Catalysing energy access in Africa through smarter energy storage management

Lead Research Organisation: University of Oxford
Department Name: Engineering Science

Abstract

This project enables the remote, automated management of distributed off-grid batteries powering solar home systems throughout sub-Saharan Africa. The research has two objectives: to examine solar home system (SHS) usage data to design smarter appliances for off-grid customers; and to optimise lithium ion battery (LIB) lifetime. M-KOPA collects daily usage data from >500,000 households in sub-Saharan Africa (SSA), however, there is a gap in effective use of the data for product management and design excellence. Using data science tools and machine learning algorithms, M-KOPA will develop approaches to maximise product battery life, and design smarter appliances. These tools will decrease the premature LIB failure rate (one of M-KOPA's greatest challenges). Oxford University will design battery failure prediction algorithms to increase the longevity, effectiveness, and reliability of future LIBs for off-grid customers. Thus, this project targets all three aspects of the energy trilemma: reducing emissions through increased life of LIBs in SSA, cost savings through optimal product design, and security of energy supply from more reliable SHS.

Planned Impact

The off-grid population in sub-Saharan Africa spent approximatively US$14.4 billion on lighting supplied by candles, kerosene and battery-powered flashlights in 2014 (Bloomberg New Energy Finance, UNEP, World Bank). All these sources represent significant damage to the environment and the health of the people exposed to them. M-KOPA is in the final stages of Gold Standard accreditation for carbon offsetting with the offsetting amount being 0.75 tons of carbon through the life of each device (ie, 375,000 tons to date).

It has been estimated that consumers save on average US$3.15 for every dollar spent on pico-solar in Africa (Bloomberg New Energy Finance). Access to adequate energy supplies and economic growth are interlinked. Having access to solar lighting is the first step on the energy access ladder.

However in order to truly benefit from the socio-economic impacts of access to electricity (such as increase health, education, productivity and leisure), it is fundamental to be able to deliver low cost appliances (Global Leap 2016). Mainstream appliances consume too much power to be cost-effectively delivered with a SHS. Therefore, the outputs of this project will contribute to financing and designing smarter, more energy efficient DC appliances (such as TVs, fans, radios, and fridges) for consumers by M-KOPA. Also, by increasing battery lifetime and hence reducing system costs, prices for SHS will be further reduced.

Customers who obtain these SHSs and appliances report more time spent studying by their children, increased incomes from micro/small enterprise activities, savings in expenditure on lighting, cooking, phone charging, travel, and other daily expenditures that are better invested in more productive activities.

In terms of environmental impact, a study has recently shown that kerosene lighting is responsible for 240 million tonnes of annual CO2 emissions (Lighting Global 2014) - equivalent to half of the UK GHG emissions per year (BEIS 2017). Enabling further penetration of M-KOPA's SHSs will contribute to the reduction of these emissions.
Additionally the replacement of traditional diesel back-up generators adopted by customers on the grid by solar systems with storage would contribute further to reduction of GHG. Finally there are other innovative solar based solutions that can also be explored such as solar powered food driers and irrigation pumps.

Publications

10 25 50
publication icon
Howey D (2019) Tools for Battery Health Diagnostics and Prediction in The Electrochemical Society Interface

 
Description We have been investigating data from a collection of solar-powered batteries used for roughly 15-24 months in actual field applications. Based on this dataset, we devised an algorithm for the estimation of battery capacity (state of health, SoH) and remaining useful life (RUL) from noisy voltage and current measurements. We have developed an algorithm to clean up data, removing corrupt data, before undertaking SOH estimates. We also developed a capacity estimation model to predict battery RUL.
Exploitation Route As we devised a novel way to track battery state of health, implementing our findings can result in cost savings by developing smarter battery products that optimise the lifetime of batteries, and therefore offering more accurate pricing for customers as the lifetime cost of the battery decreases. These ideas are being implemented by M-KOPA, the project partner.
Sectors Energy,Environment,Transport

 
Description Failure to forecast future battery degradation and end-of-life in lithium-ion batteries can lead to difficulties in scheduling maintenance and estimating the asset value of battery systems, contributing to high overall system cost. The findings here decrease the scheduling uncertainty for our industrial partner, increase battery lifetime, and reduce the overall system costs.
First Year Of Impact 2018
Sector Energy
Impact Types Societal,Economic