A thermally-driven seawater desalination system: Proof of concept and vision for future sustainability

system comprises only hollow chambers (devoid of membranes or tubes, minimal use of chemical and maintenance) where vapor generation (flashing) utilizes the enthalpy difference between the sprayed feed seawater and the saturated vapor enthalpy of the vessels. Concomitantly, vapor is condensed with spray droplets of cooler water (potable) in adjacent condenser vessels, employing a simple design concept. We present detailed design and real seawater experiments data of a DCSEC system for the first time. The water production cost is calculated as $0.52/m 3 , which is one of the lowest figures reported compared to commercial processes presented by Global Water Intelligence.


Introduction
Three-quarter of the earth's surface area is covered by seawater, which is impaired by dissolved salts, algae, etc., and less favorable for potable propose.However, freshwater is vital for human survival and it is a necessary resource for economic development of all countries, particularly in the agriculture, manufacturing, and industrial sectors.As reported recently [1][2][3], the increasing demand of freshwater globally is caused by the rapid population growth and economic development.
Owing to hot desert climate conditions in Gulf Cooperation Council (GCC) countries, freshwater availability from rivers and lakes are scare due to limited rainfall, less than 150 mm annually.However, regional variation of rainfall along the Red Sea and the Gulf of Oman may vary up to 500 mm, as shown in Table 1 [4,5].On other hand, the ground water reserves of GCC countries are estimated to be around 2300 billion cubic meters (bcm) and of which 300 bcm has already been extracted for human activities.If the current rate of extraction continues, it is projected that remaining ground water of the region will be depleted in less than 2 decades [6].
As mentioned in the introductory paragraph, the regional population growth is a major contributing factor for high potable water consumption.From 1950 to 1990, the GCC has recorded to be one of the highest population growth regions where its population has doubled from 17.8 to 33.5 million.It is projected that the GCC population will reach over 70 million by the next decade [5].Beside population growth, water demand is also a function of economic activity of countries as shown in Fig. 1 [7].Despite the increasing trends of water demand and Gross domestic product (GDP), the water use intensity per thousand GDP depicts a decreasing trend indicating improvement in water utilization by higher economic GDP, as shown in Fig. 2 [8].
In view of the increasing trend of freshwater demand, seawater desalination is considered as one of the most feasible and practical sources of potable water supply in meeting the water scarcity of GCC countries.Presently, the global installed desalination capacity is growing exponentially, producing over 100 million cubic meters of freshwater per day from the estimated 19,000 desalination plants [9,10].However, more than half seawater desalination plants of the world are installed in the Middle East and GCC countries, as shown in Fig. 3 [11].
Hitherto, two major types of seawater desalination technologies are widely installed, i.e., thermally driven and membrane processes [12,13].In thermal desalination, the separation of impurities and dissolved solids from seawater is achieved through the supply of low-grade thermal energy for evaporation and followed by the condensation processes under full saturation states of plants.Due to higher salinity, temperature, turbidity, and the frequent occurrences of algae blooms of surrounding seawater of GCC countries [14], the thermal processes, namely the multi-stage flash desalination (MSF) and the multi-effect desalination (MED), are dominantly used for practicality.On the other hand, the seawater reverse osmosis (SWRO) processes utilize high pressure seawater feed to overcome the osmotic pressures of solute concentration difference across the membranes.Although SWRO method has been widely use in other part of the world, due to favorable seawater feed conditions, but it has yet to be proven for long-term (plant life) sustainable desalination in the environment of harsh conditions of GCC seawater.
From the perspectives of plant owners and consumers, one of the major concerns is the overall unit cost of water production from the available desalination plants, that are normally expressed by capital (CAPEX) and operation (OPEX) costs.Recent reports by Global Water Intelligence [36] presented the average assorted categories of the CAPEX and OPEX, as shown in Fig. 4.Such unit water costs were analyzed by GWI over the lifespan of plants, expressed in terms of US$ per cubic meter of water produced.It can be noticed that the total unit cost of water produced for MSF, MED and SWRO are $1.023, $0.738 and $ 0.749 respectively.On closer examination of the tabulate data of CAPEX/OPEX, i.e., $0.42/$0.603for MSF, $0.29/$0.448for MED and $0.290/$0.450for SWRO, the respective itemized cost of water is relatively high for achieving the goal of sustainable desalination.Hence, the motivation of this paper is to address the unit cost of desalinated water from the viewpoints of high CAPEX and OPEX of desalination plants.The sections below describe the innovative direct-contact spray evaporation and condensation (DCSEC) approach to seawater desalination.

Direct-contact spray evaporation and condensation (DCSEC)
DCSEC is not new as it has been reported since 1980s.However, the methodology did not gain popularity in desalination industry due to two major factors.Firstly, the temperature difference between feed water and equilibrium saturated temperature of vessels is large, typically greater than 10 K.This high temperature difference limit the number of stages for a given finite available top-and bottom-brine temperatures and hence water production.Secondly, the rate of flashing phenomena from water droplets, which depends on available surface area of a droplet, has reached a local (defined by droplet size) asymptotic level giving correspondingly a low water vapor production rate.In 1981, Miyatake et al. [37,38] first reported the spray flash evaporation distillation system.In their experimentation, hot water feed at 40-80 • C were injected into low-pressure vapor zone of the flash chamber.Based on the results, an empirical formula for dimensionless temperature of spray jet was obtained, defining the characteristics of spray flash evaporation.In 2006, Ikegami et al. [39] compared the performance between upward and downward jets and observed better flash evaporation processes with shorter travel distance of feed in the evaporator chamber.Subsequently in 2009 and 2010, Mutair and Ikegami [40] conducted similar flash evaporation studies using upward jets but with larger nozzles.They found that the intensity of flash evaporation increased with higher initial water temperatures and the degree of superheat.El-Fiqi et al. [41] conducted similar experiments   with diameters of spray nozzles up to 0.4 mm, whilst an injection pressure of up to 6 bar was used.They measured the degree of superheat ranges from 2 K to 18 K with the feed water temperatures from 40 to 70 • C at assorted flow rates.Chen et al. [42][43][44] also simulated the droplet evaporation processes in a single-stage configuration and observed the relationship between increasing water productivity with initial droplet velocity.They highlighted that smaller droplet of feed are essential parameters for enhancing the evaporation processes.The above-mentioned work was, unfortunately, conducted with pure water as feed.This paper presents DCSEC design and experimentation using actual seawater as feed drawn directly from the Red Sea.The details of experimental system design and experimentation are presented in the sections to follow.

DCSEC experimental facility
A lab-scale DCSEC desalination system has been designed, fabricated, and installed at laboratory LFO 155 of King Abdullah University of Science Technology (KAUST).The schematic of pilot is presented in Fig. 5 and actual pilot is shown in Fig. 6.It comprises the hollow chambers of evaporator and condenser, along with other supporting external components such as a heater, pumps, heat exchanger, feed tank, and distillate tank.To eliminate salt carryover, demister is installed in between evaporator and condenser chamber.In this design, no physical components or interfaces were used for separation, for example, the porous membrane used in SWRO plants or heat exchanger tubes in MSF and MED.Consequently, the DCSEC has much lower CAPEX as compared to conventional desalination methods.Its simple design minimized scaling with ease of operation and hence, it has the ability to handle seawater at high salinity up to 200,000 ppm.The other advantage of DCSEC is the ability to use a low grade or temperature heat source, and consequently, it has to operate under partial vacuum pressure conditions corresponding to the saturated temperatures of the vessels.
For testing, after successful vacuum leakage test, system was prepared for real seawater experimental test.Desired vacuum pressure in evaporator, condenser and auxiliary chambers such as feed, distillate, and brine were obtained using oil driven vacuum pump.At same time, seawater and potable water valves were open to evaporator and condenser.The feed and potable water pumps were turned on for circulation together with the electric heater to raise the evaporator's feed water temperature.The evaporator temperature was controlled using a thermistor controller and a chilled water heat exchanger was installed to attain the condenser's desired temperature.A data logger was used to log all parameters such as temperature, pressure and flow rates at different location on pipelines.Distillate production from the condenser chamber was also measured with a computer logged weighing machine.Upon completing the first experiment for certain time interval, the feed seawater's supply temperature was raised by changing the heater element's electrical voltage and second test was conducted.These operating and control procedures were repeated for all test matrices and configurations.Design and operational parameters are given in Table 2.The DCSEC system was flushed with fresh water for 1 h at end of day to maintain the integrity of components (made of stainless steel 316L) from seawater's aggressive characteristics.

Results and discussion
Owing to the wide range of temperature difference (ΔT) across the single-stage evaporator and condenser, three sets of temperature differences between the chambers were adopted, namely 3-4 K, 5-6 K, and 7-8 K.For each ΔT range, the temperature of the heated seawater supplied to the evaporator was controlled from 35 ± 0.5 • C to 60 ± 0.5 • C at increasing temperature intervals of 5 o C per test run.To maintain the set ΔT of chambers, the condenser rejected heat to the ambient was achieved by regulating chilled water flowrates to heat exchanger.Due to the large thermal mass of the chambers, each experiment took 30 min to reach a steady-state, and the distillate measurement were recorded over the next 180 min.
For each feed inlet temperature to evaporator, the temperature difference of vapor between evaporator and condenser was manually controlled by adjusting the heat rejection rate of condenser to the ambient.These tests were conducted over the mentioned temperature up to 60 • C for both types of potable and seawater and concomitantly, the distillate were collected on the weighing balance.It is noted that the distillate production for seawater were generally lower up to 3.5% than those similar experiments obtained with potable water, due mainly to the effects of the boiling point elevation of seawater.For design purposes, each distillate production is normalized with evaporator volume that is L/h.m 3 .It is seen from Table 3 that at the feed temperature of 60 • C with a ΔT of 7-8 K, the maximum distillate production recorded were 8.97 and 9.28 L/h.m 3 for seawater and potable water respectively.A summary of Fig. 6.Direct contact spray-assisted evaporation (DCSEC) experiment facility at KAUST, Saudi Arabia.

R. Alrowais et al.
test results for all key parameters affecting the performance of DCSEC desalination method are shown in Table 3.
Alternatively, a pictorial and succinct trend of DCSEC distillate production with the respect to the temperatures of potable and seawater feed, but at the highest temperature difference (ΔT of chambers) of 7-8 K, can be seen from Fig. 7. Also from Fig. 8, it is observed that the distillate production from flashing (L/h.m 3 ) is linearly proportional to the temperature difference between the seawater feed and saturation temperature of evaporator chamber.Furthermore, the onset of liquid flashing into vapor could only commence after a minimum threshold of 3 K, caused by the presence of non-equilibrium phenomena such as heat transfer resistances and temperature depression of salt concentration.
Lastly, distillate water quality was also analyzed, and percentage rejection of various components was calculated as presented in Table 4.

Thermo-economic analysis
Thermo-economic analysis combines energy analysis and economic principles to determine the unit cost of water in terms of capital (CAPEX) and operation (OPEX) costs.Based on the plant design comprising: (i) daily water production capacity in cubic meter, (ii) lifespan of plant and (iii) interest rates for capital resources, the amortization of capital cost is performed with the capital recovery factor (CRF) method.The operation cost is determined by summing the yearly recurring costs of: (i) thermal energy input, (ii) parts replacement, (iii) chemical used for water treatment, (iv) electricity consumption of desalination and accompanying applications and (v) labor cost.Table 5 depicts the key design parameters of DCSEC with production of 25,000 m 3 /day.These are guessed parameters based on our pilot experimentation.Actual performance may vary as per design and operation.Of these parameters, the two salient parameters of design are the DCSEC distillate production rates [45] and the percentage of heat recovery that could be achieved experimentally.The computed performance ratio (PR: ratio of distillate produced to energy input) of six-stages DCSEC with enhanced heat recovery of condensation energy and micro-bubbles augmentation in distillate production is found to be 9.2 [45].Microbubbles enhance nucleation sites and hence boost the surface flash evaporation.The embedded microbubbles water droplets enhance surface flashing/evaporation due to increased surface area at lower temperature differential 3-4C as compared to 8-15C without microbubble.
Table 6 presents the water production cost comparison of DCSEC with three major desalination processes.The unit water cost of DCSEC has been tabulated in same manner with the categories CAPEX and OPEX as reported by GWI for other processes.In this analysis the cost of electricity is US $0.15 per kWh.To compute for the unit cost of electricity and thermal energy input, the concept of Demister: DP type, round wire 0.08 mm diameter, Density 128 kg/m 3 , surface area 403 m 2 /m. 3 .

Table 3
Summary of the distillate flow rates in liter per hour per unit volume of evaporator (L/h.m 3 ).common or standard primary energy (SPE) platform [46][47][48][49][50] is utilized where electrical and thermal energies has been converted into primary energy using appropriate conversion factors developed in mentioned literature (1 kWh elec and 1 kWh thermal would consumed 1.8 and 0.1 kWh spe respectively).It can be clearly seen that DESEC has lowest water production cost, $0.52/m 3 , as compared to other processes.One can argue that DCSEC energy should be same as MED or MSF.It is correct if DCSEC system is without heat recovery and microbubble enhancement.We demonstrated that heat recovery and microbubble injection can improve performance substantially and hence reduce energy consumption.

Future process design concept
Owning to the simplicity and low cost of DCSEC desalination design, we propose a hybrid solar-driven seawater desalination system that is totally green and sustainable for the future sustainability.Other than the components that produced CO 2 during manufacture, there is zero CO 2 emission from operating such a plant, as shown in Fig. 9.It comprises two major parts: Firstly, a solar energy harvester using an array of heliostats with a double-reflection hyperbolic reflector and compound-parabolic-concentrator (CPC).The collected solar energy is stored underground-bed using granular sand particles when irradiated by the focused rays.To maintain a uniform temperature across the sand store, a mechanical conveyor system moves the sand from bottom to top.Heat from the storage is extracted by circulating thermal oil in tube heat exchanger and delivered to the kettle boiler producing steam to drive the steam turbines for electricity generation.Secondly, the bled steam that is used to power DCSEC system.Low-grade steam is extracted from the low-pressure turbines to drive the thermal-vapor-compressor (TVC), heating the incoming seawater feed supplied to the top-brinetemperature (TBT) stage of DCSEC.Each of the multi-stage DCSEC produces liquid flashing (vapor) by exploiting the excess enthalpy between the decreasing incoming feed and the saturation temperatures of evaporators.Further enhancement of liquid flashing is achieved by injecting micro-vapor-bubbles into the seawater feed of each stage.This hybrid design of solar-driven DCSEC is a major innovation.Solar photovoltaic thermal system can also be integrated to produce additional power for components [51][52][53].

Conclusions
We have successfully demonstration a lab-scale single stage of direct-contact spray evaporation and condensation (DCSEC) with real seawater as feed and the incorporation of piece-wise technique to mimic the operation of multi-stage effect.From these test results, two major factors that significantly impact distillate production are the inlet temperature of feed spray to evaporator and the temperature differential across the evaporator and condenser chambers.It is observed that firstly, the distillate yield is nearly proportional to the temperature difference of seawater feed and saturation temperature of evaporator chamber, and secondly the minimum inception for liquid droplets flashing is about 3 K.A detailed distribution of key parameters, contributing to the OPEX and CAPEX of practical desalination and DCSEC methods, is compared with assorted desalination methods over their life span.Owing to the simplicity of DCSEC design (devoid of internals such as membrane and heat transfer tubes), the unit cost of water is the lowest amongst all practical desalination methods.From the viewpoint of sustainable solar-driven desalination, the DCSEC method is deemed the most favorable technology that could be hybridized with the solar energy harvester and boiler-steam turbines, without emitting any carbon dioxide to the ambient.
Ther thermal Elec electrical

Fig. 1 .
Fig. 1.Historical and projected water demand and GDP of GCC countries.

Fig. 2 .
Fig. 2. Historical and projected water consumption intensity of GCC countries.

Fig. 7 .
Fig. 7. Comparison of seawater and potable water in DCSEC experiments at the same boundary conditions.The feed flow rate for all tests was set at 129 L/h.

Fig. 8 .
Fig.8.Flashed water production rate for assorted temperature differences between feed and evaporator chamber.

Table 2
Detailed design and operational parameters of DCSEC experimental facility.

Table 4
Distillate water quality from DCSEC system.

Table 5
Design parameters of DCSEC with production of 25,000 m 3 /day.
**percentage heat energy recovered from brine.R.Alrowais et al.

Table 6
Comparison of CAPEX, OPEX and the total unit cost for assorted desalination methods.