Greywater contains inorganic and organic substances, nutrients, pathogens, micropollutants and microplastics. Source-separated greywater using decentralized systems can potentially provide energy-efficient and low-maintenance treatment. If effectively treated, greywater could be a source for non-potable water use in for instance urban landscaping or agricultural irrigation. The overall aim of this thesis was to investigate the treatment efficiencies of two different types of decentralized greywater treatment systems: a) on-site package plants and b) a nature-based solution (NBS) - green wall. These two different treatment systems were assessed based on their removal efficiency of organic matter (BOD, COD, TOC), nutrients (nitrogen (N) and phosphorus (P)), surfactants, indicator bacteria (E. coli and enterococci) as well as microplastics.

The study of the on-site package plants investigated eight on-site greywater treatment facilities of four different types (A, B, C and D). Systems types A-C were commercially available and type D was an onsite built sand filter. The treatment unit of type A consisted of a trickling filter fitted with geotextile resting on a sand filter bed. The treatment unit of type B included a fibrous mineral wool filter material while type C contained a series of fine-meshed plastic filters. Prior to types A, B and D, septic tanks were located to contribute with pre-treatment, whereas type C, the smallest system investigated, included a septic tank within the treatment unit. >90% removal of organic matter (BOD and COD) was achieved by types A and D, but the N removal was comparatively higher by type B (44-68%). Effective P removal was only observed in type D (56%). However, the effluent concentration from all the systems was <3mg/l. The treatment efficiency of type C was found to be relatively low.

In the green wall study, the efficiency of five filter materials (pumice, biochar, hemp fiber, spent coffee ground (SCG) and compost fiber soil) were evaluated with regards to hydraulic loading rates (HLRs) (4.5, 9, and 18 l/d). The treatment efficiency varied significantly with materials and HLRs. Biochar consistently removed 99% of BOD for all HLRs. High N removal (>80%) was observed by pumice and biochar during the high HLR (18 l/d). However, P removal by hemp was comparatively higher (75-85%) than by biochar and pumice. SCG and compost soil was tested with only 4.5 l/d, where compost soil showed effective treatment of BOD (99%), N (82%) and P (85%). SCG was the least effective material releasing more organics and nutrients in the effluent. 

Both the studies showed high concentration (>105 cfu/100 ml) of E. coli and enterococci in the influent and effluent greywater. The treatment systems were in general not effective in removing E.coli and enterococci. The most efficient system was the sand filter (D), achieving 1.4-3.8 log10  E.coli reduction and 2.3-3.3 log10 reduction for enterococci. Biochar achieved similar removal at HRL 4.5 l/d. Using thermal extraction desorption gas chromatography-mass spectrometry (TED-GCMS) technique, polyvinylchloride, polystyrene, poly-ethylene-terephthalate, polyethylene, polypropylene, and polyamide 6 were detected in the influent and effluent greywater in both studies. Even though there was variability in the influent concentrations, low concentrations were observed in the effluents, suggesting the systems were effective in retaining the microplastics.