Leroy Somer R450M AVR Problems

Why is it a <...a MUCH to big variance...>? What test criterion determines this? What were the equipment manufacturer's recommendations upon having discussed this matter by telephone?

Hi,

Firstly, I inherited these machines, that were originally installed by Daewoo (manufactured by Coelmo Italy), and have, since 2013, kept them going, largely through spending days and nights fault finding, reading manuals and applying my own electrical background.

Suffice to say that there are no benchmarks that I can work from and the test criteria is zero, only that the machines keep on running 247, 365.

Secondly, there is no way I can just call any manufacturer and ask for help, thus I came to this forum.

The problem is clear, yet the solution is not.

7 identical machines are running with practically identical AVR settings, and are all drawing similar load since they are synchronized on a Woodward system to a cummon 11KV bus in a substation, in a similar bandwidth . . . . except for the one machine, which is drawing much more current, before releasing and the current would decrease over a very short time to way below the others, as explained below. So its up and down, up and down, and not controlling like the others.

If the real output power (KW) is changing, it's not a voltage regulator problem but likely a problem with the engine driving the alternator. It could be a throttle problem or a problem with the fuel supply.

Interesting, I would have suggested it is a regulator issue...

This is what I was thinking too. There are fine gas settings and our mechanical team will check this tomorrow morning.

Hi and thank you for the excellent info. I have read your post a few times and have picked up good information about the KVAR and KW and will dig more into that perhaps in another post, but I was wondering whether you could assist with a few questions I have specifically relating to the setting up of the R450M AVR:

1. Where do I start once the machine is ON LOAD? (I monitor with my Fluke multi meter and I can confirm that voltage is stable at 400VAC / 50HZ)
2. Should I put the generator on load with a fully ant-clockwise setting of the Droop potentiometer and then start tuning, or do I leave it in the middle position.
3. Take note I only have settings for a) Voltage b) Voltage Stability c) Excitation ceiling and d) Droop
4. So my main question now is: In what sequence should I set the AVR up? I usually start with Voltage first (since this pot is always fully anti-clockwise for minimum voltage) and work my way from there.
5. In this particular case my method is not producing the required results.

Your assistance would be much appreciated.

John

let's consider 2 gensets running in parallel, for example and explanation.

if the initial voltage is not the same on the 2 units,

as far as each unit is running isolated or disconnected from the other unit, then we can measure on each unit its own output voltage.

V1 and V2 for example.

if V1<>V2 then when we connect the 2 units in parallel, the resulting voltage on the bus bar, U, will be different and the U value with be in between min and max value (V1, V2)

in fact, as soon as the 2 units are interconnected, the output voltage become the same on the 2 units; meaning that one unit is decreasing its initial voltage and the next unit is increasing its initial voltage.

the problem resulting from this situation is that one unit is going to feed the other with reactive power, and the next one is consuming reactive power; even if there is no load connected to them.

as a conclusion, the most important is first to adjust the lo load voltage or initial voltage at the same value on all units, and this adjustment must be done before the unit is connected to the others. And then this voltage potentiometer should not be moved anymore !

Hi,

And thank you once again for the excellent info.

Let me just say that for the past three days I have been heavily involved in another commissioning effort and have not had a chance to get back to our main Gas Yard Power Generation Plant and continue to try and solve this problem. I will continue there later this afternoon or tomorrow and will get back here at the first opportunity.

do not hesitate to make direct contact...

call +33(0)238609908 for assistance

someone will asl you further questions and guide you more in deep or offer you technical assistance.

Thank you, however I cannot call out from where I am without incurring exorbitant pay-as-you-go cost. Chevron is in the process of installing fixed lines and this may be possible in months to come.

We are situated deep in the Niger Delta region and there are no roads or fixed lines to where we are. Everything is brought in via ship up the river or via fixed wing or helicopter. The nearest "town" is 350km away.

So you get the idea.

OK, finally had time to give attention to the machines.

1. I have checked every machine personally, adjusted no-load voltage here and there (very small adjustments) and can confirm that they are all running at a stable 400VAC/50HZ at NO LOAD.
2. All machines also have similar excitation ceiling, stability (50%), and Droop settings. All CT's are reading correct resistance so there is current flow in all the AVR's for droop setting
3. With all of the machines running ON LOAD I can see NO.2 machine (of 8) still has a problem, even with a brand new R450M AVR.
4. I can monitor each machine's Woodward Controller in the substation and it is clear that power factor goes up and down on machine NO.2, fluctuating from i0.82 to i0.96 (over a 30 second period), with a resulting fluctuating 220KW to 380KW. So it's constantly going up and down, up and down.
5. The other 7 machines synchronized on the grid are running very stable with power factor of i0.95 and KW from 290KW to 310KW, so no problem with the others
6. It is only this one machine No.2 that has a problem that I am unable to solve.
   

PS: I have another question - What does C mean in front of the power factor on the Woodward controller. This is something unrelated but I see one of the other machines, instead o0f showing a power factor rating of i0.98 it is showing C0.98. What do I make of this and how do I rectify this, if it is a problem indeed?

Thanks in advance

John Meyer

when considering a load on a power plant, or more generally considering any electric power including active and reactive values, then regarding the reactive power it can be inductive or capacitive in respect with the main property of the load we are considering.

so the Power Factor measurement is a mean to illustrate this load property. this means that we can have both value for inductive , when a 'i' is presented before the value itself or for capacitive when a 'c' is presented before the value also

now applying this to a genset, which is providing power to the load.

when the load is inductive, the genset is providing reactive power (kVAR)

when the load is capacitive, the genset is consuming reactive power (still kVAR)

when this happens on a site, it is an illustration of unproper setting of the 'droop' system.

unfortunately, this is not really related to the load variation.

to be more precise, any load variation will lead to sudden change of hte running condition, and consequently will lead to reactive power fluctuation between gensets

why is that visible on unit n°2 ?

let's say that if unit n°2 is having different setting than the others, any running condition change will the balanced unit n°2 and the set of 7 other gensets.

so the impact of load change will be very much visible on unit n°2, but less on the other units as same value's impacr will be divided by 7.

Hi John,

Taking much consideratin to your problem, I am trying to explain as much as possible droop system, parallel operation and so on...

but I am still confused regarding the real site situation...

is kW value changing ?

in your point 4, you indicate a load fluctuation in kW...

this is important, in order to take apropriate action...

if the fluctuation on unit n°2 is regarding kW, then you should act on the speed governor (linked to speed droop!), the power factor value change could only be a consequence of the active power value change...

(same reactive power but with different activ power leads to power factor change!)

only if the active power is stable but the reactive power unstable, then you can consider it is an AVR concern.

if there is active power (kW) fluctuation, there is no chance to resolve this with the voltage regulator.

So it is very much important to identify the root cause to resolve your problem.

So far, it does not look to me that it is AVR issue

HI John,

Firstly I am no expert and am wanting to understand your problem more clearly.You have had far more knowledgeable input than mine which 'hopefully' has solved your problems, but your description sounds to me like 7 of your alternators are handling the majority of the (stable) load while most of the load variation is being supplied by your eighth alternator rather than it being shared across all of them. Is my understanding correct?

Bill

Your understanding is correct.

trying to illustrate this situation with another image...

just consider that the grid is one line... not a line, but physically a rope !

and to maintain this rope, you a team of 8 guy, each representing one of the 8 gensets...

in your situation, it looked like you have more or less 2 teams,

one team with 7 guys and one team with 1 guy only... both pulling on the rope...

even so there is no load... no action from grid side...

a small action or the "7guys" team will have bigger impact on the next team, to maintain the rope stable or static...

Hi there,

And thank you for your enlightening info.

After thorough investigation and testing I think I may have found the cause of the problem.

Background: We have 8 generators on an 11KV grid producing around 6.5MW of electricity at 400VAC/50HZ. Two alternators have been replaced over the past few years. The original alternators were Leroy Somer LSA50.1 type with PMG primary excitation and R449 AVR's, however the 2 replacement alternators were a different type in design, still LSA50.1type, but with AREP excitation and R450M AVR's. And it is these new alternators that have been producing fluctuating KILOWATT outputs, with the one I have been referring to the more problematic of the two.

Now here's the thing. I have found, in the case of both of these troublesome alternators, that the CT that is mounted around the neutral conductors within the alternator itself and connected to the AVR for droop setting, does not have all of the alternator neutral conductors passing through it, and only 1 of four conductors able to pass through the CT. The problem was, when technicians retrofitted the new alternators a few years ago they realized that the existing CT's (that were removed from the faulty PMG alternators for re-use) had an inner diameter much too small for all of the neutral conductors of the AREP alternator to pass through it. They obviously did not have the correct CT's on hand at the time and just reinstalled the old ones, resulting in the current configuration.

As an experiment I have disconnected these CTs from the AVRs in the two AREP alternators and shorted them out to prevent burnout, kept the droop settings on the AVR on default, and then started the machines and put them on load.

The result is that both alternators are running a lot more in sync with the others on the synchronized grid and Kilowatt (and power factor) is a lot more stable.

I guess my question now is: If I procure the correct CT's (400:1 Ratio) with a larger internal diameter to enable all of the neutral conductors within the alternator to pass through it, and I were to connect these back onto the AVR, should this solve my problem. Do you think that the incorrect CTs and the way in which they were mounted could have made the alternators behave the way they did?

John.