Metrics for Net Zero Energy Data Centres

October 22, 2014

RenewIT is pleased to announce the release of a public deliverable a report on relevant metrics to evaluate the performance of key energy concepts in Data Centres.

This report is in keeping with the aims of the project in taking a holistic view, covering energy efficiency alongside the use of renewable and natural energy sources, deployed capacity, adaptability of workload to the availability of renewable energy, re-use of energy, and economic impact.

RenewIT is part of the Smart City Cluster Collaboration, an initiative of eight European Union projects with the objective of defining common metrics to characterize the energy, environmental and economic behaviour of Data Centres. A selection of existing and new proposed metrics has been agreed by the Smart City Cluster for collaboration in testing, measurement and standardisation. These include basic energy efficiency metrics as Power Usage Effectiveness (PUE), measures of Energy Efficiency related to seasons of the year, and heat re-use indicators. RenewIT also makes use of other metrics alongside the common metrics adopted by the Smart City Cluster.

The report presents a methodology for cost-environmental analysis centred on energy use in Data Centres, analysing the economic and environmental impact of efficiency measures. There is a focus on non-renewable primary energy, although other metrics covering factors such as CO2 emissions or water consumption can also be considered.

Data Centres with a high ratio of renewable energy have the capability to export energy to the grid infrastructure in some seasons of the year or some hours of the day. In addition to the main metric – the renewable energy ratio – to evaluate the share of renewables in primary energy factors, new metrics are proposed to evaluate load matching and grid interaction. The term “load matching” refers to the degree of overlap between the on-site generation and the Data Centre load profiles; grid interaction refers to the energy exchange patterns between a Data Centre and the utility grid, and its impact on the overall load of the grid.

Capacity metrics, which relate peak powers with the installed and design capacities for both the IT room and the whole facility, are used to evaluate the unused capacity in a Data Centre. The unused capacity in IT and Data Centre infrastructures represents avoidable capital costs and operational costs, as well as having its impact in energy consumption.

As the best way to understand how metrics are computed is with actual numbers, some examples of different Data Centre configurations are provided, showing the method to compute the main selected metrics. Graphical representation is used to help Data Centre designers, owners, and operators to understand the influence of different factors.

Public Deliverable with the acknowledgement of the EC services. Content pending validation.