GE Proximetor vibration

K

Thread Starter

Kamel

Dear every body,

i want to know how to use the proximitor page in mark 6 to determine the different problems of the gas turbine especially the phase angles and why machine not tripping for high vibration values in proximitor.

thank you
 
Dear Kamel,

My advice is to use the proximitor page of the HMI as just another "tool" for you to judge the health of your machine, much like the seismic page of the HMI. I do not believe I have seen anyone try to diagnose a vibration problem using either of these pages of information.

As far as tripping, "most" of the GE units I have seen only use seismic sensors for tripping of the turbine.

You did not mention what type of machine you are operating, and if you are experiencing a particular problem or just looking for information.

But in my experience the seismic and proximitor informational pages are a tool to judge your machines health. If you are experiencing a vibration problem as witnessed by high seismic and or proximitor values then you need to enlist the help of a vibration expert who typically has more tools at their disposal to diagnose a specific issue. I hope this helps answer at least part of your question.
 
There are some turbines with Mark VI and Mark VIe Speedtronic turbine control systems that connect proximitors directly to the Speedtronic, and with special versions of vibration input cards they can actually display phase angle and such on Toolbox and ToolboxST displays.

Some of the F-class turbines actually do trip on high-high proximity readings, but most GE-design heavy duty gas turbines only use them for additional display values only.

Some B-N (Bently-Nevada) vibration analysis equipment can be connected to the turbine proximity inputs and used for vibration analysis and even balance adjustments.

Some B-N monitor racks have cards that can communicate via EGD directly to Mark VIe (not sure about Mark VI) Speedtronic turbine control panels via the UDH.

But, as MIKEVI said, we don't know anything about the turbine at the original poster's site, nor how the feedback/inputs are connected to the Speedtronic Mark VI. If, as MIKEVI also surmised, the proximity values are only being used for display purposes on CIMPLICITY displays (which is the most common case), then there usually isn't any data useful information to be gleaned from them. In this case, the proximity values are not used for protection (tripping), though they might be used for alarming on high indications.

So, it all really depends on the application, and the original poster didn't provide enough information to be of much more help.
 
On their heavy duty gas turbines, GE uses the seismic probes for protection and the proximity probes for monitoring and analysis. Their reasoning has been that on these machines, the mass of the rotor is nearly equal to the mass of the stator, and for this condition the seismic probes give a more reliable indication of destructive vibration.
The main exception to this practice is an insistent customer that says if you don't comply with our specification we will buy from a competitor that does.
 
The machine is 9FA using Mark 6 and proximitor page has the normal peak values for vibration in addition to vibration Fourier analysis for each sensor signal as following : a-magnitude of 1X and its phase angle. b-magnitude of 2X and its phase angle and sensor gapes including key phaser. So we use peak values to monitor vibration but my question is how to use the analysis on this page to determine the rotor condition like unbalance or bent or any other condition , generally how we use this analysis
 
Kamel,

Vibration analysis is considered to be something of both a science and an art. Meaning, that while some conditions can be quickly and obviously detected, others are very subtle and only the most experienced of individuals can look at the available data and discern what is or may be occurring when vibration in one or more planes is increasing.

There are very long, and expensive, courses to learn vibration monitoring and analysis and what data is required for what purpose and how to use that data. All of this is well outside of a forum like this.

If you are experiencing particular vibration issues (including increased hydrogen consumption), it's recommended you obtain the services of an experienced and knowledgeable individual who can generally quickly analyze the available data (which many sites do not have, by the way!) and provide you with recommendations.

Or, get yourself educated so as to be able to monitor, trend, and spot problems from the available data.
 
I understand you but i don't understand his analysis for operators it should be useful for them not only for vibration experts so i want to know if there are some guide lines to use it not that much deeply but only general monitoring .is thre is any references about this pages or i have to study vibration in general .

by the way thank you for your information and help
 
Kamel,

Other than "comparing" the proximitor 1x values to the seismic (accelerometer) readings, there is little of value for the typical operator--other than to observe some pattern or trend.

And, that is really one of the most important parts of an operator's job: to spot increasing or decreasing values (vibration; proximity; temperature; pressure; etc.) and bring them to a supervisor's attention. Most operators just think all they have to do is click on SILENCE ALARM and ACKnowledge ALARM, and RESET ALARM, with the occasional START or STOP, and that the Speedtronic will do EVERYTHING else. Worse, their supervisors do little to prevent this, only complaining when some trend is spotted AFTER the turbine has tripped, or worse some catastrophic damage has occurred.

Trending is very important, because many if not most problems can be spotted before some alarm or shutdown or trip occurs--if operators are looking for increasing or decreasing values, and then monitoring the rate of increase or decrease.

For example, if the Cooling Water system pressure is usually approximately 4.1 barg, and the Low Pressure Alarm occurs at 2.01 barg, if the Cooling Water system strainers start to be blocked the pressure will be decreasing for some time before the low pressure alarm is annunciated and Lube Oil or Atomizing Air temperatures start increasing as a result. Blockages don't normally occur quickly, and an outside operator can be sent to blow down the strainer to try to restore pressure long before the low pressure alarm is annunciated.

But, trending pressures requires more than clicking on SILENCE-, ACK- or RESET ALARM pushbuttons, and most operators can't be troubled to record (log) values and analyze them. Or, even if they do record (log) values, it's seems it's always someone else's job to see that the readings they have been recording (logging) or have been "watching" have been decreasing or increasing for hours or days.

But, we digress. Again, other than "comparing" proximity and seismic readings, and observing increasing or decreasing patterns (trends), there's little of value on these screens for the typical operator. A "bent" rotor will be easy enough to spot without the proximity display.

[I say "comparing" proximity and seismic readings, because they are not always directly compatible. So, there is some science, and some art, in looking at the two sets of data and analyzing what may or may not be happening. Again, it's not simple, and well beyond my ability to clearly, and fully, explain.]
 
Dear Kamel, I have to agree with comments from CSA. The only information I might add is this.

Trending I think is probably the second most powerful tool for monitoring any equipment. The MOST powerful tool is an observant and competant operator, especially one that has been with the plant since its beginning. One who has taken mental and written notes. He/she knows how things sound and feel as they start and run. They know what normal pressures, temperatures and other readings are for a piece of equipment at different loads, outside temperatures, starting stopping etc.

For me the value of the siesmic and proximitor pages is this. What are the values at each bearing now? What were they when the machine was first commissioned? Has there been a shift or step change in the reading either up or down? Did this shift occur after a trip?, or a start-up or shutdown? Does a high reading only occur at high load, high VARS etc.

Any of these would indicate to me that something happened to the machine during one of these events. As an operator, or even most plant technicians, I would not expect them to diagnose this type of issue. But I would expect them to notify a supervisor or plant engineer and let them determine if more investigation is needed. And lastly take notes so in the coming years you can "recall" what you saw last time whether it was last year or 10-20 years ago.
 
Top