We spoke to Lewis Dunne, Head of the Legacy Modernisation Programme at the Modernising Technology Directorate within the Central Digital and Data Office.

In 2021 the UK Government Central Digital and Data Office (CDDO) added point 12, Make your technology sustainable, to the Technology Code of Practice to coincide with the COP26 conference hosted in the UK.

The following blog posts discuss how the CDDO along with DEFRA went about adding this additional point to the Technology Code of Practice.

• Help us make government more sustainable
• Creating a new sustainability point for the Technology Code of Practice

Procurement Policy Note - PPN 06/21 - Taking Account of Carbon Reduction Plans in the procurement of major government contracts.

Lewis’ notes:

Some of the things that I'm conscious of are, when we talk about service sustainability:

• Obviously ultimately this comes to down a carbon emission / kyoto protocol emissions breakdown at a system or service level. We don't have these metrics at the moment, and it's been really hard to do, meaning that in the meantime you essentially take an organisation level view of emissions. I think system level views are really what we should be aiming for though, but naturally it's hard due to the number of overlapping systems that could make up a wider service - you might afterall be trying to piece together a picture using a mixture of available datasets.
• I think a big thing here is honestly around transparency, metrics, and visibility - any data is better than no data, so org/dept level metrics are naturally better than nothing. I think the really big driver for this type of work will come over the year ahead as we embed sustainability metrics into things like QBRs - aka the HMG/Dept review process. Getting metrics is good, but we need to compare them against forward strategy / make them visible.
• I think longer term, the move to cloud should actually help with the metrics issue - we might be able to do some nifty stuff with asset tagging in the cloud that means we could basically pull a service level view together quite easily. Cloud providers are a bit iffy with their data on this though - Azure has a tool, but its v high level and the logic structure is a little opaque. AWS I don't believe currently easily provides this kind of data, and again - if your Azure and AWS emission calculating logic is different its a bit like comparing Apples and Oranges. I think we need greater consistency there
• In the meantime, probably the easiest way of doing stuff like this is:
  1. getting that organisational level metrics
  2. that kind of cost management best practice - especially in the cloud, cost = electricity = likely emissions.
  3. I think the biggest emissions we're seeing are from those untransformed mainframes / on-prem servers etc - transforming to cloud is a key part of getting the emissions down, even if we can't track emissions on a specific service level at the moment.