Since 1980, the degradation of tubing in steam generators has driven many PWR owners to replace these large and expensive components. Initially, tube degradation necessitates plugging of the steam generator tubes which decreases thermal and hydraulic performance. As the number of plugged tubes increases, the possibility of having to derate the plant and for having to deal with additional licensing restrictions becomes significant. The final decision will be based on the lost revenue generating capabilities of the plant over its remaining life.

A steam generator replacement project is intended to restore plant performance to its original condition. Virginia Power’s Surry unit 2 was the first plant to attempt this complex operation. Since then, 48 additional plants have replaced a total of 138 steam generators at an estimated cost of nearly $6 billion.


Once the decision has been made to replace steam generators, the project team begins what is typically a four year exercise in planning and preparing. The plan culminates in an action-packed sixty day period (the actual outage) where hundreds of team workers descend upon the worksite and remove and reinstall several million pounds of hardware. What’s even more amazing is that the new 400 ton plus steam generators are typically machined to fit the existing plant’s piping systems within tolerances of less than a millimetre.

The Tihange project represented the largest steam generator replacement effort taken on to date by the Westinghouse/PCI organisation. Prior to Tihange 3, PCI had performed activities including severance cutting of the old steam generators from the reactor coolant system (RCS) piping, machining of the new steam generators and the old RCS piping, templating measurements to ensure proper fit-up of the new generators to the RCS piping, and welding of the new generators into the reactor coolant system on 20 steam generator replacement projects. Additional efforts for which Westinghouse/PCI were responsible at Tihange included the handling and transport of the old and new steam generators, RCS decontamination of the primary piping system, removal and reinstallation of all secondary system piping, removal and reinstallation of all steam generator supports, and all of the NDE associated with the scope of work.

The contract performance window was established beginning with the first cut on the primary system piping and was considered to be complete when the primary piping system was accepted for hydraulic testing, with all documentation by Tractebel. Included within this window were the cutting of the primary and secondary system piping, removal of the old steam generators, decontamination, machining of the existing primary and secondary piping, installation of the new steam generators, complete welding and NDE, as well as installation and shimming of all supports. The contractual window of 30 days was completed and accepted by Tractebel in 29 days, with the primary system window, which was the project’s critical path, completed in 20 days and 4 hours – almost 10 days ahead of schedule.


The Tihange 3 steam generator replacement project turned out to be one of the most successful efforts of its kind performed to date.

The project team (see panel) came extremely close to the current world record for 3-loop Westinghouse PWR’s, which was set at Almaraz 2 in 1997. PCI was also the primary system contractor on both Almaraz units. A comparison of major project performance objectives for all four steam generator replacements undertaken thus far in Belgium is provided in the table.

Excellent performance in the areas of project safety and personnel exposure provide additional measures of this project’s success. It is important to note that the Tihange 3 outage duration, which was completed in 77 days, also included regular refuelling and maintenance, reactor vessel head inspection, a major modification to the polar crane and a guide tube pin replacement. The success of this project was largely attributed to meticulous planning and careful integration of project team functions, as well as the excellent results of PCI’s laser metrology process which permitted extremely precise primary system fit-up and subsequently resulted in near perfect welding conditions.

PCI’s patented narrow groove welding processes enabled rapid completion of the critical welding processes. The combination of these technologies, Tractebel’s project leadership, careful project planning and experienced project implementation, which has been fine tuned during 21 projects, ensured this project’s success.


Tractebel and Electrabel are planning their next steam generator replacement project at Tihange unit 2 in the year 2001. Electrabel recently ordered the replacement generators for this effort and the company will go out for bids for the replacement contract early next year.

Since the completion of the Tihange project, PCI, as a subcontractor, has completed a steam generator replacement at Kori 1, a 2-loop plant in Korea, and as of this writing, is in process of carrying out a 4-loop replacement at Commonwealth Edison’s Braidwood station near Chicago. PCI’s order book also includes a backlog of seven additional projects which are scheduled to be performed through the year 2001. These projects are at Kewaunee, Shearon Harris, South Texas Project unit 1, Farley 1 and 2, Arkansas Nuclear One and D C Cook unit 1. It is a certainty that the lessons learned at this and the Tihange project will continue to benefit future steam generator replacement project’s.

In future, Westinghouse and PCI will take the lessons learned into these next projects, fully intent on setting more records.

Sorting problems

When technical challenges arose, special teams were immediately assigned and delegated with the task of developing solutions that would keep the project’s schedule on track and could be implemented for a minimum amount of additional radiation exposure while assuring everyone’s personal safety.
The most serious challenge to the project’s critical path schedule occurred when a seal dam bladder failed during the pipe-end decontamination process.
Seal dam bladder failure The pipe-end decontamination process is used to remove the layer of highly radioactive oxide from the inside diameters of the reactor coolant system (RCS) elbows prior to templating, machining, fit-up, and welding of the RCS. The process uses a sponge blast media delivered via high-pressure air to erode away the oxide layer. Prior to introducing the blast media, a seal dam with an inflatable bladder is installed to prevent introduction of blast media into the RCS primary piping. On the night shift of 22 June, at approximately 01.30, while the pipe end decontamination of hot leg (GV03) was in process, the seal dam failed and a small amount of blast media entered the RCS before the automatic safeguards on the decontamination equipment shut the system down. A team, which included client representatives, was assigned to develop a plan to recover the media and provide reasonable assurance that the blast media did not enter the reactor vessel or other areas not accessible. This plan included putting together a FOSAR of the hot leg elbow and piping to remove the blast media. Further, the plan included evacuating the water and possible media from the RHR suction line, and performing a visual inspection of the RHR suction piping to determine how much media may be left in the suction piping.
Investigations An inventory of all pipe-end decontamination tools and equipment was performed and all tools and materials were found to be accounted for. A pressure test of the failed seal dam was conducted. This test revealed that the cause of the loss of pressure in the bladder was due to a leak in the bladder where the valve stem enters the rubber part of the bladder. The reason that the seal dam failed was due to a manufacturing error causing a faulty bladder. The faulty bladder was removed from the seal dam that had failed and a new bladder was installed and tested.
The project continued to machine the elbows and install/weld the new steam generators while the plan for the sponge media recovery process from the RHR suction piping was developed and tested. This process included the design and fabrication of a suction effector/nozzle for the size of the suction piping. Also, a mock-up of the hot leg and the suction-piping configuration, including a channel head, was constructed and utilised to further develop the process/equipment/procedures. Finally, the owner participated and witnessed the qualification of this process prior to implementation.
After successful development of the tooling and training of personnel utilising the mock-up, the sponge media was successfully removed from the RHR suction piping. Further, visual inspections assured the project that no blast media had entered into the reactor vessel or any other areas. The critical path schedule of the project remained intact throughout this entire process.