Adoption and Use of Improved Stoves and Biogas Plants in Rural India (Authors: Somnath Hazra Jessica Lewis Ipsita Das and Ashok Kumar Singha ; SANDEE Paper.

Adoption and Use of Improved Stoves and Biogas Plants in Rural India

(Authors: Somnath Hazra Jessica Lewis Ipsita Das and Ashok Kumar Singha ; SANDEE Paper No. 86)

Presented by

Mohammad Hazrat Ali Bagati

Student No.: MSS 141514

Md. Monirul Islam

Student No.: MSS 141517

Economics Discipline

Khulna University Presented to

Mohammed Ziaul Haider, Ph.D

Professor & Head

Economics Discipline

Khulna University

Introduction

• About 40% of the global population (amounting to 3 billion people) rely on

solid biomass fuels including fuel wood, crop residues, charcoal, coal, and

dung for cooking.

• Residential bio fuel cooking is the second greatest source of black carbon, a

significant greenhouse pollutant that can lead to regional hotspots and the

melting of Himalayan glaciers and snow.

• The adverse health, livelihood, local environment, and climate impacts

generated by household biomass burning have gained increased attention in the

past few years. Improved cook stoves have been designed to alleviate these

negative impacts through increased combustion efficiency that requires less

fuel and reduces cooking time.

Stoves powered by biogas plants (biogas stoves) can deliver numerous benefits over traditional

cooking practices: 1) Mitigation of fecal-borne and parasitic diseases through the removal of openly

defecated dung;

2) Reduction in household air pollution;

3) Fuel substitution for firewood, reducing fuel collection time and easing strain on

local forests;

4) Combustion of methane reducing methane emissions, eliminating this

greenhouse gas that has a global warming potential over 20 times greater than CO2;

5) Generation of fertilizer (biogas slurry) that is more potent and of higher quality

than conventional fertilizer, which can lead to increased yields

6) A cost-benefit analysis of biogas plants in Ethiopia recently confirmed positive

net benefits to households

Challenges to biogas plant suitability(Slide 4-8)

• Recognizing these potential benefits, almost four million biogas plants have been

constructed across India through the Indian National Biogas and Manure Management

Program (NBMMP)from 1982 through 2007 (MNRE 2007), and there is potential for up

to 12 million.

• Many operational and structural problems may lead to early malfunctions and

abandonment of biogas plants, although technical studies report that the expected life

cycle of well-built and maintained biogas plants is around 25 years.

• Operational problems include “accumulation of water in the pipeline; scum formation in

the digester; clogging of the inlet and outlet; leakage of gas from the gas holder, etc.

Biogas plants have laborious operation and maintenance2010); households must

regularly add dung and water in specified proportions and stir the slurry at regular

intervals. Limited understanding of proper use and maintenance, or the inability to

diagnose or resolve reasons for failure are also barriers to plant sustainability

• Structural problems usually pertain to constructional defects, such as a crack in

a fixed dome plant or break in the pipe carrying gas to the house which will

allow all captured biogas to escape, rendering cooking with a biogas stove

impossible. Although construction requires engineering skills, many plants do

not have precise design specifications and defective plants can readily develop

cracks. Inadequate supply for spare parts can prevent the resolution of structural

problems .

Studies of biogas dissemination and adoption (Slides 4-8)

Existing literature from China, Kenya, India and Sri Lanka suggests that

households that adopt are more likely to receive a subsidy, be in the general

caste, and have higher income, more education, greater land ownership, and more

cattle than households without biogas plants .All of the studies from India focus

on partially-subsidized biogas plants. we conducted within our sample districts in

Odisha identified several issues of specific concern regarding biogas adoption:

1) No guarantee of high quality construction of biogas plants and no

maintenance provision lead to durability and repair issues.

2) Investment cost of the plants is high for the rural poor in Odisha, even after a

partial subsidy from OREDA.

3) The amount of biogas required to allow households to use biogas stove for all

meals is generally more than a small household level plant produces –

therefore, households continue to use their traditional stoves concurrently

with the biogas plant.

4) Dung is not a valueless commodity for these households; alternative uses

include production of organic fertilizer for crops or dung patties that are

burned in a traditional stove – these alternatives do not require the

significant investment of building a biogas plant.

Sampling and Data Collection Methods(Slides 4-8)Data were collected from households in rural Odisha in 2011 and 2012.Households were

selected from 8 districts in Odisha that received household biogas plants as part of a previous

campaign by the Odisha Renewable Energy Development Agency (OREDA). All OREDA

biogas plants were subsidized with subsidy levels ranging from 17-100%.

Focus groups and survey pretesting were done in June and July 2011 in four districts - Cuttack,

Jharsuguda, Koraput, Bolangir. A pilot was conducted in August-September 2011. The final

extensive panel survey was administered from November 2011 through February 2012in 503

households from 8 districts and 42 total villages: Angul (60 households), Cuttack (84),

Jagatsingpur (60), Jajpur (48), Jharsuguda (96), Keonghar (60), Sambalpur (84), Sundargarh

(11). These districts were selected to capture the geographical and cultural diversity of the state.

To ensure that the survey villages contained households who received biogas plants, blocks that

received the greatest number of subsidized biogas plants villages within these districts were

selected.

Household sampling was stratified by type of stove. After initial focus group visits

revealed that a large number of biogas plants were nonfunctional, the sample was

deliberately designed to include households with functional and broken biogas

plants, other forms of ICS, and traditional stoves, to facilitate analysis between

households using different stove types.

Within each village, approximately twelve households were selected by counting

every fifth house starting in the village center to meet pre-determined stove

ownership categories: 6 households with a biogas plant (3 households with a working

plant, 3 households with a non functional plant), 3 households with an improved

stove other than biogas (kerosene, LPG, electric, or rocket stove), and 3 households

with only a traditional mud stove were selected.

Whenever possible, the interview was conducted with the head of the household with

input from the primary cook. The survey collected data on household cooking and

fuel use behavior, demographics, socioeconomic characteristics, and consumption.

Conceptual Framework• We consider three different sets of dependent variables: 1) stove ownership, 2)

hours of stove use and 3) firewood consumption.• The simplest specification estimates the following equation to examine what

factors are significantly associated with stove ownership:• Stove Choice = β0 + β1 household demographic characteristics +β2 household

socioeconomic characteristics +β3 household health + β4 fuel use characteristics + ε

• Household demographic characteristics include gender of the head of the household (femalehead), age of the head of household (headage_HH), and household size (hhsize). Household socioeconomic characteristics include caste (caste_general), monthly expenditure (lnmonth_exp), location in an industrial area (Industrial), number of rooms (rooms), ownership of land (own_land), ownership or use of a toilet (toilet), ease of access to a 5,000 rupee loan(loan_access), and reliance on electricity as main source of lighting (electricity). Household health is represented by the number of days household members spent in the hospital due to their last episode of ARI (hospARI_HH).Fuel use characteristics include daily expenditure on traditional fuels (tradfuelcost) and fuel collection hours per day(fuelcoll_HH).

Conceptual Framework (Contd….)• A second specification adds additional demographic variables including education of

the head of household (headedu_HH), gap in ages between head of household and primary cookas an indicator for intra-household bargaining power (agegap), household status below the poverty line (BPL), and religion (hindu). We also include an indicator for household belief that household air pollution leads to negative health affects (IAP_effect).Households were asked questions as part of a thought experiment to assess their patience and preference for risk-taking.

• Finally, Based on survey responses to these questions, dummy variableswere included for households that were most willing to wait for a larger payment in the future rather than a smaller immediate payment (mostpatient) and households that prefer the chance of receiving a larger payment rather than a smaller certain payment (mostrisk). An additional variable for household illness is added – the number of people in the household who have had malaria in the past three years (malaria). In addition to self-reported ease of obtaining a loan, whether the household has received a loan in the past year (loan) was included.

• Several additional stove-specific variables were included in the regressions examining adoption of biogas: the amount of the government subsidy (biogcost_subs), the total cost of building the biogas plant (biogcost), and whether a reduction in wood was a reason considered when building the biogas plant (bg_wood).

Results and Discussion• In this portion, there have a several explanatory variables but for space

limitation only significant variables are mentioned in the tables:• ICS adoption:

Households that own any form of improved stove (71% of the sample) are compared with households that do not own an improved stove with logit regression (Table 6). Many improved stove households also own a traditional 8 South Asian Network for Development and Environmental Economics stove, as discussed above in detail. As expected, socioeconomic variables (monthly expenditure, number of rooms, land ownership, electricity as a main source of lighting, and toilet ownership/use) are significantly associated with household ownership of an improved cook stove. The results are significant when village-level fixed effects are included.

Table 6: Logit regressions on stove ownership

Results and Discussion (Contd..)

Variables (1) (2) (3)1

Log of total monthly expenditure 0.76*** 0.71*** 0.69***

Numbers of total rooms 0.27*** 0.32*** 0.20*

Household owns land 0.86***

Household uses a community, neighbor’s, or private toilet (not nec. exclusive)

0.91** 0.93** 1.10***

Main source of lighting in household is electricity 0.95** 0.70* 1.52***

Household has female head of household 0.56**

Fuel collection hours per day for entire household -0.11*

Survey Respondent was female -0.68** -0.68**

Number of years of education of household head 0.12**

Household is Hindu 0.12**

Constant -6.87*** -7.08***

Observations 502 406 370

Pseudo R-squared 0;172 0.200 0.274

ICS Ownership

Biogas adoption and sustainability• Using logit regression, we compare households that have a biogas plant (46% of

sample) with those that do not (Table 7). Households that own land, and have a greater number of rooms in their houses are more likely to have a biogas plant. Lower traditional fuel expenditure and ownership of more livestock that produce dung are significantly associated with ownership of biogas plants, signaling that dung production may be an important constraint. Lower overall monthly household expenditures is also significantly associated with biogas plants ownership, but this is likely driven by the inclusion of households with broken plants, since households with working plants have significantly greater expenditure than those with broken plants.

• Table 8 shows Higher income, lower rates of malaria, less patient households, and lower likelihood of a female head of household are all significantly associated with household ownership of a working biogas plant. A lower rate of electricity use (as the main source of lighting) is significantly and negatively associated with ownership of a working biogas plant. Lower expenditure on fuel and less time gathering traditional fuels are both significantly associated with households’ ownership of a functional biogas plant. As expected, ownership of more livestock that produces dung is also significantly associated with working biogas plants. Households that considered the reduction in fuelwood consumption when building a biogas plant were more likely to have a biogas plant that remained functional.

Table 7: Logit regressions on ownership of all biogas plants

Biogas ownership 1

Variables (1) (2) (3) (4)2

Log of total monthly expenditure -0.42** -0.44** -0.46** -0.40*

Number of rooms in house 0.13* 0.13*

Household owns land 1.14*** 0.92**

Household size 0.11*

Daily expenditure on traditional fuel (Rs) -0.12*** -0.10*** -0.08*** -0.09***

Household has taken a loan in the past year -0.58** -0.65**

Household is Hindu 0.89**

Number of milk buffalo and milk cows (> 1yr old) household owns

0.65*** 0.69***

Log of total monthly expenditure -0.42** -0.44** -0.46** -0.40*

Constant 1.53 1.55 1.27

Observations 253 201 241 229

Pseudo R-squared 0.0934 0.135 0.143 0.170

Table 8: Logit regression on ownership of only working biogas plants

Biogas (currently working) ownership1

Variables (1) (2) (3) (4)2 (5) (6) (7)2

Log of total monthly expenditure 0.85*** 0.70* 0.72** 0.61* 0.75** 0.99**

Number of rooms in house -0.23* -0.24* -0.27* -0.23*

Household private toilet -0.57*

Main source of lighting in household is electricity -1.28*** -1.41***

-1.52**

HH says easy to borrow 5000Rs 0.89* 1.25** 1.00**

Household size 0.30*

Household has female head of household -0.84*

Daily expenditure on traditional fuel (Rs) -0.05* -0.08** -0.07* -0.18**

Fuel collection hours per day for entire household -0.14*** -0.18** -0.16* -0.27** -0.12** -0.15***

-0.20**

Households take the most risks in time / risk game 2.08*** 2.06*** 2.91***

of household members reporting malaria in past 3 yrs

-0.35* -0.37* -0.70**

buffalo and milk cows (> 1yr old) household owns 0.18** 0.22* 0.17** 0.17** 0.21

Reduction in wood needed for trade. 1.15** 1.50*** 1.80**

Stove use• (Table 9). Logically, household size is significantly associated with a greater number of hours of

stove use regardless of stove type. Household expenditure has a positive significant association with hours of clean stove use (and the converse for traditional stove use), and other socioeconomic indicators (ownership of land, use of electricity as the main source of lighting) are negatively associated with hours of traditional stove use. The number of hours per day collecting traditional fuel is positively associated with use of traditional stoves, as expected. The number of days spent in a hospital for respiratory illness has a significantly negative association with hours of clean fuel use. Table 9: OLS regression of hours of stove use

Traditional stove Improved stove Biogas stove

Variables (1) (2) (3)1 (1) (2) (3)1 (1) (2) (3)1

Log of total monthly expenditure

-0.53***

-0.47*** -0.53***

0.27***

0.35***

0.40***

Household owns land

-0.27*

Main source of lighting inhousehold is electricity

-0.28*

Household size 0.23***

0.18*** 0.18***

0.11* 0.11* 0.11* 0.10***

Table 9: OLS regression of hours of stove use (cont’d)Household has female head ofhousehold

-0.74**

Household is in open/generalcaste

-0.50** -0.43*

Fuel collection hours per day forentire household

0.04* 0.06** 0.05*

Number of days householdmembers spent in hospital dueto last ARI episode

-0.09** -0.11**

Households take the most risksin time/risk game

-0.39*

Number of household membersreporting malaria in past 3 yrs

-0.12** -0.16***

Number of years of education ofhousehold head

0.04**

Household is Hindu 0.96**

Constant 6.42***

6.02*** 7.75*** -2.44** -2.89*

Observations 445 354 354 319 269 269 199 167 167

Fuelwood consumption• Model of fuelwood consumption with OLS regression (Table 10). The first analysis includes ownership of a

traditional stove as a covariate, and finds that household ownership of traditional stoves is strongly significantly associated with an increase of about 25 kilograms of fuelwood consumption per week compared to households without a traditional stove. A second model specification also finds this effect in households that own biogas stoves – those with a working biogas plant consume about 5 kilograms of firewood less per week than households with a nonfunctional plant. As expected, both fuel expenditure and time spent collecting fuel are significantly correlated with firewood consumption. Households’ general caste status, and ownership or use of a toilet is significantly associated with higher fuelwood consumption

• Table 10: OLS regression of firewood consumption (kg) per weekFirewood consumption

Variables (1) (2) (3)

Household has traditional stove - chulha, cast iron biomass, or coal

25.55***

Log of total monthly expenditure -5.12**

Number of rooms in house 0.76*

Household uses a community, neighbor’s, or private toilet (not nec. exclusive)

-3.09* -7.92***

Household is in open/general caste -3.15*

Daily expenditure on traditional fuel (Rs) 0.48*** 1.23*** 1.15***

Fuel collection hours per day for entire household 0.82*** 0.98**

Household biogas plant is working -4.52* -4.66**

Constant 28.06** 55.90***

Conclusion

This research adds to the limited evidence base of rigorous household ICS adoption and use studies. Our analysis indicates that greater fuel expenditure and time spent in the hospital for respiratory disease are significant associated with traditional stove use, while socioeconomic factors are significantly related with adoption of improved stoves. There are high rates of biogas plants facing structural and operational problems in India. Our analysis of the factors associated with continued functionality of biogas plants finds that households with greater spending capacity and more biogas-producing livestock are more likely to own biogas plants that still work. Households that spent less time gathering and money purchasing traditional fuels, and those that received a greater subsidy during plant construction were significantly more likely to own working biogas plants. The latter suggests that the subsidy may indicate a higher plant quality or greater government oversight. Village location in an industrial area and access to loan facilities were not significantly associated with stove ownership.

This analysis suggests that biogas plants have the potential to reduce firewood use, time spent gathering fuel, and respiratory disease caused by household air pollution. Future policies encouraging the construction and maintenance of biogas plants have the potential to provide tremendous health and environmental gains.

Thank You All