Tank level measurement inside a moving ship


Thread Starter


I am working as a instrumentation & control engineer in a private shipping company.
I want to measure the liquid level of a tank kept inside a big boat which is subjected to severe oscillations (boat dynamics) while the boat is in movement inside the sea.The tank level has to be kept at 50 % exact for ideal plant operation. Actually we are using a level transmitter fixed at the bottom of the tank which will give an ouput of 4 to 20 ma when the tank level is between 0 to 100%.This output of the LT is used to drive a variable speed drive of a motor which will pump more fluid so as to keep the level at desired set value.
But assuming the level is at 50%, the corresponding ma output should be 12 ma, which is alright when the boat is at standstill. When the boat is put in operation inside the sea, the waves causes oscillations in the boat and in turn causes the tank liquid to oscillate up and down , even though the mass of the liquid remains the same,which causes the LT output to osillate between 10 to 14 ma which results in the unwanted hunting nof the motor speed and this prevents the system in attaining a steady state which is undesirable.
Can anyone of you, suggest a suitable method or algorithm which helps in determining the tank level even if the ship barge oscillates.


[email protected]
[email protected]
[email protected]
This may be way off, but what if you were to place a sensor near each of the four sides of the tank? Then only run the fill motor if all four sensors trigger together. You'd have to time-delay the motor input to make allowances for any vertical movement, but...

Eesh.. This sounds like something you'd find in a textbook. I'll wish you luck in finding a solution!
Seems like you need to use mutiple measurements averaged together with the measurements located to read the corresponding high and low points of the sloshing liquid.

Jake Thompson

Just a thought, can you try using an ultrasonic level transmitter monuted at the top of the tank? We use these in waste water application and that tend to dampen out some of the oscillations in the water. Good luck
I guess the level is used as added ship ballast.

Our ship tank level gauges use an optional slosh tube to prevent unwanted level turbulence. We use a shield for both point level and continuous – 4-20 or resistive. Typically it is vented near the top and bottom as opposed to continuous vent holes. This may only slow down the slosh - putting less strain on the pumps. A inlet / outlet at each end will keep the pumps working less.

But – if you use the point level system with a latching relay it may ease the pumps even more. For example – (this can be reversed) we would set one float at the required high level. So – that activates the pump either on or off. Then – the fluid can move all over the place – float goes up and down – but nothing happens until liquid hits the other float - factory set a level you feel the level would be maintained at.

The relay is din rail with a low excite VDC to the switch with an 8 amp relay. The relay can control an alarm or contactor.

Check out the continuous, point and latching relay at www.almegcontrols.com

Please email direct form our web for more details.

Thanks Bob Hogg

PS - we can have continuous and point level in the same transducer.
The waves (and hence the disturbances on your control loop) will occur
in a narrow band of frequencies. You may simply be able to de-tune your
controller not to respond to these frequencies. Unfortunately, this will
result in a more sluggish controller. If the tank that you are
controlling the level in is large (relative to the flow rate in/out)
then this may be ok. However, this de-tuned controller could be too slow
to cope with other dynamics in the system, in which case you may need
something more complex. Kalman Filters are sometimes used in ship
steering mechanisms to reduce the effect of wave disturbances upon the

I hope this helps.

Andy [email protected]___

Advanced Control Technology Club, Industrial Systems and Control Ltd.,
50 George Street, Glasgow, G1 1QE Tel: (+44) 0141 553 1111
http://www.isc-ltd.com/actclub.html Fax: (+44) 0141 553 1232

Roderick K.Duet

Another, option is to use a averaging filter in the PLC program that is triggered by a sampling timer. The timer preset then can be adjusted to a setpoint that provides a desirable response interval. I use such a filter in batching where scales recieve a lot of plant harmonic vibration.It also works well with flow meters controlling inverter speeds for flow rate set points. If you need a sample,
E-mail me [email protected].
1) Try a hydrostatic sensor mounted in the bottom of the tank.

2) Introduce a dead band to damp out the oscillations.

3) Try averaging the signal over a period
I believe that the most reliable way to measure the water level in this scenario is to rely on the fact that in order for the water to fill the tank to a certan percentage - regardless of the turbulences- the AIR that is in that tank has to have a certain volume ( that in our case is 50% of the tank volume).

By using a valve and monitoring the volume of the air that flows in or out of the tank you can have a decent estimate of the water volume, regardless of the water movements.