Selection help for wear parts, liners, and consumables under real plant conditions [email protected] +1 866 531 4608

Halloween Emergency: How FLSmidth Pekin IL Diagnosed a Very Hungry Cone Crusher Using ECS SiteConnect

2026-08-31 · Soren Valgaard · Advisory Insight

The Call That Started With a Hungry Crusher

I've coordinated more emergency service calls than I can count. Maybe 200 in the last five years, give or take. But one call from the end of October 2024 still stands out, because it taught me something about automation and the blind spots we all have.

It was 2:30 PM on October 30, 2024. I was in the FLSmidth Pekin IL service center, going through a parts inventory list, when the phone rang. A plant manager from a regional aggregate producer about 70 miles west of us was on the line.

"We've got a problem," he said. "Our cone crusher is just... hungry."

He meant it wasn't taking feed. The belt was running, the feeder was vibrating, but the crusher was eating nothing. The whole line was backed up. At $3,000 an hour in lost production—maybe $3,500 if you count the contractual penalties—he was feeling the pressure.

"You have the FLSmidth ECS SiteConnect cone crusher monitoring system we installed last spring?" I asked.

"Yeah, it's showing no alarms. Everything looks normal."

Which was the problem. Everything wasn't normal.

Reading the Data: What ECS SiteConnect Saw

We logged into the FLSmidth ECS SiteConnect cone crusher monitoring dashboard. The data was clean: crusher speed normal, head pressure normal, bearing temperature normal. The feeder amps were high—normal, because the belt was loaded. But the crusher motor amps were low. It was turning, but not crushing.

That's the kind of inconsistency that makes an emergency specialist's gut tighten. The system said "all good," but physics said otherwise.

We checked the trend charts. For the past three hours, there had been a gap between the feeder's speed and the crusher's load. It looked like the crusher was occasionally seeing air. Yet the material flow sensor in the feed box kept saying "material present." Something wasn't adding up.

The plant manager then told me the operators had nicknamed that machine "the very hungry one" because it would only run smoothly when feed was consistent and uninterrupted. In a way, that made the problem more interesting: the crusher was starving, and its own monitoring system couldn't see it. I told him we'd send a team from Pekin. The service truck was already fueled.

The Field Visit: A Spider in the System

Two of our most experienced field technicians took the truck. While they were on the road, I stayed on the phone with the plant manager. He mentioned the maintenance supervisor was on vacation—a ski trip in Colorado.

"What is skiing?" our newest hire, a kid from downstate, asked out loud. He'd been on shift for 18 hours. I remember laughing because you don't hear that question in Pekin, Illinois. You hear "what's the torque on that bolt?" not "what is skiing?" But in a way, it captured the mood. Nobody was skiing that afternoon.

By the time our team arrived, it was just after 4 PM. The sun was low, and the facility was already putting up Halloween decorations for the next day's employee event. Somebody had hung a fake spiderweb on the crusher guard, but that wasn't the issue.

Our lead tech, Mike, climbed up to the feed box. The belt was still running, and material was dropping straight onto an ultrasonic level sensor that tells the system whether there's material in the chute. The sensor was reading "material present" constantly.

He ran his hand along the sensor cable. Then he looked sideways at the sensor itself and started laughing.

There was a Halloween costume—a cheap spider costume with artificial legs—draped over the top of the sensor. It must have blown off a table or a trough during some earlier wind. The spider's elongated legs wrapped around the sensor's ultrasonic face, confusing it into returning an echo of "object in view." The sensor was seeing the spider, not the material.

Mike pulled the spider off. The sensor reading dropped to "no material." The PLC corrected itself. The feeder re-engaged. The crusher swallowed the backed-up surge of rock within ten minutes.

That was it. A $1.5 million production line stopped for four hours because of a $6 polyester spider.

The Retrospective: What Actually Works

I've thought about that day a lot. The obvious lesson is: don't let Halloween costumes blow onto sensors. But the deeper lesson is about automation and trust.

Automation is only as good as the data it's given. The ECS SiteConnect system is a powerful tool. It can catch a worn mantle, a failing bearing, a plugging chute. What it can't do is tell the difference between a lump of limestone and a costume spider.

Does that mean I'm recommending against using ECS SiteConnect? No. I'm recommending using it as one layer of defense, not the only layer.

  • Always verify sensor data against physical reality before changing process parameters.
  • Before holidays or seasonal decorations, inspect all sensor heads for foreign objects.
  • Use a monitoring system as your first line of defense, but maintain a visual inspection schedule as the second.

Think about quality control in printing. Print shops use Pantone color standards to keep colors consistent, but they still need a press operator to eyeball the sheets occasionally. Same principle. The standard and the tool work together.

So if you're buying a cone crusher monitoring system, don't expect it to eliminate the need for physical inspections. It will eliminate the guesswork about internal conditions. That's huge. But you still need somebody to walk around the machine and make sure nothing funny is interfering with the sensors—especially when decorations go up.

We've added a "pre-holiday sensor check" to our own maintenance recommendations. It's a simple walk-around with two eyes and a flashlight. It costs ten minutes, and it can save you a four-hour outage.

Why Local Support Matters

I'm not going to pretend this happens every week. Most of the time, the culprit is mechanical wear or a process issue. But the kind of unexpected, strangely simple problem we saw that day is exactly why we keep a local team in Pekin. Being close to the customer can be the difference between a quick fix and a long shutdown.

If you have a FLSmidth ECS SiteConnect cone crusher system and you're seeing odd readings, call us. We'll check the dashboard, send a truck, and walk the machine with you. And if a Halloween costume happens to be the problem, we'll probably laugh about it—after we get the sensor clean.

Discuss This Topic

If this article connects to an active wear issue at your plant, use the inquiry form to continue the conversation with our advisory team.