Pore-size and polymer affect the ability of filters for washing-machines to reduce domestic emissions of fibres to sewage

Abstract

When clothes are worn and washed, they emit fibres into the ecosystem via discharges of sewage that have been linked to the global dispersion of clothing fibres. Facilities that treat sewage divert some fibres from sewage effluent to sludge, but no current methods of filtration eliminate their environmental release. While filters for washing-machines are sold to consumers with the argument they will reduce the emissions of fibres from clothes to the environment, there is insufficient scientific peer-reviewed evidence assessing their ability to retain fibres from washed clothes and reduce environmental contamination. To improve our understanding and develop more realistic methods to assess the efficiency of filters, we washed replicate cotton and polyester garments in replicate domestic front-loaded washing-machines with and without replicate filters (micro- and milli-meter-sized pores), and then quantified the masses of the fibres retained by the filters and those released in the effluent. Here we show micrometer-sized filters significantly reduced the mass of cotton by 67% (F2,6 = 11.69, P<0.01) compared to effluent from appliances with no filters, whilst filters in general reduced polyester fibres in their effluent by more than 65% (micrometer-sized pores) and 74% (millimeter-sized pores) compared to effluent from appliances with no filters (F2,12 = 5.20, P<0.05). While filters with micrometer-sized pores caught larger masses and total proportions of fibres than filters with millimeter-sized pores, the differences were only significant for the total proportions of cotton (t = 4.799 df = 4, P<0.01). For tests with garments of either types of polymer, the filtered effluent still contained up to a third of the original masses of fibres released from the garments. Given the diversity of clothes, polymers, appliances and filters currently sold to consumers, our work shows the value of increasing the rigour (e.g. more levels of replication) when testing filters and the need for further studies that test an even greater diversity of materials and methods in order to meet the growing demand for knowledge from governments, industry and the public.

Fig 1. Composition of the washing-machine filter.

This includes a filter-head (1), filter case (2), stainless steel filter of either 2 mm or 150 μm pores (3), rubber grommet (4), 19 x 51 mm pipe nipples (5), 19 mm female x 25 mm barbed hose with a 90° elbows (6), stainless steel clamps (7), wood-screws (8), 25 mm barbed tube (9), polytetrafluoroethylene tape (10), 25 x 1000 mm hose (11), 25 mm conduit clamps (12), metallic angled mounting bracket (13), drywall anchors (14), 19 mm machined screws (15), machined screw nut (16), 19 mm wood screws 19 mm. This picture was provided with permission from B. Jollimore at www.environmentalenhancements.com.

Fig 2. Filters connected to domestic appliance and 200 L polyethylene drum.

Fig 3. Physical structure of garments used in experiments.

Polyester t-shirts were composed of a polyester staple fibre with a hexagonal cross-section arranged at a density of 18 rows and 25 columns per cm fabric with a mock eyelet double jersey knit. Cotton t-shirts were composed of a staple fibre with a “bean” shaped cross-section arranged at a density of 15 rows and 20 columns per cm fabric with a single jersey knit.

Fig 4. The metallic cores of the filter with 2 mm (a) and 150 μm sized pores (b).

Fig 5. Emissions of debris from washing cotton and polyester t-shirts in appliances with and without filters.

Significantly different groups from ANOVA and SNK tests are indicated at P<0.05* and P<0.01**.

Fig 6. Emissions of water from washing cotton and polyester t-shirts.

Significantly different groups from ANOVA and SNK tests are indicated at P<0.01**.