A robust decision framework for strategic behaviour of integrated energy service provider with embedded natural gas and power systems in day‐ahead wholesale market

Abstract

The rampant growth in penetration of gas-fired distributed units (GFDUs) is recasting the role of the integrated energy service provider (IESP) as a price-maker player. Nevertheless, it is threatening the well-being and reliability of the natural gas networks (NGNs). In this regard, this paper investigates the strategic participation of integrated 20-node NGN and an 33-bus active distribution system (ADS) in the day-ahead wholesale market (WM), which consists of a six-bus transmission network (TN), through a bi-level optimization framework. The upper-level objective is to minimize the operational costs of the IESP as well as the cost of energy that is purchased/sold in day-ahead WM. The lower-level problem models the optimal response of the wholesale market operator (WMO) that has the objective of maximizing social well-fare by clearing the day-ahead wholesale market. To alleviate the NGN's pipeline congestion, the linepack capability of the pipelines is deployed, while demand response program (DRP) and interruptible loads (IL) are integrated into the ADS side. As a computationally efficient method, robust optimization (RO) is utilized to handle the uncertainties. The results emphasize the necessity of integrating NGN into this problem. Eventually, different case studies illustrate the impact of line pack and DRP in reducing market-clearing price.

A tactical scheduling framework for wind farm‐integrated multi‐energy systems to take part in natural gas and wholesale electricity markets as a price setter

“…The bi‐level problem formulation for the upper and lower levels is established in this section. This problem is converted into a single‐level model by KKT conditions and the theory of strong duality as depicted in [27].…”

“…The first step consists of optimal dispatching of units in MES, and at the second step, the NGM and WEM are scheduled in accordance to the MES decisions. Moreover, a bi-level framework has been studied in [26] to model the trades between retail multi-energy markets and multi-energy consumers, where retail markets and multi-energy consumers form the upper and lower levels A bi-level optimization framework with single follower was developed by [27], where the leader problem is the profit maximization of integrated power and gas energy provider. The study integrated the WEM as the lower level follower Because of their high storage capacity and low capital investment requirements, the CAES units have been the cornerstone of ancillary service markets [28].…”

“…The charge, discharge and simple cycle operation modes of CAES is established through Equations ( 24)- (26). Furthermore, the air cannot be compressed in and decompressed from CAES simultaneously, which is defined by Equation (27). Eventually, the natural gas demand of CAES during discharge and simple cycle operational states is calculated by Equation (28).…”

“…Modelling the limitations and mechanism of the natural gas network (NGN) and transmission system (TS) is vitally important in calculation of MCP. Disregarding their models renders these studies incompatible with practical circumstances. DS3 : Some of these studies (i.e., [9–13, 15, 16, 19, 21–27]) have not evaluated the CAES's presence in the tactical scheduling of MES as a price setter in various energy markets. A CAES can be operated together with a TES and NGS to enhance the role of MES in markets and manage the uncertainties derived from volatile wind power. DS4 : The studies [9–18, 21–29 31] have proposed RO and SP methods.…”

“…Disregarding their models renders these studies incompatible with practical circumstances. DS3 : Some of these studies (i.e., [9–13, 15, 16, 19, 21–27]) have not evaluated the CAES's presence in the tactical scheduling of MES as a price setter in various energy markets. A CAES can be operated together with a TES and NGS to enhance the role of MES in markets and manage the uncertainties derived from volatile wind power. DS4 : The studies [9–18, 21–29 31] have proposed RO and SP methods. However, the SP method requires complicated risk measures, and RO can only provide risk averse solutions, while in real‐world applications, the MES operator should have a choice between risk‐averse or risk seeker frameworks.…”