Endress+Hauser Ultrasonic Level Transmitters vs. Radar: A 2025 Field Comparison

Why I'm Comparing These Two Level Transmitters

I'm the quality/brand compliance manager at a process automation company. I review instrument specifications before they reach customers—roughly 200 items a year. In 2024, I rejected about 12% of first deliveries over spec mismatches. Wrong process connection, wrong wetted materials, wrong measurement principle. The last one is the worst, because you usually don't discover it until the skid is built and the customer starts asking hard questions.

The question I keep seeing in RFQs is whether to use an Endress+Hauser ultrasonic level transmitter or a radar unit. Both have strong datasheets. Both have loyal users. But the datasheets don't tell you how each technology behaves in a 3-meter sump with baffles, or a dusty silo, or a tank with a vapor space that's basically steam. So let me compare them the way I'd audit a spec—and I'll be direct about what surprised me in the last few years of doing this. And before you ask—yes, I have a bias. I like ultrasonic when it fits. But the whole point of my job is that "when it fits" actually gets verified instead of assumed.

Dimension One: Accuracy—And the Case That Surprised Me

Start with the claim you hear constantly: radar is more accurate. In theory, it is. Radar doesn't care about gas density, media dielectric, or pressure. That holds up in a lot of installations. But "a lot" isn't "all," and my job exists because of the difference between those two words.

Here's something vendors won't tell you: radar's accuracy is tied to the dielectric constant of the media. When that's low—say, a solvent in a short fiberglass sump—the radar can spend more time chasing false echoes off tank internals than actually measuring. Ultrasonic just needs a solid surface to reflect the sound wave back and an accurate air temperature reading. That's it.

In Q1 2024, we ran a side-by-side audit on a customer's solvent tank. The radar unit kept wandering across a 15-millimeter band while it hunted for echoes from internal bracing. The Endress+Hauser ultrasonic, after its echo tracking update, sat at a steady ±5 millimeters. I went into that audit expecting the radar to win. It didn't. The fun part is, both instruments were connected to the same DCS, so there was no difference in wiring or integration to explain it.

To be fair, the radar has a home field too. I've rejected specs where someone wanted to put an ultrasonic on a 15-meter steam drum or a high-temperature reactor. That's not what ultrasonic is for. But if you're measuring a typical tank under 8 meters, with ordinary air in the vapor space, a modern ultrasonic level transmitter closes most of the accuracy gap that used to justify radar's higher price.

What I conclude from this dimension: the application decides accuracy, not the sensor label. Anybody who tells you "just pick radar" is working with a decade-old rulebook.

Dimension Two: Cost of Ownership—And the Maintenance Question Nobody Answers

Ultrasonic is cheaper, and the difference isn't pocket change. Based on quotes we've compared since early 2023—and verified again in January 2025—a Prosonic ultrasonic level transmitter lands in the $900–$1,400 range; a comparable Micropilot radar runs $1,800–$3,500. Prices move with the market, so treat those as reference points, not bids.

The lifecycle side is where the "cheap" option can bite you. An ultrasonic transducer face has to stay clean. Put it in a dusty solids silo and ignore it for two years, and you'll get exactly the performance you deserve. Radar is more tolerant of some coating on the horn, but it's also more expensive to refurbish when the coating wins.

That brings up a question people ask me all the time—the same one that shows up on lab forums as "how often to change your columns hplc agilent." Everyone wants a calendar answer. There isn't one. An HPLC column gets changed when the backpressure rises or the separation degrades. A level transmitter gets calibrated when the verification says it's drifting. I've seen a clean water tank stay perfectly calibrated for six years, and a lime slurry application drift in six weeks.

What I conclude here: ultrasonic wins the upfront cost battle. Radar justifies itself when the process is genuinely hostile. The maintenance winner is the one with a real verification plan in place.

Dimension Three: Integration Is Nice, but the Loop Is What Fails

Both technologies can talk HART, Profibus, Fieldbus—so protocol compatibility rarely settles the argument anymore. The thing that isn't on the datasheet is how the transmitter performs inside a real control loop.

Take a batching unit I reviewed in late 2024. The customer flagged the Endress+Hauser ultrasonic level transmitter as unreliable. I traced the problem to a valve positioner whose feedback encoder—a DFS60 encoder—had worn so badly that the valve was hunting around the setpoint. And the temperature element in the tank was sitting in a thermowell with a thirty-second time lag. The valve hunt was feeding level variations into the logic, which the customer interpreted as a transmitter problem—it's the classic "last instrument in the loop gets blamed" pattern.

Think about how a fever thermometer gives you a temperature reading: it works because the liquid expands when it gets warm. For an ultrasonic transmitter, air temperature changes the speed of sound, and the compensation algorithm needs a temperature reading that's fresh and representative. If that element is slow, every level reading gets a constant, invisible bias.

The transmitter wasn't broken. The loop was.

That insight usually doesn't come from a datasheet. It comes from support. When somebody searches "endress hauser contact," they're often not trying to buy a spare. They've got a loop problem, and they need an engineer who can walk through it with them. In my experience, that human resource is worth more than a tenth of a percent on the accuracy spec.

Conclusion here: when it comes to integration, the device is rarely the weakness. The weak link is everything else in the loop—and the quality of the support you get when the loop misbehaves.

So: Which One Do You Actually Spec?

Here's the shortcut I use when reviewing datasheets—and it's the same framework I'd put in front of a client:

• Choose ultrasonic for tanks under about 8 meters, ordinary atmospheric conditions, and media that won't coat the transducer. That covers a lot of water, chemical, and bulk-solids applications.
• Choose radar for long ranges, hot or steamy atmospheres, aggressive media, and any process that has already proven too harsh for ultrasonic.

And one last observation from my corner of the industry: the fundamentals haven't changed, but the execution has. What was a reasonable rule in 2020—"when in doubt, pick radar"—is now an expensive habit. Signal processing is better, self-diagnostics are better, and the price gap has stayed. So if you're doing a comparison because you're unsure, that's the right instinct. Just make sure you compare the right things: the actual process conditions, the loop behavior, and the support you'll get after the purchase.