By: Tim Hysell, Co-Founder and CEO, ZincFive
Flexible, resilient, and green might be words you associate with your garden – but in 2023, they’re also exactly what clients want from their data center operations.
Each year brings exponentially higher demand for speed and power to run the metaverse, edge computing, IoT, 5G, AI, and hybrid cloud deployments. And it doesn’t take a crystal ball to see that as the digitization of global entertainment, industry, and medical centers accelerates, so will the need for data center infrastructure to support this heavy load.
The following five concepts, each proving to be more than just a trend, will define how data centers meet this increasing demand in the new year.
1. Sustainability is a necessity, not a luxury
Like the rest of the world, data centers are now facing a climate crisis as temperatures and weather events soar. This last summer saw extreme heat waves causing stress for European and Californian data centers; Twitter itself lost a key data center to extreme heat. Fortunately, many data center operators are preparing their centers to withstand such extremes.
Meanwhile, data centers are seeking ways to increase their power load and serve higher client demand without significantly increasing their electricity and emissions burdens. Solutions that help decouple power load and emissions will help data operators prepare for increased emissions regulations and clients’ growing awareness of the need for sustainable data center operations.
Solutions such as nickel-zinc (NiZn) batteries can reduce scope 3 emissions and attract clients by helping them adhere to reporting requirements. NiZn batteries’ life cycle GHG emissions are four times lower than lead-acid batteries’, and six times lower than lithium-ion batteries’. Their carbon payback time – the time required for emissions savings from the product’s use to offset the GHG from its production – is a mere 25 percent of lithium-ion and lead-acid batteries’ payback time. Other ways to increase a data center’s sustainability include incorporating clean energy into its power supply, reusing waste heat, and cooling options that use less water; the green Wyoming Hyperscale data center uses several of these options.
The green data center market is projected to grow by 14% annually through 2032, partly due to these centers’ lower costs over time. For both financial and environmental reasons, wise operators are taking heed and incorporating more sustainable measures into their centers’ operations.
2. The quality of backup power strategies matters
On-site power failure is the most common cause of significant data center outages, with the majority caused by uninterruptible power supply (UPS) failure. High expenses associated with outages make maintaining immediate power with dependable UPS systems even more critical. UPS systems must contain backup batteries that perform reliably under any circumstances. For example, NiZn batteries remain conductive even if weakened or depleted. So, unlike lithium and lead-acid chemistries, these battery strings continue to operate even with depleted cells and make what would otherwise be an expensive emergency into a simple replacement at the next planned maintenance action.
For prolonged power outages, diesel generators are typically used for temporary power production. These generators require batteries to start them, and the batteries must also be reliable to avoid instances where the backup generators fail to start when needed. For this purpose, starter batteries with high current discharge capability and reliable string operation like NiZn enable improved generator availability.
Hyperscale data centers have another option for backup power as well: rack-level battery backup can overcome some of the challenges of centralized backup, such as excess heat and conversion losses. Rack-level backup both saves significant space and power in hyperscale data centers and improves reliability. While a centralized UPS approach often impacts significant segments of downstream equipment, rack-level backup limits any effects from one faulty BBU to a single rack.
3. Costs must be reduced without sacrificing performance
Data center operators are working under economic constraints and face continual pressure to deliver more with less. Reducing costs in 2023 while simultaneously improving performance will require adopting key strategies that take advantage of new technologies and practices, with the acknowledgment that it will require a collection of tactics instead of a single silver bullet.
Space is at a premium in data centers, and the less space that equipment takes up, the more computing power and associated profits can be accommodated. Data center operators thus prioritize power capability, where a smaller machine or device can deliver the same or greater amount of service to the facility as its larger counterparts. For example, a NiZn battery system nearly halves the space needed for batteries in data centers compared to traditional lead-acid and over 30% less than newer lithium-ion solutions. These smaller footprints stem from NiZn batteries’ high power capability: they can operate at higher powers to deliver more megawatts in fewer cabinets, and so can be sized to support shorter end-of-life runtimes while removing stranded battery capacity.
Outages are costly, and although those related to power problems are thankfully becoming shorter, even outages of a few minutes are becoming more expensive. Between 2019 and 2022, the portion of all outages which cost under $100K dropped from 60% to 29%, while outages costing over $1 million jumped from 11% to 25%. Power disruptions are the biggest cause of these outages, and their prevention will continue to be a priority for operators looking to minimize costs. Fortunately, higher-reliability batteries and UPS solutions are now available to address this issue.
For data center operators looking to reduce costs, a series of strategies have emerged that should be adopted in 2023, if not already in place. Reducing the physical footprint of equipment and addressing causes of power-related outages can manage costs without hindering the ability to meet customer demands.
4. Higher operating temperatures reduce the need for cooling
Rising temperatures aren’t only happening outside data centers: 2022 saw data centers increasingly using higher operating temperatures to reduce the costs associated with cooling. Data centers dedicate large amounts of their operating budgets towards facility cooling: on average, 40% of a data center’s energy consumption goes into powering its cooling and ventilation systems. Operating at higher temperatures can help lower energy load and help fulfill sustainability goals while lowering costs.
Of course, increasing temperatures require a balancing act to ensure they don’t result in shortened equipment lifespan, safety issues, and voided product warranties. For instance, lead-acid batteries’ sensitivity to heat has often traditionally stymied efforts at raising temperatures. Fortunately, this is changing: nickel-zinc backup batteries are approved for operation at up to 10 degrees Celsius above lead-acid batteries’ recommended limit.
Temperature flexibility is an emerging strategy that has become possible thanks to the availability of alternative technologies. As these strategies are increasingly adopted in 2023, data centers can reduce cooling system investments, improve sustainability, and lower costs while maintaining operational reliability and safety.
5. Modular power solutions are becoming widespread
Each year sees a more urbanized world, with more data center operators serving densely packed populations where space is at a premium. These operators find the smaller sizes and flexible designs of modular data centers increasingly appealing. In contrast to traditional data center designs, these centers are composed of prefabricated units and preconfigured equipment. They require less space (and so lower real estate costs), are easier, less expensive, and faster to build, and can be deployed almost anywhere.
Modular data centers also allow operators to scale and replace systems as needed, instead of having to pay for expensive reconfiguration projects. This only-add-as-needed feature makes modular data centers both more cost-efficient and more environmentally friendly, which helps data centers and clients meet their above-mentioned sustainability goals. Batteries with smaller footprints, higher operating temperatures and safety in operation can significantly reduce the cost of modular power solutions for data centers.
With all these benefits, it’s no wonder that the modular data center market is projected to grow from $23 billion in 2022 to $88.5 billion by 2030. In fact, a recent survey of 228 data center executives found that over half have already deployed prefabricated modular data centers (PMDCs), and 99% plan to use them in the near future – a wise decision.
What’s Expected of Data Centers in 2023 – and Beyond
The data center of the future is powerful, sustainable, affordable, and flexible. Many characteristics that were previously seen as optional are becoming increasingly necessary, and those failing to catch up to new opportunities risk becoming laggards in a rapidly evolving industry. Data centers are facing fierce competition for clients, and must pursue every advantage they can in 2023. To do so, there needs to be an awareness of technologies, best practices and challenges, and a willingness to adopt the necessary changes and their associated benefits.
About the Author:
Tim Hysell, Co-Founder and CEO of ZincFive
Tim Hysell is the co-founder and CEO of ZincFive®, the world leader in innovation and delivery of nickel-zinc batteries and power solutions. Tim has over three decades of entrepreneurial success in founding, owning, and directing profitable business operations in renewable energy, banking, manufacturing, and medical devices. Over the past decade, Tim has expertly guided the creation and deployment of ZincFive’s nickel-zinc rechargeable battery technology.