3 Generator Load Sharing Question

[ThePunisher]

HI all,

In one the facilities of our system, I can help notice this and took a photo.

All three generators are running under-loaded and the power factor is 1.09 of one and around 0.50~0.51 on the other two. If these generators are running in load sharing mode, why are their kvars or kVA not equal?

This is for my educational purpose only.

Regards,

 

[waross]

kW output and kW sharing are controlled by the throttle positions.
KVA output and KVA sharing are controlled by the both the voltage settings and the throttle settings.
Tweak the voltage settings to balance the KVA currents.
Tweak the voltage up a little on the set with the best PF and it will take more of its share of the KVARs.

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

 

[waross]

By the way,

power factor is 1.09

This is unacceptable.
The maximum power factor is 1.00
A power factor of 1.09 is flirting with over-unity and you know that we don't do over-unity on this site,
This is probably a typo or, if you are calculating the PF from meter readings, it may be a metering error.
Did you mean 1/1.09?
That would be about 0.92 PF.


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

 

[catserveng]

Is this a shipboard system?

Some SCADA systems will sometimes show leading power factor as a "1.XX" depending on where the data came from and how it was parsed. Seems to be a poor practise but shows up once in a while on systems I work on.

Your kVAR share is poor, depending on how your control system is sharing kVAR loads, could be a simple adjustment or need some work. Do you know how the reactive power sharing is being done? Your load sharing controls could have a VAR share feature, or the units could be in cross current or in droop. Need to know how it was supposed to work before you could get a decent answer as to what the solution is.

Hope that helps, MikeL

 

[ThePunisher]

Yes its in a shipboard system, we were on-board and managed to get a photo before we left when it docked.

 

[che12345]

Dear Mr. ThePunisher

Q. " All three generators are running under-loaded and the power factor is 1.09 of one and around 0.50~0.51 on the other two. If these generators are running in load sharing mode, why are their kvars or kVA not equal? "
A 1 . Based on the display; DG2, 3 and 4 are in [parallel operation].
DG 2 3 4
% 22 23 24
kW 748 757 792
V 6731 6720 6687
A 128 69 171
Hz 60.1 60.1 60.1
Pf 0.51 1.09 0.5
A 2. I am of the opinion:
a) three generators are in operation, each at about 23% load is NOT a good practice. Run one gen would be able to handle the load economically,
b) Slight deviation in kW is NOT a major problem, needs only slight adjustment,
c) The voltage deviation is [unusual], unless it is displaying the individual DG's generated voltage. When three Gens are in parallel operation, the display should be the [busbar voltage], should NOT be the [individual} generator's voltage,
d) There is a great deviation in the current. This is NOT acceptable. Need adjustment,
e) This may be the (individual) Gen's G
generated frequency at 60.1Hz, which is fine. When three Gens are in parallel operation, the display should be the [busbar frequency], should NOT be the [individual] generator's frequency,
f) Large deviation in pf is NOT acceptable. Pf at 0.5, 0.51 are fine, but NOT 1.09 which is unusual. Need adjustable. Note: Pf should be lead 0.x .... 1 .... 0.x lag. The practical working range should be within the (lag) section, NOT on the (lead).
A 3. Propose action:
a) Shut down two Gens, run only one,
b) (Toggle) the [speed regulator], increase the speed of Gen 2 and 3 to take up more kW,
c) (Fine tune) the [voltage regulator] rheostat, increase the voltage of Gen 3 and 4,
d) Adjust b) and c) to about equal on all three Gens,
e) At very light load with busbar frequency 60.1 Hz is fine. The frequency will drop as load increases. NO action to be taken,
f) (Fine tune) the [voltage regulator] rheostat. Observe the current and pf carefully, as slight adjustment may result to marked changes. Fine turn till the pf is in the [Lag] scale, with three Gens at about the same. Caution: wrong adjustment may result to the pf going into the [Lead], which is NOT acceptable. BTW: the DG3 displaying pf [ 1.09 ] could? mean that it is [Leading], which is NOT acceptable.
Che Kuan Yau (Singapore)

 

[crshears]

Running just one generator may be unacceptable aboard ship, as this does not meet N-1 criteria [ or N+1, depending on which end you're looking at it from ] - or, to put it in simpler terms, on loss of the one generator the entire ship would go dark, whereas with two I/S it won't; if, on first contingency, the load exceeds the capability of the smallest remaining generator, it will be necessary to run three gen sets, barring the presence of some sort of load rejection facility to address this concern.

I'm no electronic governor expert, but it would seem logical to me that it should be quite possible for the real power control mode [via governors / load sharing arrangements ] to be different than the reactive control / voltage regulation mode, as, within limits of course, these can be adjusted separately and independently.

CR

"As iron sharpens iron, so one person sharpens another." [Proverbs 27:17, NIV]

 

[ThePunisher]

I have requested for the generator control settings and operational narrative (hope it is available).

I was thinking that probably, the AVR settings for DG3 is required to increase the voltage/excitation so DG3 will take some kVars and operate on lagging power factor and increase the lagging PF of the other two (increase kvar loading on them).

If by chance, all three are operating at droop mode, is it possible that maybe their droop setpoints are not equal to a large degree? The frequency is 60.1 Hz all the same but the voltage and current loading are not equal?

Is it also a case where tow of them might be in droop and one is at isochronous?

 

[catserveng]

So when working with kVAR's you need to think not only about magnitude but direction as well.

It looks like DG3 is importing some VAR's, while DG4 and DG2 appear to exporting more VAR's than they should be in a balanced system. Increasing excitation on DG3 should cause the generator to start exporting VAR's and the VAR export on the other two should drop. The total VAR load is dependent your the ship's loads, but ideally on your bus they all should share within about 2%

Knowing how you are supposed to be sharing VAR's is the best starting point. Then you should find out what kind of AVR's are installed and how their droop circuits are connected.

If all the units have digital AVR's like Basler DECS or ABB Unitrol units, then verifying droop setpoints and other programming parameters would be a good start.

If the units are using analog AVR's, and have droop potentiometers in the droop circuit, then it is a good possibility that the droop slopes may not be the same and need to be verified.

If you have some form of an active reactive power controller in the power control system, you still should verify that the droop characteristics for all the units is the same, then take a look at the reactive power sharing settings and parameters.

Hope that helps, MikeL.

 

[waross]

If you change the setpoint of only one droop governor of a parallel set, that will affect the throttle setting and the amount of load that generator picks up.
Mr. Che;
It was a long time ago, but I remember running three generators on a floating fish processing plant.
One generator would have been lightly loaded.
However, when the large refrigeration compressor started automatically, it took all three sets to carry the starting surge.
With only one or two sets online, the generator breakers would trip and the vessel would go dark.
On a ship, when the Captain needs the bow thrusters, he needs them now.
No time to ask;
"Are there enough generators online?"
CR;
Change KW sharing by changing the throttle setting. This may be done with a droop governor by changing the frequency set point.
Change KVAR sharing by adjusting the voltage on individual sets.

Is it also a case where tow of them might be in droop and one is at isochronous?

Highly unlikely.
With one set in isochronous, the iso set will act as a swing set and compensate for any changes in load.
With a load increase, all sets will share the increase.
Then the iso will become active and the iso set will pick up the load increase and the droop sets will return to their original load levels.
The only time when the loads will be equal on all three sets is when the load happens to be tree times the load setting on the droop controlled sets.


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

 

[crshears]

CR;
Change KW sharing by changing the throttle setting. This may be done with a droop governor by changing the frequency set point.
Change KVAR sharing by adjusting the voltage on individual sets.

I knew that, Bill; my point was that analogous to the way governors can be isochronous or droop and have temporary as well as permanent speed droop and load sharing capabilities, AVR's can be set up the same way, which is what Mike L was getting at, if I understood him correctly, that is.

CR

"As iron sharpens iron, so one person sharpens another." [Proverbs 27:17, NIV]

 

[ThePunisher]

Hi all please find Governor Settings attached to spice up the discussion a bit

 

[ThePunisher]

Please find screen shot of monitor...it may have changed a bit

 

[ThePunisher]

Generator nameplate

 

[ThePunisher]

Snap shot of their OPS manual (simplified instructions, no details)

 

[ThePunisher]

Looking at the DROOP setting of 5% of the WOODWARD UG-25 and based on the slope, I am a bit curious. If DG-3 is running at 757 kW or 23% of Full Load. If all three generators have the same DROOP setting, would the frequency of the DG supposed to rise up to greater than 60.1%?

I am still really concerned about DG3 not taking its rightful share of kvars and operating at a leading power factor...I am curious why is this occuring. When running at one generator or two probably, the load will force the two generators to share equally?

 

[ThePunisher]

Secondly, the output currents do not seem to add up:

DG4 (and DG 2):
792 kW, 6,687V, PF = 0.50 and the current is 171 A
Calculation shows kVA = 1,584 and should be 136.76 A at 6,687 V

DG3 seems to add up
757 kW, 6,720V, PF = 1.09 or 0.9174 leading and the current is 69 A
Calculation shows kVA = 825.13 and at 6,720V will be 70.89 A

Am I missing something or misapplying Ohms law here?

 

[crshears]

Definitely not a plethora of information there, no fault of yours. ESL issues are patently clear...but not insurmountable, either.

I stand by my original statement:

I'm no electronic governor expert, but it would seem logical to me that it should be quite possible for the real power control mode [via governors / load sharing arrangements ] to be different than the reactive control / voltage regulation mode, as, within limits of course, these can be adjusted separately and independently.

 

[catserveng]

On your governor. the UG-25+ is a standalone governor that has the ability to take remote speed adjustments, either with raise/lower inputs or a 4-20mA speed reference. On systems I have worked on where there is an overall load control system, it is usually recommended to set the droop to 5% to give the load control system more authority and helps provide more stable operation. That governor can also be enabled to use a boost sensor (MAP sensor) for load limiting. From the front it looks like a strictly mechanical governor, but it actually electronically controlled and requires a service tool (laptop) to interface to. Woodward manual 26330 for reference if you want to read up more on it.

Is it possible the manual pages you show are for "manual" operation of the system if the automated load control system is not working? Or is the system actually only manual and the engineer isn't doing his job completely?

So if this is truly a manually operated system (which I would think unlikely but not unheard of, especially if an older vessel) then the operating instruction is pretty clear and management of real and reactive load sharing and voltage and frequency control are all within the responsibility of the operator to manage.

It is more likely that the power system does have an automation system for managing power sharing and frequency and voltage control, and because it is a shipboard system manual control is available as a backup. Now the question is, does it in fact have an automated power control system and when was the last time it was calibrated and tuned/serviced?

As pointed out above, the metering for voltage is suspect, if the units are all on the same bus they are really the same voltage within a few volts. And trying to do the math doesn't seem to come out quite right, so it is possible there are some other metering issues as well.

To get better answers I think would require a lot more info about the system. Hope that helps, MikeL

 

[Bill West]

Before we critique the voltage figures maybe we should consider the VT-analog transducer-SCADA A-D converter accuracy. The 6671-6687-6720-6731 volts we see in the photo are within 0.5% of each other and so seem to me to be fair info for what is on the busbar. I had this same "multiple voltage figures on a single bus" complaint from a 500kv system dispatcher when we moved to digital displays with 0.2% resolution but had poorer than 1% cumulative accuracy sources.

The only figures that are unreasonable on this display are var/PF ones where one generator is clearly moving a lot of vars to the other two. For that I notice that the operations book mentions a busbar governor. You don't suppose the automation designer programmed the computer to do the system operator's job of fine tuning the system frequency through common adjustment of the speed settings on the 3 governors, IE maybe we have an overlay governor working here. If so then maybe there's an overlay AVR working too and it's either not switched on to DG3 or it's not starting with the 3 AVR's set equally.

The significance to frontline operators who are not expert in power system theories may not be apparent. I went to a presentation where a maintenance engineer was explaining a system solution to two interconnected diesel station operators independently fighting over volts and vars, the first audience question was why didn't he just tell them that they can't both be in charge. It demonstrated that a good operator is not necessarily a good theorist.

Bill