Transactive energy management and system design for multi-energy networks

• Main Author: Zou, Yunyang
• Other Authors: Hung Dinh Nguyen
• Format: Thesis-Doctor of Philosophy
• Language: English
• Published: Nanyang Technological University 2024
• Subjects: Engineering; Energy economics; Multi-energy systems
• Online Access: https://hdl.handle.net/10356/174618
• Institution: Nanyang Technological University

A multi-energy system (MES), also referred to as a multi-energy network or an integrated energy system, has gained widespread recognition as a cost-efficient, environmentally-friendly, and reliable energy supply paradigm over recent decades. The classical energy management methods for MESs mainly focus on supply-side management to match energy demands. However, with the increasing integration of distributed energy resources (DERs) such as photovoltaic (PV), wind, and flexible loads, the intermittent nature of renewable generation and the uncertain behaviors of flexible consumers/prosumers have posed formidable challenges to the operation and management of MESs. To tackle these new challenges, the emerging concept of transactive energy offers promising solutions. Therefore, this thesis is devoted to the design of transactive energy system (TES), also known as transactive energy management, for MESs. This thesis first summarizes and analyzes the designs of state-of-the-art TESs. Subsequently, a series of new transactive energy management methods, including both centralized and peer-to-peer (P2P) methods, are developed to meet diverse application scenarios. Finally, the ancillary service markets necessary for supporting P2P transactive energy management are also designed. Specifically, to fulfill the urgent research demand and offer a timely guideline for future TES designs, this thesis first conducts a comprehensive review of existing TES research works and provides a detailed classification from various perspectives, which include participating objects, market structure, trading commodity, clearing method, and solution algorithm. The associated principles and fundamental mathematics are elucidated, followed by discussions on the advantages and disadvantages of the different TES designs. Furthermore, two additional market tools, i.e., penalty mechanism and loss allocation mechanism, are also discussed for ensuring the feasibility and fairness of energy trading in TESs. This segment of the research can serve as a timely reference and guideline for future TES research and design. Next, to leverage the consumers’ flexibility in accommodating uncertain renewable generation locally, while considering the multi-energy microgrid (MEMG) operator’s risk aversion towards uncertainties, this thesis proposes a risk-averse transactive energy management method for a MEMG. Based on the transactive energy concept, the problem is formulated as a Stackelberg game-theoretic bi-level optimization model. The MEMG operator optimizes the energy scheduling and pricing strategies at the upper level, and the industrial, commercial and residential agents optimize their energy trading strategies at the lower level. The equivalent single-level mixed-integer linear program (MILP) reformulation is then derived for computational tractability. To coordinate the strategies made in the day-ahead and intra-day energy markets, an adaptive stochastic optimization approach is employed, by which a day-ahead stochastic MILP and an intra-day deterministic model are formed. A conditional value-at-risk (CVaR) measure is incorporated in the day-ahead stage for formulating the MEMG operator’s risk aversion towards uncertainties. To solve the models, an adaptive Progressive Hedging (PH) algorithm is developed to decompose the day-ahead stochastic MILP into multiple scenario-based subproblems which are then solved in parallel; and an outer approximation (OA) algorithm is adopted to address the intra-day bilinear problem. The simulation results confirm both the effectiveness of the proposed transactive energy management method and the efficiency of the adaptive PH and OA algorithms. Then, to prevent P2P trading from exacerbating network operations and ensure network-constraint-feasible P2P trading actions, this thesis proposes a bi-level P2P multi-energy trading framework for a coupled distribution network (DN) and district heating network (DHN). At the lower level, each nodal agent represents its intra-nodal prosumers to optimize the local energy scheduling and P2P trading strategies based on a modified Nash bargaining theory, and a distributed algorithm is then adopted to enable individual agents to make their strategies autonomously only with the sharing of trading information. Once the lower-level P2P bargaining is settled, each agent is required to submit its nodal net loads and trading adjustment tolerances to the network operators. At the upper level, the network operators minimize the line losses while satisfying network operation constraints by reconfiguring the DN and DHN as well as enforcing necessary trading adjustments from the lower-level agents when the network violations incurred by the P2P trading cannot be fully solved by network reconfiguration. Mathematically, the DN operation is modelled based on the linearized DistFlow with a set of new radiality constraints to sufficiently ensure the radial structure of the DN. The DHN operation is formulated as a quasi-linear thermal flow model independent of mass flow rate and water temperature, by which the computation complexity and limitations associated with traditional DHN formulations are addressed. The proposed framework is validated on three reconfigurable multi-energy networks, in terms of the fairness and scalability of the P2P trading, the economics and accuracy of the network operation models, and the effectiveness of the bi-level framework. Notably, the above-designed bi-level P2P trading framework simplifies each market participant as a single node, and disregards the three-phase unbalanced nature of DN and the underlying impacts of uncertainties. As a continuation of the above-designed bi-level P2P trading framework, this thesis further proposes an aggregator-network coordinated P2P multi-energy trading framework, where the aggregators and network operators coordinate to minimize the total cost and satisfy the operational constraints under uncertainties from renewable energy sources. In the lower layer, the aggregator for each region optimizes the intra-regional energy scheduling and the inter-regional P2P trading, which are formulated as an energy management problem and a double-auction market clearing problem, respectively. An adaptive robust stochastic optimization (RSO) approach is then developed to address the uncertain renewable generation modeled through a scenario-based ambiguity set. In the upper layer, based on the scheduling and trading results, the network operators optimize the DN and DHN operation decisions including network reconfiguration and Volt/Var regulation, and request necessary lower-layer adjustments if the operating violations cannot be fully solved. The DN and DHN operation problems are formulated based on a linearized three-phase unbalanced DistFlow model and a linearized thermal flow model, respectively. The simulation results verify that the proposed framework can effectively minimize aggregators’ daily costs and line loss of the networks, while removing potential network operation violations. The adaptive RSO approach is also proved to be superior to adaptive stochastic and robust approaches. Finally, to address the challenges brought about by the rapid proliferation of DERs along with the rise of P2P energy markets on DN operation, this thesis designs a multi-timescale reactive power ancillary service market, which is composed of a contracted-based day-ahead Var reserve service market and an hourly-ahead Var support service market. The day-ahead Var reserve service market enables the distribution system operator (DSO) to sign a long-term robust Var reserve contract in advance with the inverter-interfaced DERs who have an oversized inverter capacity. The day-ahead market is formulated as a two-stage robust optimization model and solved by a modified feasibility pump-based column and constraint generation (mFP-C&CG) algorithm. The robust Var reserve contract can effectively eliminate the potential market power that the DERs may gain due to system configuration deficiency and market structure flaws in the later hourly-ahead market. With the realization of uncertainties, the hourly-ahead Var support service market enables the DERs to further provide the DSO with paid hourly-ahead Var supports using their currently spare reactive power capacity. The bilateral hourly-ahead Var trading between the DSO and DERs is modelled as a non-cooperative game problem while the reactive power cost of DERs is thoroughly formulated, which thus enables the provision of more precise and realistic market strategies. To attain a generalized Nash equilibrium (GNE) of the non-cooperative game, a distributed algorithm is developed by building the connection between the GNE and the primal-dual solution of a related social cost minimization problem. The effectiveness of the proposed market framework and methods is validated on an IEEE 33-bus DN following a day-ahead P2P energy market developed in this thesis. In summary, this thesis centers around TES design and transactive energy management of MESs, which have been demonstrated as an effective energy management method in addressing the emerging challenges posed by the ever-growing penetration of DERs. The outcomes of this thesis can offer a timely guideline for future TES design and research, suggest a range of transactive energy management methods for different scenarios, and recommend ancillary service market frameworks in the context of transactive energy management.