Hi Everyone.

I'm writing up a document to serve as a knowledge base for cyber security in PowerGen, and ran into a snag.

Basically, my understanding is that 7FAs (and I'm assuming other frames) usually have a Bently 3500/3300 system installed to watch for excessive vibration. The controller is usually wired into the <P> module to provide a trip in cases of (how CSA said in a previous post) 'high-high' vibration.

My question is: What would this look like in the alarm console of an operator? My understanding of the <P> module is that it has general trip points, not points that are reserved for vibration. Would this show up as a general trip signal, or would there be an indication that it was vibration related (outside of everyone in the plant soiling themselves).

Thanks!
Mike

 

Michael Toecker,

Be careful with your assumptions, especially where GE heavy duty gas turbine control systems are concerned. ;-)

There are about as many variations and permutations of vibration monitoring on GE-design heavy duty gas turbines as eight people can imagine. And, lots of older Frame units (other than F-class) do NOT have Bently-Nevada monitors/sensors, but if they do, they won't all be connected to the <P> core, and they don't all trip the turbine.

GE's position (before they purchased B-N) was that the <b>only</b> vibration sensing they recognized for turbine protection was seismic sensors. If they provided B-N equipment at a Customer's request (and expense) they would not generally connect it trip--only to alarm, on either 'high' vibration (B-N term 'Alert') and 'high-high' (B-N term 'Danger'). Some Customers might insist the 'Danger' signal be connected to trip the turbine, but GE discouraged it.

Now, of course, they own B-N and try to sell as much B-N equipment as they can on their power generation equipment. Seismic sensors are still provided and used for vibration protection, but in some cases B-N 'Danger' signals are connected to the <P> core (in Mark VI and Mark VIe Speedtronic control systems only) to generic trip inputs.

Sometimes the generic <P> core trip inputs can be used to generate specific alarm messages, such as "Bently-Nevada Vibration Trip" (or some similar message), and in some cases it only comes as the dreaded "Composite Trip" signal (meaning the trip can be one of several inputs--go figure out which one, Technician!).

The thing with Speedtronic turbine control systems is that the standard is there is no standard, so, not all panels are configured the same, and no one checks to see that when similar signals are used to trip the turbine that the operator indications are consistent. Another way to described it is: consistently inconsistent.

A B-N 'Danger' input could be connected to the <P> core, or it could also be connected to a TMR input (on a unit with TMR controls). And, it could be used to trip, or just to alarm. It all depends. And on what is anyone's guess. Best to examine every panel to be sure exactly how it's configured.

Some Mark VI turbine control panels have B-N proximitors connected directly to the Mark VI, and sometimes also to a B-N Monitor. Again, there have been a host of iterations, permutations, variations, and configuration. I even saw panels on the same site have different configurations.

It's something of a leap to say all, most, or even just F-class GE-design heavy duty gas turbines have B-N 'Danger' (high-high vibration) inputs connected to the <P> core. And, the Process Alarm text messages could also be very different. I would hope there would be one thing which was common to the Process Alarm text messages: the inclusion of the <P> core designation, if that's where they were connected. But, even that's a stretch.

Sorry; it's sad but true.

Your basic understanding of the <P> core is valid, though. The <P> core primarily monitors shaft speed and initiates a turbine trip independently of the main controllers (<R>, <S> & <T>) on overspeed and rate of change of speed (too much acceleration or deceleration). It is quite frequently used for synchronization purposes on many turbines, as well--but that's not a turbine protection function, per se. In addition to the overspeed/rate of speed change logic, <P> has the ability to have other inputs such as E-Stop pushbuttons; DCS trips; and even B-N 'Danger' inputs (up to a maximum of seven, I believe).

Hope this helps!

 

CSA, thanks!

I guess I've gotten complacent, assuming that my personal experience with trip points and Bently is the norm, and not part of a larger ecosystem. Most of the systems I've looked at over my career (8 years, so not that long in the grand scheme) have had vibration trip settings that involved <P>. IIRC, it was mentioned in the 4108 when I was reviewing the architecture of the units I've looked at over the years as well, which reinforced my idea of a 'standard' (hohoho, what is this standard you speak of?)

I'll be editing my document to determine if a Bently exists, require an inspection of the B-N interface to the GE MARK system for trip and alarm points via both hard contact and control system protocols, and an evaluation of computer security risk associated with that connection.

Then, I'll discuss some of the potential alarm messages you mention, including the "Composite Trip". My biggest concern is that vibes were voodoo at some of the plants I visited, so providing background and impact with those plants was critical.

This discussion also reminded me that Bently's can be used on ID and FD fans as well, so I've got ANOTHER section to write about that.

Mike T