Demand response for real-time congestion management incorporating dynamic thermal overloading cost
Introduction
Driven by an effective international climate policy [1], [2], the electrical distribution networks worldwide have been hosting an increasing share of renewable energy resource (RES) based distributed generation (DG) and new forms of load consumption such as electric vehicles (EVs), heat pumps (HPs), or electrical heating ventilation and air-conditioning (HVAC) systems. Along with a greener energy mix, these Distributed Energy Resources (DERs) bring forth operational challenges including voltage violations or thermal overloading of network assets [3], [4]. Consequently, distribution system operators (DSOs) require to enhance monitoring and control ability in the network and a shift towards Active Distribution Networks (ADNs) becomes imminent [5], [6], [7].
Congestions or thermal overloadings occur when the power flow through a network asset (e.g. lines, cables, transformers) exceeds its transfer capability. Although the network assets are generally designed to withstand loads beyond a certain margin, continuous overloading results in degradation of the insulation of the distribution cables and transformer windings [8], [9]. The traditional approach of reinforcing the network assets in such cases, not only necessitates a huge investment, but will also be deemed redundant as the peak loads tend to appear only for few hours in a year [3]. To circumvent the required investment, a number of direct and indirect control approaches have been studied to tackle congestion issues in the ADNs. While the direct approaches mitigate congestions by curtailment of load and local generation [10], [11], or by influencing the voltage level at the secondary side of a smart MV/LV transformer [12], [13], their indirect counterparts motivate individual prosumers with appropriate price and/or incentive-based Demand Response (DR) mechanisms [14], [9]. The price-based methods include day-ahead dynamic tariff, distribution grid capacity market, intra-day shadow price, flexibility service market, etc. However, most of these mechanisms aim to influence the demand flexibility only, while neglect or partially address physical constraints of the distribution networks [15], [16], [17]. Consequently, applications of the DR-based methods for managing real-time congestions are still quite limited.
Different types of multi-agent system (MAS) based DR mechanisms have been deployed to exploit the flexibility from the small-scale prosumers [18], [19], [20], [21]. As highlighted in [21], [22], [23], the agent-based PowerMatcher technology optimizes the potential for aggregated individual household devices to adjust real-time operation. At the same time, Universal Smart Energy Framework (USEF) has been recently introduced as a conceptual approach to manage congestions more efficiently [24], [9], [25]. USEF combines the indirect and direct approaches of congestion management to enhance the flexibility in the distribution network and enables the DSO to obtain flexibility from a local capacity market to relieve network congestions. A more direct approach of graceful degradation complements the market-based control to curtail active power demand when adequate flexibility is not available in the market.
However, for a market-based DR mechanism, a sound methodology for real-time congestion management is significant in order to transform the realized ageing of the network assets to a corresponding monetary loss. Based on the recent developments, an integrated congestion management mechanism is proposed in [9] for the residential distribution network incorporating dynamic thermal overloading model of a distribution transformer. However, in reality, procurement of flexibility in real-time appears to be a more complex problem involving scenarios with multiple market entities in the same congested network area [26]. This work extends the market-based control proposed in [9] further, incorporating computational intelligence in a multi-actor setting. The method will take advantages of the scalable architecture of the agent-based DR technology and detailed thermal model of oil-immersed transformers. A detailed case-study, involving 229 households is presented to illustrate the impacts and expected results of the proposed mechanism. The main scientific contributions of the paper are as follows:
- •
Demand response methodology incorporating the overloading cost, calculated from the dynamic thermal model of the transformer.
- •
Application of agent-based distributed intelligence to solve the congestions locally.
- •
An appropriate local flexibility market to procure flexibility in a multi-actor setting.
The remainder of the paper is organized as follows: Section 2 presents the overview of the market-based control, Section 3 describes the thermal loading model of the transformer, the methodology of flexibility procurement is detailed in Section 4, while Section 5 provides the description of the test scenario and the assumptions adopted. Finally, simulation results are presented and analysed in Section 6, before summarizing and concluding with Section 7.
Section snippets
Flexibility in distribution network operations
Due to the increasing availability of the flexible domestic appliances and small-scale generation technologies like rooftop solar PV, market-based control of the distribution network has been drawing an extensive attention of late [27], [28], [29]. Apart from introducing new market entities such as aggregators and energy service companies, this has principally paved the way for a more decentralized operation of the future power system.
Different types of flexibility arrangements have been
Thermal overloading of transformer
Overloading of the MV/LV transformer may occur due to increased loads at connection points, for instance–simultaneous charging of the large number of EVs or switching of domestic HPs. Although thermal overloading affects the insulation life-time of the transformer windings, thermal dynamics involved in the loading allows the transformer to be overloaded for a certain duration of time. Therefore, the procurement of flexibility needs to be aligned with actual status and corresponding cost of
Procurement of flexibility
Upon determining the overloading cost, the DSO requests flexibility from the aggregators in order to resolve the congestion. The task of flexibility procurement is carried out by the LFSA that works as the interface between the network agents and the aggregators. The estimated overloading cost is taken as the basis of the calculation and adequate flexibility is procured from the available offers of the aggregators.
Simulation setup
The proposed methodology is evaluated with a simulation case study in a residential Dutch LV network. The network consists of three feeders, namely Feeder A, Feeder B and Feeder C. Total number of households connected in the feeders are 95, 42 and 92 respectively. As shown in Fig. 4, a 500 kVA MV/LV transformer feeds the network from the MV bus. All of the feeders comprise of underground power cables. The network model is built in MATLAB environment using the open-source Power System Analysis
Numerical results
Snapshots of the simulation results of the case study are first presented for two consecutive winter days in the Netherlands. Subsequently, the annual performance of the proposed approach is analysed in terms of the cost savings and reduction in congestion duration.
Each of the three aggregators matches the local supply and demand of their contracted prosumers and calculates the internal equilibrium price, for the cluster. The per unit local price signal also reflects the loading status in
Conclusions
The focus of this work has been to propose a suitable mechanism for real-time congestion management in LV distribution networks. An agent-based system architecture has been adopted to solve the problems through distributed intelligence. The residential load profiles have been generated by a bottom-up approach involving dynamic control signals. A detailed thermal model of the transformer has been used to observe the impacts on the life time of the winding insulation in a LV network with a
References (47)
- et al.
Agent-based unified approach for thermal and voltage constraint management in LV distribution network
Electr. Power Syst. Res.
(2016) - et al.
Review of congestion management methods for distribution networks with high penetration of distributed energy resources
- et al.
Exploration of dispatch model integrating wind generators and electric vehicles
Appl. Energy
(2016) - et al.
Implementing a model predictive control strategy on the dynamic economic emission dispatch problem with game theory based demand response programs
Energy
(2015) - et al.
Flexible interaction of plug-in electric vehicle parking lots for efficient wind integration
Appl. Energy
(2016) - et al.
Distributed generation integrated with thermal unit commitment considering demand response for energy storage optimization of smart grid
Renew. Energy
(2016) - et al.
Hourly demand response and battery energy storage for imbalance reduction of smart distribution company embedded with electric vehicles and wind farms
Renew. Energy
(2016) - et al.
Combining market-based control with distribution grid constraints when coordinating electric vehicle charging
Engineering
(2015) - et al.
Uncertainty management of dynamic tariff method for congestion management in distribution networks
IEEE Trans. Power Syst.
(2016) - et al.
Smart grid and smart building inter-operation using agent-based particle swarm optimization
Sustain. Energy, Grids Netw.
(2015)
DIRECTIVE 2009/28/EC on the promotion of the use of energy from renewable sources, Tech. Rep.
Multi-Agent System based Active Distribution Networks
Power Play - Impacts of flexibility in future residential electricity demand on distribution network utilisation
Voltage-Based control of a smart transformer in a microgrid
IEEE Trans. Ind. Electron.
Review on implementation and assessment of conservation voltage reduction
IEEE Trans. Power Syst.
Cited by (62)
Multi-agent reinforcement mechanism design for dynamic pricing-based demand response in charging network
2023, International Journal of Electrical Power and Energy SystemsDeveloping a novel zonal congestion management based on demand response programs considering dynamic transmission ratings
2023, International Journal of Electrical Power and Energy SystemsFind the balance: How do electricity tariffs incentivize different system services from demand response?
2022, Sustainable Energy, Grids and NetworksA novel congestion management method through power system partitioning
2022, Electric Power Systems ResearchTransactive energy for low voltage residential networks: A review
2022, Applied EnergyCitation Excerpt :Ref. [59] proposed a distribution locational marginal pricing strategy for effective utilisation of the flexibility offered by responsive consumers to alleviate grid congestion. A DR model for congestion management is presented in [60], in which the congestion cost is represented in terms of the ageing of distribution transformers due to overloading, and the Distribution System Operator (DSO) procures the flexibility of responsive consumers to shift their consumption based on the congestion cost, which, in turn, mitigates the transformer overloading. Electric springs-based grid-interactive DR schemes have also been explored extensively in the literature, especially to tackle reliability and stability concerns stemming from power fluctuations from intermittent renewable energy sources in modern power systems [69].