When you see data on people with energy access in the world, they’re based on an astonishingly low standard for “electrification.” The World Bank treats you as electrified if you have access to enough electricity to supply 25 lumens and power a phone for four hours a day.  This is referred to as Tier 1 of electrification: extremely dim light, and the power is out more often than it’s on. The bar is on the floor, but even so, the number of people in Sub-Saharan Africa without any electricity climbed until 2016, since which time it’s held about steady.

Access to quality power — that is, reliable power with voltage within the nominal range — is a bottleneck to global development and health. Today, a lot of money is spent on access projects without little to no monitoring or accountability. A lot of money is wasted on equipment that is bound to break or not work properly due to poor power quality. The good news is we can do better. I think the first step is low-cost, longitudinal monitoring, so that we can see what works and measure our progress toward appropriate, reliable power.

How big is this problem, really?

According to a 1997 WHO report, which continues to be cited — there doesn’t seem to be more recent data on this — 70% of medical equipment in Sub-Saharan Africa is broken or unused. There are many reasons for this, including user error/inadequate training, but one 2010 World Bank report cites a 2007 study of 33 hospitals in 10 developing countries, which found inadequate power supply to be the most common cause of medical device failure.

This is voltage data collected at 27 healthcare facilities in the Democratic Republic of Congo, collected by nLine as part of this study (Miles et al, 2023).  All these healthcare facilities are electrified, but there’s a lot of variety in what kinds of services each is able to provide. Most require diesel generators as backup for when the power is out, or when the voltage from their main power source is too volatile and they need to power sensitive equipment. One facility runs on diesel generator alone, 24/7, which costs tens of thousands of dollars per year (the Catholic church foots the energy bill at this particular facility).

There’s not been enough research yet into the effects of poor power quality on medical equipment, but anecdotally, poor voltage quality — whether spikes, sags, or chronically low voltage — causes equipment failures and degraded ability to deliver medical services. Autoclaves, oxygenators, x-ray machines and infant incubators are just some of the equipment that technicians in the DRC reported as malfunctioning or rendered inoperable by poor voltage. Frequent outages, even short ones, also lead to costly inefficiencies when they, for instance, interrupt lab tests and samples have to be taken over again.

Bad power quality is a symptom of issues within the grid and these issues also lead to higher energy cost and technical loss (energy that is generated by burning fuel but dissipated in the lines and equipment without ever reaching customers to be paid for and perform useful work).  If these issues aren’t addressed, the grid will remain “sick” — prone to reliability issues, frequently requiring equipment maintenance and repairs, unable to support demand growth, and highly inefficient.

Why doesn’t a solution exist yet?

You might be asking, Are there good reasons why markets or governments aren’t already making progress on this problem?

On the monitoring side, the technology to enable large-scale, affordable power sensing hasn’t existed for very long. “Universal” SIMs are a recent development — these are SIMs that work in many different countries and contexts and have enabled the development of remote sensing hardware that doesn’t require a new SIM for every new country you might like to sense in.  nLine and A2EI are two companies that develop and deploy power sensors. Both use universal SIM cards to send data over the cellular network to the cloud. (Disclaimer: I previously worked for nLine and I'm a big fan of what they do.)

Some of the obvious consumers of data like these are regulators. These are the government bodies whose job it is to hold the utilities to account, but they are often under-provisioned and don’t actually have the levers to effect change. On top of this, Western aid organizations like the World Bank hold the purse strings, and market forces don’t really work on the World Bank; they are used to doing things their way, and their way is often based on outdated assumptions or assumptions that were never right to begin with.

To make matters even worse, government-run utilities often don’t have any incentive to improve power, or even deliver power. Often, the utility actually loses money with every kWh they deliver. Utilities in LMICs are used to operating with certain kinds of data and very little in the way of “smart infrastructure.” They take spot measurements, infrequently. And in many cases, they are rightfully distrustful of Western tech companies’ promises. Incorporating new technologies is disruptive — it bears risks, and they have very little incentive to take those risks.

What can a new philanthropist do?

In no particular order, here are some ideas and general directions:

• Start companies! We need technology that supports grid operation under the resource constraints that exist in LMICs. We need new companies focused on optimizing infrastructure upgrades to be targeted and cost-effective, and informed heavily by measurement.
• Maybe it makes sense to come at this from the other direction as well: work on developing technology to mitigate the effects of poor power, e.g., cheaper, better voltage stabilizers that actually work in these contexts, or equipment that can tolerate voltage fluctuations and persistently low voltage. This would be a band-aid, but perhaps still worth working on.
• Go work for the World Bank and change the way they do things.
• Strengthen and arm regulatory bodies to do their jobs more effectively.
• Innovative financing solutions like results-based financing (e.g. for mini- and micro-grid projects, financing for utility operators should be tied to power delivered — like SEforAll’s UEF, which disburses grant payments based on number of connections, and only after the connections have been verified by an independent evaluator).
• More studies like Electrification for “Under Grid” households in Rural Kenya from researchers at UC Berkeley, which found compelling evidence that connection costs are too high for households and businesses within 200 meters of a transformer (we call these “under grid,” as opposed to “off grid”) and suggested financing solutions to take advantage of economies of scale.
• Things I haven’t thought of yet! I would be excited to hear other ideas, or learn about work that’s already happening in this space.

Thanks

Thanks Genevieve Flaspohler, Mohini Bariya, Samuel Miles, Josh Adkins and Noah Klugman, who have changed the way I think about all of this.