Power conditioners are intended to maintain the supply to the load in the event of a supply dip. There are two generic types:
- Types with no energy store – in the event of a supply dip, the device must draw all the load power from the supply, but at reduced voltage and, consequently, increased current. This means that the capability of the device must be restricted by the maximum safe input current and, possibly, by duration of operation due to thermal effects. As a result, it can only operate with a certain minimum retained voltage and for a thermally limited duty cycle.
- Types with an energy store – in the event of a dip the device provides the differential energy to top up that available from the supply. These devices may be able to operate with very low or even zero retained voltage. The differential energy is provided by the dedicated energy store, which may be, for example, an ultra capacitor or a flywheel. Performance at the time of a dip is limited by the rate at which power can be extracted from the store and converted for use. Overall performance is further limited by the time required to recharge the store from the supply following an operation cycle – the device must be fully operational by the time the next dip occurs.
This leads to a list of required performance requirements:
- The power rating of the device must be sufficient for the load.
- The current capacity of the supply must be sufficient to supply the load, the losses in the conditioner, plus (i) the power needed to recharge the conditioner energy store within the required time following a dip, or (ii) the current required by the conditioner to generate the differential energy for the load during a dip, at the minimum retained voltage.
- For stored energy devices, the energy capacity must be sufficient to compensate for the energy not supplied by the supply during a dip, plus the losses in the conditioner.
- The duty cycle capability must be sufficient to deal with the expected or predicted time and magnitude distributions of dips at the supply point.
I believe that the advice to oversize by a factor of 2.5 refers to the energy capacity of devices with stored energy to ensure that, in service, there is always a sufficient reservoir of energy to provide adequate ride through. It is a design decision, rather than a disciplined engineering calculation, based on consideration of the risks and benefits (e.g., there is no benefit in surviving one dip if the device is not ready and available to protect against the next one) and the perceived risks and costs of downtime.
On the subject of losses, increasing the output power rating of the conditioner is likely to increase the losses. However, increasing the energy storage capacity may not have too much effect; most of the losses will occur during the discharge and recharge phases, rather than the idle time when the charge is simply maintained.