Performance study of a pilot-scale multi-effect vacuum membrane distillation desalination plant

DES-12990; No of Pages 9 Desalination xxx (2016) xxx–xxx

Contents lists available at ScienceDirect

Desalination journal homepage: www.elsevier.com/locate/desal

Performance study of a pilot-scale multi-effect vacuum membrane distillation desalination plant Yu-lei Xing ⁎, Chun-hua Qi, Hou-jun Feng, Qing-chun Lv, Guo-rong Xu, Hong-qing Lv, Xin Wang The Institute of Seawater Desalination and Multipurpose Utilization, SOA, Tianjin 300192, China

H I G H L I G H T S • A hybrid system composed of MEMD and MSF was developed. The water production and GOR are 2.0 t/d and 2.76, respectively. • Water production and membrane flux exponentially increases by 55%, as feed temperature improves from 75 °C to 90 °C. • Water production increases 25% and GOR decreases 4.5%, as vacuum degree in permeation side improves from 70 kPa to 82 kPa.

a r t i c l e

i n f o

Article history: Received 31 March 2016 Received in revised form 1 July 2016 Accepted 7 July 2016 Available online xxxx Keywords: Desalination Multi-effect vacuum membrane distillation Multi-stage flash distillation Heat recovery Pilot plant experiment

a b s t r a c t A 2.0 t/d desalination plant integrated with multi-effect membrane distillation (MEMD) and multi-stage flash (MSF) evaporation was developed, utilizing polytetrafluoroethylene (PTFE) hollow fiber membranes. Performances tests and the effects of operating parameters on performance were analyzed. Results showed that the plant could run automatically and stably, and reached a gained output ratio (GOR) of 2.76. And water production and GOR were affected by operating temperature and temperature difference between effects. Water production increased by 12.7% as operating temperature increased from 60 °C to 80 °C, and reduced by 9.5% as temperature difference increased from 4.5 °C to 8.0 °C. Moreover, the feed temperature and vacuum degree also had strong effects on water production. As feed temperature increased from 75 °C to 90 °C, water production grew exponentially by 55%, while membrane flux increased from 3.95 kg/m2·h to 6.12 kg/m2·h. Furthermore, as vacuum degree increased from 70 kPa to 82 kPa, water production increased linearly by 25%, while GOR decreased by 4.5%. Lastly, feed flow and salinity had minor effects on performance. These results provide a useful reference for the application of MEMD hybrid system. © 2016 Elsevier B.V. All rights reserved.

1. Introduction Membrane distillation (MD) is developed in early 1960s pioneered by Weyl and Findley [1,2]. It is a hybrid thermal and membrane process [3], and refers to a thermally driven transport of volatile components in feed solution through non-wetted porous hydrophobic membranes by vapor pressure difference between two sides of the membrane pores as a driving force. MD contains not only evaporation, heat and mass transfer and condensation process in conventional distillation, but also membrane separation based on substance diffusion [4]. Generally, MD contains four kinds of configurations, Direct Contact (DCMD), Air Gap (AGMD), Sweeping Gas (SGMD) and Vacuum (VMD). MD has many advantages, such as gentle operation conditions, high separation efficiency, simple technological process and environmental-

⁎ Corresponding author at: 1# Keyan East Road, Nankai District, Tianjin 300192, China. E-mail address: [email protected] (Y. Xing).

friendly [5]. The major advantage of MD over other conventional processes such as thermal process is its lower operating temperature which directly utilizes low grade heat source [6–8]. MD can treat highly concentrated saline solutions of salts without suffering from a big drop in productivity compared with other membrane separation processes such as RO [9,10]. With these advantages, the MD applications are very appropriate for environmental, pharmaceutical, chemical, food, and biotechnology industries [11–13]. It gradually becomes the research hotspot in wastewater treatment [14], concentration separation [15] and desalination [16,17]. A hybrid system of MD with forward osmosis (FO) can be utilized to treat human urine and displays great prospect in some key field such as water regeneration in space station [18]. MEMD system for desalination can improve GOR obviously with multiple recoveries of condensation heat to the feed seawater [19,20]. However, in practice, the use of MD for desalination is restricted to pilot and small scale lab units due to few technical challenges [21,22], such as low thermal efficiency [23,24], low membrane flux [25,26],

http://dx.doi.org/10.1016/j.desal.2016.07.008 0011-9164/© 2016 Elsevier B.V. All rights reserved.

Please cite this article as: Y. Xing, et al., Performance study of a pilot-scale multi-effect vacuum membrane distillation desalination plant, Desalination (2016), http://dx.doi.org/10.1016/j.desal.2016.07.008

2

Y. Xing et al. / Desalination xxx (2016) xxx–xxx

and shortage of ideal membrane material [27,28]. Careful integration of MD in other separation technologies such as MED, MSF and RO could make it efficient and convenient [28]. A combined one-stage MED/MD hybrid system working at atmospheric pressure was proposed by Andre's et al. [29] and the production of distilled water is increased about 7.5% and the GOR of the whole system is improved by about 10%. Manna and Pal researched a new flash vaporization MD module which is using solar energy for phase change to drive DCMD [30]. This is a hybrid of MSF and DCMD which can be enhance the membrane flux, the highest flux up to 52.94 kg/m2·h. Fard et al. investigated the flux decline and fouling of DCMD by utilizing thermal brine from MSF desalination plant [31]. MEMD can reuse the latent heat in MD process which would reduce the energy consumption greatly [32]. Chung et al. studied a MEMD systems coupling with distillate flashing for high salinity applications, which is thermodynamically similar to MSF [33]. With advantages of lower capital requirement and scalability, the system can be an attractive alternative to conventional thermal desalination systems. In a word, the related researches by scholars have laid a good foundation for application of the MD technology. However, few industrial application projects of MD have been reported yet due to those technical challenges. To promote MD development and application in desalination, this paper developed a new desalination process through coupling MEMD and MSF, which could reuse latent and sensible heat. A desalination pilot plant based on this new process was constructed, and the influences of technological parameters and operating conditions on its performance were assessed. This research could provide references for integration, optimization, and amplification of MD systems, and guide industrial MEMD application.

2. Mathematical modeling Previous publications presented design procedures and mathematical models for different MEMD systems [29,33]. These procedures and models are the base of design and development for a MD hybrid system with other processes. The mass balance of water and salt, the energy balance equation, and heat transfer correlations constitute the complete set of equations that need to be solved for the MEMD system design. (1) The pilot test material balance flow is shown in Fig. 1. Mass and energy balance equations for each effect MD are similar for the countercurrent flow system. In this paper, we provide material balance in a single effect MD as shown in Fig. 1. (2) Calculation of material and heat balance for each effect MD.

Feed flow and its salinity balance are expressed as follows: Gi−1 ¼ Gi −Di ¼ G0 −

n X

ð1Þ

Dj

j¼iþ1

Gi−1  C i−1 ¼ Gi  C i ¼ G0  C 0

ð2Þ

where G, G0, D, C0, C are makeup feed flow, initial makeup feed flow, vapor quantity of MD, initial seawater salinity, salinity, respectively. The subscript i(i ≧ 2) is representative effect order of MD. Flashing brine flow and its salinity balance are expressed as follows: G″i ¼ G″i−1 −g i ¼ G0 −

n X

D j−

j¼1

i X

gj

ð3Þ

j¼1

G″i  C ″i ¼ G″i−1  C ″i−1 ¼ G0  C 0

ð4Þ

where G″, C″ and g are the flashing brine flow, salinity in brine and flash vapor quantity of brine, respectively. Superscripts ″ is representative the flash evaporation of brine. Heat balance is expressed according to Q i ¼ Di−1  λi−1 þ di−1  λ0i−1 þ gi−1  λ″i−1 ¼ ðGi þ Gr Þ  cp;i ðT i1 −T i2 Þ ð5Þ where Gr, Q, d, cp, λ, T1 and T2 are circulation feed flow, heat quantity of feed preheating, flash vapor quantity of distillate water, specific heat capacity at constant pressure, latent heat of vaporization, feed inlet temperature and feed outlet temperature of membrane module, respectively. Superscript' is representative the flash evaporation of distillate water. (3) The brine flash evaporation heat balance for each effect is calculated as follows: 0

gi 

λ″i

1 n i−1 h  i X X @ ð6Þ ¼ G0 − D j− g j A  cp; ″i−1  t i−1 −cp; ″i  t i þ ΔNEA″i j¼1

j¼1

where ti and ΔNEA″i are the brine temperature in the flash tank and nonequilibrium temperature rise of brine flash evaporation, respectively [31].

Fig. 1. Material balance flow.

Please cite this article as: Y. Xing, et al., Performance study of a pilot-scale multi-effect vacuum membrane distillation desalination plant, Desalination (2016), http://dx.doi.org/10.1016/j.desal.2016.07.008

Y. Xing et al. / Desalination xxx (2016) xxx–xxx

(4) The distilled water flash evaporation heat balance of each effect is calculated as follows: di  λ0i ¼

X    ðDi−1 þ g i−1 Þ  cp; 0i−1 T c;i−1 −cp; 0i T c;i þ ΔNEA0i

T i ¼ t i −BPEi

ð8Þ

T c;i ¼ T i −ΔT i

ð9Þ

where Ti, BPEi and ΔTi are the evaporation temperature, boiling point elevation of brine [31] and temperature difference loss caused by flow resistance, respectively. (6) Thermal efficiency of membrane module. Heat transfer within the membrane is due to latent heat accompanying vapor or gas flux through membrane pores and heat transferred by conduction across both the membrane material and gas-filled membrane pores. Therefore, the total heat transmitted in MD is expressed as [28]:  km  fm þ J i λi T i −T pm i δ

ð10Þ

pm where km, δ, Tfm and Ji are the thermal conductivity of membrane i , Ti material, membrane thickness, surface temperature of membrane at the liquid side, surface temperature of membrane at the permeation side and membrane flux respectively. The calculations in detail about Tfm i and Tpm i , please see [28]. According to contributions of heat transfer to mass transfer, the thermal efficiency of membrane component η is expressed as

ηð%Þ ¼

J i λi  100  km  fm þ J T i −T pm λ i i i δ

ð11Þ

(7) The membrane flux of each MD is expressed as Ji ¼

Di Am i t

ð12Þ

where Ji, Di, Am i and t are the membrane flux, condensation water quantity at the permeation side during test time, membrane area and test time, respectively. (8) The heat transfer equation of each effect preheater is as follows: Q i ¼ K i  Ai  ðLMTDÞi

ð13Þ

where Ki, Ai and LMTD are the heat transfer coefficient, heat transfer area and logarithmic heat transfer temperature difference of the preheater, respectively. (9) Lastly, the GOR of the system is expressed as X GOR ¼

Di

D0

where D0 is input stream flow for MEMD.

3. Experimental 3.1. Experimental apparatus and materials

ð7Þ

where Tc,i and ΔNEA′i are the vapor condensation temperature and nonequilibrium temperature rise of distilled water flash evaporation, respectively [31]. (5) The effective evaporation and condensation temperatures are defined as follows [32]:

Qi ¼

3

ð14Þ

The pilot scale plant coupled three effects MD in series and three stages MSF in series, with water production of 2.0 t/d. Every effect MD was composed of two hollow fiber membrane modules in parallel, the feed solution was circulated independently in every effect, and the makeup feed solution was flowing in countercurrent between effects, as shown in Fig. 1. The hollow fiber used in the MEMD system was made of polytetrafluoroethylene (PTFE). PTFE is a kind of very good membrane material because it is highly hydrophobic, and has good chemical resistance and high thermal stability [27]. The main parameters of the hollow fiber and membrane modules are listed in Table 1. The membrane scanning electron microscope (SEM) is shown as Fig. 2. In the MEMD system, the membrane modules, preheater, feed tank and circulating pump were connected through pipelines, forming a single effect MD. Then, three-effect MDs were connected in series, forming an MEMD process. Finally, it was coupled with an MSF process used for flash evaporation of distilled water and brine from the MEMD system. The integrated processes of MEMD and MSF are shown in Figs. 3 and 4. Seawater in the first-effect MD was heated using boiler steam, and the secondary steam generated in the first-effect MD was used to heat seawater in the second-effect MD. Similarly, the steam generated in the second-effect MD heated seawater in the third-effect MD. In this way, it was realized that recycling and gradient utilization of latent heat in MD pervaporation. Then, the steam generated in the third-effect MD entered the condenser to preheat makeup seawater, which was subsequently evaporated and concentrated in the third-effect MD. Concentrated brine from the third-effect MD flowed in countercurrent into the second-effect MD and back into the first-effect MD, where it was finally discharged. Then, the discharged brine flowed by pressure difference through the two-stage brine flash tanks successively for recycling additional sensible heat. Similarly, the distilled water flowed through the three-stage freshwater flash tank successively for recycling additional sensible heat. The saturated vapor from flash tanks was also used to heat the feed solution of tanks` corresponding effect MD. Implementing the above processes saw significant improvements in thermal efficiency and GOR for the MEMD system. The pilot plant control system adopted the monitoring patterns of an industrial process computer (IPC), a programmable logic controller (PLC), and field instrumentation combined. This system accomplished signal acquisition, logic control, complex operations, chain protection, and data management for the entire process system, and guaranteed the plant's automatic production and stable operation. The three-effect MD control system interface is shown in Fig. 5. Experiments measured multiple factors including temperature, pressure, flow rate, liquid level, conductivity, and pH. Temperature and pressure were measured from measuring points in the MD modules, flash tank, feed tank, and process pipes. Flow rates of seawater, brine, and product water were measured by the electromagnetic flow meters on pipelines. Heating steam flow was measured by a vortex shedding flowmeter (VSF) with temperature and pressure compensations. Measuring error was controlled at ± 1% through measuring instruments calibration. Online conductivity and pH instruments were placed at the product water pump outlet to monitor water quality. All Table 1 Hollow fiber and membrane module design parameters. Item

Unit

Technical parameters

Fiber inner diameter Membrane thickness Mean pore size Membrane Porosity Fiber number Effective Fiber Length

mm mm μm % pcs m

0.8 0.4 0.2 50 2000 0.7

Please cite this article as: Y. Xing, et al., Performance study of a pilot-scale multi-effect vacuum membrane distillation desalination plant, Desalination (2016), http://dx.doi.org/10.1016/j.desal.2016.07.008

4

Y. Xing et al. / Desalination xxx (2016) xxx–xxx

Fig. 2. SEM of the PTFE hollow fiber membrane.

operations and monitoring functions during experimentation were accomplished on the upper computer monitor screen and all measured data were automatically recorded, stored, and analyzed. 3.2. Experimental procedure and operating parameters This study was designed to test performance of large-scale MEMD desalination devices and study the effects of key technical parameters and operating factors on device performance under large operating parameters. First, the plant was initiated in a cold-state operating mode and after that, came into hot-state operation by inputting heat steam. Then, running parameters (i.e., evaporating pressure, operating temperature, heating transfer temperature difference, and flow rate) were gradually adjusted to their predetermined values. Next, the automatic operation mode was selected, and the system reached a steady state after 1 h of operation. Finally, the influencing parameters were tested. The electric valve and transducer were adjusted to alter experimental parameters. Initially, the pilot plant ran for 168 h to test the MEMD system's operating stability and water production performances. During testing, raw seawater first flowed through sodium ion exchange resin to removal calcium and magnesium ions, which can prevent scale buildup on the hollow fiber membrane. The feed temperature at the first-effect MD's

inlet, evaporating temperature in the permeation side of the first-effect MD, and the temperature difference between effects were set to 75 °C, 70 °C, and 6 °C, respectively. Conversely, the feed flow rate and salinity were set to 6 m3/h and 3.5%, respectively. Additionally, adjusting vacuum pressure controlled the evaporating temperature in the MD's permeation side.

4. Results and discussion 4.1. The performances of water production and stability test The water production and stability test results are shown in Figs. 6 and 7. Fig. 6 illustrates good full-flow performance and automatic stable operation. Operating pressure and temperature in each effect MD were approximately constant. Pressure and temperature differences between effects distributed uniformly. The third-effect was the closest to the vacuum pump and, consequently, its pressure fluctuated. However, the plant adapted to these changes and maintained stable operation. Fig. 7 reflects that the plant's water production was consistent at about 2.0 t/ d, which met the design index. Furthermore, the GOR and average membrane flux were 2.76 and 3.95 kg/m2·h, respectively. Lastly, the

Fig. 3. Schematic diagram of three-effect MD pilot plan.

Please cite this article as: Y. Xing, et al., Performance study of a pilot-scale multi-effect vacuum membrane distillation desalination plant, Desalination (2016), http://dx.doi.org/10.1016/j.desal.2016.07.008

Y. Xing et al. / Desalination xxx (2016) xxx–xxx

5

Fig. 4. Pilot plant of three-effect MD seawater desalination.

product water quality was a consistent 10 μs/cm even after running for 80 h. According to performance test results, the pilot plant for seawater desalination based on a three-effect MD achieved satisfying full flow stability. Water production met the design load and could the system run automatically and stably. These results laid a good foundation for the following pilot conditions experimental study of seawater desalination based on MEMD. 4.2. Effect of evaporating temperature on water production and GOR Conditioning tests were conducted under different evaporating temperatures (80 °C, 70 °C and 60 °C) at the permeation side of the first-effect MD. In these tests, transmembrane temperature differences the

first-effect MD's inlet, temperature differences between effects, feed solution flow, and salinity, were set to 5 °C, 6 °C, 6 m3/h, and 3.5%, respectively. Test results are shown in Fig. 8. Both water production and GOR increased as evaporating temperature increased. When evaporating temperature increased from 60 °C to 80 °C, water production increased from 1.89 t/d to 2.13 t/d (mean membrane flux increased from 3.73 kg/m2·h to 4.21 kg/m2·h), and GOR increased from 2.68 to 2.79. In MD, higher evaporating temperature resulted in smaller liquid viscosity and surface tension at both sides of membrane, but flow velocity and disturbance inside the boundary layer increased [35], which was conducive to thin boundary layers and weakened temperature polarization effects. This strengthened heat and mass transfer. In addition, as the evaporating temperature increased, the total heat transfer coefficient of the preheater and heat

Fig. 5. Process control flow chart of three-effect MD plant.

Please cite this article as: Y. Xing, et al., Performance study of a pilot-scale multi-effect vacuum membrane distillation desalination plant, Desalination (2016), http://dx.doi.org/10.1016/j.desal.2016.07.008

6

Y. Xing et al. / Desalination xxx (2016) xxx–xxx

Fig. 6. Variations of pressure and temperature of each effect with running time.

utilization efficiency of secondary vapor were enhanced. So water production increased with increasing evaporating temperature. Therefore, water production was positively correlated with operating temperature. Furthermore, the system consumed less heating steam to evaporate an equal mass of product water at high temperature. Therefore, GOR also increased as evaporating temperature increased. 4.3. Effect of temperature difference between effects on water production and GOR

Additionally, the smaller the heat transfer temperature difference, the higher the thermal efficiency. This is because there was greater irreversibility and heat loss under larger temperature differences. Considering the circulated feed flow and temperature difference loss caused by BPE, heat transfer temperature differences shall not be too small. Conclusively, the multi-effect MD process shall choose the appropriate temperature difference according to the specific experimental conditions. 4.4. Effect of feed temperature on water production and GOR

Conditioning tests were conducted under varying temperature differences (4.5 °C, 6 °C, and 8 °C) between MD effects. In these tests, temperature of the first-effect MD at the permeation side, and its transmembrane temperature difference at the inlet were set to 70 °C and 5 °C, respectively. The feed solution flow and salinity were set to 6 m3/h and 3.5%, respectively. The results are shown in Fig. 9. In Fig. 9, water production and GOR decreased as the heat transfer temperature difference increased. When the heat transfer temperature difference increased from 4.5 °C to 8 °C, water production decreased from 2.1 t/d to about 1.9 t/d, and GOR fell from 2.79 to 2.68. Optimizations of heat transfer and thermophysical properties reflect the influence of temperature difference between effects on the multi-effect MD process [33,36]. In the first-effect MD, given a fixed feed solution temperature and evaporating temperature, the larger the temperature difference between effects, the lower the feed solution temperature and evaporating temperature of its subsequent effects. This has two consequences: on one hand, water activity in saline solution reduced, which is disadvantageous for heat and mass transfer; on the other hand, larger latent heats of vaporization accompanied lower temperature. Given the same quantity of steam, it will evaporate less second vapor.

Conditioning tests under different feed temperatures (75 °C, 80 °C, 85 °C, and 90 °C) of the first-effect MD were carried out. In these tests, the evaporating temperature at the permeation side, temperature difference between effects, feed solution flow, and salinity were set to 70 °C, 6 °C, 6 m3/h, and 3.5%, respectively. Results are presented in Fig. 10. With an increase in feed temperature, water production experiences exponential growth. As feed solution temperature of the first-effect MD increased from 75 °C to 90 °C, water production increased from 2.0 t/d to 3.1 t/d (mean membrane flux increased from 3.95 kg/m2·h to 6.12 kg/m2·h). MD is a membrane separation process that uses temperature as a driving force. In the mathematical model of vacuum MD, membrane flux is proportional to partial pressure difference of transmembrane vapor. Given a fixed feed solution flow and evaporating temperature at the permeation side, membrane flux varies when partial pressure of vapor on the membrane surface at the feed solution side changes. According to the Antoine equation, there is an exponential relationship between partial pressure of vapor on this membrane surface and local temperature [37]. Therefore, membrane surface temperature

Fig. 7. Variation of product water production with running time.

Fig. 8. Effect of evaporating temperature on water production and GOR.

Please cite this article as: Y. Xing, et al., Performance study of a pilot-scale multi-effect vacuum membrane distillation desalination plant, Desalination (2016), http://dx.doi.org/10.1016/j.desal.2016.07.008

Y. Xing et al. / Desalination xxx (2016) xxx–xxx

7

Fig. 9. Effect of temperature difference between effects on water production and GOR. Fig. 11. Effect of vacuum degree on water production and water production.

and partial pressure of vapor are positively correlated with feed solution temperature, thus enhancing the driving force of transmembrane heat transfer and significantly increasing flux. Lastly, GOR increased from 2.76 to about 2.83 with the temperature rise. Therefore, the feed solution temperature is a key parameter for MEMD systems. 4.5. Effect of vacuum degree on water production and GOR Conditioning tests were carried out under different vacuum degrees (68–82 kPa) of the first-effect MD. In these tests, feed solution temperature of the first-effect MD, temperature difference between effects, feed solution flow, and salinity were set to 75 °C, 6 °C, 6 m3/h and 3.5%, respectively. Test results are displayed in Fig. 11. Vacuum degree significantly influenced MEMD water production and GOR. During vacuum MD, the mass transfer driving force was proportional to the pressure difference between the partial pressure of vapor at the feed solution side and vacuum pressure at the permeation side [37]. Maintaining feed solution temperature and circular flow, the lower the vacuum pressure, the larger the transmembrane pressure difference and the stronger the mass transfer driving force. As a result, permeation flux and water production of the system increased. As shown in Fig. 11, water production increased from 2.0 t/d to about 2.5 t/d as vacuum degree increased, improving by 25%. An increase in permeation flux indicates more latent heat of vaporization was transmitted from the feed solution side to the permeation side. This enhanced polarization effects of membrane surface temperature at the feed solution side, which is disadvantageous for heat and mass transfer. To maintain feed solution temperature and overcome this temperature polarization effect, more heating steam must be inputted. However, water production growth was smaller than steam consumption growth, thus reducing GOR. For example, GOR decreased by 4.5% when vacuum degree increased from 70 kPa to 82 kPa in the permeation side.

Fig. 10. Effect of feed temperature on water production and GOR.

4.6. Effect of feed solution flow on water production and GOR Conditioning tests were conducted under different feed solution flows (4.5–8 m3/h). In these tests, the evaporating temperature of the first-effect MD at the permeation side, transmembrane temperature difference, temperature difference between effects, and salinity were set 70 °C, 5 °C, 6 °C, and 3.5%, respectively. Results are shown in Fig. 12. Feed solution flow is another important influencing factor for MEMD processes. As shown in Fig. 12, water production increased from 1.9 t/d to about 2.15 t/d when feed solution flow increased from 4.5 m3/h to 8 m3/h. The greater the feed solution flow, the greater the flow velocity of feed solution inside the hollow fiber membrane. On one hand, flow velocity growth could disturb and thin temperature boundary layers, which could make the heat transfer coefficient inside boundary layers grow linearly [34], especially under laminar or transition flow states. This could intensify heat transfer from bulk feed solution to the membrane surface and weaken temperature polarization effects. Under these conditions, membrane surface temperature rises and permeation flux increases. On the other hand, flow velocity growth could reduce mass transfer resistance inside boundary layers, weaken concentration polarization effects, and increase membrane flux. However, to maintain feed solution temperature, higher flow requires more heating steam. Lastly, water production growth was slightly lower than that for heating steam, which led to a small GOR reduction.

4.7. Effect of feed solution salinity on water production and GOR Conditioning tests were conducted under different salinities (3.5– 10%). In these tests, the transmembrane temperature difference of the first-effect MD, evaporating temperature at the permeation side, temperature difference between effects, and feed solution flow were set to 8 °C, 70 °C, 8 °C, and 6 m3/h, respectively. Results are shown in Fig. 13.

Fig. 12. Effect of feed solution flow on water production and GOR.

Please cite this article as: Y. Xing, et al., Performance study of a pilot-scale multi-effect vacuum membrane distillation desalination plant, Desalination (2016), http://dx.doi.org/10.1016/j.desal.2016.07.008

8

Y. Xing et al. / Desalination xxx (2016) xxx–xxx

(2)

(3)

Fig. 13. Effect of feed salinity on water production and GOR.

Water production experienced a weak exponential reduction as salinity increased. Although yield changed only slightly under low salinity, it dropped greatly under high salinity. When seawater salinity increased from 3.5% to 10%, water production reduced from 2.05 t/d to about 1.9 t/ d. Salinity may affect MD in the following three ways. First, the driving force in the MD process decreased. Water activity in feed solution decreased as salinity increased, decreasing water vapor partial pressure and transmembrane pressure differences [33,36]. Second, the concentration polarization effect strengthens. Since increases in salinity enhances non-volatile solutes at the feed/membrane interface, the overall mass transfer resistance in the concentration boundary layers adjoining the membrane increases, thereby strengthening the concentration polarization effect [33,38]. Third, thermal efficiency and the temperature polarization coefficient decreases. Viscosity increases with greater salinity, which results in the convection heat transfer coefficient in the thermal boundary layer decreasing [39]. Furthermore, the temperature at the membrane surface is much lower than that in the bulk feed solution. Due to these reasons, membrane flux and water production reduces with an increase in salinity. As shown in Fig. 13, GOR also decreased as salinity increased. This is mainly due to the increase in BPE caused by greater salinity. BPE reduces both membrane flux and the effective heat transfer temperature difference driving the MD process [33]. The BPE makes the MEMD system require additional heat input for the same water production, which therefore results in GOR decrease. Lastly, the influence of salinity on conductivity of distilled water was not obvious in experiment. Conductivity was maintained around 10 μs/ cm. Therefore, the PTFE hollow fiber membrane was strongly hydrophobic and the membrane pores were not easily wet. However, the risk of the membrane pores wetting increases with greater transmembrane pressure difference and a high concentration ratio [28]. To decrease this risk, operating conditions were maintained within an appropriate range. 5. Conclusions Experimental studies were conducted on a MEMD pilot plant that couples the MSF process. Firstly, the performance tests demonstrated the feasibility and stability for 168 h about the integration, optimization, and amplification of MEMD/MSF hybrid processes. Secondly, condition tests were conducted to study the influence of operation conditions on water production. The recovery and reuse of the latent heat of vaporization in the MD process were also effectively solved. Furthermore, the sensible heat of discharging brine and distilled water was reused in MSF process. Therefore, thermal efficiency and GOR were significantly improved. Specific conclusions are as follows. (1) Both water production and GOR increased as MD operation temperature increased. When the evaporating temperature changed from 60 °C to 80 °C, water production increased by 12.7%, and the

(4)

(5)

(6)

mean membrane flux increased from 3.73 kg/m2·h to 4.21 kg/ m2·h. Water production and GOR decreased as heat transfer temperature difference between effects increased. When heat transfer temperature difference increased from 4.5 °C to 8 °C, water production decreased from 2.1 t/d to about 1.9 t/d. This was observed due to greater irreversibility and heat loss under the larger temperature difference. Given fixed vacuum pressure at the permeation side, water production grew exponentially as feed solution temperature increased. This is because maintaining temperature at the permeate side causes the transmembrane mass transfer driving force to increase exponentially as feed solution temperature increases. For example, as feed solution temperature increased from 75 °C to 90 °C, water production increased from 2.0 t/d to 3.1 t/d, and mean membrane flux increased from 3.95 kg/m2·h to 6.12 kg/m2·h, showing a 55% improvement. Although water production significantly increased with rising vacuum degree at the permeation side, GOR decreased. Increasing vacuum degree enhanced the transmembrane mass transfer driving force, but also resulted in strengthening the temperature polarization effect. When vacuum degree increased from 68 kpa to 82 kpa, water production increased by about 25%, while GOR dropped about 4.5%. Water production slowly increased while GOR slightly decreased when feed flow was increased. The feed flow increase improved flow velocity in the feed boundary layer, strengthen heat and mass transfer, and enhanced production water flux. However, the heat and electricity consumption also increased, which resulted in a decrease in GOR. Both water production and GOR experience a weak exponential reduction with an increase in salinity. Higher salinity reduced the saturated vapor pressure of the water solution and enhanced feed viscosity, which led to the transmembrane driving force decreasing, and the concentration polarization effect strengthening.

The MEMD/MSF hybrid processes was developed and assessed in this study. And the results can be a useful reference for the development and application of MD integrated technologies. There are still some key researches to be concerned in future such as membrane fouling and cleaning, energy consumption analysis and economic evaluation. Acknowledgements The authors are grateful for the support of the fundamental research funds for the Central Public Welfare Scientific Research Institution No. K-JBYWF-2013-T1. This paper is also supported by Projects in the National Science & Technology Pillar Program during the Twelfth Fiveyear Plan Period No. 2014BAB04B02. We would like to acknowledge Dalian Institute of Chemical Physics, Chinese Academy of Sciences and Zhejiang Dongda Environment Engineering CO., LTD. for the supplied helps. References [1] M.E. Findley, Vaporization through porous membranes, Ind. Eng. Chem. Process. Des. Dev. 6 (2) (1967) 66–68. [2] P.K. Weyl, Recovery of demineralized water from saline waters, US, US 3340186A. 1967. [3] K.W. Lawson, D.R. Lloyd, Review: membrane distillation, J. Membr. Sci. 124 (1) (1997) 1–25. [4] Y.F. Yang, D. Rana, T. Matsuura, et al., The heat and mass transfer of vacuum membrane distillation: effect of active layer morphology with and without support material, Sep. Purif. Technol. 164 (2016) 56–62. [5] S. Adham, A. Hussain, J.M. Matar, R. Dores, A. Janson, Application of membrane distillation for desalting brines from thermal desalination plants, Desalination 314 (2013) 101–108.

Please cite this article as: Y. Xing, et al., Performance study of a pilot-scale multi-effect vacuum membrane distillation desalination plant, Desalination (2016), http://dx.doi.org/10.1016/j.desal.2016.07.008

Y. Xing et al. / Desalination xxx (2016) xxx–xxx [6] I.C. Escobar, B.V.D. Bruggen, Modern Applications in Membrane Science and Technology, American Chemical Society, Distributed in print by Oxford University Press, 2011. [7] G.S. Wang, H. Ke, S. Gray, et al., Solar energy assisted direct contact membrane distillation (DCMD) process for seawater desalination, Sep. Purif. Technol. 143 (2015) 94–104. [8] N. Dow, S. Gray, J.D. Li, et al., Pilot trial of membrane distillation driven by low grade waste heat: membrane fouling and energy assessment, Das endokrine System und die Ƶwischenzellen, Springer, Vienna, 2016. [9] C.A. Quist-Jensen, A. Ali, S. Mondal, et al., A study of membrane distillation and crystallization for lithium recovery from high concentrated aqueous solutions, J. Membr. Sci. 505 (2016) 167–173. [10] Y. Peng, J. Ge, Z. Li, et al., Effects of anti-scaling and cleaning chemicals on membrane scale in direct contact membrane distillation process for RO brine concentrate, Sep. Purif. Technol. 154 (6) (2015) 22–26. [11] A. Alkhudhiri, N. Darwish, N. Hilal, Treatment of high salinity solutions: application of air gap membrane distillation, Desalination 287 (8) (2012) 55–60. [12] S. Gunko, S. Verbych, N. Hilal, et al., Concentration of apple juice using direct contact membrane distillation, Desalination 190 (1) (2006) 117–124. [13] A. Abdullah, N. Darwish, N. Hilal, Membrane distillation: a comprehensive review, Desalination 287 (8) (2012) 2–18. [14] M.C. Carnevale, E. Gnisci, J. Hilal, et al., Direct contact and vacuum membrane distillation application for the olive mill wastewater treatment, Sep. Purif. Technol. 169 (2016) 121–127. [15] C.N. Nguyen, T.N. Hau, S.S. Chen, et al., A novel osmosis membrane bioreactor-membrane distillation hybrid system for wastewater treatment and reuse, Bioresour. Technol. 209 (2016) 8–15. [16] L.M. Camacho, L. Dumée, J. Zhang, et al., Advances in membrane distillation for water desalination and purification applications, Water 5.1 (2013) 94–196. [17] X. Lv, C.R. Wu, Q.J. Gao, et al., Discuss about membrane distillation for application, Technol. Water Treat. 41 (2015) 26–30. [18] Q. Liu, C. Liu, L. Zhao, et al., Integrated forward osmosis-membrane distillation process for human urine treatment, Water Res. 91 (1) (2015) 31–42. [19] Y.D. Kim, K. Thu, J. Saththasivam, et al., Performance evaluation of multi-stage vacuum membrane distillation: the effect of seawater-coolant feed arrangement, Proceedings of the 7th Asian Conference on Refrigeration and Air Conditioning, 2014. [20] Y.D. Kim, K. Thu, S.H. Choi, Solar-assisted multi-stage vacuum membrane distillation system with heat recovery unit, Desalination 367 (2015) 161–171. [21] E.R. Abu-Zeid, Y. Zhang, H. Dong, et al., A comprehensive review of vacuum membrane distillation technique, Desalination 356 (2015) 1–14. [22] J. Zhang, N. Dow, M. Duke, et al., Identification of material and physical features of membrane distillation membranes for high performance desalination, J. Membr. Sci. 349 (1) (2010) 295–303.

9

[23] Y. Zhang, Y. Peng, S. Ji, et al., Review of thermal efficiency and heat recycling in membrane distillation processes, Desalination 367 (2015) 223–239. [24] X. Liu, B. Li, Heat recovery in vacuum membrane distillation, Chem. Ind. Eng. 2 (2014) 38–42. [25] L. Martı́Nez-Dı́Ez, M.I. Vázquez-González, Temperature and concentration polarization in membrane distillation of aqueous salt solutions, J. Membr. Sci. 156.2 (1999) 265–273. [26] A. Criscuoli, M.C. Carnevale, E. Drioli, Evaluation of energy requirements in membrane distillation, Chem. Eng. Process. Process Intensif. 47 (7) (2008) 1098–1105. [27] H. Zhu, H. Wang, Y. Guo, et al., Preparation and properties of PTFE hollow fiber membranes for desalination through vacuum membrane distillation, J. Membr. Sci. 446 (1) (2013) 145–153. [28] M. Khayet, T. Matsuura, Membrane Distillation: Principles and Application, Elsevier, 2011. [29] M.C.D. Andrés, J. Doria, M. Khayet, et al., Coupling of a membrane distillation module to a multieffect distiller for pure water production, Desalination 115 (115) (1998) 71–81. [30] A.K. Manna, P. Pal, Solar-driven flash vaporization membrane distillation for arsenic removal from groundwater: experimental investigation and analysis of performance parameters, Chem. Eng. Process. Process Intensif. 99 (2015) 51–57. [31] A.K. Fard, T. Rhadfi, M. Khraisheh, et al., Reducing flux decline and fouling of direct contact membrane distillation by utilizing thermal Brine from MSF desalination plant, Desalination 379 (2016) 172–181. [32] K. Zhao, W. Heinzl, M. Wenzel, et al., Experimental study of the memsys vacuum multi effect membrane distillation (V-MEMD) module, Desalination 323 (2013) 150–160. [33] H.W. Chung, J. Swaminathan, D.M. Warsinge, et al., Multistage vacuum membrane distillation systems for high salinity applications, J. Membr. Sci. 497 (2016) 128–141. [34] C. Wu, Z. Li, J. Zhang, et al., Study on the heat and mass transfer in air-bubbling enhanced vacuum membrane distillation, Desalination 373 (5) (2015) 16–26. [35] J. Orfi, N. Loussif, P.A. Davies, Heat and mass transfer in membrane distillation used for desalination with slip flow, Desalination 381 (2016) 135–142. [36] D. Zhao, J. Xue, S. Li, et al., Theoretical analyses of thermal and economical aspects of multi-effect distillation desalination dealing with high-salinity wastewater, Desalination 273 (2–3) (2011) 292–298. [37] F. Banat, F.A. Al-Rub, K. Bani-Melhem, Desalination by vacuum membrane distillation: sensitivity analysis, Sep. Purif. Technol. 33 (1) (2003) 75–87. [38] D. Singh, K.K. Sirkar, Desalination of brine and produced water by direct contact membrane distillation at high temperatures and pressures, Fuel Energy Abstr. 389 (2012) 380–388. [39] L. Martı'nez, J.M. Rodrı'guez-Maroto, On transport resistances in direct contact membrane distillation, J. Membr. Sci. 295.1 (2007) 28–39.

Please cite this article as: Y. Xing, et al., Performance study of a pilot-scale multi-effect vacuum membrane distillation desalination plant, Desalination (2016), http://dx.doi.org/10.1016/j.desal.2016.07.008