Assessing Industrial Boiler Feedwater Pump Systems

Boiler feedwater pumping systems are critical to the efficient and reliable operation of power boilers. In many plants, these pumps were installed decades ago and need to be modernized and optimized. Generally, plant maintenance staffs and aftermarket repair organizations have attributed boiler feedwater failures to internal component damage that needs to be addressed via repair, rebuild or replacement. But in addition to evaluating the components, end users should consider boiler feedwater pump reliability issues in the context of system design. Pump component failures are often the result of external conditions that can be identified, addressed and modified. Correcting or optimizing system issues can maximize the performance of the components and extend the pump’s mean time between repair (MTBR). Reviewing a typical boiler feedwater system study can illustrate the importance of identifying and validating common performance issues. In this example, an industrial plant requested an assessment of its boiler feedwater system to understand why one of two parallel pumps was delivering less flow than the other pump. The existing three-feedwater-pump system was originally designed to supply two identical power boilers that create steam for the various plant processes. In this study, the No. 2 boiler feedwater pump was evaluated in relation to the overall pump system, which also included the No. 1 and No. 3 pump units. The No. 1 unit was off and served as a backup. The No. 2 pump was considered for efficiency loss—in this case, producing less flow than the No. 3 pump at the same load. At the time of the study, the No. 2 pump had been in operation, without a rebuild, for five years and the No. 2 pump for 10 years. According to maintenance records, the automatic recirculation control (ARC) valve in the No. 2 pump discharge line had required three component replacements during the previous year. It was suspected that the bypass line remained open during periods when recirculation flow was not required, i.e., during high-demand periods. The ARC valves are designed to be fully open in high demand conditions and to protect the pump by increasing recirculation (bypassing) in low-flow conditions and preventing reverse flow. The differential pressure orifice flow meters installed in the recirculation lines were not functioning properly. Maintenance planned to repair and calibrate both flow meters to determine the bypass flow condition of the operating pumps: Is pump No. 2 and/or No. 3 recirculating flow when the ARC valve for the recirculation (bypass) line should be closed, lowering efficiency and wasting energy? Once the differential pressure orifice transmitters were recalibrated, accurate flow rates could be measured to support decisions about the operating status of pumps No. 2 and No. 3. Based on the steam production rate calculations and the flow readings from the pump suction lines, the initial indication was that flow from the No. 2 pump was being bypassed to the deaerator during high-demand periods. This scenario was reflected in the calculated efficiency loss for the No. 2 pump. If both recirculation lines were verified to be closed and pumps No. 2 and No. 3 were operating synchronously under normal boiler load, and if the No. 3 pump was producing significantly more flow (at least 20 percent) than the No. 2 pump, then the No. 2 pump could be considered for repair or refurbishment. The vibration reading indicated that the No. 2 pump had 0.18 inches per second (ips) at peak, which was in the normal range. In order to validate the flow rates, a portable clamp-on flow meter was used on the discharge line, downstream of the bypass, of the No. 2 and No. 3 pumps. Two readings, one midmorning and one midafternoon, were taken. The No. 2 pump flows were 518 gallons per minute (gpm) and 530 gpm, while the No. 3 pump flows were 606 gpm and 750 gpm. The amp draw, flow and head pressure from each pump was used to estimate pump efficiency. The calculations indicated that the No. 2 pump was about 36 percent efficient and the No. 3 pump was 49 percent efficient. The flows for the No. 2 pump (518 and 530 gpm) are 73 percent of the best efficiency point (BEP = 720 gpm), and the pump is operating in the lower range for higher-speed centrifugal pumps. The flows for the No. 3 pump (606 and 750 gpm) are 84 percent and 104 percent of BEP, respectively. An economic analysis revealed that the annual electrical cost for pump No. 2 was $746,748, while pump No. 3 cost $774,565. That equates to a net cost difference of $27,817, which represented a cost difference of 3.7 percent—a secondary indicator of efficiency loss. At the conclusion of this study, maintenance personnel calibrated both orifice flow meters to the deaerator. This showed that the No. 2 pump was bypassing flow in high-demand periods with the No. 2 ARC valve open when it should have been closed. The No. 3 pump was operating normally; the ARC valve was closed during high-flow conditions and open in low-flow conditions. As a result, the No 2 pump was not sent for repair. The loss in the No. 2 pump’s efficiency resulted from bypassing flow in high-flow demand conditions. This outcome of the study demonstrated the importance of assessing the whole pump system in lieu of attributing pump performance issues strictly to internal component damage or failure. In this case, it saved the plant from repair costs and the potential for downtime costs if a second pump failed while one of the three pumps was out of service for repair.