Desalination of Seawater, Synthetic Saline Irrigation Water and Produced Water Using Nano Zero Valent Metals: Results from a Pilot-Scale Desalination System

Abstract

Two pilot-scale desalination systems employing carbon modified nano-sized, zero valent metals (n-ZVMs) were manufactured and tested to determine (1) the degree to which high-salt water (20 to 130 mS) could be desalinated and (2) if this degree of desalination could be maintained throughout an extended treatment period. The two pilot systems (referred to as Generation 1 and Generation 2) consisted of parallel lines of four individual reactors in series, a settling tank and an activated carbon cell at the end of each reactor line. The system capacity was 300 gal in Generation 1 and 600 gal in Generation 2 with a total hydraulic residence time of 6 h per reactor line (one hour per cell/tank). A slurry of n-ZVMs manufactured from mixtures of ferrous sulfate and green or black tea extract was introduced in the first reactor on each line to yield approximately 5 to 45 g of nano metal per 100 L of influent salt water based on dosing experiments required to achieve maximum salt removal at each of the three influent salt contents used, 28 mS, 44 mS and 123 mS. Once dosing was set, continuous runs (14 days, 23 days and 9 days) were carried out. The results demonstrated that maximum removal occurred with 10 g/100 L of salt for the 30 mS salt solution, 16 g/100 L of salt for the 40 mS influent water and 40 g/100 L for the 130 mS influent. Salt removal (expressed as Na⁺ and Cl⁻ removed) approached 78% for the 30 mS influent and 41 mS influent, respectively, while removal for the highest concentration salt influent (130 mS) approached 81%. Continuous operation over the extended time-period showed no significant decrease in salt removal with a typical day to day variation of no more than 10%, suggesting that this approach to desalination could rapidly provide usable water from saline aquifers, seawater or even produced water.

1. Introduction

Engineered nano materials include zero valent metallic particles in the size range of 1 to 100 nm, referred to as n-ZVM [1]. These materials offer a great potential for water treatment due to their high surface area and high reactivity allowing for the removal of contaminants by adsorption or through reductive processes [2,3]. The application of these materials in water treatment has been reviewed [4,5,6], which highlights the diversity of application as well as the numerous types of nano materials available for removing specific contaminants encountered in water treatment systems.

Nano metals, especially nano-sized zero valent metals, including iron (Fe), aluminum (Al), copper (Cu), manganese (Mn), magnesium (Mg), silver (Ag), zinc (Zn) and titanium (Ti) (n-ZVM), have been used in water remediation and contaminant removal [7], primarily as adsorbents. Adsorption is a surface phenomenon wherein an ion or molecule forms a layer on the particle surface via chemical or physiochemical processes. Their effectiveness in treating and removing contaminants is due to the high surface area of the metals (20–40 nm) and their high reactivity due to electron rich structure [8]. This allows [1] reductive type reactions to occur such as the reductive dechlorination of many halogenated organic compounds, (2) removal via adsorption of metal(oid)s by n-ZVMs that include arsenic (As), selenium (Se), zinc, copper, cadmium (Cd) and a host of other metals and radionuclides [4,5,6].

An emerging application of nano metals is in the desalination of various saline waters. An n-ZVM can be used as an adsorptive surface for sodium (Na⁺) and chloride (Cl⁻) and other dissolved species as well as a catalytic agent for salt (as NaCl) removal far in excess of its mass or surface area predictions. Several publications and patents now describe the process of salt removal using n-ZVMs (or m-ZVMs) and several different designs for the rapid, continuous removal of salt by this method have been or are in pilot-scale operation [9,10,11] such as this work. Unlike RO and other advanced technology, our method only requires water to be pumped to the reactors and air to be supplied in low amounts to initiate the reaction. This translates to eminently mobile units, with small electricity demands usually powered by solar panels/battery systems. N-ZVM desalination systems can be used to desalinate waters including seawater, saline aquifer water, saline return irrigation water and oilfield/natural gas-produced water for industrial and agricultural re-use.

The mechanism of salt (principally Na ions and Cl ions) removal is not well understood or completely known but can be related to the following characteristics observed in the reaction between salt and particulate or slurried n-ZVMs:

Under oxidative conditions, air (primarily O₂) initiates the corrosion reaction which is also catalyzed by the presence of Cl ion. Corrosion products include a variety of iron oxides and oxyhdroxides which all have a greater surface area than the n-ZVM particles. For example, n-ZVI (iron) has a surface area of 20 to 40 m²/g but the oxide/oxyhydroxide corrosion products have surface areas between 600 and 1100 m²/g [12]. This additional surface area would provide much increased (30 to 50×) surface adsorption [12].

Akageneite, also called beta-FeOOH, is formed from the corrosion that produces Fe²⁺ which oxidizes and forms in the presence of Cl ions [13]. Chloride has been found on the surface of the mineral and also within the tunnels of its hollandite structure. The chloride content has been measured to range from 1.3% to 17% by weight [14]. Corrosion products can also include layered double hydroxides (LDH) that offer substantial interlayer space for salt removal [15,16].

The surface functional groups of iron oxyhydroxides can carry both negative and positive charges, the balance depending on ambient pH conditions. These protonated surface hydroxyl groups can bind both Na⁺ and Cl⁻, The hydration layers surrounding iron corrosion products have been found to contain Na and Cl, possibly as inner-sphere complexes [15,17].

Of particular interest is whether this technology can be scaled to perform economically under field conditions. In a previous paper [11], researchers examined the use of n-ZVMs as a means of treating saline irrigation return water. The previous process focused on the pre-treatment of the water to remove sulfate prior to desalination and was conducted in a batch mode capable of treating 200 gal per day. In contrast, this paper focuses on the application of the technology to water with very different salt contents; from low salinity (15 parts per thousand (ppt)) irrigation water to medium salinity seawater (22 ppt) to very high salt contents (70 to 130 ppt) such as those observed in produced water. The work was conducted in two pilot reactors, approximately 1/12th the size of a conceptual full-scale system with daily capacity of 7000 gallons and 10,000 gallons for the different reactors.

2. Objectives

The objectives of this work were as follows:;

-

Determine the degree to which high salt water (20 to 250 mS) could be desalinated;

-

Determine if this degree of desalination could be maintained throughout an extended treatment period;

-

Identify weaknesses in the pilot system such that improvements to the full-scale system could be made;

-

Begin to assess waste volumes and disposal options;

-

Identify and quantify the system’s materials and consumables to begin to estimate capital expenses (CAPEXs) and operating expenses (OPEXs) for the desalination process.

3. Materials and Methods

The methods used to conduct this work are detailed and explained in the sections below.

Reactors—Two pilot-scale reactors were used in the study. The Generation 1 pilot consisted of 5, 152 cm tall, 30.5 cm diameter PVC pipes placed in a series (Figure 1 and Figure 2). Influent water was fed via a peristaltic pump to the first PVC reactor where nano media slurry was added also using a peristaltic pump to deliver a precise amount of nano metals to effect desalination. Feed water was subsequently transferred to the other reactors via short (12.7 cm) stretches of polypropylene tubing. Water entering the reactor was forced down to the bottom via a 2.54 cm PVC pipe where it mixed with air delivered at the bottom of the reactor from a manifold driven by a small air pump. The air delivery rate to each reactor was set at 2 L/min continuously. Only the first reactor in the series was dosed with slurry. As the media reacts the corrosion products from iron begin to take up salt, aggregate and then fall to the bottom of the tanks. Therefore, reactor 5 had no air flow in the bottom of the tank which allowed the heavy particles to fall to the bottom. This allowed relatively particle free water to flow into reactor 6. Reactor 6 contained activated carbon to remove color and excess metal slurry.

And the real system:

The Generation 2 reactor is similar but configured slightly differently. The reactor was made from a refurbished aluminum tank that was sandblasted and then coated with epoxy paint. Dividers were welded in place to serve as the individual reactors. As with Generation 1, the first reactor was fed the media slurry with a peristaltic pump, and air bubbled through the bottom of the tank at the same rate as Generation 1. The primary difference was capacity. Generation 1 has a total capacity of 300 gallons and Generation 2 has a capacity of 600 gallons (Figure 3, Figure 4 and Figure 5).

N-ZVM Slurry—The media driving the desalination reaction was made from natural tea (black or green) extraction and metal salts as described in Walker (2023). The general composition of the media is also shown in US Patent 10919784 B2. For the different salt contents (Na⁺ and Cl⁻) the media was made to yield approximately 10 g, 20 g and 40 g of total metal/100 L of saline water. The metals consisted of iron (Fe), aluminum (Al) and small amounts of copper (Cu). The slurry was made in 50 L batches and was shown to be stable for up to 1.5 months at room temperature. It was stored in 20 L polyethylene carboys before connection to the pilot plant.

Influent Water—Due to the lack of large volumes of both irrigation-type water and produced water derived from oil/natural gas extraction, synthetic saline water was constructed in 1500 gal batches using halite, magnesium sulfate, potassium bicarbonate, calcium chloride and calcium sulfate. The seawater was also constructed from the same chemical mixtures and used in the pilot runs until seawater was obtained from Puget Sound, Seattle, WA, in November of 2023. The mixtures used to produce the various waters are shown in

Table 1. Composition of different waters used in the study.

Analyte (in mg/L Except Where Noted)	Saline Irrigation Water [18]	Produced Water 1 [19]	Synthetic Seawater	Seawater Puget Sound
Na+	5800	26,670	9766	8370
Ca2+	215	215	210	310
Mg2+	298	298	298	451
K+	420	420	420	99 ***
Cl−	8800	40,000	14,650	12546
SO42−	2100	2100	2100	1500
Alkalinity	660	660	660	460 ***
pH *	8.2	8.4	8.3	8.5
EC **	33	130	44	41
TDS	18,293	70,363	28,109	21.1

Note(s): * standard units. ** mS. *** KHCO₃ added to increase pH (approximately 200 mg/L).

along with the appropriate references.

Instrumentation, Data Collection and Analysis—During operation of the pilots, continuous (each second) measurement of pH, Cl⁻, EC, air flow and water flow occurred on the influent side of the pilot (feed tank) and the discharge side of the pilot after water passed through the activated carbon. In addition, during select times (each 8–12 h) the same analytes were determined in each reactor tank to observe chemical changes from tank to tank over time. All measurements were captured by the programmable logic controller (PLC) located on the influent side of the pilots (Figure 1 and Figure 3). The PLC controlled influent flow, air meter flow and the various probes collecting chemical information. The pH and EC (and temperature) probes were purchased from Seametrics Corporation (Model CT2X conductivity cell and Model TempHion pH, ORP/ISE smart sensor and data loggers for pH, EC, ORP and Cl⁻.) Both Na and Cl ions were determined by Ion Specific Electrodes purchased from Oakton and Environmental Express, respectively. Calibration of each instrument was carried out using Oakton purchased calibration standards. Calibration was repeated daily for pH and EC, while Cl⁻ and Na⁺ calibration was repeated every 4 to 8 h.

Pilot Operation—Both pilot systems operated similarly. Feed water rates were set via the diaphragm pumps (Generation 2) or peristaltic pump (Generation 1) and fed directly into the first reaction chamber. Flows only varied between 1.85 L/min and 3.7 L/min in both pilots. In the first reaction chamber, the media feed pumps delivered the nano slurry at a rate consistent with feed water flow and the desired amount of nano metal for each type of water. The air pumps fed 1.5 L to 2 L air/min to each reaction chamber at the bottom of the cell. The feed water entering the cells was directed to the bottom of the cell near the air flow stream using internal piping which allowed good contact between the nano media and the air flow and forced the feed water up through the cell. The water in cell 1 was transferred to the next cell via continued feed water flow and again directed to the bottom of that cell. The first four cells in each pilot had similar air delivery, but nano media was only fed to cell 1. Cell 5 contained no air flow which allowed solids to settle, while cell 6 contained 25 kg of granular activated carbon for polishing the treated water.

The reactors were run continuously for 14 days for treatment of saline irrigation water, 23 days for treatment of synthetic seawater and 9 days for treatment of synthetic produced water. Water was stored either in 2 3000-gallon poly tanks or a 1500-gallon above-ground, collapsable swimming pool. Puget Sound seawater was conveyed via a series of large water tanks in a truck bed after pumping from the Sound at a depth of 12 feet below surface. Water temperature only varied from 8 °C to 12 °C throughout the duration of the runs.

4. Results

Dosing: Prior to performing the extended runs, dosing experiments were carried out to determine the amount of nano metal required to remove between 70 and 80% of the salt. This percentage was chosen since 80% removal would remove the salt from typical seawater to a level (10 mS) conducive for RO processing at an efficiency that would produce very little brine. To determine the required dose for each water, about 500 gallons of each were prepared and run through the pilot in a single pass. During the shortened run, the delivery rate of the nano slurry was adjusted to achieve the desired removal. The nano metal in the slurry was always constant, only the feed rate changed. The data from Walker (2023) were used as a guide for initial dosing. Dosing began at 5 g of nano metal per 100 L (0.05 g/L) of feed water and increased slowly (every 2 h) until the desired removal was achieved. Replicate runs at each dose produced removal efficiencies of 73–78% for saline irrigation water, 75 to 81% for sweater and 69 to 87% for produced water. The ultimate dosing selected for each water and the relation to salt content is shown in Figure 6.

Extended Desalination Runs: The results of the extended runs are summarized in

Table 2. Summary of pilot test results.

Run Date	Dose	Reactor	Water Type	Media Type	EC	Duration	Cl initial	Cl Removed	% Reduction	Na initial	Na Removed	% Reduced
	g Fe/L				mS	hours	mg/L	mg/L	%	mg/L	mg/L	%
01/10/23	5	G1	Saline Irrigation	Tea/nano	24	12	7926	4215	53	5270	2793	53
03/10/23	6	G1	Saline Irrigation	Ch/nano	25	12	7926	4638	59	5335	3148	59
04/10/23	8	G1/G2	Saline Irrigation	Tea/nano	26	12	8400	5304	63	5616	3538	63
05/10/23	10	G2	Saline Irrigation	Tea/nano	31	336	8800	6806	77	6696	5156	77
16/10/23	10	G1	Saline Irrigation	Tea/nano	30	12	9500	7383	78	6480	5119	79
05/12/23	40	G1	Produced Water	Tea/nano	131	216	40,000	32,073	80	28,296	22,637	80
20/11/23	20	G1	Seawater	Tea/nano	44	552	14,650	11,194	76	9504	7033	74
05/2/24	20	G1	Puget Sound	Tea/nano	41	16	12,546	9410	75	8370	6361	76

. The first three rows in the table demonstrate the dosing changes and their effect on salt removal. The table can be interpreted as follows:

Column1 is the start date for each run, followed by the dose of n-ZVM used, followed by reactor type (G1 or G2). Next is the water type (saline irrigation, seawater or produced water) followed by the media mixture (usually tea extract and metal salts). The last columns show the initial EC, the initial Cl⁻ and Na⁺, the Cl⁻ removed and the % removed for both Cl and Na ions.

The effect of increasing the dose of the nano metal slurry on salt removal from saline irrigation water is summarized in rows 1 through 4 in Table 2. Increasing the dose from 5 g/100 L to 10 increased desalination from about 50% to 77% at 10 g/100 L. Once the dose was set, this reactor continued to treat saline irrigation water for 336 h, or 14 days.

The extended run with saline irrigation water (row 5) showed that an addition of 10 g of nano metal could desalinate the water by 77% over the 2-week runtime.

The sixth row is a replicate run at 10 g/100 L to ensure this dose could achieve similar results. As noted, desalination was similar or about 78% reduction compared to the 77% reduction in the previous run.

Row 7 is the extended run for synthetically produced water. At a dose rate of 40 g/100 L, 80% reduction was achieved over the entire 9-day runtime.

Row 8 shows the results for the synthetic seawater, showing that a dose of 20 g/100 L could desalinate more than 75% of the entrained salt.

Row 9 was the most recent run (2/24) using seawater pumped from Puget Sound. This comparison to the synthetic seawater in Row 8 was important for validating the approach using constructed seawater. The results for each seawater sample are similar, as a 74% salt reduction was achieved with synthetic seawater, while a 76% reduction was achieved using similar reaction conditions.

Figure 7 and Figure 8 depict additional detail of the 23-day seawater run. Figure 6 shows the NaCl removal calculated from data collected at 12 h intervals. As noted, the desalination was very consistent with only a 5 to 10% deviation from the mean removal efficiency (76%). Between hours 372 and 396 h, the media feed pump became clogged resulting in no slurry feed for approximately 2 to 3 h. Once the feed line was cleared, the desalination proceeded as before the clog occurred.

Figure 8 shows the Cl⁻ remaining in the solution during the same seawater run and demonstrates how Cl⁻ is reduced from 15,000 mg/L to under 4000 mg/L with the 20 g/100 L dose and a 4 h retention time.

Hydraulic Retention Times: One of the reasons for the use of multiple reactors in series was to allow for the independent analysis of salt removal over time. The reactors were sampled at 1 h intervals during the first 12 h of operation for the continuous runs and then every 8 to 12 h during the rest of the runs. Inspections of the samples collected from each reactor during the runs indicated that the reaction was nearly complete between reactors 2 and 3 and was complete by reactor 4. The volume of each reactor is 100 L in Generation 1 and 200 L in Generation 2 with flow rates of 1.85 L/min and 3.7 L/min, respectively, resulting in a hydraulic retention time between 2 and 3 h in which maximum salt removal occurs.

Effect of pH: The pH of the influent water has a marked effect on salt removal. Figure 9 displays the results of eight experimental runs using Generation 2 reactors. While the influent water for all runs was between 24.5 mS and 25.8 mS, the amount of media slurry varied from 5g of nano metal per 100/L to 15 g/100 L. Increasing the nano metal involved increasing the slurry delivery rate which included the tea extract polyphenols and acids which decreased the pH of the influent water within the reactor. In Figure 9 the low pH values (pH from 6 to 6.4) clearly show a much lower efficiency ranging from 39 to 51%. When the pH of the reactor mixtures was kept above 7, the efficiency increased to over 80%. The effect of pH suggests that some waters may require pH adjustment prior to treatment, usually under the following conditions:

If the source water is acidic then a neutralizing chemical such as calcium hydroxide, magnesium hydroxide, potassium bicarbonate or others may be added to the influent side of the reactor. The amount added should be based on a titration to yield enough buffering to keep the solution’s pH above 7 after the addition of the slurry.

If the source water is hypersaline and a significant increase in nano slurry is required to achieve the desired desalination, again, the neutralizing agent should be added at amounts determined by titration.

It is possible that the salt removal mechanism is different at a pH greater than 9 due to the addition of calcium or magnesium hydroxide. The reaction could be dominated via the precipitation of iron oxyhydroxides as soon as corrosion of the iron nano particles begins or from Fe (II) and Fe (III) species in the n-ZVM slurry, much like the sol gel process explored by Antia [20] that might be independent of everything except the total amount of iron in the system. Future work will follow these different possibilities using a combination of wet chemical techniques and surface spectroscopic methods.

Pilot Reactor Issues: During the extended runs, the electrical and mechanical systems were scrutinized to assess potential weak links so that better more durable replacement systems could be identified. The primary observations included the following:

The media feed pump and tubing had to be reconfigured to prevent blockages occurring due to some particle settling in the nano media. It was noticed that some settling of the nano media occurred within one hour of operation and some large particles or clumps of tea extract artifacts caused tube blockages. To prevent this, the nano media should first be filtered through a 200-mesh sieve to remove large debris and then continuously stirred during the media delivery process. Both improvements are underway.

In general, the air flow to the reactors was very steady; however, occasionally the check valves at the outlet from the quick connect valves at the bottom of the tank became clogged with nano media settling to the bottom during reactor runs. This problem resulted in increased flow readings on the inlet side of the manifold that delivers the air flow. To date, three different styles of check valve have been used to determine if one type is superior to others. The problem occurs once every 2 to 3 months, usually in one to two reactors.

The air supply and peristatic pumps have been operating continuously since 01/06/23 and none of the equipment has failed, yet. Over 80,000 gallons (302 m³) of water have been processed during that time.

Waste: Because the reactors can remove 70 to 80% of the dissolved salt load, a substantial volume of waste is produced. For example, the treatment of 1000 L of water at 80% salt removal produces the waste mass and volume described in

Table 3. Waste generated from each water type.

Water Type	Mass	Volume (Density of Solids 2 kg/L)	% Waste to Treatment Volume
31 mS Saline irrigation	12.6 kg	6.3	0.63
44 mS Seawater	17.95 kg	8.9	0.89
130 mS Produced Water	53.04 kg	27 L	2.7

.

The waste product is collected from the bottom of each reactor at cleanout intervals each month. Each of the reactors in the series contains waste, but the majority of the waste accumulates in reactors 1 and 5. However, each of the other reactors contains settled salt/metal waste but in lower amounts. To date, only the Generator 2 reactor has been examined quantitively for waste products. In this case, the reactors treated saline irrigation water with a salt content of 33 mS (similar to the water used in the 14-day extended run). After 2220 L of water had been run through the system, the reaction was stopped, the air supply shut off and all solids allowed to settle over a 24 h period. The removal efficiency was just under 75%. A transfer pump was used to remove most of the water from the individual cells (except the carbon polishing cell) without disturbing the settled solids at the bottom. Next, a vacuum hose was used to pull the solids from the bottom into a series of clean 20 L plastic buckets which had been labeled and pre-weighed. These buckets were left to evaporate to dryness or until no change in weight was observed. The mass observed in each bucket (representing each reactor) is shown below (

Table 4. Mass balance of waste collected versus waste calculated from salt removal efficiencies.

Reactor (Bucket)	Mass (kg)	Volume (Density-2 kg/L) (L)	Compared to Theoretical Mass/Volume
1	15.1	3.04
2	5.5	1.1
3	5.2	1.04
4	5.4	1.08
5	8.1	1.62
Totals *	38.3	7.88	35.1 kg/7.87 L

Note: * Total amount of nano metal added about 0.6 kg.

).

The amount of waste produced was slightly greater (38.3 kg) than predicted from an 80% efficiency (35.1 kg). The waste mass also includes the metal added from the slurry or about an extra 0.6 kg which can be subtracted from the mass observed.

The pilots are now having a waste vacuum lime installed such that the waste accumulating at the bottom of the reactor can be removed, filtered and dried for disposition or recycle/reuse. A comparison of the waste from this system to other desalination systems is not entirely straightforward since this system produces a solid waste, but using the data from these extended runs, it is likely that the “water treated (WT)” to “waste produced (WP)” ratio (WT/WP V/V) is between 80 and 347 depending on the initial salt content. For example, the WT/WP ratio for seawater is 236 or 1L of waste for every 236 L of treated water.

5. Summary

Two different but similar pilot reactors were constructed and operated in several extended runs using different levels of Na⁺ and Cl⁻ to determine (1) the removal efficiencies for each type of water, (2) the required dosages of nano media to effect 80% salt removal, (3) to assess the overall ability of the systems to continuously produce the same effluent quality, (4) to determine characteristics of the process in terms of solids removal and (5) to provide confirmation of the removal via solids mass balance. The results of the study indicated that:

Salt removal was dependent on media dose in a linear manner, with low-salt-content water (20 to 31 mS) requiring 10 g of per 100L of water, medium-level salt water (seawater 41 mS) requiring 16 g/100 L and high-level salt water (>130 mS) requiring 41 g/100 L). The resultant dosing chart can be used to predetermine the nano slurry required based on the EC of the water:

(dose (g) = 0.2926(EC) + 2.101)

The pH of the influent water was also found to be a critical factor in the removal efficiency. If the influent water pH was less than 6.5, removal efficiency was only 50%, but if the pH increased to 7 or higher removal efficiencies were always 70% or greater. The easiest solution for low pH waters would be to add a dose of potassium bicarbonate or other neutralizing agent to the influent to raise the pH to 7 or above. The amount required can be determined simply by titration with a known amount of neutralizing material.

The extended runs showed that the pilot systems could produce lower-salt-content water (70–80% less salt) throughout the duration of each run. A typical variation was less than 10% except where a single equipment malfunction occurred.

The reactor design using five separate reactors in a series appears to be a possible precursor to a full-scale system, but other designs are possible. One of the reasons for the multiple reactors was to allow the solids to accumulate across all reactors rather than a single one which could require frequent stoppages for solids removal. The solids measured in each reactor showed that most of the solids were removed uniformly in reactors 2 through 4 with reactor 1 accumulating the most (about 40% of the total). The next generation of reactors will have a vacuum line at the bottom of the reactors so that solids can be removed on a predetermined cycle or cone bottom tanks can be employed for rapid solids removal. Solids accumulation in reactor 1 was higher than the others suggesting that the solids in reactor 1 may contain a considerable amount of unreacted media in addition to the salt which was removed. This could be due to the manner in which the nano slurry is added: the nano slurry pump outlet is fed into the water inlet line and is forced to the bottom of the reactor with the water. Thus, some of the nano media may never circulate through the reactor with the incoming water. Testing is underway to determine if different placement of the feed line could improve the distribution of the nano slurry in reactor 1.

6. Conclusions

Overall, the pilot testing provided consistent reproducible results suggesting that this approach to desalination could rapidly provide usable water from saline aquifers, seawater or even produced water.

The treatment of saline waters at 30 mS or less yields effluent water in the range necessary for use in the irrigation of many crops, while the treatment of seawater or produced water with levels of salt 40 mS or higher yields effluents in the range of 6 to 10 mS, a range which is ideal for a Reverse Osmosis (RO) influent that would prevent the production of large volumes of brine.

However, aside from seawater, the synthetic waters must be replaced and tested with representative waters from saline aquifers and produced water from oil and gas exploration. If testing continues to confirm these removal efficiencies, then this technology could provide a very low cost, low energy consumption and highly mobile alternative to current desalination systems.

Finally, while the objective of the study was not to determine the mechanism of salt removal, clearly a better understanding of the process would possibly allow for the manipulation of the reaction(s) to enhance salt removal further. As described earlier, the mechanism of the salt removal is not known with certainty and while the aqueous chemistry observations during the reaction is useful, surface chemical spectroscopic methods will be necessary to identify intermediate and final reaction products. The role of layered double oxyhydroxides and the expansion of surface area during corrosion and the formation and identification of Akaganeite with its high Cl content show that these would be excellent candidates for further research.