What's the Current at the Input for a UPS?

George,

More Info Please.

Is this a single or 3phase UPS?

How big is the battery string?

What size batteries?

Do you want maximums or 25%, 50%, 75% and 90%? (never load a UPS to 100%)

Usually the UPS Input will be about 115-120% of the rated Output. This is to cater for a Full recharge after a Heavy Discharge and unavoidable losses in th system.

With the extra information I could help with a better answer.

Regards,
Sapper.

You might want to revisit the equation list you got these from. The HP of the device does not cancel out, it must be included. You must use the numbers given in the kVA/HP chart and the motor HP for this formula to work. The equations listed will give you current for the rated kVA of the UPS and nothing more.

For the thread, keep in mind not just inrush but the battery charger load when the batteries are severely discharged. The initial charging current could last 60 to 90 seconds (maybe more) and if the rectifier input breaker is undersized it will open. I have seen UPS's where the breaker would need to be reset one, two, or even three times because of the charging current demand from depleted batteries. Not the ideal set up, by any stretch but that was what we had. I do not like to second guess someone after I have seen everything in the rear view mirror but hey, that's engineering. If you don't think of everything in the planning stage someone else will after installation.

I see what you are saying - KVA/HP is a unit for a constant; not one value devided by the other for conversion. Thanks for pointing this out.

John,

I was originally going to respond along the same lines.

But, I figured that based on the wording of the question, he didn't have these available, so went for the numbers approach.

Could also say that the Supply side cable gives you an Absolute limit too.

George,

This would give you better results for Your UPS as opposed to a theoretical UPS.

Regards,
Sapper

Thanks for the answers.

What I m asking for is a practical way (rule of thumb). Let me give you an example.

We suppose that we have a 50 kVA UPS output if we suppose :

1. rectifier , n=0.95

2. inverter , n = 0.95

then ntotal = 0.95*0.95 = 0.9, thus the input power (until now) is P = 50/0.9=56 kVA

What I m missing is the power rrequired for the batteries and the losses of chraging system. So, if for example i have 50 kVA UPS and 30 min autonomy , how can I find the required power of chargng (rule of thumb, not an analytican approach) ?

Consider a 3 phase UPS 400V/50 Hz

Thank you again

George

Dear all,

During estimation stage with UPS output(kVA) is the only information, I take the following from my own experience.

MCCB and power cable UPS supply : based on twice the UPS output.

(UPS KVA + unknown battery charging KVA = 2 x UPS KVA)

MCCB and power cable for bypass (3phase input, 1phase output) : based on twice the UPS output.

(3 phase input is unbalanced, one phase has twice the current of others. Maximum phase current = 2 x calculated 3 phase current)

MCCB and power cable for bypass (3phase input, 3phase output) : based on UPS output. (In case 3phase UPS, bypass transformer is 3phase transformer )

Load calculation : UPS output

Regards

Akihito,

The current on one phase cannot be double even if you have the same load divided on only 2 phases. Usually the UPS overcurrent protection will limit the current at a lower value, so if one phase (out of 3) is lost, load shedding will have to occur, rather than shutting down the UPS.

Battery charging current is usually limited to a certain percentage of the input current (5-15%), except in cases when turbo-charge is provided.

Roughly, as pointed out in one of previous responses, assuming an "efficiency factor" to account for losses inside the UPS of 0.9 * 0.9 = 0.81, the input power [kVA] would be Pout/0.81 [kVA]. Considering the input voltage [V], the input current Iin[A] can be calculated.

To account for the necessary current to charge the battery, consider 1.15 * Iin.

The wire size for this input shall be based on this current. To size it based on a double current is too much in my oppinion.

Dear tomad,

I have been involved around 100 UPS installations. Every time, I check the main MCCB size that UPS vendor selected. My double sized MCCB comes from this experience.

I asked why their main MCCB is nearly twice the current of output KVA, they explained that their main MCCB is sized for UPS output + battery boost charge.

If, UPS has a 125A main MCCB, we must supply power with at least 125A MCCB. Supply cable must have ampacity to meet the125A MCCB. This is true even the load current is 60A.

For bypass circuit, 3phase input 1 phase output transformer (primary V connected, secondary inverted v connected) has very tricky primary currents.

2 out of 3 phase current is same as normal 3phase transformer of delta or star connection, but one phase is always twice the current of others. This comes from transformer winding arrangement and not avoidable.

So for instance, 480V 3phase in (primary) 1phase out (secondary) 50kVA has primary current of 60A in two phases, and 120A in the last phase.

So, supply MCCB may be 150A, and cable amapcity is around 150A.

Regards

Akihito,

I do not understand what you are referring to (You indicate V and inverted V connections). Are we talking about 3 phase transformers or 1 phase transformers ? For the best use of a transformer, the load should be balanced as much as possible. A balanced load is also best for the supply system.

Please attach a diagram of the transformer you are referring to.

You mentioned that you have been involved in several UPS installations. Did you do any UPS current (input and output) measurements ?

Mr. tomad,

Please disregard "inverted V", since it is not correct term.

I could not find how to attached a pdf file here, so if you do not mind, access my web page, www.akihito-shigeno.com/files/EES_E026_030.zip.

I explained how the phase currents follow in 3phase input 1phase output transformer, in EES E027 and E027.1.

Many UPSs are single phase output, and 3phase input. So as the bypass transformers. In case 3phase in/1phase out transformer is selected, the current unbalance is not avoidable.

I do not measure the input current, but I select supply MCCB and cable based on the UPS's main MCCB.

For load calculation, 50kVA UPS will consume 50kVA, since practically, UPS never become full load.

Regards

Akihito,

I looked at the attached files. The situation described in it is that of 2 single phase transformers connected together (open delta) primary and with the secondary winding in series. It is not a 3-phase input 1-phase output transformer.

I do not know why do you say that many UPS are 3-phase input and 1-phase output, since on the contrary, normal UPS are either single phase in - single phase out or 3-phase in - 3-phase out. Sometimes they include a transformer if the input and output voltages differ, or, as separation transformers (same voltage in-out).

Bypass connection is to allow separation of the UPS in case something happens without losing power supply to the load. In this case the voltage applied to the load is unconditioned, it is that of the power network.

Input power HAS TO BE HIGHER than the output power because you have losses, regardless if full-load, half-load or no-load.

Dear Mr. tomad,

Any 3phase transformer is 2 or 3 sets of single phase transformers equivalent.

So, when analyzing 3phase transformer current/voltage, you can consider there are 2 or 3 sets of 1phase transformers.

3 phase in /1phase out transformer is equivalent to two 1phase transformers, but may be constructed as a single unit.

In my experience, all UPSs (for industrial) were 3phase input/1phase output, except for one 400VA 1phase input/1phase out square wave UPS, for my PC.

What I have been saying here is the sizing of supply side MCCB and cable, not the input current itself. We need to coordinate supply side MCCB, may be in one of MCC unit, and the Main (Incoming) MCCB of UPS. If UPS is equipped with 150A Main Incoming MCCB, we must supply 150A Supply side MCCB even the our calculated full load current is 75A. In the beginning of the designing stage, we do not have this information, however, we must order our MCC units before the receiving UPS drawing. I use this "twice the UPS KVA output" as a rule of thumb.

I found the original question of this thread is "current", so my answer may not be relevant here. However, if the current is used for MCCB and cable sizing, my statement still have meaning.

Regards.

I HAVE ALWAY USED A CONSERVATIVE 150% FIGURE TO ESTIMATE LOAD FOR A UPS. EXAMPLE: 50KVA AT 240 VOLTS SINGLE PHASE MEANS A CURRENT DRAW TO THE LOAD OF 208 AMPS. IN ORDER TO REALISTICALLY SERVICE THE LOAD, ALLOW FOR POWER FACTOR LOSSES AND KEEP THE BATTERIES CHARGED, AN IPUT CURRENT OF 300 AMPS WILL DO THE JOB. THIS HAS NEVER LET ME DOWN - NO CUSTOMER COMPALINTS. IF YOU WANT TO FIGURE IN BATTERY LIFE, SURGES, TVSS, AND LINE LOSSES YOU SHOULD PROBABLY CONTACT A VENDOR THAQT SPECIALIZES IN SUCH LARGE UNITS FOR ENGINEERING SUPPORT. LIEBERT COMES TO MIND. ALSO - REMEMBER THAT MOTORS, CONDENSER UNITS, FANS AND COMPRESSORS TEND TO BLOW UP THE CALCULATIONS UNLESS YOU USE VFD'S OR SOFT-STARRS.

Please STOP SHOUTING! People round here are trying to get some sleep.

It is difficult to say with a current meter, you really need a power meter and make a graph of the power used against time.

You have losses in the charger and in the battery, the state of the charge at the beginning will also change the amount of power needed......

There are power meters around for little money that might help you....I have one but have never used it up to now.....sorry!!!

On a good day, the 150% rule mentioned by another poster would probably be very close.......

Hi Andy,

The OP asked: "What's the Current at the Input for a UPS?" - so watt's* with the power meter (vs a current meter)?

[* Sorry ]

A practical way would be to measure it using a clamp-around ammeter probe.

http://www.google.com/search?hl=en&q=clamp+on+ammeter&aq=0&oq=clamp+on+

There are 4 things you need to know to calculate maximum input current of a ups.

1) Maximum output VA

2) Reserve battery charging capacity (usually around 20-25% of va output)

3) Efficiency of ups

4) Input power factor

The first question that should have been asked is what type of UPS does he have. There are three types (unless anyone can add to the list):-

1. Power conditioning, where the input is rectified and re-inverted with a battery connected to the intermediate DC bus.

2. Fast switching, where the input is connected straight through to the output and a mains fail (out of limits) switches an inverter to the output.

3. Synchronised switching, where the input is switched straight through to the output and the inverter runs in parallel sync.

Option 1 gives the best service but is the least efficient as both rectification and inversion losses need accounting for. The charge efficiency factor of the batteries does not really come into the account as normally the inverter is supplied from a dc power supply. Thus his max I/P amps would be battery charger circuit load + O/P x 1/rectifier efficiency x 1/inverter efficiency. The efficiency given as 0.95 for each stage is I think too optimistic. Better to use 0.9 per stage, 0.8 overall. I would be surprised if a unit as big as 50 kVA did not have a soft start facility. Another issue here is the power factor presented by the rectifier circuit to the grid. As all the power is rectified in this option then, if no PF correction circuit is fitted, it will present a really bad, spiky load.

Option 2 is the cheapest choice with no power conditioning. It is however very efficient since it is in normal condition only a battery charger with a battery on float charge.

Option 3 is also very efficient dependant on whether it has power conditioning. With conditioning it may present a PF1 load to the supply grid at the expense of efficiency.

regards

Chas

Goodness, a lot of answers without actually answering the question!

Below are rules of thumb.

50kVA is the rated output of the UPS. This is quoted at a fixed power factor, usually 0.8 (but Chloride for example use 0.7 in some units). At 0.8pf then the kW output is 40.

Now the UPS will be inefficient. The manufacturer will quote figures like 95-98% - but this is at full load.

As a rule of thumb, don't load a UPS above 80% maximum. At this level the fixed losses decrease the % efficiency. 90% is a good guestimate. So now we have 10% losses which is another 4 kW.

The UPS has a battery charger for, well, guess what. Forget about the autonomy time. The charger capacity is fixed and increasing the battery quantity will not increase its output, just the length of time it takes to recharge. You should hit the manufacturers specs but 20% is my rule of thumb. So +8 kW for battery charging (from flat of course, in normal use the charger will be trickling).

So 40 + 4 + 8 = 52 kW of maximum input power.

Now we can tackle current. This depends on the power factor of the input. The manufacturer will quote figures like 0.95 - but again this is at full load. As load decreases the PF also decreases; this can go as low as 0.6 for big UPS. I like to use 0.7 at the input.

52 kW @ 0.7pf = 74 kVA

Then we have harmonics. Again, as output load decreases, input harmonics increase. I use 5% for harmonics, = 4 kVA.

So we now have about an 80 kVA load, or about 115amps. You could put this on a 125amp MCCB but you may be wiser and go for the next size 160amp. This will give some headroom for expansion.

Which, as some will gleefully point out, is about twice the original output current....

Baby Guinness,

Your algorithm is OK, but the conclusion, in my oppinion, is a little to conservative: for a 50 kVA load you indicated a 80 kVA input power (ratio = 80/50 = 1.6)

I just looked up the technical characteristics of a 80 kW (100 kVA, 3 phase, 60 Hz, 480V input, 208V output) UPS by APC (Schneider Electric), which has the 480/208V transformer included. This transformer is 112.5 kVA. So, making the ratio input ( 112.5 kVA) vs output (100 kVA), yields 1.12, which is considerably lower than the one indicated by you. Definitely, this ratio includes a combination of input-output efficiency/power factor, but this is basically what determines the currents (considering the respective voltages) and so the wire sizes, circuit breakers, etc.

Based on the attached curve, the (output I assume) efficiency would be 96% ( high efficiency mode) and 92% (double conversion mode) between 60% and 100% of load; you selected 90% (rule of thumb).

At lower load levels the efficiency decreases, but in this case the 40 kW value also decreases (assume at 50% load = 20 kW), so the 52 kW become 32 kW.

Further, the now 32 kW, even considering a (input) power factor of 0.5 (vs the indicated 0.7) yield only 64 kVA.

There was an earlier topic on what the 0.8 (output) "fixed" power factor represents: maximum or minimum, as well as how the input and output efficiencies were defined (a ratio of what).

This is a bit old thread with many good information. I think the previous thread may be helpful, it has detail calculation for the UPS design (especially the Charger Capacity Calculation section in this context):

http://cr4.globalspec.com/comment/404170

- MS

Use Tivoli Software for which you must have instslled SNMP card with your UPS sensors that will collect the data and send it to main server from which you can monitor input output charactiristics of UPS ...

It should say so in the original equipment manufacturer's information that was supplied with it. If the information is not to hand, a telephone call to it should precipitate a replacement.