Risk-constrained non-probabilistic scheduling of coordinated power-to-gas conversion facility and natural gas storage in power and gas based energy systems

Abstract

Increasing environmental concerns related to fossil fuels has led to an increased desire to utilize renewable energy in the power system. Wind power resources as clean energy have an important role in the long-term energy landscape. However, the probabilistic nature of these resources will make several challenges from the integration point of view with the modern power system. Power-to-gas (P2G) technology as an emerging storage system by converting the power generated from the wind power into natural gas (NG) plays a significant role in integrating more and more wind power into the power system. Besides, the NG-fired (NGF) units have received much attention due to faster response, higher efficiency, and far lower pollution emissions. The emergence of P2G and NGF units has led to the interdependency of NG and electrical networks. This interconnection requires the simultaneous optimization and operation of both networks considering all relevant restrictions. Therefore, this paper presents a developed security-constrained unit commitment (SCUC) problem for integrated power and NG networks in the presence of P2G, NG storage, and wind power resources. The main objective of the proposed problem is to minimize the operation cost by considering all the characteristics and interconnections between the power and NG networks. To handle the uncertainty related to wind power, the information gap decision theory (IGDT)-based non-probabilistic robust approach is applied. The proposed model is implemented on the integrated 6-bus electrical and 6-node NG networks and integrated 24-bus electrical and 10-node NG networks for different cases and numerical results reveal the effectiveness of the proposed approach.

Introduction

The growth of concerns about the financial and environmental effects of fossil-fueled power generation units has led to increasing the utilization of renewable energy sources (RES) specifically wind energy. Based on International Energy Agency (IEA) report, the annual wind power generation will be increased to 2187 TWh by the end of 2030 compared to 2009. Furthermore, the natural gas-fired (NGF) units with attribution to higher ramp-rate, and fast start-up capability less than a minute, lower gas prices, and lower carbon emission up to %60 as compared with coal fuel-based power plants have been increased sharply in the last decade [1], [2]. In the U.S.A, NG utilization specialized to electricity generation has been increased from %34 by 2011 to %39 in 2012 [2], [3]. Considering the mentioned advantages makes the NGF units more suitable to address the uncertainty of wind power and facilitate the more integration of RES into the energy system.

To compensate for the challenges of wind integration into the power system, different energy storage technologies such as hydro-pump storage, compressed air [4], batteries [5], and electric vehicles [6], thermal storage [7], and hydrogen storage [8] have been implemented by researchers. Energy storage systems (ESS) are implemented in the power system for different applications [9]. Recently, the power to gas (P2G) technology regarded as a variable energy storage method is used to accommodate the RES variation in the power system operation [10], [11].

The P2G facility, unlike the conventional energy storage system (ESS), converts surplus electrical energy (especially from renewable energy) into synthetic NG and later uses the NG network as a high potential system to transport, store and use the electricity later time.

The overall efficiency of the energy conversion for the P2G technology generally is about 50–60 percent. In comparison with other conversion processes in word, the P2G facility still has a high efficiency which provides multiple advantages especially an ancillary service market and the arbitrage opportunities. However, the increase of utilization of NGF units and the potential of P2G conversion technology intensifies the interdependency of both electricity NG networks. Therefore, both of them (NG and electrical networks) face new challenges. From the NG network view of point, NGF unit behaviors like end-user. However, the local demand including residential and commercial consumers has a higher priority concerning NGF units. Therefore, any change in the NG consumption by these loads especially in winter times can affect the fuel consumption of the NGF units and consequently their commitment as well as power generation. Furthermore, considering the fast start-up capability of NGF units, in turn, the NG system should be capable to provide more flexibility to guarantee the sustainable operation of NGF units with volatile gas consumption.

In [12], a review on economic, environmental, and technical perspectives for optimal utilization of the P2G storage is presented. An overview of different P2G network projects in Europe is studied in [13]. In this paper, the results of the effectiveness of projects through hydrogen injection on the pipeline of the NG network have been investigated. In [14], the uncertainty interval model is presented to maximize the utilization of wind farms, and also minimize the prediction error by relying on the unique capability of P2G storage. The proposed approach is formulated as a mixed-integer linear programming (MILP) model to minimize the operational cost. The authors in [15], investigated the new design for the integrated energy system by utilization of P2G technologies to enhance the reliability in extreme situations. Simulation results show that the presented model makes the system resistant to various uncertain parameters. In [16], a two-stage adjustable robust model is proposed to evaluate the performance of the interdependent power and gas systems in the presence of the P2G storages. The P2G storage is used to convert the excess electricity into natural gas.

The authors in [17] and [18] evaluate the dependency of electricity and NG networks considering the deterministic model of renewable energy sources. In [18], [19], [20] some algorithms including augmented Lagrange relaxation (LR), alternating direction method of multipliers (ADMM), and Benders decomposition were used to relax the coupling constraints into electricity and NG sub-problems and day-ahead scheduling of integrated NG and electrical networks. The NG network constraints are incorporated in the stochastic security-constrained unit commitment (SCUC) problem to enhance the impact of the NGF unit’s participation in the day-ahead energy market [21]. In [22], the effects of price-based DR on the dependency reduction of electric network and NG, as well as the reduction of system costs is studied by solving the SCUC problem of integrated NG and electric networks. In this research, transmission line outage and load prediction are considered as uncertainty parameters. The authors in [23] and [24] determine the impact of the NG network on electric power system operation solving a stochastic-based SCUC problem in the day-ahead energy market with integrated wind energy. In [25], two-stage robust scheduling of integrated power and NG systems considering the NG storage system has been investigated. The effect of P2G technology on the coordinated of electrical and NG networks in Great Britain has been studied in [26] deterministically and without considering complete characteristic parameters of both networks. In [27], the authors focus on the robust scheduling of coordinated electricity and NG networks with integrated wind energy and power to gas technology. A two-stage SCUC problem has been solved in [28] to investigate the impact of NG delivery and price fluctuating on the participation of NGF units in the day-ahead market. Authors in [29], proposed a robust flexible SCUC for reserve and energy generation of integrated NG and electrical networks in the presence of P2G storage. Authors in [30], develop an optimal design for NG and distributed energy in large/medium scale industrial parks in the NG deficiency conditions. The proposed problem was formulated as a multi-objective problem while both operation and investment costs through a chance-constrained approach. The unit commitment problem of integrated NG and electrical networks in the presence of P2G technology is presented as a multi-objective two-stage model in [31]. The main purpose of the proposed multi-objective problem is to minimize the operation cost as well as emission costs through the epsilon-constraint method.

According to the literature review, the stochastic and robust scheduling approaches have been widely implemented to handle the system uncertainties in coordinated power and NG networks. In such a problem, where the decision-makers are faced with high-level uncertainty in the input data, the membership function of probability density function (PDF) may not be available. Therefore, due to the lack of adequate data and information, the information-gap decision theory (IGDT) method can address the uncertainty as well [32]. Unlike the Mont Carlo and scenario-based stochastic scheduling, the IGDT method does not need to define the probability density function (PDF) for uncertain parameters. This method is suitable for robust decision making against severe uncertainties. Also, unlike the robust optimization (RO) algorithm, IGDT is more flexible since there is no need to know the maximum uncertainty region for uncertain parameters [33]. This technique has to determine the uncertain parameter radius in such a way that the objective function varies in the prescribed region. However, IGDT methodology mostly is used in multiple power system problems like bidding strategy problems [34], unit commitment [35], generation asset allocation [36], transmission planning [37], and restoration of distribution networks [38]. Ref. [39] applies the IGDT method to solve a self-scheduling problem to maximize the profit of GenCos. In [40], the IGDT-based robust SCUC problem has been solved with the presence of wind energy, flexible DR resources, and ESS. In [37], the authors use IGDT in a UC problem to model the uncertainty of wind power output. In [41], optimal scheduling of modern power systems such as microgrids with multi-carrier loads was investigated based on the chance-constrained programming approach. Also, the IGDT-based robust SCUC is investigated in [5] considering Lithium-ion battery storage units. An optimal bidding strategy for the intelligent parking lot of EVs using the IGDT-based risk-involved approach considering power price uncertainty is investigated by [42].

The previous studies demonstrate the importance of optimal coordinated scheduling of NG and electrical distribution systems in the presence of emerging technologies like P2G facility using the new realistic approach to handle the existing uncertainties. Therefore, this paper proposes the extended SCUC problem for the coordinated NG and electrical networks considering the P2G and NG storage technologies. In the SCUC problem, in addition to the electrical network constraints, all the restrictions of the NG system are taken into account in detail. The main objective of the problem to minimize the operation cost using the IGDT-based robust approach considering wind power generation uncertainty. To model the transmission line’s flow, the B-theta based equations [43] and DC power flow model is applied. The main contribution of the proposed in this paper can be summarized as follows:

• Risk-constrained co-optimization scheduling of electricity and NG networks without the need to probabilistic density function (PDF)

• Considering the P2G technology and NG storage system to increase the operational flexibility of electricity and NG systems with integrated wind energy.

• Using the IGDT-based method to manage the uncertainty of wind energy.

• Considering the interdependency constraint between electrical and NG networks to achieve the realistic model.

The rest of this paper is organized as follows: The P2G concept is described in detail in Section 2. The formulation of the coordinated electrical and NG networks SCUC problem is developed in Section 3. Also, the formulation of P2G and NG storage technologies as well as NGF units represent linkage between electric and NG networks are proposed in this section. Section 4 describes the uncertainty modeling based on the IGDT approach. Simulation and numerical results are presented in Section 5. Finally, Section 6 concludes the paper and drawn future works.