Forward osmosis desalination of brackish groundwater: Meeting water quality requirements for fertigation by integrating nanofiltration

Abstract

The concept of fertiliser drawn forward osmosis (FDFO) desalination lies in the premise that fertilisers that serve as draw solutions (DS) add value to the FDFO product water for fertigation. However, because FDFO desalination is concentration based, the process cannot continue beyond the concentration equilibrium, one of the major limitations of the forward osmosis (FO) process. This results in final FDFO product water that, unless subjected to substantial dilution with fresh water, exceeds the acceptable nutrient concentrations for direct fertigation. In this study, nanofiltration (NF) has been assessed as an integrated process to FDFO desalination, either as a pre-treatment or post-treatment, to reduce the nutrient concentrations in the final product water and thereby allow direct use of the product water for fertigation without further dilution. NF as pre-treatment or post-treatment was found effective in reducing the nutrient concentrations using brackish groundwater (BGW) sources with relatively low total dissolved solid (TDS). However, when using higher TDS BGW sources, the product water still required further dilution or post-treatment before fertigation. NF as post-treatment was found to be more advantageous both in terms of reduced nutrient concentrations in the final product water and energy consumption.

Introduction

The increase in fresh water demand due to rapid population growth and the expanding economies are driving water scarcity issues in many parts of the world [53]. Fresh water resources are becoming scarcer every year, with global warming and climate change further threatening water security [32], [58]. With the world's population estimated to reach 9 billion by 2050 [57], the food demand will also inevitably rise further driving the water demand for irrigated agriculture currently estimated at 70% of all global water use [53].

Water has always been a critical problem in Australia, the driest continent on this planet [50]. One of the long-standing and contentious issues has been the over-allocation of water for consumptive use, particularly for irrigation, which has left very little water for the river ecosystem need within the Murray Darling Basin (MDB). The recently adopted MDB Basin Plan intends to reduce the water consumption in the basin by 2750 GL a year [12]. The earlier guide to the plan proposed reducing water consumption up to 4000 GL of water a year which received strong protests from the basins farming community concerned about their livelihood. The fact, however, remains: the environment needs adequate water flows for ecological well-being while the economy requires water for agricultural production. MDB is known as food bowl of Australia, providing an estimated 40% of the total agricultural produce in Australia [34].

Wastewater reclamation and desalination are two reliable options for alternative water sources however, desalination remains an expensive option, particularly for irrigation [53]. Energy always remains a crucial constraint for desalination since energy constitutes up to 75% of the total operational cost of a seawater RO desalination plant [19], [52]. Although, 30 years ago, the energy required for seawater desalination was 5–10 kW h/m³, modern state-of-the-art reverse osmosis (RO) systems need only around 3.0 kW h/m³, [16], [50]. However, the energy efficiency of RO desalination has already reached a plateau and any efforts to further enhance energy efficiency are likely to be incremental [16], [50], [52]. The cost of desalinated water is still about 3.5 times higher than the cost of the natural fresh water [50] and is an inhibiting factor for large-scale irrigation.

Recent efforts have focused on developing low energy desalination technologies since energy constitute the largest component of the desalinated water cost. Forward osmosis (FO) is one such emerging technology that promises low energy consumption [11], [29], [31]. Unlike the RO process which separates water from salt using a semi-permeable membrane at very high hydraulic pressure, FO separates saline water sources by simply using a concentration gradient. A highly concentrated draw solution (DS) that generates high osmotic pressure is used to extract water from the saline water sources. Depending on the end-use of the product water, the diluted DS is usually further processed to separate and recover the draw solutes. At this stage, the FO process suffers from the need for an ideal draw solute, which can be easily separated and recovered with minimum energy. DS separation and regeneration can be an energy-intensive process which is one of the major limitations preventing greater use of the FO process for drinking water applications [24], [33].

However, FO offers a promising scope of applications when a draw solute is used that adds value to the product water. In such cases, the diluted DS can be used directly, thereby avoiding the need for additional separation and recovery steps significantly saving in energy cost for FO desalination technology [24]. This merit of the FO process has been exploited in desalination to provide a nutritious energy drink that uses sugar as the draw solute [8] and in fertigation, which uses fertilisers as the DS [36], [43].

The novelty of fertiliser drawn FO (FDFO) desalination is that the diluted fertiliser DS, after desalination, can be directly used for fertigation (fertilised irrigation) as it contains essential nutrients for plant growth. This avoids the need for the separation and regeneration of draw solutes and therefore, represents a further gain in terms of energy savings. Several types of fertilisers were tested and reported in our previous studies [43], [44]. Water flux comparable to RO process or even higher than seawater reverse osmosis process can be obtained when higher concentration of fertiliser solution was used as DS however, water flux alone cannot be used in measuring the suitability of the process. The diluted fertiliser DS after FDFO process must meet the acceptable water quality for direct fertigation. One of the limitations with the FDFO desalination process was that, the fertiliser nutrient concentration in the final FDFO product water, usually exceeded acceptable nutrient (NPK or nitrogen, phosphorous, potassium) concentrations for direct fertigation. The final fertiliser nutrient concentrations in the diluted DS after FO process depend on the concentration of the feed water (total dissolved solids or TDS) or the osmotic pressure [43], [44], [45]. Since FDFO desalination process is concentration based, the process cannot continue beyond the concentration equilibrium as shown later, and this is perhaps one of the major limitations of the FO) process. This results in final FDFO product water that, unless subjected to substantial dilution with fresh water, exceeds the acceptable nutrient concentrations for direct fertigation. Excess fertiliser is not only an economic waste, but also could cause environmental contamination, soil salinity and plant toxicity. One easy option is to directly dilute the fertiliser solution by adding fresh water before fertigation as shown in Fig. 1(a), but such an option is undesirable, especially when nearby fresh water sources are unavailable or the dilution factor is high.

The main objective of this study is to evaluate the performance of integrated FDFO–nanofiltration (NF) desalination process where NF is applied either as a pre-treatment or as a post-treatment process in order to achieve acceptable nutrient concentrations for direct fertigation, thereby avoiding the need for further dilution. The concept of using NF as pre-treatment or post-treatment to FDFO desalination process was identified in our earlier study [44] however, the detail assessment of the integrated FDFO–NF was not conducted until now. In our earlier works on FDFO desalination process, the assessments were made considering the general saline water for example, brackish water using 5000 mg/L NaCl only and seawater using 35,000 mg/L NaCl only. The assessment in this study however was conducted by simulating the actual brackish groundwater (BGW) found at Burronga salt interception scheme in the Murray-Darling basin. By simulating such conditions, the study intends to evaluate the potential application of the inland FDFO desalination process for direct fertigation. More discussion on the integrated FDFO–NF desalination process can be found in the next section. NF is chosen because of its highly selective rejection properties and also because it can be operated at lower pressure and produces higher water flux in comparison to RO. Complete rejection is not the objective of the post-treatment process as the nutrients in the permeate water are necessary for fertigation of crops and the NF process offers this alternative. Although NF has been applied for the treatment of several types of water and wastewater [7], [22], [37], [47] including separation of metals from water [3], [15] however, to the best of our knowledge, this is the first time NF has been investigated for the recovery of the fertiliser draw solutions integrated with FO process.