Calculating Transformer Losses

do it yourself ... have a try at first ,,, as you're a student .....

good luck

I have been trying to find the right equation to use. It would seem that I multiply current times voltage for each of the legs of the transformer. That would be 1.7 amps by 4800 volts for leg 1 to get 8160 watts loss for that leg and do the same for the other two legs. If I did that and add the power, I would get something like 31 kW of loss in the transformer. However, there may be some more involved calculations because of the delta and wye configurations. My class is in energy management vs. a detailed, technical class on electric transformers and such, so the textbook is not much help in this case. I would appreciate any guidance in transformer theory that I could read and use......

Matthew

Hi Mathew,

Normally any existing power supply system has KW,KVA , KVAR and PF meters to measure the energy consumption. The Low Voltage outgoing system also will have Kw meter , Ammeter and Voltmeter and PF meter. If you record these readings taken simultaneously and Calculate the Input and Out put energy consumption the difference is transformation loss. Take the readings and then do the simple calculation to find the actual loss.

Aren't most industrial power sources still metered from the low voltage secondary side?

There may be some small additional power loss from running the bigger transformers but if they still need 350 KVA of capacity the actual power losses from the bigger transformers are still going to be negligible compared to the overall power consumed any way.

Do a cost of replacement Vs energy saved comparison. If the replacement cost takes more than 10 years to offset the energy savings its probably not worth doing.

As per guidelines (of course the new NEMA and DOE) the efficiency of transformer is > 99%

The losses - partially Cu (load dependent) and partilally Iron (Independant of Load) is thus less than 1% (Infact for 1500KVA it must be >99.5%)

at 3000 KVA this will boil down to much less than 30KW at full load even at 99%

for 700 KVA (2x350) this is 7KW

So your saving is less than 20KW and in reality may be somewhere less than 10KW.

It may not really justify selling these off and buing new.

But I see another problem - why the current are unbalanced on no-load?

Thanks for the comments so far. These are very old transformers-I do know they do not contain PCBs and I have contacted the manufacturer for some build data as well.

They have two transformers and only one meter, which would lead me to believe it is on the primary side of the transformers. However, I am visiting the site tomorrow and will seek out where the meter is located.

I also observed the unbalance in the legs of the circuit. What effect does that have and how do I take more data or do more investigation as to the extent of the unbalance? They have two transformers- the split is probably about 85/15. The transformers are separated by about 200 feet and are "busbar-ed" into the respective distribution panels.

Ask to check the insulation condition. very old transformer hopefully not at the last leg of journey.

Watts are volt X amps X power factor.

Actual watts are probably much lower than calculated.

Unbalanced currents are not a problem. 3 phase transformers usually have unbalanced currents at no load.

User also probably has single phase loads. This would explain unbalanced loads.

Oil in transformer should be checked to see if it has PCB's. Even though transformer wasn't manufactured with PCB's, it could have become contaminated during repair.

Transformer efficiency probably hasn't changed since transformer was manufactured.

To make efficiency calculations, you will either have to use manufacture's data or make measurements of transformer losses.

Now I am starting to get it!!! For the three legs of one transformer, there was data taken some time ago that summed 1.7 + 1.7 + 2.8 amps at 4800 volts on the primary side at no load. I also know from utility data that the PF drops to below 0.6 at very low draws (I do not know what it is at actual no load), so the power loss would be about 17.8 kW in this case. They have two transformers, so I would multiply by two to get just over 35 kW of loss in the two transformers.

The transformers were inspected a few years ago and were certified to be free from PCB's and the oil was checked for performance.

So now I need to understand what would be the losses in a single 500 KVA transformer at 4800 v primary, or if the utility could provide it, 13,200V primary. This is an older facility and 4800 V primary is no longer common in this location.

If I contacted a transformer company, I suspect they could answer that question about a transformer?

First, thanks for the response. This has been a journey to dig into this transformer. I have contacted the manufacturer and they are looking through the archives for some data. What exactly am I looking for from them? Couldn't I use data from another 1500 KVA 4800/480 transformer or do they differ that much as built?

Is the 1.73 factor in your calculation because of the 3-phases of the transformer?

Thanks for the response!

Hi,

Since your system voltage is upgraded from 4.8 kV to 13.2 kV, you need to change the existing transformers anyway. 500 kVA Transformer will do for the requirement specified by you.

Tentative Loss figures for 500 kVA Transformer :

No load losses (Iron loss) : 800 Watts

Load losses (Copper loss) : 3500 Watts

The power factor is probably much lower than 0.6 at zero load. A halfway decently specified transformer has about 1% efficiency at full load, so the full load loss of the pair would be about 30-kW. Now, much of the loss at full load is I²R loss the windings, so this reduces significantly at low dissipation - you should expect about 7.5-kW for the two.

In short, either your 30-kW is incorrect, or the transformers are defective.

Note that winding losses will always be lower with large transformers - it's the core losses that increase. A typical rule of thumb is that no-load losses are about 1/4 of full-load losses, and load losses rise as the square of the load power*. On this basis, if the transformer is in good working order there will be little or no benefit in using a smaller transformer when it is running at above 30% capacity.
* I.e. loss ≈ full-load-power/400 + 3(VA)²/full-load-power

If you know the time profile of the load, you can roughly calculate the amount of energy lost in the transformer. My guess would be that buying and fitting new transformers is sufficiently expensive that the best bet may be to disconnect one transformer and take all the power from the other. But it's probably advisable to leave a spare in place in case of failure.

Sometimes (albeit rarely with this size of transformer) the windings can be re-jigged to reduce the field in the core. The cost is that this increases the winding resistance. The end result is that the losses resemble those of a smaller transformer - and you don't have to buy a new one.

Here is an on-line text with some pretty good basics on transformers and multiphase systems;

http://www.allaboutcircuits.com/vol_2/index.html

You may find this helpful.

Thanks for all the great comments and suggestions! I am reading through the online text on transformers. Are the losses in the transformer proportional to the power through it? Or is there some baseline losses regardless of what power the facility is drawing?

Hi Mathew,

You are to do the energy audit of the company. The purpose of energy audit is to study the present load conditions and calculate the existing pattern of the present and past energy consumption and examine th e excess energy consumption than actually required due to various reasons and to recommend possible remedial measures to reduce the consumptions to optimum level.

1. Statistical study of the load pattern for the past one year, transformer loading , power factor , KVA and KVAR and Kw ( energy consumption)

2. Study the type of load ( motors, furnace ,lighting, ) and weather optimum use of these load is achieved or wasted .

Whether the motors used are over capacity resulting in loss of energy , weather the transformers are over capacity depending on the actual load, the power factor is proper or Power factor corrector panel is to be installed.

The study should be based on present conditions and the recommendations should be based on coast analysis based on recommended investments and savings and pay back periods.

Please read the link below

http://www.nicorinc.com/en_us/commercial/planning_needs/build_strategy/energy_audit.htm

That is correct. The facility has a lot of T8 flourescent tubes for illumination and 1 75 hp air compressor which runs the majority of the equipment. They also have a small paint booth and drying oven. They run a pretty tight ship and turn things off between shifts and have reasonable climate control procedures. We have looked at the pattern of consumption over the last 18 months and it appears that the lowest consumption (on weekends when plant is not operating) is still a bit high (in my opinion). We identified these transformers as being larger than necessary so I am trying to quantify how much energy is being consumed by them. We have a quote to replace the two existing tranformers with a single smaller one. My thought is that the transformers are consuming energy 24 hours per day/every day of the year, so 25 kWatts loss results in over $20,000 per year. My goal in this thread is to understand if the transformer losses are near 25 kWatts or are they much smaller at 5 kWatts.

The losses in the transformer are approximately I²R. The resistance of the coil is not known. You need to contact a MFG rep to assist in getting this figure. You need this figure to make a close estimate of the KWH used. We are assuming that the electric service is primary metered. Is this correct?

I2R losses are always lower for high-power transformers. But the core losses increase. See post #18 for rules-of-thumb. (But you may be able to get appropriate values from the transformer manufacturer)

The facility is primary metered. The single meter is on the primary utility side and feeds both transformers. I have contacted the transformer manufacturer and they are looking through the archives for the build info on these specific transformers.

LTU Student

OK, next analysis step is to record the time-dependent data so you are ready to calculate the actual losses when the transformer data arrives. In the meantime, it would clearly be beneficial to disconnect one of the transformers from the incoming supply if this is practical.

We have made arrangements to turn off all the loads from the two transformers. We will go to the facility on a weekend and turn off all breakers, followed by all switches at the first distribution panel. This will leave only the two transformers on the meter. The utility rep has provided the instructions on how to retrieve more info from the meter on a real time basis.

So the assumption would be that whatever losses we record at no load are present all the time. The losses do not change at different loads?

Thanks!!

The losses present with no load are there all the time. But of course there are additional losses when the transformer is loaded. These are proportional to the square of the current (not the power). See the equation in my previous posting. (The form is correct, but of course the numbers will depend on the specific design of the transformer).

If when you switch back on you can arrange to drive the factory from just one of the transformers and leave the other transformer switched out you will of course achieve more than half of the savings that would be available from using a smaller transformer (assuming the smaller transformer is a scaled-down version).

I want to thank all the people who have replied to this thread. I think we have a course of action to determine the wasted energy of the transformers and compare it to the cost of an appropriately sized new unit. The previous occupant of the building used much more electricity and needed the two transformers. The current occupant needs 1/10 of the transformer capacity. They have a generator back-up unit for critical electrical loads and can allow the manufacturing process to stop for a period of time due to an outage. One of the 1500 KVA transformers is probably supplying about 35 KVA peak and the other is supplying 310 KVA peak, But much of the time they are supplying much less than peak. The transformers are separated by several hundred feet and are busbar-ed into the distribution panel so the operation to replace them is not very easy.

It would cost a lot to change to a different primary voltage so we will stay with 4800 volts primary. I also wonder if we can go to a dry transformer vs. the oil-filled. A standard single new transformer would be 500 KVA. It is outside and the ambient temp can reach 100 F for short times. I will ask the transformer company and the electrician what they recommend in this application.

I will also obtain the build specs of the new transformer so there are not as many questions in the future.

satisfied LTU student

Assuming it is not easy to supply the loads driven by the lighter-loaded transformer from the heavier one, the option of disconnecting the lighter-loaded one and replacing it with a 35-to-50kVA transformer looks to be a relatively cheap starting option. Using a small separate transformer may well be a better option overall than rewiring the site to run the 35kVA load from the 500-kVA transformer.

You should also consider the wiring losses involved in reconnection: the costs of "several hundred feet" of wiring losses at 480 Volts could well justify the cost of the smaller transformer even without taking into account an costs associated with rewiring.

N.B. Whether this is the best solution depends on the present routing of the high Voltage lines between the two transformers - if the HV input is at the heavily-loaded transformer and the HV line between the transformers is owned by the site (i.e. not by the utility company), it could be used for bussing LV from the heavily-loaded transformer to the lightly-loaded one. (The current rating of this wiring would be sufficient for a load of more than 80 kVA when distributing 450-V)

Hi Mathew,

What you assume about the energy loss of transformer under no load conditions is in the correct direction. But before reaching at the final recommendation of replacement by single transformer consider the following points.

1. For normal operating conditions and breakdown condition of one transformer weather the complete system has to be shut down. In my experience I do not recommend to replace two transformers with one. On the contrary it is better to redesign the existing transformers with a bus coupler system to run the total load on a single e transformer and shut down one transformer during lean period and holidays when only nominal general service load is to operate. The actual no load consumption can be practically found as per the method I explained in my earlier reply. This can be designed with proper CTs and automatic pre planned switching control of the transformer breakers depending on load conditions manually or automatically.

2. Find out the present actual no load loss of the two existing transformers and find out the possible no load loss of the proposed transformers from the manufactures and consider the difference as the possible gain by replacement and calculate the long term pay back on the investment. In one of the energy audit I conducted in recent past I recommended for replacement not considering the no load loss but the transmission loss. In that case the existing system. It was a very large Factory located in far away plants with a net work of HV and LV distribution and transformers .The system was 11/3.3/ 0.44 kV distribution system which I replaced with 33/11/0.44 kV and all oil filled transformers replaced with energy efficient dry type transformers and total replacement of HV and LV panel with bus couplers and automatic power factor correcting panels. The result was huge reduction of energy consumption and the total investment was recovered in five years even though my recommendation was pay back period of seven years.

3. Investments to be recommended for area where there you are confirmed that there will be substantial energy saving for the recommended investment for the new system

4. The final audit report should be technically sound and financially possible by the company. It is a time consuming job which require minute and systematic statistic study of the system, operational limitations, work schedule and type of production activity. (For example a continuous process industry cannot allow power shutdown even on lean production time or shift changes)

You are interested in learing about what energy the transformer consumes.

This is most accurately performed using infrared analysis.

FLIR teaches courses in this specifically or you can hire one of their certified IR technicians preferably a level III or a level II with at least 10 years experience.

The infrared image is analyzed along with the FLIR software and calculates energy loss. Emissivity, windspeed, time of day readings, humidy, distance, surrounding ambient temperature, transformer temperature, and other factors influence the measurement. The calculation of this loss allows you to predict how the transformer will function under heavier loads.

FLIR also has free software that allows you to predict heat losses under varying loads.

Refer to the following free FLIR documents on their website.

The relationship between current load and temperature

for quasi-steady state and transient conditions

Bernard R. Lyon Jr., Gary L. Orlove, and Donna L. Peters

Infrared Training Center, 16 Esquire Road, North Billerica, MA 01862-2598

Common misconceptions in infrared thermography condition based

maintenance applications

Robert P. Madding, Director

Infrared Training Center, North Billerica, MA

Development of a Utility Infrared Thermography Preventive Maintenance Program – With Lessons Learned II

Michael A. Kregg, Senior Engineer and T&D Infrared Thermography Program Coordinator

Commonwealth Edison, Maywood, IL 60153

Good Luck.

The 1.7-2.8A reading is not good enough to do a power loss calculation, as you don't know what the PF of that current is. (It can give you the kVA "demand" of an unloaded transformer, but it can't give you the kW power wasted by the unloaded transformer.)

To save this company money, you will have to look at the following factors:

1) the "copper" (I-squared-R) losses which will be incredibly low due to the fact that current is very low relative to the winding resistance (this would most likely go up if you installed a smaller transformer.)

2) The "iron" (hysterisis, eddy current, & harmonic) losses that are a function of the size of the transformer (this is usually small relative to I-squared-R but would go down with a smalller transformer.)

3) The power company's demand ($/kVA) charges. You would have to study how the billing is done on this, as it is a way for a power company to "punish" industrial consumers that have bad power factor or large surges in their current draw. You would have to familiarize yourself with the actuall billing contract as the demand charges are often priced differently between "peak" and "off-peak" periods.

My gut feeling is that you can probably find cheaper and easier ways to help this company save money.

Good luck with your studies.

The local utility has software that shows PF relative to KVA and KW demand. On the weekend, when most loads are turned off, the PF drops below 60%. When the plant is operating, the PF goes up well above 80%. If I multiply 1.7amps times 4800 volts times .6 PF, the result is 4896 watts. That is for one leg of one transformer. It seems to me I should multiply that power by 3 legs and multiply that result by 2 transformers to get 29376 watts. At a US dime per kWh, I would get $2.93 per hour and $25,773 per year of energy consumption. At this cost per year, it seems like there is an opportunity to change to a smaller, more efficient transformer.

On a weekend, the lowest consumption from the local utility is 60-70 kW with all of the industrial loads turned off. They shut down the air compressor, paint drying system, and turn off most of the lights. I will visit the location on a weekend to see what is left on next week. It is hard to believe a couple of computers and some security lights would be 70 kW, which is why I am looking at the losses in the transformers.

Matthew

This looks like an object lesson in how dangerous it can be to draw conclusions from partial information.

The expected unloaded dissipation of 300-kW of transformers is about 300-kW/400, or 750-watts. If more than 3-kW is being wasted this way there's almost certainly something significantly wrong.

You should alsbe very careful about that 0.6 PF - the PF of an unloaded transformer will often be very low indeed (the winding is mainly an inductor, after all). Often, this is compensated with parallel capacitance, but in practice the PF when unloaded will depend on the load and duty-cycle for which it was originally installed.

If it does turn out that this phase is actually consuming anything like 5-kW, it's worth remembering that, indvidually, small loads seldom use more than one phase. Given the geography of the site and the way people wire up to small loads, it's also possible that more than half your weekend load is on a single phase of a single transformer.

All this data should be readily available - if not from the facilities manager, then at least from purchasing or the utility. (N.B. Assuming that the facility costs are consistent with the separation between the transformers and the power levels being used, any ops manager worth his salt would consider firing a facilities manager who didn't have this sort of data on file - not that I'd recommend a student on placement to mention such a possibility)

You are right- I have partial information. That is why I asked the question. This is a reasonably small light assembly operation. The major loads are a compressor, lots of lighting, heating for the shop and HVAC for the office area. The total employment is about 120 people.

The lowest consumption recorded (in 30 minute increments across 18 months) is 75 kW on a weekend in May. My thought is that the HVAC is minimal, lights and compressor are turned off. Where is the 75 kW load? The transformers are very old 1500 KVA units and we are pulling between 200 KVA and 350 KVA, less than 10% of the capacity. So, I am trying to calculate the losses in the transformers to do a cost-benefit analysis.

If the transformers are 2% inefficient with a 60% PF, would that mean 36 kW are being consumed by the transformers? If so, that represents over $30,000 per year.

We are going there on a weekend to turn off every load and have the meter measure the transformer load. That should be the definitive answer. I will report back to this thread in early November.

LTU Student

2% loss would be 60-kW - but this should onnly occur at full load. Losses of more than 0.5% of full load (=15-kW) when the transformers are unloaded are extremely unusual - half this value or less is much more common (though at $7500/year, even that is worth saving provided the expenditure required to do it is not excessive). If wiring permits, you could save half of this by disconnecting one of the existing transformers.

BTW, age as such should not affect transformer loss unless the transformer winding insulation has been damaged (possibly by an extreme event).

By the way, at the levels you are running the additional losses in the transformers that are induced by load power factor will be negligible compared with other losses. Unless the utility company is charging a very high increment for poor power-factor (or is excentrically charging for current rather than power), you should ignore the IV in your calculations and work on actual load power (which will always be smaller than the IV).

Even if the load is all on a single transformer, the increase in losses at 350kVA would be quite small (about 1.5% of the 2% full-load loss, or 900W). In order to determine where your usage lies you should really be looking at the current in the transformer loads. In these energy-conscious days you may be able to call on the utility company to help out here - but you should in any case be able to borrow (at worst hire) a clip-on current meter that will give accurate enouh answers for your purposes.

When we turn off all of the devices in the shop, we will see what is happening. My observation is that they are consuming, at a minimum, 70 kW when the shop is in idle mode. Either we will find which loads are being left on, or we will find that the tranformers are consuming energy.

I will document in this thread on Nov. 9 what the results are.

Matthew

Very quick/short and possibly ill considered response...

a meter will measure the current. But P = VI cos p

If the current is all inductive.... then the P=VI *0

No Load current is all Magnetising (english so no Z) current and is inductive and may not be billed.

It really depends on your metering tarriff. No idea how people out of the UK charge.
Some suppliers charge for amps (KVA) some for KW.

www.11kv.com

It's simple. Put a boundary around the <...transformer...>. Then, at the boundary:

Energy out + Energy lost = Energy in.

or

Power out + Power lost = Power in.

Measure voltage, current and power factor on both sides. "Robert is your mother's brother." - Anon.