How can existing data centres be upgraded to increase efficiency and capacity?

Connectivity and data drive modern economies, with the demand for digital solutions only set to grow. At the same time, increasing awareness and concern about climate change means data centres are under attack for their energy use and high-cost operations. In response, hi-tech specialist Andrew Williamson explores innovative approaches to modernise data centres and enhance their sustainability.

The need to reduce emissions is at an all-time high. Currently, electronics account for about five per cent of total global energy usage, with the ICT sector predicting it will use around 20 per cent of the world’s electricity by 2025, contributing up to 5.5 per cent of global carbon emissions.

As technology advances and a billion more people come online in developing countries, that figure is likely to rise even higher, potentially hitting up to 14 per cent by 2040, as the Internet of Things (IoT), artificial intelligence (AI) and smart solutions, such as driverless cars, becoming part of our everyday lives.

All of this intensifies the power burden on data centres, which already consume over two per cent of the world’s electricity. Many data centres are designed with considerable redundancy to enhance uptime and availability, while handling potential peak loads that have yet to be experienced — building significant inefficiency into the facilities.

A two-fold solution

But given the urgent need to tackle high emissions, we believe there are actions data centre owners and occupiers can take now to increase their sustainability and secure a greener future.

The solution is twofold: firstly, where possible switching to renewable energy sources, including battery energy storage instead of combustion-based backup generation for short-term resilience; and second, modernising your infrastructure to improve the energy efficiency of servers, storage devices and other ICT equipment.

The benefits of modernised data centres

Upgrading legacy installations is an effective way to increase capacity without footprint increases, provided it can be achieved in an energy efficient manner. It also offers other crucial long-term benefits, such as strengthening competitiveness, reliability, safety, flexibility, environmental integration, and security and monitoring.

For example, modernised data centres are better equipped to minimise downtime and respond to incidents, via the careful design of the redundancy and resilience of power supplies and critical mechanical systems.

Through modular and scalable design, they can respond more effectively to changing customer needs and limit operating expenses. Power costs directly influence decisions to locate data centres, so accurately estimating the price of power both now and in the future is vital for modern facilities.

In addition, with increasing processing power comes an increased fire risk. modern data centres are equipped with fire detection and prevention systems and have effective evacuation and rescue built into the layout.

Using advanced technologies for cooling and heat recovery, modern data centres are better able to integrate into their community environments. Studying and redesigning — using analysis of factors, such as air flow, heat propagation, audible noise, and electromagnetic compatibility — is a key component of any upgrading initiative.

Crucially, as data becomes more and more important to customers, its target-size to criminal or terrorist organisations increases. Modernised data centres have globally integrated monitoring and security systems, both electronic and physical, to defend against all kinds of attacks and respond to incidents if they occur.

Five core steps to modernise your data centre

Working with clients, we’ve identified the following five essential steps to upgrade facilities smarter, faster and better.

1/ Mobilise an effective project team

The project team needs to include all relevant stakeholders from inside the organisation, as well as professional partners with experience in several key areas. These include finance and economics, architecture, planning and consenting, mechanical and electrical systems engineering, and utility connections, such as electric power, district heating, potable and wastewater.  An integrated, multi-disciplinary team will deliver a holistic design, which does not have built-in inefficiencies due to design margins at the interfaces between different components and sub-systems.

2/ Consider your options

There is no universally applicable process for deciding how to modernise a data centre. With many possible solutions available, it’s important that the project team evaluate their options carefully during the planning stage to make sure they select the more effective, value-led approach for their facility. Analysis and simulation tools — such as integrated safety-in-design, computational fluid dynamics, electro-magnetic transient simulations, thermodynamic models, and Monte Carlo reliability, availability and maintainability simulations —are available to support this process, quantifying reliability and resilience, safety risks and energy savings potential, along with a range of impacts on the surrounding area.

3/ Build a strategically focused business case

Although financial and economic analysis is an important part of the optioneering and planning phases, to help build a strong business case for modernisation, other factors also need to be considered. Alongside cost savings and return on investment, strategic factors, including brand-building, local community acceptance, future-proofing, and positioning in developing power and energy markets, and the so-called ‘triple-bottom-line’ framework, which measures social and ecological impacts as well as economic impacts, must also be included.

4/ Plan and deliver

Each component of the modernised data centre must be designed considering all interfaces and possible conflicts. Sophisticated tools, including BIM, finite-element analysis and numerical integration simulation solutions, can assess aspects, such as fluid dynamics, thermodynamics, electrical transient performance, electromagnetic compatibility, audible noise and arc flash risk. Our team uses these tools in planning, designing and delivering data centres for major global hyper-scalers and colocation providers to mitigate risks upfront and focus attention where needed as early as possible.

5/ Continually validate results and refine operational procedures

Rigorous and objective assessments of the real-world performance are often neglected in the euphoria of completing a project. Measuring energy savings, environmental effects and impacts on nearby uses of shared utilities are often required for regulatory approvals. They in turn, also inform improvements to operational procedures, which can lead to further savings across the whole project.

Taking a modern view

There are many hurdles for data centres to overcome in the coming decades, such as geographical implications, swamped markets and emerging regions. The challenges facing the high-tech industry are also complex: maintaining talent, minimising supply chain disruption, and implementing sustainable, environmentally sound solutions for research and development.

Success in this lightning-fast industry depends on speed, product lifecycle management and constantly improving operating processes to stay competitive. Modernising legacy facilities is one important way providers can stay ahead, increasing reliability, security and safety and building a sustainable future for years to come.

Unlocking value and efficiency

By integrating the engineering and analysis function into a single project organisation, our team has worked to strengthen the optioneering, planning and delivery stages of datacentre modernisation projects.

For us, interaction and optimisation between sustainability engineering, utility connection designs, mechanical and electrical engineering, and a broad range of technical and environmental functions is key to unlocking value, which can be difficult to achieve when you use discrete, specialist scientists and engineers.

For example, you can improve performance — when connecting electrical utilities and their interfaces with the internal power distribution system — by designing the external and internal utilities together.

 This approach helps to:

• Deepen understanding of the energy storage required for likely utility interruptions, so that the project team can assess the applicability of low-carbon technologies for resilience. And our team has assisted clients to investigate large battery-storage facilities that can reduce run-times for back-up generation, providing resilience at a lower carbon cost.
• Design the utility interface to improve the fundamental reliability of the utility power supply. We often work with clients to obtain custom supply configurations with distribution network owners, ensuring proper alignment of the internal and external assets. This means, for example, that a fully redundant internal network isn’t connected to supplies with a single point of failure, when options exist to connect to independent points of supply.
• Clarify and improve the coordination of protection systems all the way from the high-voltage network through to the low-voltage equipment, and integration of controls to give instant visibility of the status of the complete system.
• Eliminate or manage the effects of mains-borne interference, including switching surges, voltage sags, and harmonics, in both directions. This will reduce the effects of external disturbances on the data centre, as well as the emissions from the data centre into the distribution network.