UPS energy storage as the first layer of power resilience

When a power interruption or dynamic workload event occurs, UPS energy storage systems provide the immediate energy required to maintain operations while backup generators start, grid supply is restored, or the spike in synchronized power draw is controlled. UPS energy storage systems therefore act as a critical bridge between primary power loss and secondary power availability.

Data center outages are less frequent, but the impact remains significant

The latest Uptime Institute Annual Outage Analysis shows encouraging progress: for the fifth consecutive year, both the frequency and severity of data center outages have continued to decline.¹ This trend reflects industry-wide improvements in infrastructure reliability, risk management, and the adoption of advanced resiliency strategies. Infrastructure resilience has improved across the industry, even as demand for digital services grows at an unprecedented pace.

Read more: Data Center Outages Are Declining, But Here’s Why You Still Need to Prepare

However, while outages are less frequent, their severe financial impact remains a threat. According to the Annual Outage Analysis report, 1 in 5 respondents said their most serious recent outage cost over $1 million for the second year in a row. This escalation underscores a critical reality: even a single failure can have significant operational and financial consequences for business and reputation.

Read more: WHY YOUR DATA CENTER’S BIGGEST THREAT MIGHT BE THE POWER GRID

Dynamic workloads introduce new layer of complexity for UPS energy storage

Reserve power in case of an outage remains a key function of UPS energy storage systems. However, data centers with dynamic workload profiles can also require the same systems to regularly absorb sudden synchronized spikes in power demands.

For example, AI training — the process of building and refining AI models — can involve synchronized power draw across entire GPU clusters, as opposed to more dispersed user-driven demand.

Testing data by the North American Electric Reliability Corp (NERC) from a 50MW block within a 200MW AI training data center shows the fastest ramping period to be around 1.9 per unit per second for around 250 milliseconds.²

Dynamic workloads reinforce the importance of both high-rate discharge to be able to absorb spikes in demand, and high-rate recharge to be able to deal with consecutive events in quick succession.

Quick summary for data center operators

Modern UPS energy storage performance is typically evaluated across several operational criteria rather than runtime alone.

A performance-first UPS energy storage strategy focuses on:
• High-rate discharge capability — the ability to support short-duration, high-power UPS events.
• Recharge recovery time — how quickly the system returns to full readiness after an outage or power-spiking event.
• Thermal operating envelope — the temperature range in which the system can operate reliably.
• Float-mode energy consumption — the ongoing energy draw required to maintain system readiness.
• Lifecycle planning and recycling pathways — how the system is managed throughout their operational lifespan and at end of life.

Evaluating these factors provides a structured way to determine whether a UPS energy storage system aligns with the operational conditions of modern data centers.

FACTORS INFLUENCING DATA CENTER POWER RESILIENCE

Supporting performance-first UPS energy storage strategy

Legacy lead-alloy batteries are characterized by:

Faster degradation especially in high-density applications, such as AI training, HPC, etc.; risk of failure and subsequent downtime which may result in shorter service life in certain operating conditions.

Slower recharge times meaning you take longer to prepare for the next outage or may be unable to respond to a power-spiking event.

The Thin Plate Pure Lead (TPPL) advantage

Advanced TPPL from EnerSys means:

Up to 25% longer lifespan in comparison with certain conventional lead-acid designs—helping to reduce the risk of failure and unforeseen downtime.

Faster recharge times —0%-80% state of charge in 50 minutes, so you’re ready for subsequent outage events.

Track-record of powering critical infrastructure:

• Major hyperscale and colocation data centers.
• System operators of electricity supply networks.
• Mission-critical military applications.

References
1. Uptime Institute. (2025). Annual Outage Analysis 2025. Available at: https://uptimeinstitute.com/resources/research-and-reports/annual-outage-analysis-2025
2. North American Electric Reliability Corporation (NERC), White Paper: Characteristics and Risks of Emerging Large Loads, July 2025, https://www.nerc.com/globalassets/who-we-are/standing-committees/rstc/3_doc_white-paper-characteristics-and-risks-of-emerging-large-loads.pdf.
3. North American Electric Reliability Corporation (NERC), White Paper: Characteristics and Risks of Emerging Large Loads, July 2025, https://www.nerc.com/globalassets/who-we-are/standing-committees/rstc/3_doc_white-paper-characteristics-and-risks-of-emerging-large-loads.pdf.
4. International Energy Agency (IEA). (2025). Executive Summary – Energy and AI. Available at: https://www.iea.org/reports/energy-and-ai/executive-summary
5. International Energy Agency (IEA). (2025). Renewables 2025: Renewable Electricity. Available at: https://www.iea.org/reports/renewables-2025/renewable-electricity
6. Climate Central. (2024). Weather-Related Power Outages Rising. Available at: https://www.climatecentral.org/climate-matters/weather-related-power-outages-rising
7. Plautz, J. (2025). Rulemakers play catch-up as data centers multiply. Available at: https://www.eenews.net/articles/rulemakers-play-catch-up-as-data-centers-multiply/
8. CRU, Large Energy User Connection Policy – Decision Paper, https://cruie-live-96ca64acab2247eca8a850a7e54b-5b34f62.divio-media.com/documents/CRU2025236_Large_Energy_User_connection_policy_decision_paper.pdf.
9. Sky News. (2025). ‘Catastrophic failure’ that led to Heathrow power outage revealed. Available at: https://news.sky.com/story/catastrophic-failure-that-led-to-heathrow-power-outage-revealed-as-report-outlines-missed-chances-to-prevent-it-13391266
10. Insurance Times, Loss estimate for AWS outage between $38m and $581m, https://www.insurancetimes.co.uk/news/loss-estimate-for-aws-outage-between-38m-and-581m/1456730.article.