Grantee Research Project Results

2014 Progress Report: Pyrolytic Cook Stoves And Biochar Production In Kenya: A Whole Systems Approach to Sustainable Energy, Environmental Health, and Human Prosperity

EPA Grant Number: SU835548
Title: Pyrolytic Cook Stoves And Biochar Production In Kenya: A Whole Systems Approach to Sustainable Energy, Environmental Health, and Human Prosperity
Investigators: Hestrin, Rachel , Davis, Jennifer A. , Edwards, Rufus D. , Zwetsloot, Marie , Torres, Dorisel , Guerena, David , Lehmann, Johannes , Fisher, Elizabeth , Hsu, Tedman , Kakavan, Kasra , Yun, Sungwon , Wong, Philip
Institution: Cornell University , University of California - Irvine
EPA Project Officer: Hahn, Intaek
Phase: II
Project Period: August 15, 2013 through August 14, 2015
Project Period Covered by this Report: August 15, 2013 through August 14,2014
Project Amount: $87,841
RFA: P3 Awards: A National Student Design Competition for Sustainability Focusing on People, Prosperity and the Planet - Phase 2 (2013) Recipients Lists
Research Category: Pollution Prevention/Sustainable Development , P3 Challenge Area - Air Quality , P3 Awards , Sustainable and Healthy Communities

Objective:

The principal objectives of our Phase II project were to improve human health, economic prosperity, and environmental protection by reducing air pollution from cook stoves. In order to achieve these objectives, our Phase II research goals were

1. to test a combined pyrolysisbiochar cook stove that was clean-burning, efficient, and user friendly under laboratory conditions and
2. to evaluate stove usage and performance in rural Kenyan households. 

Progress Summary:

Research Goal 1: Laboratory testing of pyrolysis cook stove

Our first research goal was to test a clean-burning pyrolysis cook stove under laboratory conditions. This data provided us with an understanding of the relationship between gaseous pollutants, heat transfer efficiency, combustion efficiency and fuel usage. In addition, we used this data to compare the performance of the pyrolysis stove to other stove models currently used. 

Cornell pyrolysis stove – Figure 1 shows EFs for the Cornell pyrolysis stove with three different fuels (two types of wood pellets and switch grass pellets) and two different modes of operation (combustion and pyrolysis).  CO emissions are significantly reduced when the stove is used in pyrolysis mode, by 4.04 g kg^-1 for switch grass. However for DC hardwood pellets, CO emissions of pyrolysis were almost twice as high as those of combustion. These results demonstrate that a reduction in oxygen availability leads to improved stove combustion efficiency, however the reductions are also dependent on the type of fuel used in the stove.

Figure 1. Emission factors of the Cornell pyrolysis stove during combustion and pyrolysis mode.

In all scenarios, NO emissions were lower during pyrolysis than for combustion, regardless of the type of fuel. However, fuel containing higher levels of nitrogen led to higher NO emission factors as seen in Figure 2. 

Figure 2. NO emission factors based on fuel type and initial nitrogen content.

EFs from wood and crop residue-fueled cook stoves can range from 9-179 g kg^-1 fuel CO (Venkatamaran and Maheswara, 2001; Zhang et al., 2000; ) and 0.01-1.57g kg^-1 NOx (Bhattacharya et al., 2002; Zhang et al., 2000). As can be seen in Figure 3, the EFs measured from our Cornell pyrolysis stove are comparable to those in published literature (Jetter et al., 2012). CO emissions from the Cornell pyrolysis are at the lower end of this range, confirming that the pyrolysis stove can reduce CO emissions as efficiently as other stoves currently being used. 

Figure 3. Cook stove performance between different combustion and gasification stoves

Figure 4 presents the overall efficiency and useful firepower values, during the water boiling test.  The overall thermal efficiency for the Cornell pyrolysis stove is approximately 25% and is comparable to some of the stoves currently being used. The overall thermal efficiency takes into account the combustion efficiency and the heat transfer efficiency. In the case of the Cornell pyrolysis stove, the combustion efficiency is comparable to values being reported in the literature, however the heat transfer efficiency is much lower. Therefore the overall performance of the stove is mostly affected by the efficiency of the delivery of heat to the pot. 

Figure 4. Cook stove performance metrics 

Research Goal 2: Assessment of stove usage and performance in Kenyan Households Our second research goal was to assess stove usage frequency and performance in Kenyan households. A household characterization and energy survey was administered in 450 households to identify type of stove used, average number of household members and fuel availability/diversity. A subset of 60 households was randomly identified and contacted to be part of the traditional and pyrolysis stove emission tests. Stove usage and emissions measurements were made in 12 households that used three stone stoves during the testing phase, 27 households that received the 2-chamber pyrolysis stove, and 4 households that received the TLUD gasification stove. Consent forms were provided to households that informed study participants of the risks associated with participating in the testing of the pyrolysis stoves, and options to decline participation. Field personnel verified with each household that cooking activities on the test day were typical of normal cooking events, and households were asked not to alter their normal cooking activities.  

To monitor stove usage we installed Thermocron IButtons ® (Maxim Integrated Products, Sunnyvale, CA) to monitor stove use from temperature signatures (Ruiz-Mercado et al 2008). We programmed the monitors to record date and temperature data every 10 minutes from a period of 5 months. The stove use monitors (SUMs) were installed in the three types of stoves evaluated at the time of this study; (1) three stone fire (2) TLUD and (3) 2-chamber.                  Monitoring and collection of data was performed at 30 day intervals, ending on the day of emissions testing. Final selection of the household participating during the emissions testing was based on stove use frequency during the monitoring period, prior to testing. During this period we collected time/ date stamps for all cooking events such that we could use this information to distinguish cooking event temperature signatures from the SUMs data.        

Type of stove	Number of cooking events detected, per day, per stove	Minutes of cooking detected, per day per stove
Cornell  pyrolysis stove	0.35	36.3
TLUD stove	0.14	10.1
3-stone fire (all)	1.03	111.3
3-stone fire with Cornell pyrolysis present in household	0.75	90.5
3-stone fire with TLUD present in household	1.16	90.7
3-stone fire with no other stoves introduced	1.39	148.1

Table 1.  Summary of stove use during the sampling period

Table 1 shows the frequency with which the participating households used the two types of pyrolysis stoves that were distributed compared to the use of their traditional stove, the three stone fire. Based on the data collected we can see that households where no stove was introduced used the three stone fire for a total of 2.5 hrs a day, on average. However, in the households where either of the pyrolysis stove were introduced, there was a reduction of the amount of time that the traditional three stone fire was used. Households that received the Cornell pyrolysis stove showed a higher frequency of use of the improved stove over those households that received the TLUD stove. 

While in both cases the households used the improved stove, the overall trend is to continue using the three stone fire and their main cooking stove. Upon finalizing the study a survey was administered to each household that had an improved stove. The purpose of the survey was to determine the benefits and disadvantages of using the improved stove over their existing traditional stove. In the majority of the households, the main reasons for not using the improved stoves as frequently as the traditional stove were the ease of use, preparation of fuel and the time it took to cook the food. 

Emission measurements were performed by field personnel who visited two households during the morning hours to set up equipment, and document the weight of fuel for the given day.  For open fire stoves, a three-way inlet aluminum sampling probe was placed approximately 1-foot above the emission source of the stove.  Flow rates were checked at specified points within the sampling train (i.e. cyclone, pumps, DustTrak) before and after each emission test using a BIOS Defender 510/520 primary flow meter (Mesa Labs International, USA).  Two separate sampling trains were simultaneously used for direct measurement of the stove emissions and background concentrations of the indoor space.  Semi-continuous real-time measurements were conducted using a TSI DuskTrak 8520 Aerosol Monitor, TSI Q-Trak Monitor (TSI, Shoreview, MN), and Drager PAC 7000 Single Gas Detector (Drager, Glasgow, UK) for PM2.5, CO/CO₂, and SO₂ concentrations, respectively.  PM_2.5 gravimetric sampling was performed using 37mm PTFE (Polytetrafluoroethylene) 2.0µm pore size filters (pre-weighed and loaded into 37mm styrene cassettes) using BGI Triplex Cyclone and an SKC Airchek PCXR8 (SKC, Inc., USA) or low-flow pocket sampling pump.  Collected gas samples were transferred from an 80L Kynar bag to individual 0.5L FlexFoil metalized gas sampling bags at the end of each emission test. The FlexFoil metalized sampling bags were purged at least three times from the contents of the Kynar bags before the sample was collected and retained inside the 0.5L gas sampling bags. Gas monitors were factory calibrated prior to the study. Pre and post weights of filters for particulate matter were weighed on an electro-microbalance (Cahn Model 29, Thermo Electron Corp., USA) in an environmentally-controlled microbalance room (45±3% relative humidity and 20±2^°C), and were equilibrated for at least 48 hours.  Field blanks were collected for quality assurance purposes. 

Table 2 shows preliminary results of 1 day emissions tests calculated as a function of CO2 and CO emissions. CH4 and Hydrocarbon emissions have not been included, but CO₂, CO and PM_2.5 account for the majority of the carbon emissions. There is a significant difference in the modified combustion efficiency of the stoves, with the pyrolysis stove performing significantly better than the 3-stone fire. In addition there was a reduction of 42% CO when compared to the traditional 3 stone fire. These two metrics would lead you to expect a large difference in the PM emissions, however these differences are less pronounced with a reduction of approximately 26% in PM emissions in g/kg fuel, largely as a result of the high variability in PM emissions and the longer time the pyrolysis stoves were used (elapsed time).       

Independent samples t test

	STOVE	N	Mean	Std. Deviation	Mean Difference	Sig. (2-tailed)	Reduction
MCE	3-STONE	11	0.91	0.02	-0.03	0.001	3%
	PYROLYSIS	23	0.94	0.02
CO g substance/kg fuel	3-STONE	11	132	40	55	0.001	-42%
	PYROLYSIS	23	77	26
PM g/kg fuel	3-STONE	11	12.09	5.5	3.2	0.12	-26%
	PYROLYSIS	23	8.91	4.7
Elapsed Time	3-STONE	11	202	47	-54	0.04	27%
	PYROLYSIS	23	256	95
net carbon utilized Kg	3-STONE	11	1.8	0.5	-0.06	0.79	3%
	PYROLYSIS	23	1.8	0.8
Fuel used kg/day	3-STONE	11	4.2	1.0	-0.41	0.35	10%
	PYROLYSIS	23	4.6	1.5
overall emission g/day	3-STONE	11	49	22	7	0.43	-14%
	PYROLYSIS	23	42	27

                                                                                                                          

Table 2. Summary of 1 day emissions test

The overall emissions decreased by 14% on a daily basis, which is not as significant as expected. This could be due to the fact that even though the pyrolysis stoves were used by the households, they were not used well, and the pyrolysis fuel quality was poor. The principle of the pyrolysis stove is to achieve enough heat with a small initial fuelwood fire to catalyze the pyrolysis, after which the pyrolysis gases are combusted for cooking, and no further tending of the fire is necessary until the end of the pyrolysis burn. It seems that the householders did not fully understand the function of the pyrolysis stove and were perhaps too accustomed to feeding a stove continuously. There was little difference in the net carbon utilized or the fuel utilized, once the biochar residual was subtracted. In addition, looking at the primary fuelwood used in the stove, homes with 3-stone fires used 4.7 (±1.1)kg of fuel wood compared to homes with pyrolysis stoves that used 3.6 (±1.6)kg, with an average mass of pyrolysis fuel of 0.75 (±0.3)kg. Thus, the pyrolysis chamber did not displace much of the fuel used in 3-stone fires, and the women seemed to be tending the stove continuously with wood, as with a 3 stone fire. It is also possible that the continued use of fuelwood may have been due to the poor fuel quality used in the pyrolysis chamber, which frequently had high humidity, and leads to poor combustion. 

Future Activities:

The principle objectives of this project were to improve human health, economic prosperity, and environmental protection by (1) reducing air pollution from cook stoves and (2) conserving natural resources. Our whole-systems approach was necessary in order to achieve simultaneous benefits for people, prosperity, and the planet.  

We were successful in achieving most of our Phase II research goals, including stove laboratory, evaluation of stove usage and field performance. Three key elements were critical to our success: (1) the combined expertise of our P3 team, (2) our interdisciplinary, whole-systems approach, and (3) our long-term, “on-the-ground” experience working with NGOs and smallholders in rural Kenya. The members of our P3 team represent many academic disciplines, including Engineering, International Agriculture and Rural Development, Epidemiology, Crop and Soil Science, Horticulture, and Water Resource Management. Several team members have long-term experience working with Kenyan smallholders. Our in-country collaborators at ICRAF have extensive experience with international development and sustainability. The broad expertise of our team has been essential to our success.

Although our project focuses on rural Kenyan households, the results of our research are broadly applicable throughout the world. Our results are particularly relevant in developing countries, where traditional cook stoves are prevalent. The results from our stove design and testing will inform further academic research in engineering and combustion chemistry. They are also useful to the industrial sector, which will continue to develop more efficient and cleanburning combustion systems in response to increasing pressure from climate change, deforestation, population growth, and environmental health issues. 

This project has allowed us to leverage expertise and funds from multiple projects and organizations. The research conducted in Phase II builds upon years of stove and biochar research. Our external partners at ICRAF, as well as collaborators at Re-Char, have considerable experience in international development, climate change, and cook stove design. Funding from an NSF BREAD award and several student fellowships (including Cornell’s NSF-funded IGERT for Food Systems and Poverty Reduction, Fulbright, the U.S. Borlaug Graduate Research Fellowship Program, the Boren Fellowship, and others) have supported the personnel working on this project.

This is the first time to our knowledge that pyrolysis stoves have been tested for emissions. As mentioned before, the design of stoves is an iterative process which involves three phases (1) design and laboratory testing (2) field testing and (3) re-design based and testing based on prior testing results. Therefore, the next step in this process is to assess at what level of the IWA classification tier is the stove performing. This will allows us to rank the performance of the stove, and identify strengths and weaknesses of the current design. 

The research outcomes of this project can be used to improve the quality of life of many people throughout the world. Improved pyrolytic cook stoves can improve indoor air quality and human health. They can also reduce the amount of time and resources spent gathering biomass for cooking, allowing more time to be spent on other income-generating activities.  

Supplemental Keywords:

Cook stoves, pyrolysis, biochar, renewable feedstocks, indoor air pollution and emission reduction

Progress and Final Reports:

Original Abstract

Final

P3 Phase I:

Pyrolytic Cook Stoves and Biochar Production in Kenya: A Whole Systems Approach to Sustainable Energy, Environmental Health, and Human Prosperity  | Final Report