Supplement titlebanner ENG


In September 2017, Mount Agung, Bali, Indonesia, started to show signs of unrest, with swarms of deep earthquakes recorded on seismographs, and felt by communities, indicating movement of magma beneath the volcano. A large eruption looked imminent and, by early October, up to 146,000 people were evacuated, spread across 427 shelters [15].

Explosive volcanic eruptions generate ash which can cover many tens to thousands of square kilometers around a volcano. In Bali, the local Volcanology and Geological Hazard Mitigation Office (PVMBG) estimated that if Agung erupted there could be a layer of ash 40 cm (16 in) deep over a 30 km (18 miles) radius.

Beyond the immediate threat of lethal, proximal hazards such as pyroclastic flows and lahars within ~20 km (12 miles) of the vent, potentially hundreds of thousands of people living further afield would have concerns about inhaling ash. This is because ash particles can be small enough to enter the lung, causing irritation in healthy people and exacerbation of symptoms in people with existing respiratory diseases [1]. Ash can also contain potentially-toxic minerals such as crystalline silica, although it is not proven that volcanic silica can cause diseases such as silicosis or lung cancer [1]. In addition, uncertainties about the harm caused by breathing ash can lead to anxiety and psychological distress [16], as well as it being uncomfortable to have in the mouth and nose.

Due to these risks, agencies (governmental and non-governmental organizations with public health, civil protection, disaster management and humanitarian remits) will usually advise that those proximate to ash, especially those with respiratory problems, should stay indoors. Many agencies also distribute facemasks. A summary of advice offered around the world can be found at:

Surgical masks are commonly stockpiled for public health emergencies (e.g., for influenza pandemics) and are inexpensive and easy to procure and store in bulk. In Indonesia, PMI Yogyakarta (Red Cross), regional health (Dinas Kesehatan) and local disaster management (BPBD) governmental agencies distributed over a million surgical and ‘flat-fold/3D’ masks during the eruptions of Merapi (2010) and Kelud (2014), in Java [4]. These masks are not industry-certified, but are marketed as being effective at capturing particles sub-2.5 µm diameter – PM2.5 – and a recently-tested mask in the HIVE project confirmed this high standard of efficiency [2]. Surgical masks, although not designed for such purposes, also have a high capacity for filtration [2, 17]. Like surgical masks, ‘flat-fold/3D’ masks often have poor design which compromises facial fit [3]. Until the HIVE study, there was little-to-no evidence of the effectiveness of these masks for community use against volcanic ash; the new evidence shows that effectiveness of these masks is compromised by poor fit [3; and Figures 1 and 2].

More rarely, agencies have distributed masks to communities during volcanic eruptions which are designed for occupational use in dusty workplaces (e.g., industry-certified N95-type particulate respirators). N95 and other, similar industry-certified respirators (e.g., FFP2 in Europe) have been demonstrated, in laboratory studies, to provide better levels of protection than surgical and other masks [2, 3, 17, 18]. Occasionally, high-income countries have distributed such masks during eruptions, which were already stockpiled for healthcare workers in pandemics (e.g., Eyjafjallajökull in Iceland in 2010).

The advent of the Agung eruptions in 2017-2018 saw a shift in the approach to community protection in a low-to-middle income country (LMIC). Unlike many eruptions, the prolonged period of unrest (which culminated in the start of a magmatic eruption on 25 November 2017, with eruptions continuing intermittently throughout 2018) allowed organizational-level preparedness of public health interventions on a scale and a type not previously observed in Indonesia, or other LMICs. This resulted in a combination of changes in humanitarian practice: greater coordination of grassroots response, the use of crowdfunding platforms and a change in ethical decision-making practices around preparedness, as discussed below. These changes led to the donation, stockpiling and distribution of more than 75,000 N95 respirators, in 2017, across 52 evacuation camps, by the Mount Agung Relief (MAR) group.

The MAR group is a consortium of several local NGOs, including Kopernik (who strive to find effective, affordable solutions to poverty reduction), and concerned community members whose goal was to deliver critical supplies to the thousands of evacuated people. To our knowledge, the distribution of these types of masks had not previously occurred in a volcanic crisis setting in Indonesia or any other LMIC [4].

Coordinated agency action and responses

While coordination is not novel, it was particularly effective during the Mount Agung crisis. This was potentially because of the prolonged period of volcanic unrest, giving time for efforts at the grassroots level to be aligned. In Bali, this resulted in the founding of the MAR group. Their coordinated response allowed pooling of donations (US$ 133,000 by the end of November 2017; to deliver targeted interventions to evacuees (N95 respirators, water filters, toilet and handwashing stations and communications/ education kits on disaster preparedness). The initiative was also promoted through a Facebook page ( allowing simultaneous information dissemination and donations and to leverage support from a broad range of donors for delivery of N95 respirators which, previously, would have been inaccessible due to cost.


This specifically relates to internet-mediated fundraising of large amounts of money through many small public donations. Through web-based platforms, donations to fund specific interventions can be requested. Small NGOs, like Kopernik, harnessed this for the Agung eruption by advertising that they were seeking funding to supply facemasks to protect their socio-economically deprived communities from volcanic ash inhalation. They further stated that, based on emerging evidence [2, 3], they would only supply N95 masks ( because these provide the greatest protection. Previously, due to high cost, these masks were not considered a relevant intervention in Indonesia, but crowdfunding is a way to partially remedy the imbalance in protection offered between high and middle-to-low income countries and between rich and poor in the country concerned. The MAR group received 912 separate donations. Due to the visibility of their crowdfunding platform, Kopernik also received multiple donations of N95 masks from interested individuals and businesses.

Shifts in ethical decision making

In previous eruptions, agencies have often needed to act rapidly, in the absence of information on effective interventions for volcanic ash exposure reduction, or even knowledge of whether the ash may be harmful to inhale. Most agencies have (knowingly or unknowingly) applied the precautionary principle, that if something, such as the ash from a volcanic eruption, raises a possible threat to human health, precautionary measures should be undertaken, even in the face of limited scientific evidence that those precautionary measures will be effective [19]. This is especially so when there is an expectation that agencies will do something visible to help [12].

This approach resulted in the mass distribution of surgical and other masks during earlier eruptions in Java, Indonesia [4]. The decision by the MAR group to supply only N95 masks marks a shift towards awareness of, and use of, scientific evidence to inform their response. The decision to supply N95 masks was, at least in part, based on the HIVE project findings, which showed that N95-style respirators perform best against volcanic ash, and fit well on volunteers, even when no training on fit is provided [2, 3]. Preliminary results of the HIVE study were posted on the IVHHN website in September 2017. These results were widely used by the public during the crisis, according to postings on community Facebook groups, and the MAR group [20]. These responses demonstrated a shift from the precautionary principle towards to the principle of effectiveness [12].

Facemasks are usually designed to only fit adults, because they are meant for use in industrial and healthcare settings. However, during the Agung eruption, the MAR group supplied masks to children. This action was based on a donation of 15,000 masks designed specifically to fit children’s faces, the safety of which has been tested by clinical trial [21]. In most other volcanic crises, such masks would not be available, and adult sized masks would not fit most children’s faces. Agencies have a responsibility to advise on alternative, more-suitable interventions for children, such as keeping them indoors or moving them to a non-ashy area, rather than supplying masks which would not be protective and could even be harmful.

Ethical distribution

Kopernik donations

Figure 7. Fifteen types of industry-certified N95 masks donated to, or purchased by, Kopernik for the Agung crisis.

The decision to only supply N95 masks does raise questions about who masks should be distributed to. The MAR group did not receive sufficient funding or donations to supply masks to all those affected by ash, resulting in a need to make decisions about who should receive priority allocation. There are various ways in which people distribute scarce resources and, often, in a public health context, those judged most vulnerable and/or most exposed (e.g., emergency workers) are prioritized to access these [22], as has been seen during the COVID-19 crisis. The MAR group prioritized people over 65 years old, those with existing respiratory illnesses, pregnant and breastfeeding women, and children, when children’s masks were available.

A further allocative challenge at Agung was that at least fifteen different types of N95 masks were received from donors (Figure 7). While N95 refers to the certified standard of the filtration capacity of the material, some of the masks received are likely to have better fit and greater comfort than others, e.g., those with a valve on the front for humidity reduction in a hot climate. If some masks are more likely to be worn due to comfort factors or are likely to be more effective due to superior fit, then this too raises ethical questions about fair allocation. In future crises, requests for donations of particular mask types (brands and models) would overcome this issue.

The group also received some donations of non-disposable (i.e. reusable) masks made by a company which specializes in comfortable, N95-certified masks, of all sizes (including for children), specifically for community rather than occupational use. Deciding who will receive a reusable mask, and who will not, is also a challenge that can pose difficult ethical dilemmas. Therefore, issues of social justice associated with agencies providing the most effective form of protection against inhalation of volcanic particulates may not be completely overcome if allocative decisions are still being made.
The COVID-19 crisis has resulted in a global shortage of personal protective equipment (PPE) meaning that humanitarian provision of effective community respiratory protection in current or near-future air pollution crises (wild fires, eruptions) will be challenging and the affected communities may need to continue to use the cloth face coverings being employed as infection source control, which the HIVE research showed to be ineffective at filtering ash.

Even if effective protection is available, while NGOs are able to attract financial and other donations, governmental agencies may have to rely on tight budgets and in-kind donations from other humanitarian organizations, making provision of more expensive masks on a large scale less realistic. During the Agung eruption, governmental agencies continued to distribute surgical masks. Justice in the allocation of resources to ensure best protection, for as many people as possible in LMICs, therefore remains an issue.

Effective resource conservation and management

Figure 8

Figure 8. Perished N95-certified masks, previously donated to BPBD, Indonesia by an international NGO. The green ‘line’ towards the lower front of the mask was one of the head straps, but the elasticated material has disintegrated. Every mask had broken head straps.

If expensive masks are donated and stockpiled, it is important that they are suitably conserved. The Yogyakarta provincial disaster management agency (BPBD) in Java, Indonesia was given 13,500 N95 masks by an international humanitarian NGO. The masks were stored in an open warehouse with no temperature or humidity controls and, during a visit, the authors witnessed that the entire stock had perished as the head straps had disintegrated, rendering them unfit for use (Figure 8).

While in Europe mask manufacturers are legally required to print a use by date (usually 5 years) on their certified masks, this is not the case in Australia (where the BPBD masks originated from) making management more complex, even if stored so as to prevent degradation. There is a risk, therefore, that money will be expended on a more expensive protective intervention that will ultimately prove unusable due to degradation or unclear expiry dates.

A new age of effective protection?

Will this effort for the Agung evacuees herald a new age of provision of effective community protection during volcanic crises? In 2018, in Hawaii, a major PPE manufacturer donated over 100,000 N95 particulate respirators during the 2018 Kīlauea crisis (personal communication with manufacturer). As with the MAR group, the donation resulted from the Hawaii Department of Health deciding to recommend particulate respirator use based, in part, on the new HIVE project evidence on effectiveness of these masks for volcanic ash [2, 3]. While such donations are available to all countries (e.g., Guatemala received around 20,000 particulate respirators during the June 2018 Fuego volcanic crisis, through the same scheme), donations are upon request from non-profit organizations, which need to be aware of the potential for support, as well as the benefits of these respirators over the more easily-available surgical masks. This suggests the need for an educational endeavor to make such information known to the relevant bodies.

Even though it seems there is a wish to move towards provision of the most effective protection, economic and logistical factors are likely to continue to sway decision making for some time to come, during most crises and for most organizations, and it is unclear how the COVID-19 crisis will have changed agency and community perceptions of facemask use.

We are, however, also observing changes in respect to other types of air pollution events involving particulate matter. For example, UNICEF Indonesia and Kopernik are working on a household haze emergency kit which will include N95 masks. We hope that greater visibility of such masks within communities, the high profile of the Agung relief effort, and the influence of social media in publicizing such activities may lead to other organizations preparing for eruptions, or other forms of air pollution crises (such as wildfires), by considering, in advance, the ethical issues surrounding recommendation and distribution of facemasks and how they will raise funds or seek donations to procure the intervention of their choice. Better communication and coordination amongst relief organizations (ideally across governmental and non-governmental organizations) will also hopefully lead to pooling of resources and, therefore, decisions on equal provision of interventions across whole populations impacted by volcanic ash and other particulates.

Claire J. Horwell, Fiona McDonald, Ewa J. Wojkowska, Lena Dominelli