The Green Grid has introduced a new framework that organisations can use to assess the full environmental impact of their data centre throughout its life cycle in a uniform way.
With both public and private sector organisations under pressure from central government and industry regulators to reduce carbon footprints, environmental impact assessments are becoming increasingly commonplace.
However, the diverse range of often-incompatible approaches to assessments means that many organisations struggle to have a clear idea of exactly what their assessments should entail, according to The Green Grid.
The Data Centre Life Cycle Assessment Guidelines, based on the ISO14040 methodology, deconstruct the process of defining assessment boundaries, in order to establish which systems and operations should be included and excluded from the life cycle assessment (LCA).
For example, uninterruptible power supplies, transformers, and backup generators must always be included in the scope of an environmental impact study, but data centre lighting and fire-suppression systems should be included only if a significant impact is established.
Meanwhile renewable energy generators such as solar panels and wind turbines should only be included if renewable energy is accounted.
The report also clarifies the relevant stages of a data centre's lifespan, including manufacturing, transportation, operation, equipment upgrades, end of life and decommissioning. Each of these stages has its own associated environmental impact, which must be factored into an LCA.
"If we don't look at the complete picture, that means we hide something. In order to be honest, the study must look at everything and then focus in on something," said The Green Grid EMEA Technical Work Group member Christophe Garnier, who edited the report.
"The second important point is to try to avoid a trade-off. When you improve something, most of the time something else gets worse, so if you just focus on what's better it's not honest, unless you look at what is worse on the other side."
For example, if an organisation improves its renewable energy by adding solar panels, it can say its energy is greener, but manufacturing the solar panels and installing them on the roof also requires a lot of energy, and this process has to be factored in too.
The ISO14040 methodology also stipulates that, in order to conduct an accurate LCA, a "functional unit" must be established in order to define the scope of the system that will be evaluated by the LCA and quantify the service delivered by the data centre system.
The functional unit is defined by a combination of information technologies and facilities equipment, variations in system utilisation, and any special requirements around resiliency, serviceability, or availability.
"Within the same boundaries you can have many kinds of data centres, and if you just define the boundaries and have very different functional units, you will have different results," said Garnier.
An LCA also has to be comprehensive, according to Garnier. While lot of studies today focus on carbon footprint, there are many other factors to consider, such as water usage, biodiversity, air pollution, hazardous substances and even noise.
"If you take cold water from a river and send back hot water, there is a change in the river. There is an impact. So you must find a way to neutralise that impact.
Ultimately, the guidelines call for data centres to be assessed in a comprehensive and flexible way, so that environmental impacts can be mapped against life cycle stages to determine the relative overall impact of different environmental indicators.
Garner said that the framework will continue to evolve and improve, incorporating real-world case studies and collaboration with the organisations developing assessments to eventually arrive at common approved methods.