How Microgrids For Data Centres Increase Resilience, Optimise Costs, And Improve Sustainability

How Microgrids For Data Centres Increase Resilience, Optimise Costs, And Improve Sustainability

How microgrids use advanced analytics to intelligently manage energy assets is explored. By Carsten Baumann, Schneider Electric.

Resilient power infrastructure has been a top priority from the very early days of computing capabilities. Computer labs, data network closets, and mainframe computer rooms all evolved their power supply perspectives alongside the compute capability.

As demand for more and more compute flexibility expanded proportionately with the advent of the Internet and dot-com era, the architecture of our former computing infrastructure evolved into what we now call data centres. 


The Internet, and its associated cloud services providers, make data centres one of society’s most critical resources.

Today, the data centre industry worldwide faces a number of business and technical challenges. Of these, three energy-related concerns loom large for operators: resilience, costs, and sustainability.


Ensuring Uptime And SLA Compliance

Resiliency and reliability are often used interchangeably, yet they are distinctly different. By way of example, Microsoft explains that, “Reliability is the outcome cloud service providers strive for – it is the result. Resiliency is the ability of a cloud-based service to withstand certain types of failure and yet remain functional from the customer perspective.”

The reliability, ie: continuity, of electrical supply is paramount for data processing. From an electrical infrastructure point of view, this means not losing power, making sure that critical computing equipment is always operational and in compliance with all service level agreements. 

To help ensure this, a set of infrastructure standards were developed by the Uptime Institute to guide the design and measure the performance of data center power infrastructures.

However, increasing power demand, aging electrical transmission infrastructures, more frequent violent storms and other natural or human-caused disasters are all making grid stability issues more common in many regions. 

Ironically, bolstering generation capacity on the grid by adding renewable energy resources (such as wind and solar) is also causing greater grid power variations, as these sources are intermittent and can lead to voltage and frequency variations.

The typical power resiliency strategy for data centres has always been to provide backup generators, predominantly diesel generators. 

When grid disruptions have occurred, there have been, unfortunately, cases where data centre backup generators did not reliably start up as expected. 

Due to these factors, relying on the grid for all primary power increases operational risk. And even with traditional diesel-powered backup generation and Uninterruptible Power Supply (UPS) in place, there is a need for higher resiliency over longer, sustained periods.


Budgetary Pressures

Beyond 24/7 operation, energy cost is also top-of-mind for data centre operators. This is being driven by three realities:

  1. Data centres are huge consumers of energy – as much as two percent of grid power is deployed to support them.
  2. Many colocation providers are expanding their facilities and adding more clients and cloud services.
  3. Roughly 50 percent of data centre OPEX (exclusive of IT equipment) is the cost of electricity, and the price of energy will rise in the short-term and then flatten or decrease in the long term.

This is putting limited OPEX budgets under pressure, forcing operators to seek ways to reduce energy-related operating costs. HVAC energy consumption is itself a key contributor to the cost of operating a data centre. 

Power Utilisation Effectiveness (PUE) can only be reduced to a minimum associated with the performance of the cooling system(s) deployed. 

Even if the very most efficient cooling solutions are used to extract compute power heat, lowering the energy cost to operate a data centre cannot be less than the cost of the sum of the computing equipment power being used.

Therefore, to further reduce the cost of data center energy requires reducing the cost of energy from the grid or producing energy below the cost of grid power.


Meeting Sustainability Goals

As it is for many industries in many regions, meeting self-imposed environmental goals or, in some regions, environmental regulations, is an ongoing challenge for data centres. 

Electricity and fossil fuel consumption are both part of the formula in calculating greenhouse gas (GHG) emissions. Managing consumption and using greener energy sources is often a big part of complying with regulations. But the benefits go well beyond meeting government mandates. Minimising a building’s carbon footprint can also help achieve green building certification and establish a ‘greener’ image among prospective clients.

Additionally, customers of data centres increasingly demand greater sustainability of their operations that one can turn into competitive advantages. 

Up to now, those demands have frequently been met through the acquisition of Renewable Energy Credits (REC) and Power Purchase Agreements (PPA).


The Emergence Of The Data Centre Microgrid

A complete microgrid solution intelligently coordinates a variety of onsite, distributed energy generation assets to optimise costs and power stability, including the option to ‘island’ from the utility grid to avoid exposure to outages or disturbances. 

When the cost of grid energy rises, the microgrid can increase consumption of onsite renewable or stored energy. 

Stored energy can also be sold back to the grid when most economical. And consumption of renewable energy can be maximised to meet greenhouse gas emissions targets.

With their need for large amounts of continuous, clean, and affordable power, data centres are excellent candidates to benefit from microgrids. And there has never been a better time to take this step forward. 

Microgrid technology has reached a high level of maturity, being adopted in many types of facility and infrastructure applications, such as: utilities, community services, government offices, military bases, large industrials, hospitals, and educational campuses. The latter of these often include research facilities.

Worldwide microgrid capacity is anticipated to grow by more than 20 percent per year. Driven by previous massive growth, the overall cost of installing microgrids has dropped an estimated 25 percent to 30 percent since 2014, and is expected to continue on that trajectory.

Still, microgrid applications are unique for each organisation and, therefore, a feasibility study should be performed to determine the organisational benefits, including the investment versus estimated financial payback and potential operational gains including improvement to resiliency.


Smart Microgrid Architecture

The adoption of microgrids has grown in recent years. They have also gained a significant amount of publicity, such that their nature and purpose are much more widely understood. 

A microgrid is a localised energy system that interacts with the utility grid, encompassing one or more electric power generating resources and the necessary energy management controls to provide secure electricity to consumers.

In contrast to large utility grids, microgrids locate all energy assets – from generation to loads – in close proximity, to serve multiple buildings or even be contained within a single facility or parts of a large data centre.

A microgrid is normally connected to the main utility grid, drawing energy from the utility when economically advantageous, using a combination of utility power and onsite energy resources. 

Microgrids are also configured with the ability to disconnect and run in a self-contained mode when needed. This is appropriately termed ‘islanding,’ as the microgrid temporarily becomes its own energy island, operating separately from the main grid. 

This islanding mode is typical for many data centre operations, though the mechanism to determine islanding is purely driven by the incoming utility power quality, not by economics or resiliency factors.


First Steps In Onsite Power

All data centres around the world have some form of a backup power system supplying power to computing equipment and critical infrastructure. 

This is sometimes referred to as an Emergency Power Supply System (EPSS). Most commonly, this takes the form of one or more diesel generators, often supported by a UPS that provides power quality and backup power while generators are starting up.

For the purposes of this article, backup power systems, while essential to ensure continuity of service, are not considered a microgrid, as they are not intended to run continuously.


Moving Toward A True Microgrid

In other industries, Combined-Heat-and-Power (CHP) or Combined-Cooling-Heating-and-Power (CCHP) systems have become a popular distributed energy resource.

These systems are often configured as microgrids, as they include a local energy resource supplying – at least partially – the electricity needs of a facility, as well as delivering useful heat for absorption chillers. Colocation providers and data centres can achieve the same benefits from such systems.

To optimise costs, sustainability, and resilience, a more comprehensive microgrid solution can encompass a variety of DER, including CHP and CCHP, renewables, fuel cells, and energy storage. Choice of DER will depend on economic and environmental considerations.

At the operations level, the coordination of DER is managed by a microgrid control system. In the event of a utility grid outage, the control system is responsible for the safe disconnection from the grid and reliable transition to island mode. In island mode, the system manages all DER to maintain power stability.

With this level of digital connectivity and control, it is crucial that communication networks are secure against cyber threats. A microgrid solution should comply with end-to-end cybersecurity best practices, including alignment with standards such as IEC 62443-4-2 and IEC/ISA 62443-3-3 and the use of cyber-secure components from trusted vendors.



For colocation providers and data centres, microgrids provide value every day, and not just when the power goes out. Microgrids go beyond diesel-based power backup systems by enabling use of CCHP, renewables, fuel cells, and energy storage. 

They help increase resilience against grid disruptions, reduce energy-related operational costs, and ensure sustainability, with advanced energy analytic capabilities. 

Compliant with all applicable national and local regulations, a microgrid helps optimise and balance the use of grid versus onsite energy resources.

Ultimately, a microgrid increases a data centre operator’s confidence in uptime, ensuring tenants’ needs for computing continuity are met. Now is the perfect time for data centre infrastructure managers to adopt a microgrid solution. 

The technology is mature, making solutions more affordable and easier to implement than ever before. To ensure an optimised solution, seek a trusted expert that can offer the newest microgrid planning tools, modular architectures, along with EaaS options to reduce financial risks while maximising return on investment.








Welcome To The ‘Imagination Society’
How Digitalisation In Oil And Gas Is Creating Security Risks