Optimal Net-Load Balancing in Smart Grids with High PV Penetration

Sanmukh R. Kuppannagari, Rajgopal Kannan, Viktor K. Prasanna

Abstract

Mitigating Supply-Demand mismatch is critical for smooth power grid operation. Traditionally, load curtailment techniques such as Demand Response (DR) have been used for this purpose. However, these cannot be the only component of a net-load balancing framework for Smart Grids with high PV penetration. These grids can sometimes exhibit supply surplus causing over-voltages. Supply curtailment techniques such as Volt-Var Optimizations are complex and computationally expensive. This increases the complexity of net-load balancing systems used by the grid operator and limits their scalability. Recently new technologies have been developed that enable the rapid and selective connection of PV modules of an installation to the grid. Taking advantage of these advancements, we develop a unified optimal net-load balancing framework which performs both load and solar curtailment. We show that when the available curtailment values are discrete, this problem is NP-hard and develop bounded approximation algorithms for minimizing the curtailment cost. Our algorithms produce fast solutions, given the tight timing constraints required for grid operation. We also incorporate the notion of fairness to ensure that curtailment is evenly distributed among all the nodes. Finally, we develop an online algorithm which performs net-load balancing using only data available for the current interval. Using both theoretical analysis and practical evaluations, we show that our net-load balancing algorithms provide solutions which are close to optimal in a small amount of time.

Introduction

As solar photovoltaic (PV) installations proliferate, power grids face a significant challenge in maintaining stability: net-load balancing. Traditionally, grid operators focused on load curtailment—incentivizing customers to reduce power usage during peak demand. However, grids with high solar penetration frequently encounter the opposite problem during peak sunlight hours: a massive supply surplus that triggers over-voltages and potential equipment failure.

Existing solutions like Volt-Var Optimization (VVO) are often too computationally intensive to scale effectively. Moreover, real-world curtailment is rarely a continuous variable; modern hardware like micro-inverters allows for the discrete toggling of specific panel clusters. This transforms the balancing act into a discrete optimization problem. Mathematically, this is NP-hard, making it difficult to find optimal solutions quickly as grid complexity grows.

The Unified Net-Load Balancing Framework

To address these challenges, we propose a generalized framework that simultaneously manages both load and supply curtailment across a defined "curtailment horizon." The process begins by calculating aggregate consumption and predicting supply for each interval to identify horizons where curtailment is necessary.

The proposed unified framework for identifying curtailment horizons and selecting optimal load or supply strategies.

Our approach utilizes several algorithmic strategies:

  1. Integer Linear Programming (ILP): Defines the theoretical optimal solution, serving as a benchmark.
  2. Bounded Approximation Algorithms: To meet real-time operational constraints, we employ Dynamic Programming with scaling and rounding. This trades a negligible amount of accuracy for a significant speed boost, guaranteeing solutions within a factor of $(1 \pm \epsilon)$ of the optimal cost.
  3. Fairness-Aware Optimization: To prevent disproportionately penalizing specific nodes, we incorporate curtailment budgets using Linear Programming (LP) relaxation and a novel rounding heuristic.
  4. Online Heuristics: For scenarios where future forecasts are unavailable, a "myopic" version of the algorithm makes interval-by-interval decisions.

Experimental Evaluation

The framework was evaluated using a combination of real-world and simulated datasets. Load data was sourced from the USC Campus Microgrid, featuring 150 nodes (buildings) with discrete curtailment strategies. Solar data was simulated based on Los Angeles County hourly radiance, modeling various PV outputs and yields to create six discrete solar curtailment levels per node.

Key Results and Scalability

Practical evaluations demonstrate that our approximation algorithms achieve costs within 2% of the theoretical optimum. Crucially, the algorithms meet the strict timing requirements of grid operators. For instance, the California ISO requires generation adjustments five minutes ahead of an interval; our framework comfortably handles grids of 40-70 nodes within these limits.

Runtime performance as a function of the number of nodes for different approximation error guarantees.

As shown in the runtime analysis, the computational overhead increases near-quadratically with the number of nodes. However, even with a 10% error guarantee, the system remains highly responsive for medium-sized distribution grids. Furthermore, by adjusting fairness parameters, we successfully distributed the curtailment burden across the grid, ensuring long-term consumer participation.

Conclusion and Future Work

This unified framework provides a proactive, scalable solution for smart grids struggling with volatile solar penetration. By treating curtailment as a discrete optimization problem and utilizing fast approximation algorithms, we provide grid operators with a tool that is both mathematically grounded and practically viable. Future work will focus on addressing forecast uncertainty in weather data and exploring decentralized control mechanisms to enhance privacy and resilience in fragmented energy markets.

Summary updated

Abstract

Mitigating Supply-Demand mismatch is critical for smooth power grid operation. Traditionally, load curtailment techniques such as Demand Response (DR) have been used for this purpose. However, these cannot be the only component of a net-load balancing framework for Smart Grids with high PV penetration. These grids can sometimes exhibit supply surplus causing over-voltages. Supply curtailment techniques such as Volt-Var Optimizations are complex and computationally expensive. This increases the complexity of net-load balancing systems used by the grid operator and limits their scalability. Recently new technologies have been developed that enable the rapid and selective connection of PV modules of an installation to the grid. Taking advantage of these advancements, we develop a unified optimal net-load balancing framework which performs both load and solar curtailment. We show that when the available curtailment values are discrete, this problem is NP-hard and develop bounded approximation algorithms for minimizing the curtailment cost. Our algorithms produce fast solutions, given the tight timing constraints required for grid operation. We also incorporate the notion of fairness to ensure that curtailment is evenly distributed among all the nodes. Finally, we develop an online algorithm which performs net-load balancing using only data available for the current interval. Using both theoretical analysis and practical evaluations, we show that our net-load balancing algorithms provide solutions which are close to optimal in a small amount of time.
BibTeX
@inproceedings{kuppannagari2017optimal,
  author = {Sanmukh R. Kuppannagari and Rajgopal Kannan and Viktor K. Prasanna},
  title = {{Optimal Net-Load Balancing in Smart Grids with High PV Penetration}},
  booktitle = {BuildSys},
  year = {2017},
  url = {https://openreview.net/forum?id=NadfdX8VZd},
}
Figure