Irrigation and Drainage Systems 15: 269–279, 2001. © 2001 Kluwer Academic Publishers. Printed in the Netherlands.

Selecting the drainage method for agricultural land MARINUS G. BOS International Institute for Land Reclamation and Improvement, P.O. Box 45, 6700 AA Wageningen, The Netherlands; e-mail: [email protected] Accepted 30 August 2001

Abstract. To facilitate crop growth excess water should be drained from the rooting zone to allow root development of the crop and from the soil surface to facilitate access to the field. Basically, there are three drainage methods from which the designer can select being; surface drains, pumped tube wells and horizontal pipe drains. The selection between these techniques is not very straightforward and depends to a certain extend on the preferences of the designer. Yet, guidance is given in this paper on the basis of a set of questions that are grouped in a flowchart. Although the flowchart suggests sharp choices between alternative drainage methods, in reality the choice between the recommended methods overlaps. Partly this is because the values in the selection criteria (questions) are indicative. Further overlap is caused by the construction and operation & maintenance cost of the drainage system under local conditions. These costs are not included in the flowchart. Key words: drainage, method, pipe drains, selection criteria, surface drains, tube wells

Introduction The selection of a suitable method to drain agricultural land depends on a variety of measurable environmental characteristics and on the institutional boundary conditions (Figure 1). Data are needed on these measurables before a justified selection of a suitable drainage method can be made.

Data measurement As shown in Figure 1, data are needed on eight groups of measurables. The level of detail with which these data need to be measured is discussed below. − Precipitation: Daily precipitation should be measured with standard gauges. Data have to be made available in two ways: • Average monthly precipitation (P in mm/month) is needed for comparison with the potential evapo-transpiration (ETpot ent ial ).

270

Figure 1. Data needed to select a drainage method.

• Daily precipitation needs to be calculated for an 80% wet year. This daily value (in mm/day) will be compared with the infiltration rate of the soil. − If the considered area is irrigated, the irrigation water supply towards this area (Vc ) needs to be measured with sufficient accuracy. In open canals the use of a broad-crested weir or long-throated flume is recommended (Bos 1976; Bos et al. 1984; Clemmens et al. 1993, 2001). Accurate methods to measure the flow in pipes are described by the USBR (1997). Flow data need to show the irrigation water supply in m3 /month per hectare. − The potential evapotranspiration (ETpot ent ial ) of the major crops in the area can be estimated by use of the Penman-Monteith equation (Monteith 1965). Calculations can be made with widely available software like CRIWAR (Bos et al. 1996) and CROPWAT (Smith 1989). To calculate ETpot ent ial (in mm/month) meteorological data are needed on; temperature (minimum and maximum), radiation, humidity and wind speed. − The “water related” characteristics of the major crops grown in the area can be used to narrow the range of values of parameters that influence crop growth like; ETpot ent ial (see above), the ponding tolerance of the crop in number of days, effective rooting depth and salt tolerance. The selected values of these parameters should be sufficiently conservative to avoid the need for a different drainage system because of short-term changes in the cropping pattern.

271 − The infiltration rate quantifies the layer of water that can enter the soil. If precipitation exceeds the infiltration rate, the excess water cannot enter and either becomes surface run-off or will form a pond at the surface. The infiltration rate can be measured from a ponded field or with a “ring infiltro-meter” (Bouwer & Jackson 1974). The measured infiltration rate tends to vary widely over the area to be drained. Average values should be established for characteristic soils in the area. − The depth to thegroundwater table can be measured in an observation well. The frequency with which the depth should be measured depends on its fluctuation. For drainage selection, a frequency of once per month is recommended. The number of observation wells depends on the gradient of the groundwater table. An average of one per 1000 ha is recommended. The reference level of all wells should be levelled against each other so that a groundwater table contour map can be prepared. − Measurement of the groundwater quality concentrates on the salt content of the water. The salt content can be measured in a laboratory or can be expressed in terms of electrical conductivity. This EC can be measured in situ with a low-cost meter. The EC is expressed in dS/m at 25◦ C. Water with an EC-value below 1.5 dS/m can be (re-)used for irrigation. EC-values over ≈ 4.0 dS/m cause 20 to 40% yield reduction of many crops and thus can be called harmful (Unesco 1970). − The transmisivity (KD-value) of an aquifer is the product between the hydraulic conductivity (K) of the soil and the thickness (D) of the saturated layer. The KD-value is determined by pumping tests. Methods to evaluate pumping test data are described by Kruseman and De Ridder (1971). Computer models to analyze data are given by Boonstra (2001). Humid climates As shown in Figure 2, the first question in the decision flow chart reads: does annual precipitation exceed 1.5ETpot ent ial ? If yes, the climate may be called humid (FAO 1978). In general, precipitation then is sufficient to meet the ETpot ent ial of the cultivated crop while remaining water is available for (nonagricultural) downstream uses. In humid climates the method of drainage largely depends on the soil type. With peat soils, the infiltration rate always exceeds the daily rainfall, while the storage capacity in the peat is about 50% of its unsaturated volume. Hence, if the top 0.5-meter consists of unsaturated peat, a day-rain of as high as 250 mm can be stored before the groundwater table rises to the surface. A system with parallel open drains is needed to evacuate this stored water from the area. As a rule of thump, the capacity of this drainage system should be

272

Figure 2. The humid climate branch.

sufficient to discharge the cumulative 5-day rainfall of an 80% wet year (Bos 1994). To avoid excessive subsidence, the water level in the drainage system must be checked so that the groundwater table is not lowered below the level needed for crop growth. Series of check structures are needed to control the groundwater table. Because of the high permeability of peat the head loss over each structure should be limited to about 0.2 m (DID & LAWOO 1996). With a shallow peat layer on top of mineral soils, subsidence may result to a relatively rapid decrease of the thickness of the peat layer. If the storage capacity in the top peat layer becomes too little to avoid prolonged ponding, the drainage system should be changed. If the topsoil (mostly) consists of mineral soils, part of the rainfall during an 80%-wet-day cannot infiltrate. Thus, to avoid damages to the crop through ponding, the land surface must be graded to allow water to flow to surface drains. This requires a land slope of 0.2 m per km or more. Also all infrastructure (roads, etc.) should be designed in such a way that ponded run-off does not hamper operation. Peak run-off rates may be estimated e.g. by use of the “curve number method”. Boonstra (1994) gives examples to determine run-off rates.

273

Figure 3. The sub-humid climate branch.

Because of the high rainfall intensities and the need for a sloping land surface, the discharge capacities of drains tend to be rather high. To reduce this capacity, low-lying areas might not be drained and be used as wetland cum storage area.

Sub-humid climates Upon arriving in Figure 3 the next question in the decision flow chart reads: does annual precipitation exceed ETpot ent ial ? If yes, the climate indicator P/ETpot ent ial ranges between 1.0 and 1.5. Climate then may be called subhumid (FAO 1978). Four key questions can be used to decide on the drainage method. The period that the root system of crops can be “submerged” without significant yield reduction vary greatly. For non-rice crops it is recommended that all rain falling during an 80% wet day either should infiltrate or a surface drainage system should be constructed. Thus, the first question in this branch of the flow chart (Figure 3) asks if the cumulative one-day rainfall exceeds the infiltrate rate of the soil. If yes, a surface drainage system is required as describes for humid climates (see Section 3). If the cumulative one-day rainfall can infiltrate, the next question is related to the destination of this infiltrated water. If the depth to the groundwater table during the growing season exceeds 1.0 m, crops can develop a root system that is sufficient to attain potential crop yield while precipitation is sufficient

274 to leach salts from the root zone. Salinity usually is not a problem due to this leaching if the average groundwater table is below 1.0 m during the dry season. Hence, no drainage system is needed for sustainable agriculture. If the average depth to the groundwater table is less than 1.0 m during the growing season a drainage system is needed. The choice of the system depends to a large extend on the possibility, or the need, to re-use drainage water for irrigation (Figure 4). If the area is irrigated, it commonly is recommended to manage surface water and groundwater in conjunction. Hence, groundwater can be pumped to supplement surface water during periods with low surface water flows. For pumped tube wells to function as an effective drainage cum irrigation system, however, a reliable source of energy should be available. Otherwise the use of horizontal drainpipes, that are less dependent on pumping, are recommended. If the area is not irrigated, drainage water is not needed for re-use within the area to be drained. Pipe drains discharging water by gravity to a downstream area then is to be considered. Care should be taken not to drain too deep. This would reduce capillary rise during the growing season so that water shortage is introduced that reduces crop yield!

Semi-arid and arid climates Leaving Figure 3 via the branch towards Figure 5 leads to a climate indicator P/ETpot ent ial below 1.0. Climate then may be called semi-arid or arid (FAO 1978). The first question then is: is there an additional source of water to cover the gap between precipitation and ETpot ent ial ? In other words; is the area irrigated? If “no”, a drainage system should be used that conserves water that drains (naturally) during wet periods for re-use during dry periods. Water can be conserved by reducing surface run-off to streams (more infiltration) and by constructing storage facilities in streams. If the area is irrigated and the groundwater table is deep (more than about 3.0 m) no drainage system is needed. This depth of 3.0 m to the groundwater table is sufficient to avoid salt accumulation in the root zone of the irrigated crop. One should realise, however, that a considerable volume of irrigation water enters the area (Vc in m3 /month). The crop will not consume part of this irrigation water. This part needs to be discharged naturally or through an artificial drainage system. Water balance studies in semi-arid areas indicate that non-consumed irrigation water causes a rising groundwater table if the overall consumed ratio ETpot ent ial /(Vc + P) is less than 0.6 (Bos et al. 1991; Bastiaanssen et al. 2001). To avoid the need for a (costly) future drainage system, irrigation water should be used efficiently. In this context, the irriga-

275

Figure 4. The sub-humid climate branch with shallow groundwater depths.

tion water supply to the considered area (Vc ) must be measured and managed taking the salt balance of the area into consideration (Clemmens et al. 2001). Leaving Figure 5 leads us to irrigated areas with a shallow groundwater table and a semi-arid climate. This “water excess” situation in a “water short” climate is often a man-made problem that to a large extend can be cured by using irrigation water more efficiently (Bos & Nugteren 1974). If the toplayer of groundwater is fresh (say top 25 m) the most obvious method is to re-use this groundwater for irrigation. For this reason, the salinity of the groundwater should be sufficiently low. Most crops tolerate an EC of the groundwater up to 2.0 dS/m without significant yield reduction. If the top-layer of the groundwater cannot be re-used (either directly or mixed with fresh water), and yet needs to be discharged, the first question is; “is there a disposal site available” for the saline drainage water? In this context the benefits due to drainage within the drained area should be weighted against the damages that are inflicted downstream. In establishing these damages all downstream uses should be considered; agriculture, drinking water supply, wetlands, etc. If damages exceed the benefits, artificial drainage is not justified! If a disposal site is available, care should be taken to maintain the best possible water quality for downstream uses. Selecting a horizontal piped subsurface drainage system with narrow spacing can do this. The narrow spacing is needed because groundwater flow extends to a depth of about 25% of the

276

Figure 5. Entering the semi-arid climate branch.

Figure 6. Irrigated areas in semi-arid and arid regions.

277

Figure 7. Drainage of irrigated land in semi-arid climates.

drain spacing. Thus, with a wide spacing, a thick layer of saline groundwater would be mobilized and discharged onto the downstream environment. If groundwater could be re-used for irrigation and other uses, the first question is whether it can be pumped with a “sufficient” flow rate per tube well. In this context “sufficient” is a rather arbitrary qualification. In the selection diagram it was assumed that pumped water is re-used for irrigation without intermediate storage. To irrigate with reasonable efficiency, the flow rate at field level should not be less than about 30 l/s. Since the drawdown of the groundwater table near the pumped well is restricted, the transmissivity KD of the aquifer then should be greater than about 600 m2 /day in order to allow this flow rate from a single well (Kruseman & De Ridder 1971). With a lower transmissivity there are two drainage options; use well-fields with low capacity shallow tube wells or use horizontal subsurface drain pipes. With the latter the spacing should be such that only the top fresh layer of water is drained. The maximum spacing thus is four times the thickness of the fresh water aquifer. Thus, if the top 25 m of groundwater is fresh (Figure 6) the maximum drain spacing is 100 m. If the KD-value exceeds 600 m2 /day and a reliable source of energy for pumping is available, groundwater and surface (irrigation) water should be managed in conjunction. Tube wells then are the preferred drainage method.

278 If energy is not reliable (i.e. not available if water is needed for irrigation or other uses) a system with narrow spaced horizontal pipes can be used. Conclusions A variety of techniques are available to drain agricultural land; surface drainage, tube wells, horizontal (subsurface) drains and shallow tube wells. The selection between these techniques is not very straightforward and depends to a certain extend on the preferences of the designer. Yet, guidance can be given on the basis of a set of questions related to: − The relation between precipitation and the potential evapotranspiration. − The intensity of rainfall with respect to the infiltration rate of the soil. − The quality of groundwater i.e. if drained groundwater can be re-used for irrigation or other uses. − The reliability of a source of energy to operate tube wells at the moment that water is needed for irrigation. Although the flowchart suggests sharp choices between alternative drainage methods, in reality the choice between the recommended methods overlaps. Partly this is because the values in the selection criteria (questions in the flow chart) are indicative. Further overlap is caused by the construction and operation & maintenance cost of the drainage system under local conditions. References Bastiaanssen, W.G.M., Brito, R.A.L., Bos, M.G., Souza, R., Cavalcanti, E.B. & Bakker, M.M. 2001. Low cost satellite data applied to performance monitoring of the Nilo Coelho irrigation scheme, Brazil. Irrigation and Drainage Systems (in press). Boonstra, J . 1994. Estimating peak runoff rates. In: Ritzema, H.P. Drainage Principles and Applications. International Institute for Land Reclamation and Improvement/ILRI, Wageningen, pp. 111–144. Boonstra, J. 2002. SATEM Software for aquifer test evaluation, International Institute for Land Reclamation/ILRI, Wageningen. pp. 145. Bos, M.G. & Nugteren, J. 1974. On Irrigation Efficiencies. 4t h revised Edition 1990. International Institute for Land Reclamation and Improvement/ILRI, Wageningen. pp. 117. 4t h edition also published in Farsi with IRANCID, Tehran. Bos, M.G. (Ed.) 1976. Discharge Measurement Structures. 1st Edition 1976; 2nd Edition 1978; 3rd Revised Edition 1989. International Institute for Land Reclamation and Improvement/ILRI, Wageningen. pp. 399. Bos, M.G. 1994. Drainage canals and related structures. In: Ritzema, H.P. Drainage Principles and Applications. International Institute for Land Reclamation and Improvement/ILRI, Wageningen, pp. 725–798.

279 Bos, M.G., Replogle, J.A. & Clemmens, A.J. 1984. Flow Measuring Flumes for Open Channel Systems. John Wiley and Sons, New York. pp. 321. Republished 1991, American Society of Agricultural Engineers, St. Joseph MI, U.S.A., Spanish Edition. 1986. Aforadores de caudal para canales abiertos, International Institute for Land Reclamation and Improvement/ILRI, Wageningen. pp. 293. Bos, M.G., Wolters, W., Drovandi, A. & Morabito, J.A. 1991. The Viejo Retamo secondary canal – Performance evaluation case study: Mendoza, Argentina. Irrigation and Drainage Systems 5: 77–88. Bos M.G., Vos, J. & Feddes, R.A. 1996. CRIWAR 2.0: A Simulation Model on Crop Irrigation Water Requirements. International Institute for Land Reclamation and Improvement/ILRI, Wageningen. pp. 117 plus disk. Bouwer, H. and Jackson, R.D. 1974. Determining soil properties. In: J. van Schilfgaarden (Ed), Drainage for Agriculture, Agronomy 17, American Society of Agronomy, Madison, pp. 611–672. Clemmens, A.J., Wahl, T.L., Bos, M.G. & Replogle, J.A. 2001. Water measurement with Flumes and Weirs, International Institute for Land Reclamation and Improvement, Wageningen, pp. 386. DID & LAWOO 1996. Western Jahore integrated agricultural development project; Peat soil management study, Department of Irrigation and Drainage, Kuala Lumpur, pp 270. FAO, 1978. Report on the Agro-Ecological zones project. Volume 1: Methodology and results for Africa, World Soil Resources report 48/1, FAO, Rome. Kruseman, G. & Ridder, N.A. de. 1971. (2nd edition 1990). Analysis and evaluation of pumping test data, International Institute for Land Reclamation and Improvement/ILRI, Wageningen. pp 317. Monteith, J.L. 1965. Evaporation and the Environment. In: G.E. Fogg (Ed.) The State and Movement of Water in Living Organisms. Cambridge University Press. pp. 205–234. Smith, M. 1989. CROPWAT; Program to calculate irrigation requirements and generate irrigation schedules, Irrigation and Drainage Paper 46, FAO, Rome, pp 133. UNESCO 1970. Research and training on irrigation with saline water, Technical report of UNDP project Tunisia 5, Unesco, Paris, 256 p. USBR 1997. Water Measurement Manual, U.S. Bureau of Reclamation, U.S Government Printing Office, Denver.