A high groundwater table and soil salinity, especially in arid regions, often cause serious problems for agriculture. In irrigated areas the subsurface drainage can be an effective technique to lower the depth of the groundwater table and reduce soil salinity. In drainage systems, lateral pipes are designed to collect the free water from soil and convey it into collectors. In other words, collectors are commonly designed to convey drainage water from laterals downstream, while the laterals play an additional role in removing excess water from irrigated land. The present research was conducted to investigate the effects of collectors in discharging excess water from soil using a laboratory-tank model. The results indicated that on average 24% of drainage water was directly discharged through the collector pipe when the hydraulic gradient of the flow was sloped toward the collector pipe. Hence, it was concluded that, with proper monitoring, collectors were capable of reducing the drainage coefficient from an average of 32.5 to 24.5 mm/day, while drain spacing of the laterals can increase the results by about 15% in comparison with the present situation.

INTRODUCTION

Rising groundwater tables and soil salinization are two significant factors threatening the viability of cultivated lands, especially in arid and semi-arid regions. Today, there is increasing concern that shallow water tables and salinity may hamper agricultural productivity and in the long run endanger relevant infrastructures. Experience has shown that a subsurface drainage system is an easy and suitable way to remove the drainable surplus and maintain the water table at a desirable level below the ground surface (Singh et al. 2007). Studies showed that large areas of irrigated lands that suffer from shallow water tables, have been covered by drainage networks to provide favorable conditions for efficient agricultural operation. In this regard, in addition to technical investigations, the cost of construction and efficiency of such drainage schemes were considered in order to minimize the costs of construction.

Several studies have been conducted to investigate subsurface drainage systems at field scale and in the laboratory. For example, flow conditions and hydraulic head fluctuation in laterals were investigated by Kim et al. (2008) using a sand tank. The study revealed that increasing the velocity of flow to a certain level could lead to a reduction in the hydraulic effectiveness in the lateral pipes. In addition, an increase in the length and diameter of the laterals led to an increase in the volume flowing through the laterals and in contrast, a greater diameter of the lateral pipe resulted in reduction of flow through the laterals. Singh & Callaghan (1978) evaluated the performance at different depths of a subsurface drainage system in both the steady state and non-steady state, using a sand tank. However, the research did not report any role played by collectors in absorbing part of the soil excess water through the water table. The effect of a subsurface drainage network and evaporation on water table dynamics was investigated by Cook & Rassam (2002) using a simple analytical model. The study showed that there was no significant difference between simulated and measured values, hence the model could be used to estimate the amount of water table drawdown. Furthermore, Bilal & Sarvar (2008) studied the effect of tube-well drainage on water logging, salinity and sodicity in a field located at Peshawar, Pakistan. It was found that the electrical conductivity of the soil was significantly reduced as a result of land drainage, thus indicating that drainage practices could be a proper way to control the potential of soil salinity, as it was capable of keeping the water table at a specified tolerant level. Meanwhile, it has also been observed that a subsurface drainage system is an appropriate technique to improve crop field and remove waterlogging, especially in arid and semi-arid regions (Martínez-Beltrán et al. 2007).

Studies on the drainage coefficient rate such as that conducted by Skaggs et al. (2006) concluded that the average values of drainage intensity vary in the range from 0.58 to 1.6 cm/day. It was further argued that an increase in drain spacing can reduce the costs of construction and installation of a subsurface drainage system. It is worth noting that wider spacing normally decreases the application of more complicated and costly investigation techniques, equipment and construction. This is because this technique requires fewer pipes for operation as compared with previous methods which involve installation at a narrower spacing. Similarly, Kale (2012) explored the performance of an underground drainage network in changing the depth of the water table and soil salinity balance. The results obtained indicated that the water table level quickly decreased, as soil salinity was reduced to about 24–38% in irrigated fields under wheat and bean cultivation, respectively. Ayars et al. (2006) observed that controlled drainage treatment was effective in reducing salt load and disposal problems in arid and semi-arid regions. A controlled drainage system has been one of the accepted methods to maintain the water table at an optimal depth during the dry season (Guldaraz et al. 2002). In arid regions with a long growing season, most environmental problems are associated with a high volume of saline water in the drainage system (Akram et al. 2013). However, it was reported that deeper drains had the tendency to bring more water out from deep layers and cause the accumulation of large salts on the ground surface. Spaling (1998) highlighted the utility of the practical applications and stated that it was possible to carry out a cumulative effect analysis using available information. Kale (2012) argued that construction of a proper subsurface drainage system in irrigated areas led to a drop in the groundwater table, so that the soil salinity level reduced by approximately 24.1%, 37.9%, and 14.4% in the areas of wheat, bean and fallow, respectively. Akmal et al. (2014), in a field study, investigated how the rising water table in an irrigated area led to increased soil salinity in the root zone, and stated that in terms of economic analysis and the environment, the depth of a drainage system can be an important parameter in design and installation (Razi et al. 2011). However, deep drains largely caused the increase in salinity of drainage water. Hasanpor et al. (2010) argued that most crop plants cannot tolerate the levels of waterlogging and salinity hence crop yields will be extremely reduced.

Table 1

Summary of statistical results from measured values through both collector and lateral pipes

ParameterCollector (mm/day)Lateral (mm/day)Total (mm/day)Collector (%)
Max 10.9 35 46.2 29 
Min 5.6 17.5 23.1 22 
Average 7.8 24.5 32.5 23.9 
Median 7.8 27.02 241 23.8 
Total (mm) 508.3 1,608.6 2,116.8 24 
ParameterCollector (mm/day)Lateral (mm/day)Total (mm/day)Collector (%)
Max 10.9 35 46.2 29 
Min 5.6 17.5 23.1 22 
Average 7.8 24.5 32.5 23.9 
Median 7.8 27.02 241 23.8 
Total (mm) 508.3 1,608.6 2,116.8 24 

Obviously, economic analysis is a favorable procedure to determine drain spacing in subsurface drainage schemes, because it helps to estimate the return rate to land at maximum value for managing the drainage structure treatment in an ideal condition. Nevertheless, Smedema et al. (2004) stated that the Hooghoudt equation is an excellent formula to use for determining the spacing parameter and required drainage in an acceptable range. The approach established an inverse relationship between the drainage coefficient rates and square distance of laterals. Therefore, a higher drainage coefficient rate requires narrower drain spacing, while a lower value needs wider drain spacing, although the values are adjustable for economical spacing of drains to justify investment. As a consequence, optimization of the drainage factor can be a desirable measure to lower the costs of construction and maintenance while enhancing the performance of agricultural lands. Thus, despite various investigations into drainage, it is still necessary to be more creative to increase the capacities of drainage systems using the application of practical models. On the other hand, laboratory methods, where the natural condition of soil can be simulated to research the application of desirable techniques, can be a proper option. Hence, with respect to the function of the drainage coefficient and the importance of economical spacing of drains, a laboratory model can facilitate the estimation of the discharge rate through collectors and laterals, separately. Through this method, the effectiveness of collectors on drained land can simply be evaluated and determined. The main objective of this work, therefore, is to investigate the effects of a reticular collector on the removal of the drainable surplus in an irrigated field using a laboratory model.

MATERIAL AND METHODS

Experimental design

A rectangular steel tank constructed in a laboratory with inside dimensions of 2 m length by 0.5 m width and 1 m depth was filled with soil of different layers established to simulate natural conditions. The tank wall was 3 mm thick and supported by angle iron at the top and bottom. The inside surface of the box was painted black to prevent heat loss by radiation. Drain pipes with a diameter of 2.1 cm, were reticulated and placed at a depth of 80 cm, with lateral spacing maintained at 20 cm, while a drain pipe was placed at a depth of 5 cm from the laterals as a collector (Figures 1 and 2). A stainless steel filter was put around all the drains to prevent the entry of sand particles into the pipes. In between the laterals, piezometer semi-flexible tubes were installed at different horizontal spacing of the drains. The piezometers were installed in the soil to monitor the fluctuation of the water table during the experimental period. The piezometers were arranged between the drain spacing at specified distances from collector pipes.

Figure 1

Position of laterals and collector in the tank plan.

Figure 1

Position of laterals and collector in the tank plan.

Figure 2

A soil tank used in the laboratory.

Figure 2

A soil tank used in the laboratory.

Measuring and monitoring

Irrigation tools were attached to the tank at specified levels based on required water, soil surface parameters, and weekly irrigation using a pipe network and sensor. During the experimental period, the amount of water entering into the box was continuously monitored and recorded. In addition, the level of water within the piezometers, volume of discharged water through the laterals and the collector were measured twice daily. The collected data were analyzed using a statistical tool called ‘Statistical Package for Social Sciences (SPSS)’.

RESULTS AND DISCUSSION

Water flow through lateral and collector pipes

A summary of statistical results from measured values through both collector and lateral pipes is presented in Table 1. The investigation of results indicates that discharge values through the collector pipe ranged between 5.6 and 10.9 (mm/day), while the highest and lowest discharge values from lateral pipes were 35 and 17.5 (mm/day), respectively. The comparison between the discharge rates through the collector and laterals illustrates that about 24% of drainable water was directly received by the collector pipe during the study period, so the difference between discharge of laterals and collector as observed was significant. In other words, this measurement gives the proportion of drainable surplus which discharged into the collector pipe and then flowed toward the outlet. In design, such values are commonly neglected probably due to the lack of required parameters for subsurface drainage systems.

Equation (1) was used to determine the drained water from the collector and lateral pipes. 
formula
1
where is the outlet discharge, is the discharge through the lateral pipes, and is the discharge of the collector (mm/day). In addition, fluctuation of drainage water values from the laterals and collector is shown in Figure 3. In the graph, the amounts of discharge through the laterals and collector are shown over a 2-month period. The peak points of the curves signal the beginning of the irrigation operation and the steeper lines reveal the drainage process. A close analysis of the trend of the curves indicates the existence of a sharp fluctuation in the amount of discharge through the laterals while in the case of the collector curve, the fluctuation is gradual over the time period.
Figure 3

Comparison of daily discharge through the collector and laterals in the laboratory.

Figure 3

Comparison of daily discharge through the collector and laterals in the laboratory.

Figure 3 also demonstrates a slight increasing trend in both curves during the period of the experiment from the first day until the last day of measurement. In other words, the amount of discharge gradually increased throughout the 65 days, such that the higher values were recorded at the end of the experiment period compared with the initial values. The results can probably be attributed to dissolution and leaching out of excess salt from soil pores. The graph further reveals that the amount of water absorbed by the collector is more than the average water received by the laterals because the pipe length used for the collector is longer than each lateral pipe. The volume of water observed at the outlet was somewhat more in the collector than in each of the laterals. As a result, it may be proper to conclude that a considerable portion of the drainable surplus was received through the collector pipe throughout the period of the study.

Comparison of drainage coefficient with discharge from the lateral pipes

In accordance with the role and function of collectors in discharging a portion of the drainable surplus, the quantity of water which drained through the lateral pipes, and the drainage coefficient rate, also referred to as the discharge rate, were investigated. Figure 4 shows the difference between estimated value and the observed discharge using the laboratory method measured over the period of 2 months. With respect to the chart, the drainage coefficient rate or required discharge value was calculated as 28.5 and 35.5 mm/day in the 1st and 2nd months of the experiment period, respectively. However, the drained values via the laterals were measured at 21.8 and 27 mm/day during the same period. Since about 24% of the drainable surplus was absorbed by the collector pipe, as a result the actual drainage coefficient rate would be considered as equal to 76% of the total amount of the estimate. Thus, the drainage design coefficient rate can be reduced from 35 or 28.5 to 24.5 (mm/day).

Figure 4

Comparison between the designed drainage coefficient and observed values.

Figure 4

Comparison between the designed drainage coefficient and observed values.

A linear statistical analysis was conducted on the drainage coefficient for all measured data, leading to the following formula being derived: 
formula
2
where qa (mm/day) is the actual drainage coefficient and qe (mm/day) is estimated values of the design coefficient. This equation can be used to estimate the actual drainage design coefficient rate for a given location.
Conversely, according to the Hooghoudt and Donnan formula, the relationship between drain spacing and required discharge rate is inversed as in the following equations. 
formula
3
 
formula
4
 
formula
5
 
formula
6
where L (m) is drain distance, q (m/s) is the required discharge, k (m/s) is equivalent hydraulic conductivity, h (m) is hydraulic head difference between drain and soil, and d (m) is equivalent depth from the drain to the restrictive layer. From Equation (6), it can be seen that a reduction in drainage coefficient (24%) can result in an increase of approximately 15% in the spacing of lateral pipes in a subsurface drainage system. As a result, the cost of construction from more excavation and longer pipes can be reduced while also minimizing land loss.

Investigation of the effective radius of the collector pipe

Statistical analysis of the measured water table level is the most straightforward and reliable technique for specifying the effect of the collector pipe on water table fluctuations (Smedema et al. 2004). The average water level from the soil surface within test wells as directly measured can be seen in Table 2. Fluctuations of the hydraulic head in the observation wells are also presented in Figure 5 for the 65-day period with the application of irrigation water.

Table 2

Water head in specified distances from collector pipe

Distance (cm)Hydraulic head (cm)
10 
10 20 
15 27 
20 34 
25 38 
30 39 
35 40 
40 40 
45 40 
50 40 
Distance (cm)Hydraulic head (cm)
10 
10 20 
15 27 
20 34 
25 38 
30 39 
35 40 
40 40 
45 40 
50 40 
Figure 5

Reduction in depth of water table within piezometers at the specified distances from collector.

Figure 5

Reduction in depth of water table within piezometers at the specified distances from collector.

Investigation of the patterns of flow downward and toward drains, and comparison of water table head loss revealed that the water level in the zone adjacent to the collector dropped more than in zones located farther from collector. In other words, in the zone near the collector pipe, the water table is affected by both the collector and lateral pipes while at a distance greater than 35 cm from the collector, the water head is largely influenced only by laterals. Such a pattern remarkably showed that the hydraulic gradient of the flow tilted toward the collector pipe. Therefore, this can explain the function of collectors in receiving water from the soil, because the water levels at greater distances from collector pipes are largely affected by lateral pipes, while those close to collectors are affected by both laterals and collectors.

In addition, data from observed head loss over the horizontal flow zone, where the water table is equal between two piezometers, with estimated discharge values were compared using the correlation coefficient approach. The required discharge rate is calculated using the Hooghoudt formula, which is applicable in drainage systems, as in Equations (7) and (8): 
formula
7
 
formula
8
where q is drainage coefficient (m/s), k is hydraulic conductivity (m/s), h is hydraulic head (m), L is drain spacing (m), and d is equivalent thickness of flow zone (m).
Comparison of the results indicated that there is a reasonable correlation between the drainage coefficients obtained from the Hooghoudt approach and observed head loss. The coefficient of correlation was 0.93, which is acceptable for laboratory conditions. The statistical analysis on measured head loss resulted in Equation (9): 
formula
9
where q is required discharge (mm/day) and h is hydraulic head loss (m).

This derived equation, which is applicable only in this experimental set up, can be used to estimate the required discharge or hydraulic head rate. As a result of high water drawdown rate in the zone near to the collector pipe and the high correlation coefficient between head loss and required discharge, it can be concluded that collectors can influence the absorption of drainable water through soil layers.

CONCLUSION

This research which was conducted in a laboratory has revealed the effectiveness of collector pipes in absorbing and removing some of the drainable surplus directly from an irrigated field. With respect to the performance of collector pipes, the drainage coefficient rate can be considerably lowered from the normal condition for the same drainage network. According to the Hooghoudt equation, there is an inverse relationship between the required discharge rate (drainage coefficient) and the square of lateral distances. Thus, the reduction of the drainage coefficient rate can lead to an increase in drain spacing of approximately 15% from the initial condition. This implies that drain spacing can be increased from 40 to 46 cm under laboratory conditions, while a similar change can be adopted under field conditions. In this regard, the required materials for installing underground drainage would be reduced due to increased drainage spacing in the irrigated area. In addition, costs associated with the construction and maintenance would be minimized, hence the utility of the schemes would be effectively improved.

ACKNOWLEDGEMENT

The authors would like to express their sincere thanks to the reviewers and associate editor for their recommendations and assistance in improving the technical details and precision of this paper.

REFERENCES

REFERENCES
Akram
M.
Mazari
A.
Nahvi
A.
Bakhtiari
Z.
Hedar
D.
2013
Subsurface drainage in Iran
.
Wiley Online Library
62
(
3
),
306
314
.
doi:10.1002//ird/1774
.
Ayars
J. E.
Christen
E. W.
Hornbuckle
J. W.
2006
Controlled drainage for improved water management
.
Agricultural Water Management
86
(
1–2
),
128
139
.
doi:10.1016/j.agwat.2006.07.004
.
Bilal
S. M.
Sarvar
T.
2008
The effect of tube well drainage on water
.
Sarhad Journal of Agriculture
24
,
50
58
.
Guldaraz
K.
Latif
M.
Mahamad
J. K.
Rameshwars
K.
2002
Analysis of groundwater behaviour in a farmer controlled subsurface till drainage system in Mardan, Pakistan
.
Irrigation and Drainage System
16
,
327
345
.
Hasanpor
B.
Parsinejad
M.
Selahshor
F. K.
2010
Estimation of water table fluctuations in paddy fields by using Drainmod
.
Iranian Journal of Irrigation and Drainage
4
(
2
),
167
174
.
Kale
S.
2012
Impact of drained and un-drained soil conditions on water table depths, soil salinity and crop yields
.
African Journal of Agricultural Research
7
(
19
),
2935
2945
.
doi:10.5897/ajar12.101
.
Martínez-Beltrán
J.
González-Casillas
A.
Namuche-Vargas
R.
2007
The evaluation of the performance of subsurface drainage
.
Irrigation and Drainage
56
(
1
),
179
190
.
Razi
F.
Setodeh
A.
Danshkar Arrasteh
P.
Akram
M.
2011
The effect of deep drain on salinity of drainage water using a laboratory model
.
Iranian Journal of Water and Soil (Persian)
3
(
43
),
281
288
.
Singh
B.
Callaghan
J. R.
1978
Effect of interflow on soil drainage
.
Journal of Agriculture and Engineering
23
,
397
415
.
Singh
R.
Helmers
M. J.
Crumpton
W. G.
Lemke
D. W.
2007
Predicting effects of drainage water management in Iowa's subsurface drained landscapes
.
Agricultural Water Management
92
(
3
),
162
170
.
doi:10.1016/j.agwat.2007.05.012
.
Skaggs
R. W.
Youssef
M. A.
Chescheir
G. M.
2006
Drainage design coefficients for eastern United States
.
Agricultural Water Management
86
(
1–2
),
40
49
.
doi:10.1016/j.agwat.2006.06.007
.
Smedema
L. K.
Vlotman
W. F.
Rycroft
D. W.
2004
Modern Land Drainage
.
A.A. Balkema
,
Leiden
,
The Netherlands
.
Spaling
H.
1998
Analyzing cumulative environmental effects of agricultural land
.
Agriculture
, Ecosystems and Environment
53
,
279
292
.