Nevertheless, our experimental effluent recovery system allowed to reduce the net volume to 32 %. In addition, environmental benefits were provided in the reverse osmosis treatment by the avoided GHG emissions from the use of PV technology rather than the Spanish electricity mix  with savings of 81 % in GHG emissions. As expected, the LCA showed clear environmental benefits associated with the leachate collection and treatment. In addition, the results of the sensitivity analysis show how the impacts were considerably reduced by using renewable energies and reducing fertilizers doses. Many authors have stressed the importance of increasing renewable energy in the production of electricity to improve the sustainability of greenhouse horticulture . Torrellas et al.  reported that the specific energy of water for irrigation for tomato production in the greenhouses of the Mediterranean coast was 0.13 kWh/m3 . In that study, the authors considered the extraction of water from wells. However, in recent years, the overexploitation of aquifers in water-scarce areas, such as SE Spain, has promoted the use of unconventional water resources, such as desalinated seawater. That is why electricity consumption increased around eight-fold in the cultivation cases in which 100 % of desalinated seawater was used for irrigation . In the RE scenario, in which fossil-based energy sources are partially replaced by renewable energies , AC, GW, and EU were on average reduced by 17 %, 13 %, and 10 %, respectively. These results concurred with those obtained by Torrellas et al.  who, byincreasing the renewable energy in the production of electricity by 40 %, reduced AC, GW, and EU by 15 %, 5 %, and 5 %, respectively. Our ambitious RE scenario,dutch buckets moving from 37.5 % to 90 % of renewable energy use, is expected to become a reality by 2050, when Europe is set to become the first climate-neutral continent.Prior studies assessing the life cycle of conventional greenhouses concluded that the use of fertilizers was critical from an environmental perspective .

In Torrellas et al. , a reduction of 30 % in fertilizers led to a 15 % decrease in the EU, 10 % decrease in the GW and 5 % in the AC . Our results agree with that data since our reduction of 30 % in synthetic fertilizers led to a 13 % decrease in the EU, 10 % in the GW, and 12 % in the AU. United Nations has estimated that by 2050 nearly 6 billion people will suffer from clean water scarcity . Thus, future scenarios drive the need to improve water management practices and strategies to ensure water supply. Within this context, water reuse is a promising option to alleviate water stress, while moving towards the Circular Economy principles. Nonetheless, water reuse remains a limited practice due to barriers ranging from technical and economic feasibility to legislative restrictions and social acceptance. Therefore, to solve the imminent water crisis, it is necessary to develop efficient technologies that will make water reuse a sustainable and affordable practice to be widely implemented. With irrigation being the world largest water consumer, the application of reused water for irrigation purposes is a crucial strategy capable of significantly decreasing the demand for freshwater and therefore reducing water stress. Among agricultural techniques, hydroponics is a promising approach that can be implemented worldwide. In this soilless cultivation technique, plants grow in direct contact with water that contains the required nutrients for their development. The typical concentrations of nitrogen, phosphorous and potassium  of common hydroponic solutions are diverse , as the nutritional requirements for the plants depend on many factors, such as plant type, stage of plant growth, seasonal differences or weather conditions . Given that the hydroponic technique requires large amounts of freshwater , it is of great interest to study the potential of technologies capable of treating alternative water sources for subsequent reuse in hydroponics. Among the variety of available technologies, forward osmosis  emerged as a promising solution for water treatment and reuse, as it can recover fresh water from low quality water sources such as seawater or wastewater .

In FO, a highly concentrated solution  extracts water from a low concentration solution , and the water is transported through a dense membrane . FO membranes exhibit high pollutant rejection and have low fouling propensity, and the process does not require hydraulic pressure as it is driven by the difference in osmotic pressures between FS and DS . One of the main drawbacks of FO is the reverse salt flux  ; i.e., solute losses from draw to feed per membrane area and time . Js plays a fundamental role in the design of osmotically driven processes , since it decreases the osmotic driving force , represents economic losses  and causes difficulties with feed concentrate management , hence jeopardizing the benefits of the FO process . As pointed out by Zou et al.  in a review of approaches to reduce reverse solute fluxes in FO, it is crucial for FO operations to control and reduce Js, and they also highlighted the lack of Js data in FO studies. Therefore, detailed studies of solute fluxes in FO are of great interest to assess their impact on FO performance. One of the practical applications of FO is the osmotic dilution of soluble fertilizers for irrigation purposes , as most of them are capable of generating a high osmotic potential . In fertilizer-drawn forward osmosis , the osmotic dilution of the fertilizer DS occurs, with the aim of later being used for direct fertigation  since it contains the essential nutrients for plant growth. FDFO concept was mainly developed in the last decade and has shown promising results. FDFO is particularly interesting when applied to low quality sources as feed water, such as brackish water or greywater, avoiding the demand for freshwater for irrigation. Most FDFO studies have focused on the performance of different fertilizer salts as DS  and their interactions with different membranes . Some recent works have even used commercial fertilizers as DS . Besides, it should be noted that in most of the previous FDFO studies, authors highlight the need for further dilution because final concentration of nutrients in DS were above the threshold tolerated by the plants.

In previous cases, proper DS dilution for direct fertigation was only achieved after coupling FO with other technologies  or by applying additional pressure . Overall, most of FDFO studies were devoted to demonstrating proof of concept for the use of fertilizers as draw solutions and did not focus on the impact of FS salinity, nor on achieving an optimal dilution to the required level of nutrients for plants, especially when approaching osmotic equilibrium. Final DS concentrations suitable for direct fertigation – without further dilution of the final draw solution – are therefore essential for the success of FDFO and more studies are required on the practical application of the process . Finally, even if some studies have focused on bidirectional diffusion of the various ions present in both FS and DS , all were carried out on a very small experimental scale and under conditions far from osmotic equilibrium. For FDFO to be applicable on a full scale, relatively low DS concentrations are required, and it is of interest to achieve the desired concentrations in a single step. Given the current limitations of FDFO concerning the dilution factor of the fertilizer for direct fertigation, it is crucial to conduct more experiments close to osmotic equilibrium, as it will have a great impact on the achievable dilution rate, filtration kinetics and is expected to depend on FS initial salinity and reverse salt diffusion. Within this framework, this study aimed to evaluate the suitability of the FDFO process to achieve an effective DS dilution to generate a suitable nutrient solution for direct application in hydroponic systems . The performance of the FO process at conditions close to osmotic equilibrium, grow bucket as well as ion fluxes through the membrane were also experimentally evaluated. Experiments were performed with constant feed and draw recirculation, leading to continuous DS dilution and FS concentration. All tests were carried out with DS facing the active layer  because this configuration results in higher water fluxes .

Although this configuration of having the active layer facing the DS may lead to fouling, this negative impact was not expected due to FS nature . Additionally, although external concentration polarization may increase with DS facing the active layer, internal concentration polarization would decrease in the proposed experimental setup, since it generally used DI water as FS. The module was positioned vertically with the DS and FS circulating in counter-current , since operation in counter-current leads to better use of osmotic pressure, achieving a higher dilution rate than in co-current mode . The initial volumes were 2 L of DS and 60 L of FS. FS and DS were circulated with a peristaltic pump , with an average flow rate of 34.6 L h− 1 and 60.7 L h− 1 respectively, according to the manufacturer’s recommendations. The water flux crossing the membrane  was determined by measuring the volume extracted from FS to DS thanks to the increase in the mass of the DS with a balance  and considering 1 kg/L as density of DS. A set of tests using individual or blended draw solutes at 0.50 vs 0.05 M of initial concentrations served to evaluate the effects of DS concentration and composition on the process performance . Initial water fluxes were in the same range for all tests using the same DS concentrations and decreased significantly throughout the process due to the dilution of the DS and the consequent loss of osmotic pressure driving force. As expected, water fluxes for 0.05 M DS were low compared to those of the tests with DS at 0.50 M  due to the resulting lower difference in osmotic pressure between FS and DS. Working with low DS salinity  not only reduces Jw because of the lower initial flux, but also because of operating near osmotic equilibrium due to dilution over time. All tests at 0.50 and 0.05 M achieved the targeted dilution rate  except for KNO3 at 0.05 M . Osmotic equilibrium was not achieved in any of the tests  indicating that the water extraction capacity of the tested DS was higher than the 15 times dilution rate, which was the limit of the setup. Only when operating with KNO3 as DS at 0.05 M, a higher EC was observed in the final FS than in the DS, indicating that osmotic equilibrium was reached, at a lower DS dilution rate  than the target. This behavior results from the reverse salt flux from DS to FS, which not only leads to fertilizers losses but also limits the dilution capacity of the system.

In tests with 0.05 M of fertilizer salts in initial DS, reverse salt fluxes  did not exceed 1.5 mmol m− 2 .h− 1 in any case but were affected by the nature of the ions present in the DS . As noted above, the highest Js were observed when using KNO3, leading to the highest diffusion of both of its ions. NO3− has been widely reported as an ion with high reverse fluxes , due to its small hydrated radius. K+ passed through the membrane in equal equivalent concentration to balance the charges and keep the ionic equilibrium in both solutions . Reverse fluxes of DAP ions were much lower, confirming results from other studies . Phosphate Js was up to two orders of magnitude lower than the counter ions present in the DS . Higher FO membrane rejection of phosphate compared to ammonium and potassium has already been reported due to its bigger hydrated radius, and the stronger electrostatic repulsion with the negatively charged membranes caused by its negative multivalent charge . Consequently, phosphate reverse fluxes through the membrane are generally reported to be minimal regardless of the DS composition and concentration . The observed ions Js were different when using individual fertilizers or blended . In MIX 1, reverse fluxes followed the trend NO3>NH4>K > P , which is inversely correlated to their hydrated radii at the same charge type  , and in accordance to other studies .