When Tighter Is Wrong: The Over-Torqued Bolts That Fractured a Tennessee Plant

On November 13, 2020, a maintenance crew was doing something that looked entirely routine: torquing flange bolts on a heat exchanger at the Wacker Polysilicon North America facility in Charleston, Tennessee. Nothing about the task sounded dangerous. They were tightening bolts, the kind of work that happens in plants like that every week.

Then someone applied too much torque to the wrong bolts, and the heat exchanger cracked open.

Gaseous hydrogen chloride flooded the fifth-floor platform where seven workers were standing. Three of them, not wearing full chemical-resistant suits, panicked and tried to climb down the outside of the structure rather than wait for the cloud to clear. All three fell approximately 70 feet. One worker, Jesus Jared Aguilar Montes, was killed. The other two were seriously injured. Four other workers waited on the platform until the release stopped and then walked out safely down the single staircase.

The U.S. Chemical Safety and Hazard Investigation Board (CSB) investigated the incident. Their June 2023 report (No. 2021-01-I-TN) is unambiguous about the root cause: contractor pipefitters inadvertently over-torqued bolts on a live HCl piping flange, fracturing the graphite heat exchanger outlet pipe. One man died not because a bolt was missing, and not because a bolt had failed from corrosion or fatigue. He died because a bolt was too tight. (CSB Investigation Report 2021-01-I-TN)

What Actually Went Wrong

The heat exchanger in question was a graphite unit, a material chosen specifically because it resists highly corrosive environments like hydrochloric acid service. Graphite is excellent in that role, but it has a critical limitation: it is brittle. Unlike steel, which deforms before it breaks, graphite has almost no yield. Apply too much compressive force to a graphite flange, and it does not bend, it cracks.

The flange bolts on this exchanger were not all the same. Different connections within the assembly had different torque requirements, some joining PTFE surfaces to PTFE, others joining PTFE to graphite. Each pairing required a different maximum torque. The equipment manufacturer's manual, which the contractor crews were working from, included torque specifications for some bolt connections on the unit, but not for the specific bolts that were over-torqued.

So the pipefitters, working without written procedures from Wacker and without clear guidance in the manual, applied the torque value they knew, which happened to be too high for those particular bolts and that particular connection type. The result was a fractured pipe on live operating equipment, in a location seven workers had no fast way to escape from.

This is the part of the story that every maintenance manager, procurement professional, and fleet technician needs to hear: the problem was not a defective fastener. The problem was applying the wrong torque to the right fastener. The bolt itself was exactly where it was supposed to be. The thread engagement was fine. The material was appropriate. The torque was not.

The "Tighter Is Safer" Myth

There is a persistent assumption in hands-on maintenance work: if a bolt is supposed to be tight, making it tighter is a form of insurance. An extra turn or two gives you a margin of safety. This logic is wrong, and in the wrong application, it is lethal.

Fasteners are engineered systems, not just metal plugs. A bolted flange joint works by creating a precise, calculated clamping force on a gasket or mating surface. That clamping force seals the joint and holds the components together. But clamping force has a ceiling. Exceed it, and the consequences depend entirely on what is being clamped:

On a metallic flange with a soft gasket, over-torquing crushes the gasket permanently. The gasket loses its ability to recover and seal. Under thermal cycling, equipment heating and cooling during normal operation, the gasket cannot maintain the seal it was designed to hold. You may not see a leak immediately, but the seal is compromised.

On a non-metallic component, graphite, ceramic, fiber-reinforced polymer, or PTFE-lined equipment, the consequences are more immediate. These materials do not yield. The bolt stretches beyond specification before the material does, and then the material fractures instead. In some cases, the fracture is sudden and catastrophic, exactly as it was at Wacker.

On a standard steel flange, over-torquing stretches the bolt shaft past its yield point. A bolt that has yielded has lost its spring. Its clamping force is no longer predictable. When the joint experiences thermal expansion or vibration, that bolt cannot adapt, it has already given everything it had. What looks like a securely fastened joint may actually be a bolt that is one vibration cycle away from failing.

The Right Approach: Matching Torque to the Application

The CSB's recommendations for Wacker were specific: develop written torque procedures that clearly communicate the different torque requirements for different bolt connections within the same assembly. Use annotated photographs, color-coded bolt markers, or physical differentiation to prevent workers from applying a single blanket torque value to connections that require different values.

That guidance applies well beyond chemical plants. Any maintenance operation involving flanged equipment, pressure-containing joints, or non-metallic components needs the same discipline:

Know the OEM torque specification. A torque value that is correct for a steel-to-steel carbon flange is not automatically correct for a graphite, PTFE, or glass-lined connection on the same unit. The material being clamped determines the maximum allowable bolt load, not just the bolt size or grade.

Calibrate the tools. A torque wrench that reads 80 ft-lbs may be delivering 90. Calibration intervals exist for a reason. A tool that drifts by 10-15% is contributing to over-torquing errors on every joint it touches.

Distinguish between installation torque and retorque values. Equipment manuals sometimes specify a lower retorque value for a joint that has already been bedded in. Using an initial installation torque on a retorque cycle, as may have occurred at Wacker, applies more force than the joint is designed to accept after it has already seated.

Follow the bolt pattern. Applying full torque to one bolt before its neighbors are engaged pre-loads one side of the joint unevenly. Proper flange bolt torquing follows a cross-pattern sequence in multiple passes, bringing all bolts up gradually to the final specification.

The Right Fastener for the Job Means the Right Torque, Too

Choosing the correct bolt grade, material, coating, and length is only half the decision. The other half is applying the correct torque, no more, no less. At Wacker, the bolts themselves were not the wrong choice. The torque was.

Every fastener has a designed operating range. Below it, the joint is under-clamped and susceptible to leaks, vibration-induced loosening, or fatigue. Above it, the bolt, the mating component, or both are damaged. That range is narrower than most people assume, and on non-metallic equipment, it is narrower still.

A bolt does its job when it is installed correctly, torqued to specification, and matched to the material it is holding together. One worker at a plant in Tennessee did not go home in November 2020 because that match was not made precisely enough.

That is the lesson. Know what you are clamping. Know what the material can tolerate. Then torque to specification, and stop there.