power generation

E

Thread Starter

ESKAY

Dear CSV and mk6 experts,

ours ia a Fr5-Mk6 Alstom machine. Getting frequent alarm for the compressor inlet TC reading disagree(L86TCI). though ctif-1 and ctif-2 readings found on the screen is OK. May be there is an intermittent failure of reading in one of the TC. This alarm gets latched and we need to execute "Master Reset" to reset this alarm. while executing MR, the turbine load goes down OR up, by 1 to 2 MW, and then restores. The CTIF values are also used in CTIM-CQTC-IGV modulation- TTRX5-Trip History capture algorithms. L30FPG2L ALM also comes ON and gets reset later.

Kindly explain the reasons for this effect on LOAD and FPG low Pressure alarm. And what is input n/d means under CQTCV1 algorithm. Thanx...
 
B

Bob Johnston

If you take a look at your software, I think you will find that L3TCIFLT (Inlet temp fault) causes a change in CQTC (Comp. airflow temp. correction). When you Master Reset and L3TCIFLT clears, the CQTC value changes and this is probably causing your load change.
Sorry I can't be more specific, without looking at your software.
As for your Low Fuel Gas Pressure Alarm, it seems hard to believe that a 1-2 MW swing in load could pulse your fuel gas supply pressure low and cause the alarm but I would need to see your system to be sure. Are you running with a gas pressure which is near the low pressure alarm?
 
> Kindly explain the reasons for this effect on LOAD and FPG low Pressure
> alarm. And what is input n/d means under CQTCV1 algorithm. Thanx...

Hi,

I concur with you that the alarm L86TCI may be provoked by intermittent failure of reading of one TC. Similar cases were noticed on MKV systems: due to intermittent, internal MKV, communication disturbance sometimes are lost readings from the whole group (15 of 45) of thermocouples landed on TBQA card for one processor scan. In the case that readings are lost due to internal communication issue, value of temperature recorded by MKV for affected thermocouples is 0degF while in the case of the open circuit to thermocouple, reading is -78degC. Maybe on MKVI there is similar issue.

In typical GE code, alarm L30FPG2L is generated when P2 pressure is recorded below the limit [limit= FPRG (reference) - 10psi] for time longer than constant. On some sites this constant is 0.25sec and on some sites 2.5sec. Check the code on your site, it is probably 0.25sec. The alarm condition may happen when gas control valve opens very quickly and SRV does not respond quickly enough.
You write about load swing after MR. Arrange quick trend at the MR instant, recording at least: position of the gas control valve FSG1, SRV position FSGR, P2 pressure FPG2, FPRG, FSR, FSRN, FSRT, FSRMIN, CSGV, TNR, TNRL, DWATT, TNH, CQTC.

Result will give you explanation about P2 pressure alarm and load swing or will give further troubleshooting directions.

Regards
 
Dear sir,
Thanks for ur prompt reply to my thread. Sorry for the delay in replying.
Now our turbine running in GAS mode with 8.5 MW.The follwings are the values noted for various signals, with L86TCI alarm latched at logic"1".<pre>
CTIF1 = 72deg F CTIM=73 deg F
CTIF2 = 72 deg F FSR=40.8
CQTC= 1 n/d TTRXS=1039 degF
TTRXP= 980 deg F </pre>
CQTC is also used in FSRSU-FSRMINV2 and CTIM used in CSRGVP and TTRXV5 circuit.I,suppose, these cumulative circuit causes the reduction in LOAD and FPG2 pressure going low to the value of FPRGLMT.
What is n/d shown in CQTC ? And the value of constant CTK1 =25 degF Diff.limit for the CTIFs are same for Mk4 and Mk5.I think,sometimes CTIF1 value goes down momentarily,which causes L86TCI ALM.All the connections of CTIF TC circuit checked for loose contact and also TC itself. Kindly advice me further to solve this problem. Thanx..
 
Dear MR.GEM73,

Thanks for ur reply to my thread. Sorry for the delay in replying..

We have already one alarm for the "s" slot1VCMI (diagnostic Alarm) due to Bit error fault at the rack. This does not get reset with either "reset" button execution on the VCMI card OR with POWER on/off for the rack"S". You may advice to get rid of this ALm. Awaiting for the new VCMI card, to try to solve the problem. With this ALM existing and turbine running on load (10W), sometimes the following ALMs appear:

1.slot1 VCMI "R" diag.ALM
2.VCMI IO state exchange for "R"failed
3.compreesor inlet TC disagree(L86TCI)ALM
4."T"slot1 VCMI diag.ALM
5."X" VPRO Diag.ALM
6.Exhaust TC comm.Flt(L30Comm_io_R)

ALM comes on for the "R" and "T" momentarily and then all ALMs gets reset, except for the L86TCI and "S" slot1 VCMI Diag Alm. As the "R" comm. has failed, the TCs connected with "R" looses their values momentarily but CTIF1 (TC connected to "R") causes L86TCI ALM gets latched. This ALM needs execution of MR and while doing so, the turbine load reduces to 1 to 5MW and later returns to the original value and also causes L30FPG2L Alm to come and go with the execution of MR. All these problems are related to the Controller Communication problem. Why the comm. exchange for "R" fails all of a sudden and restores. (momentary failure). Kindly advice further. Thanx....
 
Hi,

Did you verify the supplying voltage to MKVI: voltage level, voltage unbalance v.s. ground, AC voltage component imposed over the DC voltage (ripple)?

For the diagnostic alarms analysis use Toolbox. By clicking the cards in question (VCMI or VPRO) will be presented on line parameters for these cards. Examin displayed voltages and other data on the screen.

Declared by you load swing of the range 1...5MW referred to 10MW turbine's load is very big.

Do you observe the power swing after every time MR is hit or only when MR is executed after L86TCI alarm?

Did you also record the load swing at the instant the alarm L86TCI appears or only when it is reset?

How fast is that load swing: it is like a puls or lasts several seconds?

To begin troubleshooting the root cause of the load swing please analyses data collected on trend recorded at the instant MR is executed, list of points provided earlier. You can add: FPG1 (inlet gas pressure), FAGR and FAG1 (servo current for SRV and GCV valves). Sample interval on trend should be selected very short (i.e. very frequent collection of data) in order to observe the servo current change during the event.

Regards
 
Top