Power Factor Corection - Diesel Genset 1

[SierraJ]

Hi,

I need all of you with in depth knowledge on diesel gensets to help me understand the following statement quoted by two major diesel engine genset manufacturers:

"Diesel engine gensets are rated at 80% PF (example 2000kW/2500kVA @ 80% PF). If the genset load is above 80%, lets say 95%, the genset would consume more fuel. Therefore the recommendation is to disconnect any power factor improvement capacitors when running on genset power".

This is not totally clear for me since:

1.) I am not trying to increase the capacity of the genset, which would clearly result in additional fuel consumption. The load is what it is, with or without capacitors.
2.) Using basic power system principles, any reactive power being supplied from the capacitors downstream from the genset would not need to be supplied by the genset alternator (real power being the same on both cases).

In addition, a typical operating curve for an alternator shows that the alternator could supply loads from almost 0% lagging to 99% leading.

When confronted with my question, one of the manufacturer's responded by saying that the alternator is designed for peak efficiency at 80%, therefore at any other power factor the alternator would not be operating at peak efficiency.

This was not totally true either since the alternator efficiency curve showed efficiency always improved with improved power factor from 0% to almost 100%.

Please shed some light on my analysis of this issue.

 

[alehman]

That's confusing to me as well. I've worked with a lot of DG sets and never come across that statement before. Your reasoning makes sense to me. Some people confuse power factor and load factor. I wonder if they really mean peak efficiency occurs at 80% load factor.

Some DG mfg's say to disconnect capcitors as a matter of routine, I think because there have been some problems with harmonics and when the capacitors are on with the unit lightly loaded. That can result in leading pf, which can confuse voltage regulators. Just a guess.

 

[SierraJ]

Alehman,

The information that was given to me (in order to explain the additional consumption) was that the way the alternator is designed it has the peak efficiency at 80% PF. Again, this looked reasonable to me and I expected the efficiency curve to be something like an inverted parabola with the top of the parabola at 80% PF (lowering or increading the PF always resulted in "reduced" efficiency). The efficiency curve I recieved was nothing like this.

In additiona you are correct, I know of the problems caused by leading power factor and voltage regulation problems. That's why the typical operating curve and the genset data state that the genset can only handle so much leading power factor without experiencing problems with high voltages and voltage regulation.

Maybe this is not true and it is a way of "selling" the point of disconnecting the capacitors.

 

[dpc]

A generator running at .8 pf will be producing significantly more current than at 0.95 pf (for the same kW output). This current will produce increased copper losses in the generator and everywhere else. (I'm assuming an overexcited generator that is producing vars.)

Generator armature copper losses would be higher and field copper losses would be higher. Friction and windage losses would not be affected. I'm not sure about core losses, but as long as the voltage is the same, I don't think core losses would increase.

So I'm having a hard time understanding why a generator would be LESS efficient operating at a higher power factor.

It is true that the efficiency of the generator would generally be best at full load or near full load. But losses, in kW, are also highest at full load.

What you are really concerned about are kW losses more than the efficiency as a percentage.

 

[SierraJ]

dpc,

Up to this moment we are all in the same page. That is why I didn't bought into it the first time they indicated this to me.

But just for the sake of discussion, let me tell you how this came to be.

I am doing a design for a 12MW (6 x 2000kW-13.2kV) diesel driven back-up system for an industrial plant. The plant has many low voltage (480V) power factor correction banks and the plant load right now is +/- 95% PF. At a coordination meeting, one engineer involved with the controls portion of the project, raised a flag that the capacitors needed to be removed from the system when running on back-up power. I clarified that this was the first time any body had mentioned this "requirement" but that I would check with the genset mfg.

The rest is history... I was told that the units would consume arround 2% more diesel fuel when running at 95% PF.

To this date I am not convinced, butlets see what this discussion will lead to.


Thanks everyone.

 

[subtech]

The issue of running your genset with capacitors in ckt. is a very important one.
Synchronus generators are not stable when operated with leading power factor loads. They can lose synch and if not properly protected can even be damaged or destroyed. It happens. Hence the warning from your manufacturer about getting the caps out of circuit before connecting your genset to the system. Also, the reason for increased fuel consumption at say a 95% lagging PF versus 80% could very well be because the extra fuel is being used to generate Watts that are overheating the rotor windings, stator windings, or stator iron. Blackburn explains it much better that I do.

Find a copy of J. Lewis Blackburns book "Protective Relaying-Principles and Applications".
Although this is mainly an excellent source of info on protection in general, a really great condensed lesson on "generator 101" can be found on page 262 when it comes to operating a generator with leading and lagging loads and just what happens when you step outside the norms of operation.
In the mean time, I would highly recommend following your manufacturers recommendations and stay in the 80% lagging power factor area with this machine. After some time and experience with this set, you may find that you can add some caps back into the equation and run quite well at 85% or even 90% lagging power factor, but for now, I would stick to the specs.

 

[alehman]

Reaching for my copy of Blackburn, I see the discussion of generator efficiency vs. power factor was apparently not included in my 2nd edition.

I'm not following this reasoning however. Why would there be more heating at higher power factor, when stator current is lower? You do have more field current, but the field loss would negligable in comparison to stator losses.

 

[davidbeach]

There is a variable missing in this discussion; what is constant? If we are talking about a constant output of rated kW, increasing power factor toward unity will decrease the total current, and thereby decrease the I^2R losses in the machine. Engine fuel requirement is directly proportional to kW, has nothing to do with load power factor (small errors introduced by the PMG generator supplying the field for the main generator).

On the other hand, if we are talking about a constant output of rated kVA, increasing power factor toward unity will maintain the total current approximately constant but increase the output kW. More kW is more fuel and any power output beyond rated kW is not sustainable in the long term.

So, if talking of kVA, yes it makes perfect sense to say that the generator is designed for maximum efficiency at 80% power factor, since that would be where it is at rated kW.

If talking of kW, there is no logic behind the statement of maximum efficiency at 80% power factor as improved power factor reduces current and losses.

 

[ScottyUK]

The only reasons for improvement in efficiency that I can see are:

Lower I^2R losses in the stator due to lower stator current at unity PF compared to lagging (VAR-exporting) PF.

Lower excitation current, and hence lower rotor I^2R losses and lower exciter losses, at unity PF compared to lagging (VAR-exporting) PF.

Both the above would tend to suggest that unity PF is more efficient than a lagging PF because losses are lower.

There is certainly the generator stability problem to consider, but at unity PF there should not be a problem with any modern AVR. Operation at high load with leading (VAR-importing) PF is not recommended because the unit more susceptible to losing synchronism or pole-slipping because the field is too weak to hold the rotor in synchronism. This situation could occur if there was a loss of VAR-consuming loads - motors, for example - and the PF correction caps were still in service creating a net leading load and forcing the AVR to reduce excitation in order to prevent terminal voltage from rising.




----------------------------------

If we learn from our mistakes,
I'm getting a great education!

 

[subtech]

alehman
My reference to Blackburn was not concerning efficiency vs. PF of any generator. It was towards stability and maintaining synchronism of such machines. The SSSL (Stability Limits)of the machine are dictated somewhat by its construction. Perhaps the construction of the unit in question is such that it will not tolerate being operated at all in the leading PF area. Because of the possibility of significant(relative to the machine output capacity)true power load AND lagging var requirement being suddenly lost due to any number of situations, having caps in the system that are simply tied to the bus for correction is not a good idea. Sudden shedding of a relatively large and highly inductive load could very well put the genset into a significant leading PF output situation and make the risk of protective trip of the unit a real possibility. Perhaps the manufacturer of this machine feels that a PF of 80% lagging allows for a buffer or safe zone that will keep the unit in its area of stability and allow for those unexpected (or maybe expected) load and var changes. I'll not argue anyones statement that the better the PF(close to 100%) the more efficient the machine should run. That is just a basic math thing. Machine dynamics and steady state stability limits are dictated by its construction and design. If the manufacturer says to run the machine as near 80% lagging PF as possible, there are probably many reasons why they do so. Not all the reasons may be electrical or mechanical. Monetary and future business prospects may also be part of the reason for the requirement. When unexpected trips occur, business relations get strained quickly. The statement that running the machine at 95% PF will cause extra fuel consumption may be the the manufacturers way of pressuring the owner to stay in the safe zone and avoid problems they have had in the past. Where I work, we run 12 units ranging from 22MVA to 65 MVA and the engineers(some with over 30 yrs of generating experience) insist and ensure that the units run well into the lagging var output range. They call it the "Safe Zone". It's a daily thing to see an 80 to 90 percent lagging PF on most all of the units. The reason for mentioning Blackburn was not an attempt to answer an efficiency question at all, but rather to point SierraJ towards the stabilty side of things. You're right, there is little or no discussion of PF vs. efficency of generators in Chapter 8 of Blackburns book. However, there are volumes of information about generators, protection of same, and the reasons why that protection is necessary. Protection from instability being only one. All included right there in Chapter 8.

SierraJ, I would heed the makers warnings and requests, and run the machine in the range stated.

 

[dpc]

The issue of stability, under-excited generators and power factor correction capacitors is interesting, but the OP's issue was efficiency and the claims by the engine-generator manufacturers that fuel consumption will be higher as the power factor approaches unity as opposed to running at rated pf of 0.8. I don't think there was any implication that he would be operating in a leading (under-excited) pf condition.

I'm still not seeing any reason why efficiency would suffer if power factor is improved from 0.8 to 0.95 lagging, for example.

 

[alehman]

subtech:
Perhaps I have misunderstood the intention of your earlier statement. I thought we were talking about efficiency.

I agree that it is necessary to avoid situations where the net power factor of the load may be leading unless the generator system is designed for it. If there is risk of load trip, it may be advisable to disconnect the capacitors beforehand or make sure they are connected where they are automatically disconnected if the load trips. I would think connecting the capacitors locally at the loads or sub-distribution boards would accomplish this in many cases. If the manufacturer thinks the VAR "buffer zone" is the best approach to handling this risk, his opinion should be considered in the system design. I have not been told that by a mfg. Most mfg's. do say to avoid leading load PF.

As a side note, a well known situation where this may occur is some types of UPS systems with large input filter capacitors. When the UPS is lightly loaded (or the load trips), the capacitors may remain on the system. I've seen this happen a few times. Usually the genset will trip on overvoltage or "loss of excitation" (I think the latter is a case of the monitoring system being fooled by low field current into thinking the exciter has failed). Obviously this situation must be avoided. If there is enough inductive load on the system to absorb the VARs from the UPS capacitors, you may be OK. Each situation must be carefully anayzed. Lately some UPS mfg's have recognized the problem and added automatic capacitor switching to their systems.

I'm still at a loss to explain what Sierra is being told about efficiency.

 

[mc5w]

Let's say that your system power factor with capacitors and when running on utility power is around 95% lagging. If half of your capacitance consists of small units that are switched with each motor using a contactor on the same control circuit that is attached to a particular machine, then you only need to lock out the other half of the capacitors that are instead hooked up to your switchboards.

In other words, if you change your power factor from 95% lagging to 90% lagging when running on a genset and the capacitors that are in use are switched with motors, you will stay in the safe area.

Your capacitors that are switched in response to power factor rather than by the control circuit for a motor do need to be disconnected from the system.

 

[SierraJ]

Thank you everyone, the discussion has been very productive. Altough it has mainly centered arround the leading power factor issue, which was already settled at the first few threads.

But for those of you concerned with this issue, we had already confirmed with hystorical "in-house" power readings, that the plant has never experienced a "leading" power factor condition. Most capacitors are connected at the MCC's and are siwtched "on/off" togheter with the load, which is as a totally different condition than having to turn all capacitors "on/off" with the back-up generators.

I think my initial skepticism about the "issue" was correct.