New Market Model Boosts Efficiency in Gas-Electric Distribution Networks
In an era defined by energy transition and digital transformation, the integration of distributed energy resources (DERs) into local grids is no longer a futuristic vision—it’s an operational reality. Yet, as solar panels, battery storage systems, and electric vehicles (EVs) proliferate across neighborhoods and industrial parks, grid operators face a mounting challenge: how to coordinate these decentralized assets efficiently while maintaining system reliability and economic fairness. A groundbreaking study published in Automation of Electric Power Systems offers a compelling answer by introducing a novel market-clearing framework tailored for gas-electric integrated distribution networks that actively incorporates DERs into day-ahead operations.
The research, led by Yixuan Weng of Shenzhen Power Supply Bureau Co., Ltd., in collaboration with scholars from Tsinghua University and Beijing QU Creative Technology Co., Ltd., presents a centralized trading mechanism designed specifically for medium-voltage distribution systems where electricity and natural gas infrastructures are tightly coupled. Unlike peer-to-peer (P2P) models that prioritize autonomy over system-wide efficiency, this approach leverages locational marginal pricing (LMP) at the distribution level to align individual participant incentives with broader grid objectives—maximizing social welfare, minimizing congestion, and enhancing renewable integration.
At the heart of the innovation lies a dual-system optimization architecture that respects the operational and institutional boundaries between electric and gas utilities. Historically, these networks have been managed by separate entities with distinct regulatory mandates, data protocols, and scheduling horizons. Attempting to merge their operations into a single monolithic model often proves impractical due to data privacy concerns and computational complexity. The authors circumvent this hurdle by adopting the Alternating Direction Method of Multipliers (ADMM)—a distributed optimization technique that enables iterative coordination without requiring full disclosure of proprietary information.
This method allows the electric distribution network and the natural gas distribution network to solve their respective subproblems independently while exchanging only limited, anonymized signals related to coupling devices—namely, gas-fired turbines and power-to-gas (P2G) units. These assets serve as the physical and economic bridges between the two energy carriers. When electricity prices surge during peak demand, gas turbines can ramp up generation to meet local load, reducing strain on upstream transmission lines. Conversely, when surplus renewable generation drives electricity prices down, P2G facilities can convert excess power into synthetic methane, injecting it into the gas grid for later use or storage. This bidirectional flexibility is key to unlocking system resilience and economic value, but only if market signals accurately reflect real-time conditions across both networks.
To make this vision computationally tractable, the team employed advanced mathematical techniques to linearize inherently nonlinear physics. For the electric distribution network, they used second-order cone programming (SOCP) relaxation to approximate the AC power flow equations—a common approach that balances accuracy with solvability in radial networks. For the gas network, however, they opted for a different strategy: Taylor series expansion around operating points to linearize the Weymouth equation, which governs pressure-flow relationships in pipelines. While SOCP has been widely applied to power systems, its use in gas networks remains controversial due to potential inaccuracies under certain flow regimes. By choosing Taylor expansion—a method that adapts to local operating conditions—the authors ensured higher fidelity in gas flow modeling without sacrificing convexity.
The resulting joint clearing model optimizes the dispatch of not only conventional assets like gas turbines but also diverse DERs: rooftop photovoltaics, stationary battery storage, and aggregated EV charging stations. Each participant submits bid curves reflecting their willingness to buy or sell energy at different price levels. Photovoltaic prosumers offer surplus generation; batteries arbitrage price differences by charging when prices are low and discharging when they are high; EV fleets shift charging loads to off-peak hours while guaranteeing minimum state-of-charge requirements for mobility needs. The market operator then clears these bids subject to network constraints, producing a schedule that maximizes total surplus while respecting voltage limits, thermal ratings, and gas pressure bounds.
Crucially, the model outputs distribution-level LMPs—prices that vary by node and time to reflect local congestion and losses. These prices serve as powerful economic signals: they incentivize DERs to locate and operate where they provide the most grid value. For instance, a battery installed near a heavily loaded feeder might earn higher revenues by discharging during evening peaks, not just because energy prices are high, but because it alleviates local congestion, captured in the LMP’s “congestion component.” Similarly, EV charging stations in areas with abundant midday solar might be rewarded for absorbing excess generation, helping to prevent curtailment.
The researchers validated their framework using a modified IEEE 33-node distribution system coupled with a 7-node gas network. Simulations revealed compelling dynamics. During low-price periods (e.g., overnight and midday), the system imported power from the wholesale market and charged storage assets. During high-price windows (late afternoon and early evening), local generation—including gas turbines and discharged batteries—supplied the majority of demand, reducing reliance on expensive imports. The P2G units operated strategically: they curtailed conversion when electricity prices exceeded the marginal cost of gas-fired generation, preserving economic efficiency.
Perhaps most significantly, the study compared the proposed centralized mechanism against a P2P trading baseline under varying levels of DER penetration. As the share of distributed resources increased from 20% to 100%, the welfare advantage of the centralized model grew substantially. At full penetration, social welfare—defined as the net benefit to all participants after accounting for costs—was nearly 40% higher than in the P2P scenario. This gap arises because P2P trades, while flexible, often ignore system-wide constraints. Without a central coordinator, participants may agree to transactions that collectively overload a transformer or violate voltage limits, forcing the grid operator to curtail trades post-hoc—a process that erodes trust and efficiency. The centralized model, by contrast, internalizes these constraints upfront, ensuring every cleared transaction is physically feasible and economically optimal.
From a policy perspective, the findings carry important implications. As regulators worldwide contemplate the design of distribution-level markets—especially in regions with high renewable penetration—the trade-off between decentralization and coordination becomes critical. Pure P2P models may appeal to libertarian ideals of consumer sovereignty, but they risk suboptimal outcomes in tightly constrained networks. The authors’ approach strikes a pragmatic balance: it preserves participant autonomy in bid submission while ensuring system integrity through centralized clearing. Moreover, by using ADMM, it accommodates institutional realities—electric and gas utilities can remain separate legal and operational entities while still achieving joint optimization.
The model also aligns with global trends toward integrated energy systems. The European Union’s “Fit for 55” package and the U.S. Inflation Reduction Act both emphasize sector coupling—linking power, transport, heating, and industry through flexible conversion technologies. Gas-electric integration is a foundational step in this journey, and efficient market mechanisms are essential to unlock its potential. By demonstrating how price signals can coordinate cross-sectoral assets in real distribution networks—not just theoretical transmission systems—the study provides a blueprint for practical implementation.
Industry experts note that the timing of this research is particularly apt. With the rapid rollout of smart meters, advanced inverters, and vehicle-to-grid (V2G) pilots, the technical infrastructure for DER participation is increasingly in place. What’s missing is the market layer—the rules, pricing mechanisms, and settlement protocols that turn hardware into grid services. The proposed clearing model fills this gap by offering a scalable, privacy-preserving, and economically sound framework that can be adapted to different regulatory contexts.
Looking ahead, the authors acknowledge limitations and suggest avenues for future work. Their current model focuses on day-ahead scheduling and assumes perfect forecasts—a simplification that real-world operators must address through real-time adjustments and uncertainty modeling. They also plan to extend the framework to include reactive power support, thermal loads, and hydrogen blending in gas networks. Ultimately, their vision is a hierarchical market architecture where distribution-level platforms interface seamlessly with wholesale markets, enabling end-to-end coordination from national interconnectors down to individual EVs.
For grid planners, this research offers more than theoretical insight—it provides a tested methodology for managing the energy transition at the local level. As distributed resources shift from passive connections to active market participants, the need for intelligent, integrated, and incentive-compatible mechanisms has never been greater. This study demonstrates that with the right blend of economic theory, mathematical innovation, and engineering pragmatism, such mechanisms are not only possible but already within reach.
Authors: Yixuan Weng¹, Weizhe Ma¹, Fuquan Huang¹, Zizhao Lin¹, Xue Liu²,³, Yujun He²,³, Deliang Zhang⁴
Affiliations:
¹ Shenzhen Power Supply Bureau Co., Ltd., Shenzhen 518020, China
² State Key Laboratory of Power System and Generation Equipment, Tsinghua University, Beijing 100084, China
³ Department of Electrical Engineering, Tsinghua University, Beijing 100084, China
⁴ Beijing QU Creative Technology Co., Ltd., Beijing 100084, China
Journal: Automation of Electric Power Systems
DOI: 10.7500/AEPS20230518003