- Email: [email protected]
Particle circulation in irrigation reservoirs: The role of filter backwash reject on filter clogging
Agricultural Water Management 158 (2015) 139–144
Contents lists available at ScienceDirect
Agricultural Water Management journal homepage: www.elsevier.com/locate/agwat
Short communication
Particle circulation in irrigation reservoirs: The role of filter backwash reject on filter clogging Ana Milstein a,∗ , Mordehai Feldlite b a b
Agricultural Research Organization (ARO), Fish and Aquaculture Research Station Dor, M.P. Hof HaCarmel 30820, Israel Israel Water Workers Association, Eyal, mid-Sharon 45480, Israel
a r t i c l e
i n f o
Article history: Received 1 February 2015 Received in revised form 10 April 2015 Accepted 3 May 2015 Keywords: Filter backwash reject Clogging Irrigation reservoir Particle circulation Wastewater reservoir Water management
a b s t r a c t The improvement of the quality of treated wastewater allowed increased zooplankton populations in reservoirs that store water for irrigation, causing severe clogging problems in irrigation systems. To cope with the clogging problem we started a research program on the relationships between filter clogging and particle distribution in irrigation reservoirs. The present study targets the water and particles circulation between the reservoir and its bank filters, in order to evaluate potential management procedures to avoid filter clogging. Two reservoirs with different management were selected, in one water for irrigation is removed from under the surface and in the other from over the reservoir bottom. Profiles of temperature, oxygen, time to clog filters of different pore and amount of suspended solids retained by each such filter were measured. Conclusions: (1) Returning backwash reject into the reservoir recovers important amounts of water but also re-introduces clogging-size particles. (2) In a thermally stratified reservoir where water for irrigation is removed from the epilimnion, a daily short-circuit of 10% of the large particles (>150 m) present in the deep epilimnion occurred between reservoir, irrigation filters and backwash reject. (3) In a thermally stratified reservoir where water for irrigation is removed from the hypolimnion particle concentration in removed water was notably lower and the daily short-circuit did not occur. (4) Removing particles of the backwash before returning the water into the reservoir would avoid shortcircuiting of particles and re-introducing live copepods that may reproduce in the reservoir. (5) Removing particles through sedimentation requires a retaining time of at least half an hour before returning the water into the reservoir, and a way to transport away from the reservoir and dispose the sedimented material. (6) Returning water into the hypolimnion is a management option to recover water while avoiding/reducing the backwash reject negative effect on filter clogging. © 2015 Elsevier B.V. All rights reserved.
1. Introduction In arid and semi-arid regions water availability is limited and wastewater turns into an important water resource. In Israel over 75% of the municipal wastewater is reused, most of it for agricultural irrigation (The Water Agency – Israel, 2014). During the years different regulations requested improved quality of the treated effluents, the major ones being the requirement for secondary treatment (BOD < 20 mg/l) in 1995 and for tertiary treatment (nutrient removal) in 2005 (Juanicó, 2008), the latter still being gradually implemented. The improvement of the quality of treated wastewater allowed increased zooplankton populations in reservoirs that store water for irrigation, mainly
∗ Corresponding author. Tel.: +972 4 6390651x108; fax: +972 4 6390652. E-mail addresses: [email protected] (A. Milstein), [email protected] (M. Feldlite). http://dx.doi.org/10.1016/j.agwat.2015.05.002 0378-3774/© 2015 Elsevier B.V. All rights reserved.
copepods (length about 200–1000 m) and cladocerans (length about 300–3000 m). These particles may cause severe clogging problems in irrigation systems, especially in the drip irrigation ones, which are the systems used in over half of the irrigated area of the country (OECD, 2011). At present some reservoirs deliver their water directly to the irrigation systems, which at their head have filters of 130–200 m pore. But in most reservoirs another 130–200 m pore filter battery was added at the banks of the reservoirs and the filtered water is only then sent to irrigation. This avoids/reduces clogging in the irrigation systems transferring the problem to the storage reservoirs, which must overcome it to deliver the required amounts and quality of water. To cope with the increased clogging problem we started a research program on the relationships between filter clogging and particle distribution in irrigation reservoirs that receive secondarily treated wastewater. As a first step we studied relationships between clogging and particle size distribution in a range of reservoirs with different characteristics and water management, to
140
A. Milstein, M. Feldlite / Agricultural Water Management 158 (2015) 139–144
examine the possibility of removing large zooplanktonic organisms in order to avoid clogging of irrigation filters (Milstein and Feldlite, 2014). Then we made a more detailed study of particle distribution in relation to thermal stratification development in one such reservoir, to learn about changes through the irrigation season that would affect the convenience of removing water from one depth or another in order to avoid clogging of irrigation filters (Milstein and Feldlite, 2015). The present study targets the water and particles circulation between the reservoir and its bank filters, in order to evaluate other potential management procedures to avoid filter clogging. Based on our previous studies two reservoirs with different management were selected, in one water for irrigation is removed from under the surface and in the other from over the reservoir bottom.
2. Materials and methods 2.1. The reservoirs The work was carried out during July 2011 in two reservoirs that store secondarily treated wastewater for irrigation. Both reservoirs are deep enough to develop seasonal stratification, have plastic covered bottoms, wastewater enters into the upper reservoir water layers, and removed water passes through a battery of 130 m pore discs filters installed on the bank before it is sent to the 130 m pore filters at the head of the irrigation system. Each reservoir was sampled on one day representative of the conditions developed during the peak irrigation season. That includes sunny days, high temperature, and daily western breeze that moves surface water eastward inducing a westward undercurrent over the thermocline. Galon is a 6.4 ha surface, 7.5 m depth and 480,000 m3 reservoir when full (at sampling time water depth was only 5.5 m due to water removal for irrigation). Most of its reclaimed wastewater enters all the year round in its SE side, and few amounts from an oxidation pond enter in its E side. The outlet, located in the W side of the reservoir, is made of a pipe suspended from a raft that allows pumping water out from under the surface, its opening generally being set at the beginning of the irrigation season at 2–3 m depth. The backwash reject of the irrigation filters at the bank is returned into the reservoir in its SE side, near the main wastewater input pipe. Each filter backwash event returns 9 m3 of water to the reservoir. Backwash cleaning is generally set to be performed when a fixed difference of pressure of 0.6 atmosphere is reached. The time it takes to reach to this difference of pressure depends on the amount of particles present in the water. By sampling time the cleaning episodes occurred each 30 . Seven clogging episodes due to zooplankton that required chemical treatment occurred during the 2011 season. Mezer is an 8.3 ha surface, 11 m depth and 550,000 m3 reservoir when full (at sampling time water depth was only 7 m due to water removal for irrigation). Reclaimed wastewater enters in its W side. The outlet is also located in the W side, pumping water out from over the reservoir bottom. The backwash reject of the irrigation filters is returned into the reservoir in its NE side, far from the outlet and wastewater input pipe. Each filter backwash event returns 8 m3 of water to the reservoir. Backwash cleaning is generally set to be performed when a fixed difference of pressure of 0.6 atmosphere is reached. Three clogging episodes due to zooplankton that required chemical treatment occurred during the 2011 season. During the second half of July 2011 an aerator worked continually over the output area to mix the water column and increase its oxygen content, which also re-suspended mud from the bottom. Then, backwash cleaning frequency was increased and set to be performed each 10 . After 2 weeks the aerator was stopped because it re-suspended too
much mud. Our field work was performed by the end of the mixing period. 2.2. Field work Sampling was carried out from a boat in three stations following the west-east axis of each reservoir. In each station temperature and dissolved oxygen (DO) profiles were measured each 0.5 m from surface to bottom of the water column. In three of the six stations a Clogging Potential Meter (CPS, Sagi et al., 1996; Feldlite and Yechiely, 2011) with net filters of 150 m, 100 m, 60 m and 33 m was used to measure time to clog each net. If clogging did not occur after 5–7 min filtration was stopped. Good water quality for irrigation is considered when clogging time of the 150 m and 100 m nets is at least 5 min. The CPS was also used to collect the particles retained by each filter for suspended solids analysis. Besides the samples collected within the reservoirs, backwash reject of both reservoirs and input wastewater of Galon were also sampled. Farmers provided the data of the reservoirs and their water management (volume, daily amounts entering and removed, filter backwash frequency, etc.). 2.3. Laboratory work Samples were sent to a specialized laboratory for analysis of total suspended solids (TSS) and suspended solids retained in each CPS net (SS > 150 m, SS > 100 m, SS > 60 m and SS > 33 m). Samples of backwash reject were brought to the laboratory to measure particles sedimentation time. The backwash reject was shacked, 1 l was poured into a graduated sedimentation cone at time zero, the amount of sedimented material was measured at periodic intervals, and samples of the supernatant and sedimented material were observed under microscope. 3. Results and discussion 3.1. Reservoir Galon Fig. 1 presents the temperature, dissolved oxygen and particlesize distribution with depth in reservoir Galon. The temperature and oxygen profiles show that the water column was stratified, with an upper 2.5 m deep warm oxygenated layer (epilimnion), a 1 m deep transition zone (thermocline) and a lower 2 m deep cold anoxic layer (hypolimnion). Particles smaller than 33 m dominated throughout the water column. Particles larger than 100 m, which can potentially clog the irrigation filters, were more abundant in the deep epilimnion over the thermocline. Considering the water volume and suspended solids concentration of each layer, it was estimated that there were 16 kg of particles larger than 150 m in the 1.5 m upper epilimnion, 45 kg in the 1 m deep lower epilimnion and 18 kg in the 3 m deep hypolimnion (including transition zone). Water for irrigation was removed from 2 m under the surface, coinciding with the particle rich layer of the deep epilimnion. Fig. 2 presents the circulation of water and particles of different size through the reservoir, estimated on a daily basis. On the sampling day 14,400 m3 of treated wastewater containing 95 kg of TSS (6.6 g/m3 ) entered the reservoir and over 6700 m3 of water containing 65 kg of TSS (9.7 g/m3 ) were removed for irrigation. After passing the reservoir’s bank filter battery, 94% of that water with 32 kg of particles smaller than 130 m were directed to the irrigation system. The remaining 430 m3 were the backwash reject that contained 33 kg (76 g/m3 ) of clogging-size particles. This means that half of the TSS removed from the west side of the reservoir were potential clogging particles that returned to the epilimnion in the east side of the reservoir, accumulated in the deep
A. Milstein, M. Feldlite / Agricultural Water Management 158 (2015) 139–144
141
Fig. 1. Reservoir Galon: temperature and dissolved oxygen vertical distribution in three stations (West-output, Center, East), vertical distribution of suspended solid particles of different size at the West-output station, and depth of water removal.
epilimnion over the thermocline, and were transported westward over the thermocline by the wind induced undercurrent where they were pumped out again and again. Considering only the particles larger than 150 m, 5.3 kg were removed with the water pumped out (0.78 g/m3 ), none went to the irrigation system, and the same 5.3 kg were returned to the reservoir, representing a daily shortcircuit of 10% of the large particles present in the deep epilimnion. 3.2. Reservoir Mezer At the time of sampling in Mezer reservoir an aerator had been mixing the water column in the output area for 2 weeks. Thus, samples were taken in the output area under the direct aerator effect, about 20 m toward the reservoir center outside the aerator affected area, and far East in front of the filter backwash reject entrance. Fig. 3 presents the temperature, dissolved oxygen and particlesize distribution with depth in reservoir Metzer. Away from the area under the aerator mixing effect the water column was stratified, particles smaller than 33 m dominated throughout the water column, and particles larger than 100 m were more abundant in
the epilimnion. In the output area the turbulence produced by the aerator led to homogeneous temperature and improved oxygen through the water column, but also re-suspended large amounts of all size particles from the bottom. Since water for irrigation was removed from over the pond bottom the water looked muddy and farmers increased the backwash frequency to 6 times per hour. Fig. 4 presents the circulation of water and particles of different size through the reservoir, estimated on a daily basis. On the sampling day no reclaimed water was entering the reservoir. The aerator re-suspended sediments and mixed the entire water column only locally, while about 90% of the reservoir presented a thermocline at 2 m depth. In the entire 2 m deep epilimnion there were only 4 kg of particles larger than 150 m, in the 5 m deep hypolimnion another 4 kg of them, and in the small mixed up area 2 kg. From the pond bottom of the mixed area 6500 m3 /day of water containing 172 kg of TSS (26.5 g/m3 ) were removed for irrigation on that day. After passing the filtering system, 82% of that water with almost 150 kg of particles smaller than 130 m were directed to the irrigation system. The remaining 1150 m3 were the backwash reject
Fig. 2. Reservoir Galon: circulation of water and particles of different size, estimated on a daily basis. TSS = total suspended solids, SS = suspended solids.
142
A. Milstein, M. Feldlite / Agricultural Water Management 158 (2015) 139–144
Fig. 3. Reservoir Metzer: temperature and dissolved oxygen vertical distribution in three stations (West-output, West, East), vertical distribution of suspended solid particles of different size at the West-output and East stations, and depth of water removal.
that contained 25 kg of clogging-size particles (21.6 g/m3 ). The daily volume of the backwash water poured back into the reservoir was almost 20% of the water removed for irrigation on that day, reintroducing the 25 kg of particles larger than 130 m. Considering only the particles larger than 150 m, 1 kg were removed with the
water pumped out (0.1 g/m3 ), none went to the irrigation system, and the same 1 kg was returned to the eastern stratified side of the reservoir where the particles accumulated in the epilimnion. No short-circuit occurred since water was removed from and returned to different layers that do not mix. Had the aerator not been working
Fig. 4. Reservoir Metzer: circulation of water and particles of different size, estimated on a daily basis. TSS = total suspended solids, SS = suspended solids.
A. Milstein, M. Feldlite / Agricultural Water Management 158 (2015) 139–144
in the output area the whole reservoir would have been stratified, as measured in the East side. In that case water pumped out would not have been turbid and farmers would not have increased filter backwash frequency, so that less water and particles would have been removed from and returned into the reservoir. Even assuming the high backwash frequency, the removal of 6500 m3 water from the hypolimnion would have contained an order of magnitude less particles than those measured under the aerator mixing effect: 73 kg of TSS (11.2 g/m3 ), of which only 0.1 kg of particles larger than 150 m (0.016 g/m3 ). 3.3. Particles in the backwash reject The time it takes particles in the backwash reject of the irrigation filters to sediment was measured as an important parameter to consider in potential management procedures to avoid input of such particles back into the reservoir. In one sample it took 20 for particles in 1 l of backwash reject to concentrate into a volume of 0.5 ml, 35 to reach 1 ml, and after 80 that volume still was the same. In another sample it took 30 for particles to concentrate in 1 ml and an hour to reach 1.4 ml. Thus, as a first approximation it can be assumed that in 30 most of the particles in the backwash reject would sediment. Observation under microscope of the supernatant in the sedimentation cone showed small particles of detritus, while the sedimented material had large detritus particles and many copepods (Mesocyclops arcanus) dead and alive. The copepods alive were removed and kept in a petri dish, where many of them were still alive 1 month later. Moreover, females carrying eggs were observed, which indicates that the reproduction potential of the copepods would not be affected by mechanical damage through the pumping circuit. 3.4. General discussion Our previous research in the secondarily treated wastewater reservoirs showed strong differences in irrigation filter clogging potential when water was removed from the epilimnion or from the hypolimnion of thermally stratified reservoirs (Milstein and Feldlite, 2014, 2015). In the epilimnion sunlight allowed phytoplankton development, which in turn supported zooplankton populations including organisms large enough to clog irrigation filters. These and other large particles that settle down tend to accumulate in the deep epilimnion over the thermocline, so that removing water from this layer increases clogging episodes. In different reservoirs the epilimnion depth varied between 2 and 4 m. In the hypolimnion decomposition processes dominated so that it was generally anaerobic, hence very few potentially clogging zooplankton organisms occured but badly smelling compounds and sulfur/ferric/manganese bacteria often accumulated becoming a nuisance for irrigation. Together with this, a continuous water removal from the hypolimnion led to lower amounts of decomposing particles and nutrient accumulation even under high organic loading conditions. Based on these differences we selected for the present study two reservoirs where water for irrigation is removed from different layers. The addition of aerators to mix the water in Metzer reservoir was not under our control. The effects of water mixing on filter clogging will be examined in another article including observations in other reservoirs. In the reservoirs herein studied no assessment of the particles and temperature distribution were performed before our work. Without knowing, in both reservoirs water removal was from the depth with higher concentration of particles. In Galon this happened because water was removed from 2 m under the surface coinciding with the basis of the epilimnion. In Metzer this happened because of the sediment re-suspension and water mixing effect of
143
the aerator installed in the output area. In both cases the time to clog 150 m pore nets was over 5 min in all water layers except for the pumping out layer: in Galon that net clogged in 1 40 , in Metzer in 2 12 . Again, better water quality from the clogging point of view was observed when water was removed from the hypolimnion. Returning backwash reject into the reservoir recovers considerable amounts of water (an important issue in arid and semi-arid regions) but also re-introduces clogging-size particles. Among these particles are many live copepods, which are able to reproduce and thus perpetuate the filter clogging problem. Several management options may be considered in order to recover water while avoiding/reducing the backwash reject negative effect on filter clogging. One option would be returning the backwash into the hypolimnion, which should be done far from the output area. In the anaerobic hypolimnion the living organisms in the backwash reject would die and all particles would sediment on the reservoir bottom. This option implies lengthening the backwash reject pipe, since today in all reservoirs backwash is poured out high on the reservoir’s bank from where it flows down until reaching the water surface. In reservoirs where water is removed from under the surface this option should avoid short-circuit of particles during the warm season (most of the irrigation period) while the reservoir is deep and thermally stratified. In reservoirs where water is removed from over the bottom this option should have an effect only if the backwash reject pipe opens far from the output area. Other options to recover water while avoiding/reducing the backwash reject negative effect on filter clogging imply an intermediary step allowing removing particles before recovering the water. In some cases backwash can be directed toward a nearby reservoir or pond to allow sedimentation of particles and store the water until required. If that is not possible, then a sedimentation device with retention time of at least 30 installed on the reservoir’s bank would help removing backwash particles before returning the water into the reservoir. This option implies installing a sedimentation system on the banks that can treat in half an hour the amount of backwash produced in at least two filter cleaning events, and a way to transport away from the reservoir and dispose the sedimented material.
4. Conclusions • Returning backwash reject into the reservoir recovers important amounts of water but also re-introduces large quantity of clogging-size particles. • In a thermally stratified reservoir where water for irrigation is removed from the epilimnion, a daily short-circuit of 10% of the large particles (>150 m) present in the deep epilimnion occurred between reservoir, irrigation filters and backwash reject. • In a thermally stratified reservoir where water for irrigation is removed from the hypolimnion particle concentration in removed water was notably lower and the daily short-circuit did not occur. • It is important to remove particles of the backwash before returning the water into the reservoir, to avoid short-circuiting of particles and re-introducing live copepods that may reproduce in the reservoir. • Removing particles through sedimentation requires a retaining time of at least half an hour before returning the water into the reservoir, and a way to transport away from the reservoir and dispose the sedimented material. • Returning water into the hypolimnion (under stratified conditions) is a management option to recover water while avoiding/reducing the backwash reject negative effect on filter clogging.
144
A. Milstein, M. Feldlite / Agricultural Water Management 158 (2015) 139–144
Acknowledgements This research was funded by the Israel Water Works Association (IWWA). This work could not have been done without the collaboration of the farmers that facilitated our sampling work and provided management data. The authors wish to thank IWWA workers Zeev Yecheli for his technical support and Graciela Wigorski for carrying out the suspended solids analyses. References Feldlite, M., Yechiely, Z., 2011. On-line monitoring of clogging potential in irrigation systems. In: Shoshani, M., Shaviv, A. (Eds.), Abstract Book of the International Symposium on Sensing in Agriculture in Memory of Dahlia Greidinger. Technion, Haifa, Israel, pp. 89–90, http://gwri-ic.technion.ac.il/pdf/DG/2011/1.pdf (last accessed 18.03.14).
Juanicó, M., 2008. Israel as a case study. In: Jimenez, B., Asano, T. (Eds.), Water Reuse – An International Survey of Current Practice, Issues and Needs. IWA Publishing, pp. 483–502, Sci. Technical Report 20. Milstein, A., Feldlite, M., 2014. Relationships between clogging in irrigation systems and plankton community structure and distribution in wastewater reservoirs. Agric. Water Manage. 140, 79–86. Milstein, A., Feldlite, M., 2015. Relationships between thermal stratification in a secondarily treated wastewater reservoir that stores water for irrigation and filter clogging in the irrigation system. Agric. Water Manage. 153, 63–70. OECD, 2011. Water. In: OECD Environmental Performance Reviews: Israel 2011. Part II., pp. 103–124 (Chapter 4). Sagi, G., Yechiely, Z., Shisha, A., Alkon, A., Shrem, G., 1996. Clogging potential meter – Field devise o measure water quality and filtering requirements. Water Irrig. Bull. 355, 3–5 (in Hebrew). The Water Agency – Israel, 2014. http://www.water.gov.il/Hebrew/ WaterResources/Effluents/Pages/default.aspx (in Hebrew, last accessed 18.03.14).
Contents lists available at ScienceDirect
Agricultural Water Management journal homepage: www.elsevier.com/locate/agwat
Short communication
Particle circulation in irrigation reservoirs: The role of filter backwash reject on filter clogging Ana Milstein a,∗ , Mordehai Feldlite b a b
Agricultural Research Organization (ARO), Fish and Aquaculture Research Station Dor, M.P. Hof HaCarmel 30820, Israel Israel Water Workers Association, Eyal, mid-Sharon 45480, Israel
a r t i c l e
i n f o
Article history: Received 1 February 2015 Received in revised form 10 April 2015 Accepted 3 May 2015 Keywords: Filter backwash reject Clogging Irrigation reservoir Particle circulation Wastewater reservoir Water management
a b s t r a c t The improvement of the quality of treated wastewater allowed increased zooplankton populations in reservoirs that store water for irrigation, causing severe clogging problems in irrigation systems. To cope with the clogging problem we started a research program on the relationships between filter clogging and particle distribution in irrigation reservoirs. The present study targets the water and particles circulation between the reservoir and its bank filters, in order to evaluate potential management procedures to avoid filter clogging. Two reservoirs with different management were selected, in one water for irrigation is removed from under the surface and in the other from over the reservoir bottom. Profiles of temperature, oxygen, time to clog filters of different pore and amount of suspended solids retained by each such filter were measured. Conclusions: (1) Returning backwash reject into the reservoir recovers important amounts of water but also re-introduces clogging-size particles. (2) In a thermally stratified reservoir where water for irrigation is removed from the epilimnion, a daily short-circuit of 10% of the large particles (>150 m) present in the deep epilimnion occurred between reservoir, irrigation filters and backwash reject. (3) In a thermally stratified reservoir where water for irrigation is removed from the hypolimnion particle concentration in removed water was notably lower and the daily short-circuit did not occur. (4) Removing particles of the backwash before returning the water into the reservoir would avoid shortcircuiting of particles and re-introducing live copepods that may reproduce in the reservoir. (5) Removing particles through sedimentation requires a retaining time of at least half an hour before returning the water into the reservoir, and a way to transport away from the reservoir and dispose the sedimented material. (6) Returning water into the hypolimnion is a management option to recover water while avoiding/reducing the backwash reject negative effect on filter clogging. © 2015 Elsevier B.V. All rights reserved.
1. Introduction In arid and semi-arid regions water availability is limited and wastewater turns into an important water resource. In Israel over 75% of the municipal wastewater is reused, most of it for agricultural irrigation (The Water Agency – Israel, 2014). During the years different regulations requested improved quality of the treated effluents, the major ones being the requirement for secondary treatment (BOD < 20 mg/l) in 1995 and for tertiary treatment (nutrient removal) in 2005 (Juanicó, 2008), the latter still being gradually implemented. The improvement of the quality of treated wastewater allowed increased zooplankton populations in reservoirs that store water for irrigation, mainly
∗ Corresponding author. Tel.: +972 4 6390651x108; fax: +972 4 6390652. E-mail addresses: [email protected] (A. Milstein), [email protected] (M. Feldlite). http://dx.doi.org/10.1016/j.agwat.2015.05.002 0378-3774/© 2015 Elsevier B.V. All rights reserved.
copepods (length about 200–1000 m) and cladocerans (length about 300–3000 m). These particles may cause severe clogging problems in irrigation systems, especially in the drip irrigation ones, which are the systems used in over half of the irrigated area of the country (OECD, 2011). At present some reservoirs deliver their water directly to the irrigation systems, which at their head have filters of 130–200 m pore. But in most reservoirs another 130–200 m pore filter battery was added at the banks of the reservoirs and the filtered water is only then sent to irrigation. This avoids/reduces clogging in the irrigation systems transferring the problem to the storage reservoirs, which must overcome it to deliver the required amounts and quality of water. To cope with the increased clogging problem we started a research program on the relationships between filter clogging and particle distribution in irrigation reservoirs that receive secondarily treated wastewater. As a first step we studied relationships between clogging and particle size distribution in a range of reservoirs with different characteristics and water management, to
140
A. Milstein, M. Feldlite / Agricultural Water Management 158 (2015) 139–144
examine the possibility of removing large zooplanktonic organisms in order to avoid clogging of irrigation filters (Milstein and Feldlite, 2014). Then we made a more detailed study of particle distribution in relation to thermal stratification development in one such reservoir, to learn about changes through the irrigation season that would affect the convenience of removing water from one depth or another in order to avoid clogging of irrigation filters (Milstein and Feldlite, 2015). The present study targets the water and particles circulation between the reservoir and its bank filters, in order to evaluate other potential management procedures to avoid filter clogging. Based on our previous studies two reservoirs with different management were selected, in one water for irrigation is removed from under the surface and in the other from over the reservoir bottom.
2. Materials and methods 2.1. The reservoirs The work was carried out during July 2011 in two reservoirs that store secondarily treated wastewater for irrigation. Both reservoirs are deep enough to develop seasonal stratification, have plastic covered bottoms, wastewater enters into the upper reservoir water layers, and removed water passes through a battery of 130 m pore discs filters installed on the bank before it is sent to the 130 m pore filters at the head of the irrigation system. Each reservoir was sampled on one day representative of the conditions developed during the peak irrigation season. That includes sunny days, high temperature, and daily western breeze that moves surface water eastward inducing a westward undercurrent over the thermocline. Galon is a 6.4 ha surface, 7.5 m depth and 480,000 m3 reservoir when full (at sampling time water depth was only 5.5 m due to water removal for irrigation). Most of its reclaimed wastewater enters all the year round in its SE side, and few amounts from an oxidation pond enter in its E side. The outlet, located in the W side of the reservoir, is made of a pipe suspended from a raft that allows pumping water out from under the surface, its opening generally being set at the beginning of the irrigation season at 2–3 m depth. The backwash reject of the irrigation filters at the bank is returned into the reservoir in its SE side, near the main wastewater input pipe. Each filter backwash event returns 9 m3 of water to the reservoir. Backwash cleaning is generally set to be performed when a fixed difference of pressure of 0.6 atmosphere is reached. The time it takes to reach to this difference of pressure depends on the amount of particles present in the water. By sampling time the cleaning episodes occurred each 30 . Seven clogging episodes due to zooplankton that required chemical treatment occurred during the 2011 season. Mezer is an 8.3 ha surface, 11 m depth and 550,000 m3 reservoir when full (at sampling time water depth was only 7 m due to water removal for irrigation). Reclaimed wastewater enters in its W side. The outlet is also located in the W side, pumping water out from over the reservoir bottom. The backwash reject of the irrigation filters is returned into the reservoir in its NE side, far from the outlet and wastewater input pipe. Each filter backwash event returns 8 m3 of water to the reservoir. Backwash cleaning is generally set to be performed when a fixed difference of pressure of 0.6 atmosphere is reached. Three clogging episodes due to zooplankton that required chemical treatment occurred during the 2011 season. During the second half of July 2011 an aerator worked continually over the output area to mix the water column and increase its oxygen content, which also re-suspended mud from the bottom. Then, backwash cleaning frequency was increased and set to be performed each 10 . After 2 weeks the aerator was stopped because it re-suspended too
much mud. Our field work was performed by the end of the mixing period. 2.2. Field work Sampling was carried out from a boat in three stations following the west-east axis of each reservoir. In each station temperature and dissolved oxygen (DO) profiles were measured each 0.5 m from surface to bottom of the water column. In three of the six stations a Clogging Potential Meter (CPS, Sagi et al., 1996; Feldlite and Yechiely, 2011) with net filters of 150 m, 100 m, 60 m and 33 m was used to measure time to clog each net. If clogging did not occur after 5–7 min filtration was stopped. Good water quality for irrigation is considered when clogging time of the 150 m and 100 m nets is at least 5 min. The CPS was also used to collect the particles retained by each filter for suspended solids analysis. Besides the samples collected within the reservoirs, backwash reject of both reservoirs and input wastewater of Galon were also sampled. Farmers provided the data of the reservoirs and their water management (volume, daily amounts entering and removed, filter backwash frequency, etc.). 2.3. Laboratory work Samples were sent to a specialized laboratory for analysis of total suspended solids (TSS) and suspended solids retained in each CPS net (SS > 150 m, SS > 100 m, SS > 60 m and SS > 33 m). Samples of backwash reject were brought to the laboratory to measure particles sedimentation time. The backwash reject was shacked, 1 l was poured into a graduated sedimentation cone at time zero, the amount of sedimented material was measured at periodic intervals, and samples of the supernatant and sedimented material were observed under microscope. 3. Results and discussion 3.1. Reservoir Galon Fig. 1 presents the temperature, dissolved oxygen and particlesize distribution with depth in reservoir Galon. The temperature and oxygen profiles show that the water column was stratified, with an upper 2.5 m deep warm oxygenated layer (epilimnion), a 1 m deep transition zone (thermocline) and a lower 2 m deep cold anoxic layer (hypolimnion). Particles smaller than 33 m dominated throughout the water column. Particles larger than 100 m, which can potentially clog the irrigation filters, were more abundant in the deep epilimnion over the thermocline. Considering the water volume and suspended solids concentration of each layer, it was estimated that there were 16 kg of particles larger than 150 m in the 1.5 m upper epilimnion, 45 kg in the 1 m deep lower epilimnion and 18 kg in the 3 m deep hypolimnion (including transition zone). Water for irrigation was removed from 2 m under the surface, coinciding with the particle rich layer of the deep epilimnion. Fig. 2 presents the circulation of water and particles of different size through the reservoir, estimated on a daily basis. On the sampling day 14,400 m3 of treated wastewater containing 95 kg of TSS (6.6 g/m3 ) entered the reservoir and over 6700 m3 of water containing 65 kg of TSS (9.7 g/m3 ) were removed for irrigation. After passing the reservoir’s bank filter battery, 94% of that water with 32 kg of particles smaller than 130 m were directed to the irrigation system. The remaining 430 m3 were the backwash reject that contained 33 kg (76 g/m3 ) of clogging-size particles. This means that half of the TSS removed from the west side of the reservoir were potential clogging particles that returned to the epilimnion in the east side of the reservoir, accumulated in the deep
A. Milstein, M. Feldlite / Agricultural Water Management 158 (2015) 139–144
141
Fig. 1. Reservoir Galon: temperature and dissolved oxygen vertical distribution in three stations (West-output, Center, East), vertical distribution of suspended solid particles of different size at the West-output station, and depth of water removal.
epilimnion over the thermocline, and were transported westward over the thermocline by the wind induced undercurrent where they were pumped out again and again. Considering only the particles larger than 150 m, 5.3 kg were removed with the water pumped out (0.78 g/m3 ), none went to the irrigation system, and the same 5.3 kg were returned to the reservoir, representing a daily shortcircuit of 10% of the large particles present in the deep epilimnion. 3.2. Reservoir Mezer At the time of sampling in Mezer reservoir an aerator had been mixing the water column in the output area for 2 weeks. Thus, samples were taken in the output area under the direct aerator effect, about 20 m toward the reservoir center outside the aerator affected area, and far East in front of the filter backwash reject entrance. Fig. 3 presents the temperature, dissolved oxygen and particlesize distribution with depth in reservoir Metzer. Away from the area under the aerator mixing effect the water column was stratified, particles smaller than 33 m dominated throughout the water column, and particles larger than 100 m were more abundant in
the epilimnion. In the output area the turbulence produced by the aerator led to homogeneous temperature and improved oxygen through the water column, but also re-suspended large amounts of all size particles from the bottom. Since water for irrigation was removed from over the pond bottom the water looked muddy and farmers increased the backwash frequency to 6 times per hour. Fig. 4 presents the circulation of water and particles of different size through the reservoir, estimated on a daily basis. On the sampling day no reclaimed water was entering the reservoir. The aerator re-suspended sediments and mixed the entire water column only locally, while about 90% of the reservoir presented a thermocline at 2 m depth. In the entire 2 m deep epilimnion there were only 4 kg of particles larger than 150 m, in the 5 m deep hypolimnion another 4 kg of them, and in the small mixed up area 2 kg. From the pond bottom of the mixed area 6500 m3 /day of water containing 172 kg of TSS (26.5 g/m3 ) were removed for irrigation on that day. After passing the filtering system, 82% of that water with almost 150 kg of particles smaller than 130 m were directed to the irrigation system. The remaining 1150 m3 were the backwash reject
Fig. 2. Reservoir Galon: circulation of water and particles of different size, estimated on a daily basis. TSS = total suspended solids, SS = suspended solids.
142
A. Milstein, M. Feldlite / Agricultural Water Management 158 (2015) 139–144
Fig. 3. Reservoir Metzer: temperature and dissolved oxygen vertical distribution in three stations (West-output, West, East), vertical distribution of suspended solid particles of different size at the West-output and East stations, and depth of water removal.
that contained 25 kg of clogging-size particles (21.6 g/m3 ). The daily volume of the backwash water poured back into the reservoir was almost 20% of the water removed for irrigation on that day, reintroducing the 25 kg of particles larger than 130 m. Considering only the particles larger than 150 m, 1 kg were removed with the
water pumped out (0.1 g/m3 ), none went to the irrigation system, and the same 1 kg was returned to the eastern stratified side of the reservoir where the particles accumulated in the epilimnion. No short-circuit occurred since water was removed from and returned to different layers that do not mix. Had the aerator not been working
Fig. 4. Reservoir Metzer: circulation of water and particles of different size, estimated on a daily basis. TSS = total suspended solids, SS = suspended solids.
A. Milstein, M. Feldlite / Agricultural Water Management 158 (2015) 139–144
in the output area the whole reservoir would have been stratified, as measured in the East side. In that case water pumped out would not have been turbid and farmers would not have increased filter backwash frequency, so that less water and particles would have been removed from and returned into the reservoir. Even assuming the high backwash frequency, the removal of 6500 m3 water from the hypolimnion would have contained an order of magnitude less particles than those measured under the aerator mixing effect: 73 kg of TSS (11.2 g/m3 ), of which only 0.1 kg of particles larger than 150 m (0.016 g/m3 ). 3.3. Particles in the backwash reject The time it takes particles in the backwash reject of the irrigation filters to sediment was measured as an important parameter to consider in potential management procedures to avoid input of such particles back into the reservoir. In one sample it took 20 for particles in 1 l of backwash reject to concentrate into a volume of 0.5 ml, 35 to reach 1 ml, and after 80 that volume still was the same. In another sample it took 30 for particles to concentrate in 1 ml and an hour to reach 1.4 ml. Thus, as a first approximation it can be assumed that in 30 most of the particles in the backwash reject would sediment. Observation under microscope of the supernatant in the sedimentation cone showed small particles of detritus, while the sedimented material had large detritus particles and many copepods (Mesocyclops arcanus) dead and alive. The copepods alive were removed and kept in a petri dish, where many of them were still alive 1 month later. Moreover, females carrying eggs were observed, which indicates that the reproduction potential of the copepods would not be affected by mechanical damage through the pumping circuit. 3.4. General discussion Our previous research in the secondarily treated wastewater reservoirs showed strong differences in irrigation filter clogging potential when water was removed from the epilimnion or from the hypolimnion of thermally stratified reservoirs (Milstein and Feldlite, 2014, 2015). In the epilimnion sunlight allowed phytoplankton development, which in turn supported zooplankton populations including organisms large enough to clog irrigation filters. These and other large particles that settle down tend to accumulate in the deep epilimnion over the thermocline, so that removing water from this layer increases clogging episodes. In different reservoirs the epilimnion depth varied between 2 and 4 m. In the hypolimnion decomposition processes dominated so that it was generally anaerobic, hence very few potentially clogging zooplankton organisms occured but badly smelling compounds and sulfur/ferric/manganese bacteria often accumulated becoming a nuisance for irrigation. Together with this, a continuous water removal from the hypolimnion led to lower amounts of decomposing particles and nutrient accumulation even under high organic loading conditions. Based on these differences we selected for the present study two reservoirs where water for irrigation is removed from different layers. The addition of aerators to mix the water in Metzer reservoir was not under our control. The effects of water mixing on filter clogging will be examined in another article including observations in other reservoirs. In the reservoirs herein studied no assessment of the particles and temperature distribution were performed before our work. Without knowing, in both reservoirs water removal was from the depth with higher concentration of particles. In Galon this happened because water was removed from 2 m under the surface coinciding with the basis of the epilimnion. In Metzer this happened because of the sediment re-suspension and water mixing effect of
143
the aerator installed in the output area. In both cases the time to clog 150 m pore nets was over 5 min in all water layers except for the pumping out layer: in Galon that net clogged in 1 40 , in Metzer in 2 12 . Again, better water quality from the clogging point of view was observed when water was removed from the hypolimnion. Returning backwash reject into the reservoir recovers considerable amounts of water (an important issue in arid and semi-arid regions) but also re-introduces clogging-size particles. Among these particles are many live copepods, which are able to reproduce and thus perpetuate the filter clogging problem. Several management options may be considered in order to recover water while avoiding/reducing the backwash reject negative effect on filter clogging. One option would be returning the backwash into the hypolimnion, which should be done far from the output area. In the anaerobic hypolimnion the living organisms in the backwash reject would die and all particles would sediment on the reservoir bottom. This option implies lengthening the backwash reject pipe, since today in all reservoirs backwash is poured out high on the reservoir’s bank from where it flows down until reaching the water surface. In reservoirs where water is removed from under the surface this option should avoid short-circuit of particles during the warm season (most of the irrigation period) while the reservoir is deep and thermally stratified. In reservoirs where water is removed from over the bottom this option should have an effect only if the backwash reject pipe opens far from the output area. Other options to recover water while avoiding/reducing the backwash reject negative effect on filter clogging imply an intermediary step allowing removing particles before recovering the water. In some cases backwash can be directed toward a nearby reservoir or pond to allow sedimentation of particles and store the water until required. If that is not possible, then a sedimentation device with retention time of at least 30 installed on the reservoir’s bank would help removing backwash particles before returning the water into the reservoir. This option implies installing a sedimentation system on the banks that can treat in half an hour the amount of backwash produced in at least two filter cleaning events, and a way to transport away from the reservoir and dispose the sedimented material.
4. Conclusions • Returning backwash reject into the reservoir recovers important amounts of water but also re-introduces large quantity of clogging-size particles. • In a thermally stratified reservoir where water for irrigation is removed from the epilimnion, a daily short-circuit of 10% of the large particles (>150 m) present in the deep epilimnion occurred between reservoir, irrigation filters and backwash reject. • In a thermally stratified reservoir where water for irrigation is removed from the hypolimnion particle concentration in removed water was notably lower and the daily short-circuit did not occur. • It is important to remove particles of the backwash before returning the water into the reservoir, to avoid short-circuiting of particles and re-introducing live copepods that may reproduce in the reservoir. • Removing particles through sedimentation requires a retaining time of at least half an hour before returning the water into the reservoir, and a way to transport away from the reservoir and dispose the sedimented material. • Returning water into the hypolimnion (under stratified conditions) is a management option to recover water while avoiding/reducing the backwash reject negative effect on filter clogging.
144
A. Milstein, M. Feldlite / Agricultural Water Management 158 (2015) 139–144
Acknowledgements This research was funded by the Israel Water Works Association (IWWA). This work could not have been done without the collaboration of the farmers that facilitated our sampling work and provided management data. The authors wish to thank IWWA workers Zeev Yecheli for his technical support and Graciela Wigorski for carrying out the suspended solids analyses. References Feldlite, M., Yechiely, Z., 2011. On-line monitoring of clogging potential in irrigation systems. In: Shoshani, M., Shaviv, A. (Eds.), Abstract Book of the International Symposium on Sensing in Agriculture in Memory of Dahlia Greidinger. Technion, Haifa, Israel, pp. 89–90, http://gwri-ic.technion.ac.il/pdf/DG/2011/1.pdf (last accessed 18.03.14).
Juanicó, M., 2008. Israel as a case study. In: Jimenez, B., Asano, T. (Eds.), Water Reuse – An International Survey of Current Practice, Issues and Needs. IWA Publishing, pp. 483–502, Sci. Technical Report 20. Milstein, A., Feldlite, M., 2014. Relationships between clogging in irrigation systems and plankton community structure and distribution in wastewater reservoirs. Agric. Water Manage. 140, 79–86. Milstein, A., Feldlite, M., 2015. Relationships between thermal stratification in a secondarily treated wastewater reservoir that stores water for irrigation and filter clogging in the irrigation system. Agric. Water Manage. 153, 63–70. OECD, 2011. Water. In: OECD Environmental Performance Reviews: Israel 2011. Part II., pp. 103–124 (Chapter 4). Sagi, G., Yechiely, Z., Shisha, A., Alkon, A., Shrem, G., 1996. Clogging potential meter – Field devise o measure water quality and filtering requirements. Water Irrig. Bull. 355, 3–5 (in Hebrew). The Water Agency – Israel, 2014. http://www.water.gov.il/Hebrew/ WaterResources/Effluents/Pages/default.aspx (in Hebrew, last accessed 18.03.14).