A methodology for designing electrification programs for remote areas

Abstract

One of the UN Sustainable Development Goals is the supply of sustainable energy even where no electrical grid is available. The photovoltaic rural electrification programs are the most common systems implemented in remote areas, especially in developing countries. These programs include the systems installation and their maintenance for a given period. Installation costs and even spare parts costs over time are usually well estimated. However, design and cost estimation of the maintenance systems is a difficult task, whose wrong management has often resulted in the failure of these electrification programs. In this work, a methodology for designing maintenance systems and estimating costs is presented. The methodology includes a mixed integer linear programming model and a rule based expert system. The mathematical programming model allows obtaining the optimal size and accurate cost estimation of a maintenance system, based on precise information about the installed systems. This model is calibrated and validated with real running programs and will be used to get an enlarged data set of simulated cases if needed. The rule based expert system is obtained from the data set applying classification and regression methods with general information about the region and program to be run. It can be used for designing programs or for companies making decisions about being involved in a program to be developed. The methodology has been applied to real Morocco programs.

Luces Nuevas Experience Lighting Rural Bolivia: A Way to Reach SDG 7

“…Another question that should be considered is the maintenance costs. To contribute to their estimation a model has been built and applied in Morocco [17].…”

Dimensionality reduction for multi-criteria problems: An application to the decommissioning of oil and gas installations

“…Later on, Hillerman et al (2017) proposed a ML approach to identify suspicious claims from healthcare providers and employed MCDA to rank the suspicious claims in decreasing order of auditing priority. Other applications include the combination of MCDA and cellular automata to predict land use dynamics (Quesada-Ruiz et al, 2019), the use of ML to estimate the parameters of a multi-objective optimisation model employed to design electrification plans (León et al, 2019), as well as a customer satisfaction prediction model that uses MCDA to aggregate individual preferences and employs data-mining techniques to estimate missing individual evaluations.…”

Editorial: OR for sustainable development

“…Green technologies count among them photovoltaic, wind power, hydrogen or biomass. The paper A Methodology for Designing Electrification Programs for Remote Areas (Leon et al 2019 ) aims to provide a tool that could be used by the photovoltaic rural electrification (PVRE) programs commonly implemented in remote areas, especially in developing countries. Citing the unforeseen high costs of maintaining the system that cause private energy service companies to abandon these projects, the authors offer a way of estimating maintenance costs and fee-for-service rates through the use of a combined mixed integer linear programming and rule-based expert system methodology.…”