# Electricity reform and the sequence of capacity delivery in South Africa

Policy Brief · South Africa · Energy and Infrastructure

Electricity reform changes the conditions under which projects can enter, connect, trade, and operate. It does not itself produce electricity on the date legislation commences. Reliable supply emerges through a sequence of legal, financial, technical, and operating milestones, each requiring separate evidence.

The South African government's [record of the Electricity Regulation Amendment Act 38 of 2024](https://www.gov.za/documents/acts/electricity-regulation-amendment-act-38-2024-english-isizulu-20-aug-2024) identifies commencement on January 1, 2025 and provisions concerning transmission-system operation and a competitive trading platform. These are changes to the sector framework. They should not be described as evidence that the new market and every associated asset were fully operational on that date.

## Permission and registration precede productive service

A project can be legally permitted or registered before construction begins. It can also be financed and substantially built while awaiting network access or testing. The capacity associated with each stage therefore needs a distinct label.

NERSA's [January 2026 registration statement](https://www.nersa.org.za/files/files/2026/01/MediaStatement-NERSAregisters147generationfacilitiesin3rdquarterof2025-26.pdf) reports 147 facilities with combined capacity of 1,960 MW registered in the third quarter of 2025/26. Those are registration-period figures. They should not be relabeled as 1,960 MW newly commissioned during that quarter or as firm supply available at every hour.

The distinction remains important even where a registration document identifies the intended network. A recorded connection arrangement does not establish that all construction, testing, and operating conditions have been completed. Commissioning evidence and actual output are needed to establish service delivery.

## Exhibit: capacity and energy require different calculations

Consider a hypothetical 100 MW generation facility. Assume it operates at an average capacity factor of 30 percent over a non-leap year and ignore additional losses for this illustration.

| Assumed measure | Calculation | Result |
|---|---|---:|
| Nameplate capacity | Stated design rating | 100 MW |
| Annual hours | 365 × 24 | 8,760 |
| Annual energy | 100 × 8,760 × 30% | 262,800 MWh |
| Average output over the year | 262,800 / 8,760 | 30 MW |

The facility is not a constant 100 MW supply. Nor does its annual average of 30 MW establish that it can deliver that amount during a particular evening peak. The timing and variability of output determine its contribution to system reliability.

These assumptions do not describe a specific South African project or technology. Storage, transmission, curtailment, maintenance, and demand patterns would change the usable service. Adding nameplate ratings across different technologies without those considerations can overstate dependable capacity.

## The network can be the binding constraint

Generation needs a connection with sufficient capacity and suitable technical conditions. A project may face a queue, required reinforcement, or limits on when it can export. The cost and timing of those arrangements affect its financing and commercial start.

New network investment also has a delivery sequence: planning, approvals, land and access, procurement, construction, testing, and operation. A published transmission plan provides direction and potential capacity. It is not the same as completed network service.

Location matters because generation potential and demand are not distributed uniformly. A national surplus at one time or place cannot resolve every local bottleneck. Evaluating a project requires its actual route to the customer and the constraints on that route.

Curtailment can permit earlier connection under defined limits, but it changes expected output and revenue. The project appraisal should use those operating conditions rather than assume unrestricted delivery. The appropriate comparison is the value of earlier limited access against waiting for reinforcement, with risks stated explicitly.

## Trading arrangements need functioning operations

A legal framework for competitive trade requires detailed rules, settlement systems, metering, credit arrangements, and institutions capable of enforcing obligations. The existence of a statutory mandate does not establish that every commercial process is in use.

Market access also depends on the customer's position. A large buyer able to negotiate a long-term agreement may have options unavailable to a smaller business. Network charges, wheeling arrangements, and metering can affect the final delivered price. A generation bid alone is not the customer's all-in electricity cost.

The transition should therefore be assessed through implemented rules and completed transactions, alongside service reliability and cost. A target date, consultation, approved rule, and active market are distinct milestones. Reporting them separately provides a clearer account of reform progress.

## Maintenance and distribution remain essential

Additional generation cannot fully offset weak distribution at a factory or household connection. Local equipment failures, maintenance gaps, and financial problems can interrupt service even when the national system has sufficient energy. A supply assessment should identify the actual provider and network layer involved.

Existing generation availability matters alongside new entry. Rehabilitation and maintenance can restore usable capacity without increasing the fleet's nameplate rating. Conversely, new assets can be offset by deterioration elsewhere. The net service change requires both sides of the account.

Financial sustainability also affects reliability. Tariff collection, payment obligations, maintenance budgets, and investment capacity influence whether improved operation persists. A temporary reduction in interruptions is valuable, but durable service requires the institutions and resources to maintain it.

Commissioning should also be distinguished from sustained availability after initial testing. A South African project may pass an acceptance test but later experience equipment faults, restricted dispatch, or maintenance interruptions. The delivered contribution is therefore best assessed over an operating period with actual energy and availability records. One successful synchronization establishes an important milestone, while a reliable service history establishes a different and stronger result.

## A delivery record that supports economic decisions

The useful project record identifies legal status, financial close, construction, connection, commissioning, and actual output. It should distinguish capacity from energy and nameplate ratings from dependable contribution. System-level assessment then adds demand, reserves, network constraints, and the availability of other plants.

For businesses, the economic benefit includes lower backup costs, fewer production interruptions, and more predictable delivery, not just additional units consumed. These gains can occur before output expands. They should be measured without assigning every subsequent production increase to electricity reform alone.

South Africa's electricity reforms can improve entry, investment, and coordination. Demonstrating their contribution requires following the sequence from legal change to usable service. The mature test is reliable electricity at the point of demand under sustainable operating arrangements, with each preceding milestone reported for what it establishes and the remaining dependencies kept visible.


---

American Economic Senate