Carbon payback

Carbon payback is the idea of how much CO₂ an electricity project will save or displace (including other greenhouse gases - GHG). The model is based on renewable power displacing a certain amount of power produced from other sources. It doesn't work on the basis of 1 unit of renewable electricity replacing 1 unit of conventional electricity. The actual vales cover quite a wide range. The amount of displacement is calculated using a counterfactual energy source, this gives an estimate of how long a project needs to be operational before it repays its own carbon impact. The trouble with building wind farms on peat is that the impact can continue long after the project has reached the end of its life. Each site is different. And each site needs to be treated on its own merits.

A model is a simplified mathematical simulation of what happens in real life. Provided the model is robust, suitably constructed and is supplied with accurate and pertinent raw data, the results give a way of estimating the carbon impact, and carbon payback of a project. Every project of any kinds has an initial carbon impact. For wind power the critical carbon impact is what kind of land the wind farm is built on, and what else needs to be done to get that power to consumers.

The Viking Energy wind farm uses a model created by Aberdeen University and Macaulay Institute, and is based on a mainland, Central Scotland site near a national grid connection. The model was funded by the Scottish government as an investigation to see the impact of building wind farms on peat, and for the time being, it seems to be the main model available. The original spreadsheet model is available here and the report here.

Spreadsheet models

These are the models Viking Energy have used in their wind farm carbon payback calculations. The model was not provided on their web site, but can be seen in printed form in their EIA. The source spreadsheet with default values is available from the Scottish Government web site. All spreadsheets are in Microsoft Excel format, so you'll need that to open them. The original spreadsheet doesn't automatically pull through all values from the input sheet, so check additional work sheets if you play around with the main input variables.

• See Viking Energy windfarm carbon payback model worst scenario. Basically uses 100m hydrology range. Grid mix payback period is 21 years. This is the official preferred measure, as backed up by ASA judgements. Original printed worst case version here.
• See Viking Energy windfarm carbon payback model intermediate scenario. Basically uses 50m hydrology range and a bit less ditching. Grid mix payback period is 5.2 years. Original printed intermediate case version here.
• See Viking Energy windfarm carbon payback model best scenario. Basically uses 10m hydrology range and minimal ditching. Grid mix payback period is 3.2 years. Original printed best case version here.

All scenarios use average peat depth of 1.6m. All claim 100% site and hydrology restoration, despite statements elsewhere that roads, all 118km of them, and all foundations will remain.

Change site restoration or hydrology restoration to "No" and you get a very different, carbon emitting project, in some cases into hundreds of years.

67 years carbon payback?

Viking Energy consultants, BMT Cordah, state in the Environmental Statement for Viking energy that the carbon payback could be as high as 48 years. This figure seems to come directly from using a 200m hydrology impact value, no other values were changed. So far the only response from Viking Energy is to call this a "mistype"! Perhaps what they really meant is that they forgot to remove the unwelcome news from the final version?

The input variables for this 48 years payback are on this spreadsheet. Bad as 48 years seems to be, this appears to be the "fossil fuel" equivalent. Use the proper grid mix value, and the carbon payback for 200m hydrology change becomes 67.7 years.

200m hydrology	Total payback time of windfarm (years)
Coal-fired	33.8
Grid-mix	67.7
Fossil fuel-mix	48

Key points

• Site is assumed to be upland peat - which fits the model
• Site is assumed to be Central Scotland - not an island 150 miles from mainland.
• Site is assumed to be near grid connection - Viking Energy connection is proposed to be made 150 miles away in Morayshire.
• Model used carbon displacement of power from coal, fossil fuel and estimated real (UK) electricity grid mix. Government and Advertising Standards Authority prefer grid mix value to be used as a realistic carbon displacement value. Viking Energy often quote the fossil fuel value rather than the correct grid mix.
• Model uses UK grid mix value, but Scotland actually has a far higher input of renewables power, principally from hydro power, so the Scottish grid mix may be lower than the UK grid mix value of 0.43
• Main model variables, i.e. the numbers which really count, are anything which impacts on peat and drainage. Change hydrology restoration to No and you get a very different picture, the "green" project becomes a net carbon producer! Build on suitable ground (like not peat) and you get much greener electricity.
• A number of the default values have been retained from the original model, some of the values used are "assumed" rather than actual measurements from the site
• A fundamental flaw in the model is the notion of site restoration - it is a simple yes/no. All or nothing.
• Similarly, the notion of hydrology restoration (i.e. what happens to water on the site as bogs, burns, lochs and pools) is also a simple yes/no. All or nothing.

Electricity counterfactual values

For electricity production, the basic fuel emissions are then multiplied by 2.6 to allow for inefficiencies in energy transformation, mainly due to heat loss as a whole for coal fired electricity and for transmission and other losses. A key factor is not how much electricity is produced, but the difference between production and end use. As far as we can work out, these values are CO₂ equivalent, i.e. net GHG (green house gas emissions including other green house gases such as sulphur dioxide and Nitrous oxide (N₂O)etc)

Electricity fuel source	Emission Factor (CO2/kWh)
Grid Mix1	0.43
Natural gas2	0.481
Fossil Fuel	0.607
Heavy Oil2	0.699
Coal Fired	0.86

1. For long term projects DEFRA / BERR advice is to use average grid mix displacement factor of 0.43. Some tables show annual grid average. This year it has actually increased to 0.537, meaning grid mix of each unit of electricity has actually increased its GHG content over the last few years, basically because coal has become cheeper than gas. Of all the main fossil fuels, gas has the lowest GHG emissions. Using gas at Sullom Voe power station instead of burning heavy oil at SSE owned Lerwick power station would dramatically reduce GHG emissions from Shetland electricity. Shetland currently gets up to 25% of its electricity from the gas fired Sullom Voe terminal power station, and between 0% to 20% from Burradale wind farm.

2. Heavy oil (also called bunker oil) extrapolated from fuel value CO₂ standard conversion fuel source CO₂ multiplied by 2.6 to get electricity production using that source. SNH report from Argyle Council Website. Natural gas from same calculation. Sullom Voe power station uses a similar fuel gas. See also Environmental Information Exchange, Oxford Brookes University.

See also Sep. 2009 conversion factors from the Carbon Trust or if you are feeling very brave, their source spreadsheet.