Bidirectional Pricing and Demand Response for Nanogrids with HVAC Systems

Jiaxin Cao, Bo Yang, Shanying Zhu, Kai Ma, Xinping Guan

TL;DR

This paper presents a novel bilevel energy management framework that treats household HVAC systems as flexible "thermal batteries" to balance local renewable energy. By combining Lyapunov optimization with a Stackelberg game, the authors enable a central entity to set bidirectional prices that maximize profit while simultaneously minimizing long-term electricity costs and thermal discomfort for residents—all without needing to forecast future weather or energy prices.

What Problem Does This Paper Address?

As we integrate more solar and wind power into our homes, we face a "mismatch" problem: the sun might be shining when you aren't home to use the power, or you might need heat on a calm, cloudy night. Small-scale "nanogrids" (like a smart home with solar panels) often have energy surpluses or deficits that fluctuate wildly.

The challenge is two-fold:

  1. Time-Coupling: Decisions made now affect the future. If you turn off the heater now, the house gets colder, requiring more energy later to warm it back up. This is a "time-coupled" constraint.
  2. Uncertainty & Conflict: Future weather and prices are hard to predict. Furthermore, there is a conflict of interest: a central utility (the Public Monitoring Entity, or PME) wants to maximize profit, while homeowners want to stay comfortable without breaking the bank.

Existing solutions often require perfect "crystal ball" forecasts or are too computationally heavy to run in real-time for large neighborhoods.

Methods

The authors propose a bilevel stochastic optimization approach that breaks down into two main components:

  • Lyapunov Optimization (The "Forecasting-Free" Engine): Instead of complex dynamic programming, the authors use "virtual queues" to track indoor temperature and battery states. By minimizing the "drift" of these queues, the system makes optimal real-time decisions based only on the current state and the backlog of past needs. This removes the need for accurate weather or price forecasting.
  • Stackelberg Game (The Pricing Strategy): The interaction is modeled as a leader-follower game. The PME (Leader) sets bidirectional prices (buying and selling). The nanogrids (Followers) respond by adjusting their HVAC consumption to minimize their own costs.
  • Bidirectional Trading: Unlike traditional models where users only buy power, this system allows the PME to buy excess renewable energy from nanogrids and store it in a central battery, creating a local energy marketplace.
  • Privacy-Preserving Algorithm: They developed an iterative subgradient projection algorithm. Crucially, nanogrids only need to share their "injection power" requests with the PME, keeping private data like indoor temperature and local generation secret.

Datasets

The study validated its approach using a mix of real-world and high-fidelity simulated data:

  • Renewable Generation: Solar and wind profiles were generated using typical turbine/PV models fed with real wind speed and solar radiation data from Kaggle.
  • Basic Loads: Inelastic household electricity demand was sourced from London smart meter data.
  • Environment: Outdoor temperature data was collected from the Weather Underground (Binghamton Regional-Link).
  • Market Prices: Main grid electricity prices were taken from the Independent Electricity System Operator (IESO) in Ontario, Canada.

Key Insights and Ideas

  • HVAC as Thermal Storage: The "aha moment" is treating the thermal inertia of a building as a storage device. By shifting when an HVAC unit runs, you are effectively "storing" energy in the form of heat (or cold) within the walls and air of the house.
  • The "Middleman" Advantage: The PME can offer nanogrids better rates than the main grid (buying for more, selling for less) while still making a profit on the spread. This incentivizes local consumption and reduces strain on the larger grid.
  • Stability over Prediction: The Lyapunov approach proves that you don't need to know the future to be optimal in the long run. As long as you manage your "queues" (temperature and battery) to stay stable, you will naturally gravitate toward the most cost-effective behavior.

Key Results

  • Massive Cost Reductions: The proposed method reduced the aggregate system cost by ~76% compared to standard real-time pricing models that don't account for long-term optimization.
  • Superior Comfort: Compared to "greedy" game-theoretic models that only look at the current hour, this approach reduced thermal discomfort costs by ~85%.
  • Fast Convergence: The iterative algorithm reached a stable Stackelberg Equilibrium in approximately 35 iterations, making it fast enough for real-world deployment.
  • Scalability: Computational tests showed that the system scales polynomially, handling up to 30 nanogrids with sub-second processing times.

Limitations and Future Work

  • Idealized Hardware: The model assumes the PME's battery has "unit efficiency" (100%), which ignores real-world charging/discharging losses.
  • HVAC Control: The study assumes HVAC power can be regulated continuously. In reality, many units are simple on/off systems, which would introduce "binary" variables that make the math much harder.
  • Maintenance Costs: While the authors use an amortized cost for the battery, the maintenance and "wear and tear" on HVAC units from frequent cycling are not deeply explored.
  • Future Direction: Integrating more complex "on-off" HVAC control and non-ideal battery models are suggested next steps.

Summary updated

Abstract

Owing to the fluctuant renewable generation and power demand, the energy surplus or deficit in each nanogrid is embodied differently across time. To stimulate local renewable energy consumption and minimize the long-term energy cost, some issues still remain to be explored: when and how the energy demand and bidirectional trading prices are scheduled considering personal comfort preferences and environmental factors. For this purpose, the demand response and two-way pricing problems concurrently for nanogrids and a public monitoring entity (PME) are studied with exploiting the large potential thermal elastic ability of heating, ventilation and air-conditioning (HVAC) units. Different from nanogrids, in terms of minimizing time-average costs, PME aims to set reasonable prices and optimize profits by trading with nanogrids and the main grid bi-directionally. In particular, such bilevel energy management problem is formulated as a stochastic form in a long-term horizon. Since there are uncertain system parameters, time-coupled queue constraints and the interplay of bilevel decision-making, it is challenging to solve the formulated problems. To this end, we derive a form of relaxation based on Lyapunov optimization technique to make the energy management problem tractable without forecasting the related system parameters. The transaction between nanogrids and PME is captured by a one-leader and multi-follower Stackelberg game framework. Then, theoretical analysis of the existence and uniqueness of Stackelberg equilibrium (SE) is developed based on the proposed game property. Following that, we devise an optimization algorithm to reach the SE with less information exchange. Numerical experiments validate the effectiveness of the proposed approach.
BibTeX
@misc{cao2022bidirectional,
  author = {Jiaxin Cao and Bo Yang and Shanying Zhu and Kai Ma and Xinping Guan},
  title = {{Bidirectional Pricing and Demand Response for Nanogrids with HVAC Systems}},
  year = {2022},
  eprint = {2203.00270},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2203.00270},
}
Figure