All power plants today are faced with countless challenges, whether maintenance-related, environmental, budgetary or otherwise. Personnel are accomplishing challenging tasks in creative ways using their available resources.
One resource being relied on more regularly to keep obsolete and bad acting equipment running is the technical knowledge and capabilities of quality vendors.
Like many plant personnel with a lifetime of service working on particular pieces of equipment, long-term vendors who act in partnership with end users to troubleshoot and solve challenging problems often have institutional knowledge. A vendor with knowledge of equipment unique to a site allows for consistent evaluation of performance and return on investment. This was the case involving a troublesome pump application in a nuclear facility in the Northeast U.S.
A screen wash pump application in a nuclear power plant does not typically draw the same attention as some of the more load-critical pumps. In this case, the screen wash system includes traveling screens—one for each of the circulating water and service water pumps. The screens prevent debris from entering the circulating water system. The pumps provide high velocity water to flush the debris from the screens into an engineered trough system where it can be dealt with appropriately. Each single stage pump is capable of providing enough pressure to wash three screens at the same time. If the screen wash system goes down and enough debris collects to block the screens, it creates a maintenance challenge, since the screens must be cleaned manually. These particular pumps were failing on a regular basis due to harsh operating conditions in Coastal New Hampshire.
New Hampshire is known as the Granite State due to its numerous granite quarries. This granite can also be found in seawater along the coast, becoming more prevalent after a large storm surge or nor’easter.
The screen wash pumps use this seawater with varying levels of suspended granite whenever in operation. These particular pumps were installed in the 1980s and are a single stage, double suction horizontal split-case design.
They were originally constructed using materials thought capable of handling the conditions of service with cast iron casings, 303 stainless steel shafting, bronze wear rings and bronze impellers.
After a relatively quick failure, it was evident that neither the cast iron casing nor the bronze wear rings would withstand the conditions of service. While the impeller showed signs of wear, it was within tolerance and could be reused. A decision was made to upgrade the casing to bronze and the wear rings to a material known for its corrosion resistance to seawater. While overall reliability relatively improved due to the casing upgrade, the pumps continued to fail prematurely.
After years of costly repairs on a consistent basis, steps were taken in the aftermarket to increase mean time between failures (MTBF) and reduce associated costs. In early 2000, the site partnered with an aftermarket vendor to evaluate the overall pump condition and develop a strategy to address both cost and reliability. While an initial material upgrade to all internal components would have obviously increased MTBF, budgetary and time constraints called for a more methodical plan addressing both concerns.
The first course of action in an effort to reduce repair cost was to convert the wear ring material back to bronze from the more expensive choice, as no increase in reliability was found. Second, an epoxy coating was applied to the impeller to decrease the effects of flow-induced erosion, thus increasing the life cycle. Both bearing housings were compromised and were bored out to accept a bushing insert with a 0.002-inch interference fit. The same flow-induced erosion that continually wore the impeller took a similar toll on the casing. A brazing and machining procedure was developed and carried out to bring the casing fits back to original specifications. The overall repair cost was significantly reduced but the MTBF remained roughly the same.
All dimensional information was acquired during the evaluation process and reverse engineered drawings were created for all internal components. This information allowed new pump parts to be manufactured while the pumps were still in service with a goal of reducing downtime during future repairs. A new impeller was cast out of 316 stainless steel, and a shaft was manufactured out of the same material. A duplex stainless steel Nitronic 60 material was chosen for the wear rings due to its extreme resistance to corrosion and wear but not manufactured until the pump was pulled for repair. In 2008, the new components were installed in the first screen wash pump followed by the second pump in the system a few years later.
The same brazing procedures performed during the previous repairs were again used to address casing fits. After more than doubling the MTBF, the pump was pulled in 2013 due to a mechanical seal leak. Upon inspection, both the shafting and impeller withstood the harsh conditions and could be reused. The bearing housings and casing were again eroded and required the same procedure as before to be brought back to specifications. The ability to re-use the shafting and impeller dramatically reduced life cycle cost.
The second pump was pulled in 2016 after nearly tripling the MTBF. While the impeller and shafting could be reused the casing had severely eroded in the wear ring sections and concern arose with the amount of brazing necessary to bring it back within tolerance. After careful consideration, an engineering change was made to the casing. The wear ring sections were bored out to an oversized even plane allowing for the acceptance of oversized wear rings. The oversized rings were made out of the same Nitronic 60 material with the exact original features and assembled with the pump. This eliminated the need for brazing the wear ring casing sections without affecting the fit, form, function or performance of the pump.
The second pump with the oversized rings was put back in service in late 2016 and was started up with no loss of performance. Similar upgrades will be made to all future screen wash pumps reducing cost by eliminating the brazing process. Eventually the casing wear will exceed reasonable material loss and either the entire pump or casing will have to be replaced. Drawings have been created and materials chosen if a casing replacement is required in the future.
While the screen wash system in this case may not have been as critical to the power production as many other pumps on-site, it was easily one of the highest profile maintenance concerns and consistent budget consumers. With problems since installation which spanned almost four decades, it took historical knowledge and creative engineering to keep an obsolete pump running in a cost-effective manner.