Estimates of levelised cost of electricity
The levelised cost method
The Levelised Cost Method was devised as a tool for assessing the potential for investment in different types of electricity generation. The basis of the method is first to identify the cost items that contribute to the overall cost. The cost of each item is then (a) estimated for each year of the investment period, (b) discounted to present value and (c) summed for all years. The energy generated per MW of installed generation is estimated and also discounted. The sum of all the discounted cost items is then divided by the discounted energy production to give the levelised cost of electricity (LCOE) - normally stated in £/MWhr.
In 2011, IESIS made available on its energy website a spreadsheet for calculating the levelised cost of different forms of electricity generation. The spreadsheet was written by IESIS fellow Colin Gibson who formerly held the post of Power Network Director with the National Grid Company in which role he was responsible for both the technical and the commercial performance of the National Grid
This spreadsheet includes a probabilistic method to assess the range of costs and includes estimates of what are called ‘integration costs’.
The ‘basic cost’ items are those paid directly by the generator including capital cost, operational cost and payment of debt. The integration costs include thoset for the operation of thermal plant that needs to be available to provide backup and stability control of the system and the cost of extra transmission. With high levels of intermittent generation in the system such costs cannot be other than significant. This is because two sets of generation are needed where one of these sets (the thermal generation) would suffice.
The values for the integration costs used in Colin Gibson’s spreadsheet are those that were available at the time that spreadsheet was prepared.
The covering paper gives a validation analysis of the estimates provided by the spreadsheet. This concludes that levelised cost are indicative of trends but should not be considered to be predictions of cost. Our position on this matter is that it is essential that the costs be estimated using the most reliable method that can be devised. This method would be based on a data model for the whole system.
'The Levelised Costs paper by Gibson in 2011 left many suggestions as to where it could be improved as more data became available. Dr Capell Aris has updated the 2011 levelised cost spreadsheet to include:
- a new worksheet for solar generation
- a new worksheet for carbon capture and storage (CCS) based on costs for the Canadian Boundary Dam project.
- new costs for onshore and offshore wind generation
- new costs for nuclear generation based on recent data
- a new macro that allows users to define variables that are associated with probabilities
Dr Aris worked in the Electricity Supply Industry first as reactor physics specialist at Wylfa nuclear power station, and then at Dinorwig and Ffestiniog pumped storage stations in the control and instrumentation section and later with additional responsibility for information technology systems. He is a Fellow of the Institute of Engineering and Technology.
A covering paper provides details for these amendents.
Load Factors used to Calculate Levelised Costs
The concept of Levelised Cost was developed to inform long term policy decisions. It is not appropriate to use the methodology for decisions regarding planning a particular power system. That requires knowledge of the load curve and the interaction with other plant on the particular system. For thermal plant, the plant availability is used as the load factor (there are no other limitations). For intermittent plant the plant availability is modified by the availability of the prime source of energy (wind, sunlight etc.). If the levelised cost calculated thus is used in further calculations, say, to provide an approximation of Total System Costs for a particular power system, the Levelised Cost would be modified at that point for that circumstance. For example, preferential running for renewables or nuclear would modify the estimated load factors
It should be noted that the chosen input figures and the handling of these data are those of the authors, and not necessarily those of IESIS