# PPPs: The Contract Inside the Dam

A dam is built once. The contract inside it keeps ticking for decades. Public-private partnerships are not just a financing label. They are a machine that converts a public promise to pay into private capital, and that machine cannot be held to account unless the decisions inside it are published.

By Michael Cengkuru · Interactive version: https://cengkuru.com/posts/the-second-infrastructure/

## The story

A public-private partnership, or PPP, is a long-term contract in which a private company finances, builds, and often operates a public asset, while government or users pay over time. The payment commitment is the spine of the deal. Without a bankable revenue stream, the private capital does not arrive.

Bujagali in Uganda is the case this piece walks. It is a 250 megawatt run-of-river plant. The independent power producer is Bujagali Energy Limited, or BEL, whose shareholders include private investors and the Government of Uganda. BEL sells electricity to UETCL, the public offtaker, under a 30-year Power Purchase Agreement, or PPA, signed in December 2005. The financing requirement was reported at about US$798 million. The plant became operational in 2012 and displaced more than 100 megawatts of diesel generation.

The point is not to cast Bujagali as a villain. The point is to open the black box. Every PPP is a project-finance machine with five decisions that shape what the public pays and who carries the risk.

First, the payment commitment: who pays the private company, and for how long? In project finance, the offtake contract converts a built asset into a long revenue stream the project company can borrow against. At Bujagali, UETCL buys power from BEL under that 30-year PPA.

Second, the financial terms: what makes the payment expensive or cheap? A PPP tariff is not set by supply and demand alone. It is engineered to service the special-purpose vehicle's capital structure: senior debt first, then subordinated debt, then the equity return. In 2018, more than US$400 million of BEL loans were refinanced and maturity extended from 2023 to 2032, reducing debt-service pressure. World Bank completion reporting estimated roughly US cents 5.5 per kilowatt hour cut to the Bujagali tariff for 2018 to 2023, and roughly US cents 2.3 per kilowatt hour to Uganda's end-user tariff.

Third, the guarantee: who protects the lenders if payment fails? An IDA Partial Risk Guarantee protected commercial lenders who provided about US$115.4 million, 16.4 percent of debt finance, against government-related payment failure. Covered triggers included government payment failure, termination payments, political force majeure, change in law, and currency-transfer restriction. Guarantees can lower the cost of finance. They also create public exposure.

Fourth, risk allocation: who carries construction, demand, political, and force majeure risk? At Bujagali, the EPC bid came before full geotechnical analysis. That risk drove a significant construction cost increase. Post-award negotiation raised the EPC price by about 20 percent. Risk that is not priced and published still lands somewhere. Often it lands on the public.

Fifth, the documents and data trail: can a citizen, regulator, or parliament follow the chain from tariff back to clause? OC4IDS is the project-level spine. It can carry contracting processes, finance fields such as interest rate, period, repayment priority, and guarantee-type financing, and links to the contract detail that OCDS-for-PPPs describes. Without that trail, the second infrastructure, the contract inside the dam, stays invisible while the first infrastructure is photographed at inauguration.

In 2023, Uganda's Parliament called for BEL to refund US$342 million for overpaid fees subject to an Auditor General forensic audit, and recommended PPA renegotiation. That sequence only makes sense if the public can see the machine: SPV, offtaker, lenders, guarantee, and documents.

This piece uses high-confidence, citable facts only. It does not invent a tariff decomposition, a debt-service coverage ratio, or a causal claim that the geotech overrun was capitalised into a later tariff line. Where a figure is illustrative, the interactive version labels it.

## The data

| Figure | Evidence state |
| --- | --- |
| Bujagali: 250 MW run-of-river, BEL as IPP, 30-year PPA with UETCL (Dec 2005), ~US$798m financing, operational 2012 | High confidence (IEA; World Bank completion report) |
| 2018 refinancing: >US$400m, maturity extended 2023 to 2032 | High confidence (IEA) |
| Estimated tariff relief: ~US cents 5.5/kWh (Bujagali, 2018-2023) and ~US cents 2.3/kWh (end-user) | High confidence (World Bank completion report estimates) |
| IDA Partial Risk Guarantee for commercial lenders (US$115.4m, 16.4% of debt finance) | High confidence (World Bank completion report) |
| EPC price rise ~20% after post-award negotiation linked to geotech risk | High confidence (World Bank completion report) |
| Parliament 2023: US$342m refund call subject to forensic audit; PPA renegotiation recommended | High confidence (parliamentary record) |

## Sources

- World Bank PPP Reference Guide (PPP definition)
- IEA reporting on Bujagali ownership, capacity, refinancing
- World Bank project completion reporting on finance, guarantee, geotech, and tariff estimates
- Uganda Parliament 2023 record on refund and renegotiation recommendations

## The verdict

The physical dam is only half the infrastructure. The contract machine is the second half. If payment commitments, financial terms, guarantees, risk allocation, and documents are not published in a joinable standard such as OC4IDS, the public pays for a black box.