Industrial Generator Considerations

[mls1]

I'd be interested in comments regarding two situations with an industrial distribution system. The system in question has a turbine generator with a GSU substation that steps 13.8kV up to 69kV. The system is high impedance grounded at the generator and the 13.8kV transformer connection is delta. Here are the issues:

1. There is a tap in the 13.8kV system to feed industrial loads. This was done with a high resistance grounded primary wye transformer. The transformer secondary is delta connected 2400V feeding a number of 2400V substations and loads. The 2400V system has a ground detection transformer with wye/broken delta connection. My question is, why not just use a delta/wye transformer to feed the 2400V system and put a high resistance ground on the wye for detection? Granted, when the generator breaker is open the 13.8kV system will be ungrounded but that is not that unusual and a wye connected PT bank can be used for ground detection (zero sequence voltage).

2. This same installation has a delta connected three phase regulator on the 13.8kV tap to the wye/delta 13.8kV to 2400V transformer. This is to help regulate the widely varying transmission voltage. We now need to add another tap to the system which will require another regulator. What I don't understand is why a delta connected regulator wasn't put at the substation the first time and the entire 13.8kV system was regulated rather than just the distribution tap. The generator regulator is currently PF control which wouldn't be a problem. I've seen LTCs on GSU transformers with no issues so I don't see any reason why a delta regulator would be a problem.

Comments would be appreciated.

 

[collies99]

I find it odd why the 13.8KV bus has large voltage variations. It is the job of the AVR to regulate the 13.8 Kv bus voltage. If it is not set up that way it certainly can.

What did you mean by PF control ?

The 13.8kv "feeder" regulator(s) would be required to manage the voltage drop to the delivery point at the load possibly through some form of IR compensation or other mode.

Historically, was the industrial load and generator exist as a system before and the transmission line was added in later years ?

 

[davidbeach]

The wye-delta-wye transformer with the delta buried seems to be highly underrated for these types of situations. You have a great deal of control over the grounding characteristics on each side and can easily have one side high resistance grounded and the other side solidly grounded or low resistance grounded. The day and age for ungrounded delta system has come and gone.

 

[mls1]

That is an interesting point, davidbeach. I think you are correct that buried tertiary deltas are underutilized and most likely because they are not well understood for zero sequence isolation.

To collies99 questions, the 13.8kV system is mostly governed by the transmission voltage that the transmission provider quotes as +6%/-10% so it can swing by a full 16% and nearly does within a given day. The generator governor could battle this a bit within the reactive capacity of the unit and the system impedances but my models show this would not be sufficient to provide decent regulation. The 2400V system is not very forgiving for voltage variation (it's a legacy system and really should be phased out but that would be much $).

The PF control is the regulator mode that maintains constant power factor at the unit. It's currently set for unity and largely runs underexcited. With the present transformer tap and transmission fluctuations the voltage would go very high if it were set to provide vars even thought that would correct some of the plant var load. We can fix that by tapping the substation transformer better but I'd still like to better regulate the entire 13.8kV system. Bottom line, is there a problem with adding a delta connected regulator at the GSU? I simply don't see any problem with doing that but I can't point to an example other than an LTC application.

 

[collies99]

More question:

How much power flow on the 69 Kv line to the plant ? Is there any significant var flow in the line relative to the power flow ?

Is the transmission line voltage fluctuation at the local HV bus due to/proportional to the load flow.

Is the voltage variation the same at the remote sending end bus ? Or is the remote bus fairly constant ?

 

[mls1]

The plant load is ~5MW and the generator output is typically ~10MW so it is very little for the transmission system. The transmission line voltage is what is changing, plant load has little effect on it. Most likely, it is due to the proximity with some significant interchange and large wind farms to the south. At any rate, the transmission provider has acknowledged that their voltage fluctuates heavily and that their standard is +6%/-10%.

 

[collies99]

It would seem that your plant would run much better islanded, being off the grid and solely fed from the 10 MVA generator, however export power to the system would be lost and so will revenue from power sales.

Mentioning a wind farm at the remote end suggest to me that they do not have enough var support over there. Remote VAR compensator for the wind farm system or even a remote cap bank would be an improvement for your system.

You did not mention if there was a unit CB ? I am assuming that the "tap" is between the generator and its GSU.

If you have control access to the 69 Kv incoming breaker and if power sale is not an issue then voltage control operating of-grid can cheaply be installed along with auto synchro-closer. When the line voltage return to acceptable operating level you can sync back on and when the voltage is too low you can trip and be off the grid. Far fetch idea of course and it still could be automatic. Still need line side and bus side VT for synchronizing.

You suggest that letting the AVR regulate the local bus is doable within the limits of the machine. In this mode if you operate as such, how much vars would you send into the line during low system voltage and will that compund voltage issue at the remote end or even help the remote end bus voltage if the line length is not too long. And if you set the fixed tap on the GSU at -2.5% (mid-point of its swing) would that make the generator VAR operating point, +/- vars swing better ?

Your solution of OLTC for the GSU would be best, matching the high side operating voltage to the incoming lione voltage with referenced to the 13.8 kv bus.