Impacts of Rural Electrification in Uganda 2011-2019

Systematic long-term monitoring of electrification programs is rare. This article presents the results of a six-year monitoring exercise to track impacts of rural electrification in the West Nile sub-region of Uganda. It is based on the publication

Introduction

This article aims to contribute to a better understanding of the challenges and dynamics of the implementation as well as of the strategic design of rural electrification programs in order to maximize development impacts of such interventions. For decades, grid extension has been perceived as the ultimate means for rural electrification, its high costs justified by the assumed welfare gains in rural development. Cost reductions in off-grid solar as well as sobering results of recent impact studies for grid and off-grid electrification emphasize that policy makers should re-consider impact assumptions and cost-benefit rationales for future electrification planning and strategies.

Prepared on behalf of the Ugandan Ministry of Energy and Mineral Development (MEMD) and KfW Development Bank, this report adds empirical evidence to the recent reassessments of grid electrification impacts. It is the result of a rigorous impact monitoring exercise of rural electrification in the West Nile region of Uganda. Over four consecutive monitoring surveys in 2013, 2015, 2017, and 2019, we conducted 9,750 interviews with households, businesses, secondary schools, and health centers, using a double-difference approach (with treatment and control groups) for impact attribution.

The pre-existing isolated network, operated by the local utility West Nile Rural Electrification Company (WENRECo), provided unreliable power supply to about 3,000 customers in three small towns with a peak demand surpassing the available 1 MW heavy fuel oil (HFO) based generation capacity in 2011. The West Nile electrification program invested in the

• construction of hydropower generation capacity (3.5 MW were completed in 2012, another 4.1 MW are planned) and thermal generation capacity (2.6 MW installed in May 2019, planned to be upgraded to 2 x 4 MW HFO generation by December 2019);
• rehabilitation and extension of 487 km distribution network, connecting 60 trading centers and 9 towns as well as improving the operational capacity of the local utility;
• connection of about 12,000 domestic and commercial customers including 29 health centers and 44 schools, as well as the promotion of productive, safe and efficient use of electricity.

Drivers and barriers of grid access

Grid electricity has been made available to about 1.6% of the population of West Nile, and to a significantly higher share of businesses and social infrastructure. But grid customers suffer from low reliability and poor service. 'We consider unreliability of power supply as a major factor preventing intended positive effects of grid accesssuch as investment in productive activities and replacement of generators and fossil fuel use. Poor reliability might also cause losing dissatisfied commercial and industrial customers (accounting for 75% of revenues), which directly endangers WENRECo’s financial viability. Furthermore,the planned future grid extension and densification poses an additional threat to the 'operational viability of WENRECo. We have demonstrated that additional customers are lowering the already poor income per customer ratio and that past subsidized connection offers resulted in a significant share of inactive customers. But future grid densification might even be less successful as expected, since 60% of not connected households state high costs as reason for nonconnec­tion, regardless a standing offer of a 100% subsidized free 1-pole connection, which suggests that costs for inhouse wiring remain a serious barrier for poor households.

Impacts of grid access

Impacts of grid access have been much weaker as anticipated, partially due to much higher off-grid solar electrification levels both in the control and treatment group. Grid access leads to an increased use of electrical lighting and larger appliances in comparison to off-grid solar access, but for rural households the differences are limited. Similarly, grid connected level III health centers and secondary schools show a significant increase in electric lighting and appliance use, but this does not translate into significant impacts on the provision of educational or health services. The difference for larger level IV and V health centers is even less pronounced, as almost all of them had used solar systems and generators already before getting grid connected. It appears that most households and even health centers and schools are able to satisfy their basic power demand based on off-grid solar, provided that solar systems are reliable and maintenance and repair service is available, which so far seems not to be the case. Even for rural businesses in trading centers, grid electricity does not appear to be a game changer; even though more businesses used TVs and fridges, they neither increased business hours or employment nor their turnover. In contrast, businesses in towns increased employment by 22% and more than doubled their turnover in the same period of time. Observed economic effects might be suppressed by low reliability of grid power and the limited business capacity and access to financing in rural areas. But increased competition between rural and urban businesses should be investigated to check for possible crowding-out effects of grid access in rural areas.

Lessons learned

The experiences with this program confirm that grid extension alone cannot reach the aim of universal access to electricity for all by 2030. Despite a massive intervention, only small fractions of the total population of West Nile and of the population within the reach of the electrification corridor, respectively, can be connected to the grid. Even within the electrification corridor, the current free 1-pole connection policy is not reaching the poor, as they still cannot afford the additional cost of inhouse wiring.

The cost of grid extension (without the investments in power generation) has been in the range of USD 1,200 per connection. While this is a moderate value compared to other grid extension programs, we observed that most rural household and business customers use only a fraction of the potential grid power access and accordingly reap little benefits. For these customers, adequate access could be provided by off-grid solar systems at a fraction of the cost. Increasing the cost efficiency of the electrification strategy would release funds that could be used to finance social access programs for poor rural households that are currently not reached, neither by grid extension nor by the commercial off-grid solar market. Such an integrated strategy could also avoid that better-off households are benefitting most of rural electrification programs while the poor are left behind, which is effectively increasing the perceived relative poverty in rural areas.

Although the power utility received substantial support over past years, the performance is still unsatisfying. Especially small utilities struggle to hire and keep qualified staff in remote rural areas. And the initial investment to set up an effective administration and technical operation are high compared to the low power sales and related income. The required institutional support might be higher as often anticipated.

This experience should also caution the enthusiasm of mini-grid deployment in remote rural areas. Mini-grids often require higher tariffs to be economically viable. In absence of larger customers, off-grid solar will in many cases represent the more cost-efficient option.

Another lesson is that impact assumptions should be realistic and not follow simplified linear impact chains. Electricity can greatly increase efficiency of economic activities and social services, but this does not lead automatically to relevant economic or social impacts. Especially rural economic development has complex dynamics that cannot easily be anticipated.

For the impact monitoring, the double-difference approach enabled the contrasting of patterns and trends within the treatment group to the general development trends in the region. We used four data points over a six-year period to make existing fluctuations in variables visible and even trace medium-term impacts. The combination of quantitative and qualitative surveys was crucial to be able to cross-check and discuss quantitative results with local stakeholders.

Recommendations to the power utility

• Increase the reliability of power supply by improving the distribution network, switching procedures, network balance, and monitoring of outages (SAIDI, SAIFI).
• Focus on expanding the large-commercial and industrial customers base, who usually have their peak demand during the day, to increase the power sales and income to customer ratio.
• Improve internal management procedures and record keeping, systematically collecting and processing GIS data and customer information as well as defining internal routine workflows for maintenance and customer care.
• Improve external communication by providing continued training for staff and timely information to customers and local authorities on outages and load shedding as well as changes in tariff structures or procedures.
• Encourage safe and efficient use of power, e.g. by placing posters close to the MCB sockets of the in-house installation that highlight the main safety and efficiency rules in an easily comprehensible way (using pictograms and main local languages).

Recommendations to policy makers

• Investigate reasons and trends of low water levels at Nyagak river and identify feasible remedial action. As the planned additional hydropower plant (Nyagak III) is built on the same river, a permanent drop of water levels would critically endanger the power supply in West Nile.
• Instead of focusing on expensive grid extension only, develop an integrated strategy to reach universal electricity access by grid and off-grid approaches, thereby reducing the need of subsidies for grid extension while shifting such subsidies to social access programs for the rural poor. This would also mitigate political pressure to connect more of the underserved poorer and remote rural households to the West Nile power grid.
• Boost the impacts of grid access on health and educational services by facilitating the connection of the remaining unconnected 25 secondary schools and 13 health centers in the proximity of the grid. For more remote institutions, least-cost options including off-grid solar should be analyzed instead.
• To increase economic impacts, existing commercial activities and capacities in West Nile as well as national and cross-border market potentials and linkages should be assessed to strategically target future support like e.g. vocational training, business development services and SME financing. In the best case, such prior assessment would actually guide rural electrification planning and the cost-benefit analysis of supply options such as grid extension, mini-grids or off-grid solar.