Neil Potter and Grant Bilbow at Eaton EMEA explore how a systems-engineered approach to data centre design can enhance sustainability, energy efficiency, and meet rising demands from AI and IoT technologies.
As the UK economy continues on the slow and winding road to recovery, businesses are searching for ways to grow sustainably and efficiently. Naturally, this has led them to adopting technology such as AI, IoT and cloud services – all of which depend, and are putting more strain, on data centres. So much so, the UK’s National Grid expects data centre power usage to increase six-fold in the next 10 years.
However, in light of growing environmental reporting, data centre operators find themselves under increasing pressure to find ways to meet such urgent demands sustainably. The revised European Union’s Energy Efficiency Directive, for instance, introduced an obligation for member states to monitor the performance of data centres.
It’s clear that the industry needs to rethink the way data centres are designed, built and managed. We need to adopt an approach that enables operators to minimise environmental impacts, whilst also optimising their performance and reliability in an era of increased demand.
Introducing a systems-engineered approach
A systems-engineered approach to data centre design presents a clear way forward to unlocking truly sustainable progress.
Methodical and multi-disciplinary, a systems-engineering approach considers the data centre as a complex system composed of interrelated and interdependent elements, rather than a collection of isolated components. It moves away from the conventional practices of mixing and matching, and instead ensures all parts of the data centre – such as power, cooling and IT infrastructure – work together efficiently throughout its lifecycle.
Taking this approach enables the sustainable design and management of data centres by equipping operators with four essential abilities:
- Flexible and dynamic design
Improving the environmental performance of a data centre – whether it be its carbon footprint, water consumption, or waste management – is a challenge when you cannot comprehensively review the full picture. If you don’t know how components interact with each other, let alone the external environment, any changes you make to the system could deliver unintended externalities, undermine synergies and generate trade-offs.
At the same time, the need for agility is apparent, with emerging technologies generating variable demand patterns. To effectively address challenges associated with variable loads, operators must have access to a full system view.
By looking at data centres as a unified whole, you can identify the changes required in design and the best way of making them. All while understanding their impact and optimising flexibility, meaning you can ensure design remains effective in fast-evolving environments.
- Prioritising energy efficiency
Ensuring you are able to review data centres in their entirety is particularly useful when looking at energy efficiency, as it enables the review of the power train and its associated interdependencies across the full lifecycle.
Efficient use of energy is only going to become more important, as the proliferation of emerging technologies puts even greater strain on the power supply to data centres. Morgan Stanley analysts predict that generative AI alone will drive more than three-quarters of global data centre power demands in 2027. Therefore, it is essential operators renew their focus on efficiency, to ensure they can handle increased demands whilst responding to increased environmental scrutiny.
With greater visibility of both the distribution of power and the conservation of power consumption downstream, data centre operators can understand each components’ characteristics and impact on electrical properties, voltage, current capacity and impedance.
At the design stage, for example, selecting and considering the right equipment with the entire power chain in mind, enables improved energy efficiency. For example, in low voltage systems, using copper busbars will reduce power loss by around 25% compared to aluminium ones.
Along with this, a digital software platform can monitor and manage energy efficiency through machine learning and AI to better understand where losses in the power distribution system may occur and how they can be prevented.
- Integration with renewables
Improving energy efficiency alone will not meet demand, nor progress sustainability ambitions. Instead, attention must also turn to renewable energy sources, which are ever more important in the power ecosystem.
To do so, it is essential to understand the impact of integrating renewables and alternate power sources on whole system performance and power quality. Done properly, this helps to provide resilient and reliable power, and reduces the likelihood of outages.
One such impact we need to understand is the impact of less rotating mass and less inertia within renewables, which in turn affects the quality of power flow along the power train due to less frequency control and more volatility. A systems-engineered approach means you can observe the mix of renewable on-site and off-site generation, monitoring the power being consumed and its source.
- Rethinking measurement
Whether integrating renewables or improving existing power trains, data centre operators will no doubt question how they can demonstrate the holistic impact of such changes.
Traditionally, efficiency has been determined using metrics such as Power Usage Effectiveness and Data Centre Infrastructure Efficiency. But such measures are only concerned with one aspect of a data centre’s operation, to the exclusion of all others.
In the face of greater environmental scrutiny, we must look beyond these ‘standard’ measures. From measuring OT and IT efficiencies to equipment lifecycle and water usage, there are several other metrics that must be included to calculate a holistic metric of sustainability.
With a systems-engineered approach, each individual aspect within a data centre is already measured and consolidated by a digital layer that considers the interdependencies of each aspect to determine the most efficient and sustainable data centre. From this, it’s possible to devise a new maturity metric tool that, by combining all these metrics, can report on an entire system’s environmental integrity.
Looking forward
Ultimately, a systems-engineered approach rewires data centres to become more sustainable and competitive in the digital age. Equipped with a holistic view of the data centre system, where each component is intelligently integrated and coordinated by a digital layer of data, operators can expect greater efficiency, resilience, scalability, and adaptability, while also minimising their carbon footprint, energy consumption, and waste generation.
When designed with sustainability in mind data centres can not only reduce their environmental impact, but also optimise their performance and reliability in an era of increased demand.
