Pilot scale concentration of cheese whey by forward osmosis: A short-cut method for evaluating the effective pressure driving force

https://doi.org/10.1016/j.seppur.2020.117263Get rights and content

Highlights

  • A short-cut method to determine the effective osmotic pressures.

  • Results from short-cut method agrees with fundamental forward osmosis flux model.

  • Short-cut method will be useful for filtration of complex liquid feed systems concentrated in batch mode.

  • Short-cut method will not be applicable if fouling occurs.

Abstract

Cheese whey was concentrated to a concentration factor of 2.7 in a pilot scale forward osmosis filtration system, using a commercial cellulose triacetate membrane in a spiral-wound configuration. The whey was concentrated in a batch mode, using sodium chloride as the draw solution at initial osmotic pressures of 53–75 bar. During the process, flux was shown to reduce due to the simultaneous decrease in the bulk osmotic pressure of the draw solution, increase in the bulk osmotic pressure of the whey and the effect of concentration polarisation on both sides of the membrane. The flux is known to be driven by the effective osmotic pressures of whey and the draw solution on the surface of the membrane active layer. A short-cut approach that requires minimal information in advance about the osmotic pressure of whey and the geometry of the filtration system was implemented, enabling the determination of these effective osmotic pressures. The results obtained were shown to be in agreement with the fundamental forward osmosis flux model. The short-cut approach can be utilised for estimating effective osmotic pressures of other liquid food streams to be concentrated by forward osmosis, without the need of external measurements.

Introduction

Whey, the liquid remaining after curd formation in cheese making, is a valuable resource for the production of powders of high nutritional value [1]. Such a process requires the dehydration and drying of whey, which are usually achieved by thermal evaporation and spray drying, respectively. Evaporation is applied to increase the total solids content of whey to 40–70% [2], [3] before drying to 95% or higher. Membrane filtration, mainly ultrafiltration and nanofiltration have been implemented for the separation of whey constituents and as a concentration step prior to thermal treatment, in order to reduce the total energy requirements for powder production [4]. When no separation of individual components is required, whey is concentrated simply by removing water. In this case, a membrane with the ability to reject small organic components and salts is required. Reverse osmosis membranes have been previously used for whey concentration [5], [6].

Reverse osmosis (RO) utilises high applied hydraulic pressures to overcome the osmotic pressure of whey, resulting in high energy consumption for pumping [7] and the formation of a fouling layer on the membrane surface [8], [9]. Recently, interest was shown in the use of forward osmosis (FO) for the concentration of whey, among other liquid foods [10], [11], [12], [13]. Forward osmosis utilises the chemical potential gradient of a solvent (usually water) across a semi-permeable membrane. The solvent in the feed has a high chemical potential and moves to a draw solution on the other side of the membrane where its chemical potential is lower, until an established osmotic equilibrium [14], [15]. The draw solution is a highly concentrated with an inorganic, organic or other solute and consequently has high osmotic pressure [16], [17]. Intrinsically, the maximum osmotic pressure of the draw solution depends on the solubility of the solute.

Previous studies on the application of forward osmosis for whey concentration employed draw solution osmotic pressures of 90–100 bar [18], [19], [20], surpassing the hydraulic pressure equivalent employed in reverse osmosis, which typically lies in the range of 40–50 bar [7], [21], [22]. Considering this advantage of forward osmosis, whey samples of initially high osmotic pressure can be processed and concentrates of higher concentration factor can be potentially achieved. Previous bench scale studies on the concentration of whey using forward osmosis utilised NaCl and NH4HCO3/NH4OH as the draw solution and resulted in whey total solids contents of 28% and 21% [18], [19]. However, the draw solution is diluted during the forward osmosis process and needs to be regenerated, which imposes an additional cost to the overall process [17].

Water permeate flux, Jw, in forward osmosis is proportional to the effective osmotic pressure difference across the active layer of the membrane, Δπeff, and the water permeability coefficient of the membrane, A (Eq. (1)). The effective osmotic pressure, Δπeff, is the difference between the osmotic pressure of the draw solution and the feed at the surface of the active layer, πD,m and πF,m, respectively (Eq. (2)).Jw=AΔπeffΔπeff=πD,m-πF,m

These active layer surface osmotic pressures for the feed and the draw solution are not equal to their corresponding bulk osmotic pressures, πF,b and πD,b, due to concentration polarisation. Concentration polarisation results from a change in concentration of solute present in either the feed or the draw solution close to the active layer surface. For an asymmetric FO membrane where the feed faces the active layer and the draw solution faces the support layer, external concentration polarisation (ECP) and internal concentration polarisation (ICP) are evident on the active layer and within the support layer, respectively (Fig. 1) [23].

ECP is concentrative for the feed due to the formation of a boundary layer close to the active layer, which poses a resistance to the diffusion of solute back into the bulk solution. Thus, the effective osmotic pressure, πF,m, is expected to be higher than the bulk, πF,b. ICP is dilutive for the draw solution, implying that πD,m is lower than πD,b, because as the draw solute diffuses through the support layer it encounters a resistance inherent of the structure of the support layer and is diluted by the water permeating across the membrane. A model has been developed to predict water permeate flux for an forward osmosis membrane which takes into account both concentrative ECP and dilutive ICP gradients. These gradients are dependent on the mass transfer coefficient on the feed side of the membrane, k, and the resistance to draw solute diffusion across the support layer, K (Eq. (3)). This implicit flux model requires the knowledge of the feed channel dimensions to determine k, the membrane support layer characteristics to determine K, and the feed and draw solution osmotic pressures.Jw=AπD,bexp-JwK-πF,bexpJwk

An estimate of k can be obtained from the calculation of the dimensionless Sherwood number, Sh, (Eq. (4)) by applying appropriate correlations involving the Reynolds number, Re, and Schmidt number, Sc (Eq. (5)) where dh is the hydraulic diameter of the channel, L is the channel length and D is the diffusion coefficient of the solute in the solvent [24], [25]. This determination may be a challenging task for a commercial spiral-wound configuration membrane fitted with feed spacers. Moreover, the mass transfer coefficient varies dynamically during the concentration process, due to changes in the feed density, viscosity and concentration of dissolved compounds. Solute resistivity, K, was previously defined for a forward osmosis process by Eq. (6), where B is the permeability coefficient of the draw solute [26], [27]. The value of K for cellulose triacetate (CTA) bench scale flat-sheet forward osmosis membranes has been previously calculated using this equation by assuming that the membrane has a high rejection of solute, thus B approaches zero [23], [27]. In turn, the solute resistivity depends on the structure of the support layer, expressed as the structural parameter, S, and the draw solute used (Eq. (7)). Flat-sheet CTA membranes characterised in previous studies were shown to have structural parameters in the range of 480–540 µm [28], [29]. The calculation can be further simplified if a solution of negligible osmotic pressure is used as the feed.k=ShDdhSh=aRebSccdhLdK=1JwlnB+AπD,bB+Jw+AπF,mK=SD

Information on the concentration of whey at pilot scale using forward osmosis is limited [13]. In most reported forward osmosis studies, the osmotic pressure driving force is given in terms of the bulk whey and draw solution osmotic pressures, and not the effective osmotic pressure at the membrane surface. Moreover, the effect of the osmotic pressure of whey, as it is concentrated, on the permeate flux was not discussed. In an industrial environment, an in-situ assessment of the effective osmotic pressure driving force without external measurements can be useful. Considering the above, the present work aims to assess the performance of a pilot scale forward osmosis filtration system for the concentration of cheese whey and provide a short-cut approach to quantify the effective osmotic pressures of the feed and draw solution throughout the concentration process. This simple method requires no external measurements of feed osmotic pressure and no information regarding the hydrodynamics of the spiral-wound membrane module. This short-cut method can be adopted to other complex feed streams to be treated by forward osmosis.

Section snippets

Whey specification

Liquid whey produced during the manufacture of blue cheese from pasteurised milk was supplied by High Weald Dairy, West Sussex, UK, at a pH of 4.8 ± 0.2, conductivity of 6.9 ± 0.1 mS/cm and total solids (TS) content of 6.6 ± 0.5%. The detailed composition of the inorganic constituents and the solids content of the initial whey (I) and the concentrate (C) are provided in Table 1.

Draw solution

Food grade pure sodium chloride (NaCl) (Brenntag, UK) and tap water (water hardness of approximately 230 ppm expressed

Flux for water as feed

The water permeate flux was plotted as a function of net driving force when the system is operated in FO and RO modes (Fig. 2). Generally, the net pressure driving force is the maximum pressure driving force available for filtration, noted as πD,b-πF,b+ΔP (Eq. (3)). In the FO mode (single-pass or batch), this driving force was reduced to πD,b+ΔP since water was used as feed, which has negligible osmotic pressure. In the RO mode where water was used in both the feed and draw solution sides, the

Conclusions

The present work demonstrates the potential of forward osmosis for the concentration of cheese whey in a pilot filtration system operated in a batch mode. The process produced concentrates with a concentration factor of approximately 2.7, by increasing the total solids content of whey from 6.5% to 18%. The maximum initial permeate flux was 7.2 L/m2h, obtained for an initial draw solution osmotic pressure of 74 bar. The bulk osmotic pressure of whey was determined graphically from the permeate

CRediT authorship contribution statement

Anna Artemi: Conceptualization, Methodology, Validation, Formal analysis, Investigation, Data curation, Writing - original draft, Writing - review & editing, Visualization. George Q. Chen: Conceptualization, Resources, Methodology, Writing - review & editing, Visualization, Supervision, Project administration. Sandra E. Kentish: Writing - review & editing, Visualization. Judy Lee: Conceptualization, Methodology, Resources, Writing - review & editing, Visualization, Supervision, Project

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

The authors would like to thank High Weald Dairy (West Sussex, UK) for kindly providing the whey for this work, as well as Holchem Laboratories Ltd (Bury, UK) for providing the membrane cleaning and preservation agents. George Chen and Sandra Kentish acknowledge fuding through the Dairy Innovation Hub, and Judy Lee and Anna Artemi acknowledges the assistance of Oliver P. Crossley in carrying out the ICP-OES analysis.

References (42)

Cited by (14)

  • Fabrication and characterization of inner selective antibiofouling forward osmosis hollow fiber membranes for simultaneous wastewater treatment and desalination

    2022, Separation and Purification Technology
    Citation Excerpt :

    Membrane-enabled wastewater treatment with osmotic pressure instead of hydraulic pressure as driving force has gained much interest since the introduction of osmotic membrane bioreactor (OMBR) systems [1–3].

  • Athermal concentration of apple juice by forward osmosis: Process performance and membrane fouling propensity

    2022, Chemical Engineering Research and Design
    Citation Excerpt :

    Therefore, RO is commonly employed to pre-concentrate the juice, rather than to replace the thermal method in the production of concentrated juice product (Al-Obaidi et al., 2017). FO, an emerging membrane process driven by osmotic pressure, has been recognized as a new generation of concentration technology in food industry (An et al., 2019; Artemi et al., 2020; Chen et al., 2019; Khongnakorn et al., 2020; Nayak and Rastogi, 2010; Wang et al., 2017). For example, Artemi et al. (2020) reported that cheese whey could be concentrated up to 2.7-fold in a FO system, employing a cellulose triacetate (CTA) FO membrane with spiral-wound configuration.

  • Enhancing the applicability of forward osmosis membrane process utilizing food additives as draw solutes

    2021, Journal of Membrane Science
    Citation Excerpt :

    In general, heat is used to concentrate the whey solution. Specifically, the solids content is increased up to 40–70% by evaporation, and then to 95% or more by spray drying [18,20]. Although the current method can reliably dewater the solution, evaporation using heat can affect the functional properties and digestibility of heat-sensitive dairy products [21].

View all citing articles on Scopus
View full text