We have a Frame 6 GT driving our generator. I have always noticed that during auto synchronization the slip frequency seemed pretty fast. This is my first experience with generators so I have relied on the TA for information and was told that looked normal to him. We recently upgraded our protective relaying on our generator panel utilizing a SEL-700G. We hired the relay manufacturer to translate the mechanical relay settings into the microprocessor relay. I then looked over the translation to see if it "made sense" with my limited knowledge of generator relaying. One item that stood out was the slip frequency limit was specified at 0.067 Hz (per IEEE) compared to the 0.25 Hz of my original relay. When we started up the relay wouldn't sync so our electrical engineer had me change it to 0.25 hz and we were then able to sync as before. Now it is a couple of years later and I was looking into the synchronizing control in the logic and I have found that it appears to calculate the generator frequency based on the turbine speed. The logic lists 100% speed as 5105 and elsewhere in our documentation I find that the gearbox ratio is listed as 5106 to 3600. Our actual nameplate ratio is 5094 to 3600. Which explains why we always run 99.8% speed. This seems to me would explain why our slip frequency seems high since generator is being spinning faster that the controls think it is. Based on the fact that the TA stated that the slip speed looked normal to him, is this a common set up or is something wrong here. I think the control should be reset so that 100% is 5094. The system was originally controlled by a Mark IV and is now controlled by a Mark VI. I checked the old parameter documentation and it has been running this way for 20 years. I can't believe that no one has questioned this before if it is wrong.

Can anyone provide some input?

Thanks in Advance.

 

MWO,

The load gear nameplate <b>IS</b> the proper speed-to-frequency conversion. The Mark VI (control processors and <P>) should be configured for 5106 RPM equals 100.00% speed--<b>IF THAT IS WHAT'S LISTED ON THE GEARBOX NAMEPLATE, not some piece of documentation.</b>

This specific topic has been covered before, and no matter what the turbine nameplate or documentation says, the load gear nameplate wins. Every time. The turbines--the axial compressors, actually--have a range of speed they can operate at optimally, and load gears are built to optimize the strength of the "slugs" of metal used to construct them. So, there can be slight variations of actual turbine/axial compressor speed from design and the units can still operate optimally even if they are running a few RPM higher or a few RPM lower than design. The load gear nameplate wins. (And, yes, MANY turbines operate with incorrect 100% speed scaling for decades.)

As for the "proper" slip frequency, everyone has their own idea of what's proper. Most utilities and protective relay manufacturers seem to be very nervous with higher slips, which only result in higher initial loading when the generator breaker closes. The Speedtronic control systems can close breakers at a much higher slip than most people are comfortable with, and GE likes to have a relatively high positive power output on synchronization than most people are comfortable with.

You own the machine. Whatever you are comfortable with is "proper"--presuming the utility you are synchronizing to is comfortable with it. I have always believed that one reason GE chooses a higher slip is that anything between zero and the specified slip rate will result in synchronization when all the other permissives are met. And in many parts of the world, where the grid frequency isn't very stable, a larger slip "window" is better.

Again, you--as the owner or the owner's representative--are free to choose whatever slip you desire as the proper slip. Gas turbines aren't as touchy as some other prime movers, such as steam turbines or reciprocating engines which don't do well with being jerked back to synchronous speed when the generator breaker closes at a high slip frequency.

Hope this helps!

 

MWO... slip-rate is only one parameter during synchronization. What is the voltage differential?

Regards, Phil Corso

 

The new relay settings were:
Max voltage diff = 3.3%
Voltage Level = within 10% of nominal 13.8 KV
Angle within 10 degrees.
There didn't seem to ever be any issues there.

I don't have the logic in front of me but as I recall the dead band settings to control the slip frequency were such that it would try to control the slip rate between .04 and .08 hz. I calculate this would be holding the turbine (assuming a 60 hz grid frequency) between 5097 and 5100 rpm. Assuming the gas valve is calibrated properly and not sticking, can the Speedtronics control the rpm that finely or are are those unrealistic numbers? Three rpm seem pretty tight to me.

Looking at my old Mark IV logic, it appears that they had it controlling between 0.2% and 0.4% speed. This would equate to a dead band window of about 10 rpm . This would also make their specified setting of the slip rate of the electromechanical relay of 0.25 hz make sense since it would be just beyond the upper limit of the dead band control.

Any insight on how fine I can expect the control to be? Of course I realize it depends on the condition of a lot of things, but if all were working well what could I reasonably expect?

 

MWO... which generator voltage is higher; incoming, or bus? Can you monitor the voltage-signals via an oscillograph or a storage-scope?

Regards, Phil

 

MWO,

The 700G uses the generator terminal voltage to determine the frequency of the generator, and uses a synchronizing voltage from the bus to determine the frequency of the system. The mechanical speed of the turbine is not taken into account at all. So the slip the 700G is calculating is based purely on the difference between the measured electrical frequencies

The IEEE (C50.12 and 13) standards are a reference for the manufacturers of the generators and state that the machine should be built with the capability of withstanding synchronization at +-10degrees, 0-5% voltage difference, and +-0.067hz slip. If the GE generator can tolerate 0.25hz, then that just means it can withstand more impact than the standard requires. The problem with using such a high slip comes when/if the generator breaker no longer closes as quickly as originally designed. Many autosynch controllers use the breaker closing time when deciding when to issue the close command. Most breakers will operate in 2-5 cycles (meaning the time it takes for the close coil in the breaker to energize and the contacts to physically close is 32-90ms on a 60Hz system). If the autosync controller knows this, then it can time when it issues the close command so that the phase angle difference is very close to zero when the breaker actually closes - meaning at a given slip, the angle will change x number of degrees in the time it takes the breaker to close, therefore it can time when it issues the command to try to hit zero degrees phase angle difference). If the breaker, over time, begins to slow down and not close as fast, then you run the risk of out of phase closure- i.e. the autosync controller will issue the command when the parameters are met, but by the time the breaker actually closes, the generator will be out of phase with the bus. And with a slip of 0.25hz, the phase angle is changing pretty fast.

-nic

 

nic,

Thanks for the reply. I know that the 700G uses the PT signal but it is only serving as a backup check of the synchronization (permit closure)not to control it. This is done via the Mark VI controller which does use the turbine speed to calculate the generator frequency (why they do this instead of using the generator PT signal I don't know) The slip is calculated from this measurement and the bus PT signal . My question concerns the ability of the Mark VI to control the slip frequency that close.

 

> which generator voltage is higher; incoming, or bus? Can you
> monitor the voltage-signals via an oscillograph or a storage-scope?

Really not having a voltage matching problem. Exciter is set to control at 13.8 KV when it goes to AC control before synchronization and the line voltage is generally a little high. I would say that the auto synchronizing is having to adjust the voltage up if any at all. As I had said in my original post, once we increased the allowable slip setting to match GE's relay spec, it closed in fine. It just looks to be a pretty high slip frequency to me (0.25hz). Just trying to figure out if to control the slip at 0.06Hz is asking to much of the controls. We are in ERCOT's area so our grid frequency is fairly stable.

 

MWO... I presume the SEL Auto-Sync system uses an Anticipatory Logic Scheme!

If so, then the close-signal is initiated when the phase-angle between the incomer and the bus is a specific value.

Determination of the angle requires knowledge of the closing-time of the Circuit Breaker AND response-time of its Interposing Relay!

1) What values did you select?

2) What is the calculated phase-angle for the two cases cited?

Regards, Phil

 

It's beginning to become clear to me now. Sorry it's taken so long.

Mark IV turbine control systems only had a micro-synchronizing module and logic which closed the breaker in AUTO synch mode. In the breaker close circuit there was always an electro-mechanical relay contact in series with the AUTO breaker closure from the Mark IV. Both contacts had to close for the breaker close coil to be energized.

The electro-mechanical relay was designated the 25 relay, and it drove a telephone-type relay (because of the current in the breaker close circuit), 25X, that provided the "check" to the Mark IV's 25 relay output.

In fact, when Manual synchronization was selected, the 25/25X was still in the circuit to prevent an unconscious operator from "accidentally" closing the breaker out of synch. Again, this was a "check" to the operator's closing signal to make sure the breaker would not be closed out of synch.

Now, when GE retrofits most Mark IV control systems with Mark V, Mark VI and Mark VIe control systems they recommend eliminating the 25/25X relays from the breaker close circuit because the newer Speedtronic control systems have both a synch check ("manual") capability and an "automatic" capability which can be "wired" together to replace the 25/25X contact and the Mark IV micro-synchronizer contact.

Many sites do not heed GE's recommendation and leave the 25/25X relays in the circuit and GE does not "insert" the Mark V, Mark VI or Mark VIe synch-check contact in the breaker close string. However, leaving the old electro-mechanical relay in the breaker close string can cause problems because sometimes the newer Speedtronic control systems will try to close with a very fast slip and the electro-mechancial relay contacts won't ever close when the Mark nn contacts are closing.

It sounds like the original poster's site opted not to remove the 25/25X relays during the Mark VI upgrade, and are now replacing them with a new SEL relay and are questioning the relay settings. The manufacturer's settings may or may not have been calculated with the knowledge of the Mark VI's setting and configuration.

To answer one question, the Mark VI can control speed very, very well (presuming the fuel control valves are in good mechanical condition, and the servo-valves are in good condition). The Mark VI should not be the "limiting" factor in these considerations.

I will say again: As owner/operator you can decide to have whatever settings you choose in the protective relays. If you tighten up the settings, be prepared for the possibility that you may well introduce some unforeseen issues, and you may not. In my personal opinion, the old settings will not harm the generator or the turbine in any way (or the Load Gear, because I can hear the wheels turning). Sure, they can probably stand to be tightened up some, but how much should depend on a calculated system analysis of all the components (the SEL and the Mark VI).

Hope this helps!

 

I think what MWO is asking is (MWO, correct me if i'm wrong): should the Mark VI 100% speed setting be changed to 5094rpm to reflect the name plate value, or leave it at the same value it's been at for the last 20 years (5105rpm). The present setting (5105rpm) is what is causing the such a high slip when syncing to the bus.

And because I don't know, why would someone set the 100% speed setting at anything other than the nameplate? Are there circumstances when this is desirable?

nic

 

nic,

I don't think that speed scaling was the main question MWO was inquiring about. I completely misunderstood the original question and was wondering why someone would be using a SEL relay for "synchronization" instead of the Mark VI (some SEL relays can actually send signals to the turbine governor to do "speed matching" and "voltage matching" as part of automatic synchronization). The Mark VI (and the Mark V and the Mark VIe can do this without an external relay; but some Customers actually choose not to use the Speedtronic for automatic synchronization). So, my bad for not asking for more information originally.

As for the 100% speed scaling configuration, it must not be a big deal for GE because there are likely scores, if not more, units around the world which do not have the proper scaling configuration--and they still produce power. Reliably and without manual intervention? Not usually. And, sometimes people change synchronization parameters in a misguided attempt to make speed matching work, or make it work faster, when all that would really be necessary would be to change one single configuration value (two if it's a Mark VI or Mark VIe).

But, GE doesn't teach these things to their field service personnel, and so they usually go unchecked and uncorrected--unless there is a serious discrepancy and/or synchronization attempts are very unreliable and someone does more in-depth troubleshooting.

Does it hurt anything? Not really. Is it "clean" or proper? Not really. Is it critical? Obviously not. Is there ever a condition when it's desirable to have a speed scaling configuration higher or lower than it should be? No. Not for an AC generator drive turbine.

Lastly, I believe one reason GE's synch windows (on the synch-check function and sometimes on the auto synch functions) is that in some parts of the world, where the grid frequency can be changing as fast a +/-1.5 Hz in one or two seconds it's sometimes necessary to "open" the windows or limits just to get the unit to synch under these circumstances. And, people do actually try to synchronize under these (and worse!) conditions, and wonder why it can take a minute or two for the breaker to close, blaming the Speedtronic for an inability to change turbine speed fast enough to keep up with grid frequency changes.

The upshot of all of this is still: A complete analysis of the entire site operating conditions, including grid frequency stability and expected unit reliability need to be done before just saying one value or parameter is not correct. Sure, if the grid is typically stable and the governor (the Mark VI) and the fuel valves are in good condition and the fuel valve servos are in good condition, a wide synch-check window (which is the function I believe the SEL is being used to perform) may not be ideal. Will it hurt the unit? No. Can it be "improved"? Yes. But, care should be taken to consider the possibilities of grid instability when power is needed in an emergency and that breaker just has to close. And, if the window was set for ideal conditions then it might now work as needed under less than ideal conditions when power is required very quickly.

And the Mark VI will get blamed; it usually does. And someone will usually try replacing the fuel control valve servo or re-calibrating the fuel control valve LVDTs and then the next time the unit is synchronized (when the grid is more stable), the breaker will close very quickly and that damn servo will get all the blame!

Isn't this fun?

 

nic & CSA

Yes you are both understanding my issue.

1. Original Mark IV (1979 install) setting of 100% speed was 5106 instead of 5094. My original Mark IV control specs said a 60 hz machine should have 100% speed set to 5106 and a 50 hz machines set to 5094. I think they may have had it backwards. I found a pdf which was from a combustion turbine technical forum which listed several gearbox styles (non that were exactly mine). The ratios were 5106 to 3000 and 5094 to 3600. I would think the 5106 to 3000 was for 50hz machine.

2. In 2004 upgraded to Mark VI. and translation apparently followed old specs. GE did have the sync check from the Mark VI in series with the close signal. We opted to keep the electromechanical sync check relay in the circuit in addition to the Mark VI as CSA described.
3. In 2011 we upgraded the electromechanical relays to a microprocessor relay. The relay manufacturer engineer making the translation suggested a tighter setting on the SEL to comply to IEEE settings. He was not charged with looking at the Mark VI controls. When we failed to sync, the setting was changed back to match the original electromechanical setting (0.25hz).

We have submitted a PAC to GE to see if they agree with our plan to revise the 100% speed on our upcoming outage. I'll let you all know how we come out. Thanks for all the input

 

nic & MWO,

The problem comes in only when performing an Auto synch when the Speedtronic 100% speed scaling is incorrect.

For example, if the load gear nameplate is 5106 and the Mark VI is configured for 5094, when the turbine reaches 5094 RPM the Mark VI thinks it's at 100.00% speed. The typical synch speed for Speedtronic turbine control systems is 100.3% speed, or 5109 RPM. The generator frequency will be at 100.00% when the speed of the turbine rotor driving the load gear is 5106 RPM, so at 5109 RPM the synchroscope would just barely be creeping in the clockwise (FAST) direction. Because the Mark VI isn't looking at the frequency of the generator--it's ONLY looking at the turbine shaft speed pick-up inputs and converting that to percent of speed based on the 100% speed scaling configuration--it will think that 5109 RPM is 100.3% of rated speed (since it thinks 5094 RPM is 100% of rated speed) and will not be trying to increase the speed at all.

Sometimes, some electronic relays (such as the SEL probably is) need to see more slip (a higher frequency differential between the generator frequency and running frequency) in order to energize the breaker close output relay. The older electro-mechanical relay will operate just fine in this condition as the disc is rotating slowly and the contacts will close well in advance of the Mark VI synch contacts and remain closed when the Mark VI synch contacts close. Sometimes, narrowing the slip frequency window in an electronic relay will make it even more difficult to get a closure signal when the slip is very, very low.

This is a historic problem with the way Speedtronic turbine control panels calculate generator frequency (it's been around since the beginning of the Mark IV era). Even though Speedtronic panels have an input from the generator PT to measure generator terminal voltage, they don't use that PT signal to calculate generator frequency. Instead, Speedtronic panels use turbine shaft speed feedback to calculate generator frequency (in percent of rated frequency). However, they do use the bus (running) PT feedback signal to calculate percent of rated grid frequency and make the decisions to increase or decrease the turbine speed/generator frequency. If the scaling used for either of those conversions to percent is incorrect, then there can be problems.

It's always been explained to me that electro-mechanical relays don't like a lot of slip--a large difference between incoming and bus frequency. That's because the disc can only rotate so fast and sometimes the contacts can't close before the window closes if the generator frequency is much faster (or slower) than the grid frequency.

Electronic relays, on the other hand, need to see some minimum frequency difference and the higher the frequency difference the easier it is for electronic relays to calculate the slip rate, predict the proper window and issue the breaker close signal at the proper instant in time.

Now, if the Mark VI 100% speed scaling configuration value is changed to match the load gear nameplate, 5106 in this example, then when the turbine is at 5106 RPM the generator will be at 100.00% frequency, and when the Mark VI increases the turbine speed to 100.3% for synchronizing the turbine speed will increase to 5121 RPM and the frequency of the generator will increase to 100.3% and the electro-mechanical or electronic relay will each have enough slip frequency and differential to operate properly.

Electronic relays and electro-mechanical relays don't operate exactly alike. Yes; they both serve the same function but the way they make the determination about when it's safe to close the generator breaker is different. Each one has it's own idiosyncrasies. Electro-mechanical relays don't like a high frequency differential; electronic relays don't like a very low differential.

 

Just a follow up.

Changed 100% speed to 5094 to match the gearbox.
Adjusted the slip control dead band to the range of 0.04 Hz to 0.08 Hz.

Started up (3 days behind schedule due to mechanical issues) and system successfully sync on first pass. Unfortunately I was watching the sync scope on the control panel so I didn't see the actual slip frequency nor had it recording but based on the rate of sweep of the sync scope. I figured it was about in the center of the dead band. We are also now indicating 100% speed instead of 99.8 %

Thanks to all who helped

 

Sirs,
I found this topic very interesting and informative, and have a few questions regarding “SLIP”.

1) When synchronizing with high slip, how does this affect breaker contacts?

2) Does zero slip mean no current export upon synchronization?

3) Will a generator synchronize, if no slip is introduced?

Again, this was a great question and response from all.

 

>1) When synchronizing with high slip, how does this affect breaker contacts?

In AUTO, the system will take into account the closing speed of the breaker and calculate when to issue the close command so that the breaker contacts close when the two voltages are in phase. My concern was the mechanical forces as the rotating field went through essentially an instantaneous speed change. Maybe my concern is unwarranted.

> 2) Does zero slip mean no current export upon synchronization?

Yes I believe in theory the generator and the grid are exactly the same frequency (sync scope is stationary but not necessarily in sync unless its at 12 o'clock). A positive slip means you are running the generator a little faster than the grid. Once you tie to the grid, the generator immediately slows down to the grid speed. The extra energy you have driving the generator shifts the torque angle and produces electrical power flowing out of the generator.

> 3) Will a generator synchronize, if no slip is introduced?

If it is out of sync it has to adjust the speed (creating slip) to synchronize. I'm not sure if the Mark VI allows for zero slip or not. My external relay (SEL) has a window from slightly negative to 0.25 hz. The negative range prevents the contact from chattering, once the breaker closes.