Chapter 11

Cost Estimating

If we choose we can live in a world of comforting illusion

—Noam Chomsky
It’s easy to see, hard to foresee
—Benjamin Franklin

The cost of a ship or any other large, made-to-order product must be estimated, in
the first place to determine whether the construction of that ship is a viable project.
This is necessary for both the customer who will own or manage the ship, and the
potential builder. An initial estimate of ship cost is needed when the ship is no more
than an option for the owner’s future trading plans or a potential product to be part
of the shipyard strategy.
Estimating problems exist because a ship construction project is by definition the
creation of something novel. They include the following. Many ships are prototypes
so there is limited data from similar past experience on which an estimate can be
based. The product information which is available from very early designs is limited,
but it is still necessary to create an initial cost estimate. As a rule the availability of
design data lags behind the need for it in cost estimating through the life of a project.
There are also exchange rate fluctuations over time so the value of an international
project is likely to vary from the inception to completion. In additional future cost
inflation is unknown, as is the future supply and demand for shipbuilding materials
and equipment. These factors will affect the actual cost of construction.
There is a degree of risk in any estimate because of the above. In fact the final
cost of a ship is only known after it is complete, and even then will depend on the
accuracy of accounting assumptions and data recording.
As part of the future strategy for the shipyard, the future product ranges must
be considered. Assuming this is not simply a continuation of current products, with
some minor variations then some new initial designs will be created for discussion
purposes.
At this point there is very limited design information on which an estimate can
be based, but the development of at least an order of magnitude cost is required.
There is a need to produce a credible price for potential customers so that the ship-
yard commercial function can decide whether the ship is viable as a future product.
There may be discussions with potential customers and some price indication will be
© Springer Nature Singapore Pte Ltd. 2021 123
G. Bruce, Shipbuilding Management,
[Link]
124 11 Cost Estimating

required before these can proceed. In some cases there is a close relationship between
owner and shipyard, for example for large passenger ships, for military ships and
for many classes of specialised, non-cargo ships. Both customer and contractor need
initial cost estimates to decide whether to proceed with the project or not.
At this point where the ship is just an idea, the order of magnitude cost will be
used to review concepts, from which the most promising can be selected for further
development. The accuracy at this stage for a prototype ship is limited, and the final
cost if the ship goes ahead is likely to be 20% more or less than the initial estimate.
This is significant, but will be adequate for initial product selection and will be
refined if the product is selected for further study. This estimate will also be used by
the shipyard for evaluation if there may be investment in the production facilities.
An approximate early indication of cost can be developed from main characteris-
tics of the ship. For example, it can be based on the cargo capacity of a ship. Given
records of past contracts a formula can be developed which relates the cost of a ship
to the capacity. If the proposed new product is similar to previous ships, perhaps with
a development in size, or with some novel equipment, then this is a useful starting
point for the cost.
So to take a very simple example, if the intention is to develop a larger ship for
transport of LNG, and there is a history of smaller ships then the following gives
a first indication. If an LNG ship of 160,000 m3 is proposed, and the shipyard has
a history of ships of 125,000 m3 a formula can be developed. This will be using
the ratio of cargo capacity to predict the ratio of cost. However this only works for
a reasonably consistent data set of past projects, in the same or similar shipyards
using the same or similar technology, with the same cost base. If there is more data
available, for example if steel, machinery and insulation requirements are estimated
then a more sophisticated formula can be developed. However the same caveats will
apply and the sophistication of the formula will only be useful if there is sufficient
similarity between the two ships and the same shipyard is to be engaged.
The potential cost of a ship or other project may be estimated using a variety of
estimating parameters. At the initial design stage weight is usually the parameter
chosen although capacity is an alternative. Various parameters can be and are used
in different shipyards. In all cases the value of the estimating method depends on
accurate data gathering and analysis from past projects. Alternatives used in different
shipyards include deadweight (dwt), lightship, gross tonnes (gt) and compensated
gross tonnes (cgt). The basic weights of the ship will be part of the owner’s initial
requirements and later of the contract, so are reliable. For large cargo ships the steel
weight is also a potential measure to use. All of these figures are estimated at the
early stage of a project and can be reasonably accurate.
Figure 11.1 shows how plotting ship cost against the parameter can provide a
regression curve. Cost is a function of the parameter and a formula to link the two
can be developed. The data can be plotted on a logarithmic scale to give a straight
line, which makes the information easier to interpret. However, the formula which
is developed to produce a cost is only accurate in terms of the parameter and data
selected and how good the data is in the first place.
11 Cost Estimating 125

Ship
Cost

Size Parameter

Fig. 11.1 Parametric estimating

There are dangers in using the parameters simply because there may be a high
level of variability between the costs identified for the different past ships. The data
collected generally produces a scatter of data points and the curve represents the best
fit for these data points. There will be potential, if a new point on the curve is selected
for a project, for the actual value to be significantly higher or lower. Ideally several
parameters should be used to provide some capacity to cross check the estimates.
If an initial design is to be developed, because it seems to be of interest to potential
customers then a design study will be carried out. This is where there is a potential
project and the design can be taken to a stage where a contract may be signed. How
far the design is to be developed is variable. In some cases the design if commis-
sioned by the owner will be very complete and will be sent to prospective builders
for a quotation. In others, there may be a joint development between a shipyard
and regular customer, so the design development may continue after agreement to
proceed. Typically at this stage of the design development, the estimated costs again
for a prototype ship may vary from the final cost by plus or minus 10% to 15%.
The cost of a ship is made up of several factors. The first of these is material
cost, which includes steel, machinery, equipment, pipes and coatings. Then there
are sub-contracted items usually including electrical outfitting, heating, ventilation
and air conditioning (HVAC). The cost of shipyard labour, both direct and indi-
rect, is next. Finally there are shipyard overheads which are shipyard costs not
directly attributable to the contract. Some of the indirect labour may be included
as an overhead, depending on the accounting policy.
Materials costs can be estimated providing there is a specification of requirements.
However as with other design information the specification will develop as the project
126 11 Cost Estimating

design develops, so full and detailed information is only available at the time of
purchasing. To take steel materials for a project as an example, an initial estimate
of steel may be based on a previous ship. The steel quantity for the new proposed
ship is estimated from the previous ship, usually by a formula using the principal
dimensions for the two ships. So the estimate for the new ship steel weight would be
the basis ship steel, multiplied by a function based on the ratio of the new dimensions
to the basis dimensions.
Machinery cost can be estimated using a formula relating the kilowatts for the
new ship and previous ships. Large items such as generators can be estimated by
comparing the electrical capacity of the two ships. All early estimating methods are
useful guidance, but must be used carefully.
Labour cost estimating is the main variable cost under the direct control of the ship-
builder. However because there are usually many sub-contractors and items supplied
from other parties, labour generally only represents approximately 20% of the total
cost of a ship. Nevertheless it is an important component of that cost.
Different approaches are used to estimate labour costs, usually single param-
eter estimates or time study driven methods. Single parameter estimates are gener-
ally based on past shipyard performance levels. If the past man-hours for a unit of
production are known and the unit quantity for a new ship can be estimated, then
the man-hours for that ship can be determined. The unit of production is usually a
global measure, based on an attribute of the ship which is readily calculated at an
early design stage. Typical estimating parameters to estimate the man-hours required
are described here.
Man-hours per compensated gross tonne (cgt), or cgt per man year can be calcu-
lated. The compensation factor allows different ship types to be compared. Gross
tonnes are available for any cargo ship. The compensation factors developed by the
OECD and agreed internationally are designed to allow equivalence when comparing
the man-hours required for construction of different ship types and sizes.
Man-hours per light tonne is a useful measure for non-cargo ships as it considers
the total weight of the ship. It is most used for military ships.
Man-hours per tonne of steel is a factor which is most valuable for large cargo
ships, particularly oil tankers and bulk carriers where the steel hull is the dominant
element.
Examples of other parameters are metres of pipe, square metres of coating and
there are many others, generally inverting the measure to give the quantity per man-
hour. In principle, any measureable quantity can be used so the method can also
apply later in a project to create detailed estimates at a local, work-station level.
The availability and level of detail varies as the design develops, but dimensional
parameters can generally be estimated reasonably accurately at an early design stage.
Sub-contractor costs will be based on past projects and the anticipated extent of
sub-contracting for the future period. In some cases where the shipyard and major
sub-contractors have long experience of collaboration, an estimate can be generated
in cooperation with the sub-contractor.
11 Cost Estimating 127

The shipyard overheads are estimated at an early stage by adding a percentage to

labour costs, using past information and expectations of the labour hours to be used
in the period being considered.
At the build strategy stage of the contract there is preliminary design information
available. At this stage a firm price is needed so that a contract can be considered.
The estimate produced should be within 5% of what will be the final actual cost.
As mentioned earlier, the level of design development on contract signing can vary.
For the purposes of this discussion, it will be assumed that the design is sufficiently
advanced for a specification which will include the major equipment and machinery
items. The arrangements of the ship will be sufficiently advanced to allow the lengths
of pipe and other systems to be determined with reasonable accuracy. Also the areas
and volumes of spaces in the ship will be available so that estimates of quantities of
materials can be estimated.
At the tactical planning stage, overall production information is available and this
allows more detailed estimates to be made. Accurate resource requirements must be
decided so that the project can be planned and executed. An important point here is
that the total budget set by the contract price is usually fixed, so all these estimates
must give a total within that budget. There may be agreed cost variability allowed in
the contract where the project retains some uncertainty. At this stage the estimates
should be within plus or minus five percent of an accurate figure.
The estimate determines the cost of building a ship, in a specific shipyard and at
a specific time. It is subject to the variations described earlier and to internal factors
in the shipbuilding company.
The price for the ship is set by the shipbuilding market which is affected by supply
and demand in the shipping market. This then determines the demand for ships, also
taking into account the supply of shipbuilding capacity. The capacity is influenced
by possible incentives from governments, since a lot of shipbuilding capacity world-
wide exists because of government intervention, both to create a capacity and to
maintain it. Incentives can be offered to shipbuilders and to owners, and may not be
transparent because subsidies are against trade rules in many cases.
The current and future expected freight rates also play a part, as shipowners try
to anticipate market changes. Any profit from a shipbuilding contract is the small
difference between cost and price and a typical return for a ship contract may be five
or tem percent of the contract price.
Some further influences on cost and price include whether the shipyard has expe-
rience of a ship type. Lack of experience may result in an underestimate of the
production cost. There can be a danger of the commercial imperative to secure a
contract overriding the need for a realistic estimate of cost. A poor market will
result in lower prices irrespective of costs, although there may be some lowering of
commonly used material costs. The credit rating of the customer may influence the
price which a shipyard would seek. A full order book may allow a lower overhead
rate and consequently a reduced price offer.
As the design develops an accurate quantity can be determined and at the detailed
design stage there is a precise definition of each steel part to be used. It must be
remembered that the cost of the steel is the cost of the invoiced steel, not the steel
128 11 Cost Estimating

which is actually used for the ship. There may be up to 10% wastage from part
cutting and this must be added to the net steel identified from the accurate computer
generated design information.
Other materials include pipes, cables and coatings. The situation is similar to steel
in that detail develops over time. The preliminary estimate based on past experience
and systematic variations in quantities. For machinery there will be past information
probably of limited value or an estimate from potential suppliers. The same applies
to other equipment.
For sub-contractors it is possible to seek quotations from potential contractors but
they have similar problems as faced by the shipyard, with limited early information
and an incomplete specification.
It is worth mentioning again that the actual cost of a ship may vary considerably
from the initial estimate. The actual cost is likely to rise considerably for a prototype
project. However, if further ships are expected to be constructed after the prototype,
in a series of sister ships or at least similar ships, then the later ships can be expected
to be built at lower cost. This is because there will be a learning curve effect.
Learning curves provide a means of capturing the benefits of experience. This
is a concept originally developed during the 1940s in the USA. Many aircraft were
produced in new factories using labour with no previous factory experience. Some
means of estimating productivity was sought and it was observed that there was gener-
ally a consistent improvement. Each time the number of a product made doubled,
the cost reduces by a fixed percentage. Although the theory was developed for large
numbers of aircraft, the principle does apply to ships.
The estimating process, and especially the information used and generated, is
a closely kept secret for a shipyard. However all shipyards use the same basic esti-
mating models. These generally use a spreadsheet, although there is some proprietary
software available. The estimate is carried out on the basis of the various material
categories, because materials are the largest proportion of the total cost.
A database is often generated to provide a library of past contract cost informa-
tion and current information for example on equipment costs. There are recognised
categories, especially for military ships, described as ship work breakdown struc-
tures which are used alongside the product work breakdown structure. The use of
materials, based on the ship systems, is logical since the bought in materials are large
proportion of the total ship cost. Usually each type of material or equipment has a
three digit code. For example, machinery codes may be numbers in a 400 series. A
second digit identifies a particular system, for example 410 for deck machinery, 420
for thrusters and 440 for the ballast system. The third digit identifies particular items
or materials within the systems, so perhaps 441 for pumps.
Typical systems and cost categories are listed here:
Steel as the main material used for the ship structure. It may be replaced by
aluminium for some specialised, smaller ships.
Other hull materials, which includes paint coatings and deck coverings.
11 Cost Estimating 129

Cargo systems, which vary from a minimum on simple bulk carriers to deck
cranes, container fittings, lashing systems for trailers on ro-ro ships and cargo
piping systems on tankers.
Hull systems, always ballast piping, pumps, valves and other internal fittings.
Deck fittings and equipment are also included.
Machinery for propulsion, including propellers and shafting.
Other machinery systems, which may be limited or be very extensive for a
specialised ship.
Electrical systems with generators, switchboards, lighting and cabling.
How the various systems are designated will vary, and most shipyards have their
own preferences. It is not that important provided the categorisation which is chosen
is always consistent to make the data generated over time as useful as possible.
Labour is usually treated separately, with the labour costs allocated to categories
which correspond to the materials as above. Alternatively the labour categories may
be based on the expenditure in different work stations. The allocation will depend
on the available past data and preferred practice.
Services are the final category, to take account of requirements for the contract
to be bought from suppliers. Examples are design studies, model testing, specialist
noise and vibration engineering or measurements using laser or photogrammetric
systems.
Within the estimating model, cost may be for a quantity of materials, at a known
rate. So the steel cost per tonne will be multiplied by the estimated quantity. Different
costs and quantities will be applied for steel plates, profiles and any special steels
which are required. Other materials will generally use the same principle. For
equipment the estimate may be using a supplier quotation for a specific equipment
item.
In some cases a quotation may be sought from a supplier quotation for an equip-
ment item, though this can be problematic because the supplier price will generally
depend on the state of the equipment market at the time of placing a firm order. Good
supplier relations may be of help.
Variable overheads are those mainly labour costs which can not easily or accurately
be directly attributed to a specific contract. The definition of these costs does vary
from company to company, but it is usual to include marketing, commercial activi-
ties, stores management and services such as transport and maintenance. Others are
sometimes included, such as technical staff and supervisors, although many ship-
yards attribute these costs to a contract. The costs are variable as they will increase
or decrease with the shipyard level of activity.
The motivation is often to try to keep overheads to a minimum, as they are seen
as not contributing directly to the operation of the shipyard. They may be seen as a
cost rather than a benefit. So by recording the time of some of the staff mentioned
above as part of a contract, the overheads are believed to be somehow reduced.
Realistically, if the work done is necessary for the shipyard operation, then
precisely where the costs are allocated is not so important, as they will not go away
if they are counted as direct rather than indirect. If they are not essential, then an
130 11 Cost Estimating

analysis should reveal this easily and the work can be left undone. It is not uncommon
for shipyards in financial difficulties to stop some work, such as maintenance and
training to save indirect costs. As a short term solution this may have some effect but
in the longer term it is not a good idea.
There are also fixed overheads, which are represented mainly by the costs of the
shipyard facilities. This accounts for the shipyard paying for the docks, workshops
and other hardware. These are all the costs that must be paid whether ships are built
or not and also include rentals and local taxes in many cases. The fixed overheads
are a varying proportion of the costs, depending largely on the age of the shipyard
facilities. Some of these such as docks have a very long life, and especially where a
shipyard has changed ownership, may have been completely paid for.
A brief summary is appropriate here. Firstly, estimating is not a precise science.
It is a best estimation of the costs of building a ship, with the information available at
the time of the estimate. Second, the design and other information required usually
lags behind the need for it in the estimating process. Parametric methods are useful
in early stages of a design, although not fully reliable so several methods should be
tried. Importantly, the estimated cost of building a new ship is not necessarily related
to the price that the market is prepared to pay for it.