A dirty job: dishwasher wastewater reuse and upcycle through an ad hoc engineered microbial consortium

The study targets the circular economy challenge of reducing freshwater consumption and recovering resources from wastewater in alignment with SDG 6. Dishwasher wastewater, a form of greywater typically overlooked due to small point-source volumes, contains nutrients from food residues and low levels of contaminants when eco-labelled detergents are used. Engineered microbial consortia combining photosynthetic (autotrophic) and heterotrophic partners can function as biofilters that remove nutrients and improve water quality sustainably and cost-effectively. This work evaluates whether an ad hoc engineered cyanobacteria–bacteria consortium can operate robustly under semi-continuous conditions mimicking a household Zero Mile system, both to reclaim DWW for reuse and to upcycle it as nutrient-rich irrigation water for indoor vertical gardening.

Wastewater reuse is increasingly recognized as a key component of circular economy strategies and is governed by regulatory frameworks in the EU and USA aimed at environmental protection. Greywater irrigation can reduce freshwater demand and fertilization needs but untreated greywater can pose chemical and microbial risks. Algal–bacterial consortia (microbial biofilters) have shown promise for nutrient removal and bioremediation across wastewaters via complementary phototrophic and heterotrophic metabolisms, often forming structured biofilms suitable for diverse loads and conditions. DWW is comparatively low in heavy metals, pathogens, or pharmaceuticals and largely contains tap water, eco-labelled detergents, and food residues, making it a feasible candidate for biological treatment and reuse. Prior patents and industrial approaches (e.g., reverse osmosis) face brine disposal or detergent constraints, whereas biofilters can close loops without such drawbacks. Previous work by the team demonstrated nutrient removal in DWW using a T. variabilis–based consortium in short-term batch tests, motivating the present long-term, semi-batch, replenished evaluation.

Consortium and culture conditions: An engineered one-to-three consortium was assembled from the heterocytous cyanobacterium Trichormus variabilis (VRUC168; ATCC IUCC 1444; MSU A-37) and three heterotrophic bacteria isolated from DWW (Acinetobacter sp., Aeromonas sp., Exiguobacterium sp.). Cultures grew in undiluted DWW produced by a household dishwasher (Electrolux EES69300L, eco cycle) using EU Ecolabel detergent (Everdrop) and stored at 4 °C in the dark (u-DWW). Experiments ran at 25 °C under 130 µmol photons m−2 s−1. Semi-batch tests: Two semi-batch designs simulated Zero Mile operation with periodic replenishment (every 4 days) of 25% of culture volume using u-DWW, after vigorous mixing to resuspend solids.

• S-Rep test: 50 mL working volume in ventilated 100 mL polypropylene flasks, duration 28 days, six refills. Growth monitored via in vivo chlorophyll a absorbance (665 nm) and turbidity (730 nm); final dry weight after 48 h at 37 °C.
• M-Rep test: 500 mL working volume in 1.0 L glass flasks, duration 52 days, 13 refills (total 8.125 L treated produced). Growth monitored by turbidity (730 nm) to avoid aggregate disruption. Consortium structure and microscopy: Confocal Laser Scanning Microscopy (Olympus FV1000) with DAPI staining was used for 3D biofilm imaging, member distribution, and viability. Samples were fixed (4% formaldehyde), permeabilized (0.3% Triton X-100), stained with DAPI, and imaged using UV (405 nm), green (543 nm), and red (636 nm) excitation; 3D reconstructions with IMARIS. Nutrient and pH analyses: Total nitrogen (N) and total phosphorus (P) in soluble fraction were measured in centrifuged (3400 g, 15 min) and 0.2 µm-filtered aliquots per APAT methods; calibration curves used NaNO3 (0–20 mg N/L) and KH2PO4 (0–200 µg P/L, with oxidation). pH was recorded over time. NGS community profiling: DNA from (i) 3D aggregate, (ii) planktonic medium fraction, (iii) T. variabilis culture, and (iv) DWW was extracted (PowerSoil kit). The 16S rDNA V3–V4 region was amplified (Pro341F/Pro805R) and sequenced (Illumina MiSeq). Reads were processed via QIIME2 (DADA2 denoising, SILVA 138 classifier), ASVs clustered at 97% into OTUs. Diversity analyses resampled to 2048 sequences; statistics included ANOVA (Tukey) and PERMANOVA. Data deposited under SRA BioProject PRJNA942250. Plant trial (Zero Mile demonstrator): Twenty-four Lactuca sativa seedlings were grown in a vertical garden demonstrator under continuous light (130 µmol photons m−2 s−1) at room temperature for 52 days. Four irrigation treatments (40 mL every 2 days): (i) treated DWW from M-Rep (t-DWW), (ii) untreated DWW (u-DWW, 23 mg N/L), (iii) tap water (TW), and (iv) tap water with fertilizer (TW+NP; 22.3 mg N/L, NPK 19-19-19). Biomass (fresh/dry weight) was recorded; quality indices included soluble solids (refractometry, °Brix), total phenols (Folin–Ciocalteu, GAE), total flavonoids (AlCl3 method, QE), and photosynthetic pigments (chlorophyll a, chlorophyll b, carotenoids; 80% acetone extraction; absorbance at 663, 644, 452 nm; concentrations by Lichtenthaler equations). Statistical analyses used Two-Way ANOVA with Sidak/Tukey post hoc as appropriate; Shapiro–Wilk for normality.

Consortium structure and growth: The consortium formed robust, non-adherent, reconfigurable 3D aggregates whose shape adapted to vessel geometry. CLSM confirmed healthy cyanobacteria and interspersed heterotrophs in aggregates. S-Rep (50 mL, 28 days): Periodic 25% replenishment every 4 days supported thriving growth in 100% DWW. From day 12, chlorophyll a and turbidity were significantly higher than controls in BG11 (Two-Way ANOVA Sidak, p < 0.0001). Biomass increased from 10.2 ± 2.9 mg DW (start) to 26.4 ± 1.5 mg in DWW vs 24.6 ± 1.8 mg in BG11 (ns). pH rose from 7.1 (T0) to 9.28 ± 0.12 (T28), indicative of photosynthesis. NGS across T4–T20 yielded 127,637 sequences, 243 ASVs, 105 OTUs; aggregate (71,485 seq; 82 OTUs) vs planktonic (56,152 seq; 90 OTUs), 67 shared OTUs. Aggregate composition remained stable over time (ANOVA Tukey p > 0.05) despite DWW refills; 3D aggregate vs suspended communities differed significantly (PERMANOVA p < 0.001). Consortium bacteria dynamics: Aeromonas stable, Exiguobacterium increased, Acinetobacter decreased and partly planktonic. Nutrients: accounting for inputs and discarded fractions, soluble N and P supplied totaled 2.16 mg N and 112.39 µg P; residuals at T28 were 0.25 mg N and 0.02 µg P, corresponding to 88.6% N removal and ~100% P removal. Treated DWW N/P levels fell within EU discharge limits (91/271/EEC; 98/15/EC). M-Rep (500 mL, 52 days): Aggregates remained dark green and stable throughout; turbidity values indicated sustained viability. Total soluble N decreased from 20.25 ± 0.44 mg/L (T4) to 9.22 ± 2.51 mg/L (T52); P from 958.46 ± 46.13 µg/L to 226.53 ± 59.76 µg/L. Considering total nutrient mass balances (including replenishment inputs and removed volumes), removal rates were 82.0% for N and 91.7% for P. P removal increased sharply by day 12 and stabilized near ~90% from day 16 onward; N removal stabilized ~41% by concentration comparisons but totaled 82% by mass balance. Plant performance (52 days): Lettuce irrigated with t-DWW achieved the highest fresh weight (8.11 ± 3.22 g), exceeding TW+NP (5.77 ± 0.85 g) and TW (6.63 ± 3.04 g). Dry weight was higher with both t-DWW (0.80 ± 0.28 g) and u-DWW (0.76 ± 0.13 g) than TW+NP (0.45 ± 0.03 g). Quality indices improved with DWW: carotenoids were ~4x higher in DWW-watered plants (t-DWW 44.57 ± 4.44; u-DWW 45.91 ± 8.64 µg/g FW) vs TW/TW+NP (~11–12 µg/g). Total phenols were higher with DWW (u-DWW 0.23 ± 0.02; t-DWW 0.20 ± 0.02 µg/mg FW) vs TW+NP (0.12 ± 0.02). Flavonoids were highest with t-DWW (0.22 ± 0.02 µg/mg FW) vs u-DWW (0.11 ± 0.02) and greatly exceeded TW+NP (0.02 ± 0.01). Soluble solids increased with DWW (u-DWW 5.33 ± 0.70%; t-DWW 4.80 ± 0.35%) vs TW+NP (2.60 ± 0.20%). Statistical significance indicated benefits of t-DWW across growth and quality metrics.

The engineered cyanobacteria–heterotroph consortium maintained structural integrity, community stability, and functional performance under periodic semi-batch replenishment closely resembling Zero Mile household operation. The mutualistic metabolism—oxygen and CO2 exchange, heterotrophic mineralization of food residues, and autotrophic assimilation—supported sustained nutrient removal and biomass growth. NGS confirmed resilience of the designed consortium against exogenous microbiota introduced with each DWW refill, with aggregate communities remaining distinct and stable. High removal efficiencies for total N and P across both scales and over extended durations indicate that the biofilter can consistently reclaim DWW to levels compatible with environmental discharge standards. Importantly, reclaimed DWW effectively fertilized lettuce, enhancing biomass and nutritive/organoleptic qualities (carotenoids, phenols, flavonoids, soluble solids), illustrating successful upcycling into plant production. These outcomes address the research aim of validating the consortium beyond proof-of-concept, moving toward TRL4–5, and demonstrate feasibility for integrated household water reuse and urban agriculture within a circular economy framework.

This work advances an ad hoc engineered microbial consortium as the core of the Zero Mile system for dishwasher wastewater reuse and upcycling. Under semi-batch operation with periodic replenishment, the consortium stably forms 3D aggregates, sustains growth, and achieves high nutrient removal (S-Rep: ~88.6% N, ~100% P; M-Rep: 82% N, 91.7% P by mass balance). The treated effluent effectively supports indoor lettuce cultivation, improving biomass and valuable nutritional quality indices. Together, these results support reduced freshwater consumption, decreased wastewater discharge, and added value through on-site crop production and potential reuse of microbial biomass as microbial fertilizer. Future work should scale to liter systems with full operational cycles (complete refills every ~2 days), conduct comprehensive monitoring for salts and any residual contaminants, optimize biofilter design and control, and assess long-term system robustness, regulatory compliance, and user integration for market-ready deployment.

Experiments were performed at bench scale (50 and 500 mL) with partial (25%) replenishment every 4 days, not full-volume replenishment at the typical household cadence (~every 2 days), so direct extrapolation to fully operational liter-scale systems requires validation. Interactions within the consortium governing P dynamics and early-stage nutrient removal remain to be fully elucidated. Although DWW is expected to have low toxicants when using eco-labelled detergents, comprehensive contaminant and salt monitoring under diverse real-world conditions is pending and planned before market engineering. The plant trial focused on a single crop (lettuce) under controlled indoor conditions; broader crop range and environmental variability should be tested.