long feeders & voltage drops

Status
Not open for further replies.

BrianMuir

Member
Location
Comox BC
I would like to hear about your experiences sizing long feeders >100A where you faced the problem of having to significantly up size the conductor due to voltage drop issues.

It seems to me that the voltage drop calculation should be based on realistic maximum loading and not theoretical maximums, provided that there is no inherent safety risk associated with low line voltage and the connected loads are not sensitive to slightly below acceptable voltage.

I am asking these questions because I'm looking at a feeder out to a modular classroom, subpanel breaker = 150A.

I've bumped the conductor size up to #4/0AL, which is a large as we can fit into existing conduit. My voltage drop calculation indicates this meets 3% up to 106A. I am inclined to sign off on this because I feel the likelihood of the actual load reaching this level is very low and the consequences of load line voltage in this case are mild. It could be proved via measurement I suppose.

I'll run it by my local inspector but it would be nice to hear some of your thoughts

thanks!
 

wiigelec

Member
Location
Red Desert
Not sure what voltages you are operating at but I have been in similar situations with existing conduit and such where we ended up installing transformers at both ends to increase the "transmission voltage" (if you will) to avoid up-sizing the wire...
 

G._S._Ohm

Senior Member
Location
DC area
It seems to me that the voltage drop calculation should be based on realistic maximum loading and not theoretical maximums, provided that there is no inherent safety risk associated with low line voltage and the connected loads are not sensitive to slightly below acceptable voltage.
>there are two issues here: the likelihood of exceeding a limit and the consequences of exceeding that limit.
The first is an easy calculation based on historical data or published guidelines for this sort of thing.

It could be proved via measurement I suppose.
>At least 30 samples are needed to provide a 'statistically large sample' to be used in computing the standard deviation. With a smaller number of samples correction factors need to be applied.

You might want to find Web primers on standard deviation, normal curve, Z score, tolerance limits, small sample size, non-normal distributions, statistical fallacies and so on.
I can also recommend a forum where the members know way more about this than I do.

This subject gets very complex very quickly.
 

BrianMuir

Member
Location
Comox BC
voltage is 120/240
loads include a heat pump with auxiliary electric, base board heater, lights, receptacles, hot water tanks, and as a classroom it is only used during day (warmer hours), in a mild climate.

wiglec: I have another feeder on this project where I am looking at doing exactly what you suggest (up-down). What was your circuit, conductor, transformer sizes? It take it that this was cost effective?
 
Last edited:

iwire

Moderator
Staff member
Location
Massachusetts
It seems to me that the voltage drop calculation should be based on realistic maximum loading and not theoretical maximums, provided that there is no inherent safety risk associated with low line voltage and the connected loads are not sensitive to slightly below acceptable voltage.

Currently there is no NEC requirement to size for voltage drop at all. (Other than a couple specific applications) o if you choose to upsize for voltage drop it is entirely up to you how to do so.

You can base it on the size of the over current protection, or base it on the size of the connected load, or the calculated load, or any number you want.
 

wiigelec

Member
Location
Red Desert
What was your circuit, conductor, transformer sizes? It take it that this was cost effective?
I will be mostly (educated) guessing here on the design specifics as I was the wire puller/terminator on that job and it was a couple years ago. It was a feeder for a 200A 120/208V panel with existing I believe 1-1/4" EMT with max of one 3/0. Boosted the voltage to 480V and cut amps to <100A for max of 5 #2 in 1-1/4" EMT. Transformer sizes were probably 480Ax100A ~ 50kVA.

Weighed against the alternatives of installing new conduit etc. this was the best which is why it was done, I remember the boss hemandhawing over this for a while...
 

bob

Senior Member
Location
Alabama
voltage is 120/240
loads include a heat pump with auxiliary electric, base board heater, lights, receptacles, hot water tanks, and as a classroom

It seems to me that the voltage drop calculation should be based on realistic maximum loading and not theoretical maximums
Of the load shown above what would you say was the realistic max load? Theoretical?
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
110312-1435 EST

friendlyfire:

What is the real nominal source voltage from the 240 pair? Is this maybe 250? If it is 240 or lower is there any reason not to make it 250 to 260 with a pair of autotransformers?

At what is the minimum running voltage the heat pump will work reliably? Same has to be asked of its starting voltage and at this voltage what is its peak input current?

If your normal line voltage runs on the high side of mean, and there is not under normal conditions a large variation, then using a fixed % drop as criteria may not be the optimum approach.

If the lights are not incandescent, then you need to know their lowest safe working voltage.

The resistance heaters have no problem with low voltage unless heating time is a problem.

Using a step-up followed by a set-down transformer might introduce a 6% drop with no interconnecting wires. Then on top of this you have to add the line drop. Unless you substantially oversize the transformers and/or use more efficient transformers you may gain nothing with this approach.

.
 

iwire

Moderator
Staff member
Location
Massachusetts
What is the real nominal source voltage from the 240 pair? Is this maybe 250? If it is 240 or lower is there any reason not to make it 250 to 260 with a pair of autotransformers?

Maybe I have been wrong but generally I have considered boosting the voltage higher to compensate for voltage drop to be the hack way. It will only work for a constant load, as the load changes so would the voltage.


Using a step-up followed by a set-down transformer might introduce a 6% drop with no interconnecting wires.

I believe typical transformers are wound to cover their own losses.
 

G._S._Ohm

Senior Member
Location
DC area
To take a very simple example, the design lifetime of Romex is 50 years and for every 10 C above rated temperature the lifetime halves.
With this there seems to be no catastrophic, sharply-defined, failure point but the higher you go with conductor temperature the sooner you will see some kind of failure that is noticeable by instruments or by the user.

You can get lifetime vs. voltage vs. source resistance for each load from the manufacturers if they will tell you, but setting a threshold for conductor size will still be pretty murky.

A good analogy for a difficult decision problem is with timing chains on interference engines.
You want to leave the chain in place as long as possible to get your money's worth but if it fails your engine is pretty well ruined.
The manuf. says replace at 60,000 miles but they have a financial interest in selling you parts.
What would a reasonable person do?
 
Last edited:

hurk27

Senior Member
Maybe I have been wrong but generally I have considered boosting the voltage higher to compensate for voltage drop to be the hack way. It will only work for a constant load, as the load changes so would the voltage.

I agree

Using buck boost or re-tapping a transformer to over come a voltage drop on a circuit with mutable loads, where some loads will not always be on, can and will lead to over voltages of this equipment, the results will not be good.

I agree with Bob, you should never boost voltage on a circuit or feeders that will have a varying load.
 

Hv&Lv

Senior Member
Location
-
Occupation
Engineer/Technician
voltage is 120/240
loads include a heat pump with auxiliary electric, base board heater, lights, receptacles, hot water tanks, and as a classroom it is only used during day (warmer hours), in a mild climate.

wiglec: I have another feeder on this project where I am looking at doing exactly what you suggest (up-down). What was your circuit, conductor, transformer sizes? It take it that this was cost effective?

Are your calculations made with the voltage stated above, or actual voltage at the point of origination for the feeder? Some utilities run their voltage a little hot to keep from having to install $16,000 voltage regulators. We actually run our voltages around 125 now. The customer at the end of the line still has at least 119 on the peak loads. Check the actual delivered voltage and run the calculations with that number.
 

Cow

Senior Member
Location
Eastern Oregon
Occupation
Electrician
Do a load calc.

Then do a voltage drop calc based on that, not the breaker size. If you keep it under 3% VD for the load, great. If you have room in the conduit for even larger wire and want to upsize the conductors for 3% VD based on a load closer to the breaker size then do it. This is where experience kicks in.

Doing VD calcs based on breaker size can be pretty costly and unnecessary a lot of the time. Just my opinion of course....

Make sure you always keep in mind how you'll terminate these large wires too. They won't always fit the factory lugs...
 

tryinghard

Senior Member
Location
California
?It seems to me that the voltage drop calculation should be based on realistic maximum loading and not theoretical maximums, provided that there is no inherent safety risk associated with low line voltage and the connected loads are not sensitive to slightly below acceptable voltage.

I am asking these questions because I'm looking at a feeder out to a modular classroom, subpanel breaker = 150A.

I've bumped the conductor size up to #4/0AL, which is a large as we can fit into existing conduit. My voltage drop calculation indicates this meets 3% up to 106?
225.5 The ampacity of outdoor branch-circuit and feeder conductors shall be in accordance with 310.15 based on loads as determined under 220.10 and Part III of Article 220. Check out Annex D3 for an example to calc.

Looks like you?ve got around a 340? run if this is true and your calc load equals around 150A the voltage could drop around 4.25% meaning from 240 to 230, I personally don?t think this is a big deal but the inspector might.

Something to consider 4/0 cu at this distance = 2.6% VD; if the owner is adamant to respect 3% then a change order should be available unless something about the 3% is in the specs. :grin:
 

hurk27

Senior Member
Generally I don't worry to much when the supply is 240 volts, most motor loads today operate fine on 208, and motor loads are the most acceptable to voltage drop, now with the supply at 208 volts, it kind of a different story as your already running these motors at the bottom end of their ability, and I have seen many problems when you throw in voltage drop, generally most 120 volt equipment does fine anywhere between 100 volts to 130 volts, here in Indiana voltage drop is not enforceable by an inspector, job specs, customer specs can be a different ball game.
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
110312-1901 EST

iwire and hurk:

I believe you have misunderstood what I was trying to say.

In todays time I believe the nominal unloaded design goal for power system voltage at the destination is 120 V. This in turn means that equipment designed to operate on this system would have nominal voltage design criteria of 120 V, or maybe offset a little. If I design equipment for this my criteria is for operation from 95 to 135 V. Not all equipment will be designed for this wide an input range. However, in recent years some equipment has gone much wider.

At my home lightly loaded my nominal voltage is about 125 V, from a 25 KVA pole transformer. This usually does not drop below 123 or go above 127 V under the light load conditions. This voltage variation is from the pole transformer primary side.

One town away (of no significance that it is one town away) is an Energy Store, sells "green" stuff. Their nominal voltage is about 105 V and stiffer than mine.

Stiffer means the source impedance is lower. Therefore, for a given load change there is less voltage change from that load.

Since most equipment today is designed for a 120 V nominal unloaded source voltage, and this equipment can be expected to work just fine at 125 V, then if there is low nominal source voltage, there is no reason to not increase the source voltage to 120 to 125 V. You pick your criteria.

The second point I was trying to make is that if we assume transformers with a 3% internal impedance, and we cascade these in series to allow transmission at a high voltage, then a 6% impedance has been inserted in series with the circuit to the load even before considering the voltage drop of the connecting line between the transformers.

friendlyfire has indicated his calculation is 3% for 105 A on the feeder.

If his source is 105 V, then 3% produces 102 V at the load end. This might very well be marginal. But now add two transformers of 3% each at 150 A, and suppose the net no load voltage transfer is 1 to 1, then the combined transformer drop at 105 A is close to 4% and the line drop changes from 3% to 1.5%. So the total drop at 105 A becomes 5.5%. This is worse than the feeder wires without transformers. You can go to more than 2 times the voltage and slightly reduce the 1.5% component.

If instead we increase the unloaded source voltage from 105 to 120 V, then assume a 1% additional impedance for the boost transformer, then we have a 4% or less source at 120 V and with a load of 105 A and destination voltage is about 115 V. Much better than the 102 V. In fact we are better off with respect to the airconditioner because its load current is less at 115 V input than at 102 V.

If the destination is well designed, then it is unlikely that a 150 A panel will be loaded close to its 150 rating. I think the most critical factor in this system is -- what is the minimum voltage to the airconditioner under startup conditions?

.
 
Last edited:

BrianMuir

Member
Location
Comox BC
Hi, thanks for all the replies.
I haven't had time to digest everything, but did want to comment that I am working to satisfy CEC, as I am based in Canada. I know this forum is primarily relating to the NEC but much of the code is very similar and I find the level of expertise on this forum much better than any similar site focused on Canada, so I hope you will forgive my intrusion.

CEC Rule 8-102 states:

1) Voltage drop in an installation shall:
a) be based on the calculated demand load of the feeder or branch circuits
b) Not exceed 5% from the supply side of the consumer service to the point of utilization; and
c) Not exceed 3% in a feeder or branch circuit
2) For the purposes of sub rule 1 the demand load on a branch circuit shall be the connected load if known; otherwise it shall be 80% of the rating of the overload or over current device is protecting the branch circuit, whichever is smaller

So, here, this is not optional, but I expect my AHJ is flexibile to some degree.

For me, the heart of the problem seems to be realistically estimating the connected load. I know exactly what is connected but does "connected load" mean the load when every outlet is in use, all the lights are on, and the heat is running full bore? This would exceed 80% of the overcurrent rating, so my #4/0 would be insufficient. I prefer to use instead a number based on a reasonable engineering assessment based on the nature of the loads, actual vs nameplate draws, etc. (This is a new installation so I cannot provide historical measured data)

I will ask my local inspector for clarification and let you know what I learn.
Also, I will take some time and review all of your posts.
Thank you, FF
 

Dennis Alwon

Moderator
Staff member
Location
Chapel Hill, NC
Occupation
Retired Electrical Contractor
I didn't notice you were from Canada so 250.122(B) may not apply for you. :grin:
The Nec does not make us upsize for VD unless it for a fire pump. It is mostly a design issue.
 
Status
Not open for further replies.
Top