You are on page 1of 6

A review of the use of CGT for shipbuilding performance

measurement.
Author Name(s): George J Bruce, Member

Downloaded from http://onepetro.org/SNAMESMC/proceedings-pdf/SMC05/3-SMC05/D031S005R009/2453196/sname-smc-2005-p13.pdf/1 by COTECMAR, Yeslis Guerra Ramirez on 23 May 2023


ABSTRACT
The need for a universal measure of shipbuilding performance has been recognized for many years. To meet the need, the concept of
Compensated Gross Tonnes was developed within the OECD, initially as a means of overall comparisons between regions and perhaps
nations. Over time, the use has been extended to compare individual companies and also, with a cost element, to compare ship
construction costs. More recent work has looked at extending the cgt concept to warship construction.

The extended use of cgt as a measure is accepted as less than perfect, but no generally agreed alternative has been proposed. This
paper reviews the uses of cgt, highlights some of the uses in measurement of shipyard performance, and proposes alternative means of
presenting the cgt information which are intended to improve its usefulness.

KEY WORDS: Productivity, Compensation Factor, Cost More recently, by considering the cost of labor, calculations have been
Comparison, CGT, Man-hours made of the relative cost per cgt, again comparing different nations
(Stott). As with the original comparisons, the use is better suited to
general comparisons between regions or nations, because the labor
costs may be varied, for example by overtime payments. However,
INTRODUCTION using published labor costs for nations a reasonable comparison can be
achieved.
The need for a universal measure of shipbuilding performance has been
recognized for many years. To meet the need, the concept of The most recent published development has been to use the cgt method
Compensated Gross Tonnes (cgt) was developed within the OECD, to measure warship productivity. The modest number of warships built,
initially as a means of overall comparisons between regions and over extended timescales, and the variations in outfit and weapons
perhaps nations. The principles are consistent, but the coefficients are make this a very difficult area. Several papers have been published
reviewed every few years to take account of changes in ship technology (Craggs, et al) (Lamb) which show some variations in the
(OECD). The basic principle is to take the gt for a ship and multiply by compensation figure for a given size. This is not surprising, because
a compensation factor which “corrects” the gt for differing work military ships do have wide variations in a number of features. Some
content. The start point was a basis ship, originally a 15,000 dwt cargo auxiliaries are now built to nominally “commercial” rules. On
ship, where the compensation factor is set to 1.0. In this case the gross warships, weapons fits do vary widely and the duties for which they are
tonnes of the ship are used directly as the measure of work content. designed, their endurances and speeds are all factors. The result is that
the method again becomes useful for general comparisons but not for
Thereafter, for different ship types of varying sizes, compensation specific cases. The time dimension is also important and is considered
factors were agreed internationally. The factors range from 4.0 for later in the paper.
some complex, typically passenger ships to 0.25 for very large bulk
cargo ships. The gross tonnage of the ship is multiplied by the Nevertheless, cgt remains the internationally accepted means of making
compensation factor to give compensated gross tonnes. (cgt). shipbuilding performance comparisons, and no acceptable alternative
has been developed. It therefore seems appropriate to review the
In principle, using CGT, the work content of ships which are radically subject to try and identify whether any changes can be proposed which
different in size and function can be compared. Originally used for might improve the presentation and ease of use of comparisons.
comparisons between regions, using aggregated data, the use has been
extended to comparisons between shipyards. One major use was in a
study carried out in 1992 in Europe and the Far East. (Commission for WEAKNESSES IN PRESENTATION.
the European Communities)
The conventional presentation of the cgt and cost uses curves (known
The original factors were single numbers for ranges of ship size. as iso-cost curves). However the curves do present some difficulties in
Therefore at the ends of the ranges, a small change in tonnage resulted interpretation. For very large ships, the curves become asymptotic, and
in a significant shift in cgt. So for example a bulk carrier of 49,999 dwt very small changes in compensation factor need to be identified. For
would have a factor of 0.6 whereas a ship of 50,001 dwt would have a very small ships, the curve becomes steep and in reality the cgt method
factor of 0.5. For other than large, aggregated data sets, this was a of comparison is not generally used for small ships.
potential cause of inaccuracies. This was recognized (Bruce and Clark)
and curves proposed which result in a smooth transition through the The weakness in presentation is unfortunate, because the underlying
total range of ship sizes. data can provide many useful insights for producers, and purchasers, of
new ships. The data represented by the curves can also be used to

Paper No. 2005-P13 Bruce -1-


identify future requirements in performance and to review the accurate picture is difficult, since it is really necessary to consider year
implications of changes of product mix. on year changes in various parts of the organization.

A further example (Birmingham et al), describes how the Iso-cost


curves can be used to identify the most valuable investment for a FINANCIAL ISSUES
shipyard to make, in terms of the cost improvement as well as the
productivity improvement. The paper describes the use of performance The effects of time also come into play when the money (cost of labor)
gradients, where the employee hours per cgt are reduced by an is considered. First, there is an inflation figure to be considered.
investment, but the investment results in an increase in employee cost Although this is not a major issue in Europe or Japan at present or in
per year. The performance gradient has to be steeper than the cost curve recent years, this is not the case in other countries. Inflation in South

Downloaded from http://onepetro.org/SNAMESMC/proceedings-pdf/SMC05/3-SMC05/D031S005R009/2453196/sname-smc-2005-p13.pdf/1 by COTECMAR, Yeslis Guerra Ramirez on 23 May 2023


for the investment to be useful. A similar approach can be taken to look Korea is higher. The timing of a contract, and therefore when the cgt
at future increases in employee costs, to establish the necessary measured are actually accrued needs to be consistent.
productivity improvement to compensate.
Also, exchange rates play a major role. It is usual to quote ship and
However, the problems of dealing with the curves where they tend to other costs in US Dollars or increasingly in Euros, so the exchange rate
asymptotic do apply, and the differences between iso-cost curves at the between the dollar, the Euro and the currency of the country being
ends are also difficult to identify. considered is important. This typically varies over time and in recent
years this variation has been considerable. The figure shows the
It is worthwhile seeking a simpler and more readily understandable exchange rate against the US dollar for the Yen, Won and Euro, major
presentation. currencies for shipbuilding, over the last decade. The rates are shown as
indices, with 1990 values set to 1.00.
THE EFFECT OF TIME.
3
The data to build up comparisons is taken from numbers of ships, and
are built over a period of time. The data is therefore time dependent. 2.5
Figure 1 shows the overall cgt per man year figures for Japan. It shows
the variation from year to year of the performance in cgt, but also 2
shows a steady trend of improvement. Conventionally, using a three US$ - Won
year moving average will smooth the variations, and demonstrate the 1.5 US$ - Yen
trend. Similar effects were noted for Europe. Overall, an improvement US$ - Euro
of the order of 4% annually in labor productivity can be detected. A 1
data set of ships will give a range of cgt performance, which will take
the form of a distribution. It may be reasonable to assume that this 0.5
distribution is normal.
0
1990 1995 2000 2005

100 Productivity Trend for Japan

PRODUCTIVITY

50 This is normally measured as output/input, which in shipbuilding terms


would be cgt/labor unit. In most of the papers referenced the
measurement is man-hours per cgt. (similarly examples of more
detailed measures are man-hours per tonne for steel and man-hours per
pipe). This is not a true productivity figure, and presents the anomaly
that the figure reduces as the performance improves. The conventional
1990 95 00 05 presentation shows man-hours per cgt, plotted against time to
demonstrate an improvement trend or against labor cost to demonstrate
Figure 1 Productivity Trend in Japan increasing (or decreasing) costs. It is also used for comparison purposes
to show the difference between countries. The result is a curve which is
If the distribution is plotted over time, then the trend is visible. What is perhaps most familiar as the ISO-cost curve.
also visible is the variation between individual cases, especially for
ships other than large cargo vessels. There several weaknesses in this presentation. First, as the performance
improves, the number of man-hours cgt reduces. Where the reduction is
In particular for warships, the small numbers built create problems with for numbers in tens this is acceptable, but for the very efficient
the data. Warship production also takes place over a larger number of producers of large ships man-hours per cgt are below ten and the
years, perhaps three or four for a frigate or auxiliary and up to ten for a improvements are measured to two or more decimal points. Also,
one-off large surface ship, e.g. an aircraft carrier. This compares with a unless the cost curve is mathematically defined, interpolation becomes
matter of months for conventional commercial ships. So if a shipyard difficult and the detection of a small change is not easy to plot.
producing warships is steadily improving productivity, creating an Although the curve is familiar and demonstrates trends, it is difficult to
use in a practical sense.

Paper No. 2005-P13 Bruce Page number -2-


If, as an alternative, the “true” productivity measure is adopted, that is the performance curves, an alternative presentation may be easier to
the cgt per man year, the picture changes. First, the improvement trend manage. One alternative is to invert the factor, so that the gross tonnage
is reflected in a rising performance figure, which seems more logical. is divided rather than multiplied.
Second, and much more significantly, the link between points of equal
cost becomes a straight line, so interpolation is simple. 120

A further benefit of CGT per man-year is that it eliminates argument


concerning the difference in man-hours between, for example West 100
High performance
Europe and South Korea. The actual annual man-hours are variously
quoted, but 1,600 for Europe and 2,200 for South Korea may be

Downloaded from http://onepetro.org/SNAMESMC/proceedings-pdf/SMC05/3-SMC05/D031S005R009/2453196/sname-smc-2005-p13.pdf/1 by COTECMAR, Yeslis Guerra Ramirez on 23 May 2023


considered typical. So, man-hours per cgt as a measure begs the
question of how many hours are paid for and at what rate. There is, for Profitable low
example, the question of whether some workers in Korea put in unpaid performance
50 Low perfromance
overtime. Arguably this reduces the actual productivity, but if it does
not cost the shipyard anything then it enhances the performance.

Using man-years simply requires the total annual cost of labor to be


estimated from published data.

Figure 3 shows the result. The line of equal cost per cgt is clear and any 0 20 40 60
point can be identified. The linear approach also allows the impact of
modest changes in labor cost or performance to be plotted and the Figure 4 Comparative Shipyard Performance
results are obvious. It is much simpler than handling a curve.

120 Shipyard Cost Comparison

100 100
C GT / 90
M an-year 80
CGT per Manyear

70
60
50 50
40
30
20
10
0 20 40 60 0
A nnual cost per employee (US$,000)
0 10000 20000 30000 40000 50000

Figure 3 Lines of equal cost for ship construction Cost per Manyear

The figure also shows different lines, representing different cost


performance. Although in principle the use of CGT does result in Figure 5. Data for a set of shipyards, showing relative performance
comparable work content, there are differences in the efficiency of levels.
different shipyards. Also for some, particularly high technology, ship
types, the number of shipyards able to compete in the market is small
4.0
and the available price is higher. The figure shows different costs for
these different ships.
3.0
Figure 4 demonstrates clearly that a high performing shipyard with high
labor costs is no better off in overall performance terms than a low
performing shipyard with low labor costs. Similarly, if a shipyard 2.0
continues with a low productivity then some sort of financial support
may be deduced 1.0

Figure 5 shows data from a number of shipyards, from which relative 0.8
performance and costs can be identified and compared.
0.4

5 10 20 50 100 150 200 250


COMPENSATION FACTORS D eadw eight tonnes ‘000

The factors generally range from 4 to 0.25, according to ship type and
Figure 6 Typical plot for compensation factors
size. If these are plotted, then a curve can be derived. As in the case of

Paper No. 2005-P13 Bruce Page number -3-


Given the shape of the factor curve, a straight line might be expected curve and the reciprocal of the compensation factor is close, but not so
when the reciprocal is plotted. However, as the later examples show close, to a straight line.
this is not the case.
The use of a mathematically defined factor would make comparisons
easier, and ensure consistency. Plotting the existing factors gives results
PROPOSED NEW APPROACH which suggest that this approach is feasible. However this paper is
primarily concerned with presentation and use so further changes in the
This paper therefore proposes that the compensation factor should be derivation of the factors will be left for future research.
the inverse of the current factor, used as a divisor rather than a
multiplier. The effect is exactly the same, thus: Coeficients and reciprocals

Downloaded from http://onepetro.org/SNAMESMC/proceedings-pdf/SMC05/3-SMC05/D031S005R009/2453196/sname-smc-2005-p13.pdf/1 by COTECMAR, Yeslis Guerra Ramirez on 23 May 2023


3.5

Current VLCC of 250,000 dwt has a compensation factor 0.30


If gt is 140,000 3

2.5

Then cgt is 140,000 * 0.30 = 42,000 cgt

Factor/reciprocal
2 Coeff
Reciprocal
New version, VLCC has a compensation factor of 3.33 Linear (Coeff)
1.5 Expon. (Reciprocal)

So cgt is140,000 / 3.33 = 42,000 cgt 1

The productivity is measured as cgt per man-year. 0.5

For a Panamax bulk carrier, of 65,000 dwt and gross tonnage 40,000 0 0.5 1 1.5
Log dwt
2 2.5 3

CGT is 40,000 *0.5 = 20,000 cgt Figure 8 Coefficients plotted on logarithmic scale
As above the reciprocal (2.0) gives the same result
PERFORMANCE IMPROVEMENT GRADIENTS
An example of the actual plot for double hulled Tankers is shown in
Figure 7 Returning to the issue of application of the cgt approach, the proposed
Double Hull format makes the use of performance gradients easier.
4.5

Figure 9 shows the situation for a shipyard. The current cost line shows
4
a point “A” at which the shipyard is operating. A cost increase is
3.5
anticipated, which will leave the shipyard operating at point “B”. This
3 is on a higher cost line and implies that the shipyard will become
2.5
uncompetitive. The cost increase could be the result of an exchange rate
Coeff
variation or of wage increases.
CGT

Reciprocal
2

1.5 Figure yy
1
150

0.5

0 G
0 50000 100000 150000 200000 250000 300000 350000
DWT 100

Figure 7 CGT and reciprocal of cgt for Double Hulled Tankers F


CGT/
Man year E
When the coefficients are plotted, it is apparent that they do not C
D
produce a smooth curve. This paper will not consider the issue, as the 50
derivation of the initial coefficients is outside the intended scope, but a
mathematical relationship between compensation factors for different A B

ship sizes might be expected. The development of mathematical


functions seems to the author to be a logical approach to further 00

development of the factors. 0 20 40 60 80


Annual cost per employee (US$,000)

USE OF A LOGARITHMIC SCALE Figure 9 Use of GCT to determine future required performance

To recover, the shipyard must make a productivity improvement to take


If the compensation coefficients and reciprocals are plotted against the it to point “D”. Realistically, the shipyard should aim to reach point
logarithm of Deadweight, then figure 8 shows the outcome. This would “C”, because a change in productivity will take time and there is, as
also provide a means of dealing with compensation factors for smaller discussed earlier in the paper, an underlying improvement
ships, the difficulties of which have been mentioned earlier. It can be internationally. So the performance improvement gradient required is
seen that the coefficient as plotted can be represented by an exponential as a minimum one parallel to the cost line. Ideally the gradient will be

Paper No. 2005-P13 Bruce Page number -4-


steeper than the cost line, implying an absolute improvement in
performance. Other countries can be located on the graph, to show their current
position and either past or future trajectory.
To take a second example, a shipyard is operating at point ”E” on the
same cost line. This is displaying higher costs but higher productivity Another consideration is the type of ships to be constructed. Although
and therefore equal performance. Two possible investments can be CGT is intended as a universal measure, and in principle a cost per
made. CGT which represents a competitive position can be established, it
should be recognized that the required performance for some ship types
The first will maintain the shipyard at a performance, at point “F”, is higher or lower than might be expected.
which is better than the first shipyard, if that one takes no action to

Downloaded from http://onepetro.org/SNAMESMC/proceedings-pdf/SMC05/3-SMC05/D031S005R009/2453196/sname-smc-2005-p13.pdf/1 by COTECMAR, Yeslis Guerra Ramirez on 23 May 2023


recover from the cost increase and reaches point “B”. However, there There are considerations of quality, where in some cases very low cost
will be a drop in absolute performance (equating to high performance) is combined with lower quality.
Equally, in some cases a very high quality is required and a premium
The second investment will improve performance to the point “G” might be paid for this. There are also technical barriers to competition,
where the shipyard will be on the lower cost line and therefore remain for example in some types of fast vessels or in the case of large
competitive even if the first shipyard makes the larger of its passenger ships. The relationship between owner and builder may also
improvements. be a factor, whether a purely business matter, or with some degree of
financial interdependence or political influence associated with it.
COUNTRY COMPARISONS The figure 11 indicates the relationships between some markets,
countries of build and costs, indicating how cgt might be used as part of
Returning to the original use of cgt as a performance measure, which is the product mix decision making process.
to make comparisons between countries or regions using aggregated
data, the proposed format for presentation can provide some useful
High
insights. By considering the current cost and performance figures, and
the likely future changes, the competitive positions can be reviewed Container Ships LNG Carriers
and presented in a clear format. Korea
Japan

Figure10 shows the current positions of a number of significant nations EU


Passenger vessels

in the shipbuilding market and how their positions might change in the Productivity China
future. Capability High Speed vessels
EU
250 Ro-Ro

Niche and Local Markets

Bulk Carriers

200 Low
Korea
cgt/
man-year Japan 0 20 40 60
Annual employee cost US$,000
150

Figure 11 How Countries may move in the market


China

China is moving from low value bulk carriers and basic tankers into
100
more lucrative markets, based on low cost and improving productivity
and quality.

Korea is displaced from bulk and container markets and is moving to


0
0 10 20 40 60 LNG and higher
80 value vessels.
Annual Cost per Employee US$,000
Japan is in need of a niche, but has some barriers to market penetration.
Figure 10 Relative Positions of leading Shipbuilding Countries
The EU has a small group of shipyards protecting passenger the ship
The assumption is that China has sufficient scope to improve market, and a larger group seeking niches with entry barriers.
productivity, adapting the latest ship production technology, before
employment costs rise significantly.

Korea is currently has significantly higher productivity, but is close to CONCLUSIONS


limits, in that only large scale investment can improve productivity and
this will effectively increase the annual cost per employee. The gradient This paper has reviewed the use and presentation of compensated gross
is unlikely to be sufficient to maintain competitiveness tonnes as a means of comparing shipbuilding performance between
different nations. A modified presentation has been proposed, which
Japan is in a similar position, with higher employee costs, and has the makes the use of the method simpler and which clarifies the
same problems of engineering a performance increase comparisons. Further research to review the compensation factors and

Paper No. 2005-P13 Bruce Page number -5-


to seek a mathematically derived function to define the factors is
recommended.

Some uses of CGT in managing the future of shipyards, nationally and


regionally have also been presented. There is considerably more to be
gained from the use of CGT measures in the future.

REFERENCES

Downloaded from http://onepetro.org/SNAMESMC/proceedings-pdf/SMC05/3-SMC05/D031S005R009/2453196/sname-smc-2005-p13.pdf/1 by COTECMAR, Yeslis Guerra Ramirez on 23 May 2023


Commission for the European Communities, Report of a study into the
competitiveness of European Community shipyards, KPMG, October
1992

OECD, New Compensated Gross Tonnage Coefficients, 1984, 1991

Bruce, G and Clark, J, Productivity measures as a tool for performance


improvement, RINA Spring Meeting, April 1999
Birmingham, R et al, Shipyard Technology Development Strategies,
Journal of Ship Production, November 1997

Craggs, J et al, Naval Compensated Gross Tonnage Coefficients and


Shipyard Learning, Journal of Ship Production, May 2004.

Lamb, T, A Shipbuilding |Productivity Predictor, Journal of Ship


Production, May 2002

Paper No. 2005-P13 Bruce Page number -6-

You might also like