Journal of the Kansai Society of Naval Architects,
Vol. 131, pp. 45.
32. Todd, F.H. (1951), Skin friction resistance and the effects of surface roughness. Trans. S.N.A.M.E.,
Vol. 59, pp. 315-374.
33. Loeb, G.I., Laster, D. and Gracik, T. (1984), The influence of microbial fouling films on hydrodynamic drag
of rotating disks. In: Marine Biodeterioration: An Interdisciplinary Study, J. D. Gostlow and R. C. Tipper (Ed.),
Maryland, USA, Naval Institute Press.
34. Lewkowicz, A.K. and Das, P.K. (1986), Turbulent boundary layers on rough surfaces with and without a
pliable overlayer: a simulation of marine fouling, International Shipbuilding Progress, Vol. 33, pp. 174-186.
35. El-Labbad, A.F. (1987). Techno-economic Analysis of the Problems of the Bottom Maintenance of Ships,
MPhil Thesis, Department of Marine Technology, University of Newcastle-upon-Tyne, Newcastle-upon-Tyne.
36. Picologlou, B.F., Zelver, N. and Characklis, W.G. (1980), Biofilm growth and hydraulic performance,
Journal of the Hydraulics Division, Proceedings of the American Society of Civil Engineers, Vol. 106, pp.
733-746.
37. Stoodley, P., Boyle, J., Cunningham A.B., Dodds, I., Lappin-Scott, H.M. and Lewandowski, Z. (1999),
Biofilm structure and influence on biofouling under laminar and turbulent flows. In: Keevil, C.W., Godfree, A.,
Holt, D., Dow, C. (Eds), Biofilms in the Aquatic Environment, The Royal Society of Chemistry, Cambridge, UK,
pp. 13-24.
38. Stoodley, P., Boyle, J.D., deBeer, D. and Lappin-Scott, H.M. (1999), Evolving perspectives of biofilm
structure, Biofouling, Vol. 14, pp. 75-90.
39. Schultz, M.P. and Swain, G.W. (1999), The effect of biofilms on turbulent boundary layers. Journal of
Fluids Engineering, Vol. 121, pp. 44-51.
40. Lewthwaite, J.C., Molland, A.F. and Thomas, K.W. (1985), An investigation into the variation of ship skin
frictional resistance with fouling. Trans. R.I.N.A., Vol. 127, pp. 269-284.
41. Bohlander, G.S. (1991), Biofilm effects on drag: measurements on ships. Polymers in a Marine
Environment. The Institute of Marine Engineers Third International Conference, Paper 16. London.
42. Hoyt, J.W. and Soli, G. (1965), Algal cultures: ability to reduce turbulent friction in flow, Science, Vol. 149,
pp. 1509.
43. Barnaby, K.C. and Dorey, A.L. (1965), A towing tank storm, Trans. R.I.N.A., Vol. 107, pp. 265-272.
44. Emerson, A. (1965), Model experiments using dilute polymer solutions instead of water, Trans.
N.E.C.I.E.S., Vol. 81, pp. 201-212.
45. Hawkridge, H.R.J. and Gadd, G.E. (1971), Investigation of drag reduction by certain algae, Nature, Vol.
230, pp. 253-255.
46. Rosen, M.W. and Cornford, N.E. (1971), Fluid friction of fish slimes, Nature, Vol. 234, pp. 49-51.
47. Hoyt, J. W. (1990), Drag reduction by polymers and surfactants. In: Viscous Drag Reduction in Boundary
Layers, D. M. Bushnell and J. N. Hefner (Eds.), American Institute of Aeronautics and Astronautics, pp. 413-
432, Washington, D.C.
48. Hammer, F, Hellsten, M. and Uneback, I. (1998), 10 years of experience with demonstration of drag
reduction in the district heating system of Herning, Denmark. Proceedings of the 11th European Drag
Reduction Working Meeting, Chara, Z. and Pollert, J. (Eds.), Institute of Hydrodynamics, pp. 21-22, Prague.
41
49. Canham, H.J.S., Catchpole, J.P. and Long, R.F. (1971), Boundary layer additives to reduce ship
resistance, The Naval Architect, July 1971, pp. 187-213.
50. Kowalski, T. (1998). Practical applications of dilute polymer additives for water craft. Proceedings of the
International Symposium on Seawater Drag Reduction, pp. 289-293, Newport.
51. American Standard for Testing and Materials (2000), Standard test method for measurement of barnacle
adhesion strength in shear, D5618-94, http://www.astm.org/
52. Terlizzi, A., Conte, E., Zupo, V., and Mazzella, L. (2000), Biological succession on silicone fouling-release
surfaces: long-term exposure tests in the harbour of Ischia, Italy. Biofouling, Vol. 15, pp. 327-342.
53. Denny, M. (1989), A limpet shell shape that reduces drag: laboratory demonstration of a hydrodynamic
mechanism and an explanation of its effectiveness in nature, Canadian Journal of Zoology, Vol. 67, pp. 2098-
2106.
54. Walderhaug, H. (1986), Paint roughness effects on skin friction. International Shipbuilding Progress, Vol.
33, pp. 96-100.
55. White, F.M. (1994) Fluid Mechanics, 3rd Edition, McGraw-Hill, New York, 736pp.
56. Manen, J.D.v. and Oossanen, P.v. (1988), Resistance. In: Principles of Naval Architecture, Second
Revision, E. V. Lewis (Ed.), S.N.A.M.E., Vol. 2, Chapter 5, p. 58.
57. Schultz, M.P., Finlay, J.A., Callow, M.E. and Callow, J.A. (2002), Three models to relate detachment of low
form fouling at laboratory and ship scale, 11th International Congress on Marine Corrosion and Fouling (to
be published in a Supplement to Biofouling, February 2003), San Diego, 21-26 July 2002.
58. Morgenroth, E. and Wilderer, P.A. (2000), Influence of detachment mechanisms on competition in biofilms,
Water Research, Vol. 34, pp. 417-426.
59. Van Loosdrecht, M.C.M., Eikelboom, D., Gjaltema, A., Mulder, A., Tijhuis, L. and Heijnen, J.J.(1995),
Biofilm structures, Water Science Technology, Vol. 32, No. 8, pp. 35-43.
60. Klijnstra, J.W., Overbeke, K., Sonke, H., Head, R. and Ferrari, G.M. (2002), Critical speeds for fouling
removal from a silicone coating. 11th International Congress on Marine Corrosion and Fouling, San Diego,
21-26 July 2002.
61. Haslbeck, E., Holm, E.R., Talbott, W.J., Field, A. (2002), Evaluation of hydrodynamic drag on antifouling
and Fouling Release coatings using the friction disk machine. 11th International Congress on Marine
Corrosion and Fouling, San Diego, 21-26 July 2002.
Figures and Tables Figure 1. Polishing rates of biocidal antifoulings.
42
Figure 2. Schematic set-up for the LDV boundary-layer experiments.
Figure 3. Boundary-layer velocity profiles in inner co-ordinates at Ue = 5m/s and at a streamwise

location x = 1.607m from the leading edge. A rollered and a sprayed Foul Release surface were tested
to investigate the effect of application method. A surface covered with sand grit was tested in order
to have a very rough comparison. The velocity loss function ∆U+ indicates the difference in frictional
resistance between a rough and a smooth surface. (Experimental precision uncertainty over the log-
law region: U+: ±1.72% for the uncoated steel surface, ±1.94% for the rough surfaces; ∆U+: ±14.74%).
43
Table 1. Overview of the drag characteristics
Towing tank experiments ∆CF ∆U+ Average Roughness

(compared to reference, %) (on average) (µm)
2.55m long plate 2.0·106 < Re < 4.2·106
Sprayed Foul Release 3.9 0.20 44

Sprayed SPC 23.4 2.17 75
6.3m long plate 2.0·107 < Re < 4.0·107

Rotor experiments ∆CF ∆U+ Average Roughness

Cylinder 1.0·106 < Re < 2.1·106

Rollered Foul Release 5.7 1.31 218
Water tunnel experiments ∆CF ∆U+ Average Roughness

1m long vertical plate 8.5·103 < Red1 < 3.4·104

(Emerson Cavitation
Tunnel)

Rollered Foul Release 13.1 1.54 60
Sprayed SPC 16 1.80 69
1m long horizontal plate 1.6·104 < Red1 < 4.6·104

(CEHIPAR Cavitation
Tunnel)

44
Figure 4. Two typical roughness profiles of (from bottom to top respectively) a Foul Release scheme
applied by spraying, a Tin free SPC scheme applied by spraying and a Foul Release scheme applied
by rollering. The horizontal gridlines are separated by 25µm.
Figure 5. Typical roughness measurement of a sample (sprayed) Foul Release surface.
45
University
Figure 6. Typical roughness measurement of a sample Tin free SPC surface.
Figure 7. Approximate wall shear stress against free stream velocity for three different lengths (Full
line using Eq. 3, dashed line using Eq. 4). If the release of fouling organisms within the boundary layer
on a 1m long foil is observed at a free stream velocity of 6.2m/s as in the experiments of Kovach and
Swain9, the wall shear stress in the boundary layer calculated with Eq. 3 would approximately be equal
to 55Pa. For a 90m long high speed ferry and a 250m long tanker, this wall shear stress would only be
reached if the free stream velocity is equal to 8.6 and 9.2m/s respectively. For the high speed ferry the
service speed lies well above this velocity; for the tanker this is not necessarily the case.
46
Notes
47
Notes
48
Powering
of Hulls
- a simple
discussion (2000)
Eugene Griessel, Noordhoek, South Africa.
Powering of Hulls - a simple discussion
Eugene Griessel
Noordhoek, South Africa
How fast will that hull go and how much power is needed to make it go that fast?
Easy questions - very difficult to answer exactly.
There is no such thing as a “top speed” for any given hull/powerplant combination. Many factors affect the
speed of any hull and only by stating them can an absolute “top speed” for those conditions be given. Some
factors affecting the speed are:
• Salinity/Density of water in which the hull is immersed
• Water temperature
• Roughness of hull - and the condition of it with respect to marine growth, paint used, etc.
• Air temperature and humidity
• Sea state.
• The “bend” of the hull caused by uneven distribution of internal weight.
• Displacement of hull
• The “wetted area” of the hull - affected by the length, displacement etc.
Any statement about the speed of a hull which does not define most of the above conditions is pretty
valueless. A ship in fresh water at 30˚C with an air temperature of 35˚C at light load will perform differently to
one in seawater at 8˚C with an ambient temperature of 11˚C while fully laden.
There are basically four forces that have to be overcome to move a hull through water. These are:
• Frictional resistance
• Eddy-making resistance
• Wave-making resistance
• Air resistance
In practice Air Resistance is quite minimal - accounting for a few percent of the total resistance, usually less
than 5%. However the faster a hull is moved the greater this resistance becomes.
Eddy-making resistance is quite small and in a well designed hull is negligible. It is quite often combined with
Wave-making resistance.
Wave-making resistance is extremely important and it increases dramatically with speed. This is the power
the ship expends making all those nice foamy wakes etc.
Frictional resistance is that part of the resistance created by the effect of the hull “rubbing” against the fluid
it is moving through.
contains.
© Akzo Nobel, 2003
51
Some formulas
Froude stated that frictional resistance is dependent on the following four factors:
• Area of the surface
• Type of surface
• Length of surface
• Density of fluid
Froude derived a formula : Rf=FSV1.825 that allows a close approximation of the frictional resistance to be
made. Where:
• Rf = frictional resistance in Newtons
• F = a constant dependent on length
• S = wetted surface area in square metres
• V = speed of ship in metres per second.
The values for F in seawater are:
Length in metres F Length in metres F

5 1.736 120 1.421
10 1.604 140 1.415
20 1.515 160 1.410
40 1.464 180 1.404
60 1.457 200 1.399
80 1.437 250 1.389
100 1.428 300 1.386
To derive the wetted area I have elected to use Taylor’s formula S=√C∆L because we are primarily interested
in fast hulls.
Where:
• S = moulded wetted surface area in square metres
• C = constant which differs with form but is usually between 2.56 and 2.59
• ∆ = displacement in tonnes
• L = mean immersed length in metres but usually taken as waterline length.
For slower hulls of a more fuller form and slower speed Denny’s formula can be used instead.
∆
S=1.7LT+
T
Where:
• S = moulded wetted surface area in square metres
• T = moulded mean draught
∆
• = volume of displacement in cubic metres
• L = length between perpendiculars.
Example
We have the following parameters - calculate frictional resistance:
• Speed = 30 knots or 15.432m/s
• Displacement = 90000 tonnes
• Waterline length = 312 metres
52
Wetted area will be (Taylor) 2.58 x square root of 90000 x 312 which equals 13671.6 square metres. Call it
13671 to make calculations easier.
Frictional resistance will be (Froude) 1.386 x 13671 x 15.432 to the power 1.825 or 2795336.7Newtons.
Call it 2795kiloNewtons (kN) for simplicity.
To obtain the effective horsepower to overcome this component of the resisting forces we simply multiply it
by the speed in metres per second to get kiloWatts (kW). Thus 2795 x 15.432 = 43132kW or 57842
horsepower. 1 horsepower is equal to 0.7457kW.
Effective horsepower is not shaft horsepower but that power delivered by the screws. There is a loss in
efficiency between SHP and Effective horsepower which varies according to propeller design and speed.
Having derived the formula for this hull it becomes very easy to draw up a table for the hull at various speeds.
SPEED HORSEPOWER
5 366
10 2595
15 8160
20 18393
25 34548
30 57842
35 89379
40 130335
As we can see the frictional component rises sharply as speeds increase - which was to be expected.
Of the four components of total resistance we now have derived one - frictional resistance. Frictional
resistance plays a large role in the total resistance but this gets less and less as the speed increases. At
higher speeds wave-making resistance becomes the major component in the total resistance a hull offers to
forward motion.
Wave-making resistance
When a body moves through a fluid it displaces that fluid and the displacement takes energy to effect. The
calculation of the amount of resistance that wave-making causes is fraught with difficulty. It is determined by
towing tank observations or complex modelling programs run on fast computers. We are trying to keep it
simple - so in fact what we are looking for is a simple rule-of-thumb type formula that will give us a ballpark
figure.
We know, from experience with existing ships, that the residual resistance at 35 knots averages about 60%
of the total offered and at 40 knots it is approaching 70%. Thus in our previous examples of 35 knots taking
89379 effective horsepower to overcome frictional resistance and 40 knots requiring 130335 we could expect
a total of around 223447 and 434450 effective horsepower respectively. If we take a propeller efficiency of
around 80% then we are looking at perhaps 279308 shp and 543062 shp respectively to move our example
hull at these speeds. This ties in fairly well with known hull/power/speed combinations.
53
The graph:
Is for an actual cargo vessel. It shows the wave-making resistance at various speeds expressed as a force
per tonne. Thus at about 20 knots, this particular vessel’s top cruising speed, one could expect the wave
making resistance to equal 25 Newton per tonne, or, as she was a 12500 tonne vessel, a total of 312.5kN.
Notice how rapidly the curve rises so that by 30 knots the drag now equates to 937.5kN. The power to
overcome these two forces would be 4300 horsepower and 19400 horsepower respectively. Using the earlier
formulae to calculate frictional resistance we find that at 20 knots this would be 372kN and at 30 knots
780kN. And the effective power needed to overcome that would be in the region of 5100 hp and 16100 hp
respectively. Or totals of 9400 hp and 35500 hp. Adding air and eddy-making resistance as an arbitrary 5%
we would now have 9870 and 37275 effective horsepower. Given an arbitrary propeller efficiency of 80% then
the shaft horsepower needed will be 12340 and 46600 respectively. Thus to achieve a speed advantage of
10 knots an extra 35000 shp is needed. This particular vessel, a fast reefer, was fitted with a 16000
horsepower MAN diesel engine and achieved a speed slightly greater than 22 knots on her measured mile
trials with a clean hull, standard displacement, ambient sea temperature of 12˚C and an ambient air
temperature of 19˚C. And our calculations have shown that calculated and actual powering were near enough
each other to be confident that we could predict this within a few thousand horsepower - knowing little more
of the hull than its tonnage and length.
Comparison
One of Froude’s earliest observations was that when two similarly shaped bodies of differing lengths are
moved through water they both produce the same pattern of waves but at differing speeds. Thus, for
example, a 5.5 metre hull will produce the same pattern of waves at 9 knots that a 22 metre hull will produce
at 18 knots. This was an enormously important discovery as it produced a formula that allowed tank testing
of models to ascertain their wave making resistance and then have that scaled up to a full sized ship. For the
very first time the design of hulls was put on a scientific footing and engineers could, with a great degree of
confidence, guarantee the speed/power combinations of given hulls. Today complicated computer-driven
modelling programs can eliminate much of the physical tank testing which saves much time and money
during initial design work.
All of this brings us to that old rule-of-thumb formula:
Hull speed in knots = 1.3 x sqrroot(waterline length in feet).
I want to discuss this at some length. The actual formula, as popularised by a book on Yacht design
published in the early sixties, is:
Hull speed in knots = C x sqrroot(waterline length in feet)
where C is a constant varying with form. Yachtsman usually take it as 1.34 but it varies from about 1.18 for
fine hulls to 1.42 for chunkier ones. It only pertains to displacement hulls and gives an approximation of the
maximum speed a hull can achieve before the bow wave combines with the stern wave to dig a trough out
of which the vessel would not be able to climb without using inordinate amounts of power. The formula needs
to be used circumspectly and applied to hulls within reasonable limits. It cannot be used indiscriminately to
prove absolute values for any hulls.
54
The absurdity becomes apparent when a hull of infinite length is inserted into the formula. It is not self
limiting. Thus a hull of a few millimetres could almost not move while an infinite one could move infinitely fast.
In fact the formula is not much use above about 200 feet and is in fact best utilised on small sailing dinghies.
In fact the ocean literally teems with vessels exceeding their “hull speed” according to this formula.
A formula that used to be used (and may still be for all I know) in preliminary design work to determine the
length of a ship to avoid the problem of “hull speed” was:
L = C (V/V+2)2∆ ⁄
1
3
Where:
• L = length, in metres, between perpendiculars
• V = speed in knots
• ∆ = displacement in tonnes
• C = 7.12 for slow ships
• C = 7.28 for medium speed ships
• C = 7.88 for fast ships
However this is not a very scientific way of going about finding some easy method to calculate an
approximate maximum speed of a given hull/powerplant combination. The greatest stumbling block is the
determination of wave-making resistance by a reliable yet easy method - so if anyone out there has some
rule-of-thumb type method of approximating the wave-making resistance of a hull, I invite them to add their
voice to this by e-mailing me at eugene@dynagen.co.za.
55
Notes
56
A New
Hull Roughness
Penalty Calculator
(2003)
C O’Leary, C D Anderson,
23 23
A New Hull Roughness Penalty Calculator
C. O’Leary, C. D Anderson
Introduction
The economic importance of underwater hull condition cannot be understated. Any increase in underwater
hull roughness can result in a significant rise in vessel operating costs.
There are two main types of hull roughness: physical and biological (fouling), each with their own macro (large
scale) and micro (small scale) physical characteristics.
• Macro physical roughness can be attributable to plate waviness, plate laps, welds and weld quality,
mechanical damage and corrosion.
• Macro biological roughness is typically attributable to animal and weed fouling.
• Micro physical roughness can be attributable to steel profile, minor corrosion and coating surface.
• Micro biological roughness is typically attributable to slime fouling.
Physical Roughness Biological Roughness
Macro
Micro Macro Micro
Welds Plate Waviness

Minor Corrosion Steel Profile Animal Fouling Weed Fouling
Slime Fouling
Corrosion Plate Laps
Coating Condition
Mechanical Damage
Examples of how coatings condition can influence hull roughness:
Cracking Detachment Touch-up Repairs

Any increase in underwater hull roughness will increase hull frictional resistance or vessel drag, resulting in
an additional power requirement with increased fuel consumption and cost to maintain vessel speed.
Conversely, maintaining constant power will result in decreased vessel speed and longer voyage times.
Whilst this may appear obvious, the IMO Convention on the Control of Harmful Antifoulings on Ships and the
subsequent debate on the performance merits of TBT versus TBT free antifoulings has demonstrated that
the principles, performance effects and costs of increased hull roughness on vessel operating efficiency may
not be widely understood.
contains.
© Akzo Nobel, 2003
59
New Model
International Marine Coatings have developed a new ‘Hull Roughness Penalty Calculator’ model. This is a
software programme that predicts the inevitable increase in underwater hull roughness during the specified
in service period and combines this with the risk of fouling associated with different antifouling types. The
model compares fuel usage and cost to the installation cost of different TBT free antifouling and foul release
systems to derive potential nett benefit.
The model is also able to compare the exhaust emissions (CO2, SOx) associated with the additional fuel
consumption for a particular vessel.
How Surface Roughness is Measured
Hull roughness on ships is measured as the maximum peak to lowest trough height (Rt50) expressed in
microns, in any given length of 50mm along the underwater hull.
The only equipment for measuring Rt50 values is the Hull Roughness Analyser (or gauge), supplied by BMT
(British Maritime Technology), Wallsend, UK. At each location the surface probe is manually run over a
distance of 750-1000mm to generate ~13-15 Rt50 readings, the average of which is the mean hull roughness
at that location.
750-1000mm
Surface of the coating on the hull

At least 100 locations distributed around the underwater hull (including sides and flats) are required for
statistical validity. The average of all these mean hull roughnesses gives the Average Hull Roughness (AHR).
The graph below shows how the distribution of readings typically appears.
Hull Roughness Histogram

Frequency
Hull Roughness Gauge in use
The Effect of Coating Roughness on Ship Performance

The effect of coating roughness on ship performance can be calculated using the Townsin1 formulae:
Fractional Added Resistance (∆R/R) for going from a smooth (AHR = k1) to a rough (AHR = k2) surface:
∆R/R = ∆CF/CT = 0.044[(k2/L)1/3 – (k1/L)1/3]/CT
Where:
∆ = Change in resistance, power, speed or propeller efficiency due to increased roughness
∆CF = Frictional Resistance coefficient increase
2
CT = Total Resistance coefficient = ([Total Resistance]/0.5 ρ S V )
or very approx. = 0.018 L-1/3 (if CT value cannot be found otherwise, and where L is in metres)
ρ = Sea water density
S = Surface wetted area of vessel
V = Speed of vessel
L = Length between perpendiculars of vessel
60
Fractional Power increase (∆P/P) at constant speed for going from a smooth (AHR=k1) to rough
(AHR=k2) surface:
1+ ∆P/P = (1 + ∆R/R) (1+∆η/η)-1
Where:
P = Shaft Power
η = Open water propeller efficiency
As a handy guide, the following approximate relationships hold for a Ro-Ro ship and a Tanker, which typify
Liner and Bulk Cargo ships:
For Ro-Ro ships: (1+∆η/η)-1 = 0.17 (1 + ∆R/R) + 0.83
For Tankers: (1+∆η/η)-1 = 0.30 (1 + ∆R/R) + 0.70
Figure 1 below shows the increase in power required and hence the typical increase in fuel consumption
necessary to maintain vessel speed of a fast fine ship (e.g. Container Liner) versus increasing physical hull
roughness.
Figure 1: Typical increase in power/fuel required to maintain vessel speed of a fast fine ship vs
increasing hull roughness
Note: Above 225 microns (which is undesirably rough) calculations are less precise, hence the dotted line.
How Roughness is Affected by Antifouling Type
During the period 1976 – 1986, two substantial hull roughness studies were carried out2. These studies
showed that over time, ships generally get rougher due to mechanical damage from anchor chains, tugs,
grounding, berthing, etc. and from mechanical damage, cracking, blistering, detachment, corrosion etc. of
applied surface coatings. The increase in roughness was found to differ markedly depending on which
antifouling type was used. With traditional antifoulings the increase in Average Hull Roughness (AHR) over
time was found to be 40 microns per year, with part of this increase resulting from the reasons mentioned
earlier and part resulting from maintenance painting at each drydocking (assuming no reblasting). Fouling
was removed prior to measurement of roughness.
61
For Self Polishing Copolymer (SPC) antifoulings, the average increase was found to be significantly less, at
20 microns increase in AHR per year. This reduction is a result of the polishing and smoothing action of SPC
antifoulings.
Since 1986, the traditional antifoulings have been modified with reinforcing resins and are now generally
referred to as “Controlled Depletion Polymer” (CDP) antifoulings (Interspeed® 340). The AHR increase of
these coatings is also estimated at 40 microns/year.
The three additional (to CDP), International fouling control technologies may be characterised as follows:
- Foul Release technology (Intersleek® 425 and Intersleek® 700). These products do not use
biocides to control fouling but rely on a slippery “non-stick” surface to make it difficult for fouling
species to adhere. Foul release systems provide a very smooth surface and because they are relatively
expensive to install, ship owners and operators are generally very careful to avoid damage with vessels
coated with these systems. AHR increase therefore is assumed to be only 5 microns per year.
- Patented TBT Free Copper Acrylate SPC technology (Intersmooth® Ecoloflex SPC). Proven in
service on over 6,000 vessels worldwide and proven as providing an equivalent performance level to
that achieved by premium grade TBT SPC products. AHR increase is estimated at 20 microns per year,
the same as TBT SPC products.
- Hybrid TBT Free self polishing technology (Interswift® 655). This technology offers a balance of
CDP type and SPC type antifouling properties with performance and AHR increase assumed to be
midway between the two at 30 microns per year.
The International Marine Coatings Hull Roughness Penalty Calculator estimates the increase in power
required over time for the four main antifouling technologies outlined based on their average increase in
physical hull roughness per year. The initial roughness is taken as 120 microns which is the approximate
roughness value for a typical newbuilding. Figure 2 shows the result:
62
Figure 2: Estimated increase in power needed/fuel used for a typical fast fine ship (e.g. Container
Liner) due to increasing physical hull roughness of different antifouling types over time
The Risk and Effect of Fouling of TBT Free Antifoulings and Foul Release Systems
Fouling is a biological phenomenon whose occurrence is difficult to predict and control. The type, severity
and extent to which fouling occurs varies greatly depending on the type of antifouling coating plus the
vessels’ trading pattern and operational profile (i.e. vessel speed and activity). Only by studying a large
number of vessels over extended time periods can statistically reliable information be obtained.
Slime: 1~2% increase Weed: up to 10% increase Shell: up to 40% increase

in vessel drag in vessel drag in vessel drag
International monitors the performance of all of its antifouling coatings through a system called Dataplan.
Since 1977, Dataplan has recorded the details of the underwater hull condition of over 70,000 ships. When
a ship enters drydock, the type, severity and extent of fouling on the underwater sides, boottop and flat
bottom are recorded. From this information, Dataplan can calculate an overall “Fouling Rating” on a scale
from 0 (completely clean) to 100 (completely fouled). Different weightings are assigned depending on fouling
type. By having a sufficiently large sample number of vessels, the average fouling rating over time, which is
a measure of the risk of fouling for different antifouling types, can then be assessed with a high degree of
confidence.
If 100% green grass attaches to the vertical sides of a vessel (Dataplan Fouling Rating = 100*) then the typical
experience of ship operators shows that at least a 10% fuel penalty can occur. Using this equivalence, the
% increase in fuel consumption over time due to fouling on different antifouling systems can be calculated
based on their average Dataplan Fouling Ratings (a Dataplan Fouling Rating of 10 or less denotes
satisfactory performance).
Figure 3 shows the increase in fuel consumption of two identical container vessels, one coated with a CDP
antifouling and the other with an SPC antifouling. Initially, little difference between the systems was observed
but after 18 months the CDP coated vessel picked up weed fouling on part of the vertical sides and there
was an associated, dramatic increase in fuel consumption.
63
Figure 3: Fuel Consumption of Identical Container Vessels
CDP SPC
8
% Increase in Fuel Consumption
0
10 20 30 40 50 60
-2
months
The Combined Effects of Physical Roughness and Risk of Fouling on Ship Power/Fuel Requirement
The combined effects of increased physical roughness and the risk of fouling for the different antifouling
types on the power/fuel required to maintain vessel speed is shown in Figures 4 and 5.
Notes:
1) There is usually less weed fouling on the flat bottom of a vessel because the light intensity is reduced,
hence the power/fuel penalty is lower on the flat bottom than on the vertical sides.
2) Since CDP and Hybrid antifoulings are not designed for more than 36 months in-service on the vertical
sides of a vessel the power/fuel penalty for these products rises sharply after 36 months.
3) *A Dataplan fouling rating of 10 or less denotes satisfactory performance.
Figure 4: Overall % Power/Fuel increase for a typical fast fine ship (e.g. Container Liner) vs Time for
Different Antifouling Types - Vertical Sides
CDP
Hybrid
SPC
Foul Release
Months
64
Figure 5: Overall % Power/Fuel increase for a typical fast fine ship (e.g. Container Liner) vs Time for
Different Antifouling Types - Flat Bottom
CDP
Hybrid
SPC
Foul Release
Months
Vessel Fuel Consumption

The “baseline” annual fuel consumption for a particular vessel can be calculated from the daily fuel usage
(tonnes/day) multiplied by the number of days spent at sea per year. Multiplying this figure by (1 + the average
annual % extra) fuel usage shown in Figure 4 (vertical sides) and Figure 5 (flat bottom), and then cumulatively
totalling these for the specified in service period, gives the total fuel consumption (tonnes) for each antifouling
type. The total cumulative fuel cost is then obtained by multiplying this by the price per tonne of the fuel. The
Hull Roughness Penalty Calculator predicts the inevitable increase in underwater hull roughness during the
specified in service period and combines this with the risk of fouling associated with different antifouling
types. The model compares fuel usage and cost to the installation cost of different TBT free antifouling and
foul release systems to derive potential nett benefit.
The introductory screens to the Hull Roughness Penalty Calculator are shown overleaf:
65
66
Container - 110k Post Panamax
273
87,360
Fuel per year 13,977,600 USD

Fuel per 60 months 69,888,000 USD
CO2/SOx Calculator
Extra Fuel Extra CO2 Extra SOx

Over 60 months Over 60 months Over 60 months
947,713 USD 1,492,531 Te. 27,985 Te.
2,059,671 USD 1,603,727 Te. 30,070 Te.
1,111,958 USD 111,196 Te. 2,085 Te.
Potential nett benefit for 60 months 931,893
The information contained in this summary is illustrative only and based on a number of assumptions outlined to you during the presentation.
Your International Marine Coatings contact will be happy to provide you with a list of those assumptions.
References:
1. Dr. R.L. Townsin “Workshop – Calculating the Cost of Marine Surface Roughness on Ship Performance”
WEGEMT School on Marine Coatings at the University of Plymouth, UK, 10-14 July, 2000.
2. Townsin, R.L., Byrne, D., Svensen, T.E. and Milne, A. “Fuel Economy due to Improvements in Ships Hull
Surface Condition 1976-1986", International Shipbuilding Progress”, 33, (383), July 1986.
67
Notes:
68

03 Hull Roughness

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

03 Hull Roughness

Uploaded by

Copyright:

Available Formats

HULL ROUGHNESS

The Ship Hull

© Akzo Nobel, 2003

3. Shell and weed

where the friction velocity uτ = √(το/ρ)

and το = the wall shear stress

with ρ = fluid density.

1000∆CF = 44 [(AHR/L)1/3 – 10Rn-1/3 ] + 0.125

‘Garfield’, Whitley Road, Benton, Newcastle upon Tyne NE12 8BP, UK

1. Sources and causes of antifouled hull surface roughness.

© Akzo Nobel, 2003

2. The definition and measurement of hull roughness.

3. Roughness and added drag.

4. Formulae for the calculation of added resistance.

Telfer also advocated a smooth ship extrapolater:

Subtracting the smooth from the rough yields:

5. Formulae for the calculation of added power and fuel consumption.

6. Power increase ∆P/P at constant speed.

7. Speed loss at constant power.

Rough and smooth power curves.

V1 power for rough ship = kV1n+1 + (∆P)1

8. Determination of the speed index n.

and hence n+1 = log P2/P1 (log V2/V1)-1

10. Economics issues.

© Akzo Nobel, 2003

3. The effect of slime on ship performance

4. Discussion and topics for further research

Figures and Tables Figure 1. Polishing rates of biocidal antifoulings.

Figure 3. Boundary-layer velocity profiles in inner co-ordinates at Ue = 5m/s and at a streamwise

Towing tank experiments ∆CF ∆U+ Average Roughness

2.55m long plate 2.0·106 < Re < 4.2·106

Sprayed Foul Release 3.9 0.20 44

6.3m long plate 2.0·107 < Re < 4.0·107

Sprayed Foul Release 2.1 0.21 62

Rotor experiments ∆CF ∆U+ Average Roughness

Cylinder 1.0·106 < Re < 2.1·106

Sprayed Foul Release 4.3 1.00 108

Water tunnel experiments ∆CF ∆U+ Average Roughness

1m long vertical plate 8.5·103 < Red1 < 3.4·104

Sprayed Foul Release 10.9 1.25 51

1m long horizontal plate 1.6·104 < Red1 < 4.6·104

Sprayed Foul Release 14.6 1.68 50

Figure 5. Typical roughness measurement of a sample (sprayed) Foul Release surface.

Figure 6. Typical roughness measurement of a sample Tin free SPC surface.

© Akzo Nobel, 2003

Length in metres F Length in metres F

Welds Plate Waviness

Examples of how coatings condition can influence hull roughness:

Cracking Detachment Touch-up Repairs

© Akzo Nobel, 2003

Surface of the coating on the hull

Hull Roughness Histogram

Hull Roughness Gauge in use

The Effect of Coating Roughness on Ship Performance

Slime: 1~2% increase Weed: up to 10% increase Shell: up to 40% increase

Vessel Fuel Consumption

Fuel per year 13,977,600 USD

Extra Fuel Extra CO2 Extra SOx

Potential nett benefit for 60 months 931,893

You might also like