Pump Monitoring How to Rectify Any Pump Inefficiency Before It Becomes a Problem
How can operators ascertain whether a pump is deficient and a failure imminent? The traditional approach: Installing sensors, e.g. for structure-borne noise, pressure, temperature, flow, power consumption, bearing position, at critical points and continuously evaluating the collected data. For non-critical applications, there is an easier way: A drop in energy efficiency indicates a possible problem.
In the classic movie “Das Boot” there is a scene where the engineer listens in to the marine diesel with an ear trumpet — an early stethoscope — to recognize bearing damage at an early stage. KSB has now practically replaced the ear trumpet with a Smartphone app called “Sonolyzer”. It uses an algorithm to determine the efficiency of fixed speed pumps from the noise spectrum of an asynchronous motor. Specifically, the measurement result indicates whether the operating point lies inside or outside the part-load range.
The technology behind this system is also used in “Pump Meter”, which replaces the “Pump Expert” system, which operators deemed too complex. With considerable success, as Dr. Thomas Paulus, project manager startup Industry 4.0 at KSB, reports: “An early fault detection system should cause no additional effort for operators and should already be integrated in the system — like our Pump Meter. With a few sensors, we determine the pump’s operating point and load profile. This information can be used to determine whether a pump is working well or not. Moreover, it also allows estimating the pump's expected availability. The system was very well received on the market. Approximately 35,000 pump meters have already been installed. With costs of around € 400 we are obviously within operators' accepted budget range.”
The CR monitor concept by Grundfos also uses relatively few sensors: Based on the state at new installation, the system monitors deviations from the pump system’s original efficiency. The CR monitor configures itself automatically, with a controller monitoring the data collected on site during the phase of learning. If measurements deviate from the reference value or exceeds the limits set by operators, the CR monitor emits an early warning message before potential faults occurs. These alerts can be sent via data bus to a control center, via text message or simply via signal light. When issuing a warning message, CR monitor does not transmit the complex raw data, but already provides a specific error analysis.
Ingo Landwehr from Grundfos outlines future developments: “The CR monitor will be continually expanded to be used by more pumps. To this end, we integrate the early fault detection functionality directly into our MGE motor. For example, we capture data concerning the bearing service life and sealing wear and inform operators which spare parts to order. We also expect operators in the process industry would want to feed the early fault detection system's data into their process control system for central analysis. After all, the approach is outlined in the Industry 4.0 concept.”