Lifetime engine data recording - Caterpillar (C32 ACERT)

[jfm]

re another topic,
any tests done with the new curved fins ?

Bart, the only trip I have done with curved fins is the 3 miles back to my berth from the shipyard. I haven't been to sea on the boat since - just been busy with other things for the last months or so. Normal service will be resumed soon! Hope all is well with you and the BA projects

 

[scubaman]

It's definitely possible, if you have some elementary programming skills, which I don't. There might be an off-the-shelf product too. The instantaneous load factor % is there all the time on my N2k network as well as the engines j1939 network but Garmin haven't created a function and display field on the MFD to show average and trip average. Presumably because no-one wants this at consumer level (except us on this forum!). N2K is an open standard so someone could do it. I'm about to install a Maretron USB interface to the N2k bus so that I can do more things using a laptop and Maretron's N2K view and analyser, and that might allow it. Not sure yet

This Maretron stuff is great. I have bought boxes and boxes of it, like I have $10k of it sitting on my shelf here, but I need a few weeks to install and set it up. USB interfaces (links below), genset fuel monitoring, trim tab indicators N2k version, circuit status indicators (stuff running or idle displays on helm screens with user programmable legends and colours), etc

http://www.maretron.com/products/usb100.php
http://www.maretron.com/products/N2KAnalyzer.php
http://www.maretron.com/products/N2KView.php

Pretty cool stuff! If I understood correctly, their VDR100 is made for the very purpose of recording n2k traffic on an usb drive. Their n2KExtractor-software and/or excel can be used for further analysis.

http://www.maretron.com/products/vdr100.php
http://www.maretron.com/products/N2KExtractor.php

Btw, the 'black box' -functionality might also come handy in tracking down an engine/genny/whatever problem, should one ever arise.

 

[jfm]

Yep. Once you get into Maretron there is no end to the thousands of dollars you want to give them. It's top quality gear. So far I've resisted temptation to buy VDR100 (!)

BTW, I use a pair of their j2k100s to bridge all the Cat engine data onto my n2k bus, so the Garmin screens can display engine and fuel data, and compute litres per mile and range (the garmins let you input the total tank litres into one device then they share it on their network so you can see range and trip fuel on every Garmin device and iThing)

 

[Lozzer]

Interesting info JFM. Have you actually compared this data with your fuel loading at the pump?

The reason I say this is that my actual consumption is around 10-15% less than the CAT gauges say. Although I have checked the engine life totals.

If I get time I will check my data tomorrow.

 

[jfm]

Interesting info JFM. Have you actually compared this data with your fuel loading at the pump?

The reason I say this is that my actual consumption is around 10-15% less than the CAT gauges say. Although I have checked the engine life totals.

If I get time I will check my data tomorrow.

heck that's interesting Lozzer. No, I've never checked. I do have written in my log book all fuel ever bought so I'll run a check, and also monitor it this season. Interesting comment though- thanks. The cat data doesn't actually measure fuel burned; it records what fuel the ECU has told the injectors to inject, but there could be calibration error there

 

[Latestarter1]

This post is very much apropos nothing, just for those who might be interested, and I'm not making any point but if anyone has any thoughts I'd be interested to hear. A couple of months ago a question came up on here about what did lifetime Cat engine data look like, and I said I'd post mine. So the photo below shows the monitor screen recently on one of my engines, showing the lifetime history page.

One of the useful bits of data is Cat's nice readout of lifetime fuel burn. They then quote rebuild times in terms of fuel burned, https://safety.cat.com/cda/files/891374/7/C32 Marine Engine (RNC) - Maintenance Intervals.pdf, which Latestarter will tell us is the proper way to do it. My engines are 1.75 seasons old and have burned 30k litres (each) out of 455k litres for a top end overhaul, and 760k litres for a full rebuild. So that's 24 and 42 years to go, respectively .

If I didn't know the boat better I'd think that's a lot of idle hours and a low average load factor, but fact is by the time you have pootled out of harbour to open water, and back, and faffed 15 mins while tying up needing engines running for hydraulic power to winches, you get to these high idle figures. And I never start the engines and let them warm up before leaving the berth - I drive off 2 minutes after starting up, at idle rpm of course because I'm in the harbour, and I get some load on the engines as soon as I'm in open water so as to warm them faster. Also, my average seems to accord with Cat's service requirements: eg, and ref the MAN discussion on here recently on Deleted User's boat, CAT recommend cleaning the aftercooler at 1000hrs or 91k litres of fuel. I've done 320 hours and 31k litres, which extrapolates very close to 1000hours and 91k litres

These engines are Cat C32 ACERT, and the fuel burned is obviously one engine only (Edit - I think my link above and data are for the previous "non ACERT" version of the engine - I cant find the ACERT data but I expect it wont be materially different as regards pleasure boat use. Second edit: this doc says 2m litres fuel for full rebuild, which is >100 years for me, so I can relax. I'm perplexed that Cat quote 760k litres in one doc and 2 million in another, for essentially the same engine... https://marine.cat.com/cda/files/962700/7/brochure - cat c32 acert marine engine.pdf)

JFM I will try to answer your question of accuracy of idle fuel reporting, any controls engineer would do a better job, but may be cure for insomnia.

Anyway here we go, in at the deep end!

Electronically controlled engines all work broadly along the same principals, be they common rail, EUI, HUEI, unit pump. I will use the term ECM, invented by Detroit Diesel and common term applied by all U.S diesel manufacturers. Bosch coined the term ECU so used by European engine builders.

Following key on ECM checks all engine perimeters which can be crank position, cam position, lube oil pressure etc etc and once it is satisfied allows injectors to start fueling. The ECM base calibration contains a massive look up table, all the inputs from the various sensors around the engine are distilled by the look up table which then outputs a PWM signal to individual injectors, telling them when to start fueling and the length of the PWM signal defined by the look up table tells the injector how much fuel to inject.

With zero throttle input idle speed settles to pre programmed level with ECM delivering a constant length PWM signal to injectors so life is simple or is it?

For a moment I will digress to automotive applications. If you take a modern electronically diesel with a manual transmission with say an idle speed 800 rpm, engage first gear on the level and lift the clutch WITHOUT applying any throttle, the vehicle will pull away cleanly with a dead throttle. With any decent diesel engine calibration you should be able to to do this in second and third gears. The ECM is seeing fall in idle speed and increases length of PWM signal and idle torque, enabling the engine to hold 800 rpm. In first gear ECM will deliver only a small increase in length of PWM signal, in third gear start length of signal is far greater in order to achieve 800 rpm.

Taking the forgoing into account in first gear small change in length of PWM signal required ECM assumes it is simply idling when actually pulling way. Pull away in third gear and the ECM will see the larger increase in length of PWM signal and assumes that engine is under load, however the second gear start COULD be recorded as idle or loaded.

Whether the ECM sees the second gear start as loaded or not, depends on the amount of hysteresis in the governor calibration.

Why is all this important to the fuel log? Every PWM signal relates to a defined amount of fuel, add up the sum of signal lengths and you have amount of fuel burned.

ECM calculates % engine load at any given engine speed by simply working out how much fuel AKA PWM signal length is required to hold a certain speed Vs the amount of extra fuel required to actually hold that given speed.

My guess is that at very light propeller demand at low engine speed hysteresis in governor makes life difficult for ECM to define if fuel is being burned under load or idling.

Hope this makes sense, anyway off to bed.

 

[Latestarter1]

heck that's interesting Lozzer. No, I've never checked. I do have written in my log book all fuel ever bought so I'll run a check, and also monitor it this season. Interesting comment though- thanks. The cat data doesn't actually measure fuel burned; it records what fuel the ECU has told the injectors to inject, but there could be calibration error there

Fuel log calibration is a nightmare, most manufactures shoot for over read of around 5%, however CAT calibrations used to do the maths in Imperial measure and then use factor to convert to Metric causing errors to creep in there as well.

Fuel injected is fuel burned.

 

[nimbusgb]

and there's my little tub using about 6 litres per hour with it's very old Perkins 4-108 running almost flat out and pushing me around at 7 knots.

I reckon the 30 years that engine has run, perhaps 7500 hours, it's burned about 10k litres!

 

[jrudge]

I have CAT c18s and hence get similar info. At idle it normally says fuel flow is 0.00 LPH which is probably a bit optimistic!

My average load is some 30% - so glad to see JFMs is similar!

 

[jfm]

Many thanks Latestarter. All very clear explanation, and very enlightening

The bit that isn't electronic is the "mechanical" part where the fuel is injected thru a nozzle. If at a certain speed and load the engine is running very nicely, and the look up table is nicely sending a PWM to the injector to squirt fuel for XXX nanoseconds, then life is good. But the actual physical volume of fuel that goes thru the nozzle of the injector when it squirts for XXX nanoseconds is an unknown - unknown by the ECM, I mean. Sure it can be supposedly known from a calibration lookup, by actual calibration (ie someone can physically measure that XXXns of injection event time equates to 0.0000xx litres of fuel passing thru the nozzle, but that calibration is prone to error because there are tolerances in fuel mechanical properties and nozzle dimensions etc. That must account for the 5% or 15% errors that you and Lozzer mention above, and presumably the manufacturers make sure the error over states the litres burned by rigging the calibration table accordingly. Interesting stuff. I'll see if I can measure it this season as Lozzer has done. A 5% error in say 30 or 40k litres of fuel this season should be very easy to measure

As regards the ECM knowing whether the engine is in idle or pushing the boat at idle rpm, I never thought about that. Actually I assumed that if it saw 650rpm it assumed idle. I hadn't realised that the determination of idle (or not) was taken from the length of the injection event, but I see your point. Easy to answer this one: I'll run my engines in gear at idle rpm and look to see whether the lifetime idle hours counter notches up.

Incidentally, on jrudge engines and mine the PWM you refer to, and the lookup table, create multiple injection events per firing stroke!

Thanks for the info

 

[MapisM]

As regards the ECM knowing whether the engine is in idle or pushing the boat at idle rpm, I never thought about that. Actually I assumed that if it saw 650rpm it assumed idle.

Yup, that's what I would have thought either, but I can see LS point on that.

Otoh, I'm skeptic that the difference between in/out of gear is as low as LS said:

My guess is that at very light propeller demand at low engine speed hysteresis in governor makes life difficult for ECM to define if fuel is being burned under load or idling.

I mean, of course I agree that the prop demand at idle RPM is just a tiny fraction of what it can reach at WOT.
But we are still talking of constantly pushing aside some 60T of water, at 6 knots or so.
That's bound to take much more power compared to just keep the iron bits inside the engine moving, I reckon...
But I'm also just guessing, TBH.

 

[jfm]

Yup. I'll try to get some data @weekend, on difference between pushing boat at idle rpm and just pushing the bits of iron around. I'm finding this discussion, and Latestarter's great info above, quite enlightening.

I'd like to know the 5% or 15% figure or whatever it is. If it is a constant value, I can program the size of my fuel tanks in the electronics to be 5% bigger than they actually are, to get true range.

It also makes me want to fit true flow measurement http://www.maretron.com/products/ffm100.php to the main engines, to get a better number. I actually have these FFM100s on my shelf ready to fit to gensets when I get time

 

[Deleted User YDKXO]

Very nice data and not available on MAN's AFAIK.

Its a shame that isn't it? The instantaneous fuel burn is there so you would think that the total fuel burn wouldn't be too difficult to store. I guess like my MMDS displays, you get a range figure displayed as well as the instantaneous fuel burn. Have you ever found out whether that range figure is accurate in reality or not?

 

[Deleted User YDKXO]

Well done, jfm, for posting this info. As we have discussed before, total fuel burn is a much better indication of remaining engine life than working hours. I hope that one day we can all talk in terms of total fuel burn rather than engine hours when assessing a used boat

Maybe Latestarter can answer this one. Do you know whether those figures of 455k litres for a top end rebuild and 760k litres for a full rebuild are typical of other marine engines with similar power outputs? It would be interesting to compare these figures to MAN and MTU engines with similar power ratings, if the information was publically available (which I guess it isn't)

 

[jfm]

As we have discussed before, total fuel burn is a much better indication of remaining engine life than working hours. I hope that one day we can all talk in terms of total fuel burn rather than engine hours when assessing a used boat

Yep, but alas hours is a common yardstick covering all power levels, like miles/km on a car, so it will remain the "headline" figure in used boats. If you quote total fuel burn in a boat ad you need a different measuring stick for every engine/bhp. Too much for Joe Public's brain (me included!) to deal with.

But yes indeed when you go beyond the headline hours figure it is a good data item in assessing a boat. Alas it seems from this discussion only Cat will play, not Volvo, MTU or MAN

 

[jfm]

Maybe Latestarter can answer this one. Do you know whether those figures of 455k litres for a top end rebuild and 760k litres for a full rebuild are typical of other marine engines with similar power outputs? It would be interesting to compare these figures to MAN and MTU engines with similar power ratings, if the information was publically available (which I guess it isn't)

Definitely not Mike. MTU have much shorter published intervals than Cat, AOTBE. I mean like ~50% of Cat numbers.

 

[MYAG]

Its a shame that isn't it? The instantaneous fuel burn is there so you would think that the total fuel burn wouldn't be too difficult to store. I guess like my MMDS displays, you get a range figure displayed as well as the instantaneous fuel burn. Have you ever found out whether that range figure is accurate in reality or not?

Mike I don't get range figures at all on my displays, just instantaneous fuel burn but I don't know how that figure is derived, tabled injector squirt time or fuel flow meter or some other way. Also not sure how my engine load % is calculated either, is it the CAT way described by LS1 or another way?

So for me it's the old fashioned amount I squirt back into the tank job and keeping a detailed fuel burn log. So I only have real figures but nothing to compare to. I do like the idea of the FFM100 though.

 

[jfm]

Mike I don't get range figures at all on my displays, just instantaneous fuel burn but I don't know how that figure is derived, tabled injector squirt time or fuel flow meter or some other way. Also not sure how my engine load % is calculated either, is it the CAT way described by LS1 or another way?

So for me it's the old fashioned amount I squirt back into the tank job and keeping a detailed fuel burn log. So I only have real figures but nothing to compare to. I do like the idea of the FFM100 though.

The thing about squirt back into tank is you don't know the genset burn. you can estimate it of course, but your genset hours are very high and the lph varies substnatially with electrical load, so an estimate is going to be a bit dodgy. This is why I have got a couple of FFM100s (and 4xflow sensors) here on shelf ready to install on the gensets. But sounds like I need a repeat set for the main engines haha! Petem, open that Maretron dealership quickly and give us a forum discount!

 

[Portofino]

Re Man,s I think the load % we see could be a% of the PWS time ,the injector open time -actual over max .

which if I understand kinda relates to vol -but is open to errors cos of difficulty calculating actual fuel vol ,due to bore dia issues ( temp variations in the metals ) ,pressure,density and prob other unknown variables .Hence Lozzers observation .
At Idle my 2876's around 600 rpm 8 L /hr " turning iron " only
Click in gear rpm stays at 600 but compsumption goes to 12 L/h -suggesting its squirting more in to keep Idle AND now spin gearbox -shaft -high pitched props -gives 8 -9 knots
Also % load goes up -sorry from memory I can,t remember the figure guess low teens ish ? .
So what LS says makes perfect sence for me
My torque curve is not a curve it looks like Table Mountain - 2600 Nm from 1000 rpm flat to 2000 rpm ten its drops bit to WOT 2150 .
Regarding actual fuel used ,I just use the L/H function on the MMDS as a relative indicator say from 1800 rpm to
1600 rpm 32 Knot to 26 knots .80 drops to 50 ish L/H per side

But this year for the first time we might try slow cruising D speed - 7-800rpm or low planning rpm 1100 or something ?
Interesting stuff thanks LS

 

[jfm]

At Idle my 2876's around 600 rpm 8 L /hr " turning iron " only

8 L/hr for both engines in total, or each engine?