Dynamic modeling and simulation of a multi-effect distillation plant

Transcription

1 Dynamic moeling an simulation of a multi-effect istillation plant Liia Roca 1 Luis J. Yebra 1 Manuel Berenguel 2 Alberto e la Calle 1 1 CIEMAT-PSA, Ctra. Senés s/n, Tabernas, Almería, Spain Centro e Investigaciones Energéticas, MeioAmbientales y Tecnológicas Plataforma Solar e Almería 2 Dep. Lenguajes y Computación, Universia e Almería, Ctra. Sacramento s/n, Almería, Spain {liia.roca,luis.yebra,alberto.calle}@psa.es,beren@ual.es Abstract This paper escribes a moel which simulates the ynamics of a multi-effect istillation system in ifferent operating conitions. It has been esigne to improve the operation of the process an evelop a control strategy which optimizes the istillate prouction. The physical moels are base on conservation equations of mass an energy. They also inclue experimental correlations for heat transfer coefficients. Conservation laws are applie in ifferent components, such as the heater, the effects an the preheaters. The results of the mathematical moel simulation of the whole process show promising outcomes. Keywors: solar esalination, multi-effect istillation, moeling 1 Introuction One of the challenges toay is the prouction of freshwater for those population areas with high water stress. For places close to the sea, the esalination process provies an excellent way to tackle this problem. The use of esalination plants in these regions with plentiful seawater resources is becoming a technological way to prouce freshwater. Since large-scale esalination typically requires large amounts of energy, a solution is coupling esalination plants with renewable energies [10]. This process can be performe in various ways, for instance, using solar energy in which the source that provies the heat for the esalination process is collecte in a solar fiel. Multi-effect istillation plants (MED) raise a great interest in inustry ue to its efficiency when they are couple with a solar thermal system. This kin of systems is gaining more acceptance as a result of their lower energy requirements, higher heat transfer coefficients, compactness, high prouct water quality an low pre-treatment [2, 7]. In the literature there is a wie variety of steay-sate moels for MED plants [3, 5, 6, 9]. One of the last works is the one evelope in [13], which shows a hybri system that combines a esalination system with solar an win energies. In that paper, the moel inclues the istillation unit, the flat-plat collectors an the win system. Regaring ynamic moels, the literature about multi-effect istillation systems is scarce [4, 8]. The innovation of the present paper is that the ynamic moel has been evelope with the object-oriente Moelica language using the Dymola tool an the Moelica.Thermal library. This framework has allowe us to evelop new libraries to make simulations easier an improve the operating proceure. 2 Description of the system The AQUASOL system (Figure 1) at CIEMAT- Plataforma Solar e Almería (PSA), locate in the South of Spain, proposes a solar istillation technology that consists of a compoun parabolic collector (CPC) solar fiel, two 12m 3 water storage tanks, a multi-effect istillation unit with a 3m 3 /h nominal istillate prouction, an a ouble effect (LiBr-H 2 O) absorption heat pump (DEAHP) [1]. The esalination plant at CIEMAT-PSA is a forwar-fee multi-effect istillation unit manufacture an elivere by Weir ENTROPIE (Paris, France) in It has 14 cells, or effects, in a vertical arrangement. The original first cell that worke with low-pressure saturate steam (70 C, 0.31 bar [11]) DOI Proceeings of the 9 th International Moelica Conference 883 September 3-5, 2012, Munich, Germany

2 Dynamic moeling an simulation of a multi-effect istillation plant CPC SOLAR FIELD Gas Boiler V1 DEAHP Seconary Water Tank Primary Water Tank Figure 1: AQUASOL iagram V2 Effect1 MED Brine Distillate Seawater was replace in the AQUASOL project by a new one, which works with hot water coming irectly from a thermal storage tank. For optimal operation, the inlet fee-water temperature in the first cell must be aroun 66.5 C. It is possible to reach this temperature with heat from a solar fiel as well as with steam generate by an auxiliary gas boiler couple to a ouble effect absorption heat pump that can work at variable steam loas (from 30% to 100%). Seawater is preheate on its way towar the first cell of the plant, which is at the top of the esalination tower. Vapour is prouce in this first effect (or heater) using the hot water from the storage system. This vapour flows to the preheater-1 an part of the latent heat is transferre to the seawater that flows insie this preheater, increasing the temperature of the seawater. The steam prouce in the first effect goes to the effect-2, where it is conense in a tube bunle spraye with the more concentrate brine which falls by gravity from the previous effect. The latent heat release by conensation of the vapour allows part of the seawater entering the secon effect to evaporate at a lower temperature/pressure. This conensation/evaporation process is repeate in the successive effects. Finally, the vapour prouce in the effect-14 is conense in a final conenser coole by seawater. the preheater-14 is consiere as the final conenser. Each component of the plant (the heater, the effects an the preheaters) has been moele in the Moelica language using Moelica.Thermal library. Figure 2 shows the final moel of the MED unit, which inclues the heater, 13 effects an 14 preheaters. The inputs of the moel are the feewater mass flow rate to the heater, ṁ M, the inlet temperature to the heater, T im, the salt concentration of the seawater coming into the heater, C B0, the pressure in each effect, the seawater mass flow rate, ṁ sw, an the inlet seawater temperature to the preheaters 14 an 13. The outputs of the moel are the outlet temperature from the heater, T om, an the istillate prouction, ṁ. The correlations of the heat transfer coefficients inclue in the moels were obtaine from experiments carrie out in the real plant [12]. Nomenclature an subscripts are shown in Tables 1, 2. Table 1: Nomenclature Name Description Units A Surface area m 2 BPE Boiling Point Elevation K C Concentration % C p Specific heat capacity J/kgK T Temperature ifference between successive effects K h Specific entalphy J/kg ṁ Mass flow rate kg/s M Mass kg P Pressure Pa Q Heat transfer rate W T Temperature K U Overall heat trasfer coefficient W/m 2 K λ Latent heat of vaporization J/kg 3 The ynamic moel The moel of the MED unit is base on the following assumptions: no heat losses with the atmosphere, no flash vapour is prouce, the final istillate prouction is the sum of the istillate prouce in each effect, the temperature rop in each effect is equal to the temperature ifference in the preheaters, Name B e h i k M o p Table 2: Subscripts Description Brine Distillate Effect Heater Inlet Effect ientification number MED heating water Outlet Preheater 884 Proceeings of the 9 th International Moelica Conference DOI September 3-5, 2012, Munich Germany

3 Liia Roca, Luis J. Yebra, Manuel Berenguel an Alberto e La Calle Ambi1 mmi heater preheater1 mmo effect2 ambi2 effect3 effect4 effect5 effect6 effect7 effect8 effect9 effect10 effect11 effect12 effect13 T4 V3 T3 effect14 V2 T2 T1 Figure 2: Moelica moel of the MED unit sw v Seawater Saturate vapour Distillate mass flow prouce in the heater can be estimate using the latent heat of vaporization, λ: 3.1 The heater The first effect of the MED plant is the heater. Hot water coming from a storage system enters the heat exchanger an prouces the first evaporation of the seawater. Fig.3 shows the moel of the heater. The heat transfer rate for the first effect can be calculate from the MED heating water mass flow rate, ṁ M, an the MED heating water temperature ifference in stationary conitions as follows: Q h = ṁ M C p (T im T om ) (1) Using the log-mean temperature an the overall heat transfer coefficient, U h, the heat transfer rate can be written as: Q h = U h A h (T im T v1 ) (T om T v1 ) ln T im T v1 (2) T om T v1 where T v1 is the saturation temperature of the vapour generate in the heater. ṁ h = Q h λ (3) Since the vapour pressure of the aqueous solution is lower than that of pure water at the same temperature, the boiling point of the solution will be higher than that of the water. Therefore, the temperature of the brine can be obtaine using the boiling point elevation, BPE: T B1 = T v1 + BPE (4) The BPE is a brine property an epens on the brine salinity an temperature. The mass flow rate an concentration of the brine can be obtaine applying mass an energy balances. Mass balance: Salt mass balance: t (M B1) = ṁ sw ṁ B1 ṁ 1 (5) t (M B1 C B1 ) = ṁ sw C B0 ṁ B1 C B1 (6) DOI Proceeings of the 9 th International Moelica Conference 885 September 3-5, 2012, Munich, Germany

4 Dynamic moeling an simulation of a multi-effect istillation plant TiM mm CB0 Tsw1 msw P Using the overall heat transfer coefficient: Q pk = U pk A p (T p(k 1) T vk ) (T pk T vk ) ln T p(k 1) T vk (9) T pk T vk 3.3 The effects ToM mm TB1 mb1 Tv1 m1 The vapour that has not been conense in the preheater flows to the following effect, where the seawater with a higher brine concentration flows by gravity from the previous effect. Then, the vapour conenses an transfers its latent heat to the seawater proucing a new evaporation. Fig.5 shows the moel of one effect. CB1 CB(k-1) TB(k-1) mb(k-1) Figure 3: Moelica moel of the heater T Energy balance: t (M B1 h B1 ) = ṁ sw h sw ṁ B1 h B1 ṁ 1 h v1 (7) Tvk mk P 3.2 The preheaters CB(k) TB(k) mb(k) The vapour prouce in the heater flows to the preheater-1 locate besies, where it conenses as the temperature of the seawater that flows insie the preheater tubes increases. This process is repeate in the successive effects an preheaters. Figure 4 shows the moel of the preheater. T Tvk mk Tp(k-1) msw P Tp(k) msw Figure 4: Moelica moel of the preheater The heat transfer rate for each k-preheater was calculate using the measure seawater mass flow rate, ṁ sw, an the temperature ifference between the outlet an the inlet: Q pk = ṁ sw C p (T p(k 1) T pk ) (8) Figure 5: Moelica moel of the effect The heat transfer rate equation for each k-effect evaporator is: Q ek = U ek A e (T k + BPE) (10) where T is the temperature ifference between successive effects, which is calculate in the prehetear component. The istillate mass flow rate in the k-effect is: ṁ k = Q ek (11) λ The moel of each cell or effect is base on mass an energy balances taking into account the istillate prouce an the brine mass flow rate from the previous cell: t (M Bk) = ṁ B(k 1) ṁ Bk ṁ k (12) t (M Bk C Bk ) = ṁ B(k 1) C B(k 1) ṁ Bk C Bk (13) t (M Bk h Bk ) = ṁ B(k 1) h B(k 1) ṁ Bk h Bk ṁ k h vk (14) 886 Proceeings of the 9 th International Moelica Conference DOI September 3-5, 2012, Munich Germany

5 Liia Roca, Luis J. Yebra, Manuel Berenguel an Alberto e La Calle 4 Simulation results The evelope moel can be use to improve the operation of the plant, stuying the effect of the variation in the operating conitions on the MED unit performance an prouction rate. The final istillate prouction will be the sum of the amounts of vapour prouce in each effect as follows: k=14 ṁ = ṁ h + k=2 ṁ k (15) Figure 6 shows the results obtaine simulating the evelope moel an using the following inputs: MED heating water mass flow rate, ṁ M, is 12 kg/s, MED inlet heating water temperature, T im, varies between 338 an 345 K (as shown in Fig. 6) seawater mass flow rate insie preheaters 1-13, ṁ sw, is 1.94 kg/s, preheater-13 inlet seawater temperature, T p14 is about 303 K (see Fig.6). As it can be observe in Fig.6, the MED outlet heating water temperature, T om, is about 3.3 K less than T im. Nevertheless, if the inlet temperature increases, this ifference also increases slightly. As it was expecte, higher temperatures cause higher thermal consumption. On the other han, the higher the inlet temperature, the more istillate is prouce. Therefore, the moel may be an efficient tool to estimate the thermal consumption epening on the etillate eman. This means that we can preict if the solar resource is enough to reach the prouction goals or if we shoul combine it with the use of the heat pump. 5 Conclusions In this paper, a multi-effect istillation unit has been moele. Physical equations for each one of the main components (the heater, the effect an the preheater) have been evelope using the object-oriente Moelica language. The whole plant has been efine with multiple instances of the effect an preheater subsystems properly interconnecte between them. First simulation results are promising an the moel may be use to improve the operation in the real plant. The main purpose of the moel is the preiction of the thermal ynamics of the heater as well as the preiction of the istillate prouction rate. Acknowlegements The authors woul like to thank the CIEMAT Research Centre an the Spanish Ministry of Economy an Competitiveness for funing Project DPI C References [1] D. Alarcón-Pailla, L. García-Roríguez, an J. Blanco-Gálvez. Assesment of an absorption heat pump couple to a multi-effect istillation unit within aquasol project. Desalination, 212: , [2] M.T. Ali, H.E.S. Fath, an P.R. Armstrong. A comprehensive techno-economical review of inirect solar esalination. Renewable an Sustainable Energy Reviews, 15(8): , [3] N.H. Aly an A.K. El-Figi. Thermal performance of seawater esalination systems. Desalination, 158(1-3): , [4] N.H. Aly an MA Marwan. Dynamic response of multi-effect evaporators. Desalination, 114(2): , [5] H. El-Dessouky, I. Alatiqi, S. Bingulac, an H. Ettouney. Steay-state analysis of the multiple effect evaporation esalination process. Chemical engineering & technology, 21(5):437, [6] A.M. El-Nashar. Preicting part loa performance of small me evaporators-a simple simulation program an its experimental verification. Desalination, 130(3): , [7] M.A. Eltawil, Z. Zhengming, an L. Yuan. A review of renewable energy technologies integrate with esalination systems. Renewable an Sustainable Energy Reviews, 13(9): , [8] A. Husain. Integrate Power an Desalination Plants. EOLSS Publishers Lt., [9] MH Khaemi, MR Rahimpour, an A. Jahanmiri. Simulation an optimization of a six-effect evaporator in a esalination process. Chemical Engineering an Processing: Process Intensification, 48(1): , DOI Proceeings of the 9 th International Moelica Conference 887 September 3-5, 2012, Munich, Germany

6 Dynamic moeling an simulation of a multi-effect istillation plant 305 T swi13 [K] Heater temperatures [K] T im T om m [m 3 /h] Relative time [s] Figure 6: Simulation results of the MED unit moel [10] E. Mathioulakis, V. Belessiotis, an E. Delyannis. Desalination by using alternative energy: Review an state-of-the-art. Desalination, 203: , [11] B. Milow an E. Zarza. Avance me solar esalination plants. configurations, costs, future seven years of experience at the plataforma solar e almeria (spain). Desalination, 108(1-3):51 58, [12] P. Palenzuela, D. Alarcon, J. Blanco, E. Guillen, M. Ibarra, an G. Zaragoza. Moeling of the heat transfer of a solar multi-effect istillation plant at the plataforma solar e almeria. Desalination an water treatment, 31(1-3): , [13] İ.H. Yılmaz an M.S. Söylemez. Design an computer simulation on multi-effect evaporation seawater esalination system using hybri renewable energy sources in turkey. Desalination, Proceeings of the 9 th International Moelica Conference DOI September 3-5, 2012, Munich Germany