Elsevier

Chemical Engineering Journal

Volume 187, 1 April 2012, Pages 275-282
Chemical Engineering Journal

A novel approach for SWRO desalination plants operation, comprising single pass boron removal and reuse of CO2 in the post treatment step

https://doi.org/10.1016/j.cej.2012.01.080Get rights and content

Abstract

A different approach is presented for the operation of seawater RO desalination plants in which the boron concentration in the product water should not exceed 0.3 mgB/l. The approach is based on strong acid (either H2SO4 or HCl) dosage to the feed water to attain pH  4.3, followed by CO2 stripping and subsequently strong base addition to pH 9.00–9.25. At this high pH range, a high B removal efficiency is attained, and since the water is practically devoid of carbonate species, no CaCO3(s) scaling takes place, and pH elevation is limited by Mg(OH)2(s) precipitation, expected only at pH > 9.45. The approach enables operation in the absence of antiscalants. Furthermore, CO2 stripping is effected in stripping towers in two steps: the high CO2(aq) concentration is first stripped by vacuum-operated stripping towers and the CO2-rich air is used for dissolution of calcite in the post treatment stage. The remaining CO2 mass is stripped to the atmosphere using blower-assisted stripping towers. This paper aims at introducing the new concept and providing “proof of concept”. The paper addresses experimental and theoretical aspects of the proposed process, as well as engineering and economic evaluation. The proposed approach is shown to be both technically feasible and cost effective, as compared with conventional boron removal alternatives.

Highlights

► A new design approach is presented for removal of boron within SWRO operation. ► Low B is attained by elevating feedwater pH, in one RO stage, without antiscalants. ► CaCO3 scaling is avoided by removing inorganic carbon from feed by acid stripping. ► CO2 from feedwater is reused for calcite dissolution in the post-treatment step.

Introduction

Removal of boron species (B) from desalinated water is a significant component in the process design of many seawater reverse osmosis (SWRO) desalination plants. At concentrations above ∼1.0 mg B/l, boric acid is known to damage various agricultural crops and plant species used in municipal gardening. For example, crops such as avocado and most citrus types are sensitive to B at the concentration range 0.5–0.75 mgB/l [1]. Therefore, although the world health organization (WHO) had recently updated up the guidelines for B concentration in drinking water from 0.5 to 2.4 mg/l [2] (due to strictly human health reasons), it is most likely (and also apparent from the results of recent international bids) that the demand for low B concentration (0.3–0.8 mgB/l) in desalinated waters will remain unchanged. Boron exists in natural fresh water as a weak acid with a thermodynamic pK value of 9.23. At pH values lower than the pK, the protonated, neutral, boric acid species (B(OH)3) dominates, while above it the negatively charged borate ion (B(OH)4) prevails. In seawater, which usually contains ∼5 mg B/l, the apparent pK value is ∼8.6 due to seawater's ionic composition [1]. The boric acid species, which dominates in natural seawater pH of ∼8.2, is poorly rejected by the commonly available RO membranes. While the rejection of charged ions, such as Na+ and Cl, is higher than 99%, practical B rejection using standard SWRO membranes is only ∼65–80%, corresponding to ∼0.9–1.8 mg/l B in the permeate [3]. Thus, either ion-exchange-based post treatment (PT) or the operation of a second (brackish) RO pass is typically implemented in order to meet the sometimes strict B regulations [1]. Application of a BWRO second pass includes dosage of a strong base to the first SWRO pass permeate, in order to elevate pH to 9.5–10.5, prior to its introduction into the membrane. The pH elevation diverts the boric acid species towards the borate ion, whose rejection by RO membranes is much more efficient. Ion exchange (IX) technology utilizes a resin with a high affinity towards B, which adsorbs B(OH)4 at basic to neutral conditions. Strong acid is required for the regeneration of the resin and a strong base is required for neutralization thereafter [1]. Several process configurations were developed that make use of these technologies, including combinations of the two [1], [3], [4]. New boron removal approaches have also been recently suggested [5], [6]. Cost approximations for B removal from RO permeates at the post-treatment stage [4], [7], [8], resulted in a roughly similar cost range, i.e. between 0.04$/m3 and 0.1$/m3 for either the IX- or BWRO-based methods. While energy consumption is the major cost factor for the operation of BWRO B removal, consumption of chemicals is the most significant cost item associated with the IX approach. Cost evaluations for B removal sometime disregard permeate loss associated with the removal process. Such losses are responsible for a substantial increase in the product water cost, as shown in the cost assessment section at the end of this paper. The recovery value of the second pass is limited by brucite (Mg(OH)2) and calcite/aragonite (CaCO3) scaling, which is strongly pH dependent. However, pH elevation is necessary for effective B removal. Thus, an increase in the second pass recovery or B removal beyond the scaling limits requires additional dosage of costly antiscalants. Desalination of the 2nd pass brine carried out in order to increase the recovery value, as performed, for example, in the Ashkelon desalination plant in Israel, results in higher energy investment and larger footprint.

At this point in time only a few attempts for single pass B removal have been published. Field tests conducted by Redondo et al. [9] yielded B concentration of 0.79 to 0.86 mgB/l in the 1st pass permeate, using a high-rejection Dow membrane (SW30HR-380). Redondo's results conform to Koseoglu et al. [10], who investigated B rejection with different high B rejection membranes and reported 85–90% rejection at natural-pH seawater. Such approach seems impractical not only for reducing B to below 0.3 mgB/l, but even for attaining 0.75 mg/l B in the product water, because of instabilities associated with variations in temperature and seawater B concentration. Moreover, the latter authors reported a maximum of 30% recovery, while the practical higher recovery values are bound to reduce the total B rejection. Rejection of over 99% could be achieved by adjusting pH to 10.5 however, flux reduction due to chemical scaling occurred in their experiments after an operation time of merely 14 h. Dominguez-Tagle et al. [11] examined theoretically the potential for single-pass B removal for small desalination plants, by considering 14 commercial membranes. They concluded that with the most up-to-date high B rejection membranes, a high B removal ([B] < 0.5 mg/l in the product water) can be achieved at the expense of a 12% increase in energy consumption. A work done with a Doosanhydro membrane showed that by adjusting the seawater feed to pH 9.0, a total B rejection of 96% could be obtained [12]. No permeate flux reduction was observed during 45 day operation, although a laboratory experiment showed that CaCO3 precipitated at this pH value even in the presence of antiscalants. SWRO operation at high pH was tested at full scale in the Larnaca SWRO desalination plant in Cyprus. The pH was elevated from 8.02 to 8.60 and 1st stage permeate B concentration was consequently reduced from 0.96 to 0.6 mgB/l. No antiscalants were added to the feed water and a thorough monitoring campaign showed that the membrane had not scaled; however, operation period lasted only five days [13]. Various configurations for seawater B removal were economically assessed as part of a work initiated by the U.S bureau of reclamation: Single-pass operation at pH 8.5 was found to be the most cost-effective alternative for arriving at a permeate concentration <1.0 mgB/l [14].

In the current work a new approach for B removal, applying a single SWRO pass, was suggested and evaluated. The process is aimed at separating boron species to attain B concentration <0.3 mgB/l in the product water, by using low-energy SWRO membranes. The idea is to elevate pH to beyond pH9, and at the same time avoid the risk of chemical fouling. Since the main solid that precipitates on the surface of SWRO membrane is CaCO3 (mostly aragonite, sometimes calcite) and since the total inorganic carbon concentration (which contributes most of the seawater's alkalinity) in seawater is relatively low, the approach adopted was to almost completely divert the proton accepting species in seawater (i.e. CO32− and HCO3) to CO2(aq) by strong acid dosage and subsequently to separate and reuse a required portion of the CO2 for dissolution of CaCO3(s) in the post treatment step. The rest of the CO2 mass is stripped to the atmosphere. Once devoid of carbonate species, the pH of the feed water can be cheaply elevated to pH > 9, which allows attaining the required B removal percentage with no risk of CaCO3(s) fouling the membrane. The new approach is shown to be cost-effective compared to currently implemented post treatment alternatives. Apart from single-pass B removal the proposed approach has positive economic and technical repercussions on the post treatment stage, as detailed below. The aim of the current paper is to present the new approach and establish “proof of concept”. Optimization studies are underway.

A schematic description of the process is shown in Fig. 1. The process begins with the acidification of the seawater feed to pH 4.3–pH 4.4, at which range the HCO3 concentration is reduced to below 1.0 mg/l. The acidified seawater, highly supersaturated with CO2(aq), is now subjected to a two-stage packed bed de-gasifying step (using conventional degassers). The first stage is aimed at removing 20–30% of the CO2 mass by vacuum and a low air-to-water flow ratio (Qg/Ql,). The goal is to form an air stream rich in CO2(g) which will subsequently be reused for CaCO3(s) dissolution in the post treatment (PT) stage. The rest of the CO2 mass is removed in the second degassing stage using an air blower and a high air-flow to water-flow ratio. The aim of this step is to reduce the inorganic carbon concentration (CT) to below 4% of the original value in seawater (overall CO2 removal in both stages >96%). After most of the carbonate system concentration had been removed, feed pH is raised to 9.0–9.25 by the addition of a relatively small amount of strong base, prior to entering the RO step. CaCO3 scaling in the RO step is avoided due to the water's low to negligible CT concentration, while Mg(OH)2(s), another potential scaling agent at this pH range, is intentionally maintained under-saturated. The permeate produced in the RO stage is mixed with the CO2-enriched air from the vacuum (first) degassing stage, before it is transferred to the calcite dissolution reactor. That is, The CO2-rich air is used to acidify the permeate stream, enabling calcite dissolution at a reasonable rate.

Section snippets

Chemical equilibrium computer calculations

The PHREEQC software package [15] was used to simulate chemical processes in both raw seawater and desalination concentrate. Activity coefficients were calculated using SIT (Specific Interaction Theory) and Pitzer approaches (both integrated in the software's database). Ion composition and pH values representative of the Mediterranean Sea are shown in Table 1. Desalination concentrate (with no antiscalants) was simulated by means of seawater evaporation, similarly to the procedure described in

Theoretical considerations

According to Eq. (1), at the range pH 7–pH 10 a small increase in the pH value of seawater results in a significant decrease in the B(OH)3 fraction out of the total boron concentration (BT) (see also Fig. 3). However, under normal operation, increasing the pH value of the seawater entering the 1st stage of SWRO plants (by strong base addition) is limited by CaCO3 precipitation. Although CaCO3 is, by nature, slightly supersaturated in seawater, it does not precipitate spontaneously due to a

Economic considerations

An extensive work was carried out by Busch et al. [8] to compare the cost of B removal by a 2nd BWRO pass and by ion exchange, using B selective resins (BSR). For the purpose of comparing these alternatives with the one suggested in the current paper, the same assumptions of energy prices and operational conditions which were applied in [8] were used in the cost assessments shown in Table 4. When additional data was needed it was taken from other sources or approximated by the authors.

The

Concluding remarks

The presented approach was shown feasible using theoretical and empirical tools. The cost of the suggested process was found to be significantly lower than the 2nd pass BWRO alternative, and comparable to the BSR method for B removal. Comparing the process with BSR is more appropriate since the BWRO 2nd pass alternative results also in reduced TDS, on top of B removal. The presented cost assessment may naturally vary, according to cost assumptions however, since similar assumptions were used to

Acknowledgement

The writers wish to thank Dr. Mark Wilf from RO Technology and Mr. Nikolay Voutchkov from Water Globe Consulting for their intelligent and helpful comments.

References (29)

Cited by (26)

  • Superstructure based optimization of reverse osmosis desalination systems fed by decarbonated high-pH seawater under boron restrictions

    2022, Computers and Chemical Engineering
    Citation Excerpt :

    The application of RO desalted water for both agricultural and urban gardening also requires low levels of boron. The highest threshold for the boron levels is 0.5 mg/L for blackberry and lemon, and 0.5 − 0.75 mg/L for grapefruit and peach (Nir et al., 2012a). The pH value of natural seawater ranges from 7.9 to 8.2, which is below the first acid dissociation constant of boron pKa (range from 9.23 to 8.60 when the salt concentration increased from 0 to 40 kg/m3 (Tu et al., 2010)).

  • Comparative performance of FO-RO hybrid and two-pass SWRO desalination processes: Boron removal

    2019, Desalination
    Citation Excerpt :

    Boron is a trivalent metalloid element and it is mainly present in natural water in the form of boric acid [1,2]. The average boron concentration of seawater is approximately 5 mg/L [3,4]. Also, boron is an essential element for plants and animals, and it has played a critical role in commercial activities and human development.

View all citing articles on Scopus
View full text