Home-immediately access 800+ free online publications. Download CD3WD (680 Megabytes) and distribute it to the 3rd World. CD3WD is a 3rd World Development private-sector initiative, mastered by Software Developer Alex Weir and hosted by GNUveau_Networks (From globally distributed organizations, to supercomputers, to a small home server, if it's Linux, we know it.)ar.cn.de.en.es.fr.id.it.ph.po.ru.sw

CLOSE THIS BOOKDesign and Operation of Smallholder Irrigation in South Asia (WB, 1995, 134 p.)
Chapter 14 - Groundwater development
VIEW THE DOCUMENTSmall, individually-owned, suction-mode wells
VIEW THE DOCUMENTIndividually owned and group owned force-mode wells
VIEW THE DOCUMENTLarge capacity public tubewells
VIEW THE DOCUMENTTechnical problems in design and construction of medium and large tubewells
VIEW THE DOCUMENTWater distribution from medium tubewells
VIEW THE DOCUMENTFunctions of the tubewell operator
VIEW THE DOCUMENTPower supply problems
VIEW THE DOCUMENTComparison of medium and large wells

Design and Operation of Smallholder Irrigation in South Asia (WB, 1995, 134 p.)

Chapter 14 - Groundwater development


Where alluvial aquifers are available, such as in the Gangetic and Indus basins, groundwater can make a major contribution to irrigated agricultural production. In areas underlain by hard rock the potential for irrigation by groundwater is much less, but is still an important factor. A key question that is much debated in such development is the role of the small farmer-owned tubewell versus group-owned medium sized wells and larger public wells.

In areas where ground conditions permit its use, the small well can be very attractive, both to the small farmer proprietor, as it provides a supply of water virtually on demand, and to the government, providing development with little drain on national financial resources. There are, however, physical situations in which the small well cannot operate, due to technical reasons that will be discussed below.

More relevant to the present discussion is the situation in which small wells have been operating reasonably satisfactorily for some time, but continued unregulated installation is drawing the dryseason watertable down to the physical limit for such "suction-mode" wells. Dry-season supply at the individual well becomes unreliable, and groundwater extraction becomes competitive. Some well owners construct pits at the well-head, lowering the pump 2 or 3 m below ground surface, thereby increasing the depth to watertable at which they can operate. Larger farmers or farmer groups in some areas install "force-mode" pumps which do not have a depth-towater/able limitation, further aggravating the problem for the small well owner who cannot afford such a pump.

It can be argued that such a situation could be prevented by appropriate regulation, but in the South Asian region efforts to control groundwater development, largely confined to withholding of credit for new installations, have been largely unsuccessful. On the other hand, it can also be argued that full development of the resources of the aquifer, including utilization of its storage capacity by deeper draw-down in dry years, can only be accomplished with the use of force mode pumps.

A pressing question is the fate of the small farmer where installation of force-mode pumps by larger farmers is occurring on a significant scale and is leaving the small farmer with a dry well in the critical irrigation season. Over time a water market may develop, in which larger farmers with force-mode pumps sell water to neighboring smaller farmers (Shah 1989). The process may, however, take a long time, leaving many small farmers in distress. One possible course is to form groups of small farmers to share in the ownership and operation of force-mode wells. As the construction of such a well and installation of the pumping equipment is a considerably more sophisticated task than in the case of a simple suction-mode well, at a minimum technical assistance by a government agency is likely to be required. The well could be owned and operated by the group or owned and maintained by government agency, with the group being responsible for its operation.

However, the involvement of a government agency in construction of such wells, in an area already largely developed or overdeveloped by small wells, can introduce problems of equity (Toulmin 1987). There is seldom sufficient recharge to effectively irrigate more than one-third to one-half of the surface area Where almost every farm has a small well, or access to one, a portion only of each farm is irrigated. With such density of wells the cost of distribution of water is not a factor. The position is substantially different with medium capacity force-mode wells which are spaced farther apart, and in which distribution is a significant item of cost. The logical economic solution is to supply water to an area in the vicinity of the wells, and to leave the intervening area unirrigated. However, farmers in the latter area who previously obtained limited supply from their small wells would be left without water in the dry season due to draw-down of the watertable by the adjacent medium wells.

In the broad perspective of effective utilization of the aquifer, the installation of force-mode medium wells may be desirable, and it may eventually occur through the initiative of the large cultivators. However, it can result in hardship to those left outside the service areas of the medium wells. This is a problem with no entirely satisfactory solution. Installation of medium wells with relatively large service area and correspondingly low irrigation intensity reduces the proportion of the area without groundwater supply, although at considerable cost in distribution system. This course is also likely to result in some discontent among cultivators served by a well, due to the limited supply of water to the individual. However, where groundwater is being utilized in association with largely rainfed monsoonal cropping, and where cultivators are paying the full cost of pumping, it is remarkable how economically such water is used. Perennial crops that are heavy water users, such as sugarcane, are ruled out in these circumstances, except for the large farmer who has his own well.

The remainder of this section is devoted to a brief account of the technology and application of small suction-mode wells, group operated medium-capacity force-mode wells, and larger forcemode wells that are either operated by public agency or by farmer groups.

Small, individually-owned, suction-mode wells

These wells are characterized by a pump located at ground level or in a shallow pit. Water is lifted to the surface by suction, and hence the lift is limited to the barometric height. In fact, for practical purposes it is limited to six or seven meters. The pump is a single-stage centrifugal, usually direct coupled to a small diesel or kerosine engine of 4 or 5 hp or to an electric motor, if electric power is available. The well itself (the term "well" is used interchangeably for the tubewell or for the complete installation including the pump) generally consists of a steel pipe about 10 cm in diameter with perforated lower end sunken into the ground a distance of typically 20 or 30 m. Slightly more sophisticated wells are constructed with a filter, often of coconut fiber, surrounding the perforated portion of the pipe.

Well-sinking is generally carried out by a small local contractor using a variety of indigenous methods. The yield is commonly 10 to 15 liters/sec (1/3 to 1/2 ft³/sec). Where the pump is driven by a diesel or kerosine engine, it may be permanently connected to a well, or alternatively it may be moved from one well to another (a "taxi" pump) so that four or five wells are served by the same pump. The life-span of a well is only a few years, the well being reconstructed when the need is indicated by diminished yield. In hard-rock areas the equivalent of the small tubewell is the dug-well, several meters in diameter and usually some 10 or 15 m in depth. The dug-well may be supplemented by holes drilled from the bottom of the well to increase yield. The same type of suction-mode pump is employed as described above.

A small tubewell with yield of 10 liters/sec. could supply an area of 5 or 6 ha, subject to available recharge. However, the area served is generally one or two hectares only, part of which may be by sale of water to neighbors.

The primary feature of small section-mode wells which has made them very popular in the past is their low cost, which permits a cultivator with a modest sized holding to have his independent supply and possibly to sell a limited amount of surplus water. Purchase of water has also brought the benefit of such wells within the range of the smaller cultivator who could not afford his own well, or when his holding is too fractionated to make ownership practical.

The applicability of such wells is limited by technical and socio-economic factors. Technical factors include a dry-season watertable within the suction limit, a suitable aquifer at depth reachable by low-cost well-sinking methods, and sufficient recharge. The socio-economic factors refer to the size of the holding and particularly to the degree of fractionation. A cultivator with a 1 ha holding made up of five parcels in separate locations, each of less then one-quarter hectare, could not be a candidate for a well, unless based on sale of water. The notable slow-down of the rate of installation of small wells in the eastern Gangetic basin in recent years has probably been due to the fact that most cultivators with sufficient size of holdings already have wells. The remaining cultivators have holdings too small or too fractionated to justify investment in a well.

With regard to recharge, in a monsoonal region the primary source is seepage from rainfall. Seepage from irrigated fields and from canals may also contribute. As indicated earlier, the amount of annual recharge is generally sufficient to provide irrigation to about one-third of the area, but varies considerably with amount of rainfall and soil infiltration rate.

The main constraint on the use of small suction-mode wells is undoubtedly depth to dry-season watertable and lowering of the watertable either by over development or by introduction of forcemode wells.

Individually owned and group owned force-mode wells

In the force-mode well, the pump is located at the lower end of the tubewell rather than at ground surface, and water flows up the well under pressure, rather than being drawn up by suction. Thus, there is no physical limit to the height of lift, or the depth of watertable, against which the pump can operate. The pump itself is the generally the same as the surface-mounted centrifugal, but due to the space constraints at the bottom of the tubewell, the pump is smaller in diameter. It is usually made up of a number of units operating in series, rather than a single impeller, to give the required pressure. The engine may be at ground surface and connected to the pump by a long hollow shaft extending down the length of the tubewell, with guide bearings at intervals. In the case of electrically operated pumps an option is to have a submersible electric motor directly coupled to the pump at the bottom of the well, with power cable leading to the surface.

Other types of force-mode pumps include the hydraulic jet pump, the helical rotor pump, the hydraulically driven turbine pump, compressed air "pumping" and positive displacement units including the reciprocating pump. The latter is used extensively, usually hand-operated, for domestic water supply, but the yield is too small to be used for irrigation other than for small household plots. Effectively, the options available are long-shaft centrifugal pumps, either diesel or electrically driven or electrically driven submersible-motor pumps. Where electric power is available, the submersible-motor pump is attractive, particularly for smaller-capacity units. It avoids the long shaft and supporting bearings, with attendant problems of installation and maintenance.

A key question regarding the application of submersibles to the installation of force-mode wells by individuals or small groups is their minimum practical capacity. Recent advances in the use of high-technology plastics and composites, and in manufacturing methods, have lowered both minimum capacity and cost. Facilities for servicing such units are also becoming available in irrigation areas. As a consequence, small submersible force-mode pumps are coming increasingly within the reach of the large individual cultivator. For the small cultivator, group operation of such units, or purchase of water, are likely to remain the available option.

It is noted that with the wells under discussion, generally capable of serving 10 to 15 ha or more, there is not only the well to be considered but also the distribution system. With a small well serving 1 or 2 ha distribution of water is not an important factor, as the length of run from well to farthest plot is unlikely to be more than 150 m, which is well within the range of an unlined channel. However, when the area served is 15 to 20 ha the length of run is likely to be 400 m or more, and seepage losses in distribution could be high. In this case the economic installation may be a well plus a partially lined (or buried pipe) distribution system.

There are advantages and disadvantages in group ownership or group operation of a well. Supply from a well owned by a group is presumably more secure than purchase of water, which may not always be available from a supplier. Further, in an area in which tubewell irrigation is newly coming into use, there may not yet be a developed water market, and group ownership may be the only option. However, group ownership and operation can have its problems, including allocation of responsibility for equipment maintenance and repairs, procurement of fuel and lubricants, payment for energy, and conflict resolution over for scheduling the use of the well. Experience in some areas indicates that once a water market has developed groups tend to break up, some members preferring to avoid the responsibilities of ownership by buying water from others (Shah 1987, Toulmin 1987).

Experience with joint ownership of wells, other than small suction-mode wells, has in fact been decidedly mixed. Part of the reason is the size of the area served and the number of cultivators involved. While a group-owned small well supplying 2 or 3 ha may typically serve five or six cultivators, often related to each other, a "medium" force-mode well is capable of serving a considerably larger group of forty or fifty cultivators or more, and indeed it needs to, if costs to the individual members are to be minimized. Obtaining agreement of all members of such the group to undertake the joint enterprise and accept financial responsibility is not easy. Subsequent collection of funds for day to day running expenses and for repairs in the event of major breakdown or replacement of equipment is also likely to be a problem for the group.

First preference of small cultivators is usually for government ownership and operation, usually implying subsidized rates and no capital obligation. This is obviously not a desirable solution from government viewpoint, as it requires not only large initial outlay but also continued funding of a substantial proportion of energy costs. However, in negotiating a compromise solution, the cultivator has the option of simply continuing with rainfed cultivation. With this no cost alternative available to the cultivator, the less enterprising cultivators are unlikely to be willing to assume capital obligations or the responsibility for operation and maintenance. The more enterprising might be willing to do so, but group formation is a voluntary process and requires consensus among those in the prospective service area.

In an effort to make group operation more attractive to cultivators, a number of compromises have been suggested, notably making available to the cultivator group the well and equipment at no cost and giving responsibility for running it to the cultivators, including meeting the cost of fuel or power. The reason for this course is simply expediency, rather than economic logic, there being no other way to obtain cultivator participation. However, in support of the cultivator view, it must be admitted that the government seldom attempts to recover any of the capital cost of providing canal irrigation, which is usually considerably greater per hectare served than the capital cost of tubewell irrigation.

There is, however, another factor which may deter cultivators from owning a medium well system, even if it is provided to them at no capital cost. This is the threat of major maintenance. Where a force-mode pump is required, which is the situation being discussed, both the tubewell and the pump are more sophisticated than in the case of the suction-mode well traditionally employed by small cultivators. The pump requires specialist attention in maintenance or repair. The tubewell itself may be in a difficult fine-grained aquifer with possibility of sand-pumping and eventual collapse of the formation around the screen and well failure. This situation requires sinking a new well. If cultivators have witnessed the problems of government agency in such circumstances, they are likely to prefer to leave ownership and major maintenance of well and pump in the hands of government, which has substantially greater financial and better technical resources.

The functions which government agencies would most wish to transfer to the cultivator group are in fact neither capital repayment nor major maintenance. They are the day-today operation of the well including scheduling of deliveries to individual members of the group, both of which require the services of a tubewell operator, the collection of water charges and payment of all fuel or energy costs. These are functions which the cultivator group can probably perform better than government agency. Energy costs and the salary of the tubewell operator (which can be as much as the cost of energy) are a continuing drain on government financial resources if wells are fully operated by the government, and water charges are inadequate to cover running costs (which is usually the case).

With regard to the process of group formation, it is noted that negotiation of a firm commitment to take water prior to construction of a well, although desirable, is not always possible, as the yield from a particular well and the size of the area which can be served by it may not be known until well construction is completed and yield tests carried out. This is commonly the case where the character of the aquifer is highly variable. This situation of constructing a well before final commitment by prospective users puts the constructing agency at some disadvantage in negotiation, and underlines the need for the cost sharing arrangement to be reasonably attractive to the potential group members. The process of group formation including consultation with prospective participants at all stages in the investigation and construction of a well is a key element in successful group operation of such wells.

To summarize, group ownership of a small suction-mode well serving a small group of neighbors, does not present a problem. However, where depth to watertable or the character of the aquifer preclude the use of such pumps and a force-mode pump has to be employed, generally with larger capacity well and considerably larger service area, the capital cost and the risk of incurring substantial maintenance costs with these more sophisticated pumps and wells is a considerable deterrent to group ownership. Government ownership and responsibility for major maintenance, combined with group responsibility for operation and all running costs, appears to be a practical compromise.

Large capacity public tubewells

The term, as used here, includes wells owned and operated by government agency, or owned and maintained by such agency but operated by cultivator groups. They are referred to as "direct irrigation" wells, the water being applied directly to land in the vicinity of the well. In a second category of public wells the outflow discharges into a canal which also carries surface water, the combined flow then being distributed through the surface system. The latter are referred to as "augmentation" wells. They are discussed later, under the heading of conjunctive use.

Much of the above discussion of medium-capacity wells also applies to large-capacity wells. There are, however, two additional factors to be considered in the operation of a large well. The number of cultivators in the well group is much greater and the output of the well is sufficiently large that it requires division, two or more cultivators irrigating at the same time. Such wells commonly have capacity in the range of 30 to 100 liters/sec (approximately 1 to 3 ft³/sec). They are also relatively deep, usually 100 to 200 m.

Technical problems in design and construction of medium and large tubewells

The remainder of this discussion is devoted to selected design and operational features of medium and large tubewell installations in smallholder areas. Particular attention is paid to commonly recurring problems. The stages of major well construction are drilling, installation of casing, screen, and granular filter if any, well development and testing, and finally pump installation. There is an extensive literature on the theory and practice of tubewell construction (Driscoll, 1986).

The key to tubewell performance is the design of the screen and filter system in relation to the character of the surrounding aquifer. Where the latter is relatively coarse-grained (sandy gravel) no separate filter may be needed. In the process of well development, the fine material is washed out of the formation in the immediate vicinity of the well screen, and a stable radially-graded natural filter is formed. The well is then referred to as "self-screening".

The situation is more complex where the aquifer is of finer material such as coarse to fine sands, commonly inter-leaved with silts. Even with the minimum practical width of screen opening (about one millimeter) an excessive amount of material would be washed out in the course of pumping, with subsequent collapse of the formation in the vicinity of the screen. Under these circumstances, a granular filter ("gravel-pack") needs to be placed between the screen and the formation (commonly 5 to 10 cm in radial thickness). The choice of the gradation or grain size of the granular filter is a function of the width of the screen slots and the gradation of the aquifer formation.

The problem encountered with such a filter pack installation is in washing out the drilling mud from the filter and the adjacent formation. The use of drilling mud, a mixture of bentonite clay and silt-sized cuttings, is essential to the direct rotary drilling process. This dense fluid supports the wall of the hole being drilled and also removes cuttings. It is continuously circulated during drilling, being pumped down through the hollow drill-stem and rising through the annular space between the stem and the wall of the welt Part of the mud seeps into the formation, sealing it against excessive mud loss. On completion of drilling, the mud remains in place, supporting the well, while the granular filter is poured into the annular space between well-screen and wall. The mud is then displaced from the well by water, and the process of washing the residue of mud from the filter and adjacent formation is undertaken.

This is not an easy task if the filter and the formation are fine-grained. Various techniques of surging, over-pumping, water-jetting, use of compressed air to create turbulence, and addition of detergents, have been worked out. Faced with the same problem the oil-well industry has pioneered the use of big-degradable material in mud formulation. The mud breaks down to a low-viscosity fluid after a period of time, and is more readily washed out than bentonite mud. However, such materials are sensitive to temperature and other factors and are not generally suited to tubewell construction in the region under discussion. Water-jetting involves the use of small high-pressure nozzles positioned inside the screen and discharging out through the screenslots into the filter and hopefully into the formation, producing intensely turbulent washing action. Used in conjunction with "wire-wound" screens with continuous spiral openings, this can be effective. However, the jet can lose most of its velocity in passing through the filter, before it encounters the formation. Cleaning the mud from the filter and formation and associated "well development" (washing out a proportion of the fines in the formation) can be a difficult and time consuming procedure, requiring long experience in this field.

Another drilling method which may be employed for holes of 18 inches or greater, is referred to as "reverse rotary., where the drilling fluid flows in the opposite direction, i.e. down the annular space and up the drill rod, instead of being pumped down the hollow drill rod and flowing up the annular space between rod and well, as in "direct rotary. drilling. Generally, reverse rotary drilling does not require the addition of bentonite to the drilling fluid, making the subsequent washing of the granular filter and adjacent formation simpler than with direct rotary. However, the use of the direct rotary method is restricted by the type of formation, depth to watertable (must be less than ten feet), amount of make-up water available, and other factors. Further, the minimum hole size is greater than desired for most irrigation applications. Currently, direct rotary drilling is the most commonly employed method for medium and large capacity irrigation wells in alluvial formations.

Key factors remain, including the design of the screen (wire-wound stainless steel appears to be the most effective and most durable solution), the grain size or gradation of the filter and the method of placement, and the method of washing out drilling mud and well development. The result of using excessive size of screen openings and grain size of filter, in relation to the formation, can be excessive "sand pumping" and early well failure. If sand pumping does not stabilize at a satisfactorily low level after well development, the only course is to reduce the rate of pumping (and the exit velocity from the formation) until the rate of production of sand falls to an acceptable level. The penalty for inadequate washing of drilling mud from the filter and the adjacent formation is low well yield and high pumping head (i.e low specific yield), reflecting in pumping energy required and reduced size of service areas.

The above discussion underlines the needs for care in the investigation, design, and construction of medium and large tubewells, particularly in areas of difficult formation characteristics. A well designed and constructed tubewell installation can have a life of fifteen to twenty years or more. With less appropriate design or less satisfactory construction, safe yield may be substantially reduced, and the life of the well may be relatively short.

Water distribution from medium tubewells

With tubewells of the capacity under discussion (up to 30 liters/sec) the size of service area is likely to be in the range from 15 to 30 ha, depending upon well capacity and design irrigation intensity. In a smallholder environment, with fractionated holdings, the number of individual parcels to be served may be as much as 100. Distribution of water in this situation raises a number of questions, both technical and operational.

Subject to availability of power, supply at the well-head is available virtually on demand, making the cultivation of a wide variety of crops including high value specialty crops possible, provided that a similar degree of reliability can be provided at the field boundary. The situation calls for an efficient distribution system, with close farm delivery.

Buried-pipe distribution is the desirable solution, particularly as sufficient head can be provided at the well to permit use of pipe of relatively modest size. The closed-loop type of system makes further economy in pipe size and cost possible. With the appropriate layout, the closed-loop type of system permits outlets for every 2 or 3 ha. At additional cost of pipe and outlet valves, outlets can be provided at every hectare if necessary. From the valved outlet to the field, conveyance is by short earthen channel. Technically, this is a very sophisticated system compared with supply from a canal via unlined tertiaries serving 30 or 40 ha. The question remains how to make the best operational use of it.

The ideal situation would be for any outlet to be able to take water at any time, in the same manner as from a household tap, and the pump would respond accordingly.

However, there are technical and operational limitations. The well (in the case considered) supplies at a fixed discharge of 30 liters/sec rate. This rate is, in fact, a desirable size of stream for efficient conveyance in unlined channel from the valved outlet to the field, where it may be further divided between plots or furrows. Therefore, the first operational restriction is that one outlet at a time should be in use, and one farmer at a time should use that supply (possibly two, by mutual arrangement). This ensures adequate stream size for efficient distribution, and also permits recording the amount of water used by the farmer, as it is simply the product of the well output and the time of irrigation. Opening many valves simultaneously should be avoided, except in the special circumstance of the whole area being under paddy, when a small flow from each valve may be desirable.

Operating with one outlet valve open at a time involves scheduling both supply to each of the fifteen valves, and irrigation by the several individual farmers served by each valve. The various procedures adopted in such operation are aimed at simplifying the scheduling process. One arrangement which has considerable merit is to divide the service area of the well into seven units each of 4 or 5 ha (six units if the well is to be shut down for one day each week), each unit being irrigated on one fixed day of the week. Only the outlets serving a particular unit are in operation (one at a time) on that particular day, and only the farmers within that unit are concerned in scheduling operations on that particular day.

As scheduling becomes simpler the smaller the number of cultivators involved, a variation which could be considered where holdings (specifically parcels) are very small is to divide the well command into 14 rather than 7 units, the day being divided operationally between morning and afternoon. Each unit would then be 2 to 3 ha, certainly small enough to be self-managing as far as internal scheduling is concerned.

While conceptually it would be simplest for each unit to have its own outlet valve, technically this is not essential. A particular outlet may serve different units on different days, although this involves having a branch of a distribution channel traverse the unit in which the outlet is located, enroute to a neighboring unit. This situation could invite misuse, and may be partly responsible for some of the problems which have been encountered in distribution from tubewells. Operationally one outlet value per unit would be preferable.

The above discussion refers to a medium well delivering 30 liters/second to 30 hectares, one outlet taking the whole of the well delivery at a time. A larger capacity well, for instance delivering 60 liters/sec, poses an additional operational problem, as the well discharge is too great to be handled effectively by an individual cultivator. Furthermore, distributing that flow through a single system would require large pipe size. In effect, the delivery system has to be divided, into two sub-systems each of 30 liters/sec and each seeing 30 ha. Arrangements within each subsystem with respect to division into fixed day of the week units are the same as discussed for the 30 liters/sec medium well. However, means have to be provided for dividing the output of the well into two equal parts. This cannot be done simply by bifurcating the delivery line from the well into two branches, with each seeing a 30 ha sub-system, as the frictional resistance to flow in the two branches may not be equal. It depends upon the respective distances from the well of the outlets in operation at the time. This problem can be overcome by the use of an elevated tank with two identical outflow weirs each discharging into a stand-pipe supplying one of the sub-systems. Division of flow is then made at the well, and is independent of the location (or number) of the outlets upon in the two sub-systems. There is, however, an additional problem. With the 30 liters/sec well provision had to be made for shutting off the well (automatically or manually) when the outlet taking water was closed. In the case of the 60 liters/sec well with two sub-systems it may occur that one only of the subsystem ceases to take water, the other one continuing. This requires not simply shutting off the well, but reducing its output by half. In the case of a diesel-driven pump this can readily be done by reducing the engine and pump speed. However, a conventional A C. electric motor operates virtually at one speed only. If the output of an electrically-driven pump is to be capable of being halved, there are two design options. One is to provide a control valve on the pump outlet line, the valve being partially closed when reduced flow is desired. The other is to cycle the pump at full flow on/off into a balancing reservoir, withdrawing from it at the desired constant reduced rate.

Partially closing the outlet valve is the course taken in a manually-controlled installation, when spill at the riser results from one of the two sub-systems ceasing to take water. Partial valve closure can also be automated, by float control from an elevated tank (an arrangement common in municipal supply systems). Disadvantages include energy loss inherent in operating against the additional head imposed by the partially closed valve, and increased wear on the pump thrust bearing due to the additional head. Neither factor is a serious consequence if operation with only one sub-system taking water occurs for a small proportion of time (ten to twenty percent).

The use of a balancing reservoir and cycling the pump on/off avoids the energy loss of partial valve closure. However, the viability of this system depends upon the provision of adequate capacity in the reservoir, and hence the frequency with which the pump must be switched on and off. Limits to this frequency are imposed by motor heating (five to ten minutes between starts, depending upon the type of motor), wear on the pump thrust bearing at the moment of start-up, wear on switchgear, and the effect of associated surging on the stability of the aquifer formation. Again, none of these factors are critical if operation of one sub-system, rather than two, occurs for only a small proportion of time. The proviso is adequate pondage capacity in the regulating tank.

The tank may be elevated, discharging directly into the buried-pipe distribution systems (requiring a head of four or five meters), or it may be at ground-level, requiring a second stage of lift into the two distribution systems. The elevated tank is in some respects the simplest solution, and has been widely used. It may also be employed in conjunction with a float operated pump outlet control valve, the combination of control valve and regulating tank reducing the size of elevated tank required and the degree of closure of the control valve. However, the surface tank or small earthen reservoir, with second stage of lift, provides the ultimate in flexibility as the pondage capacity can be relatively large and the cycle time with one sub-system only taking water can be an hour or more. In effect, it completely separates the operation of the well from the operation of the two distribution sub-systems, and the subsystems from each other. The total pumping head and energy requirements are not affected by dividing the lift into two stages. The second lift is by simple centrifugal pumps (about 5 h.p. each). Each sub-system may be independently automated, with float control on the riser, as previously described for the 30 liters/sec well.

Functions of the tubewell operator

A perennial question with medium and large tubewells is the role of the tubewell operator. In some cases, he is in complete control of the well and distribution system He operates the well and the outlet valves, organizes or approves all delivery schedules, and determines water charges for each cultivator. He also performs routine maintenance on equipment and maintains records. The operator or his helper must be present whenever the well is running, and conversely operation ceases if they are absent. The annual salary of the operator and helper in some government-owned and operated systems equals the annual cost of power for the well.

A question of particular interest is what the duties and responsibilities of the tubewell operator should be in the context of cultivator ownership and operation or at least of cultivator operation of a well. As the cultivators would collectively be meeting the cost of the operator, they would presumable wish to minimize that cost by transferring some of the duties to the water user group itself, or by eliminating certain functions through automation of the system. A key item in the latter category is control of the tubewell pump. The irrigation distribution systems under discussion are low-pressure pipe and incorporate an open riser adjacent to the well to limit the pressure to the pipe in the event that all outlet valves are closed and the pump is still running. With such a provision, when the operating outlet valve is closed, spill occurs at the riser and continues until the pump is shut off. This requires the presence of an operator. However, if the riser is substituted by an elevated tank with float control, the pump can be automatically shut off and restarted when an outlet valve is again opened. Various protective devices can also be provided in the switchgear to automatically safeguard the pump against power supply deficiencies. Presence of a full time operator is then no longer needed at the well. Moreover, cultivators irrigating possibly half a kilometer away from the well can open or close their outlet or change operation to a different outlet. As far as the equipment is concerned, inspection several times per day is all that is required. The operator would also record the hours of water use by the individual cultivators in the particular unit receiving water each day and check the totals against the reading of an hours-of-running meter at the well. The task of billing cultivators for water, dealing with overdue accounts, and making payments for power could be either by the operator or by a designated member of the water user group for the well.

Diesel-driven installations can also be equipped for automatic starting and stopping in response to opening or closing of an outlet valve, through provision of an elevated float chamber. However, more frequent inspection during the course of the day is required than in the case of electrically-driven units. If an operator is required to be continuously in attendance, in any case the automatic, start-stop provision can be dispensed with in favor of manual.

Power supply problems

A principal problem with operation of publicly-owned and operated large well systems in some areas has been deficiencies in electric power supply. The deficiencies have included limited hours of availability, unpredictable outages, and low voltage. With unscheduled power outages rotational irrigation schedules break down and unauthorized operation of outlet valves becomes prevalent. Attempts to ensure reliable power supply by constructing "dedicated" feeder liners, nominally reserved for supply to public tubewells, have not been as effective as hoped for.

There are obviously limits to the extent of power supply deficiencies beyond which it would be impractical to pursue further installation of electrically-driven public tubewells. However, there are means by which the consequences of power shortages can be minimized. First, the size of service area of a well should be designed with reasonably conservative regard to the hours of power supply likely to be available. Second, in anticipation of the occurrence of unplanned power outages, the scheduling of irrigation to cultivators should be kept as simple as possible, with entirely independent operation of sub-systems (if the well supplies more than one delivery system), and the delivery units kept as small as possible. The procedure scheduling irrigations lost by individual cultivators through power outages should be clearly understood by all concerned.

Comparison of medium and large wells

The above discussion has referred to medium wells with capacity up to some 30 liters/sec, supplying a single distribution system, and large wells of greater capacity supplying two or more distribution sub-systems. Choice between the two depends on economic and operational factors. With regard to the well and pump, there are economies of scale, the larger capacity well having lesser fixed cost per unit of output, particularly where a relatively deep well is necessary to reach the aquifer. With regard to the distribution system, the opposite is the case. Cost of water distribution systems (whether canal or buried pipe) increase per unit of area served with increasing size of the area supplied. In the case of tubewell systems with distribution of the type described there is a significant upward step in distribution cost when multiple (two or more) subsystems are introduced, as with large wells, rather than the single system of the medium well. This is due to the need to provide for one of the sub-systems ceasing to take water while irrigation continues in the others. The means of doing so have already been discussed (outlet control valve, elevated regulating tank, or surface pondage with supplementary lift).

A further significant cost item is the pumping head required. For a particular well, dynamic drawdown increases with rate of pumping. Where the depth to static watertable is great, differences in dynamic draw-down may not be very significant as a proportion of total pumping head, although still important in absolute terms. For a lesser depth to watertable, dynamic draw-down can be a very important item, largely determining pumping energy requirements and favoring the use of a smaller capacity well.