Power Xfmr 55/65 Deg C Rating 2

[sialac]

Historically, we have specified power xfmrs that are 55/65 Deg C rated (for U.S. use). My understanding is that this was an older designation used when the insulation was originally transistioned from the older 55 Deg C rating to the newer 65 Deg C rating. My questions are as follows:

1. It seems to me that there is no advantage to specifying a new xfmr as a 55 / 65 deg. C rating. My feeling is that I can simply specify the 65 deg. C rating at the kVA I require and I will end up with what I need. For example, if I needed to serve a 20MVA load, if I specify a 20MVA 65 deg C xfmr I will get the same xfmr as if I had specified a 16.7MVA / 20MVA, 55/65 deg. C xfmr. Does anyone agree or disagree?

2. Is there any economic or technical advantage or disadvantage to specifying the 55/65 rating vs. a 65 rating only?

3. When evaluating xfmr loading on a 55/65 degree rated xfmr per ANSI C57.91, can I simply base my evaluation on the 65 degree C rating using the kVA at 65 deg C?

 

[cuky2000]

A 55 degrees tansformer is more expensive than the 65 degrees units. According with the ANSI standard, the 55 degrees transformer has 12 % more capacity than the 65 degrees.

I will get back to you with more details.

 

[sialac]

Is it more expensive if I keep the top ratin the same? For example, If I buy a 65 deg C xfmr that is 20MVA, isn't it the sme thing as if I bought a 17.85 / 20 MVA, OA, 55/65 deg. C xfmr? Sorry, I made a mistake in my calculation of the 55 rating in my original post item 1, it should have been 17.85, not 16.7MVA.

 

[jghrist]

I don't see any advantage to specifying the 55/65°C rating. You will probably get the same transformer. You may pay more if you insist that tests be done for both temperatures. I prefer to just use the 65°C rating in the specifications. Usually, if 55/65°C ratings are desired, you would specify the full rating at 55°C. You would then get a higher capacity available at 65°C to be used as a safety margin (20/22.4 MVA). In this case, of course, you would pay more because you would be getting a higher rated transformer. There may be a difference in the amount of time that the fans would run for a given load profile; I'd have to think about this.

 

[GamezBeCJ]

sialac:

1) The way you are thinking is correct if you have the transformer with the same rating at the 65, you'll get pretty much the same thing at the same cost.

2) You're correct when you say that the paper has evolved with the years. The 55 rating (as far as I know) has been kept because it is useful in some "hot" zones, where instead of having the 30 avg 40 max temperature profile, you have 40 avg, 50 max. So it is easier to specify the transformer at the kVA you need and 55 rise and let the transformer manufacturer do the math.

3) When you are evaluating a transformer in regard to the loading, you use whatever temp rise applies to the case, the most conservative thing would be to consider the 65. Think of the 55/65 extra capacity as an extra cooling stage but without the fans. The cooling is achieved increasing the number of cooling fins in the radiators and the oil ducts in the windings.

I hope this helps. Regards

Carlos Gamez, P.E.
Industrial Consultant
Transformer Specialist

 

[mc5w]

Sialic,

You also have to consider a few other things. A Fine Print Note has been added to 2005 National Electrical Code to point out that sunlight can cause a 30 celsius rise in ambient temperature at the surface of the conductor particularly around roofs. For each 10 degree celsius increase in temperature you cut the life by half. The reason why the Itapu (is the spelling correct?) power plant has the generators in pairs that share a pair of transformers is that transformer life, efficiency, and inductive voltage drop are all much more reasonable if the transformer is normally loaded to only 50% of its pedal-to-the-metal rating.

If you take into account the sun beating on your transformer and heating it up you might actually want the 55/65 rating and specify extra cooling fans or an extra large radiator. You could even get as fancy as adding water spray to your cooling fans.

An alternative to extra cooling is that a tarpaulin made out of aluminized plastic and reinforced with nylon or aramid cords could be used to shade your transformer. Depends on how beastly your climate is.

Mike Cole,mc5w@earthlink.net

 

[sialac]

Thanks to all of you for your input. Regarding the 55 degree comments, my understanding is that the 65 degree insulation is an upgraded insulation. I thought ANSI rated a 55 deg C xfmr maximum continuous kVA based on the same 30 deg C avg, 40 deg C max peak 24h, rating as the 65 deg insulation.

If the xfmr is to be installed in a hotter ambient location, I don't see any advantage of specifying a 55 deg 20MVA xfmr instead of a 65 deg 20MVA xfmr and that specifying a 17.85 / 20 MVA 55/65 deg C xfmr would give me the same thing as a 65 deg C only rated 20MVA xfmr. (understanding, of course, that the 17.85 / 20MVA is a non-standard kVA rating - I am assuming for my argument that the 17.85/20 55/65 OA is a standard size xfmr and as readily manufactured as a 20MVA 55 or a 20MVA 65). Any thoughts?

 

[alehman]

I think the 30 deg average/40 max ambeint definition you state applies for both ratings. A 65C rated transformer will run hotter than a 55C transf in the same conditions. If the environment is hot, you might consider oversizing the tranf (or buy a 55/65 rated unit and operate at the 55 rating).

 

[tulum]

As an aside,

I deal with many large mining companies and on the majority of underground portable substations the rating required is 80^c/150^c over a 40^c ambient.

The reason being is that the ambient temp underground, 4000ft or so, is very high, 80^c ambient is not unusual.

This allows the above transformer in these conditions to run safely at 120^c (80^c amb. + 40^c rise typ). It is basically just a margin of overload safety.

Technically, you should calculate loading based on the expected ambient i.e. actual conditions.

Regards,
TULUM

 

[sialac]

Thank you tulum. My understanding, though, is that by purchasing a higher temperature rise transformer, you are simply buying a lesser weight transformer. Let me explain:

Per IEEE Std. 1276-1997, by increasing the nameplate temperature rise (initially from 55C to 65C and later for specialty xfmrs from 65C to higher temp rises), the weight and size of the xfmrs are reduced.

Specialty xfmrs are sometimes allowed to run at these hotter temperatures since they are not continuously loaded. Take the example of a mobile substation xfmr. It usually only in service for short periods of time. The mfg rates the rise as 75C (for example) and gets a few more kVA out of the xfmr. The sacrifice is that you have a shorter lifespan for the insulation. This is OK since the mobile sub is not used continuously. If the mobile sub was a 5MVA xfmr at 55C it would be heavier and larger than a 5MVA at 75C. The 55C xfmr might have a 30 year life span if it is run continuously in a substation and well over 30 years if it was used periodically in a mobile substation since it would not be in continuous service. The 75C xfmr may only last 30 years in a mobile sub and less if it were continuously in service.

In order to get the nameplate kVA out of a xfmr, it can run in no hotter ambient than 30C 24h avg with a 40C peak. That means a 20MVA 55C xfmr will provide the 20MVA at 30C ambient continuously w/o loss of life. The same goes for an 80C rise, 20MVA xfmr. It too will provide the 20MVA at 30C ambient continuously w/o loss of life. If you are planning on running in an ambient greater than 30C 24h avg, and want no loss of life, you need to increase the MVA rating of the xfmr to compensate for the additional thermal loading or accept a loss of insulation life.

I suspect that the mining xfmrs were sized with a higher temp rise in order to decrease the weight and size of the xfmr. In order to compensate for the higher ambients, however, the MVA rating must have been increased (above the MVA if it were applied to the same load in a 30C 24h avg temp.) in order to allow the xfmr to run at the higher ambient for continuous loading or the expectation must be that the xfmrs are not loaded continuously. This would allow the xfmr to have an acceptable total life span.

I might be totally wrong on all of this and I gladly accept any comment or criticism.

Thanks,
Sialac

 

[tulum]

Sialac,

You are right about the weight of the transformer being different. However, the mining transformers are infact purchased as 80^c/150^c instead of the old standard 150^c rise. Which is a lower temperature rise, and makes the transformer actually heavier than what was previously used (I was not comparing them to 55^c rated transformers).

The intent in the mining industry is not to get more KVA, it is to allow for the transformer to continuosly run hotter with no loss of life (hotter being the 80^c level design in comparison to the 150^c).

For example,

In the Past a 150^c rise,1000kva xfmr was used for X load. Now a 80^c/150^c rise, 1000kva/1333kva xfmr is used for the same X load. The deeper you go, the more stringent this rule is.

So in fact we are increasing the KVA... we used to buy a 150^c rise at 1000kva and know we are getting a 1333kva at 150^c rise to push the same load. And as long as we don't exceed the 150^c standard rating for mine duty transformers we are not decreasing the transformers life span.

I don't know if this helps or not... I find this a confusing topic... even for some more experienced engineers.

Regards,
TULUM

 

[tulum]

P.S.

I think mine duty transformers would fit under 8.3.1 of ANSI C57.12.52-1981.


"When specified, transformers designed for an 80C average winding temperature rise, as measured by resistance, shall be provided with a 150C-rise insulation system (220C limiting temperature), as defined in 5.11.3 of ANSI/IEEE C57.12.01-1979."

Meaning the life of the insulation will not be damaged unless we exceed the 220^c mark. This is why we alarm on 200^c and trip on 220^c.

Regards,
TULUM

 

[GamezBeCJ]

One additional comment:

Just to clarify. I think tulum is talking about dry transformer (please, correct me if I'm wrong) and my previous post was refering to liquid filled units. Remember that dry and liquid filled are very different products.
Regards.

Carlos Gamez, P.E.
Industrial Consultant
Transformer Specialist

 

[tulum]

Yes dry type.

Regards,
TULUM

 

[Dodgee]

Hi,

My understanding of 55/65 is that if you were going to put the TX in an enclosed area, such as a vault, you might request 55/65. Therefore, if it is placed in the vault your rated load would be the nameplate kVA. If you placed it outside per say, would could get a few more kVA out of it, about 10%. 1% per 1 degree.

 

[Dodgee]

Hi,

In addition, you are right that it is an older designation. Now a days with the improvements 55/65 no longer really seems to have a role. 65 should do just fine.

 

[sialac]

I received the following response on my original three questions (beginning of this thread)from a transformer vendor and thought I'd share them:

1. The 55/65 designation has been removed from ANSI Standards, but is still available from manufacturers.

Note: Reference IEEE Standard C57.12.00-2000, Section 5.11 Temperature rise and loading conditions, pages 22-23:

"5.11.1.1 Winding temperature rises - The average winding temperature rise above ambient temperature shall not exceed 65C at rated kVA when tested in accordance with C57.12.90-1999 using the particular combination of connections and taps that give the highest average winding temperature rise. This will generally involve those connections and taps resulting in the highest losses."

"5.11.1.4 Liquid temperature rise - The temperature rise of the insulating liquid shall not exceed 65C when measured near the top of the main tank."

I agree if you are speaking of an existing 17.85 / 20 MVA, ONAN, 55/65 deg. C transformer, than you could simply rerate the transformer for operation @ 65C rise; Losses would have to be upgraded, impedance, regulation, efficiencies would also require correction to 65C rise values.

NOTE: ANSI C57.12.90-1999 IEEE Standard Test requires correction of load losses to 20C When load losses are reported at 55C rise, they are corrected to 75C; for 65C rise, they are corrected to 85C. Reference IEEE Standard C57.12.00-2000 Loss reference temperatures:
"The standard reference temperature for the load losses of power and distribution transformers shall be 85C.
The standard reference temperature for the no-load losses of power and distribution transformers shall be 20C."

I disagree if we were to re-optimize the design of a 65 C rise 20 MVA unit versus a 17.85/20 MVA 55/65C rise. Impedance would need to be stated @ 20 MVA, regulation and losses reported at 20 MVA, and this might result in a different optimization for a new transformer; they probably would not be identical, but similar.

2. Commercial - none that I am aware of; technically - compliance with current standards when specifying a 65C rise.

3. If you correct the losses - yes.