Portal de Periódicos da CAPES

The effect of solar ovens on fuel use, emissions and health: results from a randomised controlled trial

2013; Taylor & Francis; Volume: 5; Issue: 2 Linguagem: Inglês

10.1080/19439342.2013.775177

1943-9342

Theresa Beltramo, David I. Levine,

Energy, Environment, and Transportation Policies

Abstract Inefficient cookstoves contribute to deforestation and global climate change, require substantial time (usually of women and girls) collecting wood or money for fuel and lead to just under two million deaths a year. We examined the effect of solar ovens on fuel use, time spent collecting wood, carbon monoxide exposure, and respiratory illness symptoms. A phased randomised controlled trial was run among women interested in purchasing a solar oven in rural Senegal. Of the envisioned 1000 households, 465 treatments and 325 controls took part in the baseline survey. Households randomly allocated to the control group received their stoves 6 months after treatments. Eighty per cent of our respondents typically cook for more people than the capacity of the solar oven and thus even cooks using the solar oven continue using their traditional stove. In the sixth month of owning the stove, treatments used their solar oven 19 per cent of days measured and did not have statistically significantly lower fuel consumption, time spent collecting fuel or time spent next to the cook fire. However, treatments cooking for 7–12 persons did lower their wood consumption for cooking by 14 per cent (P < .01). There is no evidence solar ovens reduced exposure to carbon monoxide or self-reported respiratory symptoms such as coughs and sore throats. This evaluation was a policy success because its results halted the proposed nationwide rollout of the solar oven, thus avoiding mass distribution of a stove which cannot reduce indoor air pollution or generate a sizeable decrease in fuel use. The results from this randomised controlled trial show that the HotPot is a poor product choice for the population as a one-pot stove cannot replace the three-stone fire for the lunch meal due to complex cooking patterns with multiple stoves, cooks and burners. A key result from our programme is stove designers – both solar and other improved biomass cookstoves – should reassess the product design to produce stoves that are affordable, durable, locally appropriate, consistent with current cooking practices (i.e., containing two burners) and large enough to accommodate multi-generational and/or polygamous households with limited incomes and no electricity. Keywords: indoor air pollutionrandomised controlled trialimproved stoveenvironmentacute respiratory illness Acknowledgements This research was supported by the National Institute of Health, the Blum Center, Sustainable Product Solutions, Berkeley Institute for the Environment, Berkeley Population Center, Solar Household Energy (SHE) and Tostan. Authors thank Darwin Curtis, Marie-Ange Binagwaho and Bridget Huttenlocher at SHE and Molly Melching, Cody Donahue, Dame Gueye and the Community Management Committee organisers at Tostan. Lamine Ndiaye, Henry Silverman, Richard Tam, Kenneth Tsang, Miyeon Oh, Cheikh Sidaty Ndiaye, Kine Seck, Magueye Ndiaye, Thierno Diagne, Awa Ndiaye, Wawounde Diop, Maimouna Sow, Ngone Sow, Aminata Bass, Anna Dione, Ibrahima Balde, Babacar Cisse and Yatening Thiang provided excellent research assistance, led by Vanessa Reed in the field. Kirk Smith, Isle Ruiz-Mercado and the UC Berkeley RESPIRE team guided the authors on measurement. Authors appreciate comments from Grant Miller, Mushfiq Mobarak, Rena Hanna, Esther Duflo, Michael Greenstone, Vijay Modi, Alexander Pfaff, Valeria Muller, Darby Jack, Robert Van Buskirk, Jason Burwen, Ashok Gadgil, Dan Kammen, Nils Tomajina, Jimmy Tran, Matt Evans, Evan Haigler, George Scharffenberger, Arianna Legovini, Abdoulaye Sy and Pascaline Dupas, and also thank participants at the World Bank, U.C. Berkeley, ASSA and Katholieke Universiteit Leuven LICOS. This randomised controlled trial passed the University of California Berkeley Office of the Protection of Human Subjects (Review #: 2007-12-50). Authors are responsible for all errors. Notes 1. Power calculations are based on data collected in similar improved cookstoves programmes from colleagues Robert Van Buskirk (2004) in Eritrea and Kilabuko Matsuki Nakai (2007) in Tanzania. From VanBuskirk, we took the variance and standard deviation for kilograms of wood used per cooking session and hours per week spent collecting fuel. From Matsuki Nakai, who took air pollution samples (PM10, NO2, CO) in 100 kitchens in rural Tanzania in 2007, we derived the variance and standard deviation of CO for cooks. 2. The sample of type of cooking structure comes from our CO sub-sample at the 6-month follow-up where we visited households' homes and has a total sample size of 281 households, though only 275 CO ppm/hour were successfully recorded. 3. We document these claims and describe the data sets below. Data from the next two paragraphs are from the baseline survey except data on cooking structures which is from the subsample of our 6-month follow-up where we measured carbon monoxide exposure. 4. We also fielded a 1-month follow-up, but solar oven usage was so low that first month (9% on average) that we only look for effects using the 6-month follow-up. 5. The ibutton is sold by Maxim. We appreciate advice from the RESPIRE team on the ibutton SUMs. 6. Forty-one per cent (236) of treatments had one or more treatment and/or controls living in the compound and 54 per cent (176) of control households have one or more treatments and/or controls living in the compound (see Appendix 4). 7. Measured CO exposure was higher at baseline, but we believe measurement problems biased up those results. 8. The other common complaints were that it cooked too slowly (35%) and was not durable (11%). 9. An additional barrier was that the solar oven required women simmer, not fry, rice for the most common lunch meal of ceebu jën (Beltramo 2010 Beltramo, T. 2010. "Peer Effects and Usage of the Solar Oven – Evidence from Rural Senegal". In Doctoral dissertation, Ca', Foscari University: Venice. [Google Scholar]). 10. We do not think leakage, where controls used solar ovens of treatments in their compound, were important for our results. First, results were unchanged when we examined only treatments and controls with no other study participants in their compound. Second, the majority of cooking was for nuclear families. Finally, usage rates were low, even including any leakage, so we cannot have missed major benefits of the solar ovens.