UPS Systems Sizing and Efficient Uninterruptible Power Supplies
Sizing uninterruptible power supplies and UPS systems is more important than ever today as an oversized UPS system will lead to increased running costs and and under-sizing can leave little enough load capacity for future expansion. Correctly sizing a UPS system is also important for efficiency. Normally the lower the load on a UPS, the less efficient it will run and the higher the running costs.
Sizing UPS systems is not a complex topic but it is one of increasing importance given the higher power Blade servers now found within datacenters and the rapid growth that IT systems can experience within a short time frame.
UPS Systems Sizing
The optimum loading for any electronic system is 80% of overall capacity. Loading a UPS system to 90 or 100% capacity is more than acceptable as well but can introduce two weaknesses. The first is that it leaves little room for expansion but more importantly, an on-line UPS may be forced into bypass by small surges in load demand.
This is a normal UPS safety feature (and one that protects the load) but it can expose the load momentarily, until the demand decreases, to the vagaries of the mains power supply itself. Most on-line UPS are capable of working for predetermined periods on overload and these are usually specified in terms of 110%, 125%, 150% and 200% of capacity. However, the higher the overload, the less time the UPS will continue to power the load from its inverter supply. Line interactive or standby UPS do not provide the same level of inherent safety as an on-line UPS and without a built-in automatic bypass facility, will generally shutdown after a short period on overload.
UPS Sizing Considerations
When sizing a UPS system it is important to start with a classification of the potential loads to be protected:
- Critical Loads: are the IT, telecoms and and electrical systems that are critical to business continuity. They can include Blade file servers, telecoms systems, PCs, storage devices, security and building management systems. These systems may be so vital that it is important to power them through even long duration mains power failures, which will affect the size of the UPS battery chosen – or even overall power solution in terms of combining a UPS and generator solution.
- Essential Loads: are vital to the organisation but not critical. These could include some lighting, heating and ventilation systems. Some of these systems may well have their own emergency standby power facilities built-into them.
- Non-essential Loads: are the systems that are non-critical. A measure of their appropriateness for UPS protection is to consider their impact on day-to-day operations when removed (for maintenance).
From the list of critical and essential loads calculations can then made of the overall load size.
Real and Apparent Power, and Load Power Factor
Most loads are actually sized in Watts (or Kilowatts – kW). This is known as the ‘real power’ of a load. Rear panel rating plates on equipment typically lists their maximum Watt rating and these can be summed to provide an overall power calculation.
The next approach in UPS sizing is to look at the Amps being drawn by a system and the mains power supply voltage (Vac) it is being powered from. Combining the two using the formula VxA=VA provides the ‘apparent power’ measure or VA (kVA = 1000VA).
As well as rear panel rating plates, manuals, datasheets and online portals can also provide useful specification data sources for sizing information. Some UPS manufacturers also maintain equipment databases that can be accessed by companies such as Critical Power Supplies to help assess a client’s demand.
Real and apparent power are linked by a term known as Power Factor (pf). If only the Watt rating is provided it is common practice to total this and then divide the final sum by a power factor of 0.8 to give an overall kVA value. However, load power factors can vary from anywhere between 0.5 and 1.0 (unity) and care must be taken with this approach. In addition power factors can be ‘leading’ or ‘lagging’ and sizing for three-phase loads and generators adds further levels of complication to the calculation.
Two final considerations include runtime and redundancy. The runtime or battery backup required will influence overall system design, configuration, logistis and floor space requirements. For runtimes lasting several hours, it is a commonly adopted approach to use additional battery packs. For large three phase UPS this approach can be limited by space availability and budget. In such instances a standby generator can be more favorable.
Business continuity is normally the primary driver for installing a UPS solution. Resilience is a key factor that is driven by the levels of redundancy required. Normally this is referred to as parallel redundancy or N+X where X equates to the level required. For example, a 100kVA in an N+1 redundant configuration would require two 100kVA UPS systems to be installed, where one module alone could power the load if the other is removed out of service for maintenance or develops a fault.
Future expansion is off course a final factor to consider. It is normal practice to add a 20% expansion factor (multiply by 1.2). In addition, if the future electrical installation is allowed for, additional UPS modules could be installed to meet expansion at a later date.
UPS sizing is a relatively straight forward process. The larger the load and system, the more complex the calculations and considerations, especially in terms of resilience and capacity. For assistance in sizing your loads, designing a resilient system and selecting a UPS system, consider a Critical Power Supplies Site Survey by one of our Power Protection consultants.