Reverse VAR


Thread Starter


I am getting a Reverse VAR fault on a Diesel Generator engine - I have looked around on this site and various others - and I have not found what would be a cause of RVAR's. If anyone has any input on what would be the cause that would be great.

RVARs. Why does everyone want to invent a new abbrevation or acronym?

What have you done to try to troubleshoot the problem?

Have you read the instruction manual for the reverse VAr detector/relay to see what it senses and how it works? Usually there is a good description of the condition being monitored and how it is monitored, and those will usually provide a very good indication of what the problem might be.

"Reverse" VArs are usually "caused" by a problem with the generator excitation system (exciter regulator; AVR; etc.) not maintaining sufficient excitation to prevent a "reverse" VAr condition from causing damage to the generator components. When connected to a grid (other generators and load) if the generator terminal voltage isn't maintained within limits, a "reverse" VAr condition can exist if the generator terminal voltage is allowed to go below the level required to make the generator terminal voltage equal to the grid voltage to which the generator is connected.

Have you noticed the generator terminal voltage lower than normal or some other abnormality with the generator excitation? Low field current? Low field volts?
I really did not try to "invent a acronym" this is how it is labeled in the CGCM - Allen Bradley's generator control module. Sorry for affending you!

AS stated above we are using the CGCM as a monitoring device for the generator. Everthing has been fine with the generator up till recently we started to get the "reverse" VAr fault from the CGCM - which in turn is bad because the CGCM cuts of excition when it sees the "reverse VAr" fault. Which of course does a good number to the system when at full load and just drops of the face of the earth.

The manual does not provide any good information on what to look for when this happens - just that it cuts excitation to protect the generator.

We are using AVR mode - the generator's voltage is about 100 - 150 volts higher then the Bus voltage to correct Power Factor.

Thanks for your help ; it is appreciated
I found the CGCM (Combined Generator Control Module) User Manual on the A-B website. In Chapter 7, Troublshooting, Protection, it says one of the most likely causes is under-excitation, and to check the UEL (Under-Excitation Limiter). Has the UEL configuration been checked? It should serve to prevent an under-excitation condition by setting a minimum excitation limit that is above the reverse VAr setpoint.

There is a procedure in the manual for simulating a reverse VAr condition.

The other is that the incorrect operating mode was selected. Has something changed in the way the unit is being operated?
This is a very clear example on what terms can come up when a process automation company does turbine controls ;)
My first suggestion would be to check all the fuse at the PPT and exciter. These guys can really cause some un imaginable tripping to occur. But since this is just a small diesel generator, I believe most of the parts are concealed within the casing and makes it hard for inspection. When the machine goes to FSNL, how does the terminal voltage, field voltage and field current behave? Once the machine is synchronized (or maybe closed on dead bus if this is for emergency diesel generator), hows does it behave to increasing load and what are the parameters just before the trip?
As far as I can think of is that the so to speak "reverse VAr" condition is normal in some grid system whereby the excess in VArs in the system makes the GSO to command the generators to absorb some of the VArs. This would in turn make the synchronous generator become an induction generator.
Was there any changes or maintenance carried out before the incident where by any chance the PTs and CTs may have been disturbed?
BTW what do you mean by "We are using AVR mode - the generator's voltage is about 100 - 150 volts higher then the Bus voltage to correct Power Factor"?
We have failed to ask when this problem occurs. Does it occur immediately after synchronizing? Does it occur at part load? Or full load?

Does it occur 20 minutes or 2 hours after synchronizing?

Does it occur at about the same time every day (say around 11:00 am, or 3:00 pm, or 7:00 pm)?

We also don't know where this diesel generator is installed and what the load characteristics are and what other generators may or may not be operating in parallel with this diesel generator, upstream or downstream of any step-up transformer.

So, there's a lot we don't know about this installation and this problem. In re-reading the original post, the request was for what would cause RVARs, and I believe we (and the CGCM User Manual) answered that question.

If one operates a synchronous generator at a relatively constant load with a constant generator terminal voltage (constant excitation) while connected to most grids, the VAr meter will fluctuate throughout the day, as will the power factor. As the system voltage goes higher, the VAr meter (which is usually operating in the Lagging condition and slightly greater than 0 VAr) will move "back" towards 0 VAr. If the excitation is held constant and the system voltage continues to increase, the VAr will start in increase in the Leading "direction" (increase above 0 VAr on the Leading side of the VAr meter--presuming it's an analog meter; many times digitial meters are configured such that Lagging VArs on a generator are positive, and Leading VArs are negative).

As the system voltage goes lower, the VAr meter will "increase" in the Lagging "direction (presuming the load is relatively stable and the generator terminal voltage is not changing).

When a generator AVR is being operated in either VAr or power factor control, the AVR will adjust the excitation to try to maintain a VAr or power factor setpoint.
I'll look more into simulating the Reverse Var condition in the manual - I did read that section before but did not believe we were in a underexcitation when this happens. Nothing has changed in the way we are operating the engine. We have 2 other installs of the same exact equipment with no problems like this.

AB's only suggestion is to replace the CGCM - like their answer for everything else. Plus AB will not answer application engineering question if we ask them about "reverse VAr" - it could be a setting but Highly unlikely since we do not get this in the other units.

We are trending everything under the sun in Bus voltage to generator phasing and currents - when we looked back at when it tripped nothing seemed out of the ordinary.

About the generators AVR mode holding the generators voltage higher then the Bus voltage to correct PF. - AB suggested we do this because when we first started the project on the first unit - our PF was off and thats what they said we had to do - Add or Subtract AVR setpoint to try to maintain PF unity.

This problem is very intermittent - they ran the other day with no problems. So I am not sure - there is very smart guys at the power plant and they don't even know what is going on --- BTW us process control guys can work on more than one thing :)
VArs and power factor are related. If the VAr meter reads 0, the power factor will be unity (1.0), which essentially says that 100% of the energy being put into the generator is going out as "real" power, watts.

Let's say the VAr meter is indicating 2.3 MVAr Lagging. The power factor meter will be indicating something less than unity (less than 1.0), also in the Lagging direction. If the VAr meter reads 1.7 MVAr, Leading, the power factor meter will be reading something less than 1.0, but also in the Leading direction.

When the generator terminal voltage is equal to the Bus voltage, the VAr meter will be equal to 0 and the power factor meter will read 1.0 (unity). This means the amount of excitation being applied to the generator rotor is exactly equal to the amount required to make the generator terminal voltage equal to the Bus voltage.

If the amount of excitation being applied to the generator rotor is less than the amount required to make the generator terminal voltage equal to the Bus voltage, then the VAr meter will be indicating Leading VArs (sometimes this is shown as a negative number on digital meters), and the power factor meter will be indicating a Leading power factor less than 1.0 (sometimes this is shown as a negative number on digital meters).

If the amount of excitation being applied to the generator rotor is greater than the amount required to make the generator terminal voltage equal to the Bus voltage, then the VAr meter will be indicating Lagging VArs (sometimes this is shown as a positive number on digital meters), and the power factor meter will be indicating a Lagging power factor less than 1.0 (sometimes this is shown as a positive number on digital meters).

Bus voltage generally goes up and down during the day depending on load conditions and on how the grid system operators respond to load changes. If one does *nothing* to the exciter controls, the VArs will change and the power factor will vary during the day (this is presuming load, power output, remains fairly constant). Why? Because the excitation is remaining constant but the Bus voltage is changing. So, the VAr meter and the power factor meter are changing.

That's what VAr and power factor control do: they adjust the excitation to maintain a VAr setpoint or a power factor setpoint, respectively, to respond to changes in Bus voltage.

It seems, from your description and from very briefly skimming through the manual, that the CGCMs are not all configured equally.

If the rVAr setpoint of one of the CGCMs is "higher" than the other(s), and not less than the under-excited limit setpoint (i.e., greater than the under-excited limit setpoint) *and* the Bus voltage was "unusually" high, it's possible that the CGCM was trying to reduce excitation to a level that was above the UEL but less than the rVAr setpoint in order to try to maintain a setpoint.

I haven't read the manual that closely, but I would suspect there is some way to "set" the PT (potential transformer) input sensing to match the PT outputs (6.9 KV equals 115 VAC, or 6.9 KV equals 120 VAC, etc.). It's possible that the CGCM having the problems isn't adjusted the same as the others.

If neither of the above is the case, then I would suspect the CGCM. Something in it's sensing circuits doesn't seem to be correct.

There are lots of other things that could be happening. We don't know how the generators are connected, whether they are on the same bus upstream or downstream of step-up transformers, whether there is one, two or three step-up transformers, or whether the transformers have any kind of tap-changers and how they are adjusted. There is cross-current ("droop") compensation on most exciter regulators and this may not be set similarly, and it's entirely likely that one or more of the cross-current compensation circuits must be set differently than the others.

We are making several presumptions about the configuration of your plant. Have you contacted the packager of the diesel-generators for assistance? A-B people are pretty good with the guts and bolts of their equipment, but not always with the processes their equipment is applied to. The packager should be more helpful when it comes to problems like this if it's not a configuration issue.

Request some insight from you on the below condition:

Generator Detail:
We have a LV Caterpillar Gas engine Generator with rated specs of 150KW, 1505RPM, 0.8PF . The engine is also Caterpillar G3406 with full load power of 138KW and full load speed 1505rpm. Local controller is CAT Digital Voltage regulator(with AVR, VAR, PF control) and remote controller is Deep Sea controller on the LV Switchgear. The load on the generator is usually low i.e. 30%. Mostly capacitive .e lighting etc.

The generator trips of reverse VARs when a 15KW motor is turned off after it is started. This started happening after a year of smooth running. No AVR set-points or anything has changed.

My thought:

It seems that when the motor is turned off, the Vars suddenly change direction from the bus to the generator and and the generator trips on reverse Var. What do you guys think? But if this case is true than it should have been happening from the start, but it started happening now.

What can be the probable cause here? What else I need to check?

Looking forward to your responses.
1) What do you mean by "when a 15kW motor is turned-off after it is started"?
2) Do you mean a 'particular' motor?
3) Does it turns off by itself during the run-up?
4) Or, do you mean it is deliberately turned-off by someone before it reaches full speed?
5) Is the motor operated with a VFD?
6) What is the motor driving?
7) What are the Gen-set's parameters just before the motor is turned off?
8) How large is the 'capacitive' load?
Phil Corso
Apologies for the delay, this Corona outbreak has us in shifts.

Kindly, find the answers as under:

1) What do you mean by "when a 15kW motor is turned-off after it is started"?
2) Do you mean a 'particular' motor?
3) Does it turns off by itself during the run-up?
4) Or, do you mean it is deliberately turned-off by someone before it reaches full speed?
5) Is the motor operated with a VFD?
6) What is the motor driving?

So for all above it is a pre-lube oil motor, only runs for about 2 mins. It has a DOL starter. No VFD on the switch-gear. It is turned ON and OFF by OPS.

7) What are the Gen-set's parameters just before the motor is turned off?
8) How large is the 'capacitive' load?

So let me elaborate the scheme.
There are 2 Gas generators and 1 backup EDG on this switchgear. GGs are of same rating i.e. 150KW and the Backup EDG standby power is 140KW.
Normally Operations is running the site(It is a remote Gas Compression site aside from our main plant at some 8KMs distance) at 1 GG. But when they have to start a bigger load i.e. an air compressor of 35KWs and few other loads, they start the other GG and sync it with running one(suppose to happen on auto through deepsea controllers). Tags are GG901A, GG901B and EDG-901C.

When this issue happened GG901A was running alone, they started the pre-lube oil motor(It is a smaller load and they do not generally start the other GG for this load), Then after the process conditions had been met, they turned it off and the running GG i.e. 901A tripped on Gen reverse VAr. Upon it's tripping, the backup EDG kicked in and sustained the site's power.

Gen's when running alone(WITHOUT lube oil motor running)have following parameters(L1, L2, L3):

V- 414, 417, 415
I- 74,73,76
KW-17.6, 17.6,18.4(44%, 44%, 46%)
Total - 53.7KW(44% loaded)
KVA - 17.5, 17.5, 18.5
Total - 53.9KVA(35% loaded)
Freq- 50.1
PF- .98(lag), 1, 1
Avg Pf -0.99
KVAr - 2.7, 2.7, 2.8
Total - 8.5KVAr(9% loaded)
Gov- 13.2%, AVR -9.9%

Gen's when running alone(WITH lube oil motor running)have following parameters(L1, L2, L3):

V- 414, 417, 416
I- 85,82,85
KW-19.5, 20.4,20.3(49%, 55%, 50%)
Total - 62.5KW(52% loaded)
KVA - 20.3 20.2, 20.4
Total - 63.1KVA(42% loaded)
Freq- 50.1
PF- .96(lag), 0.99, 0.99
Avg Pf -0.98(lag)
KVAr - 5.1, 5.6, 5.1
Total - 15.8KVAr(17% loaded)
Gov- 13.2%, AVR -8.7%

So we did some more attempts and this time there was no tripping on Gen Reverse VAR. I don't know what happened back then.

Another issue that is happening is that GG-901A does not sync with 901B. However, 901B syncs with 901A.

I have noticed that on NO-LOAD condition the RPM of GG901 varies from 1467 to 1515 and freq varies from 49.0 to 50.5, where as on GG901B on NO-LOAD condition RPM only varies from 1500 to 1515 only and freq is 50.0 to 50.4 only.

I requested OPS to get the fuel filters checked and serviced for GG-901A. They say they have had that done. I don't know what else can attribute to this variation in rpm and freq due to which I think it is not syncing as well. Can anyone shed some light? What should I do for the syncing problem?

Looking forward to your insights.



I don't respond to DMs/PMs through The big benefit of a World Wide Web forum like this is that MANY people can get to see the questions and responses and exchanges and learn from them. Going off-line, which some responders are often wont to do, defeats that purpose entirely. One person benefits from that exchange, and the really disheartening aspect (for me) is the same questions get asked over and over again (not that they don't here; everyone has a slightly different question, even though most are very similar if not virtually the same).

I'm presuming (because it hasn't been described) that there is some control scheme which is controlling the frequency the GGs and EDG are powering, because without some sort of control scheme one generator and prime mover should be in Isochrononous Speed Control, and any generator being synchronized to it should be in Droop Speed Control. That doesn't sound like the case, though, so it would appear there is some "external" third-party control scheme which is performing the frequency control by sending signals to the various generators and prime movers to maintain frequency as load changes--including synchronization.

I think you're on the right track of thinking it's something amiss with the fuel delivery system to the larger GGs, or at least one of them. By "gas generators" and "fuel filters" I'm presuming you are referring to gasoline-powered prime movers.; if it was natural gas or LPG I haven't seen many of those with changeable fuel filters. So, let's presume the motor pool has indeed at least checked, if not changed, the fuel filters. It could be a problem with the carburetor or fuel injection system of one or both of the GGs, with one being affected more than the other. (I presume they share the same fuel source, so this is a possibility.)

If it's not the fuel filters or the fuel (many times fuel can be contaminated with water), then I'm going to suspect either the GG prime mover governor is not accepting the signals from the frequency control system (including the synchronization system ("deepsea" I think you called it), OR, the frequency control system signals are somehow being compromised (electrical noise; intermittent grounds and/or shorts; etc.).

You have provided a good deal of information, but methinks there are still some details still not known (such as how the frequency of the system of the three generators are powering is being controlled if they are being operated in Droop Speed Control). It's conceivable the loads are fairly constant and well known to the operators who can anticipate changes and manually respond as necessary--but that would require lots of training for the operators, and experience, and these days Companies are more likely to spend money on automation and controls instead of training....

We're not talking about large loads here and it sounds like they are fairly stable and well-known and not subject to suddenly switching on and off. And, if the operators are well-trained then it's possible there is no frequency control being used for the prime movers and generators, but that would then throw the monitoring equipment into question (voltage; frequency; load; reactive load)--which is a possibility, also. And, it doesn't sound like the system the generators are powering has to have rock-solid frequency control, either (desirable, but not absolutely required).

As for the reverse power problem, if there's a frequency control scheme then it's possible there's also a voltage (VAr) control scheme for this small power system, that would also be sending signals to the AVRs (Automatic Voltage Regulators; generator exciter regulators) to control system voltage and try to limit VAr excursions. Seems unlikely, though, given the small nature of the power system and loads, and it's just likely that one of the AVRs was just sluggish in responding to the change in VAr load/system voltage. OR, if this is all being done manually by operators (from a remote location!) it's very possible that someone didn't respond appropriately to the loss of the 15 kW load. (Stranger things have happened; again--remotely operating multiple synchronous generators manually would take some pretty good operators with good training and a fair amount of experience.)

It's also possible that the relay which detected the reverse VAr flow was adjusted or set to a very sensitive value without properly taking into account the nature of the loads and system operation. Again, we're just talking about 150 kW generators a single 15 kW load; I know these things are all important and relative, but sometimes the settings and measurement systems for these smaller systems are not well thought-out and don't have the accuracy some larger systems get and have.

Lacking further details, that's about all I can add. In my experience it's not normal to run multiply synchronous generators in parallel in Droop Speed Control without some means of controlling frequency (which Droop Speed Control doesn't do very well and which Isochronous Speed Control does do very well). Again, I want to be clear: Without some means of controlling frequency when load changes on a small generation and distribution system (an "islanded" system not connected to a larger grid with many other generators and their prime movers) it's not normal to have all generators synchronized together and all operating in Droop Speed Control--unless the loads are fairly constant and stable and the operators have good training and experience.

If the site is remotely operated, when was the last time a crew was sent to perform cleaning and maintenance on the control system and wiring? Dust and humidity can combine to be very disruptive. And, if the site is located in a hot, humid area it probably has air conditioning which must also be maintained (filters cleaned; thermostats that can drift and not work properly; etc.). And heat is also another factor which causes electronics to not work optimally, and even fail over time. Especially when combined with dust and humidity. Cleanliness is important.

That's about all I can add at this time with the information provided. Hopefully, you have some food for thought here and you will find the cause of the problem(s) and write back to let us know what you find. Feedback is the big reason I believe is as valuable as it is to the people who use it (at least those who follow the threads in an effort to learn and grow their knowledge). Without feedback, it's difficult to know how problem(s) were solved. And, for those providing the information to try to help solve problems it's helpful for them to know if the information provided was helpful or not--many of us use the feedback to help us learn how to respond more concisely or add to future responses based on the experience and feedback provided by original posters.
1) In your very 1st post you mentioned... "Mostly capacitive, i.e.,lighting etc." My question... How much?
2) Exactly which protective device(s) "Trip" the unit?
Regards, Phil Corso
Hello and good day to all !

After i had read the differents post on the thread " REVERSE VAR" i would recommmand to have a better idea of what is called droop compensation or cross current compensation in regards of Diesel GENSETS operating in parrallel.

Can you tell us what kind of facility you working on i mean also what are the differents loads ( inductive/ capacitive ) in your plant grid...
I worked back in the days as " Generator commissioning engineer" for one of the major OEM ( gas /steam turbine projects )

Hence to my experience i could try to support you on finding what could be the root cause of the problems you are facing with.

I would suggest to who this may concern these three interesting articles found on the web :




You stated that generator terminal voltage reached from 100 till 150 than Bus voltage wich is sign of overexcitation first occured at this Genset

Then Reverse Vars (Lagging or leading Vars ) Occurs but the amount was too big and not shared by both Gensets so One of them tripped but we do not have information on what kind of controls /limiter tripped the genset ( is that by underexciation ??? ) it is an hypothese .

You should get some track on these events i pretty sure with AB AVR CGCM software .

For me i would have a look on AVR TRENDS VIA AB software see if you can catch the event wich is triggering Reverse Var from the grid ...

Also do you got type of PMS or AGC whatever it called at your plant ....??

Do you know if you got the Auto load sharing /shedding function implemented in your plant controls design???

That can be one of the track to follow in this case.

Assuming and according to you:
All parameters are all same for both AVRs on the 2 GENSETS so i presuming that:

It Looks like Load sharing/shedding is not operated properly in the state of the art as i can state first by the elements that you provided till now.
Again we do not have overall process controls of this plant to state any useful information till now.

Lets give us more answer to the details i asked for so as soon as you did that we can investigate better for a good and simple way of troubleshooting.

Hope this can be a help
Thank you very much for the responses. So here is how we solved the problem.

1- The Reverse VARs issue
I checked the Qmin trip setting in the GG controller(CAT EMCP 4.2) and it was set at 0.10pu. I checked the Generator capability curve and increased the setting to Qmin 0.15pu(leading region), as according to the curve the Gen could safely operate till 0.20pu in the capability curve.


Secondly, I asked Operations to start a 15KW motor when they want to turn off the 35KW motor. Once the 35KW is off, then after 10mins, turn off the 15KW motor as well. I did that as I reckoned that when the 35KW motor is turned off the VARs in the system return to the generator and cause the REVERSE VARs protection trip(as there is no other inductive load). Now since there is a lesser inductive load in the system, the VARs will flow towards that motor than the the generator. Then when 15KW motor is turned off, lesser amount of VARs flow to the generator and hence, does not trigger the Reverse VARs protection.
I tested the scenario two times(without increasing the Qmin set point) and it was successful. The GG did not trip on Reverse Vars on turning off 35KW motor when 15KW motor was also turned on.
What do you guys think?

2- The Hunting of RPM of GG and tripping on ''Out of sync''
I checked one by one the complete wiring of the GG. I found a one or two wires of the Governor controller loose. We cleaned and tightened these. Moreover, we also cleaned and serviced the butterfly valve that controlled gas input to the GG through the governor control card.
This resolved the problem.

Hope this helps someone.

Can you elaborate on the cross current compensation concept ?
Hello all,


As I suggested you before (seee my previous post), If you want to have a good overview and picture of cross current compensation, I strongly suggest you to read that excellent article:

I will review your last post and make comment after that.

I am glad that you solved the issue, you been faced on site.

Reur question about Cross-Current Compensation....
I suggest you search for Jojo's 2007 response in archives as #: 1026241828 ! He suggests some rather simple fact-finding procedures. I would also suggest you start a new thread .
Regards, Phil Corso