Smart questions
Smart answers
Smart people
INTELLIGENT WORK FORUMS
FOR ENGINEERING PROFESSIONALS

Member Login




Remember Me
Forgot Password?
Join Us!

Come Join Us!

Are you an
Engineering professional?
Join Eng-Tips now!
  • Talk With Other Members
  • Be Notified Of Responses
    To Your Posts
  • Keyword Search
  • One-Click Access To Your
    Favorite Forums
  • Automated Signatures
    On Your Posts
  • Best Of All, It's Free!

Join Eng-Tips
*Eng-Tips's functionality depends on members receiving e-mail. By joining you are opting in to receive e-mail.

Donate Today!

Do you enjoy these
technical forums?
Donate Today! Click Here

Posting Guidelines

Promoting, selling, recruiting, coursework and thesis posting is forbidden.
Jobs from Indeed

Link To This Forum!

Partner Button
Add Stickiness To Your Site By Linking To This Professionally Managed Technical Forum.
Just copy and paste the
code below into your site.

Helpful Member!  fms77 (Electrical) (OP)
11 Jun 04 15:23
I looking for a loading guide for distribution transformers.  At what point is a transformer overloaded?  At nameplate KVA, at 20% over nameplate?  How does the seasonal peak factor in?  Mfgs tend to recommend not loading beyond the nameplate KVA.  What about "degrees C rise"?  -- Thanks.
Helpful Member!  busbar (Electrical)
11 Jun 04 16:59

It Is not a trivial issue.  For ANSI oil, IEEE Std C57.91-1995 Guide for Loading Mineral-Oil-Immersed Transformers
  
advidana (Electrical)
19 Jun 04 21:11
If it is in the USA and it is not for a Power Utility then you cannot overloads it per NEC. For Power Utilities, check with them for what is acceptable.
DanDel (Electrical)
4 Jul 04 17:58
Actually, in some instances the NEC allows 250% overloading, while IEEE 242 recomends no more than 125%. Personally, I side with the manufacturers and their nameplate ratings and set overcurrent protection at 100% of rating.
raghun (Electrical)
29 Jul 04 1:09
The transformer and all its accessories are designed to take 150% load for not more than 30minutes.

It is important in case of parallel transformers that one transformer carries the overload when the other trips till such time the load shedding takes place to brings down the load. If the transformer over current protection is not set to take this need in to account, it is likely that the second transformer also trips and no one would like it.

This is a critical requirement in case of grid transformers and many a grid failures are as a result of conservative settings adopted for transformer over current protection.
DanDel (Electrical)
29 Jul 04 11:52
The problem with 150% overload for 30 minutes is that every O/C device available typically has a LT delay of not much more than 10 minutes, and normally this is set for much less for coordination purposes.
Even if you set the LT pickup at 150% so the trip will occur before 10 minutes, then the curve will allow overloads up to 150% (between 100% and 148% or so) for an indefinite time.
The only safe way to protect a transformer is by using the manufacturer's nameplate limits, or use temperature relays to trip out the secondary device when the temperature reaches the nameplate limits.
raghun (Electrical)
30 Jul 04 0:28
The O/C protection is not meant to protect the transformers against overloads. The temperature measuring devices are equipped to do that job very well and if overload protection is critical, the present day numerical over current protection relays come with an additional element for the purpose.

Trying to protect the transformer against overloads with O/C protection generally we get in to bigger problems and undesirable situations, this is my experience.
HCBFlash (Electrical)
4 Aug 04 11:03
Uh,  I'm no lineman, or utility power systems engineer, but I'd at least take a look at the NESC...
rbulsara (Electrical)
22 Aug 04 22:20
raghunath is absolutely correct for the type of application he is talking about...utility type systems or engineered systems not falling under NEC or similar codes.

However, if your installatin falls under NEC (in the United States speciallly), you have to follow NEC..

As for the definition, overload will be anything beyond its rating..it is a different matter how much overlaod is tolerable for how long.

Utilities can overload trasnformers for the reasons raghunath gave and also the fact they can afford to..some reduction in life due to repeated overloading is not a concern for them..it may not be the case in your case.

peebee (Electrical)
27 Aug 04 16:43
I am pretty sure that the NEC does not state transformer sizing requirements anywhere.  It does state feeder sizing requirements, and it does state maximum sizing of OC devices for transformers, but it doesn't mandate transformer sizes.

I would consider any load over nameplate rating to be an overload.  Whether or not the overload is acceptable is a different question.

For what it's worth -- I will generally size secondary transformer overcurrent protection at 125% of nameplate, which is in accordance with NEC.  With 100%-rated breakers, that would permit a 125% overload to exist continuously.  No code violation.

As mentioned above, utility requirements are different than us mere mortals.  But the utility philosophies can be interesting and enlightening none the less.  Around here, our utility wouldn't bat an eye at continuous 10 to 20% transformer overload.  And they would not lose much sleep over 40%.  

I've heard that it's more cost effective for utilities to replace transformers when they burn out from overload than it is for them to monitor the load and take care of overload situations.

Keep in mind that heat is the real problem that will kill the transformer -- so fans can be helpful in squeezing a bit more power out of a transformer beyond it's actual nameplate rating.  Similarly, if you have the transformer stuffed into a 120-degree electrical room, you may find that nameplate IS an overload condition. . . .
davidbeach (Electrical)
27 Aug 04 16:59
"I've heard that it's more cost effective for utilities to replace transformers when they burn out from overload than it is for them to monitor the load and take care of overload situations."

That works as an operating philosophy when you have a stock yard with spares.  For an owner with just a few transformers the economic picture would be just the opposite.
peebee (Electrical)
27 Aug 04 17:07
Re my statement "I am pretty sure that the NEC does not state transformer sizing requirements anywhere." -- by the way, I'd love for someone to prove me wrong.  I wish the NEC did mandate a minimum transformer size.  But, despite looking hard through the NEC for such a requirement, I've never found one.

If anyone thinks the NEC does specify transformer size, please provide a code reference.
davidbeach (Electrical)
27 Aug 04 17:34
I think that the closest NEC requirement for minimum transformer sizing is arrived at backwards through the the requirements for maximum transformer overcurrent protection.

There is also NEC 110.3(B) to consider, are the nameplate FLA values "instructions"?
cuky2000 (Electrical)
9 Sep 04 22:13
Transformers thermal loading beyond nameplate in the US and ANSI marketplace follow the IEEE/ANSI Standard C57. 91 & 92 as primarily guide while Europe and other countries follow the IEC standard 354 & 905 for the same purpose.

Statistics compiled In the pass 30 years by EPRI, IEEE and EEI shows not significant increase in transformer failure rates using the overloading criteria of 3% annual loss of life (ALOL) provided not exceeding the following constrains:

   -115 deg C maximum top oil temperature
   -160 deg C maximum hot spot temperature (55 deg C insulation)
   -180 deg C maximum hot spot temperature (65 deg C insulation)
   -LTC switch capability
   -Bushing capability
   -Insulation moisture content
   -Tank oil expansion limits

See the enclose information for aditional reference in this subject:
http://www.maac-rc.org/reference/PJM%20Ratings%20G...
http://www.usbr.gov/power/data/fist/fist1_5/vol1-5...
http://elect.mrt.ac.lk/Estimation_optimum_1998.pdf
http://www.dstar.org/Public/Transformer_loading_an...;


peebee (Electrical)
10 Sep 04 18:09
davidbeach -- no doubt -- I agree with you on that one.  But if we size the OC device at 100% of demand load, and the OC device is allowed to be 250% of transformer rating, then that means we could install a transformer sized to handle only 40% (inverse of 250%) of the load it is expected to carry and still be in full compliance with NEC.

That's kind of out of character for the NEC, where most everything else is sized at 100 or 125 or 250% of the load it carries.

And we could actually take that even further.  Only the feeders are required to be sized to handle 100% of their demand load.  There's no reason that the OC devices need to be sized as large as the feeder ampacity -- they, and their transformer, could be even further reduced in size.
davidbeach (Electrical)
10 Sep 04 19:20
peebee - I've never quite understood those rules either, but for any secondary under 600V and over 9A, not in a supervised location, the secondary must be protected at 125%, or the next standard size.  Maybe overloaded transformers fail safely, after all, the NEC only claims to provide safety, not practicality.  Practicality is why us consulting engineers exist.
NEMA6P (Electrical)
16 Sep 04 22:10
I think you guys covered this one pretty well.  For what it's worth:

1) I also size primary OC at 125% primary current at full nameplate load as a rule of thumb, whether or not I have OC protection on the secondary.  (Based on NEC table 450 -3)

2) Any loading above nameplate is definitely an overload condition.

3) Acceptability of any overload that doesn't pop the primary OC protection, in terms of severity or duration, is all determined by temp rise. (don't exceed the nameplate rise rating)

Regards,

NEMA6P

GOTWW (Industrial)
27 Sep 04 20:54
I am confused, I have not done much residential utility work, however I am faced presently with removing and re-locating some power poles.  A typical case is one transformer supplying three ~2000 sq ft residentials using a single-phase pole-mounted unit transformer. I assume that these are 100A services. So on the conservative side 80*3=240A, and at 240V is 57 Kva. Way-way to big for a typical utility to install. Now using peebees 40% gives me 22.5kva, still looks too high, considering 15kva appears to be what is currently installed. However, since it will be necessary to install new transformers to minimize the outages is there a diversity factor in play here that given peebee's 40% or 2.5*15kva = 60kva or 156A/3 = 52A expected coincident load?
peebee (Electrical)
28 Sep 04 11:28
I have never done any utility work, but I've worked with plenty of utility engineers and I can guarantee that they will use a much smaller demand factor than we would typically use on the NEC side of things.  I can't give you any good rules of thumb on a demand factor to use.  But I can give you the following tidbits which might help point you in the right direction:

+  Utilities seem to size their transformers at about 50% to 75% of what we would size them at (which is often based on connected load or 75-80% of connected load).

+  There are quite a few articles out there about the philosophy behind NEMA TP-1 transformers (Energy Star high-efficiency transformers).  A lot of research has been performed on transformer loading, and most of it seems to indicate that most small transformers are loaded to something like 30% of nameplate.

You might want to contact some utility engineers and see if they can give you some pointers on transformer sizing (just curious -- GOTWW, are you a utility engineer?  If not, how come you're sizing a utility transformer?).

Also, I'd forget about your 80*3=240A starting point for your calculations.  I'd think a better bet would be to get some good rule of thumb numbers for w/sf for residential and start there.  Unfortunately, I can't give you any guidance there, either, most of my work is in commercial/industrial.  But, my guess would be about 3 w/sf.  Assuming 3x 2500sf houses, that would be 15 kVA demand.  DON'T BASE YOUR DESIGN ON THIS.  But this does indicate that your calcs may be high.  And maybe someone else can throw out some better w/sf numbers.

Actually, after taking a look at NEC Table 220.3(A), I'm starting to think even 3w/sf might be high for residential for utility purposes. . . .
GOTWW (Industrial)
28 Sep 04 14:51
From the good-old westinghouse t&d book, "satisfactory overload protection of a distribution transformer cannot be obtained with a primary fuse, because of the difference in the shape of its current time curve od a distribution transformer......... because of this a primary fuse should be selected on the basis of providing short-circuit protection onlyand its minimum blowing current should usually exceed 200% of the full load current of its associated transformer". It goes on to say that transformer failures to overloads are unusual, with the majority of catistrophic failures are due to high impedance secondary faults, tree limbs etc.

This reminds me of this weekend, and the cascading transformer explosions in florida, probably due to high impedance secondary faults. And of course Mr Heroldo(?sp) explinating the virtues of UG electrical.
GOTWW (Industrial)
28 Sep 04 16:03
Concerning the three residential units, with an estimated NEC KVA of  10 to 12Kva, now extrapolate this to table 230.32 for 3-5 dwelling units gives a demand factor of 45%. So this gives a range of 3*(10 to 12)*0.45= 13.2 KVA to 16.2 KVA, at which nearest transformer would be 15 KVA, that by the way has 62.5 FLA. So this is my explanation.

It is funny though having to do a full blown NEC load study for a commercial building upgrade to justify keeping the 400A main service, whereas the utility company leaves an existing 10KVA unit of the pole, no questions asked.
 
And No I am not a utility engineer, I have to wear allot of hats with varying degrees of incompetence
Helpful Member!  jghrist (Electrical)
29 Sep 04 10:43
Utilities would use a variety of methods for sizing the transformers, based on their local experience.  You do need to take into account major loads such as electric heating and A/C.  As an example, an old guide from Public Service New Mexico gives 2.7 W/ft² for 1600-2000 ft² houses with gas heat and evaporative A/C, 6.9 W/ft² for houses with gas heat and refrigerant A/C, and 7.9 W/ft² for total electric houses.  They used a coincidence factor of 0.66 for 3 houses with gas heat and evaporative A/C, 0.88 for 3 houses with gas heat and refrigerant A/C, and 0.69 for 3 total electric houses.
 
GOTWW (Industrial)
29 Sep 04 13:35
Thanks jghrist, for the valuable insight into the thinking of the "evil" empire.
Bung (Electrical)
22 Feb 05 21:09
Following on from jghrist, utilities also often have a set of standard size transformers they use.  This gives economies through discounts for quantitites, allows standard padmount or pole-top designs, and reduces reuirements for spares.  So you can easily see a much larger than expected transformer in a given situation just because it is the nearest standard size, or was the only one available in time to meet the customer's required date for supply.

We also have to bear in mind the differences in "time constants" at play.  A transformer takes hours, at worst tens of minutes, to be damaged by overload.  An overcurrent relay (or fuse) takes fractions of seconds, at worst seconds, to operate.  No point in "throttling" the transformer for short term overloads (in the order fo seconds) just because that's how the overcurrent curve is set.  How would you deal with (for example) large air-con motor starting?

Bung
Life is non-linear...

rempman (Electrical)
3 Jun 05 12:59
Your question is a good one.  My response makes the assumption that your firm falls under the ANSI C2 - NESC instead of NFPA 70 - NEC.  In distribution utility applications within a residential setting you can plan to overload transformers up to 200% or more of nameplate.  This is depending of course on the coincidental loading of the services that are being served, the length of time the peak will be inplace, and the ambient air temperature.

However you may be overlooking a more important issue if you choose to follow this route.  I strongly recommend that you look at your voltage drop calculations to the service that is most at risk of being served at a voltage below your mandated requirements (Typically 114V at the meter)  You will also need to review your flicker criteria in the case of heat pumps and air conditioning.  

In the end I would wager you will be limited to around 120% of nameplate as a result of the voltage issues you may incur.

Reply To This Thread

Posting in the Eng-Tips forums is a member-only feature.

Click Here to join Eng-Tips and talk with other members!

Close Box

Join Eng-Tips® Today!

Join your peers on the Internet's largest technical engineering professional community.
It's easy to join and it's free.

Here's Why Members Love Eng-Tips Forums:

Register now while it's still free!

Already a member? Close this window and log in.

Join Us             Close