overspeed limit for hydro generators

[edison123]

For a pelton wheel hydro generator, what is the speed limit to be set on mechanical and electrical overspeed protection ? What is the applicable standard ?

The machine is 25 MW, 14 pole, 428 RPM with a 60 ton generator rotor and a 15 ton pelton turbine.

 

[ccw]

Hi edison123,
Are you matching phase with the power grid by speed control?

 

[edison123]

ccw,

We plan to test the mechanical and electrical (frequency based with toothed wheel and prox probe)overspeed protection systems in a refurbished machine, before we synchronize it.

The new ones (as claimed by the OEM) has a mech limit of 110% and elect limit of 120%, which in my opinion are too high. Why would you wait for the machine to go up to 10% to 20% overspeed before you decide to act? Given the enormous stresses involved, I would like to set them at the lowest possible limits. That is why I would like know what the standards say on this.

 

[cbrn]

Edison123,

we first have to distinguish between overspeed and runaway speed.
Phenomenon: in case of regulation failure, there are cases when the turbine will pull the generator out-of-phase and accelerate up to a limit speed given by (jet force - frictions and dissipations). This is only a transient because in the meantime other means will provide to brake the turbine (counter-jets).
Runaway: this "ultimate" condition is called "runaway speed" and has to be taken into account for structural matters (both in the turine and in the generator): all elems must withstand the centrifugal forces. It is absolutely normal having a runaway speed of 2.1 / 2.2 times the nominal (i.e. 110% / 120% of "over"). But from an electrical point of view, this is no concern because the generator normally has to be dis-excitated long before...
Overspeed: this is the "electrical" limit at which the generator dis-excitation must occur. That means that, from nominal to "overspeed", the turbine will carry the generator load, and from from "overspeed" to "runaway" it will be "freely rotating". The overspeed normally has to be discussed together with the turbine manufacturer, as setting the "over" too low would possibly lead to accidentally dis-excitate the generator as soon as a single normal regulation transient occurs. If the turbine manufacturer imposes 10% of "over" for overspeed, it means that the servo regulator is possibly not designed to track the transients faster. There is also another compromise: tracking the runner too strictly could possibly lead to oscillations in the penstock / bifurcation / ...

 

[byrdj]

For steam turbines, the protection to limit the over speed of a full load rejection (generator breaker opening) is two tiered.

First, the governor should close the controlling valves (as rapidly as possible). This action should limit the over speed (under 110%) and the turbine will settle out at a speed slightly greater than synch (normally 105%) and can be controlled by the governor to allow re synchronizing.

Second, if the first action doesn't work, the emergancy governor will trip the unit (rapidly closing isolation valves at 110%) and this should occur at a speed soon enough to limit the peak speed from reaching destructive speed (120%).

I would think the mechanical and the electrical would be within a couple percent of each other. ie back up to the other especially if either one is testable while on line.

Again for steam, there was ASME PTC20.? ,but it only defines the testing procedure. The actual trip values were supplied by the OEM which should know the stress limits of the equipment.

I have seen designs where the emergancy trip speed was higher than 110% since the governor was not fast enough to limit. A control addition of "trip anticipator" was added to slam the isolation at a speed less than 110%, but not trip unless speed reached the emergancy trip level.

I believe IEEE 600 gave the requirement that the governor will be designed to limit the over speed below emergancy trip.

 

[ccw]

Hi edison123,
After seeing what a runaway rotor did to the walls of a hydro-powerhouse (admittedly several years ago), I think it pays to be cautious. The analogy of the steam driven turbine/generator to the hydro rig is complicated by the penstock stream momentum. There is a secondary danger in "slamming" valves in a liquid stream. The OEM should be required to demonstrate the overspeed and runaway protection system. Obviously, reliability for these protection systems should be paramount. If your budget can afford it, you might consider an independent review by a recognized expert in the field.