Fuel gauge calibration

[Breakin Away]

For the last couple of seasons I have kept some logs of fuel gauge level vs. engine hours. I also keep track of exactly how much fuel we put into the boat, and top off completely once on the hard at the end of the season. My consumption has ranged from 0.55-0.60 gal/hr (depending on what date range I select). I typically cruise at 2200 RPM, though sometimes motorsail at a lower RPM.

One of the interesting results of all of this is a fuel gauge calibration chart. Once I've topped off and calculated average consumption, I can use that to back-calculate an estimate of how much fuel was in the tank at each fuel gauge reading. The end result is a calibration which is pretty much linear, but clipped at both ends. I recorded the reading in "eighths", in other words, 0=empty, 4=1/2 tank, 8=full, and 9=pegged above full. The points extrapolate to about 2.5 gallons in the tank if I ever let the gauge get all the way to "E", 10 gallons at 1/2, 17 gallons in at "F", and 18-25 gallons when pegged above F. (I assume here that Catalina's spec of 25 gallons for the MkII boats is correct, and includes all the volume in the vent and fill hoses when I top off completely.

I'm not sure if any of you have done something similar, so I'm offering it up for your comments:

[KWKloeber]

BA

Very interesting, couldn't get more linear than that data shows!
Odd though w/ the tank shape - I've seen arguments made on here that the FG isn't linear.

[Ron Hill]

Breakin : If you've read any of the posts or articles on our C34 fuel tank, I find your research interesting.

Our tanks are asymmetrical with the widest portion of the tank at the top and the narrowest portion at the bottom - wider in the front and narrower in the rear.  So when the float on the fuel gage arm is indicating that the tank is "half full"; in actuality the tank (gallon wise) is really about @ 1/3 full (8.4 gal not 12.5 gal) remaining.
I wouldn't term that as linear - as far as the gage reading against actual volume.  The MKI data plate on the top of the tank is stamped 25 Gallons so I assume it is correct.

When I first got my C34 and until the time I pulled and cleaned that tank, I only somewhat believed in that fuel gage. My 30 years of military flying experience taught me to always use fuel consumption against engine time (except for the 20 minute Warning Light - that light I BELIEVED!!).  Then I pulled the tank for cleaning and saw that the tank was an asymmetrical design with a float ("powered") switch. That confirmed to pay little attention to using the fuel gage and went to the straight fuel consumption calculated against engine running time.  I stuck with the .5 gal/hr (M25XP) .55 gal/hr (M25XPB) and seldom had run over when topping off during a refill because my 1st Mate called off the Gallons and Tenths to me - if I couldn't see the fuel pump.  I usually ran at 80% power.

A few thoughts

[Breakin Away]

I too was very surprised how linear the data looked. I expected a more parabolic shape, curved downward, to reflect the tank shape. But I did nothing to "fudge" the data to look linear. It is what it is.

And as you can tell from the points plotted, I've never run below 1/8 of a tank. I don't like seeing the needle bounce around (and the fuel slosh around) so much when it gets below 1/2 tank - though I do want to get the tank as empty as possible to avoid back-mixing of stale fuel with fresh. The one time I went down to 1/8 tank to get the data point I had a full jerry can (and extra fuel filters!) onboard.

[Jon W]

For what it's worth- I removed my 1987 fuel tank last week and brought it home to add access plates, clean the inside, and check condition for potential leaks. I was curious about this discussion so I added water until it started to leak out of the fill tube. Water started leaking at slightly more than 18 gallons. Seems to align with Breaking Aways graph that 25 gallons is the total volume of the tank plus some in the fill hose. Makes the hose clamps on the fill and vent hoses more important.

[Ron Hill]

Jon : Was the hose between the vent on the tank and the vent fitting on the transom - in a straight line or was the transom vent slightly higher or lower (when the boat was level)??

A thought

[Jon W]

The tank is in my backyard sitting level on the ground. No hoses. Filled it with water until it began to leak out the fill pipe into a bucket.

Before I reinstall in the boat, I plan to cap the fill pipe and fill the tank to the top and see how many gallons the tank alone holds.

[Noah]

Jon- i currently have 22 gallons of fuel (measured) in my new installed tank, it is not overfilled and believe it will take a few more gallons if I was so inclined.

[Breakin Away]

For the last couple of seasons I have kept some logs of fuel gauge level vs. engine hours. I also keep track of exactly how much fuel we put into the boat, and top off completely once on the hard at the end of the season. My consumption has ranged from 0.55-0.60 gal/hr (depending on what date range I select). I typically cruise at 2200 RPM, though sometimes motorsail at a lower RPM.

One of the interesting results of all of this is a fuel gauge calibration chart. Once I've topped off and calculated average consumption, I can use that to back-calculate an estimate of how much fuel was in the tank at each fuel gauge reading. The end result is a calibration which is pretty much linear, but clipped at both ends. I recorded the reading in "eighths", in other words, 0=empty, 4=1/2 tank, 8=full, and 9=pegged above full. The points extrapolate to about 2.5 gallons in the tank if I ever let the gauge get all the way to "E", 10 gallons at 1/2, 17 gallons in at "F", and 18-25 gallons when pegged above F. (I assume here that Catalina's spec of 25 gallons for the MkII boats is correct, and includes all the volume in the vent and fill hoses when I top off completely.

I'm not sure if any of you have done something similar, so I'm offering it up for your comments:

I've updated my calibration from data collected so far in 2020, using similar methodology to two years ago. My average consumption rate has been very consistent at 0.55-0.57 gph each year that I've had the boat. I use this consumption rate to interpolate some points on the chart. This year I spent the first couple of months "running on empty" since I was concerned about the age of my fuel, and trying to burn off as much fuel as possible (allowing it to slosh around to disperse any bad stuff into the Racor bowl), then adding a few more gallons and burning it off again. As a result of this, I have a lot more data points close to empty this time, which is where I want to ensure I have an accurate calibration (for obvious reasons). After a couple months of running close to empty, I added 15 gallons or so which pegged the gauge, but the calibration shows that there were only about 17 gallons in the tank at the time. The gauge is pegged from about 17-25 gallons, which means I can't really tell how much is in there (aside from measuring operating hours), but of course, there's no danger of running out of fuel when it's at that end of the calibration.

The orange line in the plot was drawn by eye to encompass the lowest locus of points, rather than doing a least squares fit. I prefer to have the least optimistic estimate of fuel content, again to ensure I don't ever over-estimate how much running time I have left before running out of fuel.

I think the scatter in the points is mostly caused by differing attitude of the boat under differing conditions. If the boat is moving, the gauge bounces due to changing attitude and sloshing fuel. If it is at rest, the gauge is steady, but also can get a little stuck, making it somewhat inaccurate.

This winter I'll pump out the fuel and store it in jerry cans like I did last winter. My fuel has been cleaner than ever this year, and I think pumping it out for the winter and filtering it back in prior to launch has made a big difference for the better. Next year at spring commissioning I'll go ahead and top it off, 5 gallons at a time, and get a good calibration of the gauge over the whole range. Of course, the boat will be on the hard then, and probably will be slightly bow-high, so that calibration will not be perfect either.

I know that my calibration may not matter to the rest of you (except maybe other MkII owners with M35B motor), but I'm offering it to you in case you want to copy my methodology on your own tank.

[Jon W]

A photo is worth a thousand words. I keep a printed copy of the attached PDF on the boat as a quick reference for me or others.

[mark_53]

Your graph does indicate you have more fuel in the tank below 1/2 gauge reading than above but not a significant difference unless your cutting it real close on a long trip with half a tank and not filling up.   Not likely. I figure as long as I see the needle move I have some fuel but usually won't go below 1/4 gauge reading.

[Breakin Away]

Your graph does indicate you have more fuel in the tank below 1/2 gauge reading than above...

I'm not sure I understand. There's always noise in the data for the reasons that I explained. But on average, it's a very safe bet that if the gauge reads over 1/2 there's more fuel than if it says below 1/2.

[KWKloeber]

25-gal capacity would clearly be the "tank" (which CTY says don't overfill) and not include any volume of hoses - fill or vent.

Ass/u/ing that the data is all correct, the plot would tell me that IIWMB I'd investigate the total fuel level system.  Either the gauge or the sender or both is suspect.  I've calibrated tons of different type monitoring equipment over 40+ years and one rule remains true.  If zero, 50% (in this case inches, not gals,) and100% isn't within the acceptable variance, I throw out the calibration and investigate the cause because it could have a shift, or a clip, or be out of scale, or be damaged that it needs to be serviced/replaced.

[mark_53]

Your graph does indicate you have more fuel in the tank below 1/2 gauge reading than above...

I'm not sure I understand. There's always noise in the data for the reasons that I explained. But on average, it's a very safe bet that if the gauge reads over 1/2 there's more fuel than if it says below 1/2.

The data does look odd unless your tank has a different shape then Ron's.  The gauge is only reading ohms, so it's only an approximation of fuel capacity based on the float level.  To get actual quantity remaining you'd have to measure displacement.

[Breakin Away]

My tank is almost certainly a different shape. It is a MkII (wider aft berth area above the waterline, so more available space for a wider tank). I previously posted a picture of the tank, and it is mounted several inches higher, so it may not be as tall. So it's a pretty good chance that it is a different shape than the MkI boats. My tank is fabricated from brown plastic (probably some variant polypropylene), which was being used by Catalina at the time, and may therefore be a different shape.

I showed pictures of all of this last winter on a different thread. https://c34.org/bbs/index.php/topic,10409.msg81154.html#msg81154  I will never remove this tank unless there is a severe reason to, because you can see from the picture that there is no way to get to the fitting on top of the tank without tearing out the locker floor directly above it. There is no reachable space above the tank to get there with tools.

There is zero reason to believe that my fuel gauge is malfunctioning. My data shows normal variability, which is to be expected from the wide variety of conditions in which I took the data. Sometimes in the slip, sometimes underway, sometimes in smooth water, sometimes in a little chop with the gauge bouncing. Sometimes right after adding 5 gallons of fuel (so float being pushed up buoyance), sometimes after consuming fuel (so float being pulled down by gravity). In my experience, all mechanical and electro-mechanical gauges have a little bit of hysteresis due to frictional effects, so how the read going up is a little different from how they read going down.

Also, the 2018 data was taken after fully topping off my tank and assuming that I was starting with 25 gallons. The 2020 data was taken after gravity draining my tank through the Racor feed hose, which I was holding roughly level with the bottom of the tank (since I didn't want to dry out my recirc loop which feeds the injectors), than adding a little fuel at a time -- but I was only adding 5 gallons at a time and burning off the older fuel by "running on empty". These differences could account for the differences in data between the two years, especially near the "full" range. Note that I haven't had more than 17 gallons of fuel in my tank this year. Next year I'll measure in 25 gallons to cover the full range of the gauge. I'm not going to top off my tank this late in the season, since I want to run it down and store it in a couple jerry cans this winter, like I did (for the first time) last winter. It seems to have really helped my fuel quality this year - absolutely no residue in my Racor bowl the last few months.

IMO, these sources of variability are to be expected in my fuel gauges, and cannot be directly compared to checking fuel gauges sitting still on land-based equipment like tanks at a chemical plant (which I have a fair amount of experience with also).