CHAPTER FIVE

CHAPTER FIVE INDEX CHAPTER FIVE - PUMPING PLANTS AND SYSTEMS .........................................................1 5.1 Introduction ............................................................................................................................1 5.1.1 General ................................................................................................................1 5.1.2 The Poor Track Record of Pumping Plant in Tanzania ...................................................1 5.2 Pump Classification and Characteristics ................................................................................3 5.2.1 Pump Classification .........................................................................................................3 5.2.1.1 Rotodynamic .......................................................................................................3 5.2.1.1.1 Centrifugal ..........................................................................................................4 5.2.1.1.2 Peripheral Pumps ..............................................................................................12 5.2.1.1.3 Special Pumps ...................................................................................................13 5.2.1.2 Positive displacement pumps ..................................................................................15 5.2.1.2.1 Reciprocating ....................................................................................................15 5.2.1.2.2 Rotary................................................................................................................18 5.2.2.1 Construction of Centrifugal Pumps.........................................................................22 5.2.2.2 Specific Speed .....................................................................................................30 5.2.3 Pump Suction Requirements ..........................................................................................33 5.2.5 Operations of Centrifugal Pumps...................................................................................34 5.2.6 Efficiencies and Fuel Consumption of Rotodynamic Pumps and their Motors.............35 5.3 Prime Movers .......................................................................................................................36 5.3.1 Electric Motors for Pump Drives ...................................................................................36 5.3.1.1 The AC Induction Motor.........................................................................................37 5.3.1.1.1 General ..............................................................................................................37 5.3.1.1.2 Electric Motor Failure Modes ...........................................................................39 5.3.1.1.3 Increasing the Power Factor..............................................................................42 5.3.1.1.4 Power Factor, Installed Capacity and Demand .................................................44 5.3.1.2 Different Starting Methods .....................................................................................46 5.3.1.3 Different Applications.............................................................................................49 5.3.1.4 Speed Control of Pumps with Frequency Converters ............................................53 5.3.2 Diesel Engines and Generators ......................................................................................59 5.4 Pumping Plant Categories ....................................................................................................62 5.4.1 Surface Water Pumping .................................................................................................62 5.4.1.1 Suction Lift Pumping ..............................................................................................62 5.4.1.2 Gravity inflow Source Pumping .............................................................................63 5.4.2 Pumping from an Inaccessible Spring ...........................................................................64 5.4.2.1 General Considerations ...........................................................................................64 5.4.2.2 Electrical power system design for remote control .................................................64 5.4.3 Pumping from Deep Wells.............................................................................................65

5-i

CHAPTER FIVE
5.4.4 5.4.5 5.4.6 5.4.7 5.4.8 5.4.9 Pumping from Shallow Wells ........................................................................................65 Booster Pumping ............................................................................................................66 Pumping Capacity ..........................................................................................................66 Pumping Period ..............................................................................................................66 Pumping Sequence .........................................................................................................66 Pump Arrangement ........................................................................................................67

5.5 Pumping Systems .................................................................................................................67 5.5.1 Single Stage Pumping ....................................................................................................67 5.5.2 Two Stage pumping .......................................................................................................67 5.5.3 Branched Pumping Systems. .........................................................................................67 5.5.4 In Line Boosting ............................................................................................................67 5.5.5 Maximum Suction Lift Calculation ...............................................................................68 5.6 Equipment Type Design and Selection ................................................................................69 5.6.1 Type Design ..............................................................................................................69 5.6.2 Selection ..............................................................................................................70 5.6.3 Metallic Materials of Pump Construction (and their Damage Mechanisms) ................71 5.6.3.1 Types of Corrosion..................................................................................................72 5.6.3.2 Types of Wear .....................................................................................................77 5.6.3.3 Fatigue .....................................................................................................82 5.6.3.4 Materials of Construction........................................................................................83 5.6.3.5 Selection of Materials of Construction ...................................................................86 5.7 Plant Pumping Systems and Equipment Protection .............................................................89 5.7.1 Dry-Running Protection .................................................................................................90 5.7.2 Water Hammer or Surge ................................................................................................90 5.7.2.1 Introduction .....................................................................................................90 5.7.2.2 Initial Water Hammer Analysis Calculation ...........................................................91 5.7.2.3 Subsequent Analysis of a Potentially Dangerous Water Hammer Situation ..........91 5.7.2.4 Flow Separation .....................................................................................................93 5.7.2.5 Pipeline Topography ...............................................................................................93 5.7.2.6 Water Hammer Protection Systems ........................................................................94 5.7.3 Protection Against Cavitation ......................................................................................100 5.7.4 Over Voltage Protection ..............................................................................................102 5.7.5 Cathodic Protection ......................................................................................................105 5.7.6 Overload (O/L), Current and Single Phasing Protection .............................................107 5.8 Erection Operation and Maintenance Instructions .............................................................108 5.8.1 Pumpset Lining up and Final Coupling .......................................................................108 5.8.2 Parallel Alignment .......................................................................................................109 5.8.3 Angular Alignment ......................................................................................................109

CHAPTER FIVE
5.8.4 Routine Maintenance Schedule Pumps .....................................................................110 5.8.5 Pipework Installation ...................................................................................................111 5.8.6 Operating Procedure ....................................................................................................112 5.8.6.1 General ...................................................................................................112 5.8.6.2 Priming ...................................................................................................112 5.8.6.3 Starting ...................................................................................................112 5.8.6.4 Stopping ...................................................................................................113 5.8.7 Maintenance Chart .......................................................................................................113 5.9 Economics of Electrical Power Systems ............................................................................115 5.9.1 Power Factor Correction ..............................................................................................115 5.9.2 Requisite Reactive Power Calculations .......................................................................116 5.9.3 Cable Sizing for the Capacitor Connection .................................................................117 5.10 Energy Source Considerations ...........................................................................................117 5.11 Instrumentation ...................................................................................................................118 5.12 Life Cycle Analysis, an Introduction .................................................................................118 5.13 Formula and Conversion Factors .......................................................................................120

TABLES AND FIGURES Figure 5.1: Radial Flow Impellers Figure 5.2: Mixed Flow Impellers Figure.5.3: Axial Flow Impellers Figure 5.4: Single Suction Close Coupled Pump Figure 5.5: Double Suction Axially Split Pump Figure 5.6: Single Suction Axial Flow Pump Figure 5.7: Section through Single Stage End Suction Centrifugal Pump Figure 5.8: Radially Split Ring Section Multistage Pump Figure 5.9: Radially Split Barrel Casing Type Multistage Pump Figure 5.10: Axially Split Multistage Pump Figure 5.11: Section through a Typical Nine Stage Ring-Section Through-Bolt-Multistage Centrifugal Pump Figure 5.12: Closed - Coupled Pump Figure. 5.13: Long - Coupled Pump Figure 5.14: Turbine Type Pump Figure 5.15: Submersible Pumpset Figure 5.16: Illustration of a Vertical Spindle Borehole Pump Figure 5.17: Electro Submersible Pump Figure 5.18: Peripheral Pumps Figure 5.19: Jet Pump 4 5 5 6 6 6 7 7 8 8 9 9 9 10 10 11 12 12 13

CHAPTER FIVE
Figure 5.20: (a) Shallow Well (b) Deep Well Figure 5.21: Gas Lift/Air Lift Pump Figure 5.22: Horizontal Single-Acting Plunger Pump Figure 5.23: Horizontal Double-Acting Piston Pump Figure 5.24: (a) Diaphragm (b) Horizontal Plunger Figure 5.25: Hydraulic Actuated Diaphragm Figure 5.26: Generalized Form of Rotary Pump Performance Figure 5.27: Sliding Vane Pump Figure 5.28: External Vane Pump Figure 5.29: Axial Piston Pump Figure 5.30: (a) Flexible Vane Pump (b) Flexible Tube Pump Figure 5.31: Single Screw Pump (progressing cavity) Figure 5.32: External Gear Pump Figure 5.33: Internal Gear Pump (with crescent) Figure 5.34: (a) Single Lobe Pump (b) Three-Lobe Pump Figure 5.35: Circumferential Piston Pump Figure 5.36: Multiple (three) Screw Pump Figure 5.37: Impellers Figure 5.38: Characteristic Curve for Radial Flow Centrifugal Pump Figure 5.39: Typical Curves for (a) Axuial and (b) Mixed Flow Pumps Figure 5.40: Casing (a) Volute Type (b) Diffuser Type Figure 5.41: (a) Single Volute Casing (b) Double Volute Casing Figure 5.42: (a) Multi-Vane Radial Diffuser (b) Axial Diffuser Figure 5.43: (a) Axial Suction Thtust and (b) Hydraulic Unbalanced Thrust Figure 5.43: (c) Thrust Balance Method Figure 5.44: Hydraulic Axial Balance with Double Inlet Impeller Figure 5.45: Axial Balance Typical Balance Disc Arrangement Figure 5.46: Axial Balance Typical Balance Drum Arrangement Figure 5.47: Axial Balance Thrust Figure 5.48 (a) 100% BEP Flow and (b) Reduced Flow Figure 5.49 (a) Radial Thrust and (b) Double Volute Figure 5.50: Units of Specific Speed Figure 5.51: Relationship between Pump Efficiency and Specific Speed Figure 5.52: Preferred Arrangement of Pumping Stations Table 5.1: Pump Type, Characteristics and Applicability Figure 5.53: Effect of Changing Speed of Pump Table 5.2: Comparative Efficiencies for Different Pumps and Motors Table 5.3: Overall Fuel Consumption Table 5.4: Degrees of Displacement and Power factor Table 5.5: Power Factor Surcharges Figure 5.54: Direct-On-Line Start 13 14 16 16 17 17 18 19 19 19 20 20 20 21 21 21 21 22 23 23 25 25 26 26 27 27 27 28 28 29 30 31 32 33 34 34 35 36 45 45 46

CHAPTER FIVE
Figure 5.55: Star-delta Start Figure 5.56: Frequency Converter Figure 5.57: Soft starter Figure 5.58: (a-d) Different Applications Figure 5.59: Speed Control of Pumps with Frequency Converters Figure 5.60: Altitude De-rating for a Naturally Aspirated Diesel Engine Figure 5.61: Temperature and Humidity De-rating for a Diesel Engine Table 5.6: Altitude and Temperature Values (Factor B) Figure 5.62: Headloss through Strainer and Footvalve Figure 5.63: Cavitation Erosion of an Impeller Figure 5.64: The Fretting Damage of a Shaft beneath an Impeller that Experience Small Amplitude Motion Figure 5.65:The three-body abrasive wear of a laser-hardened shaft sleeve in an abrasive service Figure 5.66: (a-d) Examples of Erosion Figure 5.67: Fatique fractures Table 5.7: Calculated Wear Factor Table 5.8: Material Selection Chart for Volute Casing Pumps Table 5.9: Material Selection Chart for Wet-Pit Diffuser Pumps Table 5.10: Material Selection Chart for Reciprocating Pumps Table 5.11: Combination to be Avoided When Area of Metal Considered Is Small Relative to Area of Coupled Metal Table 5.12: Combination to be Avoided when Area of Metal Considered Equal to Area of Coupled Metal Figure 5.68: Superimposition of Surge on a Pipeline Profile Figure 5.69: Pipeline Profile Illustrating Suitable Locations for Surge Protection Devices Figure 5.70: Vortex Prevention Assessment Figure 5.71: Sacrificial Anodes for Cathodic Protection Table 5.13: Typical Maintenance Chart Table 5.14: TANESCO Tariff (2008) Figure 5.72: Typical Overall Pumpset Life Cycle Cost Breakdown 47 48 49 50 53 61 62 69 69 76 78 80 81 83 86 87 88 89 89 90 93 95 102 107 114 116 120

CHAPTER FIVE
CHAPTER FIVE - PUMPING PLANTS AND SYSTEMS 5.1 INTRODUCTION

Revisions to the 1997 edition of this Chapter are generally minor except for the addition of a new Section on Pump Classification and Characteristics providing a discussion on Prime Movers in general and an introductory descriptive and illustrative evaluation of different types of Pumps; and new Sections providing an overview of materials of pump construction; and an introduction to Life Cycle Assessment of Pumping Plant. 5.1.1 General The main purpose of any pumping plant and pumping system is to lift water from a lower to a higher level. The laws of physics dictate the minimum power requirements for lifting a given mass of water through a given distance in a given time. The challenge for the pump manufacturer, system designer, and subsequently the operator is to ensure that this is done as reliably, and efficiently as possible with the minimum of energy consumption. The running and the economy of a water production line relies mainly on the success of following project planning sub-components: 5.1.2 Intake and Plant Design Pumping System Design Equipment Type Design Equipment Selection Plant, Pumping System and Equipment Protection Accuracy and comprehensiveness of Erection, Operation and Maintenance Instructions Economics of Electrical Power Systems or other power systems Energy Considerations Compliance with Instructions Observation of the Factory Ordinance The Poor Track Record of Pumping Plant in Tanzania

The recent history of Pumping Plant in this country is far from satisfactory. It has shown that smooth and economic running of most of the existing water schemes including those designed by foreign consulting firms has been impaired by deficiencies in some if not most of the areas relating to pumping systems cited above. This has in part been due to the fact that the previous edition of this Design Manual on the Electrical and Mechanical part of a water scheme design has either been ignored or inadequately followed. Hence most existing designs do not exhibit a sufficiently high degree of technical soundness in this respect. The situation has been exacerbated by the failure to effectively ensure the quality of manufacture, the competence of installation contractors and the appropriate training and provision of competent and dedicated system operators. The economic life time of water pumping equipment under normal operation and maintenance practices is at least 10 and can be 15 years. In this country however, it is not uncommon to need
Page 5-1

CHAPTER FIVE
pumps replacing much earlier following premature and permanent breakdowns, even after 5 years of operation or less. Reasons for premature breakdown include cavitation, incorrect or inadequate design, selection and pump specification, incorrect erection, silt/chemical erosion, lack of effective protection, poor operation and maintenance, etc. Furthermore water supplies are frequently interrupted due to lack of realistic principles of investment which results in cheap pumping equipment being selected, lack of adequate standby pumping capacities and systematic operating sequences. An analysis of the enumerated failures has amongst other things shown that:1. Silt and chemical erosion due primarily to poor source selection and location and poor intake design as well as to the use of clear water pumps that are not designed for such conditions. Cavitation has often been due to poor sump-pump layout which results in turbulent flow and allows air entrainment in the pump; further to this the use of complicated suction pipe lines which makes it virtually impossible to effect perfect pump priming processes (displacement of all the air in the suction line by running the pumped medium into it) and this results in poor pumping system as the pump is subjected to objectionably low net positive suction head (NPSH). Lack of sound protection has forced the water authorities to spend, untimely, millions, of shillings on rewinding or replacement motors following burn-outs due to overloads, phase failures, under-voltage, lightning and switching surges. Also lack of hydraulic, mechanical and electrical protection has exposed plants to damage and operators to accidents. Poor erection has contributed greatly towards severe and often irreparable damage to pumps and drives (prime movers), equipment shaft breakages, and burning of electric motors. It should be noted in this connection that improper equipment alignment, wrong clearances at couplings, force fitting of foundation base plates, and pipes especially at pump branches can cause excessive stress in the respective items. In such cases breakages of couplings, shafts, bearing and foundation failures, etc. are bound to occur. 5. Poor construction/installation materials or inappropriate sizes of the same have been quite predominant problems; evidently, this is because the technical know-how in this regard has not been effectively propagated to those concerned. This has also caused a lot of damage to motor-windings, foundations and drives as well as reduced expected plant outputs and life. Equipment and material handling and storage is yet another area that needs to be given due consideration at the project planning stage. Failure to do so can result into premature failures of plant and equipment and therefore affect the planned project economy. The degree of protection being specified and provided presently leaves much to be desired. Electric motor installations in areas prone to lightning especially along the shores of Lake Victoria are not provided with lightning arrest6rs. Hence motor windings have been and are still being burnt-out due to lightning surges. Failure to recognise that electricity supply, nominally of 400V, 50Hz or 11kV, 50Hz is often intermittent and below those values.
Page 5-2

2.

3.

4.

6.

7.

8.

CHAPTER FIVE
9. Poor operation and maintenance has also caused a number of schemes to stop running due to frequent breakdowns including failure to maintain adequate stocks of spare parts. It is thus necessary that with a view to minimising these problems, designers should observe the following factors as being among the important design criteria apart from ensuring that the operation and maintenance of any scheme is planned for from the design stage. i. ii. iii. iv. Technical level of education of plant operator and maintenance staff; Provision of clear and planned maintenance facilities; Comprehensive lists of likely number and timing for spare parts; and Due care and attention to the economics of project running.

10. It is rare to find factory ordinances being observed. Numerous pumping sets can presently be found without coupling guards, and some operators have been and are still likely to fall victims of even fatal accidents at pumping stations. It is important therefore to draw the attention of plant managers and operators to some clearly defined factory regulations. Even more important is the need to observe such regulations in design so as to avoid loss of human resources. 5.2 5.2.1 PUMP CLASSIFICATION AND CHARACTERISTICS Pump Classification

Although there are a wide range of pumps available for numerous applications they generally fall into two major groups. These are: 5.2.1.1 Rotodynamic Pumps. Positive Displacement Pumps. Rotodynamic

Rotodynamic pumps are essentially rotary machines in which energy is continuously imparted to the pumped liquid by a rotating impeller or rotor and consist of three types: Centrifugal, Peripheral and Special Pumps as indicated in the diagram below:

Page 5-3

CHAPTER FIVE
5.2.1.1.1 Centrifugal
This group of pumps consists of a shaft mounted impeller(s) rotating unidirectionally within a casing. The liquid enters the impeller eye and acquires energy in the form of velocity as it passes though the impeller passages. The velocity head is converted into pressure head by the volute or spiral shaped outer casing of the pump which directs the liquid from the outer perimeter of the impeller to the pump discharge. A less common method of developing pressure head is to surround the impeller with concentric diffusing passages. A well designed centrifugal pump, suitable for purpose, should achieve around 80% efficiency with not more than 20% losses in energy conversion, bearing friction and hydraulic losses. However, actual efficiency depends on several factors including size and flow rate, with efficiency generally increasing with size. It must be recognised that efficiency drops with time due to wear and tear and the all too frequent use of design flow rates to predict actual output needs to be actively discouraged as it is increasingly misleading with age. Actual operating efficiency determined from measured output and power consumed is a useful criterion in judging whether it is time for refurbishment or replacement. A major problem with all centrifugal pumps is how best to accommodate end thrust on the shaft. The pressure of the water at the inlet side of the impellor is low relative to that on the back of the impellor beyond the outlet, which is almost equal in surface area, so that an axial thrust is imparted. Unless this is balanced in some way this would quickly cause wear on the pump and shorten its life. In small pumps this end thrust is dealt with by the use of purpose-designed thrust bearings. For larger pumps, a double entry, back-to-back impellor design may be used. Centrifugal pumps can be divided into two main groups depending on the type of impeller. (a) Radial Flow Impellers. The liquid enters the impeller axially and discharges radially, in effect changing its direction by 90o. In this case the head developed is due to the centrifugal force exerted on the fluid by the impeller.

FIGURE 5.1: RADIAL FLOW IMPELLERS

Page 5-4

CHAPTER FIVE
(b) Mixed Flow Impellers. The liquid enters the impeller axially and discharges in both axial and radial directions. In this case the head developed is the result of a combination of the centrifugal force and the lift produced by the vanes on the liquid.

FIGURE 5.2: MIXED FLOW IMPELLERS

(c) Axial Flow Impellers. The liquid enters leaves the impeller in an axial direction. In this case the head developed is entirely due to the lift produced by the vanes on the liquid.

FIGURE.5.3: AXIAL FLOW IMPELLERS

Radial and mixed flow impellers can be subdivided into: (i). Single suction the liquid enters the impeller from one side only. (Fig. 5.4)

(ii). Double suction the liquid enters the impeller from both sides. (Fig. 5.5) (iii). Axial flow impellers which are of the single suction type. (Fig. 5.6)

Page 5-5

CHAPTER FIVE

FIGURE 5.4: SINGLE SUCTION CLOSE COUPLED PUMP

FIGURE 5.5: DOUBLE SUCTION AXIALLY SPLIT PUMP

FIGURE 5.6: SINGLE SUCTION AXIAL FLOW PUMP

Centrifugal pumps, irrespective of impeller type can be either singe stage, whereby the total head is the result of one impeller, or can be multistage. Standard, above ground centrifugal pumpsets will either be horizontally or vertically mounted with the prime mover either at one end or immediately above the pump. For general waterworks purposes, the maximum pressure normally developed by a single stage pump will be 80 -100 m so that for heads greater than this a multi-stage pump is usually required although increased head can also be achieved by increased speed and/or larger impellors. However, and as a general rule, the lower the speed, the longer the life of the pump. With single stage horizontal pumps, an end entry suction with side or top outlet is offered by some manufacturers whilst for vertically mounted units and for multistage pumps either side or top entry and exit ports are necessary. An end entry, single stage pump is illustrated in Figure 5.7 and a multi-stage pump in Figure 5.11.

Page 5-6

CHAPTER FIVE
outlet port

inlet port shaft for coupling to prime mover

FIGURE 5.7: SECTION THROUGH SINGLE STAGE END SUCTION CENTRIFUGAL PUMP

Centrifugal pumps can be manufactured with a split casing, especially in larger sizes. A great advantage with this type of pump is that the upper half of the casing is easy to remove giving access to the diffuser chamber and rotating element with the impellor and bearings visible for inspection and maintenance without the need to disconnect either pipework or the prime mover. In addition, new or refurbished rotating elements can be kept in stock enabling a speedy removal and replacement with such things as bearing replacement or impellor rehabilitation then being undertaken at a more leisurely and careful pace without the pressure of minimising downtime being present. In multistage units the total head is the end result of a series of impellers within the one casing. Under these conditions the required head is achieved by the summation of the heads developed by each individual impeller. Multistage unit casings can be either axially split (Fig. 5.10) or radially split. Radially split pumps can consist of a series of ring sections fastened together by external tie bolts (Fig. 5.8); alternatively for higher pressures or to cater for thermal shock, the ring sections can be encased in a barrel casing which also contains the suction and discharge branches (Fig. 5.9). Multistage pumps can be either horizontally or vertically disposed.

FIGURE 5.8: RADIALLY SPLIT RING SECTION MULTISTAGE PUMP

Page 5-7

CHAPTER FIVE

FIGURE 5.9: RADIALLY SPLIT BARREL CASING TYPE MULTISTAGE PUMP

FIGURE 5.10: AXIALLY SPLIT MULTISTAGE PUMP

An advantage of multistage pumps is that it is practicable to initially incorporate one or more dummy stages, a dummy stage simply being a diffuser without an impeller. Then as demand increases, impellers can be added to increase delivery pressure and hence flow. Pump efficiency is not much affected by the dummy stages, however the additional impellors must be considered when ensuring the prime mover will have enough power to drive the pump after they are added. A section through a typical multi-stage pump is illustrated in Figure 5.11.

Page 5-8

CHAPTER FIVE
inlet port outlet port

end thrust bearing

coupling to prime mover

FIGURE 5.11: SECTION THROUGH A TYPICAL NINE STAGE RING-SECTION THROUGH-BOLT-MULTISTAGE CENTRIFUGAL PUMP

Centrifugal pumps can be either self priming or non self priming and can have open, semi open or closed impellers depending on the specific requirements of the particular pump. Centrifugal pumps can also be identified by their basic mechanical configuration and characteristics:Overhung Impeller Type the impeller(s) is mounted on the end of the shaft which is cantilevered from its bearing supports. In addition, this type of pump can either be of the close coupled design, in which the pump casing is fixed directly to the driver frame and the impeller is mounted on the driver shaft (Fig 5.4 & Fig. 5.12) or long coupled, where the pump is mounted on a base plate and driven through a coupling. (Fig 5.13)

FIGURE 5.12: CLOSED - COUPLED PUMP

FIGURE. 5.13: LONG - COUPLED PUMP

Impeller Between Bearings Type the impeller(s) is mounted on a shaft between bearings situated at both ends.

Page 5-9

CHAPTER FIVE

Turbine Type/Submerged Centrifugal Pumps this rather ambiguous term usually applies to vertical multistage deep well pumps that are constructed with diffuser casings screwed or bolted together. The bearings are lubricated, cooled and flushed by the pumped liquid (Fig 5.14). A special type of turbine pump is the electro submersible. In these deep well units, the motor is close coupled to the pump and submerged in the well. This type of pump is used for high head applications where long intermediate shafts are undesirable

FIGURE 5.14: TURBINE TYPE PUMP

There are two types of submerged centrifugal pumps, the fully submerged or submersible pumpset and the vertical spindle pump. rising In the submersible pumpset, both the pump and its motor are submerged such that the term submersible pump is now synonymous with that of the submersible pumpset. When designed as long narrow units they are especially well suited as borehole pumps being considerably cheaper than vertical spindle pumps. The motor is located in a shroud subsequent to the pump with the water flowing around it so as to cool it. Power is supplied to the motor through waterproof cables. They also tend to be twopole driven enabling them to run at the highest available speed. There major disadvantage is that the motor is out of sight of any operator and easy to forget until it fails. Such failures are often the result of failure of the motor-cable water seals and the motor can burn-out under such circumstances. On the other hand they are reasonably quick and easy to install and remove as there is no need for a drive shaft or spindle. Nor do they have to be truly vertical in their positioning. Similarly and providing there is access at the wellhead for a mobile crane or space for shear legs, no surface structure is needed other than for the switchgear which can be mounted in a simple waterproofed housing.
main

inlet port

FIGURE 5.15: SUBMERSIBLE PUMPSET

Page 5-10

CHAPTER FIVE
They do however tend to be less efficient than the vertical spindle design because of the special design of the motor and because they usually require a larger number of stages due to the relatively small impellor diameter. Specially designed submersible pumpsets can be mounted horizontally and are then often used as in-line booster pumps. The vertical spindle pump is one where the prime mover is at the surface, mounted above flood water level in the case of river intakes, and the pump is immersed in the water. The connecting shaft or spindle rotates within the riser pipe, often within a tube or sleeve held centrally to the rising main by spider bearings which allow the water to pass through the annular space between the shaft or its containing sleeve and the rising main. Such rising mains and their contents are often in 3 m lengths allowing for slow but progressive removal and replacement. Such removal and especially replacement is a highly skilled task as true alignment is critical to avoid excessive wear and tear or even failure. An example is shown below:

FIGURE 5.16: ILLUSTRATION OF A VERTICAL SPINDLE BOREHOLE PUMP

Page 5-11

CHAPTER FIVE
In addition to the centrifugal pumps described previously, there are other types which have their own unique characteristics: Single stage electro-submersible pumps, sump pumps, glandless and canned motor pumps, sewage pumps and abrasion resistant pumps.

FIGURE 5.17: ELECTRO SUBMERSIBLE PUMP

5.2.1.1.2

Peripheral Pumps

In a peripheral pump, energy is imparted to the fluid within the cells of a vane wheel impeller. Alternatively, the cells are arranged peripherally on the outer sides of a wheel disc. Peripheral pumps come in various configurations, such as overhung impeller, impeller(s), mounted between bearings, single stage and multistage. Alternative names are side channel pump and regenerative turbine pump.

FIGURE 5.18: PERIPHERAL PUMPS

Page 5-12

CHAPTER FIVE
5.2.1.1.3
(a)

Special Pumps
Jet Pump

The term jet pump describes a pump in which a high velocity jet of fluid is utilized to create a low pressure area in a mixing chamber, causing the suction fluid to flow into this chamber. The venturi is positioned so that gradual velocity conversion occurs with minimum losses. (Fig. 5.19)

FIGURE 5.19: JET PUMP

When the jet/venturi are attached directly to the pump ahead of the suction impeller and activated by the liquid from the pump it is termed a shallow well jet pump. When the jet/venturi are set at the base of an extended suction pipe and activated by the liquid from the pump it is termed a deep well jet pump.

(a)
FIGURE 5.20: (a) SHALLOW WELL (b) DEEP WELL

(b)

Page 5-13

CHAPTER FIVE
The jet pump or ejector pump, works on the principle of ejecting a small stream of high pressure water to entrain a larger volume of low pressure water and force it to a higher elevation. Its application is normally restricted to shallow wells in combination with a single stage centrifugal pump located at the wellhead. Here a proportion of the pump discharge is directed down to the bottom of the well and through an ejector to promote the flow up the suction pipe. Lifts of up to 75 m can be realized by this method, although, in general, the higher the lift, the lower the efficiency of the pump. (b) Gas Lift/Air Lift pumps. A gas pressure source is used to lift or pump the liquid handled by mixing it with gas under pressure, usually compressed air. An air-lift pump is commonly used to test boreholes. It is based on pressurising the water surface with compressed air to force it up the delivery pipe. Lift is directly proportional to the pressure developed on the surface, i.e. the head of equivalent less friction losses in the pipe. The system involves no moving parts in the well but requires that the well be sealed above the water level to maintain the surface pressure. Such types of pump usually have a low efficiency.

FIGURE 5.21: GAS LIFT/AIR LIFT PUMP

(c)

Electro magnetic pump An electro-magnetic pump is a glandless pump without rotating parts in which a magnetic field acts upon a susceptible medium (liquid metals in principle).

Page 5-14

CHAPTER FIVE
5.2.1.2 Positive displacement pumps Positive displacement pumps are essentially rotary or reciprocating machines in which energy is periodically added by application of force to movable boundaries of enclosed fluid containing volumes, resulting in a direct increase in pressure. The various types are illustrated on the diagram blow:

5.2.1.2.1

Reciprocating

Most reciprocating pumps are either suction or lift pumps, both functioning as free delivery or force pumps, and working as single-acting or double-acting pumps, and are used in wells and boreholes. They can be manually (hand or foot) driven. In a suction pump, the plunger and its cylinder are located above the water level, usually within the pump stand itself. A suction pump relies on atmospheric pressure so that it is in practical terms limited to 5 to 8 metres primarily dependant upon elevation. (See section 5.2.5 on NPSH.) In a lift pump, the term deep or shallow refers to the depth of water in the well and not to its total depth. The pump element comprising cylinder and plunger is located below the water level and can lift water from depths as great as 180 m. They can either be driven by a lever action or for greater depths by a rotating hand wheel or pair of such mounted on either side of the headstock. For more information on a handwheel operated type of reciprocating pump, designers are advised to visit the website www.duba.com. A variation on the reciprocating well-pump is the hydrostat or hydraulic ram pump used for lifting surface water whereby a large volume of water flowing in one pipe or waterway is used to drive a ram which is connected to a smaller pump which pumps part of the water to a higher elevation through a branch pipe, with the majority of the flow continuing down the main pipe or discharging from the pump. Their use tends to be restricted to the supply of small volumes of water from steep, fast flowing streams or rivers in remote locations without an accessible power supply to hillside communities above.

Page 5-15

CHAPTER FIVE
1. Piston, Plunger or Diaphragm A reciprocating pump set generally consists of a piston / plunger (or diaphragm) inlet and outlet valves and a means by which the piston / plunger can be actuated. This is usually done by a reciprocating engine or an electric motor / steam turbine, crank and connection rod combination. These pumps can be either. (a) Single acting in which the liquid is discharged during the forward motion of the piston. (b) Double acting in which the liquid is discharged during both the forward and backward motions of the piston. Reciprocating pumps can be further classified as:(i) Simplex pumps contain one single or double acting piston / plunger. (ii) Duplex pumps contain two single or double acting pistons / plungers. (iii) Triplex pumps contain three single or double acting pistons / plungers. (iv) Multiplex pumps contain more than three single or double acting pistons / plungers.

FIGURE 5.22: HORIZONTAL SINGLE-ACTING DOUBLE-ACTING

FIGURE 5.23: HORIZONTAL PLUNGER PUMP PISTON PUMP

(a)

Piston / Plunger Type

The piston / plunger pump offers a positive means of metering a wide variety of fluids by utilising a range of gland packing materials and long life plungers. With this type of pump the flow rate accuracy and discharge pressure greatly exceed that of the mechanical diaphragm type. The wet end consists of a piston / plunger reciprocating within a cylindrical housing. (b) Mechanical Diaphragm Type

With glandless construction and simple wet end designs, the mechanical diaphragm pump offers the advantage of a positively sealed pumping chamber, in which the risks of corrosion and erosion are negligible.

Page 5-16

CHAPTER FIVE

FIGURE 5.24: (a) DIAPHRAGM (b) HORIZONTAL PLUNGER

(c)

Hydraulically Actuated Diaphragm Type This type of pump combines the glandless construction of the mechanical diaphragm pump and the accuracy, repeatability and high pressure capability of the piston / plunger pump. The diaphragm is hydraulically coupled to the piston or plunger. This type of pump can incorporate diaphragm rupture detection devices and various hydraulic fluids which are compatible with the fluid being pumped.

CONTROLLED VOLUME PUMP

PISTON DIAPHRAGM LIQUID END

FIGURE 5.25: HYDRAULIC ACTUATED DIAPHRAGM

2.

Controlled Volume Pump Another type of reciprocating positive displacement pump is the controlled volume pump which provides precision control of very low flow rates up to a maximum of about 3 l/s. Flow rate accuracy is typically within 1%. Other names for these pumps are proportioning pumps and metering pumps. Controlled volume pumps are generally available in three construction styles, piston or plunger; mechanical diaphragm and hydraulically actuated diaphragm type (piston diaphragm). Usually, the driver is an electric motor. Basically, the design criteria that applies to large motor driven reciprocating pumps also applies to controlled volume pumps. Flow rate variations are normally achieved by manual resetting of the stroke length. Automatic controls are available for stroke length resetting and motor speed. The typical efficiency of this type of pump is around 90o%.

Page 5-17

CHAPTER FIVE
5.2.1.2.2 Rotary
There are a number of different varieties of rotary pumps but in all, the essential working element is a rotor which can take a number of different forms, such as an impellor, vane or screw or a combination of these. Such pumps differ from reciprocating pumps in that the delivery is continuous and hence smoother. However, internal losses are normally somewhat higher through slip (internal leak-back). Slip increases with increasing pressure so rotary pumps are less suited to high pressure systems. Rotary pumps are physically much larger than centrifugal pumps for equivalent capacities and therefore occupy greater space.

The water output of a rotary pump is almost proportional to the rotating speed as indicated in Fig 5.26. One such pump is the helical rotary (Mono) pump which consists of a single thread helical rotor which rotates inside a double thread helical sleeve or stator, giving it the alternative names of the eccentric screw or progressive cavity pump. Both diesel engines and electrical motors are suitable for driving this type of pump, whilst some smaller units are also capable of hand operation.

Head

P = power input

P1

P2
Quantity

P3

P4

P5 P6

FIGURE 5.26: GENERALIZED FORM OF ROTARY PUMP PERFORMANCE

Being a positive displacement pump, the helical rotor pump has a head discharge characteristic that is almost vertical to the capacity and for each power input are parallel to each other. For more specific information of the commonest make of this pump, designers should refer to the web site www.mono.com. Rotary pumps generally consist of gears, screws, vanes or similar elements enclosed within a casing. They have no separate inlet or outlet valves and the liquid flows through the pump in a uniform stream as a result of the rotation of elements. These pumps are characterised by their close running tolerances. Rotary pumps can be divided into two general groups: 1. Single Rotor, and 2. Multiple Rotors.

Page 5-18

CHAPTER FIVE
1. Single rotor types (i) Vane The Vane of various forms, including blades, buckets, rollers, or slippers are radially displaced inwards or outwards by a cam which has the effect of drawing liquid into and then out of the casing. There are two types of vane pumps:

FIGURE 5.27: SLIDING VANE PUMP

FIGURE 5.28: EXTERNAL VANE PUMP

(ii)

Piston

The rotor contains cylinders in which piston like elements reciprocate. As the rotor turns, so the movement of the pistons draws the fluid into the casing, the rotation of the cylinder of the piston and cylinder being relative to the parts. There are two variations of this type of pump. In one variation, the pistons reciprocate in an axial direction and in the other the reciprocating action of the piston is in a radial direction.

FIGURE 5.29: AXIAL PISTON PUMP

(iii)

Flexible Member

The flexible member, being a vane, tube or liner is of sufficient elasticity to accomplish sealing and transfer the fluid within the casing.

Page 5-19

CHAPTER FIVE

(a)
FIGURE 5.30: (a) FLEXIBLE VANE PUMP

(b)
(b) FLEXIBLE TUBE PUMP

(iv)

Screw Progressive Cavity Pump

The progressive cavity or helical rotor pump consists of a resilient stator in the form of a double internal helix and a single helical rotor. The rotor maintains a constant seal across the stator and this seal travels continuously through the pump giving uniform positive displacement. The single helical rotor rolls in the stator with an eccentric motion. 2 (i) MULTIPLE ROTOR TYPES Gear Pump The meshing of two or more gears provides the pumping action. This meshing of gears also forms part of the moving fluid seal between the inlet and outlet ports. Gear pumps can be either:(a) External gear in which all gear teeth are cut externally and may be of spur, helical or herringbone tooth pattern.

FIGURE 5.31: SINGLE SCREW PUMP (PROGRESSING CAVITY)

FIGURE 5.32: EXTERNAL GEAR PUMP

Page 5-20

CHAPTER FIVE
(b). Internal gear in which one of the rotors has teeth cut internally and the other has externally cut teeth.

FIGURE 5.33: INTERNAL GEAR PUMP (WITH CRESCENT)

(ii)

Lobe Pump

The rotors consist of one or more lobes, the interaction of which transports the fluid from the inlet to outlet. As in the gear pumps, the shape of the rotors and their operations provided part of the fluid seal.

(a) (iii). Circumferential Piston The liquid is transported in the spaces between the piston surface to the outlet. The operation is similar to that of a lobe pump; however, a fundamental difference is that the rotors do not form any type of fluid seal.

(b)

FIGURE 5.34: (a) SINGLE LOBE PUMP (b) THREE-LOBE PUMP)

Circumferential piston pumps can either be external or internal. The internal version must have two or more piston elements and has no need for timing gears, whereas the external type may have one or more pistons and FIGURE 5.35: CIRCUMFERENTIAL PISTON requires timing gears. PUMP

Page 5-21

CHAPTER FIVE
(iv). Multiple Screw Usually the screw type rotors in this type of pump cannot drive each other and timing gears are required. The principle of operations is similar to that of the single screw pump.
FIGURE 5.36: THREE SCREW PUMP

5.2.2

Characteristics of Centrifugal Pumps

5.2.2.1 Construction of Centrifugal Pumps A centrifugal pump is a machine which moves liquid by accelerating it radially outward in a rotating impeller to a surrounding stationary housing or casing. It has two main parts; 1) a rotating element consisting of a impeller mounted on a shaft. 2) a stationary element consisting a casing, staffing box and bearings. Impeller The rotary motion of impeller imparts velocity energy to the liquid some of which is converted into pressure within the impeller passages. Impellers may be classified as radial, axial or mixed flow, depending on design.

FIGURE 5.37: IMPELLERS

(1)

Radial Flow The radial flow impeller discharges the fluid radially at 90o to the shaft axis. The characteristic curve for a typical, true, radial flow centrifugal pump is shown in Figure 5.38. The head-flow curve is relatively flat up to the design point and the power at zero flow is only about 40% of that required at design duty. As a result it can be started against a closed delivery valve which can also be closed before the pump is shut down. No problems then occur provided the valve is opened before over-heating occurs. This reduced power requirement at start up is also beneficial as it reduces the starting current when an electric motor is used as the prime mover, whilst the delivery pressure rise can be carefully controlled by slowly opening the valve thus reducing pressure transients. Unfortunately such transients are difficult to avoid when stoppage results from an unexpected power outage.

Page 5-22

CHAPTER FIVE
Whilst the maximum head developed for a given speed is not greatly in excess of the design head, the shape of the head-output curve is important. In the illustration shown here an unsatisfactory situation occurs which should be avoided because at heads higher than the duty head there are two possible outputs. The pump is therefore unstable when operating in this region and this can be particularly troublesome when two or more pumps are being operated in parallel.

FIGURE 5.38: CHARACTERISTIC CURVE FOR RADIAL FLOW CENTRIFUGAL PUMP

Designers are advised to carefully check the head-output curve of any envisaged pump and/or to clearly specify that only units with a continuously falling head are acceptable. In addition, the maximum head at zero flow, the closed valve head should not be excessive and the efficiency curve should be reasonably flat about the design duty point so that there is no great reduction in efficiency if the actual head is slightly different from that expected. (2) Axial Flow The axial flow impeller discharges fluid along the shaft axis in which the rotation of the impeller forces the water forward axially. For this reason an axial flow pump is by definition not centrifugal in its pumping action. (3) Mixed Flow Mixed flow pumps act partly by centrifugal action and partly by propeller action, the blades of the impeller being given a degree of twist. However, in practical terms, there are no precise dividing lines between radial flow (centrifugal), mixed flow, and axial flow pumps. The mixed flow impeller discharges fluid in a conical direction using a combined radial and axial pumping action as suggested by the title. FIGURE 39: TYPICAL CURVES FOR (a) AXIAL AND (b) MIXED FLOW PUMPS

Page 5-23

CHAPTER FIVE
They act partly by centrifugal action and partly by propeller action, the blades of the impeller being given a degree of twist. Characteristic curves are shown in Figure 39.However, in practical terms, there are no precise dividing lines between radial flow (centrifugal), mixed flow, and axial flow pumps. In general, axial and mixed flow pumps are primarily suited for pumping large quantities of water against low heads, whilst centrifugal pumps are best suited for pumping moderate outputs of water against high heads. Axial flow pumps in particular have poor suction capability and must be submerged for starting. They are best suited for transferring large quantities of water from a river to some nearby ground-level storage. The starting power required by a mixed flow pump is much the same as the duty power, but for axial flow pumps, the starting power is substantially in excess of the duty power. Axial flow pumps are therefore not started against a closed valve which would overload a motor correctly sized for the expected duty. It should be noted that even the smallest of pumps may be mixed flow rather than truly centrifugal, due to specific requirements or space restrictions and then mixed flow multistage pumps are deployed. Vertical Trunk Slung (VTS) or Wet Well Pumps are often such axial flow pumps that are supported from above with the drive provided by a rotating shaft within the riser part of the pumpset assembly. This rotating shaft is supported by bearings that may be contained in a separate sleeve so that they can be provided with clean water for lubrication, although some special material bearings are now manufactured which, it is claimed, need no special lubrication. Designers are however advised to seek good references for instances in which such special material bearings have been successfully used in situations similar to those proposed before opting for such. All wet well vertical spindle pumps have a number of disadvantages however including the need for a relatively lengthy down-time to remove / maintain / re-install the impeller and the bearings and require very accurate and precise realignment upon re-assembly if either rapid wear or even bearing failure is to be avoided. They are also purpose built units so are rarely available off-the-shelf and are also expensive and in general, it can be recommended that less expensive dry well axial flow pumps or centrifugal pumps are used instead as indicated in Figure 5.39 above, notwithstanding that the civil engineering costs are usually higher than the difference in price between the two types of pumps. However and when taking into account the key role played by pumping in other than gravity water supplies, the additional cost that may be involved is considered to be well worth it. (4) Inducers These are fitted on the suction side of first stage impellers. They are basically axial flow impellers with extended vanes. This enables them to operate in a cavitating condition with only a small head drop since the vapour cavity occupies only a relatively small length of passage. It thus generates enough head whilst cavitating to enable the first impeller to operate without cavitation.

Page 5-24

CHAPTER FIVE
Casing The velocity head of water leaving the impeller is converted into pressure head in the casing either by means of a volute (Fig. 5.40a) or by a set of stationary diffuser vanes surrounding the impeller periphery (Fig. 5.40b)

(a) (1) Volute Casing

(b)

FIGURE 5.40: CASING (a) VOLUTE TYPE (b) DIFFUSER TYPE

The volute is the most common form of casing (Fig. 5.41a). The volute increases in cross sectional area from the cut water to the inner end of the discharge cone so as to give a near constant average water velocity. Most of the conversion from velocity head to pressure head takes place in the discharge cone. However uneven pressure distributions around the impeller may give rise to undesirable radial loading on the shaft, particularly when operating at reduced flows. The double volute design (Fig. 5.41b) reduces this radial loading.

FIGURE 5.41: (a) SINGLE VOLUTE CASING (b) DOUBLE VOLUTE CASING

(2)

Diffuser The diffuser fits inside the pump casing and guides the flow smoothly into the discharge pipe (or the next impeller in the case of a multi stage pump).

Page 5-25

CHAPTER FIVE
The diffuser incorporates a number of vanes which form radially diverging water passages around the periphery of the impeller and recover a significant percentage of the total head. Most diffusers have radial vanes. Axial diffusers are used when it is desirable to reduce outside diameter of the casing or to increase the clearance between the impeller and diffuser vanes in order to reduce vane tip erosion in very high speed pumps.

(a) Shaft & Bearings

(b)

FIGURE 5.42: (a) MULTI-VANE RADIAL DIFFUSER (b) AXIAL DIFFUSER

The primary function of the shaft is to transmit the driving torque to the impeller. With the support of the bearings, the shaft must also locate the impeller radially and axially within the casing. Single stage overhung impeller pumps normally operate in the stiff shaft mode i.e. below the first shaft critical speed. Multistage pumps may often operate in the flexible shaft mode, running above the first critical speed. With these pumps the hydrodynamic support and damping afforded by the internal clearances normally guarantees satisfactory wet operation. Axial Thrust Axial thrust is generated by the internal pressures acting on the impeller and shaft end. One component to this axial thrust is that due to, and dependent only on, suction pressure and is shown on Fig. 5.43a. The other major thrust component for a horizontal pump is that created by a hydraulically unbalanced impeller as shown in Fig. 5.43b. This effect is due to the presence of an unopposed pressure at the impeller back shroud. For small pumps with a suction diameter similar to the shaft diameter, this effect is minimal.
(a) THRUST DUE TO SUCTION PRESSURE

Absolute suction pressure

Atmospheric pressure

Atmospheric pressure (balanced)

(b) HYDRAULIC UNBALANCED THRUST

Pressure acting on impeller shrouds

FIGURE 5.43: (a) AXIAL SUCTION THRUST AND (b) HYDRAULIC UNBALANCED THRUST

Page 5-26

CHAPTER FIVE
However, for larger pumps and / or higher head pumps the generated thrust can be very large. The most common solution to this problem is to use an hydraulically balanced type of impeller fitted with a back ring and balance holes, which reduce the pressure at the rear of the impeller hub by allowing leakage though to the suction side A similar reduction in pressure can be achieved through the use of impeller rear (pump out) vanes. Both of these methods are shown in Fig. 5.43c. Multistage pumps may have opposed impellers for axial thrust balance, or a balancing device such as a balance disc or balance drum. (2) Single Stage Pump with Double inlet Impeller Theoretically a pump of this type should be in complete axial balance; however, the presence of slight casting differences will cause the flow pattern to each impeller entry (eye) to marginally differ thus creating a residual axial thrust. This is taken by a ball bearing of the combined radial / thrust type. The arrangement is illustrated in Fig. 5.44. (3)
FIGURE 5.44: HYDRAULIC AXIAL BALANCE WITH DOUBLE INLET IMPELLER

FIGURE 5.43: (c) THRUST BALANCE METHODS

Multistage Horizontal Pump with Balance Disc The unbalanced axial thrust is approximately equal to the pump differential pressure acting on the annular area of the impeller back shroud; this being roughly equivalent to the area of the impeller eye minus the shaft area.

FIGURE 5.45: AXIAL BALANCE TYPICAL BALANCE DISC ARRANGEMENT

Page 5-27

CHAPTER FIVE
The balance disc rotates with the shaft and automatically adjusts the gap at A so that the pressure in the inner balance chamber opposes the unbalanced axial thrust of the impellers. This balance device is entirely self compensating and no thrust bearing is required. (4) Multistage Pump with Balance Drum The area of this balance drum (see Fig. 5.46) is approximately the same as the impeller unbalanced area but slightly undersized to ensure a residual unbalanced load in order to keep the shaft in tension. This axial load is taken by a thrust bearing, in this case a tilting pad bearing of the Michell type as shown in Figure 5.47.

FIGURE 5.46: AXIAL BALANCE TYPICAL BALANCE DRUM ARRANGEMENT

FIGURE 5.47: AXIAL BALANCE DEVICES

Radial Thrust When a single volute pump is operated at the best efficiency flow rate, the velocities and hence the pressures acting on the impeller are uniform around the volute. This is shown in Fig. 5.48a. At flow rates other than best efficiency point, the pressure distribution is no longer uniform. At reduced flow rates, the pressures increase spirally towards the cutwater (see Fig. 5.48b) resulting in a radial reaction F. A similar situation exists at flow rates beyond best efficiency flow rate, with an approximately opposite (in direction) reaction. Fig. 5.49a) shows typical variation of radial thrust with flow rate. Note that maximum radial thrust occurs at zero flow with the minimum at the best efficiency flow rate. The magnitude of the radial thrust is a function of the total head, impeller diameter and impeller width. Thus high head pumps with large impeller diameters will experience very high radial thrusts. If pumps of this type are operated at

Page 5-28

CHAPTER FIVE
consistently low flow rates, bearing life will be reduced and the bending stresses associated with large shaft deflections may eventually lead to shaft failure. The manufacturer may design for this problem by supplying an oversize shaft with large bearings; however, this solution leads to a more expensive unit, especially in the case of large high head pumps. An alternative is to reduce the value of radial thrust through the use of a double volute as shown in Fig. 5.49b. In such a casing the flow is divided into two almost equal streams by two cutwaters 180o apart. Although the volute pressure inequalities remain, there are now two opposing radial forces which almost cancel out. In practice the cancellation is not complete but nevertheless a major reduction in radial thrust at partial capacities is achieved. Double volute casing pumps are most often found in the petro-chemical industry where consistent operation at partial capacities is commonplace. They are also occasionally found in high head per stage water pumps. Efficiency is slightly lower due to the additional wetted area. Diffuser casings also virtually eliminate radial thrust in the same way as a double volute. This is one of the reasons behind the reliable operation of diffuser type deep well vertical turbine pumps in which radial shaft support can only be supplied by water lubricated plain bearings.

FIGURE 5.48 (a) 100% BEP FLOW AND (b) REDUCED FLOW

Page 5-29

CHAPTER FIVE

(a)

(b) FIGURE 5.49: (a) RADIAL THRUST AND (b) DOUBLE VOLUTE

5.2.2.2 Specific Speed The characteristic differences between pumps can be described by a criterion referred to as specific speed, this being the speed of an ideal pump geometrically similar to the actual pump, which when running at this speed will raise a unit of volume in a unit of time of time through a unit of head. The performance of a pump is expressed in terms of pump speed, total head and flow rate. Specific speed is calculated from the formula using data at the best efficiency point as follows: Specific Speed Ns = N Q H

Where, in SI units [USS units] N = speed of the pump in r.p.m Q = flowrate in m3/s [US gpm] H = total dynamic head in m [ft] Specific speed relates to the geometry of the pump rotor and is independent of the pump size. When applying the formula above, it should be noted that: (a) (b) With a double suction impeller, half the best efficiency flowrate should be used. With a multistage pump, best efficiency conditions for the first stage impeller only are considered in calculating specific speed.

The normal range of specific speed (SI units) is from 10 to 300 and the relationship between specific speed and basic impeller shape is illustrated in Fig. 5.50.

Page 5-30

CHAPTER FIVE
Units of Specific Speed

(specific speed in US units)

FIGURE 5.50: UNITS OF SPECIFIC SPEED

Since specific speed is used only as an index or type number, certain liberties are permissible in selecting the units used as long as the relationship between the different units are clearly understood. Consequently the numerical value of Ns will vary according to the units in which H and Q are expressed. The speed of the impeller is always given in r.p.m. The relationship of specific speed in various units is as follows: Relationship of Specific Speed in Various Units Metric Ns Q (m3/s) H (m) 1.00 Specific Speed and Pump Type By categorising centrifugal pumps in terms of specific speed, the pump user can visualise the type of pump he is likely to need. e.g suppose a pump is required to deliver 0.10 m3/s at a head of 89m at 1450 r.p.m. Using the equation: Ns = = 1450 0.10 89 15.82 (817 in US units) British Ns Q (igpm) H(ft) 47.13 USA Ns Q (usgpm) H(ft) 51.64

This is indicative of a fairly low specific speed radial flow pump. Hence the user can expect a large pump with relatively low efficiency. However, if the speed is doubled to 2900 r.p.m the specific speed of the pump will be 31.64 (1635 in US units).

Page 5-31

CHAPTER FIVE
Consequently the pump can be smaller with a higher efficiency which may allow a less powerful driver, and make the overall pumpset package much less expensive. Associated civil costs would also be correspondingly reduced. Hence specific speed can be a valuable guide to the pump user in planning and designing a more cost effective pumping station. Radial flow pumps have a specific speed between 10 and 90, mixed flow pumps between 40 and 160, and axial flow pumps between 150 and 420. Specific Speed and Pump Efficiency The maximum attainable efficiency of centrifugal pumps depends to a great extant on pump geometry as categorised by specific speed. In general, the efficiency increases as the value of Ns rises.

FIGURE 5.51: RELATIONSHIP BETWEEN PUMP EFFICIENCY AND SPECIFIC SPEED (US UNITS) Fig. 5.51 illustrates the relationship between pump efficiency and specific speed. The

trend of the curve indicates that for economical operation very low values of Ns are to be avoided; a condition that can be overcome by employing multistage pumps. Specific Speed and Characteristic Curves Specific speed not only relates to impeller geometry, but also the trend to the characteristic curves for total dynamic head, power consumption and efficiency, examples of which are shown in Fig. 5.50. The steepness of the H Q curve increases as specific speed increases, whilst the efficiency curve remains relatively unchanged, except to exhibit a narrow plateau when Ns is high. At low specific speed, power consumption is a minimum at zero flowrate and rises with flowrate. This is an overloading characteristic and the motor must be sized to meet the extreme conditions that may occur in operation. At medium specific speed the power curve has a pronounced peak at approximately the design duty. This is a non overloading characteristic and therefore the pump can work safely over the entire flow range with a motor sized to meet the peak requirements.

Page 5-32

CHAPTER FIVE
High specific speed pumps have a failing power characteristic with maximum power occurring at minimum flow. Such pumps should not, therefore, be used in zero flow conditions. Discharge throttle valves or by-pass systems should be avoided with this type of pump otherwise a motor of considerably greater power that necessary for normal duty will be required. 5.2.3 Pump Suction Requirements There are two basic arrangements of surface water pumps namely those including a suction lift and those with a positive or gravity head on the inlet side. It is preferable that any rotodynamic pump is sited so that negative pressure does not develop on the suction side. Negative pressures can result in reduced performance and may prevent the pump from being automatically primed. There is also the risk of vortexing at the suction inlet causing further problems. If negative pressures are unavoidable for a surface mounted pump then a self priming pump must be specified. Unfortunately, it is not always possible to avoid suction lift with raw water pumping on account of the costs that would be involved to do so. It is however almost always possible to avoid negative suction heads with onward or treated water pumping. It is recommended however that negative suction be avoided whenever possible, even in raw water pumping stations by using a more suitable arrangement of design as illustrated in Figure 5.52. Suction lift pumping and maximum suction lift calculations have been detailed in sections 5.4.1.1 and 5.5.5.
FIGURE 5.52: PREFERRED ARRANGEMENT OF PUMPING STATIONS

5.2.4

Common Pump Types used in water supply

Water pumps may be considered as belonging to one of six types, namely: Centrifugal pumps; Axial and mixed flow pumps; Reciprocating pumps; Rotary pumps; Jet pumps; and Air lift pumps.

Page 5-33

CHAPTER FIVE
Their principle characteristics and applicability are summarised in the following table:
TABLE 5.1: PUMP TYPE, CHARACTERISTICS AND APPLICABILITY TYPE OF PUMP 1. Centrifugal (radial flow) a. Standard (dry) b. Submerged CHARACTERISTICS AND APPLICABILITY Medium to high speed of operation (usually 880 3,500 rpm). Smooth, even discharge. Motor powered. Efficiency ranges from 50% - 85% largely depending upon operational speed and pumping head. Capacity range 25 10, 0000 l/min. Less suitable for variable heads. Requires skilled maintenance. Specific speed 10 -90. CHARACTERISTICS AND APPLICABILITY High capacity and low lift pumping somewhat similar to centrifugal pumps. Can pump water containing sand and silt. Requires skilled maintenance. Specific speeds 40-60 for mixed flow and 150 -240 for axial flow. Low speed of operation: Hand, animal, wind, solar or motor powered. Low efficiency (range 25% - 60%) Capacity range: 10 50 l/min Suitable for pumping against variable heads. Valves and cup seals require frequent maintenance or replacement Medium speed of operation. Hand, animal, solar, wind or motor powered (using gearing). Medium to good efficiency (range 60% 70%). Best suited for low capacity and high lift (submerged) pumping. Used for increasing the suction depth of small centrifugal pumps down to 75 m, thus allowing the pump unit to be placed on the ground. Relatively low efficiency. Capacity ranges 10 -800 l/min, suitable for sandy waters as the sand can be removed before entering the pump. Commonly used to test boreholes. Usually of low efficiency

TYPE OF PUMP 2. Axial and Mixed Flow

3. Reciprocating a. Suction b. Lift (deep well) 4. Rotary Helical rotor (Mono type) 5. Jet

6. Air lift

5.2.5

Operations of Centrifugal Pumps

When the rotational speed of a centrifugal pump is changed, there is little change in efficiency but the output, head developed and the power required alter according to the following relationships which are known as the affinity laws. Where N is the rotational speed, Q the output, and P the power required: Q1 / Q2 = N1 / N2; H1 / H2 = (N1)2 / (N2)2; and P1 / P2 = (N1)3 / (N2)3. These relationships are illustrated in Figure 5.53.

FIGURE 5.53: EFFECT OF CHANGING SPEED OF PUMP

Page 5-34

CHAPTER FIVE
Air in a pump will reduce its efficiency and may induce corrosion as a result of cavitation. The presence of air is usually apparent by a hard crackling noise as if the pump had some gravel inside and if such a sound does not disappear shortly after starting, the cause should be investigated. Net positive suction head (NPSH) which is largely dependent on temperature and elevation determines the maximum suction head possible. Specific speed enables the classification of geometrically similar pumps and is determined from the expression, Nsp = (N Q1/2) / H3/4 Where, N is the impeller speed (rpm), Q is the output at maximum efficiency (m3/s), and H is the equivalent head (m) 5.2.6 Efficiencies and Fuel Consumption of Rotodynamic Pumps and their Motors 5.1

Giving efficiencies and fuel consumption is difficult because they vary so widely with every individual case having something unique about it. However the figures given in the following tables may be taken as a guide to the efficiencies and fuel consumption that can be expected from a well designed installation, although wide variations occur according to the power rating and type of pump and motor. Because electricity tariffs in Tanzania are high, power costs can account to 30 40 % of total running costs. It is therefore necessary to select as efficient pumping plant as practical. It is further wise to assume an annual efficiency drop of at least 1% per year.
TABLE 5.2: COMPARATIVE EFFICIENCIES FOR DIFFERENT PUMPS AND MOTORS UNIT Horizontal Centrifugal Pump Submersible Pump EFFICIENCY RANGE THAT CAN BE EXPECTED Medium sized 80 82%, up to 85% for large size and even higher with special construction involving higher price Small size 70%, larger sizes 75 81%. Generally about 3% less than for an equivalent duty, vertical spindle pump due to its restricted diameter. Generally about 3% less than a similar duty horizontal pump. Fixed speed, alternating current (AC) induction motor, 93 95% Fixed speed AC induction motor, 90 94% Because of restricted diameter usually between 85 89%

Vertical spindle shaft driven pump Horizontally mounted electric motor Vertically motor mounted electric

Submersible pumpset motor

Page 5-35

CHAPTER FIVE
TABLE 5.3: OVERALL FUEL CONSUMPTION UNIT Electrically driven pumps FUEL CONSUMPTION About 1.0 kW for every 0.75 kW of water power output which implies an efficiency of about 75%. Can however be up to 1.3 kW per 0.75 kW water power output or higher for small pumps and for variable speed pumps. Average 0.28 kg of diesel fuel oil consumed per kWh of engine power exerted to 0.21 kg of diesel fuel oil consumed per kWh of engine power good. For lubricating oil add 5% to fuel oil cost.

Diesel Engines

5.3

PRIME MOVERS

Other than for the small pumps where wind, human or animal power is used, prime movers used to drive pumping machinery are essentially either: Electric motors, or Diesel engines. Diesel engines can either be coupled directly or through gearing or used to generate electricity which is then used to power an electric motor. However, it must be noted that there are disadvantages with all prime movers. For a water supply, the key is dependability and electric motors can only be as reliable as the supply of electricity to them. Unfortunately, much of Tanzania is yet to have a reliable virtually continuous power supply with both voltage fluctuations and unplanned power outages a not uncommon feature. In addition, there are many parts of the country and particularly rural areas that as yet have no such mains supply. Diesel engines and diesel generators require the supply of fuel, usually in 45 gallon drums or by fuel tanker for the larger schemes. Fuel drums are often prone to pilferage unless stored securely and with use carefully monitored by scheme managers. Tanker delivered fuel requires permanent storage tanks into which the fuel can be transferred. The use of diesel generators to firstly generate electricity and then for that electricity to power electric motors loses efficiency at each stage of the process. Direct coupled diesel engines do not provide power for any ancillary equipment or site lighting. Even where diesel generators are provided as a standby electrical source, they add significantly to the capital and operational costs and require periodic running as well as maintenance to keep them in an operational state to deal with mains power outages. These prime movers and their characteristics are discussed below. 5.3.1 Electric Motors for Pump Drives

An electric motor converts electrical energy into mechanical energy. The most common type of electric motor used to run pumping machinery is the three-phase alternating current (AC) motor. They can be classified as follows: AC induction motors of either the cage rotor or wound rotor type; synchronous motors; and commutator motors.

Page 5-36

CHAPTER FIVE
AC induction fixed speed cage motors are the simplest, most robust, reliable and least cost of all electric motors. Synchronous motors are more expensive because of the need for more complex control equipment as well as a DC supply for the rotor. Commutator motors provide a variable speed source and in the days before electronic variable frequency drives used to be used where variable speeds were required. Because by far the most commonly used electric motor for operating pumps are the AC induction motors, only these are described in more detail here. 5.3.1.1 The AC Induction Motor 5.3.1.1.1 General A typical AC motor consists of two parts: 1. 2. An outside stationary stator having coils supplied with AC current to produce a rotating magnetic field, and; An inside rotor attached to the output shaft that is given a torque by the rotating field.

In an AC induction motor, the phase differences between the three phases of the polyphase electrical supply create a rotating electromagnetic field in the motor. Through electromagnetic induction, the rotating magnetic field induces a current in the conductors in the rotor, which in turn sets up a counterbalancing magnetic field that causes the rotor to turn in the direction the field is rotating. The rotor must always rotate slower than the rotating magnetic field produced by the polyphase electrical supply; otherwise, no counterbalancing field will be produced in the rotor. Induction motors up to about 500 kW in output are produced in highly standardized frame sizes, making them nearly completely interchangeable between manufacturers (although European and North American standard dimensions are different). There are two types of rotors used in induction motors. Cage rotors: Most common AC motors use the cage rotor, which takes its name from its shape - a ring at either end of the rotor, with bars connecting the rings running the length of the rotor. It is typically cast aluminium or copper poured between the iron laminates of the rotor, and usually only the end rings will be visible. The vast majority of the rotor current will flow through the bars rather than the higher-resistance and usually varnished laminates. Very low voltages at very high currents are typical in the bars and end rings; high efficiency motors will often use cast copper in order to reduce the resistance in the rotor. In operation, the cage motor may be viewed as a transformer with a rotating secondary when the rotor is not rotating in sync. with the magnetic field, large rotor currents are induced; the large rotor currents magnetize the rotor and interact with the stator's magnetic fields to bring the rotor into synchronization with the stator's field. An unloaded cage motor at synchronous speed will consume electrical power only to maintain rotor speed against friction and resistance losses; and as the mechanical load increases, so will the electrical load - the electrical load being inherently related to the mechanical load. This is similar to a transformer, where the primary's electrical load is related to the secondary's electrical load.

Page 5-37

CHAPTER FIVE
Furthermore, a stalled cage motor (overloaded or with a jammed shaft) will consume current limited only by circuit resistance as it attempts to start. Unless something else limits the current (or cuts it off completely) overheating and destruction of the winding insulation is the likely outcome. Wound Rotor: An alternate design, called the wound rotor, is used when variable speed is required. In this case, the rotor has the same number of poles as the stator and the windings are made of wire, connected to slip rings on the shaft. Carbon brushes connect the slip rings to an external controller such as a variable resistor that allows changing the motor's slip rate. In certain high-power variable speed wound-rotor drives, the slipfrequency energy is captured, rectified and returned to the power supply through an inverter. Compared to cage rotors, wound rotor motors are expensive and require maintenance of the slip rings and brushes, but they were the standard form for variable speed control before the advent of compact power electronic devices. Transistorised inverters with variable-frequency drive can now be used for speed control, and wound rotor motors are becoming less common. Transistorised inverter drives also allow the more-efficient three-phase motors to be used when only single-phase mains current is available. Motor starting: Several methods of starting a polyphase motor are available. Where a large inrush current and high starting torque can be permitted, the motor can be started across the line, by applying full line voltage to the terminals (Direct-on-line, DOL). Where it is necessary to limit the starting inrush current (where the motor is large compared with the short-circuit capacity of the supply), reduced voltage starting using either series inductors, an autotransformer, thyristors, or electronic soft starters. A technique sometimes used is star-delta starting, where the motor coils are initially connected in wye arrangement for acceleration of the load, and then switched to delta arrangement when the load is up to speed. This technique is more common in motors manufactured in Europe than in those from North America. Transistorised drives can directly vary the applied voltage as required by the starting characteristics of the motor and load. Soft starters are typically computerised devices that can fault find. They are however sophisticated and need suitably qualified technicians trained in their use. Should they go unserviceable they need returning to the manufacturer for attention. Motor speed: The speed of an AC motor is determined primarily by the frequency of the AC supply and the number of poles in the stator winding, according to the relation: Ns = (120 F) / p Where, Ns = Synchronous speed, in revolutions per minute (RPM) F = AC power frequency p = Number of poles per phase winding Actual RPM for an induction motor will be less than this calculated synchronous speed by an amount known as slip, that increases with the torque produced. With no load, the speed will be very close to synchronous. When loaded, standard motors have between 23% slip, special motors may have up to 7% slip, and a class of motors known as torque motors are rated to operate at 100% slip (0 RPM/full stall). 5.2

Page 5-38

CHAPTER FIVE
The slip of an AC motor is calculated by: S = (Ns Nr) / Ns (percentage slip = (Ns Nr) / Ns 100) Where, Nr = Rotational speed, in revolutions per minute, and S = Normalised Slip, 0 to 1. As an example, a typical four-pole motor running on 60 Hz might have a nameplate rating of 1725 RPM at full load, while its calculated speed is 1800 RPM. Submersible pump motors: These are typically a two-phase AC servo motor with a squirrel-cage rotor and a field consisting of two windings: 1) a constant-voltage (AC) main winding, and 2) a control-voltage (AC) winding in quadrature with the main winding so as to produce a rotating magnetic field. The electrical resistance of the rotor is made high intentionally so that the speed-torque curve is fairly linear. Two-phase servo motors are inherently high-speed, low-torque devices.. 5.3.1.1.2 Electric Motor Failure Modes - New Methods for Low Resistance Protection
Mostly due to low insulation resistance, as many as 90% of motor and generator electrical problems occur at start-up. These are trickier than over-current or mechanical failures, but new products and techniques can reliably insure against low resistance failures automatically and safely.

5.3

When electric motors or generators fail during use, it invariably means extensive equipment repair or replacement the high cost of which is often minor compared to the impact of unscheduled system downtime. Understanding the more common failure modes is essential to minimizing their impact in an operating environment. Types of Failure: Electric motors typically fail due to one of three causes: low resistance (electric ground) mechanical (bearing failure, vibration, etc.) over-current (electrical overload)

Low resistance is one of the most common failure causes. Also known as Monday Morning or Start-up failure (these are the circumstances in which it usually strikes), low resistance is traditionally the most difficult electrical problem to protect against, since it can occur either through conventional, predictable wear or sudden, catastrophic fashion. A combination of protective devices and/or predictive maintenance tests and procedures has long helped users guard against mechanical and over-current failures
However, devices to protect against low resistance failures, are recent developments and have not established uniformly good reputations for effectiveness, or more importantly for safety.

Distinctions between Failure Modes; How is Low Resistance Different? Mechanical Failures: Mechanical failures happen for a wide variety of reasons, including inadequate lubrication, unbalance and vibration, and misalignment. What most mechanical failures have in common is that they happen gradually; they display characteristic warning signs which intensify over time. An increasing assortment of
Page 5-39

CHAPTER FIVE
sophisticated analytical techniques and services can detect impending mechanical failure at ever-earlier stages, allowing corrective measures as part of an overall, scheduled preventive maintenance program. Any good electrical apparatus shop can recommend suitable diagnostic tests and instrumentation. For all practical purposes, mechanical failures can be averted through routine preventive and predictive maintenance. Over-current Failures: Over-current failures happen most often when operating conditions cause devices to draw substantially more current than their rated load capacity. They tend to happen suddenly, and are not conducive to preventive procedures or predictive measurements. Fortunately, there are a wide range of reliable devices available to limit or prevent over-current failure. Typically, these are wired into the circuit and programmed for a specific amount of current above rated load, which they will allow for a programmed period of time after which they act quickly to shut down the current flow. Low Resistance Failures: Its been estimated that 90% of motor and generator electrical problems occur at start-up. A high percentage of these problems are caused by low insulation resistance, traditionally the most difficult type of failure to prevent. When a new or rebuilt motor or generator is first placed in service, insulation resistance can exceed 1,000 megohms (1,000,000,000 ohms). [Note that resistance should routinely be checked before initial operation.] During device operation, insulation performance can fall to dangerous levels through a) gradual decay, or b) precipitous failure. With innovative technology, carefully applied, its now possible to insure rotating electrical equipment, guarding against both types of low resistance failures. In addition, insulation performance can readily and consistently be monitored over time to be charted with unprecedented accuracy, facilitating cost-effective predictive maintenance. Such prevention methods offer the only way to assure that: A) expensive and/or critical equipment has the best possible insurance against electrical failure, and B) the protective system itself is safe. (Beware however: some cures are worse than the disease!) Gradual Decay: Insulation resistance inevitably degrades over time, eventually reaching a point at which its not safe to operate equipment. Experts disagree on exactly when repair or replacement becomes necessary, but a common rule of thumb says that insulation resistance must be 1 megohm for each 1,000 rated volts, plus 1 megohm. (e.g. 5,000 volts = 6 megohms, minimum). No electrical equipment should be operated if resistance is 1 megohm or less. The rate at which insulation resistance decreases varies with a large number of factors some controllable, some not. Easily the most significant is equipment operating temperature, generally determined by both the ambient environment and by load. With each 10 C increase, resistance typically drops 50%. More heat in device windings has a two-fold effect: resistance is lower as long as the temperature is elevated, and the useful life of the insulation is also reduced. In fact, for each 10 C rise in operating temperature, insulation life is cut by half. The good news is that the rate of resistance degradation is usually quite constant over time, and hence predictable. Thus, above minimum safe levels*, the change in resistance readings over

Page 5-40

CHAPTER FIVE
time is much more important than the resistance level at any given point. [*Accepted
industry practice is one megohm per rated kilovolt, plus one.]

In the past, the only method of measuring actual resistance was to use a hand-held megohmeter. These instruments have a number of inherent limitations, but the most serious drawback by far is that they require a trained user to take the time to make and record measurements on each piece of equipment in shop, and to remember to do so at frequent intervals. A tool and/or method to automatically test resistance each time a device is started creating accurate log of megohm levels, and calculating the rate of change is of immense value. Precipitous Failure: Despite the above comments on steady, predictable change in resistance levels, there might not be any warning signs. Windings that check out in the 50-100 megohm range can suddenly drop as much as two orders of magnitude, to less than 1 megohm overnight. A common cause is environmental change. If a facility is shut down for one or more shifts and allowed to cool off, moist outside air is often drawn in, causing condensation to form on idle equipment. Until the space is heated and humidity lowered or until auxiliary heaters can dry off the equipment it will not be safe to operate. Relatively mild climatic shifts can also cause this effect, even overnight. In recent years, a number of products and systems have been introduced as the answer to this thorny problem as a means of reducing or eliminating the risk. In concept, in execution, or both, these solutions have serious shortcomings, even to the point of compromising worker safety. No matter how through the preventive and predictive maintenance programs, and no matter how diligent the operating crew, catastrophic resistance failure can still occur. About variable Frequency Drives: Recent innovations in variable frequency and controllers have brought these devices into widespread use. Reduced price and low maintenance make variable frequency initially attractive. Their high efficiency also yields reduced operating costs, which can be a compelling incentive. The drawback to the variable frequency approach is that it reduces expected equipment life. Partly due to mechanical factors, such as increased vibration and noise, but mostly due to significantly higher operating temperatures which both lowers effective resistance, and accelerates insulation degradation. Where variable frequency drives are employed, insulation life expectantly will be reduced; its even more important here to monitor resistance levels to prevent equipment failure. Root Cause Failure Analysis on AC Induction Motors Under voltage: This root cause condition, which occurs when the supply voltage is more than 10% lower than the motor rated voltage, frequently results in motor overheating. As the motors supply voltage is reduced, the motors current draw will increase because the motor needs constant power (volts current) to supply the constant external load. However, motor heating increases with current with resultant detrimental effects on the motors insulation. These are root causes resulting in excessive heat being generated within the motor itself (windings/core iron). There is another group of conditions that, when present, interfere with the motors ability to cool itself.

Page 5-41

CHAPTER FIVE
The end result is excessive heat leading to premature insulation failure. These conditions are associated with the application environment of the motor. They include: (1) partially/totally blocked ventilation passages, (2) damaged/destroyed external cooling fans, (3) foreign substance buildup on motor surfaces, (4) motor operating at higher than rated attitude, (5) motor operating at higher than rated ambient temperature, and (6) motor operating in direct sunlight for long periods of time. Contamination is the root cause of conditions (1), (2) and (3), while misapplication/ignorance is the root cause of conditions (4), 5), and (6 Contamination: Most contamination occurs in the air surrounding the motor in the form of chemical agents, particulates, dirt, etc. Chemicals and acids can attack and destroy external cooling fans, especially aluminium ones. Also, particulates/dirt can build up on fan blades causing imbalance of the fan and eventual cracking and breaking off of the blades. Particulates/dirt can also build up on the motor surfaces themselves blocking cooling air ventilation passages as well as blanketing the motor so as to interfere with normal heat transfer from the motor to the outside air. Conditions like altitude, ambient air temperature, and sunlight tend to be overlooked when ordering motors. Standard AC induction motors meet recommended requirements of 1000 m maximum altitude. At higher altitudes, the thinner air reduces motor cooling ability so that a standard motor will run hotter under these conditions. Likewise, recommended maximum ambient temperature for a standard induction motor is at 40C. At higher ambience, a standard motor will run hotter. There is one significant cause of insulation breakdowns that is unrelated to excessive heat. That cause is high humidity conditions, which result in condensation on the motor insulation even though the motor is totally enclosed. This problem usually results from one of two scenarios: (1) long-term storage of the motor in a warehouse where there is no controlled environment and (2) alternate heating/cooling (on/off) cycling of the motor that occurs under normal service conditions. In either case, long-term exposure of the insulation to excessive moisture can gradually weaken it to the point of breakdown. A controlled environment warehouse with regulated humidity levels will solve problem (1). Problem (2) can be addressed by the use of single-phase space heaters wrapped around the stator end turns or providing a low-level DC trickle current through one of the motors phase windings during off periods. Either method helps to keep the inside of the motor dry and condensation-free. 5.3.1.1.3 Increasing the Power Factor Power factor: Power factor is the relationship between working (active) power and total power consumed (apparent power). The higher the power factor, the more effectively electrical power is being used. Essentially, power factor is the measurement of how effectively electrical power is being used. A distribution systems operating power is composed of two parts

Page 5-42

CHAPTER FIVE
Why improve low power factor? Low power factor means poor electrical efficiency. The lower the power factor, the higher the apparent power drawn from the distribution network. When low power factor is not corrected, the utility must provide the nonworking reactive power in addition to the working active power. This results in the use of larger generators, transformers, bus bars, wires and other distribution system devise that otherwise would not be necessary. As the utilitys capital expenditures and operating costs are higher as a result, they pass these higher expenses to industrial users in the form of power factor penalties and higher utility bills. Power Factor Correction Capacitors solve the problem: Low power factor problems are solved by adding power factor correction capacitors to the electrical network. As illustrated below, power factor correction capacitors work as reactive current generators providing needed reactive power (kvar) to the power supply. By supplying their own source of reactive power, the industrial user frees the utility from having to supply it; therefore, the total amount of apparent power (kVA) supplied by the utility will be less. Power factor correction capacitors reduce the total current drawn from the distribution system and subsequently increase system capacity.

Who can benefit? Water utilities and other industrial and commercial applications can benefit from improving power factor levels.

Page 5-43

CHAPTER FIVE
Not only will power Factor Correction Capacitors save money, they will: Reduce heat loss of transformers and distribution equipment. Prolong the life of distribution equipment. Stabilize voltage levels, and Increase the systems capacity

The Solution to Harmonics: Any device with non-linear operation characteristics can produce harmonics in the electrical network. Harmonic current can cause a disturbance on the supply network and adversely affect the operation of other electrical equipment. Harmonic current can cause: Excessive heating and failure of different distribution apparatus Nuisance tripping of circuit breakers and fuses Noise that leads to erroneous operation of control system components. Damage to sensitive equipment. Electronic communication interference, etc.

There is need to perform an on-site harmonic evaluation of the electrical network and come up with a detailed plan to control and reduce them.

5.3.1.1.4

Power Factor, Installed Capacity and Demand

Two factors that need to be considered in operating electric motors are the power factor and the monthly relationship between installed capacity and gross operated power demand as both can have a bearing on the power bill, and especially the former, as charges are levied for energy (TShs/kWh) and for demand (TShs/kVA). The power factor is the cosine of the phase displacement angle in alternating current circuits. It is used to calculate the active current, Jw, such that: Jw = J cos , and Pw = U J cos Where U is the voltage, and is the angle of displacement. The phase displacement in time of the sinusoidal current flow in relation to voltage is caused by the inductive and capacitive reactances. One period corresponds to 360o. The angle of phase displacement , between voltage and current is +90o for capacitance and 90o for inductance. Electrical power is the product of current and the voltage. Apparent power Ps = U x J (that is when no phase displacement and cos = 1) True power is therefore different from apparent power, and the greater the phase displacement the lower the power factor and hence for the same current, the lower the voltage. This in turn means need for a higher current to obtain the same power. 5.4

Page 5-44

CHAPTER FIVE
TABLE 5.4: DEGREES OF DISPLACEMENT AND POWER FACTOR DEGREES OF DISPLACEMENT POWER FACTOR 0 1.00 10 0.98 15 0.97 20 0.94 25 0.91 30 0.87 35 0.82 40 0.77

Hence when an electric load has a p.f. lower than 1, the apparent power delivered to the load is greater than the real power that the load consumes. Only the real power is capable of doing work, but the apparent power determines the amount of current that flows into the load, for a given load voltage. Energy losses in transmission lines increase with increasing current. Power utility companies therefore usually require that customers, especially those with large loads, maintain the power factors of their respective loads within specified limits or be subject to additional charges. TANESCO surcharges, should they wish to levy them are:
TABLE 5.5: POWER FACTOR SURCHARGES
Tariff 1 tariff for domestic loads less than 7,500 kWh per month @ 230V or 400V all other loads less than 7,500 kWh per month @ 230V or 400V more than 7,500 kWh per month, demand less than 500 kVA, @ 400V @ 11,000V minimum p.f. 0.95 surcharge formula 1 + 3%[95%-p.f.%] applied to kWH

0.90

1 + 3%[90%-p.f.%]

kWH

0.90

1 + 10%[90%-p.f.%]

kVA

3
ai T f r f i r a t o r f m nu i m pl d e i po t ch r u s ef g r a u a m r o dm i t s e oc sl d a o l t h s e 50 , 7 n a Wp k r e h0 1 . 5 9 al eo h t o r l e s d a l a h t s 50 , 7 n Wp k r e h0 1 . 0 9 3 1 +% 09 [ % f . p H W k an h t e r 0k 5 , 7 h W 0p mo r e dm , h t n d n a e0 mo 2 . 0 9 e l s n5 a h t kA 0 @ , V4 V 0 3 0 , 1 @0 V 0 . 0 9

0.90

1 + 10%[90%-p.f.%]

kVA

Power factor correction (PFC) is a technique for counteracting the undesirable effects of electric loads that create a power factor (p.f.) that is less than 1, returning an electric AC power transmission system to very near unity by switching in or out banks of capacitors or inductors which act to cancel the inductive or capacitive effects of the load. The inductive effect of motor loads can be offset by locally connected capacitors. Power factor correction may be applied both by the electrical power supply utility to improve the stability and efficiency of the transmission network; and, correction equipment can be installed by an individual customers to reduce the costs charged to them by the electricity supplier should the supplier wish to do so. A significant penalty can be charged for a power factor below a certain prescribed value with the poorer the power factor the greater the penalty. Failure to maintain an acceptable power factor can add a significant percentage to the power bill considerably in excess of the cost of supply and installation of the necessary PFC equipment. Power tariffs usually incorporate a relatively small standing charge directly related to installed capacity, adjusted by the actual monthly gross usage. Hence if the pumping plant includes standby units and whilst it is desirable that the running hours of each pumpset are kept generally similar, it may be cost effective only to swap over units on a monthly basis commensurate with the power billing cycle.

Page 5-45

CHAPTER FIVE
5.3.1.2 Different Starting Methods The following is a short description of the most common starting methods for squirrel cage motors. An overview of common problems when starting and stopping a motor with different starting methods, see page 5-57. Direct-on-line start (D. O. L): This is by far the most common starting method available on the market (Fig 5.54). The starting equipment consists of only a main contactor and thermal or electronic overload relay. The disadvantage with this method is that it gives the highest possible starting current. A normal value is between 6 to 7 times the rated motor current but values of up to 9 or 10 times the rated current exist. Besides the starting current there also exists a current peak that can rise up to 14 times the rated current since the motor is not energised from the first moment when starting. The values are dependent on the design and size of the motor, but in general, a smaller motor gives higher values than a larger one. During a direct-on-line start, the starting torque is also very high, and is higher than necessary for most applications. The torque is the same as the force, and an unnecessary high force gives unnecessary high stresses on couplings and the driven application. Naturally, there are cases where this starting method works perfectly and in some cases also the only starting method that works.

FIGURE 5.54: DIRECT-ON-LINE START

Star-delta start: This is starting method that reduces the starting current and starting torque. The device normally consists of the contactors, an overload relay and a timer for setting the time in the star-position (starting position). The motor must be delta connected during normal run, in order to be able to use this starting method. (Fig 5.55) The received starting current is about 30% of the starting current during direct on line start and the starting torque is reduced to about 25% of the torque available at a D. O. L.

Page 5-46

CHAPTER FIVE
start. This starting method only works when the application is light loaded during the start. If the motor is too heavily loaded, there will not enough torque to accelerate the motor up speed before switching over to the delta position. When starting up pumps and fans for example, the load torque is low at the beginning of the start and increases with the square of the speed. When reaching approx. 80 85% of the motor rated speed the load torque is equal to the motor torque and the acceleration ceases. To reach the rated speed, a switch over to delta position is necessary, and this will often result in high transmission and current peaks. In some cases the current peak can reach a value that is even bigger than for a D. O. L start. Application with a load torque higher than 50% of the motor rated torque will not be able to start using the start-delta starter.

FIGURE 5.55: STAR-DELTA STARTER

Frequency converter: The frequency converter is sometimes also called VSD (Variable Speed Drive), VFD (Variable Frequency Drive) or simply Drives, which is probably the most common name. (Fig 5.56) The drive consists primarily of two parts, one which converts AC (50 or 60 Hz) to DC and the second parts which converts the DC and the second part which converts the DC back to AC, but now with a variable frequency of 0-250 Hz. As the speed of the motor depends on the frequency this makes it possible to control the speed of the motor by changing the output frequency from the drive and this is a big advantage if there is a need for speed regulation during a continuous run.

Page 5-47

CHAPTER FIVE
In many applications a Drive is still only used for starting and stopping the motor, despite the fact that there is no need for speed regulation during a normal run. This will of course create a need for much more expensive starting equipment than necessary. By controlling the frequency, the rated motor torque is available at a low speed and the starting current is low, between 1 and 1.5 times the rated motor current. Another available feature is softstop, which is very useful, for example when stopping pumps where the problem is water hammering in the pipe systems at direct stop. The softstop function is also useful when stopping conveyor belts from transporting fragile material that can be damaged when the belts stop too quickly. It is very common to install a filter together with the drive in order to reduce the levels of emission and harmonics generated

FIGURE 5.56: FREQUENCY CONVERTER

Softstarter: A softstarter has different characteristic to the other starting methods. It has thyristors in the main circuit, and the motor voltage is regulated with a printed circuit board. The softstarter makes use of the fact that when the motor voltage is low during start, the starting current and starting torque is also low. During the first part of the start the voltage to the motor is so low that it is only able to adjust the play between the gear wheels or stretching driving belts or chains etc. In other words, eliminating unnecessary jerks during the start. Gradually, the voltage and the torque increase so that machinery starts to accelerate. One of the benefits with this starting method is the possibility to adjust the torque to the exact need, whether the application is loaded or not. In principle the full starting torque is available, but with the big difference that the starting procedure is much more forgiving to the driven machinery, with lower maintenance costs as a result.

Page 5-48

CHAPTER FIVE
Another feature of the softstarter of the softstop function, which is very useful when stopping pumps where the problem is water hammering in the pipe system at direct stop as for star-delta starter and direct-on-line starter. The softstop function can also be used when stopping conveyor belts to prevent material from damage when the belts stop too quickly.

FIGURE 5.57: SOFT STARTER

5.3.1.3 Different Applications All motors are used for starting and running different applications to Centrifugal Pumps. The different applications will also result in different load conditions for the motor. There are two factors to consider. 1. Braking load torque, direct braking force on the motor shaft. To be able to accelerate, the motor has to be stronger than the load. The accelerating torque is different between the available torque and the load torque. Accelerating torque - Available motor torque-load torque 2. Involved moment of inertia or flywheel mass will also affect start. The bigger inertia the longer starting time for same motor.

FIG 58 (a)

Page 5-49

CHAPTER FIVE
FIGURE 5.58(A-D): DIFFERENT APPLICATIONS

Centrifugal pump There are a lot of different types of pumps; like piston pumps, centrifugal pumps, screw pumps etc., but the most common version is the centrifugal pump. Direct-on-line start: Starting up a pump is normally not a problem for a squirrel cage motor. The problem is the wear and tear depending on pressure waves in the pipe system created when the motor starts and stops too quickly. During a D.O.L start the motor give much too high starting torque with the result that the motor accelerates and reaches nominal speed too quickly. The reason is that the braking load torque is low for a pump during start. This starting method also gives maximum possible starting current.

FIG 5.58b

Star-delta starter (Y-D): By using a star-delta starter it is possible to reduce the starting torque. The motor torque in the star position is too weak to be able to complete the start and reach the rated speed. The quadratic load torque will become too high for the motor when reaching approx. 8085% of the rated speed and the switch over to the delta position will give both high transmission and current peaks with pressure waves as a result. The current peaks can be equally high as at a D.O.L start or even higher.

FIG 5.58c

When stopping a pump: During stop is also normal to have problems. When making a direct stop by disconnecting the main supply the motor stops too quickly. Depending on high mass flow in the pipe system the water will continue with the same speed for a short

Page 5-50

CHAPTER FIVE
period and then come back gain, backwards in the pipe system. This creates high pressure shocks on valves and gives high mechanical stresses on the pipe system. Softstarter: By using a softstarter the voltage is reduced during the start sequence with the result that the motor torque is reduced. During the start sequence the softstarter increases the voltage so that the motor will be strong enough to accelerate the pump to the nominal speed without any torque or current peaks. A normal starting current with a softstarter when starting a fully loaded centrifugal pump is approximately 4 times the rated motor current. Also during the stop sequence the softstarter reduces the voltage during stop via a voltage ramp and the motor becomes weaker and weaker. Because of this the water speed slows down very smoothly without creating any pressure waves. A special function on the softstarter is sometimes available, called step-down voltage, which ensures on optimum setting to the actual need for any pipe system.

FIG 5.58d

Selection of a suitable softstarter Normal start: Starting a pump is a typical normal start condition. Select a softstarter according to the rated motor power. Heavy duty start: Not applicable for this application Recommended basic settings: Start ramp: 10 sec. Stop ramp: 20 sec. Initial voltage: 30%

Page 5-51

CHAPTER FIVE
Common problems when starting and stopping motors with different starting methods Type of problem Direct-online Slipping belts and heavy wear or bearings High inrush current Heavy wear and tear on gear boxes Damaged goods/ products during stop Water hammering in pipe system when stopping Transmission peaks Yes Type of starting method Star-delta start Medium Drives No Softstarter No

Yes Yes Yes Yes

No Yes (loaded start) Yes Yes

No No No Best Solution No

No No No Reduced

Yes

Yes

No

Auto transformer start and start of a part winding motor have similar problems to the star-delta start

Page 5-52

CHAPTER FIVE
5.3.1.4 Speed Control of Pumps with Frequency Converters When pumped volumes of liquid are variable control of pump by means of Frequency Converter saves energy
The pump is always dimensioned in consideration of the maximum flow The pump and the associated plant equipment such as pipes, valves, and tanks must always be designed to cover the maximum pumped volume. The following must be taken into consideration to determine the maximum capacity of the plant: provision for increasing demand excess demand for pumping capacity in exceptional circumstances, for example, when the tanks are being emptied or refilled in the event of emergencies, such as fire, heavy rainfall, etc. and therefore some form of control is required. The pump station being designed in view of the maximum plant capacity, a form of control must be provided to regulate the volume of flow by the variation in pumping demands. An average pumped quantity Qm may be only a fraction of the maximum pump capacity Qp. The duration curve in Fig. 1.2 illustrates how for the most part of, for example, over a one-year period the pump operates at reduced capacity. Speed control is more economical than other forms of control The pump flow can be controlled by using, for example, the following control methods: choke control by means of a valve start-stop control of the pump speed control, for example, by means of a frequency converter. Choke control (or throttling) is even today the most commonly used control form in industrial applications. Its efficiency is, however, very low when compared with speed control which in many cases gives more than a 50 per cent saving in energy. Pumps at water-works and sewage water treatment plants are normally controlled by means of start-stop control. Its efficiency is often also poor (Fig.1.3), and besides, stress due to frequent starting and stopping may cause damage to the pipes and other

plant equipment.
FIGURE 5.59(FIGS 1.1, 1.2,1.3) FIGURE 5.59: SPEED CONTROL OF PUMPS WITH FREQUENCY CONVERTERS

Page 5-53

CHAPTER FIVE
as energy is not wasted in overcoming unnecessary counter-pressure.

Choke control means that the flow of liquid in the pipes is choked by means of a valve. This results in a waste of energy because the pump is continuously working against the high pressure imposed by the valve. The power consumed by the pump can be calculated from the formula:

P = Q H g

Where: P = power [W] Q = pumped quantity [m3/s] H = pump head [m] = density of the liquid [kg/m3] g = acceleration of gravity [m/s2] = pump efficiency The above formula shows that the power requirement P is directly dependent on the product of the pumped quantity Q and the pump head H. Figs. 2.1 and 2.2 illustrate the power requirements which are represented by the hatched areas in both figures.

It can be seen that in this example the power requirement with speed control is less than half of that with choke control.

The saving obtained in energy depends essentially on the average pumped quantity. Fig. 2.3 shows how much the energy saving can be at different pumped volumes. When the power saving is known, the saving in energy can always be calculated by multiplying the power saving by the time factor. The methods of calculation are described in more detail later under this section.

FIGURE 5.59: (FIGS 2.1, 2.2, 2.3)

[SAMI is a Frequency Converter]

Page 5-54

CHAPTER FIVE
A Frequency Converter drive also reduces plant costs, Improves control effect, and reduces maintenance costs.
Fewer pumps required

Control of flow is often arranged by means of two or more parallel-connected pumps which are of different sizes. Step-by-step control is thus achieved by running the pumps in turn. Improved control with a lesser investment is achieved if the biggest pump is provided with the frequency converter control as shown in Fig. 4. Besides parallel connected pumps and motors, several valves and parts of the pipe system can be dispensed with.

Reduced need for tanks Pressure tanks and upper water tanks are used for keeping a uniform pressure in the pipes in application where the pump runs on intermittent duty as, for example, in waterworks. If the pump is provided with a frequency converter, the tanks can be smaller or may be totally dispensed with. In addition to the lower investment costs, a better control result is achieved, which means a more uniform pressure at the consumer end.

Savings in electrical distribution costs The starting current which a pump provided with a frequency converter takes from the electrical supply line is but a fraction of the starting current required by direct starting. On account of this the electrical distribution equipment can be made smaller and be purchased at a lower price. A typical object for saving purposes is a standby generator for important pumps. When a frequency converter is used for the speed control of the pump, the generator size need be only 30 to 50 per cent of that previously required.

FIG 5.59 (FIGS 4, 5, 6)

Page 5-55

CHAPTER FIVE
Compensating capacitors can be dispensed with Squirrel cage motors need reactive power which has to be generated somehow. To avoid loading the distribution network unnecessarily with reactive power, compensation is normally effected by means of capacitors near to the motor.

The frequency converter generates the reactive power required by the motor, and no compensating capacitors are needed. The cost of investment is reduced, and an optimum compensation effect achieved.

Improved control effect An improved control effect is more easily achieved with speed control than with other non-linear forms of control. A drawback of intermittent duty, for example, is the discontinuity of the control. The controlled parameter, the flow or pressure, for example, keeps varying. An accurate and linear control is achieved with the frequency converter. Fig. 8 shows the control graph of a plant with three parallel connected pumps P1, P2 and P3. When one of the pumps (P1) is provided with the frequency control, a linear control curve (R) is obtained, whilst the curve (T) of intermittent control is stepped, which can lead to abrupt variations in the pumped volume of liquid.

Reduced maintenance costs When a frequency converter control is used, the pump, pipes and valves experienced less wear, which means increased service life and reduced maintenance. Static stress reduced because the system need not operate with high pumping pressure all the time as with choke control. The pressure is as high or low as required. Dynamic stresses are far lower with the smooth frequency converter control than with an intermittent start-stop control. Pressure strokes (Fig. 9) which wear the pipes and other plant equipment can thus be avoided, and the service life may even be doubled.

FIG 5.59 (FIGS 7,8,9)

Page 5-56

CHAPTER FIVE
Calculation of the energy savings achieved by using a frequency converter: The following data will be required: Pump data Pump type and manufacturer HN m Rated head of the pump 3 QN m /h Rated capacity of the pump p % Efficiency of the pump nN rpm Rated speed of the pump Process data Range of application Hst Hmin Qmin QM ta Motor data UN PM Other data k cost/kWh Unit price of energy V kW Supply voltage of the motor Rated output of the motor m m m3/h m3/h kg/dm3 h Static head of the plant Min. required head Min. required pumped quantity Average pumped quantity Density of the pumped liquid Operating hours per year

FIG 5.59 (FIGS 10a): PUMPS CONTROLLED WITH FREQUENCY CONVERTER IN WATER WORKS
AND SEWAGE TREATMENT PLANTS

Page 5-57

CHAPTER FIVE
Calculation of the savings:
1. Power requirement is determined - either by means of the pump characteristics The power requirement with choke control P1 and with speed determined by means of the pump control P2 can best be characteristic curves provided that complete curves are available. P1 and P2 are determined according to Fig. 10. When QM and the system characteristics are known, the pump curves n1and n2 as well as the corresponding power curves P1 and P2 are found. The power requirements can be read at the point where QM and the power curves intersect. - or by means of calculation If the complete power characteristics are not available, P1 and P2 are calculated from the following formulae. Choke control P1: P1 = QM[m3/h] H1[m] [kg/dm3] g[m/s2]kW 3600 1 Speed control P2: P2 = QM[m3/h] H2[m] [kg/dm3] g[m/s2]kW 3600 2 2. Calculate the power saving The power saving obtained by means of the frequency converter control is: Ps = P1 P2 , 0.9 Where the divisor 0.9 is the approximate efficiency of the motor. 3. Calculate the energy saving The saving in energy per year is obtained when power saving is multiplied by the operating hours, that is: Ws = Ps ta = P1 P2 ta 0.9 4. Saving in money The saving in money per year is obtained when the energy saving is multiplied by the unit price of energy k.
Ks [money saved per annum] = Ws [kWh/a] x k [price/kWh]

5. Cost pay-off time The cost pay-off time is obtained by comparing the coat difference Kp between the speed control and the choke control with the achieved saving per year Ks:

t t=

Kp [cost]
Ks[money saved per annum]

FIG 5.60 ( FIGS 10)

Page 5-58

CHAPTER FIVE
5.3.2 Diesel Engines and Generators The diesel engine is a type of internal combustion engine that uses compression ignition, in which fuel ignites as it is injected into the combustion chamber. In contrast, in petrol engines, fuel is mixed with air first and then ignited by a spark plug. Diesel engines have compression ratios high enough to achieve compression ignition, whereas in petrol-burning engines, compression ignition is undesirable. A diesel engine ignites fuel by compressing gases and raising their temperature. Air is drawn into cylinders and is compressed by the moving piston at compression ratios as high as 25:1, higher than used for a spark-ignition engine. At the end of the compression stroke and at the start of the power stroke, diesel fuel is continuously injected into the combustion chamber at high pressure through an atomizer. The fuel ignites from contact with heated air whose temperature is about 700900C. Combustion causes air in the chamber to heat, rapidly increasing its pressure, which in turn forces the piston downward. A connecting rod transmits this motion to a crankshaft, which converts linear to rotary motion and delivers power to its output shaft. Scavenging (pushing products of combustion out of the cylinder and drawing in a fresh charge of air) of the engine is done either by ports or valves. To increase power, a diesel engine may, by mechanical supercharger or by an exhaust turbine, have a turbocharger to increase intake air volume and use an aftercooler/intercooler to cool intake air after compression by a supercharger to improve efficiency. Fuel injection and timing: Older injection systems are driven by a gear system from the engine and thus supply fuel in proportion to engine speed. Modern, electronically controlled engines apply controls similar to those of petrol engines and limit the maximum RPM through an electronic control module (ECM) or electronic control unit (ECU) - the engine-mounted computer. The ECM/ECU receives an engine speed signal from a sensor and controls the amount of fuel and (start of injection) timing through electric or hydraulic actuators. Controlling the timing of the start of injection of fuel into the cylinder of a diesel engine is key to minimizing emissions, and maximizing fuel economy (efficiency), of the engine. The timing is usually measured in units of crank angle of the piston before Top Dead Centre (TDC). For example, if the ECM/ECU initiates fuel injection when the piston is 10 degrees before TDC, the start of injection, or timing, is said to be 10 deg BTDC. Optimal timing will depend on the engine design as well as its speed and load. Advancing the start of injection (injecting before the piston reaches TDC) results in higher incylinder pressure and temperature, and higher efficiency, but also results in higher emissions of oxides of nitrogen (NOx) through higher temperatures. At the other extreme, delayed start of injection causes incomplete combustion and emits visible smoke made of particulate matter (PM) and unburned hydrocarbons (HC). Classes of diesel engines: There are two classes of diesel (and petrol) engines; two-stroke and four-stroke. Most diesel engines use the four-stroke cycle, but some larger diesels operating on the two-stroke cycle, Large diesel generators often use a two-stroke diesel mated to a generator, which produces current to drive electric motors, eliminating the need for a transmission drive. To achieve operational pressure in the cylinders, two-stroke diesels must utilize forced aspiration from either a turbocharger or supercharger. Diesel two-strokes are better for such an application because of their high power density - with twice as many power strokes per crankshaft revolution compared to a four-stroke. As a result, they are capable of producing much more power per displacement.

Page 5-59

CHAPTER FIVE
Cylinder banks: Normally, banks of cylinders are used in multiples of two, although any number of cylinders can be used as long as the load on the crankshaft is counterbalanced to prevent excessive vibration. The inline-6 is the most prolific in medium- to heavy-duty engines, though the V8 and straight-4 cylinder engines are also common. Small-capacity engines (generally considered to be those below 5 litres in capacity) are generally 4- or 6-cylinder types, with the 4-cylinder being the most common type found in automotive uses. Diesel engines for smaller plant machinery, generators and pumps may be 4-, 3-, or 2-cylinder types, with the single cylinder diesel engine remaining for light stationary work. Emissions: When properly maintained, diesel engines produce very little carbon monoxide as they burn the fuel in excess air even at full load, at which point the quantity of fuel injected per cycle is still about 50% lean of stoichiometric. However, they can produce black soot (or more specifically diesel particulate matter) from their exhaust, which consists of unburned carbon compounds. This is often caused by worn injectors, which do not atomise the fuel sufficiently, or a faulty engine management system which allows more fuel to be injected than can be burned completely in the available time. The full load limit of a diesel engine in normal service is defined by the "black smoke limit", beyond which point the fuel cannot be completely combusted. As the "black smoke limit" is still considerably lean of stoichiometric it is possible to obtain more power by exceeding it, but the resultant inefficient combustion means that the extra power comes at the price of reduced combustion efficiency, high fuel consumption and dense clouds of smoke, so this should be avoided. Likewise, when starting from cold, the engine's combustion efficiency is reduced because the cold engine block draws heat out of the cylinder in the compression stroke. The result is that fuel is not combusted fully, resulting in blue/white smoke and lower power outputs until the engine has warmed through. This is especially the case with in-direct injection engines which are less thermally efficient. With electronic injection, the timing and length of the injection sequence can be altered to compensate for this. Older engines with mechanical injection can have manual control to alter the timing, or multi-phase electronically-controlled glow plugs, that stay on for a period after start-up to ensure clean combustion. The plugs are then automatically switched to a lower power to prevent them burning out. Particles of a size normally called PM10 (particles of 10 micrometres or smaller) have been implicated in health problems, especially in cities. Some modern diesel engines feature diesel particulate filters, which catch the black soot and when saturated are automatically regenerated by burning the particles. Other problems associated with the exhaust gases (nitrogen oxides, sulphur oxides) can be mitigated with further investment and equipment such as catalytic converters in the exhaust. Reliability: The lack of an electrical ignition system greatly improves the reliability. The high durability of a diesel engine is also due to its overbuilt nature as well as the diesel's combustion cycle, which creates less-violent changes in pressure when compared to a spark-ignition engine, a benefit that is magnified by the lower rotating speeds in diesels. Diesel fuel is a better lubricant than petrol so is less harmful to the oil film on piston rings and cylinder bores; it is routine for diesel engines to run for considerably longer periods than petrol engines without need for a rebuild. Diesel Generators: A diesel generator is a combination of a diesel engine with an electrical generator (often called an alternator) to generate electric energy. Diesel generators are used in places without connection to mains power or as an emergency power-supply source if the mains

Page 5-60

CHAPTER FIVE
supply fails. Small portable diesel generators range from about 1 kVA to 10 kVA, while the larger industrial generators can range from 8 kVA up to 2000 kVA for large schemes. They have the advantage that the generators do not need to be coupled to the pumps, can provide auxiliary power and lighting, and can be placed in outside compounds away from electrical equipment. Diesel generating capacity is selected based on the load they are intended to supply power for, and the importance of the requirement. De-rating: It is necessary to de-rate a diesel engine to allow for elevation, ambient temperature and humidity as well as for continuous running. Altitude has a considerable effect on the efficiency of a diesel engine. A loss of output of approximately 6% per 500 m elevation above sea level will result. In addition, an output loss of 0.3% should be expected for every oC above an ambient temperature of 25oC, and for each 10% rise in humidity above zero a further loss of 0.4 1.6% will occur. When a diesel powered generator is used as a standby plant, there is normally no need to de-rate it further. However for continuous operation either as a direct coupled pump prime mover or as a generator it is considered good practice to apply an additional de-rating of 20% if being operated at more than 70% of maximum power. Curves for de-rating of a naturally aspirated diesel engine for altitude in Figure 5.60, and for humidity and ambient temperature in Figures 5.61.

FIGURE 5.60: ALTITUDE DE-RATING FOR A NATURALLY ASPIRATED DIESEL ENGINE

Naturally aspirated diesel engines struggle to maintain speed when under de-rating effects, the symptoms being black smoke heavy with particulates from the exhaust and overheating around the manifold and cylinder head. Only by lowering the speed can the problem be overcome. Turbocharged diesel engines have different de-rating requirements and the manufacturer needs to be consulted on this.

Page 5-61

CHAPTER FIVE

FIGURE 5.61: TEMPERATURE AND HUMIDITY DE-RATING FOR A DIESEL ENGINE

5.4

PUMPING PLANT CATEGORIES

Pumping plant falls into two basic categories namely surface water pumping and groundwater pumping, although there is inevitably some overlap between the two. 5.4.1 Surface Water Pumping

As discussed in Section 5.2.3, it is always preferable to avoid negative or suction lift conditions with rotodynamic pumps such as centrifugal, mixed flow and axial pumps and to design pumping stations such that there is always a positive suction. Only in the case of raw water surface pumping is this sometimes unavoidable due to the extra civil engineering costs far exceeding the costs of the different pumping arrangement. When negative suction situations are unavoidable, then the following must be observed. 5.4.1.1 Suction Lift Pumping Such pumping plant usually has the following components: An intake channel or pipe A wet pump sump A continuously rising suction main comprising flanged pipework. A pumphouse to accommodate the pumpsets according to figs 1 or 2 A surge tank complete with ancillary equipment Electrical systems wherever applicable according to the requirements of the prime movers. The flow velocity in the intake channel shall not exceed 0.45 m/s in cases where the source is populated with fish, or 0.30 m/s where there is high sand content in the water velocities below 0.l5 m/s have been very successful in lake intakes. The intake channel shall be protected by coarse and fine screens that are either inclined when to be cleaned in-situ or vertically mounted and easily removed for
Page 5-62

Considerations a.

b.

CHAPTER FIVE
cleaning. For removable screens, a set of replacement screens should always be kept in stock and be fitted immediately upon removal for cleaning of the set previously in use. The removed set should be carefully inspected and if in a suitable condition cleaned for subsequent use. However if either is in poor condition a replacement should be ordered immediately. The velocity in the suction line shall not be more than 1.5 m/s and its minimum diameter D selected accordingly. The suction line should be as short as possible and shall be always protected by a strainer and foot valve of at least twice 2 D. It shall have a continuously rising slope from the sump foot valve to the pump suction inlet. To ensure this and unless the suction pipe can be mounted vertically in profile until immediately before the pumphouse or the pump itself, as shown in Figure 5.52, two long 45 bends shall be used one immediately above the sump and the other at the pumphouse wall. The alignment of the intake channel to the pump sump shall be straight and terminate in a sump at right angles to the well. Individual isolatable sumps shall be provided for each pump or it shall be sub-divided into two separate isolatable compartment each readily accessible for cleaning and flushing from outside any pumphouse superstructure. When individual sumps are provided, means of lifting the suction pipe complete with strainer shall be provided, even if this is in the form of a removable set of shear legs. Pipework design shall be such that the removal of only one set of bolts from the uppermost flange of the lower 45o bend shall provide clear access for a clean, vertical lift of the pump strainer, riser pipe and bend. The bend shall be provided with an off-centre lifting eye positioned so that such a vertical lift is possible. The pump inlet may be preceded by a straight pipe whose length shall be not more than 8 times its diameter. Any taper needed to change the size of the suction pipe down to that of the pump suction inlet shall be an eccentric taper with the taper eccentricity on the bottom (invert) side of the pipe. Minimum sump dimensions shall be such that in plan view it is not less than the diameter of the strainer + the diameter of the suction pipe. The clearance beneath the strainer inlet shall normally be the diameter of the suction pipe, and the minimum submergence to the underside of the strainer shall be not less than, D + V2 / 2g The site NPSHA shall be greater than the NPSHR by at least 1m. (See Section 5.2.5.) Where there is a reliable electricity supply or a standby diesel generator is affordable, and the technical expertise to maintain such units can be assured, the use of raw water submersible pumps may be considered instead of surface pumps.

c. d.

e.

f.

g.

h.

i. j.

5.4.1.2 Gravity inflow Source Pumping Pumping from a spring shall be arranged such that a positive suction head is ensured on the pump suction branch. This arrangement minimizes fitting requirements on the suction e.g. no foot valve is required however a strainer must be provided. Important Considerations are: a. b. A common suction pipe shall be used in this case; however, velocity shall not exceed l.5 m/s. The dimension of the common suction line shall be such that its cross sectional area is not less than that of the total area of the suction branches leading to the pumps.
Page 5-63

CHAPTER FIVE
c. d. The common suction branch shall enter the pumphouse at right angles to the walls of the pumphouse. Where electricity is available provision shall be made to accommodate protection against dry running. Where diesel engines are used as prime movers, the diesel engine can be fitted with an alternator in order to recharge batteries. This: alternator can also be used in connection with a magnetic valve to shut the diesel supply line to the engine, so as to provide protection against dry running. Alternatively, consideration may be given to the provision of a solar powered battery charger. Where the suction static pressure is reasonably high and the total static head is relatively small the design shall be arranged such that water hammer protection is effected by a by-pass from the suction main to the delivery manifold which is fitted with a non-return valve that allows flow from the suction to the delivery line so as to prevent water column separation at the pumphouse. The type of non-return valve selected should be of the non-slam shut type.

e.

5.4.2

Pumping from an Inaccessible Spring Where accessibility is very remote and the accessible area closest to the source is at a higher level and the negative suction conditions from the source to the accessible area falls outside the permissible limits of surface pump application, pumping from this type of source can be done by a submersible pump designed for horizontal mounting. If necessary a diesel generator must be provided. Pumping shall be designed to operate on remote control principles using electrical power to be provided at an accessible point. The area at which electrical power must be produced from a diesel generating set shall be accessible to vehicular traffic but away from the spring so as to avoid the risk of pollution. The power supply to the inaccessible source from the generating set or overhead line shall be transmitted through a suitable waterproof or marine type underground cable at a distance not exceeding 200 m

5.4.2.1 General Considerations

5.4.2.2 Electrical power system design for remote control In circumstances where remotely controlled electrical power systems are unavoidable, then the following must be included. a. The control voltage across the operating coil of a remotely actuated contactor shall not be less than 85% of its rated operating voltage. b. As voltage fluctuations in any power supply system amounting to 10 % are not uncommon, the system voltage can be expected to fall to 90%. Hence the system design should be such that the voltage drop due to the control current in the control cable is limited to 5%. c. The attraction power is very important in remote control of a contactor in as far as the voltage drop is then given approximately by: DU = Ik Rv Cos.k 5%U Where, Ik = the attraction current 5.5

Page 5-64

CHAPTER FIVE
Rv = the ohmic feeder resistance Cos.k = the power factor at .the contactor U = the requisite and rated control voltage The ohmic feeder resistance, Rv = L / ( A) (ohms) Where, L = total length of the feeder (m) (the length both up and down) = conductivity of material in m / mm A = cross sectional area of the feeder cable (mm2) Hence in dealing with remote control systems these parameters must be taken into consideration to ensure that the available voltage at the remotely actuated contactor is sufficient. This is even more important where low voltage must be used d. Capacitance: Whilst the ohmic voltage drop has an adverse effect particularly with lower control voltages and especially when large contactors must be actuated, the capacitance of the control cable has an inverse influence Two main situations are frequently encountered connected with control head capacitance namely: (i) A case whereby the contactor to be actuated is at the power supply point and the switch is located remote from the supply point, and (ii) A situation which is just the reverse of the case above. For more information, reference shall be made to recent editions of Electrical Machines Standard books. 5.4.3 Pumping from Deep Wells When pumping from deep wells the following must be considered: a. For flows exceeding 45 m3/h and where the corresponding head is more than 30 m, the site structure shall be designed to accommodate a submersible pump. Only a simple house to provide waterproof accommodation for the pump starter, control and monitoring instruments need to be included as part of the site structure. For flows below 45 m3/h (12.5 l/s) and where the hydraulic conditions can be met by a helical rotor, Mono type of pump (see Section 5.2.1.1.3) For cases within the operating ranges covered by surface pumps but whereby the geotechnical stability of the soil is questionable, a submersible pump shall be given priority consideration to avoid the risk of breakage of drive shafts due to excessive eccentric shaft rotation. The system components of the installation shall include protective devices against dry running as indicated hereinafter. Pumping from Shallow Wells Where the total pumping requirement involves a high head and the hydraulic suction conditions are within the permissible operating range of a surface pump, the shallow well shall be treated as a pump sump and the design criteria under Section 5.4.1.1 shall apply.
Page 5-65

5.6

b. c.

d. 5.4.4 a.

When pumping from shallow wells the following must be considered:

CHAPTER FIVE
b. c. For a source with a very limited safe yield and where a storage tank is not provided, the shallow well shall be designed to accommodate a hand pump. In case where the suction conditions are prohibitive as far as a surface pump application is concerned, the shallow well shall be designed to accommodate either a submersible pump if electrical power is available or a horizontal reciprocating pump where the pump must be diesel engine powered. Booster Pumping

5.4.5

When the total lift is such that it cannot be achieved by a single pumping stage or where increasing demand requires that extra water is conveyed through an existing water main, then boosting is necessary. Given the limitations in waterworks appliances and pipe pressure classes, the absolute maximum, closed valve, pumping head is limited to 25 bar as beyond this special industrial valves and other components become necessary. Hence booster pumping is decided by the limiting pump characteristic where single stage pumping is not possible due to practical limitations of pump application, standard appurtenances or occasionally by complicated water hammer problems: It is also used for pumping from a distribution system to a storage tank located in a high building or to a public storage tank located at a point higher than the previously designed and installed distribution level. 5.4.6 Pumping Capacity Criteria to be considered when determining pumping capacity include: a. The maximum pumping capacity of a pump station shall, excluding standby capacity, be fixed to match with the hourly average consumption of the scheme at the end of the design period; based on a 24 hour operation in the case where a reliable electrical power supply is provided. b. For diesel engine powered pumps, the pumping capacity shall be fixed in accordance with the stipulation above but based on a pumping period of not more than 14 hours per day. c. The maximum pumping capacity from a groundwater source shall be dictated by the safe yield of the water source in cases where electricity is available. d. Where the reliability of electrical power supply is questionable, a standby diesel generating set shall be included to cater for at least 50% of the maximum pumping capacity of the station. 5.4.7 Pumping Period For electrically driven pumps the pumping period may be as long as 24 hours a day. Where conventional treatment plants are used involving settling and filtration, pumping shall be based on 24 hours/day. 5.4.8 Pumping Sequence

For plants involving more than one pump, an alternating sequence operation shall be planned so that none of the installed pumps is left idle for too long and the total number of working hours of each pumpsets should be kept more or less the same. A period for alternating between pumpsets of one month, coincident with electrical readings for monthly bills is suggested.
Page 5-66

CHAPTER FIVE
5.4.9 Pump Arrangement In surface water pumping, most arrangements comprise at least two pumps: one set as duty and the other on standby. This arrangement means that if the duty pump fails to start, the standby can be used to avoid production loss. To avoid accumulation of very old water in the standby system, the allocation of the duty and standby pumps should be alternated periodically to ensure that the standby pump remains functional and will spread the wear over both (or all) pump sets. 5.5 PUMPING SYSTEMS

Pumping systems need to be considered, dependant upon the category into which they fall. Pumping systems are however likely be in one of the following categories: a. b. c. d. e. f. 5.5.1 Single stage pumping system Two stage pumping system Branched pumping system Distribution booster pumping In line booster pumping Domestic booster pumping systems. Single Stage Pumping

Single stage pumping is used where the total pumping head required for the supply system is efficiently and economically met by the pump selected. 5.5.2 a. b. c. d. 5.5.3 Two Stage pumping The total system head of the entire water supply cannot be efficiently met in a single stage by the whole range of pumps available. Flow from intake pumps is to a water treatment plant which cannot command the supply area by gravity. The pump duty pressure exceeds the working pressure of the pipes close to and within the pumping station. The hydraulic surge at the pumping station would be dangerous even to any surge equipment and the cost for the surge protection installation would be prohibitive. Branched Pumping Systems.

Stage pumping is adopted where:

A branched pumping system may be adopted when the water supply is intended to meet water demands of isolated consumption areas located at different elevations and/or separated by a distance which make it uneconomical to make direct connection from one area to another 5.5.4 In Line Boosting If in-line boosting is adopted then: a) The incoming flow capacity to the pump is not less than 110 % the capacity of the booster pump; b) The maximum pressure at the pump suction branch does not exceed the maximum permissible suction pressure of the pump; and
Page 5-67

CHAPTER FIVE
c) The nominal pressure at the suction branch is not less than the NPSH required of the pump + 1 m whilst the pump is in operation In deciding to adopt an in-line booster pumping system the following shall be the criteria: (i) There is non-availability of adequate space to accommodate a conventional surface booster pump station at the site which provides optimum technical (electrical and hydraulic) conditions; (ii) there is need to increase the flow capacity of an existing main; (iii) noise disturbance to the surrounding area arising from a conventional booster pump station would be unacceptable to the nearby population; (a noise level of 86 decibels and above is considered unacceptable to the human ear) (iv) economy; (v) simplicity from the technical point of view; (vi) availability of reliable electrical power supply; and (vii) there are sufficient residual pressure requirements in the upstream pumping main. 5.5.5 Maximum Suction Lift Calculation

In order to ensure continuous and smooth operation of any rotodynamic pump such as a centrifugal pump there is a limit to the net positive suction head available (NPSHA) and hence the suction lift that can be achieved. The maximum suction lift shall be determined in accordance with the following formula: Hs1(max) = Pa (NPSHR. + Vp + Hfs) (m) Where, Pa VP Hfs NPSHR = Where, B = a value dependant upon the altitude and the water temperature and thus on the barometric pressure and vapour pressure of the water in metres head of water Hsuc = static height difference on the suction side of the pump in m (i.e. between the centre line of the pump element and the water level on the intake side of the pump). The value is negative if the pump element is located above the water level on the intake side of the pump and positive when the pump is below the water level.
Notes: (1) The NPSHR. curve should be provided by the pump manufacturer or his agent. Otherwise the NPSHR value must be obtained from the manufacturers catalogue or if even provided therein, it must be confirmed from the relevant manufacturer who should give a written guarantee as to the value appropriate for the design head-flow point of the pump.

5.7

= atmosphere pressure = vapour pressure at the given temperature of the water source = friction losses in the suction line 5.8 net positive suction head required = B + Hsuc - Hfs

Page 5-68

CHAPTER FIVE
(2) The term (Pa - Vp + Hp + Hfs + Hsl) = NPSHA, which for continuous operation must be at least 1 m.

Values which may be used for B assuming a water temperature of 30oC are given in the following table:
TABLE 5.6: ALTITUDE AND TEMPERATURE VALUES (FACTOR B) MASL B 0 9.4 500 8.9 1000 8.4 1500 7.9 2000 7.4

Head Loss in m

The head loss through a strainer and footvalve is a significant part of the headloss in a suction pipe system and tends to be under estimated. Suggested values for units in reasonably clean condition are given in the Figure opposite: Because it is such an important factor in determining Pump performance, but so often poorly understood or incorrectly assessed a simple example is given below:

1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2

Velocity of Water in m/s

FIGURE 5.62: HEADLOSS THROUGH STRAINER AND FOOTVALVE

A pump is to be located 4 m above the minimum water level in an inlet sump and is to be used to pump 50 l/s. The suction pipe is 200 mm dia., 10 m long and has two 45o bends. There is a strainer with a footvalve at the inlet of the pipe at the altitude is 1,100 masl. Question: What should be the NPSH of the pump? Solution: Hsuc = - 4 m Hfs = 0.40 m (strainer & footvalve) + 0.07 m (bends) + 0.20 m (10 m of pipeline) = 0.67 m B (from table 5.5) = 8.3 m. Hence, NPSHR. = 8.3 4 0.67 = 3.63 m. Therefore it is necessary to select a pump which has an NPSH of 3.63 metres or less for the capacity of 50 l/s.

5.6

EQUIPMENT TYPE DESIGN AND SELECTION

In selecting pumping equipment, type designs should be adhered to whenever possible. 5.6.1 Type Design The following equipment type designs are recommended for the respective use indicated:

Page 5-69

CHAPTER FIVE
a) Horizontally mounted surface rotodynamic pumps for surface pumping with a positive suction head whenever possible b) Vertically mounted surface rotodynamic pumps with a positive suction head where space is restricted. Also to be considered for large units where 3.3kV motors are required. c) Submersible pumps for surface water sources pumping where a positive suction head is not possible and an acceptable level of maintenance is available d) Submersible pump type for in-line booster pumping (See pumping systems) e) Submersible pumps for pumping from shallow wells or from lake intakes where an acceptable level of maintenance is available. (This type of submersible pump type differs from the aforesaid pump type design.) f) Helical rotor riser shaft (e.g. Mono type pump) for pumping relatively limited quantities from wells. This type of pump is especially suited to rural water supply schemes as mentioned elsewhere. g) Vertical spindle turbine pumps for river intakes and wells where earlier mentioned types cannot be used for whatever reason. h) For chemical dosing, horizontal (reciprocating type) metering pumps with an appropriate (GRP) shaft should be used. Chemical Dosing Pumps All dosing pumps must be of high quality design, use appropriate materials for the chemicals being handled and be accurate to enable a known precise relationship between amount being dosed and the amount of water being treated. The maximum stroking speed (rpm) of reciprocating pumps should be about 100 rpm and adjustment of dose is achieved by altering the pump stroke length. Such metering pump types with GRP shaft are robustly built and have much less maintenance, and minimal running, and spare part procurement problems than do diaphragm dosing pumps. It should be noted in this connection that the diaphragm type metering pumps have proved difficult to maintain due to fast wearing of the diaphragm, the material for which is not locally available in Tanzania. Chlorine Dosing When dosing from chlorinators, the close coupled single stage pump types are suitable where the residual pressure may not be sufficient to operate a conventional chlorinator. The V-notch, chlorinator type is suitable for chlorination of treated water or raw water from a groundwater source. 5.6.2 1) 2) 3) 4) 5) 6) 7) Selection Stability characteristics Efficiency NPSHR NPSHA Rotational speed, Na. Specific speed, Ns Overload margin on the prime mover

In selecting pumping equipment the following points need to be considered:

Page 5-70

CHAPTER FIVE
8) 9) 10) 11) 12) 13) 14) 15) Water quality i.e. pH value and silt content Power requirement and power consumption pattern Material properties of the components Capital cost Running spare parts requirement Standardization Adaptability to local site conditions De-rating and fuel consumption pattern if a diesel engine prime mover is being considered. In this connection it is advisable to have the fuel consumption curves from possible manufacturer and make a comparison. This is the best way to compare efficiencies of prime movers from different manufacturers as both would indicate the fuel consumption curves at various comparable speeds, Diesel engine cooling requirements. Air cooled engine types should be considered the best from the operation and maintenance point of view. However, there are limits as to the size which can be air cooled. Large engines e.g. 150 Hp. would normally be water cooled. The moment of inertia of the pump set and engine or pump and motor must also be considered in connection with water hammer analysis and protection requirements. Prime mover and pump torque characteristics are required for open valve, closed valve, and check valve opening conditions. Pump and Motor Identification Plates. These must be fully completed, as they provide key information regarding the equipment installed. Any pump or motor supplied with only part or no information on these plates should be rejected and the supplier instructed to remove it until the complete information is indicated.

16) 17) 18)

5.6.3 Metallic Materials of Pump Construction (and their Damage Mechanisms) The requirements for a successful pump installation are performance and life. Performance is the rating of the pump head, capacity, and efficiency. Life is the total number of hours of operation before one or more pump components must be replaced to maintain an acceptable performance. The initial performance is the responsibility of the pump manufacturer and is inherent in the pump design. Life is primarily a measure of the resistance of the materials of construction to corrosion, erosion, wear, and other factors that can influence the materials when the pump has been placed in service. The need to maximize reliability and extend the pump life makes the selection of appropriate materials of construction crucial. The selection of materials that are both cost-effective and technically suitable for the application requires a knowledge not only of the pump design and manufacturing processes, but also of the engineering properties of the material, particularly its corrosion and wear resistance properties when subjected to the conditions encountered in the pump. Sufficient information is available in the corrosion and metallurgical literature as well as from the experience of pump manufacturers to make appropriate material choices for virtually any pumping application. It is known that several factors lead to a long pump life. These include: Neutral liquids at near-ambient temperatures Appropriate material selections for pumps in aggressive services The absence of abrasive particles Continuous operation at or near the maximum efficiency capacity of the pump
Page 5-71

CHAPTER FIVE
An adequate margin of available NPSH over NPSH required as stated on the manufacturers rating curve A low velocity (developed head/rotative speed)

Pumping installations that satisfy all these criteria will have a long life. A typical example would be a waterworks pump. Some waterworks pumps with bronze impellers and cast-iron casings have a life of 50 years or more. At the other extreme might be a chemical pump handling a hot corrosive liquid with abrasive particles carried in suspension. The life of this pump might be measured in months rather than in years, despite the fact that construction was based on the most resistant materials available. Most pumping applications fall somewhere between these two extremes. Aside from straight corrosion or erosion from abrasive particles in fluid, the greatest single factor that reduces pump life is operation at flows other than the maximum efficiency or rated capacity of pump. The vane angles of the impeller are designed to match the fluid angles at maximum efficiency capacity. At flows other than the rated capacity, the fluid angles no longer match the vane angles and separation occurs with increasing intensity as the operating moves away from the maximum efficiency capacity. The destruction of the impeller vanes is particularly severe at the inlet to the impeller, as this is the point of lowest pressure in the pump. In addition to surface damage to the inlet vanes from separation, localized cavitation damage may occur during sustained operation at capacities below 50 or above 125% of maximum efficiency capacity. This is not to say that many pumps do not operate under these adverse conditions, but their life is considerably less than that of same pump operating at or near its maximum efficiency capacity. Therefore, the pump designer needs to be familiar with the various types of degradation that can affect the components of the pump and reduce its useful life. These can be grouped into the general categories of corrosion, wear, and fatigue, with corrosion and wear being the predominant life-limiting mechanisms 5.6.3.1 Types of Corrosion General Corrosion General corrosion is corrosion that proceeds without an appreciable localization of attack. This type of corrosion occurs on metals or alloys that do not develop an effective passive film on the surface. Usually, the corrosion mechanism is oxidation with the formation of metal oxide corrosion products. General corrosion is most often encountered in pumps with carbon steels and copper base alloys. Cast irons also experience a specialized form of general corrosion, known as graphitic corrosion, which will be considered separately. Carbon steel does not develop a protective oxide film and will corrode at a rate dependent upon several characteristics of the water or other fluid, including temperature, oxygen content, pH, and fluid chemistry. The rate of corrosion is also very dependent on velocity and increases with an increasing velocity. In most pump applications, with the notable exception of hydrocarbons, the corrosion rate of carbon steel is too high for this material to provide a useful life. However, carbon steel is frequently used, particularly in vertical pumps, with some form of protective coating to prevent corrosion. Epoxy is a preferred coating for many water services. Copper alloys, including both brasses and bronzes, are also subject to general corrosion in the water applications where they are most commonly used in the pump industry. The corrosion rate will be increased by the presence of small amounts of sulphides in the water. Copper alloys gradually develop a protective copper oxide corrosion film in most applications. The corrosion rate gradually decreases over time as this film develops. The
Page 5-72

CHAPTER FIVE
rate of general corrosion varies with the specific type or grade of copper alloy. Among the alloys commonly used in pumps, nickel aluminium bronzes have the lowest corrosion rate and best tolerance for higher velocities. Dealloying Dealloying is the preferential removal of one phase from a multi-phase alloy, or one element from a material. Several types of dealloying occur in the pump industry. One of the most common is the graphitic corrosion of gray cast iron. This material is low cost, easy to machine, and well suited for a variety of applications, especially in the waterworks industry. It is probably the most widely used material in the pump industry. Gray cast iron corrodes by a fundamentally different mechanism than carbon steel or ductile cast iron. The structure of gray cast iron consists of interconnected graphite flakes in a matrix that is predominantly iron. In the presence of an electrolyte, which is usually water, a galvanic cell is established between the iron and graphite. The iron corrodes, and the corrosion products are largely flushed away with the fluid passing through the pump. The original casting is gradually reduced to a porous graphite structure that may contain some iron oxide corrosion product. This is frequently referred to as graphitization. Several other types of dealloying can also occur in pumps. Brass and bronze alloys containing more than about 14 percent zinc are subject to a form of dealloying known as dezincification. The final type of dealloying that occasionally occurs in pumps is dealuminification in aluminium bronzes. Galvanic Corrosion Galvanic corrosion refers to the corrosion that occurs when one alloy is electrically coupled to another and exposed in a conductive liquid. Usually, the corrosion rate of the more noble alloy will be less than if it were exposed uncoupled. The corrosion rate of the less noble material will be greater than if it were exposed uncoupled. Several factors influence the rate of galvanic corrosion of both metals. This corrosion is greatly influenced by the conductivity of the fluid. In a fluid such as fresh water, which has a low conductivity, galvanic corrosion will be less severe and generally confined to the immediate location where the metals contact one another. However, in a highly conductive fluid, such as seawater, galvanic corrosion will be more severe and will occur over a wider area. The pump designer needs to consider the possibility of such corrosion when using dissimilar metals in a conductive fluid. Galvanic corrosion problems in seawater and other conductive fluids can be avoided by the careful use of materials. Galvanic corrosion is related to the area ratios of the coupled metals. It is always desirable to have the area of the anode, or less noble metal, equal to or greater than that of the more noble metal. The amount of corrosion that will occur in a galvanic couple also depends on the freely corroding potentials of the coupled metals. Less corrosion-resistant metals, such as zinc, cast iron, and steel will usually have more negative potentials when measured against a standard reference electrode. More corrosion-resistant metals, such as stainless steels, will have less negative potentials. The pump designer needs to be aware of the corrosion potentials of dissimilar metals used in conductive fluids in order to avoid unanticipated galvanic corrosion problems.

Page 5-73

CHAPTER FIVE
The use of coatings can decisively alter the galvanic relationships in a pump. If the more anodic component, such as a steel casing, is coated, one can expect a high rate of corrosion at those locations where the coating eventually begins to fail. This will be caused by a very unfavourable area ratio, with a small area of exposed carbon steel coupled to a large area of some more noble metal, such as stainless steel or bronze. For this reason, coatings should be employed with caution in pumps handling conductive fluids that are constructed of dissimilar metals. It is generally advisable in these applications not to coat the anodic component. Stress Corrosion Cracking Stress corrosion cracking (SCC) is a particularly dangerous form of corrosion because it is not easily detected before it has progressed to such an extent that it can cause sudden catastrophic damage. Although relatively uncommon in pumps, it can occur in several classes of materials. The pump designer should be aware of the potential combinations of material and environment that can cause SCC. Stress corrosion requires that several factors be present. These include tensile stress, which can be either residual or applied, a susceptible material, an environment capable of causing stress corrosion, and time. Hydrogen Embrittlement Hydrogen damage is a form of environmentally assisted failure that results from the combined action of hydrogen and residual or applied tensile stress. Hydrogen damage to specific alloys or groups of alloys manifests itself in many ways, such as cracking, blistering, hydriding, or as a loss of tensile ductility. Collectively, these various forms of damage are often referred to as hydrogen embrittlement. Damage caused by hydrogen is occasionally encountered in pumps. Some plating processes, such as chrome plating, which is often used to rebuild pump shafts, generate hydrogen. This hydrogen can enter the surface of the metal. Microscopic cracks can occur in higher strength steels (greater than 620-MPa yield strength). Hydrogen can also be introduced into metals during welding. In order to avoid the hydrogen damage associated with welding, ferritic and martensitic steels should be welded with low hydrogen electrodes. Microbiologically Induced Corrosion Living organisms can promote corrosion in many different environments. A variety of biological organisms thrive in both aerobic and anaerobic environments. Corrosion attributable to microbiological activity occurs most frequently in stagnant water, which remains in a pump when it is shut down for an extended length of time. Sulphate-reducing bacteria are found in many waters. They will form slimy, reddish hemispherical shaped mounds or colonies on cast iron or carbon steel. These are known as tubercles. If scraped off, there will invariably be a saucer-shaped pit beneath the tubercle. The inside of the pit will contain a wet, black deposit. The pitting is caused by traces of sulphuric acid excreted by the bacteria. This type of corrosion will usually not result in premature failure. Several more serious types of microbiologically induced corrosion afflict stainless steels. A certain class of metal ion concentrating/oxidizing microbes appears to concentrate
Page 5-74

CHAPTER FIVE
ferric and manganic chlorides, both of which are potent pitting agents. These bacteria form colonies preferentially at welds in austenitic stainless steels and are capable of causing severe pitting corrosion in a relatively short time. This problem has been encountered in a variety of equipment in both salt and fresh water. It is often discovered only when the welds begin leaking. Pumps employing welded stainless steel fabrications can be afflicted by this problem if permitted to sit idle with stagnant water, either fresh or salt, for an extended period. Biocides can be used to mitigate this problem in some instances. Finally, the decay of biological organisms can generate hydrogen sulphide, which adversely affects the protective oxide film on copper base alloys. The enhanced biological activity in warmer tropical waters, especially under stagnant conditions, can impair the corrosion resistance of bronzes and reduce the threshold velocity at which accelerated corrosion will occur. Bronzes should be used with caution in applications where macrobiological activity is anticipated and the possibility of extended shutdowns is possible. Intergranular Corrosion This infrequent type of corrosion preferentially attacks a material at the grain boundaries. This is caused by local chemical differences such as the chrome-depleted regions of an austenitic stainless steel. Bronze alloys susceptible to this type of corrosion include aluminum brasses, silicon bronzes, Muntz metal, and admiralty metal. Two things are necessary: a sensitized material and a corrosive media, such as seawater. Sensitization can occur during heat treatment or more commonly during weld repair. This type of corrosion often leads to corrosion-assisted fatigue cracks when cyclic loading is present. Austenitic stainless steels can also be sensitized during normal welding procedures. Care must be taken to avoid the sensitization range during welding followed by proper postweld heat treatment when necessary. Cavitation Erosion Cavitation erosion is primarily a mechanical process, although it acts synergistically with corrosion and is often considered with other forms of corrosion. Cavitation erosion can be defined as metal removal from the surface caused by high stresses associated with the collapse of vapor bubbles in the fluid. Cavitation occurs in a pump when the local pressure of the fluid is reduced to the vapor pressure. In a multistage pump, vapor bubbles form in the low-pressure areas at the impeller inlet and are swept by the flow into regions of higher pressure where they collapse. A great many bubbles may form and collapse in a small area, producing many microjets of high kinetic energy. The energy released by the bubble collapse is expended as impact loading on the metal surface. This situation is aggravated if protective oxide films are present because these are damaged, exposing fresh metal to the corrosive action of the fluid. This cyclic loading eventually causes the formation of microscopic fatigue cracks. These cracks propagate and intersect, resulting in the removal of metal from the surface and the characteristic spongy or porous appearance of cavitation damage. An example of a cavitated impeller is shown in Fig. 5.63. Although every effort should be made in the design and application of centrifugal pumps to prevent cavitation, it is not always possible to do so at capacities less than the rated maximum efficiency capacity of the pump. It must be recognized that at a low flow operation, the stated NPSH required curve is not usually sufficient to suppress all

Page 5-75

CHAPTER FIVE
cavitation damage. The stated NPSH required is that needed to produce the head, capacity, and efficiency shown on the rating curve.

FIG 5.63: CAVITATION EROSION OF AN IMPELLER

(as indicated by the porous appearance of cavitated regions on the surface)

At low flows, some cavitation damage should be expected. It may be impractical to supply an NPSH that would suppress all cavitation at these low flows, as it could be many times that it is required at the best efficiency point. Therefore, the possibility of cavitation damage frequently becomes a consideration when selecting material for impellers. Extensive laboratory tests of the resistance of a wide range of materials to cavitation erosion have produced data for all the materials commonly used in centrifugal pump construction. It is possible to make a good correlation between the laboratory data and field experience to develop the following tabulation of the cavitation-resistance properties of pump materials, listed in order of decreasing cavitation resistance: Stellite Chrome-manganese austenitic stainless Carburized 12% chrome stainless casting Titanium 6AL-4V Cast nickel-aluminum bronze Cast duplex stainless steel Cast precipitation hardening stainless steel Ductile NiResist Cast CF-8M Cast CA6-NM Cast CA-15 Monel Manganese bronze Carbon steel (cast) Leaded bronze Cast iron

Page 5-76

CHAPTER FIVE
5.6.3.2 Types of Wear Rotating equipment, including pumps, can suffer from damage as a result of mechanisms unrelated to corrosion. The relative motion between parts that are in close proximity to each other can produce wear when these components come into contact with one another. Catastrophic damage may occur if the parts make contact under high loading conditions or when foreign bodies are entrapped between the rotating and stationary components. An accelerated material loss or catastrophic seizure of these components can result in costly repairs or replacements. Erosion, due to the presence of solid particles in the liquid being pumped, can also limit the life of internal pump components. Wear mechanisms have been categorized into more than 20 individual processes. However, only a few mechanisms are frequently recognized as damaging to a pump: Adhesive wear: material-to-material contact Abrasive wear: solids interacting with internal components Erosion: solid particle impingement Fretting: small amplitude motion of parts causing oxidation damage Identifying the wear mechanism is somewhat difficult at times as wear, or the loss of material, within a pump can result from more than one mechanism at a time. Adhesive Wear One of the primary causes of material loss on rotating components in a pump handling clear liquids (with no solids entrained in the fluid stream) is adhesive wear. This material loss is due to material-to-material contact producing surface disruptions, material grooving, a transfer of material, and possibly galling. Two important characteristics to consider for a pair of materials that may come into contact are their adhesive wear traits and their galling threshold, galling being considered a severe case of adhesive wear. Based upon extensive empirical testing and field experiences, several sound rules of thumb have been developed through the years when selecting pump wear ring materials. Three factors are used to select materials for wear surfaces in clear liquid environments: Corrosiveness of the fluid Amount of wear allowed Galling stress Corrosion determines the class of material to be used. These classes generally fall into three groupings: non-corrosive, mildly corrosive, and corrosive. A general guide for materials in several environments is as follows:

Page 5-77

CHAPTER FIVE
Using these industry-wide accepted rules of thumb will help avoid catastrophic damage normally resulting in costly repairs. Fretting Fretting can be considered a special case of adhesive wear. It occurs when two parts in contact experience a repeated, small amplitude relative motion between close-fitting surfaces such as a loose impeller on a shaft. In a pump, there is the potential for small amplitude motion at loose fitting impellers, beneath loose bearings, and between impeller wear rings and the impeller hub. The design engineer does not intentionally create a circumstance that will generate this type of motion, but when it occurs, fretting damage can lead to other problems. Fig 5.65 shows the fretting damage of a pump shaft along the impeller-fitted area where a loose fit enables the oscillation of the impeller during operation. Since the motion necessary to cause fretting can be of a small amplitude, large vibrations in the pump may not be present. This makes the detection of fretting during any operation impossible. The impeller in this example would have similar damage along its bore.

FIG 5.64: THE FRETTING DAMAGE OF A SHAFT BENEATH AN IMPELLER THAT EXPERIENCED SMALL AMPLITUDE MOTION.
(The mottled appearance is typical of the damage caused by fretting (2.2x).)

Fretting damage can be avoided with a few relatively simple guidelines. The possibility of motion between the two components should be prevented or eliminated by either tighter clearances, or shrink fitting the assembly, which increases the clamping force. If fretting is unavoidable in a particular design, methods of mitigation can be used. These include various coatings or providing the contact zone with an appropriate lubricant. Coatings that may be used include flame-sprayed high-nickel alloys, silver plating, or possibly adding a thin, dense chrome plating to one or both of the faces in contact. Abrasive Wear Abrasive wear is often categorized into two main classifications: two-body and threebody wear. The name indicates the mechanism of wear. For the most part, three-body abrasive wear is the primary mechanism of damage in centrifugal pumps. This can occur when hard solid particles entrained in the fluid enter between ring fit areas or impeller

Page 5-78

CHAPTER FIVE
keyway faces. In fluids with high concentrations of solids, another form of three body wear is produced. Solids carried in the fluid stream can strike the internal pump surfaces. This is more commonly referred to as erosion. This type of damage is observed in the impeller and cutwaters of the casing. The degree of material damage, due to this mechanism, depends upon the bulk hardness of the material, the carbon content, and the characteristics of the solids present. Important particle characteristics include size, shape, hardness, and mass. To minimize three-body abrasive wear, a couple of variables must be taken into consideration. The wear ring clearance influences damage. The relationship between the size of the particles in the fluid stream and the gap into which they can enter is important. Wear particle hardness is also extremely important. If particles are soft and friable, such as talc, little damage would be expected to occur on metal pump components because of three-body abrasive wear. The amount of damage is expected to be greater if the particles are extremely hard. These particles include welding scale or silicon dioxide (SiO2), which is sand. The particle geometry also contributes to the amount of damage that can result in three-body abrasive wear. Often, particles of SiO2 are found in a rounded condition. Pumps used to handle river water or seawater on ships frequently encounter these configurations. Hard, round particles are less damaging than particles of equal hardness with sharp, angular configurations. Fly ash, a very hard, sharp, angular particle, is one of the most abrasive services encountered in the pump industry. Test results show that the primary property responsible for increasing resistance to abrasive wear is the hardness of a metal alloy. Small microstructural differences, alloying, and surface-condition differences within alloy groups also can influence the abrasion resistance of a material. Some of these conclusions include the following: Abrasion resistance is increased with increasing bulk material hardness. At the same bulk hardness, steels with higher carbon content have higher abrasion resistance. Cold working, which increases a materials surface hardness, does not significantly increase the abrasion resistance of the alloy. Precipitation hardening increases the bulk material hardness and abrasion resistance of an alloy. Gray cast irons show a decreasing abrasion resistance at higher hardnesses. Softer, austenitic, white cast irons exhibit improved abrasion resistance over martensitic, white cast irons. Carbides are important for the wear resistance of steels and chromium-alloyed, white cast irons. A carbide volume fraction of 30% maximizes the abrasive wear resistance for materials with a soft matrix. An example of three-body abrasive wear is shown in Fig 5.65. It shows a laser-hardened shaft sleeve after approximately one year of service in dewatering operation where abrasive wear caused a significant wear of other material combinations. The abrasive wear was caused by fine tailings in this dewatering application. To increase the life of rings in services like this, the use of hardened wear rings is a good start. This is the reason why pump producers use coated rings in applications where significant abrasive wear is anticipated. However, depending upon the severity of the service, a choice of a ring material containing carbides may be necessary.
Page 5-79

CHAPTER FIVE
For mildly abrasive services, the following materials should be considered: Ni-ResistIts resistance is due to chromium carbides in the matrix. It has good adhesive wear resistance also. Selectively hardening the surface of AISI 420 (laser hardened 50-55 Rc). Surface hardening is not susceptible to hydrogen embrittlement or SCC. Carburized and hardened 12% chromium stainless steel. For more abrasive services, the following is often considered: Hardened AISI-440C (5055 Rc) Stellite or colmonoy-coated (hard-faced) austenitic stainless steel Solid stellite Tungsten carbide Silicon carbide Partially stabilized zirconia (PSZ)

FIGURE 5.65: THE THREE-BODY ABRASIVE WEAR OF A LASER-HARDENED SHAFT SLEEVE IN AN ABRASIVE SERVICE
(Note the fine concentric scoring of the hardened surface. The helix pattern is the laser-beam overlapped zone produced by the laser process.)

Erosion Most fluids handled by pumps are considered clear liquids, meaning they do not have significant amounts of solid particulates present. The corrosive nature of these fluids dictates the required pump materials. However, many fluid-handling applications requiring pumps are far from clear liquids. Solid particulates can be removed with costly filtration systems that must work flawlessly at all times. Fabricated piping systems may introduce suspended solids from weld slag and pipe burn. Naturally occurring suspended solids are those found in water sources such as river water or seawater, as mentioned previously in the abrasive wear section. The following factors should be considered during the material and pump selection phase of the procurement process:
Page 5-80

CHAPTER FIVE
The hardness of the particles The quantity of particles Size distribution Nature (geometry) The velocity of the pumpage The angle of fluid impingement

The first four items listed deal with the suspended solids. These variables can vary from application to application. The hardness of the particles is important to understand in determining the materials necessary to yield an acceptable pump lifespan Particle velocity plays a major role in the degree of damage that occurs in a pump handling slurries. In this case, the potential energy is converted into kinetic energy, producing a material loss by the transfer of energy from the particle to the component. A full understanding of the pumpage including the fluid velocity, fluid corrosiveness, content, and nature of the solid particles present is necessary for the appropriate action to be taken in improving the life of a damaged pump.

FIG 5.66a: THE SEVERE EROSION OF A CARBON STEEL CASING IN A 17% BAUXITE AND SAND SERVICE. (Note the gouging due to the local turbulence of the slurry)

FIG 5.66b: EROSION AT THE EXIT VANE TIPS OF A DUPLEX, STAINLESS STEEL CD4MCU IMPELLER IN A BAUXITE SERVICE

FIG 5.66c: AN AUSTENITIC STAINLESS STEEL IMPELLER IN AN ABRASIVE FLY ASH SERVICE THAT SHOWS SEVERE EROSION. (Increased erosion occurs with an increasing fluid. velocity near the periphery of the impeller)

FIG 5.66d: EROSION DAMAGE OF AN AISI-TYPE, 440C STAINLESS STEEL BALL VALVE OF A RECIPROCATING PUMP IN A COAL SLURRY SERVICE

FIGURE 5.66: (a-d) EXAMPLES OF EROSION

An example of solid particle erosion in a pump is shown in Figures 5.66 a & b. The severe erosion damage of a casing is illustrated by the gouging of surfaces that were

Page 5-81

CHAPTER FIVE
directly impinged or scoured by glancing blows of the solid particles in the fluid stream. Figure 5.66 c shows that the greatest damage to the impeller is at the outer periphery, which corresponds to the highest velocity of the slurry. The least amount of damage is near the impeller inlet eye. Figure 5.66 c also shows that the lower velocity region of the impeller inlet eye has the least damage. Erosion damage can also be encountered in reciprocating pumps. Figure 5.66 d shows extensive erosion of an AISI-type, 440C stainless steel ball from a ball valve after it became stuck and unable to rotate in a coal slurry application. This caused a slurry impingement on a concentrated region of the ball. The particle velocity and impingement angle are design factors that can be used to mitigate erosion in pumps. Erosion damage, once identified, has a limited number of solutions to prolong the longevity of pump materials. This can be accomplished by the selection of hard, wearresistant replaceable liners, elastomeric liners, or, in cases where liners cannot be utilized, hard materials. Such metallic materials include white cast iron (such as Ni-Hard), high chromium (13 to 28 percent) alloy steels, cobalt-based super alloys (such as Stellite), and nickel-based alloys. 5.6.3.3 Fatigue Centrifugal and reciprocating pumps are subjected to cyclical loading, which, if not considered during design, will result in a limited life due to material fatigue. In combination with a corrosive environment, material fatigue can be accelerated due to what is commonly referred to as environmentally assisted fatigue. The one essential parameter in component fatigue is the presence of an alternating or cyclic load. In general, pumps are machines that have either fluid or mechanically induced cyclic loading on their components. Although centrifugal pumps are for the most part steady-state rotational equipment, pulsations or fluctuating applied stresses are encountered. The source of these cyclic stresses can be from fluid interaction between impeller exit vanes and diffuser vanes or, in a volute pump, the impeller vanes and the casing cut-water. Mechanically induced forces are due to bending moments acting on the pump shaft or possibly a component imbalance in the rotor assembly. Reciprocating pumps experience a cyclic loading of the internal and external components from the action of the machinery. In fact, these pumps can be thought of as large fatigue-testing machines due to the pulsating action of the pumping process. When cyclic forces are applied to materials in a pump over a period of time, a crack may initiate at the components surface. After initiation, the crack will grow with continued cyclic loading until the part finally fractures. Fractures can occur, even though the loading produces stresses that are far less than the tensile strength of the material. Corrosion, often the primary cause of pump material damage, can increase the likelihood of fatigue cracking. Corrosion-assisted fatigue is the name given to this special type of cracking. Corrosion damage can change the surface texture and significantly increase the local stresses acting on the pump component. If the corrosion damage is severe enough to produce a sharp notch in a region of high cyclic loading, then fatigue cracking of the component is inevitable An example of corrosion-assisted fatigue at two locations in the front shroud wall of an impeller is shown in Figures 5.67 a & b. The evidence of corrosion pitting on the surface indicates a strong possibility that corrosion influenced the fracture mode. Further

Page 5-82

CHAPTER FIVE
investigation shows that both shroud wall fatigue fractures were initiated at corrosion pits located in highly stressed areas of the impeller. The fluid pulsations acting on the exit vane tip result in alternating loading. Corrosion is not the only mechanism of surface degradation that can promote this form of cracking. Surface disruptions through fretting or wear can also provide sites for fatigue crack initiation. Sharp radii and defects at the material surface such as porosity and poor machining act as stress concentrations.

FIG 5.67a: OVERALL VIEW OF A CF-3M IMPELLER THAT HAS TWO CORROSION-ASSISTED FATIGUE FRACTURES IN THE FRONT SHROUD WALL

FIG 5.67b: A HIGHER MAGNIFICATION OF ONE OF THE FATIGUE FRACTURES THAT ORIGINATED AT A CORROSION PIT AT THE EXIT VANE TIP AND SHROUD INTERSECTION (Additional corrosion pitting can be seen on the impeller in this figure)

FIGURE 5.67: FATIGUE FRACTURES

5.6.3.4 Materials of Construction Impellers The pump designer needs to consider several criteria when selecting the material for the impeller: Corrosion resistance Abrasive wear resistance Cavitation resistance Casting and machining properties Weldability (for repair) Cost For most water and other noncorrosive services, bronze satisfies these criteria and, as a result, is the most widely used impeller material for these services. Bronze impellers should not be used for pumping temperatures in excess 120C. This is a limitation imposed primarily because of the differential rate of expansion between the bronze impeller and the steel shaft. Above 120C, the differential rate of expansion between bronze and steel will produce an unacceptable clearance between the impeller and the shaft. The result will be a loose impeller on the shaft. Cast-iron impellers are used to a limited extent in small, low-cost pumps. Cast iron is inferior to bronze in corrosion, erosion, and cavitation resistance. It also cannot be welded to repair damage due to wear or erosion. For these reasons, a low initial cost is usually the only justification for selecting a cast-iron impeller.

Page 5-83

CHAPTER FIVE
Martensitic stainless steel impellers are widely used where bronze will not satisfy the requirements for corrosion, erosion, or cavitation resistance. Martensitic stainless steel impellers are used in a wide range of applications, including boiler feed water, many cooling waters, and a variety of hydro-carbon applications. It does not have sufficient resistance to pitting corrosion for use in sea water. Austenitic stainless steels are used for impellers in applications requiring a higher level of corrosion resistance than can be obtained from the martensitic grades. Initial cost is a factor here that should be evaluated against the increased life. Casings The following criteria should be considered when selecting material for centrifugal pump casings: Strength Corrosion resistance Abrasive-wear resistance Casting and machining properties Weldability (for repair) Cost For many pumping applications, cast iron is the preferred material for pump casings when evaluated on the basis of cost. For single-stage pumps, cast iron usually has sufficient strength for the pressures developed. For corrosive or hazardous petroleum products, it may be necessary to specify cast steel or cast stainless steel Cast-iron casings for multistage pumps are limited to approximately 6.9 MPa discharge pressure and 177C. For temperatures above 177C and pressures up to 13.8 MPa discharge pressure, a cast steel is usually specified for split-case, multi-stage pumps. For pressures higher than 13.8 MPa, a cast or forged steel barrel-type casing is usually required. In any evaluation of cast iron versus steel casings, consideration should be given to the problem of casing erosion during operation. Erosion can occur either from abrasive particles in the fluid or from wire drawing across the flange of a split-case pump. Although the initial cost of a steel casing is higher than that of a cast-iron casing, a steel casing can often be salvaged by welding the eroded portions and re-machining. Salvaging a cast-iron casing by welding is much more difficult, and the casting usually must be replaced. The ductile irons are useful casing materials for pressure and temperature ratings between cast irons and steels. Although the modulus of elasticity for the ductile irons is essentially the same as that for cast iron, the tensile strength of the former is approximately double that of the latter. In any evaluation of the ductile irons as a substitute for the steels in the intermediate pressure and temperature range, it should be remembered that ductile iron casings cannot be effectively repair-welded. Shafts The following criteria should be considered in the selection of material for a centrifugal pump shaft: Endurance limit Corrosion resistance
Page 5-84

CHAPTER FIVE
Notch sensitivity The endurance limit is the stress below which the shaft will withstand an infinite number of stress reversals without failure. Since one stress reversal occurs for each revolution of the shaft, this means that ideally the shaft will never fail if the maximum bending stress in the shaft is less than the endurance limit of the shaft material. In practice, however, the endurance limit is substantially reduced because of corrosion and stress raisers, such as threads, keyways, and shoulders on the shaft. In selecting the shaft material, consideration must be given to the corrosion resistance of the material being pumped as well as to its notch sensitivity. Corrosion will substantially lower the fatigue limit of the material. Fatigue cracks will initiate at corrosion pits or other surface discontinuities that act as stress risers. A more detailed discussion of the effect of endurance limit on shaft design is given above under the heading Corrosion Fatigue. In the absence of corrosion, an approximate relationship exists between fatigue endurance limits and mechanical properties. The endurance limit is equal to roughly half the tensile strength of the material. Depending upon the application, the pump designer will usually select the least expensive shaft material that will satisfy the three criteria noted previously. Wear Rings The following criteria should be considered in the selection of the material for the wear rings: Corrosion resistance Abrasive wear resistance Galling characteristics Casting and machining properties Suitability for coating A centrifugal pump will often have both case and impeller wear rings. The impeller wear-ring rotates within the bore of the stationary or case wear ring. These rings provide a close running clearance to minimize leakage from the discharge to the suction of the impeller. As the rings wear with use, leakage will gradually increase, affecting the head, capacity, and efficiency of the pump. In multistage (flexible rotor) pumps, increased wear ring clearance may also affect rotor stiffness. To reduce the rate of wear of the wear-rings, and thereby increase the life of the pump, special considerations must be given to the corrosion and abrasive wear characteristics of the ring material. Since the impeller and case rings may occasionally touch one another, the combination should also be selected to have anti-galling characteristics. Bronze is a widely used material for wear-rings because it exhibits good corrosion resistance for a wide range of water services. In addition, bronze exhibits good wear characteristics in clear liquids but tends to wear rapidly when abrasive particles are present. The bronzes also have a relatively good resistance to galling. The leaded bronzes offer excellent galling resistance but use of these grades has been reduced due to environmental concerns associated with lead. The casting and machining properties of most grades of bronze are excellent. In applications where bronze is not suitable because of either corrosion or abrasive wear limitations, or where pumping temperatures exceed 120C, stainless steel rings are used.

Page 5-85

CHAPTER FIVE
Unlike bronze, the stainless steels of the 300 and 400 series have poor galling resistance. Several options are available to minimize the possibility of galling between stainless steel rings. The clearance between the rings can be increased, serrations can be machined into one of the rings, or a minimum hardness differential of 50 to 100 Brinell points can be established between the rings, if made from a hardenable grade of stainless, such as the martensitic grades. If both rings are hardened to above 400 BHN, it is not necessary to maintain a hardness differential. Increasing the clearance between rings is the least costly method for reducing the risk of galling or seizures. However, increasing the clearance will reduce the output and efficiency of the pump. In large, low-head pumps, the loss in efficiency is less than one percent, but in small, high-head pumps, the loss in efficiency can be significant. Serrated rings can be used on smaller pumps to help maintain the efficiency level, but only at an increase in manufacturing costs.
TABLE 5.7: CALCULATED WEAR FACTOR

5.6.3.5 Selection of Materials of Construction The selection of materials for pumps is often not a simple and straightforward matter because of the weight assigned to various technical and economic factors by different users. The selection of the materials for pumps is at best a compromise between the cost of manufacture and the anticipated maintenance costs. Many pump installations start out with a low service factor and, through operating experience, are gradually upgraded in material until an acceptable and scheduled replacement program is achieved. It must be anticipated that for the more corrosive services modification and replacement of the wetted parts will be necessary during the life of the pump. In the initial selection of the most economical materials for centrifugal and reciprocating pump applications, Tables 5.8, 5.9, 5.10 and 5.13 will prove helpful. When the liquid being pumped is an electrolyte, particular attention must also be directed at an evaluation of the probability of an unacceptable level of galvanic corrosion. The electrolytes most commonly encountered in pump applications are seawater, brines and mine waters. For these services the use of incompatible can lead rapid failure, and this is particularly true when a small area of a less noble metal is in contact with a larger area of
Page 5-86

CHAPTER FIVE
a more noble metal. Bronze wearing rings, for example, should never be used with a stainless steel impeller when pumping an electrolyte because the unequal area between the wearing ring and the impeller would result in a rapid preferential corrosion rate of the ring. An even more destructive combination would be the use of bronze screws or bolts in a stainless steel impeller or casing. Tables 5.11 and 5.12 list combinations of metals most commonly used in pumps that should be avoided when the application requires the pumping an electrolyte.
TABLE 5.8: MATERIAL SELECTION CHART FOR VOLUTE CASING PUMPS

TABLE 5.9: MATERIAL SELECTION CHART FOR WET-PIT DIFFUSER PUMPS

Page 5-87

CHAPTER FIVE
TABLE 5.10: MATERIAL SELECTION CHART FOR RECIPROCATING PUMPS

TABLE 5.11: COMBINATION TO BE AVOIDED WHEN AREA OF METAL CONSIDERED IS SMALL RELATIVE TO AREA OF COUPLED METAL

Page 5-88

CHAPTER FIVE
TABLE 5.12: COMBINATION TO BE AVOIDED WHEN AREA OF METAL CONSIDERED EQUAL TO AREA OF COUPLED METAL

CAUTION: The selection of materials for a pump is a complicated exercise, requiring knowledge of the engineering properties of the material, its fabrication characteristics, and corrosion and erosion resistances. Past experience is often helpful, but it is recommended that the materials selections be reviewed by a metallurgist or corrosion engineer with experience in this area, especially for critical applications or those applications with which you are not familiar. 5.7 PLANT PUMPING SYSTEMS AND EQUIPMENT PROTECTION

The design of any plant and pumping system is not complete without the inclusion of appropriate and comprehensive protective components and the economic value of such a plant or pumping system is reduced, often to an undesirable level. In order to ensure acceptable technical soundness of plants and pumping systems the following protective measures need to be considered at the design stage and adopted wherever appropriate. For the purpose of avoiding unnecessary sophistication and yet establish reliable protection of plants and pumping systems the protective systems enumerated hereunder shall be adopted in the fashion recommended herein 1. Protection against dry running 2. Protection against water hammer 3. Protection against cavitations 4. Protection against electrical surges 5. Protection against overloads/over current 6. Protection against undervoltage /phase failure 7. Protection against earth leakages 8. Protection against insulation damage 9. Protection against mechanical injury 10. Protection against cathodic corrosion 11. Protection against gas leakages 12. Protection against electrical shocks 13. Protection against rust 14. Protection against fire 15. Protections against stilling torques

Page 5-89

CHAPTER FIVE
5.7.1 Dry-Running Protection Wet installation, i.e. submerged installations are subjected to fluctuating water level dictated by varying seasonal weather changes. Pumps installed at Boreholes, (B/Hs) Lakes and Dam intakes can be expected to run dry in the course of which components, depending on water for their cooling, could be damaged if precautionary measures are not taken. Similarly surface pumps installed in series with a pressure main or submersible pump installed as an in-line booster may be expected to run dry as a result of interrupted supply into the pressure main. Also surface pumps fed from a sump may run dry especially if the sump is supplied with water from another pumping system following interruption of the system operation. For these reasons, dry-running protection is often necessary and should be provided in any of these situations. 5.7.2 Water Hammer or Surge

5.7.2.1 Introduction The subject of water hammer or surge in pipelines was briefly discussed in Chapter 4. In addition to posing risk of damage to pipelines, it is also a problem that must be addressed for pumping systems. Sudden pump shut offs due to power interruptions to electrically powered pumping plants, or an unprecedented high rate of valve opening or closing invariably results in hydraulic surges. In the case of a conventional electrically powered pumps a pressure wave is initiated downstream of the check valve which then travels to and from the pressure main termination point (storage tank into which the rising main terminates) or any other pressure wave reflection point, such as a closed valve point on a rising main branch. The pressure wave thus oscillates between the disturbances points with progressively varying amplitudes the magnitudes of which depends on the rate of flow velocity change and the hydraulic damping characteristic of the pumping system. It should be noted in this connection that hydraulic surges (water hammer) will develop only when: ts tr = 2L / a Where ts = period of the flow velocity fluctuation tr = the to and fro wave travel period i.e. the time taken by the surge wave to travel from the pumping station-to the reflection point and back again to the pumping station (seconds). L = The distance between the reflection point and the disturbance generation point (m) a = Surge wave velocity m/s Under these conditions, the reflected wave arrives at the disturbance generation point (at the pumping station) too late to be able to have any alternating influence (damping effect) on the disturbance condition. The pressure variation and the maximum pressures arising therein can lead to severe mechanical stresses on the installation in question. Furthermore the down surge 5.9

Page 5-90

CHAPTER FIVE
(minimum surge) may be as low as the vapour pressure of the pumped medium followed by cavitation and separation of the liquid columns into two separate columns. The return pressure wave following column separation then causes the two columns to move towards each other and collide at high speed. This phenomenon is in the strictest sense what is termed water hammer or hydraulic surge. 5.7.2.2 Initial Water Hammer Analysis Calculation For the purpose of determining whether or not protective devices against undue water hammer are necessary the following initial estimation procedure is recommended. Assuming a more or less uniform slope of the rising main in question between the pumping station and the discharge tank, the so called K2 value should be worked out as follows; K2 = L Va / Ho Where, L = length of the rising main (m) Va = flow velocity in the pipeline (m/s at maximum pumping capacity) Ho = total head of pump (m) at maximum pumping capacity (Qmax) For a value of K2 greater than 70, further water hammer calculations will not be required in most cases. It should be noted however that in cases where extreme apexes in the longitudinal cross section of the transmission line (pumping or gravity) exist precise calculations may still have to be carried out since the K2 value loses its meaning. For branched pumping mains, only the length from the pumping station to the branch should be entered into the equation for L. New assumptions must be made for the downstream branch pipes from the junction. These assumptions take into consideration the total head Ho less the pressure loss in the piping up to the branch and any changes in the static head as well as the flow velocities V in the new runs of the piping downstream of the branch. For this particular case. (1) (2) (3) K21 = L1 V21 / H21 K22 = L2 V22 / H22 K23 = L3 V23 / H23 5.10

5.7.2.3 Subsequent Analysis of a Potentially Dangerous Water Hammer Situation There are essentially two ways of proceeding namely: Use a purpose developed software package, or Undertake a numeric/graphical analysis in-house 1. Surge Analysis Software Except for the simplest of pipeline situations, it is strongly recommended that use be made of reputable software developed for the purpose.

Page 5-91

CHAPTER FIVE
A number of software solutions have been developed and can be purchased at varying prices, generally in the range of $3000 - $8000. Of these, that recommended for consideration is Surge 2000 produced as part of the KYpipe software package, and details can be found by logging on to their website <www.kypipe.com>. A 250 pipe solution currently (2006) costs US$ 3,000. Another alternative, especially if the designer is interested in pursuing a controlled air transient technology (CATT) approach to the use of air-valves is to contact Vent-O-Mat of South Africa by logging on to their website, <www.ventomat.com> and describing your requirements as they will often undertake a free surge analysis incorporating their anti-surge anti-shock air valves into the pipeline system as a possible full or partial solution. They also provide a considerable amount of useful technical information free of charge. Note however possible disadvantages in using this approach alone to surge protection in the section on surge protection devices. 2. In-house Numeric Graphical Analysis A procedure for undertaking a simple surge analysis in-house for a single pipeline of uniform diameter was discussed in Chapter 4, Section 4.11.7 based upon a single pipe type and diameter and hence single celerity or wave velocity value. For a single pipeline with a varying diameter and possibly different wall materials, an equivalent wave velocity must be used which can be calculated as follows: cpeq = (L1 c1 + L2 c2 + L3 a3 ..Ln cn) / LT 5.11

Where: L1 c1, L2 c2, L3 c3.Ln cn, are the respective lengths (m) of consecutive pipeline sections with their corresponding mean wave velocities, and LT is the total length of the pipeline in (m). Similarly, equivalent mean flow velocity for the pipeline with a non uniform internal diameter shall be computed using the formula stipulated above by substituting respective mean flow velocity values for each section of a constant internal diameter. Veq = (L1 V1 + L2 V2 + L3 V3 .Ln Vn) / LT 5.12 Having established the surge velocity cpeq, it is advisable to draw out a graphical solution in order to establish a complete picture of the surge pressure fluctuations corresponding to the beginning of the pressure disturbance stage down to the stabilized stage at the pumping system static condition as shown in Figure 5.68.

POSIT IVE OR UPSUR

HYDRAULIC
maximum head at pumping station

N GRA DIENT GRADIENT DU RING PUMP ING

GE RE TUR

Pipe Inver Level (masl)

STATIC HEAD
pumping head at Qmax

SU DOWN IVE OR NEGAT

TWA NT OU RADIE RG E G

VE RD W A

RISK OF COLUMN SEPARATION AT THIS LOCATION

RISK OF COLUMN SEPARATION AT PUMP LOCATION NOT KNOWN

Curves determined on the basis of an instantaneous pump stoppage and hence theoretical maximum.
Chainage (m)

FIGURE 5.68: SUPERIMPOSITION OF SURGE ON A PIPELINE PROFILE

Page 5-92

CHAPTER FIVE
This approach however only reveals the theoretically worst case magnitudes of the hydraulic surges but does not determine if in reality any flow separation takes place. In practice and assuming no flow separation at the pressure disturbance point or elsewhere along the pipeline, the magnitude of water hammer will be less than the theoretical value by an amount corresponding to the attenuating or damping effect due to the moment of inertia of the pump and the prime mover and the effects of any surge alleviating devices installed. This picture would be reflected on the pressure gauge at the pump delivery branch in the form of pressure pulsations which would gradually die off as the hydraulic condition in the pumping main returns to equilibrium. 5.7.2.4 Flow Separation In the case of flow separation either at the pressure disturbance point (pumping station) or at any intermediate high point along the pipeline, the amplitude of the first water hammer return wave at the pumping station cannot be determined by the formula [h = (c/g) Vo] as it is invalidated by the fact the separation creates new surge conditions resulting into surge magnitudes unpredictable by the formula. One thing is certain though; surge magnitudes following flow separation are much higher than those occurring in the absence of flow separation and these are the most destructive. How much higher the surge would be would depend on the speed at which the separated water columns comes back together. No attempt is made in this text to evaluate the magnitude of the upsurge proceeded by flow separation and here the use of surge software is the better option. Instead various methods to prevent such incidence are described herein under the heading Water hammer protective equipment and systems. Assuming instantaneous stoppage on the pumping system in question, the extent of flow separation may be considered to be directly proportional to the pipeline period (2L/c) i.e. the up-and down pressure wave travel period. Hence the longer the pipeline the longer would be the pipe section that would be completely evacuated by the pumped water. Since some equipment or system would be required to avoid flow separation particularly at the pumping station and at any intermediate high point along the pipeline and that the volume of the evacuated pipe section would be proportional to the travel period and therefore proportional to the pipeline length it is less problematic in short than in long pipelines. As the extent of flow separation is also a function of the flow velocity in the delivery main, it is worthwhile considering critically the possibility of selecting pipe diameters and hence flow velocities as low as possible at any critical location or locations. Needless to say cost implications on the pipes selected are determinants on the limits. 5.7.2.5 Pipeline Topography Although a graphical surge diagram gives an indication on the magnitude of the surge, this indication only refers to a particular point in question (mostly it would refer to the pumping station).

Page 5-93

CHAPTER FIVE
In order to get a complete impression on what may occur on other sections of the whole stretch of the transmission main; the slopes (upsurge and downsurge) of the water hammer waves need to be superimposed on to the plane of the pumping scheme topography. This would reveal some more information as to the magnitudes of over and under pressures on various sections of the pumping main. The topography of the pumping main at and in the vicinity of the pumping station must also be analysed in detail. In order to avoid under pressures in any section, it is desirable to have a concave pipeline topography, i.e. the alignment is selected such that a concave longitudinal profile is formed. It needs not be a perfect concave but the levels on the various apexes must not be exceedingly higher than the level of the negative transient. Unfortunately it is rarely possible to have a totally concave shaped topography due to economic or technical limitations, i.e. prohibitive cost, construction practices, etc. Nevertheless the possibility of exploiting such a valuable proposition must not be ignored in the course of determining the alignment of any pressure main. 5.7.2.6 Water Hammer Protection Systems Having made a sufficiently comprehensive water hammer analysis, preferably by use of appropriate software in all but the simplest situations, and having established the hydraulic surge characteristic for a particular scheme the necessary surge protection devices can be selected. Some of the more commonly used devices are illustrated in the Figure shown opposite.

FIGURE 5.69: PIPELINE PROFILE ILLUSTRATING SUITABLE LOCATIONS FOR SURGE PROTECTION DEVICES

One or more of these protective devices and systems may then be considered and adopted. I. Energy Dissipators i. Flywheel arrangement on the prime mover to pump shaft ii. Air vessel (surge tank). iii. Surge shaft of standpipe i. Flywheels In surge control, the action of the flywheel is to maintain the speed of the driven rotating parts in the event of a power outage. Hence, by the time the returning pressure wave travelling through the delivery main occurs, the pressure generated by the pump will have fallen because the negative surge has been counteracted to a level approaching normal duty head conditions. When a flywheel is incorporated into the system design it is vital to ensure the starting gear is correctly chosen and rated to allow a more gradual start from rest and reaching full speed and flow.

Page 5-94

CHAPTER FIVE
Flywheels rarely feature in pumping plants these days; however they can have a number of advantages. Special oil-filled flywheels can be used for the dual purpose of speed control and surge dampening. Unfortunately there is an instance in Tanzania on one of the largest pumping systems where these were removed once speed control was no longer required thereby causing serious problems of cavitation and water column separation along the transmission main that thereby lost its surge protection. ii. Air Vessels Air vessels have in the past been the commonest form of surge protection device for pumping stations as they are located at the pumping station itself and therefore under the control and supervision of the pumphouse personnel. However, unless the air water interface is maintained at the correct level they do not perform as intended and in major, high pressure situations should be enclosed within a protective earth bund to guard against damage and possible loss of life if they explode. For smaller installations, a bladder type is available which separates air from water and so obviates the need for periodic topping up of the air. However, should the bladder be damaged during maintenance then they have the same problems as larger vessels. Care must be taken when utilising these devices to ensure that the vessel is adequately sized for the application and that a strong mesh is placed over the entrance of the vessel into the pipeline to prevent the bladder from being drawn into pipeline upon pump trip and column separation. Flow out of and back into air vessels is controlled by an outward flow check valve fitted with a smaller return flow bypass and the check valve itself can create surge and water hammer of its own if the selected check valve does not react fast enough. Advantage They can be absolutely automatic, provided that the water level in the air vessel is maintained at the preset level. Disadvantages A large space is required for the air vessel, and its air compressor, slam tight closing of the check valve can be expected unless a special non-slam type such as a Hydrostop or a linear closure design type such as MaxiFlo is selected. Such check valves utilise the Venturi phenomenon as the basis of the design to economically combine the features of a non - slam and low head loss device at a reasonable initial cost. Such a valve can be fitted especially on pipes with a diameter greater than 300mm, i.e. width dia. 300mm, the application for a rubber cushion to ensure smooth check valve closing is inevitable. Dimensions An air vessel may be approximately sized according to the following formula: T = (L / 50c) [(50c + 1) / (Ho Hmin)] Va = (Co / 2) FT Vo = Va [(Hmin + 10) / (Ho Hmin)] Vges = Vo + Va Where, T = the time, rounded off in seconds, required for the surge wave the pipeline to come back from the reservoir to the pumping station.
Page 5-95

5.13 5.14 5.15 5.16

CHAPTER FIVE
Hmin = the largest value in (m) of Hmin (a) or Hmin (ND). F = the pumping main cross sectional area (m2). Va = the water volume (m3) required to flow out from the air vessel in the event of a hydraulic surge

Vo = the air volume of the air vessel for the operating condition. Vges = the total volume of the air vessel in m3, thus the aggregate size of the air vessel. = factor of safety in the height fixed from 1.5 up to 2. iii Surge Shafts or Stand Pipes

The application of a standpipe is good in small head situations up to 15m and results in small pipe friction, and therefore when the W-K diagram does not give any positive value. When entrained air in the pipeline is permissible the size of the standpipe can be reduced. Advantage Completely self operative and requires almost no maintenance. Dimensions The size of a stand pipe can be approximately calculated using the following formula: Fst = (L/g) (F Co2)/H) Where, Fst = the stand pipe cross sectional area (m2) L = the length of the rising main (m) F = the cross sectional area of the rising main (m2) Co = the discharge flow velocity corresponding to the maximum discharge flow from the pumping station at the smallest static head H = the elevation between the standpipe and the water level at the discharge tank g = 9.81 m/s/s II i. Suction Mechanisms Suction By pass 5.17

A suction by-pass is applicable whereby the permissible surge pressure in the pressure main is twice the static head of the pumping systems; and in addition the rising main is not too high above the suction water level and that there are no extremely high points along the mains. The application is therefore normally limited to small and medium head systems with a high percentage of friction head loss. Disadvantage Space requirement for the suction by-pass. At twice the static head, the vacuum pressure will cause wear to the check valve, and hence is expensive on larger rising mains. Besides the by-pass requiring a non-return valve, some additional fittings would
Page 5-96

CHAPTER FIVE
possibility be necessary for the by-pass. The system does not work like a surge protection if due to a defective check valve the reverse flow flows back through the pump. Sizing The diameter, Ds, of the by-pass may be sized from around 0.5 D to 8.8 D. The check valve on the by-pass must be arranged in such a way that the line and the valves are well below the water levels in the suction tank. ii. Suction Tank Suction tanks are only applicable whereby the permissible surge-pressure is twice the static head. The suction tank may be arranged at any height as the height of the rising main does not have any influence on the applicability of the suction tank. Disadvantage An automatic water level device to refill the suction tank is necessary and there is a large space requirement. They are expensive and for long rising mains, large capacity suction tanks are necessary. The system does not function or serve like a surge protection device in the event of a defective check valve when the back flow would run back through the pump. Advantages Self operative, simple maintenance and allows the check valve to close gently. iii. Air Valves Conventional air valves Use of a conventional air valve located near the pumping station for the purpose of limiting the effects of surge on the pumpsets is only applicable when entrained air in the pressure main seems permissible and when the static head differential 'H' is less than 10m. Without an additional check valve or a float valve at the end of the pressure main, (pressure main terminal), use of an air valve in this way is only applicable on small systems with small static heads of the order of one metre. With higher static heads, additional valves or float valves with a by-pass are necessary. In a slow reverse flow, the entrained air therein would be discharged accordingly. Placing a reliance on a conventional air-valve as a means of alleviating water hammer in moderate to high pressure water mains is unwise. Disadvantages During the return high-pressure wave, the sudden coming together of any separated water columns in the pressure main will cause transients that can be extremely large and extremely damaging. As with all appurtenances, if located in public places there is a risk of vandalism unless they are located in well secured chambers. If the outward low pressure wave causes the line pressure to drop below atmospheric at the air valve, and unless the chamber is provided with excellent drainage facilities, there is a risk of drawing in polluted ground water entrapped in the chamber.

Page 5-97

CHAPTER FIVE
Advantages Automatic with ideal flow velocities, limited maintenance; small capital investment and permits the non-return valve to close gently. Dimensions The nominal diameter of the double air valve (DAV) may be sized from around 0.15 to 0.25 D. CATT triple function air valves Unlike conventional air valves, CATT anti-surge anti-shock air valves are under certain conditions ideally suited to alleviating or even resolving the problem of water hammer as they not only provide the two usual functions of double air valves but in addition enable the rapid drawing in of large volumes of air to avoid the risk of column separation. They can protect both the pumping system and the pumping main itself. Disadvantages As with all appurtenances, when located in public places there is a risk of vandalism unless they are located in well secured chambers. If the outward low pressure wave causes the line pressure to drop below atmospheric at the air valve, and unless the chamber is provided with excellent drainage facilities, there is a risk of drawing in polluted ground water entrapped in the chamber. Advantages Performing both the usual function of air valves plus the control of transients their use will often avoid the need for the usually very expensive alternatives. Further Reading For those interested in learning more about the performance of conventional air valves compared with CATT air valves, and the control of surge by CATT air valves, designers are advised to obtain a copy of The Application of Controlled Air Transfer Technology to New & Existing Pipeline Systems by Allistair Balutto -Vent-O-Mat, 1998 III i. Other Devices Side Vents

Practically speaking, side vents can be applied where the protective measures mentioned under I and II above cannot be realized and there is the possibility of allowing a considerable quantity of water from the pressure side to the suction side during the shutting off process. Side vents, come mainly with a electrical control mechanism, compressed air or a server motor to affect slow opening and closing of the by-pass Disadvantages The reliability of operation depends on the control (adjustment) of the opening and closing time, large oil-hydraulic and electrical equipment. Therefore the control system is susceptible to control for restarting, and operator involvement is required as well as there being energy loss through the reverse flow of the discharged water quantity. Advantages With correct (adjustment) control of the opening and closing time, there are small dynamic pressure increases that help ensure that the check valve closes gently
Page 5-98

CHAPTER FIVE
Dimensions The diameter DN of the vent can be sized from around 0.3 up to 0.5 D. ii. Controlled end Valve Closing The possibilities of using controlled-end valves are the same as those for the side vent, but additionally a backflow isolator in the pump is necessary provided that the reverse speed of the pump would not rise above permissible limits. Disadvantages The operational reliability depends on the controlling of the closing time of the isolating valve, and with parallel operation of several pumps on a common pressure main, additional control is required so that on opening or closing of one pump the check valve closes swiftly. There is also energy loss through the discharged water mass. Advantages No additional by-pass line is necessary and the operation allows the check valve to close gently. iii. Burst Disks Burst or rupture disks are fail-safe devices to prevent or minimise damage when all other preventative mechanisms are unserviceable. Burst Disks are not surge protection devices as such, in the same sense as those in the previously described possibilities. Through them the appearance of hydraulic surge would or would not be prevented. Burst Disks alone are therefore recommended only where a predetermined surge danger is only very rarely expected, or when there are low water levels in the pumping system with several pumps operating simultaneously. Burst Disks alone are naturally not useful when on every closing (shut-down) operation the dynamic differential pressure could cause the disk to rupture. Disks should be mounted on an isolatable short pipe support so that only little water quantity is contained and accordingly the effect of the moment if inertia is small. They normally are supplied in packs of five and the only danger is the failure to re-order a replacement pack in time. IV Recommendation Insofar as any pumping main is likely to require air-valves, it seems logical to use those that combine both the conventional function required of them plus the added benefit of controlling transients. However, due to the need to protect, operate and maintaining any such appurtenances, total reliance on such devices is not recommended if there is the slightest risk of vandalism and/or inadequate routine maintenance. Hence the combination of CATT type air-valves together with a final fail-safe burst disc device is strongly recommended, and indeed such burst-disc devices are also recommended in conjunction with any air vessel requiring the periodic topping up by using an air compressor.

Page 5-99

CHAPTER FIVE
5.7.3 Protection Against Cavitation A. Cavitation The formation and subsequent collapse of vapour filled cavities in a liquid due to dynamic action is called cavitation. The cavities may be bubbles, vapour filled pockets or a combination of both. The local pressure must be at or below the vapour pressure of the liquid for cavitation to begin and the cavities must encounter a region of pressure higher than the vapour pressure in order to collapse. Dissolved gases often are liberated shortly before vaporization begins. This may be an indication of impending or incipient cavitation but true cavitation requires vaporization of the liquid. Bubbles or vapour cavities which collapse in a solid boundary may cause severe mechanical damage. It should be noted that pressures of the order of 100 atmospheres have been estimated by researchers during collapse of a bubble. The mechanical damage resulting from cavitation include: (a) (b) Erosions resulting directly from the collapse of the formed vapour cavities. Erosion due to forced vibrations caused by cavitation initiated shock waves.

In order that the designer may effectively avoid or minimize the adverse effects of cavitation it is necessary for him to identify the main causes for cavitation. These are enumerated below: (a) (b) (c) Exceedingly high flow velocities which result in excessive pressure drop and subsequent formation of vapour pockets and collapse on a compliant surface. Unduly low flow in the suction pipe resulting in flow re-circulation. Dynamic pulsation in the entering inlet of an impeller pump. The pulsations may create regions of low instantaneous pressure within the pump and the energy thus created then propagates throughout the system. Vortexing

(d)

Improperly designed suction sumps can lead to the formation of vortexes produced by localized eddies on the water surface, and these can progress to form a cone. Should the disturbance continue, the flow of water would carry the underwater part of the vortex down towards the pump suction bellmouth and ultimately into the pump itself. This introduces air into the impeller, due to interruption of the normal solid-liquid flow pattern. This type of disturbance apart from leading to cavitation also produces hydraulic pulsations in the pump flow and mechanical overloading of sleeve bearings and impeller guides. (see opposite and on next page) Underwater vortexes sometimes occur in round sumps, or in sumps where the pump suction bellmouth is at some distance from the rear wall. The risk of a vortex occurring can be checked by considering the Submergence Factor, F, in Figure 5.71 opposite and above, where:
cmin = d/4, optimum = d/3 to d/2

V
d

2d

FIGURE 5.70: VORTEX PREVENTION ASSESSMENT

Page 5-100

CHAPTER FIVE
F = (h/d) (c/d + 6) with the symbols as shown in the Figure. (e) Flow Separation in the Suction Line An impeller suction line may contain sections allowing the collection of dissolved or undissolved gases in the course of pumping. The gases thus collecting in such a section would increase in volume with time with subsequent separation of the flowing suction water column into two columns or several columns. Hence the flow of water into the pump would cease thereby resulting in the pump being vapour bound. As a result, cavitation would occur and cause mechanical changes. (f) Turbulent Flow In fluid dynamics, turbulence or turbulent flow is a flow regime characterized by chaotic, stochastic property changes. This includes low momentum diffusion, high momentum convection, and rapid variation of pressure and velocity in space and time. Flow that is not turbulent is called laminar flow. The degree of turbulent flow resulting from velocity can be determined from an evaluation of the dimensionless Reynolds Number (R) using the formula: R = ( V D)/ Where, V = velocity in the pipe D = internal diameter of the pipe Laminar flow has a Reynolds number of 2000 or less; transitional flow has a value between 2000 and 4000 and is a mix of laminar and turbulent flow and fully developed turbulent flow has a value of 4000 or greater. Turbulent flow results either from high flow velocities or an abrupt change in the flow due to a change in the profile of the suction line. (g) Suction lift Suction lift, exceeding the maximum permissible lift would result in an unacceptably low NPSHA compared to the NPSHR. As was mentioned earlier, the NPSHA must be greater than the NPSHR by a margin of about 1 m to ensure continuous and stable operation of a centrifugal pumps; all other things being equal. B) Measures to be taken Various measures can be considered to protect pumps from the risk of cavitation, including: a) Since almost all of the causes of cavitation discussed above result in a pressure drop at the pump suction branch, the most effective protection is the incorporation of a pressure operated relay system in any negative head suction pipe system. This entails the installation of such a device at the pump suction branch which monitors the NPSHA variation in such a manner that at a predetermined (impermissibly low) NPSHA an impulse is sent to the control circuit of the switchgear which, in turn trips the power supply and brings the pump to a stop. b) Application of special treatment to the pump material parts being or likely to be affected by cavitation may be administered using cavitation resistant materials. 5.19 5.18

Page 5-101

CHAPTER FIVE
c) Air injection into the suction branch may be administered to provide a dampening effect to possible cavitation implosions. d) If the protective measures alone cannot be applied, the best way to avoid or minimize cavitation is to reduce the flow velocities or make adjustments to any of the causes identified as being responsible for the cavitation. A number of measures can be taken following the detection of cavitation damage. Amongst them are. a) Reduction of the suction velocities by: (i) (ii) b) Increasing the suction pipe diameter Throttling the pump delivery line such that the operation is maintained at a discharge flow not exceeding the optimum flow (Qopt) .

Prevention of re-circulation flow especially on large pumps by ensuring that any necessary throttling does not go beyond the minimum permissible discharge flow from the pump. Generally the minimum permissible flow (Qmin) from a centrifugal pump ranges between 0.33Qopt to 0.45Qopt. For more accurate determination of permissible minimum flow, see the calculation of the recirculation flow magnitudes in the Operation and Maintenance Manual. Reducing the suction lift. Modifying the suction pipeline profile to avoid flow separation. Ensuring that the intake channel terminates in to the bottom side of the sump. The termination must be at a right angle. Reducing the flow velocity in the intake channel thereby eliminating turbulence in the pump sump. Avoidance of obstructing structures in front of the suction bell month. Avoiding as far as possible the use of a common sump to serve two or more pumps. This is particularly important on poorly designed large pump sump installations and more so when pumps of dissimilar suction characteristics are involved and where a suction lift is indispensable. Over Voltage Protection A) The General Phenomenon and Protection The three main causes of over voltages (surges) are switching faults, arcing to earth and lightning. Since electrical systems, i.e. transmission, distribution and their equipment are mostly designed to have their insulation capable of withstanding at least twice the operative voltage and that over voltage (hereinafter referred to as surges) due to switching rarely exceed this magnitude, it is usually unnecessary to take precautionary protection measures against switching surges. Surges due to lightning however are exceedingly objectionable as their magnitudes, depending on the lightning characteristic can be as much as 10 times the operative voltages (generally tension of several million volts can be expected, to be induced into the system) and protective measures in this connection are not a choice but a matter of absolute
Page 5-102

c) d) e) f) g) h)

5.7.4

CHAPTER FIVE
necessity. Depending on the characteristic of the surge against which the system may have to be protected, one of the following or combination of the same measures may be adopted. 1. 2. 3. 4. 5. 6. Overhead Earth Wire Horn Gap Non-Linear Surge Diverter Petersen Coil Gas Arrester Ferranti Surge Absorber

An overhead earth wire is mostly suitable for the protection of high tension overhead transmission lines. This protection reduces the possibility of a direct lightning strike reaching the main conductors and also partly shields them from electrostatically induced charges caused by charged clouds in the vicinity of the lines. An array of earth wires would be much more effective for both these purposes but this has prohibitive cost implications. The effect of an earth wire system is to reduce the magnitude of the electrostatic induction due to lightning when run above the main transmission conductors and earthed at every vulnerable point such as at a tower or at some pole. It should be noted in this connection that the voltage induced on the line due to lightning is given by; Vi = (Vc Ccl) / (Ccl + C1g) Where, Vi = induced voltage in the line Vc = the cloud voltage Ccl = capacitance of the cloud to the line C1g = capacitance of the line to the ground Hence, since the effect of the earth wire is to increase the overall capacitance, C1g, of the line to ground, the net result is a reduction in the induced voltage. The earth wire system may also provide considerable shielding of the transmission lines from direct lightning strikes provided that the resistance of the earth is kept very low otherwise flash-over between earth and the lines can occur., Generally, the earth resistance should be as low as 10-20 ohms. For the purpose of protecting a plant or equipment, e.g. a transformer, the Horn Gap, with a resistance and auxiliary electrode is one effective method. The resistance in the horn gap serves the purpose of limiting the excess current due to the surge to about 5 amps whilst the auxiliary electrode provides protection to the horn gap against cathodic corrosion. As high frequency waves tend to concentrate at the line-end turns of the transformer, resulting in severe stresses at the turns with possible puncture between them, the horn gap described above incorporated with an appropriately sized check coil eliminates this problem. One horn gap is connected between each line and the earth. Under normal operating conditions the gap is subjected to the operative phase voltage which is insufficient to send a spark across it. When however, an over-voltage of, say greater than
Page 5-103

5.20

CHAPTER FIVE
twice the rated phase voltage reaches the gap, spar- over occurs and the surge energy is diverted to earth. Arcing to earth may cause a voltage surge of as much as 3.5 times in the earth lines For the effective protection of the lines or equipment against this voltage rise, a Petersen coil with an inductance of 50-110 percent of the theoretical value should be employed The neutral point of the equipment, e.g. the transformer, should be earthed via a properly sized Petersen coil. The methods described so far can only divert the surge wave to earth. The surge energy dissipation only takes place via the resistance of the conductors themselves and the earthing resistance within the premises adjacent to the protected zone. A condenser and a resistor connected in series makes a combination capable of dissipating part of the surge energy in addition to merely diverting it from equipment. Furthermore such a combination has a characteristic that attenuates the steepness of the wave front of the transient. This combination essentially constitutes the so called a surge absorber, i.e. it is a device capable of not only diverting the surge energy but also absorbing it. In this connection, it is recommended where it may be deemed desirable to both divert and absorb the surge energy through a protective equipment to use the so called Ferranti Surge Absorber. This consists of an inductance coil which is coupled magnetically but not electrically to a metal shield and/or to the steel tank which contains it. It provides a filtering effect which prevents high frequency transient currents from passing freely through the absorber. Hence, high frequency currents are reflected away from the absorber thereby protecting the equipment from strenuous stresses that could result from such current and cause a puncturing of the insulation. For medium tension systems a gas discharge surge arrester is strongly recommended. This type of arrester employing gas as a medium in the spark gap which is capable of both modifying the wave front of the transient as well as attenuating its amplitude by almost 80%, e.g. a 5 kV transient can be reduced down to 1.1 kV. The insulation on equipment designed to operate on medium tension is capable of withstanding a surge of up to 1.1 kV voltage provided that its life time is very short, and if possible the protective equipment characteristic should be such that the spark-over time is less than a microsecond. Effects of Geotechnical Characteristics on the Effectiveness of a Lightning Arrestor It should be clear from the discussions on the various surge protective methods that their operation is based on the following four main principles. 1) Diversion of surge to Earth 2) Diversion and absorption of the surge energy 3) Increasing the line capacitance so as to reduce the magnitude of voltage that could be induced by the electrostatic effects from a charge cloud and diverting the induced surge voltage to earth.

Page 5-104

CHAPTER FIVE
4) Reflection of surge energy from the protected zone using a choke coil, applicable to high frequency transients. In order to divert the surge energy from the main path, the diversion path must have a much lower resistance compared to the overall resistance of the main path. Should the diversion path terminate to earth at a point of very high resistivity, the result would be increased resistance in the diversion path. The surge would then go to earth through the protected zone which happens to have a lower resistance. A submersible pump being protected for instance makes a very good contact with earth since it is completely submerged in ground water. If the protective device is earthed at ground, the resistivity of which is higher than that at the submersible pump, the surge energy would go to earth but through the submersible pump and therefore damage the motor windings even though a surge diverter or surge absorber was provided. Hence there are special requirements which must be taken into consideration and appropriately accommodated in the course of making an installation design and actual erection of a surge protective device intended to protect any equipment or system which always has the best contact to ground. Any other equipment or system with conditions similar to those pertaining to the above given illustrations must be treated likewise where such protection is deemed necessary. The additional capacitance of the conduction to earth needed for the prerequisite attenuation of induced voltage due to electrostatic induction must also be properly evaluated and the resistance of the earth wire must also be limited to a value that would not promote flash-over to the conductors. When a transformer must be protected against high frequency transients a properly sized choke coil must be used in series with the transformer termination lines. Under this condition the Ferranti type of surge absorber would be an ideal protective device. In all these cases however it must always be ascertained that the grounding point has a very low resistivity compare to the ground contact made by the equipment. 5.7.5 Cathodic Protection A. Occurrence Low resistivity near a pumping station (electrically powered station) would cause corrosion of metallic pipes, e.g. steel pipes, GS, DI, etc., both due to differential electrical potential as well as to stray currents from the pumping station if protective measures are not taken It is therefore advisable to take appropriate protective measures against the risk of such corrosion. However and once away from the pumping station and except when threaded joints or welding is used as a means of joining ferrous pipes, the use of barrier coated pipes in particular excludes the need for cathodic protection once away from the pumping station. One of the recommended corrective approaches is the use of natural sacrificial anodes. These should be connected to the pipeline near the pumping station together with electrical bonds across any non-continuous pipe joints. (To be noted here is that rubber gaskets between flanged joints also prevents such continuity). The most commonly used sacrificial anode cathodic protective systems are as shown in Fig 5.72.

Page 5-105

CHAPTER FIVE
Such systems are based on the principle of eliminating the corrosion build-up, occurring when the pipe metal is positive with respect to the ground. An external electrical power source adds a potential difference of opposite sign.

Bronze wire to steel plug arc welded to pipeline. On covered pipelines, insulate all joints and exposed areas. Backfill thoroughly

Slurry of Betonite and Plaster of Paris to increase the conductance of the soil and to decrease extraneous corrosion of Magnesium

FIGURE 5.71: SACRIFICIAL ANODES FOR CATHODIC PROTECTION

The system as designed will consist of some of the following elements: 1. 2. 3. 4. 5. Cathodic protection supply which may be a battery or a photovoltaic cell, grid supply, diesel generator, etc. Voltage control Dispensing pole (commonly known as sacrificial anode) Reference electrode which constitutes a cathodic Electrical connections

The positive terminal of the supply is connected to the dispersing electrode while the negative one is tied to the pipeline. In this way the supply reaches the pipe through the ground (avoiding corrosion) and comes back through the electric wire. In the process the corrosion is simply moved away from the pipe and localized at the sacrificial anode. The sacrificial anode is appropriately dimensioned and driven into or positioned in the ground at a determined distance from the pipe, ranging between 50 and 200 m, the current being supplied regulated by the voltage across the pipe and the reference electrode (cathode). The reference electrode is a Cu - CuSo4 pole placed in the ground above the pipe itself. The supplied current is regulated to maintain a voltage across the pipe and the reference electrode at a level of at least 0.85 V which is the minimal safe value to ensure cathodic protection. Voltage levels generally adopted in practical cases range between 0.85 5 V, the higher values being necessary if the pipes are insulated. The electrical resistance between the anode and the pipe may range between 0.5 10 ohms depending on the type of disperser, pipe cover, and the soil resistivity. The accuracy necessary in the voltage potential is about 25 mV. In most cases a constant current must be imposed despite ground resistance variation. Consequently, the power supply must be capable of delivering the predetermined current at maximum resistance values. Nevertheless for safety reasons, the supply voltage must not exceed 50V.
Page 5-106

CHAPTER FIVE
The intensity of current necessary is dictated by size and cover of the pipe. For each mm2 of surface of the pipe (tube) a current of between 0.2 and 1 mA is usually necessary; furthermore the current depends on the distance between each cathodic protection unit along the pipeline usually between 10 and 20 km. The power drained can vary between tens and hundreds, of watts. The power supply must be continuous and thus reliability plays and important role for such systems.

Also modularity is very important since the pipe network may have to be modified at some stage thereby requiring more power even with the deterioration of the pipe passive protection necessitates (requires) a modification to be effected over a period of years on the value of current that should be injected into the pipe protective system. Since power supply reliability both from continuity and requisite voltage tension is extremely important for such a protective system and designed to match physical characteristic variation within the premises of protection, the selection of the type of power supply to be adopted must be examined critically. Also the implications of selecting one alternative supply source or another must be analysed from the economic, operation and maintenance points of view. Where economic implications are not prohibitive the use of photo-voltaic cells is strongly recommended in remote areas. Needless to say many pipelines pass in areas where there is no national grid supply or if any, the harnessing of such a power source would be too expensive due to high costs of voltage transformation and protective equipment. Thus if screwed GS or welded steel pipelines are being considered, the cost and provision of cathodic protection must be carefully considered. It should also be remembered in this connection with respect to lightning, that any cathodic protective system must itself be protected against lightning and cost of providing. The low tension voltage requirement for the cathodic protection system needs comparing to the very high transmission supply voltage, and a very sophisticated lightning protective system may have to be improvised with no justifiable advantages. The use of solar recharged batteries is seen as a very good proposition for this purpose at pumping stations and for very short materially contiguous ferrous pipeline systems that as a result have electrical continuity in which case, only very short term periodic inspection is necessary. 5.7.6 Overload (O/L), Current and Single Phasing Protection

Electric motors operated from direct-on-line, star delta or auto transformer starters need to be protected from overload or over-current using thermal overload relays incorporated with bimetallic strips. The bi-metal strips may be designed for direct heating where small motors are involved because the current involved is small. For handling heavy currents, the bi-metal strips need to be designed to be heated indirectly and therefore a current transformer must be used. The O/L relay should be part of the motor starter and be located in the starter circuit. It must be emphasized here that if the O/L relay is located at any position other than at the correct point in the starter circuits, the protection rendered by it will be ineffective under

Page 5-107

CHAPTER FIVE
certain conditions. It should be incorporated with a manual release knob, and temperature compensating trip. Furthermore the O/L relay must be protected against short circuits. It should be noted that the release times of thermal overload relays becomes shorter as the current increases. When subjected to very high currents however, the bi -metal strips are unable to react quickly enough. Hence they are likely to suffer thermal damages if exposed to short circuits etc., and to guard them against this they must always be protected from short circuits by means of high speed strips or fuses. The maximum allowable fuse rating must be obtained from the manufacturer of the starter At this juncture, it is brought to the attention of the designer the fact that it is very widely mistaken or taken for granted that an O/L relay is capable of providing protection to the motor against single phasing and under voltage, since these two incidents automatically result in a current rise in the respective system. This misconception has led to extensive damage to motor windings and caused puzzle to those concerned as to why the properly rated and judiciously set O/L relays fail to protect the motor under such condition. Even more important is the financial losses and inconvenience resulting there from. The best way to ensure effective protection against phase failures and under-volt is to incorporate an asymmetrical relay into the starter circuit. Such a relay monitors any potential change either between each two live lines or between each live line and the neutral line. When properly connected and set, any phase failure or under-voltage would cause the relay to trip the control circuit of the starter thereby stopping the current flow to the motor windings and thus protecting it. 5.8 ERECTION OPERATION AND MAINTENANCE INSTRUCTIONS

5.8.1 Pumpset Lining up and Final Coupling (1)Initial Checks (a) All bearings should be examined for any signs of ingress of cement dust, grit dirt or mechanical damage. If necessary, ball and roller or oil filled bearings should be cleaned and regreased or refilled with oil in accordance with instructions from manufacturers. (b) The pump half coupling should be turned to ensure that the pump rotating assembly is quite free and that no distortion, or damage has occurred in transit. Note: In the case of a multistage pump with a balance disc, ensure that before rotating the shaft, the balance disc is clear of its mating renewable plate face to avoid galling. (c) Check that the drive rotor is free to turn. (2) Lining-up Pump and Driver (a) Mount the pump on the bedplate, care being taken to ensure that the feet of the pump and the bedplate pads are perfectly clean and free from burrs. Locate the pump by means of dowel pins and holding down bolts. Tighten the holding-down bolts.

Page 5-108

CHAPTER FIVE
(b) The faces and periphery of the half couplings should be cleaned thoroughly to ensure that all rough or ragged edges are removed as these would give false alignment readings. (c) Place the driver and/or gear box on the bedplate, having checked that the faces are perfectly clean and free from burrs, and insert the holding-down bolts but do not tighten up. Check the height of the pump and driver and/or gear box couplings and insert shims (which must be clean and free from oil and burrs) under the drive and/or gear box feet to obtain initial approximate coupling alignment. (3)Final Coupling (a) When running the half coupling should be separated by a gap dependent upon the type being used and this distance should be confirmed from the general arrangement drawing. Note: If the motor is fitted with sleeve type bearings, the coupling gap must be set with motor rotor in its magnetic centre. Most motors have the magnetic centre marked, but for further details consult the motor instruction manual. (b) Where dial gauges are available to check coupling alignment follow the procedure given below: Lay a straight edge across the rims or hub tops of the two half couplings and check radial alignment at quarter positions. Both couplings should be rotated and readings checked at quarter positions. Check the radial gap as-shown, again at quarter positions and rotating both couplings. Any repositioning of the motor to correct for radial misalignment must be followed by a further check on gap clearance which should be carried out again at four points, at 90o to each other using either a taper or a parallel gauge in conjunction with feelers. 5.8.2 Parallel Alignment The following procedure should be followed: (1) Clamp a dial gauge to the coupling halves and zero the gauge. (2) Link the coupling halves so that both shafts can be rotated together. If the shafts cannot be linked together, mark each coupling half with a coincident line (3) Rotate both shafts and take readings at each quarter revolution. If the shafts are not linked together, move first one shaft through a quarter revolution; then move the other until the coincident lines marked coincide. Take a reading and repeat at each quarter revolution. 5.8.3 Angular Alignment

The following procedure should be followed: (1) Clamp two dial gauges to the coupling halves. (2) Link the coupling halves or mark them as described for parallel alignment above. (3) Rotate both shafts through 180o and note the gauge readings. If both gauge readings are the same (but not necessarily zero), the angular alignment in the vertical plane is correct

Page 5-109

CHAPTER FIVE
(4) Rotate both shafts through a further 90 and take gauge readings. If the readings are the same but not necessarily zero, the angular alignment in the horizontal plane is correct 5.8.4 Routine Maintenance Schedule Pumps

Routine maintenance and servicing are essential to maintain a practical serviceable condition. The following are to be noted: (a) Where possible, couplings should be aligned within 0.025 mm ( 0.001), i.e. 0.050 mm (0.002) full indicator movement on the dial gauge. (b) It must be emphasized that since pin and bush type couplings are designed to reduce the transmission of shock to bearings, etc., and do not compensate for misalignment, they require the same accuracy of alignment as do rigid couplings. (c) Some couplings, notably the double engagement gear type, laminated metal element spacer type, and rubber tyre couplings, can operate satisfactory with a greater degree of misalignment. The greater flexibility of these coupling is intended to compensate for misalignment which may develop during operation, due to pipe work loads, settlement of foundations, etc. and does not preclude or excuse the need for accurate initial alignment. For the procedure to be followed when aligning these couplings, see the instructions supplied by the manufacturer. (d) Alignment procedures for electric motors depend on motor construction. Motors with end shield mounted bearings can be aligned by moving the complete machine. This will not upset air gaps. Motors with bedplate mounted pedestal bearings are aligned by moving the bearing pedestals. This will alter motor air gaps, and the stator must then be moved to equalize air gaps. Check air gaps at 12, 3, 6 and 9 o'clock positions, taking measurements at both ends. Turn the rotor through 180o and repeat the measurements. No measured air gap should differ the from the mean of the measured air gaps by more than 5% (e) When checking the direction of motor rotation, ensure that the pump has been disconnected from its source of power, i.e. disconnect coupling/belt drive as appropriate.

Page 5-110

CHAPTER FIVE
5.8.5 (1) Pipework Installation General It is essential that no load be imposed on the pump branches as this is liable to disturb shaft and bearing alignment resulting in excessive wear and possibly seizure. The pipework should, therefore be independently supported and anchored against pipe thrusts When erecting the pipework, the pump flanges should be the final closure joints. Check coupling alignment before and after the pump flanges have been tightened. Any alteration in alignment is indicative of pipework loading on the pump branches and steps must be taken to remove the load. (2) Suction Pipework (a) (b) Where possible arrange for the pump to operate against positive suction. Eliminate all possibility of air pockets being trapped in the suction pipework, e.g. ensure pipework rises continuously towards the pump, where the suction pipe diameter may be reduced in the direction of flow use flat topped taper pieces. The diameter of the suction pipe should be sized according to the flow and allowable head loss and may not therefore, be the same diameter as the pump suction branch. A short straight length of pipe adjacent to the pump suction branch is desirable. Where a bend is unavoidable it should be of the largest possible radius. The suction pipe should be accessible and not embedded in concrete. When choosing foot valves and strainers, bends, etc., select those which provide minimal restriction to flow. Care must be taken in the design of the suction inlet. Ideally the pipe end should be at a depth of at least 5 pipe diameters below the water surface, and at least 0.5 m from any side wall. If the pipe must be at a depth significantly less than this, it is recommended that it be positioned within 150 mm of the side wall to reduce the possibility of vortex formation. Consul t the pump manufacturer if in any doubt. It is essential to ensure that delivery pipework is adequately supported and anchored to resist hydraulic thrust. When selecting valves for delivery pipework the following points should be noted:(i) (ii) The diameter of the valve is not necessarily based on the diameter of the pipework and is dependant on flow. Non-return-valves should be sized to give a maximum flow velocity of 3 m/s (10 ft/s). Where a surge vessel is fitted to the pipeline or pumps operated in parallel, consideration should be given to fitting non-slam or linear flow nonreturn valves. In all cases where it can be afforded, linear flow non-returnvalves are recommended. Where delivery pipes do not rise continuously and air pockets can form, it is important to provide some form of air valve and again CATT type air-valves are recommended because of their additional anti-surge anti-shock function.
Page 5-111

(c)

(d) (e) (f) (g)

(h) (3) (a) (b)

Delivery Pipework

(iii)

CHAPTER FIVE
5.8.6 Operating Procedure 5.8.6.1 General The following procedure outlines the most important steps involved in pump operation. Any departure from this procedure must be based on good engineering practice. 5.8.6.2 Priming (1) Priming a centrifugal pump involves the removal of air, gas, or vapour from the suction pipework and pump casing. As centrifugal pumps are not self priming, they must be primed prior to start-up. If this is not done the pump will not operate, and in addition serious damage to water lubricated bushes could result. The simplest method of priming is to arrange the pump so that it has a positive suction pressure. The air can then be released by opening the air cock provided. Where a positive suction is not possible, it is normal practice to fit a foot valve and strainer to the suction pipe inlet and use one of the following methods of priming. (a) Back priming from the delivery main using a by-pass around the non return-valve. It is important to ensure that the suction pipework and valves are rated for the delivery pressure. Using a tundish fitted at the highest point on the pump casing to fill the pump and suction pipework using an external water supply Vacuum priming equipment such as an air or water operated ejector (jet pump) or a rotary exhauster can be used to draw water from the suction line into the pump casing.

(2) (3)

(b) (c)

(4)

In a new system, the pump should stay fully primed from shut down to start-up, making further action unnecessary. However, foot valves are liable to wear or jam and may lose the prime after a short period. It is advisable therefore, to check the state of prime of the suction main and pump before start-up

CAUTION: IT IS IMPORTANT that air does not leak into the pump during operation, because if the pump operates under such conditions, serious damage may result. 5.8.6.3 Starting (1) Before initial start-up, disconnect the drive coupling and check the rotation of the driver. An arrow on the pump casing indicates the correct direction of rotation. (2) Check the pump shaft for freedom of rotation. (3) Where external sealing water is provided, check that the supply is available at the specified pressure. (Normally 0.69 - 1.04 bar (10 - 15 psi) above pump suction pressure. (4) Ensure pump and suction pipe are correctly primed. (5) Ensure suction isolating valve (if fitted is fully open). (6) Ensure delivery isolating valve is closed. (7) Where a surge vessel is fitted in the pipeline check that the air volume is correct. (8) Where a burst disc device is fitted, check that the isolating valve is fully open. (9) Start the driving unit according to the manufacturers instructions. (10) Open the pump delivery valve slowly until the required pressure or flow is obtained.

Page 5-112

CHAPTER FIVE
CAUTION: 1. 2. 5.8.6.4 Stopping (1) (2) 5.8.7 Slowly close the delivery valve and shut down the prime mover according to the manufacturer's instructions. Shut off external sealing water supply (if fitted) to relieve stuffing box pressure and minimize wastage Prolonged running with a closed delivery valve can damage pump internal components and must be avoided If the delivery pressure does not build up, stop the pump and repeat priming operation.

Maintenance Chart

A typical maintenance chart is shown in the following Table.

TABLE 5.13: TYPICAL MAINTENANCE CHART
ACTIVITY Check Leaks for DAILY AFTER ONE WEEK AFTER ONE MONTH AFTER 3 MONTHS AFTER 6 YEARLY MONTHS REMARKS Extend to weekly if proven satisfactory during running in Extend to weekly if proven satisfactory during running in Checks can be made at flow indicators drain boxes and tundishes. Extend to weekly if proved satisfactory during running in Extend to weekly if proven satisfactory during running in Extend to weekly if proven satisfactory during running in Depending on environmental condition Extend to 2 weekly when satisfactory

Check bearing temperature Check cooling and lubricating water flow

Checks for vibration Checks for balance wear indicator Check that gland drains are clear Adjust glands as necessary to maintain slight leakages

Contd
Page 5-113

CHAPTER FIVE
TABLE 5.13: TYPICAL MAINTENANCE CHART (CONTD)
ACTIVITY Check oil level to up as required (ring oiled and thrust bearings) Check surge tank water levels (manual only) Clean and recharge ball/roller bearing with grease Lubricate ball/roller bearings Examine gland/renew packing if gland follower fully up Check couplings alignment (horizontal pump sets) Check couplings pins and bushes for water Check holding down bolts for tightness Lubricate gear couplings Renew oil in bearings Check at end of running in period then as shown Refer to maintenance instructions DAILY AFTER ONE WEEK AFTER ONE MONTH AFTER 3 MONTHS AFTER 6 YEARLY MONTHS REMARKS To be carried out when stationary except where running levels are marked Extended from running in period refer to maintenance instructions To be carried out 3 months after commissioning then yearly refer to maintenance instructions Add only small amount of grease Refer to maintenance instructions frequency can be changed depending on findings

OVERHAUL:

This to be determined by decrease in pump discharge and discharge pressure to an unacceptable level. Refer to maintenance instructions

Page 5-114

CHAPTER FIVE
5.9 5.9.1 ECONOMICS OF ELECTRICAL POWER SYSTEMS Power Factor Correction

Pumping installations qualify for TANESCO Tariff 2 which stipulates that the amount chargeable per installation fall into two categories namely: 1. Maximum Demand Charge 2. Unit Charge The maximum demand charge is based on the maximum kVA meter reading registered by the respective kVA meter for any consecutive 30 minutes during each month of the scheduled meter reading as follows (2008):
TABLE 5.14: TANESCO TARIFF (2008)
Tariff 1 tariff for Above 283 kWh, less than 7,500 kWh per month @230V or 400V (T1) more than 7,500 kWh per month, demand less than 500kVA, @ 400V (T2) @ 11,000V and above (T3) Energy (kWh) 0 Energy TZS/kWh 106.00 Demand TZS/kVA Service charge TZS 1,892.00 VAT 20%

70.00

7,680.00

7,012.00

20%

65.00

7,123.00

7,012.00

20%

POWER FACTOR (p.f) SURCHARGE

Tariff 1 1 2 3 tariff for domestic loads less than 7,500kWh per month @230V or 400V all other loads less than 7,500 kWh per month @ 230V or 400V more than 7,500kWh per month, demand less than 500kVA, @ 400V @11,000V Minimum p.f 0.95 0.90 0.90 0.90 Surcharge formula 1+3%[95%-p.f.%] 1+3%[90%-p.f.%] 1+10%[90%-p.f%] 1+10%[90%-p.f.%] Applied to kWh kWh kVA kVA

Unit charge is worked out according to the actual active energy (kWh) consumption. Since the maximum kVA demand is a function of the average power factor of an installation and there is a possibility of making upward adjustment of the same (power factor) it is strongly recommended that the economic viability of making such adjustment with a view to reducing the max. kVA demand should not be ignored as substantial saving in finance can be realized on the energy cost. Conversely, penalties can be charged. Very little can be done to cut down the active energy consumption of a fixed installation since the energy requirement is dictated by the overall efficiency of the pumping set, duty point and the efficiency of the distribution system within the installation which may be considered as fixed. For the purpose of reducing the maximum kVA demand of an installation where the overall power factor is undesirably low, central or group compensation shall be applied but limited to a group of squirrel cage motors and as far as possible shall be avoided completely as it is in most
Page 5-115

CHAPTER FIVE
cases complicates the maintenance of such installations due to sophistication in the switch gear. This is due to the fact that the power factor correction equipment using these concepts must be connected directly to the switchgear thereby necessitating incorporation of special switching arrangements for the equipment. Single compensation is strongly recommended, especially for large installations involving high tension slipring motors. For this type of installation, static capacitor banks are used and they are directly connected to the motor starter terminals. Such capacitors have built-in fuses to protect them. The capacitive voltage is discharged through the motor starter resistance No loss or almost negligible loss percentage static capacitors shall be aimed for with a loss of 12% as this can be considered acceptable. The discharge characteristic need to be such that the period corresponding to full discharge is as short as possible to avoid shocks to the operator in case the operators touch the motor housing after switching off. 5.9.2 Requisite Reactive Power Calculations

The sizing of the capacitor for the purpose stipulated above can be determined in accordance with the following formula: kVarc = (3 Io kV) Where, kVarc = the reactive capacitive power Io = no load current of the motor kV = rated operating voltage = line voltage The no-load current value, if not provided by the manufacturer of the motor may be estimated as follows: Io = 2 I1 (1-cos ) Where, I1 = full load current of the motor (this should always given on the nameplate) Cos = magnitude of the uncorrected power factor Should the size of a standard capacitor bank have a rating greater than the computed requisite kVarc then a smaller rating capacitor must be adopted. It should be noted in this connection that a capacitor bigger than the requisition size results in over-compensation at no load, i.e. the power factor would be negative at no-load. This must be avoided. It follows that the motor must be fully compensated at no-load and not a full load. The formula given for kVarc estimation will generally result in compensation of the respective motor to about 0.95 which is technically acceptable.

Page 5-116

CHAPTER FIVE
5.9.3 Cable Sizing for the Capacitor Connection The sizing of capacitor cables shall be such that any harmonics are avoided between the capacitor bank and the motor starter winding. In this connection, it should be noted that harmonics will only develop in the event of resonance and that resonance will only occur when the magnitude of the inductance due to the power transformer equals to the capacitance of the entire installation. Hence the designer of the respective electrical installation must take steps to ensure that the inductance of the transformer secondary windings is not equal to the group capacitance (aggregate capacitance) of the installation. i.e.: XLtf Xcg Where XLtf Xcg = = Inductance of the power transformer Group capacitance of the installation 5.22

It is also necessary to control the capacitive power which may induce resonance as follows: Pc = Where, Ps = rated apparent power of the transformer (kVA) n = multiple of the frequency of the harmonic in resonance (with the 5th harmonic, n = 5; with the 7th harmonic, n = 7) uk = percentage of the short circuit voltage of the transformer k = phase angle of the short circuit current of the transformer, as indicated on the nameplate of the transformer 5.10 ENERGY SOURCE CONSIDERATIONS As far as possible only conventional energy sources shall be considered. Nevertheless the use of non-conventional energies must not be ignored since some of them may prove more economical than the conventional energy. The use of biogas for small installations for instance may prove very advantageous in certain cases where animal dung is available in abundance. Another case where use of non conventional energy is ideal is where a gravity or rising main passes with a pressure residual insufficient to feed a branch main to a small low pressure zone. At the junction of the gravity or rising main and the branch line there may be a high tension but the booster pump necessary at the point would certainly require medium tension. The cost involved in the procurement and erection of the necessary step-down transformer and switchgear may be prohibitive in comparison with a non-conventional energy alternative. In this particular case it is strongly recommended to examine the possibility of adopting a mini hydro plant to generate and supply electrical power to the small booster pump. The use of windmill may also be considered in areas where wind is more or less perennial. It is impossible however to expect any area with perennial wind. Even in areas with almost perennial wind, the wind is in most cases not available for at least 4 hrs each day. Hence it should not be Ps n uk Sink
2

kVarc

5.23

Page 5-117

CHAPTER FIVE
expected that wind energy is continuous. There will also be times when the wind velocity is so low that pumping is no longer possible. Solar pumping can also be considered for small yield boreholes but is still unlikely to be cost effective for larger requirements. 5.11 INSTRUMENTATION The following instruments shall be incorporated wherever applicable. Pressure gauges to monitor suction and delivery pressure conditions where surface pumping is involved. Flow meters for monitoring production capacities from a pumping station, as well as consumption from a storage tank, input into a storage tank, and consumption in each connection in urban water supply systems. Voltmeters and ammeters for each switchgear and each motor starter panel. Pressure gauges for monitoring suction head and delivery head A pH value meter for monitoring pH fluctuation in the incoming raw water to a treatment plant and in the treated water from the treatment plant. Chlorine residual meter for measuring chlorine concentration in water either from the treatment plant or into the treatment plant in cases where pre-chlorination is deemed absolutely necessary. Chemical dosing meter mounted on the chemical metering pump. For large motors above 200 kW, a thermometer shall be incorporated to monitor the motor bearing temperature under running conditions. Hour-meters on larger but not on smaller rotating assemblies to monitor the total period of use of such items. 5.12 LIFE CYCLE ANALYSIS, AN INTRODUCTION As with pipe and pipelines discussed in Chapter 4, it is not yet possible to include a quantitative Life Cycle Analysis into the design of Pumping Systems. However, designers need to be aware that this is coming and to keep abreast of literature and developments in this important field. Pump manufacturers are beginning to pay attention to this and guidance from manufacturers can there for be sought when considering new pumping plant. The following is therefore an introduction to the subject. Effective cost reduction with preventative planning is possible when considering overall costs of a plant item and the most important objective is to identify any savings that can be realised in the course of its operating lifetime. Most pump manufacturers are continually improving their range of pumps, automation systems and services to help utilities maximise the savings. In Tanzania, energy costs are a primary consideration and the largest component of overall pumping cost and so offer the greatest area of potential savings, so the more energy efficient the pumping system the better.

Page 5-118

CHAPTER FIVE
A typical example for Europe is shown in the figure opposite, and will not be all that different for Tanzania except that the operation and maintenance % can be lower and all other areas proportionally larger. In addition, the longer the intervals between the need for servicing and maintenance the more cost effective the situation whilst the use of diagnostic systems can further improve the situation.

FIGURE 5.72: TYPICAL OVERALL PUMPSET LIFE CYCLE COST BREAKDOWN

In summary, the areas that need to be considered are: Energy savings due to maximising efficiency Reduction of maintenance and repair costs Increased productivity due to longer availability, and Process optimisation by fine tuning of pump and plant.

Areas identified that have the potential to save energy are: The use of variable speed drives Improved system design Correct pump size Selection of high efficiency pumps and motors Optimisation of installation and maintenance Possible potential savings 20% 10% 4% 3% 3% 40%

At present in Tanzania the use of variable speed drives except in large, well operated and maintained situations is not recommended. Even so the potential for reducing energy demands in other areas is still at least 20% of the overall energy demand that can be expected at present. In this regard, the past practice of buying in standard units in advance that might one day be needed and then selecting the best under the circumstances needs to be actively discouraged, whilst serious manufacturers need to be encourage to appoint technically competent local agents to be able to give the best advice possible instead of relying on salesmen whose job is primarily to sell regardless of actual suitability. The area that can be influenced the most locally is that of service and maintenance Service activities that are carried out at the beginning or during the lifetime of the plant account for a major portion of life cycle costs. All-in service packages need to be considered in this regard. However, the correct choice of service strategy depends on the type of plant, the objectives being pursued and the competence of in-house staff. Maintenance strategies are either preventive, condition-based risk-based and failure based. The first two improve the mid- and long- term availability of the plant. Purely budget driven

Page 5-119

CHAPTER FIVE
maintenance tends to be risk-oriented whilst failure oriented maintenance which has all to frequently featured in the past involves obvious risks. As the choice of service strategy has far reaching consequences for the life cycle costs, the decision should be taken during the design phase of any plant by the utility owner / operator and this made clear to the designer. Service management contracts have proven very effective elsewhere with savings of up to 25% in maintenance costs being achieved. Certain concepts of such contracts can however be readily adopted in-house including the building of a historic file for every single unit which is then used to track the work done on the unit in question. It should record all measures, the cause of any damage, mechanical analyses and spare parts needs. In this way the file serves as the basis for technical adjustments to the operating conditions and for defining the maintenance interval and the stock of spare parts required. Pump manufacturers are beginning to offer their purpose developed software that simplifies life cycle costing. A good package should: Require the entry of just a limited number of basic parameters Provide a clear overview of results in both graphical and tabular format Achieve straightforward archiving and retrieval of results, and Enable calculation and evaluation to be carried out by the utility organisation itself.

5.13 FORMULA AND CONVERSION FACTORS In this chapter some useful formula and conversion factors can be found. The formula can be used for calculating for example rated motor torque, moment of inertia, flywheel mass, etc. The conversion factors can be used to convert for example kW to HP, Celsius to Fahrenheit, km/h to miles / hour, etc.

Page 5-120

CHAPTER FIVE

nr

Page 5-121

CHAPTER FIVE

Page 5-122