Introduction
Like many countries, the UK is currently mandating the wearing of face masks by the general public in various settings, including on public transport and in shops. This policy was implemented due to a growing body of evidence that suggested that even basic face masks can be effective in reducing the spread of the virus, by reducing the range and volume of exhaled droplets containing severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) [1]. The public health benefit of masks is based on strong epidemiological evidence [2,3]; some countries that have successfully brought the pandemic under control include China and Italy [3,4]. Hence, the World Health Organization (WHO) currently recommends this public health intervention [5].
For the general public, the UK government is advising the wearing of reusable face coverings as an alternative to the single-use face masks used in the healthcare sector. The initial reasoning for this was to ensure a sufficient supply of medical-grade masks for healthcare and frontline workers and to avoid unnecessary plastic waste. The public health basis for this decision was that, although not commonly medically certified and mostly homemade, these reusable masks are effective in combating virus transmission when combined with non-clinical interventions such as social distancing and hand washing [1,6].
In this paper, we seek to clarify some of the environmental, behavioural and economic issues around the effectiveness of daily mask wearing by millions of citizens. Specifically, we explore whether single-use masks or reusable masks would be preferable as government policy if supply issues were not a problem. We start with the anatomy of different types of masks, their material composition and effectiveness in protecting the wearer from airborne viruses. Next, we investigate the behavioural factors that impact mask use by the public and community waste management requirements. We calculate the plastic waste and environmental impact of both single-use and reusable masks and carry out a cost comparison of them both. We then discuss how these results affect the different policy options open to governments. Finally, we emphasise the need for further research to clarify the assumptions and unknowns identified in our analysis.
European Standards, standalone effectiveness and anatomy
In the UK, face masks intended for use in the healthcare sector are regarded as personal protective equipment (PPE) and, under the EU Regulation 2016/425 on Personal Protective Equipment (the PPE Regulation), must meet the essential health and safety requirements set out in European Standards (ENs). There are two main Standards that apply to face masks. They are EN 14683:2019 ‘Medical face masks – requirements and test methods’, and EN 149:2001 ‘Respiratory protective devices – filtering half masks to protect against particles – requirements, testing, marking’. Depending on the standard, a face mask can be certified as a medical device and/or a PPE device, respectively. Masks employed in the healthcare sector meet either (or both) of the European Standards and it is conventionally single-use masks that are designed to meet the set requirements and have gained CE marking. Hence, the healthcare sector has built itself around using and managing these single-use devices, where they are treated as clinical waste and incinerated post-use. In June 2020, the European Committee for Standardisation (CEN) published CWA 17553:2020 ‘Community face coverings – Guide minimum requirements, methods of testing and use’, providing guidance on commercially available single-use and reusable masks designed for public use.
Both single-use and reusable masks have varying effectiveness in preventing human-to-human transmission of viruses. The nominal definition of a mask’s effectiveness is by its filtration efficiency – the ability to prevent particular aerosols, bacterial droplets and/or particles from penetrating the imminent atmosphere through to the mask wearer. This is their primary role amongst fulfilling other performance criteria as set out by various European Standards and the more recently published CEN Workshop Agreement on face masks, outlined above. These performance criteria as a whole aim to protect the user and not to protect others. But, as coughing and sneezing expels droplets and aerosols at high pressure, the general consensus is that face masks also protect others by capturing some, if not all, of the droplets and aerosols produced by the wearer. Penetration of droplets through the filter material, exhalation valves and around the face depends on the fit or seal of the mask [7]. It is therefore assumed that filtration efficiency is reflective of the mask’s efficacy in capturing droplets exerted by the mask wearer.
Bacteria and particle size used to test filtration efficiency vary among the standards. EN 149:2001 and EN 143:2000 require testing using the smallest particle sizes ranging from 0.2 to 2 μm (a median size of 0.6 μm), EN 14683:2019 requires the use of Staphylococcus aureus, a bacteria sized approximately 1 μm and CWA 17553:2020 recommends the use of 3 μm particles (the largest size). In addition, EN 149:2001, EN 140:1999 and EN 1827:1999 require adequate facial sealing and set maximum inward leakage as performance criteria. As virus particles are smaller (<200 nm) than the bacteria and particles used to test masks and filters, the classified filtration efficiencies do not convey actual efficacy against viruses.
Surgical masks (Type II and IIR) and single-use respirators (FFP2 and FFP3) are recommended for use within the healthcare sector, particularly for healthcare professionals working in environments with high risk for airborne bacteria and viruses. According to EN 14683:2019, Type I medical face masks are only recommended for use by ‘patients and other persons to reduce the risk of spread of infections particularly in epidemic or pandemic situations’ [8]. As EN 149:2001 classifies masks as PPE intended for preventing the inhalation of fine particles such as dust and pollution, the standard does not state specific recommendations concerning airborne infections. Based on filtration efficiency, masks and filters (within appropriate facepieces) classified at the highest level, under EN 149:2001 (FFP3) and EN 143:2000 (P3), respectively, provide the most protection currently available.
Unless commercially marketed for specific use, other masks are not required to comply with specific standards, and the general public can opt to use homemade or commercial reusable cloth masks that are not certified. Being uncertified means that the filtration efficiency is not confirmed. However, there are studies conveying the efficiencies of materials commonly used for making cloth masks [9,10]. As scientific evidence has emerged on the efficacies of cloth masks [10–14] and later guidelines from CWA 17553:2020, the public has gained more trust in reusable cloth masks (community face coverings and homemade/uncertified cloth masks). In Table 1 we have grouped the masks available to the general public into eight types: surgical masks, non-reusable (NR) respirators, community face coverings, reusable (R) respirators, facepieces (with replaceable filters) and homemade/uncertified cloth masks. For each mask type, we have provided the typical anatomy, the EU Standard and filtration efficiency (if applicable) and intended use. For uncertified masks, literature filtration efficiencies are quoted.
A summary of the anatomy, filtration efficiency (bacterial and/or particle) and intended use information of the masks available for the general public.
| Mask – referred name | Single-use (S) or reusable (R) | Typical anatomy/material | Certification standard | Performance | Testing notes | Use information |
|---|---|---|---|---|---|---|
Surgical mask | S | Multilayer of non-woven fabric typically pleated, with nose wire and ear loops Example: Layer 1 – PP non-woven spunbond Layer 2 – PP non-woven melt blown Layer 3 – PP non-woven spunbond | EN14683:2019** | BFE: Type I: >95% Type II/IIR: >98% (Type IIR is additionally splash resistant) | The bacteria size used for testing is approximately 1 μm [8] | Both Type II and IIR are used within healthcare settings. Type I is only recommended for public use to reduce spread of pandemic. |
Respirator – NR | S | Multilayer of non-woven fabric typically cut to give better fit around the face, with nose wire/clip and ear loops/head straps. Can include nose foam cushioning, rubber facial seal and exhalation valve for better fit and breathability Example: Layer 1 – PP non-woven spunbond Layer 2 – cotton or PP non-woven melt blown (optional) Layer 3 – PP non-woven melt blown Layer 4 – PP non-woven melt blown Layer 5 – PP non-woven spunbond | EN149:2001** | Particle filtration efficiency/max. inward leakage: FFP1: >80%/22% FFP2: >94%/8% FFP3: >99%/2% | The particle size used for testing ranges 0.2–2 μm, with a median size of 0.6 μm [18] | FFP3 are primarily recommended within healthcare, during the COVID-19 ‘peak’ period, UK government changed policies to allow FFP2 to be used when FFP3 are not available |
Community face coverings | R | Bi-layer of the same material can be pleated or cut to give better fit around the face, with ear loops/head straps. Can include nose wire Materials can vary depending on intended reuse and filtration efficiency Examples: Polyester Bamboo Cotton |
EN14683:2019
and/or CWA 17553:2020 | BFE: Type I: >95% Particle filtration efficiency: Level 1: >70% Level 2: >90% | The bacteria size used for testing is approximately 1 μm [8] The particle size used for testing is approximately 3 μm [23] Performance must be tested and maintained through to the intended number of reuses (min. 5) with 60°C washes between uses [23] | Both masks types are typically labelled with the number of reuses associated with the washing temperature Since the publication of CWA 175552:2020, the nominal washing temperature for community masks is 60°C Masks are quoted to be reusable up to 50 times (polyester) |
Respirator – R | R | Multilayer like the non-reusable version but with extra structural support (layer 2), which also gives better fit. Product involves nose clip, and ear loops/head straps. Likely to include rubber facial seal and nose foam cushioning. Can include exhalation valve. Newer masks look similar to cloth masks made of polyester or cotton outer layer with inbuilt filtering materials to increase reusability Example: Layer 1 – PP non-woven spunbond Layer 2 – polymer mix (PE-PP-PET) or cotton (sturdy support layer) Layer 3 – PP non-woven melt blown Layer 4 – PP non-woven melt blown Layer 5 – PP non-woven spunbond | EN149:2001 | Particle filtration efficiency/max. inward leakage: FFP1: >80%/22% FFP2: >94%/8% FFP3: >99%/2% | The particle size used for testing ranges 0.2–2 μm, with a median size of 0.6 μm [5] | Although labelled as reusable, they often say reused up to two or three shifts (8 hour shift) [24]. Newer masks on the market can state up to 6 months is possible [25] Specified by the standard, manufacturer must state the cleaning protocol of the mask. Most recommend wipe cleaning (with suitable disinfectants) of the surface layers between shifts |
Facepiece with designated filters | Mask: R | Mask: Typically a rigid shell/casing designed to fit as a half mask or a quarter mask made of PP, PE and/or silicone rubber. Both mask types requires exhalation valves while EN140:1999 masks are designed with an inhalation valve as well. Product involves head straps and likely requires sealant material where the facepiece meets the face. Can including cushioning for comfort |
Mask:
EN140:1999 EN1827:1999 |
Mask:
Max. inward leakage: 2% | Inward leakage for both masks measures leakage through the face seal [20,21] | For facepieces, manufacturers mostly recommend wipe cleaning (with suitable disinfectants) between use |
 | Filter: R/S | Filter: Filters are made of multilayers of non-woven spunbond and melt blown fabric. Can be either encapsulated or unencapsulated; this is dependent on the mask they are designed for Example: Layer 1 – PP non-woven spunbond Layer 2 – PP non-woven melt blown Layer 3 – PP non-woven melt blown Layer 4 – PP non-woven melt blown Layer 5 – PP non-woven spunbond |
Filter:
EN 143:2000<thinsp> |
Filter:
Particle filtration efficiency P1: >80% P2: >94% P3: >99% | The particle size used for testing ranges 0.2–2 μm, with a median size of 0.6 μm [19] | For filters, they are labelled like EN149:2001 respirators where the number of shifts they can be used for is specified. Often stated to be two or three shifts |
| Homemade/uncertified cloth masks (may have filter pocket)  | Mask: R |
Mask:
Typically two to three layers of fabric with ear loop. Can be pleated or cut to give better fit around the face. Can be made with nose wire. Can be made with a filter pocket for the insertion of filters Most common material: Cotton Most common items used: t-shirt Pillowcases | N/A |
Mask (without filter)
BFE 100% cotton t-shirt: 69.4% Cotton mix: 74.6% Pillowcase: 61.3% Tea towel: 83.2% Vacuum cleaner bag: 94.4% Linen: 60.0% Silk: 58.0% Particle filtration efficiency: Cotton quilt: 96.1% Cotton, 80 TPI, 2 layers: 49.0% Cotton, 600 TPI, 2 layers: 99.5% Chiffon, two layers: 90.0% Silk, two layers: 65.0% Silk, four layers: 88.0% | Results for tests using bacteria sized 0.95–1.25 µm are quoted. Two layers of fabric was assumed BFE for a surgical mask used in the literature study was 96.4% [9] Results for tests using particle sizes >0.3 µm are quoted. Particle filtration efficiency for a surgical mask (fitted correctly) used in the literature study was 99.6% [10] In bold are efficiencies of the most common materials. Note that the studies does not convey whether filtration efficiency are maintained after washes | Literature states that normal laundry processes are adequate in cleaning masks The number of reuses are not known due to limited studies. Clothing is assumed to withstand 20–50 washes [26] |
| Filter: S |
Filter example: PM2.5 filter
Filters are made of multilayers of non-woven spunbond and melt blown fabric. Example: Layer 1 – PP non-woven spunbond Layer 2 – PP non-woven melt blown Layer 3 – carbon-activated filter Layer 4 – PP non-woven melt blown Layer 5 – PP non-woven spunbond | ISO 168901:2016 |
PM2.5 filters
>50% | Filters claiming PM2.5 means they are sufficient in capturing particles sizes between 0.3 and 2.5 µm [22] | Filter inserts are optional |
**Note that not all surgical masks and respirators available for purchase are CE marked but may be certified according to similar standards established by other countries. Commonly referenced standards include the 42CFR84 US standard, where masks are classified to N95 (95% filter efficiency), N99 (99% filter efficiency) and N100 (99.97% efficiency) [27]. These masks are not resistant to oil unlike FFP masks. Similarly, GB2626 Chinese standard classify respirators to KN95, KN99 and KN100, which have the same filtration efficiency as the US standard [28]. N95/KN95 and N99/KN99 are deemed close equivalents to FFP2 and FFP3, respectively [29]. The modified UK government policies on face mask used within the healthcare sector allow non-CE marked single-use medical masks and respirators to be used provided that documentations of their performance meet UK and EU standards. This meant that masks certified, or in the process of getting certified, under EU and/or other countries’ standards are allowed to be used once approved by the HSE [29,30].
Filtration efficiency as required under the standard they comply to are quoted, except for homemade and uncertified commercial masks. Values from literature studies testing filtration efficiency of common fabrics following similar protocols (particularly using similar bacteria and/or particle size) as EU standards are used [8–10,18–22]. See Appendix 1 for a summary of each certification standard.
BFE: bacterial filtration efficiency.
Although current standards do not require masks to be tested against viruses and virus-like particles, there are studies on evaluating the effectiveness of mask materials against them. Davies et al., Lustig et al. and Konda et al. show that common mask materials (including the surgical N95 mask) have a lower efficiency (∼5–25% reduction) in filtering viral-sized aerosols (<300 nm) compared to filtering >300 nm aerosols [9–11]. However, other studies have shown that both surgical and N95 masks are effective in preventing the transmission of common cold coronaviruses and influenza viruses from symptomatic individuals [15,16]. There is also research to suggest that simple homemade cloth masks (two-layer cotton masks) are able to limit the spread of droplets from the wearer [17]; this is evidence that cloth masks could be used to aid the prevention of transmission in public [9].
Materials and their filtration efficiency – considerations for reusable cloth masks
The materials used to construct the filtering component of masks will determine filtration efficiency (Fig. 1). For the majority of masks, the main body of the mask acts as the filter. The exception to this is facepieces where non-porous materials form an added covering and filter. Some reusable masks also have pockets for filter inserts, which will improve their filtration efficiency. For masks without separate filters, filtration is achieved by generally layering different fabrics, or doubling (tripling) or pleating the same fabric.
Example of filter materials in their order of filtration efficiency. Typical material filtration efficiencies increase from left to right.
Synthetic non-woven fabrics are the most common category of materials used to achieve high filtration efficiency in certified masks. These masks or filters are usually made of polypropylene (PP) but some also combine PP with polyethylene (PE) or polyester (PET) (Table 1). In the manufacturing process, heat extrusion converts the polymer into submicron diameter fibres that are collected onto a rotating belt to generate a randomly laid non-woven web [31,32]. Additional processes are used to produce webs with different structures and properties. In a spunbond process, fibres bond with each other as they cool, while in a melt-blown process, high-velocity hot air is blown on the extruded fibre to obtain ultra-fine sub-micron filaments [31–33]. The resulting melt-blown web structure typically has a smaller pore size and provides for better filtration efficiency than the spunbonded web [33]. Even so, the final filtration level will depend on the combination and specific properties of the non-woven fabric used to form the mask and/or filter component. Synthetic non-woven fabrics are often thinner and lighter in weight than other non-woven, woven and knitted fabrics. This is an additional benefit for comfort but affects durability, which is why masks made of synthetic non-woven fabrics are often single use [34,35].
Examples of the types of materials used to manufacture reusable community face coverings and homemade/uncertified cloth masks along with their filtration efficiency [9,10] are presented in Table 1. Homemade cloth masks are generally made of cotton fabric. Many DIY mask designs advise using materials such as old t-shirts [36]. Two-layer cloth masks made from 100% cotton t-shirt fabric have lower bacteria and particle filtration than the surgical masks and N95 (FFP2 equivalent) masks [9–11]. In addition to poorer filtration capacities, two-layered cotton cloth masks have a higher moisture retention than surgical masks [14]. It is recommended that single-use masks are discarded once moist, which is after up to 6 hours of use [37]. This raises the concern that cloth masks may increase the risk of infection as bacteria and viruses may be absorbed and penetrate through the mask [14]. It should be recommended that these are changed more frequently to maintain effective hygiene.
Although most reusable cloth masks have lower filtration efficiencies than N95 masks, there are some now available with filter inserts. The PM2.5 air filters (ISO 168901:2016) advertised as mask inserts comply with the ISO standard and are efficient in capturing particles with dimensions between 0.3 and 2.5 μm with >50% efficiency [22]. However, filtration efficiency is not the only important factor. A recent study suggests that inserts can decrease or improve mask performance depending on their pliability [38]. Overall, the filtration effectiveness of cloth masks depends on the fit, fineness of the cloth and the number of layers [10,11].
Another issue for reusable masks is durability. Extended use and re-use of masks means that fabric will degrade with washing, which will reduce its filtration capabilities over time. THEYA Healthcare, supplier of bamboo masks, and Isko Vital™ (Eindhoven, Netherlands), supplier of cotton masks, state that their products are washable up to 25 and 30 times, respectively [39,40]. Synthetic fabrics, such as polyesters, are widely used for community face coverings available on the market. Synthetic materials are typically water resistant, dry quicker than cotton and are more durable. Suppliers such as Decathalon (Villeneuve d’Ascq, France), Maask (London, UK) and Sera Supplies (St Albans, UK) state that their masks can be washed up to 50 times [41–43]. Maask states that their masks are CWA 17553 Level 2 compliant for up to 10 washes, and change in compliance to Level 1 for 40 washes thereafter [42]. Although masks that are made of polyester are more durable, their degradation through extended use and washing will create microfibres and microplastics. An option to reduce microplastic production would be the use of wash bags that encapsulate items in a washing machine and prevent microfibres from being discharged into the waste water [44].
Table 2 summarises the filter efficiencies of viruses and viral nanoparticles obtained by Davies et al. and Konda et al. [9,10]. The table shows that cloth masks, depending on the material combination and layers, have the potential to be more effective than the conventional surgical masks and respirators. It is acknowledged that the level of protection masks offer against viruses depends on multiple factors such as the appropriate usage and fit of the mask, level of exposure, compliance, complementary interventions (such as hand washing), and early use [14,45], amongst the type and the material of construction of the mask. Greenhalgh et al. argue that, despite the uncertainty about the efficacy of the public using cloth masks, it is better to mandate them on the grounds of the precautionary principle [46]. They acknowledge that the ability of the mask to filter viruses is only one aspect of mask efficacy, and that public behaviour around mask use is equally important. This is what we consider next.
| Material | Bacteriophage MS2 (23 nm) | Nanoparticles (<300 nm) |
|---|---|---|
| Source: Davies et al. [9] | Source: Konda et al. [10] | |
| Filtration efficiency (%) | ||
| Surgical mask | 89.5 | 76 |
| N95 (FFP2 equivalent) | – | 87 |
| Cotton | ||
| 100% cotton t-shirt, two layers | 50.9 | – |
| 80 TPI, two layers | – | 38 |
| 600 TPI, two layers | – | 62 |
| Cotton quilt, one layer | – | 96 |
| Cotton mix | ||
| Unknown, two layers | 70.2 | – |
| Cotton 600 TPI × 1/silk × 2 | – | 94 |
| Cotton 600 TPI × 1/chiffon × 2 | – | 97 |
| Cotton 600 TPI × 1/flannel × 1 | – | 95 |
| Silk | ||
| Two layers | 58 | 65 |
| Four layers | – | 86 |
| Chiffon, two layers | 83 | |
| Flannel, one layer | 57 | |
| Other household materials, two layers | ||
| Pillowcase | 57.1–68.9 | – |
| Scarf | 48.9 | – |
| Tea towel | 72.5 | – |
| Vacuum cleaner bag, two layers | 86 | – |
Behavioural considerations of single-use and reusable masks
Public behaviour towards mask use is a major contributing factor towards the overall effectiveness of masks in slowing the spread of infection [6]. Masks may be able to prevent transmission of a disease by acting as a physical barrier to aerosolised pathogens. However, correct mask use, in appropriate settings, and compliance to hygiene protocols are known to be important to prevent infection (see Appendix 2 for other recommended behaviours to prevent bacterial and viral transmission). It is hygiene protocols in particular that differ between reusable and single-use masks and the ease of adherence to the protocol by the public may make one option more effective than the other from a public health perspective.
The general hygiene guidance for mask use, which includes hand washing before applying and after removing the mask, is similar [23]. Differences lie with the post-use management of each mask type. Single-use masks are regarded as contaminated waste and, in a hospital environment, there are dedicated bins for their disposal, which typically leads to incineration. These dedicated bins do not generally exist outside a hospital environment and so it is recommended that used single-use masks are separated from general waste and double-bagged. The cleaning of reusable masks poses extra hazards in terms of contamination and hygiene. The International Scientific Forum on Home Hygiene published a report on the infection risks associated with contaminated clothing [47], which states that laundering processes eliminate the contamination from fabrics. The NHS recommends that potentially contaminated clothes should be washed at 60°C with a bleach-based product [48]. MacIntyre et al. [91] performed a randomised control trial of hospital health workers in Vietnam who both hand-washed and machine-laundered reusable masks. Their research showed that hand washing reusable masks doubled the risk of contracting seasonal respiratory viruses compared to washing the masks in the hospital laundry [49]. They also showed that properly washed cloth masks were as effective as a surgical mask. Hence, the evidence at present indicates that effective washing is important if reusable masks are to protect against viral transmission [48,49]. Given that most mask washing happens in the home, then encouraging good habits for daily machine washing of masks is important.
It would be ideal to draw on targeted behavioural analyses with regards to reusable masks and intervention evaluations amongst the target population (i.e., the general public). Unfortunately, there has been little work done on this in the literature. In the absence of high-quality evidence, we can use theories, models and frameworks from the behavioural sciences to guide analyses and extrapolate from findings across other populations and related behaviours. One such model is the COM-B model of behaviour [50,51], which posits capability, opportunity and motivation as driving influences of any given behaviour, including the effective and regular washing of masks. Capability involves both the psychological (e.g., the knowledge and importance of wearing masks) as well as the physical aspects of capability (e.g., understanding how to properly put on and wash masks). Opportunity involves both social (e.g., norms) and physical (e.g., access to masks or washing facilities), and motivation involves both ‘reflective’ (e.g., making sure the masks are washed and dried regularly) and ‘automatic’ (e.g., getting into the habit of wearing and washing masks) processes that energises behaviour. To conduct the present behavioural comparison of single-use and reusable mask use, potential behavioural influences will be summarised and compared within COM-B domains.
Capability
Forgetfulness is a common barrier related to adherence to the behaviours associated with health (e.g., medication adherence [52,53]) and sustainability (e.g., using reusable bags when shopping [54]). Forgetting to take a mask when leaving the house is a likely barrier to mask use whether they are single-use or reusable. However, unlike the single-use plastic shopping bag scenario, single-use face masks are not readily available when one has forgotten to take one’s reusable mask. Therefore, forgetfulness is not likely to impact the use of one type of mask over and above the other. However, forgetfulness with regards to cleaning reusable masks can impact on the mask’s effectiveness under hygiene considerations. Establishing routines and habits have been identified as key behaviour change principles for reducing the spread of the viruses amongst the general public [55] and are also likely to help in combatting forgetfulness.
Opportunity
Sociocultural paradigms can impact adherence to mask use. For instance, the contrast between mask use as a hygienic practice (i.e., in many Asian countries) versus a practice reserved for the sick (i.e., in European and North American countries) has been identified [56]. One study with German participants demonstrated that mandating, rather than encouraging mask use, caused a higher level of compliance, with other protective behaviour correlated positively [57]. It stated further that voluntary policy caused lower compliance of mask use and could intensify social stigma between people with and without masks [57]. An Italian study suggested that wearing masks causes people to comply with social distancing rules [58], adding to a growing body of evidence of this effect [59], although there are no studies yet of the differences in these behaviours that correlate with single-use or reusable masks.
The high level of compliance shown in Asian countries without mask laws may suggest that the due diligence of using a mask is dependent on cultural or social norms of individual countries. It is observed that countries that have high compliance of mask wearing by the general public also have a near-history of other epidemics that required mask wearing [1]. A study investigating influences on mask use in Japan found social norms for mask use and conformity to social norms as powerful influences on mask use behaviour [60]. Although it is unknown how transferable these findings may be to contexts outside of Asia, creating strong social norms (i.e., sense of community duty) has been identified as a key principle for enabling behaviours that will delay the spread of coronavirus (COVID-19) [61]. This can be achieved, for instance, through media and professional advice about protective action.
Availability of masks is another key consideration with respect to opportunity. Due to PPE shortages during the start of the COVID-19 health pandemic, improper use of single-use face masks was widely observed, such as not changing and reusing them, thus jeopardising their protective effect [56]. Although disinfection and reuse of single-use masks has been studied, it is not yet known how many times this can be performed before the masks become ineffective. Shortages of reusable masks are also an issue, as is the access to machine-washing facilities. No studies have yet been carried out to compare these opportunity effects on mask wearing.
Motivation
Evidence related to other single-use and reusable hygiene products suggest that single-use may be preferable for some individuals due to the convenience it offers after use. This effect has been found with respect to menstruation products (e.g., disposable pads vs. reusable cloth pads with an antimicrobial top layer) [62]. As menstruation remains a highly taboo subject across many cultures, the generalisability of these results to the present context is questionable. However, we speculate that the considerable additional effort required from users to wash reusable masks may act as a barrier to desired adherence to mask use. Making the desired behaviour easy has been identified as a key behavioural principle for slowing the spread of COVID-19; the less effort it is to adopt a new behaviour, the more likely it is that people will do it [61].
Feeling relief from anxiety by wearing masks has also been found to promote mask use [60], although it is not clear from these results whether this effect is moderated by the type of mask worn. Perceptions of hygiene and safety may influence adherence towards use of a particular type of mask. For instance, the onset of the COVID-19 global health pandemic saw a surge in demand for food packaged in single-use packaging [63], bans on reusable cups across major café chains [64] and increased lobbying to lift bans on single-use plastic bags [65]. This suggests that an association between single-use and safety/hygiene may have entered the public psyche.
Besides hygiene and safety, it is acknowledged that environmental and cost considerations can provide motivation on which type of mask is used. The current technical guidance requires single-use masks to be discarded after one use while reusable masks are washed for reuse up to 50 times. Variability in how reusable masks are used includes the number of masks that are used in rotation; this may be influenced by their cost and subsequently influence the washing protocol followed by the user. In the scenarios below, we compare reuse models of two and four masks, and also compare machine washing with manual washing (see Table 3). Another variable is the use of additional filters within reusable masks, which is influenced by the desire to increase mask protective performance.
Summary of five UK-wide face mask adoption scenarios considered in the comparative study.
| Scenario number | Mask type | Mask use per day | Reuse model | Mask treatment | Number of masks per person per year | Addition of filters | Number of filters per person per year | Filter treatment |
|---|---|---|---|---|---|---|---|---|
| 1 | Single-use surgical mask | 1 | N/A | Disposed at the end of the day | 365 | No | 0 | N/A |
| 2 | Reusable cotton mask | 1 | Two masks in rotation | Manual washing, 50 washes | 7 | No | 0 | N/A |
| 3 | Reusable cotton mask | 1 | Two masks in rotation | Manual washing, 50 washes | 7 | Yes | 365 | Disposed at the end of day |
| 4 | Reusable cotton mask | 1 | Four masks in rotation | Machine washing, 30 washes | 12 | No | 0 | N/A |
| 5 | Reusable cotton mask | 1 | Four masks in rotation | Machine washing, 30 washes | 12 | Yes | 365 | Disposed at the end of day |
We quantify the drivers of these behavioural considerations in the next sections by developing five scenarios of regular mask wearing, washing or disposal. We calculate the environmental impact and the financial costs associated with each behavioural scenario.
Environmental assessment of face masks
Material flow analysis (MFA) and life cycle assessment (LCA) were carried out to explore the potential environmental impact of the whole UK population using either single-use masks or reusable cotton face masks for 1 year. Five UK-wide face mask adoption scenarios were modelled assuming that every person requires one mask per day (Table 3). The number of masks in use per individual per day in the UK during a pandemic depends on an individual’s responsibilities and lifestyle. Public Health England (PHE), the WHO and others, recommend surgical masks are used for maximum of 6 hours or to be discarded if found to be moist or wet (2–6 h) [66,67]. We acknowledge that the use of one mask per individual per day is likely not reflective of actual mask use. For instance, those who are exempt or shielding may not use masks and children may not wear masks as frequently as adults. However, we assumed that people who use masks more frequently (i.e., more than one per day) would balance out lower-frequency mask users. Furthermore, the average number of masks required per person per day should not affect the environmental ranking in the scenarios.
Face masks for each scenario were modelled using the cradle-to-grave approach and the functional unit (FU) employed is 1 year of mask use by the UK population (67.8 million people) [68]. Reusable masks were modelled as made of cotton and used in rotation: if an individual has two masks, it was assumed that only manual washing was possible due to the necessary frequency of washes. With four masks, it was assumed that they could be bulk washed with normal laundry. It was assumed that masks can withstand 30 washes in the washing machine [40] and 50 washes by hand washing. All other assumptions can be found in Appendix 3.
The MFA results show that the use of reusable masks significantly reduces the amount of waste entering general waste streams (Table 4). Including the mask packaging, the total waste accumulation from using single-use masks nationally amounts to 124,000 tonnes with 66,000 tonnes of contaminated plastic. Even if single-use filters are used in addition with reusable masks, the amount of waste is >50% less than using single-use masks. There is >85% reduction in waste if reusable masks without additional filters are used.
Waste arising (thousand tonnes per FU) due to face mask use in the UK for 1 year.
| S1 – single-use masks | S2 – reusable masks, manually washed, w/o filter | S3 – reusable masks, manually washed, w/ filters | S4 – reusable masks, machine washed, w/o filter | S5 – reusable masks, machine washed, w/ filters | |
|---|---|---|---|---|---|
| Waste arising per FU (kt) | |||||
| Masks | 66.2 | 6.83 | 6.83 | 11.7 | 11.7 |
| Filters | – | – | 29.5 | – | 29.5 |
| Packaging | 57.4 | 2.38 | 17.3 | 4.08 | 19.0 |
| Total | 123.6 | 9.21 | 53.63 | 15.78 | 60.2 |
FU; functional unit; w/o: without; w/: with.
Used (and potentially contaminated) face masks are considered medical waste and directed to incineration when they arise from a clinical setting. In the UK, there are currently 68 incinerators with a combined capacity of 12.2 million tonnes of waste [69]. In 2018, a total of 10.9 million tonnes of waste were processed [69], thus, on a national level, there is waste capacity to process the 124,000 tonnes of single-use mask waste. Currently, no specific incineration waste stream is accessible by the general public. At the household level, waste PPE is placed in mixed general waste, which may put waste collectors at risk of contracting infections. The Association of Cities and Regions for Sustainable Resource Management has advised keeping contaminated waste in a double bag for 72 hours before disposing into general waste. However, it is not clear how this would be monitored to prevent the risk to waste disposal workers. There may also be storage issues, both in households and at waste treatment sites, as the total waste increases.
Life cycle impact assessment (LCIA) results were generated using the environmental footprint (EF) 3.0 methodology [70] (see Fig. 2). It showed that Scenario 4, in which four masks are used in rotation without single-use filters and are machine-washed, has the lowest environmental impact in all of the impact categories analysed except water scarcity. Net impact results (Fig. 2 and Appendix 3) also show that having a higher number of masks in rotation to allow for machine washing (Scenarios 4 and 5) is more environmentally beneficial than manual washing (Scenarios 2 and 3).
Total annual cost (£) for the supply of single-use and reusable face masks per UK citizen and the percentage breakdown.
Hot-spot analysis carried out on each scenario’s impact on Climate Change indicated that, for single-use masks, Mask Transport to the UK (from China – the assumed location for the production of all masks) contributes most to this impact category (Fig. 3). Due to the higher number of masks needed in Scenario 1, the contribution of Mask Manufacture and their Transportation to the UK is higher for this scenario than in the reusable mask scenarios. This has led to Scenario 1 generating a Climate Change result that is 3.5 times greater than Scenario 4.
In Scenarios 4 and 5, each face mask would be washed 30 times before replacement. However, the masks may not withstand this amount of washing, or they may be misplaced or damaged by other means before their assumed replacement. Further analysis indicates that if the amount of machine washes per year stays constant, up to 48 reusable masks per individual can be used before Scenario 4 exceeds the impact on Climate Change indicated by the single-use masks in Scenario 1. This analysis was carried out on all other environmental impact categories and the average cap on the supply of reusable masks was calculated to be 25 masks per person (see further details in Appendix 3).
The environmental impact of using single-use masks could be lowered by changing the manufacturing location, hence reducing transport emissions. The choice of manufacturing location would depend on the supply chain available. If the UK produced its own masks in order to reduce mask transportation emissions, as it is a major importer of non-woven textiles [71], it would still incur transport emissions from importing raw materials (explored further in Appendix 3). Importing masks from China has been deemed more realistic, owing to their manufacturing capabilities and the supply chains currently in place.
Cost comparison
With a growing number of countries making the wearing of face masks outside of the home compulsory, price increases and limits on supply are to be expected. A cost analysis presented here compares the use of single-use and reusable masks for 1 year. Scenarios 1 and 4 assumed in the environmental impact study were used to form the basis of this cost study. Scenario 1 was chosen because it represents a single-use scenario and Scenario 4 was chosen because it generated the lowest environmental impact in most impact categories analysed (Fig. 3).
The comparison covers the initial cost of the respective product, the washing cost (including detergent and electricity) for the reusable face mask, and the disposal cost of incinerating and landfilling a single-use and a reusable face mask. The water cost was excluded from the calculations as it is typically a fixed cost in the UK (which represents no marginal variations). Transport cost is also excluded due to high variability of freight routes, modes of transport and types of vehicles required to deliver face masks to the UK population. Table 5 presents a summary of the cost comparison.
Summary of cost comparison (£) between a single-use and a reusable face mask, per unit and for an annual supply.
| Concept | Unitary cost (£) | Annual cost (£) | ||
|---|---|---|---|---|
| Single use | Reusable | Single use (365 pcs) | Reusable (12 pcs) | |
| Product | 0.26 | 2.04 | 94.99 | 24.43 |
| Washing | ||||
| Detergent | N/A | 0.000539 | N/A | 0.789 |
| Electricity | N/A | 0.000227 | N/A | 0.332 |
| Disposal | ||||
| Incineration (57%) | 0.000131 | 0.000706 | 0.0480 | 0.00847 |
| Landfill (43%) | 0.000123 | 0.000663 | 0.0450 | 0.00795 |
| Total | 0.26 | 2.04 | 95.00 | 25.57 |
Product
The initial cost of buying any type of face mask represents the highest cost in the face mask life cycle. The unitary cost of both types of face mask were averaged from 15 products, respectively, where the products with higher relevance were selected. The average price for a single-use face mask is £0.26, and the average price for a reusable one is £2.04 [72,73], and thus a single-use face mask is around eight times cheaper. However, the annual product cost considers 365 single-use face masks (£94.99), whereas only 12 reusable masks are required (£24.43), which makes the reusable options about four times cheaper on an annual basis.
Washing
The washing cost is only attributed to a reusable face mask. According to LCA assumptions, 0.162 g (0.156 ml) of soap are required for machine-washing one reusable mask each time. The average cost of liquid detergent (the most commonly used in the UK) is £3.46/l [74], so it will cost £0.79 to wash 12 face masks 122 times, which is the total annual cost of detergent use. Similarly, 1.58 watt-hours (Wh) are needed to wash one reusable mask once (according to LCA assumptions). The average price per kWh in the UK is £0.14 [75], and so the electricity cost of washing 12 masks 122 times results in £0.33.
Disposal
Given that waste PPE is conventionally placed in mixed general waste at a household level, both single-use and reusable face masks are assumed to follow the same end-of-life pathway. LCA assumptions suggest that 57% of face mask waste will be incinerated (with or without energy recovery), and the remaining 43% will be landfilled. In the UK, the average cost per tonne for incineration is £86, and the average cost for landfill (including Landfill Tax) is £107 [76]. The mass of a single-use face mask is assumed at 2.68 g, while the mass for the reusable counterpart is assumed at 14.4 g. Thus, the annual cost of discarding 365 single-use face masks would be £0.09, and the annual cost of disposing of 12 reusable masks would be under £0.02. The disposal cost for the reusable scenario is around 5.7 times cheaper.
The total annual cost (considering product price, washing and disposal) would be £95 for the single-use face mask scenario and £26 for the reusable option (rounding to the nearest pound) (Table 5). From a life cycle perspective, an annual provision of single-use masks for one UK citizen would cost around 3.7 times higher than the annual supply of reusable face masks. It is worth noting that the cost of washing reusable masks (122 washes) over a year is £1.22 and represents only 4% of the total annual cost for the reusable scenario (Fig. 3).
Considering a UK population of 67.8 million [68], the single-use scenario results in an annual cost of £6.4b, while the reusable option would cost £1.7b. If reusable face masks were to be made mandatory (and single-use masks restricted), this would represent annual savings of £4.7 billion for the UK economy. The cost analysis model assumes that most face masks (either disposable or reusable), or the raw materials used to manufacture them, are imported from abroad (e.g., China). Costs of finished products and raw materials are likely to spike due to high demand. As the manufacture of single-use masks typically requires specialised machinery, capital costs to begin their production in the UK would be higher than the local manufacturing of reusable masks. Reusable mask manufacture can be scaled up reasonably easily in the UK, providing a boost to the UK economy without impacting on the supply of single-use masks to the NHS. Countries such as Portugal [77] and France [78] have issued guidance for the manufacture of such masks, including a ‘stamp of quality’ in Portugal based on the Directorate-General of Health guidelines [77].
Discussion and conclusions
Many governments around the world have introduced policies that recommend or mandate the wearing of masks to slow the spread of COVID-19 (Fig. 4). Mandatory use and enforcement varies globally [79]. While many countries have introduced laws to mandate mask use, countries such as China, India, Japan, South Korea and Taiwan have made more limited recommendations [80]. Around the world, policies also differ on whether children, or children of certain age, are exempt from mask wearing; the types of places and settings where masks are mandatory (i.e., all public or selected settings); and the fines that can be incurred for non-compliance [79–83]. Some countries, such as Italy and Mexico, provided single-use masks for the general public upon mandating their use [84]. Japan provided cloth masks without imposing mandatory use [84], while the Czech Republic and the UK advocated for the public to wear reusable masks and provided information on how to make them at home [85]. There is not yet enough data to determine which combination of mask policies is most effective in slowing the spread of infection. However, with billions of people now wearing masks globally, there are potentially high environmental and economic costs.
Mask requirements and usage by countries, obtained from Mask4All, latest update on 24 June 2021 [80]. Green indicates countries where they had countrywide mandatory mask use laws in place or had reached universal usage (mask use by >80% of the population) through recommendations from the government. Yellow indicates countries where mask use was mandated by law in parts of the country and recommended by government in other parts. Red indicates countries where no mask laws and no recommendations were given by the government.
Countries such as the UK campaigned for the use of reusable masks mainly due to shortages of single-use masks in the healthcare sector. However, with better forward planning for the medical PPE supply, mask shortages may not be an issue in future. The analysis and discussion presented here can inform future mask recommendations and policy.
Over the last few decades due to decreased costs, the healthcare sector has adopted single-use masks over reusable ones in the belief they offer higher standards of hygiene both for the use phase and during waste management treatment (incineration). Availability and hygiene were also why the UK general public chose surgical masks and single-use respirators at the beginning of the COVID-19 pandemic.
Certified single-use masks have a higher filtration efficiency and fulfil higher performance criteria than most reusable masks. However, they do require new waste procedures and infrastructure for safe disposal in the form of incineration and have a higher environmental impact than reusable masks. In order to minimise the public health issues associated with the disposal of contaminated waste, and as littering of used surgical masks has been observed, local councils could install special disposal units for contaminated masks in public spaces.
As we have shown, there is growing scientific evidence that reusable masks can be made to satisfy high filtration performance criteria. From a public health perspective, what is less clear is how the level of mask filtration efficiency translates to reduced infection rates at a population level. For instance, there is anecdotal evidence of people reusing single-use masks, which may increase risk of infection, or washing them, which could compromise filtration efficiency. Similarly, if reusable masks are not laundered after every use, or laundered more than the recommended number of times, this too will challenge their filtration efficiency.
The filtration efficiency is not the only factor that determines the impact of mask wearing. The behaviour associated with mask wearing such as donning and doffing the mask, correct cleaning or disposal, and indirect impacts of mask wearing such as how it affects adherence to social distancing and hand washing, all indirectly affect the efficacy of mask wearing. We have noted that there are limited behavioural studies that differentiate and compare the impact of single-use or reusable masks on these behaviours. Extrapolating from related evidence, single-use masks may be chosen by citizens due to perceived increased hygiene and convenience of not having to wash them. By contrast, people may choose reusable masks due to the lower financial costs or the smaller environmental impact when compared with single-use masks. Further behavioural research is needed to fully understand whether our assumptions about these factors are reasonable, and also to identify barriers and enablers to wearing and cleaning reusable masks. For instance, a study comparing the use of single-use and reusable masks with the same level of standalone performance (i.e., equal filtration efficiency) by different communities will be necessary to understand whether the additional process of laundering masks affects the overall compliance and therefore effectiveness of mask use in slowing down the rate of pathogen transmission.
A simple yet plausible study could be carried out by utilising the large amount of data and engaging with participants through symptom monitoring Apps (e.g., ‘COVID Symptom Study app’ by King’s College London [86]). In addition to symptom monitoring, these platforms are also capturing mask use habit data and COVID-19 infection rates. Analysis of such data with follow-up interviews where members of the public answer specific questions on their practices related to mask use would be invaluable. For instance, interview questions could include: which type of masks (if any) they use; what individual (e.g., values, beliefs, attitudes and perceptions), social (e.g., culture and norms) and situational factors (e.g., context and cost) influence their mask use; what situations do they use them in; and, crucially, how often do they wash reusable masks? The findings of such a study would be valuable for understanding how mask-wearing behaviours impact infection rates and thus inform government policy.
A significant outcome of our analysis shows that choosing to wear reusable masks generates 85% less waste, has 3.5 times lower impact on climate change and incurs 3.7 times lower costs than single-use masks. These lower impacts depend on our assumptions, for example, a limited number of reusable masks used per person and the use of washing machines to regularly clean them. Even if these assumptions are lower than the actual number of reusable face masks purchased, for instance, the environmental significance of reusable masks is still clear. From this perspective, there is a strong argument that public policies should encourage wearing reusable masks to reduce our waste and carbon footprint.
From a cost perspective, our analysis demonstrates that there is presently a £70 per person difference between single-use and reusable masks (assuming one person uses one mask per day for a year). To prevent future shortages, governments may consider stockpiling single-use masks or ensuring sufficient local production chains, which would increase this per person cost difference, as storage and distribution costs would be necessary. Upfront costs of reusable masks may also increase as their performance levels increase. Furthermore, the actual number of reuses may be lower than we have assumed due to potential misplacement, unforeseen damage and even visual dislike, resulting in a higher number of masks being necessary. All of this would narrow the cost between single-use and reusable masks. Currently, our assumption of 12 reusable masks per person per year means that the price of these masks could increase up to £7 per mask before it matches the cost of single-use masks. An understanding of the actual number of masks, single-use and reusable, that are necessary to support an individual would inform a better cost comparison and a better environmental impact comparison.
An alternative policy to stockpiling is to promote resilience through local in-country production of masks. For single-use mask manufacture, it would be necessary to consider the potential import of raw materials since the UK does not currently manufacture non-woven fabrics on a large scale. Importing raw materials from Turkey rather than China would generate a lower environmental impact (see Appendix 3). Assessing the impact of producing reusable masks in the UK and/or other countries was not carried out but it is likely that if production is shifted from China to the UK, the overall environmental impact would reduce and remain lower than single-use masks. The UK would want to ensure that such agile manufacturing capacity is capable of scaling up quickly (in a matter of weeks). Hence a hybrid plan that combines stockpiling enough masks for the whole population coupled with a set of financial incentives for garment manufacturers to maintain the capability to retool factories for maskmaking may be the best compromise between cost and resilience.
