Generation Share - Erroneous Figures Seemingly Excluding IFA2/Other

[doig007]

The Generation component (https://api.carbonintensity.org.uk/generation) appears to have an issue relating to non-inclusion of IFA2 and Other generation.

Issue

Recognising National Grid methodology may include some embedded generation not covered by Elexon, National Grid's generation share figures diverge significantly from the Elexon actual generation mix (https://www.bmreports.com/bmrs/?q=generation/fueltype).

Diagnosis

Looking at two recent settlement periods as examples (one with interconnector imports, one without), it is possible to reproduce National Grid's Generation Share figures by excluding IFA2 and Other from Elexon data. See below tables.

To put it simply, in the first example with c.4GW of imports there is no way the import generation share in 10.9% as that would imply generation of c.37GW in settlement period 1.

Other

Notwithstanding this issue, it seems to be impossible to reproduce the actual carbon intensity figures using the methodology National Grid has published because of assumptions made (rather than sensitive data). This is not very transparent.

For example:

• Assumed T&D loss factor [references provided use 8%]
• Assumed carbon intensity of Belgian imports [Nemolink has been fully operational since early 2019] and Northern Irish imports [Moyle]

API download reference:

https://api.carbonintensity.org.uk/generation/2021-01-25T00:00Z/2021-01-25T00:30Z

[doig007]

Impact

Because I cannot reproduce the reported actual carbon intensities, it is not possible to assess whether this impacts them too.

Please can the authors confirm.