Research

Grows where the grid hasn’t.

Power from waste

Africa's sugar mills, landfills, sewage works, cement kilns, crop residues and city waste could make about 5.4 GW of power that is not built yet, much of it with heat to spare. We screened 1,321 sites and the residues of every country across all 54 African countries for what each could produce, what it would cost and what it would do for carbon emissions. Sugar mills come first: the biggest block, the cheapest power, and the fuel is already on site.

5.4 GWof power from waste not yet built across Africa in our central case (3.2 to 8.8 GW), about 35 TWh a year
1.1 GWfrom upgrading the boilers at 100 sugar mills in 30 countries, at about US$47 per MWh with no fuel bill
23 Mtof CO2e a year avoided in the central case. Landfill gas gives 31% of it from 4% of the energy
19 of 168large cities have waste dry and plentiful enough to burn for power, 15 of them in North Africa
What we found

Mills first, methane second, burning last

  • Sugar mills are the best place to start. Every mill already burns its bagasse for steam and power, but most use low-pressure boilers that make just enough for the mill. Modern high-pressure boilers and turbines could export about 1.1 GW more (0.8 to 1.8 GW) from 100 mills in 30 countries. South Africa, Egypt, Kenya, Eswatini and Uganda lead. The fuel is free and on site, so the power costs about US$47 per MWh.
  • Landfill gas is mostly about methane. The 308 dumpsites and landfills in our data could run only 160 MW of gas engines, but capturing their methane avoids 7.2 Mt of CO2e a year, nearly a third of the total. At US$12 a tonne, the carbon credits are worth about US$68 per MWh, more than the power costs to make.
  • Burning city waste rarely works. African household waste is mostly wet food scraps. Only 19 of the 168 cities with more than 300,000 people pass a basic test of heating value and volume, and 15 of them are in North Africa, where waste is drier. Even there, power costs about US$190 per MWh before the city pays a gate fee. Digesting separated food waste suits the rest of the continent better.
  • Crop residues are big but scattered. Rice husk, oil palm residue, sawmill offcuts, maize cobs and nut shells add about 2.1 GW, led by Nigeria, Côte d'Ivoire, Cameroon and Ghana. They are counted by country, not by site, and work best at the mills and sawmills that need heat themselves.
  • Carbon follows the grid. South Africa (4.9 Mt), Egypt (4.0 Mt), Nigeria, Algeria and Morocco hold most of the carbon value because their grids burn coal or gas. On the hydro grids of Ethiopia, DR Congo, Zambia and Uganda, only methane capture earns real carbon money.
Sugar mills

1.1 GW waiting in the boiler house

A tonne of cane leaves about 280 kg of bagasse. Old 20 to 30 bar boilers burn it to make the mill's own steam and power and export about 10 kWh per tonne of cane, if anything. High-pressure boilers (65 to 110 bar) with condensing-extraction turbines export 60 to 120 kWh per tonne. The chart shows the extra export power by country, from that upgrade alone.

Bars: central case (80 kWh per tonne of cane, 55% capacity factor). Lines: 60 to 120 kWh per tonne. Mauritius is left out: its mills already export about 100 kWh per tonne. Cane per country is FAOSTAT raw sugar output divided by a 10.5% sugar recovery, split across mills by size; mill-level crush figures replace this in the next edition.

Why mills come first

  • The fuel is already there. No collection, no trucks, no fuel contract. Bagasse is free at the mill gate.
  • The boiler house and the grid line exist. Most mills already run a turbine and connect to the grid; about half of the mills in our data are within 10 km of a mapped transmission line.
  • It pays the mill. Sugar prices are weak and imports are rising. Selling power gives mills a second income in the months they crush, and with stored bagasse or cane trash, beyond them.

How it compares

Our central case for South Africa's 12 mills is 255 MW (180 to 400 MW). The South African Sugar Association puts the industry's potential at about 700 MW, which assumes more cane trash, all 14 mills and running beyond the crushing season. Eswatini's Ubombo and Simunye mills already sell power to the grid, and bagasse made about 9% of Mauritius's electricity. Where mills have already upgraded, we count only what is left.

Map

Where the potential is

Each dot is a site, sized by the power it could add in our central case. Tap a dot for its range, cost, carbon and nearest heat user. Crop and sawmill residues are not on the map because we only have them by country.

Sites from Climate TRACE (landfills, sewage works, feedlots, cement and steel plants), Global Energy Monitor and OpenStreetMap (sugar mills) and World Bank What a Waste with Natural Earth (cities). Country outlines: Natural Earth.

By country

The 20 countries with the most potential, central case, MW not yet built. Known landfill gas plants are taken off.

Cost

Mill, kiln and landfill power beat industrial tariffs

Levelized cost of power by technology, against the middle half of industrial tariffs in the 26 African countries in our tariff tracker. The ring shows the cost after carbon credits at US$12 a tonne.

Central costAfter carbon credits at US$12/tMiddle half of industrial tariffs
  • Bagasse and waste heat cost US$47 to 50 per MWh. No fuel bill, so the cost is mostly capital. They pay best as self-supply that cuts the plant's own power bill, with any surplus sold to the grid.
  • Landfill gas costs US$55 per MWh, and its methane credits alone are worth more. Carbon finance, not the power price, carries these projects. Most dumpsites need capping and gas wells first.
  • Digesters cost US$97 to 124 per MWh for sewage sludge, feedlot manure and city food waste. They pay where power is dear, or where they save sludge or landfill costs.
  • Incineration costs about US$190 per MWh before a gate fee. It only stands up where a city pays to get rid of its waste, as cities in Europe and Asia do.

Real cost of capital 10% (8 to 14%), lives of 15 to 25 years, residue fuel at US$20 a tonne delivered, in 2025 US dollars. Grid connection, land and permits are not included. IRENA puts the 2024 world average for new bioenergy power at US$87 per MWh.

Carbon

Methane does the heavy lifting

Carbon avoided a year in our central case, split into what comes from displacing grid power and what comes from methane that is no longer released. Captured methane counts 27 times its weight in CO2.

Grid power displacedMethane avoidedFossil CO2 from burning plastic

Grid factors: Aloe's combined-margin factors for 2024 south of the Sahara; 2024 fleet emissions over generation in North Africa. Factors fall as grids add wind and solar, so carbon from power shrinks over time; methane does not. Heat sold to a neighbor would add more where it replaces coal or oil steam (not counted).

Cogeneration

Who takes the heat

Engines and boilers make more heat than power: about 2.3 GW of heat at our sites. Mills and kilns use their own. For the rest, a buyer has to be close: steam travels about 2 km, hot water about 5 km. We found an industrial heat user within 5 km at 89 sites, together able to sell about 180 MW of heat. The 15 largest matches:

Heat sourceCountryTechnologyHeat, MWNearest heat userDistance, km

Heat users in our data: cement, steel, aluminum, pulp and paper, chemical and food plants, refineries and sugar mills. Cold stores, breweries, abattoirs, dairies and hospitals are not yet in it, so real matches will be more. Straight-line distances between site centers.

Residues

Crop and sawmill residues

Rice husk, oil palm residue, sawmill offcuts, maize cobs, groundnut and cashew shells, coffee husk and cotton gin trash: about 2.1 GW (0.9 to 3.9 GW) if the share that collects at mills and sawmills were burned in small steam plants. Most cost US$110 to 125 per MWh, so the best projects are at a rice mill, sawmill or ginnery that also needs the heat.

ResidueMW, central

Countries with the most

CountryMW, central (range)Main residue

From FAOSTAT crop and sawnwood output (2022-24), residue ratios and a share that is realistically collected (for example 50% of rice husk, 8% of maize cobs, 15% of cashew shells because most nuts are shelled abroad). All countries (CSV).

Every site

Find a site

1,321 sites in 50 countries. Click a column heading to sort. MW is the power not yet built in our central case, with the range from our low and high cases. Confidence is low where the quantity is estimated (for example cane split across a country's mills) and medium where it comes from a site record.

SiteCountryTechnologyMW (range)Cost, US$/MWhkt CO2e a yearGrid line, kmHeat user within 10 kmConfidence

Download everything: sites (CSV) and residues by country (CSV), CC BY 4.0. Grid distance is to the nearest existing line in our atlases (World Bank 2017 lines in North Africa).

What helps

Why it can work

  • The waste is already collected. Mills, sewage works, landfills and kilns gather their fuel in one place every day. Most power projects have to build that supply chain.
  • It runs day and night. Unlike solar and wind, waste plants make firm power, and mills and landfills can follow demand with stored fuel or gas.
  • Methane is a cheap climate win. Capturing landfill methane is among the lowest-cost ways to cut emissions, and carbon markets pay for it.
  • Industry wants the power. Cement plants and mills that make their own power ride through load shedding and pay less for the power they still buy.
What blocks

Why it might not

  • No one buys the power. Mills can make more than they need, but without a power purchase agreement or wheeling rules the surplus has nowhere to go. That is the main reason South Africa's mills have not upgraded.
  • Wet waste and weak contracts. Addis Ababa's Reppie incinerator has run at less than half its planned output because the city's waste is mainly wet and organic, and the city pays the utility for the shortfall.
  • Dumpsites are not landfills. Gas can only be collected from capped, managed cells. Most African dumpsites need years of work and a city partner first.
  • Season and supply. Mills crush for six to nine months a year; residues depend on harvests; cane supply has fallen at several mills. Plants need stored fuel or a second fuel to run all year.
Reference

Method, sources and updates

How we calculate it

Sites. Landfills and dumpsites, sewage works serving at least 100,000 people, feedlots with at least 10,000 head, cement kilns and steel plants from Climate TRACE (2024). 103 sugar mills compiled from Global Energy Monitor's power tracker, company sources and OpenStreetMap. Cities over 300,000 people from Natural Earth and Aloe's atlases.

Quantities. Cane from FAOSTAT raw sugar (2021-23) at a 10.5% recovery; landfill methane from Climate TRACE emissions; sewage at 25 kWh per person a year; feedlots at 0.15 m3 of methane per head a day; city waste from World Bank What a Waste per-person rates and composition; clinker at 75% of cement; residues from FAOSTAT and residue ratios.

Power and heat. Mill export 80 kWh per tonne of cane (60 to 120) less 10 today; landfill gas 50% collected (30 to 70%) in 38%-efficient engines; incineration only where waste reaches 7 MJ/kg and 150,000 tonnes a year; cement waste heat at 30 kWh per tonne of clinker. Each site has a low, central and high case.

Cost and carbon. Levelized cost from capital cost, operating cost, a 10% real cost of capital and capacity factor. Carbon: power times the country's grid factor, plus captured methane at 27 times CO2 (IPCC AR6), plus organic waste kept out of dumps, minus fossil CO2 from burning plastic.

Screening, not a feasibility study. Quantities are estimated from public data, existing projects are only partly netted off, and costs leave out grid connection and land.

Sources

Climate TRACE, asset-level emissions, 2024 (CC BY 4.0); Global Energy Monitor, Global Integrated Power Tracker, September 2026 (CC BY 4.0); FAOSTAT crops, livestock and forestry production (CC BY 4.0); World Bank, What a Waste global database (CC BY 4.0); Natural Earth; OpenStreetMap contributors (ODbL); Ember yearly electricity data (CC BY 4.0); Aloe Energy grid emission factors and tariff tracker; World Bank Africa electricity grid 2017; IPCC AR6 (methane GWP); IRENA Renewable Power Generation Costs in 2024; South African Sugar Association via Engineering News (September 2026); GAIA report on the Reppie plant (2025); SACREEE (Mauritius electricity mix).

Public data only, indicative, not for investment decisions. The full method and source list are in the PDF note.

What changed

We rerun this twice a year, when Climate TRACE, FAOSTAT and Global Energy Monitor publish new data.

  • November 2026 - First edition: 1,321 sites and country residue totals across all 54 African countries, with cost, carbon and heat matching.

Get these changes by email

Once a month we send one short email with what changed across all our atlases and pages, each change with its source, plus new research and open data.

One email a month. You confirm by email and can unsubscribe at any time. What you get · Privacy. Protected by reCAPTCHA (Google Privacy and Terms).

Quote or embed it

Free to quote and embed with credit (see our data and terms). To embed the page:

Cite it

APA, with the date you viewed it:

BibTeX:

Common questions

What is cogeneration?

Making electricity and useful heat from the same fuel at the same time, also called combined heat and power (CHP). A sugar mill is the classic case: its boilers raise steam that drives a turbine for power, and the leftover steam boils the cane juice. Using both can turn 70 to 80% of the fuel's energy into something useful, against 25 to 40% for power alone.

Why start with sugar mills?

Because the fuel, the boiler house, the staff and usually the grid connection are already there, and the upgrade makes the cheapest power on this page. The missing piece is almost always a buyer: a power purchase agreement, a feed-in tariff or wheeling rules that let the mill sell its surplus.

Is burning waste good for the climate?

It depends on what the waste is and what the power replaces. Burning wet food waste makes little power, and burning plastic releases fossil CO2. On Kenya's or Angola's clean grids, an incinerator adds emissions. Digesting food waste or capturing landfill gas usually does more for the climate.

Why are crop residues not on the map?

We only have them as country totals from FAOSTAT harvests, not by mill or sawmill. Mapping rice mills, palm oil mills and sawmills is on our list for the next edition.

Is this a forecast?

No. It is a screening of what the waste could produce, with a low, central and high case. Whether a project happens depends on a buyer for the power, finance and the site itself.

Can I use the data?

Yes. The site table and the residue table are free to download under CC BY 4.0, including for commercial work. Please credit Aloe Energy and the sources: Climate TRACE, Global Energy Monitor, FAOSTAT and the World Bank.

Spotted an error, or something missing on this page? Report an issue or send feedback