Voltage regulator settings for OLTC 4

[bacon4life]

I am attempting to set a voltage regulator for a 15 MVA transformer with an OLTC and it seems like the tolerances have to be much tighter than I would have expected. This is my first time so please point out which assumptions I should change.

Goal: Voltage at the customer’s service of 120V +/- 5% (114 to 126V)

Sources of voltage variation:

+/- 1% voltage regulator accuracy
+/- 0.3% voltage transformer accuracy
+/- 1.2% voltage regulator bandwidth
+/- 0.5% Imbalance in high side voltage on regulated and unregulated phase due to transmission system
+/- 0.5% additional imbalance in low side on regulated and unregulated phase due to +/- 100 Amps of load imbalance
-0/+4% voltage drop on distribution system, zero at nearest customer, 4% at the last.
None Safety margin

Total result
11% tolerance

One way I could set it is at 124.8V (120V +4%) and have a bandwidth of 113.4 to 126.6.
Thus in the past one could:
A. Hope that tolerances are never at the same maximum/minimum value.
B. Hope that the customers don’t complain about the voltage. Not only does this seem unethical, we now have SCADA for the bus voltages, so we would know before the customer if the voltage was high, so this option doesn’t work.

Some other options might be:
C. Regulate all three phases
D. Strengthen the distribution for less voltage drop.
E. Better balance the feeders
F. Buy higher accuracy components

There is also a backup LTC controller in this installation. It seems like it should be set 1% higher & lower than the primary, but this would put it well outside of the allowable 5% limits.

There also does not appear to be room for the load drop compensation to operate and be sure to stay under 126V at the first customer.

Any help or comments would be greatly appreciated.

 

[davidbeach]

A voltage regulator in the line to control the far end independent of the voltage at the transformer would give you a bit of room, but adds cost.

 

[bacon4life]

I doing textbook exercises, I had thought that each regulator would give you the ability to deal with almost 10% voltage drop. The way I currently see this, each regulator would only be good for 3.5% voltage drop.

 

[Rodmcm]

In general the operation of an AVR on a tapchanger should maintain within 3 to 4% of set point. Obviously with tap of say 1.25% then the tap change setting needs to be say 2% and will be slugged with time delay before operating.
The total inaccuracy is not all that important. What is important is that the repeatability is the same at the setpoint.

 

[apowerengr]

Ignore most of the error assumptions - set it Vbase = 124 R=X=0 (compensation makes a mess of things) 2.5 V bandwidth, time delay based on utility standards (30-45 seconds), dead band (if the control is so equipped) 129-119.

4% V.D. at 119 will kick the control into a raise operation when your last customer just sees 114 volts. 2.5 volt bandwidth keeps the first customer at 126 or less (there's always some voltdrop in the dist transformer/secondary/service).

LTC and delta-wye transformer (assumption) takes care of most voltage/load imbalances (although loss of a phase on the transmission gets interesting). Make sure your sensing PT's have the correct ratio. LDC CT ratio does not matter since in this case you're not using R or X, although if you use it for SCADA load monitoring ..........
Set the no-load tap on the transformer accordingly so the LTC runs through neutral.

 

[subtech]

Agree with apowerengr.
Additionally, I'd set the backup LTC control to become active only if the other failed (went to alarm state). If both units are active, setting the backup so it doesn't interfere with the primary usually makes it pretty much worthless. We close an aux relay (latching w/ manual reset) with the alarm contacts on our primary control and the aux. enables the backup.
You can help yourself quite a bit at the far end of the feeder if you have the option of a small/medium cap bank. Eliminating 1/2 to 2/3 of the lagging var load w/caps during peak usually makes a big difference. I would look at caps first before considering regulators. Unless you've got an unusually long feeder which makes regs unavoidable, the initial purchase and maintenance cost difference between caps and regs is considerable.

 

[bacon4life]

Our electrician that programs our voltage controller regularly sees a voltage imbalance due to load imbalance on some of our feeders. The phase that is high changes on a weekday vs. weekend and winter vs. summer basis so some transformers he has been reprogramming them seasonally. We typically limit feeder imbalance to less than 100A, but I don’t know if we typically rebalance phases if there is a larger imbalance at the transformer.

It is interesting you mention that there is always voltage drop to the first customer. We have multiple feeders from most of our substations, sometimes with one lightly loaded and the rest heavily loaded. We also measure voltage at the substation bus and get an alarm if the voltage exceeds 126V. Although the customer voltage may still be ok, I am not sure I could present a justifiable calculation to reset our SCADA alarm points system wide to higher than 126V.

The primary controller we use recommends also using a backup controller, so we are using one. I guess since it is to prevent a run away situation, setting it outside the nominal bandwidth wouldn’t be a problem. It seems a bit strange to rely on the failed controller to tell you that the controller has failed. Isn’t a backup supposed to be as independent as possible?

Rodmcm
Can you explain more of what you mean that it is more repeatability than total accuracy that is important. Since we have a standardized unit substation, I had thought we would also have standard regulator settings. It turns out that we tweak them for each individual regulator after testing the regulator.

 

[RalphChristie]

Bacon:

Is the voltage-rating at the customer 120V, or is this the voltage-rating at the regulator? (or both?)

Regards
Ralph

[red]Failure seldom stops us, it is the fear for failure that stops us - Jack Lemmon[/red]

Make the best use of Eng-Tips.com
Read the Site Policies at FAQ731-376

 

[bacon4life]

Ralph:

My understanding is that US utilities usually provided 120V, with a further voltage drop allowed on the service drop and customer wiring. This is at the substation 12.47kV bus rather than customer equipment. I'm not sure that answered your question though?

 

[davidbeach]

Ralph,

Because of the long standing tradition of providing residential service at 120V, utilities will often use a 120V base for evaluating other voltages on the system. For instance we try to run our distribution substation busses at about 123V on a 120V base, or about 13.53kV on a 13.2kV bus.

 

[RalphChristie]

Ok - what I actually meant to ask, is the secondary voltage (what you are providing to your customers) 120V, or is it actually a higher voltage like 13.2kV etc.?

On our systems the OLTC is normally on a upstream transformer (66/11kV, 132/22kV, etc) with the lower voltage been held on a certain setpoint. (for 11kV normally between 11.2kV and 11.4kV, depending on the line length.) Downstream transfomers for local distrubution (11kV/400V/230V, 22kV/400/230V, etc) are been set with an off-load tapchanger to the required voltage. For instance, a local trsf near the substation will be set on a lower tap and a trsf far away will be set on a higher tap:

tap 1 +5% (11.5kV/400V) Transformers near substation
tap 2 +2.5% (11.25kV/400V)
tap 3 nominal (11kV/400V)
tap 4 -2.5% (10.75kV/400V)
tap 5 -5% (10.5kV/400V) Transformers far away from substation

Normally the off-load tapchangers are set once on the required voltage, and never been changed, except if the transformer is been taken to another location or if there are system changes.

Basicly the same parameters are been used as indicated by apowerengr and subtech, voltage setting little higher than the nominal voltage, a bandwith of 2%, and a linear based time behaviour curve. Regarding a back-up - I have never seen a voltage regulator been used as a back-up for another regulator, but if been used, I think subtech's approach would be the best.

Some additional information:
Some setting examples in the manuals of a-eberle, it might be useful.

http://www.a-eberle.de/gb/index.htm

Regards
Ralph



[red]Failure seldom stops us, it is the fear for failure that stops us - Jack Lemmon[/red]

Make the best use of Eng-Tips.com
Read the Site Policies at FAQ731-376

 

[stevenal]

We use 124 with a 3 volt bandwidth. Remember the voltage setting lies at the center of the bandwidth, so controller will call for a lower operation at 125.5 V.

In general, it is not proper to add error since the odds are that some errors will cancel others. Back in physics class we used rms addition to calculate overall error.

Also, don't forget that ANSI C84.1 allows excursions into range B that are "limited in extent, frequency, and duration." Kinda redundant, since range B defines the actual extent. On a 120 V basis, range B is 110 to 127 at the service. Range A is 114 to 126.

 

[bacon4life]

This is on a 12.47kV system represented as David explained.

I’ll have to ask about the settings on our transformers to see if they are set as Ralph suggests, but I don’t think it would work for us. Most of our feeders can be carried by a remote substation, so our one of our problems with voltage drop is with low voltage at the first customer outside the substation when the substation is offline.

This is the quote from the Beckwith instruction manual: “Use of a back-up relay, such as the M-0329B, is recommended since the M-2001C (or any independent control) cannot be expected to be its own backup.”

 

[jghrist]

one of our problems with voltage drop is with low voltage at the first customer outside the substation when the substation is offline.

The voltage at the first customer outside the station does not depend on the station OLTC settings when the customer is served from a different station.


Setting the OLTC at maximum permissible voltage (allowing for bandwidth) and load drop compensation R=0, X=0 provides maximum voltage for all customers at all loads. It also provides the worst voltage regulation (difference between light and heavy load voltage) for customers at the end of the line. The reason for setting load drop compensation R and X at something other than zero is to provide better voltage regulation to the majority of the customers.

 

[bacon4life]

It does depend on the settings of the OLTC at the remote customer and if we were to have the taps as Ralph suggested, the first customer would become the last, although the last customer in the normal configuration becomes the middle of the feeder when two feeders are carried by the remote substation.