Distributed energy trading management for renewable prosumers with HVAC and energy storage

Qing Yang, Hao Wang

TL;DR

This paper introduces a distributed optimization framework that allows smart home "prosumers" to cooperatively manage their HVAC systems and battery storage through peer-to-peer (P2P) energy trading. By using the Alternating Direction Method of Multipliers (ADMM), the system achieves a 23% average reduction in energy costs while keeping users' private operational data—like temperature preferences and battery states—strictly local.

What Problem Does This Paper Address?

Heating, ventilation, and air-conditioning (HVAC) systems are the "heavy lifters" of residential energy consumption, often accounting for the largest portion of a household's utility bill. While renewable energy (like solar) and home batteries (like the Tesla Powerwall) help, managing them in isolation is inefficient.

The authors identify a "privacy vs. efficiency" deadlock:

  1. Centralized Management: A grid operator could optimize everyone's usage perfectly, but this requires users to share sensitive data (occupancy patterns, temperature preferences, etc.), raising significant privacy concerns.
  2. Standalone Management: Users optimize their own homes privately but miss out on the benefits of trading surplus energy with neighbors who might have different demand patterns.

The challenge is to create a "transactive energy" system where neighbors can trade energy to balance the grid and lower costs without a central authority snooping on their private lives.

Methods

The researchers developed a distributed energy management model based on ADMM (Alternating Direction Method of Multipliers).

  • The Model: Each home is modeled as a "prosumer" with renewable generation, an HVAC system (with thermal inertia), a Battery Energy Storage System (BESS), and inflexible loads. The HVAC model specifically accounts for the building's thermal resistance and heat capacity.
  • The Algorithm: ADMM allows the complex global optimization problem to be decoupled into:
  • Primal Subproblems: Solved locally by each user’s smart meter. This handles internal scheduling (when to run the AC, when to charge the battery).
  • Dual Subproblems: Solved by a coordinator (the grid operator). This handles the "market clearing" for P2P trading.
  • Privacy Preservation: Crucially, users only share their trading decisions (how much they want to buy or sell) with the coordinator. They never share their battery levels, indoor temperatures, or specific appliance usage.
  • Cost Functions: The optimization minimizes a combination of grid electricity payments (including peak-load charges), battery degradation costs, and "discomfort costs" (the quadratic penalty for deviating from a preferred indoor temperature).

Datasets

The study validated the algorithm using real-world data over a one-week simulation period:

  • Renewable Generation: Real-world solar and wind power data.
  • Inflexible Loads: Residential power consumption data from the Pecan Street Dataport, a well-known industry benchmark for high-fidelity building energy data.
  • Environment: Real outdoor temperature data (specifically from Hong Kong for the summer simulation and Los Angeles for the winter simulation).
  • Hardware Specs: Parameters modeled after commercial products, such as the Tesla Powerwall 2.

Key Insights and Ideas

  • Thermal Inertia as Storage: The paper treats the home's indoor temperature as a form of "virtual storage." By allowing the temperature to fluctuate slightly within a comfortable range, the HVAC system can shift its load to times when renewable energy is abundant or P2P prices are low.
  • Synergy of HVAC and BESS: While batteries provide explicit storage, combining them with HVAC optimization magnifies the benefits of P2P trading. One neighbor's surplus solar can be stored in another neighbor's battery or used to "pre-cool" their house.
  • Peak Shaving via Trading: The distributed approach naturally results in "peak shaving." Because the grid operator charges more for peak usage, the ADMM algorithm incentivizes users to trade amongst themselves to keep their individual and collective grid draws flat.
  • Fast Convergence: Despite being a distributed iterative process, the algorithm converged in just 26 iterations, making it computationally feasible for real-time smart grid applications.

Key Results

  • Cost Savings: The proposed cooperative P2P trading method reduced total energy costs by 23% on average compared to standalone management.
  • Individual Benefits: Every user in the 10-home simulation saw a cost reduction, though the scale varied (from 14% to 83%) depending on their specific renewable capacity and demand.
  • Grid Impact: The system successfully demonstrated peak shaving, reducing the stress on the main power grid during high-demand periods.
  • Seasonal Robustness: While summer savings were 23%, a winter simulation showed a 13.4% reduction, proving the method works across different climates, though the benefit is higher when HVAC demand is more intense.

Limitations and Future Work

  • Power Loss: The current model omits electricity loss during transfer between neighbors, assuming they are physically close. In a real-world distribution network, these losses could impact the economic feasibility of certain trades.
  • Complexity vs. Scale: While it works for 10–50 users, the authors acknowledge that further work is needed to reduce computational complexity for much larger networks.
  • Trust and Security: The authors suggest that blockchain technology could be integrated in the future to create a truly decentralized, "trustless" trading platform.
  • Hardware ROI: The study focuses on operational costs but does not perform a full lifecycle analysis or calculate the payback period for the required hardware (batteries and smart meters).

Summary updated

Abstract

Heating, ventilating, and air-conditioning (HVAC) systems consume a large amount of energy in residential houses and buildings. Effective energy management of HVAC is a cost-effective way to improve energy efficiency and reduce the energy cost of residential users. This work develops a novel distributed method for the residential transactive energy system that enables multiple users to interactively optimize their energy management of HVAC systems and behind-the-meter batteries. Specifically, this method effectively reduces the cost of smart homes by employing energy trading among users to leverage their power usage flexibility without compromising the users' privacy. To achieve this goal, we design a distributed optimization algorithm based on the alternating direction method of multipliers (ADMM) to automatically operate the HVAC system and batteries, which minimizes the energy costs of users. Specifically, we decouple the optimization problem into a primal subproblem and a dual subproblem. The primal subproblem is solved by the users, and the dual subproblem is solved by the grid operator. Unlike the existing centralized method, our approach only uses the users' private information locally for solving the primal subproblem hence preserves the users' privacy. Using real-world data, we validate our proposed algorithm through extensive simulations in Matlab. The results demonstrate that our method effectively incentivizes the energy trading among the users to reduce users' peak load and reduce the overall energy cost of the system by 23% on average.
BibTeX
@inproceedings{yang2021distributed,
  author = {Qing Yang and Hao Wang},
  title = {{Distributed energy trading management for renewable prosumers with HVAC and energy storage}},
  booktitle = {Energy Reports},
  year = {2021},
  doi = {10.1016/j.egyr.2021.03.038},
}

Loading related content...

Figure