Re-use of laundry rinsing water by low cost adsorption technology

Shortage of water is a growing global problem. One way of dealing with this problem is the development of technologies for wastewater clean-up and re-use. Laundry accounts often for more than half of the daily domestic water consumption in countries like India. The major part of laundry water is rinsing water. Laundry rinsing water is relatively clean and therefore highly suitable for clean-up and re-use. The objective of this thesis is to design a rinsing water recycler (RWR) for low cost decentral recycling of laundry rinsing water. To design a RWR with an optimal performance, criteria were determined that needed to be fulfilled: removal of the main components from rinsing water, household scale, low cost, no power source needed, easy to use, portable, safe, attractive to culture, no recycling of the adsorbent and low amount of waste. The application of adsorption technology for clean-up of laundry rinsing water offers high potential. It can be low cost, applied in small devices, no power is necessary and is therefore suitable for use on low-income household scale. The project started with the removal of the main component in laundry rinsing water, namely the anionic surfactant, linear alkyl benzene sulfonate (LAS). Selection of the adsorbent is of main importance, because it determines the adsorption capacity and by that the operation cost of the RWR, the size of the RWR and the amount of waste. Furthermore, the adsorbent should be safe to use and safe to discharge in the environment. A selection of potential adsorbents with different surfactant adsorption mechanisms was investigated. The surface charge of adsorbents was found to be the most important parameter to obtain a high adsorption capacity. A positive surface interacts with the negative head group of LAS molecules and results in a high adsorption capacity. Non-ionic interactions, such as hydrophobic interactions between LAS and activated carbons, result in a lower adsorption capacity. Negatively charged materials do not adsorb LAS at all. The adsorbents were compared by LAS adsorption capacity and cost. Layered double hydroxide (LDH) was found to be very promising because of the high adsorption capacity and activated carbons (AC) were suitable because of their relatively low cost. Based on the type of material no safety or environmental issues are expected when both adsorbents are used and disposed. The LAS adsorption capacity of LDH is very promising and therefore the process parameters of the LDH production (co-precipitation method) on the LDH structure, stability and LAS adsorption capacity were investigated. The highest adsorption capacity was obtained for calcinated LDH with a M2+/M3+ ratio of 1 and 2 because of the high charge density at these ratios. LDH can be applied in a small device for re-use laundry rinsing water for short term use only. LDH aggregates are instable and the adsorption capacity of anionic surfactants reduces dramatically after prolonged use and storage in aqueous surroundings. This is probably caused by the rearrangement of the nano size crystallites of which a LDH aggregate consists. The crystallites slip past each other and form a denser structure restricting the access of the surfactant molecules. The RWR operating time depends on the adsorption kinetics. The LAS adsorption rate on activated carbon and LDH was investigated with the zero length column (ZLC) method. The influence of pre-treatment of the adsorbent, flow rate, particle size and initial LAS concentration on the adsorption rate were investigated. The experimental results were described with several models to determine the rate limiting step and accompanying parameters. The adsorption of LAS onto granular activated carbon (Norit GAC-1240) was well described by the selected adsorption model. The effective diffusion coefficient of LAS onto GAC-1240 is 1.3·10-10 ± 0.2 ·10-10 m2/s and does not change with particle size of GAC-1240 or initial LAS concentration. The adsorption of LAS onto LDH was not well described by the adsorption model or the ion exchange model. The LAS adsorption rate follows a first order decline. This cannot be caused by chemisorption because the adsorbent particle size influences the LAS adsorption rate. Surfactant molecules form a double layer or bilayer on oppositely charged surfaces resulting in a film layer resistance. A double layer model resulted in a good description of the experimental results for LAS adsorption onto LDH. The resistance of LAS adsorption onto LDH was found to be situated completely in the double layer outside the particle. The double layer mass transfer coefficient is 7·10-5 ± 2·10-5 m/s. LAS is not the only contaminant in laundry rinsing water. Other contaminates present in laundry rinsing water could influence the LAS adsorption. Sodium triphosphate (STP), sodium carbonate (Na2CO3) and sodium chloride (NaCl) present in laundry rinsing water were investigated for their influence on the LAS adsorption capacity and LAS adsorption rate onto GAC-1240 and LDH. There is no large effect of STP, Na2CO3 and NaCl on the adsorption capacity of LAS onto GAC-1240 and LDH. STP, Na2CO3 and NaCl increased the LAS adsorption rate onto GAC-1240. This is caused by an increase in ionic strength that enhances LAS adsorption. For LDH, NaCl increased the LAS adsorption rate also by increasing the ionic strength. Both STP and Na2CO3 decrease the LAS adsorption rate. CO3 2- and STP compete with LAS for the adsorption onto LDH. However, in time LAS expels CO3 2- and STP from the LDH structure. The application of a suitable adsorbent in the RWR is most practical in a column operation. The main reason is the high adsorption capacity of the bed since it is in equilibrium with the influent concentration rather than the effluent concentration. Small column experiments were performed to investigate the adsorption of LAS onto GAC-1240 in a column application. The column is designed for a long term operation and therefore LDH is not investigated. The influence of flow rate, bed height, initial LAS concentration, external mass transfer and flow direction on the breakthrough curve was investigated. In parallel a mathematical model was developed that described the experimental results well. The main deviation between the model and experimental results is caused by neglecting the effect of the particle size distribution of the adsorbent. The model assumes one particle size, where in practice the adsorbent consists of particles ranging from 315 to 500 µm. The model is used to design a column for the rinsing water recycler (RWR) to treat 25 litres of laundry rinsing water per day during an extended period. This resulted in two designs; a column (Diameter=0.06 m; Heigth=0.18 m) with a flow rate of 50 ml/min and with a flow rate of 100 ml/min. The adsorbent cost of both columns is around $12-15 per year. Three prototypes of the RWR were developed for the clean-up of laundry rinsing water. Two prototypes consist of GAC-1240 in a column operation: the bucket-tobucket and siphon. The third prototype, the permeable bag, is designed for short term operation and instantly cleans the laundry rinsing water during rinsing. The permeable bag was tested with a LAS solution and GAC-1240 or LDH. The amount of GAC- 1240 and LDH to clean one litre of rinsing water was high, which makes the cost and amount of waste too high, therefore the permeable bag is disregarded. The two prototypes consisting of the column operation were tested with model rinsing water. Model rinsing water contains a high concentration of particulate soil that does not settle and easily clogs filters and columns. Therefore, an additional step, coagulation was introduced to remove the particulate soil. The combination of coagulation and adsorption in the RWRs is very effective in removing LAS, STP, perfumes and model soil. The bucket-to-bucket and siphon prototypes meet all the initially determined criteria and were exposed to early consumer tests. The RWR prototypes were discussed in two consumer groups and successfully tested by four individual consumers in Phulera, Rajasthan, India. The flow rate is an important point for improvement according to the consumers. This can be improved by increasing the diameter of the column or by increasing the LAS adsorption rate by decreasing the particle size of the adsorbent. The consumers are interested in using and purchasing the prototypes because they are easy to use, small and clean the rinsing water to a satisfactory quality to reuse it for other household applications.