Failure of a Power Station Cooling Tower

Project Info

RTI recently investigated component failures in two cells of an eight cell power station cooling tower. During an overnight shift at the power station, workers observed that water temperatures in the tower were not decreasing sufficiently. No alarms, notifications, or out-of-the-ordinary operations were detected. When workers inspected the tower they found that fan blades and pieces of carbon fiber driveshaft were scattered in and around the cooling tower. The motor fans in cells G and H were running but no air movement could be observed in either cell. Closer inspection revealed that the driveshaft in cell G and the fan blade assembly in cell H had failed catastrophically. Investigation and disassembly of cells G and H revealed that one blade of the cell G fan assembly had a hole through it near the leading edge close to the blade tip and one of the lightning rods of the lightning suppression system was found at the bottom of cell G, bent in several places. Other detached lightning rods were found on the deck of the cooling tower.

Prior to the failure, a contractor had performed service on cell G which included the installation of a new fan blade and hub assembly and a new pedestal mount for the gearbox that attaches to the fan. The new fan assembly consisted of blades with the same gross dimensions as the original blades, but represented a new altered design. The primary visual difference was that the new blades were one piece instead of two. As well, the leading and trailing edges of the new blades were curved, thus providing an enhanced aerodynamic performance. RTI was told that upon installation and balancing of the new blade assembly, the fan’s vibration measurements were recorded during operation and the behavior was found to be within the manufacturer’s prescribed tolerances.

Upon completion of the fan blade assembly and pedestal installation, it was noted that the tips of the blades were hitting the surrounding shroud. The shroud is a circular wall with a curved profile comprised of several sections bolted together that both protects the fans and encourages efficient aerodynamic characteristics of the air flow through the cell. The tower walls and deck are attached to an internal structure while the column that supports the pedestal, gearbox, and fan blade assembly in each cell is independent of the tower structure. The contact appeared to result from vibration of the shroud as opposed to that of the fan blades. In order to alleviate this problem, attempts were made to install shims in between the sections of the shroud to effectively increase its diameter enough such that the movement of the walls due to the vibration did not result in a lack of clearance between the shroud and the blades. As the shroud is bolted directly to the deck surface of the cooling tower, the diameter increase required new mounting holes be drilled in the deck. Upon doing so, the contractor discovered that portions of the deck’s plastic-encased wooden structure had deteriorated to the point where reattaching the shroud could not be accomplished without significant repairs being made to the deck.

RTI determined that the final catastrophic failure of the cell G driveshaft and the cell H fan blade assembly were caused when a lightning rod attached to the cell G shroud vibrated out of its sleeve and fell into the cell, penetrating the fan blade and creating the aforementioned hole. Continued rotation of the fan resulted in repeated impact of the rod to the drive shaft, causing it to fail. The significant upward air flow in cell G carried the drive shaft segments up and out of the cell. Some of these segments likely fell into the adjacent cell H, hitting one or more fan blades causing the catastrophic failure of the entire cell H fan blade assembly. The vibrations causing the cell G lightning rod to detach were the result of a combination of enhanced aerodynamics of the new fan blades and the deteriorated condition of significant portions of the cooling tower deck.

Consequently, the primary cause of this sequence of failures was the deterioration of the deck structure in a very damp environment. A lesson to be learned is: Question whether the supporting structure can safely accommodate newly upgraded equipment!


  • Civil Engineering
  • Construction
  • Electrical Engineering
  • Mechanical Engineering
  • Utilities