Blockchain-Based Renewable Energy Certificate Trade for Low-Carbon Community of Active Energy Agents

Abstract

The future distribution grid is a peer-to-peer (P2P) community formed by a large number of active energy agents (AEAs), and renewable energy certificate (REC) trading is an efficient way to realize a low-carbon AEA community. AEAs can trade not only electricity but also RECs among themselves to economically and efficiently meet the renewable portfolio standard (RPS) requirements. Aiming to lower the market barrier and increase the trading benefits for market participants, this paper proposes a blockchain-based renewable energy certificate (BCREC) that supports divisible and multiple transactions. The trade process includes four stages: setup, pre-transaction, transaction, and post-transaction. A scheme based on blockchain oracles and smart contracts is implemented to achieve decentralized BCREC issuance and transaction and to support a more flexible trading market. By exploring two typical market scenarios, we verify the advantages of BCREC trading and evaluate its impacts on AEA profits and market efficiency.

1. Introduction

The future electric power distribution network in the smart grid is a peer-to-peer (P2P) community of active energy agents (AEAs) [1]. In the energy domain, each AEA is an electrical entity that can either produce or consume electricity, such as rooftop photovoltaic technology, household smart appliances, private electric vehicles, compact energy storage systems, etc.; in the information domain, each AEA is an intelligent and autonomous agent which can actively perceive the market situation and participate in electricity trade, thus maximizing their profits. It is efficient and beneficial for AEAs to locally trade energy in P2P mode; thus, the majority of the demand and supply of energy can be easily balanced among neighbors in the community rather than exchanged with the main power grid.

In the context of a low-carbon economy and sustainable development, each AEA is also responsible for greenhouse gas emission reduction, and the carbon footprint is regulated by the renewable portfolio standard (RPS) [2], which mandates a minimum share of renewable energy in electricity consumption [3]. The government enforces the RPS by penalizing electricity consumers who fall short of the target and may reward those who surpass it. The contribution of AEAs to low-carbon energy systems is generally quantified in terms of renewable energy certificates (RECs), which are issued due to the production of a certain amount (e.g., 1 MWh) of renewable energy to present the corresponding environmental value. Each AEA is mandatorily required to contribute to a low-carbon energy system by providing enough RECs. AEAs can provide RECs in two ways, either direct or indirect: (i) directly contribute to renewable energy production/consumption, i.e., generating electricity from renewable energy sources or consuming electricity with REC; (ii) pay others for renewable energy production/consumption, i.e., purchasing in REC market. Due to the structure and interaction mode of the community, AEAs can trade not only electricity but also RECs in P2P mode. AEAs can sell the surplus RECs to obtain additional profit or buy the required RECs to fulfill the RPS regulation. However, the existing RECs are intrinsically unsuitable for the P2P transaction in AEA communities:

(1) Centrality: The existing RECs are issued, transacted, and settled by a trusted center. Although this centrality ensures the authenticity and credibility of RECs, it may cause vulnerability to misconduct, the heavy computation of operations, and expensive maintenance in large-scale communities that involve a large number of REC transactions in P2P mode.

(2) Indivisibility: The current RECs are indivisible units that represent 1 MW·h of renewable energy. This indivisibility ensures the uniqueness and traceability of RECs, but it also excludes many small AEAs from trading RECs. Moreover, it is difficult for trusted centers to handle a large number of small RECs due to the high processing pressure of the centralized mechanism of issuance and management.

(3) Single-transactability: The existing RECs are issued due to renewable energy production and are disposed of after RPS verification. They can only be traded once in their lifetime. This disposability ensures that the environmental value of renewable energy is not counted twice, and one-time trades reduce the burden placed on trusted centers but may also reduce the liquidity and flexibility of the REC market.

As a result, it is necessary to develop a new REC and the relevant mechanism, which are intrinsically decentralized, so as to support the whole process of the novel trade mode of REC, including the registration, generation (issuance), transaction, verification, and destruction (disposal).

Characterized by decentralization, tamper-proofing, and traceability, the blockchain enables the specific REC trade in AEA communities [4]. Thus, this paper proposes a blockchain-based REC (BCREC) as the qualified candidate for AEA communities by using the key technologies of blockchain, such as cryptography, consensus mechanisms, smart contracts, etc. As tokenized digital assets based on the ERC-20 token standard [5], BCRECs are automatically issued and disposed of by corresponding smart contracts. Since the issuance of each BCREC is strictly related to the amount of renewable energy produced, the environmental value of renewable energy cannot be double-counted, and the flexible division and repeatable transaction of BCREC are both realized. The BCREC can be discriminated from the traditional counterpart by the following three features:

(1) Decentrality. Blockchain enables us to develop a completely decentralized BCREC, which can avoid the risk of third-party misconduct, achieve data transparency and traceability, and eliminate the cost of maintaining a trusted center.

(2) Divisibility. Blockchain helps us to develop BCRECs that can be flexibly divided, which can greatly lower the market barrier. As shown in Figure 1, a BCREC is a divisible and fungible digital token that is issued due to any amount of renewable energy and can track changes in account balances. On the contrary, since the RECs are largely denominated and indivisible, n units of RECs are issued for n MW·h renewable energy, and any amount less than 1 MW·h cannot be credited. This feature enables small AEAs to engage in the BCREC trade and avoids the trouble of issuing a large number of certificates.

(3) Multi-transactability. Blockchain allows us to develop BCRECs that can be traded repeatedly, as shown in Figure 2, which can greatly enhance the liquidity of the market. This makes it more attractive for AEAs to engage in the P2P transactions of BCRECs. The on-chain records are traceable and tamper-proof, which can ensure orderly transactions and avoid risks such as double-spending issues.

Blockchain is a promising technology for power systems [6], especially for the development of smart grids [7]. Most of the existing literature on blockchain technology for renewable energy concentrates on energy-related aspects, such as P2P energy trading [8,9,10] and resource optimization [11,12,13] among distributed energy producers, consumers, and prosumers. As for the application of blockchain in the REC trade, in [14], the authors analyzed the shortcomings of the current centralized REC issuance and tracking system and proposed a blockchain-based improvement for tokenizing RECs into non-fungible tokens. According to [15], the cross-chain transactions of RECs in the park’s integrated energy system can lower operating costs, increase renewable energy consumption, and enhance transaction efficiency and transparency. A solution for the REC trading platform based on the Hyperledger Fabric 1.1 blockchain was proposed in [16], and over 100 simulated transactions have been completed in a range that included over 40 members. The authors of [17] suggested a blockchain-based system for individual RECs to encourage the voluntary use of distributed renewable energy. In [18], several buildings participated in a small-scale experiment on REC applications, using blockchain for data recording. The buildings could exchange their electricity generation for RECs, which could cover their electricity consumption above their generation. This incentivized them to use RECs and renewable energy. In [19], RECs are tokenized into cryptocurrency by using smart contracts on the Ethereum blockchain. This enables consumers to purchase RECs directly from the producers and was simulated between 20 households using fixed and random prices. The existing studies have tokenized RECs into digital tokens on the blockchain, making REC transactions more convenient and transparent. It is also verified that REC can promote renewable energy generation. However, the limitations of traditional REC, such as indivisibility, persist even after tokenization. RECs on the blockchain should not be mere replicas of traditional certificates but should leverage the features of digital assets to make them more flexible and user-friendly.

The P2P trading of electric energy in a smart grid is thoroughly investigated [20,21], and the trend of personalization and decentralization is revealed [22]. Considering the structure of the AEA community, it is reasonable to introduce the P2P mode to REC trading in order to improve efficiency and flexibility. In [1], we proposed a hybrid REC trading system including two phases: P2P transaction and centralized settlement. However, there still exist some problems which are not well addressed. First, RECs can only be traded once and must be cleared in each period; therefore, imbalanced supply and demand can only be resolved through the centralized settlement phase, which reduces the revenues. Second, the trading mechanism is designed and verified; however, the design and specific implementation of the corresponding REC are not given.

This paper makes the following contributions:

(1) The REC trade in AEA communities is investigated and blockchain-based renewable energy certificates (BCRECs) are proposed, which are fully decentralized, divisible, and compatible with repeatable transactions;

(2) The whole process of the BCREC trade is designed, and the off-chain P2P transaction is formulated. The on-chain smart contracts are implemented, including data preparation, user registration, BCREC issuance, transaction ledgers, and settlement;

(3) By investigating two typical market scenarios, the effectiveness of the BCREC trade is verified and the performances are evaluated on AEA benefits and market trading efficiency.

The rest of this article is structured as follows: Section 2 designs the BCREC and its trade process. Section 3 implements all functions of the BCREC trade, on-chain and off-chain. In Section 4, a case study is conducted to analyze the increased trading revenue from the BCREC design and demonstrate the effect of the blockchain’s implementation. Finally, we draw the concluding remarks in Section 5.

2. Design of BCREC Trade

2.1. Preliminary

(1) BCRECs are divisible and 1 unit of BCREC corresponds to 1 kW·h renewable energy. It corresponds strictly to renewable energy generation, i.e., an electricity supplier can obtain n units of BCRECs by producing n kW·h renewable energy, where n is a non-negative real number.

(2) The BCREC market is decoupled from the electricity market, i.e., it is unnecessary for a BCREC buyer/seller to be an electricity buyer/seller.

(3) The BCREC market is a monthly market, i.e., AEAs in the community trade BCRECs once per month. The role (buyer or seller) of AEAs in the BCREC market can vary in different months but remains invariant within a single month.

(4) The BCREC market is temporally intercoupled across multiple months, i.e., an AEA can hold rather than clear the surplus BCRECs by the end of a month. Both newly issued and previously accumulated BCRECs are traded in a month. BCRECs can be traded multiple times and are disposed of after they are used to fulfill the RPS requirement.

(5) AEAs are rational market players that are motivated to participate in the P2P transaction of BCRECs due to the economic benefits, i.e., the AEAs that have surplus BCRECs can sell BCRECs at higher prices, and the AEAs that need extra BCRECs to fulfill the RPS regulation can buy BCRECs at lower prices compared to those in the centralized transaction mode.

2.2. Suitability Analysis

Due to the trust mechanism that does not rely on centralized institutions or individuals, the blockchain allows untrusted nodes in a community to collaborate, which is suitable for the P2P transaction of BCRECs. The key to the blockchain is the “code-based trust model”, which is established through various smart contracts. Each smart contract is an automated script whose code is recorded in a block and runs on the blockchain network. Triggered by predefined conditions, the smart contract is verified by all nodes in the blockchain network, and it will self-execute the actions required in an agreement or contract after consensus is reached within the community. In this paper, we design the BCREC and develop smart contracts to implement the functions of BCREC issuance, transaction ledgers, and settlements, as well as BCREC disposal on the blockchain.

2.3. Trade Process

As illustrated in Figure 3, the whole process of the BCREC trade consists of four stages: setup, pre-transaction, transaction, and post-transaction. In the setup stage, all AEAs register or log in to the blockchain network; the governmental administration verifies the identities and determines the number of BCRECs required by the RPS. In the pre-transaction stage, the blockchain oracle accesses external data sources. The BCRECs are issued to corresponding AEAs according to the data on renewable energy production. In the transaction stage, AEAs buy or sell BCRECs mainly in P2P mode and may also deal with the BCREC broker as a complement. In the post-transaction stage, AEAs confirm, record, and conduct the transactions; a governmental administration verifies each AEA and the completion of the RPS and disposes of the corresponding BCRECs. The aforementioned process is repeated each month.

3. Implementation of BCREC Trade

3.1. Platform Selection

3.1.1. Blockchain Network

A blockchain can be classified as either public or permissioned. The former has no restriction on the nodes entering and exiting the network, and the latter has access restrictions: only authorized users are allowed to run nodes and participate in the consensus mechanism. It can be a dilemma for developers to make a choice among various blockchain platforms, which are compared in

Table 1. Comparison of public and permissioned blockchains.

Type	Example	Advantages	Disadvantages
Public	Bitcoin, Ethereum	generality, complete ecosystem, existing infrastructure, mature platform, rapid development and deployment	high transaction fee unstable transaction latency (especially in large-scale applications)
Permissioned	Hyperledger Fabric, FISCO BCOS	low transaction fee high transaction performance high data privacy	dedicated infrastructure high upfront cost to build, run, and maintain

. Considering the requirement of the application and the characteristics of typical blockchains, we chose Ethereum to implement BCREC trading in AEA communities. Ethereum has completed its transition from proof-of-work to proof-of-stake through a merger. It eliminated the need for energy-intensive mining and instead enabled the network to be secured using staked ETH [23]. As a result, Ethereum’s energy consumption is greatly reduced [24], making it more sustainable.

3.1.2. Blockchain Oracle

To enable the blockchain to access and verify external data, we used the blockchain oracle, a reliable data source for on-chain smart contracts. Unlike most of the existing work that ignores this problem or relies on a centralized agent, we adopt Chainlink (https://chain.link/ accessed on 15 October 2023), a decentralized blockchain oracle that securely connects on-chain smart contracts to off-chain data and services. Chainlink creates a network of independent oracle nodes that operate in a similar way to the blockchain. This provides a decentralized and verifiable solution for the blockchain to access trusted off-chain data [25].

3.1.3. Software Tool

The Solidity programming language is used for the implementation of on-chain smart contracts. The developing environment is the Remix Online IDE v0.37.4. The MetaMask wallet software 11.4.1 and the Etherscan blockchain browser (https://etherscan.io/ accessed on 15 October 2023) are used to manage user accounts and view on-chain information, respectively. MATLAB 2020a is used for the implementation of the off-chain P2P transaction among AEAs.

3.2. Setup Stage

3.2.1. BCREC Definition

BCRECs are tokenized as ERC-20-compliant fungible tokens since it is only necessary to present the difference in quantity rather than in the quality of the renewable energy. We also can issue multiple fungible tokens if it is required to further distinguish the types of renewable energy.

Table 2. List of BCREC contract functions.

Contract Name	Function Name	Caller
BCREC	lock (address)	Administrator
	unlock (address)	Administrator
	pause ()	Administrator
	unpause ()	Administrator
	burnfrom (address, uint256)	Administrator
	issue (address, uint256)	BCREC Issuance
	transfer (address, uint256)	AEA
	approve (address, uint256)	AEA

shows the main functions that are implemented in the BCERC contract and their calling privileges. The governmental administration is assigned the admin role with higher authority for convenient supervision and maintenance. The administrator can lock an AEA’s account or pause all BCREC transfers in case of an AEA default or a market emergency. The burnfrom function is used to dispose of the AEA’s BCRECs that are used for the RPS. To ensure the orderly and accurate issuance of BCRECs, only the BCREC issuance contract can call the issue function. AEAs can only query and transfer the BCRECs in their own account. ERC-20 is a widely used token standard that is compatible with most wallet software; therefore, AEAs can conveniently manage their BCRECs with their preferred wallet software.

3.2.2. Account Management

Ethereum accounts are essential for users to interact with the platform and execute smart contracts. To ensure security and authenticity, Ethereum accounts use an elliptic curve digital signature algorithm and do not include real-world identity information. Therefore, any AEA that uses Ethereum applications is required to register an account and submit a real-world ID, the energy production data APIs, and a certain deposit. The government administration validates the registration information and activates the accounts of qualified AEAs. Only the AEAs with activated accounts can participate in the BCREC trade by calling various on-chain smart contracts. The above operation is implemented by using the account management contract including three functions which are listed in

Table 3. List of account management contract functions.

Contract Name	Function Name	Caller
Account Management	register (uint256, string)	AEA
	activate (address)	Administrator
	deactivate (address)	Administrator

.

3.3. Pre-Transaction Stage

3.3.1. Data Preparation

We use Chainlink, the blockchain oracle, to securely access data and services outside the blockchain. Once the external smart meters measure the AEA’s renewable energy generation data, they only store the data locally in JSON format and provide an access interface to Chainlink. Chainlink is responsible for ensuring that the data are credibly recorded on the chain. This process is similar to making an HTTP GET request to the data stored on a server but is a verifiable and decentralized mode.

3.3.2. BCREC Issuance

The smart contract of BCREC issuance is used to issue/generate BCRECs on the blockchain. The BCREC is intrinsic and decentralized, and the issuance completely relies on smart contracts rather than any centralized authority. Upon receiving data, the BCREC issuance contract automatically executes without any manual intervention; the corresponding amount of BCRECs (1 kW·h corresponds to 1 unit of BCREC) are issued to the AEA accounts.

Figure 4 illustrates the operations involved in the BCREC issuance process, which mainly uses the basic request model of the Chainlink oracle [25]. The process consists of two Ethereum transactions: (1) the BCREC issuance contract initiates the service request, and (2) the oracle node returns the result of the request. The contract then automatically issues the corresponding amount of BCRECs to the AEA’s account according to the data from the oracle node.

3.4. Transaction Stage

As an off-chain process, the transaction stage consists of two phases: P2P and complementary. The former is the major phase, where most AEAs become satisfied by buying or selling BCRECs, and the latter is the optional phase, where any AEA can continue transactions with the BCREC broker. Notice that it is free for AEAs to hold or sell the surplus BCRECs after fulfilling the RPS requirement.

3.4.1. Order Initialization

The BCREC market is driven by the interaction of the demand and supply of BCRECs. For an AEA ${\mathcal{A}}_i$, the demand is the amount of BCRECs required by RPS, and the supply is the total amount of BCRECs generated in this month (excluding its previous holdings in BCREC). An undersupplied AEA can become a BCREC buyer by recording its net demand as $Q_{d,i}$, and an oversupplied AEA can become a BCREC seller by recording its net supply as $Q_{s,i}$. The market demand $Q_D$ and supply $Q_S$ are the sum of the individual net demand and net supply, respectively.

The transaction order of buyer ${\mathcal{A}}_i$ is denoted as $\left[Q_{buy,i}^r,P_{buy,i}^r\right]$, where $Q$ and $P$ denote the amount and the quotation of BCRECs, respectively; the superscript $r$ indicates the index of the $r$-th round of P2P transactions. $\left[Q_{buy,i}^1,P_{buy,i}^1\right]$ is the initial order

$$Q_{buy,i}^1=\mathrm{m}\mathrm{a}\mathrm{x}(K_i{Q}_{d,i} , Q_{d,i}-Q_{acc,i})$$

(1)

If the market is globally oversupplied, $K_i>1$, and ${\mathcal{A}}_i$ purchases more BCRECs than the RPS requirement; otherwise, $K_i<1$, and ${\mathcal{A}}_i$ decreases the purchase amount and uses some accumulated BCRECs to fulfill the RPS requirement. Purchases cannot be less than the difference between the net demand and accumulation to ensure that the RPS can be fulfilled. $Q_{acc,i}$ is the amount of ${\mathcal{A}}_i$’s BCREC accumulation, and $K_i$ is determined as follows:

$$K_i=1+{\xi }_i\left(\frac{Q_S}{Q_D}-1\right)$$

(2)

${\xi }_i\in \left[\mathrm{0,1}\right]$ is determined as follows:

$${\xi }_i=\left\{\begin{array}{c}\min \left({\rho }_i\frac{Q_{acc,i}}{Q_{d,i}} ,1\right) , Q_S\le Q_D\\ \max \left(1-{\rho }_i\frac{Q_{acc,i}}{Q_{d,i}} , 0\right) , Q_S>Q_D\end{array}\right.$$

(3)

${\rho }_i$ $\in \left[\mathrm{0.5,1.5}\right]$, reflecting the sensitivity of ${\mathcal{A}}_i$ to $Q_{acc,i}$.

The quotation ${\mathcal{A}}_i$ submitted in the initial orders is set as follows:

$$P_{buy,i}^1=\check{P}+\left(\widehat{P}-\check{P}\right)\times \tau$$

(4)

where $\tau$ is a random number for simulating individual diversity among different AEAs, and $0\le \tau \le 1.$ Due to the existence of the centralized trading phase, the quotations of AEAs in the P2P phase can only vary within $\check{P}$ and $\widehat{P}$, which denote the fixed purchasing and retail prices offered by the BCREC broker during the complementary trading phase, respectively.

Similarly, as for seller ${\mathcal{A}}_j$, its transaction order is denoted by $\left[Q_{sell,j}^r,P_{sell,j}^r\right]$, where $Q$ and $P$ denote the amount and the quotation of BCRECs, respectively. The initial order of ${\mathcal{A}}_j$ is $\left[Q_{sell,j}^1,P_{sell,j}^1\right]$, and the initial amount is set as follows:

$$Q_{sell,j}^1=\mathrm{m}\mathrm{i}\mathrm{n}(K_j{Q}_{s,j} , Q_{s,j}+Q_{acc,j})$$

(5)

$$K_j=1+{\xi }_j\left(\frac{Q_D}{Q_S}-1\right)$$

(6)

$${\xi }_j=\left\{\begin{array}{c}\min \left({\rho }_j\frac{Q_{acc,j}}{Q_{s,j}} ,1\right) , Q_S\le Q_D\\ \max \left(1-{\rho }_j\frac{Q_{acc,j}}{Q_{s,j}},0\right) , Q_S>Q_D\end{array}\right.$$

(7)

The initial quotation of ${\mathcal{A}}_j$ is:

$$P_{sell,j}^1=\widehat{P}-\left(\widehat{P}-\check{P}\right)\times \tau$$

(8)

All transaction orders are broadcast on the trade platform in the AEA community. Buyers and sellers will find partners based on the transaction orders by scoring comprehensive assessments of each other.

3.4.2. P2P negotiation

During the P2P transaction phase, AEA can perform multiple rounds of counterparty selection and order-matching. To enhance the efficiency of order-matching, we designed a multi-option-based matching scheme [1]. This scheme allows buyers to select multiple sellers to attempt to trade within each round. On one hand, each buyer evaluates the scores of sellers based on two factors: BCREC quotation and BCREC abundance. Then, the buyer sorts the sellers in descending order, selects the top-N ones, and sends transaction requests to them sequentially. On the other hand, each seller evaluates the scores of buyers who request to transact with them and accepts the one with the highest score. If a buyer’s request is rejected by its first option, it will move to its second option, and so on. The multi-option-based matching scheme provides each buyer with N prior chances to ensure a successful transaction.

A transaction pair is established when a buyer and a seller mutually select each other, and then they can proceed to a transaction. The P2P market rules that each AEA can make at most one transaction per round; therefore, those who have already paired up cannot match with any other AEA in the same round. As for a transaction pair including buyer ${\mathcal{A}}_i$ and seller ${\mathcal{A}}_j$, we denote the deal price of BCREC, the deal amount of BCREC, and the market price by $P_{(i,j)}$, $Q_{(i,j)}$, and $\overline{P}$, respectively. Then, we have:

$$P_{\left(i,j\right)}^r=\frac{P_{buy,i}^r+P_{sell,j}^r}{2}$$

(9)

$${\overline{P}}^r=\frac{1}{\left|{\mathbb{D}}^r\right|}\sum _{\left(i,j\right)\in {\mathbb{D}}^r}{P}_{\left(i,j\right)}^r$$

(10)

$$Q_{\left(i,j\right)}^r=\min \left(Q_{buy,i}^r, Q_{sell,j}^r\right)$$

(11)

where ${\mathbb{D}}^r$ is the deal set, and the superscript $r$ indicates that the variable is associated to $r$-th round.

At the end of a transaction round, both parties update their orders for the next round. They deduct the amount of BCREC that they transacted from their orders. A buyer ${\mathcal{A}}_i$ updates the quotation by statically compromising two factors: the latest quotation of its own and market price:

$$P_{buy,i}^{r+1}=\delta P_{buy,i}^r+\left(1-\delta \right){\overline{P}}^r$$

(12)

where $\delta$ is a constant weight, and $0\le \delta \le 1$.

Similarly, the quotation updating of a seller ${\mathcal{A}}_j$ is:

$$P_{sell,j}^{r+1}=\delta P_{sell,j}^r+\left(1-\delta \right){\overline{P}}^r$$

(13)

No quotation updates if no deal is made. After $R$ rounds of negotiation, the AEAs, which still have remaining orders, can either enter the complementary phase to clear in one go (trade with the BCREC broker) or choose to hold as accumulation until the next month.

3.5. Post-Transaction Stage

The smart contract of the transaction ledger records the results of BCREC transactions on the blockchain. Each record corresponds to a matched order in the P2P market, identified by the trading period and the order serial number.

Table 4. Information included in a transaction record.

Name	Content	Recorder
time	Trading period of the transaction	AEA (Seller)
number	Serial number of the transaction
seller_ID	Seller’s ID in the real world
seller_Addr	Seller’s account address
buyer_ID	Buyer’s ID in the real world
buyer_Addr	Buyer’s account address
amount	BCREC amount of the transaction
price	BCREC price of the transaction
transfer_transactions_hash	The hash of the Ethereum transaction used to transfer BCREC	AEA (Seller)
payment_transactions_hash	The hash of the Ethereum transaction used for payment	AEA (Buyer)

shows the information of a complete transaction record.

3.5.1. Transaction Record

P2P transactions are based on the self-negotiation of buyers and sellers without a central authority to oversee them. Therefore, trustworthy records of transaction outcomes are essential to ensure the proper execution and settlement of orders. As mentioned before, the seller selects the deal request with the highest score and then closes the deal with the corresponding buyer. We design the system so that the seller is responsible for recording the transaction information on the chain, and the recording operation will trigger a corresponding event. The buyer will know whether their transaction request has been selected by listening to the event information. For each transaction, the seller calls the record function to record the time, serial number, two parties, amount, and price on the chain. A complete record also includes the Ethereum transaction hash for transferring the BCREC and paying the purchase price, which will be recorded on the chain by the AEA during the settlement session using the submitTransTx and submitPayTx functions (as shown in

Table 5. List of transaction ledger contract functions.

Contract Name	Function Name	Caller
Transaction Ledger	record (uint256, uint256, uint256, address, uint256, address, uint256, uint256)	AEA (Seller)
	submitTransTx (uint256, uint256, bytes32)	AEA (Seller)
	submitPayTx (uint256, uint256, bytes32)	AEA (Buyer)

).

3.5.2. Transaction Settlement

After the buyer sees that their trade request has been selected and correctly recorded on the chain by the seller, the buyer can pay the seller for the purchase of the BCRECs. To avoid the volatility of Ether, a cryptocurrency with a more stable purchasing power is issued for settlement by using the Stablecoin contract. The smart contract of Stablecoin is used to design a general equivalent that intermediates the payment and transfer of BCRECs. The cryptocurrency follows the ERC-20 token standard and its value is pegged to real-world fiat currency. This cryptocurrency is managed by the governmental administration and can only be used for the on-chain settlement of BCRECs. The cryptocurrency can be exchanged with fiat currency using a predetermined rate. The main functions of the contract are shown in

Table 6. List of Stablecoin contract functions.

Contract Name	Function Name	Caller
Stablecoin	issue (address, uint256)	Administrator
	burn (uint256)	Administrator
	transfer (address, uint256)	AEA

.

The buyer uses the transfer function of the Stablecoin contract to transfer the purchase cost to the seller and updates the transaction record of the transaction ledger contract with the corresponding transaction hash. The seller transfers the BCREC to the buyer after verifying that the payment has been received correctly and also updates the transaction record of the transaction ledger contract with the corresponding transaction hash. Only after the successful completion of the settlement can a complete and valid P2P transaction record be left on the chain. Any party with objections can reject the record and apply for arbitration.

4. Case Study

4.1. Simulation Setups

To demonstrate the operation mechanism of the BCREC trade market, we simulate a community consisting of 30 AEAs, 1 BCREC broker, and 1 RPS regulator (governmental administration) and develop the blockchain smart contracts on the Ethereum Sepolia test network. Each BCREC is flexibly divisible and represents 1 kW·h of renewable energy production. The BCREC trade is a monthly market. The issuance and disposal of BCRECs are carried out monthly before and after the transaction, respectively. The buying rate (broker buys from AEAs) of BCRECs $\check{P}$ = 0.090 CNY/unit, and the selling rate (broker sells to AEAs) of BCRECs $\widehat{P}$ = 0.130 CNY/unit. The simulation environment consisted of an Intel ^® Core ™ i7-7700 @ 3.6 GHz, 8 GB DDR3 RAM, MATLAB-2018a, CPLEX12.8.

We investigate two market scenarios—initial oversupply and undersupply—and use them for two consecutive months to test the effectiveness of the BCREC trade in balancing supply and demand. Figure 5 shows the supply and demand of each AEA at the start of the two months and the initial accumulation in the BCRECs of each AEA.

Table A1. Sensitivity factor ${\rho }_i$ of AEAs.

AEA No.	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15
${\mathit{\rho }}_{\mathit{i}}$	0.56	1.27	1.05	0.92	1.43	1.14	0.79	1.33	0.55	0.67	0.60	1.37	1.04	1.25	0.98
AEA No.	16	17	18	19	20	21	22	23	24	25	26	27	28	29	30
${\mathit{\rho }}_{\mathit{i}}$	1.01	1.21	0.69	0.92	1.42	0.90	0.84	0.92	0.63	1.41	1.34	0.78	1.02	0.65	1.15

gives the sensitivity factor of each AEA.

Table 7. Description of market information for both months.

Trade Period	The Market Supply of BCREC (Unit)	The Market Demand of BCREC (Unit)	Normalized Ratio of Supply and Demand
Month 1#	110,616.9	63,239.2	1.749/1
Month 2#	61,036.0	113,217.9	0.549/1

summarizes the overall market supply and demand information for the two months. Since both market scenarios are set to be significantly unbalanced between supply and demand, there will be a large number of untraded certificates left after the P2P phase if the AEAs do not adjust transaction orders. This, in turn, leads to a reduction in AEA profits and market efficiency. We set $R=16$ and $N=2$, i.e., the P2P transaction phase maximally includes 16 rounds, and a buyer can maximally choose two sellers to negotiate in each round.

4.2. Functional Demonstration

The P2P transaction results of the two months are listed in

Table A2. P2P transaction results of Month 1#.

Deal No.	Buyer No.	Seller No.	Amount of BCRECs (Units)	Price of BCREC (CNY/Unit)
1	16	3	6199.270	0.095
2	10	22	4881.734	0.110
3	20	29	7506.800	0.122
4	9	6	3519.386	0.111
5	24	11	3394.142	0.111
6	15	13	8422.506	0.114
7	8	28	5553.351	0.109
8	7	30	3906.981	0.106
9	21	17	1217.913	0.109
10	19	25	7603.875	0.112
11	5	1	3967.500	0.113
12	8	6	29.065	0.108
13	10	29	529.997	0.113
14	16	23	5022.131	0.111
15	14	25	1380.000	0.111
16	26	27	2279.788	0.110
17	8	12	4492.609	0.112
18	24	18	2524.318	0.112
19	21	23	3508.290	0.110
20	7	4	2329.163	0.111
21	26	2	1964.917	0.109
22	7	25	139.320	0.111
23	21	18	685.188	0.110
24	26	18	654.539	0.108
25	26	25	275.020	0.108

and

Table A3. P2P transaction results of Month 2#.

Deal No.	Buyer No.	Seller No.	Amount of BCRECs (Units)	Price of BCREC (CNY/Unit)
1	15	2	5755.3	0.108
2	17	28	11,929.8	0.101
3	22	8	6896.4	0.108
4	3	16	8423.7	0.119
5	1	23	5249.9	0.109
6	30	8	515.1	0.106
7	11	14	4934.4	0.111
8	10	7	5891.0	0.109
9	15	14	595.1	0.110
10	25	16	3562.5	0.109
11	5	4	4135.0	0.109
12	26	12	5919.7	0.109
13	10	14	102.4	0.109
14	17	16	601.0	0.109
15	17	4	168.9	0.109
16	27	12	2564.8	0.110
17	30	18	7187.6	0.109
18	21	9	3072.4	0.110
19	19	13	3424.5	0.111
20	10	4	0.2	0.110
21	6	13	2367.0	0.111
22	20	18	456.1	0.108
23	20	4	431.1	0.108
24	24	13	2310.2	0.111
25	20	12	419.0	0.108
26	29	13	1149.2	0.112
27	20	13	1168.1	0.110
28	1	9	224.1	0.112
29	1	13	831.2	0.113

, respectively. Without the loss of generality, we take the first transaction of Month 1# as an example to demonstrate the main on-chain process of the BCREC trade. In this deal, AEA16 buys 6199.270 BCRECs from AEA3 at a price of 0.095 CNY/unit. Since BCREC is ERC-20-compliant, it can be easily managed using various types of wallet software. Figure 6 shows the BCRECs held by AEA16 before the Month 1# trade as viewed using MetaMask wallet software.

4.2.1. BCREC Issuance

In this paper, we use a manually written JSON file to record the renewable energy generation data since there is no real production environment available. The data can be obtained by sending an HTTP GET request to the corresponding URL on the server. AEA3’s net supply of 9800 units in Month 1# indicates that it has generated more renewable energy than the required amount by the RPS in the current period. For simplicity, we directly issue BCREC to AEA3 for the equivalent of 9800 kW·h of renewable energy generation. The main purpose is to demonstrate the BCREC issuance process.

The issuance process is the execution of the request–response cycle mentioned in Section 3.3.2. Specifically, the request from the BCREC issuance contract should include the following parameters:

• Link token address: Since Chainlink offers its services on multiple blockchains, it is necessary to specify the address of the link token contract on the chosen blockchain. The address on the Ethernet Sepolia test chain is 0x779877A7B0D9E8603169DdbD7836e478b4624789.
• Oracle address: This specifies which oracle node operator provides the service. The test in this article used 0x6090149792dAAeE9D1D568c9f9a6F6B46AA29eFD.
• Job ID: This corresponds to the content of the task requesting the work, which is to obtain uint256 format data through HTTP GET, corresponding to the ID of ca98366cc7314957b8c012c72f05aeeb.
• Request fee: This specifies the service fee to be paid for the request, which is 0.1LINK per request in the test chain environment.
• Task parameters: These are parameters required to complete the task content corresponding to the Job ID. Specifically, retrieving the generation data consists of the following tasks: HTTP request, parsing the JSON data, multiplying the data to convert it to an integer, and encoding the data with Ethereum ABI. The required parameters are the URL of the GET request, the path to the data in the JSON file, and the multiplier to multiply the data for integer conversion.
• Callback function signature: This specifies the function that is called when the response is returned. The BCREC issuance contract has a fill function to process the response from Chainlink and uses the resulting data to call the issue function of the BCREC contract.

This process is accomplished through the on-chain smart contract in cooperation with the off-chain Chainlink nodes in strict accordance with the established rules. There are two Ethereum transactions, i.e., the transaction that triggers the request (https://sepolia.etherscan.io/tx/0x6568194be965572db85fe084a4d61fce937c02adabcc8dc24f00aa4b6163593d accessed on 15 October 2023) and the transaction that triggers the response (https://sepolia.etherscan.io/tx/0xfc7babbcb2a83fe382e02f9d393a5566b7c7d96b7e4c45feb483eb4f246569d9 accessed on 15 October 2023). The entire process can be automated using Chainlink’s oracle service.

4.2.2. Record of Deals

The P2P transaction process has multiple rounds of partner selection and transaction attempts, which cause a large amount of transaction delays and fees if conducted directly on the blockchain. Therefore, the transaction negotiation is conducted off-chain first; the BCREC transfer and the fee payment are then made on-chain for the concluded transaction orders, and the order information is recorded. As described in Section 3.5.1, the information of the concluded trade order is contained in an on-chain record made by the seller, and this operation triggers the New Transaction event in the transaction ledger contract, which can be monitored by an AEA to obtain the latest on-chain order record. For AEA3 and AEA16, after confirming that the transaction time, order serial number, counterparty, transaction volume, and unit price are correct, the BCREC can be transferred and the fee can be paid, respectively. Upon completion, the corresponding transaction hash is replenished to the on-chain record of the order, i.e., a P2P transaction order is completed. Submitting the transaction hash also triggers the corresponding event to pass information to the AEA. Figure 7 shows the event record as viewed through Etherscan:

Deal records are publicly available on the chain. The order information recorded on the chain can be accessed using the transaction time and order serial number. Figure 8 shows the complete record of the first order in Month 1#: The recording process needs to be completed by both the buyer and seller and a consensus needs to be reached to complete the recording of the order’s information. The completed record also contains the hash of the transfer and payment transactions; therefore, the complete record of the order information that can correspond to the order has been completed.

4.2.3. Transfer and Disposal

When BCRECs are used to fulfill mandatory RPS quota requirements, they need to be reclaimed and destroyed to prevent reuse. The AEAs are required to authorize a sufficient amount of BCRECs to fulfill the quota to the RPS regulator, and the regulator can only perform the disposal operation on the authorized amount of BCRECs—any excessive disposal is prohibited. Users can easily perform authorization operations using wallet software such as MetaMask. To demonstrate the underwriting process, Figure 9 shows the on-chain record that AEA16 authorizes 100 units of BCREC to be disposed of by the regulator.

4.3. Quantitative Evaluation

4.3.1. Month 1# Trading

When the market supply and demand are imbalanced, AEAs can modify the orders they submit to the market based on their net supply and demand by increasing or decreasing their BCREC accumulation, thereby achieving a market equilibrium of BCREC purchases and sales. A balanced trading market enables more transactions to be completed during the P2P phase, leading to higher profits for both buyers and sellers. Figure 10 illustrates the initial net supply and demand of the 30 AEAs and the quantities in the orders submitted after adjustment. In the trading market of Month 1#, the adjusted amount of BCREC sales and purchases are 82,606.614 units and 81,987.804 units, respectively, and the market supply and demand were essentially balanced.

Figure 11 illustrates the ratio of the seller/buyer AEA’s BCREC accumulation to the net supply/demand of BCREC as well as the adjustment magnitude in the submitted orders relative to the net supply/demand. When there is an oversupply, the adjustment means that the BCRECs are temporarily stored for future trading or use; therefore, AEAs that already have a large accumulation in BCRECs are less motivated to adjust. As Section 3.4.1 indicates, different AEAs have different sensitivity to their BCREC accumulation, which reflects their individual diversity and also affects their adjustment level.

Figure 12 illustrates the change in the number of orders for the sale and purchase of BCRECs in the market, showing an increase in the number of trading rounds during the P2P transaction phase and the number of orders concluded in each round. The P2P transaction phase ended after the eighth round of counterparty selection and order matching when all orders to buy BCRECs were cleared. Table A2 presents all P2P transaction deals completed in this phase. A total of 81,987.804 units of BCRECs were directly traded through P2P, and 618.810 units of BCRECs were not successfully sold in the P2P phase and could be sold a through complementary phase.

When there is an initial oversupply of BCRECs in Month 1#, the seller AEA reduces the amount of BCRECs it sells by keeping some of the newly acquired BCRECs for future transaction or use. The buyer AEA increases the amount of BCRECs it buys by obtaining more BCRECs than it needs to meet the RPS requirements. This way, both buyers and sellers make the market more balanced by adjusting their BCREC transactions and increasing their BCREC accumulation. As a result, the total BCREC accumulation in the AEA community rises from 83,367.443 units to 130,126.334 units at the end of the first month of trading, which includes 28,010.286 units reserved by the seller and 18,748.604 units over-bought by the buyer.

4.3.2. Month 2# Trading

Figure 13 shows how the AEA community adjusts the BCRECs sold and bought for the initial undersupply scenario in Month 2#. Figure 14 shows the ratio of the BCREC accumulation to the initial net BCREC supply or demand for each AEA and the change in the BCREC trading quantity compared to the net BCREC supply or demand. In the trading in Month 2#, the AEA sells more BCRECs and buys fewer BCRECs by using the BCRECs accumulated in previous trading periods to balance the market. The final BCREC sale quantity is 90,356.462 units, and the BCREC purchase quantity is 90,285.700 units. The more BCRECs an AEA accumulates, the easier it is to increase the amount of BCRECs it sells or to use more accumulated BCRECs instead of buying them. Some AEAs have a higher BCREC accumulation ratio in Month 2# because they increased their BCREC accumulation in Month 1#. But, the elasticity coefficient $\xi \in \left[\mathrm{0,1}\right]$ limits the maximum adjustment magnitude for the seller AEA to $1-Q_S/Q_D$ and for the buyer AEA to $Q_D/Q_S-1$.

Figure 15 shows how the number of BCREC sale and purchase orders in the market changes with the number of P2P transaction rounds and how many deals are completed in each round. After 10 rounds, all BCREC sale orders were filled and recorded in Table A3. There were 29 P2P transactions with a total volume of 90,285.700 units. The remaining 70.761 units of purchase volume moved to the complementary phase.

The AEA community’s accumulation in BCRECs dropped from 130,126.334 units to 78,015.195 units by the end of Month 2#, as sellers sold 29,249.700 units and buyers used 22,861.438 units of BCRECs. The accumulation in BCRECs decreased by 5352.249 units after both trading periods.

4.3.3. Market Trading Efficiency and AEA Profits Comparison

BCREC’s flexibility allows an AEA to adjust the amount of traded BCRECs proactively, balancing the market supply and demand, and to trade more BCRECs through P2P for higher economic returns. After two trading periods, the total sale revenue of the AEA community is 18,997.335 CNY, and the total purchase expenditure is 18,950.842 CNY. The total P2P transaction volume for both periods was 172,273.504 units and the total complementary transaction volume was 689.571 units (618.810 units sold and 70.761 units bought).

If the RECs can only be traded once and must be completely cleared out in each trading period, AEA cannot modify the initial supply/demand and can only trade directly. Simulating the trading situation using traditional REC rules with the same method and data, the AEA community earned 17,681.957 CNY from sales and spent 20,201.611 CNY on purchases after the two trading periods. The P2P transaction volume of the community was 124,275.200 units, and the complementary transaction volume was 99,559.600 units (purchase volume: 52,181.900 units; sale volume: 47,377.700 units).

Table 8. Comparison of P2P market volumes.

Trade Period	Trade Type	P2P Market Volume (Unit)	Outstanding Volume (Unit)
Month 1#	Single	63,239.200	47,377.700
	Multi	81,987.804	618.810
Month 2#	Single	61,036.000	52,181.900
	Multi	90,285.700	70.761

shows the volume and outstanding volume of the P2P market for both types of single-transactability and multi-transactability in the two trading periods. Since traditional RECs cannot be traded repeatedly and must be cleared in the current period, unbalanced trading demand in the market cannot be transacted at all via P2P. This significantly reduces the trading efficiency of the P2P market, with the outstanding volume accounting for 74.92% of volume in Month 1# and 85.49% of volume in Month 2#. In contrast, the BCREC-supported P2P trading market, with its flexible trading mechanism, enables AEA to proactively adjust trading demand and balance the market; thus, the vast majority of trading demand is met through P2P.

As shown in

Table 9. Comparison of the BCREC trade and the traditional REC trade in AEA profits.

Economic Index	BCREC Trade	Traditional REC Trade	Improvement of BCREC *
Revenue (CNY)	18,997.335	17,681.957	1154.255
Expenditure (CNY)	18,950.842	20,201.611	−716.101

* The effects of initial accumulation BCREC are removed.

, more P2P trading also leads to higher returns for AEAs. To compare the benefits of the two types, we need to consider the impact of the accumulated BCRECs. BCREC trading resulted in a net decrease of 5352.249 units in accumulated BCRECs, of which 4112.834 units were used net for purchase order adjustments and 1239.414 units were sold net for sales order adjustments. Even at the maximum price of 0.130 CNY/unit, after deducting the impact of BCRECs’ accumulation, transactions using BCRECs still resulted in an extra 1154.255 CNY of sale proceeds and 716.101 CNY of purchase cost savings for the AEA community compared to transactions using traditional RECs rules.

5. Conclusions and Future Work

In this paper, we propose a BCREC system based on blockchain technology to solve the REC trading problem in the AEA community. By converting RECs into digital assets on the chain, we enhance their trading activity. Besides P2P transactions, we also enable REC to be traded multiple times with the support of more flexible BCREC implementation, which allows AEA to adjust the trading volume dynamically, balance the market supply and demand, and achieve more deals through P2P transactions. We design the full lifecycle of the BCREC using the blockchain oracle to obtain renewable energy production data and issue the corresponding amount of BCRECs automatically through a smart contract. After the BCRECs are used to meet the RPS, we establish a disposal mechanism to remove them from circulation. The issuance, transfer, and disposal of BCRECs are all executed through smart contracts in strict accordance with the predefined rules and leave verifiable records on the chain. We create P2P transaction orders concluded by the market on the blockchain as a repository to ensure the orderly conduct of transactions.

We numerically analyze a community consisting of 30 AEAs and simulate two consecutive markets with unbalanced supply and demand. The results demonstrate that, with the support of BCRECs that can be traded multiple times, AEAs can improve the supply–demand imbalance of the market, increase the P2P transaction volume, and obtain higher trading revenue by actively adjusting the trading orders and responding to the market’s supply and demand. We deploy the relevant smart contracts on the Ethereum Sepolia testnet and verify the functionality of the whole process of issuance, trading, and underwriting. AEA can easily manage its BCRECs using wallet software such as MetaMask and view the on-chain records using blockchain browsers such as Etherscan.

Our future work consists of two aspects. First, we plan to expand the community’s size, study the trading outcomes when AEA participation is increased, and extend the trading period to examine the trading performance of the market for a longer duration. Second, we aim to improve the blockchain layer, develop the front–end interface to integrate various functional components, and conduct cost and performance analysis.