Techno-economic analysis of a hybrid electrodialysis–batch reverse osmosis process for brackish water desalination

Symbols for the ED process

C^fd

concentration at the inlet of dilute stream

C^d

concentration at the exit of dilute stream

R

recovery

Λ

equivalent conductance of solution at 20 °C

Δ

thickness of cell-pair

u

linear flow velocity

a

constant for limiting current density calculation

b

constant for limiting current density calculation

F

Faraday constant

α

volume factor

β

area factor accounting for the shadow effect

s

safety factor

ζ

current utilization/efficiency

ρ_A+ρ_C

total area resistance of membrane

Q^d

product water flow rate (production capacity)

w

width of cell-pair

L_st

length of flow path per stack

h

half thickness of the grid rods of spacer (h = 1/4 Δ)

l

distance between two grid rods or mesh size

μ

solution viscosity, kg/m-s

d^H

hydraulic diameter, m

ɳ_p

efficiency of pump

C^Δ

concentration difference of dilute/concentration stream

C^fc

concentration at the inlet of concentrate stream

C^c

concentration at the exit of concentrate stream

ΔP

pressure drop in stack, kPa

i_prac

practical limiting current density

A_prac

area required for stack

L_prac

total length required

N_cp

no. of cell-pairs in stage

I_st

current passing through stack

U_st

potential (voltage) drop across stack

E_des

specific energy consumption for desalination

E_pump

specific energy consumption for pump

E_total

total specific energy consumption for ED unit

Symbols for the BRO process

A_bro

membrane area

Q_perm

permeate flow rate

J_w

water permeability of the membrane

Q_bro

feed flow rate

SEC_ideal

ideal specific energy consumption

r

recovery ratio of process

SEC_total

specific energy consumption of real process

ɳ_{2nd law}

second law efficiency

π

osmotic pressure of feed solution

C_hpp

capital cost of high-pressure pump

P_hpp

maximum pressure of high-pressure pump

C_recir.

capital cost of recirculation pump

P_recir.

pressure of recirculation pump

Q_recir.

flow rate of recirculation pump

C_memb

capital cost of membrane

C_m

unit membrane cost

C_PrExch

capital cost of pressure exchanger with piston

C_housing

capital cost of membrane housing

C_piping

cost of piping

R_s

rejection of salt

ɳ_pump

pump efficiency

Hybridizing processes to achieve a sustainable goal is the interest of many researchers in current times. Brackish water desalination has become the most common method to supply freshwater for inland applications. Membrane technologies have been used commonly for groundwater desalination which has less energy cost and gives high-quality water for consumption (Kim & No 2012; Liu et al. 2021). Brackish water reverse osmosis (BWRO) and electrodialysis (ED) are commercially available membrane processes to treat brackish water. BWRO is suitable for salinity ranging from 1 to 10 g L⁻¹ while ED is more suitable for <3 g L⁻¹ salinity (Ali et al. 2018).

For brackish water desalination, increasing the recovery rate is an important factor due to the problem associated with high-volume brine disposal, especially in inland brackish water desalination applications. For the standalone BWRO process, it is difficult to achieve high recovery for the reason of fouling and high-pressure requirements. There are several methods proposed for brine disposal such as leaving it in surface water or a sewer system, deep well injection, evaporation ponds, and land applications (Squire 2000; Al-Faifi et al. 2010; Elsaid et al. 2012; Panagopoulos et al. 2019; Al-Anzi et al. 2021; Khan & Al-Ghouti 2021; Sahu 2021). But all these methods are associated with one or other disadvantages as not being perfect and environmental hazards such as an increase in salinity of the water body, affecting the nearby biological activities, leakage of brine in nearby aquifers, and contamination of land (Ahmed et al. 2000; Prusty & Farooq 2020).

Much research has been done to reduce the volume of concentrated brine for inland application, for example, the use of secondary treatment to recover minerals from brine (Henthorne & Boysen 2015), applying a thermal process to achieve zero liquid discharge, and chemical softening with additional membrane treatment (Bond & Veerapaneni 2008). All these alternatives require large chemicals and consumables which increase the cost of treatment. Altaee & Hilal (2015) suggested a trihybrid membrane system to reduce this cost and disadvantages of conventional methods, but still, it requires an antiscalant as a pretreatment of feed before entering the system, to reduce the risk of scaling of concentrated divalent ions. Efforts are made to increase the recovery of conventional RO using solar thermal energy and mechanism (Mudgal & Davies 2016; Raninga et al. 2021) and hybridizing with the forward osmosis process (Patel et al. 2021).

In this study, the hybrid ED–batch reverse osmosis (ED–BRO) system is used to take benefit of ED for high recovery and RO for highly pure water at low cost. The water cost in a hybrid system is compared with the standalone ED and BRO processes with the volume recovery rate. The product water from all these three setups has been maintained at the same volume flow rate and concentration to make a justifiable comparison. Many studies of hybrid ED and conventional RO processes are in literature, but to take cost advantage of the advanced RO process, here, the hybrid ED–BRO system is studied.

ED process design equations contain many proportionality constants and assumptions to calculate the process parameters. In this study, the values for these constants are taken as shown in Table 1. From our previous study (Ankoliya et al. 2021), the water recovery ratio achieved for the ED process is 70% for local water quality composition. Also, the cell-pair thickness and flow velocity values taken here are optimized for similar conditions. Other constants are related to Faraday law, NaCl solution, current density, membrane property, and pump.

The major component is membrane cost in the ED stack. All other costs of peripheral equipment are proportional to the cost of the membrane cost. The membrane cost is related to its area. Here, the cost of the membrane per unit area is taken from the literature (Zhang et al. 2017) and the total stack cost can be calculated by multiplying proportionality constants. All these assumptions and constant values are shown in Table 2 along with the life of the membrane, operational duration and electricity cost. To obtain all these costs, we require a membrane area which is calculated by using a mathematical model of the ED process explained in the previous section.

A BRO process works on batch operation – which means it generates permeate in batch and not continuously, and also, the pressure varies from minimum to peak. So, the BRO process has a lower average pressure than the conventional RO process and results in a lower SEC. The peak pressure in BRO is the same as in continuous-RO, but is only reached momentarily at the end of each cycle. For the rest of the time, the pressure is lower (Park et al. 2020). So, the purpose of using the BRO process is to reduce energy consumption in brackish water desalination.

The cycle of the BRO system consists of two stages called pressurization and purge-and-refill. During the pressurization stage, feedwater is supplied to the left end of the pressure exchanger, pushing the piston to the right. The piston forces out water from the other side mixed with the recirculation stream and goes to the feed side of the RO module through the main valve which only opens during the pressurization stage. The reject stream from the RO module is recirculated by a recirculation pump. This stage produces permeate and works until the piston in the pressure exchanger reaches the right side end. During this cycle, the pressure gradually increases to the peak value and also the concentration of water in the recirculation loop gradually increases. In the second stage, saltwater in the system becomes very concentrated which needs to be taken out and replaced by feed water, and to do that, the purge valve is opened. To divert the fresh feed filled in the left side of the pressure exchanger, the bypass valve is also opened, and at the same time, the main valve is closed. The continued operation of the recirculation pump pushes the piston to the left until it reaches its original position.

Here, the BRO process is used for 1,100 ppm salinity feedwater. The design inputs and other constant values are shown in Table 3.

The feed water used in this study is tap water from the university, which is used for general purposes other than drinking. This water is extracted from the ground; so it represents the groundwater of the university in the city of Gandhinagar in the Gujarat state of India. The details of feed water characterization are shown in Table 4.

To generate 500 LPH of clean water, it requires 714.28 LPH of feedwater flow rate. Based on this feed and permeate data of the ED stack, the membrane area, current, voltage and energy consumption can be calculated by using mathematical model equations as shown in Table 5. The cost of the whole unit with operating cost is also calculated and shown in Table 5.

In this standalone ED process, the clean water production cost was found to be 0.6 $ per cubic meter of treated water. Here, the major contributor to cost is the capital cost of the ED system because of the high price of the ion exchange membrane. To reduce the water cost, it is required to reduce the membrane area requirement in the ED stack.

To generate 500 LPH of pure water through the BRO process, it requires 625 LPH of feed flow rate having a recovery rate of 80%. This standalone BRO process having a solute rejection of over 93% can produce clean water of 73.2 ppm, which is less than the ED process. Detailed cost calculation results are shown in Table 6.

Here, the cost of clean water production is found to be 0.1 $ per cubic meter of pure water generated, which is less than the standalone ED system. Here, the capital cost contribution is also more than the energy cost, similar to the ED system.

To calculate the overall cost of clean water in the hybrid ED–BRO system, it is required to calculate the first individual process cost. The summation of both individual costs divided by the total clean water production gives the final water cost of the hybrid ED–BRO system. To calculate the individual process cost, the same process is followed as in the standalone process and it is shown in Table 7 along with the overall hybrid process cost.

The cost of the ED process in the hybrid system was found to be much less compared with the cost in the standalone system. It is because the running ED system in higher salinity in the hybrid system reduces the membrane area requirement from 8.63 to 1.72 m² which reduces the capital cost of the ED process. The SEC of ED in the hybrid system increases when compared to the standalone ED due to higher salt removal, but the energy cost is less because of running at a smaller capacity in the hybrid system.

For the feed concentration from 1,000 to 2,000 mg L⁻¹, the freshwater cost and SEC of the hybrid ED–BRO system vary from 0.20 to 0.34 $ m⁻³ and from 0.81 to 1.60 kWh m⁻³, respectively. The freshwater cost in the hybrid ED-RO proposed system in the literature using conventional RO is slightly on the higher side as shown in Table 9. In the study by Generous et al. (2021), obtaining a product water quality of 200 mg L⁻¹ and keeping a feed concentration of 5,000 mg L⁻¹, the freshwater cost was 0.37 $ m⁻³ which is comparable to our study, but at a very large scale of 23,333 m³/day production capacity. On the other hand, in a small-scale study by Thampy et al. (2011), the SEC obtained was 8–10 kWh m⁻³, which is high and shows that a high system recovery at small-scale results in high freshwater cost. The obtained result in our study at 12 m³/day of small production capacity has a freshwater cost of 0.2–0.34 $ m⁻³, which shows the cost–benefit at a smaller scale.

This study of the hybrid ED–BRO process for brackish water desalination includes cost and sensitivity analyses to understand the benefit of the hybrid process. The hybrid ED–BRO process can produce clean water at a lower cost with high recovery even at small-scale applications in brackish water desalination.

The hybrid system can achieve 94% overall water recovery and has 0.22 $ m⁻³ cost when using feed of 1,100 ppm concentration. The hybrid system extracts 34 and 18% more clean water from feed water than standalone ED and BRO processes, respectively. The membrane area of ED and feed pressure requirement of the BRO process are the main affecting parameters when feed water salinity changes in the range of 1,000–2,000 ppm. Process hybridization reduces the capital cost of the ED process due to a smaller membrane area despite having a higher SEC. Very high recovery and comparatively less cost of the proposed system make it suitable for inland applications, where brine disposal and installation of a big desalination plant is challenging.

The authors acknowledge the project grant received from the Department of Biotechnology (Government of India, Project No. BT/IN/EU-WR/40/AM/2018) and the European Union’s Horizon 2020 research and innovation program under grant agreement No. 820906. The authors are thankful to Pandit Deendayal Energy University for providing the necessary infrastructure and other facilities for the research work.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.