Generator capability whilst connected to national grid question

[scott88]

Hi,

Please could you help my understanding on the generator performance and change in p.f when connected to a grid system in the following scenario;

A generator is connected to the national grid via a step-up transformer. The maximum achievable voltage at max power that can be achieved is 420kV, in the highest tap position, with a P.F of 0.98 at the grid connection point. The grid code states that the normal operating range is 400kV +/- 5% however during abnormal conditions voltages could be between +/- 10% and should last no longer than 15 minutes. My question is if we were connected to the grid and an abnormal condition arose and grid voltage was near 440kV, what would happen to the generator? My understanding is that the AVR would try to maintain grid voltage and that a trip of the protection system would be instigated on an upper excitation limit, however I am not certain on this and would appreciate some clarification.

Thanks,

 

[ScottyUK]

You have the action of the AVR reversed.

If the AVR is in automatic voltage control, the GSU transformer tapchanger remains fixed, and the AVR is regulating the voltage on the main terminals then the AVR will reduce the field to try to reduce the machine terminal voltage as it compensates for the rise in terminal voltage brought about by the rise in grid voltage. You might conceivably hit the under-excitation limiter with the machine operating at a leading power factor as it absorbs reactive power from the grid. This is a bad area of the capability chart for a generator to be working in as the machine stability is compromised and a major grid disturbance could cause it to lose synchronism.

 

[xnuke]

I concur with ScottyUK. If the AVR weakens the field too much as it ineffectually tries to lower the grid's voltage, it can cause pole slippage that can produce major current transients. Protective relaying must be set up to prevent it from happening.

xnuke
"Live and act within the limit of your knowledge and keep expanding it to the limit of your life." Ayn Rand, Atlas Shrugged.
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[odlanor]

I do not agree!
"The grid code states that the normal operating range is 400kV +/- 5% however during abnormal conditions voltages could be between +/- 10% and should last no longer to 15 minutes."
You must adequate your AVR for that contingence.

 

[IBRCAN]

The operating voltage of a grid cannot be controlled by a relatively tiny generator which is connected to it. An AVR is used with a generator to maintain a fixed terminal voltage, making it unsuitable for use by itself in a grid-paralleling application. Instead, a VAR/PF controller is used in conjuction with the AVR, or there are some AVRs which can be programmed for VAR/PF mode. As the name implies, this device controls excitation to maintain a constant generator power factor regardless of grid voltage. Generator excitation will be increased with increases in grid voltage.

Of course there is an upper limit for generator field excitation--a system would need to include protection for over excitation.

I suggest contacting the manufacturer of the generator in question for assistance with your design.

 

[LSpark]

As the others have said the AVR maintains the generator terminal voltage and generally has negliable effect on the grid volts (we consider it to be an infintie busbar). In theory the high grid volts could occur at any of the tap positions of the step up transformer.

I'm guessing you're UK based? In the UK the large generators are given reactive instructions and they use the tapchanger on the step up transformer (bearing in mind the AVR acts to maintain the generator terminal volts) to adjust the reactive power at the HV side of the step up transformer. The generators should all have under excitation protection (I've seen a unit trip on this), posibly pole slipping protection as well as systems to prevent over excitation.

 

[waross]

Once the limit of the AVR/PF controller is reached, and absent protection tripping the unit off-line, increased grid voltage will result in less VARS delivered to the grid and a resulting shift in power factor. As the voltage rises the Vars delivered will decrease until unity power factor is reached. Further increases in grid voltage will result in the generator absorbing VARs from the grid and a dropping power factor. Real power delivered will depend on the throttle or steam valve setting.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[ScottyUK]

IBRCAN,

I guess you have limited experience in an operational power plant, because if you did then you would realise that the AVR is a key part of the generator's control. To say that an AVR is unsuitable for use with a grid-paralleled machine is quite simply untrue, as demonstrated by the numerous machines throughout the world equipped with an AVR and operating in a satisfactory manner.

 

[squeeky]

I am in Africa where anything is possible. Our AVRs are set up with feedback coming from the other AVR breakers and the incomer. The AVR of a sole alternator will see nothing else on the bars and run in voltage mode. Here the AVR maintains the voltage according to the load. If another alternator joins the bus then both AVRs see it and run in droop mode. Here they work together to control the voltage. If the operator now contects to the grid, the voltage is controlled by the grid and you can do nothing. Now all alternators move into power factor mode. Only the import and export of VARs can be controlled.

We experience high voltage ranges over long periods of time. We have had to start adding incoming transformers with On Line Tap Changes with (OLTC) 17 steps of 1.25%, giving plus / minus 10%. Now if we need to connect to the grid that is operating near or outside our limits of control, the voltage regulator on the transformer steps and allows the factory voltage to remain constant with in the new increased range. It allows for fast synchonization. Also if you are required to remain connected to the utility during the out of range periods, when it is high, you will electrically stress all your insualtion on the electrical equipment. This is seen as insulation failures. If the voltage is low, all the torque values will drop and the current will increase and possibly cause tripping if you are on your power limits with your design. The electricians turn up all the overloads to prevent "nuisance tripping' and now you have lost your protection.

 

[ScottyUK]

Squeeky,

Do your AVRs sense the high side of the GSU transformer or the low side? If they sense the low side then most generators are capable of flexing their own terminal voltage even when connected to an infinite bus because of the GSU transformer impedance. The flow of reactive power can - and does - influence the local grid voltage and is one method of controlling it. I am puzzled by this misconception that a grid can't be influenced by a generator. A grid node close to a large generator or plant certainly will be affected by the AVRs. The influence on voltage is often relatively small, but the 'infinite bus' is a convenient first-order modelling tool rather than a statement of fact, even on a dense and heavily interconnected grid like in the UK. In a larger and less dense grid like in much of the US and I suspect in S.A. too the effect of a generator on the local grid will be more significant.

 

[waross]

I can't imagine a generator large enough to connect at 420 kV being operated in voltage control mode to export power.
The generator/AVR will be operated in PF mode.
That said, I have seen the exception.
A city was supplied by a diesel generator plant.
A large hydro generating installation was constructed on the other side of the country.
A transmission line was built and the city was fed from the hydro plant. The diesel plant was mothballed.
The city grew and the transmission line became overloaded and low voltages became an issue.
The diesel plant was re-commissioned and run in either fixed excitation or voltage control mode to control the voltage at the city.
Due to the present cost of fuel, very little real power was produced by the diesel plant.
Yes, a generator can and will influence the grid voltage if the grid connection is relatively soft, and as Scotty points out; The bus is far from infinite.
scott88; Assuming that the AVR is run in PF mode, once the voltage rises past the point that the AVR is at full output, the export of VARs will be reduced. The export of real power will continue as real power production depends on throttle or valve setting and not voltage.
With a large enough voltage rise it is possible that the power factor may drop to the point that the KVA load may overload the generator and/or the GSU.
This is assuming that there are no over-voltage failures and nothing saturates.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[ScottyUK]

We only connected at 275kV, and maybe the rules are different at 400kV , but we did export over 1800MW to the grid. Our machines ran in voltage control.

The differences in mode possibly stem from differing methods of despatching units. In the UK the merchant stations are largely despatched on MW output with fixed terminal voltage, and the reactive power is regulated on the GSU transformers. In the specific instance of the plant I worked at until recently we were contracted to provide reactive power support, but rarely did in reality. The grid at that location is fairly robust and generally sat slightly high at about 282kV or so, so there was little need to export reactive power under normal conditions. The station was largely Var-neutral at the grid connection, with the Var consumption of the transformers being met from the generators. In the event that a reactive despatch was received, it would be met in the first instance by tapping the GSU transformers against the fixed output voltage of the generators with a possible trim by flexing one or more generator terminal voltages slightly to meet the despatch more precisely.

It would make a certain amount of sense for a large baseload plant to operate in PF control, and allow the merchant stations to flex both power and reactive power to suit the grid needs.

 

[bacon4life]

NERC standard VAR-002-2b requires that the "Generator Operator shall operate each generator connected to the interconnected transmission system in the automatic voltage control mode." Thus, I had assumed generators throughout the USA would have their AVR in voltage control mode. Plants in power factor mode will not contribute additional reactive power to the grid after a system disturbance.

That being said, the voltage set point given to the AVR can be adjusted up or down to reach the desired power factor by an external control loop. If the voltage set point is adjusted fast enough, the AVR might technically be in voltage control mode, but would behave as if it were in PF mode.

Modern AVRs should provide under/over excitation limiting rather than having protective relays trip the unit at the excitation limits. Any unit trips should be based on over voltage relays protecting voltage sensitive elements or volts/hertz relays protecting power transformers.

Our somewhat remote hydro plants can swing the high side bus voltages a couple of percent if they were to swing from the under excitation limit to the over excitation limit.

 

[squeeky]

Really interesting. I think at the quanities and voltages you are generating at, you are the infinate bus bar. I'm in the sugar industry and generally we would only connect to the grid in times of startup and then if we are experiencing low steam pressures on the boilers and then we import electrical energy to allow the alternators to back off and save steam until we recover the steam pressure. Exporting or co-geration requires a licience which IPPs (Independant Power Producers) are trying to get. This can be very difficult to break the monoply. There is just no way we can influence the grid exporting 10 to 30MVA at 11kV to the local grid which is connected to the national grid but at 132 and higher voltages. But your input was of interest to me. Thanks.

 

[IBRCAN]

IBRCAN,

I guess you have limited experience in an operational power plant, because if you did then you would realise that the AVR is a key part of the generator's control. To say that an AVR is unsuitable for use with a grid-paralleled machine is quite simply untrue, as demonstrated by the numerous machines throughout the world equipped with an AVR and operating in a satisfactory manner.

You are correct, I am speaking from my experience in the area of cogeneration with reciprocating-engine/generators, which are typically very tiny compared with the grid. As has been said above, the type of voltage (and frequency) control depend largely upon the size of generator vs. stiffness of the grid.

 

[ScottyUK]

I have limited experience with smaller machines connected to the distribution network. The operating conditions imposed by the distribution network operators are different to those imposed by the transmission system operator, so in the case of smaller embedded generators the requirements of controlling the generator are likely different.

Waross is very knowledgable about these smaller reciprocating engine driven plants is islanded mode or attached to small grids, you two will probably swap some ideas. I apologise if my earlier reply seemed rude.

 

[LSpark]

I had reason to look through a connection agreement for one of our DNO connected sites (132kV) and that needs to run in PF control mode. That partiular site is design to run in island mode in case of loss of connection to the main grid, so any site with blackstart capability will need to be able to operate in PF control mode even if it doesn't under normal grid conditions.

 

[ScottyUK]

PF control on an islanded unit is meaningless. The machine will run in isochronous mode - i.e. at fixed frequency - and in voltage control. The PF is determined by the load. If there are multiple units then they will likely run in droop control (simple) or have some form of active load-sharing (less simple).

 

[QBplanner]

if the machine is big let's say 800MW interconnecting to a 400Kv system radial system or wek network, it becomes a master unit in the system and it is possible to regulate the grid voltage from 440Kv down to 420kV.
if the machine is too small say 80MW interconnecting to a 400kV system, it is s slave unit and should follow the gird system and what will happen was already mentioned by a lot of experts here already.

I did not see any words mention that the unit is running in PF controll mode.

 

[crshears]

Most of the transmission-connected generators I've worked with operate with their AVR's in voltage control mode, almost exclusively with the potential source derived from the generator terminal PT's and not from the HV bus. This is particularly handy when placing hydraulic units in service, since these units can go from being at rest to fully loaded in the space of three minutes; once the unit is on line but at zero load the wicket gates can be opened up to load the unit without having to simultaneously and manually increase the excitation [due to internal voltage drop]. This is particularly desirable with Kaplan units which may experience severe cavitation at wicket gate openings between, say 40% and 70%.

In my province's system it is quite routine to use the major 'utility' [publicly owned] generating stations as reactive resources; none of their generating unit output transformers are equipped with underload tapchangers. Although some do have off load tapchangers, these are commonly selected to an optimal tap position either upon commissioning or very soon thereafter; it takes a major change in system characteristics for an alteration in tap position to even be considered.

The allusion to NERC standard VAR-002-2b by bacon4life is interesting; a number of what are coloquially referred to as 'non-utility' [private partnership] generators have connected to our grid in recent years, and although the NUGs comply with the requirement mentioned, in practice their reactive output is essentially ignored by the system operator when dispatching reactive resources...to the point where I have seen more than half of the VARs being added to the system when static capacitors are placed in service absorbed by the NUGs. As near as I can tell the only thing the NUGs respond to is their internal under- or over-excitation limits.

Ridiculous, in my view, but above my pay grade.

Smaller distribution-connected generators may more commonly, in my experience, be equipped with reactive output control as well as the classic AVR scheme. Distribution votages are much more commonly regulated with ULTCs, these being manually adjusted as needed when other reactive resources such as DOL static capacitors are being switched in or out of service; the smaller NUGs have no obligation to support voltage and prefer to operate their equipment at as low a current as possible, hence the choice to operate near zero reactive.

Hope this helps.