e have GE make Frame-9 gas turbine with MARK V control system. We have observed fire in the MARK V panel. The fire started from the DTBC panel of the <QD> core. Thenafter, we measured the resistance and meggered each solenoid and contact output from the panel. It was observed that the coil of the 20CF solenoid clutch was found to be short circuit.

I have a open question for all MARK V users that;
1. Should any short circuit in field device burn the DTBC card.
2. Have you experienced such type of fire in MARK V.
3. Is there any remedy so that such further fire incidence in MARK V is avoided.

 

This is not unheard of, since the individual solenoid outputs/contact outputs of the Mk V are not individually protected with fuses. (This was "remedied" in the Mk VI....) Likely, what happened was that there was a high current (slightly less than 15 Amps--which is what the fuse rating of the fuses for the DTBC/DTBD cards usually is; check the J8x cable fuses in the <PD> core) for a period of time that sufficient current flowed through the DTBC long enough for the traces of the card to get hot enough to burn the card--before the fuse blew. A "quick" dead short would usually result in a blown fuse, which would provide some degree of protection against "high" currents flowing long enough to damage circuits/traces. However, if a "high" current is allowed to persist in the circuit long enough it has been known to cause burning of the DTBC card.

Early Liq. Fuel Pump Clutch solenoids had coils which were only rated for approximately 100 VDC. This required the use of a resistor to limit the voltage applied to the coil. In the Mk V turbine control panel, the resistor is usually mounted on a heat sink on the door of the <PD> core, and is connected to the circuit via cables, J15 & J16. This required 20CF-1 to be connected to <QD1> Contact Output 18, DTBC-69 & -70. Figure E-5 of Appendix E of Rev. F of the Mk V Application Manual, GEH-6195F, shows the circuitry in its entirety. (If you don't have this revision of the manual, send an email to [email protected] and a pdf file of the circuit can be sent back.) If 20CF-1 is connected to these terminals, you will need to check all the cables and printed circuit cards in the circuit: J15, J16 (the cables between the DTBC card and the TCPD card in the <PD> core), JSn (the appropriate ribbon cable between the TCRA in Loc. 4 of <QD1> and the DTBC card), the DTBC card, the TCRA card in Loc. 4 of <QD1>, relay 18 of the TCRA card in Loc. 4 of <QD1>, JZ1 (the cable between the TCPD card and the resistor on the door of the <PD> core). If the DTBC card was damaged, it's likely the JSn cable and/or the TCRA card in Loc. 4 of <QD1> may also be damaged--or at the very least weakened.

If the original Liq. Fuel Pump Clutch solenoid was rated for the lower voltage and the resistor in the <PD> core was used in the original application, and the solenoid/clutch was recently replaced with a solenoid with a higher voltage rating and the resistor was not shorted out, this could be part of the problem.

About the only thing which can be done to protect against a similar problem in the future is to install additional fuses in the circuits of "large" solenoids. In this case, one will need to determine the amount of current required for normal operation/hold-in of the circuit, then choose a fuse with a slightly larger rating and install them in each leg of the circuit (positive and negative). NOTE that the circuit for the Compressor Bleed Valve Solenoid, 20CB-1, and the Liq. Fuel Forwarding Pump Stop Valve solenoid, 20FD-1, are already protected with slow-blow fuses; do not install any additional fusing in these circuits: <QD1> Contact Output 16, DTBC-61 & -62, and <QD1> Contact Output 17, DTBC-65 & -66.

In any case, once the short in the circuit is resolved, the amount of hold-in current required for normal operation of the clutch should be determined and periodically monitored.

 

The fire could have originated at the jumpers(used for extending powering the relay o/p's to drive solenoids) on the terminal board..The berg jumpers is a good source of fire as in one site in India, I had a burnt out jumper on the terminal board and failed varistor on the power supply board in PDC ,due to short circuit in field and some smoke. In your case, this could be a manifestation of similar occurrence but of more intensity casuing fire.

Remedy could be to install more fuses (one each for individual solenoid circuit, something like fused terminal block ) in mark-V which is rated as per the field device instead of relying on just one fuse for a bunch of solenoids.