DRC · Inga Hydropower · Energy · MDB Reform Platform
Does the World Bank Learn from Mega-Project Failures? (Part 6) — The Dam That Does Not Generate
Inga and the World Bank: 40 Years of Engagement. $107 Million Cancelled. 4.3% Disbursed. Rated Highly Unsatisfactory. A New $250 Million Approved — Already Deteriorating.
Inga: The Largest Hydropower Site in the World
The Inga site on the Congo River, 150 kilometres southwest of Kinshasa, has an estimated hydropower potential of approximately 40,000 megawatts — the largest concentration of hydropower potential in the world. The Congo River’s flow is remarkably stable year-round, eliminating the seasonal variability that limits other African rivers. At full development — the Grand Inga vision — the site could generate more electricity than any single power station on earth.
Inga 1 (351 MW) was commissioned in 1972; Inga 2 (1,424 MW) in 1982. Neither has ever operated at full capacity. Decades of deferred maintenance, institutional neglect, and the collapse of SNEL during the wars reduced operational output to approximately 40 percent of nameplate capacity. A 4,000-kilometre transmission line was built to carry power to the Katanga mining region — it operates at a fraction of design capacity. The DRC is a net electricity importer despite possessing the largest hydropower potential in the world.
The modern revival framed Inga as a continental energy platform: cheap baseload hydropower for South Africa (Eskom’s 2,500 MW off-take), Katanga copper and cobalt mines (critical for global battery supply chains), and eventually West and North Africa through transcontinental transmission. The Bank’s 2014 PAD called Inga “a regional game changer” that could “light up the African continent.” That framing — combining infrastructure, regional integration, private finance, and state-building simultaneously — shaped the governance architecture that subsequently collapsed.
The Collapse: S → HU in Eighteen Months
The Inga 3 Basse Chute and Mid-Size Hydropower Development TA (P131027, $107 million, FY2014) was rated Highly Unsatisfactory by IEG — efficacy and efficiency both rated Negligible. Only $3.1 million of the $73.1 million IDA grant was disbursed. Component 1 actual cost: zero. Component 2 actual cost: zero. The project collapsed before substantive implementation began.
| ISR | Date | DO | IP | Key Issue |
|---|---|---|---|---|
| 1 | Jul 2014 | S | S | Startup optimism; institutional design phase |
| 2 | Apr 2015 | S | MS | Early concerns; no outputs completed |
| 3 | Dec 2015 | U | U | ADPI-RDC presidential decrees; governance rupture |
| 4 | Jun 2016 | HU | HU | Political interference; concession crisis |
| 5 | Dec 2016 | HU | HU | Early closure; project collapse |
IEG: “Political interference to private concessionaire selection ultimately led to the cancellation of the project.”
The Governance Story
The Bank attempted to depoliticise a project whose value was fundamentally political. The Bank’s governance model required depersonalised, rules-based concession management: competitive procurement, transparent evaluation, institutional independence. The Congolese political system operates through personalised discretion over concession awards. The Bank’s own SCD states: “The top levels of the state make decisions on concession awards and partnership arrangements.”
Inga’s concession terms would define the allocation of Africa’s largest hydropower resource between South African export (Eskom’s 2,500 MW), Katanga mining companies (copper and cobalt for global battery supply chains), and 70 million Congolese without reliable electricity. The presidency’s decision to create ADPI-RDC within the President’s Office was a reassertion of sovereign authority over a project whose value was fundamentally political — a source of rents, patronage, and geopolitical leverage.
The Warnings Were Known Before Approval
The United States abstained at the Board (March 2014). Treasury warned that “the governance and environmental risks required further mitigation,” questioned whether ADEPI could realistically become operational, and flagged that bidding might proceed before environmental studies were complete. Every concern was validated within eighteen months.
Civil society groups warned. International Rivers, Bank Information Center, and Counter Balance argued that mega-dams historically concentrated political power and generated incentives for opaque decision-making. Their warnings closely paralleled — and predicted — the governance collapse.
The PAD itself flagged the risks, referring to “the inadequate investment and governance environment of the DRC.” The project was approved not because risks were unknown but because transformational ambition justified unusually high tolerance for governance risk. Classic mega-project exceptionalism.
SNEL: The Bankrupt Utility That Inga Depends On
Every Inga scenario assumes SNEL will operate expanded generation, manage transmission, collect revenue, and deliver electricity. The Bank’s own reports document that SNEL monetises less than half of the energy it produces.
36% of production lost to technical and non-technical losses. 49% of installed capacity has never been rehabilitated. 55% of low-voltage customer revenue goes uncollected. The government does not pay its own electricity bills. The energy regulator has not been operationalised. Tariffs are below cost-recovery. SNEL is caught in “a vicious cycle of mounting commercial losses, deteriorating assets and mounting debt.”
The same reform language — commercialisation, governance, tariff reform, institutional restructuring — has been recycled across every Bank engagement for twenty years: PMEDE (FY2007) → EASE ($145M) → AGREE ($600M, currently MS/MS with 18-month delay). The diagnosis has not changed because the problem has not changed.
The Re-engagement: Is This Time Different?
In June 2025, the Board approved a new $250 million Inga 3 Development Program (P506438) — Phase 1 of a $1 billion MPA. $194 million goes to community development, not dam preparation. The language shifted from continental transformation to incrementalism. That is genuine learning.
But the ISR trajectory is already repeating. Implementation progress dropped from S to MS within ten months. AGREE ($600M) reports an “18-month delay in recruiting AECOM for Inga 3 studies.” Phase 1 defers the bankability question to Phase 2.
The Verdict
The Bank failed at Inga because it fundamentally misunderstood the political economy of the DRC. There was never any realistic prospect that the largest infrastructure project in the country — the concession that would determine who controls Africa’s biggest hydropower resource for decades — would be ceded to a developer selected through a World Bank competitive bidding process. In a political settlement where the top levels of the state make decisions on concession awards through personalised discretion, the governance collapse was not an implementation failure. It was inevitable from the moment the project was approved.
The Bank is working on SNEL. The EASE project ($145M), AGREE ($600M), the Foundational Governance DPFs, and the CPF all target utility reform: revenue protection, fraud reduction, prepaid metering, MIS systems, tariff renegotiation, organisational restructuring. These efforts deserve acknowledgment. But they have been running in various forms for twenty years. The 2012 SNEL Performance Contract “did not deliver on expected outcomes and expired in 2016.” The CLR found “no evidence that their efficiency has increased.” AGREE is already at MS/MS with an 18-month delay. The diagnosis is correct. The execution has not matched it.
The sequencing is wrong. The Bank is running Inga 3 in parallel with SNEL reform, betting that SNEL will be fixed in time for Phase 2. This is the same kind of optimistic assumption that has failed repeatedly in the DRC portfolio. The rational sequence would be to demonstrate that SNEL reform has succeeded — that the utility can collect revenue, maintain assets, operationalise the regulator, and credibly commit to a PPA — before committing to a multi-billion-dollar generation expansion that depends on all of those conditions. Instead, the Bank approved a new $250 million Inga 3 that kicks every hard question — the PPA, the developer selection, the off-take framework — to Phase 2.
The new Inga 3 PAD contains no credible plan for resolving the bankability problem by Phase 2. No identified off-taker. No PPA framework. No SNEL reform roadmap with binding milestones linked to Phase 2 activation. No explanation of how the governance dynamics that destroyed the 2014 operation have been structurally addressed. The $1 billion MPA structure creates momentum toward Phase 2 approval regardless of whether Phase 1 demonstrates that the underlying constraints have changed.
Electricity Access: The Human Cost
The number of Congolese without electricity increased from 60.1 million to 70.2 million between 2013 and 2020. National electrification: 21%. Rural areas: 1%. Connected households experience outages averaging more than three hours per day for more than 180 days per year. Inga’s theoretical potential is 40,000 MW. Its operational capacity is approximately 1,700 MW, much of it unreliable. The gap between potential and delivery is the gap between aspiration and institutional reality.
Series: Does the World Bank Learn from Project Failures?
| Part | Country–Sector | Commitment | Status |
|---|---|---|---|
| 1 | Nigeria Water | $1.8bn | Published |
| 2 | Angola DPF | $2.2bn | Published |
| 3 | South Africa Energy (Eskom) | $9.13bn | Published |
| 4 | Ghana FCI | $615M | Published |
| 5 | DRC Portfolio | $6.7B | Published |
| 6 | DRC Inga | $107M + $250M | This paper |
Companion Papers
The Portfolio That Does Not Deliver — 49 projects, $6.7 billion, 6.1% to Satisfactory. The full DRC portfolio record. The companion paper.
Institutional Power Architecture — Economist dominance in Country Management and the DPO instrument bias.
Policy Without Performance — The DPO Incentive Trap. Why budget support fails in FCS environments.
South Africa ESKOM — Medupi and Kusile: the anchor off-taker that was itself in crisis.
The Energy Sector Record — The full energy GP portfolio across Sub-Saharan Africa.