Project topics and materials
Get Complete Project Material File(s) Now! »
Demand Side Management and Demand Response
Demand Side Management is a very wide concept that involves programs focusing on three differen type of issues: efficiency issues related to the isolation of homes or the efficiency of appliances, social and anthropological issues related to the change in human behavior, e.g. how to gain customer active participation and commitment through personalized feedback, and finally the most advanced level involves the introduction of information, communication and control technologies to have a fine grained management of the demand side. In this last level of management we find DR programs. Demand Response is one of the main applications in a Smart Grid environment, and is about shaping the demand of electricity during certain periods of time (particularly during peak hours) by leveraging the flexibility at customer premises in exchange of some kind of incentive. The flexibility at customer premises is based on the characteristics of elastic loads, modifying their load curves at the cost of an impact in the user experience, and on the use of distributed generation and/or storage facilities, including EV.
DSM programs allow a more efficient use of the available resources, reducing CO2 emissions and operational costs. Particularly, DR programs can improve reliability of the grid helping to fit the demand to the available/planned capacity and enable deferring capital investments by delaying the need for new peak generation facilities and further changes in the grid infrastructure. DR could provide several services at different time scales, ranging from the hour-ahead market, to the tertiary reserve, passing through the fast response reserve, but the market structure must provide incentives for aggregators/utilities to develop DR programs. This is why the roles of markets and regulation are key to understand how a realistic DR system could work. Besides characterizing the possible sources of revenue, markets and regulations define the exchanged products, imposing constrains in how those products are measured and translated into a customer environment, e.g.: regarding time issues, if customers’ flexibility will be used only in peak times or all along the day, meaning, if the product in the user side will be to follow a daily curb, to reduce the load during peak time, or both.
Functional Architecture
In an IoT-based architecture there are three fundamental functional groups. The first group is the set of functions at the edge of the architecture, enabling the merge between the physical and digital world. These Edge Functions (EF) can be divided into sensing the environment (context awareness), acting on the environment (state management) and user interaction. The second group represents functions related to data management and treatment, based on which applications execute business logic to provide customer value. We refer to this group as Cloud Functions (CF), as for most of current IoT products and services such functions are implemented over cloud infrastructure. The last functional group is called Networking Gateway Functions (NGF) and is in charge of enabling the interconnection among EF and CF function groups, by providing the functions related to connectivity, interoperability, device & resource discovery, securing communications (e.g.: secure pairing and onboarding), etc.
These functional groups can be implemented in many different ways, at different levels of the architecture. The best implementation depends on the applications for which the architecture will be applied, so an important quality of a general architecture is to enable flexible and heterogeneous implementations of such functions. In the next sections we analyze the technology options for implementing the functional architecture into physical components and to enable the communication among them.
Implementation of Networking Gateway Functions
There are three implementation possibilities for interconnecting CF and EF implementations into cloud infrastructure, edge devices and physical gateways:
• Edge devices connect to cloud infrastructure through cellular networks, by including a SIM card on each device.
• Edge devices connect to cloud infrastructure through long range & low bitrate technologies, such as LoRa (Long Range) or SigFox (Ultra Narrow Band).
• Edge devices connect to cloud infrastructure through short range technologies, with a local equipment implementing the endpoint for Physical and MAC layers.
The first two options require the implementation of CF and NGF at the cloud infrastructure or on the access network; what deters the possibility of distributing the implementation of CF by extending cloud infrastructure capabilities to the edge. Furthermore, the characteristics and requirements of cellular networks and long range & low bitrate technologies, such as costs, bitrate, energy consumption and maintenance may not be adapted to constraints from Smart Home Services. Thus, the local deployment of a physical endpoint, implementing most of NGF, would be necessary to optimize the distribution of data and intelligence for avoiding delay, reliability and privacy issues. The requirements for the implementation of the Physical Gateway device are centered on native support for the relevant protocol stacks, reliability and security features. Currently there are some IoT gateway development platforms, such as OpenPi, OpenKontrol Gateway, or even the BBB can be used with connectivity extensions, priced around $40, which highlight the low cost of building a gateway device. The implementation of a custom designed Gateway device can leverage the AM335x ARM Cortex-A8 used by the BBB, as it is cheaper (sells for $5 in quantity) and simpler (development, power management, etc.) compared to the popular Cortex-A9 multi-core platform. With respect to OSs and according to the requirements we would consider mbed OS as the main choice thanks to is broad support for connectivity technologies. In the case of Snappy, the OS has the flexibility to adapt to different technologies, but drivers are yet to be developed.
Table of contents :
I Introduction
1 Introduction
1 Context and vision on Smart Grids
2 Motivation
2.1 Structure of electricity Markets
2.2 Demand Side Management and Demand Response
2.3 Thesis objectives
3 Thesis overview
4 Summary of thesis main contributions and outline
II Novel architecture for distribution grid energy management
2 Smart Home IoT-based architecture
1 Introduction
2 Functional Architecture
3 Physical Architecture
3.1 Implementation of Edge Functions
3.2 Implementation of Cloud Functions
3.3 Implementation of Networking Gateway Functions
4 Communication Architecture
4.1 Interface with energy harvesting devices
4.2 Interface With Battery-powered Devices
4.3 Interface with the Back-end
4.4 Interoperability Considerations
5 General Architecture Improvements
5.1 Orchestration of transactions across multi-service controllers
5.2 On-premises distributed enforcing of ECA rules
5.3 Challenges of a household blockchain
6 Business model analysis
6.1 Current IoT business models are broken
6.2 New business models based on shared distributed platforms
7 Conclusions and Perspectives
3 Distribution Grid Energy Management Architecture
1 Distribution Grid Energy Management
1.1 Objectives of a distribution grid EMS
1.2 Requirements for a distribution grid EMS
1.3 Current approaches for DER management
2 DSM Architecture Proposal
2.1 Virtualized Distribution Grid
2.2 Design of a Residential VDG
3 Transactive platform implementation
4 Conclusions and Perspectives
III Innovative services enabled by the architecture
4 Renewable Energy Balancing local Market
1 Introduction
1.1 Related Work
2 System description
2.1 Local renewable energy balancing markets
2.2 Prices and auto-consumption
3 Metrics for evaluating market impact
4 Market mechanism
4.1 Price incentives
5 Households optimization problem
5.1 Determine offer quantities
5.2 Determine final flows
6 Simulations and Results
6.1 Simulation tools, procedure and assumptions
6.2 Results and discussion
7 Conclusions and Perspectives
5 Dynamic phase switching for augmenting DER hosting capacity
1 Introduction
2 System description
2.1 Local energy market
2.2 Dynamic Phase switching
2.3 Solid-state Transfer Switches
3 Related Work
4 Dynamic Phase Switch Model
4.1 Phase allocation
4.2 Choice of households to switch
4.3 Dynamic phase allocation
5 Simulations and Results
5.1 Procedure and system scenarios
5.2 Energy market simulation
5.3 Simulation tools and parameters
5.4 Results and discussion
6 Conclusions and perspectives
6 Real-time adjustment of market decisions
1 Introduction
2 Related Work
3 System Description
3.1 Local renewable energy market
3.2 Real-time control mechanism
3.3 Dynamic phase allocation
4 Model
4.1 State of agents
4.2 Agents’ strategies
4.3 Batteries
4.4 Final electricity prices
4.5 Reward allocation
4.6 Aggregative Constraint
4.7 Problem definition
5 Competitive aggregative game
5.1 Dynamic control algorithm of decentralized optimal responses
5.2 Iterative Process
6 Simulations and Results
6.1 Procedure and system scenario
6.2 Simulation tools and parameters
6.3 Results and discussion
7 Conclusions and perspectives
8 Conclusions and perspectives
IV General conclusions and perspectives
7 General conclusions and future work
1 General conclusions
2 Future work
2.1 Extensions for additional value creation
2.2 Future electricity distribution grids
Bibliography
