This article includes a fictional case study based on real conditions in the South African electricity sector. The figures are illustrative, but the legal, regulatory and commercial issues are real.
South Africa has procured approximately 6,300 megawatts of renewable generation capacity through the Renewable Energy Independent Power Producer Procurement Programme (REIPPP).[1] This rapid growth in generation capacity has outpaced the expansion of the country's electricity transmission network, leading to grid congestion, with more electricity being generated than the transmission lines can currently carry.
Curtailment[2] is no longer a theoretical risk. It is a growing operational and commercial reality. NTCSA processes payments of approximately ZAR 45 billion to independent power producers (IPPs) per annum. In 2026, the National Transmission Company South Africa SOC Ltd (NTCSA)[3] reported that a sharp rise in the volume and complexity of curtailment-related claims during April and May resulted in additional verification and assessment requirements that exceeded typical levels, creating temporary bottlenecks and leaving curtailment claims awaiting verification and settlement at approximately ZAR 2 billion.
NTCSA responded by introducing a streamlined claims-handling process and, by late July 2026, had already reduced the outstanding value to approximately ZAR 1.5 billion. By 31 August 2026, NTCSA reported that claims processing had returned to normal operating levels within its committed timeframe, with only current claims moving through the cycle, after the new process enabled it to process almost 3,000 claims in July and August.
The fictional case study
The project
The fictional Kalahari Wind Project is a 140-megawatt wind farm in the Northern Cape, built at a cost of more than ZAR 3 billion.
Its developer, GreenVolt Energy, signed a 20-year power purchase agreement (PPA)[4] under the REIPPP. The PPA's system operator instruction and deemed energy provisions[5] governed how GreenVolt would be compensated if it was instructed to stop generating power, with the precise terms and risk allocation for each project depending on the version of the PPA used.
What does a curtailment clause typically look like?
Curtailment clauses in renewable-energy PPAs typically permit the buyer or system operator to reduce a generator's output, while preserving the generator's entitlement to compensation for qualifying deemed energy.
By way of illustration:
"The Buyer may instruct the Seller to reduce generation where required for system security or network constraints, provided that the Seller is compensated for any qualifying deemed energy in accordance with the agreed calculation methodology."
In this case study, GridCo SOC Ltd (a fictional entity) acts as both the buyer of the electricity and the system operator. The project was funded by development finance institutions and commercial lenders, with monthly debt repayments of ZAR 28 million.
For three years the project performed to plan. In Year 4, curtailment began.
The curtailment
GridCo, acting as system operator, instructed GreenVolt to reduce or cease electricity generation because of network congestion and transmission capacity constraints. Under section 21(3) of the Electricity Regulation Act 4 of 2006 (the ERA), a transmission licensee is required to give generators non-discriminatory access to the transmission network. The Electricity Regulation Amendment Act 38 of 2024 (the amended ERA) goes further by limiting the system operator's ability to favour one generator over another when managing grid constraints and curtailment, requiring dispatch decisions with significant commercial consequences to be applied fairly across generators. The Grid Connection Code for Renewable Power Plants[6] further sets out the technical obligations on generators to adjust their output and comply with the system operator's instructions.
Over six months, the wind farm ran at just 41% of available capacity despite sufficient wind. GreenVolt estimated its revenue loss at ZAR 210 million.
Its debt repayments continued, forcing it to use its debt service reserve account.
The deemed energy claim
GreenVolt submitted a deemed energy claim for ZAR 140 million. GridCo disputed it on four grounds.
Ground 1: characterisation. GridCo contended that certain events were caused by GreenVolt’s own equipment failures or scheduled maintenance, not by curtailment instructions. The parties’ Supervisory Control and Data Acquisition (SCADA) system logs[7] and operational records diverged on the timing, duration and cause of several events.
Ground 2: quantum. GreenVolt based its claim on generation data from comparable prior periods. GridCo, by contrast, applied measured wind speed to the farm’s expected power curve. The gap was ZAR 70 million. The PPA did not prescribe a single methodology.
Ground 3: emergency exception. The PPA excluded deemed energy payments during a “system emergency”. GridCo invoked the exception for three events. GreenVolt countered that recurring grid congestion was a foreseeable infrastructure constraint, not an emergency within the meaning of the PPA.
Ground 4: claims procedure. One claim, worth ZAR 8 million, was submitted four days late. GridCo argued that it was out of time.
The cost of the dispute
Four months of negotiation failed. GreenVolt commenced arbitration and the proceedings lasted 22 months and cost ZAR 1 million in legal and expert fees.
The tribunal awarded ZAR 97 million, ZAR 43 million less than the claim. By the time the award was handed down, GreenVolt’s lenders had demanded a remedial plan for the depleted debt service reserve account, two senior executives had resigned and curtailment was continuing.
Four lessons for future PPAs
The dispute points to four gaps in the PPA that future contracts can close.
Failure 1: no agreed formula
The PPA contained no protocol for recording, timestamping, categorising or confirming curtailment instructions in real time.
Failure 2: an untested formula
The deemed energy formula had never been stress-tested against worked examples, leaving both parties free to adopt competing methodologies.
Failure 3: no fast-track technical formula
A dispute that was fundamentally about engineering took 22 months to resolve because the PPA required all disputes, whether technical or legal, to be resolved through a single arbitration process. There was no mechanism for early expert determination of the technical questions.
Failure 4: a broad emergency exception
The definition of "system emergency" was drafted too broadly, failing to exclude routine grid congestion and enabling NTCSA to invoke the exception and deny deemed energy payments for foreseeable curtailment.
What is the immediate solution?
Each of these gaps can be addressed within the existing contractual and regulatory framework.
Solution 1: a shared curtailment log. Every curtailment instruction should be issued through a shared electronic platform that automatically timestamps the instruction, records its stated cause and logs its duration.
Solution 2: worked examples. Before financial close, the parties should stress-test the deemed energy formula against a range of curtailment scenarios and incorporate the agreed worked examples into a binding PPA schedule.
Solution 3: expert determination. Technical disputes, for example metering accuracy, wind-resource data and power-curve assumptions, should be carved out and referred to an independent energy expert within a fixed period, with the PPA specifying whether the determination is final, provisionally binding or reviewable.
Solution 4: a narrow emergency exception. The exception should exclude foreseeable congestion and frequently recurring events, applying only to genuine, unforeseeable threats to system security expressly agreed by both parties.
Solution 5: pay now, dispute later. Undisputed amounts should be paid within 30 days, ring-fenced from any ongoing dispute over the balance of the claim.
What is an innovative solution?
The solutions above improve the existing claims process. However, a more ambitious approach, proposed here for consideration, would be to replace it entirely with a pre-agreed, automated compensation mechanism: the Parametric Curtailment Protocol.
Parametric systems trigger payments automatically when pre-agreed conditions are met, using objective data and contractually fixed formulas. They eliminate the need for a post-event claims process. The Parametric Curtailment Protocol set out below is one way such an approach could be designed for the South African curtailment context.
NTCSA has already taken welcome steps to improve the way existing claims are administered. Its streamlined process now draws operational data directly from the system operator's SCADA systems and pays the full estimated value of a claim upfront, with technical verification following afterwards, a change industry bodies such as the South African Photovoltaic Industry Association (SAPVIA) have welcomed as easing cash-flow pressure on independent power producers.
NTCSA and SAPVIA are also collaborating on longer-term, structural measures, including accelerating battery energy storage and demand-side mechanisms to encourage electricity use when renewable generation is abundant. The Parametric Curtailment Protocol proposed here would build on this progress: rather than making the current claims process faster, it would change the underlying contractual architecture by replacing post-event assessment with a pre-agreed formula, jointly governed data and automatic payment triggers.
Parametric thinking is not new to this space, though no PPA or grid framework yet combines it into a single mechanism of the kind proposed here. The African Risk Capacity[8] already uses parametric triggers to pay out automatically once a pre-defined, objectively measurable threshold, such as a rainfall or vegetation index, is crossed, without the need for a loss investigation.
A separate and fast-growing market for parametric curtailment insurance has also emerged internationally, offering renewable developers payouts linked directly to curtailment events. However, these are typically standalone insurance products bought alongside a PPA, rather than a mechanism built into the buyer-generator contract itself, and are not yet common in emerging markets.
In the UK, curtailed generators are compensated through an established constraint-payment mechanism, though this operates as an administered, case-by-case settlement process rather than a parametric trigger. Together, these examples show that objective data and pre-agreed rules can support efficient curtailment compensation.
Proposed Step 1: the Curtailment Index
At the outset of a project, an independent energy consultant could develop a Curtailment Index (a pre-agreed mathematical model) that:
- calculates the facility’s theoretical output using real-time wind or solar data;
- compares this with actual metered output;
- identifies a potential curtailment event when the difference exceeds an agreed limit and the reduction is linked to a recorded system operator instruction; and
- calculates deemed energy automatically, after accounting for availability, planned maintenance, electrical losses, grid-code limitations and other agreed exclusions.
Proposed Step 2: shared data infrastructure
Under this proposal, the underlying data could be stored on a jointly governed platform, hosted by an independent third party, with a tamper-proof audit trail and role-based access controls. Curtailment instructions and generation data would feed into this platform, creating one authoritative record that also captures the basis, sequence and allocation of instructions, so generators can verify that curtailment is applied consistently and without discrimination.
NTCSA is already using system operator SCADA data to calculate and pay estimated curtailment claims, which is a positive indication that much of the operational data needed for a parametric system is already being generated in practice. Building on this, the data could be consolidated into a single, jointly governed and independently auditable platform.
NTCSA and the future Transmission System Operator SOC Ltd (TSO)[9] could be well placed to provide an institutional home for such a platform.
Proposed Step 3: automatic payment
Under this proposal, once the Index flags a curtailment event, compensation could be paid automatically from a pre-funded curtailment escrow account, subject to agreed validation checks.
Payment could be designed to reach the generator within 48 hours. The buyer could dispute the amount within 15 business days, but the payment would not be reversed while the dispute is pending, unless fraud, manifest error or another narrowly defined ground applies. This would protect the generator’s debt-service cash flow and reduce the risk of legal proceedings.
The escrow structure will require careful design, particularly around public-finance constraints, procurement rules, governance, insolvency remoteness and security arrangements if a public entity is required to pre-fund the account. This is especially pressing given government's stated intention to move away from sovereign guarantees as a backstop for public institutions' debt obligations, which removes a mechanism that has traditionally underpinned the bankability of state-linked payment obligations in this sector; any publicly funded escrow would need to be designed around this constraint, potentially drawing on private credit enhancement instruments or a vehicle such as the Credit Guarantee Vehicle reportedly under development by National Treasury.
Realising this proposal fully would benefit from further development of the framework for recovering curtailment compensation costs, which is not yet addressed. A market, trading, system-operation or tariff mechanism that allocates these costs clearly would give investors and lenders further certainty on who ultimately bears them.
Proposed Step 4: a Standing Curtailment Adjudication Panel
Under this proposal, remaining disputes could be referred to a specialist Standing Curtailment Adjudication Panel, established under an appropriate legislative, regulatory, or contractual framework with participation of the National Energy Regulator of South Africa (NERSA)[10], the Independent Power Producer Office (IPPO) Office[11], NTCSA, and other relevant industry bodies.
A three-member panel of engineers and energy lawyers should have access to the shared data, Curtailment Index and system operator records, hearing disputes on an expedited basis and delivering binding determinations within a fixed period.
Looking ahead: the TSO transition
Inertia
Standard-form PPAs are difficult to change. Sponsors and lenders often default to familiar structures which leave little room for innovation.
Timing
Developing a Curtailment Index before financial close adds cost, complexity and time to an already compressed development cycle.
Institutional capacity
A shared platform and standing panel would demand co-ordination across NERSA, NTCSA, the responsible government department, the IPPO and private-sector participants, all of whom are already operating at full capacity.
Bankability risk
From a lender's perspective, curtailment reduces project revenue, placing pressure on debt service cover ratios and increasing the risk of finance agreement breaches. This can cause financial distress even where the project's equipment is operating properly.
The TSO transition and the growing focus on resolving curtailment compensation create a timely opportunity for the industry to build on the progress already made.
NTCSA, Eskom’s wholly owned transmission subsidiary, performs the system operator and buyer functions on a transitional basis under the amended ERA while a permanent, independent TSO is being established outside Eskom. Government has confirmed that the TSO's establishment as a standalone state-owned entity is a settled policy decision, with a transaction adviser process underway and conclusion targeted by end-2027. Most of the Act’s relevant provisions commenced on 1 January 2025, although certain provisions remain subject to separate commencement.
Separating transmission, system operation and market functions from Eskom’s generation business is intended to promote transparency, competition and non-discriminatory access to transmission and distribution systems. As with any phased institutional transition, full separation will take time and the process is expected to further strengthen confidence in the framework as it progresses.
Existing PPAs should thus be reviewed to confirm that references to Eskom, the buyer and the system operator remain operative following the transfer of the transmission business to NTCSA on 1 July 2024. Project-specific contractual analysis remains necessary.
The transition presents an opportunity to restructure the contractual and commercial mechanisms for managing curtailment risk alongside the institutional framework itself.
What this means for developers and lenders
Curtailment is a present and escalating risk for South African renewable energy projects.
The projects that navigate it successfully will be those that assessed grid-evacuation capacity before committing capital, drafted their PPAs for curtailment and built systems to record each event and calculate compensation in near-real time. This is especially true for projects contracted at lower tariffs, which generally have thinner margins and less headroom to absorb revenue shortfalls or delays in curtailment compensation.
Much of what is needed is already within reach. Shared curtailment logs, worked examples, expert determination, narrowly drawn emergency exceptions and ring-fenced interim payments can all be implemented within the current contractual and regulatory framework. The Parametric Curtailment Protocol is a more ambitious proposal that would go further, combining objective data, automatic payments and specialist adjudication into a single mechanism. It would demand careful contractual design and possible regulatory and procurement reform. But even if implemented, the Protocol would only deal with the financial side of curtailment. It would not reduce curtailment itself, which still requires grid expansion, storage and market signals that encourage demand when renewable generation is abundant.
Even so, the cost of implementing these fixes is lower than the cost of leaving curtailment compensation unresolved, in lost revenue, strained lender relationships and prolonged disputes. That cost comparison matters because South Africa needs international capital to fund its energy transition and that capital is more likely to flow where curtailment risk is clearly defined, fairly allocated and efficiently resolved.
The difference between a curtailed wind farm that remains bankable and one that becomes a dispute waiting to happen is, ultimately, the strength of its contractual protection.
This article was prepared by Webber Wentzel for general informational and educational purposes. It does not constitute legal advice. Specific legal advice should be obtained in relation to any particular transaction or dispute.
