UNIVERSITY OF CALIFORNIA COLLEGE OF AGRICULTURE AGRICULTURAL EXPERIMENT STATION BERKELEY, CALIFORNIA IRRIGATION BY SPRINKLING J. E. CHRISTIANSEN BULLETIN 670 October, 1942 UNIVERSITY OF CALIFORNIA BERKELEY, CALIFORNIA Digitized by the Internet Archive in 2012 with funding from University of California, Davis Libraries http://www.archive.org/details/irrigationbyspri670chri CONTENTS PAGE Introduction 3 The use of sprinkler systems for irrigating agricultural crops 3 General characteristics and classi- fication of sprinkler systems. 5 Stationary sprinkler systems 7 Kotating sprinkler systems 7 Arrangement and spacing of sprinklers 8 Layout of pipe lines 9 Systems with fixed sprinkler heads 10 Nozzle lines 11 Semiportable sprinkler systems .... 14 Portable sprinkler systems 15 Portable sprinkler pipe 16 Arrangement of portable sprinkler lines and methods of opera- tion 17 Use of field ditches 17 Use of pressure supply line ... 20 Drag-type sprinkler systems .". 21 Sprinklers for portable systems . . 22 Portable pumping plants 23 Pumps and accessories 24 Field ditches 26 Systems using stationary pumps and pressue supply lines. ... 26 Stationary pumping plants ... 26 Pressure supply lines 27 Hydrants 30 Low-pressure systems for field crops 30 Perforated pipe systems 31 Traveling sprinkler machines ... 32 Under-tree sprinkler systems for orchards 33 Sprinklers for under-tree sys- tems 33 Portable sprinkler pipe for under-tree systems 35 Under-tree systems using sprin- klers attached to hose 37 Portable drag -type sprinkler systems for orchards 38 Sprinkling compared with other methods of irrigation 41 Advantages of sprinkling .... 41 Conditions favorable to sprin- kling ' 43 Limitations of sprinkling 44 Cost of sprinkling 45 Depreciation and interest on in- vestment 46 Cost of operation 46 Explanation of table 2 50 PAGE Hydraulics of sprinkler systems ... 51 Discharge from nozzles 52 Coefficients of discharge of noz- zles and sprinklers 52 Discharge of small nozzles for nozzle lines 54 Pressure required for sprinkler operation 55 Flow of water in pipe 56 Formulas for friction loss in pipe lines 57 Friction loss in standard pipe. . 60 Friction loss in welded steel pipe 61 Friction loss in copper tubing. . 62 Friction loss in garden hose and hydrants 63 Flow of water in pipes with mul- tiple outlets 64 Derivation of equations for cal- culating friction losses 64 Friction loss in sprinkler lines . . 67 Friction loss in nozzle lines ... 70 Discharge, pressure, and power requirement relations for sprinkler lines 71 Distribution of water from rotating sprinklers 75 Sprinkler tests at Davis 77 Typical distribution patterns for favorable conditions .... 78 Effect of low pressure on dis- tribution from sprinklers ... 80 Effect of wind on distribution patterns 84 Effect of high speed of rotation on distribution patterns .... 86 Effect of variation in rate of rotation on distribution pat- terns 89 Desirable types of distribution patterns and proper spacing of sprinklers 93 A method of analyzing sprinkler tests for uniformity of dis- tribution 94 Distribution for geometrical patterns 95 Desirable patterns for square and equilateral-triangle ar- rangements of sprinklers ... 97 Uniformity of distribution and effect of spacing on actual sprinkler patterns 103 CONTENTS — Continued PAGE PAGE Evaporation losses 310 Interception and subsequent Evaporation from the spray 110 evaporation of water from Direct measurement of evapora- plants 115 tion losses 110 spray 112 Evaporation losses from wet sur- faces 114 Design and operation of sprinkler Indirect method of estimating systems 116 ?7!?_? rati0n l0SS fr ° m the , -, n Methods of" operation ........ 116 Capacity of sprinkler systems. . 117 Eate of application 119 Evaporation losses' from* moist " De P th of application 119 so il s m 214 Sprinkler capacity 122 Eate of evaporation from free Size of pipe 123 water surfaces 115 Acknowledgments 124 LIST OF TABLES PAGE -Sprinkler discharge required for various applications of water in a twelve-hour period 33 -Eesults of field study to determine cost of operating portable sprin- kler systems in Sacramento Valley 48 -Theoretical discharge of sprinkler nozzles 53 -Discharge of small nozzles for use on nozzle lines 55 -Friction loss in standard wrought -iron or steel pipe, for C = 100 ... 58 -Friction loss in standard wrought-iron or steel pipe, for C = 120. ... 59 -Friction loss in type M copper tubing 62 -Comparison of inside diameters and friction losses in copper tubing, types M, L, and K 63 -Approximate friction loss in garden hose and garden hydrants .... 64 -Values of the factor F by which the friction loss in pipe must be multiplied to obtain the actual loss in a line with multiple outlets 66 -Uniformity coefficients for various arrangements and spacings of geometrical patterns B and E 97 -Uniformity coefficients for patterns G, I, and K 99 -Uniformity coefficients for patterns M, O, Q, and S 101 -Summary of sprinkler tests giving pertinent data together with the calculated uniformity coefficients for different spacings between sprinkler lines 104 -Summary of data in table 14 giving average uniformity coefficients for several tests on the same sprinkler under similar conditions. . . . 108 -Uniformity coefficients for actual sprinkler patterns 109 Table 1, Table 2, Table 3, Table 4, Table 5. Table 6, Table 7, Table 8. Table 9, Table 10, Table 11. Table 12. Table 13. Table 14. Table 15. Table 16. IRRIGATION BY SPRINKLING J. E. CHRISTIANSEN 4 INTRODUCTION This bulletin will discuss the application of irrigation water by sprin- kling and will present the results of several years' research, together with general information. Although intended especially for farmers who are now operating sprinkler systems or contemplating irrigation by this method, the bulletin includes some technical material essential to an economical design of sprinkler systems — material of interest primarily to engineers, irrigation contractors, and others engaged in manufactur- ing, selling, and installing sprinkler equipment. The more technical aspects of the subject, and most of the experimental results, appear in the latter part of the bulletin. A closing section presents in nontechnical form a discussion on design and operation of sprinkler systems. THE USE OF SPRINKLER SYSTEMS FOR IRRIGATING AGRICULTURAL CROPS Sprinkling as a method of irrigation has been practiced in California and elsewhere for about forty years. Before 1920 it was limited primarily to truck crops, nurseries, and small fruits, and was practiced mainly as supplemental irrigation in the more humid regions. Stationary overhead sprinkler systems were first used in citrus orchards in some sections of California about 1920. Most of such systems were installed, however, between 1924 and 1928 in areas where surface irrigation was not entirely satisfactory because of unusually pervious soil or because of topographic features. Along the foothills east of Pasadena, the soil is gravelly in many places and the slopes are rather steep for furrow irrigation, except where orchards are laid out on contour grades; overhead sprinkling of orchards had its inception in these areas. Many of these systems were installed because of the extravagant claims made by those interested in manufacturing or selling the equipment. Sprinkling was supposed to save large quantities of water, to control insects, frosts, and diseases, and to increase production. Some of these early attempts proved unsatisfactory because of poor equipment and 1 Received for publication September 26, 1941. 2 This bulletin supersedes Extension Circular 4, Irrigation oy Overhead Sprinkling, by H. A. Wadsworth, published in 1926. 3 Some financial assistance in the preparation of this bulletin was furnished by the California Committee on the Eelation of Electricity to Agriculture. 4 Assistant Irrigation Engineer in the Experiment Station; resigned April 1, 1942. [3] 4 University of California — Experiment Station too wide spacing of sprinklers. After trial, a few of the systems were abandoned, and furrow irrigation was resumed. As was gradually recog- nized, sprinkling is not a cure-all; good results can be obtained only when systems are properly designed and installed, with the right equip- ment. Most of these original overhead systems represented investments of $300 to $500 per acre or more. Since the drop in citrus prices about 1930, relatively few stationary sprinkler systems have been installed. In 1932 the farm advisor's office in Los Angeles County made a sur- vey 6 of overhead irrigation of orchards in that area. One hundred grow- ers, representing 1,504 acres, were questioned. The majority were, apparently, well satisfied with their overhead systems. They had ex- perienced very little trouble with the equipment and believed that the change to sprinkling had increased production. Some had had trouble with sprinklers stopping ; others with packing of their soil, runoff, and the like. Only two reported abandoning the system. The survey also indicated that the use of water was practically the same as for furrow irrigation. Lightweight steel pipe with quick-couplings for portable sprinkler systems was first used about 1930. Its introduction completely changed the picture of sprinkling as a method of irrigation. Whereas with sta- tionary systems, the main consideration was the initial outlay, with port- able systems for field crops the equipment costs sometimes as little as $10 to $15 per acre and usually under $50. The expense of operation be- came the important item ; this, annually, sometimes exceeded the invest- ment in equipment. These portable sprinkler systems were better adapted to, and were used more on field and truck crops than in orchards. Ac- cording to a survey made in 1936, a greater acreage of sugar beets was being sprinkled than any other crop. Sprinklers were being used on nearly all field and truck crops as well as in orchards. Most of these systems are used in areas not adapted to surface irriga- tion. Near Clarksburg on the Sacramento River south of Sacramento, subirrigation had been practiced almost exclusively. In 1931, probably because a series of dry winters resulted in insufficient natural leaching, alkali spots were beginning to appear in many places ; they were espe- cially noticeable in fields planted to beans. To counteract the upward movement of the soluble salts, portable sprinkler systems were tried. Within a few years more than a hundred systems were in use in the Sacramento Delta, farther north in Sutter Basin, and near Meridian. Because of the high water table in these areas, the irrigation requirement was low compared with that in other places of similar climate. Sprinkler 5 Salter, A. G. Orchard overhead irrigation. Pacific Eural Press 131(11) : 346-47. Mar. 14, 1936. Bul. 670] Irrigation by Sprinkling 5 systems could therefore cover relatively larger areas at lower per-acre costs. Since surface irrigation was not feasible in many places, compari- sons were most frequently made between sprinkling and subirrigation; in general, results were better with sprinkling. Sprinkling has also become very popular for irrigating certain truck crops near Moss Landing in the Salinas Valley, where both the type of soil and topography have been reasons for changing from surface meth- ods. Another factor that makes sprinkling economical in the coastal areas is the lower water requirement due to low temperatures and high humidi- ties. Portable sprinkler systems have also been installed in many scattered areas for irrigating pastures and field crops. They have been used rather extensively in Oregon, especially for pasture irrigation. Many were in- stalled in areas where previously irrigation had not been practiced. About the same time that portable sprinkler pipe was introduced, under-tree orchard sprinkler systems also came into use. These systems consist of small sprinklers, mounted on short risers, which cover the areas between adjacent trees. Because stationary equipment of this type interferes with cultivation and with other orchard practices, most of these systems are portable. The simplest type of portable unit consists of several small sprinklers on low stands joined with %-inch garden hose. This portable unit is supplied from small garden hydrants on buried supply lines. Later a drag-type sprinkler system using lightweight steel tubing joined with special couplings was developed. The introduction of portable sprinkler pipe in small sizes led to the use of this pipe for under-tree orchard systems. The last few years have seen rapid advances in the development of under-tree equipment. Orchard sprinkling with such systems is practiced mainly in San Diego County, in the La Habra Heights section of Los Angeles County, and in some parts of Santa Barbara and Ventura counties. GENERAL CHARACTERISTICS AND CLASSIFICATION OF SPRINKLER SYSTEMS All sprinkler systems have certain features in common. They consist essentially of pipes with sprinklers or nozzles for distributing the water over the area to be irrigated. To accomplish this, the water must be dis- charged at a high velocity, which is obtained by operating the system under pressure. This pressure may be provided by gravity from a water source at a higher elevation or by pumping. Irrigation by sprinkling differs from other methods in that the water is distributed over the soil by mechanical means, whereas for other methods of irrigation the water is finally distributed by the soil itself. 6 University of California — Experiment Station The uniformity of application by sprinkling depends primarily upon the ability of the system to apply equal amounts to all parts of the area. For surface irrigation the uniformity depends upon the surface condi- tion, and upon uniform permeability of the soil as well as the ability of the system to distribute the water uniformly to local areas. Although sprinkling is frequently called an imitation of rain, it differs from rain in many respects. The distribution is never so uniform as with rain. The rate of application and amount applied are largely under control, a feature that overcomes some of the detrimental effects of rainfall. The effect on plants may be different, especially in connection with the spread of disease. Since sprinkling is commonly carried on during warm, sunny days it has little effect on the humidity except during the application, whereas rain is generally associated with cloudy weather and high hu- midity lasting a considerable time. Despite certain common characteristics, sprinkler systems have many differences. They may be classified according to their portability, or ac- cording to certain mechanical features. They are called "portable'* when most of the mechanical equipment can be readily moved from place to place over the area irrigated; "semiportable" when only a minor part of the equipment is moved, such as the risers and sprinkler heads ; and "stationary" or "permanent" when all the equipment is fixed. Obviously, definite lines cannot always be drawn between these three groups, espe- cially between portable and semiportable systems. This classification overlaps completely on any grouping according to mechanical features. Sprinkling is frequently called "overhead irrigation." This is some- what confusing because orchard systems are often classified as "over- head" or "under-tree." The overhead systems distribute the water over the tops of the trees, whereas under-tree systems distribute it near the ground between and under the overhanging branches of the trees. Under- tree sprinklers are sometimes called "low-head," or "ground" sprinklers. Either of these systems may be portable or stationary. Because stationary under-tree sprinklers interfere with cultural practices in an orchard and are also more expensive, most under-tree systems are portable. Al- though many overhead systems are portable, the high risers make them somewhat difficult to move. The pipe lines of stationary overhead systems are generally placed underground and the sprinklers directly over the trees, completely out of the way of other orchard operations. Sprinkler systems may also be classified according to method of dis- tributing the water, whether by rotating sprinklers, by fixed heads, or by nozzles along a pipe line which is rotated back and forth through an angle of about 90 degrees to cover a strip on each side of the line. This latter type is called a nozzle line. Bul. 670] Irrigation by Sprinkling 7 STATIONARY SPRINKLER SYSTEMS Stationary systems are used to a considerable extent in California. Most of the original orchard systems were of this type, with rotating sprinklers mounted on high risers over the trees. Other kinds of sta- tionary systems include those with fixed sprinkler heads, used exten- sively for lawns and ornamental plantings and to a limited extent for orchards ; and nozzle lines used principally for truck crops, nurseries, and other special purposes. ROTATING SPRINKLER SYSTEMS Rotating sprinklers have capacities of 1 to more than 100 gallons per minute. Some of the larger sprinklers are designed for pressures of 60 to 100 pounds per square inch, and will cover circles up to approximately 200 feet in diameter, whereas some of the small ones will operate on pressures of 10 pounds per square inch or less. In general, the larger the nozzles, the higher the pressure required for best performance. Some very large sprinklers, with capacities of several hundred gallons per minute, have been used for irrigating bananas in Central and South America, and for other crops in certain European countries. Because of the extremely high pressures required for their operation, these large sprinklers are not considered economical in California. For supplemental irrigation, where the annual water requirement is low, high pressures and large sprinklers may be more economical than small sprinklers be- cause of the greater permissible spacing of pipe lines. Rotating sprinklers are of two types : whirling sprinklers, that rotate rapidly, and slow-revolving sprinklers, that rotate slowly. Slow-revolv- ing sprinklers ordinarily rotate at a speed of one or two revolutions per minute ; some much slower. The slower a sprinkler rotates, the larger the area it will cover. When operating under pressures of about 40 pounds per square inch, slow-revolving sprinklers will cover areas up to about 120 feet in diameter. Whirling sprinklers of similar capacities, operat- ing under equal pressures, will cover areas 50 to 70 feet in diameter. The slow-revolving sprinklers are therefore more desirable, since they permit a greater spacing of pipe lines. Slow rotation minimizes wear, and thereby prolongs the life of the sprinkler. Various types of driving mechanisms are used to produce a slow rotation and prevent the sprinklers from stopping. Sometimes slow- revolving sprinklers vary greatly in their speed and this variation may appreciably affect the distribution of water. Whirling sprinklers tend to rotate at a more constant rate because of the momentum of the re- volving parts. They are generally less expensive than slow-revolving 8 University of California — Experiment Station sprinklers. Figure 1 shows a few of the different types of rotating sprinklers. Arrangement and Spacing of Sprinklers.— Though absolute uniform- ity is not essential, large variations in depth of application are not desirable, particularly with stationary systems where the cumulative effect of either excessive or inadequate amounts becomes serious. The distribution of water from sprinklers depends upon several factors. Careful study, however, indicates that approximately uniform applica- Fig. 1. — Different types of rotating sprinklers. Types with external driving mechanisms: A, spinning wheel type; B, D, and J, oscillating spring types; C and E, pendulum types. Keaction-drive sprinklers: F, E, and I, whirling sprinklers ; G and K have vibrators for slow rotation. Internal-drive sprinklers : L, friction-drive type, and M, gear-drive type. tion is possible with sprinklers covering circular areas. The application can be uniform only when (1) the type of pattern produced is correct for the arrangement of sprinklers, (2) the sprinklers are correctly spaced, (3) the sprinklers rotate at a uniform rate, and (4) there is no appreciable wind. For stationary installations, sprinklers are arranged in squares, tri- angles, or rectangles with a greater spacing between the pipe lines than between sprinklers on the line. For stationary orchard systems, the sprinklers are usually mounted over certain trees. With trees planted on a square, an equilateral-triangle arrangement of sprinklers is not prac- ticable. All sprinkler spacings must be multiples of tree spacings, which generally vary from 18 to 24 feet. The maximum practical spacing for medium-sized slow-revolving sprinklers operating under adequate pres- Bul. 670] Irrigation by Sprinkling sure is a sprinkler every fourth tree, in every fourth row. For this spac- ing, a triangular or staggered arrangement will probably give better distribution than a square. For most sprinklers, a spacing of three tree rows in both directions would be better, in which case staggering is not Sprinklers o Mafn ^JL Lateral o— a. - Square b. -Spacing between fines twice spacing a fang iine -&ww ^r c- - Eqaiiaterai fr/'ang/e . J - / > < p / ' f T? < 1 7 ^ ^ Mafn La £ era/-} ^ i < S < » < » d- - Eaa/faferaf frfang/e Three raws of sprinft/ers on ane fateraf Fig. 2. — Some of the various arrangements of sprinklers and pipe lines for sprinkler systems. The letters S t and S 2 denote spac- ing between sprinklers, and spacing between pipe lines, respec- tively. practicable. Other possibilities include a sprinkler in every other tree in every third or every fourth row with the sprinklers staggered, or ar- ranged in a rectangle. According to analyses of both actual and geo- metrical patterns (discussed in a later section) with fairly close spacings along the line there is little to be gained from staggering. Square and rectangular arrangements of sprinklers generally permit better dis- tribution around the borders of the field than triangular or staggered arrangements. 10 University of California — Experiment Station Layout of Pipe Lines. — The usual arrangement of pipe lines for sta- tionary sprinkler systems consists of a central main with parallel laterals at right angles to the main line, carrying the sprinklers. Each lateral is provided with a shutoff valve so that it can be operated independently. Since stationary systems are generally confined to relatively small areas such as 10-acre tracts, these laterals will seldom carry more than six or eight sprinklers. Larger areas are sometimes divided into smaller sec- tions treated separately. In all cases a main supply line through the center of the area is most desirable. Where the water comes from a well, the cheapest arrangement is to place the well approximately in the center of the tract to be irrigated ; this minimizes the pipe sizes required. Figure 2 shows various arrangements of sprinklers and pipe lines. Fig. 3. — Different types of fixed heads. The center three are lawn heads and the two end ones are commonly called shrubbery heads. Both are used for under-tree orchard systems. For an equilateral-triangle arrangement of sprinklers, about 9 per cent in length of pipe required will be saved if three rows of sprinklers are supplied from one lateral (arrangement d compared with c, fig. 2) . A similar arrangement can be used for sprinklers placed in a square. Al- though, in this case, there is no saving in total length of pipe required, and the main laterals have to be increased in size to carry three times as much water, the cost may be less, since the sublaterals can be smaller. For example, a 3-inch main lateral with 1-inch sublaterals might replace three 2-inch lines. SYSTEMS WITH FIXED SPRINKLER HEADS Fixed heads are used mainly for lawns and for portable under-tree systems for orchards. These heads have no moving parts; they cover areas 15 to 25 feet in diameter with a fine spray fairly evenly distributed. Although most fixed heads cover circular areas, some are designed to cover square areas. Figure 3 shows a few fixed heads of different types. The use of fixed heads on field and orchard systems is limited. For stationary systems, the cost is excessive, because of the close spacing required, and furthermore, the closely spaced heads are objectionable from the standpoint of cultivation. The minimum rate at which water Bul. 670] Irrigation by Sprinkling 11 can be applied with these sprinklers is often too high; the soil cannot absorb it rapidly enough to prevent accumulation of water in low places or runoff. Normal rates of application with these heads range from about 0.7 inch to 3 inches per hour, whereas with rotating sprinklers rates as low as 0.1 inch per hour are possible, and rates of 0.2 to 0.5 inch per hour are usual. Many bare soils do not absorb water rapidly enough to permit the use of fixed heads, except when only light applications are made. sWozz/e3 ^^^^nsiirp vo/ye H \\v \YI V "Potter bearing >\:-° rT ~——r~—-7^T~- h\ Fig. 4. — Detail of typical nozzle line assembly for hand rotation. The special fit- tings required can be secured from dealers handling sprinkler equipment. (From Ext. Cir. 4.) On the other hand, simplicity, absence of moving parts, and low cost make fixed sprinkler heads highly desirable where conditions permit their use, such as for portable under-tree orchard systems on permeable soils. NOZZLE LINES Nozzle lines consist essentially of parallel lines of pipe (% to 1% inches in size) equipped with small brass nozzles, usually spaced 2 to 4 feet apart (fig. 4). Being relatively expensive, they are generally used only for crops yielding a high gross return, such as certain truck crops and small fruits, and for nurseries, greenhouses, lawns, and other special purposes. They are occasionally used in orchards, especially on terraced plantings. 12 University of California — Experiment Station The pipe is generally supported 4 to 7 feet above the ground on posts spaced about 15 feet apart. Where it is not necessary to provide passage underneath for cross-cultivation, lines about 4 feet above the ground permit convenient access to the nozzles for servicing. Occasionally nozzle lines are supported by a suspension cable from much higher poles 100 to 200 feet apart, an arrangement that lessens the obstruction to cultiva- tion. To cover a strip of ground on both sides of the lines they are oscil- lated through an angle of about 90 degrees, generally by means of a water-operated oscillating motor. Figure 5 illustrates this type of system. Fig. 5. — Nozzle line used for irrigating strawberries. This line is rotated back and forth by a hydraulic oscillating motor at the head end of the line. (Six nozzles, of a special type, shown in the foreground, were being tested for distribution of water; this accounts for the difference in appearance of the jets.) A complete line of equipment is available for nozzle lines, including a variety of nozzles, ranging from those that throw a round jet for maxi- mum coverage, to others that have deflectors to break up the jet for narrower strips. Some nozzles, having orifices of triangular shape, are designed to distribute the water over a considerable area without oscil- lating the line. Special roller saddles on top of the posts carry the pipe and reduce to a minimum the friction in rotating the line. These fittings are available for either wood or pipe posts. Special hand-turning unions are also available. For best performance, however, nozzle lines should be equipped with oscillators that slowly rotate the line, resulting in a fairly uniform distribution of water when the lines are correctly spaced. These oscillators are usually double-acting piston devices that operate Bul. 670] Irrigation by Sprinkling 13 from the water pressure. Being fairly expensive, they are sometimes moved from one line to another, one oscillator being used for several lines. Sometimes one hydraulic oscillator is used to rotate a number of parallel lines by means of cables attached to arms on each line. For good results the mechanics of such systems must be carefully worked out. Because of the very small diameter of the nozzles used on these lines they are easily clogged, especially when the water is from an open source where algae might grow. Special screens are available for such condi- tions. One make of equipment utilizes small individual screens on the nozzles; but since these project into the pipe and increase the frictional resistance, they are less desirable than properly designed screens at the head of the line. Screens should have ample area in order to cause only a small resistance to flow even when much of the total area is clogged. They must also be designed to facilitate cleaning. If the source of water is free from anything that would cause clogging, screens are unnecessary. Flushing valves are used at the end of the line for flushing dirt and pipe scale from the line. Such material always tends to settle out and to clog nozzles only near the end of the line. Special union couplings with squared sockets are available to facilitate the lining up of the nozzles. They are especially important on portable lines that are moved from one set of supporting posts to another. Since the discharge of all nozzles varies with the pressure, the only way to secure a uniform discharge along the line is to limit the pressure loss to a small part of the normal pressure on the line. Frequently the friction loss is appreciable ; the pressure along most of the line may be very different from that at the head. Pressure-gauge connections with shutoff cocks at each end of the line are desirable, for they permit the pressures to be readily determined by means of a portable pressure gauge. A permanently mounted pressure gauge is sometimes used at the head end, but seldom are the pressures determined at other points. Fric- tion loss in pipe lines is discussed in detail later. Nozzle lines are usually operated at pressures of 25 to 40 pounds per square inch. The width of strip effectively covered increases with in- crease in pressure up to about 40 pounds per square inch. Above this there is little or no increase in width because of a greater dispersion of the jets. Where the pressure is adequate, the lines are spaced about 50 feet apart. Sufficient pressure should be provided to overcome friction losses and to leave sufficient pressure for effective operation. Nozzles of the type commonly used on nozzle lines have capacities ranging from about 0.15 to 0.30 gallon per minute at a pressure of 30 pounds per square inch. Special nozzles, occasionally used, have capaci- ties up to 0.70 gallon per minute. Friction loss in nozzle lines, being at 14 University of California — Experiment Station any point proportional to the square of the velocity, will vary appreci- ably with different combinations of nozzles and pipe sizes, and this must be kept in mind when designing sprinkler systems. Parta/>/e riser Va/re. A/ot reaa/red w^ea /at era/ /e snat off for n?ov//?a sprinA/ers l/l/fre t/e Post to ao/d riser p/{//n6 Stationary riser. Scmet/mes a size /aryer /ban por/nt/e riser /o increase sriffness and decrease fric//ona/ /osses. Ground ^ ' /at era/. Son?et/mes /eft on /op of yroaad and s/a/fonary riser on?// ted Fig. 6. — Details of riser for semiportable sprinkler sys- tem using portable sprinklers on a stationary pipe distri- bution system. SEMIPORTABLE SPRINKLER SYSTEMS Some sprinkler systems have stationary pipe lines and portable sprin- klers that are moved from place to place. They are especially desirable in orchards; they are less expensive than stationary systems and they obviate certain practical difficulties encountered with portable systems, such as holding high risers in vertical positions and in moving them from place to place. Figure 6 illustrates a common arrangement. A short stationary riser with valve extends 4 or 5 feet above the ground. The upper part of the riser and sprinkler is portable, and only the number Bul. 670] Irrigation by Sprinkling 15 operated at one time are required; thus a considerable saving in the cost of sprinklers is effected. For example, 6 or 8 sprinklers may suffice for a 10-acre tract, whereas a stationary system would require more than 100. In addition, there is an appreciable saving in the cost of pipe lines because smaller sizes may be used when each sprinkler is operated on a different lateral. For example l^-inch laterals may suffice, whereas 2-inch (or larger) laterals may be required where several sprinklers are operated simultaneously on one lateral. A somewhat similar type of semiportable system is extensively used for parks, golf courses, athletic fields, and other large lawns. A sta- tionary-pipe distribution system is installed with an automatic shutoff valve at each sprinkler location. Sprinklers are attached to these valves by simply inserting a special coupler and giving it a turn. These valves are sold under such trade names as "super valves," "snap valves," and "lawn valves." Such systems are also suitable for certain agricultural crops, though the cost of the automatic valves makes them rather expen- sive. When overhead sprinkling first came into use, a popular type of semiportable system utilized sprinklers mounted on a high riser on a portable stand and attached to a hydrant with garden hose. This ar- rangement permitted relatively wide spacing of laterals and hydrants, since several sprinkler locations could be served by one hydrant. The locations of the sprinklers can be varied for different irrigations to effect a more uniform seasonal application for all places in the orchard. The arrangement is not, however, entirely satisfactory because of the difficulty of moving the sprinklers. Another type of system used in orchards that might be classed as semiportable consists of stationary pipe lines supplying water to port- able under-tree sprinklers. These are discussed more fully in the section on "Under-Tree Sprinkler Systems for Orchards." PORTABLE SPRINKLER SYSTEMS The type generally known as a portable sprinkler system originated about 1930. Essentially it consists of a sprinkler line of special light- weight portable pipe with quick-couplings, together with a pumping plant. Sprinkler pipe comes in standard lengths of 20 feet ; other lengths are furnished on special order. Sprinklers are usually spaced 20, 30, or 40 feet apart. The pipe is moved across the field by carrying one length at a time. Several makes of systems are available, each with a different kind of coupling, some of which are shown in figure 7. These systems are used largely for irrigating field and truck crops, such as sugar beets, peas, beans, and onions. 16 University of California — Experiment Station Many systems operate with a portable pumping plant supplied from a field ditch along one side or through the center of the field. The entire system can be moved from one field to another. Where the water supply is obtained from wells or where, because of the topography, ditches through the field are not feasible, a stationary pumping plant and pres- sure supply line are used. PORTABLE SPRINKLER TYPE When these systems first came into use only 4-inch pipe was avail- able. Now some of the makes can be obtained in sizes of 1%, 2, 3, 4, 5, and Fig. 7. — Some of the different types of portable sprinkler-pipe couplings: A, Shur-Rane, ball-and-socket type; B, Montague; C, Calco (3-inch) ; D, Eain Storm; E, Pierce; F, Calco (4-inch); G, Wilson; H, Shur-Rane; I, Calco (4-inch, coupled). 6 inches and in any desired length. For field-crop systems, lengths of 30 and 40 feet are sometimes used. For orchard systems, pipe of the same length as the tree spacing is desirable. The longer lengths are slightly cheaper because they require fewer couplings. Most of the portable sprinkler pipe is made from electric-welded steel tubing, 16-gauge being used for 4-inch and larger pipe, and 18-gauge for 3-inch and smaller sizes. Lighter-weight sprinkler pipe, made from 22- and 24-gauge metal, has come into use recently. This is somewhat lighter to handle and is strong enough to withstand the pressure re- quired; but it is less durable than the heavier pipe. Some of the couplings (fig. 7) are attached permanently to the pipe, whereas others are separate Bul. 670] Irrigation by Sprinkling 17 and simply connect two lengths. In some, the sprinkler outlets are in the couplings; in others, the outlet is welded to the pipe a short distance from the coupling. The pipe is usually galvanized after fabrication. 1 * ♦ I V * r'-l I •11 ^ I J' It Sprin/c/er PorfaS/e spr//7k/er //he Porta6/e pc/mping p/a/r£ •D/'ic/i Fig. 8. — Single-line arrangement of portable sprinkler system supplied from a ditch along one side of the field. rv? ? ^ -«~- Spr/nA/er 1! -IV- £* \*-P{/fl7pf/7g p/tfflt D/fc/7 I i t i ♦ t \ \ i i A A 5pr//?kfer /f/?e Fig. 9. — Split-line arrangement of portable sprinkler system sup- plied from a ditch running through the center of the field. This results in a more economical operation than with the single-line arrangement shown in figure 8. ARRANGEMENT OF PORTABLE SPRINKLER LINES AND METHODS OF OPERATION Use of Field Ditches. — There are several possible arrangements of portable sprinkler lines. The most common are the single-line and split- line arrangements (figs. 8 and 9). The term "split line" is a misnomer, 18 University of California — Experiment Station but has become more or less established by usage. With the single line the supply ditch is located along one side of the field, and the line of portable pipe extends across the field. This arrangement is practical for pipe lines not exceeding 1,000 feet in length. It is generally used where an existing ditch, from which the water can be pumped, is located along the field border. An example appears in figure 10. The sprinklers are operated until the desired application is made ; then the pumping plant is shut down while the pipe and pumping plant are moved to the next position. The pipe is usually moved by two men carrying one length at a time. Under ordinary conditions, pipe can be moved at an average Fig. 10. — Portable sprinkler system irrigating beans in Sutter Basin; a single line of 4-inch pipe with sprinklers 20 feet apart. The portable pumping plant is at the far end of the line. rate of about 30 lineal feet of pipe per minute. From 5 to 10 minutes are required to move the pumping plant and start operation. Where only light applications are needed and lines are operated in each position for a relatively short period — two hours or less — a large part of the total time is lost in moving. Another disadvantage of the single-line arrange- ment is the high friction loss which is about eight times as great as with split lines of the same size and total length and same capacity. Single lines often necessitate the use of larger pipe, or smaller sprinklers, than split lines. The split-line arrangement is the most satisfactory for portable sprin- kler systems using portable pumping plants. It permits the use of mini- mum pipe sizes and allows for nearly continuous operation of the pumping plant. The pipe line on one side of the pump is shut down and moved to the next position while the line on the other side continues to operate. The pump is then shut down, moved to the next position, con- nected to the line that has been moved, and started up. The other line is then moved, connected to the pump, and turned on. With the same Bul. 670] Irrigation by Sprinkling 19 throttle setting, and with only one of the two lines in operation, the pres- sure is increased by about 50 per cent, and the discharge through the one line is about 60 per cent of the two lines. The plant is therefore in effective operation most of the time. The actual time lost by moving is slightly less than the time required to move one of the lines. The operat- ing efficiency is therefore considerably higher than for the single-line arrangement; when feasible, the split-line arrangement should be used. Sometimes two parallel lines are operated on the same side of the supply ditch as shown in figure 11. When it is time to move, the front B 'A \B' W 1 t i 2 k 2 k i M > \ \ y SprMer & i * k |* k k * t" £$ 1 i l I i > ^ k 4 \ h- ^S z -+- S 2 -*- S 2 -+- Sj,-j- 5/-i V > t t I ^4 «§t 4 * \ — ru, 1 l — £. — . Pan?/. 4f vng p/a/it < Nc Supp/y f//7e ite. Wften two /atero/s ore 1 k 1 \ 1 k 1 -V- 1 1 1 1 h i ♦ i i r- ♦ 1 ♦ 1 i - 9" 1 1 >j 1 i! 1 usee/, jtort o/?e ot opposite ere" of fte/d fro/7? ttie otter. 1 i> l A i a 1 ^i Fig. 13. — Arrangement of a portable sprinkler system operating from a pressure supply line with a stationary pumping plant supplied from a well. while both laterals are moved. Although portable supply lines may offer a saving in pipe cost, they are less convenient than stationary supply lines. Drag-Type Sprinkler Systems. — Since the main cost of operation of portable sprinkler systems is for labor for moving portable sprinkler pipe, some interest has been shown in drag-type systems, especially for orchards, in which the portable pipe is moved by dragging endwise. The power for moving the pipe may be supplied by a small tractor, a team, or one horse. Small units are sometimes moved by hand. When one or two lengths of pipe are used ahead of the first sprinkler on the line, it is not necessary to work on wet ground to move the pipe. Figure 14 illustrates a completely portable drag-type system with a portable pumping plant supplied from a ditch through the center of the field. With such systems the portable lateral is taken apart and moved forward with the drag unit. A wheel cart at the head end of the unit is 22 University of California — Experiment Station used both for holding the risers in a vertical position and for carrying the lateral forward. A more satisfactory arrangement of the drag-type system (fig. 15) illustrates the use of a stationary pressure main and laterals with portable drag units. Relatively small pipe can be used for the laterals when the drag units are distributed so that only one or two are served by each lateral. Drag-type systems for orchards are further discussed later. SPRINKLERS FOR PORTABLE SYSTEMS Slow-revolving sprinklers are most satisfactory for portable systems. Because of continuous operation, sprinklers on portable systems receive C x <*'Por£ab/e /at era/ 3 .\ „„ Mov/'ra a/7/'t ar-d ^.^. <> . <> . <> . . <> .< 8 3_l =1 ^ /atera/ A' I A" I ■0---0---0-0--- °" -0---0— jo--- 0-O—0---0- --»--■<>-■ ■©—■! r /)/ f /~h Porta6/e pu/np/ng p/ant t/n/ts /noi/ea' forward w/ta tractor or tea/77. Aa/'aceflt sect tons of /atera/ moi/ea w/tft dray aa/t. Fig. 14. — Arrangement of a drag-type sprinkler system with a portable pumping plant supplied from a ditch. much harder service than those on stationary systems. Being subjected to considerable abuse in moving operations they must, to be satisfactory, be ruggedly built. When the portable systems first came into use, very few of the sprinklers previously used on stationary systems proved suit- able. Some wore out in less than a season. On some portable systems, sprinklers may operate 2,000 hours or more during a single season, whereas on stationary systems they seldom operate more than 100 hours a season. The length of portable line required is usually determined by the di- mensions of the field, or fields, to be sprinkled ; the number of sprinklers used, by the length of line and the sprinkler spacing. The capacity of the pumping plant is limited by the source of power available. In many cases, farm tractors, used primarily for other operations, are utilized. Capacities of sprinklers range from approximately 5 to 25 gallons per Bul. 670] Irrigation by Sprinkling 23 minute, the choice depending upon a number of conditions. The capacity of some sprinklers can readily be changed by varying the size of nozzles. Sprinklers of the capacity that properly adapts the size of the pumping plant to the length of line required should therefore be selected. When the nozzles are too large, operating pressures are inadequate for proper performance ; when they are too small, inefficient use is made of the pumping unit. Efficiency necessitates operating the sprinklers at the lowest pressure that gives satisfactory distribution of water, and using the pumping plant to full capacity. The sprinklers are mounted on short risers directly over the portable lateral fyc/rartts Sta/iorfay /na//7 — 17 Pt//7?p//7tf p/a/7t Fig. 15. — Arrangement of a drag-type sprinkler system with porta- ble drag units supplied from a stationary pipe system. Solid lines indicate initial position of portable units ; dotted lines, the position after first move. Units E to L, shown in initial position, are moved in similar manner. pipe. These risers should be high enough to place the sprinkler over the top of the crop grown so that nothing interferes with the rotation of the sprinklers. For low-growing field crops such as sugar beets and peas, 12- or 18-inch risers are adequate; for higher-growing crops such as cotton, 30- or 36-inch risers may be necessary. High risers are objectionable be- cause they tend to tip over ; they also make it more difficult to move the pipe. In some cases, short bases are clamped to the pipe to prevent tip- ping. Some of the special couplings have footings to which strips of wood can be attached when required. PORTABLE PUMPING PLANTS There are two types of portable pumping plants : the tractor type that utilizes its own power both to operate the pump and to move from one 24 University of California — Experiment Station location to another ; and the independent engine type, where a separate engine is used for operating the pump. The latter may consist of a truck, a trailer, or skids on which the pump and engine are mounted. The tractor type is most common, probably because in many instances a tractor required for other farm work is available during the irrigation season. Figure 16 shows a typical plant. Tractors eliminate the necessity of other power for moving the pump from one position to another. Be- cause trucks tend to mire into mud along ditches, they are sometimes objectionable. Crawler-type tractors seldom become stalled in the mud. Fig. 16. — Portable pumping plant operating two ^-mile lines of 4-inch and 5-inch pipe with 61 sprinklers; capacity, about 1,100 gallons per minute. (From Ext. Cir. 95.) Pumps and Accessories. — Special lightweight centrifugal pumps have been designed for use on portable systems. Although not heavy, they are of rugged construction and are provided with an extra heavy shaft and bearings that will withstand operation of the pump in any position. Several sizes are available with capacities ranging from 200 to 1,200 gallons per minute, at heads of 80 to 200 feet. Detailed information con- cerning the performance of these pumps can be secured from the manu- facturers or from dealers handling sprinkling equipment. Sprinkler pumps are usually mounted on the front or rear end of the tractor, being supported by a special bracket welded or bolted to the tractor frame. For accessibility, they should be as low as possible. The pump may be driven by belts, chain, or gear. V-belts are better adapted than flat belts because of the short distance between pulleys. They should be properly selected to adequately handle the load at the required speed. Manufacturers of V-belts publish complete tables giving Bul. 670] Irrigation by Sprinkling 25 types and sizes to be used. Belts and chains should be protected against the weather as well as water from the sprinklers. Where separate en- gines are used, they are often directly connected to the pump with a flexible connection. The pump is equipped with a noncollapsible hose of adequate diameter and length. These are available in two types, rough and smooth bore ; the smooth type, which minimizes friction loss, is more desirable. Suction hose should be of the same size as, or one size larger than, the connection on the suction side of the pump, in accordance with the pump manufac- turers' recommendations. The inlet end of a hose should be provided with a screen, or strainer, to prevent debris from being drawn into the pump. These strainers should have adequate openings which, even when partially clogged, will not create an appreciable loss of head. Screens of inadequate capacity may appreciably lower the pressure (increase the vacuum) on the suction side of the pump and thereby decrease the effi- ciency. To prevent sucking mud from the bottom of the ditch, the strainer is usually designed with openings only on the top and rear end. These strainers are generally fabricated by the manufacturer supplying sprinkler pipe, or by the dealers handling this equipment. Many portable plants have a derrick for raising and handling the heavy suction hose and strainer when moving. When sprinkler systems are to be used to distribute fertilizer or spray solution, a 1-inch pipe connection between the suction hose and the pump case should be provided to permit the introduction of the solution into the system. Before using a sprinkler system for this purpose, one should consider the corrosive effect of the solution on the pump and sprinkler equipment. Ordinarily this is not serious if one takes the precaution of thoroughly flushing the equipment before allowing to stand after use. The pump discharge is equipped with a valve, which must be closed to prime the pump. Where two lines are to be operated from the same pumping plant, a Y- or T-fitting and two valves are required. This per- mits one line to be operated at a time. Flexible discharge hoses connect the pump and the sprinkler lines. These hoses should be long enough (10 to 15 feet) to allow the ends of the pipe to rest on the ground and to provide flexibility for making the angles or turns in the lines. They should be provided with quick- couplings of the type used on the sprinkler pipe. Centrifugal pumps must be provided with some means of priming. For this purpose, hand-operated suction pumps are often used. Some- times an automatic priming device is connected to the intake manifold, but it must be so arranged that the water is not drawn into the engine when the pump is primed. 26 University op California — Experiment Station Field Ditches. — Field ditches for supplying water to portable sprin- kler systems must be fairly deep to submerge adequately the strainer on the suction hose. In some places where sprinkling is extensive and the water of suitable quality, ditches constructed primarily as drains are used almost exclusively for this purpose. They are very convenient since they usually require no portable dams or tappoons to hold back the flow for the pump. The large ditches are usually constructed with a drag line or a trenching machine. Ditches constructed solely to serve a sprinkler system are usually built with a ditcher having double-wing blades mounted on a subsoiler frame. Large tractors, of at least 30 horsepower, are required to draw these ditchers. Sometimes several trips through a ditch must be made to bring it to adequate size, especially in level areas where velocities are very low. Where the ditches are constructed on any appreciable grade, portable tappoons of metal or canvas will hold back the flow and sub- merge the strainer. Usually they are constructed with a weir notch or opening of some type so that any excess flow will pass over without over- topping the banks of the ditch. Occasionally seepage softens the banks of the ditch, and the pumping plant becomes mired in the mud, especially where pumps are mounted on trailers or trucks. In some places the sprinkler system must always work down the ditch, with dams spaced at frequent intervals so that the pumping plant is always adjacent to a ditch that has had water in it for only a short time. Where sprinkler systems are used on fields of irregular topography, a strip of ground along the line of the ditch may have to be graded before the ditch is constructed. This grading can be done cheaply in comparison with the cost of leveling the whole field for surface irrigation. Since ditches must be constructed in rather definite locations to serve a sprin- kler system, they are not always practicable. Where the slopes are too steep, or where adverse grades are encountered, the best solution is a pressure supply line. SYSTEMS USING STATIONARY PUMPS AND PRESSURE SUPPLY LINES Stationary Pumping Plants. — Stationary pumping plants are em- ployed in connection with pressure supply lines. Where water is pumped from a well with a deep-well turbine pump, the additional pressure re- quired for sprinkling can be provided by additional stages (bowls) on the pump. This arrangement, which eliminates the necessity of a booster pump, is generally the most desirable. For an existing well with a pump designed to lift the water only to the ground surface an auxiliary booster pump may be more economical than changing over the turbine pump for Bul. 670] Irrigation by Sprinkling 27 the higher heads required. A horizontal centrifugal pump of the same type used on portable systems may serve the purpose. When connected directly to the discharge pipe of the turbine pump, it requires no prim- ing. The booster pump should have approximately the same capacity as the turbine pump so that both will be reasonably efficient when op- erating in series. Stationary pumping plants are often electric-driven. Though the op- erating cost may be somewhat higher than for internal-combustion en- gines, the electric plant requires less attention — sometimes an important consideration. One disadvantage of electric-driven plants is the difficulty of chang- ing the speed of the pump to accommodate the sprinkler system under varying conditions, since electric-driven pumps are usually directly con- nected and operate at a constant speed. With engine-driven pumps the speed can be varied to produce the desired pressure at the sprinklers when the length of supply line or the number of sprinklers varies. The sprinkler system must be of such capacity and so arranged that adequate pressures can be maintained under all operating conditions. Belt-con- nected electric-driven pumps are sometimes advisable so that pulley sizes can be changed to provide a change in speed, especially where large fluctuations in the ground-water level are experienced during a season, or where the number of sprinklers that can be used at one time varies because of the irregular shape of the field. Pressure Supply Lines. — As previously mentioned, pressure supply lines may be either portable or stationary. The former usually consists of extra lengths of portable sprinkler pipe without sprinklers, ordinarily of larger size. They should be large enough to carry the flow of the sprin- kler system without excessive friction loss. The approximate sizes re- quired are as follows : for flows of 50 to 100 gallons per minute, 3-inch pipe ; 100 to 200 gallons per minute, 4-inch pipe; 200 to 350 gallons per minute, 5-inch pipe, and 350 to 600 gallons per minute, 6-inch pipe. Most systems using 4-inch portable laterals will require 6-inch portable supply lines. For systems with capacities of more than 600 gallons per minute, portable supply lines are not satisfactory. Friction losses in pipe lines are further discussed in a later section. Several different kinds of pipe may be used for stationary supply lines. Welded steel lines, which are most extensively used, are less ex- pensive than any other kind of new pipe. Other pipes suitable for this purpose include casing, standard pipe, cast iron, cement-asbestos, rein- forced concrete, and reconditioned pipe and casing. Some of these mate- rials, being rather expensive, are seldom used for agricultural sprinkler systems. 28 University of California — Experiment Station Stationary supply lines are sometimes laid on top of the ground, some- times buried. When laid on the surface they must have expansion joints, such as long-sleeve Dayton couplings, at certain intervals to take care of the expansion and contraction caused by temperature changes. In some corrosive soils steel pipe will last much longer if supported above the ground on short piers than if laid directly on the ground or buried. Such piers must be spaced close enough together so that the pipe will not fail by bending, or crushing at the supports. Manufacturers can usually supply specific information regarding safe plans for any par- ticular kind of pipe. Surface pipe lines generally interfere with other cultural operations, and are therefore objectionable. Welded steel pipe is available in many sizes and weights. It is used more extensively in sizes above 4-inch diameter. Ordinarily it is supplied in 40-foot lengths and is connected together in the field by welding, spe- cial couplings, slip joints, or flanges. Welded joints are the most com- monly used. One main disadvantage of welding is that it destroys the protective coating on the inside of the pipe at the joint. Sometimes an outlet is welded on the pipe at the joint so that the inside of the pipe can be recoated after welding. These outlets are then plugged. There are two methods of field welding — eleetric-arc and acetylene. Electric-arc welding being faster, is becoming more popular. Expansion joints are usually spaced about 500 feet apart along the line, and at summits, or near structures or branches along the line. Many kinds of fittings are available for welded pipe. Dayton or Dresser couplings are sometimes used for connecting welded pipe. Although somewhat more expensive, they are especially desirable for galvanized pipe (where welding would destroy the gal- vanizing), and for small sizes, and short lines where it is not economical to provide welding equipment. They also provide for the necessary ex- pansion and contraction along the line. They are readily installed in the field by common labor without special tools. Other types of special couplings are available. Slip joints are frequently used for connecting welded pipe, especially where operating heads are fairly low. As they also eliminate the problem of recoating the inside of the pipe after the joint has been made, they are sometimes preferred to welding. Slip- joint pipe is provided with bell and spigot ends. The bell end of the pipe is heated to expand it and to soften the pipe coating, and the joint is made by either driving or jack- ing the pipe sections together. When slip-joint pipe is installed above the ground, it should be provided with hooks or lugs for wiring or bolting the sections together to prevent them from pulling apart during tem- perature changes. Bul. 670] Irrigation by Sprinkling 29 Since welded steel pipe is fairly thin as compared with other types, it should be protected against corrosion. The protection most commonly used is an asphalt or tar coating, both outside and inside. Additional protection usually consists of a wrapping; several grades are available. The type most suitable for the particular soil conditions should be se- lected. Welded pipe is sometimes galvanized. Although effective on the inside, and on surfaces exposed to the atmosphere, galvanizing protects for only a short time on the outside of pipe in contact with the soil. Under some conditions "cathodic" protection may be desirable. 6 Standard pipe with screw joints, used primarily for lines not exceed- ing 4 inches in diameter, is available either plain (black) or galvanized. Plain-end line pipe for welding or coupler connections is available in many sizes and weights ranging from 2% to 24 inches outside diameter (O.D.). Wall thicknesses range from approximately that of Merchant casing to that of extra strong pipe. Merchant casing, available in sizes ranging from 2}£ to 16 inches, O.D., differs from standard pipe in being lighter in weight and in having a finer thread. Casing may be connected either by threaded couplings or by welding. Ordinary fittings, threaded for casing, are available. Con- nections to valves and special fittings are made with flanges or by weld- ing standard pipe nipples to the casing and using standard screw fit- tings. Casing is usually installed with only an asphalt-dip coating or none. Old pipe, casing, or boiler tubing, reconditioned, is used to some ex- tent, especially for orchard systems in southern California. Usually it costs less than any other pipe. Threaded joints are sometimes used, but more often the joints are welded. Used pipe is reconditioned by cleaning it to remove rust and scale ; defective sections are cut out, and an asphalt- dip coating applied. Although the economy of reconditioned pipe may be questioned, since there are no definite standards regarding its exact condition when installed, it has often proved satisfactory. Where soils are only slightly corrosive, it sometimes lasts fairly well. Its low cost has been particularly attractive to farmers. Plain concrete pipe as ordinarly used for irrigation systems lacks strength to withstand the pressures required for sprinkling. Maximum recommended working heads for concrete pipe range from about 20 feet for large sizes to 50 feet for 6-inch pipe. It is suitable only for the so- called "low pressure" sprinkler systems, and for lines supplying water to pumping plants. Reinforced concrete pipe, however, can be made that will withstand pressures adequate for sprinkling systems. Such pipe has been used in some sections of California for irrigation systems, although 6 For a complete treatment of soil corrosion and pipe-line protection, see : Ewing, Scott. Soil corrosion and pipe line protection. 277 p. American Gas Association, New York, N. Y. 1938. 30 University of California — Experiment Station very seldom in connection with sprinkling. In 12-inch and larger sizes it is probably less expensive than steel lines and in addition is more permanent. Hydrants. — With stationary pressure supply lines, hydrants or valves must be spaced at convenient intervals along the line. The usual arrange- ment (fig. 13) shows the hydrants spaced three times the distance be- tween portable lines so that each hydrant is used for three positions of the portable line. In orchards, hydrants are usually spaced at every second, or every fourth tree row. These hydrants commonly consist of a short length of standard pipe and a valve. When welded steel lines are used, the risers are usually welded to the line. Threaded tees in the line are ordinarily used for risers on screw pipe. Several different types of valves are suitable for hydrants. For small sizes (3 inches or less) angle valves may be used; for hose connections the garden- type angle valve with hose thread is suitable. For larger sizes, gate valves are used almost exclusively. Valves are sometimes fitted with quick-couplings for connection to the portable sprinkler lines, though in many cases a portable fitting with quick-coupling is moved with the portable line and screwed into the valve before the connection is made. This arrangement eliminates a large number of special couplings and somewhat reduces the installation cost. To economize risers are sometimes provided with screw caps, and only a limited number of valves are used. These valves are then changed about once a day to the risers that will be needed for that day's operation. Although this plan reduces the initial cost, it makes operation somewhat less convenient, and necessitates shutting down the pumping plant while the valves are changed. LOW-PRESSURE SYSTEMS FOR FIELD CROPS Most of the sprinklers used for portable systems except under-tree orchard systems, operate best at pressures above 30 pounds per square inch. With lower pressures, the distribution is usually less uniform. Some sprinklers, however, distribute water satisfactorily at fairly low pressures (10 to 20 pounds per square inch). As previously mentioned, the area covered by a sprinkler increases with pressure up to a certain point; and, therefore, higher pressures permit wider spacings of sprin- kler lines with less frequent moves. High pressures also break up the jet into finer drops, which have less tendency to puddle the soil, and which permit higher rates of application. High pressures also make possible the use of smaller pipe sizes — an important feature. When the pressure is low, it is difficult to maintain uniform discharges from sprin- klers along the line, especially on hillsides. Despite these disadvantages, Bul. 670] Irrigation by Sprinkling 31 there are conditions under which low-pressure systems are desirable. Since the power requirement and the cost of pumping are approximately proportional to the pressure, the minimum pumping cost results with the lowest pressure that is otherwise satisfactory. Investigation indi- cates, however, that the power cost for pumping is usually only a small part of the total cost of operating sprinkler systems. Probably more im- portant than the cost of pumping is the fact that the lower pressure permits larger-capacity systems to be used with a given power unit, and this may appreciably lower the cost per unit of water applied, especially if the larger system can be handled without an increase in labor cost. Sometimes pressures of 10 to 20 pounds per square inch can be de- veloped by gravity, whereas higher pressures would require pumping. Sprinkling systems that can operate at these low pressures might be feasible where those requiring higher pressures would not. Even though the distribution of water with such systems might be less uniform than with higher pressures, it might be better than with surface irrigation. To meet these requirements, various types of low-pressure systems have been developed. Some sprinklers operating at 15 pounds per square inch can be spaced 40 to 50 feet apart. Low-pressure systems using sprinklers of this type resemble in many respects the higher-pressure systems. PERFORATED PIPE SYSTEMS Another kind of low-pressure system consists of lightweight, per- forated pipe that distributes water fairly uniformly over a strip of ground along the line. This pipe is available in two types having rates of application of about 1 inch and 2 inches per hour, respectively. An interesting feature is that the actual rate of application for either type remains approximately constant, while the effective width of the strip covered varies with the pressure from about 20 feet at pressures as low as 4 or 5 pounds per square inch to about 50 feet at a pressure of 20 pounds per square inch. In comparison with systems using rotating sprinklers, perforated pipe systems have relatively high rates of appli- cation which necessitates frequent moves of the pipe and may require the operator to devote full time to this task. The alternate-line arrange- ment, permitting continuous operation, is especially desirable. Because of its light weight the pipe can be readily moved by one man. There is also the advantage that a relatively large area can be irrigated with a short length of pipe. Perforated lines are used principally for irrigating pastures and other low-growing crops. The high rates of application make them most suit- able on pervious soils, or where only light applications are desired. Since it is not possible to regulate and equalize the pressure along the line as 32 University of California — Experiment Station with sprinklers, it is especially important that the system be arranged, when possible, so that the line is along the contour rather than up and down the hillside. TRAVELING SPRINKLER MACHINES To reduce the cost of sprinkling by reducing the amount of labor required for operating the system, several growers in California have constructed traveling sprinkler machines that pump water from a ditch Fig. 17. — Traveling sprinkler machine in operation. This machine moves continuously along the ditch at a speed of 75 feet per hour. Two sprinklers discharge about 500 gallons per minute under pressure of 60 pounds per square inch and effectively cover a strip 250 feet wide. The average applica- tion is about 2y 2 inches. and distribute it through large sprinkler nozzles while moving slowly but continuously along the ditch. These machines are planned for one- man operation. Figure 17 shows an example. The capacity of these ma- chines varies from about 400 to 800 gallons per minute. Most of them have been built on a tractor geared down to move 1 to 5 feet per minute with the engine operating at normal speed. To obtain maximum coverage and to break up the jets from the large nozzles into a spray, pressures of 60 to 80 pounds per square inch are used. According to field tests on two of these machines, they cover effectively a strip about 250 feet wide. They require, therefore, the construction of ditches at this spacing through- out the field, a feature which limits their use to rather flat areas where such ditches are feasible. These ditches remove about 5 per cent of the area from cultivation. Bul. 670] Irrigation by Sprinkling 33 Since portable pipe is not required, the cost of constructing such a machine is generally less than the cost of an ordinary portable sprinkler system of similar capacity. The cost of power for pumping is normally higher than for other systems because of the higher pressure required. For comparison, the total operating cost should include the cost of con- structing the ditches plus the crop value on the land removed from cultivation. TABLE 1 Sprinkler Discharge Required for Various Applications of Water in a Twelve-Hour Period Sprinkler Discharge, in gallons per minute, for various depths of application* spacing, feet 1.0 inch 1.5 inches 2.0 inches 2.5 inches 3.0 inches 3.5 inches 4.0 inches 4.5 inches 5.0 inches 6.0 inches 16 by 16 0.22 0.28 0.35 0.42 0.50 0.58 0.68 0.78 1.06 1.38 33 0.42 0.52 0.63 75 0.88 1 02 1.17 1.59 2.08 44 0.56 0.69 0,84 1 00 1.17 1.36 1.56 2.12 2.77 0.55 0.70 0.87 1 05 1.25 1.46 1.70 1.95 2.65 3.46 0.67 0.84 1.04 1.26 1.49 1.76 2.03 2 34 3.18 4.16 0.78 0.98 1.21 1.47 1.74 2 05 2.37 2.73 3.71 4.85 0.89 1.12 1.38 1.68 1 99 2.34 2.71 3.12 4 24 5.54 1 00 1.26 1.55 1.89 2.24 2.64 3.05 3.51 4.77 6.23 1.11 1.40 1 73 2 .10 2.49 2.93 3 39 3.90 5.30 6.92 1 33 18 by 18. . . 1 68 20 by 20 2 08 22 by 22 2 51 24 by 24 2 99 26 by 26 3 51 28 by 28 30 by 30 4.07 4.67 35 by 35 6.35 40 by 40 8.31 Average net application for the square area served by each sprinkler. UNDER-TREE SPRINKLER SYSTEMS FOR ORCHARDS Under-tree sprinkler systems have become very popular in some areas. Most of these systems are portable or semiportable. Essentially they consist of small, low-capacity sprinklers mounted on short risers that are located in the area between adjacent trees. The sprinklers are usually spaced the same distance apart as the trees so that each sprinkler covers an area equal to that occupied by one tree. Three different types are used for this purpose: whirling sprinklers; small, slow-revolving sprinklers; and fixed sprinkler heads of the type used for concealed lawn-sprinkler systems. Sprinklers for Under-Tree Systems. — One problem in developing this type of system has been to obtain a satisfactory sprinkler. To facilitate the moving operations and to obtain adequate penetration without run- off, these systems are sometimes operated on a 12-hour schedule, the sprinklers being moved morning and evening (6 a.m. and 6 p.m.). For the usual applications desired, this requires a sprinkler of rather low capacity, discharging 1 to 3 gallons per minute. There has been some 34 University of California — Experiment Station difficulty in obtaining sprinklers of these low capacities that perform properly. Rotating and whirling sprinklers of low capacities are not always dependable, for occasionally they fail to rotate. Also, because of the very small nozzles required, clogging is a problem. With the fixed sprinkler heads, it is difficult or impossible to secure the desired coverage without increasing the capacity beyond that desired. Most of the fixed Fig. 18. — Portable sprinkler unit illustrating small-sized sprinkler pipe and typical sprinkler. The pipe is made from iy2-inch, O.D., 18-gauge steel tubing, galvanized after fab- rication. It comes in standard lengths of 20 feet with a %-inch outlet on each length. Other lengths are also avail- able. sprinklers, such as lawn-sprinkler heads have, for example, capacities between 2 and 3 gallons per minute at 20 pounds' pressure, and are de- signed to cover effectively areas 12 to 16 feet square. Most tree spacings are 20 feet or more, and sprinkler capacities of only 1 to 2 gallons per minute are desirable. Fixed sprinklers that will effectively cover areas 20 feet square usually have capacities of 4 to 8 gallons per minute. They are suitable, therefore, only on pervious soils and where sprinklers are to be moved at frequent intervals. Table 1 gives the discharge required for average applications of 1 to 6 inches in depth over square areas 16 Bul. 670] Irrigation by Sprinkling 35 to 40 feet on a side in a period of 12 hours. This table covers the usual range of tree spacings in orchards and assumes that sprinklers are to be spaced the same distance as the trees — that is, the sprinkler is to be located in each square between four adjacent trees. For the usual tree spacing of 20 feet, an application of about 4 inches requires a discharge of only 1.4 gallons per minute. This table can be used for other periods of operation by simply increasing or decreasing the discharge accord- ingly. If, for example, sprinklers are to be moved at 4-hour intervals, Fig. 19. — Methods of determining and regulating the pressure at sprinklers. At the left is shown a sprinkler riser equipped with shut-off cock and "tank valve" which permits individual regulation of sprinklers. At the right is a homemade Pitot gauge for determining the pressure at the tip of a nozzle. The tip of the gauge is held in the center of the jet about half the nozzle diameter from the end of the nozzle, and the pres- sure, corresponding to the velocity head, is read on the pressure gauge. three times the discharge given is required. From table 3 one can deter- mine the approximate nozzle sizes for these discharges. Some orchard sprinklers have two nozzles, and others have only one. Portable Sprinkler Pipe for Under-Tree Systems. — Because of the low capacity of under-tree sprinklers, sprinkler pipe in relatively small sizes is suitable. Such pipe is now available in 1%- and 2-inch diameters. It is made from 18-gauge tubing and has special quick-couplings (fig. 18) . Although ordinarily furnished in 20-foot lengths, other lengths can be obtained on special order. Pipes of the same length as the tree spacing are preferable. Special quick-couplings are also available for other kinds of pipe, including lightweight electric conduit and hard-temper copper tubing. Before the regular sprinkler pipe was available in small sizes, many systems were made up from 1-inch and l^-inch electric conduit (sometimes called electrical metallic tubing). This tubing is available in 10-foot lengths in sizes ranging from % to 2 inches. Inside diameters are slightly greater than the nominal sizes. It is available galvanized 36 University of California — Experiment Station inside and outside or with an electroplated galvanizing on the outside and lacquer finish on the inside. One objection to its use is that welding or brazing the short lengths together destroys the protective finish on the inside. Sprinklers are mounted on ^-inch pipe risers 6 to 18 inches long, the length depending upon whether or not a summer covercrop is grown in the orchard. Risers should be as short as possible, since high risers tend to overturn the pipe. Where there is an appreciable variation in elevation along a sprinkler line, or where there is excessive variation in pressure due to friction loss, a shutoff cock on each riser, and some method of determining the pres- sure, make it possible to regulate and equalize the pressure at the sprinklers. One method is to use a small "tank valve," similar to an auto- mobile-tire valve stem, between the shutoff cock and the sprinkler (left view, fig. 19). Pressures can then be determined with a pressure gauge equipped with an air-hose chuck. With some sprinklers, pressures can be determined with a Pitot gauge (right view, fig. 19) . A whirling sprin- kler is now available which automatically maintains a constant discharge for all pressures above 8 pounds per square inch, and eliminates the necessity for manual regulation. The size of sprinkler pipe required for under-tree systems depends upon the capacity and spacing of the sprinklers and upon the length of the portable laterals. Where no pressure regulation is provided, the pipe should be large enough so that the friction loss in the pipe will not exceed 20 per cent of the average pressure ; this will limit the variation in dis- charge of the sprinklers to 10 per cent. The permissible length of pipe in different sizes for different sprinkler discharges can be determined from figure 26. In small orchards one main supply line through the center with port- able laterals extending halfway across the orchard is the simplest ar- rangement. For large orchards, it is sometimes more economical to use two parallel supply lines with shorter laterals, which will permit the use of smaller portable pipe. One can determine the most economical ar- rangement only after ascertaining what sizes of pipe are required under both plans. When sprinklers are moved only twice a day, two tree spaces can be covered with each lateral per day, and the total number of laterals required for the orchard is determined by the required frequency of irrigation. Most orchards will require two or more portable laterals. Stationary supply lines should be of ample capacity to carry the total flow with a reasonable friction loss. Pipe of minimum size can be used when the laterals are distributed throughout the orchard and arranged Bul. 670] Irrigation by Sprinkling 37 so that each takes care of the contiguous areas. Figure 20 shows a suitable arrangement for a 10-acre orchard. Friction losses in different types of pipe suitable for stationary lines are given in tables 5 and 6 and in figure 24. TJnder-Tree Systems Using Sprinklers Attached to Hose. — Many under-tree sprinkler systems consist of several sprinklers mounted on A/7/n. />/ 27 assure at fiyc/ra/rt^y te.per sp. //?. 2 'fe 'sfot/baary ma/'/T^ f'/z '/?yc/ra/?ts — ( Mm. pressure 20ft. x \ /6 spr/Mers^O' Q 5 % %% L apart . Capacity /•fyp./n. \ 1 VI 1 frees 20 ' a pari 9 & 6>6{jy0<££>iS>O &0£J[ f Pressc/re , 32 73. per sH t* 7J a oi H g m M (5 fc h5 C4 pq m fe o CO Fh I-) p GO « ■PTS »3 a> u a to a .5 «* c3 CU "o"3 is o © a ^ a o Xi O »* O oj «0 Average daily wage, dollars ©* Net operating time, per cent - Effi- ciency of system per cent o Total applica- tion, acre- inches per day OJ Average depth of applica- tion, inches OO Acres covered per day - to Pump discharge, gallons per minute >Q Minimum and maximum pressure at sprinklers, pounds per square inch "* Number of sprink- lers so Length of portable pipe, feet* ©» a - OJ CO o CO o ^H OJ oo io o o I - o d •o 00 © iO co © © CO OO OO © © CM © l-l CM oo oj © © oo © (M o OS o CO lO o o CM c OJ OJ o CM C) d o CM CM © © IO © OJ © «-H CO © © CM CM © o © 00 "*< CO r - CO o o CO O0 CD 5 IO o r- io CO CO o CI CO © © © © "* CO CO co 00 00 OO r~ co OO OO CO OO S3 CO 00 CO OO oo oj CM OJ IO CO 00 oo CO 00 •O tO CO 00 OO US 00 CO CC OJ IO CO 00 CO CO OO lO CO r^ CM OC O "0 O0 J>- CN 00 CO o O) 3 - - 0O CO OJ CM © © OO co OJ CO 00 o OJ CN iO CO to 00 o OJ 00 OO co •o t- CO •<*! OJ >o CO CO co co IO ■* CM CO CN CO CN CO CO CM r-l CO ■>»< T»< 1>- IO CO CO o CO 00 OJ CN CO CO OJ OJ CO OJ © Tj< US IO CM Tt< CO OJ CO OJ © © "tfi »o IO 00 CO CO O CM CO 5 eo CO co co OJ CO o CO CO to © CO CM CO CO o CO 00 CO CM ■* OJ CO o CO o CI o o 8S o © CM o s CC © 00 1-1 O0 1-1 © © cc oo © 00 CM CO CO OJ CO -r r- od OS CM IO CM < ^ a cn co »- oo l>- IO ■** CO 0O CO lO ■* co iO © © © © © © © lO O O0 CN © © © © CM »• rH O O © CM © N M Ol tD IO •* CO IO CO CO ■«1< CO CO © © © © o © © «o © CO © oo io co © 8°- CO IO CO CO CO CO CO >* CO 00 O CO CO OO OJ OO 00 oo oo CO oo »-i -* IO CO O) O0 N IN CO -Tt< 00 1^ CO «o © © IO CO CN OJ co CO CM ^h CM CO CO "5 CO CO IO co t»i r— cm _l US OJ CM CO CM CO CO CM MO O0 O ^H CO 5 ■* O! 00 o ^a N O O rt T|i O0 N rt © ^ © CN s 720 682 1,143 685 CO IO »-. 1-1 co oo 53-60 30-35 28-49 29-44 ^H OJ T T CO CM CM CO CO CO CO IO •**> OJ Tf< CO ■* CO CO CO 00 CO IO CO o © © © CM CO O 00 CO ^- -^ CM © © *3< © CO IO CM •O — I 0 *— CO 0 CO O © OS -i © © 1.018 712 660 461 600 603 CO CD 29-44 35-42 32-39 30-36 36-43 35-43 CO CO Ol O Ol (O lO CD *o -CP -*l CO CO CO CO o o o o o o © CO OO CD -* CM N O Ol N CO Tf CO CM N O H » «' M h N IM CO CO > < a co -h — r^ o © -^t t-~ CD © l^ t-- OS ©©—>©©© © OO o H rt f O CO tO CM CM CM CM CM CM ©©©©©© co CM © O O co to N oo K3 11 CO if} tP to ©©©'©'© o © © lO © © © © © CM © © lO © CO CM CO CO CM CO OS CM O © •— 1 © CO © lO CO N 00 N CO CM CM OO OS © OO t^ to i< to to ■* to OS 12.1 22.5 14.5 21.3 21.3 17.7 CM CD lO tO 1^ CD CM CM CO -- CM CM CO 1^ r-- -* CM OO CM to * to ffl S CO us © N N to f ID CO N rt CO rt -^ O OS o CO CO ■* CO CO K5 lO © 00 OS -^H r-* ■* lO CO ■* lO « 00 to »o OS >o «o OS OS ^ co co i* co co CO tO N Tf Ol to CM — < CM CM ^ CM © §3 CM »C © ■* CO «5 CO CO CO CO 1" N CO 1,020 1,340 920 1,320 980 760 © N ■* CO * CO if) CM CM CO > < — ' —I ~ l-H © © — I CO H ^( M IN M © © © © © © OS CO OO lO -f © ^ © © © © © © lO CO CO CO CO CO CO CM CM N N Ol O tO * to >o lO t^ »o r— tO t^ «5 © lO t^ 105 CO to OO © © OO * tO CO 11 N CO CO N N CO if) U) N OO O h CM OS lO lO © f) CO CO ■* f t~- OS -f CM © CO CO «5 -* 1 CO lO ■«* M to CO id N tO * CO CO CO CO N c3<^ CO CO fee co O Si CO in co co g co p4 a s >>CC dl co m to t-- to t-~ i-H t-i CM co co 50 University of California — Experiment Station are summarized in table 2, and following text explains details of deter- mining data in the columns. Explanation of Table 2. — Column 4 : The pressures at the sprinklers were determined with a calibrated pressure gauge attached to a special rubber fitting that was slipped over the main nozzle of the sprinkler. In doing this, the discharge of one sprinkler nozzle was shut off, momen- tarily increasing the pressure on the system. To obtain the correct pres- sure, therefore, it was necessary to determine this increase in pressure and subtract it from the gauge reading. The correction required, which varied with different systems, was usually between 0.5 and 1.5 pounds per square inch. Column 5 : The discharge of the pump was determined by summing up the discharge of the sprinklers, which was determined from the sprinkler calibrations and from the pressure measurements at the sprinklers. Column 6 : Total head includes the pressure at the pump plus the lift from the water surface in the ditch to the elevation of the sprinklers plus the velocity head at the pump discharge, all expressed in feet of head. The pressure at the pump was determined by adding the estimated friction loss in the pipe between the pump and the nearest sprinkler, to the pressure at that sprinkler. Column 7 : The area covered per day is determined by multiplying the area covered at each setting by the number of settings per day. The area covered at each setting was equal to the width of the field (approxi- mate length of the line) times the distance the line is moved. The number of settings per day was either secured from the operators, or determined from the frequency of the moves. The figure given, therefore, represents the nominal area rather than the average area actually covered. Column 8 : The average depth of application in inches is determined by dividing the total application in acre-inches per day (column 9) by the area irrigated per day (column 7). Column 9 : The total application in acre-inches per day is determined from the discharge of the system and the estimated net operating time per day (column 11). Column 10 : The efficiency of the sprinkler system is the ratio of the total hydraulic energy at the sprinklers to the energy at the pump dis- charge, expressed as a percentage. The difference between the figure given and 100 per cent represents the loss of energy in the pipe by fric- tion. Compare the average efficiency of the split-line arrangement and the single-line arrangement. Column 11 : The net operating time represents the approximate per- centage of the total working day that the system was operating at full Bul. 670] Irrigation by Sprinkling 51 capacity. For systems with a split line, where one line is operating while the other is being moved, the loss of time in moving was taken as the time required to move one line and the pumping plant. Column 12 : Where the average daily wage included board and room, an allowance was made for the value of the board and room in order to make the total wages comparable for all systems studied. In nearly all cases the daily wage was for 12 hours. These data are for 1935 and 1936, a period of relatively low wages. Column 13 : The average operating cost per acre-inch of water applied for labor is equal to the total daily wages paid divided by the total ap- plication in acre-inches per day (column 9) . Column 14: The cost of power per acre-inch of water applied was determined by dividing the total daily cost of fuel, lubricating oil, and miscellaneous expenses (as given by owners or operators) by the total output in acre-inches per day. Column 15 : The total operating cost is the sum of columns 13 and 14, no allowance being made for interest or depreciation on the equipment. HYDRAULICS OF SPRINKLER SYSTEMS Sprinkler systems are fairly complicated as compared with other ir- rigation systems. They are composed of many different kinds of me- chanical and hydraulic equipment. They require careful planning in order to fit the fields on which they are to be used, and to be able to supply the water required for the various crops. Pipe sizes, nozzle sizes, pump sizes, and power requirements must be determined. Equipment must be selected and installed. The sprinkler system as a whole must be so proportioned that it will operate efficiently. The principal factors that enter into the cost of operation must be clearly understood. The planning of a sprinkler system is essentially an engineering function, and one that requires, especially, a knowledge of the hydraulic principles involved. The hydraulics of sprinkler systems is somewhat different than the usual problems in hydraulics that are encountered in engineering practice. 10 The material covered in this and the following sections of the bulletin have been prepared primarily for engineers, and others engaged in the planning of sprinkler systems, and in the design, manufacture, and sale of sprinkler equipment. The information is presented in a somewhat 10 Some phases of this subject have been discussed in previous papers; see: Chris- tiansen, J. E. Irrigation by sprinkling. Agr. Engin. 18(12) : 533-38. Dec. 1937. Chris- tiansen, J. E. Hydraulics of sprinkling systems for irrigation. Amer. Soc. Civ. Engin. Trans. 107:221-50. 1942. Christiansen, J. E. The uniformity of application of water by sprinkler systems. Agr. Engin. 22(3) : 89-92. March, 1941. 52 University of California — Experiment Station technical manner; mathematical symbols and formulas, that would otherwise be out of place, have been employed because they add to clear- ness and conciseness. DISCHARGE FROM NOZZLES The discharge from all types of nozzles and orifices can be expressed by the orifice formula, which is derived from Torricelli's theorem : Q = CA V2gH (1) where Q is the discharge in cubic feet per second; C is the coefficient of discharge, which is approximately constant for a given nozzle or orifice; A is the area of the nozzle or orifices in square feet; g is the acceleration of gravity (32.2 feet per second per second) ; and H is the total head in feet, which includes the velocity head. This formula is not convenient to use, however, since the discharge from nozzles is commonly given in gallons per minute, nozzle dimensions are expressed in inches, and heads are expressed as pressures in pounds per square inch. Usually the veloc- ity head in the riser is negligible in comparison with the pressure head and no correction for it need be made. Expressed in these units, equation 1 becomes q = 38.00 CayfP (2) or q = 29.85 Cd 2 \fp (3) where q is the discharge in gallons per minute, a is the area of the nozzle in square inches, d is the diameter of the nozzle in inches, and P is the pressure in pounds per square inch. Table 3 gives the theoretical dis- charge (C = 1.00) of round nozzles from % 4 to Y 2 inch in diameter. The actual discharge of a nozzle can be obtained by multiplying the dis- charge obtained from this table by the actual coefficient of discharge of the nozzle. Coefficients of Discharge of Nozzles and Sprinklers. — The coefficient of discharge C of a nozzle is a ratio of the actual to the theoretical dis- charge. It is the product of two other coefficients : the coefficient of con- traction and the coefficient of velocity. The coefficient of contraction is the ratio of the cross-sectional area of the jet, at the point of minimum area, to the area of the nozzle. For most sprinkler nozzles the coefficient of contraction is nearly 1.00. The coefficient of velocity for a sprinkler is the ratio of actual velocity at the nozzle to the theoretical velocity that would result if there were no friction losses in the sprinkler or nozzle. This coefficient is an important factor, since it indicates the energy loss due to friction and turbulence within the sprinkler. It is a measure of co -* conco N OS »0 CO CM ONtDrHOO o o © —' -i O O O i-> — -' OlNtOrH^ OOOhh > OS CO OC »0 CO CM ■ ION OOONtJii CO>T|iiOrt lOTflON^" NCO<*>ON OOONMIfl ICO^^HC© Tf O CO CO t-c rtNUJOW CM O — CO CO O CO ■ w-in« co oc ex) i— i o woiom* n^*< coco>oa;rti o N n i •t< co -^ cc as CM CO •*!< "O CO — i—i NCOllO'O co oo o co »o N00 10W5 05 OOONN O CM CC O CD OOO— i^H > as co »o »o I© f-i CO »o CO OC CC N CM OC OS -1 CO "O OCM »0 OCO OOO^— OOOi-"-. •-< ■* o> CO lO CO -*tl Tj< N CM oomono o o o o —' OOlOO-HOl lO CO CM Tf CO O IN "O Oj ■* lOONON 05 0:-N' CO CM i— i CM »0 0CP3ON CM CO •<»< «0 CO N OS ^ CM ' CMCOtCIOCO NOION' CC OS tf OS cc N OS CM rf N — ^ CM CM CM rnCNNCON CM •»!< NCM 00 N OS i-l -*• CO i-HrtCNlNM OO CO CM CO N eoo-ioo COOC> rt COCO ■ -*ji io 00 CM COCMiOOO) toiomi-o NlOTf ifJN NittNON COOtC OOCM ■h/i OS CO »c ■<*< iO CX) CM CO O tONONN 0»0 »-i 00 CO *3H tJi ic »0 IO COOCCCMiO N O CO CO OS i-c — CM CM CM cmcocococo cc — ooo o eOifOoo CM CO OS CO CO Tt< •*>* i- IO IO CM CM CO ^f CO 01WO«3I co N eo os i NOOO-I OOOOH CMCMCO-£ oood-J U 3 O ^ CD — • 1.1-1 1O1CKD00H t-1 CM CO ■* »0 CO N OS O CM HOifiOOO oooo-< KJCNONtO OO lOPJ i-i i-H r-c CM CO ■*»< »0 'COOOO I •* N — CD 'NOOOrt ooiooh cm co co co co cotcoooo «<*nc;h n os cm co cm CM -*■ N OS CM CM CM CM CM CO oooo> CO OS CO CO OO t- CO i-i OS OO i-l CM COCOTJ1 IONQOOS. CM Tf iO N OS rtMlOOOO _ rt ,-( ^- — c CMCMCMCMCO CO CO 00 OS C ^f r— oo oc I OHMCD< i CM CM CO •«*< lOcONOO IO CO N CO OS O O OS TTCOlC O-* CO OOOO ^r cc ic co co ^ ic -*t< co os MiOSU5(NO) iHiHCMCIJM cnho t> coco oonnooos Tl< tO CO N00 iCOiicOCO OMtPCDOO , ,_ ,_< ,_, H CM CM CM CM CM CM IO OS -ht OS O — CM ^t< IO co co co co t-i CO -rj< CM CO i-H O i— ' CO IO OS ooooo CM i-i CO i-i ■ CO CO OS CM ■ OHNIOI ooooo iiotomo coco-*con irtWCNM . OOSTf CO i CD OCOCM 1— icMCMCO OOOOO i— . ooooo i CM t-i i-i CM l IO CON CO OS CO id CO CO CO -f »OCO N NCCWOrt CM COCO'* IO CClOOOrllO lOiiHOO csiniflioo) USOU5CDN 1KOOOON CM CM CM CO CO 00 OS O i— i CO NNCONCO NlO>nOM CO N CM N CO ififlNOOO lON^Nlfl ■f CO OS CM CD CM CO •q< CO N l i-H r-* ^- ,-. — i ,-< ,-< CM CM CM CM OOr-nm :§S S O O CO coco CTjtMNO-h CDO-.CMiOOS CO N CM I IfJCONOOOS OS IO CM OS CO TfCM^OOS OOOOO O i— i — i— 1 N M CO CC if ■* LQ CC C I OOOOC OOOOO OOOO OOOOO OOOOO ooooo IO COCO OS o ,_, ,_ _ — ,CM --i N CO cc O CD t— CO CM CO os cr. O — ■> cm cm co io co n OOOOO coco coco O^ CD CO O-^ CD Tf 00 CM IO CO fCDOscOCO MONNf ■* IO O0 i— i IO i-i N IO Tf CO >0 CO OS CO CO CC CO — CO COOSCMlOOO HlO CM CM CM CO CO CO CO -*i -*M Tj< iOIO OS O O N IO ^i CO CO CO CM CO »0 N OS i—l CO CM CM CM CM CM CO CO OS O CM CO IO N OO IO O i-i CD CM CO 0C OS do 0/ 3 Oi 3 ■H cc 3 cc "3 a m 0) 0, u cu d Tl rt £?•« cci >, rfj M «, -O d CJ IS A* >^ 10 d II ^ ^3 « 2.2 o «« 03 54 University of California — Experiment Station the sprinkler's efficiency in converting pressure to velocity — a high value indicating a small loss of energy. The coefficient of velocity for a sprin- kler can be determined directly by dividing the pressure at the tip of the nozzle, determined with a Pitot gauge (fig. 19) by the pressure at the base of the sprinkler, corrected for velocity head and difference in ele- vation. The same pressure gauge should be used for both pressure meas- urements; if different gauges are used, small differences in the calibra- tion of the gauges may cause an appreciable error in the coefficient. The coefficients of discharge of well-designed sprinkler nozzles vary from about 0.95 to 0.98. Precise determinations of the coefficient of discharge are difficult ; and unless precision instruments are used in the tests, determinations within 1 or 2 per cent are all that can be expected. According to tests, coefficients of discharge for sprinklers with two or more nozzles vary from less than 0.80 for some sprinklers to about 0.98 for others, the coefficient for any sprinkler depends upon the con- struction of the sprinkler, the driving mechanism used, and the nozzle sizes. Sprinklers with internal driving mechanisms generally have the lowest coefficients ; in some cases there is an appreciable loss of energy as the water passes through the sprinkler. The coefficient of discharge for the same sprinkler may vary considerably with different nozzle sizes. The larger the nozzles, the lower the coefficient, because of the higher velocity and the greater energy loss within the sprinkler. With a % 6 -inch main nozzle, for example, a certain sprinkler has a coefficient of dis- charge of 0.97; with a %-inch main nozzle, a coefficient of only 0.88. Discharge of Small Nozzles for Nozzle Lines. — The small nozzles used on nozzle lines are constructed differently from those used on sprinklers. They are more truly orifices in a thin wall. To minimize clogging, some of them are machined with an abrupt entrance, the base of the nozzle being slightly convex so that foreign matter will pass by without being trapped. The outer end is countersunk so that the length of the drilled hole is only about %2 inch. The abrupt entrance causes a contraction of the jet, which passes through the nozzle without touching the sides of the opening. This results in a crystal-clear stream that travels a maximum distance. The length of the hole in some nozzles is sufficient so that the jet expands and fills the opening at the discharge end. Although the coefficient of discharge is higher, the result is a fuzzy jet that breaks up more rapidly and does not carry so far. Tests on several of these nozzles gave coefficients of discharge between 0.8 and 0.9, the low values being due to the abrupt entrances. The nozzles tested had diameters varying from 0.034 to 0.045 inch, with discharges varying from a minimum of 0.10 gallon per minute at 13.5 pounds per square inch pressure, to a maximum of 0.34 gallon per minute at 42 Bul. 670] Irrigation by Sprinkling 55 pounds per square inch. Table 4, giving the discharge of small nozzles of this type at different pressures, is based on a coefficient of discharge of 0.85, the approximate average value of the nozzles tested. Pressure Required for Sprinkler Operation. — For proper perform- ance, all sprinklers require a certain minimum pressure. Some are de- signed for and operate best under fairly high pressures (more than 30 pounds per square inch), whereas others operate satisfactorily on pres- sures of 10 to 15 pounds. Pressure is needed to provide the velocity TABLE 4 Discharge of Small Nozzles for Use on Nozzle Lines* Pressure, pounds per square inch 5 10 15 20 25 30 35 40 45 50 55 60 Discharge, in gallons per minute, for various nozzle diameters 0.030 inch 0.051 .072 088 102 114 125 135 144 153 161 169 177 0.035 inch .120 .139 .155 .170 .183 .196 .208 .219 .230 0.240 040 inch 0.091 .128 .157 .181 .203 .222 .240 .256 .272 .286 .300 0.314 0.045 inch 0.115 .162 .199 .230 .257 .281 .304 .324 .344 .363 .380 0.397 0.050 inch 0.142 .200 .245 .283 .317 347 .374 .400 .425 .448 .470 0.490 055 inch 0.170 .240 .295 .343 .383 .420 .453 .485 .514 .542 .568 594 0.060 inch 0.204 .289 .353 .456 .500 .540 .577 .612 .645 .676 .706 0.070 inch 0.277 .392 .480 .554 .620 6S0 .734 .784 .832 .877 .920 0.960 0.080 inch 0.362 0.512 0.628 0.725 0.810 0.&88 959 1.025 1.087 1.146 1.201 1.255 090 100 inch inch 0.458 0.649 0.794 0.918 1.025 1.124 1.214 1.297 1.376 1.450 1.521 1.589 0.566 0.801 0.981 1.133 1.267 1.388 1.498 1.602 1.700 1.792 1.878 1.962 tested. Based on a coefficient of discharge, C = 0.85, an approximate average value for different nozzles which is required for two purposes : to secure distance of travel, or coverage ; and to break up the water into small drops that can be prop- erly distributed over the desired area. The first requirement can be satis- fied only by a pressure adequate for this purpose. The second can be satisfied by other means, such as deflectors or shape of nozzle, but only at a sacrifice in the distance the water is thrown. Many sprinklers utilize both methods, having one round nozzle for distance, and a second nozzle with a deflector or slotted nozzle that breaks up the water into small drops that fall near the sprinkler. Let us consider what happens as the pressure is gradually increased on a single round sprinkler nozzle. With very low pressure the water issues in a solid stream, and all of it strikes the ground about the same distance from the sprinkler. If the sprinkler rotates, only a narrow ring receives water. With increase in pressure, the water becomes broken up into drops and covers a larger and wider ring. Upon examining the drops that strike the ground, we find that the largest are carried to the 56 University of California — Experiment Station outside edge of the area covered, while the smallest fall near the sprin- kler. With further increase in pressure the width of the ring becomes equal to the radius, and the entire area receives water. As the pressure continues to increase, more and more of the water falls near the sprinkler, and the average size of the drops becomes smaller and smaller. The ef- fective coverage of the sprinkler, which governs the permissible spacing of sprinklers, will increase with pressure up to a certain point, and will then decrease with further increase in pressure because of the larger percentage of water falling near the sprinkler. The larger the nozzle, the higher the pressure will be at maximum effective coverage. The minimum pressure required for most sprinkler nozzles varies from about 15 to 40 pounds per square inch. The larger the nozzles and the greater the capacity of the sprinkler, the higher the pressure re- quired for satisfactory distribution. Most field and orchard sprinklers, except some under-tree sprinklers, operate under pressures between 20 and 60 pounds per square inch. Some very large sprinklers built for special purposes require pressures of 100 to 150 pounds and cover diameters up to 300 feet. Special low-pressure sprinklers are available. Some of the small under- tree orchard types operate on pressures as low as 5 pounds per square inch and will cover adequately the area between adjacent trees. The pressures mentioned above are those required at the sprinkler. When pressures are determined at any point other than the sprinkler, they must be corrected both for frictional losses and for differences in elevation to obtain the actual pressure under which the sprinkler is operating. Sometimes a user assumes that his system is operating under the pressure indicated by a gauge at the pump or source of supply, whereas some of the sprinklers may actually be working under a much lower pressure. FLOW OF WATER IN PIPE The flow of water through a pipe is always accompanied by a loss of head due to friction. Additional loss of head is also caused by sharp bends, sudden enlargements or contractions in the pipe, or other ob- structions. Figure 23 illustrates the effect of friction. The hydraulic grade line is an imaginary line joining the elevation at which water would stand in open columns connected with the pipe line. The height of the hydraulic grade line above the pipe represents the head or pres- sure at any point on the line. The friction loss is represented by a de- crease in the elevation to which water will rise in the columns, or by the slope of the hydraulic grade line. The friction loss does not always pro- duce a decrease in pressure in the pipe line. If the slope of the pipe line is greater than the slope of the hydraulic grade line, there will be an Bul. 670] Irrigation by Sprinkling 57 increase in pressure, that is (as illustrated in fig. 23), the total head or pressure Hb above the pipe line at B, may be greater than the head Ha at A. A measurement of the difference in pressure at two points on a pipe line is, therefore, not a measure of the friction loss unless the line is level, or unless corrections are made for differences in elevation. Formulas for Friction Loss in Pipe Lines. — Friction losses in pipe lines depend principally upon the roughness of the inside of the pipe, the size of the pipe, and the velocity of the water. Many formulas have been proposed to express the relation between these principal factors. ///j Head iosi /h/ friction Fig. 23. — Diagram illustrating friction loss in pipe lines. See the text for explanation. The formulas are based upon numerous tests on pipes of different kinds and sizes. The three formulas most commonly used as a basis for tables and diagrams for friction losses in pipe of the kind used for sprinkler systems are Weisbach's, Williams and Hazen's, and Scobey's. "When written in the same form, these formulas are as follows : Weisbach's: 11 H f = Williams and Hazen's : x fLV* 2gD Hi 3.022 LF 1852 (J 1.862 _£) 1.167 (4) (5) 11 This equation, sometimes called the Darcy formula, is mentioned in any standard textbook on hydraulics. 12 This equation is commonly written V = 0.001"° 04 C R°-«* S -* 1 , or V = 1.318 C B° 63 S°- 5i . See: Williams, Gardner S., and Allen Hazen. Hydraulic tables. 115 p. John Wiley & Sons. New York, N. Y. 58 University of California — Experiment Station TABLE 5 Friction Loss in Standard Wrought-Iron or Steel Pipe for C 100* Flow, gallons Friction loss, in feet of head, for each 100 feet of nominal pipe sizes t pipe in the following per minute % inch 1 inch 1M inches 1H inches 2 inches 2H inches 3 inches inches 4 inches 5 inches 6 inches 3 4.1 6.9 10.5 14.7 25.0 37.8 52.9 70.4 90.2 1.3 2.1 3.2 4.5 7.7 11.6 16.3 21.7 27.8 34.6 42.1 63.6 89.2 0.33 0.56 0.85 1.19 2.03 3.06 4.29 5.71 7.32 9.10 11.1 16.7 23.4 31.3 39.9 49.6 60.4 72 84.5 0.16 0.26 0.40 0.56 0.96 1.44 2.03 2.69 3.45 4.29 5.21 7.89 11.1 14.7 18.8 23.4 28.5 34 39.9 53 1 68.0 84.5 0.08 12 0.17 0.28 0.43 0.60 0.80 1.02 1.27 1.55 2.34 3.28 4.36 5.58 6.94 8.44 10.1 11.8 15.7 20 .2 25.1 30.5 42.7 56.8 72.9 90 5 0.12 0.18 25 34 0.43 0.53 65 0.98 1.38 1.83 2.35 2.92 3.55 4 24 4.97 6.62 8.48 10.6 12.8 18.0 23.9 30.6 38.0 46.3 70.0 98.0 ;;;;; 09 0.12 0.15 0.19 23 0.34 0.48 0.64 0.82 1.01 1.23 1.47 1.73 2 30 2.94 3.66 4.44 6.23 8.30 11.3 13.2 16.1 24.3 34.0 45.3 58.0 72 87.6 07 0.09 0.11 0.17 24 0.31 0.40 0.50 61 0.72 0.85 1.13 1.45 1.80 2.19 3.07 4.09 5.24 6.51 7.92 12.0 16.7 22.3 28.6 35.5 43.4 51.6 60.5 80.5 0.09 0.13 0.17 0.22 27 0.33 39 0.46 0.61 0.78 0.97 1.18 1.66 2.21 2.83 3.54 4.28 6.47 9.06 12.1 15.4 19.2 23.3 27.9 32.7 43.5 55.8 69.3 84.2 0.07 0.09 0.11 0.13 0.15 0.20 0.26 0.32 0.39 0.55 0.73 0.94 1.17 1.42 2.15 3.01 4.01 5.13 6.38 7.76 9.27 10.9 ' 14.5 18.5 23.0 28.0 4 5 6 8 10 12 14 18 18 20 25 30 35 40 45 50 55 60 70 08 80 0.11 90 0.13 100 0.16 120 23 140 0.30 160 0.38 180 0.48 200 0.58 250 0.88 300 1.23 350 1.64 400 2.10 450 2.61 500 3.17 550 3.79 600 4.45 700 800 900 5.92 7.58 9.31 1,000 11.44 * Computed from Williams and Hazen's formula. t To convert head in feet to pressure in pounds per square inch, divide by 2.31. Bul. 670] Irrigation by Sprinkling 59 TABLE 6 Friction Loss in Standard Wrought-Iron or Steel Pipe for C t= 120* Flow, gallons per minute 45. 50. 55. 70. 80. 90. 100. 120. 140. 250. 300. 350. 400. 450. 500. 550. 600. 700. 800. 900. 1,000. Friction loss, in feet of head, for each 100 feet of pipe in the following nominal pipe sizes t inch 2.9 4.9 7.5 10.5 17.8 27.0 37.7 50.2 64.3 80.0 97.3 ich Si, V4 inches 0.23 0.40 0.61 0.85 1.45 2.19 3.06 4.07 5.22 6.49 7.90 11.9 16.7 22.2 28.5 35.4 43.1 51.4 60.4 80.4 inches 0.11 0.19 0.29 0.40 0.68 1.03 1.45 1.92 2.46 3.06 3.72 5.64 7.90 10.5 13.5 16.7 20.3 24 3 28.5 37.9 48.5 60.4 73.4 2 inches 0.06 0.08 0.12 0.20 0.31 0.43 0.57 0.73 0.91 1.10 1.67 2.34 3.11 3.98 4.95 6.02 7.18 8.43 11.2 14.4 17.9 21.7 30.5 40.5 52.0 64.5 78.5 2V 2 inches 0.08 0.13 0.18 0.24 0.31 0.38 0.46 0.70 0.98 1.31 1.67 2.08 2.53 3.02 3.55 4.72 6.04 7.52 9.14 12.8 17.0 21.8 27.1 33 50.0 70.0 93.0 3 inches 0.06 0.08 0.11 0.13 0.16 0.24 0.34 0.45 0.58 0.72 0.88 1.04 1.23 1.63 2.09 2.60 3.16 4.43 5.89 7.55 9.38 11.4 17.2 24 2 32.1 41.2 51.2 62.3 74.4 87.3 inches 0.05 0.07 0.08 0.12 0.17 0.22 0.29 0.36 0.43 0.52 0.61 0.81 1.03 1.29 1.56 2.19 2.92 3.74 4.64 5.65 8.54 12.0 15.9 20 4 25.3 30.8 36.8 43.2 57.4 73.6 91.4 4 inches 0.07 0.09 0.12 0.16 0.20 0.24 0.29 0.34 0.45 0.57 0.71 0.86 1.21 1.61 2.07 2.57 3.12 4.72 6.61 8.80 11.3 14.0 17.0 20.3 23.9 31.8 40.7 50.6 61.5 5 inches 0.05 0.06 0.08 0.09 0.11 0.15 0.19 0.23 0.28 0.39 0.52 0.67 0.83 1.01 1.53 2.15 2.86 3.66 4.56 5.54 6.11 7.75 10.3 13.2 16.4 20.0 inches 0.06 0.08 0.09 0.11 0.16 0.21 0.27 0.34 0.41 0.63 0.88 1.17 1.50 1.86 2.26 2.70 3.17 4.21 5.40 6.71 * Computed from Williams and Hazen's formula. t To convert head in feet to pressure in pounds per square inch, divide by 2.31. 60 University of California — Experiment Station Scobey's: 13 K S LV l - 9 Hf = TmD^ (6) where Hf is the friction loss in length of pipe, L, in feet; V is the mean velocity in feet per second; D is the diameter of pipe in feet; g is the acceleration of gravity, about 32.2 feet per second per second ; / is the Weisbach friction factor, which varies with smoothness, size of pipe, and velocity, values ranging from about 0.015 for large smooth pipes to about 0.050 for very rough pipes; C is Williams and Hazen's coefficient, which varies with smoothness of pipe, values ranging from about 70 for rough pipes to about 140 for very smooth pipe ; and K s is Scobey's co- efficient of retardation, which varies with smoothness of pipe, values ranging from about 0.30 for smooth pipe to 1.0 or higher for very rough pipe. Friction Loss in Standard Pipe. — Table 5 gives the friction loss in standard iron pipe as computed from Williams and Hazen's formula. The losses are expressed in feet per 100 feet of pipe. This table is based on a coefficient, C = 100, corresponding to iron pipe that has been in use for ten to twenty years. This table is commonly employed for design purposes. Table 6 is based on Williams and Hazen's formula with C = 120. This table is believed safe for use with galvanized pipe under ordinary conditions. Coefficients for new pipe normally range from 120 to 130; for old, badly corroded, tuberculated pipe they may be as low as 50 or 60. These very low values are due largely to a reduction in the cross-section area of the pipe, since the coefficients are computed on the full pipe area. The actual inside diameters and cross-section areas of standard weight pipe are as follows : Pipe size, Inside diameter, Area, square inches inches inches V 2 0.622 0.30 % 0.824 0.53 1 1.049 0.86 1% 1.380 1.49 iy 2 1.610 2.04 2 2.067 3.35 2V 2 2.469 4.78 3 3.068 7.38 3V 2 3.548 9.88 4 4.026 12.72 5 5.047 20.00 6 6.065 28.90 1:5 Scobey, Fred C. The flow of water in riveted steel and analogous pipes. U. S. Dept. Agr. Tech. Bui. 150:1-136. 1930. Bul. 670] Irrigation by Sprinkling 61 The friction loss in pipe corresponding to other values of C can be determined by multiplying the head loss, as given in table 5 for C = 100, by the following factors : C Factor 40 5.40 50 3.61 60 2.58 70 1.94 80 1.51 C Factor 90 1.22 110 0.838 120 0.713 130 0.615 140 0.536 joo 400 600 soo /poo /soo epoo F/otv, go //oris per /nfm/te 5000 7,000 / 0,000 Fig. 24. — Friction loss in welded steel O.D. pipe. The double lines indicate differ- ences in friction loss between pipes of the lightest and heaviest wall as commonly made. Friction Loss in Welded Steel Pipe. — Figure 24 shows the friction loss in welded steel pipe of the type commonly used for main supply lines for portable sprinkler systems. This graph is based on Scobey's formula, with the coefficient K s = 0.32, which corresponds to new pipe in good condition. Since welded steel pipe is relatively thin, it is usually protected against corrosion by galvanizing or with an asphalt or tar coating, with or without a protective wrapping. The effectiveness of these coatings in maintaining a high carrying capacity depends entirely upon the material used and the corrosiveness of the water. In designing stationary pipe layouts for sprinkler systems, one may well add a 62 University of California — Experiment Station certain amount to the required flow, to serve as a factor of safety ; an allowance of 10 to 20 per cent is usual. Friction Loss in Copper Tubing. — Copper tubing (also called copper pipe) is available in both soft and hard temper. It is made in four stand- TABLE 7 Friction Loss in Type M Copper Tubing* Flow, gallons Friction loss, in feet of head, for each 100 feet of tubing following nominal sizesf n the per minute inch inch 1 inch inches . 1H inches 2 inches 1 1.7 6.1 14.0 22.2 34.0 47.0 80.0 1.1 2.4 4.0 6.1 8.5 14.4 21.9 30.5 40.8 52.1 64.8 78.8 0.7 1.2 1.8 2.4 4 1 6.3 8.8 11.8 15.0 18.7 22.7 34.3 48.2 64.0 0.42 63 0.88 1.51 2.28 3.18 4.23 5.43 6.75 8.22 12.4 17.4 23.1 29.6 36.8 44.8 62.9 0.28 0.39 0.66 1.01 1.42 1.87 2.40 2.99 3.63 5.51 7.71 10.2 13.2 16.3 19.8 27.8 37.0 47.3 59.0 71.6 2 3 4 5 6 0.10 8 0.17 10 0.27 12 0.37 14 0.49 16 0.63 18 0.78 20 0.95 25 1.44 30 2.01 35 2.67 40 3.42 45 4.26 50 5.18 60 7.25 70 9.64 80 12.32 90 15.36 100 18.68 * From Williams and Hazen's formula, C = 140. t Actual inside diameters are given in table 8. ardized types O, M, L, and K, classified according to wall thickness. The outside diameters of all types are % inch greater than the nominal sizes. One-inch tubing is, for example, 1% inches O.D., and varies from 1.055 inches I.D. for type M to 0.995 inch for type K. The lightest weight, type O, is available only in sizes above 3 inches. Since the inside diameter of copper tubing is usually less than the inside diameter of iron pipe of the same nominal size, the friction loss is sometimes greater than for iron pipe. This is particularly true of the Bul. 670] Irrigation by Sprinkling 63 heavier-walled tubing, types L and K, and of Vi-inch and l^-inch sizes. For this reason friction loss tables for iron pipe are not suitable for estimating the carrying capacity of copper tubing. Tests to determine friction losses in pipe of various kinds generally indicate a lower value for the exponent of V for smooth pipes, than for rough pipes. Dawson's 1 ' formula for flow of water in copper pipe can be written : TT 0.000315 LV 1 - 75 H ' = ^TS ' ( 7 ) This equation corresponds approximately to Williams and Hazen's formula with C = 140. Tests by A. F. Pillsbury 15 on %-inch copper tubing indicate slightly lower losses than given by Dawson's formula, but they verify this exponent of V. Table 7, computed from Williams and Hazen's formula, with C = 140, gives the friction loss in type M copper tubing, sizes 0.5 to 2 inches. Table 8 compares the inside diameters TABLE 8 Comparison of Inside Diameters and Friction Losses in Copper Tubing, Types M, L, and K* Type M (light) Type L (medium) Type K (heavy) Nominal size, inches Inside diameter, inches Friction factor Inside diameter, inches Friction factor Inside diameter, inches Friction factor y 2 0.569 0.811 1.055 1.291 1.527 2.009 1.00 1.00 1.00 1.00 1.00 1.00 0.545 0.785 1.025 1.265 1.505 1.985 1.23 1.17 1.15 1.10 1.08 1.06 0.527 0.745 0.995 1.245 1.481 1.951 1.52 5i 1.52 1 1.33 \\i 1.19 \y % 1.16 2 1 15 * The friction loss in copper tubing, types L and K, can be found by multiplying the loss given in table 7 by the friction factor given in this table. and friction loss in types M, L, and K. To determine the friction loss in types L or K it is necessary to multiply the friction loss given in table 7 by the proper friction factor from table 8. Friction Loss in Garden Hose and Hydrants. — Garden hose in %- and 1-inch sizes is used to a large extent in connection with sprinkler systems for orchards and also for lawns. In general, friction loss through hose of this type is appreciably higher than for iron pipe and probably varies with different makes of hose. According to recent tests at Davis, 14 Dawson, F. M., and J. S. Bowman. Interior water supply piping for residential buildings. Wisconsin Engin. Exp. Sta. Bul. 77:1-54. 1933. 15 Unpublished data. 64 University of California — Experiment Station the friction loss varies with the pressure, the loss becoming less as the hose swells under pressure. Table 9 gives the friction loss in 100 feet of hose and the loss through garden valves of the type generally used as hydrants for small sprinkler systems. FLOW OF WATER IN PIPES WITH MULTIPLE OUTLETS The friction loss in pipe lines with sprinklers or nozzles spaced equi- distant along the line can be calculated in a step-by-step process, comput- ing the loss for each length between adjacent sprinklers for which the TABLE 9 Approximate Friction Loss in Garden Hose and Garden Hydrants* Friction loss, in pounds per square inch, for:f Flow, gallons per minute 100 feet of %-inch hose 100 feet of 1-inch hose %-inch garden hydrant 1-inch garden hydrant 5 6.8 9.6 12.8 16.0 20.0 24 4 34.0 45.2 58.0 72.0 88.0 1.6 2 2.8 3.6 4.4 5.6 8.0 10.4 13.2 16.4 20.0 24.0 28.0 32.4 37.2 42.0 0.3 0.4 0.6 0.8 1.0 1.2 1.7 2.4 3.1 3.9 4.8 5.8 6.9 8.0 9.2 10.6 1 6 0.2 7 0.2 8 0.3 9 4 10 5 12 0.7 14 0.9 16 1.2 18 1.5 20 1.9 22 2.3 24 2.7 26 3.2 28 3.7 30 4.3 * Friction losses in hose and valves vary widely with different makes of equipment. These values, based on tests at Davis, should be used only as a guide in determining the size required. t To convert pressure in pounds per square inch to head in feet, multiply by 2.31. flow is constant. Such calculations are tedious, especially for long lines. Starting at the distal end of the line with an assumed pressure and sprin- kler discharge, one must work back toward the source to determine the required pressure and the total flow for the line. By making certain approximations, however, one can obtain a simple and fairly accurate solution. Derivation of Equations for Calculating Friction Losses. — If the line is level, the pressure will be a minimum at the distal end of the line, and will increase gradually toward the source. The discharge of the end sprinkler will be a minimum, and, as we proceed toward the pump, each sprinkler in turn will discharge slightly more water because of the in- crease in pressure. The variation in the discharge of the sprinklers is Bul. 670] Irrigation by Sprinkling 65 ordinarily not great, and to simplify the calculation of the friction loss, we will assume that the discharge of each sprinkler on the line is equal to the average discharge of all of the sprinklers, that is qi = Q2 = q* = q a = — (8) when q lf q 2 , and q s are the discharges of the first, second, and third sprinklers respectively, and q a is the average discharge of all the sprin- klers which is equal to the total discharge Q divided by the number of sprinklers, N. A relation between the friction loss in lines with multiple outlets and ordinary pipe lines, where all of the water is carried to the end of the line, can then be derived from a general equation for flow of water in pipe lines. Writing equations 4, 5, and 6 in a generalized form, but and '. - ^ A — D 2 4 F - = K ^' D 2m Making this substitution for V m , and combining K x and K 2 m , we get For pipes with multiple outlets, the total friction loss is equal to the sum of the losses between adjacent outlets. Letting q a equal the discharge at each outlet, and 8 equal the spacing between outlets, the friction loss between the last two outlets at the distal end of the line becomes j, KSqam nu and the loss between the next two outlets is _ KS(2q a )<* _ KSgSV 2 £)2m+n £)2m+n ^ ' and, similarly KSg a m N m k " = D 2m + n • ( 13 ) 66 University of California — Experiment Station The total friction loss for any number of spaces N, between adjacent outlets, becomes # / = 2(/i 1 + /i2+ • • •+h N )=KSq a m 2(l m + 2 m + Z m + • • ■+#*). (14) L Q Substituting — for S, ^ for q a , and 3>N m for 2(l w + 2 W + 3 W + • • • + N m ) we have K (L\/Q m \ _ 2N m / KLQ m \ _ ( KLQ m \ Comparing this equation with equation 10 we see that the friction loss in a pipe with multiple outlets can be determined by first estimating the TABLE 10 Values of the Factor F* by Which the Friction Loss in Pipe Must Be Multiplied to Obtain the Actual Loss in a Line With Multiple Outlets Number of outlets • m = 1.85 to = 1.90 to = 2.00 1 1.0 0.639 0.535 0.486 0.457 0.435 0.425 0.415 0.409 0.402 0.397 0.394 0.391 0.387 0.384 0.382 0.380 0.379 0.377 0.376 0.374 0.372 0.370 0.369 0.368 0.365 0.364 0.361 0.356 0.351 1.0 0.634 0.528 0.480 0.451 0.433 0.419 0.410 0.402 0.396 0.392 0.388 0.384 0.381 0.379 0.377 0.375 0.373 0.372 0.370 0.368 0.366 0.364 0.363 0.362 0.359 0.357 0.355 0.350 0.345 1.0 0.625 0.518 0.469 0.440 0.421 0.408 0.398 0.391 0.385 0.380 0.376 0.373 0.370 0.367 0.365 0.363 0.361 0.360 0.359 357 0.355 0.353 0.351 0.350 0.347 0.345 0.343 0.338 0.333 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 22 24 26 28 30 35 40 50 100 00 2A7m F = Nm+l , as given in equation 14. Bul. 670] Irrigation by Sprinkling 67 friction loss in the line, assuming that all the water is carried to the end of the line, and then multiplying this loss by a factor, F, that depends on the number of outlets on the line and the value of m used in the fric- tion loss formula. Any suitable formula, table or graph may be used, therefore, for estimating the friction loss in lines with multiple outlets. Values of the factor F, have been calculated, and are given in table 10 for m = 1.85, 1.9, and 2.0. Values of this factor for any other value of m can be readily computed from the approximate expression 18 F = -J— + -L+^™E±. (15) m + 1 2N QN 2 K J According to tests on sprinkler pipe with sprinklers uniformly spaced along the line the friction factor K is lower, and the exponent m is slightly higher, than for ordinary pipe flow, because of the partial recovery of velocity head at each outlet. For many problems, a value of m = 2 can be used. Friction Loss in Sprinkler Lines. — Most portable sprinkler pipe is made from lightweight O.D. tubing. A special graph, figure 25, shows the friction loss in 16- and 18-gauge sprinkler pipe in sizes 1 to 6 inches O.D. This graph is based on Scobey's formula, with coefficient K s as given. According to tests on such pipe, higher values of K s may be ex- pected of the smaller sizes, and the values are influenced by the type of coupling. The values given are believed conservative; most sprinkler pipe will have less friction loss than is shown by the graph. Some of the tests indicated that the pressure drop for each section of pipe between adjacent sprinklers is approximately proportional to the square of the velocity (V 2 ) in that section of pipe, as given in Weisbach's formula, instead of some lower power of V as given in Williams and Hazen's or Scobey's formulas. The reason for this is, that the velocity head is par- tially recovered at each sprinkler outlet where the mean velocity in the pipe is suddenly reduced. For estimating the friction loss in a line of sprinkler pipe with sprinklers, it makes little difference whether Weis- bach's or Scobey's formula is used ; but for estimating the loss in a line without sprinklers, Scobey's formula is preferable, unless different values of / are selected for different velocities. Where calculations from a formula are required, Weisbach's formula will be found easier to use. For sprinkler pipe, the following values of / are suggested : 17 18 This expression is exact for m = l, and m = 2, and very nearly correct for m = 3 and intermediate values. Where N is greater than 10, the last term in the expression is negligible and can be omitted. 17 Based upon field and laboratory tests on several makes of sprinkler pipe, and supplemented by published data, principally from Pigott, E. J. S. The flow of fluids in closed conduits. Mech. Engin. 55(8) : 497-501, 515. August, 1933. 68 University of California — Experiment Station 1-inch, 0.028; 1%-inch, 0.025; 2-inch, 0.023; 2%-inch, 0.022; 3-inch, 0.021; 4-inch, 0.020; 5-inch, 0.019 ; and 6-inch, 0.018. To simplify further the problem of determining friction losses in sprinkler pipe, a logarithmic chart, figure 26, has been prepared. This combines data given in figure 25 and table 10. Although based on a sprinkler spacing of 20 feet, it can be used for any spacing if the ^ 6 X •5 0.4 —f- f zl : r t- —7 ^ = ==; / r / / J z_ Z f I i ^ / T -7 ^y / 1 j / f~ A 4- \t / T 7_ / / i -/ 4/ M 7 f / / L , 7 / , V f % / / / / i \ 7 f \ i 7 A ~r ~4 / ;c Ja r / / Y V /-- f t ' ~l X / — > / -f 4 1 —4- 7=- 7- / - r 1 / ' ( -1- / t — f Sco6ey's Forma /a 1 with iro/i/es of / / f -7- / / /f s os g/ven. | / / / ~l_ 1 i f 1 II II /5 X JO 40 SO 60 r~/o»v, ga//o/?s per /nirwle 600 0OO /pOO Fig. 25. — Friction loss in portable sprinkler pipe. For mains and supply lines, where all of the water passes through all of the pipe, this diagram is sufficient. For lines equipped with sprinklers, it must be used in conjunction with table 10. friction loss obtained from the diagram be multiplied by the ratio: & To determine, for instance, the friction loss in a 20 600-foot line of 3-inch pipe with sprinklers discharging 11 gallons per minute, spaced 40 feet apart : 600 1. There will be -^r = 15 sprinklers. 2. Enter left side of diagram for q = 11 ; follow to right to intersection with 3-inch pipe. 3. Drop vertically down to a point half way between the lines for 14 and 16 sprinklers. 4. Follow to the right margin, and read friction loss of 4.1 pounds per square inch. S>?/?£//U/ JP(/ Pi/Of/pS '6 '$jfijPl/3S'/p J2/)/l//J(/s> pfcPJ^/lj/ ^ CD a rt a t-> o ft ^ -M ■+-» fcJD 3 o ft 03 OQ V CD «H r*l o 2 C 1 H3 ft CO w «H a o ft be t» g o 03 3 ft a •rH f-l ft rt t» «*> A M Z CD m fc k> CD pQ r*i fl 3 03 ft fat) OQ 03 bo ^ s o •r-i eS -t-3 on DQ a O o Ph fl «H o rr! CD ej -H M co =+H he o g OQ a O r* § Pi a> o ^ O ■=H k» ■+J ft Rl -u pi a .r-( fl bo r^ fl -^ o ctf rt ft fcJO 02 o hrl o> rW '. PI CO eg ft bo h PR ^J 70 University of California — Experiment Station 40 5. Multiply this by ~~ = 2 to obtain correct loss for 40-foot spacing : 4.1 X 2 = 8.2 pounds per square inch. When the size of pipe is reduced along the line, that is, when two or more sizes of pipe are used, the problem of determining the friction loss is more difficult. The procedure can best be explained by an example. Assume that a 1,000-foot line is composed of 320 feet of 3-inch pipe, 360 feet of 4-inch pipe and 320 feet of 5-inch pipe, with a total discharge of 375 gallons per minute. With sprinklers spaced 40 feet apart, a total of 25 sprinklers, the average discharge of the sprinklers would be 15 gallons per minute. 320 From figure 26, the friction loss in 3-inch pipe with -j- = 8 sprin- klers would be about 1.4 pounds per square inch. This is multiplied by 2 to obtain the actual loss because the sprinkler spacing is 40 feet instead of 20 feet. The corresponding loss for 8 sprinklers, spaced 20 feet apart on 4-inch pipe is less than 1 pound per square inch, and cannot be de- termined from the graph. By extending the lines beyond the lower edge of the graph, this loss is estimated to be 0.5 pounds per square inch which makes the actual loss about 1 pound. The excess loss in the 3-inch pipe, compared with the 4-inch pipe, will be 2.8 — 1.0 = 1.8 pounds. The loss in 680 feet of 4-inch pipe, with 17 sprinklers, is 2 X 2.4 = 4.8 pounds. The actual loss in the combination 3- and 4-inch line is then 1.8 + 4.8 = 6.6 pounds. The corresponding friction loss in 680 feet of 5-inch pipe would be approximately 2X1 = 2 pounds. The excess for the 3- and 4-inch line as compared with 680 feet of 5-inch line is, there- fore, 6.6 — 2=4.6 pounds. The loss in 1,000 feet of 5-inch pipe (25 sprinklers) would be 2 X 2.2 = 4.4 pounds. The total loss in the 1,000- foot combination line is, therefore, 4.4 + 4.6 = 9.0 pounds which is about the maximum permissible for efficient operation. Had 4-inch pipe been used for the entire line, the loss would have been 2 X 7.2 = 14.4 pounds per square inch. Friction Loss in Nozzle Lines. — For nozzle lines of standard galvan- ized pipe, one can estimate the friction loss in a similar manner. Assum- ing that all the water flows to the end of the line, the friction loss can be determined from tables 5 or 6. This loss is then multiplied by the proper factor F (table 10) for m = 1.85 to obtain the actual friction loss in the nozzle line. To determine, for example, the friction loss in 100 feet of 1-inch pipe with nozzles having an average discharge of 0.25 gallon per minute, spaced 3 feet apart, the steps would be as follows : 0.25 X 100 1. Determine the total discharge, which is equal to — — = 8.3 gallons per minute. Bul. 670] Irrigation by Sprinkling 71 2. Using table 6, for C = 120, we find that the friction loss Hf for 8.3 gallons per minute in 100 feet of 1-inch pipe is about 6.0 feet of head. 3. From table 10, m = 1.85, interpolating for 33 nozzles, F = 0.366. 4. The actual friction loss is, therefore, approximately 0.366 X 6 = 2.2 feet. When several pipe sizes are used in a nozzle line, the problem of de- termining the friction loss is essentially the same as for a combination portable sprinkler line. Assume that a nozzle line is composed of 100 feet of 1%-inch pipe, 100 feet of l^-inch pipe, and 100 feet of 1-inch pipe, with nozzles delivering 0.25 gallon per minute spaced 3 feet apart. The total discharge would be 25 gallons per minute; 8.3 gallons per minute at 100 feet ; and 16.7 gallons per minute at 200 feet. The friction loss in the 100 feet of 1-inch pipe, computed above, is 2.2 feet of head. Had this been l^-inch pipe, the loss would have been 0.366 X 1.6 = 0.6 feet. The excess loss for 1-inch pipe, as compared with 1^-inch pipe, is therefore 2.2 — 0.6 = 1.6 feet. For 200 feet of 1%-inch pipe, the loss would be 0.36 X 5.6 X 2 = 4.0 feet; for 200 feet of 1%-ineh pipe, it would be 0.36 X 2.6 X 2 = 1.9 feet. The excess for l^-inch pipe compared with 1%-inch pipe is 4.0 — 1.9 = 2.1 feet. For 300 feet of 1 Ms-inch pipe, the loss would be 0.35 X 5.6 X 3 = 5.9 feet. The actual loss for the combina- tion line would be the loss for 300 feet of l^-nich pipe plus the excess losses due to reducing the size, or 5.9 + 2.1 + 1.6 = 9.6 feet of head. This method of computing losses, though only approximate, is fairly reliable when the average nozzle discharge is known or can be deter- mined. In order to limit the variation in discharge of the nozzles to 10 per cent, nozzle lines should be so designed that the friction loss does not exceed 20 per cent of the average operating head on the line. Discharge, Pressure, and Power Requirement Relations for Sprinkler Lines. — Certain fundamental relations between pressure, discharge, and power requirement for sprinkler lines apply to all types of sprinkler systems. Although the friction loss in sprinkler lines depends only upon the flow of water in the pipe, this depends upon the discharge of the sprinklers, which, in turn, depends upon the pressure. We find, there- fore, a relation between the friction loss and the pressure. To clarify these relations, we will start with the sprinkler-discharge equations 2 and 3, page 52, which show that the discharge of a sprinkler nozzle is proportional to the square root of the pressure at the sprinkler. Expressed mathematically, q = KVP (16) where q is the discharge of the sprinkler, K is a proportionality factor depending upon the size of the sprinkler nozzles, and P is the pressure at the sprinkler. 72 University of California — Experiment Station Since the pressure along the line will vary because of the friction loss, the sprinklers will not all discharge the same amount of water as was assumed for the purpose of estimating the friction loss. However, the total discharge of all the sprinklers is also proportional to the square root of the pressure at any point on the line. If we assume that the exponent m = 2 (equations 9 to 15), it follows that the friction loss will be directly proportional to the pressure at any point on the line. If, for example, the pressure at the pump is doubled, the friction loss will be doubled, and the remaining pressure at the distal end of the line will also be doubled. The ratio of the pressure at any sprinkler to the pres- sure at any other sprinkler will therefore remain constant. For con- venience, the pressure ratio will be defined as the ratio of the pressure at any sprinkler on the line to the pressure at the distal end of the line, and the discharge ratio as the ratio of the discharge of any sprinkler on the line to the discharge of the sprinkler at the distal end of the line. Thus there are definite pressure and discharge ratios for each sprinkler on the line ; these remain constant regardless of the pressure. The dis- charge ratio is equal to the square root of the pressure ratio, thus (17) where q is the discharge of any sprinkler at which the pressure is P, and q is the discharge of the last sprinkler at the distal end of the line where the pressure is P . For pressure ratios less than 1.5, an approximate relation is |=l + 0. 5 (|-l). (18) Or, in words, the relative variation in the discharge of sprinklers along the line is about half of the relative variation in pressure. For example, a 20 per cent variation in pressure along the line will result in approxi- mately 10 per cent variation in discharge of the sprinklers if all have the same nozzle sizes. The average pressure along a sprinkler line, P a , approximates the pres- sure at the distal end, P , plus one fourth the friction loss in the line, P f ; or expressed mathematically, Pa = Po + 0.25P/(approx.) ] or \ (19) P a = Po + 0.25 (P n - P ) (approx.) j where P n is the pressure at the sprinkler nearest the pump. Also, the Bul. 670] Irrigation by Sprinkling 73 average discharge of the sprinklers can be expressed in terms of the discharge of the first and last sprinklers, by the equation % = % + 0.25 (Q n - %) (approx.) (20) or in terms of the pressure at the first and last sprinklers, by the equation % = Q [l + 0.12 (j? - ljj (approx.) . (21) If the discharge of any sprinkler, q ly at a given pressure, P lt is known, then the discharge of the same sprinkler at pressure P 2 , is given by the equation ff, = 8iJp (22) and the average sprinkler discharge corresponding to a pressure P 2 at any point on the line can be expressed in terms of the average sprinkler discharge at pressure P 1 in a similar manner. The total flow is the product of the average sprinkler discharge and the number of sprinklers on the line. Thus Q = Nq a = Nq [l + 0.12 (y - l) J (23) where Q is the total discharge, and N is the number of sprinklers. When the sprinkler lines are so designed that the pressure ratio does not exceed 1.2, the variation in discharge of the sprinklers will not exceed 10 per cent, and the average discharge of the sprinklers will not exceed the minimum discharge by more than 2.4 per cent. Of course, variations in elevation along the line will affect both the pressure and the discharge of sprinklers. The power requirement for pumping water for sprinkling depends upon the total pumping head and upon the total discharge of the line. The total head includes (1) the lift from the water surface in the well or ditch from which the water is pumped, (2) the friction losses in the suction pipe or hose, (3) the velocity head, and (4) the pressure at the pump. Where water is pumped from ditches, the lift and friction losses in the suction hose are small in comparison with the pressure head. For approximate calculations the velocity head may be neglected. The horsepower requirement, hp., is given by the relation where Q is the discharge in gallons per minute, P v is the total pumping 74 University of California — Experiment Station head, including lift and friction losses, expressed as pressure, and E is the efficiency of the pump, expressed decimally. The following example illustrates the use of these equations. A line of 4-inch portable sprinkler pipe has thirty sprinklers, spaced 40 feet apart. The discharge capacity of each sprinkler is 15 gallons per minute at 40 pounds' pressure. The suction lift and the friction loss in the suction hose approximate 3 pounds per square inch. The pump is 40 feet from the nearest sprinkler, and the line is level. Assume a pump efficiency of 60 per cent. What will be the total discharge, the required pressure at the pump, and the horsepower requirement, for a minimum pressure at the distal end of the line of 40 pounds per square inch ? From figure 26, the friction loss for thirty sprinklers discharging 15 gallons per minute would be about 12 pounds per square inch. This must 40 be multiplied by—- = 2 to obtain 24 pounds per square inch, the actual loss. Since this calculation was based upon an average discharge of 15 gallons per minute, it corresponds to an average pressure of 40 pounds per square inch. The pressure at the distal end would be, from equa- tion 19, 24 P = P a — 0.25 Pf = 40 — - — = 34 pounds per square inch. If the pressure at the distal end is increased to 40 pounds per square inch, as specified by the problem, the friction loss will increase in pro- portion, or 40 P f = — X 24 = 28 pounds per square jneh. And the average pressure will be (equation 19) 28 P a = 40 H = 47 pounds per square inch. The pressure at the pump will then be P P = Po + Pf = 40 + 28 = 68 pounds per square inch. The average sprinkler discharge, from equation 22, will be /47 q a = 15 a / — = 16.3 gallons per minute. The total discharge will be Q = 16.3 X 30 = 489 gallons per minute. Bul. 670] Irrigation by Sprinkling 75 The total pumping head (neglecting velocity head, but adding the suc- tion lift) expressed as pressure, is 68 + 3 = 71 pounds per square inch. The horsepower requirement will be (equation 24) 489 X 71 hp. = = 33.7. F 1,715X0.60 An appreciable saving in power could be effected by using a split-line arrangement — that is, where the source (portable pumping plant) is at the center of the line the friction loss would be much less, and the sprin- kler discharge fairly uniform along the line. According to calculations similar to those above, with two 600-foot lines, the total discharge would be 455 gallons per minute, the friction loss 3.4 pounds per square inch, the pressure at the pump 43.4 pounds per square inch, and the power requirement 20.5 horsepower. Summarizing these principles, we can state : The pressure at any point on the line is directly proportional to the pressure at any other point on the line, and a change in pressure at one point will result in a proportionate change in pressure at any other point. The friction loss is directly proportional to the pressure at any point on the line, and a change in pressure at any point will result in a pro- portionate change in the friction loss. The discharge of any sprinkler, and the total discharge, is propor- tional to the square root of the pressure ; a change in pressure at any point on the line will result in a change in discharge proportional to the square root of the change in pressure. The power requirement for pumping, being proportional to the prod- uct of the pressure and discharge, is proportional to the three-halves power of the pressure ; a change in pressure at any point will result in a change in power requirement that is proportional to the three-halves power of the change in pressure. DISTRIBUTION OF WATER FROM ROTATING SPRINKLERS The purpose of a sprinkler is to distribute the water to the soil in the form of a sprinkle or spray so that it can be absorbed without running off. Preferably, the water should be distributed uniformly over the area. Since nearly all rotating sprinklers cover circular areas, an absolutely uniform application is not possible. The degree of uniformity obtainable depends primarily upon the type of distribution pattern produced and upon the spacing of sprinklers. Numerous other factors, such as wind, pressure, and uniformity of rotation of the sprinkler, affect distribution. 76 University of California — Experiment Station Many tests have been made to determine the uniformity of distribu- tion. Manufacturers generally have facilities for testing sprinklers and making adjustments. A series of tests on both American and German sprinklers was conducted by Staebner. 18 These tests were made by catch- ing the water in receptacles placed at various distances from the sprin- kler. After the sprinkler had been operated for a specified time, the amount of water caught in each receptacle was measured to determine the depth. For most of his tests, Staebner used 112 cans and rain gauges distributed in a geometrical pattern over a square area, 100 feet on each side. Extensive tests have also been made in Germany, 19 using a large number of receptacles placed over the entire area covered. Some manufacturers and experimenters have apparently been striv- ing for a sprinkler that distributes a uniform amount over a large por- tion of the area covered, with a rather abrupt breaking off at the edges. Others have given more consideration to the effect of overlap and have tried to obtain a different pattern. Staebner judged the sprinklers tested on their ability to distribute water so that the maximum depth was not more than twice the minimum, except near the edges of the area covered ; but he did not discuss the question of overlap nor of proper spacing for such sprinklers. He states, "No matter how successfully they may dis- tribute water over a circular area, they leave much to be desired, because if circles just touch one another a considerable area is left unwatered, and if they overlap a great amount of double coverage results." He fur- ther concludes, "More uniform distribution over a large area can be obtained with the overhead-pipe system (nozzle lines) than with any other type of spray irrigation equipment now available." Before 1932 slow-revolving sprinklers were used mostly for stationary overhead orchard systems and for irrigating large parks, golf courses, and the like. In such installations, the common practice is to place the sprinklers as far apart as possible and still cover all the ground. Sprin- klers are generally mounted in either a square or a triangular arrange- ment, the distance between them being as much as 80 to 90 feet. In 1932, when portable systems first came into general use, the sprinklers, cover- ing 100 to 120 feet circles, were usually spaced 40 feet apart on the pipe line, which was generally moved 50 to 60 feet for each setup. This pro- vided more than a double overlap in both directions. With such close spacing it was assumed that the resulting distribution would be fairly uniform. "Staebner, F. E. Tests of spray irrigation equipment. U. S. Dept. Agr. Cir. 195: 1-29. 1931. 19 Studiengesellschaft fur Feldberegnung. Die Feldberegnung. RKTL Schriften heft 13:1-177, 1930; heft 30:1-176, 1932; heft 38:1-177, 1933. Paul Parey, Berlin, Germany. Bul. 670] Irrigation by Sprinkling 77 SPRINKLER TESTS AT DAVIS To obtain definite information about the distribution of water under such conditions, and especially to determine the effect of wind, speed of rotation, and spacing of sprinklers upon the distribution, a series of tests on commercial sprinklers was made at Davis. For the first 122 tests, small rain gauges made from no. 2% tin cans were placed 10 feet apart in each direction over the entire area covered by the sprinkler except on the north-south and east-west axes, where they were 5 feet apart. For sub- //' p/pe nvafer //ne & J i'p'pe pressure //ne ** I ° °/ " ° o ° ° ° / ° . . o o . . . / . O . o " o " . ' o " I " . " /" . " o . 0.0 i o / /CanJ . » o o 00/ ..jjL 00.0 00 00 0.0 ooo.ooo.o.o.^: « ^JprinMler o o o o o 000 'ZM-y////: Ud Pig. 27. — Arrangements of cans for sprinkler tests at Davis. Left, the layout for tests 1 to 122 j and right, for tests 123 to 170. sequent tests an additional can was placed in each square, one can being provided for each 50 square feet. Figure 27 shows the arrangements. Water was supplied from the domestic system under a pressure of about 40 pounds per square inch. To increase this pressure when desired, a booster pump was provided. A 1-inch calibrated water meter measured the discharge from the sprinkler. There was a return pressure line from the base of the sprinkler riser, and a calibrated pressure gauge was in- stalled near the pump and water meter. By means of a valve at this point it was possible to regulate and maintain at the sprinkler a constant pres- sure of any desired amount up to the maximum available. The arrange- ment proved highly satisfactory. A standard Weather Bureau type four-cup anemometer was installed about 10 feet above the ground near the pump house, and the wind veloc- ities given are those obtained at this location. The humidity and air temperature were determined with a sling psychrometer, readings being taken at intervals of 5 to 10 minutes during the tests. 78 University of California — Experiment Station About 130 tests on slow-revolving sprinklers have been made with these facilities. Some additional work with these sprinklers was done at Para- dise, California. Lawn-sprinkler heads and small whirling sprinklers have also been studied. Some of the slow-revolving sprinklers are shown in figure 28. For most of these tests the sprinkler was operated for one hour. The water caught in the cans was measured to the nearest cubic centimeter, equivalent to % o inch in depth. All tests were plotted as in V < Fig. 28. — Some sprinklers of the slow-revolving type used on portable sys- tems that were used in the tests for distribution of water. Some of these have been replaced by newer models. figure 29 showing contours representing points of equal depth, and cross sections in both the north-south and east-west directions. From these tests a few have been selected to illustrate typical patterns for some of the sprinklers tested; and to illustrate the effect of wind, insufficient pressure, rapid rate of rotation, and variations in the rate of rotation. Typical Distribution Patterns for Favorable Conditions. — Figures 29 to 33 illustrate typical distribution patterns for slow-revolving sprin- klers operating under favorable conditions. For all but one of these tests the wind velocity averaged less than 3 miles per hour. The speed of rota- tion did not exceed one revolution per minute, and the pressure was ample for proper distribution. The essential data for each test are given in the figures and legends accompanying them. For four of the five patterns , the average diameter covered exceeds 120 feet. These distribu- Bul. 670] Irrigation by Sprinkling 79 0ept/>, inches £. W. Sec £ '/on SO JO 40 30 20 /o o /o to JO D/sto/Jce fro/n spri/?A/er, feet 40 JO 60 Fig. 29.— Test 16. Sprinkler, B-3 ; nozzles, % 2 and % G in. Test data: pressure, 45 pounds per sq. in.; discharge, 19.7 gal. per min.; average rate of rotation, 0.7 r. p. m. Condi- tions for test favorable. SO JO 40 JO 20 /O O /O tO JO ¥0 P/stonce /ro/77 spr/nfc/er, feed Fig. 30.— Test 90. Sprinkler, F-l ; nozzles, % and % in. Test data : pressure, 40 pounds per sq. in. ; discharge 14.3 gal. per min.; average rate of rotation, 1.0 r. p. m. Con- ditions for test favorable. 80 University of California — Experiment Station tion patterns are typical of a large number obtained and illustrate the uniformity of distribution obtainable with sprinklers operating under favorable conditions — that is, adequate pressure, low wind velocities, and proper speeds of rotation. Particular attention is called to figures 31 and 33, for which the cross sections are somewhat triangular. As will be shown later, this type of pattern produces the most uniform distribution DefM, /nc/>es J \ £. VSe ct/or SO 40 SO 20 /O O /O 20 30 40 SO 60 £>/'s£a/7ce /ro/7? spr/rtft/er, / of /na/'/r /7o*z/e Fig. 31. — Test 132. Sprinkler, B-4; nozzles, % 2 and Vs in. Test data : pressure, 50 pounds per sq. in. ; discharge, 12.0 gal. per min. ; average rate of rotation, 0.29 r. p. m. Conditions for test favorable. Lower profile shows variations in average rate of rotation for different positions of the main nozzle. over a large area when the sprinklers are properly spaced. Figures 31 and 32 include a graphic record of the average rate of rotation as deter- mined for each 30-degree angle of rotation. The variations in the rate of rotation were determined for most of the tests subsequent to test number 52. There was evidence that variations in rate of rotation were largely responsible for uneven distribution. Effect of Low Pressure on Distribution from Sprinklers. — To deter- mine the effect of low pressure, several tests were made at pressures Bul. 670] Irrigation by Sprinkling 81 inadequate for proper distribution. Figures 34, 35, 36, and 37 illustrate typical patterns at pressures of 20 pounds per square inch. The most noticeable feature is the ring near the outside edge of the area covered, where the depth of application was several times greater than at a dis- tance of 10 to 20 feet from the sprinkler. This doughnut or ring-shaped pattern resulting from low pressure is one that generally produces a ^- 60 SO 40 30 20 /O O /O 20 SO 40 SO 60 ^ 0/stance /row s/>r//7/c/er, feet /Pot at /on of jpr/n/c/er £ S W Direct /'or? of /na/'n nozz/e Fig. 32.— Test 136. Sprinkler, C-l ; nozzles, % 2 and % in. Test data : pressure, 50 pounds per sq. in. ; discharge, 19.3 gal. per min. ; average rate of rotation, 0.22 r. p. m. Condi- tions for test favorable. very uneven distribution for all reasonable spacings of sprinklers. The effect of spacing on the distribution is discussed later in more detail. Another noticeable feature of the lower pressure is the smaller area covered. The average diameter of the wetted area is 90 to 100 feet, as compared with diameters of more than 120 feet for pressures of 40 to 50 pounds and otherwise similar conditions. Judging from several tests on the same sprinkler at different pressures, the reduction in area covered is roughly proportional to the reduction in discharge of the sprinkler caused by the lower pressure, while the average actual rate of applica- 82 University of California — Experiment Station Depih, //k/?ss V ft 60 so 40 so eo /O O /O SO 30 D/s/a/?ce /ram spr/nA/er, feet SO 60 70 Fig. 33.— Test 170. Sprinkler, G-3; nozzles, % and % 2 in- Test data : pressure, 40 pounds per sq. in. ; discharge, 19.3 gal. per min. Conditions for test favorable. Pe/>M, mc/res fO 40 so zo /o o /o to so 40 so £>/'si<7/?ce /ram s/>r/'/?A/er t /eet Fig. 34. — Test 11. Sprinkler, B-l ; nozzles, % 6 and % 2 in. Test data: pressure, 20 pounds per sq. in.; discharge, 14.2 gal. per min.; average rate of rota- tion, 0.6 r. p. m. Pressure inadequate for good distri- bution. Bul. 670] Irrigation by Sprinkling 83 \£ IV. J ecf/o/? / \ J \ / \ so je 40 jo 0/sfortce from spr/n/t/er, fee/ Fig. 35.— Test 31. Sprinkler, A-2 nozzles, % and %6 i n - Test data : pressure, 20 pounds per sq. in. ; discharge, 9.6 gal. per min.; average rate of rota- tion, 0.7 r. p. m. Pressure inadequate for good dis- tribution. £>epl/>, //?c/>es Til JO 20 /O O /O ZO SO fO /)/s/ trace /re/n sfr/'/rA/er, Zee/ £ S f>/rec//b/? of tf?a//7 /tozz/e Fig. 36.— Test 102. Sprinkler, G-l ; ... and % 2 in - Test data : pressure, 20 pounds per sq. in.; discharge, 16.2 gal. per min.; average rate of rotation, 0.24 r. p. m. This sprinkler is equipped with a deflector on the tail nozzle which breaks up the jet and improves the pattern. Note the uniform rate of rotation. 84 University of California — Experiment Station t>e/>t/>, /nc/res «4 55 k° £. M Sect /on > jo to so 20 /o o /o eo jo 40 so 0/stance from spr/nfi/er, feet Direction of /na/n /tozz/e Fig. 37.— Test 140. Sprinkler, C-l; nozzles, % 2 and % in. Test data: pressure, 20 pounds per sq. in.; discharge, 12.0 gal. per min.; average rate of rotation, 0.29 r. p. m. Pressure inadequate for good distribution. Pep t/?, /notes Jo 20 /o o /o to jo //', McAes *<.ss £Hf Jec t/o/> 1 & so a o /o eo 30 spr/nA/er, feet flotation of jprinA/er t=tt A/ £ S W W D/rect/b/7 of /na/n /?ozz/e Fig. 40.— Test 100, Sprinkler, G-l; nozzles, % 6 and %2 i n « Test data: pressure, 30 pounds per sq. in.; discharge, 19.6 gal. per min.; average rate of rotation, 0.20 r. p. m. This pattern illustrates the effect of wind. of pressure to any point for which reasonable distribution of water can be obtained. This reduction in area results in a corresponding increase in the actual rate of application, which exceeded 1 inch per hour in a localized area for one of the tests (fig. 43). This may be compared with an average rate of application of less than % inch per hour for the same sprinkler when rotating slowly. Obviously, sprinklers must be spaced much closer together, when rotating rapidly, to secure the same uni- formity of distribution. In addition, with portable systems, the lines must be moved more frequently. Bul. 670] Irrigation by Sprinkling 87 / \, f.W. Ject/on ,r~ J ~^ \ s 40 JO SO X? O /O SO JO 40 D/sio/xre from jpr/rA/er, feet 20 \ 1 1 1 — Potof/on of jpr/nA/er £ S h/ f//rect/b/? of mo/'/? /tozz/e Fig. 41.— Test 112. Sprinkler 1-1 ; nozzle, % 6 in. Test data: pressure, 40 pounds per sq. in.; discharge, 22C7 gal. per min. ; average rate of rotation, 1.3 r. p. m. This pattern illustrates the effect of wind. flept/!, //)c/>es 43' 4J' & 40 JO go /O O /O SO JO 40 SO f)/s&ooce fro/7? spr/aA/er, feet Fig. 42. — Test 21. Sprinkler B-2 ; nozzles, % 2 and %2 in. Test data: pressure, 40 pounds per sq. in.; discharge, 17.5 gal. per min.; average rate of rotation, 13.6 r. p. m. This pattern illustrates the effect of high speed of rotation. Note the reduction in area covered and the cor- responding increase in rate of application. 88 University of California — Experiment Station Depi/>, /ncAes %.so ; V / ' — *> U \^ \ / \ -* / £. JKJt *ct/a , \ ^ 30 so /o o /o SO SO 0/sfo/?ce /ro/n s/>r//7A/er, feet Fig. 43. — Test 118. Sprinkler, 1-1; nozzle, % 2 in. Test data : pressure, 46 pounds per sq. in. ; discharge, 22.1 gal. per min. ; average rate of rotation, 26.3 r. p. m. This pattern illustrates the effect of high speed of rotation. Dep£/>, /aches / £.tf. Sect/ov i i i V. SO 20 /O O /O tO JO fO 0/sfo/?ce /ra/7? spr//rfr/er t feet Fig. 44. — Test 149. Sprinkler, L-l ; nozzle, % in. Test data : pressure, 50 pounds per sq. in. ; discharge, 17.0 gal. per min.; average rate of rotation, 51.0 r. p. m. This pattern illustrates the effect of high speed of rotation. Bul. 670] Irrigation by Sprinkling 89 Slow-revolving sprinklers are designed to be operated at speeds of about one revolution per minute, and should never rotate faster than three or four revolutions per minute. They are sometimes provided with a device for regulating the speed. Sometimes, however, they rotate very rapidly ; in one instance a count showed a sprinkler on a portable system to be averaging 90 revolutions per minute. Rotation speeds of 20 revolu- tions per minute are common on many systems. Some sprinklers are so designed that the speed of rotation can be adjusted only within certain limits, or not at all, and high speeds are impossible. Excessive speeds increase the wear on the bearings and spindles, which in turn may cause the sprinkler to rotate unevenly. SO* o/je eomp/ete revo/e/t/o/7 I =4 " — -y~ t N — M — * — " — "— » — " — h ' \ /m/'/?i//e I *dbt/6/e /narA-sna//? />ozz/e po//7//'/?q- /tort/? *" Fig. 45. — Kecorder chart showing how variation in rate of rotation of sprinklers was determined. The pen makes a mark for each 30 degrees of rota- tion, a double mark for each revolution of sprinkler. Uneven spacing of marks shows variation in rate of rotation. Effect of Variation in Bate of Rotation on Distribution Patterns. — Before very many tests had been made, it was apparent that some of the unevenness in distribution was caused by the sprinkler's rotating at a variable rate through each revolution. With rates less than one revolu- tion per minute, this variation is not noticeable to the eye; but by timing the sprinklers through consecutive quadrants with a stop watch, it was found that there were appreciable variations in the average rate for the different quadrants. The rate was always slow or fast (as the case might be) for the same position of the sprinkler, that is, when a nozzle was pointing in the same direction, so that more water was being thrown in one direction from the sprinkler than in another. To investigate this point thoroughly, an anemometer recorder was converted into a rota- tion recorder by changing the gear ratio so that the drum turned at a peripheral speed of 1 inch in 50 seconds. A commutator arrangement was attached to the sprinkler so that an electrical circuit was shorted for each 30 degrees, or 60 degrees, of rotation as desired. For speeds ex- ceeding about 1 revolution per minute, 60-degree intervals were used; for speeds of less than this amount, 30-degree intervals. This recorder operated successfully only when the sprinkler rotated at a speed of less than three revolutions per minute. The data were tabulated by measur- ing the number of seconds for each point of contact on the chart with a scale having 50 graduations per inch. The recorder was generally 90 University of California — Experiment Station operated 30 to 45 minutes during a test ; and the data were tabulated for this period, or up to a maximum of 30 or 40 revolutions. The average rate of rotation for each 30- or 60-degree angle was calculated and plotted graphically as illustrated. These determinations were made for most of the tests subsequent to test no. 52, except those for which the rate of rotation was too high to 0epl/) t /fie/res 60 SO 40 SO ZO /O O /O ZO JO 40 SO GO £/sfo/Jce fro/n s/>r//7A/er, feet £ S W 0/rect/o/? of /now nozz/e Fig. 46. — Test 103. Sprinkler, L-l ; nozzles, % 6 and !^ 4 in. Test data : pressure, 40 pounds per sq. in. ; discharge, 22.5 gal. per min. ; average rate of rotation, 0.68 r. p. m. This pattern illustrates how a large variation in rate of rotation affects the distribution of water. Both nozzles of the sprinkler were point- ing in approximately the same direction. be determined. The apparatus for these tests and the method of analyzing the data proved very satisfactory, presenting definite evidence that uneven distribution in many tests was due largely to variations in the rate of rotation. Figure 45 is a typical record of the rate of rotation as obtained with the recorder. The variation in rate is indicated by the difference in spacing between the marks on the lines. Figures 46 and 47 illustrate how extreme variations in the rate of Bul. 670] Irrigation by Sprinkling 91 rotation affect distribution. For figure 46, both nozzles of the sprinkler were discharging in approximately the same direction; and the local areas of high concentration to the southwest of the sprinkler are due to the very slow rate of rotation of the sprinkler when discharging in that direction, as indicated by the graphic record of the rotation. For figure 47 the two nozzles were discharging in approximately opposite direc- Dep£/! t //tehes <0 JO £0 /O O /O 20 JO 40 JO 60 0/st<7/?ce /row jpr/Wfr/er, feet £ S 0/rect/b/? of m<7//? /?otz/e Fig. 47.— Test 115. Sprinkler, 1-1; nozzle, % 2 in. Test data: pressure, 50 pounds per sq. in.; discharge, 22.7 gal. per min. ; average rate of rotation, 0.80 r. p. m. This pat- tern illustrates the effect of a variable rate of rotation. The two nozzles on this sprinkler were pointing in opposite directions; the high rate of application to the northeast of the sprinkler is due to the tail nozzle, while that to the southeast is due to the main nozzle. tions ; the high concentration to the northeast of the sprinkler is from the tail nozzle, whereas that in the southwest direction is from the main nozzle. That the lowest rate of rotation occurred for both tests (different sprinklers) when the main nozzle was pointing in a southwesterly direc- tion was apparently accidental. Other tests indicated that the positions of maximum and minimum rates are apt to occur in any direction and that the variation is not due to wind, as some sprinkler users suppose. 92 University of California — Experiment Station Judging from tests on several different makes of sprinklers during strong winds — in one instance averaging more than 25 miles per hour — the positions of high and low speed bore no relation to wind direction, but could be altered at will by rotating the base of the sprinkler. One sprinkler was operated for a given period; then the base was oriented 180 degrees, and the test repeated. This caused the positions of high and low rates of rotation also to be oriented about 180 degrees, though the direction of the wind remained approximately the same (fig. 48). Fig. 48. — Variation in rate of rotation of sprinkler illustrated by plotting the rate as a radius. The similarity in shape and opposite orientation of the two diagrams indicates that the variation in rate of rotation was not caused by the wind. Variations in rate of rotation is undoubtedly due to a variation in the friction on the bearing when the sprinkler is in different positions. Since the driving force is relatively constant, a variation in speed results when the frictional resistance varies. When operating under adequate pressure the frictional resistance for most sprinklers is suffi- cient so that the unbalanced torque, due to the wind reaction on the sprinkler, has little effect on the speed of rotation. Whether the varia- tion in frictional resistance results from a lack of precision in manuf ac- Bul. 670] Irrigation by Sprinkling 93 ture, from variations in thickness of the leather washer in the bearing, or from other causes was not determined. All sprinklers tested were new and unworn. Field observations indicate that worn sprinklers fre- quently vary more in speed of rotation than new sprinklers and that there is a great difference in the performance of sprinklers of the same kind on the same line. DESIRABLE TYPES OF DISTRIBUTION PATTERNS AND PROPER SPACING OF SPRINKLERS The distribution patterns in the foregoing figures immediately suggest two questions : What is the most desirable type of distribution pattern ? Air /no/e/neni yar/ab/e 0.45 m.p./?. U^lO.Z-0.4 metes & E23 0.4 -0.6 • 1ZZA0-6-0.8 * ■H over i.0 ' Fig. 49. — Kesulting distribution of water when sprinkler patterns are overlapped to correspond to a spacing of 40 by 50 feet. A, Sprinkler pattern for test no. 16 (fig. 29) illustrates favorable conditions (pressure, 45 pounds per sq. in.). B, Pattern for test no. 11 (fig. 34) illustrates inadequate pressure (20 pounds per sq. in.). What is the proper spacing of sprinklers for the most uniform distribu- tion ? All the sprinkler tests and many geometrical patterns were ana- lyzed for the purpose of answering these basic queries. To determine the uniformity of distribution for a group of sprinklers, some of the patterns were overlapped at a spacing of 40 feet in one direc- tion by 50 feet in the other (a common arrangement for portable systems) and the resulting depth of application was calculated for points spaced 10 feet apart both ways. The application from all sprinklers affecting the net area was included. Figure 49 shows the results for two tests previously illustrated. For test no. 16, with the sprinkler operating under favorable conditions, the resulting distribution is remarkably uniform; but for test no. 11 (inadequate pressure) the variation is from about 0.2 inch to more than 1.0. This analysis, though it serves 94 University of California — Experiment Station to illustrate the distribution obtained for specific tests and spacings, does not answer the principal questions involved. A Method of Analyzing Sprinkler Tests for Uniformity of Distribu- tion. — To compare sprinkler patterns and to determine how various spacings affect the resulting distribution of water, one needs a numerical expression to serve as an index of the uniformity secured. For this pur- pose an expression called the uniformity coefficient (C u ) was adopted. The uniformity coefficient expressed as a percentage is defined by the equation , . C u = 100(1.0- — ) (25) \ mn) in which x is the deviation of individual observations from the mean value m, and n is the number of observations. An absolutely uniform ap- plication is then represented by a uniformity coefficient of 100 per cent ; a less uniform application, by some lower percentage. When the intensity of application at any number of equally spaced points over the entire area covered by a sprinkler is determined, the uniformity coefficient can be computed for any spacing (in either direc- tion) that is a multiple of the spacing of the points of observation. Thus a complete analysis of one sprinkler pattern to determine the best spacing and the resulting uniformity of application involves numerous computations. Uniformity coefficients have been determined both for actual sprinkler patterns and for various geometric patterns. The depths of application at uniformly spaced points over the net area covered by a sprinkler are first determined. Sufficient points are used so that the depth at any particular point may be considered the mean for the unit area repre- sented by that point. For the actual sprinkler patterns, the amount of water caught in each of the cans spaced 5 or 10 feet apart in parallel rows is taken as an individual observation. The pattern is then over- lapped on itself to correspond to any desired spacing, and the total amount for each point within the net area covered by one sprinkler is determined and tabulated. The mean depth of application is next de- termined, and the deviation from the mean at each point is calculated. About half of the points will have applications of more, and half of less, than the mean. These deviations are then totaled, and the uniformity co- efficient is computed from equation 25. For each sprinkler pattern, a different value of the uniformity coefficient is obtained for each spacing ; and since the spacing may be different in the two directions, many tedious calculations are required to analyze one sprinkler pattern com- pletely and determine what spacing will give the best results and how uniform the distribution will be. Bul. 670] Irrigation by Sprinkling 95 A short-cut method of determining the optimum spacings has been employed in analyzing all the sprinkler tests. This method is equivalent to spacing the sprinklers closely along the pipe line and then determin- ing the uniformity coefficients for different spacings between lines. ($ x is used to denote the spacing between sprinklers along the line, and S 2 the spacing between lines.) When sprinklers are close together (say, S 1 = 10 feet), a strip of ground will be wet so that there will be little variation along lines drawn parallel with the line of sprinklers. The profile of water concen- tration across the wetted strip along any line drawn at a right angle to the pipe line can be determined by overlapping a sprinkler pattern upon itself corresponding to the designated sprinkler spacing (&J. This is done by summing up the water caught in the cans in each of the parallel rows. The extra cans on the two diameters are omitted so that all cans used are equally spaced over the area. The tabulated sums are then combined, corresponding to various spacings (S 2 ) between lines, and the uniformity coefficient calculated. The coefficients thus determined represent a measure of uniformity in only one direction, not a measure of the uniformity for the net area covered by sprinklers spaced normal distances along the line. Since the actual sprinkler patterns are not symmetrical, because of wind and other influences, the uniformity co- efficients were determined for two directions normal to each other, and the mean value for each spacing was used. These analyses indicate the optimum spacings, which apply both along the lines and between lines. Distribution for Geometrical Patterns. — Before discussing the results of the analyses of the actual sprinkler tests, we may well consider the distribution for certain geometrical patterns. If a sprinkler is rotated at a uniform speed in perfectly still air, the resulting pattern will be symmetrical about the center. Several symmetrical patterns represent- ing different geometric shapes have been analyzed to determine the uniformity of distribution for various spacings. Six of these patterns, some being typical of actual sprinkler patterns, are shown in figure 50, together with the uniformity coefficients for different values of S 2 when S 1 equals 5 per cent of the diameter covered. The distribution for pat- tern B is nearly uniform for all values of S 2 up to 55 per cent of the diameter. Patterns A and C give fairly uniform applications for all spacings up to 65 per cent of the diameter; beyond this the uniformity drops off rapidly. Patterns D and E produce a fairly uniform distribu- tion for spacings of 75 and 80 per cent; but for spacings of 45 to 70 per cent there is an appreciable variation. Pattern F results in poor distribution for spacings of 50 to nearly 80 per cent, with only fair uniformity for spacings between 80 and 85 per cent of the diameter. 96 University op California — Experiment Station The uniformity coefficients for various spacings in both directions for patterns B and E, shown in figure 50, are given in table 11. This table illustrates how the uniformity coefficients vary with S x as well as with 8 2 . For pattern B the uniformity coefficients remain constant as 8 X is increased from 5 up to about 60 per cent of the diameter. For pattern E, however, with the spacing 8 X equal to 5 per cent of the diameter, (0.05 D), the uniformity coefficient drops from 97 when S 2 is 0.4 D, to 20 30 40 50 60 70 0O Spacing be £ ween //nes, 5 zt % of d/a/neter Fig. 50. — Geometrical sprinkler patterns and uniformity coefficients for different spacings of lines with sprinklers spaced closely along the lines. The higher the uni- formity coefficient, d, the more uniform the application. 80 when S 2 is 0.6 D, and back up to 96 when S 2 is 0.8 D. For a spacing between lines, S 2y of 80 per cent of the diameter, the uniformity coeffi- cient drops from 96 when 8 X is 0.05 D to 68 when 8 t is 0.6 D, then rises to 74 when S x is 0.8 D for a square arrangement, and to 83 when 8 X is 0.8 D for a triangular arrangement. In general, these and other patterns studied in detail indicate that the analyses for S ± equals 0.05 D or 0.1 D (the short-cut method) are reliable for determining optimum spacings, and that fairly uniform distribution results when 8 X does not exceed half the optimum values of 8 2 . If, for example, a spacing of 0.05 D by 0.8 D shows a high uniformity coefficient, a spacing of 0.4 D by 0.8 D is likely to give good results. When a fairly close spacing between pipe lines is permissible, as it sometimes is on portable systems, excellent re- sults can be obtained with patterns similar to D and E when the spacing Bul. 670] Irrigation by Sprinkling 97 between lines, S 2 , is also half the maximum optimum spacing as indicated by this method of analysis. This spacing provides in effect, a double over- lap in both directions. Desirable Patterns for Square and Equilateral-Triangle Arrange- ments of Sprinklers. — On stationary sprinkling systems, for economic TABLE 11 Uniformity Coefficients for Various Arrangements and Spacings of Geometrical Patterns B and E* Spacing between lines, S2, in per cent of the diameter ,D Spacing along line, Si, in per cent of diameter, D 40 per cent 50 per cent 60 per cent 70 per cent 80 per cent Pattern B 5 99 99 98 97 98 98 97 96 96 93 93 93 93 93 91 92 80 80 80 SO 80 80 80 77 78 67 20 67 30 67 40 67 50 67 60 67 60 Tf 67 70 65 70 T 65 80 58 80 T 60 Pattern E 5 97 96 93 96 84 84 83 83 76 80 80 80 77 77 75 67 86 86 85 83 82 71 66 66 65 96 20 94 30 93 40 88 50 81 60 68 60 T 70 70 67 70 T 73 80 74 80 T 83 * Shapes of patterns are shown in figure 50. t T denotes a triangular arrangement of sprinklers; all others are rectangular. reasons, sprinklers are usually spaced as far apart as possible in either a square or an equilateral triangle arrangement. One should know what patterns will give most uniform application, and how far apart sprin- klers can be placed with satisfactory results. To answer these questions, a set of patterns that give as nearly uniform applications as possible for different spacings have been developed. As previously shown, a conical pattern (pattern B in fig. 50) with a triangular cross section gives nearly uniform applications for spacings 98 University of California — Experiment Station up to about 55 or 60 per cent of the diameter. Slightly modified patterns of this general shape result in even greater uniformity for similar spac- ings. Figure 51 shows a series of curves illustrating half cross sections of patterns, as determined by trial and error methods, that give maxi- mum uniformity coefficients for spacings of 50 to 75 per cent of the diameter when the sprinklers are arranged in a square. When the spacing exceeds 71 per cent of the diameter, the application at the center of the square between four sprinklers will be zero regardless of the shape of the 20 30 40 £0 60 70 P/sfo/rce /row spr/rtA/er, % of rad/'i/s covered Fig. 51.— ■Half cross sections of patterns that give maximum uniformity coefficients for different spacings of sprinklers when arranged in a square. pattern. A spacing of 71 per cent of the diameter is therefore the maxi- mum possible if all the area is to be covered. Detailed analyses of three of these patterns, G, I, and K, have been made for square spacings (S 1 =S 2 ) from 35 to 75 per cent of the diameter. Table 12 gives the uniformity coefficients for these patterns both for square spacing and for a very close spacing along the line (8 ± = 0.05 D) for various spacings between lines. Data in table 12 are shown graphi- cally in figure 52 which illustrates the relation between the uniformity coefficients obtained for square spacing and for close spacing along the line. In addition, the uniformity coefficients are shown for pattern K for $ x = 0.4 D. For pattern G the curves are similar, the close spacing showing slightly higher uniformity coefficients for all spacings between lines. For pattern I there is somewhat more difference between the curves, 60 per cent spacing between lines (S 2 = 0.6 D) giving the highest uni- formity coefficients for close spacing, whereas for the square arrange- Bul. 670] Irrigation by Sprinkling 99 ment a 55 per cent spacing (S 1 = S 2 = 0.55 D) gives the best results. For pattern K, the most desirable pattern for a spacing of 70 per cent (S t = S 2 = 0.7 D), there is still more difference. When S 1 = 0.05D much better results are obtained for values of S 2 between 60 to 70 per cent of the diameter than for spacings of 45 to 60 per cent, whereas for a square arrangement the uniformity coefficients are fairly constant TABLE 12 Uniformity Coefficients for Patterns G, I, and K* Spacing be- tween lines, S2, in per cent of diameter Net area, in per cent of circular area covered Uniformity coefficients, in per cent, for the patterns: For square arrangement of sprinklers (Si = S?) 35 15.4 99 97 97 40 20.1 96 96 95 45 25.4 97 92 87 50 31.8 99 1 98 86 55 38.5 93 96 87 60 45.6 85 93 87 65 53.8 78 87 85 70 62.4 69 80 86 75 71.6 72 80 For close spacing along the line (Si = 0.05Z)) 50 3.2 100 94 89 55 3.5 95 97 90 60 3.8 89 98 93 65 4.1 82 92 97 70 4.5 74 85 94 75 4.8 78 87 80 5.1 80 * Shapes of patterns are shown in figure 51. t Italicized figures correspond to the spacing for which the pattern is designed. for all spacings between 45 and 70 per cent of the diameter. The curve for a 40 per cent spacing along a line (S x = 0.4 D) is very similar to that for close spacing (# 1 = 0.05 D) indicating that for customary spacings of sprinklers on portable systems, the results of analyses for close spacings along the line show what spacings between lines, S 2 , will give the best distribution. For an equilateral-triangle arrangement of sprinklers, patterns of a slightly different shape give better distribution. Figure 53 shows half cross sections of nine patterns that give the highest uniformity coeffi- cients for spacings between sprinklers, 8 lf of 50 to 90 per cent of the diameter. For this arrangement the spacing between parallel rows of sprinklers, S 2 , is only 86.6 per cent of the spacing between sprinklers. 100 University of California — Experiment Station 45 SO 55 SO 65 Spac/'/ig tettveefl ///res, J^, Z of rf/omete r Fig. 52. — Uniformity coefficients for patterns G, I, and K (fig. 51) for close spacing along the line and for square spacing of sprinklers. Table 13 gives the uniformity coefficients for four of these patterns — M, 0, Q, and S for both a triangle arrangement and for close spacing along the line. The uniformity coefficients for both arrangements are compared in figure 54 for patterns M and S. There is a striking similarity in the two curves for both patterns. Uniformity coefficients for the close k /oo x 30 70 — X ^ so \ *■* M so N 55 O 60 P 65 70 R 75 S 80 r 85 u SO 20 50 40 D/sta/?ce fro/7? jpr/nk/er, 50 60 70 7c of rad/t/s covered Fig. 53. — Half cross sections of patterns that give highest uniformity coefficients for equilateral-triangle arrangement of sprinklers with different spacings. Bul. 670] Irrigation by Sprinkling 101 spacing (#! = 0.052)), though somewhat higher than for the equilat- eral-triangle arrangement with 8 t = 0.866 2), do, however, indicate the spacings between lines, S 2f for triangular arrangements that give the best distribution. TABLE 13 Uniformity Coefficients for Patterns M, O, Q, and S* Spacing be- tween sprink- lers, Si, in Spacing be- tween lines, S2, in per cent of diameter Net area, in per cent Uniformity coefficients, in per cent, for the patterns : per cent of diameter of circular area covered M Q s For equilateral-triangle arrangement of sprinklers (Si = 0.866 D) 30 26.0 9.9 99 99 99 99 35 30.3 13.5 97 98 98 94 40 34.7 17.6 99 98 99 95 45 39.0 22.3 99 98 94 97 50 43.3 27.6 90t 96 89 89 55 47.7 33.4 97 96 87 82 60 52.0 39.7 91 98 88 78 65 56.3 46.6 82 94 92 78 70 60.7 54.0 72 87 97 82 75 65.0 62.0 77 90 89 80 69.3 70.6 67 79 94 85 73.7 79.7 68 86 90 78.0 89.4 73 For close spacing along the line (Si = 0.05 D) 5 40 2.5 99 98 95 97 5 45 2.9 99 97 92 91 5 50 3.2 96 98 92 87 5 55 3.5 90 98 94 87 5 60 3.8 84 93 99 90 5 65 4.1 77 86 94 94 5 70 4.5 69 78 87 96 5 75 4.8 62 72 80 90 5 80 5.1 65 72 83 5 85 5.4 66 77 5 90 5.7 69 * Shapes of patterns are shown in figure 53. t Italicized figures correspond to the spacing for which the pattern is designed. Judging from these studies, a more uniform application can be ob- tained with a triangular arrangement of sprinklers than with a square arrangement. This is particularly true for the wider spacings. A com- parison of the uniformity of distribution obtainable with both arrange- ments is shown in figure 55. The points plotted correspond to the pattern giving the highest uniformity coefficient for each spacing. To secure any added benefit from the triangular arrangement, however, the sprin- klers must be more accurately spaced than for the square arrangement. A comparison of patterns K and S will illustrate this point. Pattern S 102 University of California — Experiment Station 4S JO S5 60 65 Spac/na t>elu/eef! ///?es l S 2) % of a'/ameter Fig. 54. — Uniformity coefficients for patterns M and S (fig. 53) both for close spacing along the line and for equilateral-triangle spacing of sprinklers. Note par- ticularly that although pattern S gives a fairly high uniformity coefficient for a spacing of 70 per cent of the diameter covered by the sprinkler, the uniformity co- efficient is fairly low for spacings between 40 and 65 per cent of the diameter. has a uniformity coefficient of 94 for a triangular arrangement with S t = 0.8 Z), which corresponds to a net area of 70.1 per cent of the cir- cular area covered by the sprinkler. Pattern K has a uniformity coeffi- cient of only 86 per cent for a square spacing of 0.7 D, which corresponds •5j so ! I ^-^. -4? ■*1 /flV faa/fafe /r/a/y/t 'ra/ » arrange' T?e/}( r ? Spaa, v arra/?yt went K*\ aL 20 JO 40 JO 60 70 Net area, % of area a/" c/rc/e caverea" Fig. 55. — Comparison of highest uniformity coefficients for different patterns for square and equilateral -triangle arrangements of sprinklers. Higher uniformity coeffi- cients can be obtained for the triangular arrangement when the sprinklers are cor- rectly spaced. Bul. 670] Irrigation by Sprinkling 103 to a net area of 62.4 per cent. When the spacing between sprinklers for pattern S is changed to 0.65 D, the uniformity coefficient drops to 78 per cent, whereas for pattern K, the uniformity coefficient remains above 85 per cent for all square spacings closer than 70 per cent. UNIFORMITY OF DISTRIBUTION AND EFFECT OF SPACING ON ACTUAL SPRINKLER PATTERNS A study of geometrical patterns to determine the uniformity of dis- tribution is not entirely adequate, since actual sprinkler patterns are never symmetrical in shape because of wind and sometimes because of variations in the speed of rotation. To determine the uniformity of dis- tribution corresponding to the actual sprinkler patterns, all of the tests suitable for this purpose were analyzed. The uniformity coefficients, corresponding to a spacing, 8 lt of 10 feet along the line and spacings between lines, 8 2 , of 20 to 100 feet, were computed for both the east-west and north-south directions. Table 14 gives the mean uniformity coeffi- cient for each spacing for the two directions, together with other per- tinent data. These data are summarized in table 15, where several tests on a given sprinkler under similar conditions of wind, pressure, and rate of rota- tion are averaged. Table 16 gives the uniformity coefficients for various spacings along the line and between lines for sprinkler patterns illustrated in figures 29, 31, 33, 34, 39, and 43. This table brings out several important facts regarding the uniformity of application of water by sprinklers. A comparison of the uniformity coefficients for tests 16, 132, and 170 for specific spacings will illustrate the effect of shape of pattern. For test 16, the optimum spacing 8 2 is 80 to 90 feet; for spacings 60 and 70 feet, there is excessive overlap, and the uniformity coefficient is somewhat lower. This characteristic is not evident for tests 132 and 170 (patterns triangular in shape) but the optimum spacing is reduced to about 60 feet. The coefficients for test 11 illustrate the effect of inadequate pressure (20 pounds per square inch) . For this pattern a fairly good distribution is obtained with a spacing of 10 by 80 feet, but when the spacing 8 t is 40 feet, the results are very poor for all spacings S 2 , the coefficient being only 57 for a spacing of 40 by 80 feet. For test 170, the uniformity coefficients remain practically constant for any spacing 8 2 when the spacing #, is increased from 10 feet up to 60 feet. The uniformity coefficients for test 88 (fig. 39), showing the performance of a sprinkler with a wind velocity of 10.7 miles per hour and for test 118 (fig. 43) where the sprinkler rotated rapidly (26.3 revolutions per minute), indicate to what extent the effective area 104 University of California — Experiment Station o O P » g o S B p < m s CQ W t_j fe o h-J pa 1 E M W fe H M £ Cm 0Q c* v< w w w H t?> W r o pq Q ° ^ CM s g Ph £ 2 S 8 C— ' 0) t. O 0) S » eg fl Is o -1-5 03"^ o S3, i| t, 03 S3 3 c o > o §1 • CO (M -*f< t^. • 18 Ol 0O OO o> Ol tO T)l ijl rl N tO 00 O0 OO lO Ol if O *f CO 00 CO Ol Ol in o in m oi OO OO OO N OO CD 05 CO 00 OO 00 05 oo +1 O 03 t^ 03 ii 00 CN1 n in 00 00 00 05 00 CD CO t-» "0 O O0 OO OO t- Ol O) if O N O N O0 Ol OO Ol OO OO H 00 h OO OO Ol OO OO oi cm m N M r- Ol CO oo Ol O 03 CD 03 lO O N N N O0 CI Ol 0O CO O 05 CD Ol O O0 OO OO CD 05 00 tD N OO CO O0 OO OO OO OO Ol Tt< O O CO N W CO Ol N O CO 05 00 oo to -i o Ol Ol O 03 in 0) 0O CM rf t- Ol O O! O) OO OO Ol CM Ol O OO (» OO O0 OO 00 co m to io to 00 0100 0100 oo in oo cm to N 0O N OO OO ii CO Ol Ol Ol N m ih Ol Ol O 0) -f 0) CO rf< CD CO if OO Ol Ol Oi O) Ol 05 CO "0 CM 00 00 0100 05 N lO lO to lO 00 0105 0105 if CO N CO 01 Ol Ol 00 Ol ii CO Ol 05 00 00 CM CO Ol OS O 0) N to 00 C5 1C 00 05 05 0101 CD OO lO 05 CO O) Ol Ol OO CI tC N Ol OO 00 i- if co to in Ol Ol 05 Ol 05 f to to 00 r-- 01 Ol S l lO OO Ol N OO 05 05 05 Ol 05 00*0 00 05 05 Ol O Ol 00 Ol 05 00 0105 05 Ol 00 Ol Ol CD Ol Ol Ol Ol Ol OO OO ii 00 00 01 Ol "Hi 1| ■* o co r- in 05 O CO O <-H © Ol 1^ CD CD Ol 00 Ol Ol o N s to to m O CM n CM ^i if h m h to 05 010105 m oo ii ii ii o oi m O -h CM O iH O h M N (N ■«*! co r- to in — oi o oi o CO if CM lO if OS CM -i CM -i CD O OO -*f ■*»« O O Ol 00 o -- Ol oo ^ CD CO Pres- sure, pounds per square inch H n O O O ■* >o ■* t|i m O O O O O o o to to o ■*•*"* Di CO m o o o o CM V ■* if H< mom f if N o o if m oi" ® « 8 N OJ-C O fi 03 fc|.S t3 CD CD n >o CM CM CM CM CM co co co co co \ \ •V. \ \ to >o lO lO >o CD CD CD CD CD \ \ \ \ ^ CO CO CO CO CO CD CM CM CM CM h CO CO CO CO co m m m m CO CO CD CO CO \ -v. CD CD CD V. \ \ \ \ n i—l li5 in in CD CD CO CD CM ^i i-i i-< i-i CO in in in in t- CM CM CM CM CM co co co co co \ ^^ "\ ^^ ^^ Ol Ol Ol Ol Ol CM CM CM CM CM co co co co co Ol 05 Ol Ol Ol CM CM CM co co co Ol Ol Ol CM CM co co S "v. Ol Ol Sprin- kler desig- nation - r- • 00 O CD • N 00 O0 * OO O0 CO CO N OO t» t- VO CM • • ■ Ol TJ1 - - - CO OO • CM CO l>. • • co r- ■ Ol N CO ■ N OO OO ■ CO CO CO rH • CO (• If) N — < • S N H Ol • N oo oo CO if) Ol o tji » N 0O O) 00 I) N CO O "J 00 OO OO Ol CO i-H T-l Tfl O O CD CO 1Q CO Ol N •* lO i)i CO 0O N N N OO t^ Oi 00 © Oi CO CD OO Ol O0 t-h oo i^ r- r^ If) N OO CO 00 O) CO N O) (O OO CO OO t- OO -H -H CO CM Ol OO OO -XD Ol C^ Ol OS 00 CO o O0 N N Ol OO If) rt N If) O N N CO N OO •Jl N 00 tJI CD O0 OO 00 O0 OO Ol Ol if) if) CO N N 0O CO 00 TJI I-H if) Ol CO CO Ol OO N CO * ffl N Ol ID OO N Ol N OO if) tji Ol © CO t>. t- r- oo oo CM t~- CO CO O OO OO N OO Ol CO tJI O CO i-~ 00 0O OO OO N tJI OO N CO CO oo oo oo oo oo t— if) o -* oi 00 0O O0 00 N 00 CO O0 Oi Ttl t- Ol oo oo Ol ^i 00 CS H N a oo oo oi o> l^ i-H T-H © CO OO N OO OO Ol CO CO CO N o OO Ol OO OO Ol if) t- 00 tji if) 00 00 OO 00 00 CO TJ< T-l CM r-l oo OO 00 OO Oi if) CO Tji CM if) OO OO OO Ol OO ■* H O CO * OO Ol Oi Ol Ol « N 00 OO H O) » Ol OO Ol H l« CO U) M Ol O0 Ol Ol Ol OH (ONN Ol Ol 00 Ol Ol if) r|i OO t(< CO OO 00 OO OO Ol CO Ol CO Ol tJI Oi 00 00 00 Ol © O CO CM tji Oi Oi Ol Oi Oi TJI CO TJI S If) OO Ol Oi Ol Ol 00 CO CO Ol CO 01 Ol Ol oo oo H S !D CO N OO OO Ol Ol Ol 115 CO "* N 0O Ol Ol OO Ol Ol CO CO t)I If) oo Oi Ol Ol Ol Ol 00 CM O Oi CM 01 Ol Ol 00 Ol co 10 r~ 10 co Ol Ol Oi Oi Ol CM O CO O0 O0 Ol Oi Oi Ol Ol * CO OO ■* II) Ol Ol Ol Ol Ol t~- Ol O Ol OO Ol OO O Ol Ol N CO 00 Ol CO Ol Ol OO Ol Ol Oi Oi CO Ol OO Ol Ol Ol Ol Ol CO CD N 11) N Ol Ol Ol Oi Oi N Ol OO N OO Oi Oi Oi Ol Oi CD CO «) » CO Oi Ol Ol Ol Ol 115 Ol N 0O j< O r- O Oi Oi Oi O O --I O CD tj< O © -h — O -h CO CO CM O O O O T-H T-l O O CM CO If) N * O O •* M N C> tJI U) t^ Ol If) t^ CO f^ T-l t~- Tji O t- r-- Tfi th oo tJI CM O t- CM T-H O CO O H _ _ _ _ O >o O O tji CO 0O CM t*i CO CM CO Ol 00 Tf CO OO O CO CM CM Ol CO CM tJI CM Oi © oo t^. © * Ol N CO il CD w id tji tj< us Oi O CM T-l O Ol Oi CD Ol CO Oi O 00 t— r~- tji if) tJI CO • CM CM CO H © © CO CM O -H CM O O O CM O CM O T-l -CO CM T-l O O T-H T-H CM T-l T-H O o © • © © t-h If) CO CO if) N IN CM N CO If) t^. CM if) O CM CM CO Oi 00 CD CM CM CM CM if) O Ol CO TJ< CD CO lO If) OO CO CO r-l Ol CO CO OO O N N t— t— if) t^ t- W 1C OO OO o CD CM CM CM tji lO t-h O ■* CM CD CO tJI OO If) if) © © © © « ■* n n ■* io o o o o N rf CO ■* •>* o o o o o Tji tji tji tji if) O O O O U5 if) oi t- co o CO CM CM CM tji O if) if) O o If) TJI Ttl T)< CO © © Oi © © TJI TJI CM Tji CO 1/8 3/16 3/16 3/16 5/32 CM ■ CO CO co ■ ^ i-H \ ■ \ \ if) • co co CO CD CM CM CM t-h t-h CO CO CO \ \ \ \ \ CO CO if) if) lO CM CM CO CO CM \ \ C \ \ If) if) CO CO If) CD -H CO 00 CO oo \ \ W -v. CO T-l T-H T-H T-H CO OO CO 00 oo \ \ \ \ \. • CM CM • eo co : : :r>^- CM Tjt eo -^i •"*< Tf co 5$- > >> ^ Tfl CM CM CM CM co co co co co ^^^c?o^ CM CM CM CM CO CO tJI CO CO \ \ \ \ "V. Ol Ol T— t Ol Oi CM CM CO "tfl ""J! CO CO T-l \ \ \ \ \ T-H T-l Ol Ol If) CM CO Tji Tf Tji -Jl \ \ \ \ \ CO CD \ \ \ \ \ T-H If) If) T-H »-H CM CM CM CO CO CO TJI TJI ■*>. \ \ \ "\ H « N N ^ 6 < < -< Ph Ph Oh O O oo o o h ci CM CM CO CO CO n ■* io e s CO CO CO CO CO CO CS O tJI IO M CO TJ< lO f) CO N 00 O) r- if) lO if) if) CO 00 CO tJ< if) CD CO OO O0 O0 CO r- oc 01 o t-h CO OO OO Ol Ci cm eo tji cd r^ Oi Oi Ol Ol Ol 106 University of California — Experiment Station i CO J is si *s a s .8 J II it In CO .5 03 C O > •2 O CD ^H 0) • oo ■ . . . CO • • • IO CM CO 1 o"S OS cd « O ifl N * CO t^ CO CO CO lO CM co co CO ■ to Ol S N Ol t^ r- t^ co H « N CO CO OO O CD 00 CD »0 © CO OS O0 N O N N N t^ t~- CO OS oo t^. »o oo oo o o ** * o OS OS OS OS OS s| co »o in co cm O) C3 OO 0O OO IO 00 OO 00 co os r~ lr~ 00 « O0 ifl * * oo t^ OO 00 oo O 00 00 CO t^ OS OO OO 00 oo OS OS OS OS OS o"S ^<2 r)i OO (O Tjl M OS OS OO OS OS CM -* os os o DO t-- CO oo oo tO N N N N O0 OS 00 oo oo CM CO CO U5 OO OS OS OS OS O0 co to r~ h* 1 o OS OS O0 OS OS +3 O CD CO 0) CO OO t- 00 CM OS OS OS OS OS 00 CO os os os 00 1~- os os co s to to to OS OS OS OS OS tO N tO » N OS OS OS OS OS OO CO CO CO oo OS OS OS OS OS +3 O CD CM 0) * O0 N O0 O0 OS OS Ci Ol OS IO OS OS OS as IO oo OS OS IO t^ OO S Ol OS OS OS OS OS ia oo s to oo OS OS Ol OS Ol 0O S IO N O0 ■g g Si is S W Ol to w O CM OS O OS lO oo CM O O ■* o T-H OS OO CO CM CM O OS O OS OS OS OS CM o »o CM O CM CM O 00 M CI CO ■* H O OS T-H ^h CM r£5 1° 55 « IO N N N H © CM -< rH Ol cm r- CM O to OS CO OS — ( OS t-H CO o »o ■* OS O OS O OS o •<*< o -* © ^H CM CM © O0 CM CO IO OS CM O OS rt © CM Wind, miles per hour ■* H CO 00 CD rH Tt- 00 t-~ H O N M H to co co r- th co *tfl CM f- OS CM CM •*f CM OS O0 CM •* OS CO oo to * oi n n IO t-H CM CO •**! CO CO CM IO •** Rate of rota- tion, revolu- tions per minute OO O O OS ■* CM CM CM r-c CM OO CO H* OS oo o t- oo CM OS '-i CM CO oo O OO t- 00 t-1 CO ■«f i^ os co to tO H Ol CO O O O O O O CM CO O I-H CM O T^ t-H H © CM © th CO CM © © -h O O Dis- charge, gallons per minute © OS CO T»< CM »0 OS ^ o ->*< CO f~ Ui Ui t~- tO 0O MO h CO tH CM © CO CO CM OS CM CO -h CM i-l CM >-H CM CO CM CM N CO tO CM i-H 00 «* CO CM CM t-H t-H t-H CM CM CM CM CM t-- CM CM CM CM r-< CM IO CO rH 00 OO Pres- sure, pounds per square inch o o o o o CO -f CO 'J" CM O to co o o o o o o o >o * ■* ^ rt O O O O CO f "O to » * © © © © © tJ( M IO IO U3 o c o fc!.a -9 7/32 7/32 7/32 7/32 7/32 CO CO o CO • 1/4 5/16 5/16 5/16 5/16 CO CO CO s CO CM -H CO CM CM CM CO CO 5/16 9/32 9/32 9/32 9/32 1/4 1/4 9/32* 1/4* 1/4* Sprin- kler desig- nation 6 6 6 66 J .-) Jj (-3 i^ 2 2 22'2 2ZZ2Z •-5 •-» >"5 >"S >^3 +3 OO Ol O H N OS OS © © © co -fi o o US o co r~ o o 00 OS © tH CM O O -h ^H tH CO ■* * N OO OS © i-H CM CO r-H CM CM CM CM Bul. 670] Irrigation by Sprinkling 107 oo OS co CM us CO OS CO OS CO co . 115 • oo US ■ !>• OO • t- H (O 115 tO O OS OS CO CO 00 • © © OS CO • OO OS 00 OS t- t- CM t*i tO tO N h T-l os to to • ■«* oo • o CO ■ !D S •* CO • 0O OO N OS O OS O CO OO OS t>. OO OS © CO OO OS oo OS OO OO OO 00 CM H 115 CO t"- 00 OO OO • to © 00 CO • tO OO N 0O Tf< © t- oo O O) N ■ U5 t^ 00 OS • oo r-l ■ M OJ N OO ■ OS OO oo ■* 115 to eo N 00 OO O0 OS 00 w ra o « h 00 OO OS OS 00 h oi n a CO oo oo oo N CI tO CO N F— C^- OS OS OS r— T-i 00 OS © © OS t^ CM Ol O OO N 00 O N * O) W oo r^ os oo os U5 f- CO OS CM OO 00 OO OO 00 M H to tO H r- OO OS OS oo t|i N CO r» ■* OO 00 oo oo to CO CM 115 CO OS CO OS OO OS OO ■*f to OS OS OO 00 OO OS oo N 115 ID ID N O0 OS OS OO OS * to o oo oo os os oo r- oo 0O N CO * CM f~ 00 OS OS OS 115 OS t» CO •* CO 00 oo oo oo CO to OO to OO Ol Ol N OO 00 co t- OS OS 00 H 0O O) N OO OS 00 OO OS N ID N W * OS t^ 00 OS OS OO t^ OS OS oo OS OS OO OO OS 115 Ir^ r^ CM CO OS OS OS OS OS CO H5 TP 115 ■* OS OS OS OS OS CM CO OO O N OS OS O0 OS OS to 00 OS OS N 00 N M ID OS OS OS OS OS t» K5 0O H ■* OS 00 OS OS OS O) OO 115 O) S O) » N 0O N 00 OS OS OS OS 115 N 115 0O rt os os os os os i-l CO 115 OS 00 OS OS OS OS OS t~ 00 os os N 0O Ol OO OO OS OS OS OS OS OS CM OO OS OO OS OS OS OS OS Ol O N 00 ffl OS OS OS OS OS t- OS OS 00 OS OS OS OS OS OS OO 0O CO f- oo OS OS OS OS OS N 0O 0O Ol OO OO OS OS OS O rt O H N © -* © OS CM ■* i( h 115 to CM t~- CM OS t^ OO N N » N CM CO © O H 115 i— i © OS ic OS CM * CM CM <-H © H t~- OS O H H H N CO tO 00 © r-l OS H OS i-l T-l to CM © CM o ■* to to © CM 00 CM CO CM O -i rt H H N N CM CM © CM CO t-i © M15 CO N t- U5 CO t- © t- 00 OS OO © OS U5 CO co os ^ lO KS lO "O O "5 O O O U5 -* U5 CO ■"»< © © © © © 11) "5 CO ■* •* © © OS lO 115 CN «5 f •* ■* © © © © © 115 -CH CO US U5 © © CM © © © © CO tP • CM • CO • • • \ ■ U5 CO CO \ \ \ \ "•> M M H H H OO OO OO 00 00 \ \ \ \ ^j OO 00 OO CO to 00 00 OO OO OO \ \ \ \ \ CM CM co co CM "*i r»i tH tH CO ^ > ^> -> o^ CM CM CM CM CM CM CM CM CM CM CM CO to CO to COCOCOCOCO COCOCOCOCO M H i- I t-i t-i •*}< -ttl t* -rjl -!tl Os" Os" t^" t^" r^" OSOSOSOSOS O) 115 115 115 115 tH ^h t^ t-h rt co to to to to \. \ c \ \ co co co co co ■»P t*i t-, cm cm cm « h> i^ ^ 4 W CM CM •>* ■* 1 pq « pq pq ffl 6666 6 6s§§§ 33333 CM CM CM CM CM 6 6 6 6 6 CO CO 6 6 c\ c cn c- CV P cr cr Ifl cr o- c^ or O c ">» P •>* u- ec a o- co 1 1C ec ec a- C — . ' t-i 03 108 University of California — Experiment Station Oh ►"• O 02 | | .53 - a am o~£ O CM • O OS oo © m m t^ © © t^ M O O W U5 0O N O OO 0O OO t^ CM » O) N .—t CO © CO Tf O N h n >oo Ol OO O) M t^OOCOOJOO OOOi > cci-^ N N O N o> as co uo oo oo a cc *o co O) CO CO O) o> ot 3 q. 3 c > cp o "* cr-- *s O O ^ O «5 -H O O — 1 (M Tf 00 CO »0 CM © t^ OO ■^i CO CM CD CO M Tf CO W lO O M M ^ t» © CO CO CM CO OS M ^ OS M - CO CD CO o «o »-i . 1—1 - - CO 03 - CO Tt< CM i-l . t-~ (N «l CO N w CO t^ CD CO O 5 C3 «<«««« pqpqoot^ feOOO Bul. 670] Irrigation by Sprinkling 109 TABLE 16 Uniformity Coefficients for Actual Sprinkler Patterns Spacing along Uniformity coefficients, in per cent, for various spacings between lines, 52 line, o » oo , A o 8 °e„ o ex A <£ A A AA °aA a O i Ld° o JO o O O A i o i A ° o o ° « ( 0° A 1 • l . -J • • • • • ; : ! - 20 30 40 SO 60 70 Approximate average re/ai/'ve /iv/nid/fy, per ce/?t SO Fig. 56. — Evaporation loss calculated from sprinkler tests. This loss includes that from the spray, that from the cans during the test and period of measurement, and that represented by the amount of water clinging to the cans when they are emptied. cubic centimeters of water in each of five cans with, and five cans without, funnels. After 3 hours the average loss from the cans with funnels was only 0.75 cubic centimeter per hour, or about 10 per cent of that from the open cans. The first 122 sprinkler tests were made with open cans spaced 10 feet apart in each direction. For subsequent tests, the cans with funnels were spaced 7.07 feet apart in diagonal rows, providing one can to each 50 square feet (fig. 27). The use of the cans with funnels did not, however, materially change the calculated losses as indicated by figure 56, which shows the loss plotted against the approximate average relative humidity. The reason, probably, is the evaporation of water from the wet surface of the funnel. No way was found to eliminate this error. The measured losses should, therefore, be considered not as evaporation losses from the spray alone, but as the combined losses from the spray and from the cans during the 112 University of California — Experiment Station test and the period of measurement following the test, which averaged about 30 minutes. Indirect Method of Estimating Evaporation Loss from the Spray. — The evaporation loss from the spray may be determined approximately from thermodynamic principles. Evaporation of water requires heat. The number of calories necessary to evaporate one gram of water, called the latent heat of evaporation, varies from 539.6 for water at 100° C (boiling point) to 595.9 for water at 0° C (freezing point). It is about 585 for water at 68° F. Three sources of heat are available for evap- orating water from a spray : (1) heat from the water, (2) heat absorbed from the air, and (3) radiant heat, principally from the sun. If all of the heat came from the water, it would require a temperature drop of about 10.5° F to evaporate 1 per cent of the water. When the water is cooler than the air, as normally in the daytime, the water will absorb heat from the air, and the drop in temperature will be less than 10.5° F for 1 per cent loss. Absorption of radiant heat will also increase the evaporation for the same temperature change. When the initial water temperature is the same as the wet-bulb temperature of the air, an equilibrium condition exists; all the heat required for evaporation comes from the air, and the water remains at a constant temperature. When the initial temperature of the water is lower than the wet-bulb temperature, it will increase even though some evaporation still takes place. The evaporation would be zero, however, if the water temperature was at the dew point; and if it were lower than this, con- densation would occur, and there would be a gain rather than a loss of water. Neglecting the radiant heat and considering only that from the air and water, an equation has been derived from which the evaporation loss can be determined for any given change in the temperature of the water, provided the air temperature and humidity are known. This ap- proximate expression for the evaporation loss from the spray is = lQQCAt r p w -p a i r [P» -P a - 0.00037 B(t a - O J V ; where E is the loss of water from the spray expressed as a percentage of the amount discharged; C is the specific heat of water, calories per gram per degree Fahrenheit ; r is the heat of vaporization, calories per gram ; At is the drop in temperature of the water from the time it leaves the nozzle until it reaches the ground ; t w is the mean water temperature, degrees F ; t a is the air temperature, degrees F ; P w is the vapor pressure of water at temperature t w , inches of mercury ; P a is the pressure of the water vapor in the air, inches of mercury • and B is the barometric pres- Bul. 670] Irrigation by Sprinkling 113 sure, inches of mercury. Evaluating the constants, letting C = 0.555, r = 585, and B = 30, the expression becomes E = 0.095 At [~- J >W ~ Fa , rl . (27) \P W -P a - 0.011 (t a -t w )j The expression within the brackets varies from less than 1.0, when the temperature of the water exceeds the air temperature, to infinity when the water temperature is the same as the wet-bulb temperature ; and from minus infinity to zero as the water temperature decreases to the dew point. Equation 27 fails to take into consideration the very small drops of water which are completely evaporated or blown away by the wind, and which do not contribute to the final temperature of the water as it reaches the ground. A study of the distribution of size of drops, however, indicates that only a very small part of the water discharged is in the form of tiny drops that are lost, and it is believed that this would not cause an appreciable error in the determination of the evaporation loss from the spray. Several tests have been made to determine the temperature drop of the water between the sprinkler and the ground. The water was caught in a thermos bottle with a large funnel. To obtain a large quantity of water quickly and to avoid errors caused by changes in the temperature of the water after it is caught, the rotating sprinklers are stopped during the catch. For some of these tests, in order to increase the evaporation loss and to more nearly approach field conditions, the water was warmed several degrees above the normal temperature of the well water which was about 65° F. This was accomplished by installing two electric heaters in the pipe line supplying water to the sprinkler. The temperatures were read with a thermometer graduated to 0.2° F. Five tests on rotating sprinklers with initial water temperatures of 69.5° to 84°, and with air temperatures ranging from 75° to 101°, showed decreases in water tem- perature from 1° to 7° corresponding to evaporation losses of 0.23 to 0.81 per cent. Another test on a small spray nozzle, with an initial water temperature of 98.7° and an air temperature of 100.5°, resulted in a temperature drop of 20.7°, corresponding to a loss of about 2 per cent. From these measurements of the temperature change of the water as it passes through the air, one must conclude that the evaporation loss from the spray is negligible in comparison with subsequent losses from the wet soil and vegetation. Even when one considers the possible effect of solar radiation, and the increase in vapor pressure of the water when in the form of small drops, one cannot account for evaporation losses from the spray of more than 2 per cent of the amount applied. 114 University of California — Experiment Station EVAPORATION LOSSES FROM WET SURFACES Evaporation from soils and other wet surfaces, during and following an application of water, may be appreciable. This loss is relatively more important in connection with irrigation by sprinkling than with irriga- tion by other methods because (1) lighter applications are generally made by sprinkling than by surface irrigation, so that a larger percent- age of the water applied may be lost by subsequent evaporation; and (2) the cost of applying water by sprinkling is generally higher, so that the evaporation loss represents a greater economic loss. Evaporation Losses from Moist Soils. — The evaporation loss from moist soils has long been a subject of discussion and experiment by those interested in irrigation. Fortier and Beckett 20 conducted experiments to determine the loss from undisturbed and cultivated soils at Davis after an irrigation. Veihmeyer 21 made similar studies at Mountain View. Briefly, these experiments indicate that 1 to 2 inches of water evaporates from the soil within 3 to 4 weeks after an irrigation, more than half of which occurs during the first five days. This loss is not prevented by cultivation, partly because most of it occurs before the soil is dry enough to work. Practically all of the evaporation loss is from the first foot of soil, and most of it from the first 4 inches. The rate of loss by evaporation after the first week is negligible in comparison with the rate of extraction of water by plants. The actual loss by direct evaporation from a cropped soil is probably considerably less than from a bare soil because the ground is shaded, partially at least, by the crop, and also because the growing crop rapidly reduces the moisture content of the soil by transpiration, which lessens the opportunity for evaporation to occur. According to other experiments, the rate of evaporation from satu- rated soils is about the same as from a free water surface, or about 0.3 inch per day in the Sacramento Valley during the summer. With a water table 6 inches below the surface, the evaporation loss averaged about 0.21 inch per day during the summer; but this decreased to about 0.06 inch per day with the water table 2 feet below the surface. When relatively small amounts of water are applied to exposed soils at frequent intervals by sprinkling, a common practice in the spring, much of the water may be lost by evaporation. In some instances appli- cations of about 1 inch are made at weekly intervals to aid the germina- tion and starting of a crop and to prevent the drying out and crusting 20 Fortier, Samuel, and S. H. Beckett. Evaporation from irrigated soils. U. S. Dept. Agr. Office of Experiment Stations Bui. 248:1-77. 1912. 21 Veihmeyer, F. J. Some factors affecting irrigation of deciduous orchards. Hil- gardia 2(6): 125-291. 1927. Bul. 670] Irrigation by Sprinkling 115 of the surface soil at this time. Most of this water may be lost directly from the soil by evaporation. When, on the other hand, a growing crop completely covers and shades the soil, and when the applications are sufficient to give a penetration of 3 to 6 feet, only a relatively small per- centage of that applied may be lost by direct evaporation. Bate of Evaporation from Free Water Surfaces. — The evaporation from free water surfaces is measured at a number of Weather Bureau stations, and at many other places in California and elsewhere. The evaporation as measured at these stations has been found to depend largely upon the type of evaporation pan used, and upon its exposure. For this reason various organizations, such as the Weather Bureau, have standardized their evaporation equipment to facilitate comparisons. Tests have also been made to determine the coefficients by which the evaporation from various types of pans must be multiplied to obtain the equivalent evaporation from a lake surface. For the standard Class A Weather Bureau pan the coefficient most generally used is 0.70. 22 A Class A Weather Bureau evaporation station has been maintained at Davis since 1926. Besides the regular equipment, a recording evapori- meter is used so that instantaneous rates of evaporation can be obtained. The average daily evaporation at Davis for a fourteen-year period was as follows: May, 0.27 inch; June, 0.31; July, 0.35; August, 0.32; and September, 0.24. According to evaporimeter records, the maximum hourly rate sometimes exceeds 0.05 inch and averages about 0.04 during the afternoon. Since water is sometimes applied with sprinklers at rates as low as 0.10 inch per hour, an appreciable evaporation loss may occur during and immediately after an application. Even with application rates of 0.25 to 0.50 inch per hour, more than 10 per cent of the water may evap- orate as it is applied during the afternoon. The evaporation loss at night, however, is usually very low. Interception and Subsequent Evaporation of Water from Plants. — When crops are sprinkled, part of the water is intercepted by the foliage and later evaporated without reaching the soil. Studies of rainfall in- terception by various investigators indicate that an appreciable amount may be caught by trees and other plants, especially when the rain occurs in small storms. The determination of rainfall reaching the soil under a vegetative cover is difficult, and results obtained are not always con- sistent. Furthermore, the interception is generally reported in per cent of rainfall, and therefore depends upon the intensity and duration of 22 American Society of Civil Engineers. Evaporation from water surfaces, a sym- posium. Amer. Soc. Civ. Engin. Trans. 99:671-747. 1934. 116 University of California — Experiment Station the storm. Clark 23 made determinations of the maximum interception capacity of many plants. From his data it appears that few crops can retain 0.1 inch of water, although his attempts to measure interception by catching water in pans under the vegetation indicate much larger losses. Interception of rainfall and that for water applied with sprinklers should differ principally in the amount that evaporates during the ap- plication. The evaporation rate while sprinkling may be high, whereas during a rain evaporation rates are generally very low. The presence of water on the foliage should temporarily reduce the rate of evaporation from the soil and the rate of transpiration from the leaves. DESIGN AND OPERATION OF SPRINKLER SYSTEMS Methods of Operation, — With respect to the method of operation, por- table sprinkler systems may be divided into two general classes : first, those in which the rate of application is fairly high and the sprinkler lines are moved frequently; second, those in which the rate is relatively low and the moves are rather infrequent. Most systems with portable pumping plants fall into the first class. The capacity of the system is generally limited by the power available for operating the pump, while the length of portable pipe that can be effectively used is governed by the dimensions of the field and the general layout of the system. The sprinklers selected are of a capacity that will effectively utilize the pump- ing plant. The rate at which the water is applied receives little considera- tion ; often it is too high, and water accumulates on the ground surface before adequate penetration is obtained. The result, usually, is that light frequent applications are made where heavier and less frequent applica- tions would give better results and be more economical. Frequent moves mean continuous employment of labor together with a relatively low operating efficiency because of the larger proportion of time lost. This is especially true with the single-line arrangement. Most of these systems operate continuously day and night, the lines being moved every 2 to 6 hours. Moving pipe at night is objectionable because it takes a crew con- siderably longer to make moves at night than during the day, and because working conditions at night are disagreeable. Under the second method of operation, the water is applied at such a rate that the moves can be made at convenient times each day. The ideal arrangement is to move the lines twice a day — morning and evening so that both moves are made in daylight. From 3 to 8 inches of water is generally required to wet dry soils to the depth from which moisture is 23 Clark, O. R. Interception of rainfall by prairie grasses, weeds, and certain crop plants. Ecological Monographs 10:243-77. Apr., 1940. Bul. 670] Irrigation by Sprinkling 117 extracted by most crops. Applications of these amounts must be made fairly slowly so that the soil can absorb the water without runoff. Under some conditions it is necessary to extend the period of application to 24 hours to avoid runoff. When sprinkler systems are operated on this basis, continuous atten- tion may not be required. The operating cost will then be lower, since 60 to 80 per cent of the actual expense of operating portable systems is for labor. If this help can be employed only during the time required for moving pipe, or used effectively elsewhere during the time when they are not needed, the saving will be appreciable. There are many possible schemes for operating a system in this manner. When a portable pumping plant is used, continuous attention may be necessary to prevent damage if the pump loses its prime or if any other emergencies arise. When moves are made only twice daily, it may be practicable to employ one man for the day shift and one man for the night shift, having the shifts overlap at the time of moving. In this case, each shift would be 12 hours plus the time required to move the system. Long shifts are feasible because during most of the time no work is re- quired and the operator can rest. Where adequate applications can be made in about 8 hours, three moves a day may be desirable. With one move at dawn, another at noon, and a third just before dark, three fairly equal periods of operation can be obtained during the summer months. Two men working alternate shifts overlapping the moving period could handle a system on this schedule with less physical effort than under the conditions now common. Three men working overlapping shifts of about 9 hours may, however, be more desirable for such a schedule. For small sprinkling systems operated by the owner, a 12-hour set is especially advantageous ; it permits the farmer to carry on his irrigation work continuously, and yet take care of other farm operations. This method of operation requires a system that is free from troubles ; sprin- klers sometimes clog or stop rotating, or the pump may lose its prime. Usually it is more feasible when pressure supply lines and stationary electric-driven pumping plants are used. At present this method is being practiced to a much greater extent with orchard systems than with other types. Capacity of Sprinkler Systems. — In planning any sprinkler system one must first consider those factors that influence the selection of the size of the system. In general, the lowest initial cost results when a system is planned for continuous operation with a capacity that will just satisfy crop requirements. This may or may not be the most economical system to operate. Sprinkler systems can be designed to use effectively a smaller 118 University of California — Experiment Station flow than is desirable for any other method of operation — a feature ad- vantageous to the operator of a small farm or orchard, since he need not invest in an irrigation system with a capacity several times larger than necessary for his acreage. Whether or not a small sprinkler system, operated more or less con- tinuously, will be more economical than a larger system, operated only a few days for each irrigation, depends primarily upon the method of operation. If it requires continuous attention, a small system may not be economical because the labor cost of operating it may be very high — all out of proportion to the cost of the water or the cost of power for pumping. Where continuous attention is not necessary, and especially where the attention required can be given at convenient times, a small system operating continuously will also be the most economical to oper- ate. Such systems are especially desirable for orchards and pastures. For large field-crop portable systems with portable pumping plants, one may find it difficult to arrange the layout for slow rates of application that will permit 12-hour or even 8-hour periods of operation. Sprinkler systems designed for continuous operation must have suffi- cient capacity to meet the peak requirements during the critical part of the year. For orchards and perennial crops the maximum water require- ment occurs during June, July, and August. For some annual crops it may occur earlier or later in the season. Crop transpiration depends largely upon the climate — sunshine, temperature, and humidity. For interior valley conditions, crops that completely cover the ground — alfalfa, sugar beets, and the like — will transpire water at a maximum rate of about 6 to 8 inches per month during the period June to August. Mature deciduous trees transpire about the same amount. Citrus require- ments are a little lower. For some coastal areas, applications of 2 to 4 inches per month are adequate. One acre-inch of water in 30 days is equivalent to 0.628 gallon per minute continuous flow. To supply 8 inches in a month, therefore, re- quires a continuous flow of about 5 gallons per minute per acre. For valley conditions, this is approximately the minimum flow that will com- pletely satisfy crop requirements. For systems designed to be operated only during the day, a flow of about 10 gallons per minute per acre will be required. Where additional time must be allowed for cultural opera- tions, systems with still greater capacity must be provided. Where con- ditions are such that it is economical to provide a complete irrigation in a short period, a still larger flow may be desirable. A convenient relation to remember is that the required flow Q = ^ (28) Bul. 670] Irrigation by Sprinkling 119 where Q is the flow in gallons per minute ; d is the average depth of water applied in inches ; A is the area in acres, and T is the time in hours. For example, to apply a 4-inch application on a 10-acre tract in 48 hours requires . 450X4X10 o ^ c __ Q = — = 375 gallons per mmute. Bate of Application. — Besides the capacity of the system, one must consider the rate of application. This may be governed by the soil con- dition (the rate at which the soil will absorb the water) or by the method of operation (the period of time in which one wishes to apply a given amount of water) . Coarse- textured soils (sandy and gravelly soils) generally absorb water rapidly, usually at any normal rate at which it might be applied with a sprinkler system. Fine-textured soils (fine silts and clays) absorb water more slowly. The rate of absorption decreases with time. Some clay adobe-type soils absorb water very rapidly when dry because of the granular structure and large number of cracks that form when the soil dries. For example, when dry, such a soil might absorb the water at a rate of several inches per hour for a few minutes ; but the rate of absorption might decrease rapidly, perhaps becoming a small fraction of an inch per hour after several hours of application. The maximum rate at which the water should be applied will depend upon the amount of each application. When only 1 or 2 inches of water is desired, high rates of application may be used ; but if 4 to 6 inches or more is to be applied, much slower rates are required. A knowledge of rainfall rates and of the ability of soil to absorb prolonged rains may be used as a basis for selecting suitable rates for sprinkler application. The rate at which soils absorb water also depends somewhat upon how the water is applied. Large drops resulting from low pressures tend to puddle and seal the surface of the soil much sooner than a fine atomized spray. The condition of the soil surface is also important. A cultivated soil left somewhat cloddy will absorb the water more rapidly than one that has been pulverized to a fine dust. A hard compacted or crusted soil will generally absorb water slowly. The presence of organic matter on the surface appreciably increases the rate of absorption for some soils. Rates of application varying from about 0.1 to more than 1 inch per hour are practicable with sprinkler systems. Depth of Application. — Soils act as a reservoir for water, retaining in it the pore spaces between the soil particles until it is used by plants or lost by evaporation. When a soil is irrigated, the water penetrates rather rapidly until the moisture content is reduced by drainage to a 120 University of California — Experiment Station certain percentage, called the field capacity, after which further move- ment and penetration into dry soil becomes very slow. For practical pur- poses, the field capacity may be considered as the amount of moisture a soil will retain against the downward force of gravity, or the upper limit of the available-moisture range. It varies greatly with soil texture, and to lesser extent with other factors. In general, fine-textured soils have a higher field capacity than coarse-textured soils ; clay loams and clays range from about 20 to 30 per cent, whereas fine sands and sandy loams vary from about 8 to 15 per cent. In terms of quantity, the amount of water held by a soil at field capacity varies from about 1 inch per foot depth of soil for sandy loam to 4 inches for some clays. Not all the water held by a soil is available for plant use. The lower limit of the readily available moisture range is called the permanent wilting percentage. At this moisture percentage plants wilt and do not revive until water is added to the soil. Generally plants do not die when they wilt, but their growth practically ceases ; in some instances, they shed their leaves. The permanent wilting percentage is the same for all kinds of plants grown in the same soil, although it varies widely with different soils. Sometimes, however, shallow-rooted plants wilt before other deeper-rooted plants growing side by side. Plants differ in their ability to withstand long periods in dry soils ; some die, others go into a dormant state. The permanent wilting percentage is influenced primarily by soil texture ; for many soils, it is approximately half that of the field capacity. This rule, however, cannot be applied generally because some soils show wide departures from it. A typical Aiken clay loam, for example, with a field capacity of about 30 per cent has a wilting percentage of 21 per cent. In contrast, a typical Fresno fine sandy loam has a field capacity of about 10 per cent and a permanent wilting percentage of about 3. The Aiken clay loam holds three times as much water as the Fresno fine sandy loam; but that available to plants is nearly the same. In addition to the ability of the soil to hold water, the depth of applica- tion is influenced by the rooting depth of the crop grown. This varies with different crops, and also depends upon the soil and upon the pres- ence of hardpan or a water table near the surface. For many field crops growing in deep uniform soils, roots are active to much greater depths than is commonly supposed. Judging from recent investigation at Davis 24 , sugar beets extract moisture from a depth of 6 feet ; tomatoes from about 7 feet. Sugar beets showed little evidence of need for water until the first 4 feet of soil was dried to the permanent wilting percentage. In contrast, potatoes growing in an Aiken loam indicated a need for 2i Doneen, L. D. Studies in the irrigation of sugar beets. Pacific Eural Press 131 (8):307. 1941. Bul. 670] Irrigation by Sprinkling 121 water when the first foot of soil was dry, though there was still available water in the second foot. 25 This difference is attributed partly to the soil, as other crops growing in this soil require frequent irrigations. Studies at Shaf ter 26 showed that cotton extracts moisture from a depth of 6 feet. Alfalfa roots extend well below 6 feet. 27,28 Generally, deciduous trees are deep-rooted. Prunes 29 extract moisture from below 9 feet, and walnut roots 30 are active to a depth of 12 feet. Although the roots of citrus 31 ex- tend below 4 feet, most of the extraction is from the first 2 feet. In general, it is considered good irrigation practice to wet the soil to the depth of the rooting zone (that depth from which water is extracted by plants) each time it is irrigated. Sprinkling is practiced on many shallow soils underlaid with rock, hardpan, gravel, or a high water table. Care must then be exercised ; if too much water is applied, the soil may become waterlogged, or an appreciable amount may be lost by deep percolation. Sometimes a shallow sandy loam surface soil is underlaid with a very heavy subsoil. Because of the increased root activity in the surface soil, and also because of its lower water-yielding capacity, the moisture is extracted from the surface soil sooner than from the subsoil, and rather frequent irrigations are required to maintain available moisture in it. "When excessive applications are made, the excess water drains into the already wet subsoil, and may actually keep it in a more or less saturated condition. Roots tend to die out in the wet subsoil and become concen- trated near the surface. Sprinkling has been found especially beneficial in overcoming this condition in citrus orchards. Light, frequent appli- cations maintain available moisture in the surface soil without contrib- uting water to the subsoil. Over a period of time this permits roots to develop to a greater depth. Greatly improved tree conditions have been noted in some instances. Under different soil and crop conditions, applications varying from 1 to 6 inches or more might be desirable. Sprinkling is most extensively 25 Edlef sen, N. E. Effect of soil moisture characteristics on irrigation requirements. Agr. Engin. 18:247-50. 1937. 26 Beckett, S. H., and Carroll F. Dunshee. Water requirements of cotton on sandy loam soils in southern San Joaquin Valley. California Agr. Exp. Sta. Bul. 537:1-48. 1932. (Out of print.) 27 Beckett, S. H., and M. E. Huberty. Irrigation investigations with field crops at Davis, and at Delhi, California 1909-1925. California Agr. Exp. Sta. Bul. 450:1-24. 1928. (Out of print.) 28 Unpublished data, experiments in progress. 29 Hendrickson, A. H., and F. J. Veihmeyer. Irrigation experiments with prunes. California Agr. Exp. Sta. Bul. 573 : 1-44. 1934. 80 Veihmeyer, F. J., and A. H. Hendrickson. Soil moisture as an indication of root distribution in deciduous orchards. Plant Physiol. 13(1) : 169-77. 1938. 31 Beckett, S. H., Harry F. Blaney, and Colin A. Taylor. Irrigation water require- ment studies of citrus and avocado trees in San Diego County, California, 1926 and 1927. California Agr. Exp. Sta. Bul. 489:1-51. 1930. 122 University of California — Experiment Station practiced under conditions requiring light frequent applications, where it has a relatively greater advantage over other irrigation methods. Ap- plications of 2 to 4 inches are most common, and satisfy requirements in many places. To summarize : a sprinkler system should be planned with adequate capacity for crop requirements, and it should be capable of applying proper amounts at such rates that the soil can absorb the water without runoff. Sprinkler Capacity. — Having decided upon the general method of operation, the required capacity of the system, and the suitable rates of application, one comes to the more technical phases of the design prob- lem. The general layout of the system and the spacing of sprinklers must be considered. To determine the proper spacing for best distribu- tion of water one must understand sprinkler performance. The capacity of the sprinklers may be governed by the dimensions of the field and the total capacity of the system, or by the rate at which the water can be applied. For large portable systems, with portable pumping plants, the capacity of the system is usually limited by the power available for pumping, and the capacity of the sprinklers by the number that can effectively be used, which depends upon spacing and field dimensions. Sometimes the capacity of a system is governed by the sprinkler dis- charge required to make a given application in a certain time interval, and by the number of sprinklers for most effective arrangement. The required sprinkler discharge, to apply a given quantity of water in a certain period, can be conveniently calculated from the following expression : q = ^r (29) where q is the discharge of each sprinkler in gallons per minute ; d is the average depth of application, inches; #! is the sprinkler spacings along the line ; $ 2 is the spacing between sprinkler lines ; and T is the period of operation in hours. Consider, for example, a typical case — an 80-acre field to be irrigated by a portable system supplied by a portable pump from a field ditch running lengthwise through the middle of the field. From the standpoint of crop requirements and method of operation a total capacity of 5 gal- lons per minute per acre or 400 gallons per minute will suffice. The net field dimensions are 1,290 by 2,610 feet. Six hundred feet of pipe will reach from the pump near enough to the edges of the field for the last sprinkler to cover. Fifteen sprinklers 40 feet apart, twenty sprinklers 30 feet apart, or thirty sprinklers 20 feet apart might be used on a 600- foot line; or twice as many on 1,200 feet of pipe. Bul. 670] Irrigation by Sprinkling 123 Suppose a 12-hour operating schedule is preferred, and a 4-inch appli- cation is considered adequate. Because of wind conditions, the lines can be moved a maximum of 60 feet. To make the number of sets come out even, the lines should actually be moved only 59 feet. To apply 4 inches of water in about 11 hours, allowing for moves, with # 1 =40, and $ 2 =59 would require a sprinkler discharge of 4 X 40 X 59 q = = 9 gallons per minute . * 96X11 With a 40-foot spacing on the line, an effective arrangement necessi- tates either fifteen sprinklers on a single line, or thirty on a split line. With thirty sprinklers, the total capacity would be only 270 gallons per minute, which is not adequate. A minimum capacity of zQV-= 13.3 gal- oU Ions per minute is considered necessary. With three moves per 24 hours, allowing 7 hours' net operating time between moves, the sprinkler capac- ity would be 4 X 40 X 59 q = — = 14 gallons per minute. 96X7 * P Thirty sprinklers would have a total capacity of 420 gallons per min- ute. With three moves, or 177 feet of land covered per day, a total of 2590 __ 246 days, or 44 settings would be required to cover the field. 177 This would allow for two 4- inch irrigations per month. Size of Pipe. — The size of pipe required in a sprinkler system depends upon several factors. In general, the size should be such that the friction loss is reasonable. A high friction loss in the pipe lines makes it difficult or impossible to distribute the water uniformly over the area. Where water is supplied under pressure, pipe sizes must be ample to carry the required flow and leave a sufficient pressure for satisfactory sprinkler performance. Where pumping is required, the problem is fundamentally one of determining what sizes result in the lowest annual cost. Small pipe is less in first cost but it increases the pumping head and cost of pumping. There are two methods of approaching this problem. One is to set down, in the form of an integral equation, all items entering into the cost and differentiate to determine what pipe size gives minimum cost. 82 The other is to estimate the total annual cost for different pipe sizes that appear practicable, and in that way to determine the most economical size. Both necessitate assumptions regarding cost and life of pipe, that are at best only approximations. 32 This method of attack is discussed in : Gladding, R. D. Special design problems in distribution systems for sprinkler irrigation. A paper presented at the Pacific Coast Section Meeting, American Society of Agricultural Engineers, Corvallis, Oregon, February 16, 1940. 22 p. (Mimeo.) 124 University of California — Experiment Station Friction losses in pipe of different kinds are discussed in detail in the section "Hydraulics of Sprinkler Systems." When using friction-loss tables or diagrams for the purpose of selecting pipe sizes, one may well use conservative friction factors, or coefficients, or add a certain amount — about 10 per cent — to the required flow as a factor of safety. The same principles apply to the design of stationary systems. The uni- formity of distribution is then especially important, Careful considera- tion must be given to the arrangement and spacing of sprinklers, and to the type of sprinkler pattern. Minimum cost will result with low rates of application which permit smallest pipe sizes. Although labor costs are lower, stationary systems should also be designed for convenient operation. Many other problems enter into the design of sprinkler systems. Some of these have been discussed in connection with the descriptions of the various systems. Although the design of a sprinkler system is essentially an engineering problem, it is also one that requires a knowledge of agriculture. ACKNOWLEDGMENTS The author wishes to acknowledge the cooperation of the various sprinkler-equipment manufacturers who furnished pipe and sprinklers for tests reported in this bulletin. He also thanks Professors F. J. Veih- meyer, B. D. Moses, M. R. Huberty, A. F. Pillsbury, and Mr. J. B. Brown for reviewing the manuscript and offering many helpful sug- gestions. 16im-ll,'42(324)