In our power block we have 2 9FA CTG, 2 HRSG and 1 STG.
When we shut down unit "A"
unit "B" trip due to HHH HP drum level and also STG trip too.
I had done too many researches about that situation, some of them are listed below and none of them can be the reason.

If you have any more ideas, or if you need any more detailed information(of course you'll need many more ) I'll be glad to hear from you.

my findings are as listed.
case -1-
Runback to unit "A" and it goes to S/D.
Unit B trip due to HHH HP Drum level and STG trip to.
1-During runback, STG CTRL valves seems working good. They are throttling as usual.
2-Pressure of both HRSG's are declining as normal.
3-Unit A's "Hp steam to STG valve" close than after 20 seconds unit "B" trips due to HHH HP drum level.

these information just only for a start to this topic, I can supply more and more information if asked.
I want to attach a graph about this incident but I dont know how to do that? If anybody knows, glad to know about that.

Regards...

 

Is this a new problem? Were shutdowns of one unit previously successful, without tripping the other unit? If so, what has changed?
Are you using an autosequence program to shut the unit down? Is the plant control GE ICS (Mark VI or Mark VIe), or is it a standard DCS (Foxboro, ABB, Emerson)?

If you manually run back the load on the unit being shutdown, or if you shut it down by manually giving a stop command to the gas turbine control, it may unload too rapidly. As the gas turbine unloads, the pressure in the HRSG decays. This is OK for the unit being shut down because the heat input to the HRSG is also decreasing. However, the heat input to the other unit remains the same, but its pressure also decays. As it decays, the tendency is for the drum to swell which is what is likely why you are getting the HHH trip.

Do you have an active start-up blowdown on the HP drum? GE's current practice is to use a control valve for the HP start-up blowdown with a variable stroke limit based on HP drum pressure and then leaving the MOV isolation valve open so the valve can be used to reduce drum level under these conditions (and not just during the start-up swell).
You could also try switching the remaining unit to single element level control during the shutdown of the other unit which will give the control a more rapid response to level changes. But don't forget to switch back to 3 element control after it stabilizes.

 

i don't have more experience of power plant, but from your observation it is may be due to stg control or control system sluggish operation. after one hrsg tripping stg control valve should close partly to maintain pressure. if it not close then steam flow from running hrsg may increase and drum level may increase.

 

HP Drum Level Trip

The main reason why a sudden "swell" in the drum level occurs is because of a sudden drop in pressure of the header the Boiler is supplying to. This sudden drop in pressure is mainly due to a steam consumption increase.

I am surprised to hear such incidents happening as operators are present in the control room to prevent such occurrences. Simply giving a momentary IBD would have been enough to stop the drum level from going so high.

Please give answer to the following quiries

1. what is the normal drum operation level ( it will be most probably around 45-60%)

2. what is the drum level high trip ( around 80-85%)

3. DO you have a compensated drum level measurement (density correction for pressure and temp)

4. what is the capacity of the HRSG ( pressure , temp , t/hr)

5. what is the steam turbine rating (MW)

6. what is the average steam consumption /MW for stg( for a fully condensing turbine it will be in the range of 3-3.5 t per MW)

7. Is there a sudden loss in the pressure header the HRSG's are supplying to when the HRSG1 main outlet valve is closed

8. Is there is a sudden steam output increase in the HRSG2 (which is online) when the block valve of the HRSG-1 is closed

9. what was the steam turbine load when the block valve of the HRSg-1 was closed. the stG MW should have been reduced to the steam availability. suppose that the HRGS-2 is capable of supplying only 90T/hr then assuming a steam consumption /MW of 3T/MW , the STG load should have been around the 30MW range before the HRSG-1 block valve is fully closed.

10. do you have a automatic coordinated system / semi automatic for this or is this done manually.

11. How is the STG operated in your plant , combined cycle power plants have something like a "boiler follow" mode for STG control , the steam inlet valve is controlled only for maintaining a proper header pressure. is this how it is operated ?

12. Was there any steam flow from the HRSG-1 when the main outlet valve was closed (is main valve closure automatic or manual) ?

I know this is a lot of questions , but once you start answering yourself these questions , i believe that you would have found out the answer or at-least quite near to finding out what happened during the incident.

 

Tripping of HRSG-B is due to HP Drum Level Hi Hi, it would occur due to more steam drawn from unit -B during the process of shutdown of unit-A.

Pl. have a look into the calculation of mass balance for both HRSG & STG with respect to MW & steam production. Whether load (MW) on STG was reduced to a level which unit-B can only supply?

If load (MW)on STG kept on higher side it will try to take more steam from steam header hence more steam drawn from unit-B (reason for "swelling") , which was online at that time.

Also pl. have a look for correct functioning of HP drum level control system system (Three element) for HRSG-B.

 

mxromen,

It would be helpful to know what the control system is and who provided it. Also, something I forgot to ask before is whether or not you have any load control active at the time you are shutting down unit A. If you were shutting down from a part load condition, where Unit B was at less than base load, the load control may try to load up unit B to base load while you are shutting down Unit A, which could, in turn, make any drum level instability worse.

If the shutdown sequencing was designed by GE, I would expect it to follow steps 1 and 2 as you described.

Step 3 would be Unit A reaches what GE calls "base low flow" condition. This is minimum IGV angle, which is the minimum load before exhaust temperature (and hence steam temperature) will begin decreasing. At this point the HP steam isolation valve would close (IP and LP isolation valves may also close at this time, depending on the piping arrangement). Unit A's HP and HR bypass valves would also operate as necessary to limit the bottled up steam pressure. Once the steam isolation valve is closed, Unit A would continue unloading and go into fired shutdown after the generator breaker trips on reverse power.

Unit B feedwater control should be able to keep the drum levels under control during this, assuming the controls are properly tuned.
This is why I asked if the system has ever worked correctly, and if, what has changed since then?

You stated that you have a graph. You could try sending it to me via e-mail to:
edoengr at aol.com (delete the spaces and replace the word at with the symbol @).

 

Dear Otised,

This problem is not a new problem for us in fact.
We are suffering these kinds of STG trip for two years.
But our advantage is, we just trip three times because of HHH HP drum level.

When unit "A" trips or runback, our unit "B" HP drum goes to HHH and trip the STG, but in regular shutdowns of unit "A" we have no fluctuation neither in Hp nor the other drums of unit B. Also when we shutdown or trip or runback to the unit "B" nothing happens to unit "A" HP drum levels.

We don't have any autosequence program. We shut the machines down by operators.

Our gas turbine control systems are Mark Ve, balance of plant is controlled by DCS and STG is controlled by ABB Advant.

You are completely right about your scenario, we have totally 5 power blocks and totally 10 HRSG, all the equipments are same and control systems are same too. But we don't face with a problem like that at none of the other units.

In normal Shutdown activities we have no problem with the levels so we didn't need to switch it to single drum level. Although we switch the unit to single element, In case of unit "A" trip, but it didn't work and unit "B" went to HHH and STG trip again.

Also thanks for your efforts to help.
Regards..

 

Dear Mxromen
So if I understand correctly this problem only seems to occur when unit A trips, or is in a runback situation? Correct?

If that is the case I would be looking at controls that are particular to unit B HP drum level controls and pressure controls. For my understanding of our STG which is also controlled by an ABB Advant system, the HP pressure control can operate automatically in several different manners. Typically during normal operations the HP control valve is open 100% and the STG accepts all available steam from the HRSG's, as unit loads are reduced or if a unit trips the HP control valve throttles to maintain a programmed HP header pressure. This system should function the same way for each unit, and should not operate differently if one unit versus the other trips. I would be curious what your header pressures and drum pressures are for unit A and B during steady state operation, and what happens to them during a trip or runback of either unit.

I would be studying trends of HP header and drum pressures as well as feed rates for HP drum B to understand what is occurring. Is it really drum swell due to pressure decay, or could some be related to feedwater flow rates as well?

I would not think that the CTG controls have much to do with this situation. As well the STG controller does not seem like a likely problem since it handles a trip from Unit B ok, but not unit A. It really sounds like something is not controlling properly for the Unit B HP drum. I will be curious as to what you are able to find, please keep us updated on your investigations.

 

Dear process value,

Thanks for your detailed response.
But be sure, I’m more surprised to have an incident like that, but it seems this does not related to the operators. When I'll sent a graphic about that incident you’ll understand what I'll trying to say..

Now its time to respond your hard and detailed questions..

1. Our normal drum operation level is %50

2. Hp drum high level trip %80

3. we don't have a compensated drum level measurement (density correction for pressure and temp)

4. Capacity of the HRSG: HP pressure: 120bar Hp Sat Stm temp: 565 Celsius, 250 t/hr for each HRSG)

5. Steam turbine rating is 281MW

6. Average steam consumption /MW for stg is 2.9 t/MW

7. There is a sudden loss in the pressure header the HRSG's are supplying to when the HRSG1 main outlet valve is closed.
When main outlet valve of unit "A" start to close, unit "B" Hp drum pressure: 111bar, Hp drum lvl: 68

Main outlet valve of uit "A" fully closes in 40 seconds and unit "B"pressure drops to 101bar, Hp drum lvl :372

After 16 seconds Unit "A" main outlet valve close, unit "B" hp drum pres: 101 bar, Hp drum lvl: 665 and STG tripped.
Everything happens in 56 seconds.

8. There is a sudden steam output increase in the HRSG2 (which is online) when the block valve of the HRSG-1 is closed.
In normal operation hp stm flow is 250 t/h, but when the block valve of the HRSg-1 is start to close, it increases to 300 t/h. Block valve totally closes in 40 seconds, during that period steam flow reaches to 400t/h and exceeds that point for a few seconds that we can see as "Bad Input".

9. At the moment of HRSG-1 block valve close: STG @ 206 Mw , HRSG-2 Hp stm is more than 350 t/h. This seems a crucial case to work on it. I'll make a research about this case tomorrow and will inform you ASAP about the results..

10. You said "do you have a automatic coordinated system / semi automatic for this or is this done manually." But I couldn't clearly understand what you mean.

11. Our STG runs as boiler follow mode. When a unit's Hp steam to STG valve closes, than STG start to throttle to obtain required pressure. I couldn't observe any abnormalities during throttling.

12. There is not any steam flow from the HRSG-1 when the main outlet valve was closed (is main valve closure automatic or manual)? Main valve closes automatic during any trip of course, but in S/D operations we close it from DCS manually.

Thanks for your outstanding effort for help.
Regards..

 

I have already upload a short summary about the trip event.
you can reach it from the link below...

http://oi54.tinypic.com/nbf5o2.jpg

I'll upload more graphs about the incident.
If required I can prepare and upload any different graph.

 

Also you can access to the Hp drum level transmitters behaviour as a graph on the link below.

http://oi54.tinypic.com/5etxs3.jpg

Also I want to write about, one more critical point.
During these trips, we observe hp drum level from the "yarway" also. We realize that level on the "yarway" was normal. We can say level on the "yarway" was as the blue line on the graph.

In normal operations, "yarway" and drum transmitters are working healthy.
I'm in a great paradox about that case

Any more graphic required, please inform me ..

 

> 3. we don't have a compensated drum level measurement (density correction
> for pressure and temp)

Dear Mrxomen,

Regarding the above statement, I would be extremely surprised if your selected drum level is not corrected for density, based on drum pressure. (Since the drum is at saturation conditions except when cold, either pressure or temperature is sufficient to calculate density.) It may not be obvious when you look at the control program, but I am sure it is there. You said you have an ABB Advant control. I have used ABB Advant on one project a while ago, and they used a table lookup function based on drum pressure for the density calculation. I checked it and found it to be quite accurate.

Now, for some suggestions for your problem. Since this issue is only occurring on one of ten identical HRSG's, all with identical controls (per your earlier posting today), I am inclined to believe you may have a problem with the drum internals on the one HP drum.
I looked at your graph showing the 3 level transmitters and the selected level, along with a logic signal for the steam isolation valve. This clearly shows 2 transmitters indicating high level and 1 transmitter showing a steady level. The 2 transmitters begin showing a level increase somewhat before the isolation valve starts to close. It also shows that all 3 transmitters tracked together up until just before the isolation valve started to close, and returned to tracking together after the trip on HHH level.

The usual arrangement of the level transmitter taps on the drum is to have 2 sets of taps on opposite sides of one end of the drum and 1 set of taps at the other end of the drum. The sensor taps for the Yarway device is usually at the end with the single transmitter taps.
You mentioned that the Yarway device seemed to follow the 1 transmitter that stayed fairly steady. I am guessing that the transmitter that stayed steady is the one at the same end of the drum as the Yarway sensor, and the 2 that went high level are at the other end of the drum. If so, you may have a problem with drum internals (baffles maybe?) on this drum. This can cause a transient difference in level at opposite ends of the drum during pressure disturbances. You may want to contact the HRSG manufacturer on this.
Further supporting this theory is your statement that you don't have a problem during a normal shutdown, but only during a runback or a trip of unit A. These are both cases where there would be a more rapid pressure change.

You also stated that when you had a runback on unit B, unit A levels were steady. If you have data on this event, you might want to look at a similar plot of the 3 level transmitters on unit B, to see if there is a similar pattern of 2 transmitters swinging together and 1 holding steady or swinging out of step with the other 2.

 

you seem to have got all the answers you need in the last post itself. I checked out the graphs and i found a few irregularities in the shutdown. I do not know if how your plant is operated but i did note the following

HRSG-1 main steam stop valve close command was given when the HRSG was supplying 190 tonnes of steam into the header. normally even if you down the GT it takes some time depending on the capacity of the boiler/HRSG flow to come to zero. Only after this should you close the main steam stop valve. in certain fast stop conditions the main steam header drain is opened simultaneously. do you have such a system. this needs to be looked into

The STG was at 230 MW !!! when the HRSG-1 valve close command was given , it is a normal procedure to reduce the STG load to near half value when one of the HRSG units are shutdwon or trips. It can be clearly seen that this is the major cause for the sudden reduction in pressure. you have mentioned that during HRSG-2 down you do not face such problem. This is a paradox. does your steam turbine throttle and reduce load quickly in the o9ther case. It seems unlikely and in any case it did not happen in the graphs you had posted.

the first step is to write an SOP in which you mention that the STG load is to reduced in case of a unit down. this will keep the steam demand under control and you will not face such problems in the future

i can help/guide you in writing a SOP for a shutdown. If interested you can contact me at [email protected]

 

Process Value,

Per Mxromen's posts, this is not supposed to be a full shut down of the power block. In fact, he said they did not have this problem during a normal shutdown of unit A, but only when unit A tripped or had a load runback (a rapid unloading). The intent is to just shut down 1 gas turbine/generator/HRSG and leave the other gas turbine/generator/HRSG and the steam turbine /generator running. For this to happen, the steam isolation valve on unit A is closed before steam flow drops to zero. It is closed when unit A's IGV's reach minimum angle. This keeps the steam temperature up and avoids quenching the steam turbine. It also keeps unit A HRSG pressurized and hot, allowing for a quicker restart. The steam bypass on unit A limits the pressure rise when the isolation valve is closed.

 

Sir ,
I understand that this is not a full shutdown of the plant , but only one GT and one HRSG is being shutdown. I think it was wrong for me to assume that there will be a NRV (usually a battery of boilers system will have one) in each HRSG. Plants i have been to , the normal "hot box" up procedure is to down the GT and wait for a at least 5 min , when the steam goes down to 5-10% of the full load value. close the main steam valve and open the main header drain to limit the pressure. Yes this reduced the steam temp. But during startup the GT is loaded , the HRSG steam is drained till it reaches 80-90% (2-3min) of the spec temp value and the main steam valve is opened. the plant automatic shutdown also follows the same procedure , i have never seen a 10-20 kg/cm2 drop in pressure ( as seen in the graphs of the trip) , normally pressure drop is a negligible 2% at max. The only time i have seen such drop in pressure and swell in drum level is during a steam blow off of a boiler during commissioning.

But i still believe that excessive steam demand is the culprit in this case. having a STG at 80-90% load during one HRSG down is not a good operating procedure as per my view. but again i do not know how this particular plant is operated, whether the turbine is capable of throttling automatically to a low load and maintaining pressure. Apparently in this case it did not. some combined cycle plants have a load limit function , which is a derived max stg load depending on the GT load. what this does is , it limits the STG load when a unit is being shutdown and even when a GT unit is unloaded it reduces the STG max loading capacity. this is given as a additional back up feature , even if the normal STG does not throttle on loss in pressure , this load limit will unload the turbine keeping the steam demand in check.

 

>I have already upload a short summary
>about the trip event. you can reach it from the link below...
>
>http://oi54.tinypic.com/nbf5o2.jpg
>
>I'll upload more graphs about the incident.
>If required I can prepare and upload
>any different graph.

If you have the same data for a trip of GT B then comparing these trends from that trip should show what is different.

Otherwise compare these trends to the same trip on the other pairs of GT-HRSG. By the same trip I mean the same trip alarm on the GT unit controls.

Regards, 4-20 Camp

 

Dear All,

We have already solved the problem, I'll explain how did it done after short time.

I'm too busy to explain it now, please wait for at most 5 days more.
and extra thanks to all for your efforts and guidance..
regards...

 

Could you please explain the you solve this problem? Do you use pressure compensated in your logic diagram?

Thank you.

 

Dear Sir,

Please tell us how u solved this problem.

Regards