TPP was parallel with grid. due to malfunction of grid signal, the DCS receives the grid open signal though the TG was synchronized with the grid. since the TG was parallel with the grid it keeps on reducing the load to maintain the frequency and finally y got tripped. Both the speed controller and load controller output keeps on reducing.

I could not know the exact reason of tripping. why the TPP didn't come on the home load

 

We don't have enough information about your plant and how it is configured, programmed and designed to operate.

What is clear is that someone doesn't understand how a power plant can "come on the home load" when it's being shut down while connected to a grid. From your description it seems something initiated an orderly shutdown of the plant, reducing load. If the plant was still connected to the grid then the grid is controlling the frequency, not the plant.

If the plant was separated from the grid, something would have to tell the turbine control system to adjust fuel to maintain frequency in response to the "home" load--if the control system was designed to be able to do so, and if the control system were capable of doing so. The plant electrical system would also have to be designed to allow independent operation of the "home" load while separated from the grid.

"Tripping" generally means an immediate, emergency shutdown. It's also not clear from your post if you are referring to an immediate, emergency shutdown or an orderly shutdown.

Most power plants and control systems are configured so that an alarm is generated for every condition which results in an immediate, emergency shutdown. If not a specific alarm, then if some protective relay operated (over-voltage; reverse current; under frequency; etc.) usually "flags" itself in some way so that an operator can determine that is has operated and could be the source of the trip.

If there we in fact no alarm annunciated and no immediate, emergency shutdown, but an orderly reduction of load and opening of the generator breaker, then it's likely that someone or something initiated a STOP sequence and the plant was just "following orders" (which is not usually an alarm condition).

You said that the DCS was told the plant was separated from the grid. Does the DCS tell the generator prime mover's governor (control system) that the plant is separated from the grid and to control the "home" load frequency? Does the prime mover governor have the ability to automatically switch to Isochronous speed control? Or, does the DCS issue a command to the prime move governor to try to control load to be able to maintain frequency?

There's just too much information you haven't provided, and likely can't provide in this kind of forum. But, it certainly seems there is some misunderstanding about how the plant is capable of operating and how a plant could be operated.

 

Hi,

Thanks for the reply. let me tell u what has exactly happen. our plant capacity is 23 MW and it was running on 23 MW .due to some fault the grid signal was lost(the digital signal at DCS),though the unit was not actually isolated from the grid.

In that condition the governing system maintain the frequency as it was not actually isolated from the grid and keeps on reducing the load and finally the unit was tripped.

the generator and turbine is of BHEL make and the governing scheme is provided in the DCS.

On checking the trends I found that both the load control and speed control output keeps on reducing after the grid signal was missed

could not detect the exact reason of tripping

 

kapil,

From your description of the load reducing and the load control and speed control outputs reducing it is exactly what would happen if someone initiated a STOP while on load. A STOP from rated output will reduce the load/speed control which will reduce the power being produced by the prime mover which will reduce the power being converted to amps by the generator. Eventually, if this continues the generator output will reach zero MW, and if it continues further the generator will go into what's called "reverse power" mode.

There should be a protective relay monitoring the generator output that will open the generator breaker if the generator output drops below a setpoint. Some steam turbines actually have a slight positive MW setpoint because they should never be operated in reverse power. Most single-shaft combustion turbines have a negative MW setpoint since they are less susceptible to damage if operated in reverse power for short periods.

Yes, opening the generator breaker is generally called "tripping", which is distinguished from "tripping" the prime mover by immediately shutting off the fuel or steam input to the prime mover in response to a serious problem.

You still haven't told us if you are referring to the opening of the generator breaker or the stopping of the prime mover driving the generator. If the prime mover is a steam turbine (another thing you still haven't told us) then the steam stop valve is usually closed at the same time the generator breaker is opened (tripped). If the prime mover is a gas turbine the unit usually goes into what's called "shutdown" mode after the generator breaker is open end (tripped), reducing fuel and speed in a controlled fashion, and eventually shutting off fuel and coasting down to cooldown mode.

So, it's not clear when you keep talking about "tripping" what you mean, precisely.

Your mission, should you choose to accept it, is to determine what caused the load/speed control to be reduced when the DCS was under the impression that the tie breaker was opened. We can't help you with that because we don't have access to the elementaries and drawings and programming for the control systems (prime mover governor and DCS) at your site.

Lastly, the governing system did not maintain frequency during the load reduction while the plant was still connected to the grid. The grid controlled the frequency of the plant. There have been <b>MANY</b> posts on control.com about speed control and frequency control (droop speed control and isochronous speed control) and how they work. Please use the 'Search' feature of control.com to refer to them for more information.

 

Kapil... for those situations during which sequential event alarms are not available I suggest using the idea shown in Control.com Archives. The thread # is 1264601170!

Regards, Phil Corso

 

Ok , i have seen a very similar tripping happening at a site. They used to run the 30MW STG in load control (preselect mode) when parallel to the gird. the plant had a home load of 4 MW and during a islanding (the breaker contact to the 110KV grid was given to the plant DCS, and the plant DCS after a time of 4 sec issued a command to the governor) the governor switched to isochronous mode of operation.

now the breaker contact to the DCS malfunctioned (the 24 V dc fuse to the axillary relay got blown off in the 110 kv yard. there was a retrofit going on at a adjacent relay panel), ie the DCS got the command that the 110 KV gird breaker had opened. the set point of the iso was a fixed 50hz but the grid frequency at the time of malfunction was around 50.15. the iso governor tried to reduce the frequency to no avail (grid was very much present). the input steam was cut off in a phased manner and the machine tripped in around 2 min.

do you have a similar system in the plant. then check this possibility.

i argued with them to change the automatic iso changeover to droop. so that in case of an islanding the machine can change over to droop. so that such incidents can be avoided. They were adamant on getting the iso mode and the logic was modified such that in case of a islanding the set point of the iso controller was equal to the grid frequency. this does not solve the problem given that the grid frequency can change either way after the trip but the management was happy .

 

First, it's not very common for the grid frequency to be so high in most parts of the world, but if it were high <b>and</b> the governing function of the DCS <b>did</b> switch to Isoch control then I would expect the load to drop very quickly as Isoch is PI control, usually heavy on the I function in order to try to maintain frequency. I would not expect the unit load to drop slowly under such conditions (switched to Isoch while still paralleled with a larger grid that was at a higher than normal frequency), unless the Isoch controller was not tuned properly or had been de-tuned because of other problems/issues. Again, steam turbines don't make great Isoch machines for quickly changing loads.

Second, if the unit were operating in parallel with a grid then it is most likely already operating in Droop mode, and so switching it to Droop mode when isolated from the grid would be, well, redundant.

Third, if the unit were to remain in droop mode when islanded and the load dropped from 30 MW to 4 MW the frequency would increase very much, probably so much that the overfrequency relay would trip the generator breaker.

And, when the plant is isolated from the grid, what difference does it make what the island/plant frequency setpoint is with respect to the grid frequency? None. The only time the island/plant frequency must be "coordinated" with the grid frequency is when the plant is being re-synchronized with the grid. So, setting the Isoch frequency setpoint equal to the grid frequency setpoint at the time of the isolation event doesn't do anything as the frequency setpoint will have to be modified during re-synchronization.

There's just too much we don't, and can't, know about the specifics of the plant and how it is configured and controlled and operated. And, it's still pretty clear that some assumptions about how plants operate and respond to islanding events are not valid.

 

> "First, it's not very common for the grid frequency to be so high in most parts of the world, but if it were high and the governing function of the DCS did switch to Isoch control then I would expect the load to drop very quickly as Isoch is PI control, usually heavy on the I function in order to try to maintain frequency. I would not expect the unit load to drop slowly under such conditions (switched to Isoch while still paralleled with a larger grid that was at a higher than normal frequency), unless the Isoch controller was not tuned properly or had been de-tuned because of other problems/issues. Again, steam turbines don't make great Isoch machines for quickly changing loads." <

In India 50HZ is the nominal frequency and during lightly loaded condition especially at the night frequency reaches and sustains at a value greater than 50 Hz, mostly at 50.2 to 50.3. as it occurs each night i would say it is quite common (at least in zone II southern grid). if you know anyone in operations please ask them they will concur with the fact. and 50.15 is NOT A VERY HIGH FREQUENCY by any standards. would you say 49.85 is a very low frequency? if not the case the reverse should also be true. yes iso mode is a PI controller, but the band in which it is PI is narrow (+/- 0.1 to 0.2 depending on how it is configured) anything more, the droop reference is changed before it is switched on to PI again). please have have a look at this

http://www.2shared.com/photo/tbLeClX2/load_setpoint.html

Steam turbines don't make good iso machines is not entirely true. yes there is a restriction in it as the boiler firing rate cannot be changed quickly, and sudden change in the load will cause the master pressure to hunt, raise in case of a load throw off and lower in case of a sudden loading of the generator. But in co generation plants where these small steam turbines are used it is not a problem. In refineries, an example would be chennai refinery, they have two steam turbines of 20 MW each and three boilers of 130 tonnes. nominally a steam rate of 4 tonnes is required for 1 MW loading of the machine. thus for a 20 MW load, 80 tonnes is sufficient, and they have excess capacity to carter to the process requirement. they are operating the stg in iso mode for now 20 years without any problem. i have seen them start 2.5 MW motors independently from the gird in iso mode. yes the master pressure drops but recovers quickly. As i said with sufficient steam cushion steam turbines can take up quickly changing loads. In large utility power plants connected to the grid the steam cushion is very low as they use it only for plant purpose (deareation, regenerative feed water heater etc) which is not much but they do not use iso much, if not at all.

> "Second, if the unit were operating in parallel with a grid then it is most likely already operating in Droop mode, and so switching it to Droop mode when isolated from the grid would be, well, redundant." <

in my answer i did say that they operated the plant in load control mode ( preselect mode in GE terminology). yes it is a extension of the droop mode but it is not droop mode entirely. in the plant they were operating the machine in load control with set point at 28MW . in droop mode with the change in frequency will cause the change in load when connected to the grid. if you put the machine to the grid and put it in droop, if the grid frequency increases the machine will supply less power to the grid and vice versa, this does not happen in load control, in load control the droop reference is changed to maintain the constant load. so changing over from load control to droop during the islanding is not REDUNDANT, there is a difference between the two. and yes you cannot and should not operate the a independent machine in load control. I personally prefer changing from load control to droop in case of islanding, i have seen the same logic implemented in many places and also in the 6 machines i have commissioned till date.

> "Third, if the unit were to remain in droop mode when islanded and the load dropped from 30 MW to 4 MW the frequency would increase very much, probably so much that the overfrequency relay would trip the generator breaker." <

all machines would have a droop chara of 4-5 %, the machine i talked about had a droop of 4%, so for a load change from 28 MW to 4 MW, assuming that the grid went off line, the frequency change will be only 3.2% (increase). By default there will only be a under frequency relay in the machine the over frequency relay if present will have a value of 6-8% which will be above the droop chara. Normally the overspeed protection set at 10-12% is by default the overfrequency protection of the machine. from what i have said above, the machine will not trip. there is a reason why droop is always set at 4-5% while overspeed protection at 8-10%, a full load throw off in the machine should not cause the machine to trip each time. "

> " And, when the plant is isolated from the grid, what difference does it make what the island/plant frequency setpoint is with respect to the grid frequency? None. The only time the island/plant frequency must be "coordinated" with the grid frequency is when the plant is being re-synchronized with the grid. So, setting the Isoch frequency setpoint equal to the grid frequency setpoint at the time of the isolation event doesn't do anything as the frequency setpoint will have to be modified during re-synchronization. " <

in the above indecent the difference in the frequency between the setpoint and the actual frequency caused the machine to trip, having the frequency set point the same as the islanding frequency. in the indecent let us say that the set point was also increased to 50.15 automatically, would the iso controller have taken any corrective action. no . this was their logic behind the change. and yes this would not solve the problem, nor is it a good solution to the malfunction of the grid breaker contact. as frequency would change either way in the grid. i have mentioned that in my post too.

> "There's just too much we don't, and can't, know about the specifics of the plant and how it is configured and controlled and operated. And, it's still pretty clear that some assumptions about how plants operate and respond to islanding events are not valid." <

The indecent i mentioned took place 8 months back, i confirmed with the friends working there, the machine tripped on low forward power in 1.47 min, during that time the load was shed in a phased manner. The steam turbine is of BHEL make, the plant DCS MAXDNA also of BHEL make and the turbine governor also of BHEL make.

 

Thanks to all of u,

well the problem is much the same as faced by the Mr. Process value.
the little doubt is what is the use of taking the grid breaker signal in the DCS. as in my case False indication has given a tripping of plant.

also why the plant has not come in the island mode. also what is the exact signal/DI by which the governor will assume that know the unit is isolated from the gird. as in my case when the grid breaker opens it has still assumed that it was coupled with the grid.

Plz reply

 

let us suppose that there is a generator of 20 MW capacity. the section load is say 10 MW and it is running at droop mode. not let us say that the machine is put in load control mode with setpoint at 10MW. if there is no load change then nothing will happen. now if the section load increases by 12MW , the machine in load control mode with setpoint 10MW will try to reduce the load so it cuts out the steam/fuel , however the section load will not drop and will remain at 12MW itself ( a small reduction will take place due to lowering of the frequency if there is a sizable motor load in the section) and after sometime the machine will trip on under frequency. the reverse happens for a load reduction in a independent section.

this is the reason why in all industrial generator sets a separate tie logic is written in the DCS or SCADA to determine the presence or absence of grid. and if there is a absence of grid , the generator must switch over from load control to droop or iso depending on how it is configured.

 

The grid frequency in India is famous in the rest of the world for it's fluctuations. It is not representative of the rest of the world, and as such machines don't typically behave like they would in other parts of the world. Over- and under-frequency relays in India are not set to more typical values used in other parts of the world. I have seen trends of automatic synchronization attempts performed in India that failed because the grid frequency couldn't be "caught" by the speed matching algorithm (essentially the fuel couldn't be adjusted fast enough to match the changing grid frequency). The auto synch would sometimes eventually occur but the watt and VAr swings were pretty atrocious until some load, sometimes as high as 40-60% of rated, could be achieved.

Yes; I personally would consider 50.15 Hz to be high, and 49.85 Hz to be low. And on small grids the Customers purchasing the machines I was commissioning would not accept such deviations without detailed engineering reasoning as to why the deviations could not be limited, even if they existed for months prior to the commissioning of the new machine.

On a GE-design heavy duty gas turbine with a Speedtronic control system in the typical configuration, Isoch control has a proportional deadband of approximately 0.17%, so from 99.83% to 100.17%. This would translate into a frequency deadband of 49.915 Hz to 50.085 Hz. Anything outside of this range would be reacted to by the integral controller to (attempt) return the frequency to normal. 50.15 Hz and 49.85 Hz are both outside of this range and so integral action would quickly come into play to drive the speed/frequency back to 100%, or to a least within the 49.915-50.085 Hz range.

The above numbers are for a GE-design heavy duty gas turbine with a typically configured Speedtronic turbine control system. The original poster never told us what kind of turbine or control system he had, and did not explain that the grid frequency was not fairly stable and nearly normal. Again, the poster provided very little in the way of information. I'm glad you, ProcessValue, were able to get more information out by providing a similar <b>incident</b> as an example.

I think we agree that steam turbines and their boilers can be decent isoch machines, but that depends on a lot of factors, as we both stated. Firing rates, rates of changes, expected load changes, all of these things contribute to stable operation.

Pre-selected Load Control in GE-design heavy duty gas turbines (again, which the original poster's turbine is not) is not designed to provide what the industry calls 'secondary frequency response.' And, again, the under- and over-frequency relays in India are probably not set to values used in many other parts of the world (industrialized and non-industrialized), and the original poster didn't give us any of that information in his early posts, either. Certainly, one cannot (though <b>MANY</b> have tried!) a unit in Isochronous mode in Pre-selected Load Control.

We should be responding to the original poster's question and situation, which was very poorly described in the original posting, and subsequently as well. My response was kind of generic in nature and assumed (yes, I made an <b>ass</b>-umption in this case) the unit was being operated in a part of the world where the grid frequency was relatively stable and the unit was most likely a heavy duty gas turbine (since most of the posts here are about heavy duty gas turbines). I purposely did not respond to the first posting by another (or the same?) individual with exactly the same wording because of the lack of information; and <b>no one else</b> responded to that posting either.

I still don't believe the original poster truly understands the difference between Droop speed control and Isochronous speed control, nor how a plant operates when it is connected to a grid in parallel with other machines and when it's not in parallel with other machines. I also don't believe the recent post by ProcessValue (22 Nov 2010, 11:11 PM) adequately defines the conditions (parallel operation with a larger grid; island operation) for the scenario being described. It is possible to operate a machine in droop mode when isolated from a larger grid (islanded) if the load fluctuations are minimal and predictable, but certainly it cannot be operated in Pre-Selected Load Control or Load Control and have a stable frequency unless the load is also very stable. And, that's not usually possible for refineries or cement plants or fertilizer plants or paper mills or steel mills where large motors are started and stopped many times throughout the day.

Certainly, if the turbine governor was switched to Isoch mode when the plant was still synchronized with the grid (and I don't believe the original poster has confirmed or can confirm that) and the grid frequency was higher than nominal and below the integral action setpoint (this presumes there is some "separation" between the Proportional and Integral actions as done in some turbine control systems), then the load would be reduced until the generator breaker opened ("tripped"). It would <b>not</b> stop at the "home" load because it can't know what the "home" load is unless something tells it what the "home" load is. It would just continue unloading as long as the frequency was above the setpoint. And the rate of unloading would be a function of how the PI control was programmed, and I don't believe the original poster understands or can determine what the parameters for that function are (at least they haven't been provided to date).

If the turbine governor were in some kind of load control when this event occurred (false separation indication) it would not seem that it should respond to a speed lower signal because it would seem it would be trying to maintain the load setpoint. Again, I think it's not correct to consider that the turbine governor at the original poster's site is programmed like a Speedtronic; while it may be similar it most likely is not identical, not in the exact configuration of the speed control PI controller nor the control parameters used in the PI controller.

And lastly, there's absolutely no way we can tell the original poster what triggers the DCS to know whether or not the plant is connected to the utility/grid. Someone at the plant should be able to determine that from the drawings and software in the DCS. And then how that tells the governor function whether or not to be in droop speed control (and/or load control) or Isochronous speed control; we can't tell the original poster that either because we aren't familiar with the control system in use.

In any case, there's too many unknowns and too much 'splainin' to be done. kapil's statements clearly indicate he's not familiar with Isoch control and island operation fundamentals and can't confirm whether or not the unit switched to Isoch when the false utility tie indication was given; at least he hasn't to date. Extracting information has been difficult; all he can say for sure is that his problem seems to be like the incident described by ProcessValue.

And egos should stay out of the discussion altogether. Experience and relative observations are fair game for comment and discussion.

 

Dear CSA

Let me say this first before i start off my topic , During my freshman year (2years back) in commissioning i learnt much more from your posts than what i could get from my seniors. i used control.com and the search feature regularly and i did not post nor ask anything as all the information to which i had my doubts were clarified by the old posts. only recently did i begin contributing as only now do i fell i am competent enough to post in this very erudite community. you had replied in detail to almost all aspects of heavy duty gas turbine engineering and my overly long answers are modeled on your overly long observations and solutions . the above post was in no way meant to offend you nor did it mean any disrespect. disagreement crops up due to my experience and the lack of it. i do not know how old you are. probably my best guess the late forties or early fifties; more or less my fathers age.

now back to the topic. yes grid frequency in India does vary a lot , any where between 49 to 50.5 Hz under most conditions and personally seen it reach 48.64 to 51.2Hz. i am uploading a small pic (taken from a 110KV switch yard in southern grid ) which will show the grid frequency variation in greater detail. as you have mentioned it varies a lot.

http://www.2shared.com/photo/twMHd-NG/grid_frq_india.html

but what i find in difficulty in believing is in other grids around the world it can be maintained in the tight frequency gap you have mentioned. with GW 's of power transferred and with wide variations in load i would say a 1% tolerance in frequency would be allowed anywhere in the world. this corresponds to a frequency range of 49.5 to 50.5 Hz. most of the utility providers in the world would only agree to such a commitment. it is very much possible to maintain a frequency in the range of 49.915 Hz to 50.085 Hz in an independent machine but in a grid with hundreds of gen sets it would be very difficult if not impossible task. i have been only to Sri Lanka outside my country and they have worse statistics than India but i found that in UK the legal limit is 49.2 Hz to 50.8 Hz. they have a online frequency meter which shows the last 60 Min data. i am including a screen shot of the page and the link here.

http://www.2shared.com/photo/3R_4yW9f/grid_frq_uk.html

as for Kapils case, i do not know how the plant is configured and operated. with noting much to go on i took a guess to what might have ha penned relating it to a incident i have faced. i am familiar with BHEL plant designs. their design philosophy and the tried and tested configuration they use in governor and DCS BOP control. however not being at site this is as best as i can do, give him points to ponder so that he can make a connection to what might have happened. for my response on (22 Nov 2010, 11:11 PM) it was to give a practical example on what would happen if you put a independent machine on load control.

i will summarize what i am trying to say here

Droop mode - CAN be used for independent operation ; CAN be used for operation with the gird

Load control/preselect mode - CANNOT be used for independent operation ; CAN be used for operation with the grid

ISOCHRONOUS control - CAN be used for independent operation ; CANNOT be used for operation with the grid

well i will add the AVR control modes also here

AVR ( voltage set point tracking ) - CAN be used for independent operation ; CAN be used for operation with the gird

APFR (gen set PF set point tracking ) - CANNOT be used for independent operation ; CAN be used for operation with the grid

Reactive power control (MVAR setpoint tracking) - CANNOT be used for independent operation ; CAN be used for operation with the grid

this scheme can be used to operate any industrial / utility network. to my best of knowledge this is correct.

Plant connection to the gird and isolation from the grid is ALWAYS taken from breaker contacts. separate Tie logic is written for the same. i will give an example here. this logic is implemented in the plant DCS/SCADA or in hardwired axillary relays.

http://www.2shared.com/photo/jHaaMKFj/sample_tie_network.html

this logic is then given to the governor and the AVR for detecting the grid presence / absence. this was what i was trying to explain to kapil , who well seems to be new to this area.

as an ending note i would like to thank you once again for your sustained effort over many years in educating and sharing knowledge with young professionals at control.com. looking forward to fruitful discussions (and arguments ) in the future .

 

Thanks CSA & P. Value for the valuable suggestions u have provided for the problem since I was out of station so not able to respond.

well i have got the solution of the problem. Let me tell u the philosophy in my DCS. the dcs is supplied by BHEL with steam turbine and all the logic is provided in the DCS. During coupled with the grid it runs in the load mode and follow the load controller output. when in the island mode it comes in the speed mode (frequency), but again it follows the load controller output. this is some what similar Process value has done at one of the site.

I will search for the separate logic of breaker as recommended by the P Value. Yes don't have much knowledge regarding the different modes, but keen to know abt. them, also it is not known to a normal maintenance er. as these are done at the commissioning of the plants.

but once again thanks to both of u for the work which u r doing

 

dear all,

it appears that the control system has received False signal that the machine has been Isolated from Grid. Now, I think that you have only one Turbine-Generator in your plant Local Grid (Island). When Generator is connected to Grid, Frequency will be governed by Grid Frequency. However, when It receives the signal (even False Signal) that it has been Isolated from the grid, then Frequency Controller will come into picture and try to control the Power frequency to set point (may be 50Hz. in your case). It appears that in your case, when False signal has been reported, Frequency controller came into service and at that time Grid Frequency was more than the set value. Control system was trying to reduce the Load , to bring down the Frequency. But as ACTUALLY machine was connected to Grid, there was no change in Frequency and Control system kept on reducing the Load and ultimately TRIP occurred.

Based on your experience, why don't you take 2 such signals (Grid Isolated/Tie Breaker) and use them in AND gate , so that such spurious signal occurrence can be avoided.

 

dear mr.kapil,

please try to see Generator Protection logs, if available. it appears that, as the load kept on decreasing, either Low Forward Power Relay or Reverse Power relay has operated and Tripped the Generator.

sandeep

 

Glad to be of help . and thanks for the feedback. and yes write more of what you have done once you find the complete solution. as for regarding droop and iso mode, look at this thread

http://www.control.com/thread/1287783101

i have explained how the speedtronic turbine control is used for droop and iso mode. most of the turbine controls will be similar to it.

 

I would suggest you set your generators in "Auto droop" and use the set-point for the import/export grid power as a reference, rather than the frequency.

Should the in coming grid protection relay trip, you better have the load shedding function enabled to offload.

 

> TPP was parallel with grid. due to malfunction of grid signal, the DCS
> receives the grid open signal though the TG was synchronized with the grid. since
> the TG was parallel with the grid it keeps on reducing the load to maintain
> the frequency and finally y got tripped. Both the speed controller and load
> controller output keeps on reducing.

> I could not know the exact reason of tripping. why the TPP didn't come on the home load

Did you mean you have bad frequency signal from grid?