Generator fuel consumption list

[BioHazard]

I thought it might be interesting if we all started a list of different generators and their fuel consumption per kilowatt hour produced. I'd love to see a database of all different kinds, types, sizes, and fuels - gas, diesel, propane, alternative, whatever.

Sadly the only generator I have right now that's not in peices is my little 800w HF 2 stroke. Tonight I measured exactly 2.8kwh over 4 hours on one gallon of gasoline/50:1 oil mix at near full load. 3600 RPM.

Now what kind of data have you guys got? 

[mbryner]

Well, here's the numbers on my JKson listeroid 6/1 back when I was running it at ~480 rpm (remember when I overheated the big end?)

Fuel use:                      4.87 hrs/gal  or  0.205 gal/hr
@ Average load:         1600 watts
@ RPM:                       480 rpm
Overall efficiency:        7.76 kwh/gal  or  0.129 gal/kwh

Marcus

[mobile_bob]

changfa S195 idi, driving an st7.5 and twin 110-555jho alternators, bosch coolant pump
and thermostatically controlled rad/fan.

once warmed up to thermally stable operation and tested at max loading, with fields disabled on the 555 alternators.

7100 watts, AC   =  0.1001 gal/kw/hr

Bill Rogers reported just a bit better than 0.090 gal/kw/hr with a changfa 1115DI engine driving an ST12 generator head
again this is at full rated load, and i don't recall what that was for sure, however i am pretty sure it was over 10kwatts.

any generator that can get you to between 0.085 and 0.100 gal/kw/hr will be about as good as one is going to get, even the big boys
have trouble getting into this range with gensets in the class we are working with.

bear in mind any genset will return its best numbers at full rated load (or nearly so) and those reported consumption numbers will always
be significantly higher at part and low load operation. so unless you need 10kwatts continuously you might not want to go with a changfa 1115di engine, unless you plan on running it intermittently.

bob g

[Ronmar]

Here is a graph I did of my 6/1-ST5 at various loads on #2 offroad diesel.  Pretty similar numbers to what Marcus posted.

[Tom Reed]

Very good Ron, right at .10 at 3.5kw.

[Horsepoor]

Harbor Freight & Tool 800/900 Watt two stroke generator.

Poured 1.2 gal of RUG into a completely empty tank. Recorded the total time the engine took to consume this fuel: 7 hrs & 56 minutes rounded up to 8 hrs.

1 gal of RUG = 114000 btu/gal or abour 33.41   kWh of total energy

1.2 gal of fuel has 136,800 btu or 40.092 kWh of total energy available to the engine

Measured by (Kill-a-watt meter) a sustained constant applied load of 540 Watts for 8 hrs or 4.32 kWh      
Engine produced 3.6 kWh/gal
      
10.8%   Efficiency (Overall)      

[Jedon]

I didn't measure this myself but my main generator ( Onan 10K) is:

Typical fuel consumption
No. 2 diesel fuel, Gal/hr (L/hr) No load Half load Full load
10HDKAG  0.29 (1.10) 0.55 (2.08) 1.04 (3.90)

10K watts at full load, 1g/h so around 10KW per gallon according to the specs?

[Lloyd]

My perky/cat/volvo diesel dc bat charge http://www.microcogen.info/index.php?topic=709.msg26061#msg26061      

For the 1.5 hrs run time  the generator produced on average 3.8 kw electricity and 7-8,000 btu/hr of domestic HW cogen side. I say not to bad for about a liter of #2 DINO. Test based 30 day running average 1.5 hrs every other day. Kw measured by an bat monitor, and HW estimated based on 16 gal. tank, and temp rise over the run time to final tank temp.

Lloyd

[cujet]

Honda specs the little EU1000i at:

3.8 hours, 0.6 gallons, 900 watts. 5.7KWH/gal
or
8.3 hours, 0.6 gallons 225. watts. 3.1KWH/gal

[glort]

Harbor Freight & Tool 800/900 Watt two stroke generator.

Poured 1.2 gal of RUG into a completely empty tank. Recorded the total time the engine took to consume this fuel: 7 hrs & 56 minutes rounded up to 8 hrs.

1 gal of RUG = 114000 btu/gal or abour 33.41   kWh of total energy

1.2 gal of fuel has 136,800 btu or 40.092 kWh of total energy available to the engine

Measured by (Kill-a-watt meter) a sustained constant applied load of 540 Watts for 8 hrs or 4.32 kWh      
Engine produced 3.6 kWh/gal
      
10.8%   Efficiency (Overall)      

I have one of those little 2 stroke Gennys.
I always thought it was good on fuel but maybe I was just pulling a lot less power off it than I thought?
Rather than measure consumption, I was just mainly going on run time in practical application.  I thought it was good for a s stroke. Maybe in reality it's as poor as I assumed a 2 stroke would be.

I don't bother with consumption much. I run veg fuel so it cost me cents per gallon.

In comparison to other diesels, are listers/ listeroids as good on fuel as legend has it?

[thomasonw]

One question I have:  How do folks measure the fuel consumed?  Esp if drawing from a large tank?  I know for my little Kubota / DC generator I only have the Kubota 'spec sheet' fuel per hour at rated HP that I can draw from, and assuming I have 'fully loaded' the motor (which was the goal of the integrated controller / regulator project) I come up with:

     1,700W @ 2,600 RPMS (measured, and tuned to be 'full load' on motor via EGT, Exhaust color, and ability to respond to governor changes)
     5.0HP continuous @ 2,600 RPMs,  0.47 B.S.F.C  (from Kubota Spec sheet)
                    0.47 BSFC @5hp (assuming 7lb/gal) -->0.26 GPH (If I have done my math correctly)

So, 1.7KW/hr at 0.26 GPH -->  0.153 gal / kwh  (or, 6.5kwh/gal)
     
However, this is all calculated from the spec-sheet fuel consumption.  I have not placed a graduated tank on the small generator to measure the actual consumption.

Wish I had better numbers, but based on what I have read - 12v DC alternators are just not that efficient.  One of the things I am going to try is swapping out the current alternator for one of my spare 200A Leece Nevilles - belted to spine around 3,000 RPMS and use the amp regulation capability of my controller to limit the amps produced (so as not to stall the poor little Kubota).  I will use the EGT as a control point, and I want to see if I can get any higher overall system efficiency.  If I really care, might spring for a higher efficiency (Hair-pin??) alternator...

And if anyone in the PNW has an already-rigged up graduated cylinder they would lend me, might be interesting to measure the actual consumption.  (Or, does anyone have other ideas??)

[mobile_bob]

this is the way i measure fuel consumption, and it came after being frustrated with numbers that went all over the board.

not only from users but from some of the oem's as well

let alone problems i was having with monitoring a stable load, doing the math, measuring fuel, etc

there had to be a better way

so this is what i came up with and it works very well

i put my plastic fuel container on a gram scale, into which i suspend my fuel suction pipe
which is affixed to the support structure and has no connection to the container,, siphon feed from there down to the fuel filter on the engine to be tested

this allows me to measure with quite good accuracy exactly how much fuel i am consuming in grams weight

the next issue was variable loads, and most all loads are variable whether we want to admit it or not... resistive heaters are good for a load bank but, if they are subject to drafts or wind the elements will shift in resistance and there goes the fixed load  it will drift all over the place.  let alone if the thing gets hot enough that the thermostat unknowingly cuts off power periodically during the test... reading amps, and volts and doing the math will give you instantaneous watts consumed but it won't give you accurate numbers over time.... so

i went with a GE watt/hr meter of which i picked up a few off ebay a (new) a few years ago. they are digital readout which is very helpful and they can be set into calibration mode which enables them to read out power consumption of a load down to watt/hrs plus or minus 2 watt/hrs   that is very high resolution and it also factors in power factor, and will keep track and log variable loads, no matter how variable they are.

then you simply run the unit to be tested up to normal stabilized operating temperatures, apply the load you want to test at,  push the start button on the GE meter, note the fuel weight and run until you have produced 1 kw/hr at that load (or 2,3...kwatt/hrs) and divide the amount of fuel consumed by the number of kw/hrs produced and you have very good numbers of what the unit under test is using in fuel consumed in grams per kwatt hour @ that specific load. 

@ that specific load is important to note in that any generator will show its best numbers near its full rated load in my experience.

remember the first kwatt is the most expensive kwatt to produce when it comes to power generation, each successive kwatt lowers the fuel consumption because you dilute the base losses of the unit, the unit burns x amount of fuel whether it is producing power or not. that number is fully accounted in the first kwatt produced, however it is divided by two if you produce 2 kwatts, by 3 if you produce 3 etc...

now there will be those that pipe up and state this is only so much rubbish that you must have 50k dollar test cells to get meaningful fuel consumption numbers. to that i would argue the contrary.

reason being we can prove the numbers mathematically to be pretty damn accurate.

once we have the fuel consumption numbers from a generator, we can then set down with those numbers and with some simple math sort it all out

for instance if we have a generator that at full rated load uses for instance 322 grams of fuel per kwatt hour, we only need to look at two other factors

one is the advertized consumption numbers from the engine manufacture and the expected or advertized efficiency numbers for the generator

if the engine manufacture states the engine will use 258 grams of fuel per kw/hr mechanical

and we either know from advertising or accept without argument from the averages noted in engineering texts that single phase generators of less than 50kva of the design we are using (typical brush type ST head) we will find they average about 80% efficient

so we take the mechanical consumption of 258gr/kw/hr mechanical of the engine divided by the efficiency of the generator of 80% we get

258/.80= 322.5 gr/kw/hr

now we see that this number correlates nicely with what we actually measured in fuel consumption... however there will be those that would argue things like "how do we know that the generator is not more efficient than 80%? or the engine is more or less efficient"

either of those questions while being fair to ask, really have not bearing on fuel consumption but we can analyze this question a bit further

first of all we have reasonable confidence in our fuel consumption number of 322gr/kw/hr because we used a gram scale and we used a meter of reasonable accuracy and good resolution
and if we ran a test long enough or enough tests to prove that the numbers repeat (good testing procedure) then it is doubtful even an engineer would argue over a percentage point... so we will call this good enough

this leaves the other two factors, the engines rated mechanical kw/hr per gram consumed and the efficiency of the generator

we will find that in order for the generator to be markedly more efficient than 80% the engine would have be be markedly less efficient than advertized or visa versa.  in either even it is unlikely that the either the advertized numbers from the oem or the expected efficiency numbers of the generator in this class is going to be plus or minus a percentage point, neither of which have any bearing on the actual fuel consumption of the unit as a whole.

an interesting topic and question
i only relate the above verbiage for those that don't know how to go about doing decent testing to get numbers that are useful for comparison.

this is not to say it is the only way, the best way, or anything of the like, it is only to say it is one way that will produce numbers of reliable and repeatable accuracy that can be used for comparative measures when developing generator systems.  once one has the test equipment setup and has got things ironed out and understands how good testing is done it is surprising just what can be learned from doing the testing

we can test and get numbers for excitation, drive system losses, water pumps, other generators or alternators, etc.

it really helps to clear up some long held misconceptions on things like belt drives (how efficient or inefficient is one type vs another ) 

anyway i am preaching to the choir again

bob g

[Ronmar]

When I did mine, I added a "T" to the fuel system ahead of the fllter.  To this T I added a valve and small clear line up to a converted 1QT gear oil container placed upside down with the bottom cut out to act like a funnel.  I put a mark on the clear line to act as a start/stop point.  I would start the engine up to get it fully warmed on the main tank.  I would then add a little fuel to the test tank.  Once the engine was warmed, I apply the test load and let the load and engine temps stabelize. While this is happening, I carefully measure a precise ammount of fuel into a transfer container.  When engine and load are ready, I open the valve to the test tank and close the valve to the main tank.  As soon as that little bit of fuel in the test tank burns down to the mark on the clear line, I start the timer and pour the carefully measured fuel into the test tank.  Then the fun begins.  I have a fluke clamp-on amp meter connected to fluke 87 meter on the load feedline and a fluke 77 volt meter across the output.  I then start taking readings every minute for voltage and current output for the given load.  I do these minitue by minute readings for the duration of the test and when the fuel burns down to the mark, I either add another carefully measured volume of fuel to continue a longer test, or I stop the clock, shift the fuel valves and crunch all the numbers to get an average of my electrical readings for the duration of the test for the measured quantity of fuel burnt.  The Meter Bob describes is a better way to do it, but the way I did it, although labor intensive, yields accurate and repeatable results when the rest of the process is administered consistently.

They were accurate enough that when early in the testing process, I ran 2 tests back to back and got different readings, I had to re-evaluate my warmup procedure.  For my first few tests, I wasn't letting the engine warm fully(on a listeroid this takes quite a while)  And this has quite a large effect on efficiency.  The fully warmed engine is more efficient, and the test was able to clearly show me this.  I then started doing a very long loaded run before I started any test.  When I did this, my numbers were very repeatable.  About the only thing I didn't do was measure by weight.  All my tests were done with fuel measured by volume...

[mike90045]

I measure my day tank refilling by the quart, and use my inverter's logging function to measure what I produce. (the XW meters are bi-directional)

So for Apr, May, June, I've logged 9 hours, 16 quarts fuel and 16.5 KWh

I think I'm getting .44gal fuel per hour and 4.1Wh per gal of fuel,  roughly, if I did this right.   Long term, it should
give me a pretty decent average.  I know I'm not loading my engine fully, I don't think with only 30 hours on  it, it's quite ready for full power.