use of delta-wye distribution transformer in step up mode 4

[akm]

Would like to get some operational tips, if I need to connect a delta-wye-n(480/208) transformer in step up mode. connect to supply source of three phase line(208) from utility source and run a cyclic load of compressors at 480 volts fro the delta 480 v side.

Do I need to isolate the netral on 208 side floating, if so why? and is it a permanent feature I can have without inviting any additional problem.

Do I need to provide additional protection device on LV/HV side of the transformer.
Any additional precautions during installation.
Thanks

 

[busbar]

There are 208 delta-480Y/277V units made specifically for this application. An example is listed at:

http://www.electro-mechanical.com/7700-94.pdf

page 26.

Some earlier discussion can be found at thread238-11717 and thread238-9287

99NEC 450-3(b) dictates secondary overcurrent protection based on transformer size.

 

[akm]

Would like to get some concrete answer to the doubts. I have gone thru the said reference. Am constrained to use a transformer which is 225 kva, 480/208, DyN, in a step up mode to get 480 volt connection fot this load.
I repeat my specific doubt>> Do I hv top float the neutral in the supply 208 volt side wye. If yes, for waht specific purpose and Do I invite any additional problem by doing so.

Another corollary to above, If I have a control trafo on the load circuit, which will ultimately make the drawn load current a little unbalanve, does it create any problem.
Shall aprreciate a path forward.
Thanks

 

[230842]

It's strongly recommended to connect transform wye neutral to neutral source, if it does exist, and neutrals to ground, as well, due to protection reasons (fault detection). These connections will not affect delta operating conditios at all. Julian

 

[gordonl]

On the low voltage side your going to have to use a ground fault detection system. The 480V system is going to be ungrounded because of the wye system. It is also recommended that tou use 1000V cable for the 480V system incase you have overvoltages due to the ungrounded system.

There is no effect on the low voltage side, and the unbalanced load shouldn't be a problem as long as it isn't large enough to cause unbalanced voltage. (It will obviously have to be a phase to phase load since your 480V system has no neutral.

Gord

 

[busbar]

There’s the usual caution of triplen harmonics on the wye-side 208V bus causing potential overheating in the XO-neutral connection and currents circulating in the delta winding. The transformer becomes a source/sink for zero-sequence currents. Besides switchmode power supplies—rectifier loads, resistance welding and discharge lighting can be harmonic sources. With the XO connection floating, voltage balance on delta-served loads becomes harder to maintain and losses increase. Ungrounded-distribution warnings have frequently been discussed over the last 50 years. 480V high-resistance grounding should be implemented.

 

[akm]

Busbar's word of caution is appreciated. Please advise me, if I have to go for a 208(wye)/480(delta) step up transformer, 225 kva for two dedicated compressors load of 30+30 hp,why is the compressors tripping on huge undervoltage. The step up transformer output voltage is unbalanced and goes down to 400 volts.

 

[peterb]

Does the undervoltage occur when one of the compressors is started? What is the transfomer impedance?

 

[akm]

both the compressors are started simultaneously, impedance is 4.5%, We have run the unit in this condition for 5 months until recently, it started tripping on overload.

 

[busbar]

400/460=0.869 p.u. doesn’t seem excessive.

How far apart are the calculated/actual starting characteristics? Are the UV trips with one unit starting, one running/one starting, or both units starting? Over what time interval is the ‘400 volts’ measured? Has the 27 device dropout/time been verified accurate? Are there warrantee/reduced-life issues if the 27 settings are changed? Are the 208V-bus sags unacceptable?

 

[akm]

1)The trips are on overcurrent, due to both unit starting, since this is the only programmed mode/manual mode of starting of the freeze 7 thaw unit, we have been told by the manufacturer.
2) The measured undervoltage is immediate upon strating the unit.
3) On supply side, the voltage dip is down to 195volt.

Not to mix up various issues, I faced a situation as follows, which could not give me any answer.

Based on transformer supplier advise and some published article, I kept the step up transformer(208wye/480delta),225 kva Transformer in f;loating neutral mode and measured the open circuit voltage on HV DELTA.two of the phase to phase voltages are 470~480 volt, while third one is 27volt.(twtenty seven only)I am not able to get any clue out of it. My questiuon is does that answers/confirms that we have a fault in the supply 208 side. If yes, we have other building load in the supply circuit connected and remains unaffected, or there is a hidden message here. Request for sincere advise and help.

Thanks

 

[gordonl]

I would suggest measuring 208V and 480V windings phase to phase and phase to ground, and neutral to ground on 208V.

I'm puzzled with 27 volts phase to phase, with others still 480V. I would suggest closely checking transformer connections on both sides, including winding Wye and Delta connections.

Did the transformer run 5 months open neutral, or is that a change you recently made?

As to supply problems you'll quickly know with suggested measurements.

Gord

 

[stevenal]

Due to the reasons cited by Busbar, I suggest you float the neutral. Unbalanced voltages on the wye side can cause neutral current if not floated. This in turn causes circulating current on the delta side. Overall capacity is therefore reduced. Please note that circulating current is not detectable at the terminals, and will not open fuses or breakers. Even if voltage is balanced, triplen harmonics can cause the same problem. If your compressors do not have variable frequency drives, triplen harmonics are probably not a concern. Another problem is the possibility of an open phase on the wye side. If neutral is connected, two windings will carry the full three phase load until transformer failure occurs or source side protection operates. If floated, delta side loads will be single phased with 50% voltage on two legs. Single phase and under-voltage protection should then remove the loads. Even if the load is not removed, transformer will not be overloaded.

 

[akm]

We have run this unit with the said step up transformer for 5 months, with a few trippings on overload, remained stray events for us, untill recently the units tripped immediately on start. We contacted the trafo manufaturer and told them that the step down trafo is used as a step up, and we face this problem. The manufacturer suggested us to run the unit on floating neutral. (utility power 208 volt). After isolating the neutral connection, I started measuring the open circuit voltage on delta, 480 volt side. phase to phase voltages are 476, 453 and to my surprise 27 volt in the third phase(in spite of repetaed measureme3nt with changing voltmeters also). This measurement did not permit me to go ahead of putting the transformer in floating neutral mode and connect load. The trafo manufacturer was consulted, who tried to convince me that there is a problem in the upstream end. I have rest of the building loads connected to the upstream end and do not see any abnpormal in the 208 three phase system.
Gord, I will request you to share your views on this, while others are most welcome to provide a valuable solution.

 

[busbar]

Do the 470/480/27-volt readings occur with no/steady-state load or when a start is attempted? When that imbalance is measured, it may be worthwhile to also measure the XO-to-system neutral voltage on the 208V side. The magnitude of neutral shift could indicate the direction of the problem with respect to the lowside transformer terminals. Depending on exactly where the various ø-ø voltages are measured, readings could indicate a high impedance (e.g., failing cable termination.) The exact location of the measurements would be important; i.e., are readings essentially the same at the starter load terminals (or in the motor junction box) versus those at the H-terminal bus stubs of the transformer? As I see it, in a healthy transformer, a ø-ø ∆-side collapse should correspond to a ø-n Y-side collapse; a high degree of neutral shift in the lowside windings. Would other readers comment on this?

This may have little bearing on the problem at hand, but it’s hard to know if, in the past, the overcurrent was marginally close to the trip value and has recently increased only a small degree, or if it was always ‘normal’ and something changed in the system to cause a much-larger increase in the starting current. Does it seem like one of these conditions may be more prevalent than the other? The severely imbalanced voltages reported would tend to point to the second condition. [With motors, something like a 3½% voltage imbalance can equate to a 25% current imbalance.]

 

[akm]

Stevenal and busbar, per your advise, I did the floating neutral situation as explained above and ended up at 476, 453 and 27 volt at delta 480 v side of trafo. Waiting for your any other investigational tips and solution.

 

[akm]

I measured the open circuit voltage at 480 side in floating neutral mode (as thought to be a viable solution), measurements are 480/470/27. (load at 480 completely isolated-please note).
X0-system neutral measurement, possibly high shift could not be measured today out of panic results, will do that tomorrow morning under floating neutral condition and let you know.
I have ruled out measurement at starter load terminal and motor terminal box in view of non-satisfactory open circuit voltage at transformer 480 terminal switch fuse box (2 feet away from the 480v end).
I have no reason, but now compelled to doubt the supply end cable connection raceway (this is a 6 feet cable raceway, spliced from our existing 208 volt building supply and connected to the 208 side of the transformer through a switch fuse box.
I performed all the measurements at both side of transformer opening the Switch fuse unit box.
I am convinced about your doubt(2nd one), something changed in the system, remained unnoticed till this frequent tripping.But would like to exactly smoke out the fault with all your advise, and go for a corrective action plan to run the unit. The detection will probably become easy, once I measure the neutral shift voltage magnitude and inform you.

 

[electricpete]

I think the logical next steps have already been suggested by gordonl and busbar:
- phase-to-phase and phase to ground measurements on hi and low-side
- X0-to-ground measurements.

Beyond that I have nothing practical to add. At some point down the road you might consider a transformer test (winding resistance, ratio). Also at some point down the road might want to check for resistance to ground on hi-side and low-side buses while deenergized. But more likely there is some problem on low-side... corner grounded? not a true three-phase system? . I'd do those voltage measurements before chasing the transformer etc

 

[akm]

Electricpete: Your wrap up may be correct,as I doubt we may have a fault in the 208 side connection of the transformer, raceway and associated cable connection. This does not answer the single question that we have: how could this faulty one pump power through the step up transformer to a 70~ hp load for 5 months in normal situation(i.e. neutral solid grounded on 208 side).I am only able to see it at floating neutral open circuit voltage measurement.

 

[busbar]

It almost sounds like there is a lost-phase/open-fuse condition on the 208V feeder serving the transformer. [Somehow I made the assumption that the neutral-to-X0 conductor was not in place.] The reversed ∆-Y 225kVA transformer could then mimic an open-wye/open-delta bank, and could conceivably be serving the 70hp load, with a derated but adequate capacity of (225·0.577) ~130kVA.

I’m sure readers will be interested in your findings and resolution.