Sri Lanka's Renewable-Energy IPP Market: Grid, Procurement and Offtaker Risk

Sri Lanka’s independently produced hydro, wind and grid-solar electricity increased only modestly in 2025 despite a substantial expansion in installed capacity. CEB’s three corresponding IPP rows total 907 MW at year-end and 2,169 GWh during the year, compared with 768 MW and 2,139 GWh in the restated 2024 columns. Capacity increased 18.1%; generation increased 1.4%. Solar growth partly offset lower hydro and wind output. These are a consistently defined administrative subset, not a census of every renewable independent producer.

The commercial question is therefore how additional capacity becomes deliverable, billable and collected electricity. The August 2026 feed-in tariff decision provides a current pricing framework. Grid access, project-specific agreements, resource conditions and payment performance still determine the cash that a producer can realise. A national renewable target cannot answer those project questions.

How large is the measured renewable-IPP segment?

The market boundary is independently controlled renewable generation in Sri Lanka sold to the grid under a procurement or purchase arrangement. Utility-owned generation, thermal IPPs and self-consumed electricity are excluded. Export-only rooftop systems can satisfy the economic boundary; the available rooftop aggregate does not establish their independent ownership, so it remains separate.

Sources: CEB Statistical Digest 2024 and 2025, printed p1. Each historical comparison uses the following year’s displayed revision where available. Annual energy and year-end capacity have different timing bases; their ratio is not used as a fleet capacity factor.

The 2025 increase comprises 86 GWh more grid solar, 48 GWh less mini-hydro and 8 GWh less wind. The arithmetic describes the change; it does not identify how much arose from weather, commissioning dates, outages or curtailment. Within this subset, solar capacity rose from 173 to 307 MW, while solar output rose from 257 to 343 GWh. The three technologies supplied 12.2% of total net generation reported by CEB in 2025. That share excludes substantial renewable electricity outside the subset.

Adding CEB’s “other NCRE” row gives 971 MW and 2,372 GWh in 2025, with output growth of 2.9%. This administrative sensitivity includes dendro, biomass and municipal solid waste. Without the renewable fraction of waste, treating the whole row as clean renewable generation would overstate the precision of the boundary. Excluding it from the main series also omits eligible biomass; neither series is presented as a complete market total.

Why rooftop exports need their own account

CEB reports 1,674 GWh exported from rooftop solar in 2025. Net Plus and Net Plus Plus account for 233 and 624 GWh respectively: 857 GWh, or 51.2% of the reported CEB-network exports. These arrangements should not be described indiscriminately as systems serving the customer’s own consumption. Net Metering and Net Accounting contribute the remaining 115 and 702 GWh.

Exports are not gross rooftop production. The figures also do not classify beneficial ownership. Combining every rooftop export with privately owned power-station output would conflate different populations. The source’s broad IPP total includes thermal generation and rooftop exports; it is unsuitable as a renewable-IPP market total. Reconstructing that broad total from displayed components leaves residuals of −1 GWh in 2024 and +1 GWh in 2025. The reproduction retains those differences rather than silently balancing them.

Which companies and plants can be traced?

A purposive eight-plant check joins issuer disclosures for the year ended 31 March 2026 to PUCSL’s register dated 30 June 2026. It tests the documentary chain, not national concentration.

Issuer interests and plant capacities describe different quantities. A 49% interest does not turn a 10 MW plant into a 10 MW wholly owned asset. The two 0.4 MW discrepancies remain unresolved; spelling variations are recorded in the underlying join. Register presence does not prove current licence validity, an executed PPA or settlement of invoices. WindForce’s portfolio gives Daily Life’s detailed interest as 95.83%, while its note rounds it to 96%; this is not treated as an ownership change.

The sample extends beyond one issuer, but its selection is purposive: accessible issuer operating-asset disclosures with identifiable register rows. It cannot estimate the largest operator, ownership concentration or the share of projects experiencing payment problems.

What changed in the pricing framework?

PUCSL’s decision of 21 August sets feed-in tariffs effective from 25 August 2026, subject to its eligibility and implementation conditions. Annex 2 separates non-escalable and escalable components. For example, ground-mounted solar lists Rs18.00/kWh plus base operation-and-maintenance of Rs2.72/kWh; wind lists Rs20.80 plus Rs4.93. These are tariff components, not a quotation of every project’s realised receipt. The decision also specifies adjustments and distinguishes plants, prosumers and battery arrangements.

The financial model uses a 10.92% cost of debt, 14.19% cost of equity and 60:40 debt–equity ratio. These regulatory assumptions do not establish financing actually available to a developer or returns actually earned. Stakeholder comments in the annex challenge financing and technical assumptions; they are attributed submissions rather than measured sector outcomes. Investors must assess their own commissioning, resource, financing and collection conditions against the applicable agreement.

Separately, PUCSL requested comments on NSO’s revised SPPA submission through 29 September. The inspected consultation establishes a proposed agreement submitted for approval. It does not establish final approval, execution for a named plant or a retrospective change to every existing contract. The proposed wording is therefore excluded from claims about operative project entitlements.

Who buys the electricity, and where does risk arise?

PUCSL’s first-quarter 2026 cost-recovery report identifies NSO as the single buyer pending open access, following the restructuring of CEB’s functions. A plant’s commercial delivery chain connects permissions and land, construction and interconnection, dispatch and metering, contractual billing and payment. Each stage can delay cash generation even when installed MW has increased.

The regulator’s discussion of missing PPAs concerns utility-generation arrangements involving EGL. It must not be generalised to all private renewable IPPs. Likewise, the bulk-supply business’s revenue deficit is neither the profit of producers nor a measurement of their overdue receivables. A tariff mechanism permitting later recovery is not evidence that invoices have already been paid.

For buyers and the system, hydro, wind, solar, thermal generation and storage offer different timing and operational characteristics. A simple tariff comparison omits availability, balancing, transmission, storage losses and procurement conditions. For producers, a feed-in tariff and a tendered agreement are not interchangeable products. A defensible project comparison would match the technology, connection point, commissioning period, contract length, escalation, dispatch conditions and payment security.

What would demonstrate progress over the next six months?

The most useful indicators are commissioned capacity paired with delivered energy; attributable curtailment with the applicable compensation rule; invoices paired with collection ageing; and published procurement decisions paired with executed contracts. These indicators test four proposed critical success factors: connection readiness, dependable conversion of resources into delivered output, enforceable allocation of contractual risk, and collection sufficient to service operating and financing costs.

A favourable pathway would combine additional generation with stable collection and effective connection or storage investment. A constrained pathway would show installed additions without corresponding usable output or cash conversion. Both can occur in different parts of the market. No probability or project-return forecast is assigned. Review the next official generation vintage, final SPPA decisions and issuer interim accounts when released.

Methods, evidence and limits

All numerical changes use displayed source values. Growth equals (new / old − 1) × 100; shares use the explicitly named denominator. Source and calculation JSON retain units, periods, restatements and residuals. The licence join requires exact register rows and licence numbers and checks the original XLSX hash. Issuer accounts supply selected ownership and operating observations; they do not independently verify contract performance.

Sources: CEB 2025 digest, CEB 2024 digest, PUCSL licence register, WindForce annual report pp19–20,305–306, Resus annual report pp166,192, LVL annual report pp16,61, August tariff decision, SPPA consultation, and Q1 cost-recovery report.

The evidence supports a bounded generation series, a documented asset sample and a current regulatory account. It does not establish a comprehensive ownership census, national renewable-IPP revenue, project profitability, universal PPA validity, actual compensation or an industry-wide arrears distribution. Future work should link plant-level output, commissioning dates and collection records without substituting aggregate utility figures for producer outcomes.

No external funding. No disclosed commercial relationship with the named producers or authorities. General research, not investment or legal advice. Corrections can be submitted through the site’s correction pathway.