3 wire to 4 wire transitions 2

[DRSchmitt]

first let me say I am not an Electrical engineer (a mechanical) but I am in the Power Generation Biz so I would appreciate any thoughts on the following issue. My question begins with a small 6 MW generation plant which parallels to the grid. the design is simple: gensets are sync AC at 4160V floating neutrals connected to a bus and in turn to a step-up xfmr 4160V delta to 60KV grounded wye. Amoung all other standard Utility protection there is a Gen.Bus Ground Fault Detection Relay (59G) which is configured in a "broken delta" (I believe)to obviously detect any one of the phases going to ground etc... O.K. pretty straight forward right. Now there is added an on site power distribution system to be added to the end of the gen. bus by means of a load break switch (all 4160V). So three conductors are run the first pole to begin this circuit...years go by... finally a design and installation are complete for this on-site dist. line and it is a 4 wire system (3 phases & Neutral) the pole drops (9 in all)are single phase service with the xfmrs (high side)phase to Neutral...Oh and by the way the Neutral #4th wire is bonded to the shielding on the three conductors at pole #1 and in turn the grounding grid at the plant(this should have been the first sign of trouble by the contractor but...) and for that matter at ever pole ground also (remember no fourth wire was provided at pole #1.) so upon energization the 59G relay picked up and... yes you guessed it... we 86'd. So I would love to here some possible solutions to this problem, then I will tell you what was installed to "remedy" the situation. Any feedback would be great.

 

[jbartos]

Suggestion:
1. Normally, parallel sources (generators, transformers, etc.) with y-connection have the y-neutral grounded via the high-resistance grounded neutral bus (one source is connected to the grounded neutral bus at a time to assure the smallest capacitive charging current) for ground currents < 10A. If the 10A value is exceed, then medium resistance grounding may be considered among others.
2. The three-phase three-wire power distribution is not supposed to be grounded downstream, unless there is another conversion/transformation. Then, the next system grounding scheme is implemented. Usually, the power distribution downstream is connected to delta-connected transformers on primary and usually y-connected transformer secondary with either solid neutral grounding or high-resistance grounded neutral scheme again, depending on voltage and ground fault currents and other requirements.
References:
1. IEEE Std 142-1991 &quot;IEEE Recommended Practice for Grounding of Industrial and Commercial Power Systems&quot;
2. Beeman D., &quot;Industrial Power Systems Handbook,&quot; McGraw-Hill Book Co., 1955, Chapters 6 and 7.
The 59G overvoltage relay is normally connected to the secondary of a 1phase-distribution transformer connected in the generator neutral line to ground on its secondary side to monitor/trip faults. The 59G could also be replaced by 51G ac time overcurrent relay connected in the same location instead of 59G except it requires a Current Transformer (CT) inserted in the 1phase distribution transformer (it could be as high as 30kVA in size) secondary.
The power distribution described in your posting may be required to be modified to comply with the industry grounding standards to be functioning as it is supposed to.

 

[peterb]

Sounds like an unhappy marriage of two incompatible systems. The generator grounding that you describe is very straightforward, but is obviously not intended to have a 4-wire distribution system directly connected. Your 59G relay will be connected to the open corner delta secondary of a 3-phase set of VTs with the primaries connected line-ground - this measures the neutral displaceent (zero sequence) voltage under ground fault conditions. The 4-wire multi-grounded system is also in common use.
There are a few choices here -
1 - Solidly ground the generator neutrals and connect to the distribution neutral; this leaves the problem of how to supply the load from the utility source with the generators off - this would require a grounding transformer to establish the system neutral. This also exposes the generators to high ground fault levels, which can lead to extensive damage in case of an internal generator fault. Not preferred.
2 - Install a delta-wye, 4160-2400Y/4160V isolating transformer between the generators and the load, with the secondary neutral solidly grounded (Note that the multi-grounded 4-wire system is not compatible with resistance grounding). This is the far better solution technically. .
3- Convert the distribution system to 3-wire, by changing the service transformers to 4160V line-line primary
Keep us posted as to what your fix was.

 

[DRSchmitt]

yes Peterb, it was, and still may be a very unhappy marriage. Thanks for your thoughts. The saga continues. Your item #1 is not going to happen, since we have run for 8 years with no problems. And both your items #2 &amp; #3 were discussed as possible (and very logical)fixes. But when confronted with the problem the designer came up with something really amazing and still a little perplexing (to me anyway). A &quot;grounding bank&quot; set of transformers were installed almost at the end of the dist. line (only 11 poles and approx. 3000ft long). These xfmr's are wired wye on the three main conductor side or I guess primary 4160V and delta on the secondary which is not connected to anything just floating??? the center of the primary or neutral has become the forth wire of the 4 wire dist. system. And also the 4th wire neutral was disconnected from the conductor shielding and just grounded. Please tell me if this is going to give me any problems in the future? is my 59G still effective? what is happening in this &quot;grounding bank&quot; secondary? If one of the single phase load services develop enough current will the 59G see it as a fault or imbalance? I am concerned. It seems as though this was the most confusing solution of all. Please comment. Thanks.

 

[peterb]

What you have done is part of my item 1 - added what is effectively a grounding transformer (wye delta tranformer with the wye winding grounded). How effective this will be depends on several factors, including rating and impedance of the grounding transformer bank.
As far as future problems go, see my previous comments on increased ground fault current for generator internal faults. You will need to ensure that the single phase loads are well balanced so as to avoid unbalanced negative sequence currents on the generators, which can lead to rotor heating and may trip the units, depending on the protection provided.
The 59G relay is a neutral voltage detector. It will not be effective with what is now essentially a solidly grounded system, as the neutral will not shift appreciably for ground faults - it is now tied down by the grounding bank, rather than floating as previously.
You need to install ground overcurrent relays on the generator circuits; these can be instantaneous relays, as there is no zero sequence current supplied by the still ungrounded generators. Set these as sensitively as possible, in order to minimize generator iron damage for internal ground faults. You also need to look carefully at the ground fault protection of the grounding bank, in fact a fault study and a good look at the overall relaying are in order, if these have not yet been done.
As to what is happenng on the secondary of the grounding bank, any unbalanced zero sequence current will circulate within the closed delta. You can connect loads to the secondary, if you need to.
Technically, I still prefer my second or third option, but I realize that there are probably overriding economic constraints.