Why the University is focusing on data decluttering
DTS are running a campaign to encourage colleagues to support a data de-cluttering campaign both to help the University avoid significant cost increases in 2026, as well as decrease the environmental impact of data storage. But just what is the environmental cost of storing data? Whilst we know there is one, it can be very difficult to quantify in practical terms.
How digital data storage uses energy
Data is stored on local or remote (cloud) servers which consume energy. The server rooms or data centres in which they are housed also consume energy for building services, including cooling which is generally a significant load.
How reducing data storage benefits the environment
Reducing the data we store digitally therefore has a positive environmental impact by scaling down the number of servers needed, thereby also reducing cooling demands and potentially energy from other building services.
Why it’s difficult to quantify the environmental impact
Quantifying the precise environmental benefit and carbon savings can be challenging:
- Both servers and the cooling systems used to cool server rooms have varying efficiencies depending on the technology used and how they are managed, so their energy usage can vary significantly from one server room to another;
- Servers and data centres support more than one piece of data so their total energy use relates to all the data supported making it is impossible to quantify the kWh or carbon emissions for individual pieces of data;
- Once data storage has been reduced, there needs to be a process of reconfiguring servers to improve their efficiency, and potentially the space they are located in, in order to realise the energy, carbon and (energy) cost savings available. Where servers are housed locally in the University, this is managed by DTS. Where servers are held in the cloud, other providers such as Microsoft would be responsible for management of these processes.
Measuring data centre efficiency using PUE
Power usage effectiveness (PUE) is a metric commonly monitored in data centres which is generally expressed as the ratio of overall consumption by the data centre facility (including energy for building services) to the energy consumption associated with computer power. The closer this ratio is to 1, the more efficient the data centre is as ancillary power for loads other than computing power are kept as low as possible. Whilst many data centres use this metric to monitor the efficiency of their overall data centre, it’s not information which is generally easy to get hold of from providers.
Taking action to reduce the environmental impact of data storage
In conclusion, there is no doubt an environmental benefit to the reduction of data storage at scale, even if it’s challenging to give a straightforward tCO2 per GB of data saving.
You can take action to help improve the environmental impact of data storage by supporting DTS’s data decluttering campaign which encourages University staff to delete old files that are no longer needed. You can find full details of the campaign on the staff portal here: Data management campaign launch – University of Reading
