A community-led fire safety pilot in which residents of the BM Section A informal settlement in Khayelitsha, Cape Town, partnered with the non-profit Kindling and the Development Action Group (DAG) to assess their own fire risk and then plan and start action to reduce it. It draws on residents’ own knowledge together with fire engineering and social science, is sponsored by the Fire Safety Research Institute (FSRI) of UL Research Institutes, and has been running for well over a year (2026, ongoing).
Map
Community-based fire risk assessment in an informal settlement in Khayelitsha, Cape Town
General Information
ISIG
Residents of an informal settlement in Cape Town were trained as community-based researchers and led a fire risk assessment, then planned and began action to reduce fire risk.
The non-profit Kindling and the Development Action Group (DAG) worked with residents of the BM Section A informal settlement in Khayelitsha, with funding from the Fire Safety Research Institute (FSRI), drawing on residents’ knowledge, fire engineering and social science. Six residents were trained as researchers, and residents mapped the settlement, held workshops with youth, older people and people with disabilities, and helped analyse the results, which according to the funder’s update produced 33 problem statements. Residents then prioritised solutions, and early actions included a clean-up of a vacant plot where waste fires started, a Fire School, educational films and advocacy with city officials. Kindling’s project page reports more than 300 residents engaged and 155 hours of community engagement in the assessment phase, and the funder’s update describes workshops with 60 to 100 residents.
Khayelitsha is a large township on the Cape Flats with about 500,000 residents. A 2025 university study estimates that the BM Section has about 24,000 residents in over 6,000 shacks, next to the Eerste River and a degraded wetland. In that study’s survey of 125 randomly selected residents, 42% had experienced fires directly or indirectly. Kindling’s project sources inform that fires in informal settlements in South Africa happen almost daily, and that the systems and power structures around these settlements were never set up with their safety in mind.
A research roundtable in 2023 with more than 50 stakeholders gave rise to Cape Town’s first Community of Practice on Informal Settlement Fire Safety, which Kindling now convenes. The project source notes the pilot helped bring this group into being.
Hazard Type
Geographical Scope - Nuts
Geographical Scope
Population Size
Population Density
Needs Addressed
- Fires start and spread easily in dense settlements with flammable materials, unsafe cooking, heating and electricity, and illegal dumping.
- Fire service access is limited by narrow routes, lack of water and, at times, conflict between the service and the community.
- Official fire data and formal systems do not capture residents’ experience of fire.
- Residents face competing pressures such as food insecurity, crime and gender-based violence, so fire safety competes with other priorities.
- Residents’ own knowledge and trade-offs in daily life are rarely part of fire risk decisions.
Community-based researchers engaged children, youth, older adults and people with disabilities, and weekend sessions were held so that marginalised residents could take part.
Kindling led the work with DAG and a local advisory committee of community leaders, and residents co-led the analysis and set the priorities. Local government partners (not named in the sources) took part in action such as the clean-up, and the Community of Practice brings together city and provincial government, NGOs, universities, businesses and community representatives. The sources report no change to formal decision-making on fire safety.
A municipal fire service exists, but the case describes problems with communication, access, water supply and trust. Community response (detection, alarms, evacuation and community firefighting) was one of the six risk dimensions residents examined. Residents are now learning the correct emergency number and helping crews find their way through narrow routes.
Some provision exists (a fire service, reached through narrow routes), but access is limited, water is lacking and electrical connections are unsafe. The pilot added no physical infrastructure beyond the clean-up of a vacant plot.
To make residents the researchers and decision makers of their own fire risk work, so that solutions fit daily life.
Six residents trained as community-based researchers; community mapping and "fire memory" discussions; workshops in IsiXhosa and English with youth, older people and people with disabilities; thematic analysis done jointly by residents and Kindling; feasibility, prioritisation and planning workshops; and a decision, proposed by the researchers, to start action during the planning phase instead of only publishing reports.
Residents co-led the analysis, prioritised problems and solutions, ruled out disciplinary measures because of the risk of retaliation, and led the vacant-plot clean-up with local government partners. The community-based researchers asked for implementation to begin alongside planning, and Kindling changed its plan accordingly, with the sponsor’s support. In one later fire, workshop participants stopped neighbours from removing metal wall sheets, which would have helped the fire spread, according to the funder’s update.
Training of community-based researchers, a Fire School for residents, educational films and engagement with city officials.
Vulnerable Groups
Governance
Emergency Preparedness
Infrastructure Readiness
Engagement Level
Empowerment Level
Implementation
- Residents trained as community-based researchers who co-lead data collection and analysis.
- Residents’ knowledge, fire engineering and social science used together in one process.
- Fire risk is described by residents in six dimensions: ignition, spread, community response, fire service response, post-fire recovery and general lack of fire safety.
- “Fire memory” discussions and mapping that show gaps in official fire data.
- Planning and action run in parallel, and the plan changed when researchers asked for it.
- Plans for a Fire Risk Assessment and Fire Risk Reduction toolkit to share the method.
IsiXhosa and English. Sessions were run in IsiXhosa and translated afterwards.
Kindling, a non-profit working on fire safety in informal and humanitarian settlements, with the Development Action Group (DAG), which helped build relationships with community leaders and guided community selection and engagement.
Kindling’s website lists several fire safety projects (Cape Town, Dhaka and Hargeisa, plus fire experiments and modelling work), and the work draws on fire engineering and social science. The sources do not give Kindling’s founding date or track record.
Residents of BM Section A; six community-based researchers; a local advisory committee of community leaders; Kindling; DAG; FSRI (UL Research Institutes) as sponsor; local government partners and City of Cape Town officials (through advocacy and the Community of Practice); the fire service; and the participants of the Cape Town Community of Practice.
- Build relationships and trust with community leaders and residents, with DAG's help, and recruit local residents as researchers and set up an advisory committee.
- Hold recurring inquiry workshops to learn about the community’s pressures and daily routines (the funder reports 60 to 100 residents).
- Run the assessment over six to eight weeks (sources differ): community mapping, workshops, "fire memory" discussions and spatial analysis, with weekend sessions for marginalised residents.
- Analyse recorded discussions together with residents and write problem statements (33 are reported).
- Hold feasibility, prioritisation and planning workshops, in which residents sort ideas by problem and choose actions.
- Act on priorities while planning continues: the vacant-plot clean-up, Fire School, educational films and advocacy with city officials.
- Continue data collection and connect the pilot to the Cape Town Community of Practice and to the planned toolkit.
The Fire Safety Research Institute sponsored the work. No information on the amount or period is given. Other needs are the time of residents and researchers, facilitators, workshop venues, interpretation and fire engineering expertise. The sponsor supported the revised plan and made resources available for implementation activities alongside planning.
The work followed three phases: assessment, planning and implementation (the funder names them inquiry, participatory fire risk assessment and fire risk reduction planning). The assessment took six weeks on the project page and eight weeks in the funder’s update. Overall duration is given as well over a year on the project page and as two years in the funder’s update.
Experience of the Implementing Organisation in DRM
Target Audience
Resources Required
Timeframe & Phases
Participation Results
The main gap in this case was knowing how everyday habits lead to ignition (for example unattended batteries), not awareness of what fire does. Additionally, residents’ daily trade-offs shape fire risk more than any single technology, that iterative action builds trust and momentum, and that technical fixes introduced from outside tend not to last. The plan worked better once action began during planning.
Trust and power -> formal institutions were not built with informal settlements in mind, so the team started with relationship building and kept the process non-prescriptive, as DAG’s community facilitator coordinator describes it.
Competing priorities -> residents face food insecurity, crime and gender-based violence, so sessions were tied to daily routines and decisions.
Access to marginalised residents-> weekend sessions were added and community-based researchers reached children, older people and people with disabilities.
Language -> sessions were in IsiXhosa and translated afterwards.
Project scope -> Kindling’s original plan ended with a report and advice, but researchers asked for action, so Kindling stayed on to help put plans into practice.
Safety -> disciplinary solutions were rejected because of the risk of retaliation.
Risk: retaliation against neighbours if rules were enforced. Mitigation: disciplinary measures were ruled out.
Risk: outside solutions that do not last. Mitigation: residents lead analysis and priorities, and the plan is built around community capacity, not dependency.
Risk: exclusion of marginalised residents. Mitigation: weekend sessions and outreach by community-based researchers.
Risk: language gaps. Mitigation: sessions in IsiXhosa, translated afterwards.
Risk: end of funding. Mitigation: Kindling is seeking further support for the toolkit; no long-term funding is in place.
Risk & Mitigation Plan
Scalability and Sustainability
Community capacity combined with external support. Trained community-based researchers and residents carry the knowledge, the Community of Practice provides a continuing forum, and Kindling is seeking funding to extend the work.
The pilot is the basis for a model that others could adopt and expand; Kindling plans a Fire Risk Assessment and Fire Risk Reduction toolkit to share the method with communities, governments and practitioners, to pilot it in new communities and to make the resources openly available.
The pilot used community mapping, spatial analysis, thematic analysis of transcripts and co-created short films. Residents discuss fire safety in WhatsApp groups.
No public information on the cost of the pilot was found in the sources reviewed. Based on the available information, costs included training and payment for community-based researchers, facilitators, venues, interpretation and fire engineering time.
No public information on running costs after the pilot was found. Based on the available information, it can be inferred that continuing the work would need facilitation, the researchers’ time and meeting costs.
The work was structured around community capacity, not dependency, and continued collaboration across city, NGO, academic and community groups, through the Community of Practice, is needed; communities leading the process give more lasting results. No long-term funding or formal handover to local government was found.