Elsevier

Applied Energy

Volume 231, 1 December 2018, Pages 803-815
Applied Energy

Cost-efficient multi-energy management with flexible complementarity strategy for energy internet

https://doi.org/10.1016/j.apenergy.2018.09.152Get rights and content

Highlights

  • A multi-energy management method with energy complementarity prosumers was proposed.

  • A multi-objective optimisation framework was developed to reduce total energy cost.

  • The economic impact of the energy complementarity strategy was investigated.

  • The new method could promote the sustainable development of the urban energy system.

Abstract

The increasing complexities of energy internet integrated with distributed renewable energy resources and multiple energy infrastructures require more effective multi-energy management method. The prosumers with multiparty interaction represent major potential contributors for comprehensively improving the energy efficiency and socioeconomic benefits. In this paper, a novel multi-energy management strategy based on the complementarity of multi-energy demand was proposed to explore optimal energy scheduling problems of prosumers. The residential prosumer with a multi-energy coupling matrix and the industrial prosumer with a resource-task network were formulated to optimise the local operations. Furthermore, a joint planning for the prosumers was developed to minimise the global operating costs, where the prosumers’ interests in terms of the energy exchange process were formulated as a multi-objective optimisation problem based on the Pareto efficiency theory. In addition, an optimisation method that integrates the epsilon-constraint algorithm and the extreme points of the feasible solution space was proposed to obtain better and more diverse solutions. The proposed methodology was applied to an urban multi-energy system. Simulation results demonstrated that the proposed multi-energy management method could effectively solve the optimal energy scheduling problems. At the compromise solution point, cost reductions of 7% and 10% can be obtained by the two prosumers on a summer day, with cost reductions of 9% and 11% obtained on a winter day. The use of multi-energy management method could establish a win-win relationship for prosumers and generate substantial benefits for the whole system.

Introduction

The rapid growth of the energy demand and limited resources have led to serious global concerns about the depletion of energy resources [1]. Related research has indicated that fossil fuel combustion for energy production is a major contributor to air pollution and the greenhouse effect [2]. The literature [3] also shows that a particularly large fraction (35–36%) was from fuel combustion in power plants. The problems call for collaborative and multidisciplinary research on energy sustainability [4], [5]. As a promising next-generation power system, the smart grid (SG) has been presented, which involves the construction of an intelligent power delivery system by enabling bidirectional flows of electricity and information [6]. However, in practice, an energy system contains a variety of complex multi-energy carriers, and energy can be generated, converted, transmitted, and consumed in flexible ways. Moreover, the SG still has limited flexibility for distributed supply and demand because of its reliance on the traditional grid infrastructure.

As a feasible solution, the energy internet (EI) [7], was proposed to obtain a highly economical, efficient, flexible, and sustainable future energy system. It integrates multi-energy carriers, including renewable energy, and promotes the deep integration of the energy flow, information flow, and business flow [8]. Compared with the SG, the EI has a wide range of advantages in the following areas. (1) The individualised utilisation, optimal allocation, and targeted management of multiple energy sources are emphasized in the EI [9], while the SG only focuses on the optimal supply of a single form of energy based on consumption. (2) The SG pays more attention to the inheritance and transformation of the traditional power grid, with passive user access and centralised control primarily adopted [10]. In contrast, the EI can support energy plug-and-play during the processes of production, storage, and consumption, with the operational topology no longer limited to a specific structure [11]. In addition, decision-making units have the advantage of flexible distributed control nodes for autonomous energy deployment, which also greatly enhances the security, flexibility, and sustainability of the EI system [12], [13]. (3) A wide range of self-energy bodies or micro-power stations become the main body when using the EI, which is bound to spawn a more economical business model [14].

Despite the previously mentioned advantages, the advent of multi-energy sources and their complex coupling relationship pose great challenge to the EI. The open-ended introduction of diversified energy sources and their individualised utilisation represent a significant new supply-demand balance requirement that should be satisfied. Thus, it is especially important to clarify the complex coupling relationship between multi-energy sources. In this context, researchers solve the above issues by multi-energy management (MEM). A basic architecture for a MEM system is shown in Fig. 1. Two typical devices are engaged in the system based on the emerging information and control technology. The energy router [15], [16], as the information and control centre of MEM, can immediately communicate with other devices by using a real-time energy data acquisition and information exchange system to support the decision system of energy deployment or scheduling. The energy hub [17], [18] is defined as an input-output port model that describes the coupling in MEM, which connects the primary energy sources and energy consuming terminals to realise the indirect conversion among multi-energy carriers. For instance, natural gas-powered combined cooling, heating, and power (CCHP) technology is believed to be a key component in an energy hub, which increases the energy utilisation by capturing by-products for heating and cooling during the electricity generation [19]. Some analyses of the management, control, system optimisation, and sizing of a CCHP system were summarised in [20].

Based on the above architecture, many researchers have developed and proposed a large range of methodologies for the MEM system, which can be classified into the following categories: (1) Optimisation approaches for MEM problems, (2) The reconfiguration of MEM based on bidirectional multi-energy flow, and (3) The flexibility of energy complementarity (EC). There are various optimisation approaches for MEM problems. In [21], a MEM system with a distributed approach was used to solve the optimal energy management problem. Compared with a centralised approach, a distributed approach has better robustness, faster computation, and less communication. A probabilistic optimisation method to minimise the customer’s energy cost using an energy hub associated with the uncertain output power of solar panels in residential MEM was researched in [22]. A more complex probabilistic model verification method was developed in [23] to check the reliability and quantitative properties of energy routers. However, considering the various practical constraints and the complex coupling relationship of multi-energy sources, it is difficult to identify an accurate probability distribution for some uncertain factors. Accordingly, some robust optimisation approaches [24] have been used to allow a distribution-free model of uncertainties, which provide more convenience in practical MEM systems. Other approaches are committed to achieving coordinated optimal management that takes into account the user’s requirements [25]. An important advantage of these approaches is that they are easy to implement and cost-effective solutions.

However, because most of the energy infrastructures in MEM have limited capacity allocation and self-coordination capabilities, it is still difficult to achieve the intended purpose with the above optimisation method alone. In order to solve this problem, some researchers are committed to reconstructing MEM based on the bidirectional multi-energy flow structures. Accordingly, a more active prosumer (producer-consumer) has been proposed as an energy user who generates energy and shares the surplus with the energy buyers [26], [27]. Its characteristics are summarised in the integration of the production, consumption, purchase, and sale. Many authors have used this useful configuration. A district heating network based on solar prosumers was presented in [28] to analyse its impact on the technical parameters of distribution networks. In addition, a load shifting control for residential electricity prosumers in consideration of both designed and market indexed pricing models was developed in [29]. As a result, the formulation of prosumers implements a preliminary paradigm of distributed energy transactions.

Motivated by the above optimisation approaches and reconfiguration of MEM, it is gradually recognised that the flexibility of the EC behaviours of various energy infrastructures can be further stimulated. Therefore, a new concept of social energy was proposed in [30], which combines the inherent properties of energy, namely physicality, sociality, and informatisation. In addition, a general methodology based on parallel systems was presented to achieve the flexible management of multi-energy systems. In [31], a more complex system of interconnected infrastructures in consideration of multiple spatiotemporal scales was exemplified. In a MEM system, the integration of multi-energy areas, especially energy production areas, industrial areas, and residential areas, can further unlock the flexibility of multi-energy supply, thereby highlighting the potential socioeconomic benefits of such integration.

To obtain better economic benefits in some industrial areas, it is possible to further optimise their energy consumption patterns and explore their potential energy production capacity. On the one hand, industrial controllable load management can provide additional flexibility for electricity cost minimisation [32]. For instance, the consumption activities can be shifted to a lower price period. In [33], the potential of demand response (DR) in steel manufacturing was further discussed considering the time-based energy price, and a resource-task network (RTN) model was proposed to describe the complex manufacturing processes with multi-batch, multi-stage, and critical process-related constraints in a systematic way. It significantly enhanced the energy efficiency of the steel plant. On the other hand, a variety of waste heat recovery (WHR) technologies have already been put into practice, which are generally classified into two categories: (1) the waste heat is prevented from escaping using a steel thermal shield and (2) the waste heat is progressively absorbed by the water flow in the pipeline zone [34]. A thermal shield can effectively reduce the heat loss of industrial processes, while the water flow in the pipeline zone makes it possible for industrial waste heat to be reused in other areas.

Therefore, a flexible EC strategy for multi-energy areas can be incorporated into the reconfiguration of MEM. However, it is formulated as a relatively innovative but difficult optimisation problem, which brings many challenges. First, the modelling needs to consider the influence of hybrid factors such as the different production/consumption characteristics of multiple infrastructures, coupling relationships of multi-energy resources, and large number of complex practical constraints. Second, it is necessary to reconsider how to embed a flexible EC strategy into a mutual contradiction or coordination MEM system based on multiple local expectations and global social welfare. Third, a trade-off decision for multi-resource allocation and socioeconomic benefits is extremely important for the whole system.

In this paper, a novel paradigm of MEM that combines a flexible complementarity strategy between multiple energy prosumers was proposed to solve these problems. In this method, the multi-energy coupling matrix based residential prosumer and the resource-task network based industrial prosumer are formulated to optimise the local operations. In addition, a joint planning for the prosumers is developed to minimise the global operating costs, where the prosumers’ interests in terms of the energy exchange process are formulated as a multi-objective optimisation problem based on the Pareto efficiency theory. Furthermore, an innovative optimisation method that integrates the epsilon-constraint algorithm and the extreme points of the feasible solution space is proposed to obtain better and more diverse solutions. The methodology was applied to an urban multi-energy system. Results showed the effectiveness of the proposed optimisation method in solving the multi-objective optimisation problem. Results also showed great benefits of the MEM in energy cost reduction, supply-demand balancing, and socioeconomic benefits improvement.

The rest of this paper is organised as follows. First, the detailed configuration and operation of the proposed MEM are discussed in Section 2. Afterwards, Section 3 presents the self-scheduling formulation, including a detailed arrangement for constraints and a discussion on the objective functions of the local optimisation. As an extension of the self-scheduling model, co-scheduling that incorporates the EC between prosumers is proposed in Section 4. Moreover, the epsilon-constraint algorithm is introduced to solve the multi-objective optimisation problem. Section 5 presents a numerical analysis of the proposed method, and the economical contribution of the EC is further discussed. Finally, Section 6 highlights the contributions and conclusions of this work.

Section snippets

Proposed framework for MEM

Fig. 2 shows the proposed framework for MEM, which considers the integrated management of residential and industrial prosumers. Each prosumer plays four major roles: producer, consumer, buyer, and seller. The residential prosumer is described as an energy user who generates energy that is either renewable or produced by a CCHP system and shares the surplus with energy buyers such as the utility grid and other users. The industrial prosumer is described as an energy user who produces industrial

Self-scheduling formulation

In general, both energy bodies pay attention to the reduction in energy costs. For the residential prosumer, the key to the optimisation problem lies in when and how much of which kind of energy should be purchased, converted, or stored to maintain the supply-demand balance and minimise the local energy cost. For the industrial prosumer, the electricity cost can be optimised by shifting the time of consumption activities.

Co-scheduling formulation

In this section, we further construct the co-scheduling for MEM. A novel EC strategy between two prosumers is presented based on the buyer/seller features. Compared with Section 3, the energy costs here are further reduced by converting traditional energy consumers to prosumers.

Case study

In this section, the proposed MEM model is applied to an urban area with 100 residential districts and a steel plant. A case study that includes a comparison of self-scheduling and co-scheduling in two typical seasons is carried out to demonstrate the economic effectiveness of the method. We consider the daily scheduling problem for the whole MEM system. Based on the mixed integer programming, all of the models are carried out in GAMS and solved by the CPLEX solver.

Conclusion

Under the circumstances of the energy internet, an innovative framework for a multi-energy management system that incorporates a flexible energy complementarity strategy was proposed to improve the operation of multiple energy prosumers. Energy cost reduction, supply-demand balancing, and socioeconomic benefits improvement were taken into consideration. The residential prosumer was formulated using a multi-energy coupling matrix to describe the detailed process of energy distribution and

Acknowledgements

This work was supported by the National Key Research and Development Program of China (2016YFB0901900), National Natural Science Foundation of China (61374097, 61603083), and Natural Science Foundation of Hebei Province (F2017501107, F2017501014).

References (39)

  • M. Yu et al.

    A real-time decision model for industrial load management in a smart grid

    Appl Energy

    (2016)
  • M. Guarnieri

    A lesson from past energy crises [historical]

    IEEE Ind Electron Mag

    (2016)
  • Q. Zhang et al.

    Policy: cleaning China’s air

    Nature

    (2012)
  • L. Steg et al.

    What drives energy consumers?: Engaging people in a sustainable energy transition

    IEEE Power Energy Mag

    (2018)
  • M.H. Cintuglu et al.

    A survey on smart grid cyber-physical system testbeds

    IEEE Commun Surv Tutorials

    (2017)
  • J. Rifkin

    The third industrial revolution: how lateral power is transforming energy, the economy, and the world

    (2011)
  • A.Q. Huang et al.

    The future renewable electric energy delivery and management (FREEDM) system: the energy Internet

    Proc IEEE

    (2011)
  • Q. Sun et al.

    A multiagent-based consensus algorithm for distributed coordinated control of distributed generators in the energy internet

    IEEE Trans Smart Grid

    (2015)
  • J. Chen et al.

    A game-theoretic framework for resilient and distributed generation control of renewable energies in microgrids

    IEEE Trans Smart Grid

    (2017)
  • Cited by (57)

    View all citing articles on Scopus
    View full text