governor droop control 1

[stor138]

We have a 30 MW generator connected to a steam turbine. The turbine is a steam extraction unit, with HP inlet steam (~600 psi), MP steam extraction port (~120 psi), and LP steam exhaust (~60 psi). There is a inlet valve and an extraction port valve. There is not a valve on the outlet port.
The existing PLC-based governor has the following controls (described only when grid-connected):
1 - load control (MW); controls inlet valve
2 - speed-droop control (turbine rpm); controls inlet valve
3 - backpressure control (psi); controls inlet valve
4 - extraction port pressure control (psi); controls extraction valve
Mode 1 is a cascade control loop which provides a speed set-point to the speed-droop loop. Mode 3 is the normal operating mode and controls only on pressure - it does not consider the unit speed. Mode 4 operates independently of the other loops.
We are having issues with compliance of the governor with the local regulations - the utility that the generator is installed in requires the governor to operate in droop at all times for system stability. Since this unit is normally in backpressure control, it will not normally respond to system frequency changes.
My questions are:
1 - is droop functionality while in exhaust pressure control an operating mode that should be expected for governors on extraction/exhaust turbines? From what I can tell, I believe that this should be standard, but can stand corrected.
2 - if droop functionality is expected to be in operation while the turbine controls exhaust pressure, how does the unit react on a sustained change in utility frequency, if the back-pressure control loop is modified to be cascaded into the speed-droop loop? When the inlet valve opens in response to the system frequency change, will the corresponding pressure change seen on the header result in a counter-acting of the droop action? Are there any governor control schemes that are recommended so that the response to a change in system frequency is sustained, not momentary?
3 - how is this control mode handled in off the shelf governors? From what I can tell, a Woodward 505E handles exhaust pressure control as a cascade into a speed loop. Does a Woodward have logic to also maintain the proportional response to a frequency change once the steam outlet pressure is off setpoint, so it is not reversed by the pressure loop?
4 - any suggestions or experience on how we might achieve compliance with the utility?
I would like to better understand how systems are traditionally configured so I can better understand this specific situation.
I hope this email is clear - please let me know if there is any missing information that would help with a reply.

 

[waross]

That sounds like a recovery boiler at a pulp mill. I'll try to describe the operation of a recovery boiler, see if the description is applicable to your operation.
The control of a recovery boiler extraction turbine is not a typical control scheme.
1 The droop control may be inactive under normal operating conditions.
2 The grid frequency may change slightly under changing loads on the grid. The utility has all generation but the swing set on droop control. This allows the system to support changing loads for the short time that it takes for the swing set to restore the frequency.
The grid connection holds the turbine at grid frequency and speed.
The steam supply or flow is determined by the output of the recovery boiler.
As the steam pressure is dropped through the turbine, energy is released that is captured as electrical energy and delivered to the grid.
The 120 psi steam pressure extraction valve is controlled by the 120 psi steam pressure.
The 60 psi steam pressure extraction valve is controlled by the 60 psi steam pressure.
Whatever electrical energy is generated is delivered to the grid independently of the droop control.
A recovery boiler generates fairly high pressure steam, (~600 psi).
The mill needs a lot of steam at ~120 psi and ~60 psi).
A cheap and simple de-superheater will give you all the lower pressure steam that you need. You may have one or two de-superheaters hidden away in the upper reaches of the plant. In the event that the steam turbine goes down the de-superheater will automatically come online to maintain the supply of lower pressure steam. In the event that there is not enough steam to supply the 60 psi steam demand, the desuperheater will kick in to make up the shortfall.
Economics dictates that it is more economical to extract electrical energy from the high pressure steam and make up any shortfall in steam demand with a power boiler than to simply de-superheat. This has to do with the relative cost of electrical energy to the cost of fuel for the boilers.
The recovery boiler runs on black liquor and recovers the caustic used in the digester process. The fuel supply is determined primarily by the mill throughput.
Normally an islanded generator output varies with the load.
Normally a grid connected governor varies with the fuel or steam supply.
You have a special case where the output of the boiler varies with the amount of fuel available.
The primary use of the turbine is to supply steam at ~120 psi) and ~60 psi.
So the electrical output of the turbine varies as the lower pressure steam demand varies. (and as the boiler output varies).
The electrical energy is incidental. A pound of steam at ~600 psi has a lot more energy than a pound of steam at ~120 psi or 60 psi.
Given a supply of steam that is not under the control of the steam turbine (steam supply varies with the supply of black liquor to the boiler)
and an electrical power output that is not under the control of the turbine, the droop control is a back-up in the event of loss of the grid connection.
In the event that an upstream failure leaves your plant and part of the grid islanded, the droop control will limit the frequency rise.
In the event that you lose the electrical load the droop control will prevent a runaway.
In the event that the steam demand exceeds the production of the turbine, the de-superheater(s) will come online and supply the needed steam while robbing steam from the turbine, thus thus insuring the needed steam supply.
Under normal operating conditions the droop control may be inactive. It is a backup system that comes into play in the event of a loss of load.
A utility may have one swing set, and all other sets will be in droop control. If you want to connect to the grid, you must play by the utility's rules and run your set in droop control, even though the droop control will be inactive under normal conditions.


Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[GTstartup]

If this is in the, US all turbines above 10MW are required to operate in droop, no matter the mode. That being said they cannot make you do what cannot be done.

But..you could take the speed setpoint - actual speed deviation and (with a suitable gain) add that to the back pressure measurement (as a pressure offset so that the turbine would respond per the local regulations. You just have to correspond whatever drop/increase in pressure results in loading or deloading the turbine 100% and match that to the % droop required.

 

[stor138]

Yes GTstartup, this is in the US where droop is required at all times. I will look into your suggestion regarding creating a "power" droop by adjusting extraction pressure setpoint. The response would be dependent on header pressures at the time of the frequency deviation (eg - a frequency deviation while the inlet pressure is low would mean a SP change may not provide as much change in power as when in nominal conditions), but a reasonable estimate of the "power droop" through the operating range may be able to be determined through testing and design data.
waross - agree with most parts of your summary, with the only exception being that in our circumstance we are required to implement droop locally for support of the utility frequency. When islanded, the unit switches to isochronous.
Thank you