Basin systems - bankless channels need faster drainage for higher yields

Take home messages

• Consistent high yields can be achieved with quick draining surface irrigation systems.
• Individually draining bays into a field drain OR skip-watering should be considered if water is on bays for longer than 10 hours. This is a good option for bankless channels on flat terrain (<1:1000) which suffer from water backing up.

Background

Eighty three percent of all irrigated land in the southern Murray Darling Basin use surface irrigation. There are two main types of surface systems; border check, which predominate on lighter soils and steeper terrain (slopes > 1:1250) and basin systems, which are best suited to low permeability soils on flatter terrain (slopes < 1:1250). While considerable work has been done to improve border check systems (e.g. Campbell, 1989; Austin and Prendergast, 1997; Lavis et al., 2006), the same cannot be said for basin systems, with published design recommendations (Swinton, 1994) being outdated, not evidence based, and unable to support developments such as bankless channels.

Irrigation systems should be designed to minimise crop losses to waterlogging and water losses to deep drainage. This is achieved by applying a target depth of water in the shortest possible time. In basin systems, this means bays need to be designed so they have quick advance and recession phases. Infiltrated depth can then be managed by varying cut-off times. If this can be done, the correct target depth will be applied with control, precision and uniformity. However, this is generally not the case with contour basin systems and it has been found (North et al., 2010) that:

• Opportunity times in the commonly used bankless channel, contour basin systems are in the order of 30 to 40 hours. These long opportunity times are primarily due to slow drainage. This slow drainage is primarily caused by the hydraulic connection created between bays by the bankless channel, with water backing up in the outlet structure and impeding outflow. To minimise the risk of yield loss to waterlogging in basin systems, water should be on and off bays in less than 10 hours.

A recently completed NSW DPI and Deakin University project (Maximising on farm irrigation profitability) conducted with the irrigated cropping groups and funded through the Australian Government’s Rural R&D for Profit program, Cotton Research and Development and AgriFutures Australia examined ways of improving basin system design. One of the aims was to examine the affect bankless channels had on drainage times and see whether watering and draining bays individually (i.e. without water backing up against outflows) can reduce irrigation ponding times to 10 hours.

Methods

Water advance, flow and depth data were collected during irrigations in two adjacent 3.5ha bays (85m by 410m) in a contour basin layout at Deniliquin, NSW. Bays could be supplied from both a channel and a bankless channel in this layout, so the difference in watering and draining times between the two systems was able to be assessed. Because the two bays were ‘paired’, the effect of surface roughness was also able to be assessed by comparing irrigations with different operating conditions. The collected field data was used to validate an irrigation model (WinSRFR; Strelkoff et al., 1998; Bautista et al., 2009). This model was used to simulate irrigation ponding times where drainage was not obstructed for comparison with actual ponding times where outflow through the bankless channel was impeded by the head of water rising in the bay being filled.

Results and discussion

The field data confirmed that opportunity times of 30 to 40 hours are typical of commercial, bankless channel basin systems on flat terrain (slope 1:2000). Eliminating the hydraulic connection between bays which occurs through the bankless channel, and watering and draining bays individually cut opportunity times to around 15 hours and reduced run-off volumes by 65%.

Because there is no slope along the bay, there needs to be a head of water at the supply end to drive water to the far end. The measurements of water depth showed that this head of water increases with both increasing bay length and surface roughness (Figure 1). Depths of 75mm, 125mm and 150mm above the bay surface were needed at the supply end to drive water to the far end of the 410m long bay at Deniliquin when irrigating with a flow rate of 15ML/day across bare earth, through mown, and through standing senesced pasture, respectively.

If 15ML/day is being delivered into the block from the farm inlet, then the flow through each structure in a bankless channel within a paddock needs to be equal to or greater than 15ML/day. If it isn’t, then water will accumulate in the bay, upstream of the structure. Measurements of the head difference between the inlet and outlet sides of the 600mm pipes between bays showed that a head difference of 25mm was required to drive a 15ML/day flow rate through the pipe (Figure 2).

Figure 1. The relationship between the head of water at the supply end of an 85m wide by 410m long bay and the average distance of the wetting front during six irrigations with varying surface roughness conditions in the contour basin system at Deniliquin.

Figure 2. The relationship between discharge and head loss for the 600mm pipes between bays in the bankless channel of the contour basin system at Deniliquin.

Putting these two pieces of information together explains why drainage is often impeded in bankless channel systems. Even when the bay is bare earth and as smooth as it can be, the water height above the bay surface at the supply end will need to be 75mm to get water to the far end of the bay. Add to this the 25mm head difference required across the outlet to maintain a 15ML/day flow, and it can be seen that a step between bays of at least 100mm will be needed to prevent water backing up in the draining bay. If the bay surface is rougher, the step needed between bays will be more like 150 to 175mm. For a 100m wide bay, this equates to a paddock slope of 1:670 to 1:570, which is roughly 3 to 4 times steeper than most basin layouts, which have slopes of 1:1500 to 1:2500.

The end result of what happens when a 100mm head difference between bays is needed in basin systems which only have a 50mm fall between check banks is best illustrated by the aerial photograph in Figure 3. All bays in the contour basin system at Deniliquin were watered and drained through the bankless channel on 21 March 2017. The fall across five bays that are 85m wide on a slope of 1:2000 is 210mm. If the head of water at the supply end of the bay being irrigated is 75mm and there is 25mm of head loss through each structure between bays, then the surface of the water upstream of the bay being irrigated will meet the surface of the soil at the top of the fifth bay upstream. This can be seen in Figure 3. In effect, this means that five bays (17.5ha) need to be filled with water to irrigate just one 3.5ha bay when using a bankless channel in this system.

Figure 3. Aerial photograph of the irrigation of the contour basin system at Deniliquin being irrigated on 21 March 2017 (second autumn irrigation). The whole paddock was irrigated through the bankless channel and the image shows how water has backed up in the five bays upstream of the one being watered (in the bottom right of the photo) and is only beginning to drain from the sixth bay upstream.

The effect of the bankless channel in slowing drainage is clearly seen when the irrigation on the 21 March 2017 (Figure 4 left) is compared with the earlier irrigation of the same bay on 10 March 2017 (Figure 4 right). On the 10 March, water was supplied directly to the bay from the farm channel and drained so that water could run away freely. This irrigation was the first autumn irrigation and the soil was dry, so it took longer to fill the bay than it did for the second irrigation 11 days later. The main difference between the two events, however, is in the drainage times: 4 hours to drain on the 10 March compared to 40 hours on the 21 March. The modelling indicated drainage times for both these irrigations of around 4 hours when drainage is not impeded (lines that drop the most in Figure 4), which agrees with the data from the 10 March. Thus, it is reasonable to assume that the time water was ponded on the trial bay on the 21 March could have been reduced from over 40 hours to under 10 if the paddock had been watered using the channel rather than the bankless channel, and drained into a farm drain.

Figure 4: Actual water depths (blue lines (i.e. top lines)) at the supply end of a 3.5ha contour basin that was (1) irrigated and drained using a bankless channel (left) and (2) irrigated from the farm channel and drained to a field drain (right). Orange lines (i.e. lines that drop the most) show modelled hydrographs for the same irrigations with water supplied from a farm channel and drained to a field drain in both cases.

Conclusion

The bankless channel is the principal cause of excessively long ponding times in basin layouts. The fact that water backs up in contour bays through the bankless channel has been clearly demonstrated, and the effect of this on drainage times has been measured. A pre-condition for consistently achieving high yields and returns per ML in basin surface irrigation systems will be to design and manage them so water is on and off bays in less than 10 hours. Both the field trials and the irrigation modelling show this is clearly possible if bays are supplied individually from a channel, rather than a bankless channel, and not drained into a filling downstream bay.

If high yields are to be consistently achieved from frequently irrigated crops in basin layouts, then either the bays need to be supplied and drained individually, or bays need to be skip watered.

References

Austin, N. R. & Prendergast, J. B. (1997). Use of kinematic wave theory to model irrigation on cracking soil. Irrigation Science 18: 1-10.

Bautista, E., Clemmens, A. J. & Strelkoff, T. S. (2009). Optimal and post irrigation volume balance infiltration parameter estimates for basin irrigation. Journal of Irrigation and Drainage Engineering (ASCE) 135(5).

Campbell, B. (1989).Report on bay hydraulics studies at Kerang: 1984 to 1988. Kerang, Vic.: Rural Water Commission of Victoria.

Lavis, A., Maskey, R. & Qassim, A. (2006). Border-check Irrigation Design. In Border-check Irrigation Design(Eds A. Lavis and R. Q. Maskey, A.). Melbourne, Victoria: State of Victoria, Department of Primary industries.

North, S. H., Griffin, D., Grabham, M. & Gillies, M. (Eds) (2010). Improving the performance of basin irrigation layouts in the southern Murray-Darling Basin. Darling Heights, Queensland: CRC Irrigation Futures.

Strelkoff, T., Clemmens, A. J. & Schmidt, B. V. (Eds) (1998). SRFR. Computer program for simulating flow in surface irrigation: furrows-basins-borders. Phoenix, Arizona: U.S. Water Conservation laboratory.

Swinton, R. (1994). Designing and constructing contour irrigation layouts. In Salt Action information sheet Deniliquin, NSW: NSW Agriculture.

Acknowledgements

The research undertaken for this project was made possible by funding from the Australian Government through the Rural R&D for Profit Program, with co-funding from Cotton Research & Development Corporation and AgriFutures Australia. The work would not have been possible with the same level of confidence in water flow measurement without the generous loan of three SlipMeters™ from Rubicon Water. The generous cooperation and assistance of Michael McBurnie in conducting the field work is also gratefully acknowledged.

Contact details

Sam North
449 Charlotte St, Deniliquin, NSW, 2710
03 5881 9926
samuel.north@dpi.nsw.gov.au