28

Water Supplies B H Rofe MA(Cantab), CEng, FICE, FIWEM, FGS Rofe, Kennard and Lapworth, Consulting Engineers

Contents 28.1

Organization and management

28/3

28.2

Present consumption and estimated demand 28.2.1 Domestic consumption (details) 28.2.2 Industrial consumption 28.2.3 Agricultural requirements 28.2.4 Fire protection 28.2.5 Waste

28/3 28/3 28/3 28/4 28/4 28/4

28.3

Transmission and distribution of water 28.3.1 Pipes 28.3.2 Flow in pipes 28.3.3 Valves 28.3.4 Cost 28.3.5 Surge

28/4 28/4 28/6 28/7 28/8 28/8

28.4

Measurement of flow in streams 28.4.1 Current water 28.4.2 Crump weir

28/8 28/8 28/8

28.5

Measurement of water in pipes

28/9

28.6

Service reservoirs 28.6.1 Covered service reservoirs 28.6.2 Water towers 28.6.3 Valves and fittings for service reservoirs

28/9 28/9 28/9 28/10

The underground scheme 28.7.1 Development of a new source 28.7.2 Geology of source 28.7.3 Selection of source 28.7.4 Construction of boreholes and adits 28.7.5 Water-well casing 28.7.6 Testing boreholes and wells 28.7.7 Yield of boreholes

28/10 28/10 28/10 28/11 28/11 28/12 28/12 28/12

28.7

28.8

28.9

Surface water schemes 28.8.1 Development of a new source and requirements 28.8.2 Geology of source 28.8.3 Economics of storage and yield of reservoirs 28.8.4 Catchwaters

28/12 28/12 28/13 28/13 28/14

Formation of reservoirs 28.9.1 Valve towers 28.9.2 Floods in reservoir practice 28.9.3 Overflow weirs 28.9.4 Drawoff and diversion culverts 28.9.5 Earthen embankments and earth dams 28.9.6 Concrete dams 28.9.7 Examples of raised lakes 28.9.8 Pumped storage reservoirs

28/14 28/15 28/15 28/15 28/15 28/15 28/17 28/18 28/18

28.10 Desalination 28.10.1 Multi-stage flash distillation (MSF) (vacuum separation) 28.10.2 Electrodialysis (membrane-electrode separation) 28.10.3 Reverse osmosis (membrane pressure separation) 28.10.4 Freezing

28/18

28.11 Treatment of water for potable supply 28.11.1 Water characteristics 28.11.2 Storage 28.11.3 Algae 28.11.4 Aeration 28.11.5 Coagulation 28.11.6 pH control 28.11.7 Precipitation 28.11.8 Mixing

28/19 28/19 28/20 28/20 28/20 28/20 28/20 28/20 28/21

This page has been reformatted by Knovel to provide easier navigation.

28/18 28/19 28/19 28/19

28.11.9 28.11.10 28.11.11 28.11.12 28.11.13 28.11.14 28.11.15 28.11.16

Flocculation Sedimentation Filtration Backwashing Chlorination pH adjustment Taste control Waste products

28/21 28/21 28/21 28/21 28/21 28/21 28/21 28/22

References

28/22

Bibliography

28/23

This page has been reformatted by Knovel to provide easier navigation.

28.1 Organization and management The organization of water supplies in Britain, as in other parts of the world, has evolved from initiatives by private companies and public corporations. This situation still exists in many parts of the world, but in Britain a framework of organization and management1 has been created by successive Acts of Parliament starting with the Waterworks Clauses Act, 1874 through to the Water Act 1973 (England and Wales) and the Local Government (Scotland) Act 1973. Further modifications were introduced in the Water Bill 1983. In England and Wales, the 1973 Act created a framework of river basin management, with ten regional authorities responsible for the complete hydrological cycle including conservation, water resources, treatment and distribution, land drainage, sewerage and sewage treatment. However, in Scotland water supply and sewage are the responsibility of the district council, except in the case of the Central Scotland Water Development Board who act as a bulk supply authority. River purification is the responsibility of separate boards except in the areas of the island district councils. The Scottish pattern of management and organization of water supply, together with agent water companies (as in England and Wales) has been adopted in many English-speaking countries in local forms to comply with the form of government, but there is a movement towards the more logical format of total river basin management.

Water Undertakings in south-east England by Sharp in 1967.32 This is shown in Table 28.2.

Table 28.2 Breakdown of domestic consumption (1967) Estimated 1967 Forecast of possible average consumption average consumption in 2000

Component

gallons/ head/ day

Drinking and cooking 1 Dishwashing and cleaning 3 Laundry 3 Personal washing and bathing 10 Closet flushing and garbage disposal 11 Car washing Garden use and 1 recreation Waste in distribution 5 Total

34

litres/ head/ day 4.5

gallons/ head/ day 1

litres/ head/ day 4.5

13.5 13.5

4 5

18 22.5

45.5

13

59

50

14 1

63.5 4.5

6

27.5

8

36.5

4.5

22.5 154

52

236

28.2 Present consumption and estimated demand Demand for water varies according to the type of supply area but may generally be considered under the three headings of domestic, industrial and agricultural. Within the year, the monthly and weekly totals will vary considerably due to seasonal effect (wet or dry periods) and socio-economic effects (e.g. holiday periods, festivals, growth seasons). There will also be daily variations within each week and peak hours during each day. A peak hourly rate of around 3 times the average rate needs to be considered in the design of distribution systems, and in respect of local storage requirements. Domestic consumption assessments in Britain during the period 1976-78 are summarized in Table 28.1, together with typical figures arising from studies in other countries. In Britain domestic supplies are not generally metered2 but in most other countries domestic meters are a common feature. It is doubtful if metering offers any saving in cost, but with strict supervision and control it can assist in controlling demand using an increasing scale of charges at higher rates of consumption.

Table 28.1 Domestic consumption per head (1976-1978) Area or country

Consumption (I/head/day) average range

England & Wales Scotland London (urban) Thames (rural) Middle East (arid) Far East (tropical)

200 275 260 210 —

140-330 240-350 — — (standpipes) 50-450 120-400

28.2.1 Domestic consumption (details) A typical breakdown of the present consumption and an estimate of future demand was made for a group of six selected

28.2.2 Industrial consumption No generalization can be made as the industrial consumption in each town varies considerably according to the nature of the industry both in quantity and quality requirements. The water is generally required during the working day and this factor must be taken into account in the design of pumps, pipes and reservoirs as it affects the peak rates of flow. As general guidance the following examples are typical. (1) For brewing, the quantity of water is substantially the amount brewed, but for beer the water is preferably hard; for stout, soft; cider must be made from pure soft water without iron. (2) Canning is best done with hard water (except for peas), and iron must be less than 0.5 mg/1: anything between 20 and 40 I/kg canned. (3) The dyeing industry requires soft, iron-free water, and about 100 I/kg, mercerizing textiles takes 250 I/kg. (4) Industries such as distilling, ice-making and mineral-watermaking require large amounts of water, plus that for power purposes in steam-raising. (5) Leather requires 80 I/kg of raw hide tanned, water rich in sulphates being preferred. Rubber requires 70 I/kg processed. (6) Paper or cardboard manufacture requires anything between 60 and 360 I/kg. (7) A ton of soap requires about 22001 of water in its manufacture. (8) In the UK, sugar beet takes about 5 I/kg used in washing the beet, dissolving the sugar, transporting the material in the factory and in steam-raising. (9) In the heavier industries the following quantities may be taken as approximate, e.g. railways take about 0.221/ 1000kg of goods carried per kilometre. Cement takes 3 I/kg. Coke might consume 13 to 181 of crude water per kilogram for cooling. Electricity works take 671 of crude

water per kilowatt generated, for make up or loss in cooling towers, and 1.51 of fresh water per kilowatt for boilers. Steelworks would consume some 91 of mostly crude water per kilogram of steel manufactured. (10) The cost of industrial water or metered supplies varies considerably but is generally in the range 2p to 1Op per 1000. 28.2.3 Agricultural requirements In addition to the human population, allowance must be made in a dairy farming district for the cow population. A cow requires as much as 135 to 180 I/day and there may be special requirements such as for bottling - 1001 milk means 2001 of water; manufacturing 455 kg dried milk needs 5501,455 kg alum needs 45001, 455 kg cheese needs 7501. Where intensive fruit farming is practised a complete network of pipes is required throughout an orchard for treatment and irrigation. Similarly, considerable quantities of water are required to maintain bowling greens, golfcourses and racecourses, and overhead irrigation of crops by rotary sprinklers is on the increase. In the Thames Valley a total of 90 Ml/day has been estimated as the requirement for irrigation by the year 2000. Very high consumptions of the order of 50 000 to 100 0001/ha per day would be required by a market gardener, and for tomatoes under glass approximately 200 0001/ha per day. Again, the watercress industry4 at certain times of the year consumes very large quantities of water and may require between 2.5 and 5.7 million 1/ha per day. Paradoxical as it may seem, more water may be required in winter to keep watercress from freezing than in summer to keep it from scorching.5 28.2.4 Fire protection Generally, hydrants are spaced not more than 13Om apart. Important buildings may require additional protection, i.e. more than two hydrants within 90 m. For less important buildings one hydrant within 140 m may suffice. Hydrants should be 6 m or more away from buildings, are best placed at crossings or corners, and are usually fixed on short 80 mm branches from the main which should not be less than 100 mm. Fire mains should deliver 5501/min at each hydrant expected to be in use at the same time (generally two). As pressure in a main to command the highest buildings is generally impracticable, fire engines are used to deliver 1200 to 20001/min to a height of 50 m through a 24-mm dia. nozzle; the larger fire engines deliver up to 45001/ min. A residual pressure of 3 m at the ground is desirable to avoid the engine creating a vacuum in the main on the suction side. In towns the calculation for the distribution of water is based on very general assumptions of the amount of water required at any given moment, and it may not be practicable to design adequately for fire protection if the mains are assumed to be taking the maximum hour's domestic and industrial requirements as well. A good practical arrangement of valves and hydrants based on experience and checked occasionally by simple network analysis is of more value than any very exact calculations. Nowadays the fire authorities work closely with the water authorities to determine the positions of fire hydrants. 28.2.5 Waste Some consumption of water by waste is inevitable and few statutory undertakers can seriously claim a figure of less than 10%, whilst in some areas where pressures are higher or the mains and services are old or in poor condition, or where efficient waste prevention methods are not applied, the wastage may amount to as much as 50% or more. Waste may be due to a

number of factors including: (1) leakage from reservoirs, mains and other works of an undertaking, and from consumers' pipes and fittings through apertures, fractures, defective joints; (2) faulty washers and valve seatings; (3) bad design, failure to turn off taps; and (4) in all cases leakage and waste are intensified by unduly high pressures. Waste can be detected by detailed examination of the distribution system or house-to-house inspection, apart from a detailed check on the main reservoirs and aqueducts, etc. 28.2.5.1 Examination of the system It is best to examine the water system section by section between midnight and 5.00a.m. and check the night flow by a meter capable of reading small flows and recording them on a chart. A specific test on a 12mm lead pipe under 3.2kgf/cm 2 pressure gave a loss of 46 000 I/day for a 0.6 cm hole, 17 000 I/day for a 0.3cm hole and 1600 I/day for a 0.15cm hole. Tests on newly laid mains often call for a loss not exceeding 1 I/day per centimetre of diameter per kilometre of length. House-to-house inspections are probably in most cases the most effective way of checking waste. A dripping tap wastes up to 500 I/day and one running full as much as 10 000 I/day. The provision by the water authority of facilities for the rewashering, renewal and adjustment of taps, and repairs to service pipes at the lowest possible cost, undoubtedly encourages consumers to report leakages promptly and is an overall economy. In recent years several more sophisticated systems6 have been developed to detect leaks in mains, and to identify the precise points of leakage, thus saving a lot of abortive exploratory excavation.

28.3 Transmission and distribution of water Water may be transmitted under gravity along open or covered channels, through tunnels or through pipes.7 Open channels are often used for catchwaters, waste-water channels or for river intakes to pumping stations in pumped storage schemes. Nowadays they are not generally used for the transmission of treated water due to the danger of pollution. In some cases canals are adapted as aqueducts for the transmission of water. Some large aqueducts have been constructed with sections of covered channel constructed by 'cut and cover' methods and modern practice is to construct these of plain or reinforced concrete. Where an aqueduct is required to pass through ground appreciably higher than the hydraulic gradient, tunnelling is necessary. The general principles of tunnelling are described elsewhere but for waterworks purposes the tunnels are usually lined, even in rock, partly to ensure that a fall does not block the waterway but also to reduce the friction. The use of pressure tunnels through the centre of a congested city with modern tunnelling methods is now becoming an economically satisfactory alternative to large trunk mains laid near the surface, provided the strata below the city are satisfactory. London is fortunate in this respect and several trunk aqueducts have been constructed in the London Clay. 28.3.1 Pipes The major part of water transmission is through pipes and there has been a considerable increase in the numbers of new types of pipes and joints of all sizes in the last few years, including spun and cast iron, ductile iron, steel, concrete, asbestos cement, and their range of joints. The ducts also include unplasticized PVC and polythene pipes with their corresponding joints. Several technical factors affect the final choice of pipe material, including internal pressures, hydraulic and operating conditions,

maximum permissible diameters, external and internal corrosion, and any special conditions of laying. Joints may be classified into three categories, depending upon their capacity for movement, namely rigid, semirigid and flexible. Rigid joints are those which admit no movement at all and comprise flanged, welded and the now obsolete turned-andbored joints. The semi-rigid joint is represented by the spigotand-socket caulked lead joint which has given service for well over a century but is now largely obsolescent. Flexible joints are used where rigidity is undesirable and comprise mainly mechanical and rubber ring joints which permit some degree of deflection at each joint. Amongst the joints included in this category are the Tyton joint for cast iron and ductile iron pipes, the Johnson coupling and Fastite joint for steel pipes and the lock joint for prestressed concrete pipes and the detachable and Widnes joints for asbestos cement pipes. Victaulic joints are frequently used where longitudinal tension is required. 283.Ll Cast-iron pipes (grey iron and ductile iron) The use of vertical cast-iron pipes is now limited to the flanged pipes employed in connection to reservoirs, pumps and treatment plant, the bulk of the iron pipes in waterworks service being spun iron, centrifugally cast in metal or sand moulds. Such pipes may be of grey iron, the latter having the advantage of higher tensile strength and reduced tendency to fracture, but they are also thinner in section. Grey iron pipes and fittings of sizes 80 to 700 mm are covered by British Standard (BS) 4622 and ductile iron pipes and fittings of sizes 800 to 1200 mm BS 4772. The former classes B, C and D have been replaced, in BS 4622, by classes 1, 2 and 3 which represent (for spun-iron pipes with socket and spigot joints) recommended maximum working pressures, inclusive of surge, of 10, 12.5 and 16 bar; maximum working pressures for flanged pipes and fittings in BS 4662 are lower than those for spun-iron pipes and where necessary the Standard advises the use of ductile iron or strengthened grey-iron fittings. Pressure ratings for ductile iron pipes (class K9) and fittings (class K12) vary with size: 40 bar up to 300 mm, 25 bar for 350 to 600 mm and 16 bar for 700 to 1200 mm (BS Code of Practice 2010: Part 3). The standard length of spun-iron socket and spigot pipes to BS 4622 is 5.5 m, and available joints include Tyton (the most widely used) and mechanical flexible joints of bolted-gland type. The standard length of flanged pipes is 4 m. The range of pipes likely to be available in future differs in some respects from that of BS 4622; as British, Japanese and American manufacturers have extended their range to meet demands for larger sizes on big supply schemes overseas in developing countries. Protective coatings include bitumen sheathing and the application of centrifugally applied concrete or bitumen lining can be provided where conditions warrant these additional safeguards. Where aggressive soil conditions exist the pipe may be protected by a tubular polythene sleeve.8 It is rare to install pipes without any protection, and where this has been done it has often proved a costly mistake. 28.3.1.2 Steel pipes British Standard 534 covers the manufacturer of steel spigot and socket pipes and specials. Manufacturers of steel pipes are generally able to manufacture special pipes of any reasonable size, thickness or shape to suit customers' requirements. Pipes vary in size from 50 up to 1800mm with wall thicknesses varying from 2.5 to approximately 20 mm. They may be jointed by welding with internal sleeve welds only, or internal sleeve welds and external sleeve welds to facilitate testing of butt welds. Alternatively, if greater flexibility is required, plain-ended pipes

are used in conjunction with Johnson couplings. The pipes may be protected with bitumen, concrete, or a sheathing of bitumen wrapped in hessian plus bitumen or coated in bitumen plus asbestos sometimes reinforced with woven glass. Where the surrounding groundwater is aggressive and the soil has a resistivity of less than 5000 O/cm3 then cathodic protection is required which may be provided either by sacrificial anodes, or by the imposition of a protection current from a direct current source such as an accumulator or transformer rectifier unit. 28.3.1.3 Asbestos pipes Asbestos cement pipe5 is made of a mixture of asbestos and Portland cement to form a laminated material of great strength and density. The material is less subject to encrustation in softwater districts and is not affected by electrolytic action. Flexible joints are used exclusively throughout the range of sizes up to 900mm in diameter for working pressures up to 90 to 122m head according to size. Special bends, tees and adaptors are not made in asbestos and those of cast iron are generally used for connections to asbestos pipes; BS 486 applies to A to C pressure pipes. 28.3.1.4 Concrete pipes Standard concrete pipes of plain or reinforced concrete are made up to a diameter of about 2.3 m (or occasionally greater) and are chiefly used to convey liquids not under pressure. Sizes from 150mm to 1.8m are covered by BS 556. The joints are generally of the flexible type such as the Stanton-Cornelius. Prestressed concrete pipes can now be manufactured over a wide range of sizes, varying in diameter from 635 mm to 1.8 m (BS 4625 covers sizes from 400mm to 1.8 m), and usually having a thin steel shell with a spun concrete interior lining stressed externally by prestressing wire on the outside of the steel shell, the whole then being protected by an outer covering of cement mortar. Working pressures of up to 12Om head of water can easily be obtained. In sizes over 1.2 m, longitudinal prestressing wires are normally employed and the steel cylinder is not used. The lock joints of the simple push-in selfcentring type are completely reliable provided that the manufacturers' jointing instructions are followed precisely. 28.3.1.5 Aluminium pipes Aluminium pipes are available up to 700 mm diameter, manufactured by the helical method, but these have not been used to any great extent in water supply. The evidence would so far seem to indicate that this is a material which might be more widely used provided that suitable precautions are taken to protect the material similar to those adopted for steel. The main advantage is in the reduction in weight particularly for overground purposes. 28.3.1.6 Polyvinylchloride andfibreglass wrapped pipes Polyvinylchloride pipes are light and easy to handle, corrosion resistant, and are generally available in sizes up to 600 mm in lengths of approximately 9m. Larger pipes for waterworks purposes may require to be strengthened and this can be achieved by the use of glass fibre reinforcement. The joints are usually made by a push-on type of rubber ring joint or by a solvent welded joint, the latter only being practical where site conditions permit. It has also proved possible to mole plough long lengths of up to 200 m of this pipe up to a diameter of 300mm underground without surface trenching. It should be noted that the coefficient of expansion of PVC is 8 times greater

than that of steel and considerable movement can take place in long lengths of rigidly jointed pipelines.

In considering the design of the pipeline the external loads generally arise from the weight of the pipe and its contents, the trench filling, superimposed loads including impact from traffic, and from subsidence. The design of pipelines and the strength of the pipes required has been considered empirically and the design method commonly used is that proposed by Marston and Spangler and described by Young and Smith.9 When a pipeline has to be laid above ground over some obstruction it may either be carried on a pipe bridge or be designed as a selfsupporting arch. Special design and fabrication are necessary in these cases.

Quantity: (Ml/day)

28.3.1.7 Structural design

28.3.2 Flow in pipes 28.3.2.1 Streamline flow Reynolds found by experiment that the average velocity below which streamline flow could be maintained for various diameters of pipes is approximately given by the equation: (28.1)

Hence, in a pipe of 300 mm diameter, for example, there should theoretically be a mean velocity of under 1/45 m/s, for maintaining streamline flow; in practice, however, it would be expected to carry water at a mean velocity of between 0.7 to 1.0 m/s, so that it is not economic to use this as a criterion for design. 28.3.2.2 Turbulent flow Froude found empirically for turbulent flow that the friction consumed by water passing through a pipe varied: (1) almost as the square of the mean velocity of the water; (2) almost as the area of the wetted surface in contact with water, i.e. the circumference and length of the pipe; and (3) the nature of the surface inside the pipe. 28.3.2.3 Friction in pipes Basic formulae have been derived by Bazin, d'Arcy, Chezy, Kutter, Ganguillet, Wisbert, Hazen, Manning, Flamant, Unwin, Barnes and others. The general principles are dealt with in Chapter 5. The universal pipe friction diagram based on the Hazen Williams formula

Loss of head: (m/1000 m)

diameter (cm) x mean velocity (m/s) = \ or less

As example

Figure 28.1 Pipe-friction diagram Example of use 450-mm pipe is required to carry 9 Ml/day. What would be the velocity and head loss? On the top half of the diagram, read across from rate of flow to size-velocity; read off diagonal line giving 0.6 m/s. Strike down vertically to appropriate e value (assume 100). Read across horizontally to left scale. Therefore head loss=1.2 m/1 OO m Multiply by lengths of pipe to get total friction loss

v= 1.318C(D/4)063(///L)054 was included in the Manual of British water engineering practice as chart D and it is reproduced in a simplified form in Figure 28.1. C is a constant depending upon the type, condition and diameter of the pipe, V is the mean velocity, D the internal diameter, H/L is the head loss per unit pipe length. This diagram is useful for preliminary design purposes, but for more accurate design the charts prepared by the Hydraulics Research Station,10 based on the Colebrook-White equation, should be used. 28.3.2.4 Economic diameter of pumping main Where water is to be pumped under pressure in a rising main, there is an economic diameter of pumping main to pass a given quantity of water. If the main is reduced in diameter the cost of the main will be less but the friction will be increased and the

cost of pumping, allowing also for larger machinery required, will be more. The converse also applies. Figure 28.2 shows the combined annual cost of the sinking fund taken at 35 yr on the main together with 15yr on the machinery and pumping at a given unit rate of electricity. The curve is particularly instructive in showing that it is generally more economical to err by choosing too large a diameter than too small a diameter as the left-hand side of the curve rises more steeply than the right-hand side. Lea11 has proved mathematically that the economic diameter lies between 0.535 and 0.675 times Q1'2 (Q being the quantity pumped in m3/s). Hence the velocity for economic pumping to daily supply can be deduced as lying between 0.8 and 1.4 m/s. However, if the supply is only intermittent (as, for example, in a standby supply), then use of a higher velocity up to 2 m/s would be justified.

Relative cost of main

Total cost

Relative diameter of main Figure 28.2 Variation of cost of main with diameter 28.3.3 Valves 28.3.3.1 Standard gate or sluice valve to BS 1218 These valves, of 37 to 300 mm diameter corresponding to castiron pipes, are to be had for class 1, working pressure 90 m and class 2 for a working pressure of 12Om. Valves up to 1.2m diameter are available. Valves are usually flanged to enable them to be removed, repaired and re-inserted without disturbing the rest of the pipelines. The nonrising spindle type has the screw totally enclosed within a casing, is operated by a key, and usually opens by turning anticlockwise although this should be checked by looking at the arrow on the casing. For special purposes there are valves with spigots and sockets, double spigots, Victaulic joints, hand wheels, exposed screw rising spindle types, and anticlockwise opening or a combination of any of these. Some sluice valves of the larger sizes, such as those situated in valve towers of impounding reservoirs, are often geared to facilitate operation by one man. Sluice valves may be provided with indicators to measure the amount of opening for both rising and nonrising spindles. There are locking devices. The larger sizes are also often provided with small bypasses to relieve pressures on opposite sides of the gate and the sizes of these bypasses are a matter of calculation, but the values in Table 28.3 may be taken as typical.

Table 28.3 Diameter of valves and bypasses Main valve diameter (mm) up to 200 225-300 350-525 550-900 900-1200

Bypass diameter (mm) 10-20 25-30 50-75 75-100 100-150

28.3.3.2 Butterfly valves Butterfly valves in accordance with BS 3952 are now extensively used as they are easier to operate than gate valves, smaller in size and generally cheaper. They should not be operated at water velocities of over 5 m/s where rubber seatings are included. 28.3.3.3 Reflux valves Reflux valves are also known as nonreturn, recoil, retaining, foot and flap valves. These are made up to 1.5m diameter or more. There are many types, single door, multiple door, horizontal, vertical and tilting discs.

28.3.3.4 Air valves Air valves are put at the highest points of mains and also on flat gradients of under 1 in 500 where the distances are more than 75Om. Those having large orifices let out air in large mains when being filled and those with small orifices are used for letting out air as it accumulates in coming out of the water. The double air valve has one large and one small orifice, a vulcanite ball being used for the large orifice and a rubber ball for the small 2.5 mm orifice. This is the type most generally in use but there are other modifications, e.g. that with and without an isolating valve to enable the balls to be inspected without emptying the main, or a refined type - kinetic - which prevents the balls slamming shut through a sudden rush of air and water. In this type the air is bypassed around the ball. For large mains a double air valve (isolating, kinetic type) is the most likely to be adopted, as it is capable of inspection and cleaning, which should be done at yearly intervals, and is not liable to slam shut when the mains are filled or refilled. 2833.5 Hydrants A hydrant consists usually of: (1) an 80mm branch from the main with a duck foot on which rests the screw-down hydrant and stand pipe; (2) a screw-down hydrant and standpipe placed on the main itself; and (3) an 80 mm pipe and hose attachment. 2833.6 Washouts Washouts are usually branches of (say) 80 to 150 mm diameter, but may be larger, leading from the main to a ditch or river with an ordinary sluice valve control. Special branch tees having the invert of the branch coincident with that of the main are made to enable any sediment to be washed out of the main. Flap valves should be put on the ends of the branches. They are used to clear the main of contaminated water or sediment, and should be sized to achieve a flow of at least 0.5 m/s in the main if possible. 28.3.3.7 Valves for special purposes The pressure-reducing valve. The ordinary sluice valve, half closed or throttled, is often used for reducing the pressure in a pipeline, but should not be, as special valves of various types are made for the purpose. The common form of pressure-reducing valve provides a constant pressure downstream at less than that upstream; the downstream water presses against a piston loaded with weights, and as the downstream pressure rises it forces the loaded piston upwards, and, through a system of levers, closes the valve. Modifications in this type of valve enable: (1) the downstream pressure to vary with the rate of flow through the valve; or (2) a reduction in the head through the valve by a constant amount. The pressure-retaining valve. This type is sometimes called a sustaining valve and is an adaptation of the pressure-reducing valve. It enables a variable inlet pressure to be converted to a constant pressure or to prevent the upstream pressure from falling. The pressure relief valve. This type is also sometimes called a sustaining valve and is an adaptation of the pressure-reducing valve. It may be a loaded spring affair or an adaptation of the pressure-retaining valve by which, for important installations, the times of opening can be regulated. Flow control valves. Such valves are intended for controlling constant flows in pipelines irrespective of pressure, and modifi-

cations of them provide for dividing flows into two, or introducing other flows to make up the quantity. Some flow valves have balanced discs electrically or hydraulically operated; other forms, often of the needle type, are designed to close hydraulically in the event of a burst main, or to open and close hydraulically or electrically at stated hours, as in pumping stations. 28.3.4 Cost The cost of water mains laid complete is of the order of £2.40 per centimetre diameter per metre with 10% to be added for valves and fittings (1985), including excavation (cover 1.2m) and backfilling, but excluding road restoration. 28.3.5 Surge

28.4 Measurement of flow in streams

A pressure transient or 'surge' is caused by a sudden alteration in the velocity of flow in a pipeline or aqueduct. This surge can be transmitted through the system causing an increase in pressure. Very high pressure can be generated leading to overstressing of the pipes, joints or pumps; likewise, low pressure can lead to the creation of a temporary vacuum causing infiltration, collapse or cavitation. These conditions may be caused by: (1) opening or closing a valve; (2) stopping or starting pumps or tankers; (3) the 'slamming' of a reflux valve or tidal flap; (4) on change of load demand on a hydro-electrical generator; (5) on vibration of guide vanes or impellers; or (6) generally by any situation causing a change in velocity of flow. The phenomenon is often called 'water hammer' and can occur in small-diameter domestic plumbing. The general principles were reviewed by Lapworth12 and a graphical analysis method was described by Lupton.13 Alternatively, the mathematical 'method of characteristics', which utilizes the same equations as the graphical analysis, is more easily used on a computer and enables the pressure at several points of the system to be calculated simultaneously, whilst alternative solutions can be easily compared and refined.14 Control of surge is necessary if the rise or fall in pressure is found to be more than the components of the system are designed to withstand. The methods to be considered could include the following: (1) restricting rate of valve closure; (2) increasing pump inertia (e.g. by putting a layer flywheel); (3) providing a surge shaft; (4) fitting an air vessel; (5) providing air admission or release valves; (6) fitting a bypass round a pump; or (7) fitting a weighted surge suppressor valve or similar mechanical device. Evaluation. The velocity of a pressure wave within a system may be calculated from the formula: a 1; f l -i/

/»n *i-vrr

^LAT++ IE J

steady velocity before closure. After the end of the reflection time the pressure will drop by the same amount. As an example, consider a simple system comprising a pump lifting water through a small cast-iron pumping main to a service reservoir against a static head of 80 m. The velocity of the pressure wave through the main is 1250 m/s, the initial velocity F0 is 0.3 m/s and g is 9.8 m/s. Therefore the surge pressure on shut-off is 7/max = aK 0 /g= ± 38.4m. This gives a minimum pressure at the pump of 80-38.4 = 41.6m and a maximum pressure of 118.4 m. To avoid the cost of reinforcing the pipeline, it would probably be best to instal an air vessel. The principles of this example can be applied to a more complex system and are fully described by Thorley and Enever.14

(28.2)

where p — density of water, g=acceleration due to gravity, A:= bulk modulus of water, d= internal diameter of pipe, / = thickness of pipe, E= Young's modulus for the material of the pipe, and v=Poisson's ratio for the material of the pipe. Typical values of a, resulting from the use of this formula, would be as follows: (1) small cast-iron pipe 1250m/s; (2) large thin-wall steel pipe, 900 m/s; or (3) small uPVC pipe 300 to 400 m/s. The time taken for the pressure wave to travel to the end of the system, length L distant, and back is 2L/a and is called the reflection time. If a valve is closed in less than the reflection time, pressure on the upstream side of the valve will rise above the stated pressure by a head H max. equal to a VJg where K0 is the

For small streams or channels with flows up to about 50 000 1/hr, it is best to use a plate with a 90° Vee notch cut in, provided that there is not an excessive amount of debris and boulders (such as may be carried along in a steep mountain stream). The formula for a sharp-edged notch is g = 0.1573 tan (0/2)//2 5, where Q = flow in litres per hour with coefficient of discharge 0.585; H is the head of water in millimetres measured 1 m upstream of the notch. For permanent use the plate and notch should be made in stainless steel or bronze, but for temporary measurements a timber board is adequate. For larger flows a wide rectangular notch is more suitable using the formula Q = 0.2084L//15, using same definitions as for the Vee-notch formula except that the coefficient of discharge is 0.62 and L is the length of weir in millimetres ignoring end contractions. Although the reading of H should be taken 1 m upstream of the weir, it can be estimated by measuring it at the notch and then adding 10mm to H if the velocity is about 0.5 m/s; or 20 mm if the velocity is 1 m/s. For still larger flows, as over long overfall weirs in rivers and overflows from reservoirs, calibration by models is desirable and in any case the flow will be dependent on the particular profile adopted. However, approximate discharges can be assessed using the formula Q = 0.0579IP5; the coefficient can be increased to the order of 0.07 for an efficient crest profile. The approximate values of discharge for 90° Vee-notch, rectangular notch and the general long overflow are summarized in Table 28.4, evaluated in the usual zones of operation.

28.4.1 Current meter The flow in a wide river is calculated by measuring the crosssectional area of the river and the velocity by a current meter at several points across the section. A current meter is an instrument provided with a propeller screw which, when immersed, is turned by the velocity of the water, and the number of turns is a measure of that velocity; the mean of all the velocities multiplied by the cross-sectional area of the water is a measure of the quantity flowing. This method is often supplemented by dilution gauging for small rivers with high turbulence. Methods of dilution gauging are defined in BS 3680:Parts 2A and 2C. Further information on flow measurement is given in Chapters 22 and 31. 28.4.2 Crump weir The Crump weir15 has a sharp horizontal crest with a 1:2 slope on the upstream side and a 1:5 slope on the downstream side. Considerable research has been undertaken into the characteris-

Table 28.4 Discharges from weirs H (mm)

90° Vee-notch 0/h)

10 25 50 75 100 130 160 200 250 400 600 800 1000 1500

50 490 2780 7650 15750 30300 50950 89000

Rectangular (long, per metre length) Overflow per metre length 0/h)

26050 73680 135350 208400 308900 421 750 589450 823 750

in m3/s 0.16

0.46 0.85 1.31 1.83 3.36

tics of this type of weir, and this is summarized in Water Resources Board Publication TN8.16 It is capable of measuring high discharges and can function in partially drowned conditions which means that the size of the structure can be kept down, thus avoiding objections by amenity and fishery interests. Compound types with side section separated by piers can be used to improve discharge of low flows, and a 'flat Vee' type with the crest sloped in at 1:10 also provides for this. Each type should be individually calibrated for reliable results. The results can be read directly on a staff gauge or recorded on a chart or punch tape for later data processing.

from the source, such as pumping, on the other. All service reservoirs must be at the highest possible level necessary to serve the houses which are to be supplied. If there is no natural ground sufficiently high for the purpose the water must be raised by being perpetually pumped (i.e. boosted), or the reservoir must be elevated and so becomes a water tower. At 1982 prices, the cost of pumping is about 2.5 p per 10001 per 100 m lift. 28.6.1 Covered service reservoirs All service reservoirs must be covered to keep the clean water which is put into them from being fouled by exposure to the atmosphere, which encourages algal growths (especially with hard water), by dirt from the air, or by vermin from the ground. Such reservoirs are usually built half in and half out of the ground, the material of excavation being used for banking material for the walls and covering the roof. About 300 mm of topsoil is usually placed over the whole expanse of roof to improve the appearance and also to keep the concrete of the roof at an even temperature to prevent expansion and cracking. Walls may be of mass, or reinforced, concrete, brick, masonry or puddled clay; columns may be of plain, or reinforced, concrete, brick, steel or masonry. Roofs may be of reinforced concrete, brick arched, or asbestos sheeting on mild steel trusses. Floors may be of plain, or reinforced, concrete, or bricks on puddled clay. Internally, reservoirs may be lined with asphalt. Different circumstances and different persons dictate the choice; mass or reinforced concrete walls, reinforced concrete roof and floors, concrete or coated steel columns being the author's usual choice. The principles of design follow orthodox practice for masonry, concrete, or steel structures. For preliminary design purposes the dimensions, capacities and costs in Table 28.5 may be used.

Table 28.5 Service reservoirs on level ground 28.5 Measurement of water in pipes For measuring the flow of water within a pipe the most useful apparatus is the Venturi meter, named by the American inventor Herschel after an Italian of the eighteenth century who experimented on the flow of water in tapered pipes. As the heads in the pipe and at the throat vary with the velocity through the pipe the quantity passing through the meter is proportional to the square root of the difference of the pressure at the inlet and at the throat. By pressure pipes, connected to the upstream side of the meter and the throat, the rate of flow is recorded either visibly as in a manometer or transferred to a pen and paper chart. This rate of flow can also be integrated instrumentally to show the total quantity. Friction losses are 300 to 600 mm for good design. A truncated version called a DaIl Tube is now generally used. On longer pipes fast electromagnetic flow meters are available with external fixings that provide no head loss and flows can easily be transmitted for remote reading.

28.6 Service reservoirs An important item in the distribution of water is the covered service reservoir, not to be confused with the open impounding or pumped storage reservoir. Whereas the function of an impounding reservoir is to store crude water for use in dry months or years, the service reservoir stores the drinking water for immediate use. The covered service reservoir is an integral part of the distribution system and its object is to balance the daily fluctuations of demand on the one hand and the method of delivery

Capacity (Ml) Dimensions depth (m) Side of square (m) or circular dia. (m)

2.25 4.5

6.75 9.0

11.25 13.5

3.7 25

4.6 32

5.5 35

6.4 38

7.3 39

8.2 41

28

36

40

42

44

46

Approx. cost (1987) (£000) 195

350

490

600

715

780

28.6.2 Water towers The cost of water towers is balanced against the cost of boosting, i.e. the capital cost of the tower and its maintenance against the capital cost of boosting plant and its maintenance. If a reduction of capital expenditure is of paramount importance, boosting may be chosen, but in many cases greater security is felt with an elevated tank. Towers are of several designs. For a given capacity the globular form, of steel, seen in the US, is the most economical in the weight of steel. The cylindrical form with 'dished' bottom is next-best. The rectangular form conveniently built up of steel plates is often used in the UK for industrial applications and in developing countries due to ease of design and fabrication. The majority of towers for waterworks purposes are of reinforced concrete, consisting of cylindrical structures, either on legs or totally enclosed; they are often lined with asphalt, and firms specializing in reinforced concrete are usually employed for their construction.

A form much favoured is the cylindrical steel tank surrounded with a thin external shell of reinforced concrete having a space of 1 or 1.3 m between the steel tank and the shell. The shell lends itself to architectural treatment to harmonize with the surroundings, can always be maintained to present a pleasing appearance, and can often be constructed by a local contractor. The steel tank inside the concrete shell can be inspected externally for leakage, and preserved by painting; this is not so easy with the reinforced concrete tank, particularly if it shows signs of leakage. It is not economically justifiable to provide standby elevated storage and the largest towers seldom exceed 5 Ml in capacity or 30 m in height. They are subject to significant wind loading, and usually require substantial foundations. As a general guide a 1 Ml storage at 30 m height would cost approximately £600 000 (at 1985 prices). 28.6.3 Valves and fittings for service reservoirs The appurtenances of a reservoir consist of suitable inlet, supply and washout valves, overflow arrangements, roof air inlets, a depth indicator and recorder, and an outlet meter and recorder. Flow through the reservoir is desirable, and to facilitate this the inlet is placed at one end and the outlet at the other. Air ventilators are provided to prevent accumulation of gases and accommodate the rise and fall of the water level. The overflow pipe is usually a standpipe with a bellmouth on top. In nine cases out of ten, and particularly where underground water is the source of supply, the service reservoir should rarely (e.g. 5 or lOyr) need cleaning and, on these occasions, as the operation need not take more than a few hours, the supply can usually be maintained by bypassing perhaps through a small tank or at night; partition walls with all the necessary duplication of pipework are only required for the larger sizes where there is no alternative storage available during the period out of service. Some form of telltale or automatic level recorder is necessary, and there are many types, to inform the operator, at the source, of the water level. Although there may be a meter at the source to regulate the quantity of water flowing into the reservoir it is often desirable to have a meter on the outlet side to ascertain the rate of draw-off hour by hour in the day.

28.7 The underground scheme

The capacity of the borehole pumps must be determined by the water requirements of the district and the capacity of the underground works. Some pumps may be left to themselves, others may require men in charge. Pumps may have to pump the whole daily supply of 24 h in 8 or 16 h to fit in with shifts, or even more during weekdays so as to close down at weekends. Hence the capacity of the pumping machinery, and any treatment works and pumping mains, if designed for 8 h pumping, must be 3 times as large as those designed for 24 h pumping, although the total daily or annual quantity pumped to supply is the same. The water, after being lifted by the well pumps to the surface, is pumped on either by the same pump or a modification thereof, or by a surface pump, with or without a balancing reservoir. Automatic pumping machinery for waterworks pumping is now generally adopted owing to its improved reliability and saving in cost. The water, where necessary, may be softened, filtered (rare for underground supplies), or treated for removal of iron and/or manganese. The pumping main, which must be designed in accordance with the pumping rate, may be of asbestos, iron, steel or PVC with various coverings, and is best left free from tappings for house services and taken direct to the covered service reservoir. The service reservoir from which the water is distributed should hold at least 24 h and preferably 3 days' supply: the distribution mains from the reservoir should be designed to carry a rate of about 3 times the average consumption to allow for peak hourly demands. 28.7.2 Geology of source17 The source of water for a pumping scheme which depends upon obtaining water from an underground source, in England particularly, rather than in Wales or in Scotland, is based on the following geological formations, in order of merit. 28.7.2.1 Chalk" This includes all three divisions, Upper, Middle, and Lower, with overlying gravels and crags, and occurs in Yorkshire, Lincolnshire and East Anglia. Chalk, which is associated with overlying porous gravels, Thanet Sand (and other Lower London tertiaries) is present in the Home Counties (Buckinghamshire, Berkshire, Surrey, Middlesex, Hertfordshire and Essex), London, Kent, Sussex, Hampshire, Isle of Wight, and with underlying Upper Greensand in Wiltshire and Dorset.

28.7.1 Development of a new source The origin of the water for an underground water scheme is rain which has fallen on the surface of the ground and has sunk in; to retrieve this water a well is dug in the fissured formation into which the rain has penetrated. If the strata are not sufficiently open to yield enough water an adit or heading may be driven in the bottom of the well. If the strata are well fissured and yield water readily, one or more boreholes are adopted. Whatever the 'hole' in the ground may be, its function is to accommodate a pump of the right dimensions for the economical pumping of water; this pump being low enough to reach the water when the water level in the well is at its lowest. Where the water level never falls or is never likely to fall more than 8m below the surface, as in areas of alluvial estuarial plain with monsoon recharge occurring every year, it is usual to use surface pumps for ease of maintenance. However in most situations the fully submersible pump is now used. To drive the pump there are the alternatives of oil and electricity; oil is seldom used except as a source of power. Whatever the pumping machinery may be, however, for waterworks practice it must be absolutely reliable, and it is desirable to have provision for standby equipment and boreholes.

28.7.2.2 Bunter and Keuper Sandstones These are present in South Lancashire, Cheshire, Yorkshire, Nottinghamshire, Staffordshire, Warwickshire and parts of Somerset and Dorset. 28.7.2.3 Oolites" Lower oolites include somewhat arenaceous deposits in East Yorkshire; the Lincolnshire Limestone of Lincolnshire; thin beds in Northamptonshire and Oxfordshire; the Inferior oolites and Cotswold Sands of the Cotswold hills of Gloucestershire and Worcestershire and the somewhat arenaceous limestones of Somerset and Dorset. Upper oolites are found in Yorkshire, Oxfordshire and Wiltshire. 28.7.2.4 Lower Greensand The chief development of the Lower Greensand is all around the foot of the Chalk beneath the Gault of the North and South Downs, which enclose the Weald of Kent and Sussex, and parts of Surrey and Hampshire. Some of the Lower Greensand

provides useful soft water at a depth of over 300 m in the Slough area. Other divisions include spreads of Greensand at the northern foot of the Chalk in Bedfordshire and Hertfordshire, and in Lincolnshire. 28.7.2.5 Carboniferous Under this general term may be included the Carboniferous Limestone which collects water in large fissures as in the Mendips of Somerset, Derbyshire and Yorkshire; the Millstone Grit and other grits with small fissures of the Yoredales and Coal Measures of Lancashire, Yorkshire and Wales; grits in the Upper Coal Measures for Coventry, and small supplies in the culm of Devonshire. 28.7.2.6 Permian and Magnesian Limestone This occurs in the north-eastern counties. 28.7.2.7 Ashdown Sand and Tunbridge Wells Sand These are present in the Kent and Sussex Wealds. 28.7.2.8 Old Red Sandstone Old Red Sandstone is present in South Wales, the Forest of Dean and Herefordshire. In addition there are water-bearing gravels in proximity to rivers or other so-called water-bearing formations in juxtaposition with many of the above water-bearing strata. The following may be taken as a rough indication of the extent and quantities of water that are pumped from the main four formations (see Table 28.6).

Table 28.6 Underground supplies in the UK

Strata

Exposed surface area (km2)

Chalk/Upper Greensand 13000 New Red Sandstone 4500 Oolites 6500 Lower Greensand 2600

Underground extent (km2)

Quantity pumped (Ml/day)

18000 3000 2600 13000

1800 400 150 120

spot and local depressions given an indication of possible overextraction. The proportion of rainfall which percolates into the underground strata is affected by surface land use and topography which dictate the amount of evaporation and surface runoff. As a general guide over porous strata, the average percolation is from 200 to 300 mm per year with an average rainfall of 600 to 900 mm per year. Lapworth has suggested that percolation in the Chalk is 0.9 of the average rainfall less 340 mm per year. The round figure of 250 mm per year is often assumed for a working basis and this is equivalent to 0.7 million I/day per square kilometre of gathering ground, the area being determined from the extent shown by the underground contours which may be assumed to flow (or be drawn) to the selected source. However, in a dry year with low rainfall and high evaporation, this could be reduced to almost nil. For an accurate assessment of the reliable yield of a particular aquifer, monthly or even daily figures should be assessed using the Penman22 or similar formula which takes into account such factors as solar radiation, the drying power of the air, wind speed, vapour pressure and temperature. Application of these formulae have been programmed to incorporate a 'root' factor for different types of land use and soil moisture deficit, which can be obtained from the Meteorological Office in the UK. With regard to natural factors affecting adversely the cost of underground water these are mainly: (1) hardness where rain acidified by vegetation dissolves calcium and magnesium from limestone; (2) chlorides, which come from contacts with rock salt as in Cheshire, or from rocks containing highly mineralized water from ancient seas or other sources and which have been protected throughout the ages by a clay layer; and (3) iron from acidified rain dissolved from the rocks, ground or peat through which it passes as in the ironstone of the Weald. The waters for new sources should be analysed chemically much more fully than those already in use, and it may be necessary, for example, to look for such a substance as fluorine which affects teeth. Man-made factors affecting the quality of water are those of proximity of buildings, sewage, and noxious effluents, e.g. gas liquors on the gathering ground, refuse tips and manure, all objectionable in varying degrees according to their distance from, and the potential fissures underground leading to, the well. The site of the well should, of course, be chosen as near as possible to the place where the water is required; every kilometre of main adds greatly to the cost, e.g. a 300mm main at a pumping rate of 5 million I/day costs £50000 per kilometre. Among other engineering considerations, the highest underground water level when pumping is significant, because for every 30m in elevation saved about 3Ip/15001 is saved.

28.7.3 Selection of source20 21 The natural factors governing the quantity of water obtainable at any one spot are: (1) the direction of flow of the water underground; (2) the general geological arrangement of the strata; (3) the area of strata exposed to the skies; (4) the thickness of formation; (5) the extent of the formation underground; (6) the nature of the fissures and formation and porosity of strata; (7) the amount of rainfall; and (8) the evaporated and dissolved salts from the strata. The factors governing the final choice of site are those enumerated above, with conditions of proximity of buildings, proximity to places to be supplied, cost of development and other engineering considerations, elevation and acquisition of site. The direction of flow of water underground can be ascertained by plotting contours of underground water levels from the records of water levels, referred to Ordnance datum, in existing wells. The arrangement of the strata such as faults and/ or rolls, anticlines, synclines, thinning out, and change in the lithology may modify profoundly the potential yield at any one

28.7.4 Construction of boreholes and adits The diameter of the boreholes for water are much larger, and the depths are much less, than for oil wells. Boreholes under 300 mm diameter are seldom made for a permanent source of public water supply in the UK. Steel lining for boreholes is standardized to BS 879 with diameters up to 1.2 m. Bored wells, however, may be made up to 3.6m diameter under water without pumping. Wells and adits when made by hand are dug in the dry, to rest water level, and by pumping below rest water level, but are very expensive and seldom justified in economic terms. Boreholes are made by rotary drilling or by percussion; drilling enables exact cores to be obtained, whereas the percussion method pounds up the material. Both methods are used according to the conditions of hardness of the strata and any special requirements as to necessity of cores for the identification of strata. In sandy formations, boreholes are sunk by the mud flush system, the mud keeping back the sand as drilling

proceeds through the soft sand; this .process enables gravel to be inserted outside the perforated tubes, which is a very satisfactory way of keeping back the sand when the site is brought into commission. A modified method of drilling is known as the 'reversed' flow system where water is pumped under pressure into the borehole while being drilled. Sites are often developed by two boreholes so that duplicate machinery may be inserted in each. Often a sandstone site may be developed by several boreholes spread over the site with a small pumping unit in each in order not to pump too large a quantity at any one spot and so draw in sand. Boreholes in the Chalk are risky, for although the site may be geologically a good one the borehole may quite easily miss water-bearing fissures. A geophysical survey of the site to ascertain where there is least electrical resistance to indicate the place most likely to be the most fissured may be useful and the use of aerial photographs and satellite images (LANDSAT) is useful in establishing fissure patterns and the presence of swallow holes. The cost of drilling and testing a lined 600mm diameter borehole over 50 m deep with a neighbouring observation hole would be approximately £400 per metre depth (1982). 28.7.5 Water-well casing The lining of a borehole may be plain for lining out clay or other non-water-bearing material. Usually the top 15 m is lined out to prevent surface contamination. The water-bearing portion of the borehole, depending on the capacity of the rocks to stand up and not collapse, is often unlined. At some boreholes, even those in the Chalk which were thought to be safe, have collapsed it is often considered prudent to line a borehole throughout its depth with perforated or slotted tubes in the water-bearing horizons and plain tubes in the Clay or unstable sections. The tubes in the water-bearing portion of the borehole may be perforated with holes at centres or by slots 3 to 12 mm wide, 150mm or more long spaced at 100 to 150mm centres. The British Standard casing (BS 879) for water wells includes lap welded and welded steel tubes. The main points of general difference are in the joints which may be screwed and socketed with: (1) V thread screwing; (2) square thread screwing; or (3) screw flush butt joints with square form thread parallel screwing. Nowadays uPVC tube and slotted lining is extensively used and is replacing the use of steel in many applications where structural strength is not required. 28.7.6 Testing boreholes and wells The testing and development of newly developed sources is often complained of as being costly, but it is extremely necessary, as it is the basis upon which any pumping scheme rests. Some newly constructed holes show stationary pumping levels almost within a few minutes of commencement of the test; whereas in others the water levels are not stabilized for some time, even after 3 days or more. The frequent stopping and starting of pumping often improves the yield of a newly drilled borehole and the rate of rise at the end of the test gives the measure of the inflow into a well. Yields may be increased in Chalk and other limestones by treating the boreholes with hydrochloric acid. Fissures, clogged by boring, are cleared and the well losses reduced. Development can also be achieved by hydraulic fracturing of the strata. 28.7.7 Yield of boreholes For comparing the yield of one borehole with another it is convenient to compare the specific yield, the quantity pumped divided by the difference of level between rest and stabilized

pumping level. Typical specific yields range from 75 million 1/h per metre (poor) to 750 milion 1/h per metre of lowering (good). Theoretically, the yield of wells varies as the logarithm of their diameters, but the mathematical theory of wells cannot be applied, except broadly, since conditions in the fissuring of strata are too variable and uncertain in practice. Geology does not lend itself to any mathematical treatment which presupposes that every cubic metre of strata over many square kilometres extent and many metres depth is of absolute uniform composition throughout. Measuring devices for the testing of wells include, of course, a flow recorder, or method for measuring the quantity, and a depth recorder for measuring the depth of water; both instruments are provided with charts or digital recorders connected directly into the logger system. Pumping tests must be continuous day and night, and the commonly accepted period for a public water scheme is 14 days. Frequently, longer periods are necessary where the water level is not stable or additional information about the aquifer is required, and this would normally include a series of 2 to 4 h step tests at different rates with equal rest periods between each test. The yield of a borehole can be assessed by plotting on a logarithmic scale and comparing with a number of theoretically produced curves prepared by Theis, Jacob and others which indicate if it is behaving as a confined, semi-confined or homogenous aquifer, and from which the sustained yield over annual or several months' pumping can be extrapolated from the base of the 7- to 14-day test.

28.8 Surface water schemes 28.8.1 Development of a new source and requirements23 A surface water scheme usually includes the construction of a reservoir to store river water at times of high flows for use in times of low flows in order to give a uniform daily rate during a design drought period consisting of consecutive drought years, with a given probability of occurrence - often taken as once in 50 or 100 years. Such a scheme basically depends on: (1) the quantity of rain falling on the gathering ground, and (2) the evaporation or loss in quantity after it falls on the gathering ground to the dam; (3) the storage of the reservoir consistent with cost and its reliable yield, and (4) the suitability of the geological conditions for safety of the submerged area and the dam site. Thus, meteorology, engineering and geology are interdependent for ensuring the safety and cost of a surface water scheme. 28.8.1.2 The main uses of reservoir conservation The main uses of reservoir conservation are: (1) (2) (3) (4)

Domestic and industrial water requirements. Hydro-electric generation. Irrigation. Regulation of the flow of a river by increasing dry weather flows and by reducing floods. (5) Recreation (fishing, sailing).

28.8.1.3 The chief types of reservoirs The chief types of reservoirs are: (1) Impounding reservoirs, so called because the sides of a valley, with the dam, impound the natural flow of the river. (2) Pumped-storage reservoirs, formed by a dam or bund

remote from the river from which they are filled only by pumping. (3) Some impounding reservoirs may also be used partly for pumped storage. (4) Both types could be used for river regulation, i.e. regulating the flow of a river to maintain abstractions lower down.

28.8.1.4 The chief types of modern dams The chief types of modern dams are: (1) Gravity concrete dams. Triangular in section where the line of pressure passes through the 'middle third' of the dam; socalled 'gravity' because any cross-section could stand without overturning. (2) Modifications of concrete dams, i.e. curved gravity, prestressed, reinforced, thick arch, thin arch, double curvature. (3) Buttress, multiple arch concrete dams. (4) Earth dams with various forms of clay cores which are supported by sands, gravels, soft sedimentary and all other soft rocks. (Various types of construction, e.g. hydraulic fill dams, can be included in this category.) (5) Rockfill dams with impervious cores, or faced with asphaltic concrete. Types (1), (2) and (3) concrete dams are usually adopted on rock foundations, while (4) and (5) earth rock dams are generally the best solution on soft alluvial or sedimentary deposits.

runoff, the storage is based on a proportion of the annual runoff. Thus if the proportion is under 75 or 80% of the average annual runoff (generally known as the flow of the three driest consecutive years, which has been determined by records of gauging for 35 yr), economic reservoirs of reasonable size are assured. For proportions greater than 80%, the size and cost of the reservoir may be doubled for only a very small increase in yield. 28.8.3.1 Storage and yield from river flow records Where the flow of the river is known over a series of years the storage necessary for the different yields can best be calculated from plotting the flows cumulatively as a mass diagram. Figure 28.3 is a typical diagram for a small scheme in which the wavy line OX represents the cumulative runoff from a gathering ground of 425 ha for 2 yr. The straight line OX represents the uniform yield of 5.5 million I/day during 22 months. The vertical distance between the two lines represents the quantity by which the total actual runoff is below the average at any time.

Storage 1000 Ml Yield 4.3 Ml/day

28.8.2 Geology of source The chief valleys in Britain where surface waters are impounded are on the following geological formations, which are relatively impermeable:17 (1) The Ordovician, Silurian and Old Red Sandstone. The grits and shares of the Ordovician and Silurian in Wales present many developed sites in Wales (e.g. Claerwen, Llandegfedd) and in the Lake District, the artificially enlarged lakes, Thirlmere, Haweswater and Crummock. (2) The Carboniferous Series, Yoredales, and other grits and shales below the Coal Measures underlie most of the sites which have been developed in Yorkshire, Lancashire and Cheshire (Scammonden, Erwood and Lamaload). (3) Other Clay formations, e.g. the Keuper Marl (Chew) and Forest Marble of Somerset (Sutton Bingham), the Lias Clay of the Midlands (Eyebrook and Draycote), the Ashdown Sand and Weald Clay of Sussex (Weir Wood and Bough Beech). (4) The granites of Dartmoor and Cornwall (e.g. Meldon, Siblyback) and the igneous and metamorphic rocks of Scotland, support the pre-stressed AlH na Lairige, the thin double curvature arch Monar and very many buttress dams for hydro-electric power. The geographical requirements are that the valley should be wide and flat for the size of the reservoir but narrow at the site of the dam, sufficiently elevated to command the town and large enough to provide an adequate yield. 28.8.3 Economics of storage and yield of reservoirs The storage of a reservoir is related to the annual runoff of the gathering ground. If the required daily quantity of water is less than the driest weather runoff and can be acquired, there should be no necessity for a reservoir. If the required quantity is more than the daily

Yield 5.5 Ml/day Storage 455 Ml

Figure 28.3 Cumulative runoff and yield

Thus the maximum deficiency during the 22 months is 100 million 1, which occurs in September 1949, hence the storage required to maintain the guaranteed yield of 5.5 million I/day is 1000 million 1. Similarly, for a smaller yield, i.e. for 4.25 million I/day, represented by the line OY the storage required would be 450 million 1, based on a period of 14 months. 28.8.3.2 Storage and yield from rainfall and evaporation records Where the runoff of a stream or river is unknown, the relationship between storage and yield may be assessed for the gathering ground of the dam in the following steps: (1) (2) (3) (4) (5) (6)

The average annual rainfall. The average annual evaporation and other losses. The average annual runoff ((I)-(2)). Various proportions of (3), known as yields. Storages corresponding with yields (4), from Table 28.7. Finally, the storage consistent with economy and site conditions nearest to giving the required yield is chosen.

Table 28.7 Lapworth yield-storage relationships

becomes sensitive to the dry weather flow (DWF) as in the first two examples above and special droughts may 'wreck' the calculations. The percentage of runoff taken for the yield and storage depends on the physical conditions of each site and on the daily requirements of each town.

Storage (cm) Average runoff from gathering ground (cm) 72.5 12.5 26.3 37.5 46.3 52.5 57.5 62.5 67.5

25 50 75 100 125 150 175 200

25

37.5

50

18.8 22.5 25.0 35.0 40.5 45.0 49.3 56.3 61.3 61.3 71.3 77.5 72.5 84.8 92.5 80.8 96.8 108.3 88.0 107.3 120.5 95.5 115.0 132.5

62.5

75

50.0 66.3 82.5 100.0 117.5 132.5 145.0

70.0 87.5 105.5 125.0 142.5 157.5

Conversions: Storage: 1 cm of water on 1 ha= 1000001 Runoff: 1 cm per annum on 1 ha = 274 I/day

The following notes may be useful in making a preliminary assessment.

Example of use of Lapworth Chart and Table 28.7. Assume a gathering ground of 1000 ha with an average annual rainfall of 175 cm. Assume average annual evaporation is 50 cm so that average annual runoff= 175 - 50= 125 cm. From Table 28.7 for average annual runoff of 125cm, consider storage of 25, 50 and 75 cm and tabulate as below. Storage related to 125 cm Corresponding yield (cm) From conversion Storage (1 per h) Therefore Ml/1000 ha Yield (Ml/day)

25 72.5

50 92.5

2.5 x 106 5 x 106 2000 5000 20.0 25.3

75 105.5 7.5 x 106 7500 28.9

26.8.4 Catchwaters

(1) The average annual rainfall. The records of the Meteorological Office should be referred to for any area in Britain. Gathering grounds in the South and Midlands, 750 to 1000mm, Wales, Lake District to Scotland 1000 to 1500mm per annum. In other parts of the world best use must be made of often sparse data. (2) The evaporation or loss of the rainfall. Penman22 has compiled a map of Britain showing the average annual losses, and similar maps can be compiled by application of the formula using local factors in other parts of the world. (3) The average annual runoff is the difference between rainfall and evaporation on the gathering ground. (4) For the yield-storage relation, reference for a first approximation should be made either to the Lapworth chart,24 from which Table 28.7 below gives extracts of the yield-storage relation for various runoffs. (The rainfall less evaporation is also known as the available yield.)

In order to augment the yield of a reservoir, catchwaters are often resorted to so that additional water is led into a reservoir from another gathering ground. As it is not economical to design catchwaters to take maximum floods, only up to 90% of the water available is taken. The proportion taken is regarded as the 'efficiency' of the catchwater and can be assessed by drawing up a flow frequency curve25 relating to the stream at the intake point if such records are available. A catchwater may be a tunnel or, more often, an open channel graded to suit the contour of the land. This is the cheapest form of structure and, provided that the length is not too great, it is cheaper than a pipe. The design of an open channel can be based on the Chezy formula.

The determination of the size of a reservoir required to balance 50% or less of the available yield is difficult as the problem

Reservoirs may be formed with earthen (or rockfill) dams, concrete dams (gravity, arch, multiple arch, cupola, buttress,

28.9 Formation of reservoirs

Table 28.8 Dam categories and design flood factors Category

Result of failure

Initial reservoir condition

Flood inflow (general standard)

Wind speed

Minimum wave surcharge (m)

A

Lives in a community endangered Extensive damage, or lives not endangered in a community Negligible risk to life. Some damage No loss of life likely. Limited damage

Spilling

PMF

1 in 10 yr Max. h

0.6

Just full

0.5 PMF or 1/10 000 yr

1 in 10 yr Max. h

0.6

Just full

0.3 PMF

Av. annual

0.4

Spilling

0.2 PMF or 150-yr flood

Av. annual Max. h

0.3

B

C D