Cummins Onan genset - how to measure fuel burn - Q for ARE?

[jfm]

Out of curiosity, doesnt sea state account for more than 10 pct? I mean, if A to B will use 1000 litres, are you happy to go with 1100 litres ?

given that fuel consumption is going to be dependent on sea state?

Generally the sea state makes no difference to 80 feet/60 tonnes. If for some reason I am going to punch into a big sea then I can allow the extra fuel, but my objective is to get the raw data to begin with, and then I can do the maths including the seas state fuel burn factor. Gjgm, no I wouldn't be happy with100 litres spare but if you multiplied all your figures by say 5 then I probably would (with the right instrumentation and a "plan B" if something when wrong as opposed to pure open water)

Is it not more prudent to buy say 1000 litres of fuel in Bonifacio to be on the safe side and not worry about your gennie fuel burn?

Yup, agreed, prudence and various things go into the maths when you make the buy fuel/don't buy fuel decision, but what I'm trying to do is get the best quality raw data in the first place

If the baffling in the tank is good, then the level will not surge (much). Would an ultrasonic or optical meter give you a precise enough output after calibration ?

And even small aircraft need fuel flow system monitors with the kind of precision jfm is seeking, with a greater impact if they get the figures wrong. Perhaps something here ?...The Microflo device (other totalizers are available ) gives fuel flow, fuel used, fuel remaining, and endurance, though whether the output can be integrated into one of the big Match screens is a known unknown.

Thanks. See below re baffling. I'll take a look at those. Calibration is very difficult, practically speaking, on a genset that shares its fuel with main engine tanks of this size. I really need n2k output though other outputs can be used with converters

so a 38% variation....I keep 20% up my sleeve as a minimum, running in bad weather and pulling air instead of fuel on a roll,

I honestly get nothing like 38% variation. I have instantaneous read out on the dash of litres/mile and litres/hour and would see even a 10% variation immediately and could allow for it or fix the problem

A lot of people rely on electronic equipment which gives you three decimal places of accuracy but on flawed date. Miles run and fuel burnt per NM is the only real data worth looking at. Work back from tank fill/miles run (through the water) data.

I'm not suggesting relying on anything. I'm suggesting getting the best quality raw data to do the maths on, but given that it is electronically measured data I'd always do another "sanity check" calculation too

Well, I wasn't suggesting that the mains had anything to see with the genset fuel burn math.

Yep, sorry, forums eh. It would be much easy to discuss this in a pub

Anyhow, all that aside, I did understand that you're after something better than an estimate - my comment re. the Floscan was in that direction.
Maybe worth a look, if you can't get the "real" numbers out of the Onan ECUs. Which I agree would be the easier/better way, if feasible.

Yup, ultimately I'm just trying to get the raw data displayed easily on a screen so that I can then apply whatever maths is appropriate in the circumstances. Thanks re floscan. My understanding (but only afaik so I'm happy to be corrected) is that you have to calibrate them. You fill the boat's tanks, then drive the boat for an hour or 2, then fill the tanks again and measure the fuel used from the fuel pump at the dock. Then compare that reading with the floscan measurement, then you enter a correction factor. Obviously that process cannot practically be done with a genset. The Maretron system uses positive displacement metering so this calibration is not necessary. The Maretron gear is also much smaller to fit. TBH, the Maretron stuff looks better in every respect than Floscan. (But a direct link to j1939 data would be even easier than Maretron!)

Morning,
Working at PYI today, I will reply when I get back to hotel later
Anthony

Thanks ARE.

***************************************

Just as regards tank design, baffling, and possibility to suck air in a big sea, the pics below are my last-resort central fuel tank. As regards orientation, the second picture shows the starboard side of the tank. The bottom of this tank is 750mm or so lower than the bottom of the two other (bigger) tanks. So when I'm down to my last fuel I'm pulling fuel from the bottom of this tank. As you can see, it is quite a nice design and allows me to pull fuel down to say the last ~150 litres (the foot print at bottom of tank is about 1.5m x 750mm so every 100mm of tank depth is ~110 litres), so I do not need to allow much "sloshing around" factor in my fuel maths

 

[Bandit]

My point is you have built in fuel flow meters on each main engine and you also want this for the generator.

I assume each of them have to measure flow and return to give a reasonably accurate burn figure.

I don't know how accurate fuel burn is over a journey +/- say 10% or possibly 20%, probably the % inaccuracy is greater than the generator actual fuel burn?

I looked very seriously at fitting Floscan meters to my engines which are mechanical VP 63P's but I have not done so as by monitoring my fuel burn over the years I have come up with good approximations which I feel happy with. Perhaps if I was looking to do long range cruises I might do that. I Know what I burn at at normal cruise rate , and I know very accurately what is left in each tank.

It does concern me using electronics to say what has been drawn from the tank values. less fuel return line values, to produce fuel burn values , then do the same for the other engine and the generator to come up with a "quantity of fuel that should be in the tank" and the inherent inaccuracy of doing this.

I prefer to know what is in each fuel tank to within 15 litres before I set off and to keep an emergency reserve of 20% up my sleeve.

 

[MapisM]

Thanks re floscan. My understanding (but only afaik so I'm happy to be corrected) is that you have to calibrate them....

Aha, I see.
Actually, I'm a bit surprised, 'cause I never heard about such need so far, also from folks in the US (where Floscan it's much more widely known and used than here) which installed that stuff on boats whose range is measured in days (if not weeks), rather than hours....
...And the only complain I ever heard is about the cost of the thing, never its accuracy or reliability. In fact, it's also used for big commercial boats.
Then again, I never installed one myself, so what do I know?

Btw, nice setup for the lower center tank. Just curious, is there any reason why it wasn't built with some sort of V-bottom, to fully exploit the space that I guess you should have where it's placed?

 

[MapisM]

It does concern me using electronics to say what has been drawn from the tank values. less fuel return line values, to produce fuel burn values, then do the same for the other engine and the generator to come up with a "quantity of fuel that should be in the tank" and the inherent inaccuracy of doing this.

Having 100% mechanical engines (both mains and genset) onboard, I sympathise with your concern.
But the accuracy of fuel burn data in modern electronic engines is simply in another league, compared to any other measure you can think of.
You don't actually need to measure any flow, neither drawn nor returned, because the ECU decides in any given moment how many drops of fuel squeeze in the combustion chamber, and if you can simply "read" that number out of the ECU, you've got the most reliable figure you can get.

That said, I'm happy to live with my sight gauges and zero instruments on the helm station.
Also because, even in the longer passage I made so far, what I had left in the tanks upon arrival was sufficient to cruise all the way back to the starting point....

 

[jfm]

My point is you have built in fuel flow meters on each main engine and you also want this for the generator.

I assume each of them have to measure flow and return to give a reasonably accurate burn figure.

I don't know how accurate fuel burn is over a journey +/- say 10% or possibly 20%, probably the % inaccuracy is greater than the generator actual fuel burn?

I looked very seriously at fitting Floscan meters to my engines which are mechanical VP 63P's but I have not done so as by monitoring my fuel burn over the years I have come up with good approximations which I feel happy with. Perhaps if I was looking to do long range cruises I might do that. I Know what I burn at at normal cruise rate , and I know very accurately what is left in each tank.

It does concern me using electronics to say what has been drawn from the tank values. less fuel return line values, to produce fuel burn values , then do the same for the other engine and the generator to come up with a "quantity of fuel that should be in the tank" and the inherent inaccuracy of doing this.

I prefer to know what is in each fuel tank to within 15 litres before I set off and to keep an emergency reserve of 20% up my sleeve.

No, my main engines do not measure flow and return. The engine's ECU is constantly computing a value for amount of fuel to inject in each cylinder burn cycle, ie 12 data values for each 2 revs of the engine. This data is simply added up and a value for instantaneous litres per hour (at something like a 10Hz refresh rate as a guess from looking at how fast the number is tweaked on the screen) and total fuel burnt is then calculated via a clock in Caterpillar's software. Total fuel burnt with trip functions etc is also calculated (based on Caterpillar's lph value) by Garmin's software independently, and I can see both Caterpillar's and Garmin's data on my screens. So long as the injectors are working (and my engines are newish) that data is better than 1% accurate, it seems. I therefore have a very accurate figure for fuel burnt by the main engines, as well as very accurate read out of litres/mile underway. I also have CAD drawings of fuel tanks and know the volume precisely, so I know how much fuel I have left with very high accuracy, apart from the fuel burnt by gensets which "let the side down" by requiring an old fashioned calculation.

I have no problem with old fashioned calcs and do them anyway when I am tight on range, as a back up/sanity check to the electronic data. I'd just like to get easy access to electronic data for the gensets as well. BTW, genset burn is often 7 litres per hour btw; each genset has a 3.3 litre engine.

EDIT: I see I have repeated what MapisM has said on the ECU data - sorry

 

[jfm]

Aha, I see.
Actually, I'm a bit surprised, 'cause I never heard about such need so far, also from folks in the US (where Floscan it's much more widely known and used than here) which installed that stuff on boats whose range is measured in days (if not weeks), rather than hours....
...And the only complain I ever heard is about the cost of the thing, never its accuracy or reliability. In fact, it's also used for big commercial boats.
Then again, I never installed one myself, so what do I know?

Btw, nice setup for the lower center tank. Just curious, is there any reason why it wasn't built with some sort of V-bottom, to fully exploit the space that I guess you should have where it's placed?

Yup but I think Floscan was the leader for many years because they kinda had no competition. People tolersted its shortcomings because it was much better than guesswork. I have not tried either, but gut feel says that the new Maretron fuel flow hardware is quite a lot better (and Maretron is very much a leading company - no back street operation as you know!)

I read the manuals online and you definitely have to do trial and error calibration of floscan, but not maretron

Yes I suppose my tank could have had a V bottom for the price of an extra weld or two. I'll do that on Match 3! Actually it could have had 300 litres more if I had a shower rather than a bath in the m/cabin heads. The bottom of the bath/shower is below w/line, so has to drain into a grey water tank/sump then be pumped overboard. Because of the volume of water in a bath, this tank needs to be bigger than it would have been if I had just a shower, and so this tank/sump "eats" some space that this fuel tank could have used. I'm not sure I should be admitting this to you MapisM because I know you will be horrified at my priorities here . Thing is, I wanted to be able to do Antibes-Bonifacio return, and Antibes-Palma fuel dock, at P speed and with some reserve fuel, and the arrangement I have delivers that whereas most other production P boats won't. Any longer trip needs a fuel stop anyhow

 

[MapisM]

I read the manuals online and you definitely have to do trial and error calibration of floscan

Understood. You live and learn, as they say....

I'm not sure I should be admitting this to you MapisM because I know you will be horrified at my priorities here

LOL, naah, I can actually understand that. I also like Bisazza tiles a lot, as you know.
If anything, for the jacuzzi I would have considered to place it on the f/b instead, if feasible. Which in turn could have drained directly overboard...


On a separate note, re...

BTW, genset burn is often 7 litres per hour

As I said, the 4 l/h pan estimated seemed reasonable to me, and your number now made me curious, so I just had a quick look at Onan spec sheet:
https://powersuite.cummins.com/PS5/...nary_Asset/pdf/Consumer/specsheets/a-1494.pdf
And in a sense, both numbers are correct, if you look at 2/3 of page 2: the fuel burn ranges in fact from 1.9 l/h with no load, to 7 l/h at full load.
4 l/h is exactly at 50% load, and incidentally the 5 l/h I assumed for good measure is at about 3/4 load.

I'm saying this because if you are experiencing an AVERAGE fuel burn of 7 l/h, either you're always using a helluva lot of electric power (which I wouldn't expect while anchored - i.e. no a/c, surely?), or your genset is burning more than it should.
Just a thought...

 

[MapisM]

EDIT: I see I have repeated what MapisM has said on the ECU data - sorry

Haha, not really. We both posted exactly at the same minute, so it's fair to say that we just thought along the same lines....

 

[jfm]

I just had a quick look at Onan spec sheet:
https://powersuite.cummins.com/PS5/...nary_Asset/pdf/Consumer/specsheets/a-1494.pdf
And in a sense, both numbers are correct, if you look at 2/3 of page 2: the fuel burn ranges in fact from 1.9 l/h with no load, to 7 l/h at full load.
4 l/h is exactly at 50% load, and incidentally the 5 l/h I assumed for good measure is at about 3/4 load.

I'm saying this because if you are experiencing an AVERAGE fuel burn of 7 l/h, either you're always using a helluva lot of electric power (which I wouldn't expect while anchored - i.e. no a/c, surely?), or your genset is burning more than it should.
Just a thought...

Ah, ok, you are right. 4-5 litres/hour would be right then. Yep I'd normally expect to be running one genset60% loaded most of the time -stabs, fridges, small batt charger load for 24v draw, a/c, and then cooking now and again

Going back to your earlier comment, yes it would take 100 hours to burn 400 litres and I would notice that. I guess I'd still just like to have the data though! On principle!

 

[burgundyben]

Going back to your earlier comment, yes it would take 100 hours to burn 400 litres and I would notice that. I guess I'd still just like to have the data though! On principle!

I'm going to assume that as you have asked the question that the engine in the genny is a common rail one, in which case it knows exactly how much fuel its used, so all you require is the data collection and presentation.

Can't be hard?

 

[jfm]

I'm going to assume that as you have asked the question that the engine in the genny is a common rail one, in which case it knows exactly how much fuel its used, so all you require is the data collection and presentation.

Can't be hard?

The existence of the electronic data is nothing to do with common rail; the rail is upstream of the injectors. The data could be equally be available for an engine with no CR and with unit injectors instead. It all depends on whether the injectors (whether fed by a rail or unitised pressure source) are electronically controlled. If they are, the data must be there. It doesn't follow necessarily that the data is in networked j1939 format of course.

I agree it ought not to be that hard but nevertheless I'm still trying to find out how to do it!

 

[ARE]

This may be a question for ARE but if anyone else knows the answer I'd be interested to hear please

I'd like to get Cummins-Onan genset fuel burn data onto N2k network. Does anyone know how? My gensets are model code MDKBT, 23kva, though I do not think this Q is specific to a particular genset.

My research so far comes up with following possibilities:

1. Fit maretron J2K100s . These take the gensets' j1939 data and put it onto the boat's N2k network. Perfect, you'd think, but Maretron tell me this wont deliver fuel data onto the network, presumably because the Onan electronics don't "talk" this data http://www.maretron.com/support/know...000+network?. ARE is it really true that so little data can be grabbed off the genset's j1939 network? Is there a way to get fuel data?

2. Fit Maretron flow meters into the fuel flow and return lines. These then create the correct data for N2k network. These would work well, but cost is about £1000 per genset and I'd like to see if #1 above can be made to work before installing these. http://www.maretron.com/products/ffm100.php

3. Fit floscan metering. Very expensive, and impossible to calibrate for a genset, so this one is a non starter.

Any better ideas?

John,

The Kubota engine fitted to your gensets is a mechincial injected engine and the generator controller does not monitor fuel burnt or even fuel flow, so the Cummins Insight software or the optional J1939 NIM interface would have no benefit to you.

As stated the fuel burnt would be between 1.9Lph to 7Lph depending upon load. However the genset has a internal electric fuel pump which run's all the time, so fuel flow though the system is around 80Lph at 6psi, this fuel is used to lubricate and cool the injection pump. so typically you could see around 78Lph to 72Lph returned to tank.

I don't know how accurate the maretron units are in calculating the difference between flow and returned fuel? but you can see there could be very small quantity deferences.

Anthony.

 

[burgundyben]

The existence of the electronic data is nothing to do with common rail; the rail is upstream of the injectors. The data could be equally be available for an engine with no CR and with unit injectors instead. It all depends on whether the injectors (whether fed by a rail or unitised pressure source) are electronically controlled. If they are, the data must be there. It doesn't follow necessarily that the data is in networked j1939 format of course.

I agree it ought not to be that hard but nevertheless I'm still trying to find out how to do it!

Ah yes, fair point, I use 'common rail' as a generic term for electronic diesels, where in fact its note quite the same thing.

Interestingly, a colleague is ex Ricardo, they have injectors with pressure sensors in the tip that monitor cylinder pressure and modify fuel delivery as its happening on each cycle. Very cool.

 

[RAI]

Would not a kWh meter on the generator output not be a fairly direct measure of the fuel burnt?

 

[jfm]

John,

The Kubota engine fitted to your gensets is a mechincial injected engine and the generator controller does not monitor fuel burnt or even fuel flow, so the Cummins Insight software or the optional J1939 NIM interface would have no benefit to you.

As stated the fuel burnt would be between 1.9Lph to 7Lph depending upon load. However the genset has a internal electric fuel pump which run's all the time, so fuel flow though the system is around 80Lph at 6psi, this fuel is used to lubricate and cool the injection pump. so typically you could see around 78Lph to 72Lph returned to tank.

I don't know how accurate the maretron units are in calculating the difference between flow and returned fuel? but you can see there could be very small quantity deferences.

Anthony.

Thanks Anthony. That definitively answers the question - if the Kubota's governor is mechanical then there exists no electronic data for fuel injected. "End of", so to speak.

The only sensible answer then is maretron then I think. These need a flow sensor installing on the flow and return to the diesel engine, then they connect easily to the boat's existing N2K network. These feel like they will be highly accurate. They are positive displacement, sort of gear pumps in reverse in fact, so they should measure the volume quite accurately, plus they measure the fuel temperature and adjust for it in the electronics. Maretron is generally top quality gear as you know. They claim +/- 0.25% accuracy so with a pair of them the accuracy should be comfortably better than 1%, which is pretty good. I'll investigate the mechanicals of where to install these and report back - as they use 1/4npt fittings for the fuel lines I think it will be quite easy to install them close to the gensets using a (say) 200mm long new fuel line to go between the genset and the flow sensors, and then connect the existing 1/4npt fuel lines to the other side of the flow sensors, and all done

 

[jfm]

Would not a kWh meter on the generator output not be a fairly direct measure of the fuel burnt?

Yes it would give a good approximation, though perhaps no more accurate than reading off the genset hours and multiplying by 4-5 litres as Mapis M suggested. Someone mentioned the kwh meter idea above. It is sort of "plan B", if I cannot get "plan A" to work

 

[rafiki_]

Thanks Anthony. That definitively answers the question - if the Kubota's governor is mechanical then there exists no electronic data for fuel injected. "End of", so to speak.

The only sensible answer then is maretron then I think. These need a flow sensor installing on the flow and return to the diesel engine, then they connect easily to the boat's existing N2K network. These feel like they will be highly accurate. They are positive displacement, sort of gear pumps in reverse in fact, so they should measure the volume quite accurately, plus they measure the fuel temperature and adjust for it in the electronics. Maretron is generally top quality gear as you know. They claim +/- 0.25% accuracy so with a pair of them the accuracy should be comfortably better than 1%, which is pretty good. I'll investigate the mechanicals of where to install these and report back - as they use 1/4npt fittings for the fuel lines I think it will be quite easy to install them close to the gensets using a (say) 200mm long new fuel line to go between the genset and the flow sensors, and then connect the existing 1/4npt fuel lines to the other side of the flow sensors, and all done

Apologies if I am being a bit slow here, but does the Martron measure supplied and returned fuel? Without this, the info is incomplete. As a previous post stated, wouldn't you be able to derive lph from kW per hour generated, given you don't need lph to many decimal places?

 

[Portofino]

Isn,t £ 2K better spent on fuel ?rarther than a a gadget to show on yet another " screen " 4-6 L/ h consumption of the geny- which the rest of us calculate in a man maths kinda way ,and seem to manage - with passage planning ? sleep well at night .Dreaming of the last read of Rod Heikels Med pilots ?
Topping up turns into an excuse to show off close 1/4 handling and send wife + kid (s) off foraging into the hinterland eg a Sunday Times or more " vino colapso "or meteo print out ( another version if in doubt ) .
There are more reasons to call in somewhere even -first time to top up when on " tour " - then to " plan it with 1% accurate base data " down to the last 10 % .
My Geny 0.8 to 1 L/h. With a 650 L tank - same No,s as JFM,s ( proportions ) but @ only 10 % volume wise ,
But same issues .
Most I have ever put in in one go is 500 L ie left / down to last " sloshing " 150 L .As soon as it drops below 1/2 I,am thinking / planning the next top up - not necessarily the same day but soon - usually put in 3-400 L ish ( out of 650 )
Includeds trips to Corsica -Elba etc
It's just that 99.9 % of Geny equipped med boats seem to manage without knowing down to the last 1% what the fuel burn is - man maths works .

 

[jfm]

Apologies if I am being a bit slow here, but does the Martron measure supplied and returned fuel? Without this, the info is incomplete. As a previous post stated, wouldn't you be able to derive lph from kW per hour generated, given you don't need lph to many decimal places?

Yep you install a maretron flow meter into the flow and return (ie you buy 2 of them per genset) and then configure them on the screen as flow/return respectively, and the maretron little black box squirts fuel burn rate data onto the n2k network. On the garmins I then set up/recognise the gensets as "engine 3" and "engine 4" and my total fuel burn is then computed by Garmin as the sum total of all 4 engines

Yes you can derive the figure. I just want it to display in a slick way on the screens and I want the fuel remaining figure to be right on my screens. I never set out to justify this, any more than any of us can justify a boat in the first place

 

[rafiki_]

Yep you install a maretron flow meter into the flow and return (ie you buy 2 of them per genset) and then configure them on the screen as flow/return respectively, and the maretron little black box squirts fuel burn rate data onto the n2k network. On the garmins I then set up/recognise the gensets as "engine 3" and "engine 4" and my total fuel burn is then computed by Garmin as the sum total of all 4 engines

Yes you can derive the figure. I just want it to display in a slick way on the screens and I want the fuel remaining figure to be right on my screens. I never set out to justify this, any more than any of us can justify a boat in the first place

Fair enough. I guess most of us have our degree of OCD-ness.