How To Select Batteries for Datacentre Energy Storage Applications

The UK like many other countries is experiencing an energy revolution. Energy storage is seen as one technology that can help countries to meet their energy demand needs, low carbon and climate change obligations. Energy storage technologies have been developed to cover a wide range of applications from grid scale to commercial buildings and domestic properties. Often referred to as local energy storage (LES) systems, the purpose is very similar to that of an uninterruptible power supply but in a primary ‘active’ rather than a secondary ‘standby’ role.

Energy Storage Applications

Storing energy for later use is a growing industry with expert reports proposing 5-year compound growth figures of over 22% for the domestic market alone. More info: There are also several technologies available ranging from pumped-hydro to flow batteries and compressed air storage in underground caverns at the grid level – more info: to smaller building-scale applications using some form of battery array.

Datacentre Power Backup Systems

Within datacentre and server room environments there is a growing interest in how best to apply energy storage. There is still the need for protection from local power outages, but the emphasis is more on how to:

  • store locally generated power from renewable sources
  • use off-peak electricity price periods (low £/kWh) to charge a battery set
  • use the power as source of revenue generation by selling excess energy back to the grid

The most common battery type for energy storage applications is a Lithium-ion (Li-ion) or Lithium Polymer battery. These are more suited to this type of primary application than lead-acid batteries due to the more rapid and frequent charge/discharge cycles required from the battery set. Lead acid batteries are more suited to standby cycles and in for renewable applications lead acid OPzS batteries are the preferred choice if a lead acid battery is specified.

Virtual Power Plants (VPPs)

Energy storage systems can act as a virtual power plant where the energy storage during off-peak periods or from renewable power sources, can be used as the primary source when mains power is available but at peak-rates. The difference in electricity kWh costs can be used to recover the spend on the capital investment. This type of arrangement could benefit domestic as well as some commercial properties.

For datacentre operators a virtual power plant arrangement provides the option for the facility to become more grid independent. The stored energy could be used to support a UPS system and reduce the need to run on local standby power generators which are typically diesel powered. In addition the stored energy could allow the site to take part in a National Grid demand side response type program or feed-in-tariff arrangement.

Demand side response programs are aimed at balancing demand on the grid and maintaining the frequency within set tolerances to the National Grid to avoid Ofgem penalties. As part of a DSR program, a datacentre could agree to disconnect itself from the grid for a set period thereby removing potentially demand for up to several hundred kilowatts or megawatts of power from the grid. Stored energy could also be exported back to the grid in much the same way as wind turbine and solar PV plants operate.

Sizing Energy Storage Batteries

The way to size and select batteries for use in energy storage systems is much the same as that for an uninterruptible power supply. There are common steps but with some that are more specific to the rapid charge and discharge cycles for energy storage systems:

  • Load: this is the total amount of power to be drawn from the system and should be calculated by listing all the critical loads and their power demands in watts or kilowatts (real power) or VA or kVA (apparent power). The real power and apparent power loads are linked by the load power factor. Most loads today draw a near Unity (equal to 1) power factor but some older types including old computer power supplies can be as low as 0.5.
  • AC Output Voltage: knowing the load requirements is important in terms of their supply requirements which can be single phase (230Vac, 50Hz) or three-phase (400Vac, 50Hz). UK and European voltages are shown but these will be different in other countries.
  • Inverter DC Voltage: the inverter is the section of the UPS or energy storage system that supplies AC (alternating power) to the connected loads. The DC voltage will typically be sized in multiples of 12 i.e. 12, 24, 48, 96Vdc or greater. The overall number defines the quantity of battery strings and blocks required.
  • Backup Runtime: the period the battery set is to be sized to run form. This should take into account the amount of time for a power outage, the time required to start a backup generator (with a safety margin) and the time the system us to run as a virtual power plant.
  • Preferred Battery Type: in a standby power role lead acid UPS batteries are the preferred choice. In an energy storage system then application requires lithium-ion batteries. As well as being more suited to rapid charge/discharge cycles, Li-ion batteries are also more temperature tolerant. The choice of battery type also places more demand on the battery charging and monitoring system. For lithium batteries this is a more complex and costly device in order to balance out charging to the cells and prevent potential safety issues such as thermal runaway.
  • Depth of Discharge DoD: this is the lowest level of discharge the battery should reach without damage to its capacity. The depth of discharge can run from 50-10% dependent. The lower the DoD the longer greater the runtime and expected working life of the battery. A battery having a DoD value of 50% may be capable of 1000 cycles before its capacity reduces. At 20% DoD, the battery may last 3000 cycles or more.

A UPS system is designed for use with a specific battery type, whether this is lead acid or lithium. There is no dual-source UPS system currently available. This means a decision as to which battery technology is right for the application must be taken at the project design stage.

Most UPS suppliers and manufacturers now include lithium-ion powered uninterruptible power supply alternatives. Small single-phase UPS systems using lithium-ion batteries are more suited to remote Edge computing and higher temperature server rooms applications.  The UPS are more compact due to a higher power density than lead acid batteries and higher temperature tolerant. For larger three-phase monoblock or modular UPS applications including use as virtual power plants, the lithium-ion batteries are supplied as a separate battery cabinet system including the battery monitoring and management system.


UPS systems have always stored energy for later usage. The difference between this type of arrangement and that of a local energy storage system is that the charge/discharge cycles are more frequent and the wider applications available including taking part in grid balancing roles. Within the next five years lithium-ion will become the dominant battery choice over lead acid for datacentre and server room backup power applications. This will then give operators the choice as to how they configure their UPS. Whether as a standby power source for emergency protection from power outages or in a more proactive energy storage role.

How To Select Batteries for Datacentre Energy Storage Applications was first seen on Server Room Environments