Measurement results with different pumps

The effect of cavitation occurrence on the estimated rotational speed and shaft torque was also tested with two other pumps in the laboratory in order to have more general results. For the laboratory measurements, the pumping systems were installed in the same water tank system as the Sulzer pump. Since different pumps have their own operational characteristics and NPSH_R curves, these tests allow more generalised conclusions concerning the use of a frequency converter in the detection of cavitation occurrence, although the specific speeds and BEP values of these pumps are relatively close to each other.

Serlachius pumping system. Firstly, tests were conducted with a pumping system consisting of a Serlachius DC-80/260 radial flow centrifugal pump, a four-pole 15 kW Strömberg induction motor and an ABB ACS 800 frequency converter (see Fig. 4.33 and Appendix A). The operating point location of the pump was monitored with the pressure measurements for the pump head and flow rate. The mean values of the rotational speed and shaft torque were determined with an HBM torque measurement transducer on the motor shaft. The estimates provided by the frequency converter were stored with the ABB DriveDebug software similarly as with the Sulzer pumping system. Compared with the Sulzer pump, the required NPSH of the Serlachius is notably higher, 2–7 metres at 1425 rpm (see Fig. A.3). This allows the occurrence of cavitation in the laboratory without additional suction-side throttling, which may have affected the test results of the Sulzer pumping system.

Fig. 4.33: Serlachius pumping system. It consists of a Serlachius DC-80/260 centrifugal pump and a Strömberg induction motor. The rotational speed of the pumping system was controlled with an ABB ACS 800 frequency converter.

Five measurement sequences were carried out with the Serlachius pumping system. Each measurement sequence had a specific system curve shape: the valves were adjusted so that the flow rates were approximately 50, 70, 100, 120 and 150 % of 22 l/s at 1425 rpm that is theQ_BEP of the Serlachius pump. With these valve settings, the pump was driven at rotational speeds ranging from 1050 rpm to 1800 rpm. At each rotational speed, pressure measurements for the pump operating point location were carried out, and estimates for the rotational speed and shaft torque were read from the frequency converter.

Values ofn_RMS andT_RMS were calculated for each measurement from the pre-filtered estimates of the rotational speed and shaft torque, as they have been feasible in the case of the Sulzer pump. Normal valuesn_RMS,N and T_RMS,N were determined by calculating the mean ofn_RMS and T_RMS values, which have a sufficient NPSH ratio to avoid the occurrence of 3 % head-decreasing cavitation: as the suction energy and impeller inlet velocity values of the Serlachius and Sulzer pump are almost the same, the corresponding NPSH criterion of 1.35 was applied for this pumping system. The resultingn_RMS/n_RMS,N ratios are shown in Fig. 4.34 as a function of NPSH ratio. Respectively,T_RMS/T_RMS,N ratios are shown in Fig. 4.35. These figures clearly show the impact of fully developed cavitation occurrence on then_RMS/n_RMS,N andT_RMS/T_RMS,N ratios.

They also support the assumption that feasible threshold values forn_RMS/n_RMS,N andT_RMS/T_RMS,N are case-specific, which is why a further study for the determination of these threshold values should be carried out. The occurrence of cavitation in the measurement sequence with a 150 % flow rate was also noticed during the measurements from the rattling noise of the pump and varying measurement values of the rotational speed and shaft torque, which is captured by n_RMS/n_RMS,N andT_RMS/T_RMS,N.

0 1 2 3 4 5 6 0

1 2 4 6 8 10 12 14 16

NPSHA/ NPSH

R

n RMS/n RMS,N

1.5*Q_{BE P} 1.2*Q_{BE P} QBEP

0.7*Q_{BE P} 0.5*Q_{BE P}

Fig. 4.34: n_RMS/n_RMS,N ratios as a function of NPSH_A/NPSH_R ratio for the Serlachius pumping system.

The chosen NPSH ratio of 1.35 for the result classification is shown by a dashed vertical line. In this case, the occurrence of cavitation has a strong impact on the relative values.

0 1 2 3 4 5 6

0 1 2 4 6 8 10 12 14

NPSH_A / NPSH_R T RMS/T RMS,N

1.5*Q_{BE P} 1.2*Q_{BE P} QBEP

0.7*Q_{BE P} 0.5*Q_{BE P}

Fig. 4.35: T_RMS/T_RMS,N ratios as a function of NPSH_A/NPSH_R ratio for the Serlachius pumping system.

The chosen NPSH ratio of 1.35 for the result classification is shown by a dashed vertical line.

As a more detailed example of the results obtained for the Serlachius pumping system, n_RMS/n_RMS,N andT_RMS/T_RMS,N ratios are given in Fig. 4.36 for the rotational speed of 1500 rpm.

The NPSH ratios for each measurement are also given in the figure. The figure shows that the RMS ratios are clearly higher when the NPSH ratio is 1.18.

5 10 15 20 25 30 35 40 0

1 2 4 5

Flow rate (l/s) n RMS/n RMS,N

5 10 15 20 25 30 35 40

0 12 4 6 8 10

Flow rate (l/s) T RMS/T RMS,N

4.74 4.75 3.27 2.13

1.18

1.18

2.13 3.27

4.75 4.74

Fig. 4.36: nRMS/nRMS,N andTRMS/TRMS,N ratios as a function of flow rate at the rotational speed of 1500 rpm. NPSH ratio of the Serlachius pump is also given for each measurement.

Grundfos pumping system. Another sequence of test measurements was carried out with a pumping system consisting of a Grundfos LP 100-125/130 mixed flow centrifugal pump, a two-pole 5.5 kW Grundfos induction motor and an ABB ACS 800 frequency converter (see Fig.

4.37 and Appendix A). This pumping system was tested to have results also with a different type of centrifugal pump, and consequently to analyse, if the proposed method could be also feasible with other pump type than just with the radial flow centrifugal pumps.

The operating point location of the pump was monitored with the pressure measurements for the pump head and flow rate. Also this pump has a larger NPSH requirement than the Sulzer pump, 2–9 metres at 2900 rpm. In addition, this pump has higher n_q and n_ss than the Sulzer pump because of its mixed flow impeller. For this reason, a larger margin of NPSHA may be required to avoid the occurrence of cavitation.

Fig. 4.37: Grundfos pumping system. It consists of a Grundfos LP 100-125/130 centrifugal pump and a Grundfos induction motor. The rotational speed of the pumping system was controlled with an ABB ACS 800 frequency converter.

Five measurement sequences were carried out with the Grundfos pumping system. Each measurement sequence had a specific system curve shape: the valves were adjusted so that the flow rates were approximately 50, 70, 100, 120 and 155 % of 24 l/s at 3000 rpm (Q_BEP of the Grundfos pump). With these valve settings, the pump was driven at rotational speeds ranging from 2100 rpm to 3180 rpm. At each rotational speed, pressure measurements for the pump operating point location were carried out, and estimates for the rotational speed and shaft torque were read from the frequency converter. Correspondingly, normal values n_RMS,N and T_RMS,N were again determined by calculating the mean of n_RMS and T_RMS values, which have a sufficient NPSH ratio to avoid cavitation (NPSH_A/NPSH_R ratio of 2 was selected because of the larger suction specific speed compared with other pumps).

The resulting n_RMS/n_RMS,N ratios are illustrated in Fig. 4.38 as a function of NPSH ratio.

Correspondingly, T_RMS/T_RMS,N ratios are shown in Fig. 4.39. In this case, the occurrence of cavitation has a clearer effect on the shaft torque than on the rotational speed. A measurement sequence with the 120 % relative flow rate has a notably highn_RMS/n_RMS,N, although the pump was otherwise operating without signs of cavitation during the measurement. This may be caused by the design of this pump, possibly by incipient cavitation occurring at NPSH ratios of 2.5–3, or by some other pump or process-related issue. Otherwise, also in this case the occurrence of head-decreasing cavitation has generally an increasing effect on the n_RMS/n_RMS,N andT_RMS/T_RMS,N ratios, when the pump is prone to a fully developed cavitation having an effect on the pump performance.

0 1 2 3 4 5 6 7 8

Fig. 4.38: n_RMS/n_RMS,N ratios as a function of NPSH_A/NPSH_R ratio for the Grundfos pumping system. The chosen NPSH ratio of 2 for the result classification is shown by a dashed vertical line. According to n_RMS/n_RMS,N ratios cavitation may already occur with higher NPSH ratios.

0 1 2 3 4 5 6 7 8

Fig. 4.39: T_RMS/T_RMS,N ratios as a function of NPSH_A/NPSH_R ratio for the Grundfos pumping system. The chosen NPSH ratio of 2 for the result classification is shown by a dashed vertical line.

As a more detailed example of the results obtained for the Grundfos pumping system, n_RMS/n_RMS,N andT_RMS/T_RMS,N ratios are given in Fig. 4.40 for the rotational speed of 3000 rpm.

The NPSH ratios for each measurement are also given in the figure. The figure shows that the RMS ratios are at highest when the NPSH ratio is 1.09. At this rotational speed, however, the difference inT_RMS/T_RMS,N ratios is small between the NPSH ratios of 1.09 and 2.78, which could indicate that the pump may be cavitating also at NPSH ratios over 2. However, this was not detected during the measurements.

5 10 15 20 25 30 35 40

Fig. 4.40: nRMS/nRMS,N andTRMS/TRMS,N ratios as a function of flow rate at the rotational speed of 3000 rpm. NPSH ratio of the Grundfos pump is also given for each measurement.