Power Generation Based on Pressure Retarded Osmosis: A Design and an Optimisation Study

Transcription

1 Poer Generation Based on Pressure Retarded Osmosis: A Design and an Optimisation Study Mudher Sabah 1, Ahmed F. Atan, Hameed B. Mahood 3,4 and Adel Sharif 3,5 1,4 University of Misan university of Al-Mustansiriya, Baghdad, Iraq 3 University of Surrey, CORA Centre, Dept. of Chemical and Process Engineering, UK 5 The Qatar Foundation, Qatar Energy and Environment Research Institute Abstract A design procedure for an open system osmotic poer generation plant utilising the salinity gradient beteen to liquid streams has been developed. Seaater (high salinity) as a dra solution and brackish ater (lo salinity) as a feed stream ere selected. The high applied hydraulic pressure as substituted by implementing a pressure exchanger to reduce the energy consumption ithin the plant. The effect of the dra solution flo rate, the feed stream flo rate and the applied hydraulic pressure on the net poer production, membrane area and specific energy consumption ere examined. The calculations revealed that the optimised values of the dra solution flo rate, the feed flo rate and the applied hydraulic pressure ere 35000Kg -0000Kg respectively. In addition, the optimal membrane area corresponding to the maximum net poer production as m. Keyords : osmotic pressure, forard osmosis,pressure retarded osmosis Introduction: The increasing demand on energy due to global population and technological development, as ell as to environmental problems, requires the identification of ne energy resources. One of these resources is poer production utilising a salinity gradient via Pressure Retarded Osmosis (PRO). This technique simply exploits a difference in salinity beteen salt ater and fresh ater, hich appears as a difference in a chemical potential, to produce a high volume and a high pressure stream. The amount of fluid volume that can be obtained is dependent on a salt concentration in a salty stream, hich qualifies by osmotic pressure. Hoever, an increase in the osmotic pressure results an increase in the amount of the ater or fluid floing across a membrane. Accordingly, it is possible to harness the difference in salinity beteen the oceans and rivers or seas and rivers to produce electricity, hich considerably assists in reducing the demand on fossil fuels. Salinity gradient poer, hoever, in addition to PRO, can be exploited by different technologies, for example, by reversed electrodialysis (RED), hydocratic generator and vapour compression [1] Among these technologies, the PRO technique has received more attention. Sideny Loeb (1950) as the first scientist inventor of this process hen he produced drinking ater from seaater by using a high pressure pump. This process as hindered during the seventeenth of the nineteenth century due to the lack in producing a suitable membrane until1997, hen Statkraft Noregian Poer Producer Company established a PRO plant to produce 5 kw electricity for the first time. Nevertheless, many studies took place before and after Norman [] as the first investigator suggests a diagram of an osmotic energy convertor, hile the first experimental data for the PRO process as published during 1976 [3]. utilising a hollo-fibre RO membranes. On the other hand, the concept of PRO, hich suggested before by Lobe and Norman [4], has been proved by subsequent studies carried out [5,6,7]. In these investigations, the total output membrane poer density obtained as loer than the expected values, based on the osmotic pressure difference across a membrane (1.56 to 3.7 W/m ). Furthermore, internal concentration polarization has been observed and its effect on the flux through the membrane and consequently on the PRO economic has been ensured. Lee [8] developed a theoretical reference model for PRO performance based on FO and RO experiments. Water flux and concentration polarization ere researched. The mechanical efficiency of different theoretical PRO models as investigated quantitatively [9]. The possibility of using the PRO technique to produce electricity from the Great Salt Lake as investigated by Lobe [10,11] and he found the cost could be 0.15$/kWh. Seppala [1] demonstrated ith evidence that a non-linear relationship controlled the transport of ater and solute ith osmotic and hydraulic pressures. A very high membrane poer density (more than50 W/m) and 16% Carnot efficiency has been found by McGinnis [13] hen using ammonium-carbon dioxide as a dra solution. Sam [14] studied numerically the feasibility to produce electric poer by PRO. Utilising a hydrodynamic mass transfer model, the real size membrane poer output as about 40% less than the lab scale size. In addition the counter-current flo configuration produced more poer than the concurrent by 15%. Remon [15] Elimelech [16] separately carried out an extensive revie of osmotic poer production by salinity gradient. The impact of different parameters on the poer production capacity and the possible techniques hich could be exploited ere illustrated. Volume, Issue 1, December 013 Page 68

2 Theoretical part In this paper, an optimisation study and a design procedure for a large scale open system osmotic poer station has been developed utilising natural salinity and brackish ater resources (sea and river). The effect of dra and feed mass flo rate as ell as to the hydraulic pressure on the plant output poer and the membrane poer density have been researched. Water flux across the membrane in PRO processes, J, is usually represented by the folloing phenomenological relationship: J A ( P) (1) Equation (1) shos that J is determined as the product of the system permeability to ater, A, and the net transmembrane driving pressure, hich is the net difference beteen the osmotic pressure, ΔΠ, and the net hydraulic pressure, ΔP. The density of the poer obtained from the PRO process, W, can be estimated as the product from multiplying ater flux by the hydraulic pressure [17,18] W J P A ( P) P () The poer density of the membrane that is required to obtain a profitable PRO process as determined to be beteen 4 6 W/m [18]. A suggestion as made by another study [17],that the first derivative for equation () ith respect to ΔP assuming A as a constant may specify the maximum value for W: W max A (3) 4 Hence, from Eq. (), it can be indicated that W max hich can be reached hen ΔP equals 0.5ΔΠ. The ΔP is normally estimated by assuming linear pressure drop alongside the OMU, i.e.: PVDS in PVDS out PVFS in PVFS out P (4) and similarly for : VDS in VDS out VFS in VFS out (5) Where the subscripts VDS and VFS refer to the Dra Solution (the high concentration side) and the Feed Water (the lo concentration side), respectively. Mass balance In commercial modules, here cross flo regimes are normally utilised, the concentrations change ith membrane length. In processes that utilize co-current cross-flo modes, the loest solution permeability may occur at the concentrated DS inlet position here the FW enters at its loest concentration. As the concentration difference across the membrane continuously decreases, the solution permeability to ater transfer continuously increases. Finally, many practical parameters and indices can be identified for the PRO process in an open hydro-osmotic poer (HOP) plant. A material balance around the Osmotic Membrane Unit (OMU) gives the folloing: Q Q Q Q Q VDS out VDS in VFS in VFS out here Q is the volumetric flo rate and Q refers to the permeated ater flo rate across the membrane. Figure (A) Schematic representation for ater flux and solute concentration across a pore of a symmetric membrane in the PRO process, (B) electrical analogy for the system resistance to ater flux. Volume, Issue 1, December 013 Page 69

3 Osmosis Hydro-Poer Plant: The schematic diagram of the proposed Osmosis poer plant via PRO technique can be shon in Fig. (1). It consists of three main components; membrane (1), hydro-turbine (unit no. 4) and pressure exchanger (unit no. 6) in addition to the other auxiliary units. Seaater is pumped from the sea via a lo pressure pump to a pre-treatment unit (unit no. 5), by a stream, (no. 13) for removal of suspension materials. This is used as a dra solution (stream no. 16) after it is pumped to the FO osmosis, then processed (unit no. 3) via a high pressure pump or pressure exchanger (unit no. 6) and lo pressure pump (no. 7). Meanhile, brackish ater is received at the pre-treatment unit (unit no. 9) via a lo pressure stream (stream no. 10). This stream is implemented as a feed to the FO process, here it is pumped via a lo pressure pump (no. 8) via a stream (no. 11) to the FO unit (unit no. ). According to the osmosis pressure difference beteen to streams (no. and no.3), permeation occurs from lo osmotic pressure side (unit no.) to the high osmotic pressure side unit (no. 3). Hoever, the salty stream is mitigated and leaves the FO unit ith lo salt concentration and high pressure via stream (no. 17). This stream (no. 17) is expanded at a hydro-turbine (unit no. 4) to produce electric poer and depressurised and dran outside the station (stream no. 0). Part of the high pressure stream (no. 17) is sent to the pressure exchanger (unit no. 6) via stream (no. 18). This, of course, helps to reduce poer consumption through the plant hich makes the process viable and economic. On the other hand, the lo osmotic pressure stream (no.) is concentrated and leaves the FO via stream (no. 1). As shon in Fig.(1), the process exploits the open system technique to avoid the separation process, hich definitely makes the process uneconomic and unviable. Furthermore, the availability of natural resources, especially in south of Iraq, helps the location of cheap and reliable additional energy sources. Fig. (1): Schematic diagram of open system PRO plant Table 1: Revie of PRO studies from the early days to the most recent Investigations, [19,0,1,]. Feed/Dra Solution Membrane Hydrauli Poer Osmotic c Densit Research Reference Pressure(bar Pressure y Group/year s ) (bar) (W/m ) Hollo Fiber seaater RO Loeb et al.(1976) [3] FRL Composite Loeb, seaater RO Mehta.(1979) [4] Hollo Fiber seaater RO Mehta.(1978) [5] Hollo Fiber Mehta, Loeb seaater RO (1979) [6] seaater spiral Jellinek,&Masuda ound RO (1981) [7] (3.5%) CA& PA& PBIL flat sheet seaater RO Lee, Baker.(1980) [8] DI ater /Brine (3.5 CTA flat sheet Achilli et al %) seaater FO (HTI) (009) [9] Freshater/Seaater Lab TFC (flat Gerstandt et al. sheet/hollo fiber) Statkraft (008) and CA seaater FO [30] River ater/brine (3.5 TFC Hollo Fiber Fane et al.(011) [31] Volume, Issue 1, December 013 Page 70

4 6%) seaater FO Cellulosic River ater/ seaater membrane, Hollo (3.5%) Fiber seaater RO (M NaCl) Wasteater(0.5M) /Brine (M) CTA commercial flat sheet seaater FO(HTI) CTA commercial flat sheet seaater FO(HTI) YIP et al.(011) [3] Tang et al. (01) [33] Kim & Elimelech (01) [34] Optimization Procedure: A design procedure has been developed based on mass balance to find the best conditions for making operation of the PRO plant smooth and economic. Many parameters affect the PRO performance; hoever, a computer program has been ritten to optimise them and make quantitative comparisons to specify the best values that give constant poer production. The first effective parameter is the membrane area. Fouling is a common problem hich appears at PRO open system operation plant. Hoever, decreasing the membrane area ill considerably reduce fouling as ell as reducing the capital cost. Figure () shos the variation of the membrane area ith applied hydraulic pressure at a constant feed volumetric mass flo rate. An increased of the membrane area ith increased applied hydraulic pressure can be clearly seen in the figure. The hydraulic pressure act practically as an additional resistance hich impedes permeate or ater flux from the feed side to the dra side. Therefore, for a constant output poer, increasing the membrane area is a solution to overcome this obstacle. At the same time as, the membrane area is increased, the the feed side volumetric flo rate is increased at a constant applied hydraulic pressure. This could be justified reasonably by the implicit relationship that connects the feed side flo rate and the total flux, hich reflects directly on the membrane area. The higher the feed side flo rate, the higher the flux subsequently increases the membrane area required. Meanhile, a similar relationship can be shon beteen net poer produced and the applied hydraulic pressure at a constant feed flo rate. Although the hydraulic pressure impacts negatively on the total permeate volume, it directly enhances the net poer produced. The net poer production at the hydraulic turbine, of course, depends on the total inlet turbine feed volume in addition to the applied hydraulic pressure value. On the other hand, at constant hydraulic pressure, the increase of the feed flo rate leads to n increase in the net poer produced as a result of increasing the total volume permeating through the membrane. No considerable effect can be shon from the hydraulic pressure on the specific energy consumption value, here its value is still very lo even at the maximum hydraulic pressure applied (Fig.(3)). Similarly, studying the dra side effect on the PRO process revealed the same effect of the hydraulic pressure on the total membrane area, net poer production and specific energy consumption as shon in Figs. (4, 5 &6) respectively. Fig. : Membrane area versus applied hydraulic pressure Fig.3: Net poer production versus applied hydraulic pressure Volume, Issue 1, December 013 Page 71

5 Fig.4: Specific energy consumption versus applied hydraulic pressure Fig.5: Membrane area versus applied hydraulic pressure for different dra solution flo rates. Fig.6: Net poer production versus applied hydraulic pressure at different dra solution flo rates. Fig.7: specific energy consumption versus applied hydraulic pressure at different dra solution flo rates. Volume, Issue 1, December 013 Page 7

6 Conclusions: A theoretical design procedure for open system PRO plant is developed. According to the results the folloing conclusions can be made: - Total membrane area increases hen increasing the applied hydraulic pressure for both feed and dra side flo rates. - Net poer production increases ith increased applied hydraulic pressure. - Specific energy consumption increases hen increasing applied hydraulic pressure and decreasing the volumetric flo rate of feed. References: [1] Berrouche Y. and P. Pillay (01), Determination of salinity gradient poer potential in Quebec, Canada, J. Reneable Sust. Energy, V.4, Pp [] Norman, R.S. (1974), Water Stalinisation: a source of energy, Science, V.186, Pp [3] Lobe S., Van Hessen F. and Shahaf D. (1976), Production of energy from concentrated brines by pressure-retarded osmosis, II. Experimental results and projected energy costs, J. Memb. Scie., V.1, Pp [4] Loeb S. and Norman R.S. (1975), Osmotic poer plant, Science, V. 189, Pp [5] Loeb S. and Mehta G.D. (1978), Internal polarization in the porous substructure of a semi permeable membrane under pressure-retarded osmosis, J. Memb. Scie., V.4, Pp [6] Loeb S. and Mehta G.D. (1979), A to coefficient ater transport equation for pressure-retarded osmosis, J. Memb. Scie., V.4, Pp [7] Mehta G.D. and Loeb S. (1979), Performance of permasep B-9 and B-10 membrane in various osmotic regions and at high osmotic pressure, J. Memb. Sci., V.4, Pp [8] Lee K.L., Baker R.W. and Lonsdale H.K. (1981), Membrane for poer generation by pressure retarded osmosis, J. Memb. Sci., V.8, Pp [9] Loeb S., Honda T. and Mehta G.D. (1990), Comparative mechanical efficiency of several plant configurations using a pressure-retarded osmosis energy converter, J. Memb. Sci., V.51, Pp [10] Loeb S. (001), One hundred and thirty benign and reneable megaatts from Great Salt Lake? The possibilities of hydroelectric poer by pressure-retarded osmosis, Desalination, V.141, Pp [11] Loeb S. (00), Erratum to: One hundred and thirty benign and reneable megaatts from Great Salt Lake? The possibilities of hydroelectric poer by pressure-retarded osmosis ith spiral module membrane [Desalination, V.141, Pp.85-91], Desalination, V.14, Pp [1] Seppala A. and Lampinen M.J. (004), On the non-linearity of osmotic flo, Exp. Them. Fluid Sci., V.8, Pp [13] McGinnis R.L., McCutcheon J.R. and Elimelech M. (007), A novel ammonia-carbonate dioxide osmotic heat engine for poer generation, J. Memb. Sci, V.305, Pp [14] Sam van der Zan, Pothof I.W.M., Blankert B. and Bara J. (01), Feasibility of osmotic poer from a hydrodynamic analysis at module and plant scale, J. Memb. Sci., V.389, Pp [15] Ramon G.Z., Feinberg B.J. and Hoek E.MV. (011), Membrane-based production of salinity-gradient poer, Energy & Enviro. Sci., DOI: /c1ee01913a and.rsc.org/ees. [16] Logan B.L. and Elimelech M. (01), Membrane-based processes for sustainable poer generation using ater, Nature, V.488, Pp [17] A. Achilli, T. Cath, A. Childress, Poer generation ith pressure retarded osmosis: An experimental and theoretical investigation, Journal of Membrane Science 343 (009) 4-5. [18] K. Gerstandt, K.V. Peinemann, S.E. Skilhagen, T. Thorsen, and T. Holt, Membrane processes in energy supply for an osmotic poer plant, Desalination 4 (008) [19] R.E. Pattle, Production of electric poer by mixing fresh and salt ater in the hydroelectric pile, Nature 174 (1954) 660. [0] S. Chou, R. Wang, L. Shi, Q. She, C. Tang, A. G. Fane, Thin-film composite hollo fiber membranes for pressure retarded osmosis (PRO) process ith high poer density, Journal of Membrane Science 389 (01) [1] G. Z. Ramon, B. J. Feinberg, E. M.V. Hoek, Membrane- based production of salinity- gradient poer, Energy & Environmental Science 4 (011) [] R. Semiat, J. Sapoznik, D. Hasson, Desalination and ater treatment 15 (010) [3] S. Loeb, F. Van Hessen, D. Shahaf, Production of energy from concentrated brines by pressure-retarded osmosis, II. Experimental results and projected energy costs, Journal of membrane Science 1 (1976) [4] S. Loeb, G.D. Mehta, A to coefficient ater transport equation for pressure retarded osmosis, Journal of Membrane Science 4 (1979) [5] G.D. Mehta, Further results on the performance of present-day osmotic membranes in various osmotic regions, Journal of Membrane Science 10 (198) Volume, Issue 1, December 013 Page 73

7 [6] G.D. Mehta, S. Loeb, Performance of permasep B-9 and B-10 membranes in various osmotic regions and at high osmotic pressures, Journal of Membrane Science 4 (1979) [7] H.H. Jellinek, H. Masuda, Osmo-poer. Theory and performance of an osmo-poer pilot plant, Ocean Engineering 8 (1981) [8] K.L. Lee, R.W. Baker, H.K. Lonsdale, Membrane for poer generation by pressure retarded osmosis, Journal of Membrane Science 8 (1981) [9] A. Achilli, T. Cath, A. Childress, Poer generation ith pressure retarded osmosis: An experimental and theoretical investigation, Journal of Membrane Science 343 (009) 4-5. [30] K. Gerstandt, K.V. Peinemann, S.E. Skilhagen, T. Thorsen, and T. Holt, Membrane processes in energy supply for an osmotic poer plant, Desalination 4 (008) [31] S. Chou, R. Wang, L. Shi, Q. She, C. Tang, A. G. Fane, Thin-film composite hollo fiber membranes for pressure retarded osmosis (PRO) process ith high poer density, Journal of Membrane Science 389 (01) [3] N. Yip, M. Elimelech, Thermodynamic and Energy Efficiency Analysis of Poer Generation from Natural Salinity Gradients by Pressure Retaded Osmosis, Energy & Environmental Science 46 (01) [33] Q. She, X. Jin, C. Tang, Osmotic poer production from salinity gradient resource by pressure retarded osmosis: Effect of operating conditions and reverse solute diffusion, Journal of Membrane Science (01) [34] Y. Kim, M. Elimelech, Journal of Membrane Science 49 (013) Author Mudher sabah received the B.S.degrees in physics science from Basrah university During , M.S. degrees in physics science(theoretical physics) from Al-Mustansiriya University. During , he stayed in university of Basrah college science-department of physics 005,he no in university of misan. Volume, Issue 1, December 013 Page 74