Pump Performance

Affinity Laws. Pump performance is affected by the rotating speed. When speed increases, the flow increases linearly, and the head increases as a square of the speed (14). 

Figure 10-32 shows the schematic of a pump, moving a fluid from tank A to tank B, both of which are at the same level. Tne only force that the pump has to overcome in this case is the pipe function, variation of which with fluid flow rate is also shown in the figure. On the other for the use shown in Figure 10-33, the pump in addition to pipe friction should overcome head due to difference in elevation between tanks A and B. In this case, elevation head is constant, whereas the head required to overcome fric tiou depends on the flow rate. Figure 10-34 shows the pump performance requirement of a valve opening and closing. 

FIG. 10-31 Typical pump performance curve. The curve is shown for water at 85 F. If the specific gravity of the fluid is other than unity, BHP must he corrected. 

The same procedure maybe used at other pump flows to permit plotting the series of balance-point curves as has been done in Fig. 29-61. From such curves, one can establish the maximum lean pump at any total tower outflow, and combining this with the semilean-pump performance curve results in Fig. 29-55. Bypass flow plotted in Fig. 29-55 is obtained by adding simultaneous lean- and semilean-pump flows and subtracting the recovery pump-turbine flow required to make the balance point at that lean-pump flow. 

Motivation Unit tests require a substantial investment in time and resources to complete successfully. This is the case whether the test is a straightforward analysis of pump performance or a complex analysis of an integrated reactor and separation train. The uncertainties in the measurements, the likelihood that different underlying problems lead to the same symptoms, and the multiple interpretations of unit performance are barriers against accurate understanding of the unit operation. The goal of any unit test should be to maximize the success (i.e., to describe accurately unit performance) while minimizing the resources necessary to arrive at the description and the subsequent recommendations. The number of measurements and the number of trials should be selected so that they are minimized. 

Centrifugal pumps perform the same fimetion as PD pumps, but they do it differently. The.se pumps generate pressure by aeeelerating, and then deeelerating the movement of the fluid through the pump. 

Some pump companies will promote and tout their low Nss values. Sometimes a specification engineer will establish a maximum Nss limit for quoted pumps. Let s consider the.se examples of operating parameters of pumps, and determine the Nss. These values are lifted from the pump performance curves at the BLP. 

Pump performance curves are the least used, least consulted, least appreciated, and least understood aspect of the world of industrial pumps. The plant personnel who most need their pump curves, meehanics and operators, generally don t have the curves and accompanying information at their disposal. The people who control the performance curves store them in a file, in a drawer, in a cabinet that s almost never opened. They don t share the information contained in the curves with the people who need it. Maybe it s because they themselves don t understand the information to share it. In the next few paragraphs and pages, we re going to explain the pump performance curves. This might be the most important chapter of the book. 

It is interesting to note that some pump users literally know that their pumps will fail after a specific time period. They understand that the running time of the pump should be maximized to have an acceptable yield in the process. This type of strategy is expensive since it raises a doubt of the continuity of the pump performance. To compensate, some plants install back up or redundant pumps. 

Figure 1.4.1 Pressure drop calibration and pump performance measurement. ...

Piston, or positive displacement pumps, are well known and much used. Centrifugal pumps are not as well understood. Consequently, piston pump performance is sometimes expected from centrifugal blowers. The main difference is that positive displacement or piston pumps generate flow, whereas centrifugal pumps produce pressure. With a piston pump, the pressure will increase to the level needed to maintain the flow set by the piston volume and stroking speed. In contrast, centrifugal pumps produce pressure the flow will increase until the pressure drop, produced by the flow, matches the pressure produced by the pump. 

In the left upper corner of Figure 3.4.1, the centrifugal pump performance is shown. As ean be seen, the head generated depends on RPM but is independent of the flow, within a 10 % error, up to a eertain limit. The pressure staits to deeline when that point is reaehed at whieh the flow is high enough that the pump itself limits the flow beeause of its eross-section. 

Penney, R. W., Inert Gas in Liquid Mars Pump Performance, Chemical Engineering, July 3, 1978, p. 63. 

Balje, O. E-, Study on Design Criteria and Matching of Turbomachines, Part B Compressor and Pump Performance and Matching of Turbocomponents, ASME Paper No. 60-WA-231, ASME Transactions, Vol, 84, Journal of Engineering for Power, January 1962, p, 107. 

The effects of impeller shape for the usual centrifugal process pump performance are given in Figure 3-34. The only part the process designer can play is in the selection... 

Figure 3-32. Comparison of impeiier types for centrifugal pump performance. (Adapted by permission from Pic-a-Pump, Aliis-Chalmers Mfg. Co.)...

This available value of NPSHa (of the system) must always be greater b) a minimum of two feet and preferably three or more feet than the required NPSH stated by the pump manufacturer or shown on the pump curves in order to overcome the pump s internal hydraulic loss and the point of lowest pressure in the eye of the impeller. The NPSH required by the pump is a function of the physical dimensions of casing, speed, specific speed, and type of impeller, and must be satisfied for proper pump performance. The pump manufacturer must ahvays be given complete Suction conditions if he is to be expected to recommend a pump to give long and trouble-free service. 

Based on handling pure liquids, without entrained air or other non-condensable gases, which adversely affect the pump performance. 

Absolute pressure at the pump inlet must not be low enough to release non-condensables of (2). If such release can occur, then the NPSHr would need to be increased above that of the cold water requirements to avoid cavitation and poor pump performance. 

Now the pump selected reads NPSHr on its pump performance curve of 12 feet for cold water ser dce. 

Electric motors in pump application never run at the standard rotative design speeds noted above, but rotate at about (with some deviation) 3450, 1750, and 1150 rpm, which are the speeds diat most pump manufacturers use for their performance curves. If the higher numbers were used (motor designated or name plate) for pump performance rating, the pumps would not meet the expected performance, because the motors would not be actually rotating fast enough to provide the characteristic performance curves for the specific size of impeller. 

Example 3-16 Pump Performance Correction For Viscous Liquid... 

When a pump performance is defined for water, the corrected performance for a viscous fluid can be developed using Figure 3-56 or 3-57. In order to develop the curves for viscosity conditions of 100 SSU or 1,000 SSU as shown in Figure 3-58, the following general procedure is used [17]. 

McKelvey, J. M., U. Naire, and F. Haupt, Flow Gear Pumps and Screw Pumps Perform in Polymer-Processing Applications, Chem. Eng., V. 83, No. 20, 1976. 

The pressure p used in Equation 3-32 is the differential developed pressure (across the pump inlet and outlet). Since the inlet suction pressure is usually small compared to the discharge pressure, the discharge pressure is used. Thus, this is the application resistance pressure in most cases. Figure. 3-54 shows a typical reciprocating pump performance. 

Figure 3-54. Typical power-pump performance (courtesy Ingersoll-Rand Co.).

Pump Installation 627. Pump Operation 630. Pump Performance Charts 631. Mud Pump Hydraulics 631. Useful Formulas 645. 

---