After eleven years of development, IEEE Std 1415-2006 "IEEE Guide for Induction Machinery Maintenance Testing and Failure Analysis" has been officially released as of May 1, 2007. The standard covers a majority of testing technologies used to troubleshoot and analyze rotating machines including the types of testing performed by ALL-TEST Pro, Areva, Baker and PdMA, including pass/fail limits, whether the data is trendable or the testing is potentially distructive. A large number of end-users and the manufacturers were all represented through its development.

Attached is a copy of the cover and table of contents of the new standard.

Howard

Pages_from_1415-2006-2.pdf (173 Kb, 97 downloads)

Welcome back, MotorDoc!

I read your paper on the same subject but couldn't really work out which tests can be used to monitor conditions of submersible motors driving submersible pumps.

Josh:

Thank you. Needed a break and have been opening additional offices.

The standard really doesn't expressly address submersible motors. However, many of the tests can be performed on submersibles without too much trouble.

Now, it depends on the design and size of the submersible machine. Some smaller submersibles have contactors mounted in the motor/pump assembly. The problem is that you cannot test the winding in a de-energized state.

Another issue is the length and potential condition of cables going to the submersibles. One of the primary reasons for the development of the ALL-TEST Pro instruments, for instance, was to test their own (BJM Corp, the parent of ALL-TEST Pro, LLC and the manufacturer of BJM Submersible Pumps) pumps. The objective of using the low voltage test was to identify potential or existing winding faults without potential damage to cable or giving the opportunity to clean and dry a winding if there was seal failure. In 2002, I worked on the Areva/Summit/ALL-TEST Pro partnership for Electrical Signature Analysis so that we could test the dynamic condition of the pumps in the test tanks (while I was with BJM Corp). Ie: You cannot put a vibration probe in the submerged test tank, but you sure can see vibration and mechanical/electrical issues through current and voltage signature analysis.

The ALL-TEST Pro online, PdMA EMax and Areva Empath systems can provide this capability in a similar manner. Baker's Explorer system evaluates the torque of the pump and, in a steady-state system, can provide some valuable data, as well. Many vibration analyzers also have the capability of taking current data and providing current FFT's (ie: CSI). While there is a little work to the analysis, the results are pretty good and vibration shows as a 1X RPM signature, just like vibration.

Basically, if the submersible is in place, a low voltage test, insulation to ground test and an ESA or torque-style dynamic test can provide most of the data you need.

Do you have a specific issue with your submersibles?

Sincerely,
Howard

Yes, we would like to do condition monitoring for our submersible pumps and motors where offline vibration monitoring is not possible.

There are 2 units A and B with one being on duty and the other on standby. The spec is 140 KW, 415 V. 3 ph, 50Hz.

At the moment all we can monitor is discharge pressure and flow rate which are 270 m3/h and 18 barg resp'ly.

Another issue is the mechanical seal for the wet winding type of the submersible motor is suspected to leak out externally supplied cooling, potable water from an overhead tank.

Josh:

You make a good point, as well. When considering mechanical condition, vibration analysis would be the preferred method. However, when you cannot perform vibration, then ESA/MCSA would be the next best method. For static condition of the cable and winding, then MCA testing should be added.

So far as the condition of the mechanical seal, in the work I did with ESA at BJM and with the US Coast Guard on bow and stern thrusters, there is a signature that can help identify bad mechanical seals. However, I have not performed any analysis on a wet-winding pump with ESA, so cannot say if the signature will work for your condition. As far as other methods of detecting leakage out of the pump, I don't personally know of any solution. We used moisture sensors in our oil and air filled submersibles.

For vibration condition of a machine using current signature, if you are using the PdMA, Areva (Empath) or ALL-TEST Pro online instruments you would identify the running speed in the demod spectra, then look at the current spectra. If the running speed sidebands around line frequency are less than -65dB, or at least -10dB above the noise floor, then you will most likely have a signficant vibration issue (the demod value cannot give you condition as the condition has to be related to the actual current value). If the condition increases over time (periodic testing), then you will want to investigate.

The seal signature that we identified in air and oil filled submersibles, as well as some standard pumps and the bow and stern thrusters, was sidebands of line frequency that fell within 45 to 55Hz in 60Hz line frequency, and 35 to 45 Hz in 50Hz line frequency. If there are two evenly sized peaks, then there is some wear, if either peak increases greater than the other, then the wear is significant. We were, and are, actually not sure why these signatures exist, the results were obtained by experiment.

Personally, I use the ATPOL unit for most data-collection conditions and the EMPATH system for detail analysis if I am looking for a specific point in operation. The main reason that I use both units is that both units use the same analysis software, so the data can be shared and compared directly for trending purposes.

The MCA test will be able to identify the condition of the wire-to-wire condition of the cabling and the condition of the windings. The PdMA will focus on ground wall insulation condition and you can trend the capacitance values to get a general idea of condition. The ATPro units focus on inter-turn degradation with some focus on ground-wall insulation value.

Hope this is helpful.

Howard