Six popular questions about fuel tank calibration.

1. What is the difference between calibrating the sensor and calibrating the fuel tank ?

Many people confuse these concepts. Sensor calibration is done after the sensor is trimmed and consists of fully immersing the sensor in fuel to “teach” the new length. Fuel tank calibration is needed to convert the fuel level into volume

2. Why do I need to calibrate the tank?

All existing gauges measure fuel level, but not fuel volume. To convert fuel height to volume you need to know the shape of the tank. The tank calibration procedure is the determination of the dependence of the fuel volume in the tank on height, in essence it is the determination of the geometric shape of the tank.

Therefore, do not look for other models of fuel level gauges in an attempt to avoid this procedure as it relates to the fuel tank and not the sensor. Tank calibration is usually the filling of the tank in measured portions from empty to full and is a relatively time-consuming procedure.

3. Why do I need to calibrate the sensor? And how do you do it?

If the fuel level sensor electrode is trimmed or lengthened, its “default” calibration becomes incorrect, which may result in a dead zone “at the bottom” or “at the top” of the sensor. And to calibrate the fuel level sensor does not require a lot of fuel – a piece of plastic pipe closed on one side, in which 1-2 liters of fuel is poured. The length of the tube must be longer than the length of the sensor to be calibrated. (Figure 1).

This operation is not always required. For example, eurosens Dominator fuel level sensors do not require calibration if the fuel tank will be calibrated using batch refills. Because the first and last tank calibration points are also the calibration points of the sensor. In addition, there is a “dry” calibration option.

If the calculated tank calibration is to be used (discussed below), then calibration of the fuel level sensor in the pipe is necessary.

4. tank calibration by filling the tank in portions from empty to full is very time consuming. Is there any way to avoid it?

There are a few instances where this is possible.

  • We will be satisfied with the level, not the volume of fuel. For example, we need to know if it’s time to fill up the fuel tank. In this case, the fuel level information will suffice.
  • We have a height to volume conversion chart for this tank. We can use this data to calculate the tank calibration with the eurosens configurator. These tables are often seen by owners of fuel storage facilities and gasoline tankers.

Therefore, if you work with such objects, the first thing you should do is to check with the customer about the availability of tare tables.

  • We have such a tank shape that we can calculate the fuel volume for any altitude using a formula.

For example, a perfectly rectangular tank (parallelepiped with no rounding) or a vertical cylinder (Figure 2).

Fig. 2. Ideal tank shapes for which calibration is not required.

The results of tank calibration can be plotted on a graph (sensor output signal – fuel volume in the tank). For the tanks shown in Fig. 2, the graph will be a straight line (Fig. 3). Thus, it makes no sense to tare the tank by filling portions, because on the graph all these points will fall on this line, and we can count them without filling portions.

Fig. 3. There is no point in taring the entire tank as all points fall on the same straight line

Other popular tank shapes are the horizontal cylinder (Figure 4) and the tank (Figure 5).

Fig. 4. Horizontal cylinder
Fig. 5. Tank shape - tank

The calibration table for these tanks can be calculated with the eurosens configurator. All you need to do is calibrate the sensor to the correct length using a pipe (see step 3) and enter the tank dimensions (Fig. 6). For a horizontal cylinder, we introduce the tank dimension E=0.

Fig. 6. Tank calibration table calculator

5. We have a calibration chart from a very similar tank. Can it be used?

For example, we install fuel level sensors on 20 identical cars with identical tanks. Immediately our first wish is to calibrate one fuel tank with one sensor, so that we can use this data on the next cars.

But that would be a mistake, because the fuel tanks have slightly different volumes, and the gauges between each other also have some difference in readings and are trimmed slightly differently. We will end up with improper refills, especially when refueling to a full tank.

But we can use the tool to compare and process the tare tables from the eurosens configurator (Fig. 7).

Fig. 7. Converting the tank calibration table to a similar tank

To get the tank calibration tables for our 20 cars we need the following:

  1. Determine the exact tank volume of each machine (by filling to full).
  2. Calibrate all sensors in the pipe (see point 1).
  3. Edit the sensor calibration curve by stretching and compressing along two axes so that the sensor values “full”-“empty” coincide with the calibration values according to item 2, and the tank volume coincides with the measured one.
  4. Then save the resulting table to a file and load it into the sensor.

This method works ONLY for tanks of identical shape but slightly different volumes.

6. How do I select the correct number of portions to calibrate the tank?

Many people think that the more portions used to calibrate the tank the more accurate the result. This is true in most cases – for example, for tanks whose cross-section changes continuously from bottom to top (Fig. 8).

Fig. 8. The tank is asymmetrical, irregularly shaped. The more tare points the better.

In this case, if we draw a graph of volume dependence on height and overlay on it the results of calibration by 8 points, it can look like in Fig. 9. Red line – tank calibration on the basis of calibration table by 8 points. It is noticeable that the actual shape of the tank (black line) is not reliably conveyed.

Fig. 9. Determination of the shape of an asymmetrical tank by 8-point taring (inaccurate).

When taring with 16 portions, the difference with the previous result is significant (Fig. 10). The shape of the tank communicates much better.

Рис. 10. Determining the shape of an asymmetrical tank by taring by 16 points (not bad anymore).

If we are taring a real rectangular tank (Fig. 11), it usually has roundings at the bottom and top, and is symmetrical in the middle part with almost vertical walls.

Fig. 11. Widespread fuel tank

It is advisable to tare such tanks with variable dose volume, in the upper and lower part of the portion volumes to make less, and in the middle part it is possible to make them more. This way we do not lose in accuracy (as you can see in Fig. 12 we can recreate the graph from these points), but we increase the speed of calibration.

Fig. 12. This savings on portions will have almost no effect on accuracy

Therefore, it is difficult to answer the question of the best number of portions to calibrate the fuel tank. It depends on the shape of the tank. This is usually 20-30 points, but the number can be reduced by constant cross-section fuel tank parts.