They Kept Replacing the Level Sensor. A Starrett Tube Micrometer Found the Real Problem.

They Kept Replacing the Level Sensor. The Level Sensor Was Fine.

In my role coordinating emergency precision measurement deliveries for industrial customers, I've handled a lot of rush orders. Most start with a panicked phone call and a deadline that has already started moving. This one started with a level sensor.

On a Thursday in May 2024, at 4:47 p.m., I was finishing a calibration log when an engineer from a food processing plant called. 'We need a Starrett tube micrometer and a valid calibration certificate here by 6 a.m.,' he said. I asked what happened. It had started, like too many plant emergencies do, with an alarm.

The glycol expansion tank in the cooling line kept tripping its low-level alarm. An operator added four gallons of glycol, reset the system, and continued his rounds. Two hours later, the alarm went off again. A technician replaced the level sensor with a brand-new unit. That bought them five hours. By lunch, the alarm was back, and someone was about to replace it again. The engineer stopped them.

It is tempting to blame the new level sensor. But the level sensor was telling the truth. Glycol was leaving the tank. Somewhere behind a wall, under pipe insulation, there was a leak.

How the FLIR One Thermal Camera Pointed Us to the Leak

A service tech had a FLIR One thermal camera attached to his phone. Late the night before, he scanned the corridor where the return line ran and spotted a cold streak along the insulated pipe chase. It was faint, but it did not match the surrounding thermal pattern. That was the first clue.

If you look for 'how to use FLIR One Thermal Camera,' you will find a lot of pretty pictures and surface temperature readings. The real process is more careful than that. On this job, the practical steps were:

1. Set the emissivity before comparing surfaces. The FLIR One preset menu makes this easy, but you must change it when you move from painted steel to wet insulation.
2. Take a reference image of a known dry area. It gives you a visual baseline, like zeroing a micrometer before taking a reading.
3. Shoot from multiple angles. Straight-on thermal shots can hide reflections, especially near electrical panels and warm equipment.
4. Use the thermal image as a map, not as a diagnosis.

That last point is the one that saved us. The FLIR One showed where to open the wall. It did not show the remaining wall thickness of the damaged line, and it did not show how far moisture had traveled. Those answers required other tools.

The MR176 Moisture Meter Showed Where to Stop Cutting

After the crew opened the first section of insulation, the engineer asked a very practical question: how far does this damage go? If he only repaired the visibly wet part, the leak could be sitting in saturated insulation further down the line. That is where the MR176 moisture meter earned its place on the job.

The crew used the MR176 in pinless mode first, walking it along the surfaces near the pipe chase. When the relative reading jumped, they switched to pin mode on a piece of material that was already being removed, just to confirm the reading. It was not a laboratory measurement, but it was repeatable, and it gave the repair crew a defensible place to stop cutting.

I don't have hard data on how many emergency repairs are delayed because someone stops at the first wet patch. But based on the rush jobs I've coordinated, my guess is that it happens more often than people admit. The moisture meter doesn't make the decision, but it makes your thinking visible.

Why the Starrett Tube Micrometer Was the Critical Tool

Once the insulation was peeled back, the metal underneath looked bad. Pitting, rust scale, and a few spots that made the engineer go quiet. He had two options: patch the line and watch it, or cut out the damaged section immediately. A visual inspection was not enough to decide.

We had already shipped the Starrett tube micrometer and a current calibration certificate traceable to NIST. The micrometer let him measure the remaining wall thickness directly. He zeroed it on a certified reference, took readings at the worst pits, and wrote the numbers down. The numbers made the decision for him: replace the section.

I've seen a lot of repair debates die on a measurement like that. When somebody asks, 'How bad is it?' there is no better answer than a direct reading from a calibrated precision tool.

The Starrett 0 12 Micrometer Set Was Overkill Until It Wasn't

The original request included the full Starrett 0 12 micrometer set. I almost questioned that request because the damaged line was one nominal size. I'm glad I kept my mouth shut.

Here is what that full set did on site:

• It let the crew measure the outside diameter of replacement tubing and fittings without waiting for another tool delivery.
• It covered different line sizes in the same chase. The first leak wasn't the only one.
• It kept the repair moving through the night with certified measurements, not estimates.

Before dawn, the crew found a second questionable section on a different size run. Because the Starrett 0 12 micrometer set was already there, they did not lose three hours waiting for an additional tool.

That is when I learned, again, not to assume. I assumed one micrometer would be enough because I was thinking in single measurements. The crew was thinking in entire systems. Those are different categories.

The Real Lesson Was About Specialists

The line was back in service before 7 a.m. At nine, the engineer called to say the level sensor was reading steady. It had done its job all along.

I'm not a process engineer. I don't sign off on minimum wall thickness numbers or repair procedures. I help get the right measurement tool to the right place, with valid calibration, and I make sure the people using it understand what it can and cannot show.

I'd rather work with a specialist who knows that boundary than a generalist who promises to handle everything. The vendor who says, 'This part is outside my lane, here is who I trust for that,' earns my respect. When the stakes involve an entire plant shutdown, you don't want false confidence. You want someone who can measure the truth.

On that night, the truth was a set of numbers from calibrated instruments. The level sensor knew it first. The equipment helped us act on it.