Skip to content
problem genome

One of 592 unsolved problems in the Problem Genome, a collection written up for student design and engineering teams.

health · manufacturing

marked approved, faulty at home

South Africa requires every paraffin stove type to be approved against a national standard, but when 150 approved stoves were given to households in a Johannesburg informal settlement, 43 were found defective a year later, failures the researchers trace mostly to non-compliant manufacture, and 23 of those households kept using the faulty stove

Source: Kimemia et al., “Burns and fires in South Africa's informal settlements: Have…” Open ↗source: primaryLast updated: 2026-11-17

Problem statement

Illuminating paraffin (kerosene) is the main cooking fuel for about two million South Africans. Energy poverty is concentrated in informal urban settlements, which remain largely unelectrified. There, a paraffin stove that leaks, flares or tips over can set off a shack fire, and fire can spread from one dwelling to many. Under a compulsory specification (VC 9089, in its 2013 form), every type of non-pressure paraffin stove must comply with the national standard SANS 1906 and be approved before it is sold, with the approval number marked on the stove or its packaging. An intervention project gave approved stoves, with safety education, to 150 households in a Johannesburg informal settlement. A follow-up survey 12 months later found 43 stoves (of the 135 still in the study; 15 households had moved) that had malfunctioned, and 23 of those households were still using them. Laboratory tests on seven of the faulty stoves, run at the end of 2015, about 18 months after the stoves were handed out, found fuel leaks in five of them (a brand-new stove of the same design and maker, tested for comparison, also leaked, which the authors take to point to manufacturing defects as the possible cause) and defects "in critical safety features such as flame control and the self-extinguishing mechanism." The same researchers later wrote that the failures "occurred within days to weeks of first use" and that the lab tests showed they were "mostly due to manufacturing shortcomings and non-compliance with the compulsory paraffin stove specifications." The 2018 paper also faults the standard itself: it calls the 10° tip-over limit "too lenient," says the stability requirement covers the stove alone and should cover the stove with its pot, and says that rating durability in hours of burning "favours the industry" at the expense of poor households, who have to buy more than one stove a year. In the 2021 paper's words, "non-compliant products continue to be manufactured and distributed with the official mark of approval," and the regulator's role "appears ineffectual." Approval checks a stove type, on a specimen, before sale. What is missing is a cheap, repeatable way to catch stoves that carry the mark but do not meet the standard, in shops and in the first weeks and months of use. In April 2026 a fourth edition of SANS 1906 was issued in draft for public comment. The unsolved problem is making the approval mark mean that the stove in the home actually meets the standard, and giving households who cannot afford a replacement a way to know when theirs does not.

Why this matters

According to the 2022 commentary by the same research group, about 600,000 households (2 million people) relied on paraffin for domestic energy in 2018. Paraffin fuels cause 53% of informal-settlement fires (candles 30%), and paraffin stove failures that end in an explosion "are often implicated" in the roughly 5,000 shack fires emergency services attend each year. The commentary adds that the resulting flame burns, the most severe burn outcomes, "likely comprise a significant proportion" of the 100,000 burn injuries and 2,000 burn deaths reported each year. At least ZAR 490 million is spent caring for patients with paraffin burns, up to ZAR 180 million is estimated to be lost each year in destroyed homes, and survivors carry long-term social, economic and psychological costs. The 2021 South African Medical Journal paper describes the fuel as "burnt in poorly made stoves that are prone to malfunction," and still used because it "is easily decanted, widely available in neighbourhood outlets, perceived as affordable, and often the only available option."

What’s been tried and why it hasn’t worked

  1. A compulsory standard with type approval. The compulsory specification VC 9089 (as amended in 2013) requires stoves to "comply with the requirements of SANS 1906." Each appliance type must be approved by the National Regulator for Compulsory Specifications (NRCS) before it is imported, sold or supplied. The application includes test reports from a recognised conformity-assessment body and "one specimen of the appliance type." After approval, "ongoing proof of compliance" must be made available to the regulator (conformity of production), with records kept for three years; the specification defines the applicant for approval as "The manufacturer or importer", and the NRCS may withdraw approval if requirements "have not been met or maintained." Nothing in the specification's approval annex describes testing stoves after a period of use in homes. SANS 1906 itself was not read; by the researchers' account, its 2012 edition (Ed 3.1) requires a self-extinguishing mechanism that puts the flame out if the stove is tilted beyond its stability limit, a leak-proof tank, a steady flame and sturdy construction (their 2017 SAMJ paper).
  2. Approved stoves plus education, followed up in homes. The follow-up study (full text read) found 43 of the 135 approved stoves still in the study had malfunctioned by the 12-month survey (15 of the 150 households had moved). The faulty stoves had defects in flame control and the self-extinguishing mechanism. Households reported four flare-ups; in two, clothing or hair caught fire, and none caused an injury that needed hospital care. Of the 23 households still using a faulty stove, 21 were interviewed. Continued use was "significantly associated with the time from receipt of the stove to detection of first failure": stoves that failed later were more likely to stay in use. In the interviews, the average times to a detected failure of the five safety features, taken separately for households that kept using the stove and those that stopped, ranged from about three to eight and a half months; the industry rates the stove's design life at 500 hours of burning, about six months of use. The authors conclude that enforcement of appliance standards, public education on stove use and "structural change for the energy-poor" are all needed. Their 2021 account adds that the 150 households were a random sample, that the failures involved auto-extinguishing mechanisms, flame control levers and "leakproof tanks," and that lab tests traced them mostly to manufacturing shortcomings and non-compliance. Their 2017 paper says the self-extinguishing mechanism on approved wick stoves is "prone to malfunction and failure within a few days to weeks of use" and that leaks occur "even in a new, out-of-the-box stove."
  3. Enforcement in shops. Raids do happen. In a 2016 blitz the NRCS and the National Consumer Commission, with the City of Cape Town and police, confiscated 44 non-compliant paraffin heaters and three paraffin stoves from 61 Western Cape retailers. The 2021 paper still judged the regulator's role "ineffectual," because non-compliant stoves keep being sold with the approval mark. No read source describes systematic sampling of marked stoves from shops and testing them against the standard.
  4. The researchers' call for review and enforcement. In 2021 they asked for SANS 1906 to be "urgently reviewed to make it more stringent and directed towards enhanced performance of critical safety features such as the self-extinguish mechanism, flame control, mechanical stability and a leak-proof tank." In 2022 the same group wrote that "the current compulsory paraffin stove standard is dated," asked the standards body (SABS) "to institute an urgent review to address its shortcomings," and called on the NRCS to increase enforcement and curtail, for example through harsher penalties, the manufacture and distribution of "risky sub-standard stoves." They also asked for safety education and compulsory safe-use information on stoves and fuel packaging. Their main recommendation is a longer-term phase-out of paraffin in favour of electricity or LPG. That is an energy-access question this page does not take up.
  5. What happened next. On 17 April 2026 the SABS issued SANS 1906 Ed 4 in draft for public comment (closing 2026-06-16). The gazette notice gives only its scope: the "construction, operation and safe use" of non-pressurised paraffin appliances "primarily intended for domestic use." Whether the new edition addresses these failure modes, or how approval and market surveillance will change, was not found in any read source.

What would unlock progress

A way to tell, cheaply and repeatedly, whether a stove that carries the approval mark actually has working flame control, self-extinguishing and a tank that does not leak, both when it is bought and in the first weeks and months of use, when the researchers say failures appear. That could be a short screen that an inspector can run on stoves bought from shops (a tilt check, a flame-control check, a leak check, and a few days of use cycles, knocks and refuelling), with a matching field check that a researcher, a fire-safety officer or a household can run. The follow-up study also shows a human side. Stoves that fail later stay in use, so a household needs a clear signal that a stove has crossed from worn to dangerous, and a replacement route it can afford. Other product-safety regimes that pair type approval with market sampling are the obvious comparison.

How we’d know it’s being solved

The share of approved stoves that households report as having malfunctioned after 12 months of use, measured the way the Johannesburg follow-up did (43 of 135), should fall sharply, and so should the share of marked stoves bought from shops that fail a screen of flame control, self-extinguishing and leaks. The share of households who keep using a stove after it has failed should fall too (23 of 43). At the national level, the burn and shack-fire counts the commentary cites would be the eventual measure. They are not a fair test of a stove standard alone, because fuel, housing density and candles drive them as well.

Where to start

Facility-free door (design the field check): using the failure modes the researchers name (flame control, the self-extinguishing mechanism, leaking tanks), design an inspection protocol: a sampling plan for marked stoves bought from shops, a check sequence, a pass/fail sheet and a pre-registered analysis plan. A South African research group or inspector could use it on stoves at purchase and in homes at 2 weeks, 3 months and 12 months. The deliverable is the protocol, handed over; it is not run. This door needs no facility.

Make door (a check card that stays with the stove): make a printed, durable card or tag that hangs on the stove and lets a household check, in pictures and few words, the same failure modes: a leaking tank, a flame that will not turn down, a self-extinguishing part that is broken or does not move freely. Every check on the card is done with the stove cold and unlit; it never asks anyone to tip a lit stove. It gives the plain "stop using it" warning the follow-up study asks for when a key safety feature is faulty. A design studio can make and print it. Testing it with households is designed, not run: hand over a test plan for a South African partner.

Make-and-test door (screening rig, access line): build a bench rig that puts a wick stove through repeated heat cycles and controlled knocks and tilts, and records when flame control and the self-extinguishing mechanism stop working, so that an inspector could screen shop-bought marked stoves in days rather than waiting for failures in homes. Running it with fuel needs a fire-safe laboratory or test hall and a supervisor trained in fire testing. A team without that access should build and check the mechanical part (cycling, tilt, timing) with the stove unfuelled, and hand the fuelled test protocol to a lab that has that access.

Access note: the people who would test these results are households in a Johannesburg informal settlement, product inspectors and fire-safety officers in South Africa. None of them can be reached from a US or European campus within a term. Any work with them goes through a South African partner, such as a university injury-research group or a municipal fire-safety office, and through that partner's ethics review. On campus a team can only check that the card is easy to read; that is not a test with real users.

Relevant skills: product-safety and test engineering, mechanical design, survey and inspection design, graphic and information design, product design for printed and durable objects.

Genes

Related problems

Stuck the same way, in other fields.

field: food & the living worldthe sorbent works until real air touches itDirect air capture sorbent degradation under real atmospheric conditionstested in the wrong worldsource: secondaryfield: food & the living worldsensors degrade in real soilReal-time soil nutrient sensors can be unreliable in field conditionstested in the wrong worldsource: primaryfield: digital & spaceevery tool call compounds the errorAI agents that chain tool calls suffer exponential reliability decaytested in the wrong worldsource: secondaryfield: making & materialsa robot lab made 36 of 57; mixtures trip it upA self-driving lab ran closed-loop for 17 days and made 36 of 57 inorganic targets, but reading multi-phase samples remains hardtested in the wrong worldsource: primary

Sources

  • Kimemia D, van Niekerk A, Govender R, Seedat M. "Burns and fires in South Africa's informal settlements: Have approved kerosene stoves improved safety?" Burns: journal of the International Society for Burn Injuries 2018, 44(4): 969-979, doi:10.1016/j.burns.2017.11.006 (full text read through library access; abstract record, PMID 29395395), accessed 2026-11-17. Link ↗ Link 2 ↗ Link 3 ↗
  • Van Niekerk A, Kimemia D, Seedat M, Annegarn H. "Energy impoverishment and burns: The case for an expedited, safe and inclusive energy transition in South Africa." S Afr J Sci. 2022;118(3/4), Art. #13148, accessed 2026-11-17. Link ↗ Link 2 ↗
  • Department of Trade and Industry, "Amendment of the Compulsory Specification for Non-Pressure Paraffin Stoves and Heaters - VC 9089," Government Notice R. 552, Government Gazette No. 36725, 8 August 2013, accessed 2026-11-17. Link ↗
  • Department of Trade, Industry and Competition, "Standards Act, 2008: Standards Matters," Notice 3890 of 2026, Government Gazette No. 54520, 17 April 2026 (Schedule A, drafts for comment: SANS 1906:20XX Ed 4), accessed 2026-11-17. Link ↗
  • Kimemia D, Van Niekerk A, Seedat M. "Paraffin dangers, health and socioeconomic consequences: Urgent need for policy action." S Afr Med J 2021;111(1):17–19, doi:10.7196/SAMJ.2020.v111i1.15095 (full text, archived publisher PDF), accessed 2026-11-17. Link ↗
  • Kimemia D, van Niekerk A. "Energy poverty, shack fires and childhood burns." S Afr Med J 2017;107(4):289-291, doi:10.7196/SAMJ.2017.v107i4.12436 (full text, archived publisher PDF), accessed 2026-11-17. Link ↗
  • SAnews, "Paraffin heaters, stoves confiscated in W-Cape," 9 December 2016, accessed 2026-11-17. Link ↗
  • Kimemia D, van Niekerk A. "Why unsafe paraffin stoves are still being widely used in South Africa." Elitsha News, 10 August 2018 (the authors' own news article; supporting only), accessed 2026-11-17. Link ↗
Revision notes

This brief gets checked against its sources and contains the following revisions. Future revisions get added here.

Worth knowing

  • Five of the seven faulty stoves tested in the laboratory leaked, and so did a new stove of the same make; the authors call manufacturing defects the 'possible' cause of the leaks.
  • The central evidence is one follow-up of 150 households in one Johannesburg settlement.

Full check

Sources read

  • Kimemia et al., approved kerosene stoves in South Africa's informal settlements (Burns, 2018): full text (read through library access)
  • Kimemia, Van Niekerk and Seedat, paraffin dangers (S Afr Med J, 2021): full text (from an archived copy of the publisher's PDF)
  • Kimemia and van Niekerk, energy poverty, shack fires and childhood burns (S Afr Med J, 2017): full text (from an archived copy of the publisher's PDF)
  • Van Niekerk et al., energy impoverishment and burns (S Afr J Sci, 2022): full text
  • Department of Trade and Industry, amended compulsory specification VC 9089 for paraffin stoves (Government Gazette, 2013): full text
  • Standards Matters notice listing the draft fourth edition of SANS 1906 (Government Gazette, 2026): full text (the relevant entry)
  • SAnews, paraffin heaters and stoves confiscated in the Western Cape (2016): full text
  • Kimemia and van Niekerk, the authors' own news article (Elitsha News, 2018): full text (searched; supporting only)
  • SANS 1906, the national paraffin stove standard, and its draft fourth edition: could not be opened (paid or not public)

What changed

  • Numbers: the 43 of 135 stoves are now described as stoves the households themselves judged to have malfunctioned, for any fault, not as stoves with a defect in named safety features.
  • Credit: the words about non-compliant stoves carrying the official mark, and the regulator's role being 'ineffectual', are now credited to the 2021 paper, not the 2018 one.
  • Citation details: the 2022 paper is now described as an invited journal commentary, not a peer-reviewed paper.
  • Entry points: the check card now gives the plain warning to stop using a faulty stove that the follow-up study asks for, rather than a signal households were said to lack.

Left open

  • what the national standard SANS 1906 requires [the standard was not read; described as the researchers report it]

Separate check of specific claims: A loose paraphrase of the 2018 paper's criticism of the stability test was brought back to its wording: the requirement covers the stove alone and, the authors say, should cover the stove with its pot.

Separate check of specific claims: The page now says the laboratory tests were run about 18 months after the stoves were handed out and that five of the seven faulty stoves leaked, as did a brand-new stove of the same design, which the authors take to point to manufacturing defects as the 'possible' cause; the sentence on ongoing proof of compliance now follows the specification's own wording; and citations were added to the source list.

Earlier full check: Two central figures were corrected from the full paper: the 43 faulty stoves are of the 135 still in the study, and the four flare-ups caused no injury needing hospital care; the page now places failures in the first weeks and months of use, with the interview averages, and adds the paper's criticisms of the standard itself.

Separate check of specific claims: Entry points: the household check is now its own entry point, a printed check card used only on a cold, unlit stove, with household testing designed for a South African partner; the access note now says who the real users are and that they cannot be reached from a US or European campus within a term.

Earlier full check: After the authors' full papers were read, the page was reframed from wear over months of use to stoves made below the standard but sold under the approval mark, both titles were changed (display title 'Approved, Then Broken' → 'Marked Approved, Faulty at Home'), and the researchers' account of the standard, their call for a stricter one and a 2016 enforcement raid were added.