load, ammeter and black smoke questions

[Jimc]

so we had a real nice day today. fired up the 003 because i wanted to run it for a couple hours under a higher load and i really wanted to see what it could do at its max. gen is running 1ph 240v. i messed with it adding and removing loads from my shop. i first fired up my compressor and put that on constant run. thats 7.5hp and is rated at around 34 amps running. i slowly started turning on lights in the shop and watching the ammeter. i have 30 - 4 tube ho fixtures in the shop so i just surn them on a few at a time until i get it where i want. here is where i am finding something not right. gen is running at 80% load and all is beautiful (according to meter). 85% i get slight black smoke from the exhaust, 90% is a little more noticeable. 95% its chuggin black smoke. quite a bit of it. enough to make clouds. at first i am thinking this thing is supposed to be able to handle 125% atleast for a short time. ill tell you now there is no way i could ever get this thing to that level. my first thought is something is not right with the engine but since i rebuilt it this thing runs so awesome its not funny. i dont have a clamp on ammeter or i wouldnt be guessing here. i started thinking about the loads i had on it. at 90% load i had the compressor running, figure 32-36 amps there and i also had 20 of the fixtures on. the ballast says each fixture is rated at 2 amps ea. thats a total of 72 amps and even if i stay conservative and say realistically the lights were 1.5 amps then that would still be a total of 62-66 amps. so i think the ammeter on the gen is not right. i believe a 72 amp draw is far more than 90% rated load for an 003a.....am i correct? that seems like alot. i have another non running 003 here and it has the same ammeter so i did a swap out to see and i get the exact same reading with the second gauge as well. real quick note i flipped on all 30 fixtures and was able to stall the gen out. i quick shut them off to avoid brown out to anything in the shop for more than a couple sec. i thought the breaker would have tripped but thats another discussion. with both gauges reading the same i looked over the wire schematic and i see the ammeter is wired to the voltage selector switch but after that it loses me. one wire seems to goto the sw and just end. where does this gauge pick up the current draw and how does it read the draw and come up with a total for both legs? right now i assume where its picking up the signal is the problem. any of you guy with more exp on the electrical side have any input?

 

[doghead]

Capitalization would sure make that easier to read.

 

[Triple Jim]

72 amps X 240 volts = 17,280 watts = 170% of rating. Are your lamps running on 240V? If they're running on 120, you'll need one on each of the two 120V legs to provide the load at 240. Also, you need to check the load with some kind of ammeter of known accuracy to verify it. Just because the ballast says "2 amps" doesn't mean that's really what it's drawing. On top of that, fluorescent lights are not purely resistive loads, so you're adding the complexity of a non-unity power factor. Just some thoughts.

 

[Jimc]

Yes the lights are running on 120 jim. They are pretty well balanced on each leg in my main panel. Yes i need i get ahold of a clamp on ammeter to check the load to be absolutely sure. Overall it just seems like alot of stuff running. More than a 90% load and enough to where i dont think its and engine issue.

 

[johnray13]

Could your loads have been out of balance? With out an ammeter, your kinda pissing in the wind. I would start there so you know what you are actually drawing at the gen set terminals.

FWIW, I load tested mine on a Dayton 3P, 208, 15kw job site heater and it didn't blow any black smoke. I believe it metered around 120%. I know it wasn't pegged.

I agree that something is out of whack. A new Amprobe is in order.

John

EDIT: Jim answered before I could type my response...

 

[Jimc]

Yeah ill dig up an ammeter maybe tomorrow and check it out.

 

[Grega]

Jimc Of all that stuff you have in your shop and no clamp on ammeter!!!!

 

[Jimc]

Haha i know i know!! Ive always done good with my trusty fluke dmm .....until i started messing with generators lol.

 

[Jimc]

Ok i picked up a clamp on meter. Its not a real high dollar one and varies a bit depending i
on the location of the cable through the loop but assuming this thing is correct, average its giving me a reading of about 55amps and its showing only 95% load on the meter on the generator. If my calc are right thats about 130% of rated load. At this load i have a little bit of exhaust smoke. Not real bad or overwhelming. At what would be 120% i have no smoke at all.

 

[steelypip]

I bet your unmeasured factor is the reactive load of the air compressor. It's a big inductive load, so your power factor is not very close to 1 any more. Depending on the ballasts, the fluoro lights may or may not be a neutral load either.

The generator is rated at a power factor of 0.8. The power meter is calibrated assuming that same power factor of 0.8. The load banks used load testing are all resistive loads, which means they have a power factor of 1.0. THAT'S why the manual says to run the generator at 125% indicated on the load bank, because 1/0.8= 125% - the generator must be over-rated by 25% to be able to put the nameplate power into a load with a power factor of 0.8.

I really, really don't want to go into a long theoretical discussion of real versus imaginary power, but simply put - you're probably making power that the systems can't effectively use because you need some capacitors in the line to get the power factor closer to 1.

Or you could just use resistive heating elements for a test load bank like the army and everybody else does.

 

[Jimc]

Ok so basically what your saying is depending on what is drawing the power it will make a difference in how accurate the ammeter on the gen reads?

 

[Crawdaddy]

Not to drag the thread off-topic too much, but for a power factor of .8, do the meters on these generators indicate 80% for the full load or 100%?

 

[steelypip]

I'm pretty sure that the output meter is calibrated assuming a load power factor of 0.8. So it will indicate 100% when it is pushing rated power into a 0.8 PF load. When it is pushing rated power into a 1.0 PF load (like a resistive load bank, lots of incandescent light bulbs, space heaters, an electric stove, a toaster oven (but not a microwave)), then it's going to indicate 125% load at rated output.

 

[Jimc]

thats interesting. You would think if there was say 45 amps on the legs that the meter would read the same no matter what its powering. Im not to bad with elec but never had to deal with pf or alot of elec theory.

 

[Triple Jim]

I'm pretty sure that the output meter is calibrated assuming a load power factor of 0.8. So it will indicate 100% when it is pushing rated power into a 0.8 PF load. When it is pushing rated power into a 1.0 PF load (like a resistive load bank, lots of incandescent light bulbs, space heaters, an electric stove, a toaster oven (but not a microwave)), then it's going to indicate 125% load at rated output.

I had assumed otherwise, but you're right, that's what the -12 manual says:

A .8 power factor load equals a 100% meter reading for maximum rated load. A 1.0 power factor load equals a 80% meter reading for maximum rated load.

 

[Triple Jim]

thats interesting. You would think if there was say 45 amps on the legs that the meter would read the same no matter what its powering.

It will. The problem is when the current and voltage are not in phase you have a non unity power factor, and the actual power transferred is not found by simply multiplying volts x amps, but is less than that. This means the meter would read higher than you would expect for a given % real power. This is where VA comes in... that's volts x amps even if the PF is not 1. You've probably see UPSs that are rated at something like 600 VA but only 500 watts.

The meter is actually sensing current. According to the TM, the meter is calibrated to read 100% with a PF = .8, at 10,000 real watts. With a PF = 1 and 10,000 real watts, the current is lower, so the meter reads 80%.

 

[Jimc]

That is some really good info Jim. Thanks to you and steely for sharing. I'll have to do a little reading now on pf to educate myself a little more on it.

So another question, is the ammeter on the gen just a regular analog ammeter like the clamp on one i have with a gauge face printed in % of load or is it an actual different type of gauge?

 

[Triple Jim]

It's an analog ammeter that uses a current transformer in the generator output to provide a signal to the meter that's proportional to the main output current. Current transformers usually have very few turns of big wire on the primary, and in many cases are just one turn made by passing the conductor once through a hole in the transformer. The winding of the secondary connects to the meter, and the two are designed to work together to give a calibrated reading.

Your clamp-on meter is similar, with a 1-turn primary formed by the wire passing through the iron structure of the "clamp". A display then converts the signal from a secondary winding inside to something you can read.

 

[Jimc]

Thanks Jim

 

[steelypip]

People have trouble with the ideas behind real and imaginary power mostly, I think, because they're usually presented as mathematical abstractions rather than with a physical metaphor (remember how successful the hydraulic metaphor is for DC systems). The names used don't help either - 'reactive load' isn't very descriptive, and 'imaginary power' sounds weird, stupid, or both. But those are the terms we're stuck with.

I tend to think of a massive resonating object - a child on a playground swing. if you apply energy in phase with the motion of the swing, then the amplitude of the swinging increases. By comparison, applying energy out of phase results in wasted work - you pushed harder than the amount of amplitude boost would suggest. That 'lost work' was imaginary power. You expended energy, but the load didn't respond as desired. The meter works by doing a very small amount of work (moving the needle) based on the amount of real power being made. It can't measure imaginary power created. The army compromised by assuming that the average load on an MEP has a power factor of 0.8 and calibrated the meter to read accurately when that was true.

In the ideal world, all loads have a power factor of 1.0. In the real world, current flow frequently leads or lags voltage. When it does, we waste energy. If I'm buying and schlepping diesel fuel up the hill to to burn to make that power, I'm highly interested in maintaining a power factor as close to unity as possible...

FYI: A Kill-A-Watt is the cheapest power factor meter I know of. Handy little beasts - I wouldn't run a generator without one.