Review on low carbon planning and operation of integrated energy systems

Abstract

The integrated energy system is an important prerequisite for the sustainable transformation to the low‐carbon power system. Therefore, this paper aims to provide readers with insights into the existing research about the planning and operation models of integrated energy systems. First, the general structure of integrated energy systems is elaborated, and the new issues faced by the low‐carbon planning and operation for the integrated energy systems are represented. Next, the technical framework and the general model are proposed. Furthermore, the typical models and solution technologies of the low‐carbon planning and operation are summarized based on specific models operating in multiple scenarios. Finally, based on the characteristics and framework of the integrated energy system, the future research direction is given, which is expected to guide the sustainable development and low‐carbon transformation of energy systems.

INTRODUCTION

At present, the excessive combustion of fossil energy has brought many problems, such as serious environmental pollution, excessive greenhouse gas emissions and energy inefficiency. Because of the excessive emission of greenhouse gases, the climate has changed dramatically, and extreme weather events have increased day by day, affecting the supply and consumption of energy. Improving energy efficiency, ensuring energy security, promoting renewable energy consumption, and promoting environmental protection has become the inevitable choice for sustainable energy development.^1–5 In the process of the global low-carbon transition, developed countries and regions such as the United States, Japan, Canada, and the European Union have formulated climate strategies and committed to achieving net-zero CO₂ emissions by 2050. China continues to improve its nationally determined contributions, announcing in September 2020 that it will strive to achieve “carbon peak” by 2030 and “carbon neutrality” by 2060.⁶ In this context, integrated energy systems (IES) emerge. Globally, several integrated energy demonstration projects such as the EU ElECTRA Demonstration Project, Japan's Baiye Smart City, Sino-Singapore Tianjin Ecological City, Jiangsu Tongli Integrated Energy Service Center, and Shanghai Chongming Island Demonstration Project have confirmed the importance of IESs in carbon reduction.^7,8 In addition, the low-carbon development strategies of many countries and regions, such as the “European Green Agreement” proposed by the European Union in 2019 and the “Green Growth Strategy” proposed by Japan in 2020, have also defined the important role of the IES.⁶ In the context of carbon reduction, how to promote the low-carbon operation of the IES has become an important issue at present.

However, there is no comprehensive model framework focusing on the low carbon viewpoint considering optimal environmental and economic benefits and supply-demand coordination in the existing literature. In this paper, the technical framework and the general model are proposed, and the typical models and solution technologies of the low-carbon planning and operation are reviewed. The contributions of this paper can be summarized as follows.

• (1)

  The planning and operation of IESs are reviewed innovatively focusing on the low carbon viewpoint considering optimal environmental and economic benefits.

• (2)

  The new challenges of IESs with respect to high penetration of renewable energy sources are investigated, and crucial technologies to achieve the successful coordination of energy supply and demand are summarized.

• (3)

  The new challenges, general models and typical solution technologies associated with the low-carbon planning and operation are summarized within the modeling process, which give more insights for the future researchers.

GENERAL STRUCTURE OF IES

To achieve the mutual transformation and fusion application of thermal, solar, natural gas and electric energy, IESs require a wide range of energy input, output and conversion equipment. The operation diagram of an IES is shown in Figure 1.

Figure 1. The operation diagram of an integrated energy system.

In addition, it is necessary to design subsystems and build a functional structure from the perspective of functional zoning when establishing IESs. The general subsystem functions of the IES should focus on the following aspects:

• (1)

  Energy supply system: The main role of this subsystem is the energy input, where external energy can enter the IES and maintain its smooth operation. In practical operation, the energy supply subsystem is the backbone of the IES, which supplies a variety of energy resources to the latter. For example, the subsystem can directly supply primary energy such as natural gas and coal to the IES. It can also support the operation process of the IES based on electric energy.

• (2)

  Energy conversion subsystem: In the IES, different types of energy need to be converted to each other to meet the uncertain demand, which is achieved by the energy conversion subsystem. Generally speaking, there are two types of energy conversion subsystems: The first category is renewable power generation system, including photovoltaic, wind and hydroelectric power. The second type is combined heat and power (CHP) or combined cooling, heating and power (CCHP) system with the internal combustion engine, gas turbine, micro-fuel engine, fuel cell and Stirling engine as the prime motor.⁹

• (3)

  Energy transmission network: In the application of IESs, all types of energy will be transmitted to the demand side in different types based on the energy transmission network, which includes power grid, heating network, cooling network and gas network.

• (4)

  User subsystem: In the operation of the IES, the user subsystem is the main subject of energy consumption, and the direct users of the energy system are mainly electricity users, including industrial, commercial and civil power users. Meanwhile, load reduction, load transfer and load conversion are realized through the flexible load in the IES.

The topology structure diagram of general IESs is shown in Figure 2.

Figure 2. The topology structure diagram of general integrated energy systems.

NEW ISSUES AND TYPICAL MODELS OF THE LOW-CARBON PLANNING AND OPERATION FOR THE IES New issues

IES changes the mode in which each energy subsystem plans and operates separately, and realizes coupling and complementation of different energy subsystems. So IES can make full use of the advantages of various energy sources to enhance the efficiency and flexibility of energy utilization. However, the characteristics of IES such as tight coupling of multiple energy sources and increasing source-load uncertainty, and the demand for environment-protection performance of the system under the low-carbon background causes new issues of IES planning.

• (1)

  Increasing uncertainty of users' energy demand on the load side. Different from the single power demand in the traditional power system, users in IES have multiple energy demands such as electricity, cold and heat, leading to increasing uncertainty of users' energy demand. As a result, traditional dispatching methods are difficult to adapt to the dynamic changes of the load of the system accurately.

• (2)

  Increasing energy fluctuations on the source side. The proportion of renewable energy in IES is greatly increased while the output of renewable energy power generation such as wind power and photovoltaic power generation is intermittent and uncertain. Consequently, the energy fluctuations of the source of the system increase, which makes it more difficult to balance the supply and demand of the system.

• (3)

  Carbon level assessment. Traditional system planning aims at economic optimization under the premise of ensuring reliability. But under the low-carbon background, system planning needs to consider both economic and environmental benefits and thus carbon level is included in the model assessment system. How to compare economic benefits and environmental benefits and realize the optimization of the economy and environment-protection performance of the model is a new issue faced by system planning.

• (4)

  The development of low-carbon power marketization. The carbon trading mechanism puts an economic value on carbon emissions and overcomes the inconsistency between economic benefit and environmental benefit evaluation. It's one of the effective methods to assess carbon emission. But under the carbon trading mechanism, the overall cost of the system also needs to consider the additional environmental cost caused by carbon emissions. To calculate this cost, new carbon trading cost models are needed.

• (5)

  IES operation simulation. Due to the diversity of energy sources in IES, the operating environment of the system is more complex. To obtain the operation data of the planning model in a large number of different scenarios and assess the rationality of the planning model, it is necessary to realize the operation simulation of IES through simulation technology.

• (6)

  Multiagent collaborative planning. IES is no longer the aggregation of independent subsystems, but the tight coupling of each subsystem. Due to the tight coupling of each link in the system, the disturbance of one link will propagate through coupling elements. Hence, independent planning for the single agent is not suitable for IES anymore.

Typical models

Similar to the traditional energy system, the optimization planning of the IES is to establish a mathematical model that can describe the operating characteristics of the system on the basis of the known operating characteristics and boundary conditions of the planning objects, and then obtain the planning scheme that meets various indicators. It is the prerequisite to clarify the model application scenario, the planning objective and the constraints. The application scenario of the IESs mainly focuses on the coordination of source-grid-load-storage and multi-energy flows. The goal of source-grid-load-storage collaborative planning is to strengthen the mutual support capacity between the energy network and load. The IES planning considering the cooperation of multi-energy flows is to improve the clean energy absorption rate and multi-energy complementary capacity on the premise of ensuring the reliability of energy supply. The optimization framework of the IES is shown in Figure 3.

Figure 3. The optimization framework of the integrated energy system.

Decision variables

Select equipment operation status and energy consumption rate as operating variable O_i,j,t, carbon emission level and renewable energy consumption level as environmental variable E_i,j,t and the dynamic change of renewable energy power generation output on the source and energy demands on the load as fluctuation variable U_i,j,t.

Typical objective functions

Traditionally, the optimization objective function of an energy system is to minimize the total cost of the system, which economically indicates the affordability for the power sector.¹⁰ To cope with climate mitigation and implement the sustainable development concept of energy conservation and emission reduction, a supplement of environmental concerns is included in the objective function, such as minimizing carbon emission and maximizing the renewable energy share of the generation mix.^11,12

The low-carbon planning and operation of IES comprehensively consider both economic and environmental benefits, and generally aim at the optimization of economy and environment-protection performance. Meanwhile, they consider meeting users' energy demand and enhancing the consumption level of renewable energy as much as possible.

The systems without considering carbon trading generally adopt multi-objective functions as follows: [Image Omitted. See PDF]where F is the total cost and F_i represents each part of it except carbon trading cost. E is the total carbon emissions of the system. E_i represents the carbon emissions of each power generation equipment and E' represents the carbon emissions reduced by carbon reduction equipment. η₁ … η_i  are the utilization rate of various renewable energy sources.

Subsequently, environmental objectives are generally combined with the cost objective.^13,14

In the existing research of IESs, the basic ideas of low-carbon planning and operating are as follows: Carbon emission or carbon trading prices are converted into corresponding factors, variables, or constraints, and incorporated into the planning and operation model of the IES, and then put forward a system operation optimization strategy considering both environment and economy.

The typical single objective function for IESs is as follows: [Image Omitted. See PDF]where F is the total cost. F₁ represents the total operation cost, including equipment operation and maintenance cost. F₂ represents environmental cost, including the cost of carbon reduction measures (e.g., carbon capture), carbon trading cost, and the penalty costs for unused renewable energy. F₃ is the supply-demand coordination cost, including flexible load dispatching cost and demand response compensation cost.

Typical constraints

The low-carbon planning and operation of IES should also take environmental protection constraints and supply-demand coordination constraints into consideration after considering the reliability and operation constraints of the system.

Typical constraints are as follows: [Image Omitted. See PDF]where g(O_i,j,tU,_i,j,t) are the operation constraints, including power generation and transmission constraints and fuel supply constraints. h(O_i,j,tU,_i,j,t) are the reliability constraints, including power balance and capacity expansion constraints. k(O_i,j,tE,_i,j,t, U_i,j,t) are the environmental protection constraints, including carbon emissions and renewable energy rate constraints. m(O_i,j,tU,_i,j,t) are the supply-demand coordination constraints, including dispatchable generation constraints and demand response constraints. $${\delta }_{1}$$is the operating boundary associated with the operation parameters of the equipment and the energy network. $${\delta }_{2}$$is the stable operating boundary related to the energy balance within the system. $${\delta }_{3}$$is the environmental boundary considering carbon quota limits and renewable energy utilization rates. $${\delta }_{4}$$is the boundary related to supply and demand response coordination.

Solution methods

IES planning is usually a nonlinear programming problem. The solution methods of it are mainly mathematical programming and heuristic algorithms.

Typical models define the basic forms of the low-carbon planning and operation of IES, but to solve specific problems, decision variables, objective functions and constraints should be modified according to the application scenarios such as carbon trading, multiple energy synergies, demand response, renewable energy consumption, and flexible load.

SOLUTIONS TO LOW-CARBON PLANNING AND OPERATION OF IESS Support technology based on the application scenario

Combined with the new issues, the planning framework and the typical models of the IESs, precise prediction of energy supply and demand and low carbon electricity market are considered as the crucial technologies to achieve the low-carbon planning and operation of IESs.

Precise prediction of energy supply and demand

Research shows that carbon emissions cannot be curbed without conducting a comprehensive shift from fossil to renewable resources.¹⁵ Meanwhile, due to the application of renewable energy sources, the randomness of the new generation of IESs will be greatly enhanced compared with traditional energy systems. Research indicates that variability and stochastic variation of renewable resources have a high cost, which can be considerably reduced due to an efficient prediction of the source fluctuation.¹⁶ Therefore, taking the randomness of new energy sources into account during the modeling process is a prerequisite for current planning research. As the main clean energy sources, wind energy and solar energy are affected by wind speed and light intensity respectively. Thus, combining the probability distribution characteristics of light intensity and wind speed, building a random output model of clean energy is a common method at present.

The energy supply of IESs is affected by climate, geographical environment, regional layout and other factors. Extensive and in-depth research has been carried out on various renewable energy supply forecasting. Mathematical models such as gray models and intelligent algorithms such as neural networks play a significant role in the energy supply prediction of IESs. A new research model based on the exponential smoothing and GM(1,1) methods is carried out in [17], which puts forward two new ideas to forecast the sustainable level of China's energy supply, enriching the theoretical study of energy security forecasting. Artificial neural network(ANN) provides a powerful means for in-depth study of modeling solar energy systems. The commonly used ANN types and architectures are overview in [18], including multilayer perceptron neural network, Elman neural network and so on. In terms of wind-generated power supply prediction, a wind speed prediction model based on GA-ANN improved by variational mode decomposition(VMD) is proposed in [19], which can efficiently improve the accuracy of wind-generated power supply prediction.

Different from the traditional distribution network with only fixed load, a variety of new loads will participate in power grid dispatching on the consumption side of the integrated power system. The capacity allocation of energy equipment will be affected by the user's differential load demand and the ratio of energy consumption to heterogeneous energy flow. Therefore, it is necessary to analyze the load characteristics of IESs in large-scale application scenarios. Mathematical models such as gray models and intelligent algorithms such as neural network play an important role in the energy demand prediction of IESs. Refs. [20–23] predict energy demand in macro scenarios using a gray model method. Ref. [24] ensembles various artificial neural networks to improve the prediction accuracy of energy consumption in regional scenarios, such as feed-forward back propagation neural network (FFNN), radial basis function network (RBFN), and adaptive neuro-fuzzy interference system (ANFIS).

With the rapid development of artificial intelligence and data science, energy demand forecasting methods represented by deep-learning and reinforcement learning are widely applied in the field of grid load prediction, which provide a new way to promote forecasting accuracy in integrated power systems. A reliable and efficient power system integrating alternative sources of energy is proposed based on distributed intelligent multiagent technology in [25]. Ref. [26] develops a mathematical model of a hydrogen-based self-sustaining integrated renewable electricity network (HySIREN) employing a supply-demand forecasting model and deep-learning (DL) algorithm, and constructs a self-sustaining energy system with enhanced flexibility. A novel real-time incentive-based demand response algorithm for smart grid systems with reinforcement learning and deep neural networks is proposed in [27], which uses a deep neural network to predict future price and energy demand and adopts reinforcement learning to obtain the optimal incentive rates.

Accurate forecast of renewable energy supply and integrated energy demand on the multi-time scale is the basis of integrated power system planning and modeling. A summary of the typical prediction models of energy supply and demand is shown in Table 1. It can be foreseen that artificial intelligence theory will further enhance the practicability of the IES planning and operation model in the future.

Table 1 Typical prediction models of energy supply and demand.

Prediction model types	Prediction models
Mathematical models	Gray models
	Auto-regressive integrated moving average (ARIMA) models
Machine learning models	Regression tree
	Random forest
	Support vector regression
	Neural network
	Bayesian forecast
	Ensemble learning
	Deep learning

Low carbon electricity market

The low carbon electricity market is the basic application scenario of the integrated power system. Low carbon technologies will fundamentally reshape the electricity sector. Thus, the future low carbon electricity market and associated regulation scheme require a comprehensive new design.²⁸ Analyzing the interaction mechanism and operation mode of carbon market and electricity market is the current research hotspot. Since the Kyoto Protocol became effective, many countries in the world have explored and built the carbon trading market, but there is still no uniform trading rule for the carbon market in the world. Carbon quota allocation plays a fundamental and crucial role in the carbon emission market. There are three kinds of allocation methods: benchmarking, historical emission method and auction. By March 2020, one international, five national, sixteen provincial/state and seven city-level carbon markets have been established around the world, covering more than 20% of global greenhouse gas emissions.²⁹ There is a positive and symmetrical correlation between the electricity market and the carbon trading market. Therefore, analyzing the interaction between them, constructing a low carbon electricity market mechanism with a suitable operation mode is a crucial prerequisite for the planning and operation research of integrated power systems. The new characteristics of the low carbon electricity market compared with the traditional electricity market are summarized in Table 2.

Table 2 The new characteristics of the low carbon electricity market.

Aspects	Types	Illustrations
Carbon quota allocation	Benchmarking method	Quota allocation according to the industry unified carbon emission intensity, the most widely used
	Historical emission method	Quota allocation according to the historical emission level of the enterprises
	Auction	Enterprises obtain quotas through auction according to their own needs
Carbon trading prices	Single carbon price	Constant price representing the medium and long-term supply and demand
	Ladder-type carbon price	Dynamic price changing with the carbon emissions, representing the real-time supply and demand
	Carbon prices based on market clearing	Constant price representing the real-time supply and demand
Carbon trading commodities	Carbon emissions commodities	Carbon emission permission, encouraging enterprises to reduce carbon emission cost by emissions reduction
	Carbon offset commodities	Carbon emissions offset or participating in market trading, encouraging the utilization of emission reduction technologies such as renewable energy sources and carbon capture systems

At present, the research on integrated power system planning and operation under the background of carbon trading is gaining more and more attention in the academic community, and related scientific research results are abundant. At the planning level, an electric-gas-IES planning model that takes wind power and carbon trading into account is established in [30]. At the operational level, a hierarchical low-carbon operation framework considering a carbon-constrained locational marginal price (LMP) is proposed in [31], which employs carbon emission flow (CEF) to calculate the emission of the upper supply layer and the lower regional consumption layer. At the dispatch level, Ref. [32] establishes a stochastic economic model considering both the uncertainty of wind power and the carbon trading mechanism, which can provide reference for the scientific allocation of the load demand of wind power IESs.

Other theories are often combined with carbon trading mechanisms to optimize the planning, operation and dispatch of energy systems. Ref. [33] establishes a dispatching model with carbon-green certificate coordinated trading mechanism for virtual power plants (VPP) system integrating carbon capture power plants, power-to-gas, wind and solar generators, and price-based demand response, which breaks the barriers of traditional carbon emission trading (CET) and green certificate trading (GCT) mechanism. To meet the demand of low-carbon emission and waste reduction, an IES containing an Energy-from-Waste (EfW) plant and a Simple Integrated System Management (SISMan) decision-aid model is established in [34], which has significant reference value for the realization of “waste-free cities.” In the future, the development and utilization of clean energy will show the trend of combining centralized and decentralized. Therefore, it is necessary to promote the new mode and new business development of decentralization and support the large-scale development and utilization of distributed energy. Ref. [35] establishes a novel decentralized optimal multi-energy flow (OMEF) of large-scale IESs in a carbon trading market, which is of great reference value for the decentralization of energy systems.

As for the future research direction, the current research focuses more on the operation and dispatching level of IESs modeling in the low carbon electricity market, while there is less research on the planning level considering carbon trading. It is still urgent to establish a relatively complete IES planning model based on the comprehensive consideration of the low-carbon power market.

Planning and operational model

The general structures of objective functions and constraints have been concluded in Section 3, and their detailed factors will be elaborated as follows.

The key factors and their further illustration of objective functions in the IES planning, operation and scheduling model are listed in Table 3. Existing models mainly aim at minimizing cost and carbon emissions to form a multi-objective programming problem, whose cost functions comprehensively consider investment, operation, carbon, renewable energy, demand response, and other integrated factors.

Table 3 Key factors of objective functions.

Objective functions	Illustrations
Minimize total cost	Investment costs	Annual value cost of the initial and expansion investment of the system equipment
	Operation costs	The operation and maintenance cost, fuel cost, transmission cost, and so on
	Carbon costs	The sum of carbon trading cost, carbon capture and storage (CCS) cost, and so on
	Renewable energy costs	Penalty for nonconsumption or storage of renewable energy, including green certificate trading cost
	Demand response costs	Supply-demand coordination cost, including flexible load dispatching cost, demand response compensation cost, and so on
	Integration costs	Backup cost, storage cost, employment cost, subsidy
Minimize carbon emissions		For multi-objective optimization model
Maximize the renewable energy share of the generation mix		For multi-objective optimization model

The key factors of constraints in the IES planning, operation and dispatching model are listed in Table 4. On the level of constraints, the existing models mainly consider carbon trading constraints and market-clearing model on the basis of operation constraints, such as power balance constraints and power generation constraints.

Table 4 Key factors of constraints.

Constraints	Illustrations	Formulas
Power balance constraints	Supply-demand balance	Power generation + (imported power − exported power) + (discharge − charge) ≥ power demand
Power generation constraints	Upper and lower bound constraints	Lower bound ≤ power generation ≤ installed capacity × operation hours × capacity coefficient
	Ramp rate constraints	Hourly power generation ≥ minimum generation level
		Generation (t + 1) − generation (t) ≤ maximum ramping up rate × installed capacity
		Generation (t) − generation (t + 1) ≤ maximum ramping down rate × installed capacity
	Reserve constraints	Reserve requirement = installed capacity − potential reduction − peak load
Energy network constraints	Power, thermal, and gas network constraints	Power balance constraints, storage capacity and tie lines constraints
Tie lines constraints	Power transmission constraints	Power flow ≤ transmission capacity
		New transmission capacity × installation cost unit capacity ≤ the upper limit of investment
		Exported electricity ≤ the total power generation
	Tie lines transient stability constraints	Transient stability margin ≥ Ɛ > 0
Environmental constraints	Carbon quota constraints	Generation × fuel consumption rate × (1 − elimination rate) × emission factor unit fuel consumption ≤ carbon quota
	Renewable power constraints	Renewable energy consumption rate = (renewable power capacity)/(total power capacity) ≥ target rateGreen certificate shall be purchased for the insufficient part.
Energy storage Constraints	Upper and lower bound constraints	Reserve capacity allocation ≤ energy to storage ≤ maximum storage level

From the planning level, on the basis of traditional factors affecting system planning, the carbon emission quota during the planning period is taken into account to make a comprehensive plan for the power and network structure of the system. A typical IES planning model takes the lowest cost of investment, operation, and carbon trading as its objective function to establish a multi-objective programming model considering economy and environment. In the field of coupled natural gas and electricity systems and application of power-to-gas (P2G) technology, a two-stage low-carbon operation planning model that considers a bilateral carbon trading mechanism with active demand side management is proposed in [36]. To further reduce carbon emissions, Ref. [37] verifies that electric-gas interconnected IES considering carbon capture can efficiently take into account economic and low-carbon efficiency, and meanwhile enhance the renewable energy accommodation.

From the operating level, many studies build a low-carbon operation model of IES based on the background of renewable energy. A regional energy system optimization model considering the responsibility of renewable energy consumption, green certificate trading and carbon trading mechanism is established in [38]. The International Energy Agency predicts that the power generation of renewable energy will reach 25%–41% of the total energy by 2040.³⁹ In response to the complexity, randomness and limited dispatchability that variable renewable energy sources bring to power systems, the system elasticity considering integrated flexible sources is becoming increasingly crucial. Among the technologies that improve the elasticity of IESs, dispatchable power plants can compensate system output in case of solar irradiation and low wind speed⁴⁰; transmission grids provide spatial smoothing to match the power supply and demand⁴¹; energy storage technologies can overcome the fluctuation of renewable energy to some extent⁴²; and demand-side management controls the flexible load resources to respond power fluctuation.⁴³

Existing IES models have taken into account renewable energy, carbon trading and other emission reduction technologies and mechanisms. However, in the following studies, the model strategies can still be further improved by extending the flexibility options in the four dimensions of source-grid-load-storage. The impact of the future electricity market on the power operating and dispatching process and the interaction with other energy systems within the Energy Internet should also be taken into account to cope with the complexity and dynamic uncertainty of IESs.

Model solving techniques

The models established in [44] are mixed-integer linear programming models, which means the problems that need to be solved belong to Mixed-integer Linear Programming (MILP) problems. Based on the MATLAB 2019b software and YALMIP, Gurobi 9.0 is called in [44] to calculate the above model, while the CPLEX solver is called in^45,46 based on the MATLAB platform. In the MATLAB environment, based on the YALMIP toolbox, an IES optimization operating model is established and solved using CPLEX. The reason is that lots of external optimization solvers can be called by the YALMIP toolbox, which allows people to use unified modeling and solving language. The CPLEX demonstrates the advantages of flexibility, speed, and reliability in solving linear programming problems.⁴⁶

The model established in [47] contains nonlinear terms derived from carbon trading models, which are converted into linear models using segmented linearization, and are solved by invoking the CPLEX solver via GAMS software.

Another model that is often encountered in this type of study is the Mixed-integer Nonlinear Programming (MINLP)model. The latest progress of methods to solve the MINLP problem is summarized, and these methods are classified and summarized in [48].

The proposed models in [49,50] need to solve nonlinear programming problems, which are suitable for solving by invoking the convergent and robust interior point method (IPOPT) in GAMS. In [51], the fmincon solver in MATLAB is called to solve the optimal scheduling of IES for economic and environmental benefits, a nonlinear optimization problem.

Facing the optimization problem of the hybrid system, the Benders decomposition method is used to solve the problem in [52]. The original optimization problem is decomposed into the main problem of the thermal system and the sub-problem of the power system, solved iteratively. The Benders decomposition method is widely used in combination optimization problems. The current application of the Benders decomposition method is discussed, and the scope of application is summarized in [53].

Intelligent optimization algorithms have also made significant developments. The Bacterial Community Chemistry (BCC) algorithm is an intelligent optimization algorithm inspired by biological behavior that has the advantages of global search, fast convergence, and high accuracy. It achieves good results in the power system. The energy planning problem proposed in [54] is a multicycle discrete optimization problem, which is optimized using a discrete BCC (DBCC) algorithm. The Multi-objective Optimization Bacterial Colony Chemotaxis (MOBCC) algorithm is proposed and applied to the multi-objective dispatch problem of power systems in [55].

To solve the problem of nonlinear optimization, many scholars have proposed algorithms that simulate the intelligence of biological populations, such as genetic algorithms, ant colony algorithms and particle swarm algorithms. Ref. [56] proposes that because these algorithms do not rely on the gradient information of the objective function when solving, they are particularly suitable for large-scale complex optimization problems that cannot be solved by traditional methods. However, due to the diversity and complexity of optimization problems, these swarm intelligence algorithms will still have some shortcomings in different applications. To solve the problem of the model complex in structure, containing multiple variables and nonlinear constraints, these algorithms need improvement. On this basis, combined with the simulated annealing algorithm, the above algorithms are improved in the following three aspects: fitness function, improved selection operator, crossover operator adaptive adjustment in [57].

As a result of the expanding reform of the energy market, many emerging market organizations have emerged in the energy market, which has led to fierce competition between them. Therefore, how to deal with the issue of conflicts of interest between multiple stakeholders has become an important research direction in [58]. A mathematical model of each stakeholder is built then solved by using a two-stage optimization algorithm, that is, using the Adaptive Differential Evolution (ADE) algorithm combined with the GUROBI toolbox in [58]. To prevent the leakage of information and protect the privacy of all parties in the IES, an iterative solution algorithm is proposed to solve the problems of low-carbon economic scheduling and energy sharing in a decentralized manner in [59]. The algorithm describes an iterative dynamic process between interactive price setting (performed by ESP) and energy collaborative sharing (performed by IES). The model is solved by using the Differential Evolution (DE) algorithm and the Optimization Toolbox.

A review of the planning and operation models of IESs is shown in Table 5. It can be concluded that renewable energy, carbon capture, carbon trading and other emission reduction technologies and mechanisms have been considered in the planning and operation models of IESs. The research on the interconnection and interaction of multiple IESs will further strengthen the robustness and reliability of the IESs.

Table 5 The review of planning and operating models of integrated energy systems.

Ref.	Time	Classification	Scenarios	Objective function	Constraints	Solution method
[54]	2019	Planning	Electric–gas-integrated energy system planning model that considers carbon trading	Investment cost + operation cost + carbon trading cost	Power generation, gas network, tie lines, and power balance constraints	Bacterial colony chemotaxis algorithm
[57]	2021	Planning	Electric–gas interconnected integrated energy system considering carbon trading, carbon capture and P2G	Investment cost + operating cost + carbon cost (carbon trading, carbon capture)	Energy conservation, energy network, and unit output constraint (P2G)	Swarm intelligence algorithm combined with simulated annealing algorithm
[50]	2017	Operation	Integrated electricity and natural-gas energy systems considering peak load shifting	Operating cost + demand response cost (considering peak shaving and valley filling target)	Power balance, power generation, tie lines, and electricity-gas network constraints	IPOPT solver based on the interior point method (IPOPT) in GAMS
[52]	2018	Operation	Low-carbon economic dispatch of integrated electrical and heating systems considering wind power accommodation	System economic cost + system environmental cost	Power balance, power generation, tie lines, and storage capacity constraints	Benders decomposition
[60]	2019	Operation	Probabilistic models of uncertain multiple loads of electricity, heat and cold	operation cost + carbon trading cost	Power balance, power generation, tie lines, and energy storage constraints	Mixed integer linear programming model solved on Gurobi platform
[46]	2021	Operation	Electricity and natural gas energy system considering carbon capture device	Operating costs + carbon trading costs	Power balance, energy network, power generation, and tie lines constraints	Mixed-integer Linear Programming(MILP)
[51]	2021	Operation	Considering renewable penetration, carbon capture, and pollutant reduction	Operating costs + the renewable curtailment penalty cost + pollutant treatment cost	Power balance, power generation, pollutant emission, and energy storage system constraints	Fminconsolver in MATLAB
[59]	2021	Operation	A new model of carbon quota allocation and trading considering carbon balance and carbon sharing among multiple systems; Multiple integrated energy systems; Energy service provider	operation cost + carbon trading cost	Carbon quota constraint; Energy balance; Operating conditions of equipment	Bi-level optimal framework; Differential evolution algorithm; Stackelberg Game Approach

Simulation technologies of the IESs

For model simulation and solution, many scholars use the YALMIP toolbox or GAMS in MATLAB to model and call the solver to solve.^44–46 MATLAB software is used to generate scenarios and SCENRED GAMS software is used to reduce scenarios. Taking into account the uncertainty of load forecasting, the LHS method is used to generate a large number of scenes for electricity, heat loads, and cooling loads, and the K-means method is used to reduce scene numbering in [60].

Ref. [61] studies in detail the economic dependence of the US IES. The network flow modeling simulation technique is applied. It realizes the simulation time and step size of different subsystems, and the solution efficiency is high, which is of great significance in describing the high-dimensional characteristics of the IES.

The collaborative model of multi-agents

Two important features of an IES are multi-energy coupling complementarity and multiagent game.⁵⁹ Therefore, the collaborative programming model between multiple agents is an important research direction.

Ref. [54] studies electrical synergy systems. Carbon trading cost models are improved. A planning model is established for electrically IESs that considers carbon trading. Ref. [46] also studies electrical synergy systems, but the difference is that in this study carbon capture is considered, hoping to reduce carbon from the power supply side, while introducing carbon trading. An integrated electrical system (IEGS) that considers carbon capture devices is proposed.

The multi-energy interconnection system formed by the coupling of natural gas and electricity has attracted a lot of attention because of its sustainability and efficiency. However, the operation of the natural gas system and the operation of the power system belong to two different time scales, and the coupling system operation must consider the influence of this factor. The dynamic power flow equation of the natural gas system is derived in [62]. The bilateral energy conversion relationship of the gas-electric combined system is considered, and a multi-period optimization operation model of the coupling system is established.

Ref. [59] focuses on low-carbon economy dispatch and energy sharing approaches for multi-IESs (IES). An optimized model is proposed based on Stackelberg games for Energy Service Providers (ESP) and IES Group. Aiming at the multi-energy optimal power flow problem, a distributed optimization method based on multiagent architecture is proposed, and it is solved by a genetic algorithm in [63].

To achieve the optimal allocation of energy equipment capacity, Ref. [64] optimizes the overall optimization design of the integrated energy microgrid including the CHP generation unit, and verifies the model effectiveness by comparing the studies of the two operating modes of thermoelectric power and the electric heat.

SUMMARY AND PROSPECT

The sustainable development and low-carbon transformation of energy systems is an important research direction of energy conservation and emission reduction. Based on existing research, it can be concluded that current planning and operation models have been greatly improved to achieve an overall optimization. Renewable energy, carbon trading, and other emission reduction technologies and mechanisms play crucial roles in the operation process of IESs.

Based on a brief review of the general structure of the IES, this paper briefly introduces some new issues faced by low-carbon planning and operation according to the technical characteristics of the new power systems. As mentioned, minimizing the total cost is the common objective function shared by traditional and IESs. To take both economy and environment into account, carbon emission or carbon trading prices are converted into corresponding factors, variables, or constraints, and incorporated into the planning and operation model of the IES. To cope with the increasing fluctuations and uncertainty on the source and load side, the technologies for precise prediction of energy supply and demand have become the basis of integrated power system planning and modeling.

Existing IES models have taken into account renewable energy, carbon trading and other emission reduction technologies and mechanisms. However, they need to be further improved by extending the flexible options in the four dimensions of source-grid-load-storage. Furthermore, the optimization model and carbon emission reduction technologies discussed in this study have not been widely applied in the form of a complete and comprehensive system due to the economic factors and lack of available data. Thus, more practice regarding the low-carbon IESs should be conducted to implement the research achievement and realize the zero-carbon planning and operation of IESs.

To improve the operation process, the interaction mechanism of heterogeneous energy flow under different temporal-spatial scales needs to be further studied in the future. The main influencing factors of carbon emission should be further identified with the dynamic correlation of energy type and operation characteristics, and the applicability of multi-energy complementary technology in different low-carbon scenarios needs to be further analyzed in combination with the operation characteristics of each energy flow and low-carbon electricity market theory.

In future studies, the influence mechanism model of the carbon market and other energy markets should be investigated, due to the impact of carbon trading on the operating cost of various energy units and the competitive pattern of the energy market such as the electricity market, the natural gas market, the coal market, and the green certificate market. Moreover, a diversified carbon trading mechanism with respect to the regional energy market should be established to realize the mutual coordination of carbon resources, and fully utilize the market and energy allocation ability of carbon trading.

The IES is of great significance to promote the sustainable transformation of the global energy structure and help to achieve the goal of energy conservation and emission reduction. However, the current research on low-carbon planning and operation of the new-generation energy system is still in its infancy and needs further improvement and development.

ACKNOWLEDGMENT

This study is supported by the Zhejiang University Research Funding (No. 110000-521542).