Hybrid Membrane Systems for Secondary Effluent Polishing for Unrestricted Reuse for Agricultural Irrigation Gideon Oron1,2,3, Amos Bick2, Leonid Gillerman1and Yossi Manor4 1) Ben-Gurion University of The Negev, The Institute for Desert Research, Kiryat Sde-Boker, 84990, Israel. (E-mail:
[email protected],
[email protected]>, respectively) 2) Ben-Gurion University of The Negev, The Department of Industrial Engineering and Management, BeerSheva, 84105, Israel. (E-mail:
[email protected] ) 3) The Grand Water Research Institute, Technion Haifa 32000, Israel 3) Central Virological Laboratory, Sheba Medical Center, Tel-Hashomer 52621, Israel. (E-mail:
[email protected] )
Abstract Field experiments are in progress for secondary wastewater upgrading for unrestricted utilization for agricultural irrigation. The integrative approach of secondary effluent polishing is based on using a hybrid UltraFiltration (UF) and Reverse Osmosis (RO) 3 membrane pilot system with a capacity of around 1 m /hr. The UF effluent is used to feed the RO membranes. The RO permeate is subsequently applied for vegetables irrigation. Field results indicate the importance of the UF component in the removal of the organic matter and the pathogens that are still contained in the secondary effluent. Under specific conditions, when the dissolved solids content is relatively low, regarding sanitary and health aspects, the UF effluent can be applied for unrestricted irrigation. During the RO stage most nutrients are removed, allowing applying the effluent without jeopardizing the soil fertility and the aquifers. Preliminary economic assessment indicates that the extra cost 3 for effluent polishing via the UF stage only is in the range of 5 to 15 US cents/m . The 3 extra cost for the RO stage is as well assessed at 10 to 25 US cents/m . The additional cost depends to a large extent on the quality of the incoming raw secondary effluent and local requirements at the command region. Keywords Effluent; Hybrid membrane systems; Irrigation; Unrestricted Reuse
INTRODUCTION High quality water is one of the main production factor in most sustainable agricultural production systems. However, due to the spiraling shortages, mainly due to intensive exploitation of groundwater from aquifers and steady reduction in natural recharge, particularly in arid zones – alternative sources and improved water treatment methods are searched (Asano and Levine, 1995; Oron, 1996). The gap between supply and demand can be closed primarily by several strategic directions: (i) importing water from external sources; and (ii) further development and reuse of the additional water sources and under specific conditions, treatment of the low quality waters to acceptable levels (Oron et al., 1996). Potential advanced water treatment includes the use of membrane technology (Bick et al., 1996), mainly for saline water and seawater. However, membranes can also be used for advanced wastewater treatment. Membrane treatment of effluent appears attractive since
it is a steady source of water. Brine disposal is of serious concern due to potential environmental nuisances (Mickley, 2001). Effluent treatment and reuse has attracted a great deal of attention during the past decade (Oron et al., 1998). Improved technologies for the removal of particles, turbidity, bacteria and cysts reduction without disinfection are based on the use of membranes, and include mainly Micro-Filtration (MF) and UltraFiltration (UF). The advantages of MF or UF for organic matter removal with selected salt removal by Reverse Osmosis (RO) membranes make integrated systems promising (Bick and Oron, 2000). When saline water is adequately used for irrigation, it can improve agriculture production of crops, particularly in orchards (Hoffman et al., 1986; Oron et al., 1999; Hill and Koenig, 1999). Its’ use for crop production offers several benefits: (i) reuse of large amounts of saline water during the entire year, with minimal environmental risk to groundwater, and; (ii) a premium market price for the fruits and vegetable products coupled with a high content of total soluble solids and an extended shelf life, due to the plants’ adaptation to the stressful growing conditions (Mizrahi and Pasternak, 1985). RO membrane technologies are commercialized at a rate that has far exceeded their deployment. There is scarce data concerning optimal membrane selection and facility design for effluent treatment (Heller et al., 1998). Implementing of RO knowledge selection becomes an important issue in attaining the goals of cost reduction and quality improvement.
MATERIAL AND METHODS The experiments are in progress in the field of Kibbutz Chafets-Chaim, located 40 km west of Jerusalem, Israel. Due to local restrictions for reuse, the secondary wastewater is upgraded using a mobile membrane pilot unit. The pilot plant consists of two ring filters (100 and 20 micron respectively), a feeding pump, one module with a single Spiral Wound (SW) UF membrane, two modules with four Spiral Wound (SW) RO membranes, a flushing system and a chemical cleaning unit. A centrifugal pump delivered the filtered secondary effluent to the UF system. The secondary effluent is treated in a UF membrane stage and fraction of the concentrate is re-circulated and blended with additional feed effluent. A bleed valve was located in the circulation loop to control the concentrate flow. The UF permeate, which was collected in a reservoir, was used as a feed to RO unit. Backwash was employed to control the UF fouling. A programmable controller (PLC) implemented the sequence of events associated with backwash frequency and duration. Sodium hydroxide was applied at end of the day (pH=14) in order to clean the UF element. UF membrane type was NIROSOFT RM10-8, 8040 spiral wound. The UF pilot plant utilizes cross flow and dead-end processes and the membrane surface area is 35.5 m2. The experiments were performed at pH 8, feed pressure was 4 Bars and the operating temperature was 30C. The feed water (secondary effluent) is taken directly from 100-micron filter at Chafets-Chaim agricultural region (3.5 million m3 of recycled wastewater. The RO unit consists of a two-stage array with four 4" elements (Filmtec type FT30 4040 thin film composite membrane in spiral wound configuration). The membrane coating is remarkable in that it has surface pores controlled to a diameter of approximately 150 angstroms (Huisman, 1993). Except for the chemical cleaning process, the pilot runs automatically and the feed was pretreated with acid (HCl) without antiscalant. There was no post-treatment: permeate was delivered to a reservoir and supplied to the customers. The RO elements were not back-washed: the membranes were cleaned at 70 hours of operation with citric acid.
2
IRRIGATION Several issues were examined during the irrigation study: (i) verifying the effects of Onsurface Drip Irrigation (ODI), applying intermittently secondary effluent and RO permeate on the yield, and;(ii) assessing salinity distribution in the soil as function of the depth and effluent quality application. The possibility of using saline secondary effluent and RO permeate combined with ODI technology for vegetable irrigation was examined in a bell pepper field experiment. The peppers plants were arranged in beds of 3 rows, and a row spacing of 0.93 m. The research began on March 2000 with the following treatments: (i) irrigation with secondary effluent (ODI technology), (ii) irrigation with secondary effluent [Subsurface Drip Irrigation (SDI) technology]), and; (iii) irrigation with RO permeate and secondary effluent (ODI technology). Soil properties were characterized by a standard gravimetric method. Soil salinity was determined by a common method of measuring the Electrical Conductivity (EC) of saturated extracts. Soil samples for moisture and salinity assessment were taken at 0, 30 and 60 cm depths from the equidistant on both sides of the plants, near the emitters and in the middle between two adjacent emitters. The emitter spacing was 1.0 m and one drip lateral served each pepper row.
Table 1. Pilot plant water quality parameters Quality
Selected recorded flow UF UF RO Permeate Brine Permeate Cpufi Cbufi Cproi 8.2 8.3 7.3 0 63 0 15 42 4 24 600 0 370 370 8.5 87 99 7.1 357 382 15 5.6 5.6 0.3 20.2 20.6 0.4 275 350 12 32 39 1.4 28.9 35 1.2 42 42 0.7
i, RO Brine Cbroi 8.2 0 12 100 910 168 710 12.7 41.3 605 58 63 75
Parameter i, mg/l
Feed Cfri
pH )-( TSS BOD COD ClSO4= HCO3= NH4+ PO4 Na+ K+ Ca++ Mg++
8.48 22 27 180 370 99.4 378 5.6 20.2 275 32 35.3 42
EC, dS/m
2.3
2.3
2.3
0.001
3.4
99
Fecal Coliforms, CFU/100ml* Coliphages F+, PFU/100ml Somatic Coliphages, PFU/100ml
23 *103
0
34*103
0
3.3*102
100
9.0*101
1.0*101
4.0*101
0
1.0*101
91
3.8*103
5.0*101
3.4*103
0
2.8*101
99
* - CFU and PFU – Colony Forming Units and Plaque Forming Unit, respectively
3
Total RO Rejection, % -100 85 100 98 93 96 95 98 96 96 97 98
RESULTS Effluent quality Table 1 presents typical water quality data of the pilot plant performance (Bick et al., 2001). UF treatment provided effective pretreatment for the RO unit. It should be noted that the UF membrane did not removed total organic compounds but on the other hand it is a disinfecting step, removing completely fecal coliforms. Based on the data recorded it seems that UF membrane barrier guarantee over 4 log fecal- coliforms removal. The tests concerning Somatic Coliphages and Coliphages may indicate the existence of bacteria colonies on the membrane surface that can be removed by sodium hydroxide solution to prevent bacterial re-growth. RO permeate quality was fairly constant and ionic removal range was 93 -98%: this is superior to agricultural irrigation regulation (WHO, 1989; EPA, 1992; Shuval et al., 1997). Odors were detected in the RO permeate and it is speculated that it indicated the presence of H2S compounds. Total rejection was calculated according to equation 1. i =100( Cfri – Cproi)/(Cfri)
(1)
where i is the rejection of i parameter, %; Cfri is the feed concentration of i parameter, and Cproi is the RO permeate concentration of the i parameter. Membrane performance Figure 1 presents the UF permeate flux performance relative to start up. Figure 2 and figure 3 show the changes in normalized RO permeate flow and salt rejection in the first 100 hours of operation. After 67 operating hours the membranes lost about 35 percent of it’s permeate flow because of too high system recovery, and citric acid cleaning recovered the system performance. 4.0
UF Performance
Nor malize d P ermeate Flux relative to start up
3.5
Chafets -Chaim, Is rael 3/2000-10/2000
3.0 2.5 2.0 1.5 1.0 0.5 0.0
30
40
50
60 70 80 90 Filtration Time, hours
100
110
120
Figure 1. Normalized UF permeate flow relative to start up
The general expression for the i normalized effluent quality parameter Gi is given by (2): Gi = Goi/Gti
(2)
Where Goi and Gti are the values of the specific quality control and/or operating parameters at the beginning of the system operation (t=0) and at time t, respectively
4
RO 4040 Normalized Performance Chafets-Chaim, Israel 3/2000 - 10/2000
Normalized flux relative to start up
1.2
Citric Acid Cleaning
1 0.8 0.6 0.4 30
40
50
60
70
80
90
Filtration Time, hours
100
110
120
Figure 2. Normalized RO permeate flow relative to start up
After 100 operating hours the RO membranes lost about 2 percent of its salt rejection. The following conclusions ensued from the study: (i) UF membrane is an excellent pre-treatment alternative for secondary effluent feed to RO, (ii) cleaning of the RO membrane was necessary because of too high recovery operation mode, (iii) after cleaning, the RO membranes kept the original normalized permeate flow, and; (iv) the change in normalized salt rejection is not necessarily associated with biological growth and the suspected cause is the adsorption of colloidal heavy metal matter on the membrane.
RO 4040 Normalized Performance Chafets -Chaim, Is rael 3/2000-10/2000 Normalized salt rejection relative to start up
1.05 1.04 1.03 1.02 1.01 1.00 0.99 0.98 0.97 0.96 0.95 30
40
50
60
70
80
90
Filtration Time, hours
100
110
120
Figure 3. Normalized RO salt rejection relative to start up
Salinity distribution in the soil Salinity distribution in the soil as expressed by the EC is shown at 27 July 2000 (Figure 4). This figure shows the change in salinity with depth and time: (i) the relatively low salinity in the active root zone of the crop with SDI technology compare to ODI technology, (ii) after ODI and SDI, salt accumulated in the top layers, leading to a salinity higher than initial status, (iii) there is a leaching process with time of irrigation by RO permeate and secondary effluent with ODI technology (iv) there is a relative low salinity in the active root zone with ODI technology by using of RO permeate and secondary effluent compare to secondary effluent.
5
, 2000
27 July, 2000 Soil Salinity-Electrical Conductivity
uctivity (EC), dS/m
Electrical conductivity (EC), dS/m 15
10
0
2
4
6
8
Soil depth, cm
0 SDI
10 Membrane Initial 20
ODI SDI
30 40 50 60
Figure 4. Soil salinity subject to various effluent applications on 27 July 2000. Initial(Initial soil salinity), Membrane(intermittent RO permeate and secondary effluent using under ODI), ODI (secondary effluent applied under ODI) and SDI (secondary effluent applied under SDI) Crop yield 2
The pepper yield was assessed by taking three samples of each treatment in an area of 2 m (Table 2). These preliminary results indicate the trend of leaching process effects attained during irrigation by blending RO permeate and common secondary effluent the relative low salinity in the active root. The yield with the RO permeate mixture and secondary effluent is higher by more than 100% relatively to applying secondary effluent. Additional work is in progress to confirm these findings. Table 2. Red pepper yield under various effluent qualities and application methods Effluent quality and application method Secondary effluent under ODI Secondary effluent under SDI Secondary effluent intermittently with RO permeate under ODI
Yield, Kg/ha 9,600 14,000 19,000
ECONOMIC ASSESEMENT Economic assessment of the proposed production system is based on defining an objective function (in this work an expression for the water cost) to be optimized, subject to a series of technological, environmental, chemical and operational constraints. The components of the objective function include the selection of the pretreatment method and membrane type, pretreatment costs and RO costs necessary to attain a definite permeate quality, transportation brine disposal and permeate storage costs, cost (or profit) for operation and maintenance expenses, design and contingency expenses. The objective (cost) function for UF and RO plants are given by the following general expressions: UF permeate Effluent Cost of Cost of O & M Cost of Return Optimal Water UF UF UF Brine for Cost Cost Pretreatme nt Unit Expenses Disposal Permeate
6
(3)
RO permeate Feed Cost of Cost of O & M Cost of Return Optimal Water RO RO RO Brine for RO Cost Cost Pretreatment Unit Expenses Disposal Permeate
(4)
Selection of the pretreatment method and membrane type takes into account the designed plant capacity, permeate salinity and experimental results in pilot plants. Commonly, selection of the treatment method and successively the membrane type is associated with defining of a set of Boolean variable, receiving 0,1 values only (Oron, 1979). The constraints define a feasible domain in the decision space and refer to the capacity of the system (both storage and flow rates) and energy losses. The objective function and the constraints are given by a set of linear equations. Consequently it allows using commercial available PC software to obtain an optimal value for the objective function and the decision variables (Tora, 1992). Total UF cost, cent/m3
29.0 28.5 28.0 27.5 27.0 26.5 26.0 25.5 25.0 15
20
25
30
35
40
liter/m2-hr
Flux, Figure 5. The extra UF cost for a treatment plant with a capacity of 20,000 m3/day
The field data was used to define management model for economic assessment. The model is based on production of 20,000 m3/day UF permeates at 95 percent recovery (membrane life span 4 years, interest rate 3.5%, electricity cost 0.062 $/kWh). Sensitivity analysis and assuming wastewater treatment to a secondary 3 level at a cost 15 US cent/ m ) indicates that the extra cost for effluent polishing via the UF stage only is in the range of 5 to 15 US cents/m3. Similarly, and using RO software (Rodesign, 1998; Rosa, 1993) shoes that the extra RO treatment cost is in the range of 10 to 15 US cents/m3. These findings coincide with previous works (Wilf et al., 2001; Laine et al., 2000). CONCLUSIONS The performance of hybrid UF and RO membrane systems for treatment of secondary effluent proved that the technology is promising towards efficient removal of pathogens and nutrients. Several directions and tendencies can be emphasized: (i) UF membranes are very effective for removing soluble organic particles including coliform bacteria and control of RO fouling, and (ii) the UF permeate quality meets health regulations for unlimited irrigation. Consequently, the risk of consuming agricultural products irrigated with UF reclaimed effluent is minimal. The UF stage appears to be as effective as disinfection for the removal of pathogens from secondary effluent. Applying RO permeate for irrigation has contributive leaching effects in the active root zone along with improved yields. Using different mixtures of secondary effluent, UF and RO permeates will still allow applying the effluent for unrestricted use and maintain low extra expenses.
7
ACKNOWLEDGMENTS The work was supported by the INCO-MED Project ICA3-1999-00014 (Proposal No ICA3-1999-10035) on Water Resources Management Under Drought Conditions: Criteria and Tools for Conjunctive Use of Conventional and Marginal Waters in the Mediterranean Region (WAM-ME); and The Beracha Foundation, The Stephen and Nancy Grand Water Research Institute, The Technion, Haifa Israel, The PalestinianJordanian-Israeli Project (PJIP) (1996), on Membrane Technology for Secondary Effluent Polishing: From Raw Sewage to Valuable Waters and Other By-Products, The Levin Family Foundation, USA.
REFERENCES
Asano, T. and Levine, D. (1995). Wastewater reclamation, recycling and reuse: past, present and future. Paper presented at Second International Symposium on Wastewater Reclamation and Reuse. Iraklio, Greece. 17-20 October 1995. pp. 517. Bick A., Oron G. and Shye S.(1996). Nanofiltration technology facts and figures. Journal of Israel institutes of chemical engineering, 25, 8-11. Bick A. and Oron G. (2000). Desalination technology for optimal renovation of saline groundwater in a natural reservoir. Desalination, 131, 97-104. Heller M., Garlapati S. and Aithala K. (1998). Expert Membrane System Design and Selection For Metal Finishing Waste Water Treatment, Expert Systems with Applications, 14, 341-353. Hill, R., and Koenig, R. T. (1999). Water salinity and crop yield. Electronic publishing, AG-425.3, Utah State Universty Extension, Logan, UT 84322, May, 6 pages. Hoffman G.J., Mead R.M., Ziska L.H., Francois L.E. and Gatlin P.B. (1986). Salt tolerance of mature plum trees yield. Water Management Research Laboratory Annual Report, USDA, Frenso, CA, 62-63. Huisman J.(1993). Filmtec Membranes Technical Manual Dow Europe Separation Systems, Germany. Jolis, D., Hiran, R. A., Pitt, P. A., Muller, A., and Mamais, D. (1996). Assessment of tertiary treatment technology for water reclamation in San Francisco, California. Water Science and Technology, 33(10-11), 181-192. Laine J.M., Vial, D. and Moulart P. (2000). Status after 10 years of operation - Overview of UF technology. Desalination, 131(1-3), 17-25. Mickley M.C. (2001). Membrane concentrate disposal: practices and regulation. U.S. Department of the Interior, Bureau of Reclamation Technical Service Center Water Treatment Engineering and Research Group, Program Report No 69, 2001. Mizrahi Y. and Pasternak D.(1985). Effect of salinity on quality of various agricultural crops. Plant Soil, 89, 301-307. Oron G., Campos C., Gillerman L. and Salgot M. (1998). Wastewater treatment, renovation and reuse for agricultural irrigation in small communities. Agricultural Water Management, 38(3), 223-234. Oron G., DeMalach Y., Gillerman L., David I. and Rao V.P.(1999). Improved saline water use under subsurface drip irrigation. Agricultural Water Management, 39(1), 19-33. Oron G., Mehrez A. and Brimberg J.(1996). Optimizing multiple sources, variable quality water use. International Water & Irrigation Review, 16(1), 39-47.
Oron, G. (1979). An algorithm for optimizing nonlinear constrained zero-one problems to improve wastewater treatment. Eng. Optimization, 4(May,109-115.
Rodesign. RO system design software (version 6.4©).(1998).Hydranautics 8444 Miralani Drive, San Diego, California, USA Rosa, Filmtec RO design software.(1993). (version 2.0) Dow Europe Separation System. Rhinmunster, Germany. Shuval H., Lampert Y. and Fattal B.(1997). Development of a risk assessment approach for evaluating wastewater reuse standards for agriculture, Water Sci. Technol., 35(11-12), 15-20.
TORA. Optimization system-version (1992) (1.044 software). U.S. Environmental Protection Agency (EPA).(1992). Guidelines for water reuse (manual), EPA/625/R-92/004, Washington, D.C., 247pp. Wilf, M., Pearce G., Allan, J. and Suarez, J. (2001). Reclamation of sand filter backwash effluent using capillary MF/UF membrane technology. Paper presented at the 4th annual IDS conference “Integrated Desalination Systems”, 12 – 13 December, Haifa, Israel. World Health Organization (WHO).(1989). Health guidelines for the use of wastewater in agriculture and aquaculture, W.H.O Tech. Rep. Ser., 778.
8
