Power consumption information on computer loads is often not specified in a way that allows simple sizing of a UPS. Besides, if you know the exact power consumption of your PC, buying a correctly sized UPS can prove to be a daunting task. It is possible to configure systems that appear to be correctly sized but actually overload the UPS.
REASON: Many people are confused about the distinction between the Watt and Volt-Amp (VA) measures for UPS load sizing. According to Neil Rasmussen, the founder and the Chief Technical Officer of American Power Conversion, this confusion has been created by the many UPS manufacturers due to their failure of distinguishing between these measures. In this tech note I will try to explain the difference between Watt and VA and how the terms are correctly and incorrectly used in specifying power protection equipment.
Before proceeding with the explanation, first, we need to understand that Electricity is actually made up of three parts.
TRUE or REAL Power (WATT),
Apparent Power (VA),
Reactive Power (VAR).
Don’t worry about the explanation of these three parts at the moment as it will follow in the article. Ok, let’s start,
WATT is a unit of POWER and POWER is a product of VOLTAGE and CURRENT (P=VxI) or for making it easier for readers (P=VxA)
V or VOLT is used to represent VOLTAGE unit and AMPERE (I or A) is a unit for measuring CURRENT so theoretically (Volts) V x A (Amps)=WATTs, but it’s not always the case.
The terms VA (volt-amps) and watts are frequently used interchangeably when discussing the power consumption of an electronic device. VA is an expression of “APPARENT POWER” and WATT is an expression of “TRUE POWER” in AC circuit. In AC circuits, VA is referred to as APPARENT power or Power that appears to be flowing in the circuit. WATTs are referred to as TRUE power or an indication of the power that is truly being dissipated by the load. When the load is resistive, power dissipation in VA and watts will be the same.
For example, suppose we have a simple AC circuit. In this circuit, the power source is 220 volts, and the load is a simple light bulb with 440 ohms of resistance. The circuit current (A) can be calculated using Ohm’s Law by dividing the voltage (V) by the resistance (R).
A = V/R
A = 220/440
A = 0.5 amps
The power (P) consumed by the light bulb may be calculated using one of these formulas:
P = V xA
P = 220 x 0.5
P = 110 watts OR
P = A(square) x R
P = (0.5)square x 440
P = 110 watts
As the load is resistive, we can see from the above example that VA and Watts are absolutely same, that is 110.
Things don’t stay the same when the load becomes electronic. The constantly changing amplitude and polarity of AC power gives rise to reactive components in an electronic load. There are two types of reactance – inductive and capacitive – and they are opposite in nature. Together with resistance, they represent an opposition to AC current flow called impedance. VA and watts change because circuits with impedance exhibit a characteristic called power factor (pf). Power factor is the ratio of true power or watts to apparent power or volt amps. (Both are identical when the voltage and current are in-phase that seldom happens in AC circuit).
In addition to the power that reactive loads actually dissipate, a certain amount of power is absorbed by the reactive load and then once again released to the circuit. The power that is absorbed and then released again to the circuit is known as REACTIVE POWER, and it is the difference between apparent power and true power.
For example, consider the same AC circuit now with a computer as a load. The computer’s impedance is roughly known to be 60 ohms. So, if we simply apply Ohm’s Law, the current flowing in the circuit would be equal to:
A = V/R
A = 220/60
A = 3.66 amps
Again applying Ohm’s Law, the power consumed in the circuit would appear to be:
P = VxA
P = 220 x 3.66
P = 805 VA
Since the computer is a reactive load and not a resistive one, the power factor of the computer must be considered in order to determine the watts dissipated by the computer
P = V x A x pf
P = 220 x 3.66 x 0.60 (Computers with Power Factor Corrected supplies can have the power factor greater than 90% or equal to 1)
P = 484 watts
The difference between the 805 VA apparent power and the 484 watts of true power is the reactive power or
805 – 484 = 321 VAR or volt-amps-reactive.
Why Watt and VA rating may differ for a computer
Computers use an electronic switching power supply. There are two basic types of computer switching power supplies,
1) Power Factor Corrected supplies (PFC)
2) Non- PFC or Capacitor Input supplies.
It is not possible to tell which kind of power supply is used by just inspecting the equipment, and this information is not commonly provided in equipment specifications. Power Factor Corrected or PFC supplies have the characteristic that the Watt and VA ratings are equal (that is they have a power factor of 0.99 to 1.0).
Non-PFC or Capacitor Input supplies have the characteristic that the Watt rating is in the range of 0.55 to 0.75 times the VA rating (power factor of 0.55 to 0.75). Computer equipment made prior to 1996 also typically used this type of power supply and exhibited a power factor less than one.
UPS Power Rating
UPS have both maximum Watt ratings and maximum VA ratings, though not always mentioned together. Neither the Watt nor the VA rating of a UPS may be exceeded. It is a de-facto standard in the industry that the Watt rating is approximately 60% of the VA rating for small UPS systems, this being the typical power factor of common personal computer loads. In some cases, UPS manufacturers only publish the VA rating of the UPS. For small UPS designed for computer loads, which have only a VA rating, it is appropriate to assume that the Watt rating of the UPS is 60% of the published VA rating. For larger UPS systems, it is becoming common to focus on the Watt rating of the UPS, and to have equal Watt and VA ratings for the UPS, because the Watt and VA ratings of the typical loads are equal.
Example where a sizing problem can occur
Consider the case of a typical 1000VA UPS with no mentioning of power factor (pf). The user wants to power an 800VA computer with the UPS. If the computer has a Non-Power Factor Corrected power supply with 0.6 pf, then it will have a Watt rating of
800VA x 0.6 = 480 WATT
which is within the VA and WATT rating of the UPS, which is 60% of 1000VA or around 600W.
Now consider the same computer with a Power Factor Corrected power supply, it will then have a Watt rating of 800W and a VA rating of 800VA due to pf being 1. Although the VA rating of the load is 800VA, which is within the VA rating of the UPS, the UPS will not power this load because the 800W rating of the load exceeds the Watt rating of the UPS, which is 60% of 1000VA or around 600W.
CONCLUSION
Most UPS products are rated in VA and good brands always have a power factor rating that is prominently published as part of the product specification (Standard UPS usually have a power factor of 60% -65% and good brands may have a pf of more than 80%). In many cases, UPS power factors are designed to approximate computer power factors. A 1000 VA UPS with a power factor of 60% or 0.60 would deliver 600 watts, which means you are buying a 600VA UPS at a price of 1000VA. At low power levels, the differences between VA and watts are often slight. However, understanding the difference between VA and watts is very important to make sure the power protection device is compatible with the load.
If using equipment nameplate ratings for sizing, a user might configure a system, which appears to be correctly sized based on VA ratings but actually exceeds the UPS Watt rating. By sizing the VA rating of a load to be no greater than 60% of the VA rating of the UPS, it is impossible to exceed the Watt rating of the UPS. Therefore, unless you have high certainty of the Watt ratings of the loads, the safest approach is to keep the sum of the load nameplate ratings below 60% of the UPS VA rating. Note that this conservative sizing approach will typically give rise to an oversized UPS and a larger run time than expected.
