PRES22_0231.docx DOI: 10.3303/CET2294120 Paper Received: 17 April 2022; Revised: 13 May 2022; Accepted: 23 May 2022 Please cite this article as: Voutetaki A., Papadopoulos A.I., Plakas K., Seferlis P., 2022, Integration of an Electrodialysis Process for Selective Nitrate Removal with Renewable Energy Sources, Chemical Engineering Transactions, 94, 721-726 DOI:10.3303/CET2294120 CHEMICAL ENGINEERING TRANSACTIONS VOL. 94, 2022 A publication of The Italian Association of Chemical Engineering Online at www.cetjournal.it Guest Editors: Petar S. Varbanov, Yee Van Fan, Jiří J. Klemeš, Sandro Nižetić Copyright © 2022, AIDIC Servizi S.r.l. ISBN 978-88-95608-93-8; ISSN 2283-9216 Integration of an Electrodialysis Process for Selective Nitrate Removal with Renewable Energy Sources Alexia Voutetakia,b, Athanasios I. Papadopoulosa*, Konstantinos Plakasa, Panos Seferlisb a Chemical Process and Energy Resources Institute (CPERI), Centre for Research and Technology Hellas (CERTH), 57001 Thermi, Thessaloniki, Greece b Department of Mechanical Engineering, Aristotle University of Thessaloniki, 54124 Thessaloniki, Greece spapadopoulos@certh.gr Intensified agricultural activities with an excessive fertiliser utilisation result in high penetration rates for several chemical compounds through soil to the ground water reservoirs. One of the main components in fertilisers are nitrate ions, which at high concentrations have harmful effects on human health. There is a number of available technologies for nitrate removal, but the majority of them use chemical additives and generate waste that has to be treated before their release to the environment. One promising alternative is the electrodialysis (ED) process. To make the process more sustainable and allow the deployment of such systems to remote areas where limits on electric power consumption may be present, the ED system can be powered by photovoltaic (PV) solar panels. The challenge is to operate the system efficiently under the intermittence of the solar radiation and the variations in the water concentration of nitrates. In this work, a mathematical model describing the removal of nitrate ions, under a variety of meteorological conditions, and for a variable daily demand of drinking water has been developed. Solar radiation and nitrate feed are real data from an area with drinking water contamination problem. This simulation model predicts the removal rate of nitrate ions, the sizing of the solar panels and the capacity of the battery with respect to the battery’s state of charge. Efficient energy management strategies are developed to ensure that the targeted pure water demand in terms of volume and concentration of nitrates is met and the use of non-renewable, externally provided electricity is minimised. 1. Introduction As freshwater resources are drastically reduced by increasing pressures from climate change, population growth and rapid urbanisation, the use of unconventional water resources, is one of the most important solutions to confront water scarcity and increase the availability of water resources, especially in arid regions of the world. This need has resulted in the rapid development of desalination technologies, most notably the membrane filtration technologies. Among them, electrodialysis (ED) technology has shown particular precedence in recent years due to its competitive advantages, such as high clean water recovery, short treatment times, flexibility, and lower operating and maintenance costs (Strathmann, 2004). Another important advantage is the direct use of electricity to achieve ion separation, which favours its combination with renewable energy sources (RES), such as solar energy (Plakas et al., 2018). The hybrid ED − PV system could be applicable to areas with excessive solar energy and electric power limitations. The challenge of implementing solar energy with ED is the intermittent production of electricity, in combination with the brine management. Another important challenge is the treatment of nitrate-contaminated (NO3-) brackish water in areas with agricultural activity. Groundwater in those areas is considered unsuitable for drinking due to high NO3- (> 50 mg/L) concentrations. ED research over the last 10 y focused on the recovery of high value-added materials., such as lithium with desalination (Ji et al., 2017) as well as the removal of heavy metals (Gherasim et al., 2014). The use of ion exchange membranes for the recovery of nutrients has also found successful applications, such as municipal waste treatment (Vineyard et al., 2021), and the recovery of useful compounds from manure (Shi et al. 2021). However, there are only a few works regarding the selective removal of NO3- ions with ED, most of them treating 721 wastewater instead of fresh water and lacking model-based investigations (Ye et al., 2019). Regarding the combined application of ED with renewable sources, all the publications reported systems where the RES (wind or solar energy) was directly connected to the ED unit, as indicated in the review of Mir et al. (2021). For instance, Cipolletta et al. (2021) studied a hybrid RES (wind and PV energy- ED) system for irrigation of banana cultivation. The lack of energy storage has very significant disadvantages as non-renewable sources should be used during power outages. There are also a few studies where energy storage is used, but only in a recent one has the overall system been optimized using systematic methods that take into account energy management (Bian et al., 2019). The previous work however is not implemented on nitrate removal. In the following presented work, an integrated system of ED and PV-battery was theoretically studied. For the application of the system to a case study for nitrate ions removal from a ground source, weather, feed and water demand data were collected, for the application of the ED − PV system. The electrodialysis unit size, the PV and the battery numbers were estimated with a dynamic model, so that the process could be controlled to meet energy and water demands. 2. Materials and methods 2.1 Electrodialysis energy consumption Electrodialysis is an electrochemical process in which ions migrate through ion-selective membranes as a result of their attraction due to two electrically charged electrodes. To calculate the energy demand of electrodialysis (EED, kW) the applied voltage (VED), current density (IED), and process time (t) are needed, as shown in Eq(1). The applied voltage and the process time were input variables to calculate the current density output parameters of the ED model. 𝐸𝐸𝐸𝐸𝐸𝐸 = 𝐼𝐼𝐸𝐸𝐸𝐸 ∙ 𝑉𝑉𝐸𝐸𝐸𝐸 ∙ t (1) 2.2 Photovoltaic and battery model A PV cell has the ability to convert photon energy into electrical energy in the form of direct current. Modeling for PV panels takes into account its electrical characteristics. The relationship between the voltage I and the voltage V of the equivalent circuit is given by calculating the currents IL, ID and Ish, as shown below. Variables IL, I0, Rs,, α, are obtained by algebraic equations and are discussed in literature (Ulleberg, 1998). This equation demands from PV manufacture data the values of current and voltage at maximum power, the current at short circuit current conditions, the voltage at open voltage conditions, the values of temperature coefficient at short circuit and open voltage conditions, and the number of solar cells. 𝐼𝐼𝑃𝑃𝑃𝑃 = 𝐼𝐼𝐿𝐿 − 𝐼𝐼𝐸𝐸– 𝐼𝐼𝑠𝑠ℎ = 𝐼𝐼𝐿𝐿 − 𝐼𝐼0 ∙ (𝑒𝑒 V+𝐼𝐼∙𝑅𝑅𝑠𝑠 𝛼𝛼 − 1) − V + 𝑅𝑅𝑠𝑠 𝑅𝑅𝑠𝑠ℎ (2) The output power from the 𝑃𝑃𝑉𝑉 array is given by: 𝑃𝑃𝑃𝑃𝑃𝑃 = 𝐼𝐼𝑃𝑃𝑃𝑃 ∙ 𝑛𝑛𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 ∙ 𝑉𝑉𝑃𝑃𝑃𝑃 (3) where nconv is the efficiency of a DC/DC converter (90-95 %). A lead-acid accumulator is an electrochemical device that converts electrical energy into chemical energy during the charging process and vice versa during the discharging process. The mathematical model calculates the charging and discharging current, voltage and the state of charge (SOC) of the accumulator. The SOC is the fraction of the current capacity of the accumulator at each time instant, divided by its nominal capacity (Manwell et al., 1993): 𝑆𝑆𝑆𝑆𝑆𝑆𝑡𝑡 = 𝑆𝑆𝑆𝑆𝑆𝑆𝑡𝑡−1 ∙ (1 − 𝜎𝜎) + 𝐼𝐼𝑏𝑏𝑏𝑏𝑡𝑡 ∙ 𝑛𝑛𝑏𝑏𝑏𝑏𝑡𝑡 ∙ 𝛥𝛥𝛥𝛥 (4) where 𝜎𝜎 is the self-discharge rate (2.5 %), 𝑛𝑛𝑏𝑏𝑏𝑏𝑡𝑡 is the efficiency factor (95 %) and 𝛥𝛥 is the time in h. 2.3 Power management strategy A block diagram of the presented system is shown in Figure 1. The PV panels collect the solar irradiance and convert it to electrical energy, which is transferred to the battery accumulator, and then to the ED unit, where the current is used to process ground water. Depending on the battery’s SOC the energy is managed differently. When the solar irradiance is sufficient, the current that is transferred to the battery charges the accumulator, and the battery’s SOC is increased. On the other hand, when the irradiance is not enough, as the battery only provides load, it discharges, and its SOC decreases. In the 1st case of Figure 1, the battery’s SOC ranges between minimum and maximum values, (SOCmin < SOC < SOCmax) and the current load is transferred to the ED unit, where it is used to process groundwater. In the 2nd case, the battery’s SOC is less than the minimum value (SOC 722 < SOCmin), and the current can’t meet the ED unit demand and the ED unit is stopped. Τhe battery is charged by the PV panels, until it’s SOC exceeds SOCmin and the 1st case is implemented. In the 3rd case, when the solar irradiance is too high, and the battery is fully charged, there is excessive current produced that can be either sold to the grid, or serve other loads, such as hydrogen production (Ipsakis et al., 2009), although the latter is not investigated here. In this case, the system is oversized, and such energy surplus should be minimized. Figure 1: Block diagram of the proposed system 2.4 Proposed approach In this work the goal is to investigate a scaled-up system for a real case in Thessaloniki, Greece. Firstly the ED size (membrane area Am, number of cells Ncell) and operating characteristics (separation efficiency SP and voltage V) were evaluated based on the nitrate feed concentration, nitrate target concentration, and drinking water demand for about 20 houses of the nearby region (10 m3/d). In the second step, the calculated output current density from the designed ED was used to determine the PV (NPV) and battery (NBAT) numbers, in order to meet the ED load demand. The different options for PV and battery numbers, that would cover the energy demand without connection to the electrical grid, were evaluated to attain the maximum daily water production (WProd), and the minimum renewable energy loss (Eloss) of the system, with respect to the battery SOC (%), as follows: Both problems were addressed through exhaustive numeration. The dynamic ED model, used as described in previous work (Voutetaki et al., 2020), was validated with experimental data from literature for the nitrate ions (Mohammadi et al., 2021). The current load demand for ED, and the solar radiation data of the case study’s region, were used to design the PV-battery system. A PV-battery model was developed (Ipsakis et al., 2009), and the power management strategies were obtained from Giaouris et al. (2013). Electrodialysis is usually a batch process. Under current application a specific water volume is recirculated for enough time to meet the drinking water specifications. Then, the ED unit stops (load demand=0), and the ED batch is replaced with another batch. The challenge of this work was to design enough batches, for the battery’s SOC not to decrease more than SOCmin and so that the drinking water produced from the batches would be maximized. The diluted water stream produced from ED is for consumption, and the concentrated water stream, is for irrigation. 3. Implementation The study was implemented in a region near Thessaloniki, with nitrate contamination of ground water. The initial nitrate concentration was 100 mg/L, and the target was for the water to meet drinking level, which means that nitrate has to be reduced to < 50 mg/L. The ED size and operating parameter estimation was for V= [20- 40 V], Ncell= [45- 65], Am= [0.03- 0.06 m2]. The PV and battery estimation were made with solar radiation data of the 𝑚𝑚𝑚𝑚𝑛𝑛𝑉𝑉, 𝑁𝑁𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐, 𝐴𝐴𝑚𝑚 𝑚𝑚𝑚𝑚𝑚𝑚 𝑆𝑆𝑃𝑃 𝑠𝑠. 𝛥𝛥. 𝑉𝑉𝑚𝑚𝑚𝑚𝑐𝑐 < 𝑉𝑉 < 𝑉𝑉𝑚𝑚𝑏𝑏𝑚𝑚, 𝑁𝑁𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 𝑚𝑚𝑚𝑚𝑐𝑐 < 𝑁𝑁𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 < 𝑁𝑁𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 𝑚𝑚𝑏𝑏𝑚𝑚, 𝐴𝐴𝑚𝑚𝑚𝑚𝑚𝑚𝑐𝑐 < 𝐴𝐴𝑚𝑚 < 𝐴𝐴𝑚𝑚𝑚𝑚𝑏𝑏𝑚𝑚 (5) 𝑚𝑚𝑚𝑚𝑛𝑛𝑁𝑁𝑃𝑃𝑃𝑃, 𝑁𝑁𝐵𝐵𝐵𝐵𝐵𝐵, 𝐸𝐸𝑐𝑐𝑐𝑐𝑠𝑠𝑠𝑠 𝑚𝑚𝑚𝑚𝑚𝑚 𝑊𝑊𝑃𝑃𝑃𝑃𝑐𝑐𝑃𝑃 𝑠𝑠. 𝛥𝛥. 𝑆𝑆𝑆𝑆𝑆𝑆𝑚𝑚𝑚𝑚𝑐𝑐 < 𝑆𝑆𝑆𝑆𝑆𝑆 < 𝑆𝑆𝑆𝑆𝑆𝑆𝑚𝑚𝑏𝑏𝑚𝑚, 𝑁𝑁𝑃𝑃𝑃𝑃𝑚𝑚𝑚𝑚𝑐𝑐 < 𝑁𝑁𝑃𝑃𝑃𝑃 < 𝑁𝑁𝑃𝑃𝑃𝑃𝑚𝑚𝑏𝑏𝑚𝑚, 𝑁𝑁𝐵𝐵𝐵𝐵𝐵𝐵 𝑚𝑚𝑚𝑚𝑐𝑐 < 𝑁𝑁𝐵𝐵𝐵𝐵𝐵𝐵 < 𝑁𝑁𝐵𝐵𝐵𝐵𝐵𝐵 𝑚𝑚𝑏𝑏𝑚𝑚 (6) 723 case study’s region that were collected for 1 y. The number of batteries and PV panels were ranged within NBAT = [2, 6] and NPV= [9, 20], so that the load demand of the ED was satisfied. The battery capacity and PV panel model values were 866 Ah and 0.64 m2, with 9.85 % efficiency for the PV and 95 % for the battery, making the overall system efficiency 9.3 %. The daily water production was estimated using a water tank with initial volume of 5 m3, that fills with 2 m3/batch, and empties at a constant hourly rate. The energy loss (surplus) is calculated from the multiplication of the time duration at which the SOC is above SOCmax, to the respective load that exceeds the battery’s capacity. The SOC range is SOC =[50, 90 %] of the total battery capacity. 4. Results and discussion 4.1 Electrodialysis unit validation and size estimation The ED unit validation for nitrates was made with experimental data from Mohammadi et al. (2021), and is shown in Figure 2a. Clearly, the developed model matches the experimental data sufficiently, as the deviation was estimated to range between 1- 10 %, with an average of 5 %. For the determination of the ED unit characteristics, as per the first step of the proposed approach, the unit processed 4 m3/batch, so the concentrate and dilute volumes were 2 m3 each. Several simulations were made for different ED unit parameters, and the optimum solution was selected for minimum energy consumption and maximum nitrate ions removal. The ED unit was estimated to have 68 cell pairs, and membrane sizes of 0.038 m2. For treating 4 m3 of ground water with initial concentration of 100 mg/L, the final concentration for nitrate was [NO3]- =40 mg/L, that was achieved with 25 volts, 150 L/h of flow rate, and 1 h of treatment. The nitrate removal with ED is shown in Figure 2b. The ED load demand was calculated at 130 kW for one batch based on Eq(1). (a) (b) Figure 2: a) model validation for nitrate b) nitrate removal results with 𝐸𝐸𝐸𝐸 4.2 Photovoltaics and battery number estimation The estimation was made for 6 consecutive cloudy days, (1,680-1,829 h of Figure 3a). Ιn Figure 3b, the PV energy for the 6 cloudy days is presented. Ιn Table 1, the minimum number of PV panels, produced water flowrate, and energy loss for the different battery system values are presented. a) b) Figure 3: a) Annual solar irradiance of Thessaloniki, b) solar irradiance of 6 cloudy days 724 Table 1: Different battery-PV systems evaluation 𝑁𝑁𝐵𝐵𝐵𝐵𝐵𝐵 𝑁𝑁𝑃𝑃𝑃𝑃 𝑊𝑊𝑃𝑃𝑃𝑃𝑐𝑐𝑃𝑃, m3/d 𝐸𝐸𝑐𝑐𝑐𝑐𝑠𝑠𝑠𝑠, MWh/y 2 20 10.8 1,920 3 15 9.6 854.4 4 12 8.88 367.2 5 10 11.52 384 6 9 10.56 2.1 The 6-battery/10-PV panel system has high water production (10.56 m3/d) and the minimum energy loss (2.1 MWh/y). The battery 𝑆𝑆𝑆𝑆𝑆𝑆 % and hourly water production are shown in Figures 4a and 4b. In this 6- battery system every hour 0.44 m3 is drawn from the water tank. On the other hand, the 5-battery system, whose SOC and water production are presented in Figure 4c and 4d, has the highest fresh water production (11.52 m3/d) and relatively low energy loss (384 MWh/y). In this 5-battery/10 PV system, a quantity of 0.48 m3 is drawn hourly from the tank. Both cases represent viable systems that meet the drinking water target. Note that the energy loss values are high, because the system capacity is small and the area has intense solar irradiation. Apparently, the energy loss-battery cost trade-off is an important feature in this area that is not considered in this work. a) b) c) d) Figure 4: a) 𝑆𝑆𝑆𝑆𝑆𝑆 and b) water production of a 6 battery/10 𝑃𝑃𝑉𝑉 panel system, c) 𝑆𝑆𝑆𝑆𝑆𝑆 and d) water production of a 5 battery/10 𝑃𝑃𝑉𝑉 panel system 5. Conclusions In this study, a hybrid ED − PV system was theoretically studied to treat nitrate-contaminated water. The sizing of the ED unit and PV/battery systems were made with dynamic models properly adjusted to the case study’s water characteristics, and water demand data using annual information of solar irradiation. The results showed that an ED unit of 68 cell pairs, 25 V, demands 130 kW/batch for treating 4 m3 of water with 100 mg/L of nitrate. To produce enough energy, different systems of PV/Batteries were used, that were compared to each other on higher water production and lower energy surplus. The more suitable options were a 5 battery/10 PV system and a 6 battery/9 PV system. The first had the highest water production and the second the lowest energy surplus. For the selection of the most advantageous system, more factors have to be studied, such as equipment cost factors, and surplus energy management. The presented ED − PV system is regarded to be a sustainable, environment friendly solution for treating nitrate contaminated waters and freshwater production, for areas with limited access to electrical grid. 725 Nomenclature 𝑆𝑆𝑆𝑆𝑆𝑆 – state of charge of accumulator, % 𝐼𝐼𝑃𝑃𝑃𝑃 – operation current for the 𝑃𝑃𝑉𝑉 system, A 𝐼𝐼0 – diode reverse saturation current for the 𝑃𝑃𝑉𝑉 system, A 𝐼𝐼𝑠𝑠ℎ – shunt current for the 𝑃𝑃𝑉𝑉 system, A 𝐼𝐼𝐿𝐿– light current for the 𝑃𝑃𝑉𝑉 system, A 𝑅𝑅𝑠𝑠ℎ – shunt resistance for the 𝑃𝑃𝑉𝑉 system, Ω 𝑅𝑅𝑠𝑠 – series resistance for the 𝑃𝑃𝑉𝑉 system, Ω 𝑚𝑚 – curve fitting parameter for the 𝑃𝑃𝑉𝑉 system, V 𝑉𝑉– operation voltage for the 𝑃𝑃𝑉𝑉 system, V 𝐼𝐼𝐵𝐵𝐵𝐵𝑡𝑡– charging/ discharging current, A 𝐼𝐼𝐸𝐸 – diode current for the 𝑃𝑃𝑉𝑉 system, A 𝛥𝛥– time, h 𝐸𝐸𝐸𝐸𝐸𝐸 – ED energy consumption, kW 𝐼𝐼𝐸𝐸𝐸𝐸 – ED current density, A 𝑉𝑉𝐸𝐸𝐸𝐸 – ED voltage, V 𝑃𝑃𝑃𝑃𝑃𝑃– 𝑃𝑃𝑉𝑉 power, W 𝑉𝑉𝑃𝑃𝑃𝑃– operation voltage of the 𝑃𝑃𝑉𝑉 system, V 𝑁𝑁𝑃𝑃𝑃𝑃– 𝑃𝑃𝑉𝑉 panels number 𝑁𝑁𝐵𝐵𝐵𝐵𝐵𝐵– battery number 𝐸𝐸𝑆𝑆𝑆𝑆𝑃𝑃– energy surplus, W 𝑊𝑊𝑝𝑝𝑃𝑃𝑐𝑐𝑃𝑃– water produced, m3 Acknowledgments This research has been co‐financed by the European Regional Development Fund and the European Social Fund (project code: ΚΜΡ6-0083181). References Bian D.W., Watson S.M., Wright N.C., 2019, Optimization and design of a low-cost, village-scale, photovoltaic- powered, electrodialysis reversal desalination system for rural India, Desalination, 452, 265-278. Cipolletta M., Moreno V C., Dallavalle E., Zanuttigh B., Cozzani V., 2021, Aiding water production in isolated islands using integrated renewable energies sources, Chemical Engineering Transactions 86, 205-210. Gherasim C.V., Krivcik J., Mikulasek P., 2014, Investigation of batch electrodialysis process for removal of lead ions from aqueous solutions, Chemical Engineering Journal, 256, 324-334. Giaouris D., Papadopoulos A., Ziogou C., Ipsakis D., Voutetakis S., Papadopoulou S., Sefelis P., Stergiopoulos F., Elmasides C., 2013, Performance investigation of a hybrid renewable power generation and storage system using systemic power management models, Energy, 61, 621-635. Ipsakis D., Voutetakis S., Seferlis P., Stergiopoulos F., Elmasides C., 2009, Power management strategies for a stand-alone power system using renewable energy sources and hydrogen storage, International Journal of Hydrogen Energy 34, 7081-7095. Manwell J.F., McGowan J.G., 1993, Lead acid battery storage model for hybrid energy systems, Solar Energy, 50, 399–405. Mir N., Bicer Y., 2021, Integration of electrodialysis with renewable energy sources for sustainable freshwater production: A review, Journal of Environmental Management, 289, 112496. Mohammadi R., Ramasamy D. L., Sillanpaa M., 2021, Enhancement of nitrate removal and recovery from municipal wastewater through single- and multi-batch electrodialysis: Process optimisation and energy consumption, Desalination, 498, 114726. Plakas K.V., Sioutopoulos D.C., Karabelas A.J., 2018, Energy consumption and minimization, Chapter 14 In: A. Bazargan (Ed.), A multidisciplinary introduction to desalination, River Publishers, Gistrup, Denmark, ISBN 978-87-93379-54-1. Shi L., Hu Y., Wang Y., Bei E., Lens P. N., 2021, In situ electrochemical oxidation in electrodialysis for antibiotics removal during nutrient recovery from pig manure digestate, Chemical Engineering Journal, 413, 127485. Strathmann H., 2004, Chapter 1 Overview of ion-exchange membrane processes, Ιn: Elsevier, Membrane Science and Technology, vol. 9, Amsterdam, Netherlands, 1-20. Ulleberg U., 1998, Stand-alone power systems for the future: optimal design, operation & control of solar- hydrogen energy systems, PhD thesis. Norwegian University of Science and Technology, Trondheim, Norway. Vineyard D., Hicks A., Karthikeyan K. H., Davidson C., Barak P., 2021, Life cycle assessment of electrodialysis for side stream nitrogen recovery in municipal wastewater treatment, Cleaner Environmental Systems, 2, 100026. Voutetaki A., Gkouletsos D., Papadopoulos A., Seferlis P., Plakas K., Bollas D., Parcharidis S., 2020, Efficient modelling of electrodialysis process for waste water treatment through systematic parameter estimation, Chemical Engineering Transactions, 81, 841-846. Ye Z.-L., Ghyselbrecht K., Monballiu A., Pinoy L., Meesschaert B., 2019, Fractionating various nutrient ions for resource recovery from swine wastewater using simultaneous anionic and cationic selective-electrodialysis, Water Research, 160, 424-434. 726 PRES22_0245.pdf Integration of an Electrodialysis Process for Selective Nitrate Removal with Renewable Energy Sources