Many data centre organisations have an opportunity to implement sustainability measures in quite a cohesive way in order to optimise their operations and minimise both carbon emissions and energy usage. 
 

Some of the components we have seen within such organisations include:

• Higher density compute, using for example next generation GPU chips
• Optimising energy usage in cooling, for example through precision cooling as one of the toolkit of equipment modifications or new infrastructure deployed
• Moves to implement next generation networks within the data centre
• Changing set points on equipment controllers to improve efficiencies and longevities
• Implementing monitoring of equipment, utilities and emissions, potentially with sophisticated algorithms driving optimisations, with tight control loops, predictive maintenance, and even AI options
• Minimising common space electricity usage, such as motion sensitive lighting, installing LED etc
• Integrating waste heat recovery from data centre cooling into district heat schemes
• Contracting renewable electricity supplies, even supplementing locally with solar panels on roofs
• Understanding and influencing choices around supply chain transparency, optimisation and resilience
• Minimising use of water, including harvesting water from buildings
• Using live walls, campus-wide tree planting, and broader diversified landscaping
• Adoption of sustainable construction materials and methods, including circular economy approaches
• Working smartly, with elements like remote monitoring, remote hands on incident management, remote meetings

Digging further, innovative techniques to reduce energy usage and carbon emissions have included spreading data processing and storage load patterns over periods of time. Also, deep modelling has been deployed (e.g. computational fluid dynamics) perhaps even combined with new measuring devices to capture new operational data.

Quite a list! 

So how can all this be organised?

Well, organisations’ annual budget planning cycle is one of the ways in which such sustainability measures can be endorsed and funded within tight budgets. The cases for each of the above components and others can then be tested, argued, prioritised, sanctioned, funded, and tracked to completion. As well as the impacts on energy usage and carbon emissions, resilience of the service will be of prime concern.

During this planning process, each adopted line item can be owned by the Operations and Environment teams as appropriate. Furthermore, line items can be explicitly tied back to operational and environment/sustainability strategies and roadmaps.

At the heart of a lot of the options we have seen, is a sophisticated and reliable monitoring system. Not only is that in its very design, but also in terms of the hardware needed to implement that monitoring.

In our view, monitoring can usefully be made to drive an automated management reporting process, bringing key information to the management team in an effective and timely manner. That way, decision making is supported by deeply relevant data, presented in a digestible way. As an added bonus, that management reporting process can serve ever tighter external reporting requirements, particularly in the environment compliance space.

There’s lots more to consider and explore to minimise energy usage and carbon emissions from a data centre. And, new technology, particularly in the AI-arena is continually coming forwards. 

Let us know if you think we might be able to help you consider some of the above.