Band A Is Not Expensive. Generators Are.

7. Why “Band A is expensive” survives the arithmetic

Because the arithmetic is not what people are objecting to.

The objections are real, and they are about service and governance, not about price per unit:

• Reliability gaps. Band A promises 20 hours. Where it delivers 14, the customer has paid a Band A price for Band B service and still has to run the set. That is a contract failure, and it should be enforced as one.

• Estimated billing and metering distrust , decades deep.

• Cash flow. Tokens demand a lump sum up front. Fuel can be drip-fed. For a household without savings, that difference is decisive even when it is expensive.

• The 2024 memory. Band A went from roughly ₦66 to roughly ₦225 in a single step. That is a 240% overnight increase in a line item. People do not process that as a subsidy reform. They process it as a shock, and they are still processing it.

Every one of these is a legitimate grievance. Not one of them makes ₦209.50 a high price for a kilowatt-hour. Fix the metering. Enforce the hours. Do not respond to a service failure by buying the most expensive electricity in the country.

8. One big set, not twenty small ones

There is a second arithmetic mistake sitting next to the first one, and it costs Nigerian clusters more than the tariff debate ever will.

Walk down a line of twenty shops in any Lagos or Aba market. You will hear twenty engines. Each shop bought its own 0.9 kVA set. Collectively, they are running the most expensive electricity system that can be assembled from commercially available equipment.

Four separate economies are being thrown away.

First, the fuel curve. We established it in Table 1: consumption per kilowatt-hour falls with engine size. Twenty sets at ₦2,510/kWh versus one properly loaded 15–20 kVA diesel at roughly ₦540/kWh in fuel and maintenance is not a marginal improvement. It is a different price band.

Second, diversity. This is the one engineers understand, and shop owners never see. Those twenty shops do not all peak at once. The tailor’s iron, the barber’s clippers, and the pharmacy’s fridge compressor hit at different moments. Sum of individual peaks: about 8 kW. Coincident peak: about 5.6 kW, a diversity factor near 0.7. You buy 30% less machine to serve the same customers. Twenty individuals cannot capture this. One shared plant captures it automatically.

Third, load factor. A shared set runs at 60–75% load most of the day instead of 30%. That is precisely the difference between the ₦2,300 row and the ₦540 row.

Fourth, fixed costs spread. One operator, one service contract, one bulk diesel delivery at a negotiated price instead of twenty jerrycans bought at retail with the informal-market markup, one properly earthed installation, one AVR protecting twenty shops’ equipment.

Worked case: twenty shops, 2,080 kWh/month.

Pooling cuts the cluster’s electricity bill by about ₦3.7 million a month — a 71% reduction — without a single kilowatt-hour of grid supply and without anyone changing what they run. That is ₦185,000 a month back into each shop, which in most of these businesses is a material share of profit.

Worked case: a forty-home residential estate, 16,000 kWh/month.

Forty homes, each averaging 400 kWh a month off a 5 kVA petrol set. Individual peaks of about 2.5 kW sum to 100 kW; the coincident peak is nearer 50 kW, because forty households do not iron, pump, and cool simultaneously. Note the installed-capacity line especially.

The estate saves about ₦6.5 million a month, roughly 42%, or ₦162,000 per household — and it does so having bought half as much generating plant. That is the diversity factor turned into cash. Add the things that do not appear in the table: no fuel stored in forty kitchens, no exhaust at forty bedroom windows, and a quiet estate.

The service-charge arithmetic is what actually sells this to a residents’ association. ₦9,064,000 across forty homes is ₦226,600 a month each, prepaid-metered by consumption rather than levied flat. Households that currently spend ₦350,000–400,000 on petrol and servicing see the difference immediately, and the ones who use less pay less — which is precisely what a flat estate levy fails to deliver and why so many estate schemes collapse.

Worked case: an industrial estate, eight firms, 96,000 kWh/month.

Eight light-manufacturing tenants, each running its own 100 kVA set at about 12,000 kWh a month.

The direct saving is about ₦9.7 million a month, or 18% — a smaller percentage than the market or the estate, because these firms were already near the efficient end of the curve. At this scale, the fuel economies are modest, and three other things matter more:

Redundancy becomes affordable. Two 300 kVA sets in N+1 configuration cost less than eight private sets and give every tenant a reliability none of them can buy alone. Today, one firm’s engine failure stops that firm’s line. In a shared plant, it stops nothing.

Gas becomes possible. This is the prize, and it is only available above roughly 500 kW of aggregated load. A gas-fired or CNG-converted plant at this scale delivers somewhere near ₦250–300/kWh — below the pooled diesel figure, below several West African grid tariffs, and within sight of Band A. No single 100 kVA tenant can justify the gas connection, the compression, or the offtake commitment. Eight of them together can. Aggregation is the entry ticket to a fuel that individual tenants simply cannot reach.

The estate becomes a counterparty. Ninety-six thousand kilowatt-hours a month is a load worth a DisCo’s attention. That is a dedicated feeder negotiation, a bilateral contract or an eligible-customer application. Eight separate firms each complaining about their feeder are eight complaints. One estate metering 96,000 kWh is a commercial proposition.

Note the pattern across all three cases: the smaller and poorer the cluster, the larger the percentage saving from pooling. The market shops save 71%, the estate 42%, the industrial park 18%. Aggregation is worth most exactly where it is least often attempted.

Note the ceiling too, because it is the honest part: even the well-run shared plant at ₦737, ₦567, or ₦475 is still two to three-and-a-half times Band A. Pooling is not an alternative to the grid. It is a way of losing less money while you wait for it.

Why it does not happen. Not engineering. Every one of these installations is routine work for a competent contractor. The binding constraints are institutional:

• Collections. One shop that stops paying still gets power unless someone can disconnect it. This is a metering and enforcement problem, and it is solvable with prepaid sub-meters — which now cost a small fraction of the monthly savings above.

• Governance. Someone must own the asset, hold the fuel money, and be accountable for the overhaul fund. Market associations, estate management companies and cooperatives are the natural vehicles; the ones that work have a written tariff and a bank account, not a WhatsApp group.

• Capital. ₦4.5m up front against ₦3.7m of monthly savings is a payback of under two months. Any bank will lend against that, and several already do — the obstacle is that nobody presents the arithmetic.

• Regulation. NERC’s mini-grid framework covers exactly this: registration for small systems, permits above that threshold, with defined tariff-setting and a compensation mechanism if the DisCo later arrives. The rules are not the reason clusters do not consolidate.

And there is a strategic bonus that dwarfs the fuel saving. An aggregated cluster is a creditworthy customer; twenty individual shops are not. A market association metering 2,080 kWh a month, or an estate metering 40,000, can negotiate a dedicated feeder, sign a bilateral contract, qualify as an eligible customer, and hold a DisCo to a service standard. It can also finance solar at commercial scale, where a single shop cannot. Consolidation is how the cluster stops being a price-taker.

9. For commercial and industrial users

Generator power is a hidden tax on unit cost, on export competitiveness and on working capital.

Unit cost first: a manufacturer at ₦500–650/kWh all-in cannot price against a regional peer at $0.10–$0.20. In an energy-intensive process, that gap alone can exceed the whole net margin.

Working capital second, and this is the one finance directors underweight. Diesel is prepaid, held in a tank, subject to shrinkage, and exposed to a naira fuel price that moves with FX. A firm running 2,200 litres a day carries millions of naira of working capital in a tank that a bank will not lend against.

The operating conclusions:

1. Treat generators as standby insurance , priced accordingly, not as a base-load tariff.

2. Pool before you buy. One correctly sized shared plant beats a yard full of private sets — by 71% for a market row, 42% for a residential estate, 18% for an industrial park, and it is the only route to gas. See section 8.

3. If your feeder cannot deliver Band A hours, the economic project is to get onto a feeder that can — dedicated feeder, bilateral contract, eligible-customer route. That project is worth spending real money on. At ₦1.1 billion a year of avoidable cost, the 400 kVA plant above can fund substantial network investment out of one year’s savings.

4. Hybridise for daytime load. Solar’s marginal kilowatt-hour is near zero once the capital is spent. Against diesel at these fuel prices, payback is commonly two to four years. I have no brand to sell you and no view on which panel. I have a view on the arithmetic.

5. Standardising on diesel is not a strategy. It is the absence of one.

10. Externalities, briefly

Noise. PM2.5 in residential compounds. Exhaust at ankle height outside shops where traders sit for ten hours a day. Foreign exchange demand to import the fuel. These are real costs, and they fall mostly on people who did not buy the generator. They are not the argument. The argument is that self-generation is two to ten times more expensive per kilowatt-hour, and it would still be the wrong choice if it were silent and clean.

Box: How to audit your own generator in one week

You do not need my numbers. You need yours.

1. Fill the tank to a marked level. Dip it or weigh it. Write it down.

2. Log running hours. Use the hour meter if you have one, a notebook if you do not.

3. Log the load if you can. A ₦15,000 clamp meter on the output cable will do it. If you cannot, list what was actually running and estimate the watts.

4. Refill to the same mark and record the litres.

5. Divide: (litres × ₦ per litre) ÷ (kW × hours) = your true ₦/kWh.

6. Add 15% for oil, filters and servicing. Add your set’s purchase price divided by its realistic lifetime kilowatt-hours.

Compare the result to ₦209.50. In twenty years, I have never seen this exercise come out in the generator’s favour. I have seen it change several capital budgets within the week.

The final comparison

The rational Nigerian energy stack, in order: grid first wherever the hours exist. Pool your generation if you must self-generate. Solar for predictable daytime load — and for the 0.9 kVA class, solar first, because nothing else is worse than what they are paying now. Generators last, and only as insurance.

I am not telling you the grid is reliable. It is not, and the Band A service standard is honoured unevenly. I am telling you something narrower and harder to argue with: when the kilowatt-hour arrives, it is cheap. Two to fourteen times cheaper than the one you make in your compound — and the multiple is worst for the smallest set, owned by the customer with the least money.

Calling Band A expensive while running a petrol set is not just a political position; it is an arithmetic mistake — and repeated across sixty million households and every industrial estate in the country, it is an arithmetic mistake with macroeconomic consequences.