LETTER • OPEN ACCESS Recent citations Dynamics of sustained use and abandonment of - Association of personal network attributes with clean cooking adoption in rural South clean cooking systems: lessons from rural India India Praveen Kumar et al

To cite this article: Nishesh Chalise et al 2018 Environ. Res. Lett. 13 035010 - The refill gap: clean cooking fuel adoption in rural India Bodie Cabiyo et al

View the article online for updates and enhancements.

This content was downloaded from IP address 170.106.35.93 on 28/09/2021 at 08:48 Environ. Res. Lett. 13 (2018) 035010 https://doi.org/10.1088/1748-9326/aab0af

LETTER Dynamics of sustained use and abandonment of clean OPEN ACCESS cooking systems: lessons from rural India

RECEIVED 3 October 2017 Nishesh Chalise1, Praveen Kumar2, Pratiti Priyadarshini3 and Gautam N Yadama2,4 REVISED 1 Department of Social Work, Augsburg University, Minneapolis, United States of America 8 February 2018 2 Boston College School of Social Work, Boston College, United States of America ACCEPTED FOR PUBLICATION 3 Udaipur Regional Office, Foundation for Ecological Security, Udaipur, India 20 February 2018 4 Author to whom any correspondence should be addressed. PUBLISHED 14 March 2018 E-mail: [email protected]

Keywords: clean cooking, community based system dynamics, adoption, household air pollution, , sustained use Original content from this work may be used under the terms of the Abstract Creative Commons Attribution 3.0 licence. Clean cooking technologies—ranging from efficient cookstoves to clean fuels—are widely deployed Any further distribution to reduce household air pollution and alleviate adverse health and climate consequences. Although of this work must maintain attribution to much progress has been made on the technical aspects, sustained and proper use of clean cooking the author(s) and the title of the work, journal technologies by populations with the most need has been problematic. Only by understanding how citation and DOI. clean cooking as an intervention is embedded within complex community processes can we ensure its sustained implementation. Using a community-based system dynamics approach, we engaged two rural communities in co-creating a dynamic model to explain the processes influencing the uptake and transition to sustained use of biogas (an anaerobic digester), a clean fuel and cooking technology. The two communities provided contrasting cases: one abandoned biogas while the other continues to use it. We present a system dynamics simulation model, associated analyses, and experiments to understand what factors drive transition and sustained use. A central insight of the model is community processes influencing the capacity to solve technical issues. Model analysis shows that families begin to abandon the technology when it takes longer to solve problems. The momentum in the community then shifts from a determination to address issues with the cooking technology toward caution in further adhering to it. We also conducted experiments using the simulation model to understand the impact of interventions aimed at renewing the use of biogas. A combination of theoretical interventions, including repair of non-functioning biogas units and provision of embedded technical support in communities, resulted in a scenario where the community can continue using the technology even after support is retracted. Our study also demonstrates the utility of a systems approach for engaging local stakeholders in delineating complex community processes to derive significant insights into the dynamic feedback mechanisms involved in the sustained use of biogas by the poor.

Introduction technology side there is a vigorous effort to develop cleaner and efficient cooking technologies, and renew- Substantial empirical evidence points to the harm- able energy technologies to reduce exposure to harmful ful environment and health impacts of ambient and emissions from both HAP and ambient air pollution; household air pollution (HAP) [1–9]. Reducing HAP, (2) social, behavioral and public health researchers which impacts almost 41% of the global (mostly poor) have focused their efforts on understanding a range population [10], improves both health and environ- of determinants including technical aspects, which mental outcomes. The UN has therefore committed could drive the adoption and sustained use of these to providing affordable, reliable and sustainable mod- technologies in energy-poor communities. Technolog- ern energy for all as one of its sustainable development ical efforts to develop better cooking alternatives and goals (SDG Goal 7) [11]. Two parallel efforts are under- clean energy systems have received considerable atten- way to address this challenge of HAP: (1) on the tion and support [5, 12]. Commensurate emphasis

© 2018 The Author(s). Published by IOP Publishing Ltd Environ. Res. Lett. 13 (2018) 035010 has not been given to understanding social, ecological use [17–19]. The issue of adoption and sustained use and behavioral factors driving the adoption and sus- of biogas is complex. This complexity arises not only tained use of clean cooking practices in energy-poor from a high number of factors [32] but how they inter- communities [5, 13, 14]. act with each other. Social, economic, environmental Improved cookstoves (ICSs) have received con- and technological factors interact in nonlinear relation- siderable focus as technologies to address HAP in ships to form feedback processes with time delays. The energy-poor communities [15–18]. Most of these ICSs cause of the problem behavior, in this case a persistent are low cost and targeted to poor communities [5, 12]. trend of low uptake or high rate of abandonment of However, adoption and use of ICSs are saddled with clean cooking technologies such as biogas over time, their own set of challenges. HAP exposure-response cannot be attributed to a few independent factors but curves are non-linear in nature [1, 2, 19]. Health rather to how the processes are structured in a set benefits can be derived only at very low levels of of interacting feedback mechanisms. Methods with a exposure [2, 4]. For substantial health benefits from unidirectional approach and inability to incorporate clean cooking, the World Health Organization (WHO) contextual processes are less suitable for developing recommends reduction in PM2.5 exposure levels to insights into problems embedded in complex commu- 35 𝜇gm−3 [20, 21]. Most ICSs exhibit poor perfor- nity systems [33, 34]. A systems perspective is necessary mance against the WHO’s recommended indoor air to understand the dynamic interplay of social, ecologi- quality guidelines (IAQGs) in the actual household cal and technological factors driving the adoption and scenario. Emissions performance of multiple mod- sustained use of clean cooking systems such as biogas. els of ICSs against the ISO’s International Workshop There is a paucity of literature deploying systems Agreement’s (IWA) tiers have shown that none of thinking in exploring the determinants of adoption these could be placed in tier 4 in terms of and sustained use of clean cooking technologies in emissions performance [22]; they are mostly placed energy-poor communities [14, 20, 35]. More specif- in tier 1 and tier 2 [22]. Health-related benefits are ically, these determinants in the context of biogas for thus compromised despite switching to ICSs. Further, these communities have been barely studied, and merit this shift does not insulate communities from relying closer inquiry [20]. Using biogas as a representative on biomass as a fuel. Anthropogenic degradation of clean cooking technology in rural poor communities forests and the drudgery of collecting biomass con- of India, the current study and its results bridge this tinues [23]. Adoption of ICSs also does not ensure a gap in our understanding of the drivers of the uptake complete abandonment of traditional biomass stoves. and sustained use of clean fuels for cooking. With- Use of multiple stoves (also known as stacking) com- out a robust assessment of the determinants impacting bining traditional cooking stoves and ICSs is routine the adoption and sustained use of biogas, the chal- in such poor communities [15, 24–26]. There has lenge for energy-poor communities will persist despite been a recent emphasis to develop strategies to push biogas being a viable technological solution for HAP. IWA’s tier 4 cooking systems in energy-poor com- The study also provides key behavioral insights, which munities. Liquefied petroleum gas (LPG), induction could be tailored and adapted to promote adoption and stoves and biogas digesters are tier 4 clean cooking tech- sustained use of other clean cooking systems in many nologies with emissions below the WHO’srequired energy-poor communities. IAQGs [22]. LPG and induction stoves are mired with multiple issues of affordability and accessibility, especially in poor communities of rural India [27]. Bio- Methods gas digesters (biogas) could provide a viable solution for low-income rural communities, where affordabil- To understand the complex community processes ity and accessibility to LPG or induction stoves is still shaping the sustained use of biogas, our study uses the a challenge [20]. However, adoption and sustainment community based system dynamics (CBSD) modeling (with no stacking) of biogas in these rural communities approach [36, 37]. System dynamics is a computational remain significant impediments [28–30]. modeling approach which focuses on understanding Tricket and colleagues [31] argue that sustaining the relationship between the feedback structure of the evidence-based practices to achieve intended outcomes underlying system and the resulting system behavior is one of the most important aspects of community level [38]. The relationships in the model are defined as interventions. Such interventions are often embed- a system of differential equations and solved com- ded within complex community processes, further putationally. We use Vensim DSS software (Ventana complicating the goal of sustainability. The dynamic Systems, UK) to develop our model and its simulation. interaction of the intervention with community pro- System dynamics has been used in studies of diffusion cesses underscores the need to work collaboratively of innovation [39] and community level interventions with communities and recognize the importance of [40]. their knowledge and involvement in intervention pro- The primary goal of this pilot study was to explore cesses. A collaborative approach also helps to close the the applicability of the CBSD approach in understand- gap between knowledge development and knowledge ing the dynamics of sustained use of clean cooking

2 Environ. Res. Lett. 13 (2018) 035010

technology. Although this letter will describe the Table 1. Demographic and socioeconomic attributes of the two approach, it is important to note that details includ- villages. ing a description of the study design and steps involved Sustained use Abandoned in implementing CBSD have been described elsewhere Number of households 31 189 [37].Thedatacollectedinthisstudyareattheaggregate Total population 120 706 level and do not include any individually identifiable Adults 53 415 information. Therefore, the Human Research Protec- Male 42 217 Female 26 198 tion Office at Washington University in St. Louis has Children 43 291 determined that this study does not involve activities Male 23 171 that are subject to Institutional Review Board oversight. Female 20 120 Caste distribution (Households) Scheduled caste (SC) 018 Data sources Scheduled tribe (ST) 21 143 Other backward caste 028 The model and simulation results presented in this (OBC) paper are based on group model building sessions with Type of house Concrete 13 34 two communities, one of which was able to sustain the Adobe 18 155 use of biogas and another that abandoned the tech- Semi-concrete 00 nology. The two communities are located in Bhilwara Occupation district in rural Rajasthan, India. The local partnering Agriculture 15 136 Agricultural labor 2 4 agency implemented a biogas project with both com- Other labor 13 44 munities at the same time, creating an ideal situation Service 0 4 for a comparative study. These communities are also Self employed 11 (non-agricultural) adjacent to each other and therefore share similar geo- Other 00 physical conditions. The study site is a drought-prone Households below poverty line 26 123 region with an average yearly rainfall of 690 mm. Most of the households in both villages have small land- holdingsanddependonrain-fedagricultureastheir primary livelihood, which is supplemented by wage participants to identify the number of families using labor. Households are dependent on nearby forest for biogas at the time of each event. For every time point, fuelwood, non-timber forest products, timber for con- the facilitator checked whether the data from the imple- struction and fodder for livestock. Table 1 highlights menting agency corroborated the participants’ recall. some of the demographic and socioeconomic charac- At times when there was a discrepancy, the facilitator teristics of the two villages. would ask questions regarding the particular house- Data for the reference mode and the model struc- holds that installed biogas that year. This allowed us ture were collected duringthe CBSD modelingsessions. to develop trends of biogas over time in each commu- Reference mode refers to a graph depicting change in a nity while orienting the group to conversations around variable measuring a behavior of interest over time. biogas installation and use. Instead of presenting the The model structure refers to the feedback mecha- agency-based data to the volunteers we were able to nisms causing behavior change over time. A day long develop the total number of biogas units in a commu- model building session was conducted with groups nity through participatory methods. from each community. Each session had a minimum of ten participants from the community. The number Building the system dynamics model of participants reflects the best practice in group-based During the CBSD process, participants shared a nar- participatory modeling [36].Theparticipantswerenot rative based on their lived experiences with the selected in a representative manner. The field staff rep- technology. In real time, the research team developed resenting the local agency partner visited the villages a model that represented the participants’ narrative prior to the CBSD session and asked for female vol- as a set of feedback mechanisms. These mechanisms unteers who would be interested in taking part in the delineate a set of interrelated hypotheses explaining the study. While the model was built with the women who sustained use of biogas over time. These interacting volunteered, the reference mode and the model were mechanisms, non-linear in nature with time delays, subsequently presented to non-participating members make it difficult to cognitively infer the behavior of of each community, for member checking [41], to the system over time [43]. Therefore, mathematically ensure that the narrative was accurately represented defining the feedback mechanisms to simulate the in the feedback mechanisms. We adapted the ‘timeline’ model is crucial to determine whether the set of mech- exercise used in participatory rural appraisal [42]to anisms identified in the model can reproduce the develop the reference mode. We asked the participants reference modes of interest in this study. to re-construct a timeline based on major events in the The core of the model represents the movement of community’s history. Thereafter, the facilitator asked families through various stages of their experience with

3 Environ. Res. Lett. 13 (2018) 035010

Families with non-functional biogas Rate of abandonment by adopters Complete adandonment Rate of abandonment by stackers

Families using both Families Families using traditional stoves and predominantly using traditional stoves Installation rate biogas Adoption rate for biogas of biogas exclusively biogas + + B1 R1 + Advocacy and Spread of biogas + Promotion through word of mouth Total families + using biogas Installation through advocacy and promotion Installation through + word of mouth + + + Effectiveness of advocacy and Social interaction promotion among community Fraction of interactions members resulting in installations

Figure 1. Stock and flow diagram depicting the different stages of the families’ transition during the adoption process. biogas (figure 1). Families at each stage are represented due to clogged pipes. Participants described a period in stocks, which are variables that can accumulate over between encountering a malfunction and deeming a time, and their values depend on the rate of inflows biogas plant non-functional. The length of this period and outflows. If the inflows and outflows are equal was based on their motivation to use the biogas unit and (i.e. netflow = 0) the level of the stock will be in knowledge of the technology. If the time to solve issues equilibrium. was longer than they were willing to wait, they would The first stock is families using traditional cook- give up and consider that biogas plant non-functional. stoves, which represents the initial number of families in At this stage, the technology was not beyond repair, a community using traditional cooking methods. They and could still be made functional. However, once the transition into the stock of families using both tradi- participants transitioned out of this stage, repairing the tional and biogas as the flow increases, represented in technology was not an option. It should also be noted the installation rate of biogas. This stage is referred to as that families who predominantly shifted to using bio- ‘stacking’ where, instead of replacing the traditional gas might also abandon the technology as it became , families use it simultaneously with a cleaner non-functional over time. cooking stove. In the stove stacking stage, families are Each of the transitions (i.e. flows) is driven by experimenting with and learning about the new tech- a structure of feedback processes. The transition, nology with the security and backup of a tried and tested installation rate of biogas, is driven by two feedback traditional stove. From this stage, they either transition processes. First, through the process of advocacy and to join other families predominantly using biogas or promotion, government and non-government organi- abandon the technology and transition to being one of zations convince families and communities to install the families with non-functional biogas. Table 2 pro- biogas. In our study area, the Foundation for Ecolog- vides an example of how the variable installation rate ical Security (FES) worked with community members of biogas was formulated. to spread awareness of biogas. They even arranged The term ‘predominantly’ highlights the reality for families to visit a nearby village to demonstrate of rural households where traditional cookstoves are how others were successfully using it. Some families never completely abandoned but are used on occasion installed the new technology following such promo- and in emergencies. Households in this state, however, tional campaigns. Second, as families install the new use biogas for their daily cooking. Similarly, the term technology, those without biogas are motivated to ‘non-functional’ highlights that the primary reason for install it through word of mouth. This is a reinforc- abandonment is due to malfunction of the new tech- ing process—as the total number of families using nology. The participants identified various reasons for biogas increases the rate of installation also increases. a malfunctioning stove, from broken valves to water Social networks and contagion are significant driving seepage in the digester, and lack of methane production factors, initially.

4 Environ. Res. Lett. 13 (2018) 035010

Table 2. Description of variables, units, and formulation used in defining the installation rate of biogas.

Variable Definition Units Installation rate of biogas Installation through advocacy and promotion + Installation through Households/month word of mouth Installation through Families using traditional stoves ∗ Effectiveness of advocacy and Households/month advocacy and promotion promotion Effectiveness of advocacy Fraction of households currently using traditional stove that would Household/household/month and promotion install biogas based on advocacy and promotion by implementing (Households agency installed/households reached/month) Installation through word Total families using biogas ∗ Social interaction among community Households/month of mouth members ∗ Fraction of interaction resulting in installations Social interaction among Number of contacts an average household makes with other Contact/month community members households in the community Fraction of interaction The probability of installing a biogas based on contacts made with 1/contact resulting in installations households who have already adopted.

As the number of families using biogas rises, it This is a costly and time-consuming issue to solve. contributes to the knowledge needed to properly main- To reproduce the abandonment behavior, our model tain the technology and troubleshoot problems in a is calibrated by varying the values for initial motiva- timely manner. External recognition of a community’s tion to use biogas and the initial time it takes to resolve efforts to shift to biogas is also an important mecha- issues. nism. External recognition can come from local and Reproducing the behavior from the system is the national government and non-government organiza- first step in gaining confidence in the model. Our tions. Our modeling indicates that this is an important model reproduces the sustained use behavior well. driver in motivating community members to resolve It underestimates the maximum number of families issues and continue using new cooking technologies. using biogas in the village where it was abandoned but Both knowledge of the new technology and external matches the overall pattern of initial uptake followed recognition drive the processes to reduce the number by abandonment. Reproduction of data is considered of issues and the time it takes to successfully resolve a weak test in examining the logical consistency of them (figure 2). the model [38]. Therefore, we performed a dimen- sional consistency test and sensitivity analysis to build Model testing confidence. Testing a system dynamics model involves reproduc- Each variable in the model has a defined unit. ing the behavior of the real system and ensuring the As an example, installation through advocacy and logical consistency of the feedback mechanisms. The promotion is represented as households/month.This goal of testing is to increase confidence in the model variable is a product of families using traditional stoves and uncover errors [38, 44]. In figure 3,weshow and installation through advocacy and promotion. both the sustained use and abandonment behaviors When the units of these two variables, households and over time based on data and then compare them 1/month, are multiplied the resulting unit is house- to our simulated data. The goal for the model is to holds/month. Units on both sides of the equation elicit an underlying structure of feedback mechanisms are consistent. A built-in function tests the dimen- that produce qualitatively similar behaviors to the real sional consistency of all the equations and provides world. In other words, the goal is to reproduce general a list of errors. The final model passed the dimensional scenarios showing either initial uptake and sustained consistency test. use or initial uptake and abandonment. Therefore, There are numerous parameters with uncertain val- statistical comparisons of the data and the simula- ues in the model. An important test is to understand tions are not performed. The variation in behavior the sensitivity of model behavior to the uncertainty (i.e. sustained use versus abandonment) is a result of in the parameters. To conduct the sensitivity anal- changing parameter values that represent two distinc- ysis, we first calibrated the model parameters to tions in the communities. First, firewood was more produce the ‘sustained use’ scenario and varied each accessible in the community that abandoned the tech- parameter (+/− 50%) individually and graphed the nology, which is represented in the model as a driver resulting behaviors. As shown in figure 4, 100% of lowering motivation to solve issues and continue the simulation runs based on the sensitivity analy- using biogas. Second, this community also faced more sis result in a sustained use pattern of behavior. In challenges in the introduction phase. For instance, other words, the model behavior is sensitive to vari- one of the families abandoned their new biogas ation in the parameters. The overall model behavior unit because water had seeped into the digester. (i.e. sustained use), however, is robust.

5 Environ. Res. Lett. 13 (2018) 035010

Time households use biogas Fraction- of families - - abandoning biogas

Time to solve issues - - Families with + non-functional Rate of R5 biogas abandonment by + adopters Improved perceived Complete benefits to reduce adandonment Rate of abandonment abandonment by stackers

Families using Families Families using Perceived benefits Motivation to both traditional predominantly traditional stoves of biogas use biogas Installation rate stoves and biogas Adoption rate for using biogas + + of biogas exclusively biogas + B1 + R1 R4 Advocacy and Promotion Spread of biogas + through word of mouth Solving issues faster Installation through + with enhanced + advocacy and Total families motivation Installation through promotion using biogas R3 word of mouth R2

Solving issues faster Increased installation Knowledge of with gained knowledge through percieved biogas benefits +

External recognition + and encouragement

Figure 2. Simplified model highlighting the feedback loops.

30

25

20

15

Families using biogas 10

5

0 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Time (Years)

Sustained (D)Abandoned (D) Abandoned (S) Sustained (S)

Figure 3. Overlay of simulated results (S) with data trends (D) on cumulative number of families using biogas in sustained use and abandoned communities.

Testing intervention scenarios they would use the technology if it was repaired. Therefore, the repair scenario was tested to under- The value of developing a mathematical model is the stand what would happen if all the non-functional ability to test interventions and examine counterfac- technologies were repaired. The cost or time taken tual scenarios. Such experimentation at a community to repair all the non-functional biogas units was not level can be very expensive or is often not feasible. taken into account. The goal was to test what would A simulation environment provides a platform for happen in the best case scenario where all repairs were quick experimentation and learning that can pave the completed instantaneously. If the outcome was not way for future studies. Three interventions were cho- positive in this scenario, constraints such as cost and sen for testing based on discussions with community time would further diminish the outcome. Similarly, members and the local implementing agency (table 3). the local agency partner mentioned how a commu- For example, community members mentioned that nity becomes motivated by visits from governmental

6 Environ. Res. Lett. 13 (2018) 035010

Parameters combined 50% 100% Families using biogas 30

22.5

15

7.5

0 2004 2007 2009 2012 2014 Time (Month)

Figure 4. Sensitivity of model behavior due to variation in model parameters.

Table 3. Three interventions tested in the model to alter system went up as more families in the community adopted behavior from abandonment to sustained use. and sustained the use of biogas. The intervention Interventions Period Nature of the intervention created additional encouragement regardless of the Repair non-functional 2014–2016 Moves 100% of the number of families using the technology. In the model, biogas families with the ‘external recognition and intervention’ variable is non-functional biogas to changed to one between the years 2014 and 2019. This families using both biogas representsa100%increaseinthe naturalrateofexternal and traditional stove for recognition. the time period The time it takes to solve issues also plays a crucial Provide external Increases external 2014–2019 role in the sustained use of the technology. When issues recognition and recognition and support of encouragement community shift to biogas are not resolved in time the new technology becomes by 100% for the time non-functional and leads to lowering of perceived ben- period efits. Our intervention reduces the time to solve issues, Improve effectiveness 2014–2019 Reduces the time it takes which is possible, according to the female volunteers, of technical support to solve issues such that through technical assistance and as knowledge about problems are solved the intervention becomes embedded in the communi- almost immediately for the ties over time. In the model, the ‘average time to solve time period issues’ is reduced such that issues are solved almost immediately between the year 2014 and 2019. To determine how the magnitude of interventions and non-governmental agencies and universities. The affects the sustained use of biogas, we vary the levels of ‘external recognition and encouragement’ intervention encouragement and technical support of interventions was tested to understand the impact of visits from from non-existent to the highest level (100%). The outside groups. repair interventions and the timing are kept similar to The first intervention repairs all the non-functional previous scenarios. Simulation results (final year: 2030) biogas technologies. The families whose biogas were are exported to Microsoft Excel and organized as a 2 × 2 non-functional and were not repaired, under this sce- table with one intervention as a column and another as nario, receive a one-time intervention transitioning a row. Finally, we developed a surface plot to highlight them to families using both traditional and bio- how the number of families using biogas changes with gas. In the model, the ‘fractional rate of repairing variation in the two interventions, independently and biogas’ was changed to one between the years 2014 in combination. and 2016. The second intervention increases external recognition and encouragement of community and household behavior to shift to biogas. The success- Simulation results ful community, in our group model building session, confirmed that recognition from various governmental As shown in figure 5, repairing all the non-functional and non-governmental organizations increased their biogas technology has an immediate impact as the motivation to keep using biogas. This recognition number of families using it increases. However, the

7 Environ. Res. Lett. 13 (2018) 035010

Figure 5. Impact of combination of three interventions on system behavior. processes involving delays in solving issues that lead community based system dynamics can derive insights to abandonment have not been altered. As issues arise from key stakeholders—women in rural areas. Our and families are unable to resolve them in a timely model highlights the various feedback mechanisms manner, biogas digesters are abandoned. Supplement- salient in understanding the way women eventually ing the repairs with encouragement has some impact, havetonegotiatethemanycomplexitiesoftheinterven- but is unable to qualitatively change the system behav- tion deployed to improve their health and well-being in ior over time. Although the motivation to sustain the the context of their own lives. As a result, even minor use of the technology increases, it still takes families differences in the intervention can result in sustained a long time to solve issues, resulting in the disuse use or abandonment of a technology. One of the advan- of biogas. Only with the addition of technical sup- tagesofusingsystemdynamicsmodelingishighlighting port do we see a shift in shortening the time it takes how similar feedback structures, such as those in these to resolve issues and eventually in the system behav- communities, can result in drastically different behav- ior from abandonment to sustained use. We define iors due to an accumulation of small variations over ‘sustained use’ at thecommunity levelasthecumulative time. number of families using biogas which includes those The model building process and analysis show who have come back to it and households that have that introducing biogas to rural communities is a adopted it for the first time. When the model shows complex process that spans the social, ecological and sustained behavior after the interventions, it is again technological domains [20]. Understanding the struc- showing a combination of families returning to biogas ture of feedback mechanisms across these domains that had once abandoned the technology and the new is critical to enabling rural communities to adopt families that have adopted it for the first time. and sustainably use these technologies. Clean cook- The system behavior of biogas in our simulation ing interventions fail when understood in isolation is insensitive to changes in the magnitude of exter- of household and community dynamics. Engaging nal encouragement intervention (figure 6). The plane women and key community stakeholders to develop of the surface plot on the x-axis is nearly flat. The the initial model improves its representativeness and reduction in time to solve issues as a consequence of creates an environment for sharing insights between improved technical support has a nearly linear rela- researchers, local implementing state and non-state tionship with the outcome but the effect of technical agencies, and communities. The study tested three support plateaus such that a further increase in techni- concurrent interventions and found that technical cal support would not have any impact on number of support and assistance can play a crucial role in families using biogas. shifting households from traditional stove use to adopting and sustainably using biogas digesters. The other two interventions, including fixing disused bio- Discussion and conclusion gas and motivating households to use it without proper technical support, may not yield positive results. Our study is one of the first to develop a dynamic These findings reiterate previous research on factors model of clean cooking adoption, implementation and impacting installation and use of biogas in rural low- maintenance. We also show how a novel approach like income communities. Bond and Templeton [28]and

8 Environ. Res. Lett. 13 (2018) 035010

Figure 6. Surface plot showing the impact of varying two interventions on the outcome.

Lewis et al [20] have noted that biogas offers significant Our findings not only validate the existing line of promise for addressing HAP in rural poor commu- insights from Mirza et al [45], Lewis et al [20] and Bond nities. However, it faces operational and technical and Templeton [28] but also extend these arguments challenges in routine practices [20, 28]. Low aware- by emphasizing that merely encouraging communities ness to fix operational issues on a timely basis and lack to use clean cooking is not sufficient. Time to experi- of technical training (especially for women) exacerbates ment with, understand and experience the advantages the scenario [28, 45]. of clean cooking are also important. Increasing techni- Our findings show that as biogas use overall in a cal knowhow and embedding it within communities, community changes, it also drives the level of techni- particularly in women, ensures that even as problems cal knowledge embedded in that community. As this arise on the technical front, they are resolved in a timely knowledge is embedded, it helps resolve issues. So, manner. Resolving issues in the use of new technolo- in part the level of necessary technical support in a gies not only increases the technical knowhow but also community is a function of how much technical knowl- the confidence to implement and maintain new clean edge has accumulated due to the number of biogas cooking methods. One of the challenges for commu- users. How additional technical support is provided nities to experiment with new technologies including through external funding beyond the already accu- clean cooking is the various sources of risk that are mulated knowledge of biogas use in communities is ubiquitous in their lives. Many communities in low- therefore an important question to examine. We do this and middle-income countries, including those in our through simulating the introduction of technical sup- study that are the focus of clean cooking technology port. Our simulation demonstrates that higher levels interventions, rely on agriculture and informal labor of indigenized technical knowledge is more sustainable markets for their livelihoods. Community members in combination with small amounts of technical sup- are uninsured against the multiple risks these liveli- port from the outside as needed. Our results highlight hoods are subject to, such as monsoons, erratic rainfall that providing timely repairs and technical knowl- and other risks. This inclines such communities to edge (especially for women) can sustain biogas use be risk averse, especially when it comes to invest- in these communities, although funding mechanisms ing in new technology and experimenting sufficiently and willingness to pay among users for such exter- with its use [46]. Therefore, the time to experiment nal technical support need further exploration, and and see the outcomes of a new technology before this is beyond the scope of our model. Anecdotally we committing are significant processes influencing a observed that social networks of women and opinion household’s decision to adopt [47]. Not every shapers in these communities are instrumental in pro- household, however, need depend on their own exper- viding motivation and encouragement to users. Future imentation. They benefit from the experience of other studies should expand the scope of the research to households that enables to them to make a commit- incorporate the role of gender-based social networks ment to new technologies [48, 49]. In a systematic in clean energy transitions. review of the literature on adoption of ICSs, Rehfuess

9 Environ. Res. Lett. 13 (2018) 035010 and colleagues [50] found that social influence was a current work by testing how other variables would major factor in swaying a household’s decision to take impact model behavior. up clean cooking. In our study, we show that house- holds communicate the perceived benefits of the new technology through word of mouth. Such a mech- Acknowledgments anism can both spread benefits or disappointment with new interventions. As we show in our study in Research reported in this publication was supported the community that stopped using, multiple techni- by the Washington University Institute of Clinical cal problems during the early stages resulted in fewer and Translational Sciences grant UL1TR000448 from adoptees. Similarly, Agurto-Adrianzen [51]findthatin the National Center for Advancing Translational Sci- the initial stages of the adoption of ICSs, the number ences (NCATS) of the National Institutes of Health of adopters facing problems with the technology nega- (NIH). The content is solely the responsibility of the tively impacted the household’s likelihood of adoption. authors and does not necessarily represent the official Therefore, in the early stages of the roll-out of clean view of the NIH. We are grateful to the communities cooking, additional effort is necessary to ensure long- that participated in this study and to the Foundation term use of the technology in that community. for Ecological Security for enabling community access System dynamics models provide a simulation plat- and helping to facilitate model building with commu- form to explore scenarios to develop insights about the nity. In particular, we thank Alka, Giridhar, Laxman, system and inform program development. However, Pratiti, Rahul, Shantanu and Vaibhav for their field and the scenarios are based on assumptions about the sys- community support in Bhilwara and Udaipur districts. tem and when these assumptions are no longer true the Thanks to Kelsey Werner and Smitha Rao for a care- scenariosbecomeinvalid.The model suggests that,with ful reading of the manuscript. Dr Peter Hovmand’s adequatetechnical support, sustained useof anew tech- work and his insights on Community Based Sys- nology is possible. However, the current model does tem Dynamics, especially in resource-poor settings, not take into consideration the changing landscape of were crucial in developing this model and analysis. energy technologies. For example, what would happen if the households could access LPG? The community- based approach ensures that the assumptions in the ORCID iDs model closely reflect the realities on the ground. LPG was not a factor for the communities in the study, and Gautam N Yadama https://orcid.org/0000-0002- thus was not included in the model. This does not mean 5907-1132 that in the near future LPG would not be a part of the energy mix of the households in these communities. References In future studies, the boundary of the model can be expanded to include the influence of access to LPG on [1] Brauer M et al 2016 Ambient air pollution exposure household decision making. estimation for the global burden of disease 2013 Environ. Sci. The results from a CBSD approach reflect the Technol. 50 79–88 [2] Burnett R T et al 2014 An integrated risk function for context of the community and the particular energy estimating the global burden of disease attributable to ambient technology used by it. The results of this study may fine particulate matter exposure Environ. Health Perspect. 122 not be readily generalized to other settings. The goal 397 would be to replicate this study in different community [3] Bonjour S et al 2013 Solid fuel use for household cooking: country and regional estimates for 1980–2010 Environ. Health and technology contexts and understand what aspects Perspect. 121 784–90 of the model apply in different settings. For exam- [4] Smith K R et al 2014 Millions dead: how do we know and ple the Urban Dynamics [52] model has been applied what does it mean? Methods used in the comparative risk in several contexts including large cities like Boston assessment of household air pollution Annu.Rev.Publ.Health 35 185–206 to growing towns like Palm Coast, Florida [53]. For [5] Yadama G N 2013 Fires, Fuel and the Fate of 3 Billion: The this model, the general process by which households State of the Energy Impoverished (Oxford: Oxford University transition from using traditional cooking technology, Press) to stacking, and finally to predominantly using the [6] Yadama G N, Schechtman K B, Biswas P, Chalise M and Castro P 2013 Indoor air pollution and respiratory health: a new technology, could be applied in multiple settings. transdisciplinary vision Transdisciplinary Public Health: Similarly, processes like the word of mouth effect are Research, Education, and Practice ed D Haire-Joshu and well established in the diffusion of technology litera- TMcBridevol49(NewYork:Wiley) ture and could be applicable in other contexts. Finally, [7] Smith K R, Mehta S and Maeusezahl-Feuz M 2004 Indoor air pollution from household use of solid fuels Comparative biogas was used as a representative technology to quantification of health risks: Global and regional burden of understand determinants of sustained use or abandon- disease attributable to selected major risk factors ed M Ezzati, A ment of clean cooking systems. There might be other D Lopez, A Rodgers and C J L Murray (Geneva: World underlying factors that were not discussed during com- Health Organization) pp 1435–93 [8] WHO 2014 7 Million Premature Deaths Annually Linked to munity engagement and consequently not accounted Air Pollution (Geneva: World Health Organization) (Accessed: for in the model. Future studies could expand on the 1 February 2015)

10 Environ. Res. Lett. 13 (2018) 035010

[9] Kumar P and Mehta S 2016 Poverty, gender, and [29] Surendra K C, Takara D, Hashimoto A G and Khanal S K empowerment in sustained adoption of cleaner cooking 2014 Biogas as a source for developing systems: making the case for refined measurement Energy Res. countries: opportunities and challenges Renew. Sustain. Soc. Sci. 19 48–52 Energy Rev. 31 846–59 [10] GACC 2011 Igniting Change: A Strategy for Universal [30] Mwirigi J et al 2014 Socio-economic hurdles to widespread Adoption of Clean Cookstoves and Fuels (Washington, DC: adoption of small-scale biogas digesters in sub-saharan Africa: Global Alliance for Clean Cookstoves) areviewBiomass Bioenergy 70 17–25 [11] UN 2015 Sustainable Development Goals (www.un.org/ [31] Trickett E J et al 2011 Advancing the science of community- sustainabledevelopment/) level interventions Am.J.PublicHealth101 1410–9 [12] Yadama P J, Sahu M, Biswas P and Dyda V 2012 Social, [32] Lewis J J and Pattanayak S K 2012 Who adopts improved fuels Economic, and resource predictors of variability in household and cookstoves? A systematic review Environ. Health Perspect. air pollution from cookstove emissions PLoS ONE 7 120 637 e46381 [33] Luke D A 2005 Getting the big picture in community science: [13] Bailis R, Cowan A, Berrueta V and Masera O 2009 Arresting Methods that capture context Am. J. Commun. Psychol. 35 the killer in the kitchen: the promises and pitfalls of 185–200 commercializing improved cookstoves World Dev. 37 [34] Hirsch G B, Levine R and Miller R L 2007 Using system 1694–705 dynamics modeling to understand the impact of social change [14] Masera O R, Saatkamp B D and Kammen D M 2000 From initiatives Am. J. Commun. Psychol. 39 239–53 linear fuel switching to multiple cooking strategies: a critique [35] Lewis P 2012 Who adopts improved fuels and cookstoves? A and alternative to the energy ladder model World Dev. 28 systematic review Environ. Health Perspect. 120 637–45 2083–103 [36] Hovmand P S 2014 Community Based System Dynamics [15] Pillarisetti A et al 2014 Patterns of stove usage after (Berlin: Springer) introduction of an advanced cookstove: the long-term [37] Kumar P, Chalise N and Yadama G N 2016 Dynamics of application of household sensors Environ. Sci. Technol. 48 sustained use and abandonment of clean cooking systems: 14525–33 study protocol for community-based system dynamics [16] Mukhopadhyay R et al 2012 Cooking practices, air quality, modeling Int. J. Equity Health 15 1 and the acceptability of advanced cookstoves in Haryana, [38] Sterman J D 2000 Business Dynamics: Systems Thinking and India: an exploratory study to inform large-scale interventions Modeling for a Complex World (Boston, MA: Glob. Health Action 5 19016 Irwin/McGraw-Hill) [17] Kar A et al 2012 Real-time assessment of [39] Milling P M and Maier F H 2011 Diffusion of innovations, pollution in Indian households due to traditional and system dynamics analysis of the Complex Systems in Finance improved biomass cookstoves Environ. Sci. Technol. 46 and Econometrics ed A R Meyers (New York: Springer) 2993–3000 pp 136–62 [18] Mortimer K et al 2017 A cleaner burning biomass-fuelled [40] Hovmand P S, Ford D N, Flom I and Kyriakakis S 2009 cookstove intervention to prevent pneumonia in children Victims arrested for domestic violence: unintended under 5 years old in rural Malawi (the Cooking and consequences of arrest policies Syst. Dyn. Rev. 25 161–81 Pneumonia Study): a cluster randomised controlled trial [41] Creswell J W and Miller D L 2000 Determining validity in Lancet 389 167–75 qualitative inquiry Theory Pract. 39 124–30 [19] Forouzanfar M H et al 2013 Global, regional, and national [42] Kumar S 2002 Methods for Community Participation: A comparative risk assessment of 79 behavioural, environmental Complete Guide for Practitioners (Leamington Spa: Practical and occupational, and metabolic risks or clusters of risks in Action) 188 countries, 1990-2013: a systematic analysis for the global [43] Homer J and Oliva R 2001 Maps and models in system burden of disease study Lancet 386 2287–323 dynamics: a response to Coyle Syst.Dyn.Rev.17 347–55 [20] Lewis J J et al 2017 Biogas stoves reduce firewood use, [44] Forrester J W and Senge P M 1980 Tests for building household air pollution, and hospital visits in Odisha, India confidence in system dynamics models System Dynamics Environ. Sci. Technol. 51 560–9 ed A A Legasto, J W Forrester and J M Lyneis (New York: [21] WHO 2014 WHO Indoor Air Quality Guidelines: Household North-Holland) pp 209–28 Fuel Combustion (Geneva: World Health Organization) [45] Mirza U K, Ahmad N and Majeed T 2008 An overview of [22] BerkeleyAir 2012 Berkeley Air Monitoring Group: Stove biomass energy utilization in Pakistan Renew. Sustain. Energy Performance Inventory Report Berkeley (Berkeley, CA: Rev. 12 1988–96 University of California Press) [46] Yesuf M and Bluffstone R A 2009 Poverty, risk aversion, and [23] Smith K R and Sagar A 2014 Making the clean available: path dependence in low-income countries: experimental escaping India’sChulhaTrapEnergy Policy 75 evidence from Ethiopia Am.J.AgricEcon.91 1022–37 410–4 [47] Rogers E M and Shoemaker F F 1971 Communication of [24] Ruiz-Mercado I, Canuz E, Walker J L and Smith K R 2013 innovations; A cross-cultural approach Quantitative metrics of stove adoption using stove use [48] Bandiera O and Rasul I 2006 Social networks and technology monitors (SUMs) Biomass Bioenergy 57 136–48 adoption in northern Mozambique Econ. J. 116 869–902 [25] Ruiz-Mercado I and Masera O 2015 Patterns of stove use in [49] Conley T G and Udry C R 2010 Learning about a new the context of fuel–device stacking: rationale and implications technology: pineapple in Ghana Am. Econ. Rev. 100 35–69 EcoHealth 12 42–56 [50] Rehfuess E A, Puzzolo E, Stanistreet D, Pope D and Bruce N G [26] Ruiz-Mercado I, Masera O, Zamora H and Smith K R 2011 2014 Enablers and barriers to large-scale uptake of improved Adoption and sustained use of improved cookstoves Energy solid fuel stoves: a systematic review Environ. Health Perspect. Policy 39 7557–66 122 120–30 [27] Kumar P , Kaushalendra Rao R and Reddy N H 2016 [51] Agurto-Adrianzen M 2009 The Role of Social Capital in the Sustained uptake of LPG as cleaner cooking fuel in rural India: Adoption of Firewood Efficient Stoves in Northern Peruvian role of affordability, accessibility, and awareness World Dev. Andes Munich Contract No.: MPRA paper No. 15918 Perspectives 4 33–7 [52] Forrester J W 1969 Urban Dynamics (Cambridge, MA: MIT [28] Bond T and Templeton M R 2011 History and future of Press) domestic biogas plants in the developing world Energy Sustain. [53] Alfeld L E 1995 Urban dynamics—the first fifty years Syst. Dev. 15 347–54 Dyn. Rev. 11 199–217

11