The physical and financial losses in Nigeria’s power sector (Part III)

•5.11 TWh unbilled, N637.5 billion uncollected

IN Parts I and II, we followed Nigeria’s grid electricity from installed capacity to available capacity and actual generation, using Nigerian Electricity Regulatory Commission (NERC) data from April 2025 to March 2026. Of 13,625 megawatts (MW) installed, only about 5,174 MW was available on average and about 4,312 MW was generated. Part III now examines the electricity that reached the eleven Distribution Companies (DisCos): how much was accounted for and billed, how much of the bills became cash, and why some DisCos perform much better than others.

Electricity received, billed and cash implications

During the 12 months, the DisCos received about 30.31 terawatt-hours (TWh) of electricity at their trading points but billed customers for only about 25.20 TWh. About 5.11 TWh—more than five billion kilowatt-hours, or “units” as consumers commonly call them—did not become billed energy. In other words, for every 100 units received by the DisCos, about 83 units were billed to customers and roughly 17 were not. The unbilled electricity reflects a combination of physical losses through distribution lines and transformers and commercial losses arising from inaccurate metering, unmetered consumption, meter bypass, erroneous billing, poor customer enumeration and electricity theft. Across the four quarterly reports, NERC’s reported values for electricity supplied to the DisCos sum to about ?3.69 trillion. Dividing that value by the 30.31 TWh received gives an implied weighted-average value of about ?121.68/kWh. This is not a wholesale tariff or an NBET invoice rate, but an analytical average derived from NERC’s reported energy and monetary values.

Customer bills totalled about ?3.00 trillion for 25.20 TWh of billed energy, implying an average billed rate of about ?119.07/kWh. Because customers pay different tariffs according to service band and customer category, this is also a weighted system-wide average rather than a uniform retail tariff. Roughly ?688 billion of potential billing value therefore did not become customer bills. Of the approximately ?3.00 trillion actually billed, only about ?2.36 trillion was collected, leaving roughly ?637.5 billion uncollected. The DisCos therefore converted about 81.4 per cent of the value of electricity supplied into customer bills and collected about 78.8 per cent of those bills. Put together, only about 64 per cent of the original value of electricity supplied became collected revenue. Put simply, for every ?100 worth of electricity supplied to the DisCos, only about ?81 became customer bills. Roughly ?17 of the original ?100 was billed but not collected, leaving only about ?64 in cash.

These two stages are combined in a common distribution measure called Aggregate Technical, Commercial and Collection (ATC&C) loss: ATC&C loss = 1 − (Billing Efficiency × Collection Efficiency). Using the 12-month figures, Nigeria’s eleven DisCos recorded an aggregate ATC&C loss of about 35.9 per cent.

How does Nigeria compare?

India provides a useful international benchmark, although the methodologies are not perfectly identical. In FY2024/25, India recorded a national AT&C loss of 15.04 per cent, with billing efficiency of 87.59 per cent and collection efficiency of 97 per cent. The larger difference is in collections: about 79 per cent in Nigeria against 97 per cent in India. Ghana’s two principal distribution utilities report ATC&C losses of about 25 per cent. In Pakistan, K-Electric’s FY2024/25 T&D-loss and revenue-recovery figures imply an AT&C loss of about 22.8 per cent. Reporting systems differ, but all three benchmarks are substantially better than Nigeria’s 35.9 per cent. Nigeria’s distribution losses are therefore not an unavoidable feature of supplying electricity in a developing economy.

DisCos performance classification

The aggregate figure conceals enormous differences among the eleven DisCos. I calculated a day-weighted twelve-month average of each DisCo’s quarterly ATC&C losses and classified them as Strong at 25 per cent or lower; Moderate at above 25 per cent up to 40 per cent; Poor at above 40 per cent up to 55 per cent; and Critical above 55 per cent. These are my analytical classifications, not NERC ratings.

Only two DisCos were Strong: Eko at 15.4 per cent and Ikeja at 18.2 per cent. Three were Moderate: Abuja at 32.1 per cent, Port Harcourt at 36.3% and Enugu at 39.8 per cent. Three were Poor: Benin at 43.1 per cent, Kano at 43.8 per cent and Ibadan at 44.2 per cent. And three were Critical: Yola at 58.0 per cent, Jos at 61.1 per cent and Kaduna at 70.3 per cent.

Six of the eleven DisCos therefore lost more than 40% of the potential revenue represented by electricity supplied to them. Kaduna’s roughly 70% ATC&C loss means, broadly, that only about ?30 of every ?100 of potential revenue became collected revenue.

Why such wide differences?

Metering provides one important clue. As of March 2026, 88.50 per cent of Eko’s active customers and 87.96 per cent of Ikeja’s were metered—the two highest rates among the eleven DisCos. By contrast, the three Critical DisCos had much lower metering rates: Yola 33.16 per cent, Jos 34.58 per cent and Kaduna 36.43 per cent. Nationally, only 59.13 per cent of active registered customers were metered.

Geography and economics also matter. Eko and Ikeja serve densely populated Lagos, with concentrated commercial and industrial demand and a strong network backbone. Yola, Jos and Kaduna cover much larger territories containing many less densely populated and lower-income communities. Lower purchasing power can affect collections, while dispersed settlement requires longer networks and greater infrastructure investment.

But the relationship is not absolute. Ibadan serves Ogun and important industrial corridors yet falls in the Poor category; Benin serves economically important states including Ondo, Edo and Delta but is also Poor. Performance therefore appears to reflect an interaction among metering, customer density, purchasing power, economic activity, distribution-network condition, transmission access and management effectiveness.

Conclusion and recommendations

The forensic chain is becoming clearer. Of Nigeria’s 13,625 MW installed capacity, only about 5,174 MW was available and about 4,312 MW was generated on average. The DisCos then received 30.31 TWh but billed only 25.20 TWh. They issued bills worth about ?3.00 trillion but collected only about ?2.36 trillion. For every ?100 worth of electricity supplied to Nigeria’s DisCos, only about ?64 ultimately became collected revenue.

Nigeria’s 35.9 per cent ATC&C loss compares poorly with India’s 15.04 per cent, K-Electric’s approximately 22.8 per cent in Pakistan and Ghana’s roughly 25 per cent. Within Nigeria, the disparity is equally striking: only two of eleven DisCos were Strong, while six were Poor or Critical. These comparisons show that such losses are neither inevitable nor simply a consequence of Nigeria’s size or development challenges. There is substantial recoverable value in electricity already generated. The remedies should follow the causes: network rehabilitation for technical losses; enforcement against meter bypass and theft; metering and accurate enumeration for unmetered customers; and credible billing, customer service and stronger collection systems.

Accountability must also improve. Under the Service-Based Tariff, Band A customers are promised at least 20 hours of supply daily, Band B 16 hours, Band C 12 hours, Band D 8 hours and Band E 4 hours. Yet the quarterly reports do not provide a simple DisCo-by-DisCo account showing what share of electricity went to Bands A–E and whether each band received its promised hours. While Band A tariffs are much closer to cost-reflective levels, Bands B–E remain supported by Federal Government tariff subsidy. Such reporting would help the public see what service taxpayers are subsidising and whether it is actually being delivered. Nigeria therefore needs a transparent Distribution Recovery Account tracking electricity from each DisCo’s trading point to customers. NERC should require publication, by DisCo, of energy received and billed, technical and commercial losses, billing value, revenue collected, metering rate, ATC&C loss and principal causes. It should also publish, by service band, electricity allocated, actual average hours delivered and the percentage of feeders meeting their service commitments.

That takes us to Part IV, where we follow the money upstream.

After these losses, how much did the DisCos remit to the electricity market, how much did the GenCos invoice, how much did the Federal Government subsidise—and who ultimately carried the burden?

Oyedele should tell us whether Dangote’s tax explanation remains valid under the new tax regime. Adedeji should quantify the tax burden on a bag of cement. Umahi should tell us what became of government’s promised engagement with manufacturers. Darma should explain what expensive cement means for affordable housing. Oduwole should explain where the consumer dividend from decades of cement-industry protection has gone.

And the FCCPC must follow the evidence wherever it leads.

Nigeria has the limestone, the factories and excess installed capacity. Nigerians should not pay more for Nigerian cement than consumers in countries to which we export it.