Fuel tank Level puzzle

[catmandoo]

I have a Tank Tender fuel tank level indicator which works on the principal that if you pressurise a small bore plastic tube dipping into the fuel tank with air until bubbles come out of the end you can measure tank level by means of the air pressure required to keep the fuel out of the tube . All very well and quite accurate . ( Also expensive instrument )

However I have a problem . I get a different reading when the boat is moving about 1 inch of level !!! . It recovers when the boat is stopped . This gives me a problem on fuel consumption rate during long journeys

The 40 gallon tank is about 5 feet long , flat topped,semicircular , radius 10 inches , with 2 baffles . It runs fore and aft .

Fuel is extracted at one end and recycled back to the other end from the diesel engine injectors etc

Three possibilities

1 When the boat is moving it tilts towards the return end . ie the stern dips .

2 The return end gets hotter than the supply end due to heat from the diesel engine injectors and with no mixing due to the baffles and poor convection currents the difference in density results in a depression at the level gauge end ( U tube effect )

3 a combination of both

Any comments as as to cause and suggestions on how I can get an accurate reading when moving without relocating the position of the tubing ???

Any instrument engineers out there ?????

 

[pvb]

[ QUOTE ]
I get a different reading when the boat is moving about 1 inch of level !!! .

[/ QUOTE ]Sadly, I've always been quite pleased if a boat fuel gauge works well enough for me to know whether the tank's full, half-full or nearly empty. /forums/images/graemlins/wink.gif

 

[chuzzlewit]

I don't think 2 is good physics.

 

[catmandoo]

yes but with 24 hours or more of continuous motoring particularly into a head wind one gets a little sensitive regarding fuel consumption and what's left ?

 

[Guest]

To be accurate the tube sensor would have to be dead centre of the tank to negate level changes with aspect / motion of boat. You also get surges in a tank only reduced by baffles - not stopped.

Combination of above will lead to poor gauging.

I assume that your gauging is taken of from the aft end of tank ? Then it could be subject to serious buffeting of tank contents causing momentary increases of pressure due to momentum of the fluid in movement in the tank.

 

[catmandoo]

[ QUOTE ]
I don't think 2 is good physics.

[/ QUOTE ]

Why not

Fact

a Tank is in middle of of one hull of catamaran low down in bilges hence at start same temperature as sea
b 10 minutes after engine start and normal speed obtained there is no change in level
c one hour later level anomaly appears
d There is definitely a temperature difference engine oil temperature at one end and sea water temperature at the other
e tank is baffled but good mixing is not obtained hence hot stuff accumulates at hot end and cannot flow quickly to cold end . process is continuous

f cold fuel is heavier and hot lighter so there must be a difference in levels
g after boat comes to rest and a reasonable time level, expected fuel level is obtained


I must admit I have only noticed this on short runs of 3 to 4 hours . On a trip of say 10 hours or more I suppose the whole tank gets hot and the problem disappears /forums/images/graemlins/smirk.gif

 

[catmandoo]

Sensor is in forward end of tank approximately in middle of boat . Ideal location Is middle of tank but not practical due to floor directly on top . Current location is on hand hole cover which is recessed . I was thinking of some sort of compensator on instrument . No problem with surges but concerned that holes in tank baffles may not be big enough to get good convection or even reasonable level equalisation

May be I should cool the fuel return or

Must check with Tank Tender manufacturer in Texas !!

 

[Guest]

............ lets say your diesel (just figures only - not specs !) is .8500 kg/ltr at 15C ... and you raise its temp 20C ... to 35C ........ the dens will only change to 0.8362 kg/ltr at 35C.

so assuming you have 100ltrs of fuel at 15C ... = 85kg ......... that same 85kg will now be 101.65 ltrs .... but that is assuming that the whole tank has increased in temp. But as you say it is only one end ........ the likely increase then is probably less than 1/2 a litre - probably not even registered on your gauge ........

(before anyone catches the deliberate ommission - I have used weight in vacuum and ignored buoyancy for air ... 0.0011 kg/ltr correction ..... and of course VCF etc. etc for simplicity)

/forums/images/graemlins/cool.gif

 

[VicS]

[ QUOTE ]
............ lets say your diesel

[/ QUOTE ] Not only is the change in volume so small as to be insignificant but the physics involved in this type of instrument would not produce a change in reading even if the expansion was significant.

They work on a pressure principle and since pressure is proportional to the product of the depth (at which the reading is taken) and the density the change in density with temperature and the perceived rise in level at the "hot" end of the tank will conspire to cancel each other out precisely.

Elementary physics!

 

[Guest]

I forgot he said a Pressure sensing gauge ... but even so - I did say in my little example calc - that the difference was probably not registered on his gauge.

 

[catmandoo]

[ QUOTE ]
[ QUOTE ]
............
They work on a pressure principle and since pressure is proportional to the product of the depth (at which the reading is taken) and the density the change in density with temperature and the perceived rise in level at the "hot" end of the tank will conspire to cancel each other out precisely.

Elementary physics!

[/ QUOTE ]

Sorry dont agree with you regarding your last two paragraphs .

I accept that if the data quoted regarding thermal expansion of diesel over 65degC (which I did not have at the time ) is as quoted then expansion is negligeable .

However the measurement is from the cold end and at that position there would be no change in density and the analogue reading on the air gauge would correspond to the manufacturers calibration at 15 or 20 Degrees C or whatever

I cant even get a coherent answer from the manufacturer so I accept the problem is one of trim during motion and so must take fuel consumption readings at the same state ie during motion or at the beginning and end of each journey /forums/images/graemlins/laugh.gif

 

[VicS]

[ QUOTE ]
Sorry dont agree with you

[/ QUOTE ] I was assuming that the reading was at the warm end! So if the gauge is at the cool end as you say the density remains the same.

If you know the reading is high by an inch when underway can't you just allow for that. You've got 1" less fuel than the gauge says. Anyway it would be a very unsound practice to run the fuel so low that an inch became critical!

 

[rogerthebodger]

As the draw off and return are at opposite sides of the tank with (question) baffles between. Could it be that the baffles are restricting the flow from one side of the tank to the other thus cresting a difference in levels between the draw off side and the return side when the engine is running at speed but would equilise out when engine is not running or running slowly.

The temperature difference could also indicate this restricted flow past the baffles

 

[DavidMcMullan]

Is the fuel tank air vent clear? If it was partially blocked it could have an effect on the pressure in the tank.

 

[catmandoo]

This did occur to me .
Unfortunately I can't check the baffles without draining the tank . It is supposed to have two baffles equally spaced creating 3 compartments

However it is a stainless steel tank which according to the supplier last year is identical to the old Mild steel tank baffles and all that I look out last year with much huffing and puffing .
I have just now looked at a sketch I made of the old baffles and note that there were triangular cut outs at the top corners about 1 .5 to 2 inch wide and one centre line 1.5 to 2 inch triangular cut at the bottom . This arrangement might restrict convection flow particularly if the level in the tank was below the top two cut outs leaving the bottom one to balance each chamber and also more likely to have cold diesel flowing through it. which means that most of the hot diesel was contained in the aft compartment

 

[catmandoo]

Unfortunately I have to calibrate the reading by dividing the tank into inch slices and draw a graph of inches versus litres so as it gets more empty there is less volume available as the bottom 5 inches is semicircular the top 5 inches being tangential.

Gets more complicated if I assume the tank or the fuel level slanted . This is what the manufacturer suggested .

practically speaking I keep about 80 litres in reserve in jerry cans and can top up on the run . I never let the tank get less than half full but one of these days I might need to
take advantage of the lower levels . ( Eg how many times have people run their car on the red line trusting in the reserve ) Fuel consumption is based on engine hours so fine if I am doing 5 knots or more but not so if I unexpectedly have 22 knots of wind on the nose and can only steam at 1 knot as I did in 2006 down the Messina straight against tide and wind

 

[catmandoo]

Yes fortunatley .

I can appreciate the advantage of returning the fuel to the opposite end of the draw off as it would not be a good idea to recirculate hotter fuel back to the engine - don't want vapourisation in the filters in the engine compartment or injectors . Already cooling the oil due to hot Med weather . no space to cool the return fuel