Eng-Tips is the largest engineering community on the Internet
Intelligent Work Forums for Engineering Professionals
-
Congratulations waross on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!
Fast de-excitation brushless exciter
- Thread starter nelvex
- Start date
What is your application/problem?
I have spec'ed, installed and serviced a lot of generators in the 15 KVA to 1500 KVA range. The natural characteristics of the AVR were fast enough for load loss. Never had a problem. Standby sets would sometimes trip on over-speed when the load was abruptly dropped when switching back to the mains after an outage. That was a governor and setting issue. With 3% droop and no load the target frequency is 61.8 Hz (51.5 Hz). A little fast and the load dump would cause over-speed trips. A little slow and the UFRO would act prematurely.
A brushless exciter generates AC. This is rectified by the rotating diodes to supply the field. When the AVR cuts the output to the brushless exciter the output of the brushless exciter will drop almost instantly. The field and voltage will decay as the magnetic field collapses. The time to decay to zero is generally taken as 5 time constants, one time constant, in seconds being the product of the induction of the field coil in Henries and the resistance of the discharge path in Ohms. If that's not fast enough for you, you may have to come up with some type of rotating field forcing circuit to reverse bias the field current and force the current down.
In one time constant the variable will reach about 63% of the final value.
Bill
--------------------
"Why not the best?"
Jimmy Carter
I have spec'ed, installed and serviced a lot of generators in the 15 KVA to 1500 KVA range. The natural characteristics of the AVR were fast enough for load loss. Never had a problem. Standby sets would sometimes trip on over-speed when the load was abruptly dropped when switching back to the mains after an outage. That was a governor and setting issue. With 3% droop and no load the target frequency is 61.8 Hz (51.5 Hz). A little fast and the load dump would cause over-speed trips. A little slow and the UFRO would act prematurely.
A brushless exciter generates AC. This is rectified by the rotating diodes to supply the field. When the AVR cuts the output to the brushless exciter the output of the brushless exciter will drop almost instantly. The field and voltage will decay as the magnetic field collapses. The time to decay to zero is generally taken as 5 time constants, one time constant, in seconds being the product of the induction of the field coil in Henries and the resistance of the discharge path in Ohms. If that's not fast enough for you, you may have to come up with some type of rotating field forcing circuit to reverse bias the field current and force the current down.
In one time constant the variable will reach about 63% of the final value.
Bill
--------------------
"Why not the best?"
Jimmy Carter
FreddyNurk
Electrical
Basler has a number of excellent tech notes, though I don't recall if they specifically deal with 'reverse excitation' or 'fast de-excitation'.
If you look at the time constants in datasheets for machines in the size range that Waross mentioned, they're usually quite short, thus its not really a requirement.
I attended a Basler presentation last year where they discussed the aspects of 'reverse excitation', a feature only found in the larger specification AVRs. I admit I don't recall if it was specific to do with brushless excitation, as they also did a large presentation on static excitation (i.e. removing the exciter rotor / stator and connecting direct to the main field winding, normally with slip rings).
I would think it'd only come in for larger machines (probably about the size that ScottyUK deals in), certainly in the size range I deal in its not a requirement (which is about the same scale as Waross).
Have a look at the Basler website and see if they have any tech notes or application notes for it.
If you look at the time constants in datasheets for machines in the size range that Waross mentioned, they're usually quite short, thus its not really a requirement.
I attended a Basler presentation last year where they discussed the aspects of 'reverse excitation', a feature only found in the larger specification AVRs. I admit I don't recall if it was specific to do with brushless excitation, as they also did a large presentation on static excitation (i.e. removing the exciter rotor / stator and connecting direct to the main field winding, normally with slip rings).
I would think it'd only come in for larger machines (probably about the size that ScottyUK deals in), certainly in the size range I deal in its not a requirement (which is about the same scale as Waross).
Have a look at the Basler website and see if they have any tech notes or application notes for it.
- Thread starter
- #5
I'm am referring to a large 180MVA power generator which is equipped with an brushless exciter. At a fault in the stator winding arc must extinguished quickly. Due to the large time constant of generator excitation circuit the electric arc in the stator winding can damage high enough until the electric arc is extinguished. Therefore a method is required to decrease the time constant.
For waross..please give me more details about what is "you may have to come up with some type of rotating field forcing circuit to reverse bias the field current and force the current down."
For waross..please give me more details about what is "you may have to come up with some type of rotating field forcing circuit to reverse bias the field current and force the current down."
Try googling "negative field forcing" (Basler has several papers freely available, amongst others), although I'm not aware of this being used to shorten the time that a generator contributes to a fault as excitation is typically just turned off by the protection system. Negative field forcing primarily is used to speed up the response of the excitation system during operation of the unit, e.g., minimising overvoltages after a load rejection.
Most generators will have quite a bit of remnant flux, so the generator will continue to feed the fault as it spins down regardless of excitation time constants.
Most generators will have quite a bit of remnant flux, so the generator will continue to feed the fault as it spins down regardless of excitation time constants.
FreddyNurk
Electrical
Actually, time constants were one of the reasons that Basler were plugging their 'static excitation systems', i.e. removing the brushless excitation system and going direct, though it wasn't to minimise fault damage, it was to speed up the response of the unit.
As for fault response, I suspect mgtrp is right.
As for fault response, I suspect mgtrp is right.
LionelHutz
Electrical
I keep thinking about the typical brushless exciter and I don't believe you can do negative field forcing with one.
The typical exciter is a 3-phase pilot generator winding with 6 diodes attached feeding into the main field winding. You control the main field winding by controlling the field on the pilot generator. Overall, they rather simple setups.
To force the field off, you'd have to switch the rectifier to use 12 SCR's and a controller to for SCR's. Instead of dropping the pilot field to remove the excitation you'd have to be able to signal this controller to force the current to zero instead. In other words, you'd basically need a 4-quadrant DC drive rotating on the rotor shaft which you can control externally. It seems very unreasonable to expect to have that level of electronics rotating on the shaft.
The typical exciter is a 3-phase pilot generator winding with 6 diodes attached feeding into the main field winding. You control the main field winding by controlling the field on the pilot generator. Overall, they rather simple setups.
To force the field off, you'd have to switch the rectifier to use 12 SCR's and a controller to for SCR's. Instead of dropping the pilot field to remove the excitation you'd have to be able to signal this controller to force the current to zero instead. In other words, you'd basically need a 4-quadrant DC drive rotating on the rotor shaft which you can control externally. It seems very unreasonable to expect to have that level of electronics rotating on the shaft.
TheBlacksmith
Mechanical
Yes, thinking about it, negative brushless excitor current would be blocked by the rotating rectifier and have no effect on the main generator field.
A brushless exciter is basically an alternator. The AVR feeds DC to the field and the rotor generates AC. This AC is rectified by the rotating diodes to DC for the field coil. All the AVR can do is cut the DC to the field of the brushless exciter. ANY DC to the field of the brushless exciter will result in an AC output to the rotating diodes. If If the time constant of the brushless exciter field is where the issue is, no problem, A suitable AVR can reverse force the field effectively to kill the pilot field. (Well maybe some problems but doable). However if it is the time constant of the main field that is the issue, you may need some type of rotating circuit and some way of it knowing when to reverse bias the field. Not a problem with a brush type exciter.
To extinguish the arc, I would consider isolating the stator and interrupting the ground path. There may still be enough capacitive current to maintain an arc, but hopefully a less severe arc.
Bill
--------------------
"Why not the best?"
Jimmy Carter
To extinguish the arc, I would consider isolating the stator and interrupting the ground path. There may still be enough capacitive current to maintain an arc, but hopefully a less severe arc.
Bill
--------------------
"Why not the best?"
Jimmy Carter
- Status
Similar threads
- Locked
- Question
- Question
- Question