Open AccessArticle

A Three-Layer Scheduling Framework with Dynamic Peer-to-Peer Energy Trading for Multi-Regional Power Balance

Tianmeng Yang

¹,

Jicheng Liu

¹,

Wei Feng

¹,

Zelong Chen

¹,

Yumin Zhao

¹ and

Suhua Lou

^2,*

Northeast Branch of State Grid Corporation of China, Shenyang 110180, China

State Key Laboratory of Advanced Electromagnetic Engineering and Technology, Wuhan 430074, China

Author to whom correspondence should be addressed.

Energies 2024, 17(24), 6239; https://doi.org/10.3390/en17246239

Submission received: 18 November 2024 / Revised: 7 December 2024 / Accepted: 10 December 2024 / Published: 11 December 2024

(This article belongs to the Section A1: Smart Grids and Microgrids)

Download

Browse Figures

Figure 1
Multi-regional smart grid framework for energy interaction and coordinated scheduling. "> Figure 2
The model solution flowchart used in this paper. "> Figure 3
Preliminary balance analysis of renewable energy and load demand for all regions. "> Figure 4
Power balance analysis for all regions. "> Figure 5
Regional pricing and power trading situations across different time periods. "> Figure 6
Success rate assessment of iterative trading process between buyers and sellers across different time periods. "> Figure 7
Typical day 2 energy trading results. "> Figure 7 Cont.
Typical day 2 energy trading results. "> Figure 8
Typical day 3 energy trading results. "> Figure 9
Typical day 4 energy trading results. ">

Versions Notes

Abstract

This paper addresses the critical challenges of renewable energy integration and regional power balance in smart grids, which have become increasingly complex with the rapid growth of distributed energy resources. It proposes a novel three-layer scheduling framework with a dynamic peer-to-peer (P2P) trading mechanism to address these challenges. The framework incorporates a preliminary local supply–demand balance considering renewable energy, followed by an inter-regional P2P trading layer and, ultimately, flexible resource deployment for final balance adjustment. The proposed dynamic continuous P2P trading mechanism enables regions to autonomously switch roles between buyer and seller based on their internal energy status and preferences, facilitating efficient trading while protecting regional privacy. The model features an innovative price update mechanism that initially leverages historical trading data and dynamically adjusts prices to maximize trading success rates. To address the heterogeneity of regional resources and varying energy demands, the framework implements a flexible trading strategy that allows for differentiated transaction volumes and prices. The effectiveness of the proposed framework is validated through simulation experiments using k-means clustered typical daily data from four regions in Northeast China. The results demonstrate that the proposed approach successfully promotes renewable energy utilization, reduces the operational costs of flexible resources, and achieves an efficient inter-regional energy balance while maintaining regional autonomy and information privacy.

Keywords:

P2P energy trading; renewable energy integration; dynamic price mechanism; regional power balance; flexible resource schedule

1. Introduction

In recent years, the energy landscape has faced various uncertainties, both domestically and globally [1]. There is a significant imbalance between energy supply and demand. Developing renewable energy is a powerful complement to optimizing the energy mix and achieving green, low-carbon development goals [2,3]. However, a renewable energy output is closely tied to geographic and climatic factors, resulting in varying production capacities across different regions [4,5]. Furthermore, the scale of renewable energy infrastructure that is built in different regions also leads to variations in output [6]. Renewable energy’s intermittent nature or suboptimal infrastructure design can cause oversupply in some regions, leading to curtailment, or undersupply, requiring other resources to meet local load requirements [7,8].

In this context, leveraging flexible resources to regulate the power system is an effective solution for balancing supply and demand. Within a regional grid, various methods can be used to coordinate the variability of renewable energy: utilizing energy storage [9,10] such as pumped storage [11,12] and batteries to shift surplus renewable energy from peak production to periods of high demand [13,14]; and flexibly dispatching conventional thermal power plants to supplement shortfalls in the renewable energy output [15,16]. This diversified power structure can ensure the reliability of an electricity supply and improve the integration of renewable energy.

However, the operational costs of flexible resources during this balancing process can be significant, especially when there is a substantial renewable energy surplus or load deficiency. This single approach may not be well suited to the trend of large-scale renewable energy development. Currently, encouraging regions with renewable energy surpluses to sell their excess to regions with shortages, through a peer-to-peer (P2P) trading mechanism, can not only promote the utilization of renewable energy and increase the revenue of the selling regions but also reduce the operating costs of flexible resources and the costs for the purchasing regions (as the cost of renewable energy is generally lower than the operating cost of flexible resources). To implement this method, P2P trading has emerged as a new energy management model for the next-generation smart grid due to its scalability and intelligence [17,18].

The P2P trading concept has been widely applied in the field of distributed computing for resource sharing. P2P trading is a protocol for exchanging products or information between two peers, which can be further divided into centralized P2P coordination trading and decentralized P2P distributed trading. The former requires a third-party centralized structure for coordinated trading [19,20], while the latter avoids the drawbacks of centralized trading [21], such as potential privacy infringements due to regional information sharing [22,23] and decreased computational efficiency due to data centralization [24].

Additionally, differences in the geographic environment and internal resource characteristics across regions lead to heterogeneity. The designed P2P energy trading market should respect the differentiated preferences of market participants, allowing for bilateral energy trading with varying transaction quantities and prices. Since economic considerations are the driving force behind P2P trading, appropriate bidding strategies need to be designed [25,26]. Some researchers have introduced other participants [27] or upstream agents [28] to unify pricing, but in the actual trading market, regional managers may be reluctant to provide their own energy data to others. Therefore, this paper first uses the historical trading price data of each region as the initial trading price. As the entire trading process is continuous, using a fixed trading price or directly setting a price range while guaranteeing the fairness and continuity of the trading would be difficult.It may also lead to low trading success rates and returns due to arbitrary initial bids. Hence, a suitable price update mechanism must be developed to facilitate the continuous auction process, making the entire P2P market more dynamic, stochastic, and able to adapt to the energy demand situations of different regions in different time periods. This paper adopts a price update mechanism that aims to maximize the trading success rate.

In summary, the main contributions of this paper are as follows:

1. We designed a three-layer energy dispatching process to address the energy regulation needs of regions with heterogeneous internal resources. First, we used local renewable energy and load demand values for preliminary supply–demand balancing; second, we used a continuous P2P trading mechanism for power exchange between regions with surpluses or deficits; and third, we employed flexible resources to determine the ultimate supply–demand balance.

2. We designed a dynamic trading price mechanism. Initially, we used historical trading price data for P2P trading and then updated the prices based on the goal of maximizing the trading success rate for the current time period.

3. We designed a dynamic continuous P2P trading mechanism, where different regions can autonomously choose to be sellers or buyers during different time periods based on their own preferences and internal energy balance situations, effectively reducing the risk of leaking regional energy information.

4. We designed simulation experiments using representative and typical daily data generated through k-means clustering of actual data from four regions in Northeast China, demonstrating the model’s effectiveness.

2. Regional Models Incorporating Differentiated Characteristics

Geographic variations across regions lead to distinct characteristics in their renewable energy installations, power generation capacity, and energy consumption patterns. Each region’s energy configuration model is designed based on its unique power generation potential, development conditions, and energy demands. The generation resources vary by region:

Region 1: Coal, hydropower, pumped storage, wind, and solar power;
Regions 2 and 3: Coal, pumped storage, wind, and solar power;
Region 4: Coal, electrochemical storage, wind, and solar power.

To maximize the renewable energy utilization while minimizing the reliance on coal-fired generation, this paper proposes a three-tier energy allocation model. The model prioritizes meeting load demands through wind, solar, and hydropower; facilitating inter-regional power exchange, prioritizing renewable energy transfers between regions; and achieving a supply–demand balance through local resources (coal plants, pumped storage, and battery storage). Figure 1 shows a schematic diagram of the energy interaction and cooperative scheduling framework for a multi-region smart grid.

2.1. Objective

Each region optimizes its 24 h operations to minimize costs while maintaining a supply–demand balance. For each region, the objective is to minimize costs over a 24 h day. The objective function for the specific supply–demand balance of each region i is as follows:

min c_{i}^{t o t a l} = c_{i}^{e x c h a n g e} + c_{i}^{i n n e r}, i = 1, 2, 3, 4

(1)

c_{i}^{e x c h a n g e} = \sum_{t = 1}^{T} λ_{i, t}^{e x c h a n g e} p_{i, t}^{e x c h a n g e}, i = 1, 2, 3, 4

(2)

λ_{i, t}^{e x c h a n g e} = \{\begin{matrix} λ_{i, t}^{e x c h a n g e, b u y} & p_{i, t}^{e x c h a n g e} > 0 \\ λ_{i, t}^{e x c h a n g e, s e l l} & p_{i, t}^{e x c h a n g e} < 0 \end{matrix}

(3)

\begin{matrix} c_{1}^{i n n e r} = \sum_{t = 1}^{T} (c_{t}^{p s} + c_{t}^{c} + c_{t}^{w t} + c_{t}^{p v}) \\ s . t . p_{t}^{e x c h a n g e} + p_{t}^{c} + p_{t}^{h} + p_{d, t}^{p s} - p_{c, t}^{p s} + p_{t}^{w t} + p_{t}^{p v} = p_{t}^{l o a d} + p_{c u r, t}^{w t} + p_{c u r, t}^{p v} \\ c_{2}^{i n n e r} = c_{3}^{i n n e r} = \sum_{t = 1}^{T} (c_{t}^{p s} + c_{t}^{c} + c_{t}^{w t} + c_{t}^{p v}) \\ s . t . p_{t}^{e x c h a n g e} + p_{t}^{c} + p_{d, t}^{p s} - p_{c, t}^{p s} + p_{t}^{w t} + p_{t}^{p v} = p_{t}^{l o a d} + p_{c u r, t}^{w t} + p_{c u r, t}^{p v} \\ c_{4}^{i n n e r} = \sum_{t = 1}^{T} (c_{t}^{e s s} + c_{t}^{c} + c_{t}^{w t} + c_{t}^{p v}) + \frac{1}{360} c_{i n}^{e s s} \\ s . t . p_{t}^{e x c h a n g e} + p_{t}^{c} + p_{d, t}^{e s s} - p_{c, t}^{e s s} + p_{t}^{w t} + p_{t}^{p v} = p_{t}^{l o a d} + p_{c u r, t}^{w t} + p_{c u r, t}^{p v} \end{matrix}

(4)

where

c_{i}^{t o t a l}

represents the total operational cost for region i;

c_{i}^{e x c h a n g e}

denotes the cost of electricity trading (buying and selling) with other regions, where positive values indicate power purchases and negative values indicate power sales;

c_{i}^{i n n e r}

represents the generation cost of flexible resources within each region (detailed in Equation (4));

λ_{i, t}^{e x c h a n g e, b u y}

and

λ_{i, t}^{e x c h a n g e, s e l l}

represent the electricity buying and selling prices for region i’s transactions with other regions, respectively; and

p_{i, t}^{e x c h a n g e}

indicates the power exchange requirements (Step 2) for transactions with other regions. In Equation (4), the additional constraints for objective function

c_{1}^{i n n e r}

are Equations (11), (12) and (17)–(24). For objective functions

c_{2}^{i n n e r}

and

c_{3}^{i n n e r}

, the additional constraints are Equations (11), (12) and (17)–(22). The additional constraints for objective function

c_{4}^{i n n e r}

are Equations (11)–(16), (21) and (22).

Inter-regional power exchange offers distinct advantages over other flexibility sources in power systems. These advantages include enhanced renewable energy integration, reduced generation costs, and environmental benefits. Therefore, our approach prioritizes P2P energy trading between regions before utilizing local generation units for residual supply–demand balancing. The specific bidding strategies and trading procedures are detailed in Section 3. Equation (4) defines the objective function and constraints for balancing the residual supply and demand using local generation resources across regions.

2.1.1. Annual Investment Cost Analysis of Distributed Energy Storage Systems

The annual investment costs for distributed energy storage systems are influenced by factors such as the installation capacity and rated transmission power, which can be defined as follows:

c_{i n}^{e s s} = [\sum_{m} (λ_{e, m}^{e s s} E_{m}^{e s s} + λ_{p, m}^{e s s} p_{m}^{e s s})] \frac{τ {(1 + τ)}^{T_{L}}}{{(1 + τ)}^{T_{L}} - 1}

(5)

where

c_{i n}^{e s s}

represents the annual investment cost of electrochemical storage m;

λ_{e, m}^{e s s}

represents the unit capacity cost;

λ_{p, m}^{e s s}

denotes the unit power cost;

E_{m}^{e s s}

indicates the rated energy capacity;

p_{m}^{e s s}

represents the rated power capacity;

T_{L}

denotes the system lifetime; and

τ

represents the discount rate.

2.1.2. Cost of Wind and Solar Power Curtailment

Wind and solar power, as clean renewable sources, have minimal operational costs; their main economic impact stems from curtailment, which can be quantified as follows:

c_{t}^{w t} = λ^{w t} p_{c u r, t}^{w t}

(6)

c_{t}^{p v} = λ^{p v} p_{c u r, t}^{p v}

(7)

where

c_{t}^{w t}

and

c_{t}^{p v}

represent the curtailment costs of wind and solar power;

λ^{w t}

and

λ^{p v}

denote the penalty coefficients; and

p_{c u r, t}^{w t}

and

p_{c u r, t}^{p v}

represent the curtailed power.

2.1.3. Generation Cost of Coal-Fired Power Units

Coal-fired units offer high flexibility and strong load-following capabilities, enabling the intelligent adjustment to grid load variations to maintain supply–demand stability. Their generation costs are primarily operational.

c_{t}^{c} = a^{c} {(p_{t}^{c})}^{2} + b^{c} p_{t}^{c} + c^{c}

(8)

where

c_{t}^{c}

represents the fuel cost of coal-fired units;

p_{t}^{c}

denotes the operating power; and

a^{c}

b^{c}

, and

c^{c}

represent the cost coefficients.

2.1.4. Operational Cost of Pumped Storage and Electrochemical Energy Storage Plants

Different regions employ either pumped storage or electrochemical storage stations based on their geographical conditions. While these storage facilities effectively manage peak shifting and valley filling, frequent cycling leads to energy curtailment and equipment wear. Therefore, the objective function incorporates the operational costs of storage stations.

c_{o p, t}^{e s s} = λ^{e s s} (p_{c, t}^{e s s} + p_{d, t}^{e s s})

(9)

c_{t}^{p s} = λ^{p s} (p_{c, t}^{p s} + p_{d, t}^{p s})

(10)

where

c_{o p, t}^{e s s}

and

c_{t}^{p s}

represent the operational costs of battery storage and pumped storage respectively;

λ^{e s s}

and

λ^{p s}

denote the cost coefficients; and

p_{c, t}^{e s s}

p_{d, t}^{e s s}

p_{c, t}^{p s}

, and

p_{d, t}^{p s}

represent the charging and discharging power of battery storage and pumped storage.

2.2. Constraints

2.2.1. Abandonment Constraints for Wind and PV Units

0 \leq p_{c u r, t}^{w t} \leq p_{t}^{w t}

(11)

0 \leq p_{c u r, t}^{p v} \leq p_{t}^{p v}

(12)

2.2.2. Operational Constraints of Electrochemical Energy Storage

s_{t}^{e s s} = s_{t - 1}^{e s s} + \frac{p_{c, t}^{e s s} η_{c}^{e s s}}{E^{e s s}} Δ t - \frac{p_{d, t}^{e s s}}{E^{e s s} η_{d}^{e s s}} Δ t

(13)

\{\begin{matrix} 0 \leq p_{c, t}^{e s s} \leq p_{m a x}^{e s s} \\ 0 \leq p_{d, t}^{e s s} \leq p_{m a x}^{e s s} \\ p_{c, t}^{e s s} \cdot p_{d, t}^{e s s} = 0 \end{matrix}

(14)

s_{m i n}^{e s s} \leq s_{t}^{e s s} \leq s_{m a x}^{e s s}

(15)

\{\begin{matrix} 0.3 E^{e s s} \leq s_{T}^{e s s} \leq 0.7 E^{e s s} \\ s_{0}^{e s s} = 0.5 E^{e s s} \end{matrix}

(16)

where

s_{t}^{e s s}

represents the state of charge (SOC) of battery storage at an interval t;

η_{c}^{e s s}

and

η_{d}^{e s s}

denote the charging and discharging efficiencies;

E^{e s s}

represents the rated capacity; Equations (14) and (15) represent the constraints of power and SOC variations; and Equation (16) ensures the constraints of the initial and final SOC.

2.2.3. Operational Constraints of Pumped Storage Hydroelectric Systems

s_{t}^{p s} = s_{t - 1}^{p s} + \frac{p_{c, t}^{p s} η_{c}^{p s}}{E^{p s}} Δ t - \frac{p_{d, t}^{p s}}{E^{p s} η_{d}^{p s}} Δ t

(17)

\{\begin{matrix} 0 \leq p_{c, t}^{p s} \leq p_{m a x}^{p s} \\ 0 \leq p_{d, t}^{p s} \leq p_{m a x}^{p s} \\ p_{c, t}^{p s} \cdot p_{d, t}^{p s} = 0 \\ - Δ p_{m a x}^{p s} \leq p_{c, t}^{p s} - p_{d, t}^{p s} - p_{c, t - 1}^{p s} + p_{d, t - 1}^{p s} \leq Δ p_{m a x}^{p s} \end{matrix}

(18)

s_{m i n}^{p s} \leq s_{t}^{p s} \leq s_{m a x}^{p s}

(19)

\{\begin{matrix} 0.3 E^{p s} \leq s_{T}^{p s} \leq 0.7 E^{p s} \\ s_{0}^{p s} = 0.5 E^{p s} \end{matrix}

(20)

where

s_{t}^{p s}

represents the SOC of pumped storage at an interval t;

η_{c}^{p s}

and

η_{d}^{p s}

denote the charging and discharging efficiencies;

E^{p s}

represents the rated capacity; Equations (18) and (19) represent the constraints of power and SOC variations; and Equation (20) ensures the constraints of the initial and final SOC.

2.2.4. Output Constraints of Coal-Fired Power Plants

θ_{t}^{c} p_{m i n}^{c} \leq p_{t}^{c} \leq θ_{t}^{c} p_{m a x}^{c}

(21)

\{\begin{matrix} p_{t}^{c} - p_{t - 1}^{c} \leq Δ p_{u}^{c} \\ p_{t - 1}^{c} - p_{t}^{c} \leq Δ p_{d}^{c} \end{matrix}

(22)

where

p_{m i n}^{c}

and

p_{m a x}^{c}

represent the minimum and maximum power output of coal-fired units;

θ_{t}^{c}

denotes the unit commitment status; and

Δ p_{u}^{c}

and

Δ p_{d}^{c}

represent the ramp-up and ramp-down power constraints.

2.2.5. Output Constraints of Hydropower Plants

0 \leq p_{t}^{h} \leq p_{m a x}^{h}

(23)

\{\begin{matrix} P_{t}^{h} - p_{t - 1}^{h} \leq Δ p_{u}^{h} \\ P_{t - 1}^{h} - p_{t}^{h} \leq Δ p_{d}^{h} \end{matrix}

(24)

where

p_{m a x}^{h}

represents the maximum power output of hydroelectric units;

Δ p_{u}^{h}

and

Δ p_{d}^{h}

denote the ramp-up and ramp-down power constraints.

3. Inter-Regional Peer-to-Peer Energy Trading Mechanism

3.1. Bidding Strategy Optimization in Energy Markets

After meeting the initial local demand with their own renewable energy resources, regions engage in power trading to address remaining supply–demand imbalances.

In each bidding round, regions determine their bid/offer prices with the objective of maximizing expected economic benefits. The optimal bidding objective function is established as follows:

max_{{\hat{λ}}_{i, t}^{s e l l e r} \in [\underline{λ}, \bar{λ}]} π_{i, t} = {\hat{λ}}_{i, t}^{s e l l e r} (1 - \frac{{\hat{p}}_{i, j, t}^{s e l l e r} - p_{i, j, t}^{s e l l e r}}{{\hat{p}}_{i, j, t}^{s e l l e r}}) ρ ({\hat{λ}}_{i, t}^{s e l l e r})

(25)

max_{{\hat{λ}}_{j, t}^{b u y e r} \in [\underline{λ}, \bar{λ}]} π_{j, t} = {\hat{λ}}_{j, t}^{b u y e r} (1 - \frac{{\hat{p}}_{i, j, t}^{b u y e r} - p_{i, j, t}^{b u y e r}}{{\hat{p}}_{i, j, t}^{b u y e r}}) ρ ({\hat{λ}}_{j, t}^{b u y e r})

(26)

\{\begin{matrix} ρ ({\hat{λ}}_{i, t}^{s e l l e r}) = \frac{1}{N} \sum_{n = 1}^{N} \frac{p_{i, j, t, n}^{s e l l e r}}{{\hat{p}}_{i, j, t, n}^{s e l l e r}} \\ ρ ({\hat{λ}}_{j, t}^{b u y e r}) = \frac{1}{N} \sum_{n = 1}^{N} \frac{p_{i, j, t, n}^{b u y e r}}{{\hat{p}}_{i, j, t, n}^{b u y e r}} \end{matrix}

(27)

where N represents the total number of trading rounds;

{\hat{λ}}_{i, t}^{s e l l e r}

and

{\hat{λ}}_{j, t}^{b u y e r}

denote the optimized selling and buying prices, respectively;

\underline{λ}

and

\bar{λ}

represent the lower and upper bounds of price variations;

{\hat{p}}_{i, j, t}^{s e l l e r}

and

{\hat{p}}_{i, j, t}^{b u y e r}

indicate the expected trading power of selling and buying regions;

p_{i, j, t}^{s e l l e r}

and

p_{i, j, t}^{b u y e r}

represent the actual successfully traded power of selling and buying regions; and

ρ ({\hat{λ}}_{i, t}^{s e l l e r})

and

ρ ({\hat{λ}}_{j, t}^{b u y e r})

denote the probability of successful transactions for selling and buying regions.

As demonstrated by Equations (25) and (26), regional managers must balance two key priorities: maximizing revenue through strategic bidding, while ensuring successful transaction completion. Before initiating new renewable energy trading requests with other regions, the system adjusts the price parameters to maximize the potential transaction volume. Specifically, sellers reduce their asking prices, while buyers increase their bid prices. The price adjustment can be expressed through the following equation:

\{\begin{matrix} {\hat{λ}}_{i, t, τ + 1}^{s e l l e r} = {\hat{λ}}_{i, t, τ}^{s e l l e r} + α_{τ} Δ λ \\ {\hat{λ}}_{i, t, τ + 1}^{b u y e r} = {\hat{λ}}_{i, t, τ}^{b u y e r} + α_{τ} Δ λ \end{matrix}

(28)

α_{τ} = \{\begin{cases} - 1 & p_{i, j, t}^{s e l l e r} > {\hat{p}}_{i, j, t}^{s e l l e r} \\ 0 & p_{i, j, t}^{s e l l e r} = {\hat{p}}_{i, j, t}^{s e l l e r} or p_{i, j, t}^{b u y e r} = {\hat{p}}_{i, j, t}^{b u y e r} \\ - 1 & p_{i, j, t}^{b u y e r} > {\hat{p}}_{i, j, t}^{b u y e r} \end{cases}

(29)

where

{\hat{λ}}_{i, t, τ}^{s e l l e r}

and

{\hat{λ}}_{i, t, τ}^{b u y e r}

represent the trading prices at iteration

τ

;

α_{τ} \in {- 1, 0, 1}

indicates the direction of price change, as detailed in Equation (29); and

Δ λ

defines the step size of the price adjustment.

3.2. Implementation Framework of Peer-to-Peer Energy Trading

This section presents a bilateral auction-based P2P energy trading framework for inter-regional energy exchange. The trading framework operates on a 24 h cycle, with uniform hourly intervals (

Δ t = 1 h

Initially, regional managers conduct internal supply–demand equilibrium analyses, incorporating renewable energy generation and load consumption patterns. Based on these assessments, regions declare their market positions as either buyers or sellers. Each region i formulates distinct strategy sets for selling (

s_{i, j}

) and buying (

b_{j, i}

) operations.

s_{i} = {t, p_{i, t}^{s e l l e r}, λ_{i, t}^{s e l l e r}}, λ_{i, t}^{s e l l e r} = {\hat{λ}}_{i, t}^{s e l l e r}

(30)

b_{j} = {t, p_{j, t}^{b u y e r}, λ_{j, t}^{b u y e r}}, λ_{j, t}^{b u y e r} = {\hat{λ}}_{j, t}^{b u y e r}

(31)

Given the heterogeneous geographic distribution of resources across regions, power capacity offerings and bidding scales exhibit natural variations. To safeguard strategic information and prevent the exploitation of regional supply–demand characteristics, we introduce a novel multi-segment bidding mechanism. This approach enables regional energy trading agents to distribute trading information stochastically among multiple counterparts, enhancing privacy protection.

s_{i, j} = {t, p_{i, j, t}^{s e l l e r}, λ_{i, t}^{s e l l e r}}

(32)

b_{j, i} = {t, p_{i, j, t}^{b u y e r}, λ_{j, t}^{b u y e r}}

(33)

\{\begin{matrix} \sum_{i \neq j} p_{i, j, t}^{s e l l e r} = p_{i, t}^{s e l l e r} \\ \sum_{i \neq j} p_{i, j, t}^{b u y e r} = p_{j, t}^{b u y e r} \end{matrix}

(34)

Equation (34) demonstrates that the total power that is bought and sold must be balanced across all regions.

0 \leq p_{i, j, t}^{s e l l e r} \leq p_{i, j, m a x}^{s e l l e r}

(35)

0 \leq p_{i, j, t}^{b u y e r} \leq p_{i, j, m a x}^{b u y e r}

(36)

Regional trading limits are defined by

p_{i, j, m a x}^{s e l l e r}

and

p_{i, j, m a x}^{b u y e r}

, representing the maximum selling capacity from region i to j and maximum buying capacity from region i to j, respectively.

During each time segment, regional agents process incoming trading announcements dynamically. The mechanism implements two optimization strategies:

Selling regions prioritize high-price buyers through an optimization function that maximizes their revenue potential.

$max \sum_{i \neq j} \sum_{b_{j, i}} (λ_{j, t}^{b u y e r} - λ_{i, t}^{s e l l e r}) p_{i, j, t}^{s e l l e r}$

(37)

$\{\begin{matrix} 0 \leq p_{i, j, t}^{s e l l e r} \leq p_{i, t}^{s e l l e r} \\ (λ_{j, t}^{b u y e r} - λ_{i, t}^{s e l l e r}) p_{i, j, t}^{s e l l e r} \geq 0 \end{matrix}$

(38)
Buying regions preferentially engage with low-price sellers through an optimization function that minimizes their procurement costs.

$max \sum_{i \neq j} \sum_{b_{j, i}} (λ_{j, t}^{b u y e r} - λ_{i, t}^{s e l l e r}) p_{i, j, t}^{b u y e r}$

(39)

$\{\begin{matrix} 0 \leq p_{i, j, t}^{b u y e r} \leq p_{i, t}^{b u y e r} \\ (λ_{j, t}^{b u y e r} - λ_{i, t}^{s e l l e r}) p_{i, j, t}^{b u y e r} \geq 0 \end{matrix}$

(40)

The bilateral trading process concludes either with (1) a successful match, resulting in the publication of negotiation outcomes (

S_{i, j}

for sellers,

B_{i, j}

for buyers) or the (2) termination of the auction round if no viable bilateral solution exists. This framework ensures efficient market clearing while maintaining competitive price discovery and information privacy.

S_{i, j} = {t, p_{i, j, t}^{s - t r a d i n g}, λ_{i, j, t}^{s - t r a d i n g}}

(41)

B_{i, j} = {t, p_{i, j, t}^{b - t r a d i n g}, λ_{i, j, t}^{b - t r a d i n g}}

(42)

\{\begin{matrix} p_{i, j, t}^{s - t r a d i n g} = p_{i, j, t}^{b - t r a d i n g} = min {p_{i, j, t}^{s e l l e r}, p_{i, j, t}^{b u y e r}} \\ λ_{i, j, t}^{s - t r a d i n g} = λ_{i, j, t}^{b - t r a d i n g} = (λ_{j, t}^{b u y e r} + λ_{j, t}^{s e l l e r}) / 2 \end{matrix}

(43)

where

p_{i, j, t}^{s - t r a d i n g}

and

p_{i, j, t}^{b - t r a d i n g}

represent the final agreed trading power for selling and buying regions, respectively;

λ_{i, j, t}^{s - t r a d i n g}

, and

λ_{i, j, t}^{b - t r a d i n g}

denote the final settled trading prices for selling and buying regions, respectively. As shown in Equation (43), the final trading price is determined through an average price clearing mechanism, where the traded power volume is limited by the smaller of the two quantities proposed by the buyer and seller.

4. Results

4.1. Simulation Parameters and Conditions

This study utilizes hourly wind power, solar generation, and load data (8760 h) from four regions in Northeast China. To enhance the computational efficiency, we employed k-means clustering to condense these data into four representative days for simulation purposes. While our current analysis focuses on four regions, the proposed P2P trading model is designed to be scalable, accommodating future expansion as more urban areas adopt P2P energy trading. The complete solution methodology is illustrated in Figure 2. All simulations were performed using MATLAB R2023b with YALMIP toolbox and CPLEX 12.10 solver.

4.2. Analysis of Energy Trading Results Across Regions

Figure 3 shows the preliminary balancing results for each region with and without P2P. It shows that implementing P2P energy trading between the different regions leads to smoothing of the overall load profile. Regions 1 and 4 have surplus renewable energy generation, which can be effectively utilized through trading with Regions 2 and 3. This allows the use of renewable energy in those regions to be maximized. Region 2, on the other hand, does not have enough renewable generation to meet its own demand. The P2P trading helps satisfy the majority of this demand from the surplus in other regions. Region 3 acts as a seller and a buyer at different times, buying energy during periods of high demand (16–24 h) and selling surplus during other times.

However, the inter-regional trading plan is still unable to perfectly balance supply and demand in each region at all times. For example, there is still an imbalance during hour 18. To address this, the regions would need to utilize flexible generation resources to achieve final supply–demand equilibrium.

The different regions have varying internal resource profiles due to their distinct geographical environments. Figure 4 provides the detailed results of the supply–demand optimization within each region. The symmetry between the upper and lower portions of the graph illustrates the balance between generation and consumption. Viewing the four regions as an integrated system, the overall supply exceeds demand. This suggests that, without the addition of new energy storage facilities, the required investment in renewable energy generation infrastructure could potentially be reduced.

Region 2, after its initial attempt to balance renewable generation with local demand, still experiences a supply shortfall even after the P2P trading plan. To meet this remaining demand, Region 2 utilizes pumped storage hydropower and coal-fired generation as supplementary resources. In contrast, Regions 3 and 4 are largely able to achieve supply–demand equilibrium using flexible, lower-emission resources. However, some degree of renewable energy curtailment persists across the four-region system as a whole.

Nonetheless, the comparison of Figure 3 and Figure 4 indicates that the P2P trading scheme reduces both renewable energy curtailment and the need for flexible generation, providing economic benefits. This underscores the value of coordinated, cross-regional energy planning and optimization.

4.3. Analysis of Bidding Strategies Among Regions

The energy trading situation is described in Figure 5, which shows the buy and sell dynamics across four time periods—4 h, 10 h, 16 h, and 22 h. This includes the trading prices, as well as the transacted power amounts. The trading prices are influenced not only by historical prices but also by the broader energy trading conditions across multiple regions, which are optimized.

In this example, the entire day is divided into 24 time slots. During each slot, the different regions determine whether they will be a seller or a buyer based on their preliminary supply and demand conditions from internal renewable energy sources and load. The sellers and buyers can vary across time slots—for instance, Region 3 acts as a seller during the 4 h and 10 h slots but becomes a buyer during the 16 h and 22 h slots. This flexible approach, where the sellers and buyers are dynamically generated based on each region’s supply–demand situation rather than being fixed, helps protect the privacy of the internal dispatch planning within the different regions.

Figure 6 illustrates the evolution of the trading success rates across different time periods (4 h, 10 h, 16 h, and 22 h). The results demonstrate that the proposed model effectively avoids local optima and converges to a global optimum. This convergence reflects the model’s ability to simultaneously maximize the power trading volume while minimizing the transaction costs, achieving an optimal balance between these competing objectives.

4.4. Energy Trading Results on Different Typical Days

Figure 7, Figure 8, Figure 9 illustrate the energy supply–demand patterns on representative days. The analysis demonstrates that across all scenarios, the P2P energy exchange plan consistently enhances renewable energy integration and reduces generation costs from flexible resources, validating the model’s robustness.

5. Conclusions

This paper presents a P2P trading mechanism for multiple regions in local energy markets. Case studies demonstrate three key findings: First, the proposed trading mechanism enables regions to sell surplus renewable energy for a profit while allowing energy-deficient regions to purchase renewable energy from others, thereby reducing generation costs from flexible resources. Second, the model eliminates centralized third-party structures, addressing inherent challenges in centralized management systems. Regions can independently choose to act as buyers or sellers based on their real-time supply–demand conditions. Trading partners and power volumes can vary across different time periods, ensuring operational flexibility and transaction privacy for regional managers. Third, regions initially set trading prices based on historical data, then engage in bilateral negotiations to optimize the pricing. This iterative process continues until both parties maximize their benefits in terms of successful transaction rates and trading costs.

In conclusion, the proposed model effectively meets the privacy requirements of regions with diverse environmental factors in the energy trading market. It demonstrates significant potential for promoting broader adoption of renewable energy across society while reducing generation costs from flexible resources. The framework provides a robust technical foundation for future urban low-carbon transitions, offering a dynamic and autonomous approach to inter-regional energy exchange.

Author Contributions

Conceptualization, T.Y. and S.L.; methodology, T.Y.; software, J.L.; validation, T.Y., W.F., Z.C. and Y.Z.; formal analysis, Z.C.; investigation, Y.Z.; resources, S.L.; data curation, T.Y.; writing—original draft preparation, T.Y.; writing—review and editing, S.L.; visualization, J.L.; supervision, S.L.; project administration, T.Y.; funding acquisition, T.Y. All authors have read and agreed to the published version of the manuscript.

Funding

The work was funded by the science and technology project, guided by the northeast branch of the state grid corporation of China. Project Number: (529926240007, SGDB0000DJJS2400100).

Data Availability Statement

The data presented in this study are available from the authors upon request.

Conflicts of Interest

Authors Tianmeng Yang, Jicheng Liu, Wei Feng, Zelong Chen, Yumin Zhao were employed by the company Northeast Branch of State Grid Corporation of China. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

References

Allan, G.; McGregor, P.; Swales, K. Greening regional development: Employment in low-carbon and renewable energy activities. Reg. Stud. 2017, 51, 1270–1280. [Google Scholar] [CrossRef]
Siddik, A.B.; Khan, S.; Khan, U.; Yong, L.; Murshed, M. The role of renewable energy finance in achieving low-carbon growth: Contextual evidence from leading renewable energy-investing countries. Energy 2023, 270, 126864. [Google Scholar] [CrossRef]
Huang, J.; Yuan, M.; Wang, L.; Zou, Z.; Sun, Y. Enhancing low-carbon building operations leveraging demand response driven by renewable energy consumption contributions. J. Build. 2024, 95, 110206. [Google Scholar] [CrossRef]
Liang, C.; Umar, M.; Ma, F.; Huynh, T.L. Climate policy uncertainty and world renewable energy index volatility forecasting. Technol. Forecast Soc. Chang. 2022, 182, 121810. [Google Scholar] [CrossRef]
Pata, S.K. Comparative impacts of energy, climate, and economic policy uncertainties on renewable energy. J. Environ. Manag. 2024, 370, 122494. [Google Scholar] [CrossRef]
Alonso-Travesset, À.; Coppitters, D.; Martín, H.; de la Hoz, J. Economic and regulatory uncertainty in renewable energy system design: A review. Energies 2023, 16, 882. [Google Scholar] [CrossRef]
Villamor, L.V.; Avagyan, V.; Chalmers, H. Opportunities for reducing curtailment of wind energy in the future electricity systems: Insights from modelling analysis of Great Britain. Energy 2020, 195, 116777. [Google Scholar] [CrossRef]
Diab, I.; Damianakis, N.; Chandra-Mouli, G.R.; Bauer, P. A shared PV system for transportation and residential loads to reduce curtailment and the need for storage systems. Appl. Energy 2024, 353, 122131. [Google Scholar] [CrossRef]
Yang, H.; Wang, J.; Ma, Y.; Zhang, D.; Guerrero, J.M. Battery energy storage system planning for promoting renewable energy consumption using a constraint boundarization technique. IET Renew. Power Gener. 2023, 18, 2867–2879. [Google Scholar] [CrossRef]
Kumar, J.; Parthasarathy, C.; Västi, M.; Laaksonen, H.; Shafie-Khah, M.; Kauhaniemi, K. Sizing and allocation of battery energy storage systems in Åland islands for large-scale integration of renewables and electric ferry charging stations. Energies 2020, 13, 317. [Google Scholar] [CrossRef]
Huang, Z.; Liu, L.; Liu, J. Multi-Time-Scale Coordinated Optimum Scheduling Technique for a Multi-Source Complementary Power-Generating System with Uncertainty in the Source-Load. Energies 2023, 16, 3020. [Google Scholar] [CrossRef]
Lou, N.; Zhang, Y.; Wang, Y.; Liu, Q.; Li, H.; Sun, Y.; Guo, Z. Two-stage congestion management considering virtual power plant with cascade hydro-photovoltaic-pumped storage hybrid generation. IEEE Access 2020, 8, 186335–186347. [Google Scholar] [CrossRef]
Yang, Y.; Bremner, S.; Menictas, C.; Kay, M. Modelling and optimal energy management for battery energy storage systems in renewable energy systems: A review. Renew. Sust. Energ. Rev. 2022, 167, 112671. [Google Scholar] [CrossRef]
Islam, M.M.; Nagrial, M.; Rizk, J.; Hellany, A. Dual stage microgrid energy resource optimization strategy considering renewable and battery storage systems. Int. J. Energy Res. 2021, 45, 21340–21364. [Google Scholar] [CrossRef]
Wang, G.; You, D.; Lou, S.; Zhang, Z.; Dai, L. Economic valuation of low-load operation with auxiliary firing of coal-fired units. Energies 2017, 10, 1317. [Google Scholar] [CrossRef]
Lei, C.; Wang, Q.; Zhou, G.; Bu, S.; Zhou, N.; Lin, T.; Wei, F. Probabilistic wind power expansion planning of bundled wind-thermal generation system with retrofitted coal-fired plants using load transfer optimization. Int. J. Electr. Power Energy Sys. 2023, 151, 109145. [Google Scholar] [CrossRef]
Sousa, T.; Soares, T.; Pinson, P.; Moret, F.; Baroche, T.; Sorin, E. Peer-to-peer and community-based markets: A comprehensive review. Renew. Sust. Energ. Rev. 2019, 104, 367–378. [Google Scholar] [CrossRef]
Tushar, W.; Saha, T.K.; Yuen, C.; Smith, D.; Poor, H.V. Peer-to-peer trading in electricity networks: An overview. IEEE Trans. Smart Grid 2020, 11, 3185–3200. [Google Scholar] [CrossRef]
Liu, N.; Yu, X.; Wang, C.; Li, C.; Ma, L.; Lei, J. Energy-sharing model with price-based demand response for microgrids of peer-to-peer prosumers. IEEE Trans. Power Syst. 2017, 32, 3569–3583. [Google Scholar] [CrossRef]
Alam, M.R.; St-Hilaire, M.; Kunz, T. An optimal P2P energy trading model for smart homes in the smart grid. Energy Effic. 2017, 10, 1475–1493. [Google Scholar] [CrossRef]
Zhang, Z.; Li, R.; Li, F. A novel peer-to-peer local electricity market for joint trading of energy and uncertainty. IEEE Trans. Smart Grid 2019, 11, 1205–1215. [Google Scholar] [CrossRef]
Wu, Y.; Wu, Y.; Cimen, H.; Vasquez, J.C.; Guerrero, J.M. P2P energy trading: Blockchain-enabled P2P energy society with multi-scale flexibility services. Energy Rep. 2022, 8, 3614–3628. [Google Scholar] [CrossRef]
Moghadam, A.Z.; Javidi, M.H. Designing a two-stage transactive energy system for future distribution networks in the presence of prosumers’ P2P transactions. Electr. Power Syst. Res. 2022, 211, 108202. [Google Scholar] [CrossRef]
Meng, Y.; Qiu, J.; Zhang, C.; Lei, G.; Zhu, J. A Holistic P2P market for active and reactive energy trading in VPPs considering both financial benefits and network constraints. Appl. Energy 2024, 356, 122396. [Google Scholar] [CrossRef]
Zou, Y.; Xu, Y.; Li, J. Aggregator-Network Coordinated peer-to-peer Multi-Energy Trading via adaptive robust Stochastic Optimization. IEEE Trans. Power Syst. 2024, 39, 7124–7137. [Google Scholar] [CrossRef]
Wang, Z.; Hou, H.; Zhao, B.; Zhang, L.; Shi, Y.; Xie, C. Risk-averse stochastic capacity planning and P2P trading collaborative optimization for multi-energy microgrids considering carbon emission limitations: An asymmetric Nash bargaining approach. Appl. Energy 2024, 357, 122505. [Google Scholar] [CrossRef]
Kuno, S.; Tanaka, K.; Yamada, Y. Effectiveness and feasibility of market makers for P2P electricity trading. Energies 2022, 15, 4218. [Google Scholar] [CrossRef]
Woo, C.K.; Sreedharan, P.; Hargreaves, J.; Kahrl, F.; Wang, J.; Horowitz, I. A review of electricity product differentiation. Appl. Energy 2014, 114, 262–272. [Google Scholar] [CrossRef]

Figure 1. Multi-regional smart grid framework for energy interaction and coordinated scheduling.

Figure 2. The model solution flowchart used in this paper.

Figure 3. Preliminary balance analysis of renewable energy and load demand for all regions.

Figure 4. Power balance analysis for all regions.

Figure 5. Regional pricing and power trading situations across different time periods.

Figure 6. Success rate assessment of iterative trading process between buyers and sellers across different time periods.

Figure 7. Typical day 2 energy trading results.

Figure 8. Typical day 3 energy trading results.

Figure 9. Typical day 4 energy trading results.

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

© 2024 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).

Share and Cite

MDPI and ACS Style

Yang, T.; Liu, J.; Feng, W.; Chen, Z.; Zhao, Y.; Lou, S. A Three-Layer Scheduling Framework with Dynamic Peer-to-Peer Energy Trading for Multi-Regional Power Balance. Energies 2024, 17, 6239. https://doi.org/10.3390/en17246239

AMA Style

Yang T, Liu J, Feng W, Chen Z, Zhao Y, Lou S. A Three-Layer Scheduling Framework with Dynamic Peer-to-Peer Energy Trading for Multi-Regional Power Balance. Energies. 2024; 17(24):6239. https://doi.org/10.3390/en17246239

Chicago/Turabian Style

Yang, Tianmeng, Jicheng Liu, Wei Feng, Zelong Chen, Yumin Zhao, and Suhua Lou. 2024. "A Three-Layer Scheduling Framework with Dynamic Peer-to-Peer Energy Trading for Multi-Regional Power Balance" Energies 17, no. 24: 6239. https://doi.org/10.3390/en17246239

APA Style

Yang, T., Liu, J., Feng, W., Chen, Z., Zhao, Y., & Lou, S. (2024). A Three-Layer Scheduling Framework with Dynamic Peer-to-Peer Energy Trading for Multi-Regional Power Balance. Energies, 17(24), 6239. https://doi.org/10.3390/en17246239

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Menu

A Three-Layer Scheduling Framework with Dynamic Peer-to-Peer Energy Trading for Multi-Regional Power Balance

Abstract

1. Introduction

2. Regional Models Incorporating Differentiated Characteristics

2.1. Objective

2.1.1. Annual Investment Cost Analysis of Distributed Energy Storage Systems

2.1.2. Cost of Wind and Solar Power Curtailment

2.1.3. Generation Cost of Coal-Fired Power Units

2.1.4. Operational Cost of Pumped Storage and Electrochemical Energy Storage Plants

2.2. Constraints

2.2.1. Abandonment Constraints for Wind and PV Units

2.2.2. Operational Constraints of Electrochemical Energy Storage

2.2.3. Operational Constraints of Pumped Storage Hydroelectric Systems

2.2.4. Output Constraints of Coal-Fired Power Plants

2.2.5. Output Constraints of Hydropower Plants

3. Inter-Regional Peer-to-Peer Energy Trading Mechanism

3.1. Bidding Strategy Optimization in Energy Markets

3.2. Implementation Framework of Peer-to-Peer Energy Trading

4. Results

4.1. Simulation Parameters and Conditions

4.2. Analysis of Energy Trading Results Across Regions

4.3. Analysis of Bidding Strategies Among Regions

4.4. Energy Trading Results on Different Typical Days

5. Conclusions

Author Contributions

Funding

Data Availability Statement

Conflicts of Interest

References

Share and Cite

Article Metrics

Article Access Statistics

Further Information

Guidelines

MDPI Initiatives

Follow MDPI