**1. Introduction**

Fundamental changes are expected in the power systems in terms of increasing shares of renewable energy resources (RES)-based local distributed generation (DG) units and price-responsive residential flexible loads. Due to the intermittent nature of the DG units and reduced diversity of the loads, more and more operational challenges are appearing in residential LV networks in terms of voltage variations, thermal overloading and power quality issues [1]. Congestion or thermal overloading result from power flows through a network asset (lines, cables, transformers) exceeding its transfer capability. Even though the network assets are usually designed to withstand power flows beyond a defined margin, continuous overloading leads to ageing and subsequent failure of the distribution cables and transformer windings [2,3]. The conventional approach of reinforcing the network assets to tackle such issues necessitates a huge investment, and it is also deemed redundant, as the frequency of such issues is unpredictable [4]. In order to avoid the huge costs, different types of demand response (DR) mechanisms have been studied and developed worldwide for the procurement of flexibility from end-users in order to solve these operational challenges. These mechanisms can be roughly subdivided into direct and indirect load control methods. In direct methods, end-users are compensated through an appropriate contract, which enables automatic or manual shift in consumption of electricity based on the requirements of the network operator. Indirect methods, on the other hand, invoke

flexibility through a market-based coordination and/or dynamic price signals [5,6]. An overview of such mechanisms has been presented in [7–12] for various types of smart grid domains such as, for instance, microgrids, load balancing services, smart electric vehicle (EV) charging and residential energy management.

A centralized coordinated home energy management system, which implements daily power based network tariffs to mitigate network congestion due to the high demands of heat pumps (HPs) and EVs, is described in [13]. In contrast to conventional centralized methods, a number of decentralized approaches have been proposed in order to adequately address the comfort of the end-users for scheduling the energy consumption of dynamic inputs [14–17]. Such decentralized approaches aim to enable the end-users to take autonomous actions with no or limited communication with other entities. Contrary to the centralized approaches, they offer a more robust and reliable operation due to the lower vulnerability for communication failure.

A local market based framework has been presented in [18] that supports the network operators and prosumers to fully utilize the available flexibility and, thus, enhance the operational efficiency of the network. Dynamic tariff schemes that consider the high level penetration of EV and HP have been discussed in [19–21]. In order to tackle congestion in distribution networks, a dynamic subsidy (DS) method is presented in [22] to motivate aggregators to adjust the load profiles and to take network constraints into consideration. DS is inherently an incentive or reward that the distribution system operator (DSO) is willing to pay to the aggregators who would reschedule their energy profiles. Sensitivity analyses and linear approximation methods have been applied in [23] to develop a pricing model, considering EVs as a flexible load and accounting for three-phase and single-phase voltage variation, network losses and overloading of network components. For the identification of barriers, potential solutions and opportunities, a generic method is applied in the Netherlands [24]. The applied method focuses on market integration aspects and tries to access the flexibility procurement mechanism through the aggregators. As an outcome of the study, an analysis on relevant issues for market integration with reference to the Dutch system and European Grid codes is presented and a lack of proper communication system between users is identified. Authors of [25] have performed an analysis to understand the network pricing methodology most suitable for prosumers in the future, and they concluded that contracts would be economically efficient when retailers have access to the full range of devices, which helps them to convert wholesale electricity pricing in retail contracts, and second, the wholesale price associated with customers must represent the short term marginal cost of electricity production.

However, for a market-based DR mechanism, a sound methodology for congestion management is important to reflect the realized ageing of the network assets to a representative financial loss. An integrated congestion management mechanism has been proposed in [2,16] for the residential networks that involve the dynamic loading model of a distribution transformer. In reality, procurement of flexibility can be a more complex problem involving multiple involved actors in the same network. This work extends the market-based control proposed in [16] for day-ahead scheduling of the residential appliances that considers the dynamic network tariff to avoid thermal overloading of the MV/LV transformer. The tool developed in [16] has been used for estimating the overloading cost for different scenarios of imminent day-ahead loading, and a more realistic approach has been adopted of changing the network tariff dynamically for various scenarios of loading. Thus, the new approach becomes robust for uncertainties associated with day-ahead scheduling. The method will take advantages of Multi-Agent Systems (MAS) for a scalable architecture and distributed intelligence in a multi-actor setting. In principle, the method allows for local decision making and dynamic scheduling of the appliances with limited communication and a smaller number of iterations among the actors. Key contributions of this article can thus be listed as follows:

• Day-ahead market-based congestion management through agent-based scheduling of the residential appliances;


The rest of this paper is organized as follows: Section 2 highlights the overall approach and system architecture; Section 3 presents the problem formulation along with the description of the proposed methodology; Section 4 provides the description of the simulation test case and the assumptions adopted. Finally, simulation results are discussed in Section 5, before summarizing and concluding with Section 6.
