I am a recent mechanical engineer graduate and I am lucky enough to find myself employed. I have been awarded the task of developing a weekly performance report for a 1MW Stationary Gas Turbine (strictly used for electrical generation). I am new to developing reports. What all should I place in the report? I was thinking of mostly including graphs such as Exhaust Temperature versus load. I am not sure of what include or what graphs to include. I could really use any help you all can provide me.

 

To me a performance report is all about performance. I would be looking at things like:

fuel usage
up time vs. down time
power generated

--
Bob

 

Hi Ryan,

Take a look at OPC Trender. It's a quick and easy tool to start trending your Gas Turbine data:

http://www.matrikonopc.com/products/opc-archiving/opc-trender.aspx

Let me know if you have any questions.

Cheers,
Wassim

 

Bob Peterson, is, as usual, 100% correct. You would do well to understand what factors affect GT performance, and then what instrumentation is present on the machine at your site that can be used to reflect or calculate performance. Heat rate is one factor, but not every machine has the instrumentation to accurately calculate heat rate. Having said that, one can still establish a formula and calculation method, and then trend the calculated heat rate over time to determine if performance is decreasing and by how much.

Some turbines monitor axial compressor discharge pressure, which is a good indicator of performance. A dirty compressor will not flow as much air as a clean compressor--at the same ambient conditions. And GTs are "mass flow" machines, so lower air flows mean lower performance (efficiency and output). So, some times it's necessary to correct operating conditions to some reference value to compare apples to apples. Some GT manufacturers have graphs and formulae to do this; some don't.

Some turbines monitor turbine inlet temperature for control (as well as protection) and this can be a good indicator of performance. Some monitor turbine exhaust temperature for control and protection.

The cleanliness of the axial compressor inlet filters can also have an impact on performance.

So, you really need to better understand the factors that affect turbine performance and then try to understand what instrumentation is available to help you quantify performance.

There are many companies which sell software of various sophistication to calculate GT performance. So, this is not any small undertaking. You would do well to understand how this information is going to be used so you can determine how accurate the calculation and monitoring need to be.

Or, is someone just wanting to automate the collection of operating data for trending and historical purposes, sometimes mistakenly referred to as "performance monitoring"? Most operators are loathe to record operating data manually, or even by simply pushing a button to print a screen (though in fairness, they may have to push that button multiple times to record different screens). And, then, the recorded data is just "stored" somewhere, never reviewed or even looked at unless there is some problem, and then it's discovered that if someone had been trending the data they would have seen in impending problem.

As with most turbine operating conditions, trending is a very important aspect. Is a particular parameter increasing or decreasing with time? Is it increasing or decreasing rapidly or slowly? So, the rate of change is also important.

As to Wassim's contribution, OPC is a means of collecting data. You still have to understand what the purpose of the data collection is, what parameters are available, and then you can choose a method of collecting the data.

So, best to understand how the information is going to be used, exactly what "they" mean by "performance data" (actual performance, efficiency and/or output, or historical data-gathering), and then what instrumentation is available, and then how to "format" that information (how to collect and present it) so that it can be used in the desired manner.

Hope this helps!

Welcome to the biz, also!

 

Thank you all for your help. I am currently trending data from the Gas Turbine and storing the information in a .csv file. This I will probably link to Crystal Reports or Microsoft Share Point so as to produce weekly reports.

Now with regards to the type of weekly report. I am calling it a performance report because I want to see, if any, the degradation of the gas turbine and catch any impending problems. My other concern would be the actual graphs. Should I plot every data point that I poll? I mean the trending is pulling data every 0.5 seconds and for a weekly report that would be 1,209,600 data entries per parameter. Or is it preferable to plot a sample. And if so how big of a sample? I ideally I would like to compare the graphs to the benchmark data. Currently I would like the report to contain Fired Starts, Fired Hours, Graphs; Exhaust temperature v.s Power Output (load), Fuel flow v.s load, Computed Inlet Turbine Temperature v.s load, efficiency v.s load, and compressor efficiency, if I have the data available.

Am i going in the right direction? Thanks for your feedback.

 

Hi Ryan,

Now that I get the full picture of what you are trying to do, how can I get in touch with you so I can discuss this further?

Here is my email: [email protected]

Cheers,
Wassim

 

I have sent you my e-mail. My e-mail address is [email protected]