Cost estimation

methodology
The Marine Energy Challenge approach to estimating
the cost of energy produced by marine energy systems

Commissioned by The Carbon Trust

and carried out by Entec UK Ltd
May 2006
Disclaimer
This report has been prepared by Entec UK Ltd. solely for use by
the Carbon Trust by whom it was commissioned. It is not
addressed to and may not be relied upon by any person or entity
other than the Carbon Trust without the prior written permission of
both the Carbon Trust and Entec UK Ltd.

In producing this report, Entec UK Ltd. has relied upon

information provided by third parties.

Neither the Carbon Trust nor Entec UK Ltd. nor their respective
directors, employees or affiliated companies give any
representation or warranty, express or implied, as to the accuracy,
completeness or fairness of the contents of the report, nor accept
any responsibility or liability for any loss, whether direct, indirect
or consequential, arising from reliance on it.

Any use of this report by any third party for whatever purpose is
solely the responsibility of that party who should use such due
diligence to verify the report's contents and consult its own
advisers as may be appropriate.

The Carbon Trust does not give investment advice and nothing in
this report constitutes, or should be taken as, a recommendation to
enter into, or the giving of any advice in relation to, any
investment.

This report may be copied and distributed only with the consent of
the Carbon Trust.

©The Carbon Trust 2006

 i

Contents

1. Introduction 1
1.1 Why is this report needed? 1
1.2 Marine Energy Challenge approach to cost of energy 1

2. Cost of energy 3
2.1 The cost of energy equation 3
2.2 The present value approach 3

3. Performance 6
3.1 Device characteristics 6
3.2 Resource 6
3.3 Losses 6
3.4 Availability 7
3.5 Annual energy production 8

4. Cost 10
4.1 Key factors affecting cost of energy 10
4.2 Capital costs 10
4.2.1 Structure costs 11
4.2.2 Mechanical and electrical costs 11
4.2.3 Mooring costs 13
4.2.4 Installation 13
4.2.5 Grid connection 14
4.2.6 Project management 14
4.3 Deployment location 14
4.4 Operating costs 15
4.4.1 Planned maintenance 15
4.4.2 Unplanned maintenance 15

5. Uncertainty in estimates 18

B/15992/C001/065 May 2006

 ii

Table 2.1 How investors might think about discount rates 3

Table 4.1 The influence of the deployment location on the cost of energy 15
Table 4.2 Cost centres and cost drivers 17

Figure 2.1 Range of discount rates applicable to different types of project 4

Figure 3.1 Simplified process for calculating the energy production of a marine energy project 9
Figure 4.1 Structural complexity 11
Figure 4.2 Moorings 13
Figure 4.3 Cost breakdowns for an example wave energy device a) capital b) O&M 17

Appendix A Sources of further information

B/15992/C001/065 May 2006

 1

1. Introduction

1.1 Why is this report needed?

The Marine Energy Challenge studied many different marine energy concepts, each at a
different stage of development. The most important comparator is the cost of energy. The
process of calculating the cost of energy is relatively straightforward. However, because each
device and deployment location is different and each is at a different stage of development,
obtaining meaningful estimates for each can be complicated.
This document explains a simple approach to estimating the cost of energy applicable to all
marine energy devices. It shows how costs and performance can be calculated in a consistent
way and compared between devices. It allows for more sophisticated methods to be used, if
available, by highlighting both where known quantities are used and where only estimates are
available.
The Marine Energy Challenge showed that the cost of energy from most marine energy devices
is currently too high, but that with further consistent development these costs can reduce. This
method helps clarify both the costs and the stage of development to which the costs refer.
This method brings the following benefits—
• Demonstrates how the cost of energy can be calculated from cost and performance
information
• Brings clarity to the cost estimation method

• Shows some ways that costs can be estimated in the absence of accumulated
experience
• Provides a framework for collecting information on devices at different stages of
development

• Provides a start point of an assessment method from which discussions on status

and potential can be made between investor and technology developer.

1.2 Marine Energy Challenge approach to cost of energy

The cost of energy produced by a marine energy device is related to the incurred costs and
quantity of energy generated by the device. The Marine Energy Challenge aimed to lower the
devices’ costs of energy; this can be done by reducing the costs of building and operating a
device and by increasing the device’s energy production. The Marine Energy Challenge looked
at all options for reducing the cost of energy.
The costs include capital costs and operation and maintenance (O&M) costs. The performance
of a device is related to the amount of electricity it produces. These are all interrelated and an
improvement in one may require a trade-off with another. This means that before a device’s cost
of energy can be estimated, it is necessary to define a sensible basis of design. This describes

B/15992/C001/065 May 2006

 2

the fundamental operation of the design but might not describe the optimal configuration.
Nevertheless the design should be one that could actually be built (i.e. using certain materials
and known construction techniques), could be deployed (i.e. using certain vessels and moorings
or foundations) and will work (i.e. produce electricity reliably and survive the marine
environment).
During the Marine Energy Challenge, the first stage of evaluation was to define a baseline
design. The costs and performance were then determined and the baseline cost of energy
estimated. In some cases, this indicated that the costs were too high to justify the performance,
and subsequently ways were sought to either decrease the costs, improve the performance, or
both. An iterative design process followed during which different design possibilities were
explored and their potential benefits were quantified. This resulted in improved designs with
lower costs of energy and/or greater confidence that certain cost and performance levels could
be reached.
The cost of energy is a moving metric. As designs evolve the costs will change. Ultimately
developers need to reduce the cost of energy to as low a level as possible. In the meantime this
may mean trying slightly more expensive options that have greater long-term cost-reduction
potential. Whatever the approach, the cost of energy should always be kept under review and
any previous estimates and assumptions revised.

B/15992/C001/065 May 2006

 3

2. Cost of energy

2.1 The cost of energy equation

An installation’s cost of energy is determined by a discounted cash-flow calculation. Given a
certain discount rate and period, the cost can be estimated using the following equation, where
PV indicates the present value over the service life—

PV (Capital cost ) + PV (O & M costs ) + PV (Decommissioning costs ) ⎡ £ ⎤

Cost of energy[£/kWh ] = ⎢ kWh ⎥
PV (Energy production ) ⎣ ⎦
Capital cost + PV (O & M costs ) ⎡ £ ⎤
≈ ⎢ kWh ⎥
PV (Energy production ) ⎣ ⎦
Equation 1
A more complex version of this equation would allow for the capital cost expenditure to be split
over several years and perhaps could include a mid-life refit. It also includes the costs of
decommissioning the scheme. The simplified equation above contains only the most important
variables.

2.2 The present value approach

The ‘present value’ approach underpins many different financial investments. It is routinely
used to assess and differentiate investment opportunities. There are many ways to interpret the
present value of an investment, but the calculation process is roughly the same regardless. The
interpretation of the present value depends on who you are, or rather what type of investor you
are.
The basic principle of the present value approach is to recognise that the value of £1 today is
more than £1 in the future. To account for this we ‘discount’ the value of money spent and
income generated in the future. The level of discounting can be interpreted in several ways—

Table 2.1 How investors might think about discount rates1

Risk An investor who puts money into a project expects a return on the investment. If the investment is
risky then they expect a higher return. For a given level of risk the investor might set a minimum return
that they are willing to accept. This could be called their ‘hurdle’ rate.

Opportunity Investors often have several choices of where to place their money. For example an investor may
choose to leave their money in the bank. This would provide them with a modest return. If they place
their money elsewhere then they lose the ‘opportunity’ to make the return from the bank. An investor
would then compare other investments against the ‘opportunity’ that they would lose. In most cases
they would expect a greater return.

1
Further reading: Corporate Finance, Brealy, Myers, Allen, McGraw-Hill International Edition,
ISBN 0-07-111551-X

B/15992/C001/065 May 2006

 4

Discount rates can represent either the project return or risk. Different industries carry different
levels of risk and are often assigned different discount rates. We consider that the discount rate
for the first commercial marine energy schemes would be around 15% whilst 8% might be
applied to it when the technology matured and became established in the market.

Figure 2.1 Range of discount rates applicable to different types of project

Adapted and reproduced with permission from Tom Murley at HgCapital

This means that when we calculate the cost of energy using this present value approach we
incorporate the project return into the cost of energy. Thus a scheme that produces energy at a
cost of 5p/kWh at 8% discount rate might produce electricity at a cost of 6p/kWh at 15%2. This
is the same scheme working in the same way, but with two different interpretations of project
risk or investment return.
Alternatively, if this same scheme secured a power purchase contract and was paid 5p/kWh for
all the power it produced over its life then it would make a return of 8%, but if the contract paid
6p/kWh it might make a return of 15%.

2
These figures are purely to illustrate the point about sensitivity to rate of return. Details of the costs of
marine energy systems can be found in the Carbon Trust report ‘Future Marine Energy’.

B/15992/C001/065 May 2006

 5

The Marine Energy Challenge used a range of discount rates between 8-15%. These covered the
likely range of financing options for technically proven products first entering the market
through to well-proven products in wide deployment.

B/15992/C001/065 May 2006

 6

3. Performance

3.1 Device characteristics

The first stage in assessing the performance of a marine energy project is to characterise the
performance of the device in terms of the wave or tidal current loading. For some devices this
calculation can be achieved using linear mathematical models and coefficients derived from
computer codes. Test data can be used to verify the model tests and provide better estimates of
the coefficients. In some cases, perhaps because the wave-device interaction is highly non-
linear, testing may be the only way to determine performance.
The Marine Energy Challenge used test data when these were available and developed
numerical models in all other cases. Mathematical models are important even where test data
are available because if they are well understood they can be used to calculate loads, design
parameters, find optimum sizes and to try different configurations. Validated numerical models
make good design tools.

3.2 Resource
The available resource is then measured or estimated. At a tidal stream site the most important
measurements will be of tidal current velocity. For wave sites both wave height and period are
important. For both technologies the amount of time for which each sea condition occurs must
be estimated. An estimate of the annual energy output can be calculated by multiplying the
power output of a device in each sea condition to the annual occurrence of that condition.
The energy production will vary depending on the site and, in general, devices placed in more
energetic climates will produce more energy. However, if the devices are poorly matched to the
climate then the energy output could be lower. Alternatively, the benefits of additional income
from increased output may be offset by the cost of increasing the device strength to tolerate
more severe wave loads or increased maintenance requirements. Either case would result in a
higher cost of energy.

3.3 Losses
Once the theoretical energy output has been calculated some account is taken of the various
energy losses. The power curves and power surfaces include some account of inefficiency.
These include, for example, the conditions when the device is deliberately turned off when the
incident energy is too low. Other losses must also be accounted for; these include losses in the
electrical cable to shore, and lost energy due to an imperfect match between incident energy
direction and device orientation. Many of these losses can be accounted for by applying a
simple reduction factor to the energy generated by the device. For example if the cable losses
are assumed to be 2% then for every 100kWh supplied to the cable 98kWh are delivered to the
electricity grid on the shore, the remaining 2kWh is generally lost as heat to the sea. Many of
the losses of this kind apply equally to all levels of power produced and are often assumed
constant in the high-level cost of energy predictions used in the Marine Energy Challenge.

B/15992/C001/065 May 2006

 7

3.4 Availability
The availability of a device is a measure of the amount of time that it is running without fault.
This time does not include periods when the device is working but is not generating due to a
lack of waves or tides. Availability is a measure of the reliability of the device.
During the Marine Energy Challenge, several methods for calculating availability of marine
energy devices were considered. These used various combinations of the following—

Estimates of servicing requirements

Most technologies will require routine servicing. Servicing involves a combination of inspection
and replacement of consumable parts (such as oil and brake pads). Mature technologies have
well-defined service intervals. These might be fixed time periods, such as every six months, or
they might be conducted after a certain amount of electricity is produced. Many modern systems
have sophisticated systems to calculate when servicing is required, depending on the time that
has elapsed, the load on the system during the interval, the environmental conditions and other
data from the monitoring of the condition of the system, e.g. of wear rates.

Estimates of failure rates

A reliability model is built from known or estimated failure rates for each of the components in
the device (MTBF). The failure analysis of systems is a well-refined and widely used technique.
In the Marine Energy Challenge the failure rates analysis was simplified in many cases so that
different configurations of system could be assessed readily. Modelling such systems can often
be conceptually straightforward but obtaining sensible operational data is impossible without
building the entire system.

Estimates of time to repair

The mean time to repair (MTTR) depends on the nature of the fault. For example some faults
can be rectified using remote control systems, whereas others will require visiting the device.
When visits are required, repair time is dependent on the nature of the work, for example
whether access for personnel is needed or whether heavy equipment or parts must be
transferred. Each type of intervention is affected to a greater or lesser extent by the prevailing
weather. Remote intervention might be immune to weather conditions whereas transferring
personnel with cumbersome loads may incur significant ‘waiting on weather’ delays.
The overall availability is a function of the rate of anticipated failures and the time to repair
them. Because the number and timing of failures cannot be predicted, and because failures are
often interrelated, the assessments make use of various statistical techniques to estimate
availability. For example, Monte Carlo analysis allows simulation of numerous scenarios with
different random combinations of faults and fault interactions and prevailing weather
conditions. This produces an estimate of the most likely availability for the complete system
whilst accounting for the main knock-on effects of the various faults on the overall operation of
the farm.
The methods for estimating reliability can rapidly become very complex and they rely on much
accumulated operational experience. When a large number of complete systems have been in
operation for many years the availability can be observed.
Ultimately the availability of the system is shown as a single representative number. An
availability of 100% would mean that the machine was always able to generate electricity when

B/15992/C001/065 May 2006

 8

the conditions were right. A more realistic availability level would be say 95%, when about 5%
of the time the system would be under repair or suffering a fault that stopped it generating.
A more sophisticated, and less conservative, measure of availability would consider the time
variation of energy production. This would mean that if all servicing could be undertaken at
times when the device would not be generating anyway, such as when there are no waves, the
availability could reach very high levels.

3.5 Annual energy production

The annual energy production is calculated by combining all of the above analyses.
1. The device characteristics for a range of sea conditions are defined based on numerical
modelling and testing.
2. The number of times that each sea condition occurs in an average year is estimated for a
given site and these are combined with the performance characteristics to estimate the gross
energy output.
3. The likely losses incurred transmitting the energy within the farm and to the shore are
calculated based on the location and distance to shore. The estimated energy output is
reduced to account for these energy losses.
4. The sea conditions, system complexity and repair procedures are considered to estimate the
amount of generation lost to planned and unplanned maintenance. The energy output is
reduced once again to account for this lost generation.
5. The energy produced by the farm is then calculated as a long-term yearly average.

B/15992/C001/065 May 2006

 9

Figure 3.1 Simplified process for calculating the energy production of a marine energy project

Measured sea conditions Numerical model of the device

Characterised sea conditions Predicted power characteristics

Seastate scatter diagram Power characteristics for the whole sea

9

5
8
3
5
7
5 2
Significant wave height [m]

3 3 5
6
5 5 2 5
3 8 2 5 3 3 5
5
3 5 5 3 10 2 3 2

3
8 10 5 5 5 10 5 5 5
4
7 8 3 5 2 5 5 3
6 15 8 18 7 10 3 3 2 5 5 2
3
10 13 35 22 28 15 17 2 5 5 5 3 5 3
10 13 22 20 25 30 12 12 8 8 7 2 5
2
2 18 15 30 20 13 10 12 15 8 5 2 3
3 3 7 10 15 15 22 22 15 8 7 2 5 3 2
1
3 2 5 3 8 10 12 10 10 3
3 2 2 3
0
4 5 6 7 8 9 10 11 12 13 14 15
Energy period [s]

Availability Efficiency

4 5
Some energy is lost in
Some energy generation is
transmission
lost when the device is
unavailable to generate

6
Electricity delivered to the grid

B/15992/C001/065 May 2006

 10

4. Cost

What affects the costs of marine renewables, and at what costs can electricity be generated
from waves and tidal streams today?
These questions were the starting point for our assessment of cost-competitiveness. This section
summarises the findings based on data gathered during the Marine Energy Challenge.
The cost of energy is an important parameter, but it is not a static one. As technologies develop
their cost of energy will alter. Some technological advances may reduce the overall cost of
energy, some may increase it. The aim is to produce an overall reduction over time. There can
be many routes to achieving this. Whichever is taken the cost of energy should be reviewed and
managed routinely. This means that technology developers and energy farm operators need to
revise their estimates often, as part of continual product development.

4.1 Key factors affecting cost of energy

The costs of energy of marine renewables technologies depend on several factors. Principally,
these include capital costs, operating and maintenance (O&M) costs and the amount of
electricity produced (performance). There will also be costs of decommissioning. Current
estimates indicate these will be small compared to initial capital costs, and because they fall at
the end of a project, the present value in a discounted cash flow analysis is low and has only a
marginal effect on cost of energy. Like wind energy, wave and tidal stream energy are free at
source so there is no fuel cost.
Essentially, capital costs and O&M costs must be weighed against performance, since this is the
saleable output and represents income to the generator. A high-performance device can afford to
be expensive if its costs are more than met by the value of electricity sold. However, if the costs
are so great that they exceed the income from generation, the device will not be economically
viable. The balance of costs and performance is manifested in the cost of energy, and the target
for this is the cheapest alternative: another form of renewable or conventional power generation.

4.2 Capital costs

The capital cost of marine renewables technologies can be broken down into: the cost of the
generation device itself (materials, components and labour in manufacturing and fabrication
processes); the costs associated with installing it (deployment); the costs of keeping it on station
(foundations or moorings); and the costs of connecting it to the grid (electrical cables and
switchgear). Some of these costs are more dominant than others, and the relative distribution of
cost centres varies between different device concepts and site locations.
It should be noted that the capital costs of wave and tidal stream energy devices are not static
and will change over time due to developments in technology, the costs of raw materials and
components and experience gained in manufacturing and deployment. As might be expected,
the total capital cost depends strongly on the number of devices built and installed, and also
where they are deployed.

B/15992/C001/065 May 2006

 11

This section describes the approach taken in the Marine Energy Challenge to estimate these
costs. Figure 4.3a shows an example of a cost breakdown for a wave energy device.

4.2.1 Structure costs

In many cases the structure forms the largest cost of any marine energy device. It usually has to
both interact with the waves or tides and support the power conversion equipment such as
gearboxes, hydraulics and generators.
The cost of the structure can be a simple product of the material weight and the cost of that
material. For example a large steel structure with a mass of 20 tonnes might cost around
£20,000 based on a steel cost of say £1000/tonne.
This approach is useful in some cases but quite often the complexity of the structure has a large
overall effect on the cost. Thus a simple lump of material costs less than a complicated
fabrication. Most often the unit costs for fabrication of relatively straightforward shapes are
collated and then the number of each type of shape calculated. This usually results in the
majority of the structure costing roughly the same per unit mass, but with a few components
costing significantly more. Such components might include joints between parts of the structure
or flanges and interfaces with other items of equipment.

Figure 4.1 Structural complexity

Simple Complex
100-tonne 100-tonne
structure structure

The Marine Energy Challenge used engineering experience and a build-up of unit costs, as well
as quotes for the construction of the whole unit from experienced fabricators. Both approaches
are needed. The first is simpler and allows the trade-offs between cost and performance to be
made more readily and the second is suited better to refined designs.
The experience of the fabrication industry is essential to developing low-cost solutions. Often
the most expensive manufacturing operations can either be simplified or eliminated completely
without significant changes to the overall concept. The advice of experienced fabricators was
particularly useful in the refinement of the Marine Energy Challenge devices’ structural designs.

4.2.2 Mechanical and electrical costs

The mechanical and electrical costs include all the items that convert the movement of the
device or the surrounding water to electrical energy. For example these can include the
hydraulic systems between two moving parts of the structure, the blades of a water turbine, the

B/15992/C001/065 May 2006

 12

gearbox that increases shaft speed, and the electrical generator that converts the motion to
electrical power.
The mechanical and electrical costs are strongly device specific. These components are sized
mainly according to the peak power output of the device. Thus a gearbox is sized such that the
product of its rated torque and its rated rotation speed is roughly equal to the power of the
device. Likewise the product of rated generator current and the rated generator voltage is
roughly the rated power of the device.
Determining the optimum power level of the device is an iterative process. It is generally not
economic to install a power conversion system that can convert the highest levels of energy in
the sea. This is because these high levels do not often occur and thus the device would be
underutilised most of the time. Instead a compromise is made and the device is rated at less than
the highest power likely to be seen. Occasionally some of the energy is shed and not extracted
by the device.
The ratio of the mean power output to the peak power is known as the capacity factor3. A
capacity factor of 30% means that the generator will produce the energy equivalent to 30% of
that produced if it could run at full load. The device cannot run at full load all the time because
there is insufficient power in the sea.

Energy output[kWh]
Capacity factor[%] =
Energy available[kWh]
Annual energy output[kWh]
=
Rated power of device[kW ]× time in the year[h]
Equation 2

Mean power output[kW ]

=
Rated power output[kW ]
In the Marine Energy Challenge the simplest approach used was to estimate roughly the best
balance between installing a more expensive and larger generator with the additional income
from the energy produced. This balance is different in each case. Once it is known, the sizing of
all the mechanical and electrical systems can be calculated from the overall device size.
In many cases after the design had been developed the balance of output to installed capacity
was reviewed. This led to either increases or decreases in capacity factor.
In reality the capacity factor of a scheme is not a design driver, but a consequence of the
balance of output and cost. Thus some devices had capacity factors of below 10% and others
above 50%. It is not possible to conclude which has the better overall economics from these
figures alone.
In many instances the design of the mechanical and electrical equipment requires new systems
and indeed some of the intellectual property inherent in the device designs involves these novel
systems. Often only guesses of the performance of these components can be made. In several

3
In conventional power systems the concept of the capacity factor is also applied. In this case it is
termed the load factor. The difference is that in renewable energy systems it is the fuel supply (e.g. the
waves or tides) that varies, whereas for conventional system the fuel supply is constant but the load
varies. The load varies as people turn lights and machines on and off.

B/15992/C001/065 May 2006

 13

cases, manufacturers of similar systems were contacted for advice. This advice included advice
on novel electrical systems, hydraulics, control systems and new materials.

4.2.3 Mooring costs

The moorings include all the components required to hold the device in place. For some designs
the moorings and the structure are effectively the same item, for example the monopile used in a
tidal energy turbine could also be considered as the main structure. However, in many cases the
moorings are quite separate systems that allow the device to move independently and are
required only to hold it on station and prevent it drifting. For many mooring systems their
design depends on the mean and extreme (storm) loads placed on them by the sea. This
contrasts with the design of the power take-off system that is optimised towards the average
conditions in the sea. The water depth, the tidal current and the tidal range also all affect the
design and cost of the mooring system. Moorings are therefore designed to suit both the
technology and the deployment location.

Figure 4.2 Moorings

4.2.4 Installation
The method of installation will depend on the nature of the device. The choice of vessels, for
example, will also change. Some devices can be towed to site using a tug and their anchors
placed using an anchor-handling vessel. Others might be carried on a heavy-lift vessel or a
barge. Piles and other structures might be positioned using stable platforms such as jack-up
barges that can float out to the site and then ‘jack’ themselves up out of the sea to form a
temporary platform.

• Heavy-lift vessels • Tug

• Jack-up • Anchor Handler

• Barge
The offshore and oil and gas industries have developed a large number of specialist vessels
(such as those above) for completing all types of work at sea. Tugs and other non-specialist
vessels are widely available and can be procured at less cost, whereas some of the specialist
jack-up barges are much more expensive and are harder to charter. Generally the more
specialised the vessel the greater the cost.

B/15992/C001/065 May 2006

 14

The costs associated with deployment are usually estimated with vessel charter rates. These are
usually daily rates. The day rates for vessels change in response to the demand for the vessels.
During the Marine Energy Challenge (2004-2005) costs for anchor handlers increased several-
fold due to increased demand for their services in the oil and gas fields of the North Sea.
The Marine Energy Challenge used long-term average rather than spot rates for vessel costs.
This means that some prototypes may well incur dramatically lower costs, or indeed unfortunate
higher costs depending on their timing and their ability to negotiate good prices.
Marine energy systems though are slightly different to other offshore operations. Many marine
energy systems are designed to be deployed in large numbers. This means that some will need
longer vessel-hire periods and will ultimately be able to either negotiate lower day rates or
indeed justify procuring their own dedicated vessel. Many of these options were considered in
the Marine Energy Challenge.
Deployment costs are also strongly related to the location. The sea conditions, tidal ranges, etc.
will dictate the choice and thus the cost of the vessel. The distance to port will also affect the
transportation time and the duration of vessel charters.

4.2.5 Grid connection

Estimates during the Marine Energy Challenge included the costs of all electrical connections to
the shore. They also included the necessary shore-based facilities to join the output of the
marine energy system to the land-based electricity grid. In reality other costs may be incurred
too, the shore based grid might need reinforcement or adaptation to absorb the new generation,
but this is certainly not true for every site around the world. Thus shore-based electrical system
upgrades were not included in the costs.
The grid connection costs include any cables, transformers and switchgear needed to connect
the offshore farm to the land. Generally the costs depend on the distance to shore, the ground
seabed conditions along any cable route and the power being transmitted.

4.2.6 Project management

All capital projects, such as marine energy systems, require project management. All the Marine
Energy Challenge cost of energy estimates allowed for some level of project management for
their schemes. Such costs are extremely difficult to estimate and so simple fixed proportions of
the overall costs were allocated to management in most cases. These allowances cover the costs
of managing the project as well as insuring some of the construction risks. Early projects are
likely to have very high project management costs, though these will probably reduce to levels
found in similar industries.

4.3 Deployment location

Many of the elements of the cost equation depend on the deployment location of the marine
energy system—

B/15992/C001/065 May 2006

 15

Table 4.1 The influence of the deployment location on the cost of energy

Cost centre Influences

Energy production Energy density at the site

Foundations and moorings Water depth, ground conditions, tidal streams, tidal ranges, energy density

Grid connection Distance to shore, ground conditions along any cable route

Installation Water depth, tidal streams, tidal ranges, distance to suitable port

The deployment location has a strong influence on the energy production and the costs of a
scheme. Thus suitable target locations for each device were decided early on in the Marine
Energy Challenge.

4.4 Operating costs

The O&M costs of marine renewables can be considered in several parts, including:
maintenance, both planned and unplanned; overhauls, where it is most economic to re-fit
components during the service life; licences and insurance to allow the devices to be kept on
station and to manage the associated risks; and ongoing monitoring of wave or tidal conditions
and the performance of devices.
Figure 4.3b gives a breakdown of O&M costs for a specific wave farm envisaged. Like capital
costs, O&M costs also depend on the size of the installations and the location, and are also
likely to vary from year to year. At present, it is much more difficult to estimate O&M costs
than capital costs due to the lack of experience in operating wave and tidal stream farms,
although it is possible to infer costs from experience with upstream oil/gas installations and
offshore wind farms.
Below we discuss planned and unplanned maintenance and how these affect energy generation
and thus project income. Here we describe how we consider the costs of these activities.

4.4.1 Planned maintenance

The costs of planned maintenance include—
• Cost of consumable replacement parts (e.g. new brake pads, replacement oil)

• Cost of servicing the vessel in terms of the time and personnel required

• Cost of waiting on the weather conditions to be right to allow servicing to take

place.

4.4.2 Unplanned maintenance

The cost of unplanned maintenance is considered in a similar way. Here though the costs are not
known, it is not known which parts will fail and when. Therefore the costs are an estimate of the
likely average for a project. The costs include—

B/15992/C001/065 May 2006

 16

• Cost of replacement parts (e.g. new brake pads, replacement oil)

• The costs of spares kept in case of failure (for small schemes only small spares are
carried (e.g. a puncture-repair kit), but for larger farms maybe entire devices can be
kept as spare (e.g. a spare racing car)
• Cost of servicing the system in terms of the time and personnel required

• Cost of having service equipment and personnel on stand-by in case of fault

• Cost of waiting on the weather conditions to be right to allow servicing to take

place.
Of course there is much that can be done to minimise these costs, and as the industry develops
equipment will become more reliable, the parts will be more readily available, more personnel
and equipment will be available at short notice to make repairs and better operation efficiencies
are possible.
The Marine Energy Challenge estimated some of the likely costs of unplanned maintenance, but
for the first few devices these costs were often very high. Since the Challenge was to estimate
not only the costs today but the costs in the future some assumptions about increased reliability
and operational efficiencies were needed. These assumptions were made by observing the
development of other industries, by theoretically trading off capital and operation costs (e.g.
including spare/redundant systems) and by investigating sophisticated options such as the
purchasing of bespoke vessels rather than relying on hiring available systems.
Wherever possible evidence for these areas of efficiency improvement was found and
demonstrated. Usually this resulted in either a reduction or an elimination of unplanned
maintenance costs. It must be stressed though that these assumptions do not apply to the first
few devices where we will still be learning about the reality of operating these new devices in
the real sea environment.

B/15992/C001/065 May 2006

 17

Table 4.2 Cost centres and cost drivers

Cost centre Main drivers Example measures and variables

Capital cost
Structure cost Material cost Cost per tonne, tonnes of material
Extreme loads
Mechanical and Rating of the machine (installed capacity) Peak power output, mean power output
electrical costs
Moorings Water depth
Tidal range
Tidal flow
Storm conditions
Compliance
Installation Type and availability of vessels required Vessel day rates
Distance to port
Time taken for installation
Time waiting on weather
Grid connection Power transmission level Cost per kilometre
Distance to shore
Project management Project management Proportion of the total capital cost
Insurance
Permissions
Operating cost
Planned maintenance Cost of replacement parts
Component design duty and known service
intervals
Time to complete service
Distance to port
Time waiting on weather
Unplanned Cost of replacement parts
maintenance Cost of spares
Time to complete service
Time waiting on weather
Cost of personnel and materiel standby

Figure 4.3 Cost breakdowns for an example wave energy device a) capital b) O&M

a) b)

Notes: Based on data gathered during the Marine Energy Challenge. The charts refer to specific types of
wave energy converter and are not representative or typical of wave energy technologies as a whole.
There are considerable variations between different technologies, project locations and project sizes
(numbers of machines installed). Also, future design improvements, performance/cost optimisations and
learning effects could change the relative weighting of some cost centres. The O&M chart shows annual
average costs evaluated over the entire life of a wave farm.

B/15992/C001/065 May 2006

 18

5. Uncertainty in estimates

Only when much experience has accumulated, both of developing and running marine energy
technologies, can good estimates of the cost of energy be made. Even then, estimates for
particular projects will still be uncertain. It is good practice when making any estimates to
estimate also the uncertainty in them. For example we might conclude that a particular item
might cost about £1000, but in reality could cost anything from £900 to £1200. We would say
that the £1000 middle estimate was our ‘best’ guess.
Some times we assign probabilities to these figures. For example we might conclude that there
is a 90% chance that the item above would cost less than £1200, but only a 10% chance that it
might cost less than £900. Sometimes we can work out these estimates using statistical
techniques, but often we have to make our own estimates of the uncertainties instead.
Whichever method is available, assigning a range with a nominal ‘confidence’ level provides
much more useful information.
Narrow confidence bands show more certainty in an estimate than wide ones. For example if we
were to estimate the cost of readily available raw material (e.g. steel) we might find it was in the
range £995-£1010, whereas a similar value of a less common material (e.g. silicon) might cost
£800-£1200. Both will cost the project around £1000, but we are surer of the steel costs than the
silicon costs.
Uncertainty information helps us identify which areas we need to investigate further, e.g. where
costs need to be calculated in more detail perhaps. It also shows where uncertainties will always
exist and where these need to be managed differently. For example the estimate of the long-term
energy output of a wave energy farm will always be uncertain because it depends on the weather
and the weather is extremely difficult to predict for the whole life of the project. This means that
project investors need to find other ways to deal with this uncertainty. One way is to treat it as a
project risk, in much the same way as they treat the ‘market’ risk of other products.
Combining estimates and their uncertainties can be difficult. If the estimates are calculated with
rigorous statistical techniques, then the best estimates and their uncertainties can be combined
statistically too. However, more often the estimates are not based on statistical observations and
additionally there are complex interactions and trade-offs between the different options. For
example, it is quite unlikely that all the costs (if estimated properly) will all be at their lower
end, neither is it likely that they will all be at their higher end. Thus adding all the low costs
together does not give a reasonable low overall cost.
Additionally, if one parameter reduces the cost of energy it might influence another to increase
it again. For example, if the wave resource turns out to be at the high end of the estimate, more
energy might be produced and the cost of energy lowered, but this might also cause an increase
in wear rates and thus more maintenance is required, hence raising the cost of energy again.
For very complicated systems models of the cost of energy can be constructed that allow for all
of these interactions. Once again, techniques such as Monte Carlo analyses can be used to
calculate both the best combined estimate and the uncertainty in that estimate.
It is good practice to make estimates of the uncertainty in all individual estimates as well as in
combined estimates.

B/15992/C001/065 May 2006

Appendix A
Sources of further information
1 Page

For more information see—

Carbon Trust
(http://www.carbontrust.co.uk)
Future Marine Energy – Results of the Marine Energy Challenge: Cost competitiveness and
growth of wave and tidal stream energy
Available from (http://www.carbontrust.co.uk)
Carbon Trust Marine Energy Challenge Newsletters
(http://www.carbontrust.co.uk/technology/technologyaccelerator/marine_energy.htm)

B/15992/C001/065