Is 480v more "efficient" than 208v?

[hmspe]

Correct me if I am wrong on this, but a kwh is a kwh. At the higher voltage for the same load, would you not use less kwh, therefore have a less expensive electric bill for any given load? I recently came across this exact situation on a job and gave the owner the choice. He went with the 277/480.

How do you get the KWh being lower? If you have a 10KW heater it draws 10 KW regardless of the voltage, right? This assumes the heater is rated at the voltage it is installed on. Run it for an hour and you have 10KWh Take a 5 horsepower motor. At 208V the Table 430-150 (the 1999 NEC is what's handy) FLA is 16.7A. 16.7A X 120V X 3 phase = 6012VA. At 480 the FLA is 7.6. 7.6A X 277V X 3 phase = 6316 VA. There's a lot of rounding error in these calcs, but for practical purposes the VA (and KVA) is the same for a 5 HP motor regardless of the voltage it's wound for. When the volts go up the amps come down for equivalent loads.

There are far too many unknowns to make absolute statements, but for equvalent equipment sizes (same wattage on lights, same tonnage on AC, etc.) the KWh drawn will be the same regardless of voltage. If the rate the utility charges per KWh is the same at all voltages the bill won't change because the power usage won't change.

There are areas in a system that might make a difference. Because the current will be higher at 120/208V the heat losses in the wires will be higher. On the other hand, when you add a step-down transformer to a 277/480V system you have losses in the transformer that would not exist on a 120/208V system.

I tend to stay with 120/208 whenever I can. In my opinion the total installed costs are about the same [larger wire and conduit, and higher ampacity on the service at 120/208; added transformers and a second set of (more expensive) panels at 277/480]. For a properly sized and installed system the overall I2R losses will be about the same. You don't pay for space or cooling for stepdown transformers. 120/208 is much safer for the janitor who gets told to swap out a fuse.

All that said, the building described would be a good candidate for 277/480V because of the lengths of the runs. The issue will be whether there is room for transformers and panels. The only way we could make things work on one elementary school I did was to put the transformers on the roof. Not ideal in Arizona.

Martin

P.S. Hi, Don.

 

[iwire]

The only way we could make things work on one elementary school I did was to put the transformers on the roof.

Good way to go if you have to, I have seen 13.8 primaries brought to the roof so the entire service gear could be located up out of the rentable space.

Was there no room to hang them from the ceiling near the the panels they serve?

 

[fido]

the KWh drawn will be the same regardless of voltage. If the rate the utility charges per KWh is the same at all voltages the bill won't change because the power usage won't change.

Thank you for taking the time to explain this. I knew there was something plainly obvious I was missing with my thinking on this.
Paul

 

[donw]

Thanks everyone for your help. There are some "bump-outs" (unairconditioned structures added on to the exterior walls) in which the current design houses the SES and panels. It would be possible to add a couple of transformers into two of these. I wonder if this would be alright in AZ, if I specified 80 deg rise xformers?

 

[ramsy]

The school has an odd (horse shoe) shape - ..The bottom of it is 170' long; one side is 300' long and the other is 244' long. ..service entrance is on the long side ..a 600A distribution panel on the other wing.

I see one 208/120 xfmr vault burried dead center on horse-shoe lot, feeding both wing panels within ~85 ft.

If the single 208 xfmr gets its 480 supply, and sends its 208 feeds overhead, to the wing distro's, Tbl. 310-17 permits smaller conductors. An additional savings in equip. capital costs. The other 480/277 loads pass directly thru the service entrance wing.

 

[iwire]

If the single 208 xfmr gets its 480 supply, and sends its 208 feeds overhead, to the wing distro's, Tbl. 310-17 permits smaller conductors.

What wiring method could you use at a school that would be under 310.17?

 

[ramsy]

What wiring method could you use at a school that would be under 310.17?

Interesting point.

If Tbl. 310.17 prohibits messenger supported single conductors, and the utility or lineman sub-contractor can't do it, then NEC Tbl. 310.20 (messenger cable) doesn't save us much in material cost.

In that case, I'd compare overhead & underground costs to route 1 vault, central-stepdown xfmr, against 1 pole-mounted-stepdown xfmr, per wing.

Maybe pole-mounted step-downs are already nearby, and seperate svc.drops for hi & lo voltage could be arrange in advance for the school.

 

[donw]

Okay, I did some analysis. I do believe that the I2R savings due to lighting will be a wash with the Z losses and energizing of the transformer primaries, so I concentrated on the A/Cs. I took one of my worst-case A/Cs. It is a 5-ton and is currently fed from a 200A panel. The panel feeder is 136' #3/0 and the branch is 84' #8. I assume a replacement 480V panel will be 100A, #1 with #10 branch. Using Table 9 Z at .85 PF, I get Ztotal (feeder+branch) of 0.07 ohms for the present 208V unit and 0.114 ohms for the new 480V unit. Currents respectively are 27.1A and 11.7A.
((27.1)**2)x(0.07)=51.4W per phase
((11.74)**2)x(0.114)=15.7W per phase
relative percent I2R inefficiency is: (51.4-15.7)/3252W = 1.09%
3252W is the phase wattage of the motor
Owner says monthly electric bills in other similar facilities is are around $5K. Air conditioning costs about 40% of total energy cost in hot/dry climates, so savings by going 480V would be about $22 per month average. Now, this is with one of my worse runs. What do you think?

bob, as for your lighting analysis, you assumed the 15A load per circuit at the end of the load. Obviously, this is distributed, so I tried estimating my 32 some odd lighting circuits with an average of 45 ft. each with the 15A load at the end. Using your numbers for #12 wire and $.10 per KWh, I came up with around $150 for the 9 month period. Now, this does not include the feeders...

 

[Jraef]

The thermostat will be set to whatever setting is required to make the setter comfortable. The higher the voltage reaching the unit, the less time (smaller duty cycle) the unit will have to run to maintain temperature.

When we're talking about incandescent lighting and the such, the savings can be realized by being able to use a lower-wattage lamp to generate a given amount of illumination. I'm not sure about discharge lighting.

Nope. Watts is watts is watts is watts. Voltage is irrelevant.

The ONLY issues are I^2R losses and material costs (and with the cost of copper, that is an issue now).

 

[benaround]

If that school was in Buckeye, Please excuse me for the names I used in vane.

 

[LarryFine]

Nope. Watts is watts is watts is watts. Voltage is irrelevant.

The ONLY issues are I^2R losses and material costs (and with the cost of copper, that is an issue now).

I really was talking about losses in a 'roundabout way. Let's say we're talking about resistive heating. A person will set the thermostat to feel a certain comfort level. In an all-other-things-being-equal world, a given heat gain will require a given KW-per-hour amount of power.

Now, that's KW/H at the heater terminals. If the voltage is below nominal level, the instantaneous heat gain will be below design specs, and the heater will run with a larger duty cycle, which is where the 'hour' part of KW/H comes into play. Of course, we pay by the KW/H, right?

However, we also pay for the power that heats up the wires between here and there, which is a larger part of the total power purchased. Even if we size the wire for the same percent of loss, it's still a greater total quantity of power, because a lower-voltage system has more current.

So, while the math says they're equal, it seems to me that over time, the system with the lower current will run for a lower numbers of hours per month to maintain a given heat gain, or comfort level. Likewise, with lighting, the benefit of lower voltage drop is seen over time.

Disclaimer time: all of the above is not based directly on anything I've read. It really based on how my brain sees it. When someone asks whether higher voltage is "more efficient", I have to answer that it is, because of how I extend the instantaneous theory over time in my mind.

Scary, ain't it?


Now, as for this specific situation, the only way to be sure is to price the job both ways.

 

[rattus]

Cost analysis:

Cost analysis:

I tend to think that the higher voltage would be more efficient overall, but this question can only be answered by a careful cost analysis. Seat of the pants judgements won't get it. My gut feel is to keep it simple. There are too many ramifications to make a snap judgement. If I were the customer, I would not want extra transformers taking up space and generating heat not to mention the possibility of a failure.

Now I will be the first to admit that I am not the one to make that cost analysis.

 

[haskindm]

As electricians, we are trained to think in amps. We size conductors and overcurrent devices in amps. However, we pay for electricity in Watts or KW. As others have said, Watts is Watts and remain fairly constant regardless of voltage. There are some advantages from using higher voltage due to there being less voltage drop. There are also financial advantages, at installation, for using higher voltages as the wiring can be smaller because the amperage loads are smaller. If 277-volt lighting is used, more light fixtures may be installed on a 20-amp lighting circuit than if 120-volt fixures were used, which also reduces installation costs. Much of this savings may be offset by the cost of the transformers and distribution equipment that will need to be installed to provide power for things like 120-volt convenience outlets (unless you are in an area where the power company will provide both 480 and 208 volt power to the building, in which case you will still have two distribution systems). You will need to look at the big picture. Will the HVAC equipment be more or less expensive if purchased for 480-volt rather than 208-volt? Are there other motor loads for which there may be an equipment cost difference between the two voltages. The long term cost diferences between the two systems will be small, remember we pay for watts not amps. Does the utility charge more or less for providing 208 volt service versus 480 volt.
Many factors will need to be taken into consideration.

 

[Bob NH]

"I really was talking about losses in a 'roundabout way. Let's say we're talking about resistive heating. A person will set the thermostat to feel a certain comfort level. In an all-other-things-being-equal world, a given heat gain will require a given KW-per-hour amount of power.

Now, that's KW/H at the heater terminals. If the voltage is below nominal level, the instantaneous heat gain will be below design specs, and the heater will run with a larger duty cycle, which is where the 'hour' part of KW/H comes into play. Of course, we pay by the KW/H, right?"

Unless the wires are where they don't contribute to heating the building, it doesn't matter if the power is dissipated in the wires or the resistance heaters.

Now if you are talking about A/C, every kWHr that is dissipated in the wires takes another 0.25 kWHr to pump it out of the building. That is why low energy lights, computers, computer monitors, and other equipment are so important in places where A/C is a significant part of the power consumption. That is also why transformers should be located outside the air conditioned area.

 

[hmspe]

Okay, I did some analysis....

You're way overthinking this. There are far too many assumptions for the calculations to be meaningful. About the only thing we can say is that the losses are in the same ballpark. The only valid calc would be a complete calc on a finished 120/208 design vs. a complete calc on a 277/480 design.

If energy use is all that matters you could probably make a case that you could save maybe 1% by going to 277/480 because of reduced I2R losses. If total cost of ownership over 10 years is the standard you'd probably be 1% to 2% better off with 120/208. My experience is that the installed cost is higher for 277/480.

Just my opinion -- YMMV.

Martin

 

[hmspe]

Thanks everyone for your help. There are some "bump-outs" (unairconditioned structures added on to the exterior walls) in which the current design houses the SES and panels. It would be possible to add a couple of transformers into two of these. I wonder if this would be alright in AZ, if I specified 80 deg rise xformers?

This is done all the time, but don't kid yourself that just because you spec a low temperature rise that they won't put in a 150 deg rise. they'll call it "value engineering", which is actually neither. It's first cost reduction, and the owenr will pay for that reduction many times over You probably should have an exhaust fan in each electrical room (and louvers for intake air).

Martin

 

[hmspe]

If that school was in Buckeye, Please excuse me for the names I used in vane.

Consider yourself excused....

If I had had any options on that project I would have done otherwise, but there was no space inside for transformers except to take storage space from the janitors, and toilet paper storage won the space wars.

Martin

 

[hmspe]

I really was talking about losses in a 'roundabout way. Let's say we're talking about resistive heating. A person will set the thermostat to feel a certain comfort level. In an all-other-things-being-equal world, a given heat gain will require a given KW-per-hour amount of power.

Now, that's KW/H at the heater terminals. If the voltage is below nominal level, the instantaneous heat gain will be below design specs, and the heater will run with a larger duty cycle, which is where the 'hour' part of KW/H comes into play. Of course, we pay by the KW/H, right?

However, we also pay for the power that heats up the wires between here and there, which is a larger part of the total power purchased. Even if we size the wire for the same percent of loss, it's still a greater total quantity of power, because a lower-voltage system has more current.

So, while the math says they're equal, it seems to me that over time, the system with the lower current will run for a lower numbers of hours per month to maintain a given heat gain, or comfort level. Likewise, with lighting, the benefit of lower voltage drop is seen over time.

Disclaimer time: all of the above is not based directly on anything I've read. It really based on how my brain sees it. When someone asks whether higher voltage is "more efficient", I have to answer that it is, because of how I extend the instantaneous theory over time in my mind.

Scary, ain't it?


Now, as for this specific situation, the only way to be sure is to price the job both ways.

Larry,

I think you're mixing up UNDERVOLTAGE and DIFFERENT DESIGN VOLTAGES. For undervoltage: Let's take a residential 4500W water heater. At 240V it will draw 18.75A. Froms Ohm's law the resistance of the heating element is 12.8 ohms. Run it on 208V and it will draw 16.25A. That makes it a 3380W water heater at 208V. At 208V it will take longer to heat the water, but the load is less, so the power used (KWh) should be the same either way. For different design voltages: If you compare a 10KW heater designed to run at 208 to a 10KW heater designed to run at 480 there will be no difference in how long they run -- KW is KW. That means KWh will be KWh.

As to savings on voltage drop, it tends to be very small on a properly designed system, so it takes a long time to see any appreciable cost savings. If you look at installed costs ($6 switch at 277, $3 switch at 120, $85 breaker at 277, $26 breaker at 120) the costs for the I2R losses can start looking very insignificant.

Martin

 

[kingpb]

Just for reference, motors nameplated for either 208 or 480 volts do not change in efficiency depending on voltage. Therefore the only savings would be in lower current to the motor. OK, so you run #12 AWG to a 5 Hp motor for either voltage, which means no conduit or cable savings. Say the run is 50 ft to the motor, the losses are 23W and 5.7W respectively. A savings of 17.3W for running on 480. A 75kVA 480-208Y/120V Xfmr will have approx. 143W of losses. So, if you have more then 9 motors ( 9 x 17.3W = 155W) for each 75kVA transformer your installation would pay for itself by going to 480V instead of using just 208V. Increased light fixtures on a circuits, plus lower overall heat, would reduce cooling some.

All in all, I think if you have very many motor/motor loads, or large square footage or lineal distribution footprint, the 480V would be better, especially if any of your air handlers is very big. 208V for large loads becomes non-standard and expensive.

 

[donw]

You're way overthinking this...

I just needed some energy cost estimate to give to the owner, because an electrician was telling her how much more energy would cost at 208v than 480v. Here's another monkey wrench. SRP said they could provide a second 480v service in addition to the current 208v transformer...for a fee. That would dispense with the extra 208v transformer/heat problems, but up front costs are likely high and monthly meter costs of another service would outweigh any energy savings. And then there's the cost of redesign...and some of the A/Cs have already been purchased. I think I have her convinced to keep the design as it is.