advertisement
from the Automation List department...
Gas Turbine Trip Response
Continuous process industries, DCS questions. topic
Posted by arocon on 27 October, 2005 - 12:35 am
work in GE Frame 9 Gas Turbine of which control system is Mark IV. We have 5 Gas Turbines. One of the GT if trip then another GT is tripping whether all GTS are in Base load or Part load. Could anyone tell me what is the problem? and tell me where in Mark IV system i have to check. As i knoww if all GTs are running in Base load if any GT trip then it should not try to increase the load of all other GTS as they are already in Base load. Also If GTs are running in Part load they should behave with Droop control.

Could anyone tell me specifically where to look in control system, load control, droop control or frequency control in the Mark IV system? Also could you a little bit explain about droop control. It will be a great help.


Posted by markvguy on 28 October, 2005 - 10:46 pm
It's virtually impossible to answer your question about tripping without a detailed understanding of YOUR system, its control schemes, and how all the units are connected to each other and/or to the grid. There are just too many possible configurations.

Also, you are not supplying the various process alarms which are being annunciated whenever one of these trips occur. GE's heavy duty gas turbine control philosophy is that ANY and EVERY condition which would result in a trip (emergency shutdown) of the turbine and/or generator MUST have a process alarm associated with it.

Now, having said that, there are times when this has been violated during installation, start-up/commissioning and after the unit has been put in service--but it is the expectation that any condition which would result in an emergency shutdown (trip) would be annunciated on the process Alarm Display. (It should be every operator's and technician's and owner's expectation, anyway, that there should never be an unannunciated trip.)

[Actually, under very extreme conditions on Mk IV there can be unannunciated trips--but they would be preceded by numerous Diagnostic Alarms which were not resolved and allowed to persist. A subject for another discussion, though,...]

There are also some very confusing--and complicated--process alarms which are annunciated, and some individual process alarm conditions can be the result of one of several conditions, but that's what the technician is supposed to do: troubleshoot each and every alarm condition, understand every individual condition which could cause the alarm under investigation to be annunciated, and resolve the appropriate condition.

For example, the alarm "Generator Differential Lockout" is usually caused by the differential lockout relay, 86G, actuating. But, the 86G lockout relay can be actuated by any one of several other relays in the generator protection scheme, and even by other relays in the high-voltage and/or transformer protection scheme.

So, without drawings of how the generators are connected to the grid (each generator through its own step-up transformer, or two generators per transformer, or all generators through a single step-up transformer, etc.), without understanding all the protective relay settings and schemes, and without detailed alarm printouts (process and Diagnostic alarms) it's impossible to troubleshoot this problem via this forum. It could be that if the units are running on gas fuel and one of them trips there is an upset in supply pressure to all the units such that the (physically) closest units trip on overtemperature because the SRV didn't respond fast enough. Or, that a protective relay setting erroneously detected "excessive" fault current flowing in some phase because of an improperly adjusted breaker contact. Or....or....

As for Droop (and Isochronous) here goes...There are as many different definitions and metaphors and thoughts about Droop Speed Control (and Isochronous Speed Control) as there are instrument technicians and plant managers and plant engineers; this is just _this_ writer's understanding.

On any electrical grid/network, the LOAD is not a function of the number and rating and power output of the GENERATORS, it's the aggreagate total of the motors and lights and devices which CONSUME power. The power SUPPLIED must equal the power CONSUMED in order for the grid/network to operate as designed (desired voltage, and, for AC grids, frequency). If one just keeps connecting generators to the grid and "loading" them, the voltage and/or the frequency will increase above the setpoint.

It is very important to control the frequency of an AC electrical grid/network. On an AC electrical grid/network with multiple generators all supplying the "load" (really, each one is just 'accepting' some of the load) theory has it that there is one generator which is controlling the frequency of the grid and its governor is operating in Isochronous mode. This means that the governor will adjust it's fuel/steam/water/blade pitch (depending on the type of prime mover driving the generator) in order to maintain the frequency setpoint.

Assume the Isoch unit is running at approximately 50% of rated output and a motor is started somewhere on the grid. The effect of the load of the motor would be to cause the grid frequency to decrease--but the Isoch unit very quickly increases the torque input to the generator to maintain the frequency. If somewhere else on the grid another motor is shut off, the decrease in load would tend to cause the grid frequency to increase--but, again, the Isoch governor will immediately decrease the torque input to the generator to maintain the frequency setpoint.

If the load is allowed to increase to the rated power output of the Isoch unit and then more load were to be added then the grid frequency would start to drop--unless other generator(s) were connected to the grid to reduce the load on the Isoch unit such that it could then respond to increases/decreases in load and frequency.

Theory has it there can only be ONE generator whose governor is operating in Isochronous mode--two or more governors configured to operate in Iscoh will generally "fight" each other for control of the frequency. (There are governor schemes to prevent this, depending on the size of the grid/load.)

On very large electrical grids there are so many generators connected to the grid (and to each other) there probably isn't a single unit controlling the frequency. In this case, the system operators are responsible for adjusting the load on several units, and for bringing more generation on line when necessary and for reducing the number of generators when necessary--all in order to maintain grid frequency within a very tight tolerance.

All the other generators connected to the same electrical grid "should" (read "must") have their governors configured to operate in Droop Mode. Droop governor mode is not as concerned about maintaining frequency as Isoch mode; in fact, the word "droop" means that the governor is going to allow the speed (frequency) to droop from its setpoint--because the frequency is being controlled somewhere else (by an Isoch unit or the grid operators).

When you are increasing the load being 'supplied' (really, being accepted) by a gas turbine operating in Droop mode, even though you are watching the MW meter what you are really doing is telling the SpeedTronic to increase the speed of the unit by increasing the fuel being admitted to the unit. However, if the unit is connected to an AC grid with many other generators--or even with just one other generator whose governor is operating in Isoch mode--the speed/frequency of the generator being loaded doesn't change, and that additional increase in fuel becomes additional torque input to the generator which becomes increased amps flowing out of the generator. So, you're not changing the load setpoint as you load/unload a unit--your're changing the speed setpoint, but the speed isn't changing because it's being controlled somewhere else on the grid.

At the same time you are loading a unit in Droop mode, if no other changes are made the grid frequency will start to increase. So, the Isoch unit must back down its torque output to its generator to maintain the frequency, which means the load of the Isoch unit will drop as the load of the Droop unit is increased. (Same thing when you unload a Droop unit--the Isoch unit must increase its load to make up for that being "shed" by the Droop machine, otherwise the grid frequency will begin to drop.)

Both of these governor modes--Isoch and Droop--are speed control modes. Both governor modes increase or decrease the load being supplied by the unit in response to changes in speed setpoint. Isoch governors very quickly respond to very small changes in frequency; their setpoint is grid frequency (remember, generator rotor speed is proportional to generator frequency). Droop governors don't respond very quickly to speed setpoint changes--the design of the system is such that another unit/"entity" is in control of the frequency. In fact, at rated power output most prime movers operating in Droop mode are operating at a speed that is 4- or 5% less than the setpoint. In other words, the actual speed is being allowed to "droop" below the speed setpoint--hence the term "droop speed control."

Think of a train with multiple engines being individually controlled, each by a train engineer. pulling a load and being to run at a constant speed across a flat plain. To maintain that speed, the train engineer has adjusted the throttles of the engines such that the speed of the train is not changing. Now, imagine that a train car is added to the train (or, that it starts up a hill--in either case, the load on the engines increases and if all the train engineers increase their engine power outputs, the train will start to speed up excessively. If just one train engineer increases his engine's power output to maintain the speed, then train speed will remain constant.

If the train starts down a hill (or train cars are "removed"/disconnected), the speed would tend to increase. If all the engineers reduced the power output from their engines the train might slow excessively. However, if just one engineer reduced his engine's power output it would be easier to maintain the train's speed. If too many cars were "removed"/disconnected and the one engine's minimum power output was reached, then the train's speed would start to increase--unless the power output of one or more of the other engines were reduced to allow the one engine to control the train's speed.

If the train engines have governors, one of them can be configured to be the "Isoch" unit and the others to operate in "Droop" mode. The Isoch engine will adjust it's power output to maintain speed as the terrain changes (or as train cars are "added" or "dropped off"). If the load increases such that the Isoch engine reaches its maximum power output and the hill becomes steeper (or more train cars are "added"), the train speed would tend to decrease. But, if one of the other engine's is "loaded" then the Isoch unit can back off in order to respond to any further increases in load.

You can also think of it as several people working together to hold a heavy beam at a certain height. If one person lifts more than the others, the beam will rise and the others will have to reduce their share of the load to maintain the height. It's all very similar.

Remember, all those who are reading this and possibly taking exception: You are all free to explain it in your terms (just please don't unduly criticize this writer's descriptions unless something is grossly in error!).


Posted by Anonymous on 4 November, 2005 - 10:34 pm
The alarm message for causing trip is Exhaust temperature overtemperature. But immediately we started GT and synchronized without any problem.

Hope this information will be helpful to explain the topic.


Posted by markvguy on 6 November, 2005 - 3:32 pm
An 'Exhaust Overtemperature-Trip' process alarm while operating on CPD-biased exhaust temperature control indicates that the actual exhaust temperature has exceeded the reference exhaust temperature by more than 40 deg Fahrenheit. This alarm should be preceded by an 'Exhaust Overtemperature Alarm' process alarm, which indicates the actual exhaust temperature exceeded the reference exhaust temperature by more than 25 deg F.

If you truly want a well-reasoned and thoughtful answer, take a few minutes and answer the following questions--including any information you might think pertinent.

Also, please provide the entire list of alarms for 30 seconds prior to and at least 30 seconds after the trip. You can use the Trip Log/Trip History feature to view the alarms prior to and for a few seconds after the trip (you must print the Trip Log/Trip History BEFORE restarting the unit).

Has this problem just started recently, or has it existed for some time?

Are the units running on gas fuel, liquid fuel (distillate), naptha???

Has something recently changed or been modified in the fuel supply system? (A new gas fuel pressure regulator or gas fuel compressor, or new liquid fuel forwarding pumps, or a new or recently "rebuilt" liquid fuel forwarding system pressure regulator, any recent modifications to the fuel supply piping/filtering, etc.?)

Have any fuel system-related Control Constants been changed recently (i.e., prior to this problem beginning to occur)?

Does the unit which trips after the first unit share a connection to a main step-up transformer?

Are there Diagnostic Alarms present prior to the trip on the second units to trip?

What are the servo currents for the fuel valve(s) for all three processors while running at Base Load on the units which are experiencing the highest number of conicidental trips?

Are _ALL_ three 4R, 4S, and 4T relays picked up prior to the trip of the second unit?

Does the Generator Lockout Relay (86G) or Transformer Lockout Relay (86T) require a manual reset prior to restarting?

The easiest "thing" to blame for (perceived) improper operation is the SpeedTronic turbine control system. Assuming no "serious" Diag. Alarms are present prior to the trips of the second units--and, based on the LIMITED information provided to date--and that the fuel valve servo currents of the second units to trip are all reasonably well balanced, the most likely cause of the trip is a sudden increase in fuel flow to the second unit which is resulting from a sudden decrease in fuel flow to the first unit--and that increase is not being properly detected and minimized by the Mk IV.

Many, many, many hours have been spent trying to prove that (perceived) problems are being or have been caused by the SpeedTronic (Mk II, Mk IV, Mk V, and Mk VI). Most of those same problems were usually found to be OUTSIDE of the SpeedTronic turbine control panel.

In general, it's easier to analyze a problem and eliminate all potential "external" causes before turning one's attention to the SpeedTronic. (This is assuming that no "serious" Diag. Alarms are present.) This is primarily because of the lack of documentation about GE heavy-duty turbine control philosophy as well as a general lack of documentation about how the various versions of the SpeedTronic turbine control systems implement that philosophy.

But, it's also because it's more common that such problems--especially if they are occurring on multiple units--are outside the SpeedTronic turbine control panel.

Ultimately, please write back and let us know if and when you resolve the problem. Many people monitor these exchanges, and we can all learn from other's experiences especially if there is feedback--positive or negative.


Posted by martin young on 29 October, 2005 - 12:41 am
You really need to detail what alarms you get on the SpeedTronic to enable anyone to tell you what's causing the trip. You probably also need to provide details of the electrical system your 5 machines are connected to.

My guess would be loss of control volts or generator undervoltage.

The speed difference percentage between full load and no load is commonly known as droop. Commonly 4% is used in the UK.


Posted by markvguy on 29 October, 2005 - 1:00 am
It's virtually impossible to answer your question about tripping without a detailed understanding of YOUR system, its control schemes, and how all the units are connected to each other and/or to the grid. There are just too many possible configurations.

Also, you have not supplied the various process alarms which are being annunciated whenever one of these trips occur. GE's heavy duty gas turbine control philosophy is that ANY and EVERY condition which would result in a trip (emergency shutdown) of the turbine and/or generator MUST have a process alarm associated with it.

Now, having said that, there are times when this philosophy has been violated during installation or start-up/commissioning or after the unit has been put in service--but it is the expectation that any condition which would result in an emergency shutdown (trip) would be annunciated on the process Alarm Display. (It should be every operator's and technician's and owner's expectation, anyway, that there should never be an unannunciated trip.)

[Actually, under very extreme conditions on Speedtronic Mk IV turbine control panels there can be unannunciated trips--but they would be preceded by numerous Diagnostic Alarms which were not resolved and were allowed to persist. A subject for another discussion, though,...]

There are also some very confusing--and complicated--process alarms which are annunciated, and some individual process alarm conditions can be the result of one of several conditions, but that's what the technician is supposed to do: troubleshoot each and every alarm condition, understand every individual condition which could cause the alarm under investigation to be annunciated, and resolve the appropriate condition.

For example, the alarm "Generator Differential Lockout" is usually caused by the differential lockout relay, 86G, actuating. But, the 86G lockout relay can be actuated by any one of several other relays in the generator protection scheme, and even by other relays in the high-voltage and/or transformer protection scheme.

So, without drawings of how the generators are connected to the grid (each generator through its own step-up transformer, or two generators per transformer, or all generators through a single step-up transformer, etc.), without understanding all the protective relay settings and schemes, and without detailed alarm printouts (process and Diagnostic alarms) it's impossible to troubleshoot this problem via this forum. It could be that if the units are running on gas fuel and one of them trips there is an upset in supply pressure to all the units such that the (physically) closest units trip on overtemperature because the SRV didn't respond fast enough. Or, that a protective relay setting erroneously detected "excessive" fault current flowing in some phase because of an improperly adjusted breaker contact. Or....or....

As for Droop (and Isochronous) here goes...There are as many different definitions and metaphors and thoughts about Droop Speed Control (and Isochronous Speed Control) as there are instrument technicians and plant managers and plant engineers; this is just _this_ writer's understanding.

On any electrical grid/network, the LOAD is not a function of the number and rating and power output of the GENERATORS, it's the aggreagate total of the motors and lights and devices which CONSUME power. The power SUPPLIED must equal the power CONSUMED in order for the grid/network to operate as designed (desired voltage, and, for AC grids, frequency). If one just keeps connecting generators to the grid and "loading" them, the voltage and/or the frequency will increase above the setpoint.

It is very important to control the frequency of an AC electrical grid/network. On an AC electrical grid/network with multiple generators all supplying the "load" (really, each one is just 'accepting' some of the load) theory has it that there is one generator which is controlling the frequency of the grid and its governor is operating in Isochronous mode. This means that the governor will adjust it's fuel/steam/water/blade pitch (depending on the type of prime mover driving the generator) in order to maintain the frequency setpoint.

Assume the Isoch unit is running at approximately 50% of rated output and a motor is started somewhere on the grid. The effect of the load of the motor would be to cause the grid frequency to decrease--but the Isoch unit very quickly increases the torque input to the generator to maintain the frequency. If somewhere else on the grid another motor is shut off, the decrease in load would tend to cause the grid frequency to increase--but, again, the Isoch governor will immediately decrease the torque input to the generator to maintain the frequency setpoint.

If the load is allowed to increase to the rated power output of the Isoch unit and then more load were to be added then the grid frequency would start to drop--unless other generator(s) were connected to the grid to reduce the load on the Isoch unit such that it could then respond to increases/decreases in load and frequency.

Theory has it there can only be ONE generator whose governor is operating in Isochronous mode--two or more governors configured to operate in Iscoh will generally "fight" each other for control of the frequency. (There are governor schemes to prevent this, depending on the size of the grid/load.)

On very large electrical grids there are so many generators connected to the grid (and to each other) there probably isn't a single unit controlling the frequency. In this case, the system operators are responsible for adjusting the load on several units, and for bringing more generation on line when necessary and for reducing the number of generators when necessary--all in order to maintain grid frequency within a very tight tolerance.

All the other generators connected to the same electrical grid "should" (read "must") have their governors configured to operate in Droop Mode. Droop governor mode is not as concerned about maintaining frequency as Isoch mode; in fact, the word "droop" means that the governor is going to allow the speed (frequency) to droop from its setpoint--because the frequency is being controlled somewhere else (by an Isoch unit or the grid operators).

When you are increasing the load being 'supplied' (really, being accepted) by a gas turbine operating in Droop mode, even though you are watching the MW meter what you are really doing is telling the SpeedTronic to increase the speed of the unit by increasing the fuel being admitted to the unit. However, if the unit is connected to an AC grid with many other generators--or even with just one other generator whose governor is operating in Isoch mode--the speed/frequency of the generator being loaded doesn't change, and that additional increase in fuel becomes additional torque input to the generator which becomes increased amps flowing out of the generator. So, you're not changing the load setpoint as you load/unload a unit--your're changing the speed setpoint, but the speed isn't changing because it's being controlled somewhere else on the grid.

At the same time you are loading a unit in Droop mode, if no other changes are made the grid frequency will start to increase. So, the Isoch unit must back down its torque output to its generator to maintain the frequency, which means the load of the Isoch unit will drop as the load of the Droop unit is increased. (Same thing when you unload a Droop unit--the Isoch unit must increase its load to make up for that being "shed" by the Droop machine, otherwise the grid frequency will begin to drop.)

Both of these governor modes--Isoch and Droop--are speed control modes. Both governor modes increase or decrease the load being supplied by the unit in response to changes in speed setpoint. Isoch governors very quickly respond to very small changes in frequency; their setpoint is grid frequency (remember, generator rotor speed is proportional to generator frequency). Droop governors don't respond very quickly to speed setpoint changes--the design of the system is such that another unit/"entity" is in control of the frequency. In fact, at rated power output most prime movers operating in Droop mode are operating at a speed that is 4- or 5% less than the setpoint. In other words, the actual speed is being allowed to "droop" below the speed setpoint--hence the term "droop speed control."

Think of a train with multiple engines being individually controlled, each by a train engineer, pulling a load and being run at a constant speed across a flat plain. To maintain that speed, the train engineers have adjusted the throttles of the engines such that the speed of the train is not changing. Now, imagine that a train car is "added" to the train, or that it starts up a hill--in either case, the load on the engines increases and if all the train engineers increase their engine power outputs, the train will start to speed up excessively. If just one train engineer increases his engine's power output to maintain the speed, then train speed will remain constant.

If the train starts down a hill (or train cars are "removed"/disconnected), the speed would tend to increase. If all the engineers reduced the power output from their engines the train might slow excessively. However, if just one engineer reduced his engine's power output it would be easier to maintain the train's speed. If too many cars were "removed"/disconnected and the one engine's minimum power output was reached, then the train's speed would start to increase--unless the power output of one or more of the other engines were reduced to allow the one engine to control the train's speed.

If the train engines have governors, one of them can be configured to be the "Isoch" unit and the others to operate in "Droop" mode. The Isoch engine will adjust it's power output to maintain speed as the terrain changes (or as train cars are "added" or "dropped off"). If the load increases such that the Isoch engine reaches its maximum power output and the hill becomes steeper (or more train cars are "added"), the train speed would tend to decrease. But, if one of the other engine's is "loaded" then the Isoch unit can back off in order to respond to any further increases in load.

You can also think of it as several people working together to hold a heavy beam at a certain height. If one person lifts more than the others, the beam will rise and the others will have to reduce their share of the load to maintain the height. It's all very similar.

Remember, all those who are reading this and possibly taking exception: You are all free to explain it in your terms (just please don't unduly criticize this writer's descriptions unless something is grossly in error!).

Your use of this site is subject to the terms and conditions set forth under Legal Notices and the Privacy Policy. Please read those terms and conditions carefully. Subject to the rights expressly reserved to others under Legal Notices, the content of this site and the compilation thereof is © 1999-2014 Nerds in Control, LLC. All rights reserved.

Users of this site are benefiting from open source technologies, including PHP, MySQL and Apache. Be happy.


Fortune
Warp 7 -- It's a law we can live with.
Advertise here
Advertisement
our advertisers
Help keep our servers running...
Patronize our advertisers!
Visit our Post Archive