IT Sustainability Think Tank: How IT leaders can balance innovation with sustainability

Cloud computing underpins countless applications, services, and company initiatives. However, the environmental footprint of these activities has started to come under scrutiny.

Today’s IT leaders find themselves at a crossroads. On one hand they must embrace innovation - which includes using AI, machine learning, and big data analytics - to propel their organisations forward. And they must also grapple with the environmental consequences of datacentres, energy consumption, carbon emissions, and e-waste.

So how can IT leaders strike the right balance? 

The first step involves drilling into the finer details of the organisation’s cloud and datacentre assets. Whether on-premises or in the cloud, datacentres are the beating heart of modern IT infrastructure. However, their energy demands, emissions, and potential environmental impact are substantial.

Huge amounts of electricity are needed to power datacentres and their all-important air-conditioning systems. In North America, this power is often drawn from “dirty” power grids which are powered primarily by fossil fuels – thus leading to particularly high carbon emissions. While not all datacentres are identical, some now consume the equivalent electricity of 50,000 individual homes.

As of 2022, the global emissions from cloud computing now exceed those from the commercial aviation industry.

The overall carbon emissions from cloud computing are estimated to comprise 2.5% - 3.7% of the global total compared to 2.4% from commercial aviation. However, the considerations are wider than solely electricity and carbon emissions.

An oft-overlooked part of the datacentre sustainability equation is the use of water within cloud datacentres – but this is equally critical and impactful and requires concerted focus from business leaders too.

Hyperscale datacentres are increasingly adopting water-based cooling systems as they are more effective, more efficient, and more scalable. However, while they may reduce carbon emissions, there are other environmental concerns about water consumption too.

Datacentres can consume astronomical amounts of water. For example, Google’s datacentres are believed to use approximately 550,000 gallons of water every day – and that’s over 200 million gallons each year. This is especially concerning in regions experiencing water scarcity, and where climate change is driving up global temperatures.

Datacentres also often use chemicals (such as biocides and corrosion inhibitors) in their cooling systems to prevent microbial growth and maintain system efficiency. These chemicals can potentially leach into wastewater, causing issues for surrounding land and populations.

Datacentres generate wastewater through cooling tower blowdown, equipment cleaning, and other necessary activities. The chemicals present in this wastewater can impact local ecosystems if not treated effectively and require proper disposal to avoid environmental harm.

Generative AI, meanwhile, carries a significant environmental burden - with a generative AI web search using five times the power of a traditional web search. As organisations test and adopt generative AI technologies, both in the datacentre and on the desktop, it is imperative to track and analyse how this affects an organisation’s environment impact. A cost and benefit analysis should also be conducted to ensure the use of Gen AI is still worthwhile.