GT tripping on lightning

M

Thread Starter

Manoj Gupta

We are operating 9 nos of Frame-6 GTs with Mark V/VI control system. Recently we faced tripping on one machine during lightning. Interested to know the means to avoid damage to the cards and tripping of machine.

We had observed the tripping of one of GT on "High Exhaust Temp" during lightning. Following were our observation.

1)Readings of all thermocouples (of exhaust,bearing metals, wheel space, Lube oil headers, Bearing drains) were showing either extreme high or extreme low reading. Exhaust thermocouple readings normalised after 6 sec. And other reading got normalised after replacing the TCCA card only.

2)R –Processor P2 (Tag on: 96FG2A) transmitter channel component found burnt at TBQB card.

3)CPD transmitter found damaged.
Integrity of complete shielding cable/termination and earthing pit found OK.

My questions are:
1) Is someone using Surge Protector Devices with Mark-V/VI system and how is the performance of the SDP. Can SDP protect against false tripping?

2)Are some other users facing problem of tripping of GT during lightning? What are the actions taken to avoid tripping of GT during lightning.

3)GE normally provides T/C cables running to Mark System Panels. Does any other site use 4-20 mA signal for T/C readings?
 
Tripping during electrical storms is difficult to protect against. Unfortunately, the best protection is one that divides, or separates, earthing into two separate "systems". Usually called "protective earth" and "functional earth" these two separate earthing systems are most often created during construction of the plant. And even then, there is the odd occasion where a lightning strike will still cause an unwanted trip.

The protective earth system is the one that motor frames are grounded to, and grounded transformer neutrals are connected to; this is the grounding system that provides safety to personnel from shocks and electrocution.

The functional earth system is used to connect control system earths/grounds to provide a common potential point.

When lightning strikes a piece of structural steel in a plant with a single earthing system and that strike is relatively close to the control system's connection to the same earthing system this can momentarily elevate the control system common point and cause havoc with the control system. Some older analog systems were particularly susceptible to lightning strikes causing unwanted trips and problems. Some digital systems, too, seem to have more of a propensity than others, but just about any control system can experience problems that are very difficult to protect against.

Now, in a divided earthing system that is properly constructed, sufficiently robust, and where field devices and instruments are properly grounded the reported incidence of lightning-induced trips seems to be lower.

One of the problems with separate earthing systems is that in order for them to work properly the functional earth system has to be properly constructed and fairly substantial. Most times, functional earths are poorly constructed ("It's just for the control system--and they don't carry much current!") and relatively small in comparison to the protective earth system.

Also, when field devices and instruments are being installed great care has to be taken to enure they are connected to the functional earth, and not to protective earth--and this is the other major problem with a divided earthing scheme. People just don't design or supervise the construction of a functional earth system properly, and that can lead to unintentional cross-connection of the earthing systems.

The problem in a running plant with a single earthing system is that it's very difficult to construct a new earthing "grid" or "pit" in an existing, running plant, and then to separate the grounds around the plant as required for this type of system to have the best chance of working.

I'm not aware of any plants that have used or are using any static protection devices to try to reduce trips caused by lightning.

As for using 4-20 mA transmitters for T/C signals, you would be introducing a time lag into the control system that might set up an unwanted catastrophic failure.

And Speedtronic panels are not constructed to substitute mA inputs for T/C inputs for exhaust temperature sensing. Speedtronic panels are designed to accommodate certain inputs in certain configurations for reliability and to match GE control philosophies and practices.
 
M
Thanks CSA.

Coming back to protective earth and functional earth, I understand you are talking about dirty earth and clean earth, as per terminology we are using. All instrument shield/screens are connected to clean earth pit and all other earthing cables (including Power supply, motor frames, transformer neutral) are connected to dirty earth pit. Finally all earthing pits (Clean or dirty) are connected to each other.

In our set up, we have all instrument shield/screen cables are connected to the earthing strip in Mark Panel and from there a common earthing cable is laid to Clean earthing pit. No other earthing cables are connected this pit except similar earthing cables from other Mark Panels.

So I understand that our earthing scheme is similar to what you are describing as divided/separate earthing system.

I was searching internet on this topic and what I can understand from it that, a very high current flow in the ground due to lightning can increase the ground potential. This in turn causes EMF in cables running in the ground causing induced false readings of instruments. In our case, all cables from GT to Mark panels are running buried for about length of 150 meter. Can this be probable cause of GT tripping? If cables are not laid underground/buried , does it make any difference in the effect of lightning. We had a couple of exactly similar tripping few years back also but in some other GTs. In all cases, it is our observation that only those field devices readings are getting affected where the output is either mV(T/C) or V( CPD/FPG2 transmitters are having 0-5 V output). There is no change in the readings of the other instruments.

Can someone explain this? And what we can do to avoid such type of tripping.
 
So, you're saying that thermocouples are never grounded in the field, on "dirty" earth, and that transmitters and the interconnecting cable shield drain wires are never connected to dirty earth in the field?

I don't know as I've ever measured a T/C to ground, but I do know that Speedtronic panels can accommodate grounded or ungrounded T/Cs, and RTDs as well.

I also know that a lot of cable shield drain wires are improperly grounded in the field, and I would ask where the shield drain wires are terminated in the Speedtronic panels: clean earth or dirty earth?

Again, this whole business of divided/separate earths has to be very carefully thought out during the design phase--and even more importantly implemented and supervised during the construction phase--of a plant.

I'm also of the opinion that the armour-shielded cables used in many parts of the world are a source of much electronic noise in control panels, simply because the armour shields are properly insulated from earth at one end or the other.

There are a lot of ways lightning strikes can wreak havoc on a plant/process. And, it requires a lot of methodical checking and correcting to try to eliminate all of the potential problems, and, still--there's going to be the odd strike at the least likely thought-of location that's going to cause a trip.

There's no simple answer to this dilemma. If there were, everyone would have implemented it long ago--or some manufacturer would have patented a product and made a fortune by now. This whole business of dirty earths and clean earths, or whatever someone wants to call it, is one attempt at trying to remedy this (and other) problem(s).

That's not said to try to discourage you from your efforts--you may find the answer everyone's been looking for and make a fortune yourself!
 
Top