Planning for a Data Centre in London: Are you Air Quality Neutral?

Air pollution in London is widespread and as a significant public health issue it sits high on the political agenda. Part of a number of measures designed to tackle this, the London Plan dictates that major developments (over 1000 m²) must be at least Air Quality Neutral.

Failure to demonstrate that the data centre is Air Quality Neutral runs the risk of planning permission being denied. So understanding the air quality impacts and how to mitigate these is a crucial step in ensuring a resilient and efficient planning strategy.

The challenge

Data centres do not fit comfortably into any particular Use Class and no definitive class has ever been established. Most local planning authorities (often influenced by precedent) therefore tend to categorise data centres as B8 (storage or distribution) and use this to set the benchmark for the Air Quality Neutral assessment.

The issue is that a typical storage or distribution facility will have much lower building emissions than a data centre can realistically achieve—mainly due to high-emitting diesel back-up generators. Required to avoid the risk of any power outages and for use only in emergency, these generators spend the majority of the year switched off but are fired up at regular intervals for mandatory testing. Unfortunately, even when only used for these short periods diesel generators account for the vast majority of data centre emissions.

So when assessed against a typical B8 Use Class, a data centre’s emissions will almost certainly exceed the benchmark. And for data centre developers—because emissions are tied to the mandatory testing of the generators—it’s often not possible to reduce them, creating a real risk of planning refusal.

Is this method appropriate for data centres?

The ‘Air Quality Neutral’ policy (the basis for how the benchmarks are calculated) in the London Plan is designed to address the problem of multiple new developments that individually add only a small amount to pollution but together lead to baseline air pollution levels creeping up. This suggests that the policy primarily relates to residential, mixed-use developments, storage/distribution centres, hotels and schools—which each only generate small emissions but can have a cumulative effect.

For those developments, it’s reasonable to assume a relationship between the floor area (or the number of dwellings) and the emissions produced by heating and powering the space i.e. the bigger the space you have to heat, the higher you can assume the emissions will be. You can then set a benchmark accordingly. But in our experience, data centres don’t fit neatly into that assumption. 

Data centres are (and should be treated by the planning system as) a unique class. They have entirely different and intensive power requirements to run the network infrastructure and servers as well as cooling processes. Their power usage is not directly related to heating a certain floor space. So is it appropriate to require a data centre to meet emission benchmarks set for very different types of facilities with significantly different power requirements? Arguably not.

How we are helping clients

On behalf of our clients we’ve questioned whether the Air Quality Neutral calculation is designed to be used for data centres. Instead, we point to the more critical issue—and core premise of air quality assessments—of whether emissions from the development are likely to have an adverse impact on surrounding land-users.

This will of course vary and be specific to each site. Using an atmospheric dispersion model, we can advise clients on the optimum height for the generators’ stacks to adequately disperse pollutants; ensuring no adverse impact on the surrounding area. With this approach, we’ve been helping clients to overcome the challenges of the Air Quality Neutral Assessment and secure planning permission.

For a chat about how we can help you, use the form below.