Abstract
The importance of reactive power compensation for power factor (PF) correction will significantly increase with the large-scale integration of distributed generation interfaced via inverters producing only active power. In this work, we focus on co-optimizing energy storage for performing energy arbitrage as well as local power factor correction. The joint optimization problem is non-convex, but can be solved efficiently using a McCormick relaxation along with penalty-based schemes. Using numerical simulations on real data and realistic storage profiles, we show that energy storage can correct PF locally without reducing arbitrage profit. It is observed that active and reactive power control is largely decoupled in nature for performing arbitrage and PF correction (PFC). Furthermore, we consider a real-time implementation of the problem with uncertain load, renewable and pricing profiles. We develop a model predictive control based storage control policy using auto-regressive forecast for the uncertainty. We observe that PFC is primarily governed by the size of the converter and therefore, look-ahead in time in the online setting does not affect PFC noticeably. However, arbitrage profit are more sensitive to uncertainty for batteries with faster ramp rates compared to slow ramping batteries.
Introduction
As rooftop solar (distributed generation) proliferates, it creates a significant challenge for grid operators: power factor (PF) degradation. Most small-scale solar inverters are designed to produce only "active power" (P), which performs the actual work. However, the grid also requires "reactive power" (Q) to maintain magnetic fields in inductive loads like motors and transformers.
When solar panels provide the active power but not the reactive power, the ratio between the two—the Power Factor—drops. A low PF leads to higher currents, excessive thermal losses, and potential equipment damage. While large-scale plants have strict PF rules, small residential systems often do not, leading to a systematic degradation of the grid's load profile. This study investigates whether residential energy storage can provide the missing reactive power while simultaneously performing its primary economic function: energy arbitrage.
Methodology
The authors frame the challenge as a co-optimization problem. The core difficulty is that the joint optimization of active and reactive power is non-convex, making it computationally intensive to solve efficiently. To address this, the following strategies were employed:
- McCormick Relaxation: The non-convex PF constraints are reformulated into a mixed-integer convex problem, allowing for rapid solution by standard industrial solvers like Gurobi or Mosek.
- Penalty-Based Schemes: A penalty function for PF violations is introduced, enabling the system to prioritize PFC or arbitrage based on the severity of grid requirements.
- Model Predictive Control (MPC): For real-time application, MPC is used to make decisions over a rolling horizon, accounting for future uncertainties.
- Forecasting: Auto-Regressive Moving Average (ARMA) models predict future load and solar generation, while ARIMA models forecast volatile electricity prices.
Experimental Setup
The study utilizes high-resolution, real-world data to validate the proposed models. This includes 15-minute averaged household consumption and solar generation data from Madeira, Portugal, as part of the EU SMILE project.
Additionally, real-time electricity pricing data from the New York ISO (NYISO) and the California ISO (CAISO) were used. Various battery profiles were simulated, ranging from slow-ramping systems (e.g., Tesla Powerwalls) to fast-ramping technologies (e.g., flywheels).
Results and Discussion
The research yields several key insights regarding the synergy between arbitrage and power factor correction:
- The Decoupling Effect: Active and reactive power control are found to be largely decoupled. Because the battery's inverter often has spare capacity when not charging or discharging at maximum power, this capacity can be utilized for PFC at virtually no cost to the arbitrage mission.
- PFC is Myopic: Unlike arbitrage, which requires a look-ahead window to anticipate price changes, PFC is primarily a real-time requirement. Look-ahead windows do not significantly improve PFC performance.
- Sensitivity to Forecasting: Batteries with faster ramp rates are more sensitive to forecasting errors. Incorrect price forecasts can lead to greater profit loss for fast-ramping batteries compared to slow-ramping ones.
- Converter Sizing: The capacity to correct the power factor is limited primarily by the size of the inverter (converter) rather than the battery's energy capacity.
In deterministic simulations, arbitrage profit remained nearly identical regardless of whether PFC was performed. Furthermore, while solar integration without storage dropped the minimum PF from 0.82 to 0.15, the proposed storage control successfully maintained PF at or above the 0.90 limit required by many utilities.
Conclusion
This work demonstrates that energy storage systems can effectively co-optimize for energy arbitrage and local power factor correction. The decoupling of these tasks allows for grid stabilization without sacrificing economic returns for the consumer. Future work may involve integrating more detailed battery aging models, expanding to feeder-level control for neighborhood coordination, and analyzing the financial incentive structures necessary to encourage residential reactive power provision.
Summary updated