Skip to content
Sign inGet Started

Capacitive vs Ultrasonic Fuel Sensing: Contact Changes What Can Go Wrong

By Saeed Baatiah · Published:

Short answer

A capacitive probe sits in the fuel and reads its dielectric constant; an ultrasonic unit stays out of the liquid and times an echo to the surface. Their published accuracies are close, around ±0.5%, so the decision rests on contact, what fouls each, and whether the tank must be opened.

Key takeaways

  • A capacitive probe measures the fuel's dielectric constant, so a change of fuel blend changes the reading.
  • Diesel measures about 2.2 and biodiesel about 3.3 at room temperature, a 50% shift a capacitive probe reads as a level change.
  • Published accuracies are close: the OMNICOMM listing states ±0.5%, Eurosens states ±1%, and Teltonika's UL202 ultrasonic states ±0.5%.
  • Contact is the real axis: an immersed probe can be fouled by insulating films of dirt that falsify capacitance signals.
  • A non-contact unit avoids opening the tank, which matters under warranty and where an immersed element would be contaminated.

01Two physical principles, one question

Both technologies answer the same question, how high is the fuel surface, by measuring something entirely different. A capacitive probe is a pair of concentric electrodes lowered into the tank, and fuel between them changes the capacitance because fuel and air have different dielectric constants. The device therefore reports level indirectly, by reporting how much of its length is surrounded by liquid rather than air. An ultrasonic sensor emits a pulse and times the echo returning from the fuel surface, converting that flight time into a distance. The OMNICOMM listing is capacitive, specifically the LLS series, and covers a measurement range of 0 to 4,000 millimetres; the ULS-NC-001 in our marketplace is ultrasonic and non-contact. Understanding which quantity each device actually senses is what makes their failure modes predictable, because each one fails when its own measured quantity is disturbed.

02The accuracy numbers do not decide it

A buyer comparing the two usually starts with accuracy, and the figures refuse to separate them. Our OMNICOMM capacitive listing states ±0.5%. The Eurosens Dominator specification states ±1%. Teltonika's UL202 ultrasonic fuel sensor is published at ±0.5%. So the best capacitive figure and the ultrasonic figure are the same number, and the spread within the capacitive family is wider than the spread between the two technologies. That is a useful thing to establish early, because it means the accuracy row cannot carry the decision. It also means a vendor presenting one technology as inherently more precise is arguing past the published specifications. The choice has to be made on other grounds, and those grounds turn out to be about installation and about what disturbs each measurement rather than about how finely either can resolve a level under favourable conditions.

03What a capacitive probe is really sensitive to

Because a capacitive probe infers level from the dielectric constant of the liquid around it, anything that changes that constant looks like a change in level. The hydrocarbon fuel class a capacitive probe measures sits between 2 and 3, which is a narrow band and the reason the technique works at all. But measured room-temperature values put diesel at 2.2 and biodiesel at 3.3, a 50% shift. A tank refilled with a materially different blend therefore reports a step change that no fuel movement caused. Temperature is kinder than expected here: patent measurements give diesel a dielectric coefficient of 0.032%/°F against a density coefficient of 0.033%/°F, and because a decrease in fuel density also decreases the dielectric constant, the two nearly cancel across roughly −10 °F to 190 °F. So blend is the exposure, not heat.

04Fouling, and why immersion is the cost

The second exposure a contact probe carries is contamination on its own surface. It is a documented failure mode rather than a theoretical one: patent literature describes insulating films of dirt falsifying capacitance signals, and coating stability against water exposure is written into probe designs as an explicit requirement, one patent specifies holding the coating dielectric constant constant over many years of water exposure, with the coating constant itself between 2 and 7. That tells you two things. Manufacturers know immersion means deposits, and they engineer against it rather than deny it. It also tells you where a slow, quiet accuracy loss comes from in a tank carrying water or sediment. The probe does not stop working, it drifts, which is harder to notice than an outright failure and is the reason a periodic sanity check against a delivery volume is worth the effort.

05What non-contact buys, and what it costs

A non-contact ultrasonic unit measures the distance to the fuel surface from above, without any part of the sensor entering the liquid. That removes the entire fouling category at once, and it removes the installation itself as an intervention: nothing has to be immersed, and in the non-contact configuration nothing has to be cut into. That matters in three concrete situations, a tank still under manufacturer warranty, a site where cutting an opening is not permitted, and fuel carrying contamination that would foul an immersed probe. The cost is that the measurement now depends on a clear acoustic path. An echo has to leave the transducer, reach the surface and return, so anything in that path is in the measurement: internal structure, foam, heavy vapour, or a surface too disturbed to reflect coherently. A moving vehicle tank is the harder case for exactly that reason.

06Movement is a shared weakness, measured

Neither principle escapes the fact that a moving fuel surface is a poor thing to measure. Technoton publishes the degradation explicitly for a capacitive sensor: not more than 0.2% stationary, against 1% for moving vehicles, a fivefold change from motion alone, and it is quoted by the manufacturer rather than inferred by a critic. Ultrasonic sensing faces the same physics from a different angle, because a tilted or broken surface scatters the echo instead of returning it. The regulatory world takes this seriously enough to legislate against it: OIML R 117-1 requires an anti-swirl device in each road tanker compartment unless a compliant gas separator is fitted. So when comparing two datasheet accuracy figures, check whether either was measured stationary. A stationary figure describes a static storage tank well and a truck in motion considerably less well.

07Choosing from the tank, not the datasheet

Since accuracy does not separate them, decide from the installation. Ask whether the tank can be opened at all, warranty status and site rules answer that before any specification does. Ask what the fuel carries, because water and sediment argue for non-contact while a clean supply removes that concern. Ask whether the surface will be moving, since a stationary storage tank is the friendly case for both principles and a truck in motion is not. Then check the interface against the tracker, which is where many otherwise sound choices fail: the OMNICOMM listing is RS-485 Modbus RTU and needs a serial port with matching protocol support, while our Teltonika FMC003 listing has one RS-232 port, a different electrical standard, not interchangeable. Both technologies still need a calibration table, because both measure level rather than litres.

Capacitive contact probe versus non-contact ultrasonic sensing
PropertyCapacitive probeNon-contact ultrasonic
What it measuresDielectric constant of the surrounding liquidEcho flight time to the fuel surface
Published accuracy±0.5% (the OMNICOMM listing); ±1% (Eurosens)±0.5% (Teltonika UL202)
Enters the fuelYes, immersedNo
Affected by a fuel blend changeYes, diesel 2.2 versus biodiesel 3.3No
Affected by deposits on the elementYes, insulating films falsify the signalNo immersed element to foul
Needs a clear path above the liquidNoYes, internal structure, foam or vapour interfere
Needs a calibration tableYesYes, both report level, not litres

Figures and sources

Questions

Which technology is more accurate?

Neither, on published figures. Our OMNICOMM capacitive listing states ±0.5% and Teltonika's UL202 ultrasonic is published at ±0.5%, while Eurosens states ±1%. The spread inside the capacitive family is wider than the gap between the two principles, so accuracy cannot carry the decision.

Will a capacitive probe misread if I change fuel supplier?

It can, if the blend genuinely differs. Measured room-temperature values put diesel at 2.2 and biodiesel at 3.3, a 50% dielectric shift the probe reads as a level change. Temperature is a much smaller concern, because the dielectric and density coefficients nearly cancel.

When is non-contact clearly the right choice?

When the tank cannot be opened, under manufacturer warranty or where site rules forbid cutting, and when the fuel carries contamination that would foul an immersed element. Patent literature documents insulating films of dirt falsifying capacitance signals, so that concern is real rather than theoretical.

Does either technology work well on a moving vehicle?

Both degrade. Technoton publishes not more than 0.2% stationary against 1% for moving vehicles on a capacitive sensor, a fivefold change from motion alone. A disturbed surface also scatters an ultrasonic echo. OIML R 117-1 requires an anti-swirl device in each road tanker compartment for the same reason.

Sources

Related insights