Today is...
Wednesday, May 23, 2018
Welcome to, the global online
community of automation professionals.
Featured Video...
Featured Video
A demonstration of EtherCAT control of linear motors using the CTC EtherCAT master.
Our Advertisers
Help keep our servers running...
Patronize our advertisers!
Visit our Post Archive
Unequal load sharing of generators
I have a question about unequal load sharing of generators.

Hi there,

Can anyone please answer this question? When generators are connected in parallel then they begin to share unequal loads, and this results in flow of high reactive currents. Please tell me, why do they share unequal load? Thanks in advance. :)

I'm sure this will prompt a swift attack from the Exclamation Pointer, but here goes anyway.

We don't know the nature of the load at your site, or if you are supplying a small load independent of a larger grid with several (small) generators. We also don't know if this "problem" just started or has been existing for some time. Below are some general principles for multiple synchronous generators (alternators) being operated in parallel with each other and connected to a larger grid with other generators.

Reactive current flow on a synchronous generator (alternator) is a function of field excitation, not load. Properly adjusted exciter regulators ("AVRs") will minimize reactive current flow, and properly trained operators will adjust excitation to maintain desired power factor or VAr setpoint.

Most generators will require an increase in excitation as the unit is loaded to maintain unity power factor; without it, power factor will usually be less than 1.0, leading.

I would suspect there is something wrong with the reactive current compensation circuit(s) of the exciter regulator(s), or the tuning/adjustment of the exciter regulators.

By Phil Corso, PE on 26 July, 2008 - 1:05 pm

CSA, your consternation about the Exclamation Pointer is understandable. However, you probably already know that despite different droop characteristics, paralleled generators will properly share a load at one operating point!

Regards, Phil

As both of you know, this is why we maintain power system operators. They are responsible for ensuring that VAR (sorry Phil), MW load, and voltage are balanced.

AS usual: Delicate balance thing that I preach!


Are you saying that generators having different droop characteristics can only be operated stably at one operating point? Is that a power- (Watt) or reactive (VAr) load point?

How does one calculate that point?

Are you referring to reactive- or real droop characteristics?

Inquiring minds want to know.

CCTT & CSA, I don't know if your comments were meant to test my experience, veracity, age, or some combination therof. Also, I hoped this topic didn't start another brouhaha!

But if you insist, then, a separate more focused topic entited, "The Physics of... Droop", should be initiated!

Regards, Phil


I don't think anyone is trying to initiate a brouhaha nor incite anyone to any enhanced use of punctuation.

That would indeed be interesting, reading about droop from your perspective. But, don't expend the effort for me; it's safe to say we agree to disagree on some things, and this might be yet another one.

My questions were completely legit: Are you saying that generators with different droop characteristics can only be operated stably at one operating point, and if so, is that a power- (Watt) or reactive (VAr) load point? Further, are you referring to the governor droop or the exciter regulator droop, or both? Can this single operating point be calculated?

I would appreciate it greatly if you would provide the details of the proper load-sharing of generators with different droop characteristics being operated in parallel. I'm really curious because, technically speaking (and you tend to be very technical) generators don't have droop. The governors of the prime movers of generators have droop (speed-frequency control), and the generator exciter regulators have droop (reactive current compensation), but generators don't have droop (though we generally refer to the droop of a unit when we mean the droop characteristic of the prime mover's governor). Because I've seen tens of generators and their prime movers and their exciter regulators stably share load at any load point.

Or, are you saying that prime movers and synchronous generators being operated in parallel can only provide power to a load at the synchronous frequency of the generators? Because there are lots of prime movers and synchronous generators in operation in many parts of the world which are supplying loads at frequencies greater than synchronous and much less than synchronous. And, while it's not ideal, they are all keeping the lights on and the motors spinning, albeit at more or less than rated or optimal conditions. Even the European and North American grids experience frequency excursions from time to time, and while it can be scary and even lead to brown-outs or black-outs, the grid usually remains fairly "stable" and eventually returns to normal at or near synchronous frequency.

So, exactly what are you saying, or alluding to?

My response related to your specific question about just two alternators in parallel has been covered in a previous post.

CSA & CTTech,
Past history suggests that to provide a "Physics of... Droop" topic would be futile. There have been some 300 threads covering droop. I contributed less than a hand full. One of them did express droop as the slope of a straight-line representing speed (in rpm or frequency) vs load (in kW). Unfortunately, it was presented in a mathematical format, thus its probably of little value to either of you.



There's nothing wrong with mathematical explanations, after the general theory is properly explained and understood. As has been said many times on this site (by myself and others), too many times an explanation of VArs begins and ends with vector diagrams and formulae--and that doesn't explain what's happening to the system to operators and technicians who need to understand what the effect of VArs is on the systems they are working on.

Engineers and people who have an aptitude for maths and formulae quite often have difficulty understanding and explaining VArs and power factor amd droop because they can only cite the formulae and the vectors associated with the phenomenon, and can't explain what's happening and why it's important to monitor and control VArs. I've encountered regulators in Europe and North America who were absolutely ignorant of droop and its effects on a grid, but were "in charge" of testing and approval of droop settings! (And, in the Middle East, and Asia, as well--it's not a well understood concept, made more complicated by the nature of gas turbines and the environments in which they are operated).

I have seen some very nice charts and graphs that attempt to explain droop, along with some formulae, and none of them really explain what droop does or why it's important. In spite of the reference to the slope of a line, there have persisted to be lots of questions about droop, so formulae and math references haven't seemed to alleviate the need for written explanation.

I've always found that once I understand the principles and the "hands-on" aspects of something, that I can much more easily understand and apply the formulae and understand the importance of various aspects of the phenomenon under consideration.

You're absolutely right; there is a linear relationship. But does that explain what droop is? To someone who's not been trained in reading graphs and charts and using formulae to operate or troubleshoot a process or piece of equipment? I have been at many refineries and power plants around the world, and there are many people who don't understand y=mx+b, or gain and offset. And we all know that many primary school classes all teach this, but it's never explained how this is used in the "real world."

I may be remiss by not always referring to formulae (I did use one repeatedly whey trying to describe synchronous operation and droop), but I try to keep their use to a relevant minimum.

And to provide all pertinent information when alluding to solutions or theory, which is something that is not always present in a lot of the responses on And if you want to take that last comment personally, please feel free to do so. Brief and nebulous responses to technical questions which require and need more are not always beneficial. You do generally offer to help off-line, but that's not always helpful to others who read the posts and don't benefit from your knowledge and experience when you are providing help "one-on-one" off-line. So, what I'm saying is: Many of us would appreciate more meat in the bun, or, in other words, "Where's the beef?"

CSA, "droop" doesn't necessarily have to be


Dear all,

If you can suggest some good reference material on governors and their response (especially in context of steam turbines), I would be much indebted.


By blackstoneboy on 19 September, 2008 - 2:01 am

Hi Nimra.

What is the controller you are using when you do paralleling?? You should have master control for load sharing if your individual generator controller doesn't communicate.

By Ritinkar Sen on 5 October, 2008 - 11:25 am

As a non-electrical engineer, may I dare to attempt an answer to Nimra's question as to why two alternators in parallel share unequal load?

The answer must lie with the governor characteristics. Ideally, the two alternators should have similar droop characteristics i.e. the two droop curves should be parallel. In such case, it is a simple matter to increase fuel supply to the engine of the alternator that is taking less load. This may be done from the main switchboard by operating a switch that sends signal to the governor to increase fuel supply. This way, the droop curves will coincide and the alternators will share the load equally.

The situation becomes quite different if the droop characteristics are different i.e. the droop curves(rpm - load curves) of the two alternators have different slopes. In this case the load sharing will depend on the common rpm. If a horizontal line parallel to the base is drawn at any rpm, the respective loads may be determined by the points where the line cuts the droop curves. The one with a large droop (i.e. large slope) will be nearer and will take less load than the alteranator which has a smaller droop.

Hope no one is laughing !!

By Mohammed Haseeb on 13 December, 2008 - 4:39 am

:) if the generators are identical,they share equal loads

This can happen when the droop characteristics of the turbines (running these generators)are dissimilar. This leads to unbalanced MW loading of the generators. Regarding high reactive currents (i.e. MVAR) the problem lies with the excitation system.

The excitation system maintains the terminal voltage of the generator in spite of the changes in the reactive loading & hence the MVAR loading (Reactive current increases/decreases to maintain the same)changes. This is the reason why Automatic Voltage Controllers are used & the excitation changes automatically (with the reactive load coming on the generators) to maintain the terminal voltage of the generators (and in turn the reactive loading of the generators).

Hope this clarifies. Kindly also check the excitation system. Pl.let me know when & how the problem is rectified.