Janus: A Next-Generation Grid Optimization Tool for Cost Reduction and Renewable Integration

SUMMARY

Janus is a next-generation grid optimization tool designed to reduce redispatch costs and increase renewable energy integration. Unlike traditional methods that treat redispatch and topological adjustments separately, Janus optimizes both within a unified framework. This integrated approach improves grid efficiency, reduces costs, and minimizes renewable curtailment. Using advanced relaxation techniques, Janus solves complex, non-linear problems with high accuracy and speed, achieving a 0.2% tolerance gap—far better than existing tools. Currently in its proof-of-concept stage, Janus has shown promising results and is being prepared for pilot studies. 

confi

10Introduction

As power grids transition to higher shares of renewable energy, the challenges of maintaining grid stability while minimizing costs are becoming increasingly complex. The rapid expansion of wind and solar generation, coupled with the need to ensure grid security, has led to a sharp rise in redispatch costs and renewable energy curtailment. To address these challenges, we introduce Janus, an advanced grid optimization tool designed to simultaneously optimize topological actions and redispatch strategies. By integrating both elements within a unified optimization framework, Janus delivers a significant reduction in system costs while enabling greater renewable energy utilization.

The Challenge: Rising Redispatch Costs and Renewable Curtailment

Redispatch refers to the process of adjusting power plant output to alleviate transmission congestion. As renewable energy penetration increases, transmission bottlenecks become more frequent, requiring costly redispatch measures to maintain grid reliability. Studies suggest that by 2030, redispatch costs could exceed €11 billion annually, escalating beyond €34 billion by 2040. Additionally, renewable curtailment—where excess generation is wasted due to grid constraints—remains a significant issue worldwide. For example:

• The Southwest Power Pool (SPP) in the USA experiences curtailment rates of ~10%.
• California reports a renewable curtailment rate of approximately 7%.
• Germany currently faces curtailment levels of around 4%.

These figures highlight the pressing need for solutions that can alleviate congestion, reduce redispatch costs, and maximize renewable energy utilization. Traditional grid optimization methods treat redispatch and topological adjustments separately, leading to inefficiencies and suboptimal decision-making.

Introducing Janus: An Integrated Optimization Framework

Janus is designed to overcome these limitations by optimizing both redispatch actions and topological adjustments within a single framework. This integrated approach ensures that all available grid management tools are considered simultaneously, leading to more effective congestion management and cost reductions. Key innovations of Janus include:

1. Unified Optimization Methodology: Instead of treating redispatch and topological optimization as separate problems, Janus solves them jointly, ensuring a holistic approach to grid operation.
2. Advanced Relaxation Techniques: Janus employs an innovative optimization-based constraint relaxation strategy to solve the inherently non-linear, non-convex problem with high accuracy.
3. Superior Solution Accuracy: While traditional approaches exhibit cost gaps exceeding 4%, Janus achieves a tolerance gap of only 0.2%, significantly enhancing cost efficiency.
4. High-Performance Design: By reducing need on computationally expensive AC sensitivities, Janus minimizes processing time without compromising accuracy.
5. Compatibility with Acceleration Technologies: The tool has been developed with performance in mind and is designed to leverage advanced acceleration techniques pioneered in Ceres, another high-performance optimization tool for power system planning.

Key Advantages of Janus

Janus provides several key advantages over traditional redispatch and topological optimization approaches:

1. Cost Reduction in Redispatch

By jointly optimizing topological actions and redispatch strategies, Janus identifies lower-cost solutions compared to conventional methods. Given the projected increase in redispatch costs, even small percentage reductions translate into significant financial savings for system operators.

2. Increased Renewable Utilization

By alleviating transmission constraints more effectively, Janus minimizes renewable curtailment. This leads to better utilization of available wind and solar resources, increasing grid sustainability and reducing reliance on fossil fuel-based power plants.

3. Enhanced Computational Efficiency

The tool is designed to minimize computational overhead by avoiding expensive AC sensitivity calculations. This efficiency makes Janus a practical solution for real-time and operational planning applications, allowing grid operators to make optimal decisions faster.

4. Improved Grid Stability and Reliability

By simultaneously considering both redispatch and topological reconfiguration, Janus helps improve overall grid stability. The tool can proactively mitigate congestion issues before they escalate into major reliability concerns, reducing the risk of blackouts or emergency interventions.

5. Future-Proof Design for Emerging Grid Challenges

As power systems continue to evolve with higher shares of renewables, flexible loads, and new storage technologies, Janus is built to adapt. The optimization framework can be extended to incorporate emerging grid management techniques, making it a future-proof tool for system operators.

Proof of Concept and Next Steps

Janus is currently available as a Proof of Concept, demonstrating the potential of its novel optimization approach. Initial results highlight its ability to significantly outperform existing methodologies, offering superior cost efficiency and enhanced grid management capabilities. The next steps in the development of Janus include:

• Pilot Studies with Grid Operators: Engaging with system operators to validate the tool’s performance under real-world conditions.
• Scalability Enhancements: Expanding the tool’s capabilities to accommodate larger and more complex power networks.
• Further Computational Optimization: Leveraging additional advancements in acceleration technologies to further enhance processing speed and efficiency.

Conclusion

Janus represents a breakthrough in grid optimization, offering a powerful tool for reducing redispatch costs and maximizing renewable energy utilization. By solving the non-linear, non-convex optimization problem with an unprecedented accuracy of 0.2%, Janus sets a new benchmark for efficiency and effectiveness in grid management. As energy systems transition towards higher renewable penetration, innovative solutions like Janus will be essential for ensuring a cost-effective, reliable, and sustainable electricity supply. With its strong foundations and promising initial results, Janus is poised to become a key component of next-generation power system operations.

Supporting analysis

The charts below illustrate the relaxation approach (Figure 1), where new cutting planes are added in each iteration, the old one are relaxed selectively with a novel approach developed at Pharoes, and the solution progressively evolves  towards the optimum (point 10 in green). It also shows the economic impact of each of the solutions, which after an exploratory phase where costs oscillate heavily, it ends up converging to the best solution found (Figure 2). Finally the results in Figures 3 and 4 show that the model results, which are limited to 10 iterations due to runtime limitations, are heavily dependent on the initial topology configuration. Whereas starting with a solution where all topological actions are closed leads to results which are close to the optimal results, starting with a topological configuration where grid violations are minimized does not improve the quality of the results, it actually worsens them. The same results are found in case market reoptimization is allowed as well as in cases where it is not allowed by setting a high enough market reoptimization threshold. 

About the authors 

This article was written by Diego Luca de Tena, Managing Director of Pharoes, who is based in Madrid.