2024 Data Center Energy Storage Industry Insights Report

September 4, 2024

Description

The data center industry is evolving rapidly with unprecedented speed and innovation, with battery storage solutions emerging as a key focus. To help industry professionals navigate these changes, ZincFive and Data Center Frontier have collaborated to produce this report, offering insights into the current landscape and future trends as predicted by their peers. Featuring contributions from 117 diverse industry professionals worldwide, this report examines the state of data center energy storage, covering usage, perceptions, priorities, challenges, future predictions, and the impact of AI.

Select data highlights include:

  • Only a third of respondents (34%) said they completely trust their existing backup system.
  • When evaluating energy storage solutions, industry professionals prioritize safety (69%) and total cost of ownership (64%). Nickel-zinc (NiZn) batteries, along with valve-regulated lead acid (VRLA) and lithium iron phosphate (LFP) batteries, ranked highly for both safety and cost considerations.
  • Energy storage technology limitations (50%) and sustainability targets/mandates (44%) were driving the changes respondents considered for their energy storage technology.
  • Four in five respondents (81%) said that sustainability criteria were important in their power system buying decisions.
  • Two-thirds (63%) said that their organization’s sustainability programs have resulted in some cost reductions, with one in five (19%) seeing significant cost reductions.
  • 26% of respondents expect their runtime to decrease, 38% foresee no change, while the remaining respondents believe it could increase.
  • When asked what they were not getting out of their current battery backup/energy storage technology, respondents listed the following four top priorities in order of mention frequency: long life, reliability, sustainability, and cost reduction.

Register to read full paper

Executive Summary

The data center industry is evolving rapidly with unprecedented speed and innovation, with battery storage solutions emerging as a key focus. To help industry professionals navigate these changes, ZincFive and Data Center Frontier have collaborated to produce this report, offering insights into the current landscape and future trends as predicted by their peers. Featuring contributions from 117 diverse industry professionals worldwide, this report examines the state of data center energy storage, covering usage, perceptions, priorities, challenges, future predictions, and the impact of AI.

The key data highlights include:

  • Only a third of respondents (34%) said they completely trust their existing backup system.
  • Energy storage technology limitations (50%) and sustainability targets/mandates (44%) were driving the changes respondents considered for their energy storage technology.
  • Two-thirds (63%) said that their organization’s sustainability programs have resulted in some cost reductions, with one in five (19%) seeing significant cost reductions.
  • 26% of respondents expect their runtime to decrease, 38% foresee no change, while the remaining respondents believe it could increase.

A reliable UPS system with high-quality batteries is crucial for continuous power and protecting data center operations. When evaluating energy storage solutions, industry professionals prioritize safety (69%) and total cost of ownership (64%), with nickel-zinc (NiZn) emerging as a notable battery chemistry.

The study highlights that safety is the top priority for data center backup power, with seven in 10 respondents prioritizing the safety of battery chemistry. Cost is also critical, with lifetime cost and initial CAPEX closely correlated. Nickel-zinc batteries, along with valve-regulated lead acid (VRLA) and lithium iron phosphate (LFP) batteries, ranked highly for both safety and cost considerations.

Sustainability is important to 81% of respondents, with many data centers (64%) assessing supply chain sustainability and tackling Scope 3 emissions.

The study reveals that newer battery chemistries are gaining traction, driven by the combined set of important benefits that data center operators value.

When asked what they were not getting out of their current battery backup/energy storage technology, respondents listed the following four top priorities in order of mention frequency: long life, reliability, sustainability, and cost reduction. 

When discussing AI’s impact on power requirements and energy storage technology, respondents highlighted its influence on several areas: dynamic load management, predictive maintenance, intelligent Battery Management Systems (BMS), advancements in battery technology and power density, sustainability, integration with renewable energy, AI-driven cooling solutions, and energy-efficient algorithms, among others.

Demographics

Respondent Demographics

A third of respondents (32%) identified as a Senior Manager, Vice President, Director or Department Head while a quarter (24%) said they were project, technology or team. Four in five (79%) said their organization was in North America, with 30% also operating in Europe.

Organizational Demographics

Nearly a quarter (23%) said that their organization’s primary role was a customer of data center services, while 22% were data center providers. Two in five respondent organizations (39%) had more than 1,000 employees across all locations.

Data Center Demographics

More than half of respondent organizations identified as an enterprise data center (55%) and colocation center (52%). Three in ten (30%) said that their data centers use less than 5 megawatts across all campuses, while more than a third (36%) are using more than 100 MW with 23% using over 500 MW.

Data Center Demographics

Respondents were primarily operating between zero to four data centers in each tier. Half of respondents operated between one to four Tier 1 (48%) and Tier 3 (51%) data centers. Three in five (59%) operated between one to four Tier 2 data centers.

Energy Storage: Usage and Outlook

UPS Size

Half of all respondent organizations (51%) were using less than 500 kW UPS size while one in ten (11%) were using more than 2 MW as their typical deployment UPS size.

Centralized UPS Energy Storage

More than a third of respondents (37%) were using valve-regulated lead acid (VLRA) cell batteries for centralized UPS energy storage, followed by lithium-ion manganese oxide (LMO) batteries (26%)

Energy Storage Solution Feature Priorities

Seven in ten respondents (69%) said that the safety of chemistry was a priority (top priority + high priority) when selecting an energy storage solution. Two in three (64%) respondents said the lifetime cost consideration/total cost of ownership was a priority.

Safety of Battery Types

VRLA cell batteries (64%), LFP batteries (63%), LMO batteries (63%) and Nickel-Zinc batteries (62%) were rated the highest on safety (excellent + very good + good) among battery types.

Footprint of Battery Types

LFP batteries (66%), LMO batteries (66%) and Nickel-Zinc batteries (54%) were rated the highest (excellent + very good + good) among battery types based on footprint.

Sustainability of Battery Types

LFP batteries (60%) were rated the highest (excellent + very good + good) based on sustainability followed by VRLA cell batteries (55%).

Cost of Battery Types

When it came to cost, VRLA cell batteries (59%) were rated the highest (excellent + very good + good) followed closely by Nickel-Zinc batteries (58%).

Energy Storage Technology Drivers

Energy storage technology limitations (50%), sustainability targets/mandates (44%) and the transition from centralized to distributed UPS or energy storage (41%) were driving the changes respondents considered for their energy storage technology.

Energy Storage Backup Run Times

Two in five respondents (38%) said that their backup run times were more than 10 minutes, while three quarters (73%) expected their run times to either go up (35%) or not change (38%).

Existing Backup System Evaluation

Only a third of respondents (34%) said they completely trust their existing backup system. Of those remaining two-thirds, more than half (57%) felt that the failure point of their existing backup system was human error, followed by the transfer switch (36%) and generator (24%)

Importance of Sustainability

Four in five respondents (81%) said that sustainability criteria were important (very important + somewhat important) in their power system buying decisions.

Sustainability Efforts

Nearly two-thirds of respondents (64%) said that they are assessing the sustainability performance of their supply chain. Two thirds (63%) said that their organization’s sustainability programs have resulted in some cost reductions, with one in five (19%) seeing significant cost reductions. Their sustainability efforts have also resulted in better product choices with 59% seeing significantly better product choices or somewhat better product choices.

Open Ended Questions and Answers

How do you feel artificial intelligence (AI) will impact power requirements and energy storage technology at your data center(s)?

Enhanced Energy Efficiency

Dynamic Load Management: AI can optimize the distribution of workloads across servers, reducing overall power consumption by ensuring that servers operate at optimal efficiency. Predictive Maintenance: AI can predict when components are likely to fail, allowing for timely maintenance that avoids energy wastage from malfunctioning equipment.

Improved Energy Storage Management

Intelligent Battery Management Systems: AI can enhance the performance and lifespan of battery systems by managing charge and discharge cycles more effectively and predicting battery health and replacement needs. Integration with Renewable Energy: AI can manage the integration of renewable energy sources with energy storage systems, optimizing the use of renewable energy and reducing reliance on grid power.

Reduced Power Consumption

AI-Driven Cooling Solutions: AI can optimize cooling systems in real-time, adjusting temperatures and airflow to minimize power consumption while maintaining optimal operating conditions for servers. Energy-Efficient Algorithms: AI can develop and implement more energy-efficient algorithms for data processing, reducing the overall power requirements of the data center.

AI is likely to have a mixed impact on power requirements and energy storage technology in data centers. While it may increase power demands due to the computational intensity of AI workloads, it also offers opportunities for optimizing energy efficiency and driving advancements in energy storage technologies. It will be important for data centers to balance these factors and leverage AI to achieve sustainable and efficient operations.

Artificial intelligence (AI) will significantly impact power requirements and energy storage technology at data centers by increasing power consumption due to the intensive computational needs of AI workloads and specialized hardware. However, AI can also drive improvements in energy efficiency through optimized operations, intelligent power management, and advancements in battery technology. Additionally, AI will enhance the integration of renewable energy sources and smart grid management, contributing to a more sustainable and efficient power infrastructure in data centers while reducing their carbon footprint.

AI will require more power. Energy storage will become more important to our operations.

AI can optimize energy usage within data centers by dynamically adjusting cooling, lighting, and computing resources based on real-time data analytics. This can lead to more efficient use of power, reducing overall energy consumption.

Ability to help identify predictive power consumption and availability, and dynamic balance of power to meet hi-usage times.

AI will be a game changer for the power requirement and energy storage.

Better assessments, recommendations, and choices.

Higher demand and more stress on supply chain

Higher density of power

Increasing AI usage will result in a corresponding increase in power demands and uptime requirements increasing power requirements significantly.

It has already made a large impact for us; nearly doubled UPS requirements.

Power requirements and energy storage will go up

Require 10x electricity and therefore power needs. Need to further ensure there isn’t a single point of failure in backup systems.

We think that AI will substantially impact operational efficiency

Will certainly make things easier for moderating power.

Will increase consumption

Will increase it 3x time

A lot. I would recommend a human element as a control.

Do not see any additional impact. We primarily use AI IoT nodes to connect to mobile data center

I believe it will be huge impact. It will have much more capabilities and less human error

I fear that AI will consume a majority of the power grid at an exponential rate

It will eventually and continue to make bigger impact and controls of data center usages, all areas

Some of it is good, and some is bad. It still needs much research and development.

That would be great if we could have AI in place to analyze the efficiency.

Enhance power management and restart times

More rack KW loading

Only increase power needs

Should increase power requirement.

THAT’s a very good question!

To soon to know.

unknown

Va a ser mas consumo de energia y recursos

Very much

Will demand higher power in the near future

AI will be an aggregated value

Are you kidding me? It’s a huge impact.

Could help

Do not know

I think these are scary times.

Increase

Increase

N/A

N/A

Added power and cooling will be required. As reliable higher capacity and density UPS units are available, they will be added over time.

AI is growing exponentially; it is hard to say how much but it will have impact

AI obviously increases power needs which in turn increases the load on UPSs. It also increases the cooling needs.

AI will be a useful tool in power analysis and trending that will increase efficiency in power utilization.

AI will demand a similar amount of power overall in a data hall but concentrated in a smaller amount of IT space. The load profile of AI processing creates issues back to the utility and very large issues with a generator. These loads can also put strain on batteries that may have to absorb the load changes. We are looking at alternatives and supplements like supercaps to mitigate load changes.

AI will impact DCIM (Data Center Infra management) in a very strategic manner in terms of PUE, Power Efficiency

AI will require more electricity load and energy storage technology to cater for the higher loads of GPUs

AI will require more networking and computing equipment which as always will require more power

Being able to predict quicker possible power loss or failure due to climate.

Clients want more power that we do not have.

Demand is growing exponentially

Densities moving toward 50+ kw/rack and time to power becoming a huge constraint; power costs are increasing and will have to figure out microgrid solutions

AI has not been implemented in our systems

AI is here to stay and AI in future is oil

I just feel like it will make datacenters more necessary.

Increase demand significantly.

Increase in demand for precision cooling

Increase power requirement

It will drive cost and all other factors up.

It will have a major impact on the power requirements

Load balance calcs and automated stops

Load will go up; backup time will stay the same or not be required.

Low impact

The high-power drives strong demand on copper as a natural resource and will have copper supply in a tight balance for years to come.

Training systems may not need backup power, since runs can be checkpointed. Significant savings by leaving out UPS and generators

I believe that it is too early to allow deep AI integration but believe that it is inevitable that there will be AI used for efficiency and test scenarios.

It is going to be very enormous, and I think it will see data centers resulting from sourcing energy from a nonrenewable energy because of the high-power consumption that AI needs.

Need for availability certainty (six nines plus) driving need for both new primary power, backup power and storage; also beginning to use availability and risk assessment to identify power failure weak spots

No answer at this time. We are still exploring.

no lo se aun.

No thoughts in that area unless major cost savings

Not sure

Not used

significant impact

significantly

Significant

What do you need out of your battery backup/energy storage technology that you are not getting today?

Out of 78 responses, the top priorities listed were, in order of frequency: long life, reliability, sustainability, cost reduction, increased backup capacity, smaller footprint, and improved monitoring, among other factors.

Scalability:

Ability to scale up energy storage capacity as AI workloads grow.

High Energy Density:

Higher energy density for sustained power during computationally intensive AI workloads.

Fast Charging and Discharging:

Rapid charging and discharging capabilities to meet dynamic power requirements.

Long Cycle Life:

Longer cycle life to withstand frequent charge and discharge cycles.

Enhanced Efficiency:

Improved energy efficiency for reduced energy costs and environmental impact.

Intelligent Management and Monitoring:

Integrated intelligent management and monitoring for optimized utilization and proactive maintenance.

Safety and Reliability:

Reliable and safe operation to ensure uninterrupted AI system performance.

Environmental Sustainability:

Sustainable solutions with low environmental impact and minimal reliance on rare or hazardous resources.

Better cycle life and quicker recharge after a discharge. More choice in fire-safe chemistries.

Existing facilities do not have available space for significantly greater storage capacity. Higher density energy storage may reduce the modifications necessary to the buildings to accommodate the higher power levels required for AI and HPC.

Out of our battery backup and energy storage technology, we need higher energy density to store more power in a smaller footprint, faster charging and discharging capabilities to handle peak loads efficiently, longer lifespan to reduce replacement frequency and costs, improved reliability to ensure uninterrupted power supply, and better integration with renewable energy sources to enhance sustainability and reduce our carbon footprint. Additionally, we seek more advanced management systems that can use AI to optimize energy use and maintenance, ensuring maximum efficiency and performance.

10,000% assurance it will work

Availability and cost reduction on mandated technology

Better Battery Management Metrics

Better capacity

Better lifetime cost/performance

Better monitoring

Better monitoring solutions for 3rd party integration

Compact sizes

Ensuring that energy storage solutions have a long lifespan with minimal degradation over time is crucial. Many current technologies may face issues like capacity fade or reduced performance after repeated charge-discharge cycles.

Getter storage capacity for a longer period of time

Less issue on UPS and more issue on primary power sourcing; for UPS needs to be able to scale to support higher densities with out encumbering footprint

Less physical space

Long life

longer durations

Longer life

Longevity

Low maintenance cost

Greater reliability

Mayor capacidad menor costo

High end “push notifications”

N/A

Lower cost, grid compliance

N/A

Maybe less FP’s

N/A

More backup capacity

N/A

More capacity, longer runtimes, higher voltages and amperage

N/A

More energy with more stability

New technology is on the way 😉

More interaction with utility services

No soy experto en energia.

Need to be upgraded with the latest and efficient

None

Need to regiment and monitor a little better

Not sure

More environment friendly. Longer life. 

Reduced costs. More sustainability.

Sustainability

Backup power is okay – more concerned of long-term blackout

Scalability and limited OEM support in the region

Uptime

Cost modulation

Efficiency, longevity and cost

More storage

Space saving

Automatic load change

Data Lakehouse and electric vehicle, charging power

Remote management control unwavering performance

Better communication and issue management

The ability to at a bare minimum detect issues before they become larger problems

Reliability

Sustainable roadmap for future deployments

Somehow, the charging time is a bit longer than expected

Nothing

Technology upgrades

Nothing satisfied with current setup

All good

Nothing. Our UPSs always perform when needed. We have 100% power availability in all our Datacenters.

Currently nothing

Currently we are not in need of the OOB very urgently, it can be a supplementary option in the next 5 years.

I’m good right now.

I’m not sure

In the short term we have no requirements

It will be very high in the Priority list

Just more of it

Our cost are low, that’s our priority.

Smarter way

We are getting what we need but like to so innovation and growth is sustainability

We are good with Wet cells.

We are happy with our highbred approach

We have continuous power system. Need improvement to coupling design for higher reliability and extended life.

Other (please specify) responses

Which of the following best describes your job level? Other responses

Account Director

Construction/CX Engineer

Power, Space, Cooling Engineer

Assistant Sales Manager

Order Management Coordinator

Sales Operations Manager

Business Development

For CENTRALIZED UPS, what type of energy storage does your organization use? Other responses

Don’t know

N/A

unaware

Gravity storage

not sure

unsure

I’m not sure

Proprietary system

We supply copper bus bar only

Please rate each battery type based on the following criteria. Other responses

N/A

unaware

Exs

not na

Unaware

Which of the following is driving the changes you are considering to your energy storage technology? Other responses

Cost; ethical material use (avoid cobalt)

N/A

How do you see your UPS or energy storage backup run times changing in the future? Other responses

N/A

Where do you believe the failure point is? Other responses

N/A

Methodology

Endeavor Business Intelligence and Data Center Frontier conducted data collection and analysis on behalf of ZincFive Inc. The data was collected from June 26 to July 19, 2024, resulting in 117 qualified survey responses. The methodology adheres to standard marketing research methods, practices, and procedures.