To be or not to be: On system dynamics and the viability of mini-grids in rural electrification

Doctoral thesis, 2018

One to two billion people are expected to receive electricity access in developing countries in the coming decades. Many of these people will live in rural areas in developing countries where the existing grid will not be able to reach. These people will therefore rely on off-grid technologies to gain electricity access. One off-grid technology that is expected to grow significant is mini-grids. In addition, the number of mini-grids in developing countries are expected to growth significantly. Furthermore, renewable energy sources are increasingly used in mini-grids, putting larger emphasises on dimensioning and management of the technical system. However, previous experiences with mini-grids in rural electrification have been mixed, and many systems have failed or been abandoned prematurely. The many interactions between technical, operational and social elements make it difficult to attribute the failures to specific causes.

The main purpose of this thesis is to investigate why mini-grids in rural electrification fail. The investigation focuses on reliability of electricity and how it is impacted by and impacts operation of mini-grids, the technical system and the community. The investigation is made through the implementation of system dynamics and load assessment.

Rural electrification consists of many actors with different goals, it concerns the behaviour of people and is affected by technology. As, such, formulating relevant problems in rural electrification is difficult. As shown in Paper II, qualitative system dynamics can aid the process of tackling this complexity and therefore also in formulating problems. Results from Paper IV show that initial dimensioning of mini-grids is important for long-term viability. However, the dimensioning is dependent on estimations of electricity usage or electricity usage in similar areas, which are often done through collection of data through interviews. As shown in Paper III, interview-based load profiles might not be an accurate estimation of measured load profiles. Thus, estimates from interview-based load profiles might provide misleading estimations resulting in non-optimal sizing. Results from Paper I show that long-term reliability in mini-grids is affected by operational practices and community behaviour. Even though poor reliability is associated with the failure of specific components in the technical system, they are subject to operational practices and are thus influenced by the overall functioning of a mini-grid. As such, long-term reliability in mini- grids needs to be considered from a system perspective.