OK try to picture this, i'm on a ship with 4 2282 kVA Generators,Gens 1 & 2 feeding one switchboard and 3 & 4 feeding another. Both switchboards can be connected together via a bustie.

The whole ship can easily run off 2 Generators, but whilst sailing we run 3 as a precautionary measure in case one of them fails. The law states that when we are in DP ("Dynamic Positioning" a setting which hold the ship in one position) we must have all 4 Generators connected with the Bustie open.

Everything has been running fine until the company decided to install an AGS system which will Automatically stop & start Generators as per load requirements. Apparently this bypasses the law requiring the 4 Generators and results in significant fuel savings.

The problem is this system has quite a few fail-safes, one being it will disconnect any Generator with more than 10% kVar variance. As stated in my summary we have one with approximately 30% variance, which renders the system unusable.

Any suggestions help would be greatly appreciated.

 

Tharris...
the reason for the 30% difference in kVAr can be found in the following Control.Com thread:

http://www.control.com/thread/1419896241, "The Physics of Tap-Change."

Specifically, read the Jan-08 post. When applied to your case, it illustrates that lagging-kVAr is produced when the generator's air-gap voltage is higher than its terminal voltage.

Of course, terminal voltage is "fixed" and air-gap voltage can't be measured directly. But, it can be "inferred" by noting one of parameters, field-excitation. That is, power to the rotor-winding, i.e., Vr x Ir.


So, if you haven't already compared the Operating Parameters of the four generators, I suggest it as a good start.

Regards,
Phil Corso

 

Tharris,

> Of course, terminal voltage is "fixed" and air-gap voltage
> can't be measured directly.
> But, it can be "inferred" by noting one of parameters, field-excitation. That is, power
> to the rotor-winding, i.e., Vr x Ir.

What the responder was trying to say is: the problem lies in the difference in excitation being applied to the machines. Of course that was intuitively obvious from the maths and unexplained variables--oh, and the immeasurable air-gap voltage reference.

It would seem the AGS doesn't try to control VArs or power factor. What's the point of an "automatic" generator control system that doesn't do that?

 

Thanks for the reply guys,

Phil,

Excitation being a determining factor did enter my mind, but here's where it gets wierd. This morning we tested all possible combination of the 4 Generators and found that;

The following combinations share quite nicely,
1 & 2
1 & 4
2 & 4
2, 3 & 4
1, 3 & 4

While these ones do not,
1 & 3
2 & 3
1, 2 & 3
1, 2 & 4

If DG3 excitation was a problem, would it not occur everytime DG3 was used?

I also checked the excitation currents and they are all approx 1 amp regardless of whether there is a kvar fault or not.

CSA,

I should have been a bit clearer, the AGS is a piece of add on software to the Power Management System (Kongsberg Kchief). Kchief controls via Increase/Decrease Pulses an Engine Governer (Woodward Easygen) which controls the Engine itself via an analog bias.

 

Please visit our webpage www.brush.eu, I will be glad to visit you and discuss Power Management System to solve all your problems and upgrade your system.

 

Tharris...

Q1) Was the 3-4 combination inadvertently omitted?

Q2) Did kVAr imbalance start after AGS installation?

Q3) If answer to Q2 is 'Yes', then were CTs added and/or reconnected?

Q4) Have observed/metered Ampere readings been corroborated using a Clamp-On CT?

Phil ([email protected])

 

Thanks again for your reply, to answer your questions

> Q1) Was the 3-4 combination inadvertently omitted?
Q1..Yes it was, 3,4 Combination shares fine

> Q2) Did kVAr imbalance start after AGS installation?
Q2..Yes it did


> Q3) If answer to Q2 is 'Yes', then were CTs added and/or reconnected?
Q3..Yes CT's were added

> Q4) Have observed/metered Ampere readings been corroborated using a Clamp-On CT?
Q4..No we haven't, our clampmeter recently went walkabout, I'm not really comfortable running an excitation current through a multimeter, but if you think its the right track in finding this fault it's our only option until we can replace the clampmeter.

 

Tharris...

kVAr imbalance can be solved using parameter values. It's simple approach, but take real work. A world of information can be gained from numberology. I found that Comparative Analysis usually leads to the solution. Hence, my usual first step is to request 'data'.

Returning to your problem... I asked abut CTs because I found that many 'sudden' imbalance problems are caused by improper CT connections. That said, remember the phrase " ... for the want of a nail, a Kingdom was lost."

Regards,
Phil

 

Further to my earlier post... prioritize your investigation. First, concentrate on data-collection. Then, corroborate that CT secondary-circuit leads are correctly phased.

Lastly (and least)... although possible, the probability of an error in rotor-current measurement is nil.

 

Tharris,

First, DP configuration rules can only be revised with the consent of the vessel's classification agency.

What type of VAR control is used with the voltage regulators? Are they using droop, cross current compensation, or is the voltage bias controlled by a load share controller?