PTQ Q1 2024 Issue

Pump operating issue case study: Identifying the poor performance root cause

Problem An alkylation unit’s main fractionator reboiler pumps were experiencing operational problems. Two 100% pumps were available: one online and one 100% spare P-900/901. A rotating equipment engineer described the problem: When placed online, the P-900 control valve is 70% open at 130,000 BPD flow. Shortly after start-up, the flow starts decreasing, and the control valve opens. It gets to a point where the control valve is wide open, and the flow is still dropping off. When the pumps are switched, the same behaviour is observed. The pump starts up per design, but the flow quickly deteriorates. It was concluded that the pump behaviour was indicative of a net positive suction head (NPSH) issue. It was thought that the net positive suction head available (NPSHa) was lower than the net positive suction head required (NPSHr). A new pump with a lower NPSHr was specified. Analysis A replacement pump was selected with input from the origi- nal equipment manufacturer (OEM) and was about to be pur - chased. Management asked the process engineering group to approve the new pump before the order was placed. In Stratus’ three decades of providing service to the hydro - carbon processing industry, the most common misdiagnosed and undiagnosed engineering issue seen has been hydrau- lics. Often, this is due to a lack of effective engineering tools, but it can also be due to improper application of available engineering tools. For this problem, the process engineer used Process Engineering ToolS (PETS) software to analyse the proposed replacement pump. Physical properties, hydraulic pressure drop, and NPSH evaluation were performed. The analyses

showed that no pump suction flashing should occur at the operating conditions. No NPSH issue was evident (see Figure 1 ). The analyses showed that the NPSHa was well above the installed pump’s NPSHr. Additionally, the proposed replace - ment pump had very similar NPSH requirements to the installed pumps. The Fittings Equivalent Lengths Tool was used to obtain the equivalent lengths for the pump suction and discharge piping. The PETS Pump Curve Tool was used to enter the pump curve data. The control valve datasheet information was entered into the Control Valve Tool. The Pipe System Hydraulics Tool was used to put it all together and model the system (see Figure 2 ). The model was used to check the system at various conditions. The model showed that the system was process capable, that the existing pump had no NPSH issues, and that there was no flashing in the system. PETS showed the pump opera - tional issues were not due to suction flashing, assuming typical pipe condition. However, the pump’s operational issues were real. What was causing it? The pump was taken apart to investigate. A pump suction witches hat strainer (unknown and with no visible handle) was installed backwards in the system. Debris was found inside the witches hat (see Figure 3 ). Figure 2 The PETS Pipe System Hydraulics Tool modelled the system and checked it at various conditions

Figure 1 PETS showed no NPSH issue was evident

Figure 3 Debris found inside the witches hat

PTQ Q1 2024