hello again,

We at our site are trying to understand how the Speedtronic Mark V incorporates the same equation for all modes of governor control. The problem we are facing is that how the MarkV control the variables for different modes of turbine in the ONLY equation. Can we model the system with the same equation for all modes and is there a rate of change controller within the Mark ( eg. for TNR rate limit, FSR rate limit etc).

Kindly please suggest some nice paper or site link where i can find information or e-books about the governor modeling.

Thanks in advance.

 

Hello again,

As was said before, or in other words, again, the information you seek about how your turbine is controlled is in the CSP for your unit.

The equation for Droop- or Isochronous Speed Control is basically the same. When in Droop Speed Control, the control loop is straight proportional control. When in Constant Settable Droop Speed Control, there is a integral element on the load component. When in Isochronous Speed Control, the control is very tight proportional-plus-integral control.

You need to review your CSP and find TNR (Droop Speed Control Turbine Speed Reference) and TNRI (Isochronous Speed Control Turbine Speed Reference) and sketch them out on a sheet of paper to see how they are derived (calculated) and related. If the turbine has Constant Settable Droop, there is usually a term/signal TNRL (which means Turbine Speed Reference - Load-biased when used in this context; not Turbine Speed Reference - LP Shaft, as most Longname descriptions have it).

When it comes to calculating Droop, that depends on whether or not the unit is operating in "straight" Droop Speed Control or in Constant Settable Droop Speed Control.

Yes; there are loading (and unloading) rates for GE-design heavy duty gas turbines. They can be found in the CSP, also. They basically define how much change in TNR occurs in a period of time for different conditions, such as manual load changes (from the buttons on the display), or for automatic load changes (when one selects Base Load, or Pre-select Load, or STOP).

The loading (unloading) rate Control Constants are usually defined in the array TNKR1_n, where "n" is 0 through 8 or sometimes, 0 through 9.

I want to caution everyone reading this: Do NOT completely trust the Longname description in the CSP for these Control Constants. The Longname description *DOES NOT* define what the signal does or how it's used. The Longnames for these Control Constants (TNKR1_n) are not usually correct, and the usage of the Control Constants in the CSP for every particular turbine defines when the rate is used for that particular turbine.

This has also been said many times before about many different signal names in Speedtronic turbine control systems: The Longname descriptions are usually correct for most signals, but for some, they are not usually correct. The actual usage defines exactly what the signal is or does, not the Longname description. It's unfortunate that the Longname descriptions weren't always properly updated, and that requisition engineers didn't follow any standards when choosing rates or editing CSPs, but that's the way it is.

You're probably reading the best descriptions, other than what you may find in the manuals provided with the unit and in the Control Specification document, that there is. There really isn't a lot written, other than what's in the CSP. It's like a foreign language, but it's not difficult to learn. It's a graphic language, so that makes it a little easier.

We're willing to help you try to understand how your turbine operates and is controlled, but we don't have access to your CSP and documents. The CSP is an ASCII text document, so you can copy portions of it to your post/response, and describe what you think is happening in the snippet, and we will confirm or correct your analysis.

I want to caution you though, that I'm very reluctant, as was a previous contributor to control.com, to specify how to calculate Droop. This is because almost every time I've done this on a job site, I've lived to regret it. People have gone and changed the Droop setpoint based on how they think it will affect the operation of their turbine, and their belief or understanding is not usually correct. Just because something can be changed, doesn't mean it should be changed. And, when it comes to Droop setpoints, it's so intertwined with many other turbine control operations that it can very serious consequences if all relevant operations and parameters aren't also properly adjusted.

Almost every prime mover in the world, regardless of type or manufacturer, usually has a droop setpoint of either 4% or 5%. And, with gas turbines, there are a lot of factors which affect turbine operation and give the *impression* that the unit is operating properly because it's not achieving it's design droop setpoint. Just because a gas turbine doesn't reach 4% droop when it's at Base Load *DOES NOT* mean that Droop Speed Control isn't working properly!!!!!

I haven't looked at wikipedia.org recently, but their description of droop speed control isn't very good. I have gone through most of my old university texts, and have realized that most of those descriptions contribute to a misunderstanding rather than a comprehension of the principle, and the most were probably written or edited by people with little or no practical experience.

One of the best sites I have come across for basic power generation fundamentals is candu.canteach.org. They have a lot of really good, useful, and well-presented information on that site. It's not all in a consistent format, but what's there is better than most anywhere else that's free on the World Wide Web.

Good luck with your modeling endeavor! I would be very interested to learn how much it costs you to develop your own model, versus how much one could have been purchased for. My experience is that it seems relatively simple, but the reality is quite another thing. I know of several plants that embarked on developing their own model in-house, and then just gave up when they started looking at the monies spent to date. Some, went and spent nearly as much to buy a modeling package, but others just cut their losses and dropped the project.

 

Dear CSA,

I your above reply, in third paragraph, you have mentioned a bit about constant settable droop control (by saying that there is a integral control on load component). If possible could you elaborate more on this? That how does it work and how the calculation is?

Thanks & Regards
SSV