For the 40% of the world's families living in energy poverty today, energy services are provided almost exclusively by the same three-stone fires that have been used for millennia. The pollution from the pervasive use of these fires represents the second leading cause of death for women worldwide and contributes significantly to local and global climate change. Improving access to clean energy services can facilitate improved health and livelihoods and serve as a precursor to other economic and social development. Yet within these diverse, complex, and highly-localized communities, the most effective strategies to provide clean energy are not clear; and success of programs to provide technologies such as biomass cookstoves or subsidize fuels such as LPG or electricity has often been limited. This is because an energy carrier or conversion technology is only a small component of a much larger energy system that includes a complex set of needs, constraints, and other variables at the household, community, and global scales. Within this system exists a range of technical, economic, social, and environmental objectives that often conflict between these scales to create an imbalance between stakeholders; and outcomes vary widely based on technology design choices and local conditions. As a result, development of effective solutions requires a clear understanding of the direct and indirect impacts of design decisions that are rooted in the fundamental interactions between energy, the environment, and people.

In order to assist in understanding these interactions in a systematic fashion, this dissertation develops a probabilistic unified modeling approach that seeks to facilitate energy system design by predicting outcomes in terms of a set of multi-disciplinary considerations and objectives. This approach incorporates a large parameter space including local energy needs, demographics, fuels, and devices to create a comprehensive analysis of potential strategies in terms of a range of technical, environmental, economic, and social outcomes. While recognizing that there is no single 'best' solution, this methodology allows the designer to investigate and understand trade-offs between conflicting and competing objectives, the effects of usability and multi-functionality, sensitivities of input parameters for identification of prominent and critical factors, the impacts of uncertainty in decision-making, and the potential for compromise and integrated strategies that provide sustainable and effective energy services.

The model is used to explore a number of scenarios to provide energy services in a remote off-grid village in Mali for which detailed measures of disaggregated energy use are available. In addition to detailed analysis of the baseline situation, strategies investigated include the introduction of (1) general improved biomass cookstoves, (2) advanced biomass cookstoves, (3) communal biomass cookstoves, (4) LPG cookstoves, (5) solar water heaters, and (6) community-charged solar household lighting. Following this and other analyses, an integrated strategy for energy services is developed.

The results show that the factors with the largest impact on the outcome of a technology strategy include the rate of user adoption, value of time, and biomass harvest renewability; in contrast, parameters such as cookstove emission factors may have less impact on the outcome. This suggests that the focus of village energy research and development should shift to the design of technologies that have high expected user adoption rates. That is, the results of this study support the hypothesis that the most effective village energy strategy is one that reinforces the natural user-driven process to stack technologies while moving toward efficient and convenient energy services. A comprehensive strategy that provides the current state-of-the art technologies to optimally meet each specific energy need in the Malian village with a population of 770—including advanced cookstoves, LPG cookstoves, solar water heaters, and solar battery lighting systems—is expected to annually create 2.5 TJ of energy savings, 500 metric tons of CO2e savings, a 40% reduction in health risk, and offer substantial improvement of quality of life. Moreover, this strategy will reduce operating costs to the users including time by an estimated $1,000 (US) each year. Such a strategy is expected to cost $12-$13 per person per year to purchase and maintain the necessary technologies if supplied by outside financing, a figure which might double or triple when implementation costs are included. This is a relatively small expense in comparison to the projected cost of $110 per person per year to provide the necessary agricultural, health, and educational inputs needed for the Millennium Villages, a figure reported to be well within the range committed by international aid organizations.