APPENDIX 34
Supplementary memorandum submitted by
Mr Peter Fraenkel, Marine Current Turbines Ltd, following the
Evidence Session of 21 March
RESPONSE TO QUESTIONS FROM THE COMMITTEE
RE. THE
COSTS OF
ELECTRICITY FROM
TIDAL STREAM
TURBINES, NOW
AND IN
THE FUTURE
It is impossible to give a simple short answer
to this question, since costs are effected by numerous parameters,
not least being the site conditions, the scale of the project
and the financial assumptions. However we have developed a techno-economic
model that can calculate the economics on the basis of calculations
that require several dozen parameters to be defined. The model
can synthesise a typical tidal stream resource on the basis of
several key factors, notably the "mean Spring peak current
velocity" and it can also synthesise a tidal turbine system
sized correctly to resist the loads and with all the components
having cost-functions that allow the overall cost of the systems
to be determined in relation to their performance. To make life
simpler we have a "baseline model" that utilises what
were judged as "typical" parameters, this model was
recently scrutinised by independent consultants, Binnie, Black
and Veatch under contract to the DTI, and their report broadly
endorsed our analysis and could presumably be obtained from the
DTI for any members of the Committee that wish to look into this
more deeply (although the report is still in final draft form
at the time of writing and the DTI may wish to delay releasing
it until it is in its final form).
Projected cost variation with time for tidal stream
electricity
Our detailed costing model results are plotted
against time in the graph below.
The capital costs will be reduced through the
R&D programme up to 2004 and the first commercial projects
thereafter. This is achieved partly through economies of scale;
eg installing a batch of 20 to 30 turbines typically reduces unit
costs by up to 50 per cent as a result of spreading the project
planning, design and installation cost overheads over more machines.
Hence the single rotor turbine in R&D Phase 1 will cost over
£4,000/kW but the twin rotor Phase 2 machine will be significantly
more cost-effective at £2,500/kW. A single Phase 3 machine
would be marginally more cost-effective than the Phase 2 one,
but, by installing a batch of five machines, economies of scale
bring unit costs down to £1,500/kW at the end of the R&D
programme—a level of cost-effectiveness which was only surpassed
by wind turbines since 1990.
After the R&D programme is completed, the
initial projects would involve batches of 20 to 30 systems and
these will have unit costs in the range £600 to £1,000/kW
which makes them immediately competitive with both offshore wind
and onshore biomass generation at the present time. These are
of course costings for early projects and they are conservative
since they are based on the initial design concept and make no
assumptions on possible further cost savings that may be achieved
as the technology is further developed. Most new technologies
can improve in cost-effectiveness quite rapidly; wind generated
electricity today is 25 per cent of the cost, in real terms, that
it was in the early 1980s when commercial wind farms were first
installed.
The above chart also gives the electricity production
unit-costs to be expected and compares these with examples of
costs of other generation technologies. It can be seen how the
capital costs translate so that the three phase R&D programme
yields electricity at 22, 12.5 and 8p/kWh for Phases 1, 2 and
3 respectively. It is expected that the first commercial projects
that follow will deliver electricity at between 4 and 6p/kWh and
that electricity costs will continue to fall to around 2 to 3p/kWh
for large scale projects, ulitmately making the technology cost-competitive
with conventional fossil fuelled power generation, but with the
added commercial advantage of all the credits that can be gained
for clean power generation. Costs of less than 2p/kWh could be
achieved in favourable situations given the kinds of improvements
experienced with other technologies.
The question obviously arises as to why we can
expect this technology to be so much better than other renewables
such as wind. The answer is primarily because this uses a high
intensity energy resource (typically capturing 6 to 10MWh/m2 per
annum compared with less than 1.5MWh/m2 per annum for wind for
each square meter of rotor in the flow); this results in a relatively
small size for a turbine of a given power rating combined with
what is potentially a low cost technology fabricated from steel.
A single 1MW tidal turbine rotor will typically be less than 20m
in diameter compared with a 1MW wind turbine which needs a rotor
of about 60m diameter. Apart from being small, tidal turbines
are built like ships, while wind turbines, where weight is critical,
need much more exotic and costly components. Low capital cost
leads to low electricity costs.
A key cost assumption for the longer-term commercial
product is that we can do better than £2,700/tonne in manufacturing
costs and that tidal current turbines need be no heavier than
150 tonne/MW (ie a production cost for the systems of £405,000/MW)
and that installation can be kept at an average level of £100,000
per turbine. These are reasonably modest targets since the "going
rate" for large ships manufactured in SE Asia at the present
time is in the order of £1,500/tonne including everything
(engines, electrical systems, radar, soft-furnishings for the
Captain's cabin, the lot) and a tidal turbine production run involves
a mix of steel fabrication and electrical components not dissimilar
from that in a ship (but lacking any soft furnishings!).
Given success with developing the technology
beyond the targets that can be envisaged at this early stage,
then even better cost-effectiveness than this may be envisaged,
to yield what may well be one of the least cost methods of electricity
generation so far invented—ie something comparable with
large scale hydro power. This is not entirely surprising since
marine current turbines are in fact a form of large scale hydro
power, but one where the environmental impact can be minimal.
Conclusions and comments
To answer the questions more specifically, it
can be seen that:
1. The initial unit costs of electricity
from single prototype systems will be in the order of 12 to 22p/kWh
(which is incidentally immediately competitive with electricity
costs from small diesel generators as used on small islands lacking
a grid connection).
2. In five years when the first commercial
projects are developed, probably around 20MW, we expect costs
to fall to around 4 to 6p/kWh (depending on scale of installation,
distance to the grid and strength of the resource at the chosen
location).
3. In less than 10 years we expect 100MW
scale projects with costs in the 3 to 4p/kWh range (or less, if
significant as yet unknown improvements in the technology occur).
4. In the longer term it seems reasonable
to expect unit electricity costs to "bottom out" at
around 2p/kWh for "Gigawatt-sized" projects.
It should be noted that all these cost assumptions
are based on:
— locations with mean Spring peak current
velocities in the 2.5 to 3m/s range (5 to 6 knots);
— a 3km primary "umbilical"
connection to the onshore grid;
— key financial parameters are 8 per
cent discount rate, 20 years amortisation;
— all costs are included to make the
results realistic, including profits for the developers, manufacturers
and contractors, plus insurance, seabed rental to the Crown Estates
(based on 2 per cent of margin from sale of electricity), EIA
costs, etc;
— O&M costs nominally set at 3
per cent of installed capital cost per annum.
To summarise all the cost assumptions would
make this explanation excessively long but we would be happy to
enter into more detailed explanations for anyone wishing to have
more detail on this.
It should also be noted that most of the cost,
as with all non-fuel burning renewables, is the repayment of capital.
The DCF analysis used assumed a 20 year life, but this technology
is analogous to hydro plants and once effectively developed the
technical life should be in the order of 40 years or more (although
certain components will of course need more frequent replacement).
If this seems surprising it should be noted that offshore defence
structures built during the 2nd World War still survive off the
UK's east coast and many offshore oil and gas industry structures
date back to the 1960s and 70s. The mechanical and electrical
component is technically similar to a low head hydro plant where
operational lives of 30 years or more are not uncommon. Hence
the electricity costs after the capital cost has been written
off will be very much lower, less than 1p/kWh, since the only
costs outstanding at that stage are O&M and decommissioning.
An important additional point is that a major
part of the cost of smaller projects relates to the overheads
involved in starting the project (surveys, permissions, engineering
and design, etc as well as the connection to the grid which can
be costly); it follows from this that the marginal cost of extending
an already existing project is also very much lower, since a large
part of the overhead costs will already have been invested.
20 March 2001
| 
