Effluent salinity of pipe drains and tube-wells : a case study from the Indus plain

Keywords: anisotropy, aquifer, desalinization, effluent salinity, groundwater, irrigation, salt-water upconing, soil salinity, stream-function, subsurface drainageIrrigated agriculture in arid and semi-arid zones often suffers from waterlogging and salinity problems. Sub-surface drainage systems can be used to control the groundwater table and to facilitate the leaching of salts from the rootzone. In the Indus plain, pipe drains and tube-wells are used for this purpose. Regional water management requires that the development of the effluent salinity with time of these systems is known in advance. Numerical models based on the Darcy equation and the mass balance equation for water flow and the advection-dispersion equation for solute transport are powerful tools to predict the effluent salinity of pipe drains and tube-wells at field level. In advection-dominated transport problems, however, solute impulse response functions based on stream-functions constitute a more computationally efficient approach.A new modelling approach is presented that combines the one-dimensional vertical finite-difference SWAP model for the variably saturated zone with a solute impulse response function for the saturated zone. This approach is applied to the Sampla experimental pipe drainage site in Haryana, India, the S-I-B-9 pipe drainage unit of the Fourth Drainage Project, Punjab, Pakistan and the Satiana tube-well Pilot Project, Punjab, Pakistan. Results show that the effluent salinity of pipe drains and tube-wells changes only gradually with time due to the low percolation from the irrigated fields and due to the large quantities of salts stored in the groundwater. Areas with relatively high percolation and a shallow depth of the impermeable layer (pipe drains at Sampla) still require 10 years before the effluent salinity has reduced to equilibrium levels. In contrast, desalinization of the rootzone generally takes only 1-3 years. The implication is that farmers will benefit quickly from the installation of a drainage system. However, for the safe use and disposal of the effluent, long term solutions are required.In the Indus plain, groundwater salinity usually increases with depth. In water scarce areas, the shallow fresh groundwater may be an important source of irrigation water. In waterlogged areas, where sub-surface drainage is installed to control the groundwater table, the presence of fresh groundwater bodies may result in a relatively low effluent salinity. The finite-element model SUTRA is used to study the behaviour of skimming wells and pipe drains in fresh-saline groundwater systems. The model is calibrated on two documented experiments with a skimming well and a scavenger well at Phularwan research farm, Punjab, Pakistan. Salt water upconing below the skimming well is particularly sensitive to the anisotropy factor of the aquifer. The relationship between aquifer anisotropy and the Electrical Conductivity ( <em>EC</em> ) of the pumped water is non-linear. The skimming well simulations show that water with an <em>EC</em> of<img src="/wda/abstracts/ongeveer_enkel.gif"/>1.7 dS m <sup>-1</SUP>can be pumped from a thin fresh groundwater body, provided that the pumping rate is low. Under the same circumstances, pipe drains yield a better effluent quality ( <em>EC</em> of 1.2-1.3 dS m <sup>-1</SUP>). With pipe drains, flow is restricted to the shallow fresh groundwater. The deeper saline groundwater is left untouched. The better effluent quality for pipe drains as compared to skimming wells, must be evaluated against the considerably higher installation costs for pipe drains.