This site retreives 5-minutely feed from BMRS which provides data on the output by fuel type for the GB electricity supply grid. The data only include generating capacity connected to the transmission network.
Methodology – CO2 intensity
To determine an estimate of the current carbon intensity of the electricity supply we apply intensity factors (g/kWh) to each fuel type as follows:
| Fuel Type | Electricity Info | National Grid |
|---|---|---|
| coal | 937 | 937 |
| ccgt | 394 | 394 |
| ocgt | 651 | 651 |
| oil | 935 | 935 |
| nuclear | 20 | 0 |
| wind | 0 | 0 |
| hydro | 0 | 0 |
| biomass | 120 | 120 |
| pumped storage | 0 | 0 |
| other | 300 | 300 |
| Interconnectors | ||
| France | 68 | 53 |
| N Ireland | 458 | 458 |
| Netherlands | 475 | 474 |
| Ireland | 458 | 458 |
Most of these values are taken from Staffell, 2017. [1]
Other
From November 2017 the BMRS data feeds have separated out biomass from other.
For data prior to this, based on yearly aggregated government energy data, “other” is assumed to be 85% biomass with a carbon intensity factor (cif) of 120 and 15% other with a cif of 300.
Nuclear
These factors do not take account of emmissions from plant construction or fuel transportation let alone a full Life Cycle Analysis (LCA).
Given the widely varying data for LCA, particularly for nuclear power, it is beyond the scope of this project to attempt to apply LCA values to the carbon intensity.
In addition, this project is not designed to determine the lowest carbon energy sources, but to report the carbon intensity of the current fuel mix, so CO2 already produced and commited is outwith its remit.
While most non-direct emissions have therefore not been included, for nuclear power, where CO2 emissions from ‘burning’ the fuel are zero, we felt we could not ignore the considerable CO2 emissions directly related to the front end fuel cycle, which includes mining and milling uranium ore, and the conversion and enrichment process.
Arguably, transport emissions for biomass should similarly be included, but these vary widely. And the environmental case against transporting biomass fuel long distances is well made elsewhere.
The emissions related to mining and transportation of coal are not insignificant, but they are not the major element of coal’s carbon intensity.
Our cif for nuclear aims to make a conservative consideration of the front end of its fuel cycle.
Taking a figure of 66 gCO2e/kWh [2] for the LCA of nuclear, with 38% of that attributable to the front end of the fuel cycle (which includes mining and milling uranium ore, and the conversion and enrichment process) gives a value of 25gCO2e/kWh and a cif of 25. We have used a value of 20 to ensure that nuclear is not being treated unfairly compared to other fuel types where there are some emissions in fuel preparation and transportation that we have not included.
We have also adjusted the interconnector values from Staffell’s figures to allow for the nuclear component within these.
[1] Staffell, Iain. “Measuring the progress and impacts of decarbonising British electricity” Energy Policy 102, pp. 463–475 (2017)
[2] Sovacool, Benjamin. “Valuing the Greenhouse Gas Emissions from Nuclear Power: A Critical Survey” Energy Policy 36, 2950–2963 (2008)
Methodology – embedded generation
The BMRS feed does not include embedded generation.
In order to do so we have made a number of assumptions and simplifications.
Solar Power
We use the half-hourly feed from the Sheffield Solar project at Sheffield University to get the estimated GB solar output.
Wind Power
In an initial attempt to account for embedded wind turbines, we assumed that embedded wind power generates at similar levels to transmission-connected wind power, allowing for the higher load factors of offshore turbines.
Installed capacity for offshore (non-embedded) is 5471MW while onshore is 10105MW, of which 5143MW is embedded.
Taking an average load factor of 0.369 for offshore and 0.273 for onshore (source: renewableUK 5 year rolling average, July 2017), to estimate a multiplier for total output from the non-embedded output:
( 5471 x 0.369 + 10105 x 0.273 ) / ( 5471 x 0.369 + ( 10105 – 5143 ) x 0.273 )
gives an approximate value of 1.42.
However, checking this against historical data it overestimates total wind output, indicating that embedded turbines have a lower average load factor than non-embedded onshore turbines, with the average being around 1.3.
As an improvement on using a fixed multiplier, we compare the most recently available half-hourly data from BMRS for total wind output and compare that with the real-time data published for the same period and calculate a multiplier from the ratio of all wind to non-embedded and apply that to the current real-time data. If the most recent data is more than four hours old, we default to the average value of 1.3.
Hydro Power
Around 15% (240MW) of GB hydro capacity is embedded but we do not currently include this in our data as there is no clear relationship between the outputs from different hydro schemes, and the output from embedded hydro is not made available in a timeframe which would allow an approach like that used for wind power.
Other embedded capacity
We are not able to include any other embedded output in our data. These are relatively small compared to solar and wind power, only around 100MW in total is listed in National Grid’s Embedded Generation Register. Even if these generating sources could be incorporated in our data, it would not significantly alter the carbon intensity or fuel mix values.