Anonymised live project exposure · Southern Africa · Utility solar IPP

The headline tariff looked investable. The real question was how much revenue survived the contract chain—and whether the debt case could live with what remained.

Felix supported model review, financing analysis and bankability work for utility-scale solar projects using a private C&I offtake route, a grid and wheeling counterparty, an operational agent and a payment-agent structure. Client, project and counterparty identities are withheld here.

~50 MWac / ~58.5 MWdcP90 revenue case25-year PPA structureCFADS · DSCR · LLCR · reserves

The mandate

Reconcile the model with the transaction—not the other way round

The projects were being advanced through a multi-buyer private-power structure. A generation model could show attractive returns while still omitting the commercial deductions and contractual conditions that determine cash available for debt service.

The work therefore moved between the financial model, technical yield material, EPC and BOQ assumptions, draft PPA terms, wheeling and grid questions, operational-agent terms, payment security and financing requirements.

Analyst note

The tariff was never the answer on its own. The decision depended on the net tariff after wheeling, balancing, trading, losses, taxes and agent fees—and on who carried non-payment and curtailment when the structure failed to perform.

What Felix was responsible for

  • Rebuild and review revenue, CAPEX, OPEX, debt and equity logic.
  • Test P50/P75/P90 generation and downside cases.
  • Reconcile BOQ and project-cost changes into model returns and coverage.
  • Translate draft contract provisions into model questions and bankability risks.
  • Prepare decision notes, investor materials and lender-oriented questions.

What changed during the work

The model was not static; commercial and cost changes had to flow through every return and covenant

One project iteration moved from an earlier cost and tariff basis to a revised BOQ and a 6.5 USc/kWh tariff case. The change improved project IRR, equity IRR and minimum DSCR, but did not remove the need to close the offtake, wheeling, payment-security and termination package.

Model issueWhy it matteredReview treatment
Energy caseDebt sized on an optimistic yield can create false covenant headroom.P50, P75 and P90 cases were separated; lender discussion centred on the conservative case.
Developer premiumIncluding sponsor value in senior-debt sizing overstates debt-eligible uses.Developer premium was separated from project cost and funded as a sponsor/equity item.
Net tariffHeadline PPA price can hide agent, wheeling, balancing, trading, loss and tax deductions.Each deduction was required as an explicit input or unresolved information gap.
Repayment shapeA flat amortisation profile can force a weak minimum DSCR despite adequate lifetime cash.Repayment sculpting and covenant headroom were tested against CFADS.
Grid interruptionSolar availability does not create revenue when the network cannot take power.Curtailment, deemed energy and network responsibility were treated as contract and model issues.
The point

A stronger IRR after a BOQ update did not make the project “bankable.” It made the unresolved contract package more important because the investment case was now worth protecting.

Questions taken from model to contracts

Bankability review

Payment security
Is the LC funded, on demand, replenishable and available to lenders after a missed payment?
Termination
Do offtaker and political-force-majeure termination payments cover debt, break costs and equity?
Curtailment
Who pays for network-driven curtailment and how is deemed energy calculated?
Allocation
How is generation allocated across buyers, and who bears shortfall, imbalance and replacement-power cost?
Wheeling
Which party carries network losses, changes in charges, metering errors and interrupted paths?
Cash controls
Are payment agent, waterfall, DSRA, lock-up, cure and lender step-in provisions aligned?

The financing view

Coverage ratios were used to locate risk—not to decorate the dashboard

Across project iterations, Felix tested debt quantum, tenor, interest, grace, reserve funding and repayment shape. Minimum DSCR was read alongside average DSCR, LLCR, cash waterfall and the timing of weak years.

The practical questions were: how much debt could the project carry without relying on uncontracted upside; what happened after a six-month payment delay; whether the DSRA could bridge the stress; and when distributions should lock up.

75%Debt cap tested in lender-style structures
6 monthsForward debt service considered for DSRA
1.20x–1.31xMinimum DSCR range across model iterations
P90Conservative generation basis for debt discussion
Why the metrics are ranges

The work covered more than one project and model iteration. Publishing one polished “final” number would be less truthful than showing the range and the conditions that caused it to move.

What the project team received

Outputs designed for sponsor, investor and lender conversations

Model and QA

Integrated assumptions, calculations, debt schedule, sensitivities, audit checks and reconciliation of changed inputs.

Decision materials

Investor executive summary, business plan, finance commentary and clear conditions before external reliance.

Bankability workplan

Prioritised issues around offtake, payment security, wheeling, grid, EPC, land, E&S, tax and financing evidence.

Decision reached

Proceed with staged capital and lender engagement, but tie progress to evidence conversion: executed buyer allocation, payment security, grid and wheeling documents, locked EPC scope, land and E&S, tax confirmation and lender-accepted model mechanics.

Disclosure boundary

This page is based on genuine project-team exposure. Project, client and counterparty names are withheld. Figures are rounded and combine project iterations. It does not state financial close, lender approval or execution of the outstanding agreements.

Model and contract review

Need the financial model tested against the actual commercial structure?

Share the model, contract list, current decision and deadline. The first step is to identify which assumptions are supported and which are carrying the conclusion.

Request a scopeInspect public architecture