Timing of Field Firing for Self Excited Generator 3

[eeguy]

I understand that an SRC based AVR must conduct (or fire the field) for some "apropriate" portion of a half wave to regulate the generator output voltage, but...For a "generic" half wave AVR on a self excited 3 phase generator, is there a particular instant in time with regard to the phase of the output voltages that is optimal for firing the field? Or, can the firing be totally asychronous to what's going on in the generator (as would be the case if the AVR derived its power from some "unrelated" AC source)? Why does, or doesn't it matter? Also, Would the answer be any different if the generator were paralleled to another generator vs driving a load standalone?

 

[jbartos]

Suggestions/Comments on: </images/new.gif> eeguy (Electrical) Jun 10, 2001 posting marked by ///\\I understand that an SRC based AVR must conduct (or fire the field) for some &quot;appropriate&quot; portion of a half wave to regulate the generator output voltage, but...For a &quot;generic&quot; half wave AVR on a self excited 3 phase generator, is there a particular instant in time with regard to the phase of the output voltages that is optimal for firing the field?
///Generally, the AVR must control its output (DC) to the speed governor and to whatever else this output may be used for. Therefore, the AVR output is sensing the generator output voltage (or possibly from other source providing a correct and suitable voltage reference) and processes it by changing the firing phase angle proportionally to the generator voltage output. There are control loops involved with control blocks with various constants to have the genset delivering the generator output nominal voltage at nominal frequency within designed tolerances.\\Or, can the firing be totally asynchronous to what's going on in the generator (as would be the case if the AVR derived its power from some &quot;unrelated&quot; AC source)? Why does, or doesn't it matter? Also, Would the answer be any different if the generator were paralleled to another generator vs driving a load standalone?
///Generally, as in any feedback control loops, the reference point or setting must be used. How it is accomplished, that is a different question. However, the generator output voltage and frequency monitoring for its control purposes is mandatory, to achieve the specified generator output voltage and frequency. There is some manufacturer literature available at:

http://www.basler.com

http://www.thomasregister.com

type Regulators: Voltage the will return 194 companies to obtain more info about voltage regulation.
Books:
1. Bergen A. R., &quot;Power Systems Analysis,&quot; Prentice-Hall Inc., 1986, Chapter 11 &quot;Voltage Control System&quot;
2. Fitzgerald A. E., Kingsley Jr., C., Umans D. D., &quot;Electric Machinery,&quot; Fifth Edition, McGraw-Hill, Inc., 1990, Page 581 Par. b. Turbine-Generator Excitation Systems
It addresses both options, i.e. either input voltage reference from the generator output or from a separate voltage source for voltage reference purposes.
If generators are in parallel, then additional requirements are usually in effect, e.g. load sharing, voltage and speed(frequency) regulator droops, etc.\\

 

[peterb]

The generator field winding requires a DC excitation source. In principle, the method of providing this DC is transparent to the generator - there is no connection between the firing pulse timing and the generator output phasing. Remember that the generator output voltage is determined by the flux density of the rotating magnetic field, which is set by the DC excitation current. The firing pulse determines the LEVEL of DC current supplied to the field, in response to the AVR setpoint and actual voltage.

For standalone unit operation, the AVR will operate in voltage control mode. For parallel operation, the AVR must allow for reactive load sharing. Several options are available, such as PF control, VAR control, voltage droop control, cross current compensation etc. The choice will depend on operational considerations and will be built into the AVR control circuitry. In either case, the main generator field will still be supplied by DC current, so that the process is not dependent on synchronizing the thyristor firing with the output voltage.
You are correct that the source of the power supply for the AVR can be from a separate source, but it is normally taken from the generator output so as not to be dependent on external systems. There may also be an external DC supply for field flashing (if the generator remanent voltage is not high enough to build up to full voltage when starting). Another desirable optional power supply source is a current source fed from generator CTs - this will provide excitation power when the generator is subjected to a fault or large motor starting inrush - these conditions can result in output voltage decay with a voltage-only AVR supply, with consequences for system operations.

 

[eeguy]

peterb,

Thanks for the informative reply. Is the pulsating nature of the field drive (basically 1 pulse every 16mS for a 60HZ half wave AVR)reflected in the output voltage of the generator?

 

[bru]

Yes, sort of.
The &quot;pulse&quot; will have to happen for every cycle. The timing is to do with the point in the AC cycle rather than the frequency. If the &quot;firing&quot; happens early in the cycle (or sine wave), the full &quot;available power&quot; from the source will be allowed to proceed thru the rectifier circuit and into the generator field. This would provide a stronger field and thus higher output voltage (or under load, more current at the same voltage, or higher VAR output or whatever the current configuration requires). (That is what peterb was referring to).

In a typical installation, the source for the field is derrived from the AC system (in a stand alone, ultimately from the generator output) and requires this kind of controlled rectification in order to supply the generator field.

If the &quot;firing&quot; happens later in the ac cycle then less available power is allowed into the field and the output will be reflected in this.

This is typically referred to as the &quot;firing angle&quot; (see jbartos' commentary above, and good links by the way) and is dependent on the output requirements at the output terminals of the generator, and is controlled by the AVR circuitry as peterb has well described above.

The method of bringing the generator on line or sychronizing is quite different from that of synchronous motors, which does require a type of timing for introducing the field. There is sometimes confusion in this. Generators typically synchronize at the breaker and sychronouse motors shronize at the field circuitry itself, with the supply breaker already closed.

 

[eeguy]

bru,

Thanks. I understand that more forcing requires an earlier firing point, but my question had to do with how quickly the gererator can really respond. peterb mentioned that it is the DC field current that matters, indicating that the field averages the firing pulses. How often must we fire for the signal to &quot;look&quot; like DC. Whats the typical time constant of a generator field? Thanks again.

 

[jbartos]

Suggestions: Visit

http://www.basler.com/downloads/shb1bull.pdf

for a sample of &quot;SSE-N Rectifier Chassis&quot; which may be controlled by firing circuit chassis (Basler DECS-300 Digital Excitation Control System). It provides DC output to Generator Exciter Field. Clue is in the following: The SSE (shunt static exciter) consists of a control chassis, power rectifier chassis (SSE-N), and electrical power isolation transformer: all the elements required to maintain generator terminal voltage within ±0.5%. It implies that the power rectifier must be accurate enough including any ripples and spikes. This requires adequate quality filtering and may require more than 6-pulse rectification. Consequently, it resolves any worry about the generator output quality dependent on the automatic voltage regulation and static exciter engineering and design quality including &quot;field averages and firing pulses.&quot;

 

[djs]

The rotating field on most AC generators is a VERY large inductor. Therefore the current and the magnetic field produced by it can not change very quickly. In fact this is in a lot of cases the limiting factor in how fast a generator can respond to a load change.

Many AVR's can over-excite the field to try to compensate for this. For example, a common field voltage is 120 VDC or 180 VDC. Most AVR's will put out >200 volts in field &quot;forcing&quot; conditions.

So to answer you question, ripples in the field current generally will not show up on the AC output.

 

[Siya]

Can anyone explain how drooping works in AVRs and why is it necessary, i.e what would happen if the generators are in parallel without AVR drooping? How is drooping achieved.

 

[jbartos]

Suggestions:
1. Drooping is necessary for load sharing in the parallel mode of approximately same size generators.
2. The voltage regulator experience voltage drop as its generator load increases. This phenomenon is utilized for a device called loadsharer.
3. If generators are in parallel without AVR drooping then they would not share load according to programmed scheme. They would however generate power and not necessarily share load proportionally.
4. Drooping cannot be achieved without AVR drooping. However, there are other principles for load sharing of parallel generators, e.g. isochronous mode of operation.
5. There have been some excellent postings placed in this Forum covering generator droops. They can be obtained over Keyword search in this Forum.
6. Visit

http://www.kepcopower.com/hbkrdnc.htm

http://www.oz.net/~coilgun/levitation/powersupply.htm

http://www.onsemi.com/pub/Collateral/AND8028-D.PDF

http://www.pioneermag.com/power/model.html

http://www.elantec.com/pages/apppdf/d40950.pdf

http://www.marconi.com/html/glossary/glossaryv.htm

http://www.intersil.com/data/FN/FN9/FN9027/FN9027.pdf

etc. for more info

 

[genRman]

Siya,
in my meager experience, the generator droop circuit is there to control reactive current, the engine guys take care of the kW load sharing through their power droop control ( engine governor system)
The method for controling reactive current my company uses is called quadrature droop control and is a function of a current sensing device, usually a C/T being placed in a phase of the generator output winding.
The C/T provides a signal proportional to the primary current passing through it.
When connected across a burdon resistor of a given value, it generates an AC voltage.

The interesting bit of the application is due to the AVR's voltage sensing circuit, normally sensed across two or three phases of the main stator winding.

For instance, if the sensing circuit in a 12 wire, 400V generator was connected across 1/2 U phase and 1/2 V phase, then a sensing voltage in the order of 220V would be seen.

If you draw out the vector relationship between the stator phases, then draw a vector across the sensing points, that will represent the AVR supply.

Going back to the C/T output, if we put this C/T in W phase, then the AC output across the burdon resistor will be in phase with the W phase current at unity power factor.

A vectorial summation of the C/T output voltage and the AVR sensing voltage will result in only a very small net increase9 theoretically zero) in the total sensing voltage.

If however, the generator supplies a power factor tending towards lagging, the vector summation of the C/T output and the AVR sensing voltage will provide a net increase at the AVR sensing terminals. Since the AVR regulates by comparing a preset value to that of it's sensing circuit, a higher sensing value than the machine terminal voltage would peroduce is seen by the AVR, which then marginally decreases the gnerator excitation level to produce a 'drooping' charactoristic is. effectivly this circuit makes the generator AVR control system sensitive to reactive current and has the effect of softening the AVR, hence reactive current load sharing is achieved.
A lack of reactive current load sharing would mean that paralleled generators would 'fight' each other to supply the required reactive load. Thr results of this can be quite scarry, ammeters swing around and genereal geneset load instability ensues as each genset ships and then rejects load.

 

[jbartos]

Suggestion: The AVR and Load Sharer can physically be distinct hardware. The AVR is aligned with the generator exciter and Load Sharer with the motor governor. Visit

http://selcousa.com/t4800.pdf

http://selcousa.com/t4300.pdf

for the Load Sharer info