Autonomous Decision-Making in Dispatch Systems

1. Introduction

Autonomous Decision-Making in Dispatch Systems refers to the use of computer-based systems, artificial intelligence (AI), algorithms and automated control technologies to make or implement decisions about the dispatch of electricity without requiring a human to approve every individual decision.

Electricity dispatch means deciding which generating units should produce electricity, how much they should produce, and when they should produce it. In a modern electricity system, automated dispatch can respond to demand, generation availability, grid congestion, frequency, prices, weather conditions and system emergencies.

Autonomous dispatch can improve speed and reliability. However, electricity dispatch decisions have legal consequences. They can affect generators, municipalities, consumers, electricity prices and the stability of the national grid. Therefore, autonomous decision-making must operate within a clear legal and regulatory framework.

2. Meaning and Operation

An autonomous dispatch system generally follows several stages:

Data collection – information is received from generators, consumers, meters, weather systems and the grid.

System analysis – software evaluates demand, available generation and network conditions.

Decision-making – an algorithm determines the appropriate generation or dispatch response.

Automatic execution – instructions are communicated to generating units or other grid resources.

Monitoring – the system checks whether the desired result has been achieved.

Human intervention – authorised operators can intervene when necessary.

For example, if electricity demand suddenly rises, an automated system may increase generation from available units or activate battery storage. During an emergency, it may reduce generation or disconnect certain resources to protect grid stability.

3. South African Legal Framework

The Electricity Regulation Act 4 of 2006 (ERA) is central to electricity regulation in South Africa. It establishes the regulatory framework for generation, transmission, distribution, trading and related electricity activities.

The South African Grid Code is also important. It establishes technical and operational requirements for electricity-system participants. NERSA administers the regulatory framework and approves relevant codes and licence conditions.

Autonomous dispatch therefore cannot be based only on technical programming. Its decisions must comply with legislation, licence conditions, grid-code requirements and lawful directions from the relevant authorities.

4. Relevant Case Laws

Eskom Holdings v Sonae Arauco (2024)

In Eskom Holdings SOC Ltd v Sonae Arauco (Pty) Ltd, the Supreme Court of Appeal considered the legal importance of the South African Grid Code. The court recognised that the relevant Codes form part of licence conditions and discussed the System Operator's responsibility to take appropriate remedial action where abnormal conditions threaten reliable electricity-system operation.

This case is highly relevant to autonomous dispatch. If an algorithm automatically changes generation or takes remedial action, its programming must be consistent with applicable Grid Code obligations.

Eskom Holdings v Vaal River Development Association (2022)

In Eskom Holdings SOC Ltd v Vaal River Development Association, the Constitutional Court considered electricity supply, the Electricity Regulation Act, licence conditions and the constitutional and administrative-law framework governing electricity decisions.

The case demonstrates that electricity-system decisions cannot be separated from public-law responsibilities. An autonomous dispatch system must therefore remain accountable to the legal framework governing the electricity sector.

Eskom Holdings v Lekwa Ratepayers Association (2022)

The Supreme Court of Appeal considered the relationship between Eskom's bulk electricity supply and municipal electricity distribution in Eskom Holdings SOC Ltd v Lekwa Ratepayers Association. The judgment helps explain the different institutional responsibilities within the electricity system.

This is important for autonomous dispatch because an algorithm operating within one part of the electricity system should not improperly interfere with responsibilities legally assigned to another institution.

Pharmaceutical Manufacturers Association v President (2000)

The Constitutional Court's decision in Pharmaceutical Manufacturers Association of South Africa v President of the Republic of South Africa established the importance of legality and rationality when public power is exercised.

For autonomous dispatch systems, this means that automated regulatory decisions must have lawful authority and must be rationally connected to the purpose for which the power exists. An algorithm cannot create its own legal authority.

Earthlife Africa Johannesburg v Minister of Environmental Affairs (2017)

In Earthlife Africa Johannesburg v Minister of Environmental Affairs, the court emphasised the need to consider relevant environmental and climate-change factors in energy decision-making.

The case is relevant where dispatch decisions affect environmental obligations. An automated system should therefore be capable of incorporating legally relevant environmental requirements rather than considering only cost or technical efficiency.

5. Major Legal Issues

Algorithmic Transparency

Generators and other market participants may need to understand why a particular dispatch decision was made. Secret or unexplained algorithms can create accountability problems.

Human Oversight

Human operators should be able to intervene when an automated decision creates an unsafe or legally problematic situation.

Liability

If an autonomous system incorrectly dispatches a generator and causes financial loss or grid instability, responsibility may arise for the system operator, licensee, technology provider or other responsible party depending on the circumstances.

Cybersecurity

A compromised dispatch algorithm could manipulate generation levels and potentially destabilise the electricity system. Strong authentication, access controls and cybersecurity testing are therefore necessary.

Fair Market Access

Autonomous dispatch should apply approved rules consistently. Algorithms should not unlawfully favour particular generators or technologies.

6. Importance for Future Electricity Systems

Autonomous dispatch will become increasingly important with the growth of solar power, wind generation, battery storage, distributed energy resources, electric vehicles and flexible demand.

These resources can change rapidly. Automated systems can respond much faster than manual decision-making. They can also coordinate thousands of distributed resources simultaneously.

However, greater automation creates a corresponding need for stronger audit trails, explainability, cybersecurity, human oversight and regulatory supervision.

7. Conclusion

Autonomous Decision-Making in Dispatch Systems can make electricity grids faster, more flexible and more responsive. Algorithms can continuously analyse demand, generation, network conditions and emergencies and automatically implement dispatch decisions.

Nevertheless, autonomous dispatch must remain subject to the Electricity Regulation Act, South African Grid Code, licence conditions, constitutional principles and administrative law.

The cases of Eskom v Sonae Arauco, Eskom v Vaal River Development Association, Eskom v Lekwa Ratepayers Association, Pharmaceutical Manufacturers and Earthlife Africa demonstrate the importance of lawful authority, reliable electricity-system operation, rational decision-making and consideration of relevant factors.

The appropriate legal approach is therefore “autonomous technical decision-making under accountable regulatory supervision.” Automation may execute dispatch decisions rapidly, but the responsible electricity institutions must retain legal responsibility for the rules, system design, oversight, review and consequences of those decisions.