Kleen Energy: How a Routine Pipe-Cleaning Method Became a Fatality Pathway

A pipe-cleaning task at the Kleen Energy power plant in Middletown, Connecticut, killed six workers and injured at least 50 people on February 7, 2010. The U.S. Chemical Safety and Hazard Investigation Board concluded that the explosion was not simply a failure to control ignition. It exposed a decision system that allowed a familiar construction practice to outrun hazard analysis, permit discipline, and regulatory attention.

The case still matters because the dangerous choice did not look dramatic at the moment it was made. The work was described as cleaning new fuel-gas piping. The fatal pathway emerged when natural gas was released during the task, accumulated in the building, and found an ignition source. That sequence is a useful warning for any operation in which commissioning, maintenance, or temporary work is treated as less hazardous than routine production.

Initial scenario: a temporary task with permanent consequences

Kleen Energy was a combined-cycle natural-gas power plant under construction. Workers needed to remove debris from newly installed fuel-gas piping before the system could be placed into service. The method used a high-pressure natural-gas release, commonly known in the industry as a gas blow, to push material through the pipes.

The U.S. Chemical Safety and Hazard Investigation Board documented that the release created a large flammable gas cloud inside the plant. The cloud ignited, producing the explosion that killed six workers and injured at least 50 others. The CSB final report was approved in 2010, and the Connecticut Governor's Commission also examined the event.

The initial scenario therefore contained more than one hazard. It combined stored energy, flammable material, temporary work conditions, incomplete construction, and the presence of multiple contractors. Yet the task was not governed with the same level of visible control that many organizations reserve for production operations.

That distinction is the first lesson. A temporary activity can be more exposed than a stable process because barriers are still being installed, responsibilities are moving between organizations, and the normal operating envelope has not been proven.

Decision: why the familiar method was allowed to continue

The critical decision was not only to use natural gas for pipe cleaning. It was to accept that method as a normal way to complete the work without demanding a stronger demonstration of how the gas would be contained, vented, monitored, and kept away from ignition sources.

The CSB reported that the practice was common in the power-generation industry. Familiarity made the method appear controlled, even though common use is not evidence that a hazard is acceptable. A recurring practice can become invisible to the people who should be challenging it.

The Governor's Commission found that the construction project was regulated in many respects, while the specific process used to clean the natural-gas piping was not clearly governed. That gap matters because a permit system can be extensive and still fail to control the decisive step when the task sits between engineering, construction, commissioning, and operations.

Andreza Ara�jo's book The Illusion of Compliance is useful here because it separates the presence of rules from the quality of the decisions those rules produce. Kleen Energy shows how a heavily documented project can still carry an unexamined exposure when nobody owns the question, "What could make this temporary method unacceptable today?"

Execution: where the barriers lost contact with the work

The work environment was changing as construction progressed. Contractors were completing installations, equipment was being commissioned, and different teams needed access to the same physical space. In such conditions, a control that exists on paper can lose its meaning when the people executing the work do not share one risk picture.

A strong control sequence would have required the project team to identify the gas release as a major hazard, define exclusion zones, establish a written method, verify the atmosphere, control ignition sources, confirm who had authority to stop the task, and coordinate every contractor affected by the release. The point is not to create a longer form. It is to make the most consequential assumptions visible before the energy is introduced.

The CSB's findings also showed why engineering and management controls must be considered together. A technical choice, such as how to remove debris from a pipe, creates organizational consequences. It determines the size of the hazard, the area that must be cleared, the competence required, the monitoring needed, and the quality of communication between teams.

In Safety Culture: From Theory to Practice, Ara�jo argues that culture becomes visible through repeated decisions, especially when schedule, cost, and convenience compete with risk controls. In this case, the work method was not just a technical detail. It was a culture signal about what the project considered normal during a high-consequence transition.

Measured result: six deaths and a control failure larger than one mistake

The immediate result was six fatalities, at least 50 injuries, and major damage to the plant. Those figures come from the CSB's Kleen Energy investigation and the Connecticut Governor's Commission. They describe the consequence, but they do not fully describe the management failure.

The result was not caused by the absence of every safety rule. It came from the distance between the rules, the selected method, and the physical conditions created by the task. That is why incident investigation should examine how work was authorized, not only which individual action occurred immediately before ignition.

Why the control system failed before ignition

The failure became visible at ignition, but the control problem began earlier, when the project treated the pipe-cleaning method as a narrow technical choice instead of a temporary major-hazard operation. That framing limited the questions asked by managers, contractors, and regulators before the release began.

This is a recurring investigation pattern. The final event is often fast, while the conditions that make it possible accumulate slowly through accepted assumptions, unclear ownership, and weak verification. A useful report should reconstruct that accumulation so the organization can remove the pathway instead of merely reminding workers to be careful.

Generalizable lessons from the Kleen Energy case

Temporary work deserves a higher challenge threshold

Commissioning, shutdown, construction, and maintenance often involve unfamiliar combinations of people, equipment, and energy. A task should receive more scrutiny when the normal barriers are incomplete, not less scrutiny because the activity is temporary.

Familiarity is not a risk assessment

When a method is common, the organization may stop asking what has changed since the last time it was used. Supervisors should require a current explanation of the hazard, the credible escalation path, and the conditions that would make the method unacceptable.

Ownership must follow the hazard

Projects often divide responsibility among the owner, engineer, general contractor, specialist contractor, and operations team. That structure is not automatically unsafe, but it becomes dangerous when every party assumes another party owns the temporary process that creates the exposure.

Permits should expose assumptions

A permit should force the team to state how gas will be released, where it can accumulate, what will prevent ignition, who will verify the atmosphere, and who can stop the job. If the form only records signatures, it is evidence of administration rather than control.

Investigation must follow the decision chain

James Reason's work on latent failures helps investigators look beyond the final act and examine the conditions that made the act plausible. The question is not whether a worker should have noticed danger in the last seconds. The stronger question is why the system allowed the work to reach that point with the exposure still unresolved.

What to apply in your operation

Start with the temporary tasks that introduce energy, flammable material, pressure, or simultaneous contractor activity into spaces that are not yet operating normally. Do not begin by asking which forms are missing. Begin by asking which work methods could create a major hazard before the formal process recognizes it.

For each selected task, require a short decision review led by the person who controls the work method. The review should identify the hazard source, the credible release scenario, the affected area, the people who must be removed or protected, the verification that proves conditions are safe, and the exact authority to stop the task.

Then test the control at the worksite. A supervisor should be able to point to the isolation, monitoring point, exclusion boundary, communication route, and emergency response without searching through a folder. If the control cannot be seen or verified, it has not yet reached the work.

Finally, treat repeated temporary work as a management signal. When the same exception appears every shutdown or commissioning cycle, the organization should redesign the method rather than normalize the exception. Ara�jo's Safety Culture Diagnosis offers a practical way to examine whether the organization rewards visible control or merely rewards completion.

Kleen Energy was not a warning about one careless worker. It was a warning about a familiar method that was never challenged at the level required by its potential consequence. Safety is about coming home, and that standard must apply most strongly when work is temporary, shared, and changing.

Explore Andreza Ara�jo's work on safety culture, leadership, and incident prevention.