I.G. Kamphuis

Distributed Control Concepts using Multi-Agent technology and Automatic Markets: An indispensable feature of smart power grids

Multi-agent technology is state of the art ICT. It is not yet widely applied in power control systems. However, it has a large potential for bottom-up, distributed control of RES and DER in future power systems. At least two major European R&D projects (MicroGrids and CRISP) have investigated its potential. Both grid-related as well as market related applications have been studied. This paper will focus on two field tests, performed in the Netherlands, applying multi-agent control by means of the PowerMatcher concept. In the PowerMatcher concept (http://www.powermatcher.net/) software agents are used as representatives of the power producing and/or consuming installations. Via market algorithms a strategy is determined to ensure, that their operational schemes are coordinated in order to balance supply and demand according to the business case. The algorithms in the PowerMatcher use a bottom-up electronic market mechanism. Building such a system, controlling primary user processes on one hand, assuring local autonomy, and operating on the electricity market on the other hand, appears to be feasible with mainstream ICT-components. We will describe and discuss a number of results from two field tests performed with the PowerMatcher concept, and discuss further developments.

Capacity management within a multi-agent market-based active distribution network

Normal operation of an active distribution network (ADN) requires simultaneous optimization of different objectives of the various involved actors. This results in a multi-objective optimization problem which has not yet been treated completely. This paper considers a particular relationship between commercial and technical coordination, involving capacity management of the distribution network. First, the market-based ADN, its actors and their objectives are described. An agent-based approach is desirable to handle the complexity of this ADN. Then, several technical issues for integrating capacity management within a multi-agent market-based ADN are pointed out. After that, the developed agent architecture and coordination mechanism are further elaborated upon, along with a formulation of the multi-objective optimization problem. Finally, a decentralized approach for integrating capacity management is introduced and demonstrated.

Local DER driven grid support by coordinated operation of devices

In the traditional operation of electricity networks the system operator has a number of ancillary services available for preservation of system balance. These services are called upon near real-time, after the planning phase. Ancillary services consist of regulating power, reserve capacity and emergency capacity, each with their own characteristics. Regulating power is deployed via load frequency control. Reserve capacity is used to release regulating power and can be called upon to maintain a balance or to counterbalance or resolve transmission restrictions. Both are traded at the Dutch energy market under an auction model with a single buyer (TenneT). Emergency capacity is rewarded on the basis of accessibility/availability within 15 minutes. In local electricity networks neither planning nor ancillary services exist. Planning is done by aggregation into large customer groups. For ancillary services one relies on the system operation as sketched above. In local electricity networks with a large share of distributed generation the costs of keeping the electricity system reliable and stable will increase further and technical problems may arise. The European SmartGrids initiative responds to these challenges in their strategic research agenda. One of the issues addressed in this agenda is the changing role of the distribution grid in which users get a more active role. One opportunity is the introduction of ancillary-type services at the distribution level, utilizing different types of producing and consuming devices in the local network, in order to make the total system more dependable. Distributed generation has a number of characteristics that are similar to characteristics of consumption. Part of it is intermittent / variable, although to a large extent predictable (PV, wind versus lighting, electronic devices). Another part is task-driven (micro-CHP versus electrical heating). Yet another part is controllable or shiftable in time. And storage can behave both ways. The main key words here are flexibility and variability. This flexibility provides a virtual storage capacity within the electricity grid that can be utilized for balancing services at the local grid. We will present how the PowerMatcher concept, developed by ECN, supports the setting up of local balancing markets in a flexible and logical way. The ICT is already available as an enabling technology. The concept has been demonstrated in several field tests.

The development of demand elasticity model for demand response in the retail market environment

In the context of liberalized energy market, increase in distributed generation, storage and demand response has expanded the price elasticity of demand, thus causing the addition of uncertainty to the supply-demand chain of power system. In order to cope with the challenges of demand uncertainty under the unbundled electricity market, the concept of Market-based Control Mechanism (MCM) in retail market environment has been emerging. This paper presents the concept considering demand elasticity as an opportunity in retail market environment for inventing a new bid mechanism. This work formulates demand elasticity model as a Markov decision problem and implements pursuit algorithm as a machine learning technique to evaluate the price elasticity of demand by predicting the price. The performance of the algorithm is compared with the numerical calculation of price elasticity of demand for the given simulation settings.

Integration of agent-based functions to facilitate operation of Smart Distribution Networks

In addressing various issues in the transition to a sustainable and low carbon society, Smart Grids can play a significant role. Smart Grids use both state-of-the-art as well as emerging technologies in the field of Power Electronics (PE) and Information and Communication Technology (ICT) as core enablers. In such a context an ICT agent-based approach to deploy artificial intelligence into distributed system operation is an appropriate and promising technology to handle the system complexity and can enhance the overall performance of the grid. To adapt to the unpredictability of the myriads of small-scale generation and controllable loads, various agent-based technologies to facilitate the operation of Smart Grids have been developed. The agent-based functions are highly autonomous and operate independently while their mutual influences on each other are currently are understood and studied. This paper discusses the possibility of mutual interaction in a unified framework with the objective to facilitating both ancillary services and energy trading. An integrated platform for harmonizing the agent-based functions is proposed under the context of Smart Grids, with specific focus on Smart Distribution Networks.

Limitations in the design of ancillary service markets imposed by communication network delays

Real-time balancing of the European electricity grid will become increasingly dependent on market-based control mechanisms that are enabled by connecting millions of prosumers to an open communication network. The use of communication systems inevitably introduces delays in the energy balancing control loop, which could endanger market operation and stability of the electricity grid. By investigating the interaction between price-based control algorithms for real-time balancing and information and communication technology, we aim to provide systematic design rules for unrestricted ancillary service markets.

Demand dispatch for different demand response schemes: A consolidated solution

Advance infrastructures have changed the passive consumers into active because now they can share information, perform automatic control as well as directly influence the electricity market via demand response (DR) programs. Till today, many DR Programs are proposed in Smart Grid (SG) paradigm and are facing enormous challenges. This paper concerns the DR scheduling problem of the dispatchable loads at the end consumer premises. In DR Programs, the electricity prices vary over time and users receive the reward payment by the energy service providers, if the users modify the consumption during DR events. Thus, the paper devises the independent atomic dispatchable model which could be used for investigation of different kinds of DR Programs. The another objective of the paper is to formulate the generic scheduling problem i.e. how to schedule the operation of the dispatchable loads, taking in account the overall energy cost, the comfort level and the timeliness. Paper also performs the simulations for two different DR Programs namely Time-Of-Use (TOU) and Energy Bidding (EB), and results show that the scheduling problem is suitably formulated for real-life instances and is applicable to other types of loads

Energy Flexometer: An effective implementation of Internet of Things for market-based demand response in an energy management system

Market-based control mechanism (MCM) needs the IoT environment to fully explore flexibility potential from the end-users to offer to involved actors of the smart energy system. On the other hand, many IoT based energy management systems are already available to a market. This paper presents an approach to connect the current demand-driven (top-down) energy management system (EMS) with a market-driven (bottom-up) demand response program. To this end, this paper considers multi-agent system (MAS) to realize the approach and introduces the concept and standardize design of Energy Flexometer. It is described as an elemental agent of the approach. Proposed by authors Energy Flexometer consists of three different functional blocks, which are formulated as an IoT platform according to the LonWorks standard. Moreover, the paper also performs an evaluation study in order to validate the proposed concept and design.

A study of congestion management in smart distribution networks based on demand flexibility

Electrical distribution networks worldwide are facing frequent capacity challenges due to the widespread roll out of various distributed energy resources (DERs). A number of demand response (DR) mechanisms have been developed in order to circumvent the problems and enhance the flexibility of the distribution network. While the existing centralised control system remains its crucial role for reliable and secure grid operation, distributed intelligence is a complement technology with a focus on dividing the control task into a number of simpler problems and solve them with minimum exchange of information. Based on the recent developments of distributed intelligence, this paper investigates a set of different congestion management approaches to effectively regulate the overloading issue of transformers or conductors. The study is validated with simulations for representative Dutch low-voltage (LV) networks.

Learning technique for real-time congestion management in an active distribution networks

During the last few decades, the concept of demand response (DR) in the energy sector has gained substantial momentum. Research has led to a range of DR solutions. These solutions mostly differ in their applications, the hosting power system, the energy market etc. Moreover, as per the EU directive, DR aggregators should be allowed to trade DR alongside supply in both day-ahead and real time electricity markets. Meanwhile, independent aggregators do not consider physical limitations of a network, thus setting up new a challenges for network operation. In this paper, an active learning technique for real-time congestion management is proposed to tackle this challenge. This enables distributed system operator (DSO) to incenticize independent aggregators efficiently in order to use DR for overloading mitigation. Lastly, a case study is simulated which verifies the performance of a new approach for congestion management.