Click to learn more about author Carsten Baumann.
Once upon a time, data center reliability was all that mattered. The conversation has now evolved to include sustainable resiliency. Grid power disturbances are becoming more frequent due to higher demand, aging infrastructures, intermittent renewable energy sources, and more frequent storms or other disasters.
Combining this with an accelerated trend of digital transformation, such as remote work and teaching environments, IoT, and big data applications, we are experiencing the need for more reliable and green data center capacity.
Buying electricity is projected to become more expensive. The continued growth in demand and capacity is driving the need to find a more sustainable way forward. Microgrids can offer sustainable energy onsite generation that leads to cost savings and improved resiliency.
The Trend Toward Data Center Sustainability
Sustainability is a concept already embraced by many sectors and is now being progressively adopted by the data center and colocation industry. Data Center Frontier recently reported that cloud builders are deepening their commitment to green energy. Gil Santaliz, the President and CEO of the NJFX data center in Wall, N.J., is quoted as saying, “Renewable energy is becoming a prerequisite. If you can’t get hold of renewable energy, you’ll be at a disadvantage.” iMasons Founder and Chairman Dean Nelson envisions a future in which “digital infrastructure will contribute to the global economy and society without harming the planet.”
Many digital infrastructure leaders share that vision and are acting on it today. For example, Microsoft boldly announced that it would be carbon negative by 2030, a message reinforced by a recent successful proof of concept using hydrogen fuel cells to power a row of data center servers for 48 consecutive hours. Another example is the partnership between Switch and Capital Dynamics that is driving the Gigawatt 1 solar and battery project in Nevada.
Whether pursuing a competitive advantage or simply meeting local environmental regulations, reducing one’s carbon footprint has often been achieved using the offsets of renewable energy credits or “virtual” power purchase agreements that enable access to remote green energy sources, such as solar or wind farms. These are positive steps, yet they do not help support greater onsite resilience.
The Emergence of the Data Center Microgrid
The newest microgrid designs are helping data centers address all three goals by coordinating a variety of onsite, distributed energy resources (DER), including renewable generation. A microgrid will help optimize costs while also enhancing power stability and includes 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 the consumption of onsite renewable or stored energy. Stored energy can also be sold back to the grid when most economical. And the consumption of renewable energy can be maximized to meet greenhouse gas emissions targets.
While backup power systems — typically comprising diesel-generators and a UPS — are ubiquitous and essential to ensure continuity of service, they are not intended to run continuously. To move toward a true microgrid requires additional energy resources with intelligent control.
There is now a wide range of choices for onsite energy resources. Some examples include:
- Renewables: Depending on availability, costs, policies, incentives, and local market factors such as electricity pricing and regulations, a number of onsite renewable options are worth considering. This can include solar, wind, and biomass.
- Fuel Cells: Due to their small footprint, fuel cells are becoming a popular choice at data centers. Depending on financing, incentives, and fuel costs, these systems can also deliver significant energy savings. And if the hydrogen used can be produced from electrolysis powered by a renewable energy source, such as solar, the resulting fuel can be considered a renewable resource.
- Energy Storage: Onsite energy storage not only supports resilience in the event of a grid power outage (in coordination with other DER), it also helps maximize the value of renewable energy generation and can be used to support peak demand management.
- CCHP: Combined-cooling-heating-and-power (CCHP) systems that run on natural gas offer great efficiency by providing both energy and cooling. They typically run continuously and are often better maintained than standby generators, meaning they are ready and able to provide power in the event of a blackout.
Microgrid control occurs through a combination of onsite microgrid controller and software, as well as cloud-based analytics and services. Intelligent control manages grid connection and islanding, all aspects of safety, and how DERs are used while connected or disconnected from the grid.
Advanced energy analytics are used to achieve cost savings through a dynamic model of “utility interdependency” that enables the flexibility and functional value of DER to be fully monetized. KPIs ranging from forecasting energy production to energy market and fuel pricing are used by AI modeling to help determine the best times to generate, use, store, or sell energy.
The greater the variety of DER, the more optimization opportunities can be taken advantage of, and the greater the resiliency against a single mode of disruption. The choice of DERs depends on geographical opportunities and constraints, access to fuel resources, and the overall project financial attractiveness.
A New Era in Microgrid Design and Affordability
Design, construction, and operation are enhanced by a new breed of microgrid solutions based on standardized, prepackaged system components and modular pre-validated architectures. These enable configured-to-order microgrid systems that minimize delivery time and costs, increase reliability, and maximize ROI due to simplified design, installation, support, and maintenance.
Implementing a data center microgrid will include a variety of financial considerations. Fortunately, the growth of microgrids in recent years has driven down costs by an estimated 25 percent to 30 percent since 2014, and this is expected to continue on that trajectory. There are also many options for financing and operating microgrid infrastructures and a variety of government incentives available. As microgrids tend to be capital intensive, Energy-as-a-Service (EaaS) models offer a flexible ownership structure, in which the data center operator pays a monthly fee for operational expenses to the third-party owner.