kVA to Watts Calculator
Work out what a UPS, inverter or power supply labelled in VA will actually run — the watt figure on the box, and why it is always the smaller of the two numbers printed there.
Last checked against the code
kVA to Watts Calculator
Updates as you type- Nameplate apparent powerThe larger number on the box
- —
- DifferenceCapacity the rating carries that a load cannot draw
- —
W = VA × PF · W = kVA × 1,000 × PF
How this is calculated
Pick up a UPS and there are two power figures on it: something like1500VA / 1350W. Neither is a typo and neither is optional. The first is apparent power, the total the output stage can pass. The second is real power, what your equipment can actually consume. Divide one by the other and you have the unit's design power factor, which is the only thing this conversion needs:
W = VA × PF
This site keeps a separate page for the same relationship at generator and facility-transformer scale, and the split is about equipment rather than arithmetic. Large plant is quoted in kVA and kW, and it is rated at 0.8 by convention, so one assumption covers most of the market. Small equipment is quoted in VA and watts, and there is no convention: the ratio is a design choice that varies from 0.6 to 1.0 across products sitting on the same shelf. A converter that defaulted to 0.8 here would be borrowing a number from a class of machine this page is not about.
What drives the ratio is the unit's topology. Older standby units, the cheapest kind, are commonly built at 0.6 — the origin of the familiar 1000VA/600W pairing that surprises people at the checkout.Line-interactive units, the volume of the market, sit around 0.7 to 0.8. Online double-conversion and newer lithium units reach 0.9, and a growing number are rated one-to-one, printing a single number because the two are equal.
Which figure binds depends on what you connect. A load with a poor power factor exhausts the VA rating first. A rack of modern IT equipment, running at 0.95 or better, exhausts the watt rating first — which is why the watt number is the one to compare your kit against and the one this page produces.
Two things the answer does not tell you. Duration is one: that comes off the runtime curve in the datasheet, which is a property of the cells rather than of this ratio. The other is what the unit pulls from the wall, which needs a supply voltage this conversion never asks for.
Worked example
One full calculation with real numbers, so you can follow along and check the tool by hand.
A 1,500 VA unit, rated at 0.9. The arithmetic is one line: 1,500 × 0.9 = 1,350 W. The remaining 150 VA of the rating is capacity the output stage can pass but no load can draw as real power.
Put a realistic setup behind it. A desktop workstation at 300 W, two monitors at 60 W each, a four-bay NAS at 60 W and network gear at 30 W comes to510 W. That fits inside 1,350 W with room to spare, and the unit is the right choice.
Now the same box, from a different product line. A budget standby unit with the identical 1,500 VA on the front, but built at 0.6: 1,500 × 0.6 =900 W. The 510 W setup above still runs — there was plenty of margin — so far, nothing has gone wrong.
The failure comes at the next step up. Add a second workstation and a monitor and the setup reaches around 1,000 W. Against the 0.9 unit that is comfortable, with 350 W of headroom. Against the 0.6 unit it is100 W over the limit, and the unit will refuse the load or drop it on transfer. Two products with the same headline number on the front, one of which does the job and one of which does not.
That is the entire argument for reading the second figure. And it is why the preset chips on this page start at 0.6 rather than at a comfortable middle value: the cheap unit is a real product that real people buy on the strength of the number printed largest, and assuming a modern ratio for it is the specific mistake that leaves someone 100 W short.
Visual comparison
| Design power factor | Bar | Real load supported (W) | Against a 1,000 W setup |
|---|---|---|---|
| PF 0.60 | 900 | Fails a 1,000 W load by 100 W | |
| PF 0.65 | 975 | Fails a 1,000 W load by 25 W | |
| PF 0.70 | 1050 | Runs a 1,000 W load | |
| PF 0.75 | 1125 | Runs a 1,000 W load | |
| PF 0.80 | 1200 | Runs a 1,000 W load | |
| PF 0.85 | 1275 | Runs a 1,000 W load | |
| PF 0.90 | 1350 | Runs a 1,000 W load | |
| PF 0.95 | 1425 | Runs a 1,000 W load | |
| PF 1.00 | 1500 | Runs a 1,000 W load |
Source: A fixed 1,500 VA rating evaluated across the ratios small equipment is built at.
Reference tables
This table turns the previous chart into a shopping tool. Find the VA rating you are looking at down the left, then read across to the column matching that unit's design power factor — the ratio of its own two printed figures — and the cell is the real load it supports.
The most useful way to read it is diagonally. A 1,000 VA unit at 0.9 supports 900 W, and so does a 1,500 VA unit at 0.6. Those are the same capability from ratings that differ by half, which means comparing units on VA alone is close to meaningless across product tiers. Shop on the watt column and the comparison becomes honest.
Use the table rather than the converter when you are choosing between several candidate products, since it puts them side by side. Use the converter when you have one unit's exact ratio in front of you and want the precise figure rather than the nearest column.
These ratings are what product lines cluster around, not a standard series. One manufacturer's 1500 VA rung is another's 1440 VA or 1600 VA, and rack-mount ranges step differently again. Confirm against the actual product before ordering.
| Rating | At 0.6W | At 0.8W | At 0.9W | At 1.0W | Typically bought for |
|---|---|---|---|---|---|
| 500 VA | 300 | 400 | 450 | 500 | Entry-level desktop standby unit. Enough for a router and a modem through a short outage, and little else. |
| 650 VA | 390 | 520 | 585 | 650 | A single desktop computer and one monitor, sized for a clean shutdown rather than for riding through. |
| 750 VA | 450 | 600 | 675 | 750 | The smallest size commonly offered with pure sine wave output, which some equipment and most active PFC power supplies want. |
| 1,000 VA | 600 | 800 | 900 | 1000 | A workstation with two monitors, or a small network cabinet. The first size where meaningful runtime becomes possible. |
| 1,500 VA | 900 | 1200 | 1350 | 1500 | The volume seller. A workstation plus peripherals, or a home lab, or a small rack with a switch and a NAS. |
| 2,200 VA | 1320 | 1760 | 1980 | 2200 | Rack-mount territory. A small server or two, or an AV rack, usually with an external battery option. |
| 3,000 VA | 1800 | 2400 | 2700 | 3000 | The top of what runs from a standard 15 A or 20 A receptacle. Above this a unit generally needs its own dedicated circuit. |
Common mistakes
Buying by the VA number on the front of the box
A 1000VA unit that turns out to be 600 W will not run a 700 W setup, however prominently the first figure is printed. The watt rating is on the specification sheet or the back label, it is the binding one, and on budget standby units it is roughly 60% of the headline.
Assuming a power factor instead of reading the two ratings
You almost never have to guess here. Divide the unit’s watt rating by its VA rating and you have its exact design power factor. A preset is only for the case where you are comparing products you do not have the sheets for yet.
Confusing the watt rating with runtime
Two different components set them. The output stage fixes the watts and decides whether your kit runs at all; the cells fix the minutes. Every specification sheet lists runtime as a separate curve against load, and a unit sitting comfortably inside its watt rating can still expire in ninety seconds.
Putting a laser printer or a motor on it
Inrush is judged against the inverter’s instantaneous capability, not the steady rating you just calculated. A laser printer waking its fuser is the classic offender — it overloads units that are correctly sized on paper, which is why printers are the one thing manufacturers tell you to leave on the unprotected outlets.
Forgetting the unit itself has to be plugged in somewhere
A 3000 VA unit drawing near its full rating is close to the working limit of a 15 A receptacle, and charging its batteries after an outage adds to that. At the top of this range the unit generally wants a circuit of its own.
This tool provides planning estimates. Always verify final values against your local code and a licensed electrician.
Frequently asked questions
How many watts is a 1500VA UPS?
Between 900 W and 1,500 W, depending entirely on which unit you bought. A budget standby model rated at 0.6 gives 900 W. A mid-range line-interactive one at 0.8 gives 1,200 W. A modern online or lithium unit at 0.9 to 1.0 gives 1,350 W or the full 1,500 W. The VA figure alone genuinely does not answer the question — that is why the box prints two numbers.
Why does my UPS list both VA and watts?
Because they are limited by different things and either can bind first. The VA figure reflects how much current the inverter stage can pass; the watt figure reflects how much real power it can actually supply. A load with a poor power factor will hit the VA ceiling first, and a modern near-unity IT load will hit the watt ceiling first. Manufacturers print both because they cannot know in advance which kind of load you will connect.
What power factor do modern UPS units have?
The market has been moving steadily toward unity. Standby units from the 2000s were commonly 0.6; line-interactive units settled around 0.7 to 0.8; current online and lithium models are frequently 0.9, and an increasing number are rated 1:1 with a single figure on the box. If you are replacing an old unit, do not assume the new one shares its ratio — you may be able to buy a smaller VA rating for the same real load.
Is this the same calculation as kVA to kW?
The arithmetic is identical and only the scale differs — watts are simply kilowatts times a thousand. The pages are separate because the equipment is: a generator nameplate is quoted in kVA and kW and conventionally rated at 0.8, while a desk-side unit is quoted in VA and watts and can sit anywhere from 0.6 to 1.0. Using this page for a 100 kVA generator would work, but you would be typing 100,000 into a field built for four-digit numbers and reading an answer in a unit nobody quotes at that size.
Do I add up idle draw or peak draw for my equipment?
Use normal running draw, measured or estimated with everything switched on and doing what it usually does. Idle figures understate a workstation badly, since a modern CPU can triple its consumption between desktop and full load. Genuine peaks matter less than people expect for IT equipment, because they are brief and the unit rides them out — the exception is anything with a motor or a heating element, which should not be on the unit at all.
Does this tell me what current the unit draws?
No, and that is a separate question with a separate tool. Current depends on the supply voltage, which this conversion never asks for. If you need to know what the unit pulls from the wall — to check it against a receptacle or a circuit — take its VA rating to the kVA to amps converter along with your supply voltage, or take the watt figure to the watts to amps converter.