Transformer Temperature and Voltage Management in High Solar PV Distribution Systems: A Convex Optimization Approach

Kernekoncepter

This paper proposes a centralized model predictive control (MPC) strategy to manage voltage and transformer temperature in distribution systems with high solar PV penetration, minimizing PV curtailment while ensuring system reliability.

Resumé

Bibliographic Information:

Ghorbansarvi, A., Hamilton, D., Almassalkhi, M. R., & Ossareh, H. R. (2024). Transformer Temperature Management and Voltage Control in Electric Distribution Systems with High Solar PV Penetration. arXiv preprint arXiv:2410.09011.

Research Objective:

This paper addresses the challenge of managing both voltage regulation and transformer temperature in electric distribution systems with high solar PV penetration while minimizing PV curtailment.

Methodology:

The authors propose a centralized MPC strategy that incorporates a linearized distribution system model, inverter constraints, and a transformer temperature model. To handle the non-convexity of the transformer temperature dynamics, a convex relaxation is applied, and conditions for its tightness are explored using Karush-Kuhn-Tucker (KKT) analysis. The performance of the proposed MPC is evaluated through numerical simulations on a radial distribution network.

Key Findings:

• High PV penetration without proper control can lead to overvoltage and transformer overloading issues.
• The proposed MPC strategy effectively manages voltage magnitudes and transformer temperatures within acceptable limits.
• The convex relaxation of the transformer temperature model provides a conservative temperature prediction, ensuring safe operation.
• KKT analysis identifies conditions under which the relaxation is tight, guaranteeing optimality.
• Including a penalty term for reactive power in the MPC objective function improves closed-loop performance with short prediction horizons.

Main Conclusions:

The proposed centralized MPC approach effectively manages voltage and transformer temperature in high solar PV distribution systems while minimizing PV curtailment. The convex relaxation and KKT analysis provide theoretical guarantees for the controller's performance.

Significance:

This research contributes to the development of advanced control strategies for integrating high levels of solar PV into distribution grids while ensuring system reliability and efficiency.

Limitations and Future Research:

Future work includes investigating decentralized control strategies, extending KKT analysis to cases with binding voltage constraints, and studying the impact of forecast uncertainty on MPC performance.

Statistik

The total PV curtailment for the case when only PV curtailment is considered in the MPC objective function is 12.4%.
The total PV curtailment for the proposed MPC approach is 4.84%.

Citater

"While voltage regulation and temperature management have been studied individually, few works have proposed a unified framework that addresses both challenges simultaneously."
"This significantly reduces the computational burden while still achieving good results in terms of minimizing curtailment and maintaining system reliability."

Vigtigste indsigter udtrukket fra

Transformer Temperature Management and Voltage Control in Electric Distribution Systems with High Solar PV Penetration

by Amirhossein ... kl. arxiv.org 10-14-2024

https://arxiv.org/pdf/2410.09011.pdf

Dybere Forespørgsler

How might the increasing adoption of electric vehicles and their charging patterns impact both voltage regulation and transformer temperature management in distribution systems with high solar PV penetration?

The increasing adoption of electric vehicles (EVs) introduces a significant new load on distribution systems, and their charging patterns can exacerbate the challenges of voltage regulation and transformer temperature management, especially in the presence of high solar PV penetration. Here's how:
Impact on Voltage Regulation:

Increased Load: EVs, particularly during concurrent charging periods, can significantly increase the load on distribution feeders. This can lead to voltage drops along the feeder, potentially violating the lower voltage limits (vmin) and impacting power quality.
Reverse Power Flow: While high PV generation during the day can cause overvoltage, the increased evening peak demand from EV charging might coincide with reduced or absent solar generation. This can lead to rapid voltage fluctuations and challenges in maintaining voltage stability.
Location Dependency: The impact of EV charging on voltage regulation is highly dependent on the spatial distribution of charging stations and their proximity to the distribution transformer. Clusters of EV charging stations can create localized voltage sags, especially if located far from the substation.
Impact on Transformer Temperature Management:

Higher Peak Loads: The combined load of EVs and existing household demands can lead to much higher peak loading on distribution transformers, especially during evening hours when solar PV output is low.
Increased Loading Duration: Unlike many traditional loads, EV charging often involves prolonged periods of high power draw. This sustained high load can lead to a faster build-up of heat in the transformer, increasing the risk of overheating and accelerated aging.
Thermal Inertia: Transformers have significant thermal inertia, meaning they heat up and cool down slowly. If not managed properly, consecutive cycles of high load due to EV charging, interspersed with periods of lower load, can lead to a gradual temperature increase in the transformer, eventually exceeding safe operating limits.
Mitigating the Impacts:
Addressing these challenges requires a multifaceted approach, including:

Smart Charging Strategies: Implementing time-of-use tariffs or incentivizing off-peak charging can shift EV charging demand away from peak hours, alleviating stress on the grid.
Coordinated Voltage Control: Utilizing smart inverters in both PV systems and EV chargers to provide reactive power support can help regulate voltage and mitigate voltage fluctuations.
Dynamic Transformer Rating: Employing real-time monitoring and dynamic transformer rating systems can allow for greater utilization of existing transformers by considering their actual operating conditions and thermal capacity.
Grid Modernization: Investing in grid modernization efforts, such as upgrading transformers with higher capacity or deploying distributed energy resources like battery storage, can enhance the grid's ability to handle increased and fluctuating loads.

Could a decentralized control approach, where individual inverters make local decisions based on limited information, achieve comparable performance to the proposed centralized MPC strategy while potentially offering advantages in terms of scalability and communication overhead?

A decentralized control approach, where individual inverters make local decisions based on limited information, could potentially offer advantages in terms of scalability and reduced communication overhead compared to the centralized MPC strategy. However, achieving comparable performance in terms of voltage regulation, transformer temperature management, and minimizing PV curtailment presents significant challenges:
Potential Advantages of Decentralized Control:

Scalability: Decentralized approaches are inherently more scalable as they don't rely on a central controller to process information from and send control signals to all distributed energy resources (DERs). This is particularly beneficial for large distribution systems with numerous PV systems and EV chargers.
Reduced Communication Overhead: By relying on local measurements and communication between neighboring devices, decentralized control can significantly reduce the amount of data transmitted and the complexity of communication infrastructure required.
Faster Response Times: Decentralized controllers can react more quickly to local grid conditions as they don't need to wait for instructions from a central controller, potentially improving the system's dynamic response to disturbances.
Challenges in Achieving Comparable Performance:

Limited Information: Decentralized controllers operate with limited information about the grid, typically relying only on local voltage measurements and potentially communication with neighboring devices. This lack of global visibility can lead to suboptimal decisions that might not effectively address system-wide constraints.
Coordination Challenges: Ensuring proper coordination between numerous independent controllers to achieve desired system-level objectives like minimizing total PV curtailment or maintaining transformer temperature within limits is a complex task. Without proper coordination, decentralized control actions might conflict with each other, leading to instability or inefficient operation.
Tuning Complexity: Tuning numerous decentralized controllers to achieve optimal performance can be significantly more challenging than tuning a single centralized controller. This complexity increases with the size of the distribution system and the number of DERs.
Achieving Comparable Performance:
While challenging, achieving comparable performance with decentralized control might be possible through:

Advanced Control Algorithms: Developing sophisticated decentralized control algorithms that can effectively utilize local information and communicate with neighboring devices to make informed decisions.
Hierarchical Control Structures: Implementing hierarchical control structures where local controllers operate within a broader framework defined by a higher-level controller that provides some degree of coordination and sets system-wide objectives.
Learning-Based Approaches: Utilizing machine learning techniques to enable decentralized controllers to learn from past experiences and improve their performance over time.
Conclusion:
Decentralized control approaches hold promise for managing the increasing complexity of distribution systems with high DER penetration. However, addressing the challenges of limited information, coordination, and tuning complexity is crucial to achieving comparable performance to centralized MPC strategies. Further research and development of advanced decentralized control algorithms are essential to fully realize the potential of this approach.

How can the insights from this research on managing transformer temperature be applied to other critical infrastructure systems facing similar challenges from increasing renewable energy integration and fluctuating demand patterns?

The insights gained from this research on managing transformer temperature in distribution systems with high solar PV penetration have broader applicability to other critical infrastructure systems grappling with similar challenges posed by increasing renewable energy integration and fluctuating demand patterns. Here's how:
1. Thermal Modeling and Monitoring:

Transferable Thermal Models: The principles of thermal modeling used for transformers, such as the regression model based on experimental data, can be adapted to other infrastructure components with thermal constraints, like power electronics in solar inverters, battery energy storage systems, or even electric vehicle charging stations.
Real-Time Monitoring: The emphasis on real-time temperature monitoring in this research highlights its importance for any infrastructure system with thermal limits. Deploying robust sensing and data acquisition systems allows for proactive management and prevents unexpected outages.
2. Predictive Control Strategies:

Adapting MPC for Other Systems: The core concepts of Model Predictive Control (MPC), such as using forecasts of renewable generation and demand to optimize control actions, can be applied to other infrastructure systems. For example, MPC could be used to manage the charging and discharging of a large-scale battery energy storage system to smooth out fluctuations from wind power generation.
Constraint Handling: The research demonstrates the importance of explicitly considering operational constraints, like transformer temperature limits, within the control framework. This approach is crucial for ensuring the safe and reliable operation of any critical infrastructure system.
3. Managing Fluctuating Demand:

Demand Response Programs: The concept of curtailing PV output to manage transformer temperature can be extended to other demand-side management strategies. For example, incentivizing flexible loads to reduce consumption during peak periods can alleviate stress on infrastructure components.
Energy Storage Integration: The research highlights the potential of energy storage to mitigate the impacts of fluctuating renewable generation and demand. This insight is applicable to various infrastructure systems, such as using pumped hydro storage to balance the grid with high solar PV penetration.
Specific Examples:

Water Distribution Systems: Managing water levels in reservoirs and pumping stations under fluctuating demand and solar-powered pumping systems can benefit from similar thermal modeling and predictive control strategies.
Building Energy Management: Optimizing HVAC systems in buildings with rooftop solar and varying occupancy patterns can utilize similar approaches to manage energy consumption and thermal comfort within constraints.
Transportation Systems: Managing traffic flow and congestion in smart cities with increasing electric vehicle adoption and charging demands can leverage similar forecasting and control techniques.
Conclusion:
The core principles of thermal modeling, predictive control, and demand management employed in this research provide valuable insights for managing other critical infrastructure systems facing similar challenges from renewable energy integration and fluctuating demand. By adapting these approaches and developing tailored solutions, we can enhance the resilience, reliability, and sustainability of critical infrastructure in a future powered by clean energy.