Challenges in Forward Osmosis of Seawater Using Ammonium Bicarbonate as Osmotic Agent

Jian-Jun Qin, Gayathri Danasamy, Winson C.L. Lay, Kiran A Kekre

  Open Access OPEN ACCESS  Peer Reviewed PEER-REVIEWED

Challenges in Forward Osmosis of Seawater Using Ammonium Bicarbonate as Osmotic Agent

Jian-Jun Qin1,, Gayathri Danasamy2, Winson C.L. Lay1, Kiran A Kekre1

1PUB, Singapore’s National Water Agency, Singapore

2Imperial College London, UK


This study aimed at exploring whether product quality, membrane fouling and salt reverse flow would be challenges in forward osmosis (FO) of seawater using NH43 as an osmotic agent. Experiments were conducted with a lab scale FO system containing effective membrane area of 95 cm2. Synthetic seawater (SSW) with 3.5-7.0 mg/L boron and a real seawater (RSW) were used as feeds and 1.5-2.5 M NH43 as draw solutions. The experimental operation could be stablized within 0.5 h. For the SSW, boron rejection ranged of 47-85% and increased with increasing water flux while boron in the permeate was greater than 0.8mg/L. Water flux with RSW was 3 times lower than that with SSW, indicating that there might be serious membrane fouling with RSW. It was surprisingly observed that non volatile organic in the FO permeate was 8-10 mg/L, which was from the draw solution although NH43 used was analytical grade. Additional water cost would be $0.4/m3 because of NH43 loss. It was concluded that product quality in terms of high TOC contaminant in NH43 and low boron removal, serious fouling with RSW and salt reverse flow could be challenges for the FO process using NH43 as osmotic agent for seawater desalination.

At a glance: Figures

Cite this article:

  • Qin, Jian-Jun, et al. "Challenges in Forward Osmosis of Seawater Using Ammonium Bicarbonate as Osmotic Agent." American Journal of Water Resources 1.3 (2013): 51-55.
  • Qin, J. , Danasamy, G. , Lay, W. C. , & Kekre, K. A. (2013). Challenges in Forward Osmosis of Seawater Using Ammonium Bicarbonate as Osmotic Agent. American Journal of Water Resources, 1(3), 51-55.
  • Qin, Jian-Jun, Gayathri Danasamy, Winson C.L. Lay, and Kiran A Kekre. "Challenges in Forward Osmosis of Seawater Using Ammonium Bicarbonate as Osmotic Agent." American Journal of Water Resources 1, no. 3 (2013): 51-55.

Import into BibTeX Import into EndNote Import into RefMan Import into RefWorks

1. Introduction

Forward osmosis (FO) is known as the net movement of water across a semi-permeable membrane driven by a difference in chemical potential across the membrane, i.e., passage of water is from lower concentration side to higher one because the lower the concentration, the higher the water chemical potential. Although the FO phenomenon was observed in 1748, the interests in research on FO and in engineered applications of FO have been increasingly attractive since 2000 [1, 2]. However, there are still three major challenges of FO in municipal scale seawater desalination: a) fabrication of FO membranes with high water flux and high draw solute rejection; b) being lack of appropriate draw solutions; and c) cost-effective post-treatment on a total system approach [3].

In desalination application, the energy consumption for seawater desalination using reverse osmosis (RO) is currently 3-3.5 kWh/m3. Newly developed FO technology could potentially lower desalination energy consumption since osmotic pressure, not hydraulic pressure as in the case of RO, provides a natural driving force for water transport through the FO membrane. It is reported that the Manipulated Osmosis Technology developed by University of Surrey could reduce energy consumption to <1.6 kWh/m3 using proprietary osmotic agents as a draw solution. The proprietary process had also been demonstrated using a 1 m3/h pilot unit and is now ready for commercialization [4]. Khaydarov [5] developed a solar powered direct osmosis desalination process with a pilot testing of 1 m3/h capacity. He reported the specific power consumption of the desalination process could be less than 1 kWh/m3 excluding the requirement on solar energy.

In recent years, there has been increasing interest in an integrated FO process for seawater desalination developed and intensively studied by Elimelech and his research group at Yale University [6, 7, 8, 9]. In the process, ammonium bicarbonate (NH43) is utilized as the osmotic agent and a thermal process such as the distillation is proposed as the post-treatment to strip and recover the osmotic agent. Potential advantages of NH43 osmotic agent include high osmotic pressure which can result in high water flux and utilisation of waste heat (if available) to recover the osmotic agent at <60 ºC. Other researchers [10-16][10] have also studied the FO process with NH43 as the osmotic agent. However, existing studies on the NH43 concept focus mainly on effects such as osmotic agent concentration, temperature, and role of internal concentration polarization on water flux. Study of this process using real seawater as a feed has not been reported so far and there is also a lack of data on product quality. Practical issues such as loss of osmotic agent in the FO process for implementation in the municipal scale have also not been adequately discussed. Objective of this study is to address the above issues and discuss on challenges in using NH43 as osmotic agent in a FO process for seawater desalination.

2. Experimental Materials and Methods

The lab scale FO unit used in experiments is shown in Figure 1 [17]. Two FO membranes from Hydration Technologies, Inc. were used in the study. The FO membrane with a diameter of 110 mm (effective membrane area of 95 cm2) was positioned vertically between the two compartments in a membrane holder without any additional support. The active layer of FO membrane was orientated to face the draw solution in order to reduce internal concentration polarization (CP) and thus obtain high water flux [18].

The experiments were carried out at temperature of 25±1ºC. During experiments, feed and draw solution were re-circulated using Pump I and Pump II with a flow rate of 9 l/min (MD-6Z, supplied by Iwaki Co. Ltd), respectively. Due to the intensive mixing, external concentration polarization phenomena were expected to be negligible.

Where, (1) Feed side (2.6L), (2) Draw solution side (2.6L), (3) FO membrane, (4) Recirculation Pump I, (5) Recirculation Pump II, (6) Measuring tube, (7) Draw solution inlet, (8) Draw side sampling, (9) Feed inlet (open compartment), (10) Feed sampling, (11) Sealed compartment, (12) pH reader.

Two types of feed water were used: 0.5 M NaCl solutions with different boron concentrations (3.5-7.0 mg/L) were used as synthetic seawater (SSW), and real seawater (RSW) which was collected from the Variable Salinity Plant of Singapore. The RSW was pre-treated by ultrafiltration and was solids free. Different concentration of NH43 was used as draw solution for the experiments. When the feed water was SSW, the draw solution was 2.5M NH43 solution with osmotic pressure around 100 bar. When the feed water was RSW, draw solution was 1.5M or 2.5 M NH43 with osmotic pressures around 60 bar and 100 bar, respectively. For overview, the designed experimental conditions are summarized in Table 1. Characteristics of the real seawater (RSW) are given in Table 2.

The permeate flow rate, concentrations of solutes in feed and draw solution were measured during experiments. The experimental duration was 1-2 hrs. The parameters pH and conductivity were measured by a benchtop pH/conductivity meter (model WTW). Boron and metal ion analyses were performed using a Perkin Elmer model 2100 Inductively Coupled Plasma Emission Spectrometer. TOC was measured using Shimadzu TOC analyzer model 5000A as per USEPA 415.1 standard method.

Analysis on water flux, solute flux and rejection of the referred solute (Boron) is provided as per Eq.1, Eq.2 and Eq.3, respectively.


Where: Jw is water flux, (l/m2/h or LMH); ΔV is the volume increase measured from the measuring tube, (L); Am is effective membrane area, (m2); Δt is the time recorded, (h); Js is solute flux, (g/m2/h); ΔW is the mass increase of the solute measured from draw solution side, (g); Cf and Cp are the solute concentration in the feed and permeate (NOT the draw solution), respectively.

3. Results and Discussion

3.1. Characteristics of Feeds and Draw Solutions

Characteristics of feed and draw solutions in experiments are summarized in Table 3. The duration of each experiment was about 120 minutes. It is interesting to note that about one unit increase in pH of the feed was observed from start to end of the experiment in all tests except Test 8. This could be attributed to the reverse flow of 3 ions from the draw solution of 2.5 M NH43 to the feed, resulting in a significant increase of OH- concentration (3- à OH- + CO2) in the feed. Meanwhile, the loss of 3- ions from the draw solution resulted in an increase of H+ concentration (CO2 + H2O à H+ + 3-), and hence a reduced pH. This observation is not well understood. However, it has implication on ions transport in a FO process and should be studied further. No significant change in conductivity was observed at both feed and draw solution sides during the experiments. Although the experiments were run in batch, the dilution of the draw solution due to permeation of feed water could be considered negligible because max net volume increase over the experimental duration was about10 mL compared to the draw solution compartment volume at 2.6L (Less than 1% in volume). Quasi-constant concentration in the draw solution may therefore be assumed.

Table 3. Characteristics of feed and draw solutions in experiments

3.2. Determination of the Steady State

A couple of experiments were conducted prior to the test to ascertain the effect of varying Boron concentration on water flux (Figure 2). The results indicate that the experiment could reach a steady state (in terms of water flux) within 30 minutes. During this time, the water flux decreased with the time before levelled off to a stable value. This observation could be explained as the draw solution was pre-filled in the experimental system while the feed was filled just before the experiment started. Internal concentration polarization (ICP) occurred within the support layer of the membrane which faced the feed side. The development of the ICP resulted in a decrease in driving force and hence the flux within the first 30 minutes. The ICP stabilized thereafter and flux became stabilized correspondingly [19].

Another set of experiments that tracked the steady state water flux under different Boron concentration indicates that indicates that boron concentration in the feed of SSW had no significant influence on water flux when 2.5 M NH4HCO3 was used as the draw solution (Figure 3).

Figure 3. Water Flux vs. Boron Concentration in the Feed
3.3. Challenges on Water Quality and Fouling

Analysis of boron concentration and rejection at different water fluxes in experiments are summarized in Table 4. Water flux in Group 2 (Tests 4-6) was much higher than that in Group 1 (Tests 1-3) due to different types of FO membranes. As a result, boron rejections (76-85%) in Group 2 were higher than that in Group 1 (57-67%). Large difference in water flux between Group 2 (the synthetic seawater) and Group 3 (Tests 7& 8, real seawater) could be attributed to membrane fouling by compounds found in the real seawater and not in synthetic water. Water flux in Test 8 was lower than that in Test 7 due to the reduced concentration of draw solution used. Again, Boron rejection in Test 8 was lower than that in Test 7.

Table 4. Analysis of boron concentration and rejection at different water fluxes

A few discussions could be made. Consistent with this study, Tang et. al. [20] also found higher fouling potential with NOMs in the feed when the FO membrane active layer was faced to draw solutions. The water flux has significant impact on the boron rejection observed. With increasing water flux, higher boron rejection was achieved in an FO process due to dilution effect that is similar to RO process. This observation is consistent with Jin et. al. [21].In that study, boron rejection in the range of 29-62% was observed with NaCl as draw solutions. The reduced boron rejection observed when using RSW as compared to SSW could be attributed to the greater fouling encountered under the RSW condition due to fouling-enhanced concentrative ICP effect. [20, 21].

Mane et al. [22] associated the increase in boron rejection with the dilution effect at higher flux in RO studies. Similar trend was reported by Cengeloglu et al. [23], Koseoglu et al. [24], Prats et al. [25] and Magara et al. [26]. Based on the results of Group 3 in this study, boron concentration in the permeate could be dramatically reduced if a new FO membrane with much high water flux can be developed in the future, which could be a challenge for FO desalination [3].

The average water flux achieved with synthetic seawater (0.5 M NaCl) as feed and 2.5 M NH43 as draw solution was 5.7 LMH in Group 2. McCutcheon et al. [6] obtained water flux of 13.5 LMH using similar feed and draw solution at 50ºC and the temperature-corrected flux could be 6.4 LMH [19]. In terms of water flux, this study is consistent with the work of McCutcheon et al [6], and the key difference is the experimental temperature.

For all the tests, boron concentrations in the permeate were greater than the previous WHO guideline for drinking water (0.5mg/L) but would be within the current one (2.4 mg/L) [27]. The results indicate that boron removal may be just adequate for seawater desalination using FO, and under specific circumstances, a further post-treatment for boron removal would be required.

Furthermore, it was also observed that the TOC in the draw solution of Test 7 was 9-10 mg/L, which was greater than the TOC in RSW feed at 2.1 mg/L as shown in Table 5. Analysis indicated that the TOC was non volatile organic and it was due to the NH4HCO3 draw solution (2.5M). Such high TOC value would not be able to meet drinking water guideline.

Table 5. Challenge on water quality for high TOC contaminant

The results demonstrated that the product water quality using NH4HCO3 as the draw solution could be a serious issue, which so far had not been discussed extensively in the literature. Another issue of concern is the membrane fouling for practical FO application in seawater desalination.

3.4. Challenge of Loss of NH43 Osmotic Agent

Another practical issue for implementation of FO with NH43 osmotic agent is the loss of draw solute through the membrane into the feed (Js’/Jw). Table 6 shows the daily amount of NH43 draw solute that needs to be replenished for a FO process due to loss through the membrane for different Js’/Jw scenarios. The implication is that a municipal scale FO plant with capacity of 100,000 m3/day will lose at least 200,000 kg of NH43 on a per day basis, which need replenishment. Additional production cost for topping-up of the loss of NH43 will be $0.4/m3 if NH43 cost is $0.2/kg. From a logistics point of view, a FO process that uses NH43 draw solute will not be feasible for a municipal scale application at the current membrane development. The limitation must be addressed via great enhancement of FO membrane performance with high water permeability & high solute selectivity such that Js’/Jw ≤ 0.01 g/L [3]. This is a challenge to the FO research community. Future development needs to take into account the inter-dependent relationship between membrane development and draw solution selection, and under the consideration of cost-effectiveness.

Table 6. Loss of NH43 osmotic agent in FO

4. Conclusions

From findings in this study, the following conclusions can be drawn:

1) When synthetic seawater solution was used as the feed water with boron concentration between 3.5 and 7.0 mg/L, boron rejection by the FO process was in the range of 57-85%. Boron rejection could be increased with increaseing water flux.

2) When real seawater that was pretreated with UF was used as the feed water, serious membrane fouling was observed (Jw was 3 times lower compared to that with synthetic seawater) and boron rejection reduced to 39-56% due to fouling-enhanced concentrative ICP effect.

3) The impaired product water quality in terms of TOC and huge loss of osmotic agent when NH4HCO3 is used as the draw solution are challenges for application of FO process for seawater desalination.

It is recommended that future development needs to take into account the inter-dependent relationship between membrane development and draw solution selection, and under the consideration of cost-effectiveness.


The authors would like to express appreciation to Dr Htun Oo and Dr Sze Sze Chin for their interactions in this study.


[1]  Cath, T.Y., Childress, A. E. and Elimelech, M., “Forward osmosis: Principles, applications, and recent developments,” J. Membr. Sci., 281, 70-87, 2006.
In article      CrossRef
[2]  Qin, J.J., Chen, S., Oo, M. H., Kekre, K. A., Cornelissen, E.R., Ruiken, C.J., “Experimental studies and modeling on concentration polarization in forward osmosis,” Water Science Technology, 61, 2897-2904, 2010.
In article      CrossRefPubMed
[3]  Qin, J.J., Lay, W. C.L. and Kekre, K. A., “Recent developments and future challenges of forward osmosis for desalination: A review,” Desalination & Water Treatment, 39, 123-136, 2012.
In article      CrossRef
[4]  Global Water Intelligence, Water Desalination Report, Vol. 43 No. 23, page 1, 18 June 2007.
In article      
[5]  Khaydarov, R. A., Khaydarov, R. R., “Solar powered direct osmosis desalination,” Desalination, 217,225, 2007.
In article      CrossRef
[6]  McCutcheon, J.R., McFinnis, R.L. and Elimelech, M., “A novel ammonia-carbon dioxide forward (direct) osmosis desalination process,” Desalination, 174, 1-11, 2005.
In article      CrossRef
[7]  McCutcheon, J.R., McGinnis, R.L., Elimelech, M., “Desalination by a novel ammonia–carbon dioxide forward osmosis process: influence of draw and feed solution concentrations on process performance,” J. Membr. Sci., 278,114, 2006.
In article      CrossRef
[8]  McGinnis, R. L. and Elimelech, M., “Energy requirements of ammonia–carbon dioxide forward osmosis desalination,” Desalination, 207 370-382, 2007.
In article      CrossRef
[9]  Elimelech, M. and Phillip, W. A., “The future of seawater desalination: Energy, technology, and the environment,” Science, 333 712-717, 2011.
In article      CrossRefPubMed
[10]  Ng, H.Y., Tang, W., Wong, W.S., “Performance of forward (direct) osmosis process; membrane structure and transport phenomenon,” Environ. Sci. Technol., 40, 2408, 2006.
In article      CrossRefPubMed
[11]  Tan C.H. and Ng, H.Y., “A novel hybrid forward osmosis – nanofiltration (FO-NF) process for seawater desalination: Draw solution selection and system configuration,” Desalination and Water Treatment, 13,356, 2010.
In article      CrossRef
[12]  Low, S.C.,Preliminary studies of seawater desalination using forward osmosis,” Desalination and Water Treatment, 7, 41, 2009.
In article      CrossRef
[13]  Cath, T. Y., “Osmotically and thermally driven membrane processes for enhancement of water recovery in desalination processes,” Desalination and Water Treatment, 15,279, 2010.
In article      CrossRef
[14]  Achilli, A., Cath, T.Y., Childress, A.E., “Selection of inorganic-based draw solutions for forward osmosis applications,” J. Membr. Sci., 364,233-241, 2010.
In article      CrossRef
[15]  Hancock, N.H. and Cath, T.Y., “Solute coupled diffusion in osmotically driven processes,” Environ. Sci. Technol., 43,6769, 2009.
In article      CrossRefPubMed
[16]  Phuntsho, S., Shon, H. K., Hong, S., Lee, S, Vigneswaran, S., “A novel low energy fertilizer driven forward osmosis desalination for direct fertilization: Evaluating the performance of fertilizer draw solutions,” J. Membr. Sci., 375, 172, 2011.
In article      CrossRef
[17]  Danasamy, G., “Sustainability of Seawater Desalination Technology – the study of forward osmosis as a new alternative,” Master Thesis, Imperial College London, 2009.
In article      
[18]  McCutcheon, J.R., Elimelech, M. “Influence of concentrative and dilutive internal concentration polarization on flux behaviour in forward osmosis,” J. Membr. Sci., 284,237, 2006.
In article      CrossRef
[19]  Qin, J. J., Oo, M. H., Tao, G., Cornelissen, E.R., Ruiken, C.J., de Korte, K.F., Wessels, L.P., Kekre, K. A. “Baseline study on osmotic membrane bioreactor: Optimization of operating conditions in forward osmosis,” The Open Chem. Eng. Journal, 3, 27, 2009.
In article      CrossRef
[20]  Tang, C. Y., She, Q. H., Lay, W. C. L., Wang, R., Fane, A. G., “Coupled effects of internal concentration polarization and fouling on flux behavior of forward osmosis membranes during humic acid filtration,” J. Membr. Sci., 354, 123-133, 2010.
In article      CrossRef
[21]  Jin, X., Tang, C. Y., Gu, Y., She, Q. and Qi, S., “Boric acid permeation in forward osmosis membrane processes: Modeling, experiments, and implications,” Environ. Sci. Technol., 45, 2323-2330, 2011.
In article      CrossRefPubMed
[22]  Mane, P.P., Park, P.K., Hyung, H., Brown, J.C. and Kim, J.H., “Modeling boron rejection in pilot- and full-scale reverse osmosis desalination processes,” J. Membr. Sci., 138, 119-127, 2009.
In article      CrossRef
[23]  Cengeloglu, Y., Arslan, G., Tor, A., Kocak, I. and Dursun, N., “Removal of boron from water by using reverse osmosis,” Separation and Purification Technology, 64,141-146, 2008.
In article      CrossRef
[24]  Koseoglu, H., Kabay, N., Yüksel, M. and Kitis, M., “The removal of boron from model solutions and seawater using reverse osmosis membranes,” Desalination, 223, 12-133, 2008.
In article      CrossRef
[25]  Prats, D., Chillon-Arias, M.F. and Rodriguez-Pastor, M., “Analysis of the influence of pH and pressure on the elimination of boron in reverse osmosis,” Desalination, 128, 269-273, 2000.
In article      CrossRef
[26]  Magara, Y., Tabata, A., Kohki, M., Kawasaki, M.and Hirose, M., “Development of boron reduction system for sea water desalination,” Desalination, 118, 25-34, 1998.
In article      CrossRef
[27]  WHO, Boron in drinking water. 2009.
In article      
  • CiteULikeCiteULike
  • MendeleyMendeley
  • StumbleUponStumbleUpon
  • Add to DeliciousDelicious
  • FacebookFacebook
  • TwitterTwitter
  • LinkedInLinkedIn