A research-industry perspective of battery systems technology for next-generation data centers

In this modern digital infrastructure era, data centers are the place where huge amounts of data are stored, processed, and managed. They support key services, such as Artificial Intelligence (AI), and online transactions, making it possible for most of the sectors, industries, and businesses. With that much data growing rapidly, secure, scalable, and reliable data centers are crucial for the global economy and technological innovation. Consequently, one of the key factors in an efficient data center is its energy concerns, which involve maintaining continuous energy support with high sustainability and reduced CO₂ emissions. In this regard, batteries are devices that are considered important in modern data centers. Their importance lies in stabilizing the renewable energy by storing surplus power and keeping the data center's power stable. At the core of batteries is the Battery Management System (BMS) that helps to enhance performance, safety, and life span by tracking different battery states and analyzing them. Algorithms in advanced BMS technology can increase energy efficiency using accurate charge/discharge control, reduce failures, and extend the battery's Remaining Useful Life (RUL) of the data center facilities. Consequently, the proposed paper overviews the battery technologies and BMS for data centers and highlights their power architecture and energy management topology. It starts with power modeling of data centers using Uninterruptible Power Supply (UPS), Power Supply Unit (PSU), and different power topologies. Then, it presents the energy efficiency metrics that are standard for planning data center energy performance. The important types of batteries are discussed, explaining why they are a common part of today's data center operations. The BMS architecture and its different control layers are further presented in relation to state estimation, such as States of Charge (SoC), Health (SoH), and Energy (SoE). In this regard, the main analysis of the paper is on the integration of AI into BMS and how it affects performance metrics to have energy-efficient next-generation data centers with advanced BMS. In the end, net-zero standards, industrial standards, a literature review of industrial white paper, recently funded projects and challenges are presented. Then, future research directions, and industrial perspectives, are provided.