Trending of test data

[sngpl]

We operate a power plant having 5 hydroelectric generators. Over the past 6 years we have been carrying out the following tests, in addition to several others, on these units:

1- Insulation resistance of stator with power cables connected.
2- Insulation resistance of stator with out power cables.
3- Insulation resistance of each phase.
4- Stator winding resistance.
Of course PI has been calculated for all the insulation resistance tests.

What sort of trending should be done in order to produce a formal report for the customer? I mean we can trend the insulation resistance of each unit for these six years, trend the PI of all the units with respect to each other etc. How can we trend the insulation resistance of the units with respect to each other??
Any input shall be highly appreciated.

 

[5P10]

sngpl,
Trending the insulation resistance might not be relevent, as insulation resistance measurement is very much dependent on temperature, weather condition etc. I would think that trending PI for each unit is the only data that would give relevent reference to your customer on how good is the system. Trending the stator winding for each unit might be good also.
Just my tuppence...

 

[sngpl]

5P10,

IEEE standard 43-2000 states that in order to avoid the effects of temperature in trend analysis, all tests should be corrected to a common base temperature of 40 degrees C using the following equation:

Rc = Kt * Rt

So we can trend insulation resistance.
Thanks for the reply anyway!! I appreciate it.

 

[josephkendall]

What exactly are you trying to convey with the data? Are you showing the breakdown of the Stator insul.? Are you showing the breakdown of the cable insulation? The data in order to be correlated to each other must be tied to some common factor between the units. IE: Enviromental: Temperature, humidity, airflow, etc.

 

[jbartos]

Suggestion: The lower insulation resistance/impedance in time will indicate the insulation aging. An insulation analyzer may help. There may be a lot of empiricism involved initially until a trend is established. This is a normal trend finding procedure in other areas too. If any service firm has some trends developed, they will probably be proprietary since their business depends on them.