RPM drop in Engine

Hello ,
We have a diesel CAT gen set.. The gen-set is operated in parallel to another 4 units of same ratings .
The gen-set drops RPM suddenly to Zero in fraction of a second and comes back to its rated RPM 1200.
Hence there is an condition of reverse power for a short duration.
I have attached a record of RPM vs throttle % with respect to time.
When load is demanded quickly this happens ..but when not it smoothly ramps up and there is no drop in RPM.

If this gen-set is swapped to another control cubical there is no issue. Any gen-set connected to this particular cubical shows the same symptoms.
Speed controller is replaced with the one working from another gen-sets, but the issue remains the same.
I have rapped up the fuel set-point to the engine too. But loads are shared properly and progressively loads are accepted, I can see an increase in Amps. with load . But at one point suddenly the Speed drops to Zero and back to 1200 RPM. it happens whatever may be the Fuel set-point.

Any thing to check to resolve this ?



It's virtually impossible for what you say is happening to happen. When a synchronous generator is synchronized with other generators on a grid of any size the speeds of all the generators are fixed by the grid frequency. Frequency and speed are directly related by the formula: F=(P*N)/120, where F=Frequency (in Hz); P=Number of poles of generator rotor; and, N=Speed of the generator rotor (in RPM). Now most diesel generators are directly coupled to the synchronous generators they drive, and that means that generator speed--and diesel speed--are a function of the frequency of the grid. Two-pole generators operating at 50 Hz would spin at 3000 RPM; four-pole generators operating at 50 Hz would spin at 1500 RPM; and six-pole generators operating at 60 Hz would spin at 1200 RPM (if my math is correct).

So, if the unit in question, connected to the questionable cubicle, is operating in parallel with (synchronized to) three other diesel-generator sets then one diesel CANNOT suddenly drop to zero RPM and suddenly increase back to rated speed (1200 RPM in your case). It's simply not possible.

Reverse power occurs when there isn't enough fuel to keep the unit spinning at rated speed--so, in your case that means the fuel would not be sufficient to maintain 1200 RPM. In this case, power from the grid will keep the generator spinning at 1200 RPM--because the generator will actually become a motor, a synchronous motor spinning at 1200 RPM using current (amperes) from the grid to do so.

So, it's just not physically possible for the unit to go to zero speed while the generator breaker is closed.

I suggest there is some problem with instrumentation or speed pick-ups or PTs (Potential Transformers) associcated with the cubicle in question. I know next to nothing about CAT diesel-generator controls, but I'm pretty sure the unit cannot physically go to zero RPM and back to rated speed in one or two or even five seconds. The laws of physics just aren't that flexible.

Current can be VERY fast in "changing" direction, but that's just going to keep the unit spinning at rated speed as long as the generator breaker is closed. I don't know how the units are connected and disconnected from the cubicle(s) but I suspect something is very amiss with this connection/disconnection. Something is loose or intermittent.

Please write back to let us know what you find--and good luck!

Thanks for the reply .. Yes that what is troubling me ..
I hope u have seen the attached CAT ET report and Power recording.
Power Recording shows Reverse Power condition , Sometimes the Breaker trips , sometimes it will not, as the condition for tripping the breaker is >-120Kw for more than 3 sec . The event in question happens for 1 Sec max .
The RPM ,throttle Vs Time chart extracted from CAT ET shows that the RPM reaches Zero, but i can see the Digital RPM meter on engine drop to as low as 24 RPM when sudden load is demanded .
Both Control cubical and Engine have their independent Speed sensor. In actual the speed of the shaft is not dropped , but somehow the ECM of engine is getting the speed signal wrong when sudden load is demanded .
Cubical end the speed feedback remains constant ,1200 RPM . AS ECM senses from its ckt wrong reading , the Injectors are stopped firing and hence the Negative KW .
If the load demand is linear and progressive there is no issue there is no speed drop perceived by ECM and reverse power condition.
But the Question is how come it works without any issue on another cubical :) . Under what circumstances the ECM can read wrong reading from its sensor . the speed sensor cable is without any shield , what can affect it to malfunction with this cubical that to for a fraction of a second ,under sudden load demand ...


A question I would have is how the RPM sensor works. Is RPM determined by pulses per unit of time or time between pulses? In the first case you would not see a large drop in speed that occured very quickly, while in the second case it would be easy to see a large drop in speed from a momentary change. If the encoder generates lots of pulses it would likely be pulses per unit of time, whereas a sensor that got one pulse or few pulses per revolution might use time between pulses. This might be a programmable setting to allow the controller to work with different pulse rate encoders.
The sensor is common for timing wheel and Engine RPM. It's fixed on the timing wheel rather than on the flywheel. Its a proximity sensor. The O/P is a PWM signal to the engine ECM. I think its the second case. But that how the CAT system is.

The puzzle is, when the full engine is swapped to another cubical there is no issues. If sensor was faulty it should have the same symptoms irrespective of the cubical its connected to.

What can corrupt the signal on way to ECM? Shielding /Grounding requirements to take care of ...
Why it doesn't happen when linear load is present? Why only in some cases when sudden demand presents?
Gentlemen ,
Just a Question , ..what would happen if 3 engines are sync and one of the engines don't accelerate at the same pace as the other two . As CSA and grayhlucas said , the ECM may perceive the drop higher due to the sensing method ,but in actuality the drop is relatively low.
The above is the status of vitals as seen by PLC ..
Any clue why the speed of this engine is dropping so low when the other engines are not dropping like this ..
is this is acceleration issue of the engine / the throttle is not rising fast compared to the rise in load demand ?



You don't seem to understand. It's pphysically impossible for a single synchronous unit synchronized to (operating in parallel with) other synchronous units to operate at a speed that is anything other than the speed that is proportional to the frequency of the system. Full stop. Period.

The machine at your site IS NOT IMMUNE to the laws of physics.

If the unit suddenly tried to stop and then just as suddenly jumped back to rated speed there would be physical damage like you cannot imagine.

It didn't happen. What the "data" seems to suggest simply did not--and cannot--happen. Not even at your site, no matter what you believe or want to believe happened.

Unless maybe your site is in the middle of the Bermuda triangle, and even then it didn't happen.

The data is simply wrong, either because the sensors are not working correctly, there is a wiring issue or the data capture mechanism is not working properly.
Agreed with your point 100% ....
Two recording systems the Engine CAT ET:- CAT ET shows all over 50 parameters via its software , sampling rate is 20 per second.
PLC S7-300 Siemens :- , having independent speed feedback systems via its proxi switch . Monitoring what the PLC sees .
The drop between two sensors are also of different magnitude . ( may be due to the way they sense)
I will record two engines data simultaneously on next opportunity,

My basic issue is this , the engine is receiving only Throttle Input from the cubical 0-300mA . .. CAT ET says there is a throttle present and its rising ..but still there is a reverse power condition . This is only on one cubical ...not on any other cubical..

You have NOT provided an information on these "cubicals" and how one engine can be connected to different cubicals.

If, as has been said, this problem is only occurring on one engine when connected to one specific cubical, then you have your direction: Examine the interconnection (ALL of them!) between the cubical and the engine which are, together, experiencing the problem.

Reverse power is just that: Instead of power being "pushed" out of the diesel-generator it is being "pushed into" the generator, and the diesel. Now, that can ONLY happen when the fuel being supplied to the diesel is less than the amount required to keep the diesel and the generator it is driving running at rated speed (the speed which is proportional to the frequency of the system the generator is synchronized to). Amperes are directly proportional to fuel flow-rate--plain and simple. "Follow the fuel flow-rate" and you will find your answer(s)--if your description of the problem is correct. Reverse power occurs when a generator remains synchronized to the grid but the power being developed by the generator's prime mover (a CAT diesel engine in your case) is not sufficient to keep the generator spinning at rated speed (I'm presuming the grid frequency is also at rated). AND, NO generator can run at any speed other than the speed proportional to the grid frequency it is synchronized to--because the grid will keep it spinning at rated speed even if the prime mover can't. It's simply not possible; the laws of physics and magnetism just will not allow it to happen. Not at your site; not at any site.

This isn't rocket science. It's plain AC power generation fundamentals. Either your site is immune to the laws of physics and magnetism or there's a problem with the instrumentation or the control systems recording the instrumentation. If the generator is truly going into reverse power then there is something which is causing the fuel flow-rate to the diesel to momentarily decrease significantly and then increase significantly--which is NOT good for the equipment (the diesel or the coupling between the diesel and the generator or the generator rotor). But, it's just NOT possible for the unit to be going to any speed, especially zero speed, while the generator breaker is closed, and then return to rated speed.

Full stop. Period.
- All generators are connected via Plug and Sockets to cubical bay . We can swap any gen-set to the cubical .
- This issue is from 2013 , All components in that cubical has been swapped with other working cubical.- No change
- We know the issue is from the cubical , but all that is coming to engine is throttle signal , it may be less than what required or More than whats required , or its not ramping up at a rate required.
- but same controller works good in other cubical ,and other cubical controllers gives the same result on this cubical.
-All wrings checked by others in past , i am the 10th person now on this issue.
- Attached ;- load on engine will give u the idea of loads before the repairs. these are from standalone Fluke power meter.
- if i run a direct cable from controllers to Mag and throttle , engine is unstable.
- We worked on the Shielding& power supply issue to the controllers and it was OK . Attach :- Engine-OK. It was running Ok since last 2 years , but now its back to same old status.
-what is pinching is .. for 1-5 years if its working, why back to same status , that time also the fuel settings were not changed. still it was able to pickup the sudden demand .
- the speed and voltage controller from other cubical where they work have been swapped , but still no use.
- The ramping and throttle settings are on speed controller . Have already ramped up the throttle % on Run from 26% to 32% .. fuel available is also higher.
On shielding front , both mag and throttle cables shield are grounded on both ends , if shield ground is removed from anyone end engine becomes unstable and reverse power condition becomes more leading to tripping of breaker . Under the grounded shield condition the magnitude of reverse power is low and short , but exists.


There is something amiss with the wiring or sensors in the suspect cubical.

If the gen-set actually dropped load in 1 or 5 seconds and then just as quickly picked-up load again there would be a clanging/banging noise at the engine and eventually the engine OR the coupling between the engine and the generator would physically break. Or the generator rotor would be severely damaged.

Or all of the above would occur.

AC power generation involves magnetic forces--from two magnetic fields--that work to keep the magnetic field of the generator rotor locked into synchronism with the apparently rotating magnetic field of the generator rotor when the generator is synchronized to other synchronous generators. Any attempt to cause the generator rotor to slow down or speed up is thwarted by the attraction of the two magnetic fields of the rotor and the stator.

You know what happens when you take even two small magnets and try to push the North poles (or the South poles) of the magnets together. You may succeed in making them touch, but if you release them they IMMEDIATELY separate from each other, and in many cases the North pole of one magnet and the South pole of the other magnet attract each VERY quickly.

It's exactly the same inside the generator. Only the strength of the magnetic fields is many times stronger than the ones you tried to force together. If the generator rotor did actually manage to go faster or slower than the apparently rotating field of the generator stator what would happen is that as like poles of the two magnetic fields approached each other they would try--violently--to repel each other. And, that means that the rotor would try to suddenly--and violently--stop. Or, the rotor would actually try to spin backwards. OR, it may actually be repelled forwards. But in either case at some point opposite poles would again align with each other and THAT would make the rotor try to--violently--stop. The phenomenon is called "slipping a pole" and it can be very destructive to both the generator and the prime mover--and the coupling between them. Physically destructive. It's not a pretty sight to see afterwards.

That's why I maintain what you are describing is not physically happening. It's something amiss with instrumentation or wiring or something like that. If it were happening, it would eventually cause great physical damage to the generator, coupling and/or diesel. And, it doesn't seem to have caused that much damage (at least not yet!).

This is about all I can offer. I still do not understand the concept of switching cubicals and gen-sets, but I can be very obtuse at times and a picture is often worth a thousand words. (Hint. Hint.) What I can offer is that what you have suggested is happening is most likely, and almost certainly, not happening. Physics just don't allow for such things to happen.

Have you ever asked yourself this question: If the diesel accelerates and decelerates in speed as fuel is increased and decreased when the generator breaker is NOT closed, why doesn't it change speed when fuel is increased and decreased when the generator breaker IS closed??? (Because it doesn't change speed as fuel is changed when the generator breaker is closed. Because the magnetic forces at work inside the generator prevent that from happening.) And, speed and frequency are related--directly related. Increase speed and frequency increases; decrease speed and frequency decreases--EXCEPT when the generator breaker is closed. EXCEPT when the generator is SYNCHRONIZED to other generators. Increasing fuel increases amperes being produced by the generator when the generator breaker is closed, and decreasing fuel decreases amperes being produced by the generator breaker when the generator breaker is closed. Changing fuel doesn't change speed when the generator breaker is closed, it changes amperes. Which changes load (MW). And that's because of the magnetic forces at work inside the generator keeping the generator rotor LOCKED INTO SYNCHRONISM with the apparently rotating field of the generator stator when the generator breaker is closed and the generator is SYNCHRONIZED to other generators. (Look it up; it's true. There are many YouTube videos even about how generators work.) It's not magic--it's just magnetism.

Best of luck. Let us know what you find out!

Pls find a typical pug bay picture . Each phase has 3x 646MCM cables and rest are Cables . u are seeing 5 plug bays for 5 cubical interconnected inside with a common Bus .
Mag and Throttle are in single plug and cable (shielded Twisted pairs ).

I have a recording of the issue in hand , if u have a drop box site , i will be able to drop u the video. U will then understand how engine behaves in 1-2 seconds .

All engines under same Throttle input should deliver the same KW . If the throttle i/P is different then the load shared will be unequal .
Same way the rate of change of throttle , will define how fast the engine is capable to adapt to sudden load , throttle rise to prevent drop in RPM of sync engines , when sudden power is demanded ( 300Kw in 1-2 sec) from the engines, the throttle response should be fast enough according to the change . Else all engines RPM will drop .
If 3 engines are online , and all 3 have different throttle response , then each engine will raise at a different rate, is it possible to have a reverse power condition then ? as one of the engines is not getting enough fuel at that fraction of a second .
Thus the throttle will change a lot to compensate for the drop in KW.
A stable engine will not have plenty variations in RPM , but it will also not have a fast changing throttle I/P .
But we require the dynamic response of the engine to be faster. ... as the same engine under slow changes in load have no issue . If the loading demand of 300Kw is spread across 3 seconds the engine has no issues , No - Kw condition . but if the same is demanded in 1 sec then it has a reverse power condition .
Hence i said , will plug in two CAT ET on two gensets , see what is the change in throttle when same load is demanded from these two engines , the throttle magnitude and rate of change ,slope should change at same rate, as at stable conditions they were delivering same KW @ sync speed.
Now Have changed settings from cubical to match the readings on two CAT ET.
Now both engines in theory should get same fuel at same rate under same change in throttle (almost) . Thus -Kw condition should not exist . ... . This is what in theory should work . ... yet to try in practical . Let me know what u think .


Okay; I'm beginning to understand the cubical thing. And, I'm beginning to believe these units are operated independently of other units--so the common bus in the big white thing with the five plug bays (sometimes called cubicals), and I'm presuming that big white thing is powering a load independently of a grid and other units/big white things. (I'm making this presumption based on some of the questions you've asked--but it's just a big SWAG (Silly Wild-Arsed Guess.)

I also presume you can see the total power output of the big white thing with however many units connected and supplying power to the common bus in the big white thing.

So, what happens to the total power output of the common bus of the big white thing when the unit connected to this questionable cubical does it's un-natural thing? Does it also drop by the same amount as the lost power from the unit connected to the questionable cubical? Do the other units suddenly increase their power outputs to compensate for the temporary loss of power from the unit connected to the questionable cubicle? Presuming this load being powered by the common bus in the big white thing is independent of other units or a grid, does the frequency suddenly decrease and the increase when the power output from the unit connected to the questionable cubicle drops?

If the common bus in the big white thing is powering a load (loads) independently of other units or a grid if the power was lost from one of the units the frequency of the output of the big white thing would drop--and the other units connected to the big white thing would GROAN and lots of black smoke would come out of their exhaust as their fuel racks were suddenly "opened" to increase the fuel to try to maintain the frequency.

But, the fact remains--whether or not the big white thing is powering a load independent of other units or a grid--one diesel generator set can't go to zero speed in an instant and come back to rated speed in an instant. The power output could drop precipitously and even go to reverse power--if the fuel were suddenly reduced--but the speed can't go to zero if the generator remains closed to the common bus in the big white thing. Not no way. Not ho how. Nada. Niente. Zippo. Nope. Just can't happen.

If the power output of the common bus of the big white thing drops by an amount equal to the amount of lost generation of the unit connected to the questionable cubical of the big white thing but the frequency of the output doesn't change by much--then the big white thing is either connected to a larger grid, OR the remaining units connected to good cubicals of the big white thing are STRAINING, and their exhaust is thick and black, to try to maintain frequency and load. But, still the fact remains: The speed of the unit connected to the questionable cubical of the big white thing CANNOT go to zero and come back in one or two or even three seconds without damage to the diesel or the coupling between the diesel.

Even if the generator excitation system (the "AVR") was not working properly, it still couldn't go to zero speed IF THE GENERATOR BREAKER REMAINED CLOSED DURING THE SPEED DEVIATION. Not no way. Not no how. Nada, Niente. Zippo. Nope. Not on planet earth, anyway, unless the laws of physics on planet earth are suspended at your site.

There are just so many things which you can be looking at: total power output of the big white thing (never mind the division of the load--the load division is probably the result of some load sharing scheme of the big white thing, and it's entirely possible for each and every unit to be putting out a completely different amount of power and still be producing power at the desired frequency and level. That's just "load sharing" and nothing more--and it's probably the thing that's sending the demand signal that drives the throttle which controls the fuel rack which controls the amount of fuel being injected into the cylinders.

You can be looking at the exhaust stacks of all of the units when this is happening.

You can (probably) be looking at the fuel flow-rates to all the units when this is happening.

You can be looking at the frequency of the common bus of the big white thing. And, the unit connected to the questionable cubical.

I'm sorry; but I must be missing something here. And so must you. Because this just can't happen and the diesel and coupling between the diesel and the generator and the generator rotor just can't be withstanding this beating. Or, the description is just totally incorrect. I firmly believe the unit connected to the questionable cubical IS NOT losing power and it's CERTAINLY NOT going to zero speed and back to rated speed in one or two or even three seconds. I firmly believe the problem is what's being recorded and reported that's the problem. Because if this were really happening, either those are some DAMN FINE CATs and generators and couplings, or there's something wrong with the sensors/instrumentation/data recorders.

Or physics laws are suspended inside the fence at this power plant.