Our power plant (GE 9FA with Mark VI controls) had a trip last week on G3 purge pressure high.

After many checks we found that 2 of the 3 pressure switches (normally closed contact) had a high contact resistance - The contact resistance of switch A and C were in the K ohm range and seemed quite high in comparison to B which was 0.4 ohm. During operation we would get an alarm from one of the switches for some milliseconds before clearing. We have fitted a pressure gauge and checked the vent opening so we are sure it is not a real high pressure.

We have now replaced the pressure switches with new and the alarm has not been received since.

My question is - how / why would the Mark VI interpret the higher resistance of the A and C pressure switches as being and open contact?

Thanks for reading and hope you can assist.

Regards
Alan

 

Congratulations on finding and resolving the problem.

The Mark VI interprets open and closed contacts just the same as all control systems interpret closed or open switches: current flow. When current is flowing above a certain level the switch is deemed closed, and when the current flow is less than some level the switch is deemed open.

No; I don't know precisely what that current level is for the Mark VI. It's not a configurable parameter.

Why did the switch resistances increase? Usually when this happens, it's because of moisture getting into the switch and causing a "growth" of crusty, spiny white matter that is not as conductive as shiny metal contacts are. And measuring these resistances (of this matter) is very difficult, and probably doesn't result in the same current flowing when the Mark VI is impressing 125 VDC across the contacts.

I would be very interested to know what the 125 VDC split is on the unit, and if there were any 125 VDC Battery Ground alarms annunciated intermittently prior to this problem.

 

CSA,

Thanks for the feedback. After much pressing of GE they advised that a contact resistance above 75kohm would be sufficient for the switch to be deemed closed.

The plan is to open the switch to see if we can determine the cause - not so easy as it is an encapsulated switch but I'll let you know what we find. The units are about 8 years old but its and indoor system so not easy to see how moisture would get to them but like you say it would appear to be the only rational explanation. Our instrument techs are now more aware that they need to check for open & closed contact resistance when doing their calibration checks.

On your other point about earth fault alarms - there weren't any prior to or during the trip. We actually spent over 24 hours pressure testing the purge valves and debating with GE about increasing the switch setpoint to allow a test run to gather more data. At one point we were contemplating replacing the valves as we believed it was a real high pressure event. These valves as you are most probably aware are located in the worst possible place so removing them is something of my worst nightmares.

Thanks
Alan

 

IrishAl_0,

It's presumed you have meggared the wiring between the Mark VI TBCI terminals and the switch contacts, and between any intermediate junction boxes, looking for any possible source of high resistance between the Mark VI input terminals and the switches themselves.

The worst switch location ever was the mounting of 26AA-1 on single-base Frame 5s back before the mid-1980's. The switch was virtually impossible to access, but once someone did get into the space, under the #1 Bearing drain piping and on the back side of the L.O. Tank under the Accessory Gear Box, one would find about 5 meters of capillary tube coiled up near the switch. That switch could have been mounted anywhere else with a capillary that long, but NOOOOO!!, it was mounted in place that was inaccessible for anyone weighing more than 10 stone and taller than 1.7m. And even if you could get to the switch, you couldn't get your head 'round to see the adjusting wheel or to the sub-junction box to see the terminals or loosen or tighten the screws.

Most switches GE or their packagers use are explosion proof. Some are easier to "disassemble" than others, and some are virtually impossible to get to the switch contacts themselves even if you destroy the switch in the process (being hermetically sealed).

Pressure-testing those valves can be very difficult to do in-situ. But, getting them out and re-installing them is almost as difficult and time-consuming.

However, you definitely don't want either of those valves to leak--ever.

My money is still on some kind of moisture having gotten into the switch, either during operation (which you say is likely), or during shipping from the factory, or during storage at the construction site, or during construction. There are a lot of poor construction practices, both on site and at vendor facilities (I presume the switches are or were mounted in a skid that was likely built by a vendor to GE or the packager.)

Anyway, please write back to let us know what you find.

 

The wiring and switches were meggered together and independently. The wiring was okay but 2 of the 3 switches had high resistance. Unfortunately the switches were hermetically sealed when opened it was hard to tell if what we saw was normal or just as a result of damage caused when attempting to open.

The skid, like you suggest, was factory assembled and shipped to site. It is located indoors and in a dry temperate climate - not too humid, nor cold. The unit has also been in place for 8 years so one would expect failure before now if it was due to moisture ingress during shipping or pre-installation storage conditions.

We have included checks now for our calibration guys to record contact resistance during switch calibration to hopefully prompt closer check and prevent recurrence.

Thanks for your help.