K-rated transformer question

Status
Not open for further replies.
J

JES2727

Senior Member
Location
NJ
I'll be installing a new transformer in the near future. I've tried to up-sell the customer a K factor rated trans rather than a standard one. I almost had the deal all sealed up, then they came back with this :

"In our data center, the UPS conditions the electrical supply, including harmonics. In fact, that is the UPS component we replaced several weeks ago (the "choke"or "core"). I'm not sure how that affects the transformer performance and type. If you can still make a case for the k-rated transformer, I'd say we get it."

So, does the UPS negate the need for a k rated trans?

Thank you.
 
I do not think I have ever worked on a project that actually required a K-rated transformer. The only advantage I see to them is the 200% neutral.
 
jim dungar said:
I do not think I have ever worked on a project that actually required a K-rated transformer. The only advantage I see to them is the 200% neutral.
Click to expand...

Same here actually. But I took a CEU class recently on transformers and left there thinking that a standard transformer is practically obsolete, considering core losses etc. So I see a data center, I think harmonics, and I think K-rated trans. But I'm no expert on transformers.
 
Modern data centers do not have extremely high harmonics. The need for a k-rated transformer should be evaluated based on the expected harmonic distortion in the load. Without knowing the distortion of the load, how is selling the customer a possibly unnecessary k-rated transformer an "upgrade"?
 
JES2727 said:
Same here actually. But I took a CEU class recently on transformers and left there thinking that a standard transformer is practically obsolete, considering core losses etc. So I see a data center, I think harmonics, and I think K-rated trans. But I'm no expert on transformers.
Click to expand...

And who taught this calss..Oh right the manufacture?
 
jghrist said:
Modern data centers do not have extremely high harmonics. The need for a k-rated transformer should be evaluated based on the expected harmonic distortion in the load. Without knowing the distortion of the load, how is selling the customer a possibly unnecessary k-rated transformer an "upgrade"?
Click to expand...

Done all the time! Not saying it is right, but a lot of BMW's are driven by salesmen upselling based on perceptions not needs.
 
JES2727 said:
So, does the UPS negate the need for a k rated trans?

Thank you.
Click to expand...

All transformers are K rated a standard transformer is a K-1. The transformer downstream from the UPS will be affected by LOAD harmonics if there is a REAL harmonic issue, no matter what the UPS is equipped with, unless there is NO LOAD on the UPS.
 
To answer the question of the UPS issue, there are some double conversion UPS designs that have what is called "active filtering" where they have more complex front-ends that produce very low harmonics on the incoming line. They are a lot more expensive because instead of using a passive diode bridge rectifier on the front end, they have transistors there as well, so it's almost like having two separate inverters. If they paid extra for one of those, a transformer with a high K-factor might be redundant.

On the other hand, it might also be that the UPS salesman sold them on a "premium" product that just has a DC link reactor. While this does reduce harmonics somewhat, it does not eliminate them. The only way to know for sure it to measure the line with a harmonic analyzer and see what you have with the UPS in full charge mode. If there are more harmonics than the UPS guy claims there will be, you can make a case.
 
JES2727 said:
I'll be installing a new transformer in the near future. I've tried to up-sell the customer a K factor rated trans rather than a standard one. I almost had the deal all sealed up, then they came back with this :

"In our data center, the UPS conditions the electrical supply, including harmonics. In fact, that is the UPS component we replaced several weeks ago (the "choke"or "core"). I'm not sure how that affects the transformer performance and type. If you can still make a case for the k-rated transformer, I'd say we get it."

So, does the UPS negate the need for a k rated trans?

Thank you.
Click to expand...
The UPS output is conditioned but that isn't particularly relevant.
The transformer in question probably feeds the UPS input and should be rated accordingly.
 
I thought for the past few years that modern servers in data centers have near unity power factor and almost no harmonics. Any manufacturered since about 1995.
APC white paper #26

So a data center needs just a K-1 (standard) transformer. And here's probably more than anyone wanted to know about the types of UPS. White Paper #1.

An office or school with a high concentraction of personal computers on a transformer should get a K-9 or 11 transformer. And this transformer does nothing to treat the harmonics. Just beefs up the insulation to survive higher heat levels from harmonics bouncing around inside.

If you want to treat the harmonics then a zig-zag transformer should be sold. Seems to also save about 2% of electricity thru the loaded transformer. But it's a cost adder. I think it's about double a standard K-1 transformer. Kinda funny how some companies' marketing information shows a quick payback for customers. Even claimed to earn LEED points. Was going to link it but can't find that company any more. Seems others picked up on it since Eaton also claims it will earn points in EAc1 for LEED. 2 MB file. Too bad there's no way to document this small energy savings in LEED's energy model.
 
brian john said:
THe UPS exist which leads one to believe it is on the UPS output?
Click to expand...
If the UPS exists and failed some weeks ago, see the OP, then that infers that it has been in service for at least some time.
You wouldn't generally choose a UPS with an output voltage that doesn't match the requirements of the load which is why I took the transformer to be on the input. But I don't know.
Perhaps Jes2727 will clarify this point.
 
Xfmr.

Xfmr.

What kind of UPS is it does it have an output Xfmr.? How old is the UPS does it use PWM with IGBT or SCR's? What K-Factor are you trying to sell them a K-4, 13, or 20? you can get a 200% neutral on a K-4 but how much harmonic content do you have / are you trying to fiter.
 
The UPS is existing. We are installing a new distribution panel which will be backed up by a generator. When complete, the UPS will be transferred to this new distribution, and served by the new transformer. I suggested that the customer consider a K-13 transformer. After reading all the input here, I realize that there is more to consider than just the fact that the transformer will serve a data center.
Thanks for all the advice.
 
JES2727 said:
After reading all the input here, I realize that there is more to consider than just the fact that the transformer will serve a data center.
Click to expand...
As far as the transformer rating is concerned, surely it isn't altogether relevant that it is a data centre installation?
Doesn't the data centre get its power from the UPS output?
The UPS output is effectively decoupled from its input.
 
K-rated Transformers Vs. Harmonic Mitigating transformers

K-rated Transformers Vs. Harmonic Mitigating transformers

K-rated transformers are just oversized transformers with larger nuetrals. The harmonics are not really eliminated just "trapped" in the nuetral. To truly negate the harmonics use a harmonic mitigating transformer which cancels out the harmonics. No need to increase the size of your transformer just get the correct size harmonic mitigating transformer. We used them at a casino which had large computer loads (slot machines) and harmonics were of great concern. These transformers also are more energy efficient as they reduce energy losses that harmonics cause. We used PQI, Inc. transformers. Just my two cents worth.
 
Status
Not open for further replies.
Top