1. Why GE is using single T/s for tripping on warren pump L.O. Temp. high trip..when at so many other parameters it is using 2-O-O-3 logic.

2. Why temperature indication of warren pump L.O. is not provided which is very important parameters as far as plant reliabilty is concerned?

The last question is is it O.K. to trip the turbine based on single input which has high probability of failure as it is installed in high vibration zone of accessories compartment?

 

You haven't said how old the unit is nor how long the switch has been in service. I've been to sites where, when a device failed after 23 years, the Customer was aghast at the "poor quality of devices provided" with the unit. It was his feeling the device should not have failed at that time, and that, even though it hadn't been calibrated in years, it should have provided indefinite service. Everything will fail at some time; most failures will be accelerated by poor maintenance practices or lack of attention to impending problems (heat, vibration, etc.). How long since the switch had last been calibrated or replaced?

To my knowledge, there are no GE design engineers contributing to this forum and this kind of inquiry would best be directed at that organization. What kind of a response do you expect? An award for pointing out a flaw in the design? Engineering is a series of compromises--some are better than others. Do you know of a perfect turbine? A perfect car? A perfect mattress? A perfect stove? All engineers are human, and engineers work for managers who work, these days, for bean counters who scrutinize every cost to to save as much money as possible. Sometimes it really is a poor design or a poor application of an instrument for a specific function.

It would be *very* unlikely that a GE-packaged heavy-duty gas turbine with a TMR control panel would utilize a single instrument to trip the turbine. A GE-designed heavy-duty gas turbine packaged by others might be more likely to have such a configuration. Who packaged and provided the GE-design turbine? What control system is it using? Is it SIMPLEX or TMR?

Another possible reason for tripping on a single input would be a Customer requirement, or a change made in the field at the Customer's request, or without factory knowledge.

Another possibility would be that there is simply not a sufficient location for multiple sensors on the pump, though that seems a poor excuse for not making provisions for multiple sensors. GE isn't shy about asking their vendors to make changes to improve the reliability of their units--and adding the cost to the cost of the unit. Some packagers cut just about every cost possible, without regard to reliability.

If you are saying that there is no "remote" indication of Warren Pump L.O. temperature on some display, it may be for the above reason: insufficient location for the sensor required for the application. Is there no local temperature indicator? Are the operators not required to periodically check gages and indicators on the unit for operating parameters not available on remote displays?

As to the last question, it's part and parcel of the first two questions, is it not? Most temperature switches have sensing bulbs and capillary tubes which allow the switch to be remotely located from the parameter being monitored. Most failures of this type that I have seen have resulted from poor maintenance practices (not properly running capillary tubes in the clips provided which should greatly reduce the effects of vibration and rubbing/chafing); poor design practices (mounting the devices in such a location that the capillary tubes are subject to high traffic conditions--people stepping on them, tools being dropped on them, etc.), and just plain age.

Depending on the control system on the unit, you may be able to replace the temperature switch sensing bulb with a dual-element T/C or RTD and connect them either directly to the turbine control system or have them connected to redundant switches which would then be connected to the turbine control system for added reliability. Of course, either of these alternatives would require new conduit and wire, and a control system modification. But, if this single point failure is causing continued reliability issues, then one of them is probably worth the cost and effort. By connecting the T/C or RTD directly to the turbine control panel, one could also add the value to a display to relieve the operators of having to get up out of their chair and walk out to the noisy, hot unit to check a temperature indicator (heaven forbid).

It's been said before on control.com: Once the unit is out of warranty, the owner and the owner's operators are free to make whatever changes they deem appropriate or necessary. Some changes (exhaust temperature control; IGV control; etc.) would not be recommended without first consulting GE or the packager. Others--well, if it makes sense and it doesn't adversely affect unit operation or it positively affects unit operation, then why not? In such cases, all that's really required is to think the modification through, choose a reliable vendor/supplier of the device, instrument, make the appropriate modifications, and the issue becomes a non-issue.

Why make a mountain out of a small pile of dirt?