AI/ML cools Telehouse North DC

08 May 2023

Telehouse is a leading global data centre service provider, bringing together more than 3,000 business partners including carriers, mobile and content providers, enterprises, and financial services companies. The company provides reliable, secure, and flexible colocation, enabling organisations to accelerate speed to market and create business opportunities through fast, efficient, and secure interconnections.

Telehouse puts sustainability at the heart of its data centres. The company strives to enable best practice operational performance across its data centre estate, concentrating resources where the most significant environmental improvements can be achieved. In addition to taking advantage of 100% renewable energy and complying with ISO standards for environment and energy management, Telehouse already takes full advantage of cloud, virtualisation, and innovative cooling to contribute towards efficiency. The company recognises that improving cooling efficiency is one of the most effective ways to optimise performance and secure environmental improvements.

Expanding insightful data
Telehouse needed an environmental strategy to deliver against two challenges – enabling improved data centre performance while securing carbon reductions.
Accordingly, Telehouse engaged its services partner CBRE to identify next steps for its data centre performance optimisation programme. CBRE advised that Telehouse engage EkkoSense to deploy the EkkoSoft Critical AI-powered monitoring software. Backed by EkkoSense’s specialist cooling optimisation skills, EkkoSoft Critical was considered an excellent solution to help Telehouse accelerate cooling and airflow optimisation across its UK data centres.

The goal was to secure continuous performance improvements around visibility, efficiency, and resilience, with the initial phase at Telehouse North – the company’s oldest facility that opened in 1990. Target project outcomes included optimising cooling performance, removing potential thermal risk, and unlocking quantifiable carbon savings to support Telehouse’s carbon reduction goals for 2022 and beyond.

The ML and AI-powered EkkoSense optimisation software works by monitoring, visualising, and analysing the performance of data centre facilities. It analyses thousands of temperatures and cooling points across the site in real-time to identify where levels of cooling can be tweaked, and dramatically increases the level of insightful data available to the operations team to remove risk and improve resilience.

Sensor deployment across the data halls provided Telehouse with the ability to monitor and identify performance improvements. If the temperature in a section of the data centre is outside the normal range, the sensors will flag this. The EkkoSoft Critical 3D visualisation and analytics platform continuously provides advice to the Telehouse team about adjusting cooling settings such as fan speed adjustments, cooling set points, floor grille placements etc. – resulting in quantifiable cooling energy savings and a reduction in carbon emissions.

Presenting all this ML data in a comprehensive 3D view makes it much easier to visualise the complex thermal performance of the thousands of racks deployed. With the EkkoSense software now collecting 3,000 data points every five minutes, the millions of data points already collected contribute directly to the effectiveness of ML algorithms to support continuous improvement in cooling energy usage and overall energy savings.

Addressing corporate sustainability requirements
Deploying EkkoSoft Critical at Telehouse North has performed ahead of target and succeeded in improving quantifiable energy savings and carbon emission reductions.

“It’s very difficult to manually inspect every element of a data centre to identify inefficiencies and make improvements,” said Telehouse’s senior operations director, Paul Lewis. “EkkoSoft Critical provides us with a highly granular level of data and visualisation to help support our green agenda - and ensure our customers meet their sustainability targets. We’ve already made significant carbon emission reductions from our initial rollout of EkkoSoft Critical at Telehouse North, and we’re eager to implement the software around our wider campus to extend these capabilities.”

On first engagement, EkkoSense identified that a potential 10% saving against cooling load saving was achievable from optimising data centre cooling at Telehouse North. This cooling energy saving will enable Telehouse to achieve an anticipated 461 tonnes reduction in CO2 emissions.

The success of the trial forms a new chapter in a long line of sustainability wins for Telehouse. The organisation powered the world’s very first multi-storey adiabatic cooling system in 2016, its North Two data centre is certified to the BREEAM excellent standard, and all its sites are supplied with 100% renewable energy. Rolling out EkkoSense to Telehouse’s other London sites will help extend the company’s focus and commitment to ongoing sustainability.

EkkoSoft Critical also proved to be especially valuable for Telehouse at a time of record-breaking temperatures in the UK. During July and August 2022, the software helped Telehouse’s data centre team to monitor and protect equipment and maintain uptime. The 3D visualisation enabled Telehouse to identify potential thermal hotspots before they became an issue during the 2022 heatwave. EkkoSoft Critical was also able to highlight areas that were being over-cooled, presenting further opportunities for energy saving and carbon reduction.

Telehouse is now exploring how the solution can be used to improve capacity management processes, identify any capacity constraints, and better quantify available capacity.