Standing in the bustling peri-urban settlement of Ngwerere, where nearly 1,000 Zambians live on the outskirts of Lusaka, you can see the power lines that follow the main road. But the residents here have never had access to grid electricity. That has not stopped enterprising business owners from setting up three video arcades and a movie theater, thanks to a 12-kilowatt solar array and battery bank installed by Standard Microgrid, which also provides power to grocery stores, hair salons, and nearly 100 homes.
Developers, donors, and customers are increasingly interested in the potential for microgrids to provide power to hundreds of millions of people who lack it, particularly in South Asia and sub-Saharan Africa. Grid infrastructure is expensive to construct and often subject to routine load-shedding and outages that can last for days or longer, harming homes, businesses, and public facilities. Increasingly popular solar home systems provide power for common uses such as lighting, mobile phone charging, and even refrigeration, but are insufficient for many commercial applications. Microgrids offer the right combination of affordability, reliability, and capacity to service areas that need more power than a home solar panel can provide, but do not have enough load density for the central grid. So investments in microgrids are on the rise, with over $250 million in public debt commitments in sub-Saharan Africa alone, set to leverage over $4 billion in total investment.
Senior Policy Associate - Energy Access Project, Duke University
PhD Student - Department of Electrical and Computer Engineering, University of Massachusetts Amherst
Assistant Professor - Department of Electrical and Computer Engineering, University of Massachusetts Amherst
Special Faculty - Department of Engineering and Public Policy, Carnegie Mellon University
At the same time, microgrids still face a number of barriers. Capital requirements are high and revenues per customer tend to be low, so most private projects require low-cost capital with loan durations of 10 to 15 years. Government policy does not always favor microgrids; for one thing, the same conditions that make communities attractive for microgrid expansion also make them attractive for the centralized grid, and government planners are typically not obliged to plan around private off-grid developers. So, in addition to the usual risks of investing in emerging economies, private developers risk stranding assets when centralized grids encroach. Nor is public ownership a panacea; community-owned and public microgrids often have similar issues concerning maintenance and long-run sustainability. Rigorous research on how to overcome these barriers is sorely lacking.
Plugging the power gap
To help fill the gap, researchers from Duke University, the University of Massachusetts Amherst, and Carnegie Mellon University are partnering with private microgrid developers; CrossBoundary, a financial services firm with offices in Nairobi; and Energy4Impact, an NGO that helps to develop energy businesses; to evaluate the effects of various business model experiments in several dozen microgrids. The initiative is the Mini-Grid Innovation Lab, the first of its kind. One experiment involves offering low-cost financing for customers to take up appliances for home or business use—from televisions and speaker systems to refrigerators and hair clippers—inspired in part by research demonstrating latent demand for home appliances among rural consumers and similar state-sponsored efforts in the newly electrified rural United States in the 1930s. Unlocking latent demand could increase consumption and average revenue per user, the critical metric driving economic sustainability for microgrid operators.
Another experiment evaluates the effects of subsidies on consumer choice, to see if developers might increase revenues by lowering the price for power. The first results from this study found that for every dollar that customers saved in prices, they spent $0.93 on increasing their energy consumption. That is, despite substantial price reductions, developers saw revenues fall by just 7 percent.
Other interventions in progress will evaluate the effects of high-speed wireless internet, choosing between different metering technologies, and proactively connecting customers to the microgrid rather than waiting for households to sign up.
Results from the first experiment—market-rate loans for households and microenterprises to purchase appliances—also offer important insights. Among the nearly 2,000 customers in the program, about one in six bought at least one appliance, the most popular being televisions, refrigerators, speaker systems, and “decoders” set up to receive satellite television signals. Somewhat surprisingly, differences in income or wealth have little to do with the choice to buy appliances. Instead, prior electricity consumption has the largest impact, and having a bank account also helps. Among residential customers, larger households are more likely to purchase appliances, up to a point: The very largest households are less likely to do so.
Among the households that buy new appliances, median consumption more than doubles immediately (see Figure 1), though the long-run response is more measured. The consumption pattern suggests households may initially use the new appliances heavily, scale back somewhat as they adjust to higher bills, and then eventually readjust to a “new normal” level of usage that is moderately higher than before the intervention.
Figure 1: Median consumption increases after customers receive appliances
Connect and finance, and even the poor will pay
Most importantly, the results suggest there is latent demand among even poor households to use more electricity-intensive appliances, provided they can finance the purchases and have access to electricity at an affordable cost. In our current research, along with P.P. Krishnapriya, we are studying consumption patterns after the appliance loans are repaid, which may help to relieve household budget constraints.
Notably, the increase in consumption is far from uniform, and depends a lot on the appliances that people buy. The biggest increases are among customers who buy refrigerators and speaker systems—average consumption more than triples when the appliances are delivered. By contrast, customers who buy televisions, decoders, and hair clippers increase their daily consumption by much less. People who don’t buy any appliances also consume more. This may be because of spillover effects; customers can still buy appliances from other sources even if they did not under the original program. Alternatively, it may represent consumption growth that would have happened without the program.
A number of questions remain to be answered, not least of which is whether the economics would work out in communities where developers lack access to grants to pay the upfront costs of appliances; for this program, these capital costs were covered by a generous grant from the Rockefeller Foundation. In the absence of a similar grant, it is not clear if developers should divert precious capital to customer loans instead of building more microgrids, or connecting more households. Another unanswered question is whether microgrids enable the creation or expansion of micro and small enterprises.
An idea worth supporting
Nonetheless, the initial results are promising. Programs to increase average revenue per user can bring value for developers seeking to prove their business models in thin markets. Demonstrating the viability of these business models can also help convince policymakers that microgrids are economically sustainable and worthy candidates for supportive policies and programs. The mix of private profit and public support can help developers, funders, and governments build better infrastructure. That, in turn, can serve customers starved of the services that power provides, and improve the development prospects of communities such as Ngwerere.