Facebook
Google
GitHub
Linkedin
Open Loop vs Closed Loop
- Thread starter Chemsouhd
- Start date
Chemsouhd,
In later Mark V turbine control panels, and all Mark VI and -VIe turbine control panels, FSRACC, FSRN & FSRT all have a reference other than fuel valve position (and I'm referring primarily to natural gas fuel). For FSRACC the reference is acceleration rate, TNHAR. So, when the unit is accelerating the Mark* positions the fuel valve to maintain the acceleration rate reference--to make TNHA equal to TNHAR.
For FSRN the reference is turbine speed, TNR. The Mark* positions the fuel valve(s) in proportion to the error between the turbine speed reference, TNR, and TNH. The larger the difference, the more fuel is required (the gas valve(s) are opened more); the smaller the error, the less fuel is required (the gas valve(s) are closed more).
For exhaust temperature control, Based Load or Peak Load, the reference is exhaust temperature, TTRX, and the gas fuel valve(s) are positioned to make TTXM (the actual exhaust temperature) equal to TTRX.
For FSRSU_FI and FSRSU_WU, there is no "reference" in the sense of acceleration or speed error or exhaust temperature; it's just valve position and nothing else. The fuel valve(s) are sent to positions that <i>should</i> equal a factory calculated fuel flow-rate that is sufficient to establish flame or warm up the combustors and hot gas path after flame is established. If the LVDT calibrations are incorrect, or the fuel make-up (heating values; BTU content) are different from those used by the factory to make the valve position calculations then there may not be enough fuel to establish flame (or there may be too much!), or there may not be enough fuel to maintain flame during warm-up (or there may be too much!).
So, during firing and warm-up the gas fuel control valves are sent to positions, and the feedback is simply position--not acceleration or speed or exhaust temperature. FSRACC, FSRN & FSRT all position the fuel valve(s) to whatever position is required to make the feedback (acceleration rate; speed; exhaust temperature) equal to the reference--whatever the position is that makes the feedback equal to the reference. FSRSU_FI & FSRSU_WU just put the valve at some position and change the servo current to keep the valve position (LVDT feedback) equal to the position reference. (What would one use for feedback when trying to establish flame or warm up the combustors and hot gas path after flame is established?)
Liquid fuel is different only the sense that FSR is converted to an actual liquid fuel flow-rate, not a liquid fuel control valve position. The liquid fuel control valve is moved to whatever position it needs to be at in order to make the actual liquid fuel flow-rate equal to the liquid fuel flow-rate converted from FSR. Everything else is exactly the same as described above for all the FSRs.
It surprises many people--and shocks some--to learn that fuel control valve position is not a function of actual gas fuel flow-rate reference or feedback. Many units monitor gas fuel flow-rate but it's NOT used in any fuel valve control calculations. Gas fuel control valves and fuel nozzles are sized based on calculated flow-rates based on expected gas fuel characteristics--so if the gas fuel make-up changes over time, then problems can begin to occur. All gas fuels are not the same.
In later Mark V turbine control panels, and all Mark VI and -VIe turbine control panels, FSRACC, FSRN & FSRT all have a reference other than fuel valve position (and I'm referring primarily to natural gas fuel). For FSRACC the reference is acceleration rate, TNHAR. So, when the unit is accelerating the Mark* positions the fuel valve to maintain the acceleration rate reference--to make TNHA equal to TNHAR.
For FSRN the reference is turbine speed, TNR. The Mark* positions the fuel valve(s) in proportion to the error between the turbine speed reference, TNR, and TNH. The larger the difference, the more fuel is required (the gas valve(s) are opened more); the smaller the error, the less fuel is required (the gas valve(s) are closed more).
For exhaust temperature control, Based Load or Peak Load, the reference is exhaust temperature, TTRX, and the gas fuel valve(s) are positioned to make TTXM (the actual exhaust temperature) equal to TTRX.
For FSRSU_FI and FSRSU_WU, there is no "reference" in the sense of acceleration or speed error or exhaust temperature; it's just valve position and nothing else. The fuel valve(s) are sent to positions that <i>should</i> equal a factory calculated fuel flow-rate that is sufficient to establish flame or warm up the combustors and hot gas path after flame is established. If the LVDT calibrations are incorrect, or the fuel make-up (heating values; BTU content) are different from those used by the factory to make the valve position calculations then there may not be enough fuel to establish flame (or there may be too much!), or there may not be enough fuel to maintain flame during warm-up (or there may be too much!).
So, during firing and warm-up the gas fuel control valves are sent to positions, and the feedback is simply position--not acceleration or speed or exhaust temperature. FSRACC, FSRN & FSRT all position the fuel valve(s) to whatever position is required to make the feedback (acceleration rate; speed; exhaust temperature) equal to the reference--whatever the position is that makes the feedback equal to the reference. FSRSU_FI & FSRSU_WU just put the valve at some position and change the servo current to keep the valve position (LVDT feedback) equal to the position reference. (What would one use for feedback when trying to establish flame or warm up the combustors and hot gas path after flame is established?)
Liquid fuel is different only the sense that FSR is converted to an actual liquid fuel flow-rate, not a liquid fuel control valve position. The liquid fuel control valve is moved to whatever position it needs to be at in order to make the actual liquid fuel flow-rate equal to the liquid fuel flow-rate converted from FSR. Everything else is exactly the same as described above for all the FSRs.
It surprises many people--and shocks some--to learn that fuel control valve position is not a function of actual gas fuel flow-rate reference or feedback. Many units monitor gas fuel flow-rate but it's NOT used in any fuel valve control calculations. Gas fuel control valves and fuel nozzles are sized based on calculated flow-rates based on expected gas fuel characteristics--so if the gas fuel make-up changes over time, then problems can begin to occur. All gas fuels are not the same.
Dear All
first of all thank you Mr CSA for you explanation which i find very accurate and i want to confirm that by sharing an experience about the open loop of the FSRSU (start-up FSR).
Our machine after Firing and warm-up it started accelerating (which means the gas valve started opening from the FSKSU_WU value to the FSKSU_AR value with a ramp fixed by the acceleration ramp rate FSKSU_IA. and this loop has no feedback of the HP speed TNH, so once we reached the 60% of the hp speed and the starting system disengaged the HP shaft started losing speed (even though the valve continued its opening rate normally) and the machine tripped by flame loss.
After inspection we found that the intervalve pressure was the problem and it was lower than it should be, so the fuel flow rate was not enough to increase the shaft speed combined with the stroke of gas valve as it was calculated by the OEM.
The whole point here is that while the HP shaft was loosing speed during start up, the machine was not responding to that (exactly the gas valve), the GCV was following a startup ramp continuously, in an other words it was blind of the HP speed.
I hope this helps understanding the situation.
waiting for your feedback.
first of all thank you Mr CSA for you explanation which i find very accurate and i want to confirm that by sharing an experience about the open loop of the FSRSU (start-up FSR).
Our machine after Firing and warm-up it started accelerating (which means the gas valve started opening from the FSKSU_WU value to the FSKSU_AR value with a ramp fixed by the acceleration ramp rate FSKSU_IA. and this loop has no feedback of the HP speed TNH, so once we reached the 60% of the hp speed and the starting system disengaged the HP shaft started losing speed (even though the valve continued its opening rate normally) and the machine tripped by flame loss.
After inspection we found that the intervalve pressure was the problem and it was lower than it should be, so the fuel flow rate was not enough to increase the shaft speed combined with the stroke of gas valve as it was calculated by the OEM.
The whole point here is that while the HP shaft was loosing speed during start up, the machine was not responding to that (exactly the gas valve), the GCV was following a startup ramp continuously, in an other words it was blind of the HP speed.
I hope this helps understanding the situation.
waiting for your feedback.
Isulamu,
As I prefaced my first reply to this post, later Mark Vs, Mark VI and VIe turbine control panels used acceleration rate for acceleration after warm-up was complete, comparing TNHAR to TNHA.
Early Mark V turbine control panels operated as you described--the acceleration was presumed to be increasing if all operating parameters were normal.
As I prefaced my first reply to this post, later Mark Vs, Mark VI and VIe turbine control panels used acceleration rate for acceleration after warm-up was complete, comparing TNHAR to TNHA.
Early Mark V turbine control panels operated as you described--the acceleration was presumed to be increasing if all operating parameters were normal.
CSA
I haven't understood what you mean by early Mark Vs, because what I have described is the same in Mark II to VIe. After warm up is complete, the machine starts acceleration following the VCE start-up (Mark II) or FSRSU (open loop) for latest Control systems, but in the same time if the acceleration rate of the HP exceeds the HP acceleration limit TNHAR value the FSRACC would limit the fuel (closed loop). The only difference in Mark II, the fuel signal VCE would be driven to the minimum value directly and not limited like the FSRACC.
I haven't understood what you mean by early Mark Vs, because what I have described is the same in Mark II to VIe. After warm up is complete, the machine starts acceleration following the VCE start-up (Mark II) or FSRSU (open loop) for latest Control systems, but in the same time if the acceleration rate of the HP exceeds the HP acceleration limit TNHAR value the FSRACC would limit the fuel (closed loop). The only difference in Mark II, the fuel signal VCE would be driven to the minimum value directly and not limited like the FSRACC.
Isulamu,
I don't have access to any CSP or application code at this writing, but in my experience about halfway through the production of the Mark V turbine control panels the acceleration reference after warm-up was changed to use TNHAR instead of TNHA, and FSRACC was used as the controlling FSR. If I recall correctly (and I may be wrong--I have been once or thrice before!), the units for FSKSU_IA and FSKSU_AR were not %/sec (% TNH/sec) or %/min (% TNH/min).
And, after it was changed in the Mark V, it was also duplicated in the Mark VI and Mark VIe. Prior to the change in Mark V, the acceleration was not controlled, in the sense that the fuel was just increased at a rate (predetermined at the "factory") to cause the unit to accelerate at a desired rate (which was never specified). After the change, the acceleration rate was specified in an array (can't recall the name of the array at this writing), and fuel was increased or decreased to maintain the rate in the various array values (some of which were the same value).
The FSKSU_IA and FSKSU_AR values were left in the code (CSP; application code), but they didn't have the same effect on control. Just because they still exist in the code doesn't mean they are used, or that they are used in the same way in all versions of Mark* control panels (isn't this fun????).
Again, I've been wrong before, but as recently as last year I adjusted the TNHAR array to increase acceleration rate (decrease time from warm-up to FSNL)--and that wouldn't have been easily possible with the CSP in early Mark V turbine control panels ("Mark Vs"--plural). The acceleration rate in the early Mark V panels was always an approximation and never very precise; it got much more precise and predictable with the change to using TNHAR and TNHA to control acceleration after warm-up. And, I believe the two FSKSU Control Constants you keep mentioning just were used as limits after the change, whereas before they were used to increase FSRSU to try to make TNH increase at some pre-determined, but not specified, rate.
I can't get to more documentation for a few days.
I don't have access to any CSP or application code at this writing, but in my experience about halfway through the production of the Mark V turbine control panels the acceleration reference after warm-up was changed to use TNHAR instead of TNHA, and FSRACC was used as the controlling FSR. If I recall correctly (and I may be wrong--I have been once or thrice before!), the units for FSKSU_IA and FSKSU_AR were not %/sec (% TNH/sec) or %/min (% TNH/min).
And, after it was changed in the Mark V, it was also duplicated in the Mark VI and Mark VIe. Prior to the change in Mark V, the acceleration was not controlled, in the sense that the fuel was just increased at a rate (predetermined at the "factory") to cause the unit to accelerate at a desired rate (which was never specified). After the change, the acceleration rate was specified in an array (can't recall the name of the array at this writing), and fuel was increased or decreased to maintain the rate in the various array values (some of which were the same value).
The FSKSU_IA and FSKSU_AR values were left in the code (CSP; application code), but they didn't have the same effect on control. Just because they still exist in the code doesn't mean they are used, or that they are used in the same way in all versions of Mark* control panels (isn't this fun????).
Again, I've been wrong before, but as recently as last year I adjusted the TNHAR array to increase acceleration rate (decrease time from warm-up to FSNL)--and that wouldn't have been easily possible with the CSP in early Mark V turbine control panels ("Mark Vs"--plural). The acceleration rate in the early Mark V panels was always an approximation and never very precise; it got much more precise and predictable with the change to using TNHAR and TNHA to control acceleration after warm-up. And, I believe the two FSKSU Control Constants you keep mentioning just were used as limits after the change, whereas before they were used to increase FSRSU to try to make TNH increase at some pre-determined, but not specified, rate.
I can't get to more documentation for a few days.
| Thread starter | Similar threads | Forum | Replies | Date |
|---|---|---|---|---|
| A | Closed Loop/Open Loop Capture Buffer EX2100e | Power Generation | 1 | |
| A | Simple Examples of Open Loop and Closed Loop Control | Process Control | 4 | |
| G | Open Loop Control System | Sensors & Calibration | 1 | |
| L | Open loop control tuning? | Motion Control | 4 | |
| T | Open Loop Temperature Controller | Process Control | 2 |
