Energy costs associated with powering the increase in artificial intelligence and cloud computing are top of mind as the industry looks to power data centers more efficiently and effectively. Operational expenditure (OPEX) costs for data centers are considerable, with the massive amounts of electricity needed to power and cool the facility contributing greatly to its energy costs. Solutions to hedge against the rising costs of energy are needed.

According to the U.S. Department of Energy (DOE), some of the largest data centers require upwards of 100MW of power, equivalent to the energy needed to power 80,000 homes. That number is only expected to increase, as the total power consumption of data centers is anticipated to more than double between 2022 and 2026 globally.

As data center executives consider how to power their upcoming projects, a multi-phased approach in energy planning has emerged as a strategy that can significantly impact OPEX costs while securing the required energy to run a facility. This multi-phased approach looks at energy resources in two primary stages, while utilizing microgrids to provide reliability, energy security, and improved power quality. Microgrids can optimize cost-efficiency and grid independence through advanced energy management systems. By looking at microgrids in a phased, multi-year approach, data centers can obtain reliable power, cost-savings, and emissions reductions:

  • Phase 1. Microgrids with renewables, CHP, and energy storage
  • Phase 2: Small Modular Reactors incorporated into the microgrid

A Multi-Phased Approach Accounts for Future Generation Technologies

The first stage looks to incorporate microgrids powered by currently available distributed energy resources (DERs) to immediately reduce dependency on centralized grids and to integrate generation from multiple fuels and storage. This approach looks at the immediate available resources to meet energy demands efficiently, harnessing renewable energy sources including solar and wind, battery storage, and combined heat and power (CHP).

The second stage brings in newer future technologies such as small modular reactors (SMRs). While not yet commercially available, planning for their future integration ensures they can seamlessly be incorporated to bring scalable, low-carbon baseload power. The multi-year approach to energy planning provides the route from currently available DERs to newer technologies like SMRs in order to future-proof the data center for steady growth.

Commercial scalability for SMRs has become a top investment priority as new energy resources are needed. Consider how the Department of Energy (DOE) issued a $900 million solicitation to support the deployment of SMRs. In Canada, the federal nuclear safety regulator has approved the first commercial small modular reactor. The promise of reliable, scalable power from SMRs for energy-intensive sectors has catapulted their advancements significantly.

Looking Beyond the Utility-Only Limitations 

While data centers may choose to source their energy needs with a utility-only approach, forgoing any investment into DERs and relying on existing utility infrastructure will present  challenges. The current options and energy planning scenarios available for data center operations include the following:

  • The Utility-Only Case: With no investments made into DERs, energy demands are met solely through existing utility infrastructure for the next two decades. This approach assumes that utilities can provide the power, which can be a challenge with utility upgrade lead times of 7 to 10 years.
  • Utility with SMR-Only Case: Energy demands are met through existing utility infrastructure until 2035 when SMRs are expected to be deployed at commercial scale. No investments are made into DERs before the integration of SMRs in this case. This strategy exposes the system to increasing utility costs and potential carbon penalties as electricity prices rise.
  • Multi-Phased Approach: DERs are immediately deployed to meet current energy demands, followed by the introduction of SMRs in 2035 to amplify the DERs in use. This approach minimizes operational costs by reducing exposure to rising utility rates, while aligning with any stated decarbonization goals.

Multi-Phased Approach is Region-Agnostic at Reducing Costs

In Xendee’s recent whitepaper exploring how data centers’ OPEX would respond to a multi-year energy planning approach compared to that of the utility-only and utility with SMR-only options, modeling was done into two distinct regions: Santa Clara, California and Ashburn, Virginia.

Sata Clara is a popular location for data center development but is plagued by high electricity costs, while Ashburn, or Data Center Ally as it has now become coined due to the sheer volume of data center development happening, has much lower electricity costs but growing energy infrastructure concerns due to the demand.

Despite these differences, the study found that a data center located in Santa Clara utilizing the mutli-year energy approach would experience 79.66% OPEX savings, and an 8.69% reduction in emissions. For a data center in Ashburn, the OPEX savings are also achieved at a rate of 59.51% and emissions reduction of 23.86% with the multi-year approach. Both savings are higher with the multi-year approach compared to that of the utility-only or utility and SMR-only approach.

Early investments to adopt the multi-year approach in Santa Clara were higher, influenced by the dependence of DERs to offset the region’s high electricity costs. Asburn experienced more flexibility with early adoption of DERs thanks to the lower electricity costs of the region. Yet, both gain the long-term OPEX savings.

Strategizing for Future, Sustainable Energy Growth 

Data center executives and developers require confidence in the accuracy and adaptability of their strategies to ensure the near and long term success of their projects. Energy complexities exist in any-region, but one fact remains true: Relying solely on the utility for a data center’s energy demands is not cost effective or reliable.

# # #

About Dr. Michael Stadler

Dr. Michael Stadler is the Chief Technology Officer and co-founder of San Diego based Xendee, a software company supporting the design and operation of complex distributed energy systems.

Before that, Michael Stadler was a Staff Scientist at Lawrence Berkeley National Laboratory and led their Grid Integration Group. He is a recipient of the White House’s Presidential Early Career Award for Scientists and Engineers (PECASE), which is the highest honor bestowed by the U.S. government on science and engineering professionals in the early stages of their independent research careers. Over the course of his career, Michael has published more than 270 papers, journals, and reports to date and holds 17 copyrights and patents.