I Practical Irrigation and Pumping Water Requirements, Methods of Irrigation and Analyses of Cost and Profit BY BURTON P. FLEMING, M.E. Associate Member American Society Civil Engineers; Head of Department of Me- chanical Engineering, State University of Iowa ; formerly Irrigation Engineer, Office Irrigation Investigations, U. S. Department of Agriculture FIRST EDITION FIRST THOUSAND NEW YORK JOHN WILEY & SONS, INC. LONDON: CHAPMAN & HALL, LIMITED 1915 Copyright, 1915, by BURTON P. FLEMING PUBLISHERS PRINTING COMPANY 207-217 West Twenty-fifth Street, New York ACKNOWLEDGMENT The following firms have kindly extended the use of cuts for illustrations: American Well Works, Aurora, 111. Buffalo Steam Pump Co., Buffalo, N. Y. Byron Jackson Machine Works, San Francisco, Cal. De La Vergne Machine Works, New York, N. Y. Fairbanks Morse & Co., Chicago, 111. Keystone Driller Co., Beaver Falls, Pa. Stover Engine Works, Freeport, 111. 111 CONTENTS PAGE INTRODUCTORY NOTE vii CHAPTER I THE AMOUNT OF WATER REQUIRED i The difficulties in the solution of the problem. Amount used affected by skill of irrigator. Amounts of water actually used. Periods of irrigation for different crops. Amount of water used at each irrigation. Best size of irrigating stream. Acres per day irrigated. Amount of pumped water to allow per acre. Dry-farming methods an aid to irrigation. CHAPTER II SOURCES OF SUPPLY n Legal considerations surrounding use of surface sources. Adequacy of supply, surface sources. Legal considerations, underground sources. Adequacy of underground supply. Geology of deep wells. Geology of shallow wells. CHAPTER III THE FLOW OF UNDERGROUND WATER 20 The rate of flow. The ground-water surface. The draw- down. Theory of flow into driven wells. Equation of time of pumping. Limitations of formulae. Interference of wells. Practical limits of draw-down. Size of well tube. CHAPTER IV STRAINERS 37 Definition. Special cases governing depth of strainers. Kinds of strainers. Conclusions on strainers. CHAPTER V WELL SINKING -45 Practical suggestions. Well-drilling machinery. Spudding machines. Jetting machines. Rotary machines. Operation of well-sinking. The well pit. Weights of pipe. Tapering of borings. Stove pipe casing. VI CONTENTS CHAPTER VI PAGE PUMPS, PUMPING MACHINERY, AND APPLIANCES 58 Caution needed in selection of pumps. Pumps which have been proposed. What should decide make and type of pump to use. Standard types of irrigation pumps. CHAPTER VII CENTRIFUGAL PUMPS 63 The centrifugal pump described. Specifications for centri- fugal pumps. Characteristics of centrifugal pumps. What the plant designer or operator must know. The efficiency of the pump. Pump curves. The selection of a pump. Size of engine or motor. Pump builders' diagrams. Pump equa- tions. Locations and conditions suitable for centrifugal pumps. CHAPTER VIII DIFFERENT TYPES OF INSTALLATION FOR CENTRIFUGAL PUMPS . 108 Plant No. i. Friction effects and example. Drive. Prim- ing. Ejector primer. Discharge pipe. Water hammer. Fittings. Pump. Plant No. 2. Pump pit and arrangement of belt drive. Well pipe and strainer. Pump foundation. Fittings. Drive. Plant No. 3. Advantages. Pump and motor speeds. Wiring. Attention required. Electric drive the ideal arrangement. Plant No. 4. Application of the low-lift plant. Plant No. 5. Suspension frame. Step bearing and end thrust. Stages. Priming. Discharge pipe and details. Driving pulley. Plant No. 6. Vertical electric drive. Means of water measurement. CHAPTER IX TYPICAL PLANTS NOT USING CENTRIFUGAL PUMP 130 The question of sand. Duplex and triplex pumps. Capac- ity limited. Advantages and efficiency. Vertical rods. Deep- well pumps. Speed. Drive. Important details. CHAPTER X COST OF PUMPING 135 Importance of knowledge of pumping costs. Plant owners' statements unreliable. Factors affecting cost of pumping, (i) Cost of power. Head and quantity pumped determine power requirement. Steam. Gasoline. Crude oil and dis- tillates. Producer gas. Electricity. (2) Interest on first cost of plant and depreciation. (3) Maintenance and repairs. (4) Attendance. CONTENTS VU CHAPTER XI PAGE THE QUESTION OF COST AND PROFIT ON A SMALL FARM IRRIGATED BY PUMPED WATER 152 Elements of the problem. A fair estimate of yield. Cost of crop production. Shipping costs and market rates. Money rates, etc. Demonstration of a problem. CHAPTER XII RESERVOIRS 160 Their necessity. Water tightness. Construction. Rodents. Capacity. Depth. CHAPTER XIII PRIME MOVERS 165 Steam engines and boilers. Throttle-governed engines. Fly-wheel governor engines. Boilers. Auxiliaries and fittings. Boiler insurance. Gasoline engines. Difficulties in opera- tion. Over-rating. Oil fuels. Distillate engines. The gas producer and engine. Principles of operation. Conditions warranting adoption of gas-producer plant. Electric motor drive. CHAPTER XIV THE CENTRAL STATION PUMPING PLANT . . . . . . .185 Locations suitable for central station plants. Conditions governing feasibility, (a) Adequacy of supply, (b) Head. (c) Suitability of tract for agricultural purposes, (d) Ship- ping and marketing facilities, (e) Size and shape of tract, (f) Ownership, (g) Possibilities of co-operation, (h) Fuels and prices. Pumping season. CHAPTER XV WINDMILLS 197 The field of the windmill in irrigation. Kinds. Size. Governing. Selection of a mill. Power of a mill, wind records, and power diagrams. The problem of determining best size of pump cylinder, concrete case. Size of pump for direct-acting mill. Size of pump for geared mills. The pump cylinder. INTRODUCTORY NOTE To people living anywhere on the American continent west of the hundredth meridian, the practice of irrigation as a fundamental necessity in the production of crops is so common a matter and so thoroughly a part of their daily observation and experience, that they scarcely appreciate the viewpoint of the farmer of the humid regions, who stands more or less aghast at the idea of spending a large sum per acre to secure for the growth of the most common crops that moisture which under his conditions, nature itself provides. The author does not agree with those, who in the at- tempt to make a virtue out of a necessity, go so far as to maintain that the lack of moisture which makes irrigation necessary over perhaps one-third of the area of the United States is in reality a blessing. It cannot be denied that the irrigation farmer of the west cultivates a soil in which many important plant foods, leached out by natural rain- fall from the soils of the more humid regions, are still retained, making it in general, therefore, a rich and pro- ductive soil when water is applied; he is not, for some years at least, and in many locations never will be, bothered by the drainage problem; he enjoys a climate in which the abundance of sunshine makes not only for rapid crop growth, but also for the physical well-being of himself and family while most important of all, perhaps, from the standpoint of agronomy, he can control the moisture supply and he is comparatively free from those seasonal variations in rainfall and temperature which make farm- ing in the great valley of the Mississippi, for example, X INTRODUCTORY NOTE so uncertain an enterprise. On the other hand, it can neither be denied nor evaded that these advantages are secured and the region made habitable, only by the expen- diture of comparatively large sums of money in developing the natural water resources, in preparing the land for irri- gation and applying the water. All of these expenses the irrigation farmer must bear over and above those numerous and sometimes heavy financial burdens attending actual crop production in more humid regions, while securing also the advantages of education, public improvements, and the protection of government. This additional financial burden under which the irrigation farmer labors is therefore a real one, and that he bears it complacently, is able to pay a comparatively high rate of interest on farm mort- gages, finance extensive public improvements, such as good roads, and maintain school systems quite the equal of those in more fully settled localities in the humid regions, is evidence that the advantages accruing through irriga- tion, as above noted, are tangible and have a money-earning or real economic value. The cost of irrigation is enormous. It is estimated by the census of 1910 that up to July i of that year, about $308,000,000 had been spent by private and public or quasi-public enterprises in reclaiming the 14,000,000 acres under cultivation by irrigation in the West. It is this tremendous investment, simply in the means of supplying moisture for growing crops, that excites the wonder of the farmer or banker of the humid sections, at the ability of the Western communities to stand the strain, for under whatever arrangement the various irrigation projects are or have been financed it must not be forgotten that all even- tually are paid for by the products of irrigated land. Cer- tainly an appreciation of the meaning of the figures above cited, should impress upon the Westerner the tremendous INTRODUCTORY NOTE XI importance and extent of the subject of irrigation which he usually takes so much as a matter of course. The sum above mentioned has been spent almost entirely in the development of means of supplying farms with water by gravity methods of distribution, but it is significant of the rapidity with which development is pro- ceeding in the arid West that the pumping of water for irrigation purposes is attracting more and more attention each year and already it is estimated that 250,000 H.P. of pumping engines and motors are engaged in this work, that nearly $9,000,000 have been expended in the necessary plants, while the acreage capable of being irrigated amounts to over 260,000 acres. In California and certain other sections favorable to the growth of citrus fruits the pumping of water for the irrigation of lands not otherwise susceptible of irrigation has been a common practice for many years and the means and appliances have been well worked out. In other parts of the West, however, it is only comparatively re- cently that farmers have thought seriously of attempting to irrigate on a commercial scale by any other means than that of conveying surface water to the land by gravity. The water was diverted from a surface stream or taken from a storage reservoir. Land irrigated or susceptible of irrigation by gravity canals has become in the course of time so high in price, or so scarce, or the means necessary for gravity irrigation have become so expensive, that at present the chance is very small for the man of strictly limited resources to secure a foothold in any of those parts of the West already well developed. There still remain, of course, vast areas of land whose latent agri- cultural possibilities merely wait the touch of water to make such land immensely productive. Much of it lies either on the higher benches or mesas adjacent to irrigated Xii INTRODUCTORY NOTE valleys and above high-line canals, or it is found in numer- ous localities where topographic and climatic conditions preclude the presence of surface streams. Such land, in many cases, may be homesteaded, or if in private owner- ship may be bought at prices ranging from $5 to $40 per acre, depending upon how successfully it or other lands in the neighborhood have been "dry farmed. 77 Where such land has beneath it a water-bearing formation, and general economic conditions are favorable, we have a location suitable for the profitable development of a scheme of irrigation by pumping. Other locations equally favorable may often be found where it is possible to pump water from a high-line canal upon land lying above the canal, and in other cases, where a long and expensive canal may be neces- sary to reach a suitable location for head-works, a careful study of the problem may show that a decided saving in first cost and operation as well as maintenance, will result from the installation and operation of a pumping plant to place water upon high-lying land adjacent to a surface stream. Particularly may this be so if there is a possi- bility of generating the power necessary for pumping at some near-by point where the conditions favor an inex- pensive hydro-electric development and transmitting the power thus developed to the most feasible location for the pumping plant. It is not improbable that, if there were any assurance that the necessary skilled attendance for such a plant would be always provided, it would pay some irrigation concerns with which the writer is familiar to abandon their present long and costly canal lines, now so difficult to maintain, install a pumping plant and either buy the necessary power or build a simple hydro-electric plant themselves. If properly installed and maintained such plants would be sure to obviate those serious losses frequently sustained through canal breaks, which are a INTRODUCTORY NOTE Xlll common occurrence in the very midst of the irrigation season, on canals located in canyons with very steep cross slopes and in unstable or porous material. In this connection, it may be stated that a remarkable development has occurred within the past few years in the installation of irrigation pumping plants in Idaho and Oregon, along the Snake River, where, because of its slope FIG. 1. A high-grade pumping plant on the Snake River, eastern Oregon. This plant is supplied with current at 66,000 volts and has eight motors of total capacity of 1,150 horse-power driving centrifugal pumps with aggregate working capacity (normally five pumps in operation) of over 22,000 gallons per minute. The water is raised to three elevations, 55, 84, and 110 feet above the supply canal which conveys water to the plant about 3,600 feet inland from the river. This plant exhibits probably the most advanced example of mechanical design in irrigation work in the West. and the configuration of the valley, gravity systems are very expensive. Electrical power is being generat- ed in large amounts on the river and its tributaries by several competing companies and is sold at from $20 to $28 per horse-power for the irrigation season. Several very large electrically driven plants have recently been con- x] v INTRODUCTORY NOTE structed which pump water from the river upon lands lying as much as 150 feet above, and in areas as large as 15,000 acres, requiring as much as 3,000 H.P. in one plant. In the vicinity of Payette, Idaho, a very unusual develop- ment of this phase of irrigation is found, there being about 1 60 plants in a length of 20 miles along the river, utilizing from 2 to 1,000 H.P. each. A large number of the plants lift water from a high-line canal upon orchard lands lying above the canal, while others divert directly from the river. Nearly all of these plants have electrically driven centrifugal pumps and many fine fruit ranches depend entirely upon them for water supply. Before any decision can or should be reached with regard to the feasibility of a pumping-plant project, a careful and systematic study should be made of the matter in all its phases. This applies to the small individual plant as well as to the large central plant, perhaps with greater force to the former since such plants often represent the entire working capital of an individual who stakes his future upon the success of his venture. Ill-considered plans, wrong and costly types of machinery, too high a "pumping head," and a disregard of such simple factors as nearness to markets and suitable crops to grow with high- priced water have contributed, in the many cases which have come under the writer's observation, to the lack of success attending a pumping-plant investment. It is not enough that there shall be cheap land and a water-bearing stratum beneath it, or some other adjacent source of water supply. This water must not be at great depth below the surface to be irrigated or the source be so far away that long pipe lines are required; otherwise it will not pay with any ordinary crop to develop the water supply, however extensive or unfailing it may be. The crop grown must in any case be one requiring relatively little water, and must INTRODUCTORY NOTE XV give high crop value per acre, while the machinery used should be of such type as will give the service desired with a maximum of running economy and a minimum of main- tenance charges. It will be the endeavor of the writer in the present volume, to consider the irrigation problem chiefly from the pumping standpoint, treating of those matters which interest the man considering the installation of a small pumping plant from both the standpoints of design and operation. It is hoped further that the author's suggestions will be helpful to the contractor who specializes in the machinery of pumping plants and be of some assistance to the engineer who is called upon to design the central station plant. Beginning with estimates of water requirements of different crops on different soils and in different localities, the writer will consider in turn: the matter of wells and well-sinking; pumps and pumping machinery suitable for different depths and volumes together with typical designs for certain assumed conditions; prime movers including a discussion of oil engines, gas producers, etc.; windmill irri- gation, chiefly from standpoint of co-relation between wind velocities and pump size; the question of cost and profit in pumping and a method of estimating the latter for certain conditions; and, finally, the central station plant and its possibilities. An attempt will be made to make the discussion as general as the nature of the subject will allow, but where specific instances or trade names are thought helpful they will be given. The writer brings to his aid in the preparation of this volume an experience of over eight years in irrigation work, during which time he has covered most of the Western States, and has had much opportunity to observe and study irrigation conditions. He has had the benefit of consider- able direct personal' experience (some of it rather bitter, XVI INTRODUCTORY NOTE indeed) in the matter of irrigation pumping and on several questions connected therewith is able to give the results of experimental work. The writer has drawn his data freely from government reports, experiment station bulletins, particularly those written by himself, and various other sources specifically mentioned in the text. PRACTICAL IRRIGATION AND PUMPING CHAPTER I THE AMOUNT OF WATER REQUIRED The Difficulties in the Solution of the Problem. A necessary preliminary to the consideration of any problem in the water supply for irrigation, is a more or less definite knowledge of the amount of water required in the irrigation of the particular crops it is desired to grow upon the par- ticular lands it is desired to irrigate. At the outset, one is confronted with a difficulty in securing definite information upon the matter, due to the fact that most of the data we have on what has been called "The duty of water," has been obtained by investigations made on gravity systems. Except under exceptionally well-managed canal systems, especially those where the water user is charged on the basis of the amount used, at a certain price per acre-inch* or acre-foot,* there is always a temptation for the irrigator to use more water than is really necessary because it costs him little or nothing and he usually works under the time- honored delusion that "The more water the more crop." Moreover, most measurements have been taken where but little or no attempt has been made to eliminate seepage losses in distribution, and where but little attention is paid * The acre-inch is the amount of water which without loss of any sort would cover a level area of one acre to the depth of one inch. The acre-foot is twelve times this amount. I 2 t .\ 1ft PRACTICAL ^IRRIGATION AND PUMPING to the prevention of loss through leaky and imperfect ditches or field laterals. Consequently results obtained under these conditions are not to be considered comparable with those bound to follow when the irrigator realizes that every revolution of the pump by which the water is raised, represents a definite amount to be deducted from the returns of the crop grown and where consequently we are apt not to find pumped water escaping from fields into adjacent roadways or pouring through breaks in poorly built laterals while the irrigator discusses politics with his neighbor. Although therefore, so far as the crops them- selves are concerned, the water requirement is the same whatever the source of the water supply or the method of distribution, the human element which enters into the problem makes it possible in estimates to allow a very much higher duty for pumped water than for water de- rived from surface sources. If in connection with the pumping plant a reservoir is used and losses in distribution are prevented by employing pipe or concrete distributaries and if, also, great care is observed in laying out the fields in such a way as to reduce evaporation and needless seepage to a minimum, the amount of water needed for a crop may be but little more than its absolute water requirement. Amount Used Affected by Skill of Irrigator. The amounts of water found to be used in the irrigation of crops, when supplied by the gravity systems, present considerable varia- tions in different localities due to differences in climatic conditions, topography as affecting surface or subterranean drainage, soil porosity, the character of distribution, and finally the skill of the irrigator. It is the opinion of the writer, based on his own measurements and upon observa- tions made in various parts of the West under considerable range of altitude and latitude, that the irrigator's skill has probably a greater effect upon duty of water measurements THE AMOUNT OF WATER REQUIRED than have any or perhaps all of the other conditions men- tioned above. That is to say that a careless irrigator in a northern climate may use more water in the irrigation of a crop grown upon a dense soil than would a careful irrigator use for the same crop on a deep sandy soil in the intensely hot valleys of New Mexico or Arizona. The best the engineer can do, therefore, especially when designing a gravity system, is to base his estimates upon averages, not forgetting that local irrigation customs and practices may change easily the values subsequently obtained in the practice of the water users by 25 to 50 per cent. Amounts of Water Actually Used. The writer has measured the duty of water on fields of alfalfa in western Nebraska and southern Wyoming where 52 acre-inches per acre were used and he has seen abundant alfalfa crops grown in New Mexico with 40 inches. Likewise he has also observed conditions in New Mexico where irrigators thought it impossible to grow a crop of alfalfa on less than 60 acre-inches. In the following table are given limiting values of the duty of water from various crops as determined in different localities with gravity systems. TABLE I DUTY OF WATER, VARIOUS CROPS, GRAVITY SYSTEMS ACRE- INCHES PER ACRE Alfalfa Corn Wheat Oats Orchards Small Fruits Garden Truck Patches 36 to 60 24 to 30 18 to 26 18 to 24 18 to 20 30 to 36 The quantities in the table are based upon averages secured by measurements made at the edge of the field. The duty at head-gate of the main canal will increase the 4 PRACTICAL IRRIGATION AND PUMPING above values by 30 to 50 per cent, due to seepage and evap- oration losses in distribution. Periods of Irrigation for Different Crops. Probably as satisfactory a basis of estimate as it is possible to obtain in figuring upon water requirements is as to the number of times a crop must be irrigated and the probable amount supplied at each watering. This is for the reason that practice in regard to such common crops as are included in the above table is pretty well standardized, and the number of irrigations necessary or desirable will not be found to vary appreciably from the mean for any given locality. Thus in the case of alfalfa, in the northern climates this crop will make two and sometimes three cuttings; in the southwest, four generally and sometimes five cuttings are secured. The number of cuttings is also of course influenced by altitude. The common practice is to give this crop a thorough irrigation at the beginning of the season to get it well started, a second previous to the first cutting, a third shortly after the crop has been removed, not until the fresh growth has attained a height suffi- cient to give it some protection against excessive evapora- tion and scalding, a fourth about a week or ten days before the second cutting, and so on for each succeeding cutting. Thus each cutting will secure at least two irrigations, or the number of irrigations will be about double the number of cuttings. It might be said in passing, that usually in the growing of alfalfa, rainfall, unless most unusual in amount, has but little effect upon the practice of irrigation, and rarely will cause the irrigator to miss a regular watering or will diminish appreciably the amount of water which should be used during a perfectly dry season. Practice in regard to grain crops is more variable, de- pending upon locality, amount of rainfall, etc. In most cases an irrigation is necessary either just before or imme- THE AMOUNT OF WATER REQUIRED 5 diately after plowing and planting in order that moisture conditions may be proper for germination. A second irri- gation will follow possibly at the end of two weeks or seventeen days, and the third and usually the final is given while the grain is in the milk. An extra watering may be necessary between the second and third mentioned, in the absence of normal summer rains, thus giving a minimum of three and a maximum of four irrigations for wheat, oats, barley, flax, etc. Corn, sorghum, and Kaffir corn are crops usually re- quiring less water than broad culture crops, since they are cultivated more or less frequently, thus preventing rapid soil evaporation, besides which there soon is formed in the process of growth a dense shade further reducing soil evaporation. Three irrigations are usually found ample for such crops. Orchards require less water per acre than most crops, due to the fewer number of plants per acre and to the fact that greater care ordinarily is taken in distribution in orchards. Usually three and not to exceed five irrigations are given and the tendency is towards the lower limit when proper cultural conditions are maintained. Truck gardens, owing to the greater sensitiveness of the plants to unfavorable moisture conditions, must be irri- gated with care and with not excessive amounts of water. Thus, although the truck garden may need to be irrigated over the entire area from six to eight times or possibly more during a season, yet the amount of water used will probably be less than that required for a broad culture crop or one which is deep-rooted. Amount of Water at Each Irrigation. Nothing is better known by the irrigator of extensive experience than that "it does not pay" to use a small stream of water or a "small head of water" in irrigation. A small stream seems to dissipate and lose itself when one attempts to spread it 6 PRACTICAL IRRIGATION AND PUMPING over an extensive area, so that if a certain volume is avail- able, as for example, i acre-foot, it might be found im- possible to spread this amount uniformly over 2 acres with a small stream, no matter how carefully the land had been prepared or what its character. On the other hand, with '"a good irrigation head" as he would call it, a skilled irrigator would without difficulty distribute the acre-foot over the acre and probably secure great uniformity in its distribution, so that no one part would be soaked and another part be left practically dry. The difficulty in dis- tribution may be said to increase even on well-prepared land as the total quantity applied decreases. Experiments conducted by the writer on sandy open mesa soil, showed that it was next to impossible to secure uniform distribution even on small carefully prepared plats, with less than 3 acre-inches of water per acre at an irrigation, and that very much greater success was attained when 4 acre-inches were applied at an irrigation. In actual experience under gravity irrigation systems where water is used with but little thought of economy, 5 and 6 acre-inches are usually applied per acre at an irri- gation of alfalfa, 5 acre-inches with grains, 3 to 4 with orchards, and about 3 with truck patches. It will be seen, therefore, that in general, 3 acre-inches is the probable minimum which may be allowed per irrigation, even under the careful system of irrigation which must be assumed as existing or will exist under a pumping project, and for most cases, probably 4 acre-inches would represent the value attained by the average irrigator even when im- pressed, as every one using pumped water should be, with the supreme necessity for economy in its use. Best Size of Irrigating Stream. Although, as suggested in a previous paragraph, a small "irrigating head" is un- economical, on the other hand, large streams are equally THE AMOUNT OF WATER REQUIRED 7 conducive to waste when too large for an irrigator to handle properly. The size of stream which one man may handle, will be determined entirely by the character of crop being irrigated, the thoroughness with which the land is prepared for irrigation, its slope, and the character of the soil. Each of these conditions is more or less dependent upon the other; thus with a grain crop the same care in preparing the land would not be necessary or expected as in the case of a melon crop. However, in the case of alfalfa or grains, where well-defined furrows do not exist, a larger stream could be used to advantage than would be desirable for furrow irrigation, and again a crop on land carefully leveled and prepared, could be irrigated by one man with a larger stream flow than where the surface is so uneven as to require considerable of the irrigator's time and skill to conduct water to the high spots. Also on sandy, open soil, one man could handle a large stream to less advantage than a small ofie, due to the greater tendency of the water, in the former case, toward erosion both of field and laterals, a condition not existing on an adobe or dense loam soil. In general, one man may handle streams of the sizes given by the following table on different soils and with the special methods of irrigation pertaining to the crops indi- cated. TABLE II MAXIMUM NUMBER OF GALLONS PER MINUTE, WHICH ONE MAN MAY HANDLE SUCCESSFULLY IN IRRIGATION Sandy Soil Dense or Heavy Soil Alfalfa Grains Orchard Truck Garden 450 60O 400-500 300-450 250-300 600-900 500-700 400-500 300-350 8 PRACTICAL IRRIGATION AND PUMPING The minimum-sized stream with which a man may do good work is 200 gallons per minute under usual conditions, and if a flow no greater than this can be secured from a pumping plant it will be better, in general, to store several hours' or even days' supply in a tight reservoir and use a large stream for a short time rather than attempt to accom- plish anything with so small a stream. The question of reservoirs will be more fully considered in a subsequent chapter. Acres per day, Irrigated. It is of some importance in figuring costs of irrigation and in estimates on sizes, to know how much acreage the average irrigator may cover with irrigation streams of different sizes when applying various quantities per acre. The following diagram will enable this to be determined graphically. I- o a* 2 I ?/ 100 800 800 400 500 600 700 800 900 Size of Irrigating Slream-gals./inhi. DIAGRAM 1 SHOWING THE ACREAGE WHICH ONE MAN MAY COVER WITH IRRIGATING STREAMS OF DIFFERENT SIZES AND APPLYING VARIOUS QUANTITIES It will be noted by the diagram that when applying 4 acre-inches per acre with a poo-gallon-per-minute stream, one THE AMOUNT OF WATER REQUIRED 9 man may cover nearly 5 acres. This acreage is to be con- sidered about the average maximum performance of one man on well-prepared ground and the table will be found to correspond with the experience of practical irrigators generally, though some variation is to be expected according to local conditions. Amount of Pumped Water to Allow per Acre. Sum- marizing what has been said in previous paragraphs, we may state the case briefly as follows : 1. Pumped water will be used with greater care than water supplied by gravity because of its recognized greater cost. 2. For a successful practice the quantity applied per irrigation cannot be less than 3 acre-inches per acre, but need rarely exceed 5 acre-inches. 3. The number of irrigations during the season will depend upon the crop grown, the locality, the soil conditions, the rainfall with some crops and to a considerable extent will de- pend upon local irrigation customs and practices. 4. Using the average number of irrigations and assuming 5 acre-inches per acre per irrigation for alfalfa, and 4 acre-inches per acre for all other crops, we may prepare the following table to show the probable duty of water for various crops, grown under average climatic and soil conditions with pumped water. The following quantities are to be regarded as ample only where the greatest care is taken to prevent losses in dis- tribution and where the most advanced ideas in cultivation and soil treatment are adopted and put into practice to prevent unnecessary losses by evaporation and seepage. 10 PRACTICAL IRRIGATION AND PUMPING On very gravelly soils newly put in cultivation double the quantities below given might scarcely suffice TABLE III PROBABLE DUTY OF PUMPED WATER Crop Alfalfa per Cutting Small Grains Season Corn Season Sorghum Kaffir Corn Season Orchard Small Fruits Season Melons Season Truck Garden Season Acre-inches per acre . 10 12 20 26 12 24 24-30 Dry Farming Methods an Aid to Irrigation. We cannot too strongly emphasize the fact and shall refer to it from time to time, that financially successful irrigation farming with pumped water is only possible when the idea is thor- oughly ground home that pumped water is expensive and the same methods and practices cannot be allowed in irrigation with pumped water, as prevail on farms supplied from gravity canals. Only by adopting and following the best practices of dry farmers in the conservation of soil moisture, will the farm ledger show a satisfactory profit when pumped water is used for irrigation. This is true un- der any circumstances, but applies with special force to those cases where the total lift of pumped water equals or exceeds 50 feet. For orchard fruits, of course, or special crops such as melons, it is profitable to use water pumped from any reasonable depth, but considerable exercise of good judg- ment and first-class business management are necessary to wring a profit out of a pumping plant under other conditions. CHAPTER II SOURCES OF SUPPLY Legal Considerations Surrounding Use of Surface Sources. Careful investigation should be made of the proposed source of supply, to determine its adequacy for the proposed scheme and the nature of any legal or physical difficulties likely to be encountered, before any serious con- sideration is given to the construction or economic details of a pumping project. In those instances where it is pro- posed to pump water from a flowing stream or an existing canal, the legal right to the use of the water should first be looked into. If the source proposed be a natural flowing stream or river, much care and attention should be paid to securing this lawful right to the use of water, if the proj- ect lies in any of the states where water laws other than the doctrine of riparian rights exist and are enforced. Where the old common-law doctrine of riparian rights exists it is doubtful if any scheme having in view the abstraction of water from a flowing stream and its use in the irrigation of adjacent land would be regarded by the courts as lawful. In the arid states, on the other hand, where the right to appropriate and use the water of flowing streams and other natural sources is recognized, the procedure usually consists in making application to the State Engineer (in certain prescribed ways) for the right to appropriate a definite quantity of water, at a definite point, for a desig- nated purpose. If the proposed scheme does not conflict with other rights on the same stream, the application will be granted under certain conditions as to diligence in con- struction of the necessary works and bona-fide use of the ii 12 PRACTICAL IRRIGATION AND PUMPING water so obtained in beneficial ways. In those instances where it is proposed to pump water from an existing canal or reservoir, a contract may be arranged with the owners of the same for the right either to a continuous flow of a definite number of cubic feet per second or a certain num- ber of acre-feet during a season. In the latter case certain stipulations should be made as to the periods in which may be secured the fractional amounts making up the total quantities for the season. It is suggested that a contract calling for a certain number of acre-feet during the season is likely to be the more satisfactory to the operator of the pumping plant, since, unless a reservoir is provided in connection with the plant, it will be greatly to the advantage of the pumping- plant operator to secure relatively large flows for several short periods during the season, than a small continuous flow throughout the season, although the aggregate amount in acre-feet will be the same. Such a contract is, however, difficult to secure in most cases, and might better be avoided altogether unless some accurate and reliable means of measurement be provided which will be respected by both parties to the contract. Adequacy of Supply Surface Sources. In general, no extensive study need be made of the question of adequacy of supply in the above cases, since the most casual inquiry (except where large areas are involved) will satisfy the engineer or prospective owner as to whether there is likely to be a sufficient water supply for the purposes contem- plated. Of course, if the project involves the use of waters of a torrential stream flowing only in times of excessive rainfall, the case is one deserving careful study of all avail- able information as to rainfall (yearly normal, maximum, minimum and periods of fluctuation, rate in heavy storms, run-off, etc.), and as to the physical possibilities and cost SOURCES OF SUPPLY 13 of storage works or reservoirs. It may occasionally be found practicable, in unusually favorable locations, to store flood waters cheaply and pump from the storage basin or reservoir onto adjacent lands, when for any reason a gravity distribution is impossible. Such an undertaking, if of any extent, needs careful study from the stand- point of cost, since, if to the cost of a pumping plant be added that of storage works, the land must be fertile and the crops profitable to pay a reasonable return on the in- vestment. Legal Considerations Underground Sources. The le- gal side of the other method of securing a supply, namely, by pumping from wells, is not as important as when water is obtained from surface sources; the only apparent legal necessity at present is: to be in lawful possession of the land upon which the plant is constructed. Doubtless, with an increase in the number of plants in any given section, many interesting legal questions will arise, since there is no doubt whatever but that every additional pumping plant drawing water from the com- mon underflow, impairs the capacity of every other plant within a circle whose radius will be larger as the capacity of the plant is increased. This point is illustrated in many sections of the West, both with flowing or artesian and pumped wells. In the Pecos valley of New Mexico, the number of wells tapping the artesian source has grown so large that nearly all of the first wells sunk, which formerly spouted water many feet in the air, have ceased to flow at all and pumps are necessary at present to bring to the surface, water which now may stand in the well tubes some distance below the ground level.* * In many instances the well casings have been eaten through by corrosion at various depths below the ground level, with the result that 14 PRACTICAL IRRIGATION AND PUMPING In some districts of California, where pumping has been carried on extensively and thousands of plants are in oper- ation, the level from which water must be pumped has lowered tremendously, indicating, in the absence of struc- tural defects in the wells, that the field has been over- developed. In justice to the original and older plants, it is evident that some legal restriction should have been placed upon the construction of others, in case the evident failing and overdevelopment of the field did not of itself deter further exploitation. Adequacy of Underground Supply. The location of an underground supply and its probable adequacy when found are both matters largely of guesswork and upon which no one should venture to give an unguarded or definite opinion. There is no subject about which less is definitely known or in which the rules are subject to more exceptions. Although a geologist perfectly familiar with a region may give cer- tain opinions as to the probability of an artesian supply being found and its probable depth, unseen faults or fissures in the underlying strata may completely upset his calcu- lations. On the other hand, there are many successful artesian wells drilled upon the advice and under the inspi- ration of a local seer or some expert with a " divining rod" which may puzzle the geologists to account for at all. To a greater extent is this true of the shallower subsurface water strata which do not, as in the case of artesian sup- plies, depend necessarily upon conditions determined by much water now fails to reach the surface and leaks away underground. This not only impairs the capacity of the entire artesian field, but is helping to cause saturation of the soil and subsoil, which has made the alkali and drainage problem in this district a most urgent one. The gradual failure of this heretofore abundant artesian field may therefore be quite as much due to structural defects in the old wells as to the presence of later borings. SOURCES OF SUPPLY 15 geologic formations covering va^t areas of country, but rather upon conditions more local in their nature as regards rainfall, surface drainage, and porosity of soil and sub- soil; conditions which may vary tremendously in a single township. Beginning in Colorado and extending far into Texas, is a vast extent of territory which as recently as ten or fifteen years ago was considered to be practically without water supply of any character aside from the seasonal rains. Now in this region wells are becoming more numer- ous each year and gradually a vast country is being trans- formed from a rather dubious cattle range (due to lack of water) into a country suitable for the habitation of man. One who visited the plains of eastern New Mexico and the Texas Panhandle as recently as ten years ago would scarcely deem it possible that now in this region, serious, sober-minded American farmers by the hundreds should be attempting to make their fortunes by mixed farming, util- izing water pumped from an apparently inexhaustible underground supply for their domestic and stock needs and the few acres of crops. Although few extensive pumping plants are in operation, windmills are seen on every hand and water is encountered at such depths as make it appear reasonable to expect that in the not-distant future, small farms irrigated by water pumped by power supplied from a central source will be a commercial reality. The source of this particular supply is not well determined. The general theory accepted and advertised by real-estate boomers and others interested in the disposal of these lands or in their colonization, is that there is a vast underground river flow- ing from the distant Rocky Mountains of Colorado toward the Gulf of Mexico and that beneath every acre of this region is to be found sufficient water for its irrigation. Such a theory is doubted by geologists who are inclined to account for the widespread existence of underground water in this i6 PRACTICAL IRRIGATION AND PUMPING region upon either a slowly moving underflow in connection with some adjacent surface stream or as being a natural underground reservoir filled in past ages and the level of which may be lowered more or less rapidly according as the amount withdrawn by pumps and natural outflow ex- ceeds or is less than the amount reaching the underground reservoir by downward seepage of the natural rainfall. Geology of Deep Wells. A cross section of a valley which may be regarded as typical of many western valleys, having both artesian and shallow underground supplies, is Pervioua Stratum Impervious Strata FIG. 2. A cross section typical of many valleys in the West. given in Fig. 2. It will be noted that the source of the artesian flow lies in outcroppings of the pervious rock stratum in adjacent mountains and this water, percolating slowly through the porous stratum, will rise to the surface and perhaps give considerable pressure at the outlet of a pipe driven from the valley floor deep into the porous stratum of rock, giving a flowing or artesian well at point A. Other similar wells at B and C will give less flow or may indeed not be flowing wells at all and must be operated with pumps, because the outlet of such wells lies near the hydraulic gradient. Eventually, if a large number of wells SOURCES OF SUPPLY 17 are sunk into the porous stratum the natural flow from the first wells sunk will decrease or may cease altogether, due to the lowering of the hydraulic gradient. Eventually, if pumping is resorted to, in a large number of such wells the demand is likely to exceed the capacity of the porous stratum, the depth from which water must be pumped will become excessive, and pumping will no longer be profitable. The capacity of the artesian field is determined by the areas exposed at K K through which the rain-water may per- colate, the degree of porosity of the water-bearing stratum, and the degree of water-tightness of the upper and nether strata. A pervious stratum of relatively high density will allow of but very slow percolation, due to fluid resistance, so that though considerable pressure may be developed at the surface level of a plugged boring, the amount of water yielded by the well upon removal of the plug will be very meagre. The artesian-well question is largely one of geology and the favorable opinion of a geologist familiar with a region should be obtained before wells are sunk, for although local and unforeseen conditions may, as they have in the past, entirely invalidate a geologist's conclusion and judgment, yet it is only where a hydraulic condition exists similar to that represented by the figure that artesian wells are possible, and before putting money into a hole like an artesian boring it is advisable at least to know whether the probabilities are for or against the success of the venture. Geology of Shallow Wells. In a valley similar to Fig. 2 there may be a second body of ground water overlying the first and occupying the porous river debris deposited in the valley trough above the bed rock or impervious strata. This debris is that deposited by the river in past geologic ages, and may consist of alternate horizontal layers of gravel, sand, clay, etc., with occasional pockets of sand or gravel deposited by the ancient river in the same way as 1 8 PRACTICAL IRRIGATION AND PUMPING gravel beds and sand shoals are now continually being formed by our modern rivers. This ground water, as it is termed, extends across the valley trough at a surface level corresponding usually to the mean stage of the water in the river, though the level at which ground water may be found is often greatly affected by local conditions such as the presence of large canals on near-by benches, the seepage from which may cause local elevations of the ground-water plane much above the mean level of near-by watercourses or drainage channels. Under normal conditions the ground water will have a surface slope in the direction of the axis of the valley which will be approximately the same as the surface slope of the river and there will be a progressive movement of the ground water downstream, but at a very slow rate, due to the resistance of the materials through which it passes. In case of a rocky ledge or impervious barrier across the valley at any point, a large underground reservoir will be formed above this point and after a long- continued drouth, when the surface stream may have entirely disappeared, ground water will be found at approxi- mately the level of the lowest crest of the barrier. Where there is no barrier, the ground water will continue to flow long after the surface stream is dry and there will be a progressive lowering of level of ground water due to this flow as long as the drouth continues independently of any draught upon the underflow by pumping. Many streams in the West are torrential and may be very large rivers immediately after heavy rains. During the time of flow of the surface stream a large amount of water percolates downwards and to a less degree laterally into the sandy bed and banks of the stream, and continues its downward course until it reaches an impervious stratum, joins existing ground water, or fills the interstitial spaces in the porous material of the valley trough. Where large SOURCES OF SUPPLY 19 amounts percolate downward, as in sandy stream beds during long-continued floods, the valley bed becomes sat- urated for great distances from the river, and subsequent to the passage of the flood there will be a subterranean flow continuing for long periods and giving rise to what are truly called "underground rivers." Again, streams may debouch from the mountains onto a sandy plain across which it may flow in times of freshets in a well-defined stream bed, but at other times the stream simply disappears completely into the sands a short distance from the point of debouchure onto the plain and continues its onward move- ment very much more slowly as an underflow, thus giving rise to such conditions as are found in the Mimbres Valley of New Mexico, where underground water is plentiful, although the Mimbres River, during the greater part of the year, is represented merely by a broad strip of white sand winding across the plain. Other conditions in which we may find ground water are best represented by a great saucer-like basin filled with pervious materials which absorb the greater part of the yearly rainfall, and in which the ground-water level will rise eventually to the lowest point in the rim of the basin. Such basins usually have a surface topography so flat as to be devoid of extensive or important surface streams, the run- off during periods of heavy rainfall merely running into depressions where it remains until absorbed by the soil or is evaporated. In localities in which such conditions exist water may be encountered at shallow depths, but it is likely to be alkaline, due to the leaching out of soluble salts in the surface layers by the passage of rain-water into the ground water and the gradual concentration of these salts due to lack of drainage. Many such basins are known in the Southwest, of which the most notable is the great Estancia Valley. CHAPTER III THE FLOW OF UNDERGROUND WATER Rate of Flow. In the preliminary investigations which should precede any underflow pumping project of impor- tance, an inquiry into the adequacy of the supply and the determination of the size and number of the borings, or a decision as to the character of the works by which the supply shall be developed, must necessarily take place first. Such an inquiry must be based first of all upon some knowl- edge or information as to the rate at which water will percolate through the water-bearing materials to the gathering works. The rate of flow of the ground water is a matter upon which there has been much speculation and investigation, and although the factors governing the phenomenon are too numerous and too indefinite to make possible a satis- factory prediction for a local case, at least two conditions are known to have a very important effect in determining the rate of flow, namely, the surface slope of the under- ground stream and the porosity of the materials in which the flow occurs. Turneaure gives the following table as being applicable to the determination of rate of flow of underground water. TABLE IV RATE OF FLOW IN FEET PER DAY Material Slope of Water Surface Ft. per Mile IO 20 30 40 50 IOO Fine sand O.2 0.4 0.6 0.8 I.O 2.O Medium sand i-5 3-0 4-5 6.0 7-5 15.0 Coarse sand 40 8 o 12 O 16 o 20 o 4.O O Fine gravel free from sand 20-40 40-80 60-120 80-160 I 00-200 200-4OO 20 THE FLOW OF UNDERGROUND WATER 21 If the width of the underground stream and the average depth to the underlying impervious stratum can be deter- mined, and if the slope of the ground-water surface be meas- ured, by the relative elevation of water standing in wells sunk some known distance apart along the axis of flow, some idea of the amount of water passing in the under- ground stream may be gained by use of the above table, using as arguments slope and character of water-bearing materials. Unfortunately, it is seldom possible to ascertain, even approximately, the vertical and lateral limits of an underground stream and such calculations are apt, there- fore, to be of little real value. The Ground-Water Surface. When, as is shown in Fig. 2, a number of wells are sunk into the shallow under- ground supply, there will be during pumping a local lower- ing of the surface of the underground water which in the absence of pumping is a plane surface, which transversely to the axis of the river extends horizontally at about the level of its mean stage and which in a direction parallel to the river takes its mean slope. In the case of an underground reservoir the surface of the ground water will of course be level in all directions except during pumping. While pumping is going on, however, the surface will be relieved by a series of approximately cup- like depressions, at the centre of each of which will be a well from which water is being pumped. A vertical section through the centre of one of these depressions will show the intersections with the water surface as a pair of curved lines, hyperbolas A B, C D Fig. 3, which gradually ap- proach the ground-water plane as they recede from the welL Revolving one of these curves through a circle will sweep out a volume of which a section is G-A-B-C-D-F. While pumping is in progress, this volume will contain material from which water has been drained, and below 22 PRACTICAL IRRIGATION AND PUMPING this volume the water-bearing materials will be saturated and a flow of water will be occurring from every direction into the well tube. The rate at which water flows through MM Level of Material E Standing Water Water Bearing Water Bearing Material Saturated Bearing Material Saturated FIG. 3. Illustrating condition of water-bearing materials surrounding well tube, during pumping. the water-bearing materials towards the well tube will be a measure of the capacity of the well. This rate depends upon two conditions, namely, the extent to which the water level has been lowered next the well tube or the distance C E and upon the porosity of the material. THE FLOW OF UNDERGROUND WATER 23 " The Draw-down." The amount of lowering C E or the draw-down, as it is called, determines for a given area of strainer opening, the amount of water which the well will yield for a given porosity of materials. Thus, with the porosity constant, if the draw-down is increased the amount discharged will be increased, and vice versa. Again, if we have two wells, each having the same size of strainer area and yielding the same amount of water, the one located in a porous bed of water-bearing materials will have less draw- down than one in dense materials. The amount of water which may reach the well depends upon the velocity of water through the water-bearing material, but the laws and constants governing this velocity are not well determined. Theory of Flow into Driven Wells. Some investigations into the theory of flow into driven wells reveal several considerations of interest and importance in a practical way, and it may be profitable, therefore, to consider the case where a well is sunk into a water-bearing stratum re- mote from interference by other wells and in which there is no general horizontal flow of the ground-water, i.e., its surface is level. Let C D E F, Fig. 4, be a pit sunk to the level of standing water A B, and let G represent a suction tube driven into the water-bearing stratum and then withdrawn to expose a strainer K M. With such an arrangement it is evident that the water may not be drawn down below K and the maxi- mum "draw-down" will be L. The total depth of the stratum from which the supply is drawn is H. Since in general the lines of saturation K N extend steeply upwards at first and then slope away gradually, no very great error is introduced by assuming that the area through which flow occurs has a depth H and that since this area is the product of H and the circumference of a circle whose FIG. 4. THE FLOW OF UNDERGROUND WATER 25 radius is r, evidently the area will increase directly as r. Hence, if Q is the discharge in any convenient unit, the Q velocity V at any radius r will be V = _ T-,-* It has been shown by various experimenters that, for a given character of material, the flow of water through such material is a function of the surface slope, and the porosity. At a radius r from the well tube, let the slope be -r- then V = K j where K = the porosity coefficient and from the Tr dh above K = d r 27rrH From this we find by integration: Where H 1 = length of strainer d = diameter of strainer H = distance of bottom of strainer below the level of standing water K = a constant depending upon the material. According to Lembke the values of this constant are as follows : *See Turneaure and Russell's "Public Water Supplies " (Wiley & Sons) for a formula based on the theoretically correct assumption that the annular space through which water flows is governed by the vertical distance from a plane M to the saturation line. Such an assump- tion is avoided in the present discussion, since it leads to an equation extremely difficult to integrate. 26 PRACTICAL IRRIGATION AND PUMPING Material. K. Sand and gravel 9,400 Coarse sand 2,800 Medium sand 760 Fine sand 150 This equation is approximately that of the curve of satura- tion and for given values of Q, H, and K the value of h H 1 may be determined for various values of r, thus enabling the saturation curve to be plotted as roughly shown in Fig. 3. By this equation but little of practical value can be deter- mined with regard to the probable yield of the well, since too many factors are involved. The yield Q is the quantity desired, but this cannot be found from the above equation except when r is known, and this is usually indeterminable. However, it is possible to work out a few useful relations based upon the time required to attain a certain draw- down with a given discharge, which will help to clear up the question of how deep to bore a well and what draw-down will be necessary to secure different discharges. Referring to Fig. 3 we see that a solid of revolution G F D C B A (where G, A, F, and D are located at points where the curve of saturation practically coincides with the water plane) represents at any given time the amount of water which has been pumped up to that time. This volume may be determined by using the equation just de- termined and integrating for volume. In Fig. 4 consider the differential volume swept through in one revolution by the elementary section of length Z and width d r. Thus we have : dU= 2 TT r Z dr But Z = L - (h - H 1 ); hence we have: dU=27rLrdr-27rrdr 27rKH log e - d Let = unity and integrating we have : THE FLOW OF UNDERGROUND WATER 27 In this equation r is taken between limits of R and where R is the radius of the circle of influence when h is 2 27rKHL practically equal to H or R = e Q Neglecting values multiplied by we have: where S = per cent, of pore spaces in the water-bearing ma- terial. The Equation for Time of Pumping. If Q = volume drawn continuously through the well tube in a unit of time, we have: T = ^r where T will be in same units of time as Q. 47TKHL-, |-T Q -, Hence : T = = -^ 2.3 Q DIAGRAM SHOWING ADVANTAGE OF DEEP BORING AND LONG STRAINER IN PROLONGING PRODUCTIVE PERIOD OF WELL AND REDUCING "DRAW-DOWN" 40 e iso I 10 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Days Pumping DIAGRAM 2 Using this equation we may compute the draw-down L corresponding to different lengths of time of pumping with 28 PRACTICAL IRRIGATION AND PUMPING fixed values of Q and soil porosity, and using different values of total depth of well. In Diagram 2 is shown a series of curves plotted with draw-down against time of pumping for a discharge of 1,000 g.p.m. with a value K = 300, a percentage of po- rosity of .30, and various draw-downs and depths of well. A series of values for the same quantities are given in Table IV (a). In this table is seen very clearly the effect of sands of different degrees of fineness upon the yield of the well and the " draw-down." The advantage of deep wells and long strainers in fine material is also shown by this table. It is to be understood that the table is based upon the assumption of a limitless body of underground water with no accessor^ supply. The deductions to be made from such a table are as follows : 1 . The deeper the well the less the draw-down for a given flow. 2. The longer the strainer the greater the time the well will give the required flow without ex- ceeding the desired or allowable draw-down. 3. In fine material, deep wells and long strainers are essential. The above equation may also be used to determine approximately the flow which may be expected under a given set of conditions. Thus let it be assumed that it is desired to know what flow can be secured from a well driven 30 feet into the water-bearing stratum and pro- vided with a lo-foot strainer. It is desired that it be pos- sible to pump continuously during six months without the draw-down exceeding the allowable 20 feet. Also the water-bearing material has a porosity of 30 per cent, and an effective size of grain of .2 millimeters. THE FLOW OF UNDERGROUND WATER H w CM i ;'.!* |- i. '!."! ;'! 1 = 0) ; to O O ;;; rH [> i n oo ! ! ! I i * w 2 .. H 9 r g S cd o i ; t : : ! i ! : : M * *a> <*H * | 7- o ::::::::: 33| S 1 ' ^ ' '. '. '. '. '. '. '. d> S UJ water g : | ':";': .':.'-,!':' I :: o m g I :-:::::::: i-T > a, H S ,5 > j ^ a ! a ! i i 1 ? - s i M W | H 3 n ^ - H Z $ 5i s ; ! ! ! ? ^ ! j i > \ 8 a ~ g g B W ^ < J S 2 w pp 1 a i ; i i ? * I i i i W , 65 ^ W P en H^ 09 C/5 & o 3-^9 53^ _3 1 1 "o : : : : 5 | : : : M y K> J ^ 55 * " 1 i O . ^ B 4 3K 4-45 14 4.00 3K 3X 4.78 H 4-25 3K 4 5-56 14 4-63 4 4X 6.00 14 4-69 4^4 4>^ 6.36 14 4-94 4/^ 4X 6-73 H 5-22 4^ 5 7.80 14 5-56 5 sA g 8.20 8.62 14 14 5-78 6.06 5% 6 10.46 14 6-63 6^ 6^ 11.58 14 7-13 6^8 7 12.34 14 7-69 7>i 7/^ 13-55 14 8.22 7% 8 I5-4I II# 8-63 8> 8^ 16.07 II# 9-31 8% 9 17.60 H> 9-75 9% 10 21.90 '# 10.81 5 inches and external diameter is 5^ inches. It is advisable, therefore, in ordering casing that the words "external diam- eter" or " internal diameter, " as the case may be, follow the particular size desired. The above tables give the sizes and weights of standard piping and casing which will show WELL SINKING 55 the distinction in sizes very clearly, as well as the compar- ative weights per foot for the same nominal sizes. Tapering of Borings. If a well has been started with standard pipe and after reaching a certain depth cannot be driven further because of friction, the boring is usually continued by dropping down inside the pipe a smaller- sized casing or pipe (usually the former), of such diameter that couplings on the inner pipe will allow it to pass freely but at the same time give a fairly close fit. Thus 8-inch standard pipe has an actual internal diameter of 7.98 inches so that 6fi inches (inside diameter) casing, the couplings >f which are 7.69 inches outside diameter, will be able to enter the pipe without difficulty. Special joints may be secured on piping such as " in- serted joints," "flush joints," etc., the object of such joints being to make a smooth exterior surface and thus lessen the friction in driving which arises with piping having the usual coupling joints. Such special joints cost extra and since they are weaker than standard pipe couplings should not be used except in special cases. They can only be obtained on the larger sizes. Where a pipe is to be sub- jected to heavy driving, a special coupling joint can be obtained which by the use of specially long threaded ends and special couplings allows the ends of adjacent pipe sections to butt together, thus preventing threads being stripped and couplings split, as sometimes occurs during heavy driving with ordinary couplings. For very deep drilling, grades of pipe known as heavy or extra heavy should be used with special joints if the expense is not prohibitive. As may be seen from the above tables, standard pipe weighs about double the artesian-well casing of the nearest equivalent nominal internal diameter, and it usually costs about one and one-half times as much as the casing. Where 56 PRACTICAL IRRIGATION AND PUMPING the well is a great distance from a point of shipment the freight charges on casing will be very much less than on the same length of standard pipe, so that casing is much the cheaper delivered at the site of the well. It must be remem- bered, however, that casing cannot be driven, without great risk of injury, to a depth greater than 100 feet by " spud- ding," that because of its lighter weight it does not sink as readily as pipe, and it is much more easily injured in assembling. Another feature of importance, particularly in steel casing where the water is alkaline, is the ease with which electrolytic action causes serious holes to be eaten in the thinner casing, although, of course, it will only be a matter of somewhat longer time when standard pipe will be similarly injured. In such water the only safety consists in specifying and being sure one gets the purest kind of wrought-iron pipe. Stovepipe Casing. A method of well drilling familiar to Californians, but not much used outside of that state, is that known as the "stovepipe" method. It is used with considerable success in alluvial material, but cannot be recommended where coarse gravel or boulders are likely to be encountered. The casing consists of riveted steel sec- tions built up in a double layer, the outer telescoping over the inner and breaking joints, thus making the riveting of adjoining sections unnecessary. A cutting section of heavier sheet steel is provided on the lower end, and the whole is sunk by ordinary spudding methods, except that the casing is forced downwards by hydraulic jacks. As a section is lowered, more of the stovepipe sections are added. Depths in excess of 1,000 feet have been attained by this method. The most interesting^feature of the method is that by means of a special cutting tool lowered inside the casing upon completion of the well, vertical slits are cut in the casing opposite those strata which the driller's log indicates are WELL SINKING 57 water-bearing, thus solving the question of a strainer, and insuring that the strainer will be exactly where wanted. It is not unlikely that this same idea might be applied success- fully to the perforation of artesian-well casing which would not be very much more difficult to cut than two layers of No. 12 sheet steel, particularly in the larger sizes of casing. It may be added that the "stovepipe method" is not used for borings less than 1 2 inches in diameter, the most common size being 14 inches. CHAPTER VI PUMPS AND PUMPING MACHINERY AND APPLIANCES IN the foregoing chapters we have discussed methods of arriving at an approximate estimate of the amount of water required, we have discussed the possibility of estimating the probable capacities of wells, when the supply must be taken from an underground source, and we have given, somewhat briefly and possibly inadequately, an idea of the method by which the well may be constructed and the water supply developed. In the present chapter and one or two following, we shall describe the pumping machinery which it is advisa- ble or necessary to use to bring the supply to the surface or to pump it from the source to the point at which it is desired the water shall be used. Caution Needed in the Selection of Pumps. At the out- set we desire to sound a note of warning and caution to those who feel themselves competent to select and install their own machinery. Nothing is more commonplace than a pump, and probably every American who has had any experience with one, no matter what the type, has an in- ward conviction that he could invent a better one, the natural result being that inventive geniuses by the score have invented and patented pumps, while occasionally some inventor more courageous or fortunate than the rest succeeds in getting his ideas into concrete form and on the market. The curious thing about all such devices is the absolute assurance of those interested in their introduction that they will surpass anything now on the market in effi- ciency, durability, and general excellence. It is no uncom- 58 PUMPS AND PUMPING MACHINERY AND APPLIANCES 59 mon thing for mechanical efficiencies of 95 per cent, to be claimed, and an enthusiastic salesman once informed the writer that he was sure 100 per cent, efficiency could be secured in the use of his pump if the water passages and pipes were nickel-plated to reduce the friction of water on iron. One needs reflect but a moment to appreciate the absurdity of such claims, for it is one of the elementary principles of physics that no machine can be 100 per cent, efficient, while any one familiar with the ordinary processes of manufacture will realize that even relatively moderate efficiencies in machinery can only be attained by refinements in design, materials, and manufacture which are only war- ranted when the saving of power incident to the use of the very efficient machine will pay interest on the difference in cost between it and one of less efficiency. Other important considerations which would justify or condemn the use of the highly efficient machine are convenience in installa- tion and use, durability, space occupied, weight, etc. Pumps which have been Proposed. Among the many pumps which at one time or another have been brought forward by hopeful geniuses for the solution of the problem of irrigation pumping are: Screw pumps, propeller pumps, bucket elevators, air lifts, air displacement pumps, and various modifications of centrifugal and turbine pumps, rotary pumps, balanced plunger pumps, etc. The last named was conceived, by an inventor who effected by means of weights on the end of a lever a balancing of the column of water on top of the pump plunger, so that ac- cording to his idea it would require no more power to pump a given quantity of water through 5oo-foot head than through 50 feet. All of these pumps will actually pump water, and many of them have some merit, but the great majority are found by mechanical test to fall far below the expectations of their inventors, while many are extremely 60 PRACTICAL IRRIGATION AND PUMPING wasteful of power and have efficiencies of from 25 to 50 per cent. Such pumps waste much of the power applied to them in shock and churning effects of the water, besides purely mechanical defects causing friction and eddy losses. It rarely pays, therefore, for the man who desires an efficient, serviceable, and satisfactory pumping plant to be led aside and induced to purchase some recently devised and comparatively untried pumping device, whose manu- facturer attempts to catch the unwary by attractive guar- antees of low cost of pumping. Before being induced to purchase such machinery the intending purchaser would do well to demand a mechanical test of the pump and a report upon the same made by a competent and reliable mechanical engineer whose position and reputation enable him to give an unbiased opinion.* What Should Decide the Make and Type of Pump to Use? A decision as to the type of pump to adopt for a partic- ular set of conditions should be governed somewhat by its reputation. The standard pumps now on the market have been slowly evolved through years of experience and experi- mentation on the part of skilled designers who apply the * It is indeed surprising, not only how quickly the news of a new device spreads, but how eager Western people are to apply every advance in the art of pumping to the problem of cheap irrigation pumping. The writer was surprised, recently, to receive a letter from a gentleman in Utah asking the opinion of the writer as to whether, in his judgment, the Humphrey Gas Pump could be used with success in the irrigation of his farm of about 100 acres. The writer was compelled to reply that, however interesting and successful the Humphrey pump might be, it seems scarcely possible in the present stage of its development to adapt it successfully to the uses of an irrigation pump on a comparatively small farm in a region remote from coal suitable for the production of producer gas. We have no doubt, however, that eventually this pump or one acting on the same principle will, when made in the right sizes, be of value in the solution of the question of cheap water supply for moderate lifts. PUMPS AND PUMPING MACHINERY AND APPLIANCES 6 1 results of experience in years of practical operation to the design and construction of their pumps. It is true, of course, particularly in the case of centrifugal pumps, that there are stock sizes which are sold, like shelf hardware, with only the most meagre attention to the particular conditions under which they are to operate. Even in such cases, however, the purchaser may be sure, if the pump is made by a reliable manufacturer, that it will be durable and serviceable, which assurance is lacking in the case of many of the so-called "freak" pumps already described. The writer must not be considered as decrying or discouraging originality in pump design, but the pumping-plant operator whose profits depend upon the reliability and durability of his equipment can better afford to adopt and use standard machinery than act as an experimental agent for some new and comparatively untried device. The expense of experi- ments and the burden of failures had much better be borne by those engaged in its manufacture than by the purchaser, whose livelihood is likely to depend upon its successful operation. If a purchaser desires the high economy and efficiency usually claimed by promoters of new styles of pumps, he can secure the same or higher economy in more refined machinery, for which, however, he will be required to pay a correspondingly higher price. The fact that a pump will actually raise water is no guarantee of its efficiency, that pump being most efficient which raises a given quantity of water through a given head with a minimum power consumption. Until a new pump can be shown by reliable and thorough mechanical tests to ex- ceed the efficiency of standard machinery, it should be let alone by the irrigation farmer, unless conclusive evidence can be produced that it is more reliable, more durable, and very much lower in first cost than a standard machine which will do the same work with the same power consumption. 62 PRACTICAL IRRIGATION AND PUMPING The Standard Types of Irrigation Pumps. There are three standard types of pump which Western practice has shown are suitable for irrigation work. These are centrif- ugal pumps or turbine pumps (which are a special form of centrifugal), single or multi-cylinder reciprocating pumps, and well cylinders. In rice irrigation certain forms of rotary pumps have been used with such success that they might properly be called standard equipment. Under certain special conditions a water-lift or bucket-and-chain pump might also be included. Each of these types of pumps is best adapted to a \ certain set of conditions, the limits of which are very well recognized, though unquestionably the limits which we shall hereinafter mention will be called in question by those manufacturers who make but one type of pump and who profess to believe this type suitable for any set of conditions without regard to economy of operation, capacity, head, or practical difficulties of operation. CHAPTER VII CENTRIFUGAL PUMPS THE, centrifugal type of pump enjoys a well-deserved popularity with those who have to solve the problems of irrigation pumping, because of its extreme simplicity, its low price, the comparative ease with which it may be installed, and its freedom from some of the annoyances which are encountered in the operation of other types. It is, however, in the small and stock sizes a machine of low efficiency, as will presently be pointed out, and for that reason, where the cost of power is an important consider- ation, it may be well to study its characteristics with some care before choosing it for any given case. The Centrifugal Pump Described. In its simplest form, the centrifugal pump comprises a casing, D, inside of which FIG. 12. The parts of a simple, cheap centrifugal pump. is rotated a runner or impeller. The impeller, I, as shown in the figure has side plates between which are cast back- wardly bending vanes V. the water entering the impeller 63 64 PRACTICAL IRRIGATION AND PUMPING through an opening at the centre on the far side of the impeller (not shown in the sketch) and being ejected at high velocity from the openings around the periphery of the impeller, thence passing through the flanged outlet O into the discharge pipe. The diameter of the opening determines the nominal size of the pump. Thus in a 6-inch pump, the opening is approximately 6 inches in diam- eter. In the 'cheek plate F is usually a babbitted bearing to support the shaft of impeller and a stuffing box S by which air is prevented from entering the pump and de- stroying the vacuum. To the left of the parts shown would be a pulley mounted on the shaft (if the machine be belt- driven) and an outboard bearing. There is also usually a thrust bearing provided at or near the outer end of the shaft to take up any unbalanced thrust due to the action of the water on the blades of the impeller as it enters the latter. The form of pump shown in the sketch is called the horizontal type; those in which the shaft is vertical belong to the vertical type, and each has its proper sphere of usefulness, depending upon location, as herein- after explained. The type of impeller shown is called the closed type, but some excellent makes of pumps have what is termed the open type as shown in Fig. 2 1 . The relative advantage of the two types of impellers is not well determined, although it is to be noted that all of those pumps in which a determined effort is made by the designers to secure a higher efficiency than is obtained in the ordinary centrifugal pump have enclosed impellers with carefully moulded and shaped water passages. For total heads or lifts in excess of 75 or 100 feet the speed at which the simple centrifugal pump must be run becomes so ex- cessively high that two or more simple pumps must be operated in series; that is, the discharge of one pump is led to the suction of the next, and thence into the CENTRIFUGAL PUMPS discharge pipe. The quan- tity of water obtained by such a combination is about the same as that from a single pump and the head through which water may be lifted for a given speed of rotation is roughly as many times that for a single pump as there are stages, thus in a two-stage pump the head attainable would be about twice that in a |p single-stage pump. Multi- stage pumps may be built as two or more separate simple pumps connected together by suitable pip- ing and run by the same shaft or there may be a number of runners in the same casing with suitably shaped passages to con- vey the water from around the periphery of one im- peller to the centre of the next. Such pumps, be- FIG. 13. Phantom view of a special type of belt-driven multi-stage vertical centrifugal pump arranged to be lowered inside a large- sized casing sunk by rotary methods, and tak- ing water from a driven well of depth within 200 feet. Is entirely self-contained and does away with open pit. A later type, made by same manufacturer, may be installed in 12-inch casing and has delivered 1,000 G.P.M. FIG. 13. 00 PRACTICAL IRRIGATION AND PUMPING cause of the complexity of the patterns from which they are made and the complex cores and difficult castings in- volved are somewhat more expensive than the combina- tions of simple pumps. In following pages are views of horizontal simple and multi-stage, as well as of the ver- tical types. Specifications for Centrifugal Pumps. For the ordinary CENTRIFUGAL PUMPS , 67 small individual installation, it is idle, perhaps, to suggest specifications which should govern the purchase of pumps, for several reasons : First, the average individual buys upon the reputation of a pump or upon its durability and serviceability in some instance which has come under his direct observation, frequently regardless of mechanical details or efficiency; second, the pump he buys is usually of such size that it comes within the range of stock sizes and it would be very difficult and expensive to secure a single machine embracing other than the usual stock details. Except, therefore, in cases where a large number of small pumps are bought at one order for some co-ordinated scheme of pumping it is useless to attempt to frame speci- fications to which pump builders may be expected to care to conform. It may, however, not be out of place to enumerate certain essential features which an efficient and durable centrifugal pump should comprise even in com- paratively small sizes and which should govern the selection and purchase of machinery for any but the cheapest central station plants. Pump Case. To be of close-grained cast-iron, free from blow-holes and shrinkage cracks. Should be suitably reinforced. with ribs and flanges in sizes over 6 inches and with discharge heads greater than 75 feet. The pump case should be divided through the centre line of the shaft in a plane affording the greatest ease of removal of half of case, to permit inspection and cleaning of interior of case and of impeller, without disturbance to the suction or discharge piping or connections. This is of special im- portance in the case of pumps taking water from a canal or river where weeds, fish, and trash of all kinds may be taken in with the water through the suction pipe and cause serious clogging and interference with operation of pump. With solid-case pumps the job of taking apart a I . ft SI CENTRIFUGAL PUMPS 69 pump to clean and inspect it, is one involving much time and labor. With split-case pumps this is comparatively simple. Suction. For all except very small sizes, the suction opening of single-stage pumps should be double, allowing water to enter impeller from both sides. This avoids end thrust as with side suction pumps, in which it must be resisted by thrust bearings and is a constant source of power loss in spite of the various hydraulic balancing arrangements in use. The suction passage should be of ample size with gradually curving flow lines and be self- contained within the case. Impeller. For large pumps this should be of bronze, preferably of the composition known as Government Bronze. This is resistant to corrosion and under the scouring action of water acquires a smooth surface which greatly reduces the energy loss in friction of water passing at high velocity through the impeller. Packing Joint and Clearance. The clearance between the pump case and impeller at the packing joint should be a running fit and at this joint preferably should be bronze rings that may be replaced when worn. In the more elaborate pumps a labyrinth packing is sometimes introduced at this point. Shaft. The shaft should be of forged, open-hearth steel and of ample strength to resist torsion, bending, and other stresses. Where it passes through the stuffing box or boxes and on the inside of the pump case it should be protected from wear by a removable bronze sleeve. Stuffing Boxes. These should be arranged for water seal and soft packing and the glands should be of bronze. In the larger pumps glands should be split to facilitate complete removal without disturbing other parts. Bearings. In the bearings only the best grade of babbit should be used. In horizontal pumps the babbit 7O PRACTICAL IRRIGATION AND PUMPING should be in split shells, which may be removed without disturbing other parts. The bearings should be ring oiling with ample oil reservoir provided with glass oil- gauge. The bearing pedestal construction should be rigid, and so formed as to prevent throwing of oil. Ample thrust collars should be provided in all horizontal double- suction pumps to take up any slight unbalancing due to clogging of one side of impeller. For side suction, single- stage pumps of large size the thrust bearing should be of marine type and provided with water jackets. In vertical pumps a ball- or roller-bearing of ample size should be provided to carry full weight of shafting, impeller, and also motor armature, in electrically driven types, in addi- tion to any unbalanced thrust due to water pressure. Flexible Couplings. These should always be pro- vided between motor and pump in direct-connected units, to eliminate wear on bearings due to slight inaccuracies in alignment. Inspection and Tests. For a large pumping-plant proj- ect careful specifications covering points above enumerated should always be drawn up by a competent mechanical engineer and the contract should permit inspection of the machinery during construction, to see that mechanical de- tails are built according to specifications. Finally, it should be stipulated that the pump or pumps should be tested at the factory previous to delivery under the conditions of speed and head at which they are to operate and, if pos- sible in the case of electric drive, should be tested with the same motors by which they are to be operated. The results of the test should be plotted on cross-section paper and curves drawn showing the characteristics of the pump. These curves should be considered as guarantee of per- formance and should be checked by test under running conditions subsequent to completion of plant. CENTRIFUGAL PUMPS 71 Characteristics of Centrifugal Pumps. Those who now operate, as well as those who expect to operate centrifu- gal pumps should have some familiarity with the mechani- cal characteristics of such machinery, for it is undoubtedly true that much of the dissatisfaction we find among those who use these pumps has been due to an attempt to operate them under conditions for which they were not designed or well adapted. As will be noted from the brief description already given, the centrifugal pump differs from a reciprocat- ing pump, with which most of us are familiar, in having no piston or plunger and no valves. Its action evidently, there- fore, depends upon the whirling motion imparted to the water by the rapidly-rotating impeller. Simple as this action may seem, the fact remains that although many voluminous works have been written upon the subject and many ^ and some- times conflicting^ theories advanced, no formula or method 4ias yet been devised by which, with the speed of the impel- ler and the size and proportions of the impeller and casing given, it is possible to figure the capacity of a pump or its efficiency. What is known delTnitely of the action of centrif- ugal pumps has been determined almost entirely by experi- ment and in designing a new pump to give certain desired characteristics, the designer is helped but very little by theory, and must project largely his knowledge of the action of existing pumps to the new design. By this process of evolution, centrifugal pumps are continually being im- proved upon and higher efficiencies attained, but even in those of most advanced design there occur obscure losses in energy due to friction, shock, or impact, secondary whirling effects or eddies and leakage of water through clearance spaces which materially cut down the mechani- cal efficiency of the machine and either entirely upset the theoretical notions which attempt to explain its action, or by reason of their obscurity make the constants impossible 72 PRACTICAL IRRIGATION AND PUMPING of determination in any formula in which these losses are recognized and allowed for. What the Plant Designer or Operator Must Know. So far as the operator of a centrifugal pumping plant is con- cerned, and particularly the man designing such a plant or intending to use centrifugal pumps, he is interested not so much in questions of the design of the pumps themselves as in what the pumps that he can buy will do under a given set of conditions. Unfortunately he is not greatly assisted in securing such knowledge by the information which may be gleaned from a manufacturer's catalogue containing descriptions and ratings of their stock pumps. What a purchaser desires to know is what speed is required to force water through a given head, what discharge may be ex- pected at that head, and what actual horse-power is re- quired under those conditions to operate the pump. The manufacturers' rating as given for a particular size of pump is usually what is known as "the economic capacity," which is a term of doubtful meaning and of very little use to a pump purchaser, for nothing is usually said as to whether this capacity is the maximum attainable, which may be at high speed and low head or is that capacity which is secured under conditions of maximum efficiency. It is the peculiar characteristic of centrifugal pumps that the capacity is a variable depending upon the speed at which it is run, and the total head against which it operates. Conversely, every centrifugal pump when run at a certain speed will give a certain discharge at a certain head. If the speed be increased with the head constant, the dis- charge will be increased according to a definite law, and if the speed be maintained constant and the head be decreased, the discharge will generally increase. Furthermore, every centrifugal pump has a definite head for different speeds at which it operates most economically from the standpoint CENTRIFUGAL PUMPS 73 of power consumption and in order to force the water through this head this particular speed should be used, if, as is frequently the case, the cost of power is the most important factor in the cost of operation. to T3 S tlh The Efficiency of the Pump. The efficiency of a centrif- ugal pump, as of every other pump or any mechanical con- trivance transforming mechanical work into another form 74 PRACTICAL IRRIGATION AND PUMPING of energy, is the ratio of the effect to the cause, or in the case of a pump we may define it as: Energy in moving water Efficiency = Energy supplied pump. The energy in the moving water is the product of the weight of water elevated in a given time multiplied by the A VIEW ROM SUCTION END SHOWING SECTION OF CASING IMPELLER EXPOSED SHOWING SECTION THROUGH LINE A-B FIG. 17. Cross-sections of 2K-inch Pump, Serial No. 1. (See Diagram 5.) total head in feet through which the water is raised. Thus a pump elevating continuously 450 gallons of water per minute through a total head of 8.8 feet gives energy to the water equivalent to one horse-power or 33,000 feet-pounds per minute. If the motor or engine by which the pump is run delivers two horse-power to such a pump the efficiency CENTRIFUGAL PUMPS 75 will be 50 per cent, and one-half of the power delivered by the engine will be lost in useless churning and fluid friction effects in the pump and by mechanical friction in its bearings and stuffing box. Just as the capacity of a SHOWING UCTION END OF CASING REMOVED WITW THE IMPELLER EXPOSED SECTION AT A-B FIG. 18. Cross-sections of 2-inch Pump, Serial No. 2. (See Diagram 6.) centrifugal pump is a variable, so also is the efficiency, and for different heads and capacities we have different efficien- cies, all varying according to certain laws. For any given speed there will be a certain head at which the efficiency will be a maximum, and a knowledge of these facts is 7 6 PRACTICAL IRRIGATION AND PUMPING necessary to intelligent selection of a pump for a given purpose and to insure efficient operation after it is installed. Pump Curves. In the following pages are given a number of diagrams which show the characteristics of sev- A- VIEW FROM SUCTION END SHOWING SECTION OF IMPELLER AND CASING SECTION THROUGH A-B FIG. 19. Cross-sections of 4-inch Pump, Serial No. 3. (See Diagram 7.) eral pumps in a series of tests made under the direction of the writer in an investigation to determine various facts relative to the performance of stock sizes of centrifugal pumps such as are commonly used in irrigation work. CENTRIFUGAL PUMPS 77 itf-pBOH TOOJ 8 8 g. 8 3 3 % 7 8 PRACTICAL IRRIGATION AND PUMPING CENTRIFUGAL PUMPS 79 8o PRACTICAL IRRIGATION AND PUMPING CENTRIFUGAL PUMPS 81 82 PRACTICAL IRRIGATION AND PUMPING CENTRIFUGAL PUMPS 8 4 PRACTICAL IRRIGATION AND PUMPING The preceding diagrams give the following information : (i) The discharge in gallons per minute for various speeds when total heads are given. (2) The mechanical efficiencies at different speeds. (3) The horse-power required to be delivered to the pump in order that different quantities per minute may be discharged through different total CENTRIFUGAL PUMPS 85 heads. Some of the diagrams also give the speeds which should be used to give maximum efficiencies at SHOWING SE'CTICN OF CASING WITH IMPELLER EXPOSED FIG. 21. Cross-sections of 6-inch Pump, Serial No. 8. (See Diagram 9.) different total heads and the corresponding horse-power input.* * The names of the makers of the pumps corresponding to the dia- grams are withheld for obvious reasons. 86 PRACTICAL IRRIGATION AND PUMPING The Selection of a Pump. To make use of the diagrams, let us assume that it is desired to pump a supply of 300 gallons per minute from a well in which water stands 30 feet below the surface, and in which the draw-down will not SECTION THROUGH- PLANE OF ! IMPELLER, VIEWING FROM SUCTION END SECTION AT A-B rCQ FIG. 22. Cross-sections of 4-inch Pump, Serial No. 10. (See Diagram 10.) exceed 15 feet at this discharge, making the total hydro- static head at this discharge about 45 feet. In the follow- ing figure, Fig. 23, is illustrated the meaning of the terms relating to "head" or "lift." CENTRIFUGAL PUMPS As will be apparent from the figure, the total distance through which the water must be elevated, or the hydro- static head, is the distance of standing water below the level of water at the outlet, plus the "draw-down"; and the Pressure Obs./8q.in.) x 2.304=" Discharge Head-Ft^A id-A+Bjlfe = H t-fri ction^head h ^difference in velocity Bead between discharge anVsuction Suction (inches of Mercury) xK133= action HeadjFt.=B FIG. 23. total head, which has the same meaning as the term used in the pump curve diagrams, is the hydrostatic head, plus the friction head, plus or minus a small correction called the difference in velocity head. The hydrostatic head may be determined from a knowledge of the depth of water below 88 PRACTICAL IRRIGATION AND PUMPING the surface and the assumed "draw-down," but the friction head must be calculated in those cases where it cannot be allowed for in the reading of a pressure gauge. Thus as shown in Fig. 23, if a pressure gauge and a vacuum gauge be attached to a pump at the points as shown, the readings of these gauges, multiplied respectively by the proper con- stants as indicated in figure, give the total head in feet, when to the sum is added the vertical distance between the centre of the discharge gauge and the point of attachment of the vacuum gauge to the suction pipe (disregarding the cor- rection for change in velocity head). Since, however, for a pump not yet installed, the gauge readings are not avail- able, the total head must be calculated by adding to the hydrostatic head a correction for the friction head. It is important that this factor be allowed for, since it may in unusual cases amount to as much as the hydrostatic head. The friction head may be denned, for the benefit of those not familiar with the science of hydraulics, as that head which would be necessary in a perfectly level pipe-line to cause a flow through the pipe of the desired quantity of water. Thus, if it were desirable to cause a flow of 100 gallons per minute through a level pipe 2 inches in diameter and i oo feet long, this water would have to enter the pipe at one end through a large vertical riser at least 22 feet high in order that such discharge might occur. In other words, the friction head in a zoo-foot length of pipe with a flow of 100 gallons per minute is about 22 feet. It is seen from this one example that friction head may be an extremely important item in the factors effecting flow of water in pipes, and in the design of pumping plants it must be reduced to a minimum, as will later be explained, by the use of pipes and fittings of proper size, in order that the cost of pumping may be reduced. CENTRIFUGAL PUMPS 8 9 The hydrostatic head in the instance cited on page 86 was 45 feet. If the discharge be assumed to occur at the ground surface, the vertical length of pipe involved may be assumed to be the same as the hydrostatic head, and upon this assumption for the general case, Diagram 12 is figured, showing for various capacities and depths of wells 5 100 200 300 400 500 600 700 800 900 1000 Gallons per Minute DIAGRAM 12 SIZES OF PIPE SUITABLE FOR DIFFERENT HEADS AND DISCHARGES BETWEEN 100 AND 1,000 GALLONS PER MINUTE up to a total hydrostatic head of 100 feet the commercial size of pipe which should be used in order that the friction head may be kept below a certain maximum. In using the diagram for the case being considered, find 300 gallons per minute on horizontal scale and trace vertically upwards to the curve marked 40 feet hydrostatic head. The lift in the present case is 45 feet, but since we can use only the nearest commercial size, which as seen by the 9 o PRACTICAL IRRIGATION AND PUMPING scale at the left is 4-inch pipe, it is unnecessary to inter- polate. Hence, provisionally a 4-inch pipe for the discharge will be adopted. Referring now to Diagram 13 we find a means of determining the head to be allowed for friction. Following vertically upwards from 300 gallons per min- ute, on the horizontal scale at the bottom, to the curve 100 200 300 700 800 900 1000 400 500 601 Gallons per Minute - DIAGRAM 13 FRICTION HEADS FOR DIFFERENT SIZES OF PIPE 100 FEET LONG DIS- CHARGING VARIOUS QUANTITIES OF WATER marked 4-inch pipe, we find the corresponding friction head for loo-foot length of pipe is about 6 feet. For an allowed length of 50 feet in this case, the friction head will be 3 feet, and consequently the total head will be 48 feet. An examination of the characteristics of various pumps as given in the diagrams (pages 77-83) will enable us to construct the following table: CENTRIFUGAL PUMPS 9 1 TABLE VIII SPEEDS AND EFFICIENCIES OF VARIOUS PUMPS FOR 300 G. P. M. AND 48 FT. TOTAL HEAD, AS DERIVED FROM DIAGRAMS OF PUMP CHARACTERISTICS, PAGES 77 TO 83 Pump No. Speed R. P. M. Eff'y % Size Discharge Pipe I 960 48 2^ 3 I,I2O 45 4 7 860 27 6 8 650 35 6 10 900 53 4 ii 560 45 4 This table gives two criteria which may be used as a basis of selection: first, the speed; second, the efficiency. In general, it may be said that a slow-speed pump is prefer- able, unless it is to be driven by an electric motor (in which a higher speed is desirable, particularly in direct- connected sets), because of the longer life and greater durability of bearings of the slow-speed machine. From this standpoint it would appear that pump No. n is preferable, since it combines a very moderate speed with reasonably good efficiency. From the standpoint of effi- ciency, however, it is evident that pump No. 10 is superior to the others, particularly since the speed, 900 R.P.M., is by no means excessive. In case fuel cost is high, No. 10 should be chosen; if fuel cost or power cost is relatively unimportant, choose No. n, assuming that the pumps themselves sell at about the same price. Under some cir- cumstances it may be desirable to operate the pump for short periods at a greater or less capacity than that above given. It must be understood that this is a poor policy in general, for it will usually mean a very considerable lessening of efficiency unless the efficiency curve is quite flat over a considerable range of speed. The relation between speed, PRACTICAL IRRIGATION AND PUMPING horse-power input, and efficiency for any given head may readily be deduced from the characteristic curves of the pump and for the purpose of illustration this has been done for pump No. 10 when working at 48 feet head. The characteristic curves for this case are shown in Diagram 14. Although the total head will vary as the discharge changes, due to change in frictional head, it is difficult to show this 1300 1200 60 g'llOO 50 a llOOO 40 20 r/j fi a -G | 900 30 1 15 I 800 20 10 700 10 5 4 IN. CENT. PUMP SERIAL NO. 10 CHARACTERISTIC CURVES FOR 48 FT. TOTAL HEAD X 100 200 300 400 500 Gallons per Minute DIAGRAM 14 600 on a diagram and consequently Diagram 14 merely shows how a change in speed affects efficiency, horse-power input, and discharge when the head is constant. The conditions represented by this diagram show very closely what actu- ally takes place when a pumping-plant operator for any reason changes the speed of his engine or motor or alters the ratio of driving to driven pulleys. As may be seen from the diagram, this pump will deliver 300 G.P.M. through 48 feet total head when operated at a speed of 905 R.P.M. CENTRIFUGAL PUMPS 93 It will require 6.7 horse-power delivered at the pump pulley to accomplish this and the pump will have an efficiency of about 52.5 per cent. This is not the highest efficiency which may be attained by this pump, but it is that at which it will operate when fulfilling nearest the required conditions. Now as may be seen from Diagram 14, the discharge may be increased to double the amount, or to 600 gallons per minute, by increasing the speed to about 1,220 R.P.M., but it will be noted that the efficiency drops rapidly so that at this speed and discharge the efficiency is only slightly over 40 per cent., the horse-power input not being proportional to the discharge, but increasing more rapidly than the discharge increases. Consequently, about 20 per cent, more fuel will be used, or power will be con- sumed, per unit of water pumped at the higher speed than at the lower speed, and it would evidently be unwise, from the standpoint of economy, to operate the pump for any length of time under these conditions. This will be espe- cially true in case an electric motor is used, since if it is rated at a horse-power corresponding to the 300 gallons per minute discharge, it will be seriously overloaded and will heat badly at the higher discharge, if indeed it can be made to develop the greater power required, belt drive being assumed and the speed change being secured by change in pulley ratio. Size of Engine or Motor The foregoing naturally brings up the question of the size of engine or motor to use. For convenience, the case above given will be considered further. As may be seen in Dia- gram 14, the actual horse-power required at the pulley of the pump for a discharge of 300 gallons per minute against 48 feet total head is 6.7 horse-power. Allowing for belt slippage and other losses, an engine or motor would be required 94 PRACTICAL IRRIGATION AND PUMPING rated to deliver from 7 to 8 horse-power. Although in general a gasoline engine should give its most economical fuel consumption at or near its rated power, the tendency of most engine manufacturers seems to be to over-rate gasoline engines, and it would probably be better, therefore, to provide an engine rated at 9 to 10 horse-power. This will leave a slight overload capacity available, and the engine would probably be found to govern better and work more reliably than would the engine which is loaded up to its rating. With an engine capable of delivering 10 horse-power it would be possible, as may be seen from the curves in Diagram 14, to secure 400 G.P.M., approxi- mately, in case the necessity arose and provision was made for changing speed. A properly designed engine or motor will, of course, have a definite and fairly constant speed, and although it is possible to change this speed over a slight range, the engine-driving pulley and the pump pulley should be so selected that the proper pump speed will result as based upon the mean speed of the engine or motor. Generally the size of the pump pulley is fixed, and cannot be altered because of the design, and consequently the size of the engine pulley must be specified. Ordinarily it will be found expedient to use a clutch pulley on a gasoline engine, this being supplied at a reasonable price as an extra by the engine builders. Let it be assumed that an engine of the size contemplated runs at a speed of 300 R.P.M. The pump in question is to have a speed of about 900 R.P.M. and it has an 8-inch pulley. It is evident, therefore, that the engine clutch pulley should be = 24" diam- eter, and this size should be specified when the engine is purchased. Too much emphasis cannot be placed on the proper selection of engine as to type and size and to secur- ing the proper speed for the pump, since the economical CENTRIFUGAL PUMPS 95 operation of the plant depends upon securing the highest efficiency possible from the pump and working the engine under such conditions as will promote the use of a minimum quantity of gasoline per unit of power developed. Pump Builders' Diagrams The above desirable conditions of operation will only be secured when those operating pumping plants or con- 5"CENTRIFUGAU PUMP CLASS A CAPACITY-HEAD AND ISO-EFFICIENCY D.IP. CURVES CONSTANT SPEED 10 I I i ' I I I. 1 -l *-i T-> rl -pi Capacity in Gallons per Min. DIAGRAM 15 templating their erection give their attention to the characteristics of pumps similar to those described and illustrated in the preceding pages and pump builders are willing to supply authentic curves in connection with 96 PRACTICAL IRRIGATION AND PUMPING their catalogues similar to those in Diagrams 15-16-17. These were furnished the writer through the courtesy of one of the largest builders of pumping machinery in the country, and are particularly valuable in the light they throw upon the characteristics of the various sizes of pumps considered. The use of these diagrams is essentially the same as CapaeUjjy JQ Gallons perJlin. DIAGRAM 16 those already described, although by them it is somewhat easier to interpolate between curves. To illustrate the use of the diagrams, let it be supposed that it is desired to obtain 800 gallons per minute, pumped through a total head of 70 feet. The questions to be solved are (i), the best size of pump to use; (2), the efficiency at which it will operate; (3), the horse-power necessary to be pro- CENTRIFUGAL PUMPS 97 vided; (4), the speed at which it must operate. Let it be assumed that the pump is to be operated 90 days of 24 hours each during the year and that power is supplied by an electric motor, current for which costs 2.8 cents per K.W. hour.* The costs of the three pumps whose characteristics are 130 4"CENTRIFUGAL PUMP CLASS-A CAPACITY-HEAD AND ISO-EFFICIENCY-B. H.P. CURVES CONSTANT SPEED 100200300400500600700800900 Capacity in Gal. per Min. DIAGRAM 17 shown in the diagrams and as arranged for connection to an electric motor are: 4-inch pump $215.00 net F.O.B. cars factory. 5-inch pump $255.00 net F.O.B. cars factory. 6-inch pump $300.00 net F.O.B. cars factory. * The cost of electric power is frequently fixed by a sliding scale so that there is a certain minimum charge whether power is or is not used, and the greater the amount of power used the less is the charge per K.W. hour. (See page 144.) 9 8 PRACTICAL IRRIGATION AND PUMPING Referring to the diagrams, the following information is given as shown in the following table: Table showing characteristics of three pumps delivering 800 gallons per minute through 7o-foot total head. Pump Speed R. P. M. Eff'y % H.P. Required 4" 1,075 41 35 5" 785 60+ 23 6" 690 62 24 Although a fairly slow speed is desirable, none of the speeds above given are to be considered excessive, but, other conditions being the same, the 6-inch pump would be selected from this standpoint alone. In the estimate of depreciation, however, it must be assumed that the depre- ciation of each pump increases in almost direct proportion to its speed. Thus letting depreciation on the 6-inch pump be 8 per cent., on the 5 -inch pump it will be 9.1 per cent., on the 4-inch pump 12.5 per cent. Assuming that money commands 8 per cent, interest and that the yearly length of service as before stated is 90 days of 24 hours each, we may construct the following table giving yearly cost. Pump Motor Yearly Total Pump Interest and Interest and Cost of Yearly Depreciation Depreciation Power Cost* 4" $44-10 $l6o.OO $2,820 $3,024 5" 43.60 I2O.OO 1,850 2,013 6" 48.00 120.00 1,935 2,103 * Not including attendance and cost of lubrication, or interest and depreciation on other parts of plant which would in every case be about equal. CENTRIFUGAL PUMPS 99 The final column in above table should be used as a basis of judgment in a decision as to size of pump, since it includes interest and depreciation on the plant and the cost of power. It is also evident that, since the 4-inch pump requires a larger motor because of the greater horse-power required, it will be more expensive in first cost. It is evident that the 5-inch pump is the one which should be selected for the assumed conditions. It will be noted that yearly cost of power in above example is really the decisive factor, but, as may readily be seen, this factor will decrease in importance as the number of days of the year during which the plant is operated is decreased. The foregoing method of selection of a pump is suggested as being advisable when curves similar to those of the diagrams are available and when some approximate estimate of length of time of pumping and cost of power may be obtained. Pump Equations An examination of the characteristic curves and an in- vestigation of the efficiencies of the various pumps as given in the preceding pages, must lead one of an inquiring mind to wonder, first, why an average of 50 per cent, of the work or power applied to pumps of the types tested should be lost in transforming that work into energy of flowing water, and second, since the curves showing relation between head and discharge for constant speed seem to follow the same law for all pumps, why it would not be possible to express this law by a mathematical expression containing factors which properly would take into account the peculiarities and proportions of the pump. To the solution of the first question the present investigation, unfortu- nately, can offer no clew beyond confirming the fact that power losses are unquestionably (and apparently unnec- essarily) large in pumps of the size of those tested. It IOO PRACTICAL IRRIGATION AND PUMPING would seem to offer a very fruitful field of investigation and experimentation, particularly on the part of manufacturers, to devise some simple shape or arrangement of impeller and casing to do away with the losses from shock and friction which now accompany the change from velocity head to pressure head in small centrifugal pumps. And experiments seem to show that, in sizes of over 6 inches and in a few cases of that size, the designs of some manufacturers have been so worked out that efficiencies of 75 per cent, and over have been attained. There seems to be, in sizes below 6 inches, the opportunity for some manufacturer to produce a simple centrifugal pump so designed that it will equal the efficiencies attained by larger sizes and yet will not be un- reasonably high in price. To return to the subject of loss of power in the centrifugal pump, it may be said that it undoubtedly occurs through a combination of friction, shock, and eddy effects in all pumps, and to the eddy effects, etc., must be added the leakage through clearance spaces in the more poorly de- signed and built pumps. The nature of these losses, how they vary, their relative or absolute amount, and the best means of preventing them, all remain yet to be discovered, and the subject offers a most engaging field to investigators provided with the requisite equipment and resources. The second point referred to, namely, the question of rinding some expression governing the relation between discharge and head at constant speed, can be answered to some extent by a consideration of the theoretic relations in the light of the curves given and the proportions of the various pumps. The chief factor in connection with the theory of cen- trifugal pumps is the head or vertical distance through which the water may be pumped. This is shown by theory to be directly proportional to the square of the CENTRIFUGAL T speed of the impeller or rotating part of the pump. The head actually realized is less than the theoretical by the effect of water friction losses in the passageways surround- ing the impeller and leading to the pump outlet, and in shock or impact effects at entrance to and upon leaving the impeller. We may therefore write: Actual head = h = H - hi - h 2 - h 3 - h 4 = theoretic head head lost in friction, etc. Where h = actual head realized in feet. H = theoretic head in feet, hi = head lost in impeller. h 2 = head lost in discharge chamber. h 3 = head lost by impact at entrance to impeller. h 4 = head lost by impact at exit from impeller. Now, as is well known, water friction losses are propor- tional to the square of the velocity. Hence: Let S = velocity of water passing through impeller. T = f velocity of water in discharge chamber. Vi = velocity of outer periphery of impeller. V 2 = velocity of inner periphery of impeller. R = radial velocity of water at entrance to impeller. Then we may write: H = JVr> hi = M S 2 h 2 = N T 2 Where J, M, and N are constants of proportion. Impact at exit from impeller may be considered to be proportional to some quantity P Vi (S + Vi), since it is zero when Vi is zero, it increases as S increases, and becomes merely a frictional effect proportional to Vi 2 when S = o. In this equation, P again is a constant of proportion. The IO2 V PiJ-keXlCAL J&EIGATION AND PUMPING effect of impact at entrance to impeller may be written in a similar way, since at impending delivery there is no im- pact and the loss at entrance to impeller is then a frictional effect which during discharge must be proportional to the combined effect of radial and peripheral velocities. Hence for the loss of head at entrance to the impeller we may write: h 3 = CV 2 (R + V 2 ). But since R is proportional to S and since V 2 is propor- tional to Vi we may write h 3 = B Vi (S + Vi) or the loss at entrance from friction and impact is proportional to the loss at exit. Combining various terms we may write: h =JVi 2 -MS 2 -NT 2 -BVi (S+Vi)-PVi (S+Vi) Now T may be taken as proportional to Vi, since the water in the discharge chamber evidently rotates at some less velocity than Vi and we may therefore write T 2 = L Vi 2 where L is a constant. Expanding and substituting, we may write: h=JV 1 2 -MS 2 -NLV 1 2 -BViS-BVi 2 -PViS-PVi 2 or h= (J-NL-B-P) Vi 2 -MS 2 -(B+P) ViS. Let (J-NL-B -P) = K! Let M = K 2 Let B + P = K 3 and we have h = Ki Vi 2 - K 2 S 2 - K 3 Vi S as representing the equation for the actual head realized by a centrifugal pump, taking into account losses by friction and impact. In this equation S may be represented by the term where Q = the discharge in cubic feet per second a CENTRIFUGAL PUMPS 103 and a = the total area of water passages through the im- peller normal to the flow line at exit. Hence: In this equation the constants Ki, K 2 , and K 3 will evidently be applicable to only one type and size of pump, but it is of interest to determine what is the absolute value of such constants and how they compare for different pumps. To determine the constants, use was made of the method of least squares, which, although very laborious, is the only reliable method of so combining a series of obser- vations that the resulting equation will be the best possible average of all the observations. The method was applied to the head-discharge curves of pumps Nos. 1,3, and 10, and from these observations the following equations were derived: D 2 N 2 Q 2 DNQ Pump No. i, h = .oo 3 66--9.i5 2 -.oo 9 i^-cosa D 2 N 2 Q 2 DNQ Pump No.3,h = .oo 33 55 --9.31 -.00947-^- cosa D 2 N 2 Q 2 DNQ Pump No. io,h = . 00336 --- 2.740 - .0208 -- cosa 2g y 2ga 2 2ga It will be noted that the equations have been slightly changed from the general form above given and that a factor, "cos a" has been inserted in the last member to take account of the angle of the vanes. In these equations: D = Diameter of impeller over vane tips in feet. N = Speed of impeller in R.P.M. 2g = Twice the acceleration due to gravity = 64.4. The agreement of these equations, or the curves which 104 PRACTICAL IRRIGATION AND PUMPING they represent, with the actual curves found from the tests of the pumps is shown in Diagram 18. As will be seen, the general form of the curve given by the equation seems sufficiently close to the actual curves to warrant the belief that the assumptions upon which the original equation is based are essentially correct. It would, therefore, be possible by such an equation to predict the Pump* 3 Curves from Actual Tests Formulae 110 .100 100 200 300 400 500 Gal. per Min. 100 200 300 400 500 600 Gal. per Min. DIAGRAM 18 100 200 300 400 500 600 Gal. per Min. performance of the particular pump at any speed, head, or discharge if two out of these three conditions are known. As will be seen, however, the correspondence between the constants of the three equations is not sufficiently close to allow the equation determined for one pump to be used as a basis for predicting the performance of another. It would seem, therefore, that there is still lacking some factor de- pendent upon the size or proportions of the pump which must be used in the original equation to make it really general, that is, applicable to any series of pumps. It is thought, however, that the fact that an equation may be made to apply, within a reasonable degree of accuracy, to CENTRIFUGAL PUMPS 105 the performance of even one pump throughout the range of working conditions, is sufficiently interesting to be worthy of notice. Locations and Conditions Suitable for Centrifugal Pumps The head against which water must be pumped and the capacity desired are the two factors which largely decide the question of whether a centrifugal pump is or is not suitable in irrigation work. So far as their construction is concerned it is now possible to negotiate almost any lift under 1,000 feet and the maximum quantity which can be lifted under such limitation of head is only limited by the amount of power available and the difficulties in making large-sized castings. Centrifugal pumps are in successful operation where 3,000 horse-power is absorbed by a single unit, but such an example has no bearing upon those prob- lems connected with irrigation from wells where 1,000 gallons per minute is the maximum quantity which may be developed successfully from a single well and where heads of over 100 feet belong only to those situations where fruit growing make such lifts profitable. Where the supply is pumped from an open water source, as a river or reser- voir, and is to be delivered into canals or distributaries, 5,000-6,000 gallons per minute have been handled with success in a single unit, but at comparatively low heads. In general it may be said to be feasible to use centrif- ugal pumps under the following conditions: In single units (1) In pumping from open water source for supply of under 10,000 gallons per minute. (2) In pumping from wells where the pump may be placed not over 5 feet above standing water level, where total depth to water does not exceed 100 feet, total head II Ii 3 a M iJ -s 12 -Si CENTRIFUGAL PUMPS 1 07 does not exceed 125 feet, and quantity pumped does not exceed 1,000 gallons per minute. The limitation in depth in (2) arises from the difficulty and expense of sinking a pit of the size necessary for a pump installation to a depth greater than 100 feet. The limita- tion as to head in the same instance arises from the prac- tical difficulties in operation inherent in centrifugal pumps when heads as much or greater than 125 feet are encoun- tered. The writer bases this statement upon his belief, based upon experience, that the multi-stage pump used in such cases is not free yet from serious objections due to end thrust and internal leakage, the first due to high pres- sure and weight of shaft in case of vertical pumps and the second due to the fact that well-water is likely to be heavily charged with fine sand which under high pressure forces its way into intermediate bearings and stuffing boxes and quickly causes serious wear and leakage. Doubtless means will be evolved in time which will eliminate these difficulties, but the writer has yet to learn of a centrifugal pump for these conditions in which the design has been so far per- fected that when operated by the average man under practical conditions, the difficulties above mentioned will not arise. To obviate the necessity of a pump pit and yet use the centrifugal principle in very deep wells, a form of centrifugal multi-stage pump of such size that it may be lowered inside of a large size well-casing has been put on the market in recent years and under favorable conditions of operation has given much satisfaction. It is, however, an expensive type of pump and of limited capacity. Its use is more especially recommended for water- works' use, or for locations where water has acquired a high value for purposes of irrigation, so that the initial expense is justified. See Fig. 13, page 65. CHAPTER VIII DIFFERENT TYPES OF INSTALLATION FOR CENTRIFUGAL PUMPS THE head to be pumped against, the depth to water, and the character of power to be used are determining fac- tors in the decision as to the character of pump and general arrangement of plant, and every plant design must to a certain extent be based upon a study of local conditions. It is possible, however, to give a few examples illustrating certain types of installation which have been found satis- factory in practice and the designer may then vary the details to suit local or special requirements. PLANT No. i The plant shown in Fig. 24 is that which it is customary to adopt where a limited quantity of water is to be lifted to adjacent high lands from an open water source. (In the case shown in Fig. 24 this is supposed to be a canal.) In such a case the pump is set up on a firm foundation at such elevation that the suction is not, say, over 10 feet vertically above the lowest water-surface elevation of the source. The design shown in Fig. 24 may be modified and materially improved by providing a concrete-lined sump or pit beneath the pump house with a passageway leading to the canal. In this passageway grooves should be left in the concrete for the insertion of screens or trash racks and possibly stop planks or a gate, to keep water out of the sump in an emergency or for cleaning. This scheme 108 INSTALLATION FOR CENTRIFUGAL PUMPS 1 09 also allows the pump and motor to be placed further away from the canal on a more secure foundation and permits the use of a short vertical suction pipe which is a very desirable consideration. In some cases where a long discharge pipe is necessary to reach the point of gravity distribution, it may, according to the nature of the ground, be cheaper to dig a supply ditch or canal some distance inland from the source and place the plant closer to the point of discharge, eliminating more or less expensive discharge pipe, and possibly enabling a smaller size to be used. Friction Effects. It must be remembered that the friction effect tends to increase the suction lift just as it does the lift on the discharge side and the distance from the inlet of the suction pipe to the inlet of the pump must not be great enough to cause a friction head which together with the vertical suction lift will much exceed 25 feet, which must be considered the economical limit of suction lift. Many turns and valves also increase the friction effect. Example. An example will probably make this more clear. Suppose a 4-inch pump is to be used. This size refers, as previously explained, to the diameter of the discharge opening of the pump. The suction opening is usually at least one size larger or say 5 inches. The suc- tion pipe should in no case be less than the size of the suction opening of the pump and if the suction pipe is to be of considerable length it should be at least two sizes larger than the nominal size of the pump. Let us say that in this case 5-inch pipe should be used on the suction line. Assume that the pump is placed 10 feet above water level and that the distance from the strainer to the pump is 100 feet measured along the axis of the pipe. Also assume that there is a foot-valve on the strainer and two elbows in the line. If the pump discharges 500 gallons per minute, 110 PRACTICAL IRRIGATION AND PUMPING the friction head for a 5-inch pipe of the length assumed will be about 5.5 feet. (See Diagram 13.) The loss of head in the foot- valve will be about 2.8 feet and in the two elbows about 2 feet. The velocity head will be i foot. The total suction head will therefore be found as follows : Hydrostatic head 10.00 ft. Loss of head foot -valve and strainer 2.80 Friction head in 100 ft. 5-inch pipe 5.50 Friction head, two elbows 2.00 Velocity head i.oo Total suction head 21.30 ft. It is apparent from this example that the effect of friction in a long suction pipe may be to increase the suction head by more than double the actual vertical lift and for this reason, if for no other, it is advisable to have a suction pipe as short and direct as possible. Of course the use of a large pipe will much reduce the friction head, but it is advisable, if possible, to limit the length of the suction pipe to a mini- mum from the standpoint of operation, since where we have a long and crooked pipe we also have many joints which it is always difficult to keep perfectly tight, as they must be with any centrifugal pump, a slight air leakage into the suction cutting down the flow tremendously, and when serious enough, stopping the flow entirely. Drive. The drive, in the case illustrated by the figure, is by gasoline engine, though steam or electric power could be used as well. It is advisable to use a generous length of belt for centrifugal-pump drive, the distance between the engine and pump centers being from 15 to 20 feet. Priming. For priming the centrifugal a common pitcher pump may be used. This should be connected by %-inch pipe to the pipe connection to be found on the top of the centrifugal pump casing. An ordinary globe valve, or, INSTALLATION FOR CENTRIFUGAL PUMPS III what is better, a good check opening outwards should be placed in this line and after the priming has been accom- plished the operator should be sure that no air leaks into the centrifugal through this line. If the pump is to be primed each time it is started, a check valve must be placed in the discharge pipe immediately above the pump, and in any case where the pipe discharges into a tank there must certainly be provided a check valve to prevent the emptying of the pipe or tank when the pump is stopped. If no check valve is used the priming pump is useless and a foot-valve at the strainer is necessary. In this case, to prime the pump when it is started the first time, the plug on top of pump casing must be removed and water poured in through the opening till the suction pipe and pump are full. This presupposes, of course, a tight foot-valve, and if it remains so there will be no subsequent necessity for priming. Foot-valves are, however, subject to derangements, and af- ter a stop of a week or so it is no uncommon experience to find that all the water has leaked back through the foot- valve and the tedious process of priming must be repeated. Ejector Primer. Where a check valve is used, the discharge pipe is large or long and the head over 50 feet, thus providing a considerable supply of water stored at the necessary pressure, a very convenient method of priming is by the use of a water ejector attached to the centrifugal similar to a priming pump and using for its operation water taken through a small pipe connection from the discharge pipe. As long as water remains in this pipe other means of priming will not have to be resorted to, but must be provided and held in reserve. For power- driven priming pumps the use of a small water pump with a discharge pipe carried sufficiently high to insure that the valves will always be immersed, is preferable to an air-pump which must be stopped as soon as the centrifugal is primed 112 PRACTICAL IRRIGATION AND PUMPING to prevent damage to piston and valves by water caught in the clearance. Discharge Pipe. The same statement as to size and alignment of pipe from the standpoint of friction applies to the discharge as to the suction. A small pipe with many elbows and bends causes unnecessary friction losses, and it is a direct saving of power to use large pipe and as few elbows as possible. Water Hammer. An important matter in connection with design of discharge pipe is water hammer. When a pump is suddenly stopped, as in the case of a motor- driven unit by the circuit breaker tripping, there is a surging back and forth set up in the discharge piping which for a few seconds may multiply by many times the usual working pressure. If a check valve is used this excess pressure comes upon it and the discharge pipe, but when a foot-valve is employed the hammer effect is also expended upon the pump case and suction pipe. With very long discharge pipes and heads of over 50 feet some reliable means should be employed to relieve excess pressures. This may take the form of quick-acting relief valves with a free water passage at least one-tenth the area of dis- charge pipe, or it may be a large air chamber or a verti- cal surge pipe connected to the discharge above the check valve. Fittings. Where, as is illustrated, the water is to be conveyed to a point immediately above and adjacent to the plant, it will be found an economy both in the saving of pipe and of power, through lessened friction losses, to use 45-degree elbows in the discharge line. "Long Sweep" fit- tings should be used wherever right-angle turns are made either in suction or discharge pipe, and if stop valves are used they should be gate valves rather than globe valves. The use of the latter cannot be too strongly condemned in INSTALLATION FOR CENTRIFUGAL PUMPS FIG. 25. A common type of plant for pumping from driven well, where water is found at shallow depths. 114 PRACTICAL IRRIGATION AND PUMPING water-piping in general, but particularly in any case where economy in power is a consideration, since it is to be re- membered that a globe valve causes the same friction head as would be caused by 100 feet of the same nominal size of pipe. Pump. As to the type of pump to use for a plant such as is illustrated in Fig. 24, there is no question as to the eminent desirability of a horizontal centrifugal. If the total head is less than 70 to 100 feet the single-stage pump should be used, but for greater heads the necessary speed of the single-stage pump will be excessive for successful and long- continued operation and a multiple stage should be used. Where electrical power is available at reasonable cost a pump direct connected to a motor will be found most con- venient. If a gasoline engine is used, it should be selected of a make known to be reliable. Some suggestions on this point will be found in Chapter XIII. PLANT No. 2 The pumping plant shown in Fig. 25 is of a type very common in those sections of the West where water is found at shallow depths, as in river valleys immediately adjacent to streams, and water is pumped either to supplement a gravity supply or is used by some fruit or truck grower for an independent and dependable water supply. Pump Pit and Arrangement of Belt Drive. In such a plant an open pit 6 to 8 feet square or round is first dug to water and lined with timber or in some cases with con- crete. Owing to the use in this case of an inclined belt reaching from the engine at the surface to the horizontal centrifugal pump in the pit, such pits are limited to a depth of about 25 feet. With greater depths it is not feasible to INSTALLATION FOR CENTRIFUGAL PUMPS 115 use the inclined belt, owing to its excessive length. It might be suggested that in order still to use the horizontal pump with deeper pits, a counter-shaft could be placed across the top of the pit from which a vertical belt might extend to the pump. This has, however, a serious disad- vantage in the use of the vertical belt, which seldom works satisfactorily, besides which there is a loss of power in the use of the countershaft and two belts instead of one. With the inclined belt, the sag of the belt increases the arc of contact on the driving and driven pulleys and the weight of the belt gives the necessary adhesion to the pulleys without the excessive initial tension necessary for a vertical belt. Rope transmission has been suggested (and used in a few instances), for deep pits where it was desired to use a horizontal centrifugal pump, but the complication and expense of this form of power transmission do not recom- mend it for irrigation work. Since for reasons just given this type of plant is limited to locations where water is found not deeper than 25 feet below the surface, it may be possible in firm ground to dig the pit without lining same as excavation proceeds, but as soon as the pit and belt chute have been completed, the lining should be put in without delay. For a really perma- nent structure and where the expense is not prohibitive, a 4- to 6-inch concrete lining should by all means be provided and, in case concrete is used, it will be found more economi- cal in excavation and in use of concrete and but little more expensive in forms to make the pit circular. The inside diameter should not be less than 6 feet and the same mini- mum dimensions hold for a square or rectangular pit. In most cases a wooden lining will be used which under aver- age conditions should last seven to ten years before it needs renewing. A common practice in building a wooden lining is to use 2" x 6" or 2" x 8" vertical sheathing inside Il6 PRACTICAL IRRIGATION AND PUMPING of which a 4" x 6" horizontal framing is spaced every 4 feet vertically. Well Pipe and Strainer. The pit having been lined, the next thing in order is the sinking of the well-tube, a job which should be left to the professional well-driller, but which may be negotiated by the layman if he has the necessary tools and the large amount of patience required, together with considerable ingenuity. The matter of well sinking has been considered in Chapter V. If the top of the strainer is from 22 to 25 feet below the level of ground water, the suction connection of the pump can be made directly to the top of the well-pipe, but in case of a shallow water-bearing formation or where for any reason the top of the strainer is less than 22 feet below the level of standing water, then a suction pipe or draft tube should be dropped down inside the well-pipe and strainer to at least 25 feet below standing-water level and the pump connected to this suction pipe. Pump Foundation. No floor is required usually in a well-pit, but the pump should rest on heavy timbers attached to the pit lining, since the material for some dis- tance around the well-pipe is apt to settle considerably soon after pumping begins, particularly if much sand is removed, and unless the pump is securely supported inde- pendently of the floor, it is likely to settle out of align- ment, not only making it difficult to run the belt properly, but possibly cracking a suction flange connection. Fittings. The connection at the top of the well-pipe or draft tube should be a tee capped with a blind flange, rather than an elbow, since in case there is considerable fine sand around the strainer it is likely to accumulate inside the strainer and considerably reduce the capacity of the well. With a tee connection on the suction pipe it is a very easy matter to remove the blind flange, lower a INSTALLATION FOR CENTRIFUGAL PUMPS 1 17 sand bucket, and bail out the sand. In general, it will be found that it is decidedly more convenient to use flanged instead of screw fittings in all pipe work in the pit, due to the narrow quarters in which work must be done and the difficulty of using large pipe wrenches in making up screwed connections. Immediately above the pump a check valve should be placed in the discharge line, since, there being no possibility of using a foot- valve in a driven well, a priming pump must be used each time the pump is started, to fill the centrifugal with water. In some cases, flap valves are used on the end of the discharge pipe and we have seen some small plants provided with nothing better than a large, tapered plug covered with a piece of rubber belting, which was driven into the end of the discharge pipe to make an air-tight end. The disadvantage in this, aside from its crudeness, is that air in the long discharge pipe must be partially exhausted by hand pumping before water will rise into the centrif- ugal pump to such an extent that it will prime itself and establish the flow. A check valve, suitable for the use named, may be purchased at a reasonable figure in any size desired from any machinery or well-supply house, and is a justifiable expense. This valve should, if possible, be of the "increaser" type, that is, one end should have a flange connection the same as the flange on the pump, but the other end should be provided with flange for the next larger pipe size, since in all but the shallowest wells and lowest lifts the discharge pipe should be a size larger than the outlet of the pump, in order to decrease the friction head. It is also the part of wisdom to attach a vacuum gauge to the suction pipe near the pump. This may be done by boring and tapping out a hole for a J^-inch pipe connection. The hole should be bored on the horizontal axis of the pipe Il8 PRACTICAL IRRIGATION AND PUMPING near the suction flange of the pump. The vacuum gauge will indicate the " draw-down," and will show whether the strainer is open and in good condition, and whether the underground supply is holding out. In case the strainer is not clogged, a gradual increase in the reading of the vacuum gauge will indicate that the supply is failing and that the ground-water level is being lowered by pumping or other causes. A sudden increase in the " draw-down" usually indicates a clogging of the strainer by sand or other material, and is a condition requiring immediate attention. Drive. Little need be said here regarding the belt drive, since it is a very simple matter where a chute, as indicated, is used. Figure 25, illustrating such a plant, shows a gasoline engine drive, but steam engine or motor drive might be substituted, if conditions warranted and it was not desirable to use a direct-connected plant, such as is indicated in Fig. 26. PLANT No. 3 This plant is similar in every way to that just discussed, except that the pump is direct connected to a motor mounted upon the same bed-plate. This is not recommended where the pit is likely to be very damp, or where there is a possi- bility at any time of the ground water rising to such an extent as to submerge the pump and motor. The depth of the pit in this case is limited merely by the fact that a pit rather large in dimensions is needed and consequently the limiting depth for the direct-connected pump may probably be placed at about 50 feet. Advantages. This type of plant has many advantages from the standpoint of convenience over an engine-driven plant, and is what might be termed a standard type for central station systems of pumping where electric power is INSTALLATION FOR CENTRIFUGAL PUMPS 1 19 FIG. 26. An electric-driven plant for pumping from depths within 50 feet. A stand- ard type of installation for the individual plants in a central station project. I2O PRACTICAL IRRIGATION AND PUMPING generated at some central point, and distributed over a considerable area to numerous individual plants of the type described. The power being alternating current of standard frequency and voltage in all cases, the motor is of the induction squirrel-cage type, and is practically fool-proof. Pump and Motor Speeds. It is of very great importance in direct-connected sets to have the pump so built or selected that at the motor speed, which is practically unchangeable, the greatest efficiency will be realized from the pump for the particular conditions of head and discharge that prevail at the given plant. Unfor- tunately, since stock pumps are used in these installa- tions, and the motor speed will vary between 900 and 1,700 R.P.M. according to size and type, it only now and then happens that a pump is running at the speed at which it should run for the total head prevailing if a minimum current consumption is desired. By the use of characteristic curves as described in Chapter VII, this situation might be changed by choosing a pump which at the motor speed, the head, and discharge desired, would give a maximum efficiency. Lack of attention to this matter has discouraged several plant operators known to the writer, who, in paying excessive bills for electric current, were unknowingly paying for their lack of knowledge of the characteristics of a centrifugal pump. They were finally forced to abandon the use of electric drive in favor of gasoline engine or steam where inefficient plant operation does not so quickly make itself felt or noticed as is the case where one can see the dollars slipping away with every unnecessary revolution of the index of the watt-hour meter. Wiring. In the installation of an electric plant all wiring should be put in by an experienced electrician, and INSTALLATION FOR CENTRIFUGAL PUMPS 121 the wires should be laid in conduits both in the pump house over the pit and in the pit itself down to the motor. Attention Required. If the priming pump is at the sur- face, as is possible where the pumping set is not over 25 feet below the surface, about all the attention required of the operator besides seeing that the pump is oiled (and this may be done from, the surface, if desired), is to operate the hand pump until the centrifugal is primed and then by starting box to start the motor. No attention should be required by the plant during an eight- or ten-hour run, at the end of which time, of course, re-oiling is necessary. Electric Drive the Ideal Arrangement. Electricity very closely approaches the ideal power for irrigation pumping, and unless the local rates for current are excessive, it will pay one to consider its adoption very closely before deciding upon steam, producer-gas, gasoline, or distillate drive, since with these the expense for attendance is a consider- able item in the total cost of pumping, but it is almost en- tirely obviated with electric drive. With some of the best gasoline engines, attendance may be a very small expense, but the writer has yet to learn of any gas-engine-driven plant in which the operator could leave it entirely to itself during an 8- or lo-hour day, thus being free to give his entire attention to the distribution of the water. PLANT No. 4 This, as shown by Fig. 27, is introduced to show a design similar to that used on one of the largest low-lift central station pumping plants in the West, in which water is pumped from a river and distributed over a large acreage by means of concrete-lined canals and ditches. As shown by the figure, A is the centrifugal pump of very large size, driven by an induction motor, B. There may be, and in the instance cited, are, several such sets in one pump-house. 122 PRACTICAL IRRIGATION AND PUMPING INSTALLATION FOR CENTRIFUGAL PUMPS 123 The drive is by a silent chain, C, which enables a relatively short distance between shaft centres to be used. E is the priming pump, and F is the valve in the discharge line used when priming the centrifugal. L is a strainer set in a block of concrete, and D is the starting box. It will be noted that practically the entire lift in this case is suction lift. In pumps of this size, relatively high efficiencies were attained in tests by the builders of the pump and are probably con- stantly maintained so long as operating conditions remain the same as those for which the pumps were designed. In some cases, indeed, pumps in use some time show better efficiencies than new ones, owing, doubtless, to the lessened water friction as the cored-out passages become smoother under the scouring action of the water. Direct-connected units would probably be preferable to chain drive if the proper motor speed could be secured. Applications of the Low-Lift Plant. The type of plant illustrated by Fig. 27 is, of course, of limited application, since it is seldom that conditions are found similar to those for which this plant was designed. It may, however, in exceptional cases, be found cheaper to build a pumping plant rather than a gravity system, owing to slight river slope which makes a long canal necessary in order to reach the lands to be irrigated, or unfavorable conditions for a headgate, such as a shifting river bed, floods, etc., may sometimes justify the building of a pumping plant at some point adjacent to the lands to be irrigated. When either water-power or cheap coal is available, a careful study of relative costs may show a decided advantage in favor of a pumping plant similar in type to that shown in Fig. 27. PLANT No. 5 When the depth to standing water exceeds 25 feet, and for any reason an electrically-driven plant, as shown in 124 PRACTICAL IRRIGATION AND PUMPING Fig. 26, is inadvisable or impossible, it is customary to adopt the vertical centrifugal pump, which is driven by a shaft extending from the surface. Suspension Frame. The pump itself is suspended in a framework of wood or steel, which is held at the surface by a trussed frame resting on the top of the curbing. By this method of suspension, the pump is kept in true alignment with the shaft and no difficulties are encountered, due to the sinking or displacement of the pump, as might occur if the pump were supported independently of the frame on a foundation built in the bottom of the pit. The framework is always provided with cross and diagonal bracing on a 6- or y-foot spacing, and the cross-braces support self- aligning bearings for the vertical shaft, see Fig. 30. Step Bearing and End Thrust. At the surface is usually a cast-iron frame provided with ample bear- ings, to take the side thrust due to the pull of the belt, and a step bearing at the top, which takes the weight of the entire shaft, pulley, and impeller, and any unbalanced end thrust due to the action of the pump. This bearing is the most important bearing in the entire installation, and is the one which, if poorly made, is apt to give more trouble than all the rest of the installation combined. In some makes of pump, the end thrust due to the action of the pump is supposed to balance the weight of the shaft pulley and impeller, and elaborate means are provided to accom- plish this end. Usually, however, it is found that such schemes are more or less of a failure, since a slight change in operating conditions, such as speed or head, cause an unbalancing of the system and the necessity for re-adjust- ment. In other pumps no attempt is made to balance the weight of the shafting pulley or impeller, the hydraulic end-thrust being eliminated by the construction of the pump, and the weight of the rotating parts is taken up by INSTALLATION FOR CENTRIFUGAL PUMPS 125 ball or roller bearings in a well-designed step bearing. When well made, this latter type is likely to prove the more satisfactory under the conditions of irrigation work, al- though under stable operating conditions there is likely to be less mechanical friction loss and consequently a higher efficiency attained with the balanced step type. Stages. The number of "steps" or stages to adopt for the pump will depend upon the head. If the discharge is to occur at the surface and ground water is encountered at less than 50 feet below the surface, the single-step or single- stage pump may be used satisfactorily, but for greater dis- tances below the surface a two-stage, or multi-stage pump should be used. Although there is no reason why greater depths should not be negotiated (as indeed have been in various parts of the West), the limiting depth for really satisfactory operation of this type of plant may be said to be reached when the pit reaches a depth of 75 or 80 feet. Vertical shafts longer than this increase, rapidly, the difficulties of operation, and for greater depths it will be advisable to use either electric drive with a vertical or horizontal direct-connected motor-driven pump or some other type, as will be noted later. Priming. For gasoline or electric drive, hand priming is necessary, unless a small electric-driven air-pump aux- iliary can be afforded. If a hand pump be used for exhaust- ing air from the centrifugal, it must be located at the bottom of the shaft, for it will be found practically impos- sible to make joints in a line reaching to the surface suf- ficiently tight to enable the pump for priming to be placed there. With steam-driven plants, a 24-inch or i-inch steam line may be extended down into the pit to operate an ejector, which if used with a check valve between the ejector and pump, will obviate the necessity of going down into the pit 126 PRACTICAL IRRIGATION AND PUMPING at all for priming, since the steam may be admitted into the priming line by a valve at the surface. Discharge Pipe and Details. The same remarks as to well-pipe connections, valves, and discharge pipe apply to this plant as to those previously considered. A large-sized discharge pipe should be used and long radius tees and elbows rather than common fittings. Driving Pulley. The pulley at the top of the vertical shaft should be so placed that a horizontal plane through the crown of the pulley will be about on a level, possibly FIG. 29. a little above, the centre of engine pulley, when this pulley rotates, as shown in Fig. 28. When the centres of the driving and driven pulleys are from 1 6 to 20 feet apart, as they should be in such case, the weight of the belt and INSTALLATION FOR CENTRIFUGAL PUMPS 127 the pull of the tight side will cause it to lower on the ver- tical pulley as far as the tension will allow. It not infre- quently happens that the vertical pulley has to be re- adjusted in position after the plant is put in operation and sometimes it will be found necessary to use an idler for this type of drive, but its use should be avoided if possible. It should when used be placed not less than 3 feet from the vertical pulley, and the highest part of its circumference should be placed on a level with the centre of the vertical pulley. In a vertical plane, that side of the circumference of the vertical pulley from which the belt leaves, should be tangent to a plane passing through the crown of the driving pulley of the motor or engine. When the driving pulley rotates as in Fig. 29, a hori- zontal plane through the crown of the vertical, or driven, pulley should pass tangent to or below lowest portion of circumference of driving pulley. The same condition with respect to the vertical position of the vertical pulley holds as in the former case. PLANT No. 6 Vertical Electric Drive. Where electric power is avail- able, a very satisfactory type of centrifugal plant fulfilling the same purpose as stated for Plant No. 5 is shown in Fig. 30. The underground portion of this plant is in all respects similar to the one last described, but the drive is by an electric motor mounted vertically on a framework at the surface, and connected to the vertical shaft by a flexible coupling. A vertical thrust bearing takes the weight of shaft and couplings and the motor is self-contained, the weight of the revolving armature being taken up by a thrust bearing in the motor itself. Such a plant put out by an experienced and reliable manufacturer, although expensive, has many 128 PRACTICAL IRRIGATION AND PUMPING FIG. 30. A type of installation for deep well pumping using multi-stage centri- fugal pump in open pit. INSTALLATION FOR CENTRIFUGAL PUMPS 1 29 advantages in point of durability and convenience over the type of drive illustrated in Fig. 28, there being no belt- ing or idlers and the power being instantly available. This plant, like the one preceding, is, in the judgment of the writer, limited to pits not much exceeding 75 feet in depth. MEANS OF WATER MEASUREMENT In connection with an irrigation pumping plant, the im- portance of providing some means of measuring the dis- charge can scarcely be sufficiently emphasized, especially when centrifugal pumps are used. It enables a constant check to be made upon the performance of the plant, indicating when the pumps are falling off in efficiency or becoming clogged, or, in the case of the driven well plant, may indicate a fouling of the strainer or increase of draw- down. It enables efficiency tests to be made upon com- pletion of plant, and facilitates such tests at intervals during its life to determine if efficiency is being maintained. The plans for a plant should, if possible, therefore, always provide some accurate means of measurement of pump dis- charge. This generally takes the form of a trapezoidal weir at or near the discharge outlet, though, if the slight increase in head thus caused is objectionable, a rating flume might be used. In some cases, a Venturi tube may offer the only feasible solution, as where the water is distributed over the area irrigated in underground conduits. CHAPTER IX TYPICAL PLANTS NOT USING CENTRIFUGAL PUMPS WHEN the depth to water is from 75 to 100 feet, the multi-stage centrifugal pump, while not at all impractica- ble, becomes increasingly difficult to operate, satisfactorily, in the hands of the average man who is not a mechanical expert or without long experience in this work. The Question of Sand. It must be understood, of course, that where much sand is apt to be pumped with the water, as is always the case when the strainer is landed in a body of water-bearing sand, the only economically feasible method of pumping is by the centrifugal pump. The air lift need not be considered in this connection, due to its well-known lack of economy. In case the sand problem does not enter in, then it may be well to adopt the type of plant shown in Fig. 31. Duplex and Triplex Pumps. In this case we employ a duplex or triplex pump with the working head at the sur- face and the pump cylinders in a pit a short distance above the level of standing water. The pump plungers are oper- ated by rods extending between the pump cylinders and the working head, the rods being held in vertical alignment by roller guides attached to timbers extending across the pit at points spaced from 5 feet to 6 feet apart vertically. Drive. The working head may be driven by a steam or gasoline engine, but, where electric current is available, may be actuated by a motor connected either through gears or a silent chain belt. Capacity Limited. The capacity of such a pumping set is limited, since the number of strokes per minute cannot 130 TYPICAL PLANTS NOT USING CENTRIFUGAL PUMPS 131 FIG. 31. Showing use of triplex reciprocating pump in deep pit. exceed a certain maximum, owing to inertia effects. With a given size of cylinders, the capacity of the pump will be the volume displaced per minute by the two or three cylinders multiplied by a certain correction factor for slip which, in a well-designed pump in good condition, should not be less than 85 per cent., that is, the amount of water lost by slip should not exceed 15 per cent. Such pumps are made in sizes having piston displacement of 132 PEACTICAL IRRIGATION AND PUMPING from 50 to 300 gallons per minute at 40 working strokes per minute. These pumps for irrigation work should be pro- vided with leather-packed pistons and rubber valves, since when so equipped they are less likely to be injured by sand. Brass-lined cylinders, although better in other ways, are likely to be more quickly scarred and ruined by sand than are those less expensive. The suction and discharge piping should be installed according to suggestion already given for other plants. Advantages and Efficiency. This type of pump has the evident advantage over a centrifugal in requiring no priming, it can be used in a pit of a little smaller dimen- sions, the difficulties attendant upon the use of long vertical shafting are absent, and finally, the mechanical efficiency is somewhat higher than in most centrifugal pumps and should in a well-designed and installed plant amount to between 70 and 80 per cent., figuring between the power input to the working head, and the energy of the moving water. Vertical Rods. The vertical rods must be in perfect alignment for satisfactory operation, and the pump cylin- ders must be very securely anchored, since the alternate pull of the rods in a deep pit is a very severe stress which must be resisted by the anchor bolts holding down the cylinders. In a plant of this type installed by the writer, the cylinders were bolted to a very heavy timber bedded horizontally in the concrete wall of the pit. Deep Well Pumps The limit of open-pit construction may be said to be reached at a depth of 100 feet, and if an irrigation supply is to be obtained from depths greater than this, it is prob- able that a study of the problem will limit the solution to TYPICAL PLANTS NOT USING CENTRIFUGAL PUMPS 133 the use of some type of deep- well apparatus, i.e., a pump cylinder in a driven well with a pumping head at the sur- face, or in exceptional cases a deep-well turbine pump might be recommended, although this is limited to wells of large bore and is expensive equipment. Deep- well apparatus, either in the single and familiar single-acting pump cylinder or in the more complicated double-acting cylinders, and with various valves, etc., are made by a considerable number of makers, and competi- tion in this line has developed a type of machinery which in the better grades affords a striking evidence of the attention now paid to details. In the pump heads we now find such details as white metal bearings, oiling systems, drop forgings, massive and well-braced frames, and so on, where formerly it was simply put together to be sold rather than to run. Capacity Limited. Deep-well pumps are not particu- larly desirable for irrigation work (aside from the fact that deep-well pumping is expensive) since the quantity of water developed by such pumps is usually far below the most modest requirements and the flow of many days' pumping must be stored in a reservoir in order to provide for one day's irrigation. The capacity of such pumps depends upon the diameter of pump cylinder, the length of the stroke, the number of strokes per minute, and the slip. Speed. In general, the number of double strokes or, in other words, the revolutions of the crank, should not exceed 40 per minute, and for extreme depths probably not more than 25, in order that the stresses due to the reciprocation of the long sucker rod and heavy plunger, and the inertia of the water column may not be excessive. Drive. Where electric power is available, the most sat- isfactory drive is a motor mounted on the same base as the pumping head, and geared to it by a rawhide or cloth pinion. 134 PRACTICAL IRRIGATION AND PUMPING Belt drive from a steam or gas engine is equally satisfactory in case electric power is not available, and in this case it is advisable in selecting a pumping head to choose one in which the belt wheel is mounted as low as possible. Important Details. If a pumping head is desired in which there shall be freedom from annoying and costly breakdowns, attention should be paid to the selection of a pumping head which is massively built, in which there are substantial guides and a babbitted cross-head, all bearings should be babbitt- or brass-lined and those machines should be given preference which do not have overhanging bear- ings, and in which the gear teeth are cut rather than cast. Such attention to details of construction will mean very much more reliable machines, and one in which stoppages for hot boxes and repairs will be much less frequent, although of course the first cost of the machine is going to be higher than the machine in which not so much attention is paid to the refinements mentioned. With this type of plant, the capacity is so relatively small, even in the largest sizes, that it will be found that a reservoir is an absolute essential to its successful use in irrigation. CHAPTER X COST OF PUMPING Importance of Knowledge of Pumping Costs. The matter which most immediately interests, or should inter- est, the man who expects to take up the practice of irrigation by pumping, is that of cost, and it is upon this point par- ticularly that he should take special pains to become thor- oughly and reliably informed. There are, unfortunately, too many projects now in the West which probably would never have been carried through had the owners been careful to acquaint themselves beforehand with reliable information from unprejudiced sources on the various details of cost. Plant Owners' Statements Unreliable. It might be said at this point that the prospective pumping-plant owner will do well to accept with some reservations the statements of the owners of existing plants, both as to the cost of oper- ation and the capacity of their plants. Since many, if not most, of such plants are the product of the owner's labor and thought, he is bound to have a certain pride in his achievement which blinds him to its faults and leads him to entertain a possibly sincere conviction that his only pumping expense or charge is for power, that he uses less of this as compared with the amount of water pumped than any of his neighbors, and usually also makes him, when in public, estimate the capacity of his plant at just about double what an actual measurement will show it to be. There has long been a need for some authoritative tests to determine the actual cost of pumping and although a 135 136 PRACTICAL IRRIGATION AND PUMPING beginning has been made in the matter by several investi- gators, including the writer, there yet remains much to be done before reasonably reliable estimates are possible for a given set of conditions. Factors Affecting Cost of Pumping. The cost of pump- ing depends upon the following factors: (1) Cost of Power, which involves (a) The quantity of water pumped. (b) The total head through which this quantity is pumped. (c) Efficiency of pump. (d) Efficiency of transmission of power between engine or motor and pump. (e) Cost of steam, gasoline, distillates, or elec- tricity. (2) Interest on first cost of plant, and depreciation. (3) Maintenance and repairs. (4) Attendance. These several items, numbered i, 2, 3, 4, enter into the cost accounts of any enterprise involving power or manu- facturing a product by the use of power, and should, there- fore, enter into the calculations of the man who proposes to run a pumping plant on a businesslike basis. In order that each may be properly understood, the items will be discussed in order. (i) Cost of Power Head and Quantity Pumped Determine Power Require- ment. A pump running steadily raises a certain quantity of water through a certain distance every minute. Since each gallon weighs about 8K pounds, the pump lifts a certain weight every minute through a certain height, and consequently performs work just as does a laborer in a COST OF PUMPING 137 trench who elevates to the surface every minute or two a shovelful of dirt weighing a certain amount. Now, in the case of the laborer, if he throws out larger shovelfuls at the same intervals of time as before, and from the same depth, he is doing more work than before, as is he, also, if he con- tinues to throw out the same sized shovelfuls in the same intervals of time, but from a greater depth. This illustrates the two facts frequently ignored by pump operators, first, that if the quantity of water discharged per minute is in- creased when the head remains the same, the amount of work done by the pump increases in direct proportion; second, that if the quantity is not changed, but the head increased, the work will be increased in direct proportion. From this it follows that the work done by the pump varies as the product of head times discharge. Consequently, it will require twice as much power and the power cost will be double for a 5o-foot head what it would be for a 25-foot head for the same quantity of water pumped. It is of con- siderable importance, therefore, that prospective pumping- plant operators realize fully the fact that the attempt to pump against a high head jeopardizes their chances of suc- cess, since power used up in overcoming head, represents an outlay for which there is no return, whereas power used in increasing the discharge results in greater acreage irrigated and greater profits. It being seen, therefore, that head and capacity are the two factors governing the amount of power required, it is of importance to calculate or estimate closely the probable requirement of the specific design in question. The method of estimating the power requirement is given in Chapter VII, page 86. The power required being known, it is then in order to estimate the cost of power. This will depend, in turn, upon the type of engine adopted, for it may be said that that engine will or should be used, which under the local con- 138 PRACTICAL IRRIGATION AND PUMPING ditions will give power most cheaply. Without going into the mechanical details of the machines, which will be dis- cussed elsewhere, it may be said that a choice must be made from the following: (1) Steam Engines, (2) Gasoline Engines, (3) Crude Oil or Distillate Engines, (4) Producer Gas Engines, (5) Electric Motors. The cost of power will now be discussed for each of the several prime movers mentioned, as based upon informa- tion from tests made by the writer, and as drawn from sources known to be authoritative, as well as conservative. Steam. The cost of power developed in a steam plant varies considerably with the size of the plant. This follows from the fact that the smaller the plant the greater in pro- portion are the heat losses, due not only to lack of refinement in the details of the equipment, but also to certain physical laws which it is unnecessary to discuss here. The cost of power in a steam plant may be said to depend wholly upon the amount of coal used, since, under most circumstances, the boiler feed water is a minor expense, and may be neg- lected. In all centrifugal-pump plants an automatic, simple or compound engine should be used, since the rotative speed of the pump being high the engine speed may also be high. Corliss type engines are practically eliminated from con- sideration in connection with the size of plants which it is the purpose of the author to treat. For very large plants, such as used in the rice belt, Corliss engines are probably economically justifiable. The steam consumption of the automatic high-speed engine will vary from a minimum of 30 pounds delivered horse-power per hour in the larger engines to 50 or 60 pounds in the smaller sizes. The amount COST OF PUMPING of coal required per delivered horse-power hour of engine and boiler feed pumps will depend not only upon the size of engine, but also upon the size and type of boiler. In the smaller plants the amount of coal burned per delivered horse-power per hour of engine will vary from a maximum of nearly 15 pounds in the small plants, to an average of 8 or 9 pounds in the larger plants below 200 horse-power. The following diagram shows for different sizes of plants and different prices of coal, the fuel cost per delivered 100 90 80 |TO ^60 W50 r 30 20 10 HOURLY FUEL COST HIGH SPEED STEAM ENGINES AUTQMATIC-NON-CONDENSING II 1 / H 1 / " # / 7 */ 7 & c / c f / / / / 10 20 30 40 50 60 70 80 90 100 Fuel Cost- Cents per Hr. DIAGRAM 19 horse-power hour and the fuel cost per hour of operation. For plants operated but ten hours per day, about 10 per cent, should be added to the values given by the diagram for the standby losses. This diagram shows the basis upon which a reasonably safe estimate of the cost of power may be based in steam plants of sizes within the limits given by the diagram. To illustrate use of diagram, suppose that it is found, for a 140 PRACTICAL IRRIGATION AND PUMPING given set of conditions, that an engine capable of deliver- ing 70 horse-power is desired. Following horizontally across to the curve representing the price of coal, the cost per hour in dollars is found vertically beneath on the horizon- tal scale. The cost per hour for coal at prices different from those of the diagram may be obtained by direct proportion. Gasoline. The gasoline engine is most useful in pump- ing plants requiring less than 30 horse-power. At and HOURLY FUEL COST GASOLINE ENGINES / I / 1 / ^ / a ^' / $ / q ^ ? c f f / >1 ^ e> x " - X 10 20 30 .40 50 60 70 80 90 100 Fuel Cost -Cents per Hr. DIAGRAM 20 below that power, it makes the most convenient and cheap, if not always the most reliable power we have. In the hit- and-miss governed gasoline engines, the full-load fuel con- sumption for an engine in good order varies between about i pint of fuel per delivered horse-power per hour in the large engines, to about double that amount in the smallest. The COST OF PUMPING 141 power cost per hour is shown graphically in Diagram 20 for gasoline costing 16 and 24 cents per gallon. Crude Oil and Distillates. Crude oil has not come into use for engines of small power, say below 10 to 15 horse- power, and its combustion is attended by difficulties of so serious a nature that it can only be used in engines of special type which are very much more expensive per horse- power than ordinary internal-combustion engines. Crude oil from the California fields is difficult, if not impossible, to use in such engines without previous distillation, owing to its heavy asphaltum base, but oils from the Texas and Louisiana fields may be and are being used with fair success. Distillates, that is, oils from which the lighter hydro- carbons, such as gasoline and engine naphthas, etc., have been distilled off, and the heavy bases removed are very commonly used both in special engines and in engines of the ordinary type fitted with special carbu- retors. Kerosene is rapidly taking the place of gasoline in localities where gasoline is expensive, but kerosene, like other lower-grade distillates, cannot be used in gaso- line carburetors. The tremendous growth in the demand for gasoline for use in pleasure and commercial auto- mobiles, to say nothing of the great number of farm engines, tractors, etc., and its uses in the industries, is doubtless responsible for a rapid increase in its price, so that unless the condition is relieved, it will soon have a very serious effect upon the fuel-cost factor in pumping. One attempt to meet the situation has resulted in the evolution of a new fuel known as "Motor Spirit," which is said to be a combination of gasoline and kerosene and which sells at a few cents under the market price of gasoline. This fuel is claimed to have a higher heating value per pound than gasoline and to have no objectionable qualities as an engine 142 PRACTICAL IRRIGATION AND PUMPING fuel except a rather pungent odor from the exhaust. The introduction of this fuel is not thought, however, to have any very deterrent effect upon the rise in price of fuel oils in general. Distillates could, until recently, be purchased at the refineries at about 2^ cents per gallon F.O.B. in tank-car lots and rarely cost more than 4 cents per gallon at the nearest railroad point in tank-car quantities. The recent advances in crude oil have, however, somewhat increased these prices. Since a gallon of distillate weighs about 7 pounds, it follows that i pound of distillate at 4 cents per gallon costs 0.57 cents. The fuel consumption of an oil engine is usually stated in pounds of fuel oil per delivered horse-power per hour rather than in pints or gallons, owing to the variable weight of oil per gallon. The makers of the Hornsby-Akroyd type of engine in this country guarantee a fuel consumption at full load of i pound per delivered horse- power per hour. A test of a 17 -horse-power engine of an engine of this type at near full load, made by the writer, gave a fuel consumption of i.io pounds Solar oil per delivered horse-power per hour. This engine was used for driving a centrifugal pump. It probably is wise to count on a fuel cost of from 0.6 cent to 0.7 cent per delivered horse- power per hour for such fuel in an engine specially designed to use it, with oil at 4 cents per gallon. For oil at greater or less price per gallon than that mentioned the price per horse-power hour would vary accordingly. Crude oil engines of the high pressure or Diesel types, when in first- class condition, will operate on from 0.50 to 0.75 pounds crude oil per hour, but such engines in this country are not made in small sizes, i.e., less than about 125 horse-power. Producer Gas. Undoubtedly the cheapest power on earth to-day is producer gas when made in an efficient pro- COST OF PUMPING 143 ducer and used in an efficient engine. There is no question whatever but that in the gas producer lies the solution of the smoke nuisance and the problem of producing power cheaply from coal. In the course of time, when engines and gas-producing processes have become more perfected, it is doubtful if the steam engine will be used in any enter- prise where we have continuous operation. In other cases, also, the gas producer will be used when producer-gas equip- ment is so far reduced in price as compared with steam equipment of equivalent power that the fuel saved by the producer will more than equal the difference in the fixed charges on the two types of plants. Producer gas may be made from a great variety of materials, among which are anthracite coal, charcoal, coke, bituminous coal, lignite, oak bark, sawmill refuse, or wood of various kinds. Of these, the first three named are the most commonly and success- fully used, and those alone, indeed, which it is possible to use satisfactorily in the ordinary suction producer. Al- though producers for bituminous or soft coal have been devised for commercial use, they are much more compli- cated, and therefore considerably more costly, than the hard-coal producer. Unfortunately, anthracite similar to that of Pennsylvania is unknown in the Western States, except by importation, and its use is, therefore, out of the question, since it sells for from $6 to $12 or $15 per ton, depending upon the grade. There are some semi-anthracites, however, such as the product of the Cerillos Mines of New Mexico, which it is claimed have been used with consider- able success in suction gas-producers and which upon test have been found to yield i horse-power hour on 1.5 pounds of coal. This coal may be purchased at not to exceed $7.50 per ton at the nearest railroad point in most parts of New Mexico, Arizona, West Texas, and southern Colorado. In Mexico, mesquite charcoal has been used 144 PRACTICAL IRRIGATION AND PUMPING successfully, this costing about $10 per ton (gold). Lignites, of which great deposits are found in the Dakotas, have been tried in producers by representatives of the United States Geological Survey with apparently favorable results, though it is not believed that this fuel has been tried com- mercially to such an extent that much can be said about it so far. The other materials mentioned in the above list are of relatively low heating value per pound, and unless they can be used in the immediate vicinity where found or pro- duced, it is not likely that they are of commercial value as fuels. Electricity. As stated previously in discussing the various types of plants, electric power is by far the most convenient when it can be obtained, and plants run by electric current have a further advantage in that cost of attendance is reduced to practically nothing. It may be more expensive than power generated on the premises, although it is not improbable, in some cases where plant owners found electricity bought from a central plant more expensive than the power they could generate themselves by a gasoline engine or some other type of prime mover, that the owner of the plant ran the same himself, and therefore made no charge for attendance, and further that he made no allowance for interest, depreciation, or maintenance, basing his inference entirely upon fuel cost. It is customary for companies supplying electrical power to base their tariffs upon a certain monthly charge based upon the size of the motor, which must be paid regardless of whether the machine is or is not used. This charge may vary, say, from $i to $2 per horse-power per month, and is justified upon the grounds that the company must main- tain a certain equipment ready to supply this power when- ever needed, and since this equipment is subject to interest and depreciation charges it is reasonable that the customer COST OF PUMPING 145 should share this expense. In addition to this fixed charge, there is a power charge which varies according to a sliding scale, the greater the number of kilowatt hours used, the less being the charge per kilowatt hour. This may vary from as much as 7 cents per kilowatt hour to as low as 2 cents. Another method is to have a minimum and a maximum charge per kilowatt hour. Thus, say, for below 500 kilowatt hours per month it might be 5 cents per kilowatt hour, but above 500 kilowatt hours per month it might be 3 cents per kilowatt hour. In addition to this would also be added the monthly fixed charge based upon the size of the motor. In some cases with a pumping load and for plants of above 25 horse-power capacity, the charge is based upon a certain amount per horse-power, as determined from a peak load indicated by a graphic or curve-drawing watt- meter. The writer knows of contracts where the power charge for the season is $20 per horse-power upon a seasonal half -hour peak, the season being five months, and of others with a charge of $4 to $5 per horse-power per month upon a monthly peak of one hour. Such contracts are drawn up with the object of protecting the power company against excessive peak loads in large pumping plants with many units, and are supposed to encourage the plant operator to exercise judgment and discretion in the use of power. Peak-load contracts are, however, not at all justifiable in plants of only one or two units and they lead to much trouble and misunderstanding unless the power company furnishing alternating current is prepared to maintain a very uniform frequency and voltage. Power consumers entering into such contracts should install in their plants their own graphic frequency meters, by which they can ascertain if high seasonal or monthly peak loads are co- incident with periods of high frequency. If such be the 146 PRACTICAL IRRIGATION AND PUMPING case, there is considerable reason to believe that the peak load is due as much to poor operating conditions in the generating plants of the power company as to excessive load conditions in the pumping plant, due consideration being given, of course, to changes of head and to regime of pumping machinery. Still another method of charging for electrical service is known as the "flat rate." This varies from $20 to $40 per rated horse-power of motor for a season of five months. The amount charged varies according to locality and is usually based on a sliding scale, the larger the motor the less the rate. Where pumps are operated continuously and the motors are loaded up to their rating, this often proves the most satisfactory, both for consumer and power company. The figures given will vary widely with different companies, depending upon the size of the central station, upon whether it is hydraulic or steam, and, if the latter, the cost of coal. (2) Interest on First Cost of Plant and Depreciation It is very rarely, indeed, that the practical operator of a pumping plant considers interest and depreciation in with the cost of fuel as contributing to the total cost of pumping, the reason for this being, probably, that they do not appear as evident an outlay as the fuel bill. A moment's reflection, however, should make it apparent that the farmer who owns a pumping plant has invested in it a certain capital which he may have borrowed and upon which he may be paying current rates of interest, or it may be that he has tied up in the pumping plant certain savings which other- wise might be loaned at local rates. However this may be, it is certainly a good principle to figure interest at current rates upon the cost of the plant and add to this the cost of fuel, etc., since the pumping plant must earn this interest COST OF PUMPING 147 and enough to pay for fuel, otherwise it is certainly a poor investment. In other words, good accounting would sug- gest that the pumping plant be credited with the amount of water it produces at the value of this water either upon the farm to which it belongs, or upon its value if sold to surrounding farms at current water rates (the matter of water rates being entirely a matter of location, character of crops grown, etc.), and the plant should certainly be debited with the cost of production of this water. The cost of pro- duction will include fuel and supplies, repairs, attendance, interest, and, possibly, depreciation. If, at the end of a year, a balancing of the ledger shows that the plant has not produced enough water of sufficient value to cover these items, then it is a business failure. It may, of course, be changed in certain respects, where experience has shown that defects interfering with its reliable or efficient oper- ation exist, but if the plant is as reliable as others of its kind, then the best thing to do is to sell the plant for what it will bring. Depreciation has been mentioned in the above as a pos- sibility, since it depends upon whether the pumping-plant owner regards the plant as "a going concern" or an experi- ment. If he regards it as the latter, then it is not a per- manent improvement or asset, and no thought need be given to its renewal after it is worn out. If, on the other hand, it is to be regarded as a necessary appurtenance to the land and as something which gives to the farm its value, then it is wise to make immediate provision against the time when it will be consigned to the junk heap and a new equipment installed. This means, then, that in addition to paying for fuel, supplies, repairs, attendance, and interest on first cost, it should earn yearly such a sum as will, when put at interest, have accumulated at the end of such time as it may be expected the plant will last, a sum that will be; 148 PRACTICAL IRRIGATION AND PUMPING sufficient to pay for a new plant. It will be seen that it is a difficult matter to estimate the yearly sum which should be set aside, since, in the first place, it is a problem in annuities and compound interest, and, in the second place, it is very difficult to tell just how long the plant will last. In other words, what number of years is it reasonable to assume must elapse before the plant is so badly worn out that it will be cheaper to replace it entirely than to attempt to repair it? The problem is rendered all the more difficult, since not all parts of the plant will depreciate uniformly. Thus the probable life of a boiler is about fifteen years; of a steam engine, ten years; of a gas, gasoline, or distillate en- gine, eight years; of a cheap centrifugal pump, five years (unless the water is unusually free from sand, when it might easily be double that) ; rubber belting, three years, leather belting, five years; reciprocating leather-packed, rubber valve pumps, five years; deep- well ball- valve pumps, five years; wooden curbing and well timbers, five to seven years. Exceptionally good operating conditions and intelligent care may increase the above periods by double in the case of high-grade pumps, and for plants in use but a month or two out of the year the depreciation may be very slight, if the machinery is properly housed, covered, and greased while idle. The above periods may be taken as a basis, however, and either the depreciation figured on each item separately or the depreciation charge based upon the plant as a whole. In general, ten to twelve years will be the life of the plant as a whole, except in the case of electrically driven very high-grade plants, which with proper care should have a life of 25 years at least. The following table abstracted from Kent's Handbook shows the sum which must be put away at the end of each year at various rates of interest to accumulate $1,000 at the end of different intervals of time. COST OF PUMPING DEPRECIATION TABLE 149 Years to Run Rate of Interest, Compounded Annually 3 per cent. 4 per cent. 6 per cent. t; . $188.35 130.51 112.46 87.24 53-77 $184-63 I26.6I 108.53 83.29 49-94 $177-39 119.13 101.03 75-87 42.96 7 . 8 10 IS . As an illustration, suppose that in a district where money may be compounded at 4 per cent, annually, a pumping plant costs $2,500. Then for an estimated depreciation period of ten years there should be added to the yearly cost of operation account 2.5 X $83.29 = $208.22. If at the end of each year this sum is placed in a bank on time de- posit at 4 per cent., it will at the end of ten years amount to the original cost of the plant. This scheme, known as an amortization or depreciation account is regularly adopted and followed by business firms employing machinery sub- ject to wear and tear, and it is certainly worthy of being imitated by the man who regards a pumping plant as a business proposition and not as an experiment or play- thing. In the above illustration it will be noted that the depre- ciation is about 8> per cent., and if the prevailing rate of interest in a locality on real estate is, say, 8 per cent., then 1 6^" per cent, of the original cost of the plant must be earned by it annually in addition to fuel and other items of expense in order that it may be considered a financial success. (3) Maintenance and Repairs In every plant, of whatever description, there are things constantly needing to be replaced or purchased new, and 150 PRACTICAL IRRIGATION AND PUMPING even in the best plants small repairs will need be made from time to time. Thus in gasoline-engine plants, batteries will need replacing, spark plugs will become short-circuited, valves may warp beyond possibility of grinding, and need replacing, etc., while in a steam plant valves must be re- seated, gaskets replaced, leaky boiler tubes repaired, etc. Then occasionally some carelessness in operation may result in a ruined bearing, a cracked cylinder jacket, and so on. All such repairs and replacements must be charged against the plant and enter into the cost of operation, for they will occur from year to year and no plant is free from them. The same is true of such items as waste, lubri- cating oil, etc. None of these items of expense can be neglected by the man who really desires to know how much it costs to pump water. (4) Attendance An oil-engine plant (distillate or gasoline), when in good condition, requires a relatively small amount of attention, although it is scarcely true, as some engine manufacturers claim, that such a plant can be started in the morning and require no further attention till it is shut off at night. While this may be true in theory, the practical operator finds that a gasoline or distillate plant requires from one- third to one-fourth of a man's time, when including in the course of a season all those vexatious little delays due to faulty ignition, choked or wet carburetor, hot boxes, etc. For a steam plant or producer-gas plant, the constant attendance of a more or less skilled man is required during the entire season of pumping, and in a large plant the man will need the occasional services of one or more helpers. Pumps driven by synchronous or induction motors of under 25 to 40 horse-power will need very little attention except an occasional oiling and possibly replacement of COST OF PUMPING 151 packing. Larger plants up to 500 horse-power, these being in general those which pump from surface sources, will require a regular attendant who may, however, also act in many cases as ditch rider and water master. Plants above 500 horse-power will require the constant services of a skillful operator and a night helper in case of 24 hours' operation. The matter of skilled attendance is a most difficult one for large electrical pumping plants, since the season is usually only five months, and it is impossible to secure the services of really capable operators for merely that length of time. Much of the success of these projects will depend upon the management being able to provide 12 months' employment at an attractive salary for men able to operate the pumping machinery on an efficient basis and maintain it in the best condition for reliable operation. It is evident, therefore, that to the cost of power, interest and depreciation, maintenance and repairs, there is still another item of importance to be added in order to arrive at the true cost of pumping. If a man is employed as attend- ant who devotes a portion of his time to the care of the plant, the total number of hours during the pumping season when he is so occupied should be kept recorded and the proportional part of his season's pay charged against the plant. In case the owner of the plant attends to it himself, his natural tendency is to regard the charge for attendance as nil. This is fallacious, however, for the time so occupied might be given to other equally or perhaps more profitable work. If possible, therefore, the owner should estimate the number of hours devoted to the plant during the season and charge it up at the same hourly rate as would be paid competent help hired for the purpose. CHAPTER XI THE QUESTION OF COST AND PROFIT ON A SMALL FARM IRRIGATED BY PUMPED WATER Elements of the Problem. The most vital question confronting every individual, whether he be an irrigation farmer considering the advisability of building a pumping plant or a business man venturing upon an enterprise of any sort, is whether the project will pay. This question is paramount to any question of design, installation, or oper- ation, but unfortunately it involves in its consideration and solution more or less definite knowledge on each of these three points since, until some definite figures are available on the actual cost of obtaining water, one very essential element in the problem is lacking. For this reason, there- fore, a consideration of the question of cost and profit has been deferred until amounts of water needed and types of plants to secure this water could be discussed and some better idea secured, perhaps, of those elements which enter into the cost of water. The latter, it must be recognized, is, however, only one of a number of elements which de- termine the feasibility of a project from the financial standpoint, and preliminary estimates, which are to deter- mine whether there is a reasonable expectation of profit, should, also, involve the following: FIRST A Fair Estimate of Yields, This should not be based upon any exaggerated idea of the fertility or richness of the soil, or upon results secured, possibly, by some farmer of the district who, by unremitting toil, fertilizers, and special knowledge and methods, has secured phenomenal returns from an acre or two of ground. Neither should one swing to the other extreme and take as representative THE QUESTION OF COST AND PROFIT 153 the results of the less skillful farmers or those who through lack of water, unskillful cultivation, or insect or rabbit depredations made little or no crops. The endeavor should be made to gauge as nearly as may be the average crop- producing ability of the land when given intelligent care and attention, and taking results over a series of years if such information be available. Care and discrimination are often necessary when crop information for the exact locality is not available, in projecting such data for other districts than the one in question, for although climatic and soil conditions, altitude, etc., may be thought exactly the same, it is frequently found that the productive capacity is widely different. SECOND The Cost of Crop Production. This will in- clude all those expenses incident to the growth and har- vesting of the crop, and are fairly well defined and under- stood by those having any practical knowledge of irrigation. THIRD Shipping Costs and Market Rates. Both of these items enter into the question of returns, and should not be neglected by the individual who is making a careful study of the possibility of profit. It is a local matter entirely, and is one requiring careful and full investigation. The matter of suitable market and one convenient of access is very largely a determining factor as to whether it will pay to invest in an irrigation scheme, for however cheaply the water may be secured, if it is in a region remote from rail- roads and suitable markets it is evidently of little use, except, possibly, in the development of a cattle-raising industry. FOURTH Money Rates. Labor Conditions. Cost of Materials. That these are also involved in any local set of conditions, and will enter directly or indirectly into the whole problem of cost and profit, is too evident to need emphasis or discussion. 154 PRACTICAL IRRIGATION AND PUMPING Demonstration of a Problem in Cost and Profit. To demonstrate a consideration of the problem, a tract of 20 acres will be assumed in a locality where alfalfa, grains, melons, truck crops, and orchard fruits (not citrus) may be grown. It will be assumed, for convenience, that it is proposed to drive the pumping plant by electric current, although, of course, the kind of power used might be gaso- line or steam engine, and the method of attacking the problem would be the same. The type of plant is that shown in Fig. 26 (Page 119), and its capacity will be taken at 300 g.p.m. The cost of such a plant for various total lifts (static, suction, and friction head) is shown by the accompanying diagram. The costs given are based on Total Head in Feet , 8 S 3 8 S g ESTIMATED TOTAL COSTS DRIVEN WELL-DEEP PIT TYPE ELECTRIC DRIVEN PUMPING PLANTS CAPACITY 300 G.P.M. HORIZONTAL CENTRIFUGAL PUMP DIRECT CONNECTED TO 2 OR 3 PHASE -60 CYCLE- 220 VOLT INDUCTION MOTOR x f s' ' y x" ^ S* X ^' ' s' '" x X ^ < J x" x x f ** , X X s' x x x X x "* ,x yf COSTS INCLUDE Pump and Motor Pipe -"Valves -Fittings Transformer Starting. Box Switches -Fuses -Wiring Excavation and Lining of Pit ' Well Sinking ' Installation Housing x ~x x' X ^s' X x jX ** X x^ .X x /' X /' s'' X NC )Th ' - Maxin 3De 3th 'of I 'f 1 1 -I 3 600 800 1000 1200 1400 1600 Cost in' Dollars DIAGRAM '21 1800 2000 2200 actual Quotations of machinery jobbers, and include esti- mated freight charges for 100 miles and labor in erection. The cost of a 3-phase 6o-cycle 2 20- volt mo tor "and starting box is also included. THE QUESTION OF COST AND PROFIT 155 Assuming that 30 acre-inches will be required by the tract per season (this including all distribution losses), it will be found that the plant must operate about 900 hours. Estimating the energy cost at 3^ cents per kilowatt hour (in advance of more exact knowledge of cost of power when total power requirements for a given project are known), and allowing an interest charge of 8 per cent, and depreci- ation, taxes, etc., of 8 per cent., we find that the total fixed and operating charges of an electric-driven pumping plant are as shown in Diagram 22. With the operating and fixed charges known, the prob- Total Head Ft. _, >_ t? C * D 1 ' W 5 O o o-o o ESTIMATED TOTALS OF FIXED AND OPERATING CHARGES Pl/MPING PLANTS OF 300 G.P.M. CAPACITY WORKING AT VARIOUS TOTAL HEA.DS INDUCTION MOTOR DRIVE DRIVEN-WELL DEEP PIT TYPE / * s / / / ' / / /* /\ / / / ENERGY COST AS FOLLOWS 6600 Volt current to customers transformer 3.5 cents per K.W. hr. plus 50 cents per month of irrigation season per motor H.P. OTHER CHARGES VIZ: Interest, taxes, amortization =.16$ Attendance, repairs, supplies=$5 x H.P. NOTE i For plant cost see previous Diagram / s / s 100 200 300 400 500 600 700 800 Yearly Total Cost Dollars ' DIAGRAM. 22 900 1000 able returns may be calculated. The table on page 156 the basis of estimate of net returns on various crops. A farm of 20 acres was selected as representing probably the limiting size for a'man of average means starting in a new country. The division of this acreage for the first few years is a matter upon which considerable difference of opinion may arise. Some would doubtless attempt a large 156 PRACTICAL IRRIGATION AND PUMPING a o; o S3 co S = fcH $<<[>'-> 10 10 10 o o * o o o 00 cO OO HH QN Tj- ri" \O ^ CO iO (S -H (N co ^ M Iiil ^O *-O 10 O O O O O ity ^ iff OOOOO -QOO iOOiOOO -OOO *]! ** Net Return per Acre. 1 lOOiONOiO -OQ <^ iO t^** CO cO co O O *fc ' CO -. Total Gross Return per Acre. 8 8 8 8 S-5 8 : : o 10 o o oo co o 1 a i i 5 5 ? J3 ^ ^ C O 10 O ^ . ^ Tt ,3 2 o > s 2 S2 H2 c ' : Q 3 3 3 Q 10 o .0 o rh : : ^ per cent, of water. No electric-ignition system is used, but a component part of the engine is an air compressor for starting and for injection of the fuel. A slightly modi- fied and less heavy engine of this type put out by an l8o PRACTICAL IRRIGATION AND PUMPING American manufacturer is in successful use in California on the crude oil of that State, with its troublesome asphal- tum base. The Diesel engine uses less than i pound of oil per horse-power per hour and is destined to have an important future in the Southwest and in California, where oil fuel is abundant and cheap. Its use in relatively small pumping plants is, however, exceedingly questionable in view of the more or less expert attendance required and in view of the fact that it costs about double the price of a gasoline engine or steam plant of equivalent power. The Gas Producer and Engine. Principles of Opera- tion. The elementary action of the gas producer is familiar to all in the example of the ordinary heating stove, which, after being well filled with fresh coal, will sometimes experi- ence a mild explosion due to the formation and ignition of gas. The gas is formed by slow combustion without sufficient draft or air supply to produce complete oxidation or burning of the carbon of the coal. The same principle is made use of in the gas producer in which we have a large retort lined with fire-brick in which coal is burned in the presence of an inadequate air supply, causing the formation of an explosive gas which, however, is caused to explode in an engine cylinder and do useful work in driving forward the piston of the engine. Since the gas formed in the process is used directly and its heating power given out directly in the engine cylinder, it is evidently a much more economical method of using the coal than burning the gases given off by this coal under a boiler, generating steam thereby, which is transmitted a greater or less dis- tance through pipes and finally used in a steam engine. Aside from any questions of practical operation, therefore, it would be apparent to any one that a gas-producer plant would be much more economical in fuel than a steam plant. While the process of generating the gas as above PRIME MOVERS l8l outlined seems simple, the difficulties in cooling, cleaning, and regulating the gas flow, and of firing, cleaning, and poking the producer, to say nothing of finally utilizing the gas in the engine cylinder, all have to be overcome and solved before the plant can be called a commercial success. Up until the last half-dozen years, the producer plant was a thing to be avoided by the man not wilfully seeking trouble, but the genius of American designers and engineers, backed up by several years of practical experience, has finally evolved plants made by several manufacturers, which under the right kind of management and when provided with the right kind of coal, give no more difficulty in operation than many steam plants. The type of producer best adapted for plants of the size usually adapted for small pumping plants is the suc- tion producer. In this, the air passing through the producer is caused to flow by the suction effect of the engine, which upon the suction stroke takes in a charge of gas. There is, therefore, no gas storage and the amount of air passed through the producer varies directly with the speed of the engine. The plant consists of (i) the producer proper, with its accessories, such as a blower and feed-hopper; (2) a vaporizer (usually) in which the hot gas coming from the producer is cooled by coming in contact with water- cooled surfaces; the air taken into the producer by suction being made to pass through the vaporizer, takes up moisture besides being heated, and thus increases the efficiency of gas production. After leaving the vaporizer, the gas passes into (3) the scrubber, where tarry substances and dust are removed by the gas coming in contact with a spray of water, and being made to pass through thick layers of moist coke. Leaving the scrubber, the gas next passes through (4) a purifier in which is excelsior, sawdust, and the like, intended to remove the moisture and any re- 1 82 PRACTICAL IRRIGATION AND PUMPING maining suspended impurities. Anthracite coal and coke or charcoal are the fuels best adapted to be used in this type of producer, but bituminous producers are being experimented with by many of our foremost engineering establishments and it will doubtless be only a matter of time till the chief trouble now experienced in the use of bituminous coal will be removed, namely, the dust and tarry products formed which, unless effectually removed, prevent the satisfactory operation of the engine. The bituminous producers so far devised have proven quite reliable, but their cost is very much greater than the ordinary suction anthracite producer plant. The latter are made in sizes of not less than 25 horse-power, a com- plete plant and an engine of this size costing, approxi- mately, $1,800 at the factory. Conditions Warranting Adoption of Gas-Producer Plant. The writer does not recommend this type of power plant except under these conditions: 1. When anthracite, coke, or charcoal are available, or a semi-anthracite, the equivalent of New Mexican Cerillos. 2. When the power required is in excess of 50 horse- power. 3. Where steam coal costs over $3.50 per ton, gasoline over 1 8 cents per gallon, distillate oils over 9 cents per gal- lon, and electricity is either not available or will cost on the average over 5 cents for horse-power hour. 4. When attendance of intelligence and skill is avail- able to run it. We caution the prospective plant owner against adopt- ing a producer-gas power plant except after a most careful inquiry into its merits and defects as compared with other kinds of power, and he should be particularly careful not to base his judgment upon fuel economy alone, for PRIME MOVERS 183 reliability or ability to pump water when crops most need it is worth more than all the coal which might be saved in a year's operation by the producer plant as compared with steam. The average pumping-plant operator can ill afford to lose a crop because he has a plant which fails him, largely for want of knowledge of how to run it, at the most critical time, probably, in the whole season. Men given to enthusiasm over new mechanical devices and methods are apt to regard the producer-gas plant as the key to the whole problem of cheap pumping and either purchase such a plant themselves or induce their neigh- bors to do so, although they have but the most meagre and superficial idea of the kind of fuel needed, the practical operating difficulties, and of the economic conditions which make such a plant needed or advisable. Electric Motor Drive. Practically all irrigation pump- ing plants with electric drive use, or will use, three-phase alternating current of frequency of 60 cycles. Below 25 horse-power the voltage commonly used is 220, but above that power 440 volt motors are usually adopted. For primary distribution, 66,000 and 44,000 volts is common practice over extensive districts, and there are a few large pumping plants taking power direct from such primary lines. The transformer-room equipment is, however, very complicated and expensive, consequently most pumping installations of moderate size should be designed, if possible, to receive current at 2,200 volts, the usual plan being for the plant to own the necessary transformers to step down this voltage to that of the motors, current being measured on the low-tension side. Below 50 horse-power ordinary "squirrel-cage" induc- tion motors are commonly used, but above that power "slip-ring" induction motors with starting rheostats should invariably be employed, owing to the excessive starting 184 PEACTICAL IRRIGATION AND PUMPING current required by the " squirrel-cage " type in large sizes. Where a large number of induction motors are on a system, operating conditions frequently require the power company to insist on the adoption of synchronous motors in plants of large size. Such motors have desirable characteristics for pumping-plant operation, but require special starting apparatus, which makes them less conveni- ent than an induction motor, and more expensive. As to make, there is but little difference between small- size induction motors of different manufacturers, the chief consideration being one of sufficient provision for ventila- tion. It is of importance, therefore, with any make, that in locating motors in the plant, they be placed where there is ample air circulation, avoiding corners and small, boxlike shelters, etc. Although induction motors stand severe abuse they sometimes burn out through being too heavily overloaded and through prevention of air circulation by accumulations of oil-laden dust on the coils and in the ventilating ducts. As to whether the motor should be direct- or belt- connected to the centrifugal pump, there is some chance for argument. It is sometimes impossible to get the correct speed when using a direct-connected small stock pump and motor, and to get this a belt-connected set is the only solu- tion. A belt, however, is a constant source of trouble, expense, and energy loss. These are completely avoided in the direct-connected sets. CHAPTER XIV THE CENTRAL STATION PUMPING PLANT Locations Suitable for Central Station Plants. In various parts of the West are to be found large tracts of land in every way suitable for agriculture, but too remote from surface streams to make gravity irrigation possible. Not infrequently such tracts are found to overlie a sub- terranean water supply of sufficient magnitude and at such depth as to make it possible to irrigate all or a part of the tract by pumping. Such tracts are: The country surrounding Portales, N. M., part of the Estancia and Mimbres Valleys of New Mexico, the Santa Clara Valley, Arizona; the Riverside district and the San Joachim and Sacramento Valleys of California; the Willamette Valley, Oregon, the Valley of the Arkansas in Kansas and Colorado, and various others. Where topographic, hydrographic, agricultural, and economic conditions all favor the development of such sections or tracts by pumping, it is generally recognized as probable, at least, that much the cheaper and more economical scheme of rec- lamation is by a series of pumping plants scattered over the tract, each driven by power generated at a central plant rather than the same number of individual plants, each generating power by a small engine, perhaps very wasteful of fuel. At the central plant large power units being used, advantage may be taken of th"e most up-to-date and eco- nomical machinery and thus power may be generated and distributed to the separate plants at considerably less than it would cost the individual owner to generate it himself. 185 1 86 PRACTICAL IRRIGATION AND PUMPING Conditions Governing Feasibility. The feasibility of such a plant depends upon a number of conditions, of which we may mention the following: (a) Adequacy and chemical character of water supply. (b) Depth of water supply and probable total head to be pumped against. (c) Suitability of tract for agricultural purposes. (d) Shipping and marketing facilities. (e) Size and shape of tract as governing cost of dis- tribution lines and transmission losses. (f) Ownership of tract. (g) Possibilities of co-operation in case of private ownership of land and desirability of local ownership and control. (h) Fuels obtainable and price delivered. (i) Possibility of utilizing power at times other than during pumping season, e.g., beet sugar and canning industries, city light and power, inter- urban transportation, general manufacturing, and other purposes requiring power. It will be seen at once from a consideration of these conditions that a decision as to the feasibility of the cen- tral pumping plant must be based upon a broad study of a large number of more or less closely related subjects, all of which have an important bearing upon the feasibility of such a project. It is, therefore, not a matter to be decided offhand, and parties interested in the development of such a project should call to their aid the services of a compe- tent engineer, -who is familiar not only with the engineering features of such a scheme, but who also understands thor- oughly the irrigation side of the problem. He should be required to inquire carefully into all sources of information and present a report covering essentially the points men- THE CENTRAL STATION PUMPING PLANT 187 tioned on page 186, besides giving his opinion as to the practicability and feasibility of the project. Upon such a report, if favorable, local parties will be warranted in vigorously furthering the scheme and with such a report, if it is desirable or necessary to introduce outside capital, there is much greater chance of interesting the careful in- vestor or capitalist than by setting before him a mass of glittering generalities, guesses, and assumptions. (a) Adequacy of Supply. Naturally the first question of importance arising is as to the source of supply and its adequacy to the purposes of the project. Where a surface stream is to be used, simple measurements and records of stream flow will give the desired information as to available supply, but where an underground supply is to be developed the matter is largely mere conjecture when considering the possible demands for a large acreage. Existing wells in the tract, if any, should be investigated, and, if possible, their maximum capacity ascertained. If no suitable wells are available, test wells should be sunk and such temporary machinery be installed as will enable a test to be made not only of the flow, but of the draw-down at various flows, and of the saline ingredients at various depths. If the saline contents exceed 4,000 parts in 1,000,- ooo, it is doubtful if the water could be used with success in irrigation. In the event of an excess of salts in the water at one level, it is frequently possible, by going deeper, j to strike other strata in which the salty ingredients are so small as to be harmless. Regarding the total amount of water required over the tract at any given time, it will be found an extremely dif- ficult matter to arrive at any very satisfactory estimate. If the tract be divided up into a large number of small holdings and only one crop is grown, it is not unlikely, unless a very unusual degree of co-operation and organi- 1 88 PRACTICAL IRRIGATION AND PUMPING zation exists, that all will require water simultaneously, making a very serious draft upon the underground supply and a severe demand upon the power station, not only by reason of the large number of pumps in operation at the same time, but also because the draw-down is likely to be increased considerably, shortly after pumping be- gins. Diversified crops and a centralized distribution, whereby the periods of demand for power will be more constant, is something which is imperative in economical operation of a central power station for pumping and is a matter which should be well understood by all interested parties. The value of diversified crops in increasing the yearly load factor can best be shown by an example. Let us assume an area of 10,000 acres planted in crops as follows: 3,000 acres in alfalfa, 3,000 acres in orchard, 3,000 acres in small grains, etc., "i,ooo acres in melons and truck gardens. These areas may be assumed as divided into a large number of small holdings, but since, under ordinary cir- cumstances, each crop over the entire district will need water simultaneously, it is probably equivalent to a single holding. This statement may need some modification in practice, since not always will the judgment as to a crop's moisture requirements be unanimous. Among the more skilful and experienced irrigators there is, however, sur- prising unanimity of opinion as to the time for irrigation, as has been proven time and again in the management of irrigation enterprises to the dismay of managers, who for some reason might happen to be ill-prepared for large and sudden demands for water. Upon the above assumption, however, we may construct the following diagrams, which represent days of the irrigation system on the horizontal THE CENTRAL STATION PUMPING PLANT 189 axis and acre-feet on the vertical axis. Sections i to 4 rep- resent the individual weekly requirements of the different crops, while section 5 represents the total amount to be supplied the entire area from week to week. As will be seen from the diagram, the total water requirement for the entire acreage under diversified crops 00 ioo r^\ r I \( ^ E -^ /LUX \ if : \ 1 km . I 4 1 : 2ioo ^200 ALFALFA ORCHARDS GRAINS MELONS, TRUCK GARDENS 500 \ \ / f \ .400 * ->. \ \ \ /* /" "^i \ \ S \ r~ '^\ \ i \ -<200 100 1 1 ; J E: ^ "Ts ^ CSS 4i-/ i ' / 8 12 16 20 24 28 32 3o 10 41 48 52 56 60 64 68 72 76 80 84 88 92 96 lOODays SUMMATION OF ENTIRE DEMAND OF DISTRICT^ DIAGRAM 24 is much less irregular than the individual requirement of any particular crop, and the maximum total amount re- quired, which governs the maximum power requirement and therefore the size of the central station equipment, is 1 6 per cent, less than though the entire area was in alfalfa. Such diagrams, compiled from the best available I go PRACTICAL IRRIGATION AND PUMPING information as to number of irrigations and periods between same and the probable water requirement for such crops as may probably be grown successfully in the district in question, should always be used to assist in forming an intelligent estimate of the total water requirement, and, therefore, the probable power requirement when the head is determined. (b) Head. We have already considered the question of draw-down and of static and friction head, so that when the average depth to water is known, it may be possible to estimate with some degree of accuracy that total head against which the water must be pumped. This will, of course, be an average for the district, since the ground- water plane being practically level, the static head will naturally vary with the topography and the total head on some quarter sections will be very much different from that on others in the same section, particularly in a rolling country. Now arises the important question as to what crops may be grown with profit, using water pumped through the average head just estimated. This is really the crux of the whole matter, and involves in its solution not only a fair estimate of yields, producing costs, shipping costs, and market rates, but also an approximately correct idea of the cost of producing and distributing electric power over the area in question. The best way of attacking this problem is to determine first the approximate cost of delivering an acre-foot of water at the surface for an acreage of a specified amount, using an average power cost, and taking into account all legitimate expenses chargeable to pumping plant. In es- timates and calculations on this problem, it should be considered from the standpoint of the individual owner of small means, who is the one most likely to be attracted THE CENTRAL STATION PUMPING PLANT IQI by the irrigation pumping proposition. This problem has been analyzed in Chapter XI, headed "The Question of Cost and Profit on a Small Farm Irrigated by Pumped Water." (c) Suitability of Tract for Agricultural Purposes. This had best be determined, if in an absolutely new country, by the character of the native vegetation and by the depth and physical character of soil as determined by borings and test pits in different localities. Samples of the soil should be sent to the State Agricultural Experiment Station for chemical analysis, and an opinion obtained as to its value for agricultural purposes, while if time and opportunity permit it may be well, also, to have an agri- cultural expert go over the ground and give his opinion as to its suitability for the proposed purpose. In case agri- culture of any kind has been carried on in the tract or upon adjoining tracts of like character not too far distant to present materially different climatic and soil conditions, it will be necessary to obtain, if possible, full information as to feasible crops, yields, action of soil under irrigation, i.e., whether easy or difficult to irrigate, etc. The opinion of experienced irrigators on the tract or in vicinity should be given great weight, usually, on general questions of methods of irrigation. (d) Shipping and Marketing Facilities. A very im- portant consideration in connection with the success of any project, except where it is so fortunately located as to be within hauling distance of some market big enough to absorb the entire product, is evidently the provision for getting the product to market cheaply and quickly. In the case of perishable products, like melons and garden veg- etables, some well-organized shippers' association is neces- sary to handle the product successfully, and of course it goes without saying that, in such cases, the tract must 1 92 PRACTICAL IRRIGATION AND PUMPING have railroad facilities. A wagon haul of more than 3 miles is usually fatal to melons and a haul of as low as 5 miles with alfalfa will pare down the profits materially. It is well, therefore, to canvass the question of markets and shipping facilities pretty thoroughly in a report, and state what market is available or may be developed and how reached. This matter, while more one of economics than engineering, cannot be neglected by any engineer who attempts to present a fair and unbiased report upon the merits of an irrigation enterprise, either for his clients or for the general public. The history of the West is replete with instances where engineers have considered merely the engineering features of irrigation enterprises, and have ignored the equally important question of whether it will probably pay their clients to finance the undertaking, and the general public to attempt to farm the lands. (e) Size and Shape of Tract. Where electricity is to be generated and transmitted from a central station, it is of very considerable importance that the points at which the power is to be used shall be compactly grouped. This requires, first, that the lands irrigated shall preferably be in one body and not in isolated or individual tracts ; secondly that the tract shall not be in the shape of a long, narrow strip. The reason for this, obviously, is that the more closely grouped the pumping-plants are, and the closer they are to the power-station, the less is going to be the cost of transmission lines and the less the transmission losses. (f) Ownership of Tract Central station pumping plants may be financed and installed in one of several ways. First, by the co-operation of a group of farmers who actually own the land. In this case the number of farmers must be sufficient, the land controlled sufficient in acreage, and suf- ficiently contiguous to make central station pumping evi- dently worth while, by which we mean that unless the THE CENTRAL STATION PUMPING PLANT 193 acreage combined is in excess of 1,000 acres in one body, a central station proposition is, at least, doubtful. Second, the scheme may be promoted by a company which en- deavors to furnish power to individual pumping-plant owners, charging upon the basis of the power used. In this case, the same caution must apply as has been found neces- sary in the promotion of gravity irrigation schemes, namely, that the only safe course for any irrigation development proposition is to own at least one-half of the land to which the water is to be applied, making profits sufficient to pay for the initial investment out of the rise in the value of the land to which water may be applied. The company which has no load other than a pumping load and which depends for its revenues solely upon the sale of power will soon find itself in the same financial condition as have many irriga- tion companies which put large sums of money into dams, reservoirs, and distribution systems, and merely sold water to irrigators at so much per acre-foot. Very few, if any, such enterprises paid interest on the capital invested. The only entirely satisfactory basis upon which a company may undertake a central station pumping project, is first to acquire at, say, not over $15 per acre, a suitable tract of dry land, and develop same ready for occupancy and irri- gation by prospective farmers to whom it is sold on a basis of double or treble the original cost under a contract providing for payment in a term of years, and providing further that the central power-station property shall eventually pass into the control of the landholders. Where, however, the pumping load is merely incidental or additional to the existing load, or that which it is proposed to develop, it is not so essential for the company to own a tract of land, but, nevertheless, it is always a wise precaution. (g) Possibilities of Co-operation, etc. Where a com- pany already established proposes to furnish power to the 194 PRACTICAL IRRIGATION AND PUMPING owners of lands for pumping purposes, it will be found essential in more or less extensive projects to draw up con- tracts in which the time of day during which power will be supplied for pumping is expressly agreed upon. It will also be essential to enlist the co-operation of these owners to reduce the peak load by dividing themselves into dis- tricts, each of which will use power when irrigation is needed, on certain days of the week, thus avoiding an overlapping of demands, which it might exceed the over- load capacity of the central plant to supply, and which would invariably result in dissatisfaction among the pumping-plant operators, and an unfortunate resentment against the company. (h) Fuels and Price. The question of fuel is an im- portant one in connection with central station pumping, and is a matter to be considered carefully in deciding upon the most suitable central station equipment. The various fuels have been discussed previously, as well as have the prime movers in which they would be used, hence it is un- necessary to more than mention here the connection of the fuel question with the more general study of the feasibility of a central station for pumping. Where water power may be developed within 50 miles of the irrigation project this may be by far the more economical power to use under conditions favoring easy construction of the electric generating plant. Pumping Season. The greatest danger to the success of an electrical pumping-plant project is in the brevity of the season during which power for pumps is required, and excessive peak loads. From 90 to 105 days of the year is the period during which pumps may be run for irrigation purposes and except under the most skilful management, even with the most helpful co-operation of the plant oper- ators, the load factor during these 90 to 105 days is likely THE CENTRAL STATION PUMPING PLANT 195 to be ioo per cent., that is, the demand for power may at times (when every one is irrigating) be equal to the com- bined horse-power of every plant in the project. If, for instance, there were 40 plants, each requiring 25 horse- power, the total load might be about 1,000 horse-power for most of the time, or, as shown by Diagram 24 on page 189, there would, even with diversified crops, be occa- sions for a week at a time when the entire series of plants would be in operation. Thus we have a condition which every experienced power engineer seeks, if possible, to avoid, namely, a plant of large station capacity, but of exceeding low yearly load factor. If the plant is used for irrigation alone it will stand idle for over two-thirds of the year. Common business prudence, therefore, suggests that some use be found for the plant during the remainder of the year. In well-settled localities there will likely be demand for power in lighting, in power purposes on the farms, and possibly in industrial enterprises of various kinds, and in street or interurban railway service. The lighting load is one which will not necessarily add to the peak load during the pumping season, but it must be noted that industrial power and railway requirements will add to the maximum capacity of the plant, since the re- quirement is likely to be constant throughout the year. The really desirable kind of load to develop is one which is not coincident with the pumping load, and this may be found in such service as supplying power for beet-sugar manufacture and canning factories. Indeed, it would seem a very desirable phase of the work of a beet-sugar factory to utilize its boiler plant during the period in which it would otherwise be idle in furnishing power for pumping, lighting, and general power purposes, if situated in a region where water-power is not available, and where lands desir- able for beet raising, but above existing canals, could yet 196 PRACTICAL IRRIGATION AND PUMPING be irrigated by pumps driven by power generated at the factory. The same observation holds in regard to the usual canning factory or any concern using steam and power in large quantities for only a brief period each year. In the Snake River Valley where hydro-electric power is used in large amounts for pumping from the river and canals, the power companies make a special rate for heating of residences and buildings by electricity, thus providing a source of income during the fall and winter. CHAPTER XV WINDMILLS The Field of the Windmill in Irrigation. Any discus- sion of pumping for irrigation would be incomplete without some reference to the use of windmills. Although wind- mills cannot possibly be regarded as feasible or economical of use for the areas, quantities of water, and heads con- templated by our previous discussion, the windmill has a very useful field and is a most important feature of certain classes of Western agriculture. Its most conspicuous service in recent years has been in the aid of dry-farming; indeed, it seriously may be doubted if without the small truck or garden patch and the domestic water supply made possible by the windmill, even a bare existence would have been the reward of those hardy pioneers who have shown up the possibilities of dry farming. It is an unquestioned fact, as proved by the experience of all who have tried it, that to make life endurable on the dry farm and to have some means of tiding over the unusually dry and unpro- ductive years, there must be some independent means of water supply sufficient to irrigate three, and preferably five or seven, acres of truck garden and alfalfa. The products of such an acreage may be the sole support of the dry farmers' family and stock during those off years which must be expected, however firm may be the prevailing belief as to a permanent change in climate. For the development of the water supply, some have installed power-pumping plants, but the majority favor a windmill plant, because of its simplicity and apparent small operating expense. Another important field of use- 197 198 PRACTICAL IRRIGATION AND PUMPING fulness for the windmill in irrigation is in connection with the development of small orchard tracts in the suburbs of many of the Western cities. Here suitable tracts may frequently be acquired on the mesas or benches adjacent to the town, which are not supplied by high-line canals, and yet where a water supply may be developed within a depth of 100 feet. Such tracts possess the evident advan- tage of convenience to market, and many men are trying the experiment of putting small tracts of 5 to 10 acres into orchards of peach, apple, pear trees, etc., irrigating by water pumped by windmill. Wind power is used almost exclusively for pumping the water supply of stock on the great ranges of the Southwest. This is a development of the last 10 or 15 years, and has enabled cattle to be grazed on vast areas formerly untouched because of distance from nearest water hole or surface-water supply. The discovery of water beneath portions of the great plains of Texas, and other sections long considered hopelessly dry, with the subsequent rapid dry-farming development, is the result of the successful attempt of cattlemen to erect windmills for stock watering, and thus extend the ranges. Altogether, it may be said that wind power offers a very interesting and possibly the ultimate solution of the problem of devel- oping the agricultural possibilities of the great plains country by pumping. Undoubtedly, as fuel increases in price, as it will, at an increasing rate as the years go on, manufacturers will be justified in introducing improve- ments in wind engines, which will increase their power and general suitability to the requirements of pumping, these improvements now being delayed, owing to the impossi- bility of constructing windmills of the power of small gaso- line engines, which will anywhere near approach the price of the latter. Kind of Mills. It is not our purpose to enter into any WINDMILLS 199 technical discussion of windmills, and it will suffice merely to mention some of the more important types and classes of mills as respects their structure. Size. Windmills are rated according to the diameter of the wind wheel, which in standard American machines may vary from 8 to 16 feet by intervals of 2 feet. Sizes larger than 1 6 feet are in use by railroad companies, and in some localities, as, for instance, in certain parts of California, very large home-made wheels upward of 25 feet in diameter are in use. The Dutch type of mill has a four-vane wheel of very large diameter, but there are only one or two examples of this type in this country, and it is of no com- mercial importance whatever. As to the material, the more common examples of modern windmills have galvanized pressed-steel vanes in the wheel, but there are also on the market very serviceable mills with wooden vanes. Under modern methods of construction, using light structural steel shapes and pressed-steel vanes, etc., there is no reason why a steel mill cannot be made as light as the wooden type, besides being stronger and more durable. Governing. In order to prevent the mill from being blown down in high winds, which would happen if the full sail area of the wheel were opposed to the wind at all times, various schemes of governing have been adopted by different makers. The most common scheme is to turn the wheel with its edge to the wind in high velocities by either a governing vane which functions when the air pressure on its area exceeds a predetermined limit as measured by an opposed spring; or in one make of mill the axis of the wheel is set eccentrically to the vertical axis of rotation, and the air pressure on the wheel area itself forces the whole head to rotate on the vertical axis against the opposition of the tail as determined by a spring, the tension of which can be adjusted. Another method is one in which the vanes are 200 PRACTICAL IRRIGATION AND PUMPING rotated about their long axis, so as to oppose merely their edges to high winds, while in still another scheme, the vanes are hinged at the periphery of the wheel, and in high winds the vanes move into a position with their axes more or less closely parallel with the direction of the wind. Obviously in the last two cases the plane of the wheel remains per- pendicular to the direction of the wind, and the mill is not usually provided with a tail or rudder. Another important difference between types of mills is in respect to the manner in which the power is transmitted. With some mills the power is transmitted to a pump rod directly by a pitman connected to a crank on the main shaft of the wind wheel. In another class, known as geared mills, the pump rod is attached to a pitman connected to a crank pin on a gear which meshes with a pinion on the wind-wheel shaft. The usual gear ratio is from 3 or 4 to i, thus making the number of pump double strokes from one-third to one-fourth the number of revolutions of the wind wheel. In another type of mill, the wind wheel is connected by a bevel gear and pinion to a vertical shaft, which by means of a second set of bevel gears at the base of the tower may transmit power to a horizontal shaft to which the pump- ing machinery may be attached. This is known as a power mill, and is useful for other purposes than pumping, though there is some loss of power through friction in the two sets of gearing necessary, which is avoided with the pump- rod types. The Selection of a Mill. It is useless to make many specific recommendations concerning the selection of a mill, since it is to be expected that the makers' statements in advertising literature very frequently will be taken without the necessary " grain of salt/' and just as often a mill will be bought upon its reputation or upon the en- WINDMILLS 2OI dorsement of a neighbor. We may, however, lay down these general rules, which should govern one in buying any mill. (1) POWER OF MILL. A mill is wanted that will deliver the maximum possible quantity of water pumped from a predetermined depth with such wind conditions as prevail in the given locality. (2) GOVERNING. A mill is wanted which will govern properly in high winds with a minimum of personal atten- tion. (3) STRENGTH AND DESIGN. A mill is wanted in which the parts are of ample strength, but not necessarily heavy and massive. (4) BEARINGS AND OILING DEVICES. A mill is wanted which has ample bearings properly fitted and provided with oiling devices of sufficient capacity and such type that the mill will not have to be oiled oftener than once a month. (5) PUMP CYLINDER. A pump cylinder is wanted of capacity that will load the mill properly and which is so designed and made of such material as will give a maximum length of service with the least wear of working barrel and valves. The points above cover the general specifications of importance to consider in the purchase of a mill as regards the size of wheel and the pump cylinder, together with the design and construction of the essential working parts. Many windmills in standard sizes and prices are now on the market which conform to the desired character of construc- tion indicated. When the intending purchaser has decided upon the size and type of mill he needs, his only care will be to see that he does not buy a low-priced mill in which the construction does not conform to what has come to be regarded as standard. Power of Mill. We have in this matter many factors 2O2 PRACTICAL IRRIGATION AND PUMPING involved which it has been the endeavor of numerous experimenters to discover and correlate. Among the pub- lished researches which concern the power of wind wheels, we may mention those of Wolf, Perry, E. C. Murphy, Hood, Fargusson, King, and Fuller, in this country, and of Chat- ter ton in India, and Ringelmann in France. The endeavor of these various experimenters has been to derive some relation between the power of a mill measured in horse- power, the diameter and type of wheel, and the velocity of the wind. Their work has very clearly demonstrated the extreme complexity of the problem, but some important facts have been established which have an immediate bearing upon the practical problem of so proportioning the load and size of a mill for a given locality that the maximum possible amount of work may be accomplished by the mill in a given time. The first matter to which attention should be given is the wind record for the specific locality, or at the nearest Weather Bureau station where records are kept. Such a record for Cheyenne, Wyoming, taken from Farmer's Bulletin 394 on Windmills, shows that v the average wind velocities during a period of five years for the months April-September, inclusive, were as follows: Hours per Month during which the Wind's Velocity per Hour was Miles' per Hour. . . Hours. o to 5 209.9 6-10 283 6 11-15 16-20 62 ^ 21-25 26-3O 31-35 36-40 40 and oven A similar record may be obtained from any Weather Bureau observation office, and should be obtained if a study similar to the following is intended for a specific locality. The following is a table similar to that above, and gives results of observations of wind velocities at WINDMILLS 203 Dodge City, Kansas, for seven years, as compiled by Mr. Murphy: Hours per Month during which the Wind's Velocity per Hour was Miles per hour. . Hours oto 5 140 6-10 1 08 11-15 157 16-20 IOQ 21-25 72 26-30 -14. 31 and over 22 The most conspicuous fact apparent from the above tables is the great preponderance of time during which comparatively low wind velocities prevail even in so notably windy a climate as that of Wyoming. It is evident, there- fore, that the mill will be best fitted for accomplishing use- ful work which will make use of the low wind velocities, for, if these can be determined, then some estimate is possible of the amount of water which a mill may pump in a season if the depth to water and the friction and hydrostatic head are known, as well as the size of the pump cylinder. The diagrams on page 204, taken from Farmer's Bul- letin 394, show certain characteristics of a 1 4-foot power mill. The first of these diagrams shows the most important fact to be noted in connection with a study of windmills, namely, that the horse-power which the windmill may deliver is dependent upon the load placed on the mill. Thus in Diagram 25, the curves marked A, B, C, etc., correspond to what would in effect follow a progressive increase in the size of pump cylinder, where the mill is used for pumping. Take, for example, curve A. This would cor- respond to a pump cylinder of small diameter, so that the number of pounds of water lifted per one stroke of pump is small. It will be noted that a mill with this load would start in a wind of about 4 miles per hour, and, as the wind velocity increased, the horse-power output would 204 PRACTICAL IRRIGATION AND PUMPING 5.5 14 FT.STEEL MILL POWER DEVELOPED AT DIFFERENT LOADINGS AND WIND VELOCITIES 6 8 10 12 14 16 18 20 22 Wind Velocity -Miles per Hour 28 30 DIAGRAM 25 14 FT.STEEL MILL 60 l tc ' s CO .. l / 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 Wind Velocity Miles per Hour DIAGRAM 26 WINDMILLS 205 increase very slowly. As will be seen from curve A in Diagram 26, the speed of the wheel would increase very rapidly at this load. Refer now to the curve marked H, in both diagrams. This would correspond to a very much larger pump cylinder and a greater weight of water lifted each pump stroke. As will be seen in Diagram 26, the mill would not start with this load until the wind had attained a velocity of nearly 20 miles per hour, but a slight increase in wind velocity would be accompanied by a very rapid increase in delivered horse-power, so that at a veloc- ity of 26 miles per hour the mill thus heavily loaded would deliver nearly seven times the horse-power of the same mill loaded as in the first case and at the same wind velocity. The speed of the mill would also be 40 revolutions less per minute for the heavier load. These diagrams, therefore, illustrate a very important fact or principle, viz., that the power output of a mill in a given wind velocity varies with the load upon it and that at the higher velocities of wind the heavily loaded mill gives a greater horse-power at a more desirable speed for pumping purposes. On the other hand, the heavily loaded mill will deliver no power what- ever in moderate winds, and requires a comparatively high wind in which to start. This fact has been appreciated for some time and various devices have been invented which would automatically vary the load of a mill according to the wind velocity. A perfectly acting device of this sort would evidently enable the mill to deliver a maximum possible power output in the course of a season, the mill always operating at the most efficient load no matter what the wind velocity. Unfortunately, such an ideal device has not been made, and it involves mechanical difficulties not likely to be surmounted for some time to come. The best that can be done under the circumstances, therefore, in view of this peculiar characteristic of the wind 206 PRACTICAL IRRIGATION AND PUMPING wheel, is to select a pump of such size that for the partic- ular wind conditions of any given locality the mill may be expected to deliver the greatest possible quantity of water in a season. The selection of this pump size involves a rather extended series of calculations and comparisons, which probably may best be illustrated by a concrete example. The Problem of Determining Best Diameter of Pump Cylinders, Concrete Case. Let it be supposed that it is desired to know what pump should be used with a 1 2-foot mill of the same type used in the experiments from which the above diagrams were deduced; where it is desired to pump water from a depth of 40 feet and discharge it into a reservoir 6 feet deep. Assuming a draw-down of 5 feet and a friction head of 3 feet, the total head to be pumped against would be 40 + 6 + 5 + 3 = 54 feet. The following work, upon which will be based the de- termination of the most effective pump size, will rest upon certain published results of Professor Murphy on a 12 -foot mill, to which he attached a friction brake in such a way that the performance of the wheel at different brake loads and at different wind velocities could be accurately de- termined. These results, which give the actual power of the wheel, can be used for a comparison of direct stroke and geared mills as well. The diagrams which follow are de- rived from the results and diagrams given by the authority mentioned. Diagram 27 gives the relation between wind velocity and the revolutions per minute of the wind wheel for different loads as expressed by the number of foot-pounds of work accomplished per one revolution of the wind wheel. The Size of Pump for Direct-Acting Mill. Windmills are usually arranged for different lengths of pump stroke, and for the present example for a 1 2-foot mill this will be WINDMILLS 207 taken at 10 inches. The sizes of pump cylinder from which a selection may be made are assumed as of diameters of 4, 5, 6, and 7 inches, for the direct-stroke mill. A slip of 15 per cent, will be assumed for the pump, which value would represent one in very good condition, and a mechan- ical efficiency of 80 per cent, for the mill that is, 80 per 12 FT. STEEL '"'POWER" WINDMILL 8 10 12 14 16 18 20 22 24 Wind Velocity . Miles per Hour DIAGRAM 27 cent, of the possible work of the wheel will be assumed as available at the pump. Using the value for slip just men- tioned, the following table shows the capacity of different diameters of pump cylinder in acre-feet per hour at differ- ent numbers of strokes per minute: 208 PRACTICAL IRRIGATION AND PUMPING TABLE XI CAPACITY OF DIFFERENT SIZES OF PUMP CYLINDERS, WITH 15 PER CENT. SLIP ACRE-FEET PER HOUR Pump Strokes per Minute 2O 30 4 50 60 70 80 7x10 inches .0052 .0038 .O027 .OOI7 .0078 .0057 .0039 .0025 .0104 .0076 .0054 .0034 .0130 .0095 .0065 .0042 .0156 .0114 .008l .0051 .0182 0133 .0091 .0059 .0208 .0152 .OIO8 .0068 6 x 10 inches 5x10 inches 4x10 inches It is, of course, obvious that the higher speeds, that is, those much above 40 strokes per minute, are impracticable, owing to severe inertia effects. The whole range of speed indicated will, however, be used for illustration. Using the sizes of cylinders above given, the following table gives the foot-pounds developed per revolution of wind wheel for a 54-foot total lift with 15 per cent, slip and 80 per cent, mechanical efficiency. TABLE XII FOOT-POUNDS OF WORK PER ONE REVOLUTION OF WlND WHEEL Size of Pump 4 x 10 ins. 5 x 10 ins. 6 x 10 ins. 7x10 ins. 8 x 10 ins. Ft.-lbs. per rev . 260 430 58o 800 1,050 Using the values of foot-pounds per revolution of wind wheel as an argument, we may now by use of Diagram 27 determine the wind velocity which is necessary to give the required speed of the wind wheel. These values are shown in the following table: WINDMILLS 209 TABLE XIII TWELVE-FOOT DIRECT-STROKE MILL WIND VELOCITIES IN MILES PER HOUR WHICH WILL TURN WIND WHEEL AT VARIOUS SPEEDS WHEN USING VARIOUS SIZES OF PUMP CYLINDERS Size of Pump Revolutions of Wind Wheel per Minute 20 30 40 50 60 70 80 7 x 10 inches 17 14 II-5 9 19 15-5 13-5 ii 20 I? 15 12.6 22 18.5 16.5 14-5 24 20.5 i8-5 17 6 x 10 inches 23-5 20 20 25 5 x 10 inches 4x10 inches Using the values of wind velocity as above given, we now refer to the weather records and determine the number of hours per season during which these velocities are found to occur. These are shown in the following table: TABLE XIV SHOWING NUMBER OF HOURS PER SEASON DURING WHICH WIND WHEEL WILL ROTATE AT GIVEN SPEED WITH GIVEN LOAD TWELVE-FOOT DIRECT-STROKE MILL Revolutions of Wind Wheel per Minute Size of Pump 20 30 40 50 60 70 80 7 x 10 inches ..... 653 653 653 4^0 4^0 6 x 10 inches 044 044- 653 6S3 653 4^0 c x 10 inches 04.4. Q44 044 6er Season Size of Pump 1,187 944 653 430 Totals 7x10 inches 5. 1 6.1 112 6 x 10 inches 4-5 6.2 5-7 l6.4 5 x 10 inches v 3-7 4.7 7.0 12 T> 4x10 inches I T>.2 7. c: 2.Q II 5 The fact now becomes apparent, as seen from an inspec- tion of the total quantities pumped in a season, given in the last column of the above table, that the 6- by lo-inch pump is much the most efficient size to use, since by it, under the local conditions of wind movement and head pumped against, the largest quantity of water is delivered. As a matter of fact, this would have to be modified in the practical case, since if a pump of this stroke and diameter were used, the mill would have to be arranged to cut out at lower wind velocities to avoid the danger of operation at high pump speeds for considerable periods. If this size of pump were used and it were proposed to prevent the pump from operating over 50 strokes per minute, then, as shown by the above table of wind velocities versus loads and speeds, Table XIII, the governor would have to be adjusted to throw the mill out of action at wind velocities of over WINDMILLS 211 19 miles per hour. This limitation on rotative speed would cause a loss of 5.7 acre-feet per season with the 6- by zo-inch pump, 4.4 acre-feet for the 5- by lo-inch, and 6.4 acre-feet for the 4- by lo-inch. The relative quantities pumped per season would then be shown by the following table: TABLE XVI TWELVE-FOOT DIRECT-STROKE MILL. 54-FEET TOTAL HEAD Size of pump. . . . Acre-ft. per season 7 x 10 ins. II. 2 6 x 10 ins. 10.7 5 x 10 ins. 7-9 4x10 ins. 5-1 Upon this basis it appears that there is but little differ- ence between the 7- by lo-inch and the 6- by lo-inch pumps. Doubtless an investigation similar to the preceding for pumps of different strokes would show some combination of stroke and diameter, which, without exceeding the safe pump speed, would give a greater seasonal capacity than that just found. One fact shown by the above tables con- firms that already well known by pump and windmill manufacturers and those who have investigated the per- formance of windmills, namely, that the direct-stroke mill is best adapted to localities where the average wind ve- locity is high and of long duration. Size of Pump for Geared Mills. The same methods and the same data will be taken for this example as in the preceding investigation for the direct-stroke mill. It is understood that in the geared type of mill the pump is driven by a pitman and pump rod connected to a crank shaft which rotates at less speed than the main shaft to which the wind wheel is attached. The speed reduction varies with different makers, but in the present case the speed of the pump shaft will be taken at one-third the 212 PRACTICAL IRRIGATION AND PUMPING speed of the main shaft. Then the number of pump double strokes will be one-third the revolutions of the wind wheel. The introduction of the gearing thus necessary causes a friction loss both in the gearing and in the journals so that 60 per cent, of the power possible from the wind wheel will be assumed in this case as being available at the pump. For this investigation a greater number of pump diameters will be considered, but of the same stroke with one exception. The table on page '2 13 gives the sizes of pumps which will be tried, together with the foot-pounds of work corresponding to one revolution of the wind wheel, for a total pumping head of 54 feet. Averaging the acre-feet discharged over the range of wind velocities which occur for the same number of hours per season and adding together the discharges for the several ranges, we obtain the total for the season. The 8" x 10" pump is found upon this basis to give the maximum, for wind velocities below 25 miles per hour. By referring to the discussion of the direct-stroke mill, it will be seen that for the weather conditions of the ex- ample the difference between the performance of the direct and geared mills in this instance is not striking. Placing the results side by side, we have. Direct Stroke. Geared. Size of cylinder 7 x 10 ins. 8 x 10 ins. Water-pumped acre-feet per season 11.2 10.5 Both mills would be adjusted to be thrown out of action at a wind velocity of about 25 miles. While there is a slight advantage, as above shown, for the direct-stroke mill, it is to be understood that this is due to the relatively high average wind velocity. For localities with lower average wind velocities the advantage would be quite WINDMILLS 213 CO G 04 vg j X u Ea o CO G Q X ^ o 1 a.' c/) O ^0 1 X 00 w 1 CO Q O %$ 1 X -1 g ft ^ ^ o j > *s ffi CO 2 1 w^ ^ g X PM fl\ w ^ VO s% ? 1 " 1 2 g X fc o 1 o 1 | G w '5 ^ c/5 > fa o a o 1 ! 1 U o r ^ a; *T .2 04gH * 2 SgS IM s S ^ o o so JO 3ZJS 10 O O 10 O ON N -i W 1010 t-i NH r}- 1-4 P< : O U. O >- O HHTi- Hi-3-n- oo t^ . oo oo 10 M . 00 "I ' 00 CN 00 sui oi x S 'sui oi x 9 'sut oi x i WINDMILLS 215 o ^t- co rf ON CO 10 CO O VO i-i lO rf M CO O vO -^- 1-1 O CO CO O O ON VO CO Tf ** <* CO O I s * co vO O ^O O n i-i 10 ^ n 10 ci \O . vQ CO CO TJ- corj-io vOco^J- ONCOO TriO i-iiot^ i-iioco >O rh lO 00 CO ON f-c Tj- Tt M IO O ON vO f^