Hello to all,
By searching this and some other forums I have found on many places that closed loop control must be very fast when controlling turbine. For example I was reading about PCS7, Telperm XP and some other DCS solutions (like ABB's 800xA) and in each description there was something like: "for fast closed control loop application such as turbine controllers..." I saw that in each case cycle update time for this applications is significantly smaller than for normal application. For example, typical cycle time for controlling turbines is 5 - 10 ms and for other parts of the plant 100 ms.

I don't understand why gas and steam turbines require such high speed of execution.

Why this is important?

Thanks

 

I believe the issue is that most PLCs don't read speed at a rate sufficient to prevent excess steam or fuel from resulting in an overspeed. It's been my experience that when PLCs are used as turbine controls, the speed sensing circuits are usually one of many special circuits, sometimes employing proprietary cards (which kind of negates one of the alledged benefits of using a PLC as a turbine control, the one that you can buy parts "anywhere").

Think about a steam turbine and how fast it can accelerate if load is suddenly lost! Heavy duty single-shaft gas turbines have this big "brake" called an axial compressor but can still accelerate fairly quickly on loss of load. Aero-derivative gas turbines (multiple shaft units) can also accelerate quite fast due to low inertias (light weight components).

I believe this is what many people refer to as 'fast turbine closed-loop control'.

 

Thank you for your reply CSA.
So, generally cycle update time of PLC is "slow" comparing to how much turbine can accelerate in case of generator tripping (when generator breaker is disconnected)? That's why it is needed to use fast cycle CPUs.

I have seen such application where S7-400 is used to control turbine, but actual protection system and logic is stored in AddFem module which has by amount of 100 faster cycle time.

 

Just a few adder to earlier communications.

Turbines are normally high speed machine and are expected to operate at rated maximum speed where overspeed is a subject of concern. If it reaches "yield" speed, a permanent deformation in rotor is a possibility which calls for inspection. At "burst" speed, the turbine is expected to fail catastrophically. Overspeed limit for a turbine is guided by manufacturer and code like API612.

Overspeed protection of turbines are generally achieved through dedicated hardware. It used to be (in earlier days) a completely mechanical system (MOST) operated through trip bolts at defined speed but is gradually being replaced by electronic overspeed system.
As the overspeed is a fast acting phenomenon and degree of overspeed depends of nature of problem/failures, overall response time for the overspeed detection system is very important. It involves:

1. Overspeed detection response time. It depends on design of multitooth speed wheel, type of speed sensor, input response time at card, voting time (2oo3) etc.

2. Logic processing time by the overspeed control hardware. Generally it is lesser than 50 msec.

3. Execution time. It depends on contact bounce time for interposing relays, latencies of hydraulic system, soleniod valve, turbine trip valve etc.

Vendors like Triconex, Woodward, Bently etc supply dedicated overspeed hardware for big turbines.

Other than this protection system, turbines may have speed or performance control in separate control hardware which also need to be intelligent or fast acting to avoid wide overshoots/undershoots in speed and should be able to handle regular downstream disturbances.

Overspeed trip devices come into picture when the control system can not restrict speed within safe upper limit.

Regards.

 

dear sir,

just in simple and few words. the control system for any process/machine(here it is turbine) should take care of dynamic behavior of the process/machine. infact to control the process/machine effectively ,all the devices in the closed loop system(ie cumulative effect of process change sensing system, control system, field actuating device etc.) should be faster than the dynamic behaviour of the process/machine.

now, in the present case, turbine system is very dynamic and requires very fast response devices,right from the speed sensing system to control system to final actuating devices .this is why every time you come across turbine control system, speed of control system becomes important.

secondly, the protection system should be still faster to take care of the issues of Overspeeding of turbine.

hope this is useful to clarify your doubts regarding requirement of fast control system for turbine.

thanks
sandeep

 

Dear Sir!

As an electronic engineer recently works in a v94.2 type power plant, i am very glad to see text of an expert engineer. Unfortunately i have not any information and documentation about philosophy, basics, design, operation and maintenance of governor and protection software and hardware. For an example your text about addfem was an new thing. I don't find anything about it. Can you send me some general documents that explain this system and work of its components? It is your kindness to send me some more things so i can study from your mastership.

Thanks a lot,
Darvish