Generator Output vs. Power Factor. 13

I work for a small municipal light Dept. in Massachusetts, USA. We have three 3412 Catarpillar generator sets (325kW each) for peak shaving. The 3 phase generators are rated at: 406kVA, 325kW, @ .8 power facter. The generators are brushless, revolving field, with a solid state exciter. When we run these units the plant power factor jumps from .85 PF to .99PF. I run these units at 325 kW each for a total of 975kW going back on to the grid. The mechanic that we are interviewing to maintain these units says we are overloading each unit by 20%. He claims that if the power factor is above the nameplate 0.8, you have to de-rate the output kW. Now this seems backwards to me, if anything I would think you could run one of these units at 406kW due to the .99PF. He insists that I'm wrong and because this is generation it is different. Also note that, the name plate on each unit specifies that 489amps at 480 volts is the maxumum. When I run them at 325 each, the amperage per phase is approx. 405 via the fluke 43 meter.

Any Thoughts...

Thanks all.

Chris

 

[jbartos]

Suggestion: Obtain the "generator capability diagrams (P-Q diagrams)" for your generators from the manufacturer and try to stay inside the loop curve. This will guarantee that the generator is not overloaded.

 

[SooryaShrestha]

As rightly responded by jbartos the exact condition whether active power can be above the name plate rating depends upon the the shape of Capability Curve. I have, however, tried to present here some of the pertinent points. In the Capability Curve, the active power at rated power factor is the real rating of the generating set. It may be maintained the same or even higher active output at the unity power factor depending upon the type of prime mover, because the controlling factor for the active output above the rated power factor depends upon the capacity of the prime mover. Generally, the diesel generating sets are not rated above the required output capacity, which is in terms of the active power. So when the gen. set is run above the active power output the prime mover is overloaded or will not be able to supply the required power. Now if the prime mover has the capacity above the generator rating plus the losses, then it can be loaded above the name plate active power rating as in the case of the hydrogenerators, which will have the turbines as prime movers in the majority of cases rated higher. This will be very clear with the capability curve.
So I would say your mechanic is not wrong with his experience, because the prime movers he dealt did not have overloading capacity.

 

[peterb]

I'm not too sure that you will be able to get the capability curve for such a relativley small generator, but jbartos & powereng are both correct. In reality, the generator capability above rated power factor is limited by stator heating (read limited to rated stator current, or kVA), and also by the prime mover output capability. Operating at rated kW and higher power factor (read lower kVAR and also lower kVA) will not overload the generator. Be guided by the generator metering for kW, PF and A.
From your readings, with the generators at 325 kW, 480 V, 405 A are operatiing at a power factor of 0.96 (325 kW/336.7 kVA).
You didn't indicate where the plant power factor is being measured. If the meter is on the HV side of the supply transformer, the PF will include transformer reactive losses. These will be decreased when the plant load is supplied in part by the in-plant generation, hence the increase in PF.

 

[rhatcher]

With respect to the Diesel Generator Set load capabilities, several excellent post have addressed that. I am very interested in the statement:
"When we run these units the plant power factor jumps from .85PF to .99PF."

Before I go off on a what may be a tangent, please clarify a few points.
- Is a "small municipal light department" a power generation station? If so, what is the capacity of the primary units (kW and kVA)? Also, are you the sole supplier to your grid or are you interconnected with other stations?
- If you are a generation station, when you say that the "plant" power factor increases to 0.99, are you referring to the combined output of the primary units and the peak shaving units or to just the output of the primary units?
- When you bring the peak units on line and the "plant" PF becomes 0.99, what are the kW and current outputs of the individual primary units? How are these values different from before?
- Are all of the peak shaving units running at the same kW and current or did you just measure one?

Please note that my interest is based on real and reactive power load sharing between interconnected generator sets (or interconnected power generation stations). I am wondering that if you are a generating station then where is the reactive power you were supplying (implied by the 0.85PF) before bringing the peak units on line being supplied from now? Have you shed that load onto another station or is the power factor of the combined output of the primary and peak shaving units of your plant in fact still .85? (If it is and the peak units are at .96PF, you may be exceeding the kVA ratings of the primary units to supply the additional kVARs, ie. primary unit PF went down in the balance...)

 

[jbartos]

Suggestion to peterb Feb 14, 2001 posting.
1. Please, notice that the generator capability curve also has a field current limit and associated heat with it. This is a portion of the capability curve in its upper capacitive KVAR region.
2. Also, the capability curve has its stability limitation which is in the inductive KVAR region.
3. The above 1. and 2. are in addition to the mentioned stator current or stator heat limit adjacent to the stator end-iron heating limit.
4. Visit

http://www.usbr.gov/power/data/fist/fist1~4/1~4_21.htm

for more info
5. The capability curves are normally supplied with larger generators, however, the smaller generators may have them too, since they are not so difficult to produce. Also, the manufacturer site is available for more info:

http://www.cat.com/products/engines_n_power_systems/spec_sheet_library/EPG/_epg/60.html

 

[peterb]

To correct two inaccuracies in jbartos's post of Feb 19, the field current limit applies to the overexcited (inductive kVAR) portion of the curve and the stability limit applies to the underexcited (capacitive kVAR) portion of the curve.
For the case in point, I did not imagine that underexcited operation was contemplated, but for the record it is probably not a good idea nor necessary for a plant of this size to operate in the underecited (leading power factor) region to absorb kVAR from the utility. This would imply a high utility system voltage that needs to be reduced by the peak shaving units - not a credible operating scenario.
Given that we were originally talking about operation at a higher than rated power factor, I do not think that field current limitation in the overexcited (lower power factor) region of the curve is a factor in this discussion.

 

[rhatcher]

I have to go with jbartos here. I did not visit the provided web links but I support his statements as true to fact. In my experience I find the terms capacitive and inductive ambiguous when referring to generator power factor. However, I understand jbartos in the context of his response to mean that an overexcited generator is capacitive in that it has a leading power factor. His assertion that this is apparent from the generator V-curves is correct. I define, as do I think he does, an overexcited generator as one whose terminal voltage is fixed by interconnection or synchronization with other generators (basic premise of V-curves) and whose power contribution to the load has a power factor of greater than 1.0.

I also disagree with the idea that a generator whose excitation (power factor) differs from that of the bus or utility to which it is connected will absorb or supply (circulating) KVARs as a rule. The load power supplied from interconnected generators will be SUM of that provided by the individual interconnected generators. Any degree of real or reactive load sharing is possible between the generators. Of course, in the extreme it is possible to overexcite the system such that KVARS circulate between generators. It is also possible to overdrive (prime mover throttle) the system such that real power circulates between generators (ie one becomes a motor driving its' prime mover). However, this is not a rule, it is an extreme.

Finally, there is such a thing as a generator whose only purpose is to supply reactive power. That device would be called a "synchronous condensor".

 

[peterb]

Responding to rhatcher -
Let's try to clear up a couple of things here, which are supported by the usbr link that jbartos refers to -
- When a generator is supplying lagging power factor load, it is operating overexcited. The generator appears to the load to have a capacitive reactance for this condition (it is supplying kVARs).
- When the generator is supplying leading power factor load, it is operating underexcited and appears to the load as an inductive reactance (it is absorbing kVARs).
- Operation of the generator in parallel with a utility system has analogies between the exciter (controlling reactive power flow) and the prime mover (controlling real power flow). If the prime mover power is increased, the real power delivered to the utility bus increases. Similarly, if the excitation current is increased the reactive power delivered to the load is increased - this increase in excitation current is in the direction of overexcitation. Conversely, decreasing the prime mover input power decreases the real power delivered to the utility bus and decreasing the excitation current (in the direction of underexcitation) decreases the reactive power delivered.
The generator will not be supplying circulating kVARs - each generator will supply the amount of kVARs determined by its excitation current and system conditions. If the excitaton current on a machine is decreased, there will be a point where the generator is absorbing kVARs from the utility system - this is an underexcited condition and will give a leading power factor measurement at the machine terminals. Reference to any generator capability curve will confirm that the machine has some capability to operate underexcited, but this is limited by stability considerations. Note that on an interconnected system it may be necessary to operate some machines in this part of the curve in order to reduce system voltages under light load conditions.
The synchronous condenser is essentially a synchronous motor that supplies lagging kVAR to the system. It operates on the positive kVAR axis (overexcited), with minimal kW input from the system to run the machine. Note that the term "condenser" is synonymous with "capacitor", and the synch condenser appears as a capacitor to the system - it supplies kVARs, just as a static capacitor does.

 

[don01]

hi guys
Just a few small thoughts here. The pf rises to .99 when the sets are brought on line?. A synchronous or salient pole motor (read alternator not supplying ) operates as a pf correction device when run with excitation > shaft load.

If the pf of the load is climbing to .99 when sets come on line are the cat motors driving above load requirements and working in a leading pf range?
Something doesn't feel right

What have I missed?

 

[pjs]

I am agreeing with you, that as the power factor is increased, up to 0.99 lagging, you can draw more useful power , i.e. active power in electrical terms. It may be noticed that the current rating of the DG sets are generally rated at the power factor of 0.8, but it does not mean that you can not operate the system at higher power factor, till the maximum limits of lagging side. As long as your power factor is nearing 0.99 lag, you can draw the power at maximum efficiency & maximum output torque. If you are interested in the curve of torque & efficiency against power factor, please ask for the same.

 

[rhatcher]

I have been away and need to correct my post of 2/20. With parallel generator sets, increasing the excitation of an individual set will cause it to assume more of the reactive load and therefore have a more lagging power factor. That is a fact which I erroneously contradicted. I must have left my brain at work that day(haha!). The only thing worse than posting an error is to do so while "correcting" another post. I applaud perterb for catching it right away and setting the record straight (I had to give a star for that!). I apologize to the group for any confusion which resulted from my post.

To get back on track, the situation described is not right. In an effort to be brief (which I am not good at) I will refer to my post of 2/16 and Don01's post of 2/23. Does anyone else see it?

 

[davidm]

The mechanic that we are interviewing to maintain these units says we are overloading each unit by 20%. He claims that if the power factor is above the nameplate 0.8, you have to de-rate the output kW.

I would take a look at the vendor documents, and especially a capability curve if available as suggested by others, but I would be VERY suprised if your mechanic wasn't in error about this. The generator curves I have seen decrease MVAR limits as you go away from a unity power factor (either lagging or leading). In other words, a power factor between 0.8 lagging and unity at the nameplate KW rating should NOT be overloading your machines. A power factor LESS than 0.8 at rated MW WOULD be overloading your machines. Verify, but I believe that your thinking is very correct on this Chris.

Now this seems backwards to me, if anything I would think you could run one of these units at 406kW due to the .99PF.

Chris, this appears sound, i.e., from the generator point of view you haven't exceeded its kva and amp rating and a pf approaching unity should let you run higher KW - BUT! This is where you really DO need your vendors gen capability curves, and maybe ratings of your diesels. I would give Cat a call on this one. If I had no supporting information, I would never exceed my generator nameplate KW even above 0.8 lagging pf. And since you haven't been exceeding 325kw per your description, I think you are good to go, but your mechanic might not be....

Some excellent info already by many in this forum, including an great description of overexcitation vs. underexcitation by peterb. I recommend the very good online references for voltage regulation/excitation at the Basler web site. Not sure of which document, but one has a good tutorial on generator basics, including capability curves, protection, etc:

http://www.basler.com/html/dwntech.htm

 

[SooryaShrestha]

In this thread, I think the question regarding the power factor jumping from 0.85 to 0.99 is still unanswered. It seems Experts in this forum raised doubts why the power factor jumped from 0.85 to 0.99. I am trying to answer the question. I expect comments from our experts.
First of all, let me put this way. The active power flows from one node to the other if there is difference in phase, as active power flow is directly related to the Sine of the power angle, which is nothing but the phase angle difference between the voltages of the two nodes. Of course, the active power flow depends upon the voltages of the two nodes also and the impedance between the two nodes.
But the case of the reactive power flow is different. The reactive power flow is equal to the difference of the voltages between the two nodes, the grid voltage and the impedance between the grid and system from which the reactive power is going to flow.
In the question raised by Courville, it seems that there was one unit of 325 kW running at 0.85 pf connected to the grid. This means it is supplying 325 kW and 201 kVAR. This will give fixed terminal voltage independent of the generator and dictated by the grid voltage. The excitation is also fixed, which makes the internal voltage also fixed. Now when another generating unit is brought to synchronism at the same internal voltage with the same terminal voltage (a small change can take place, which will give actual value a little different from calculated value) and the grid voltage.Constant voltage difference will dictate the same reactive power supply. Of course the synchronous impedance of the machines affects little bit to the power supply, the machines being of very small rating.
With two units supplying 650 kW and 201.4 kVAR, the overall power factor will be 0.955 and with the three units supplying 975 kW and 201.4 kVAR, the overall power factor will jump to 0.979.
That rhatcher posted as Fact is thus explained above and the raised question of power factor jumping is explained.
The explanation made by davidm will be true only from generator point of view, but not from prime mover point of view. This has been explained in my earlier response.

 

[jbartos]

Suggestions: There appears to be a lack of References in this thread. Therefore, I am adding some to support my postings and their compliance with some References.
References:
1. Slemon G. R. "Magnetoelectric Devices Transducers, Transformers and Machines," John Wiley and Sons, Inc., 1966
2. Stevenson Jr. W. D., "Elements of Power System Analysis," Third Edition, McGraw-Hill Book Co., 1975
3. Fitzgerald A.E., Kingsley Jr. C., Umans S. D., "Electric Machinery," Fifth Edition, McGraw-Hill, Inc., 1990
4. IEEE Std 100-1984 "IEEE Standard Dictionary of Electrical and Electronics Terms"
5. Say M. G., "Alternating Current Machines," A Halsted Press Book, John Wiley & Sons, New York, 1976

My posting content is based on Reference 1, Figure 5.44 "Circle Diagram Showing Power Ps and Reactive Power Qs into Synchronous Machine at Various Constant Values of Field Current. Terminal Voltage Es Constant," that is based on complex power entering machine
Us=Ps+jQs=Is x Es*. The left half plane (Ps<0) covers generator and right half plane (Ps>0) covers motor. The upper plane is capacitive Qs and lower plane is inductive Qs.
This is consistent with Reference 2 Figure 2.6 &quot;Capacitor considered (a) as a passive circuit element drawing leading current and (b) as a generator supplying lagging current.&quot; Please, notice that the &quot;lagging current does not necessarily imply the inductor when it comes to generator. If the current flows into the capacitor, then it leads voltage and it is related to the passive capacitor element. If the current flows from the capacitive generator, then the current lags voltage. However, the Reference 2 Paragraph 2.5 &quot;Complex Power&quot; defines it as S=P+Q= V x I* which is different from Reference 1. It states that &quot;To obtain the proper sign for Q, it is necessary to calculate S as VxI*, rather than V* x I, which would reverse the sign for Q.&quot; The Reference 2 defines the complex power in agreement with ANSI standard. Reference 2 Table 2.1 shows:
Generator action assumed:
If P is +, emf supplies power
If P is -, emf absorbs power
If Q is +, emf supplied reactive power (I lags E)
If Q is -, emf absorbs reactive power (I leads E)
Motor action is assumed:
If P is +, emf absorbs power
If P is -, emf supplies power
If Q is +, emf absorbs reactive power (I lags E)
If Q is -, emf supplies reactive power (I leads E)
Reference 3 Fig 5-15 &quot;Capability curves of an 0.85 power factor, 0.80 short-circuit ratio hydrogen-cooled turbine generator.&quot; Shows the Per-Unit Reactive Power (lagging) versus Per-unit Power with field heating limited curves and armature heating limited curves in the first quadrant with per-unit scale positive. This first quadrant generator capability curves correspond to the second quadrant of Reference 1 Fig. 5.44 that incidentally is marked as capacitive Q. This is in agreement with Reference 2 Figure 2.6 (b). Reference 2 shows a construction used for the derivation of a synchronous generator capability curve in Figure 5-16 (in the first quadrant) based on P-jQ=V x I, which is consistent with Reference 2 Table 2.1 &quot;If Q is +, emf supplied reactive power (I lags E).&quot;
Reference 4 defines the phasor power as S=P+jQ=E x I*. IEEE Std 100-2000 (seventh edition, current edition) does not include that definition.
Notice that the capability curve will have the field heating limit in upper plane for S=I x E*=P+jQ power definition (P<0, Q>0 and capacitive), Reference 1, and in the lower plane for S=E x I*=P+jQ power definition (P<0, Q>0 and capacitive), Reference 2.
Reference 5 states in Paragraph &quot;Operating Characterisctics&quot; on page 376 that &quot;For underexcitation the input current may lag for all loads, and the maximum power will obviously be reduced.
Some clarification is needed to the Peterb posting on February 19, 2000 that I am attaching below marked:
peterb (Electrical)
Feb 19, 2001

To correct two inaccuracies in jbartos's post of Feb 19, the field current limit applies to the
overexcited (inductive kVAR?????) portion of the curve and the stability limit applies to the
underexcited (capacitive kVAR????) portion of the curve.

 

[peterb]

jbartos, please read my Feb 21 post for clarification of the overexcited vs. underexcited condition.

 

[SMLD]

Hello everyone,

I'm Chris Courville, the one who initially started this thread. The responses to this thread have been very helpful, and I personally thank each of you. The mechanic I referenced in the question, as it turns out, was indeed talking about the prime mover being overloaded, not the generator. The prime mover is rated at 483hp and 325kW
(483/1.34(.9))=324.40. I was thinking about the generator only, disregarding the prime mover. The mechanic and I have settled this matter and are in good relations. He was wrong about the gen set being overloaded, but right in saying I could not get more kW out of the set due to the prime mover rating. He just didn't know how to explain it so I would understand.

As far as the power factor goes, I ran some tests with a Fluke 43 and got a .99 pf at three test points: the station xfmr(1.5mVA), the plant bus, and directly at the generators. I got this pf with one, two, and three generators paralleled. Also note that these generators are backfeeding through the station xfmr(1.5mVA 480vac to 13.8kvac) onto our distribution circuit. This is what we would do in a peak shaving situation. When I shut the gens down, the pf goes to .76, .56, .82 for phase A,B, and C respectfully.( These measurements were taken at the station xfmr). After about 5 minutes, the generator cooling tower motors shut down. With this inductive load off line, the pf goes to .94, .63, and .98 for A,B, and C phase respectfully. It seems that the genertors are correcting the pf when they are online.

 

[don01]

SMLD
hi chris,
just looking at your pf(s) when generator is off line is all your lighting on the b phase???
Something is pulling the pf on that phase down hard. I suggets a bit of an investigation of the loads at the distribution boards and tried to balance that pf as well as current load (if not amps are balanced). This imbalance is costing you money when you run the generator and from the supply co. when you don't.
(it may also limit some of your output)
Also I would investigate the sizing of the cooling tower fan motors (and pumps?) they are causing far too much change in the pf. They must be running at a VERY low value.
Any way this thread was just getting fired up!! so why stop the fun.


Best to all Don

 

[jbartos]

Suggestions/Comment:
1. It is nice to read that the solution was found and was acceptable. As far as the power factor increase of the paralleled generators is concerned, it is the result of overall generator control or regulation concept (or strategy) that tends to deliver as high watt power as practical to the load since nature of the load is sensed by the generator regulators that may interact with the speed governor. One may have a considerable influence over the generator set operations, in case those load sharers are installed. The load sharer may have adjustable &quot;load gain,&quot; &quot;adaptation,&quot; &quot;droop,&quot; and &quot;drive,&quot; e.g. Load Sharer T4300 by Selco USA, Inc. Therefore, any significant change, i.e. an increase, in the power factor, while the generators are &quot;on,&quot; is not uncommon. However, the previous posting concern, and verification of the control concept for a possible error appear to be justified.
2. PeterB posting Feb 21, 2001 and Mar 5, 2001 somehow omitted the alignment and location of the field heating limit at the generator capability curve. The field heating limit is located in the opposite half plane (e.g. upper half-plane) to the stability limit part of the capability curve, unless the whole curve is upside down, which some manufacturers show.

 

[jack6238]

For Chris
I have found 12 pages of data in my files for a Caterpillar 3512 engine driving a 1135 KW, 1800 RPM, .8PF. 480 volt, wye connected 689 frame generator. This reactive capability curve for this machine shows the machine output in KW is limited to 80 % of the rated KVA. In other words the portion of the curve which connects the overexcited region to the the underexcited region is a straight line from .8 lagging to .996 leading and it passes through 80 % of rated KVA. This is a clear indication that the engine has less capability than the generator. I also have a note in my files from Cat. stating that the engine is the limiting factor for this unit and has a rating of 1135 KW.
I suspect that your unit has a similar shaped reactive capability curve. Send me your fax number and I will forward this curve. The curve is give in % rated KVA and I believe it will be suitable for your unit.