With the rising adoption of Electric Vehicle (EV) technology and Renewable Energy Sources (RES), electric distribution grids are facing new challenges regarding congestion management. The present work steps into the topic of controlled charging mechanisms to reduce physical grid extension by utilizing flexible loads from EV. Although, existing research concludes a positive impact on congestion relief, less attention is given to a holistic and light system that is implementable under current circumstances. This thesis develops a novel system towards micro-auctions for local flexibility allocation amongst EVs to reduce grid congestion. A functional software prototype simulates a virtual market and grid environment. Each EV acts as an autonomous agent submitting bids to the local flexibility market, offering 15-minute charging breaks. Based on individual risk preference and state-of-charge, bidprices vary amongst EVs.
The Distribution Grid Operator (DSO) constantly asses grid status and contracts positive capacity during critical phases by accepting current bids. It can be shown, that regardless of the penetration rate of EVs, the proposed model balances the tested grid topology below the maximum workload and within a predefined range. According to simulation assumptions, a ninefold increase of EVs can be accommodated with the proposed model. Although, with monotonically increasing penetration rate, average charge-increase converges to zero. Due to the proposed intervals, EVs are grouped to continues batches with demandresponse latency. Once contracted, EVs remain charging or not-charging for 15 minutes. The assignment to certain batches does not change over simulation time. Based on the proposed request control mechanism, critical grid conditions can be reduced by 49%. Whereas quantitative results are limited to the proposed simulation assumptions, qualitative effects are generalizable to a certain extend.
Contents
1 Executive Summary
2 Introduction
2.1 Motivation and Problem Identification
2.2 Research Question and Objectives
2.3 Structure of the Thesis
3 Theoretical Foundations
3.1 BriefHistoryoftheGermanEnergyGrid
3.2 Distribution and Transmission
3.3 Energy Markets
3.3.1 Energy-Only-Markets
3.3.2 Capacity Markets
3.3.3 Contractual Balancing
3.3.4 Auctions in Competitive Electricity Pools
3.4 The Problem of Balancing and Congestion
3.4.1 Grid Congestions and Frequency Balance
3.4.2 Congestion Mechanism
3.4.3 Balancing Mechanism
4 Local Flexibility: A Novel Approach
4.1 The new importance of Distribution System Operators
4.2 Definition of Flexibility
4.3 The Problem of Low Demand Flexibility
4.4 Flexibility from Electric Vehicles
4.4.1 V2G
4.4.2 Current Limitations of V2G
4.4.3 Controlled Charging
4.4.4 Current Research to Controlled Charging
4.5 Local Flexibility Markets
4.6 The Traffic Light Concept
4.6.1 Historical Background
4.6.2 Concept
4.7 Focus of this work
5 Simulation Software: Concept and Design
5.1 Problem Definition and Goal
5.2 Conceptual Design
5.2.1 Applied Grid Environment
5.2.2 Traffic Light Thresholds
5.2.3 DSOandFlexmarketConcept
5.2.4 EV Concept
5.2.5 Communication and Information Concept
5.3 Architectural Design
5.3.1 Simulation Framework
5.3.2 System Design
5.3.3 Input values
5.3.4 Agent Behaviour Modeling
6 Evaluation
6.1 Evaluation process and tools
6.2 Time-Series Simulation
6.2.1 Grid Workload
6.2.2 Completed Charging Requests
6.2.3 Bid Acceptanc Ratio
6.2.4 Price of Bids
6.2.5 Charging pattern of Example EV
6.2.6 Summary Time-Series Simulation
6.3 Sensitivity Simulation
6.3.1 Auction Cost, EV Profit and SoC Increase
6.3.2 Completed Charging Requests
6.3.3 State of Charge and Optimal EV penetration
6.3.4 Summary Sensitivity Simulation
6.4 Identified Effects
6.4.1 Request Control
6.4.2 Latency
6.4.3 Batching
6.5 Robustness against Dynamic Inputs
6.5.1 Initial Sate of Charge
6.5.2 Arrival and Departure Time
6.5.3 Risk Preferences
6.5.4 Random Seed
6.6 Limitations
6.6.1 Forecasting Demand Structure
6.6.2 Dynamic Charge Power
6.6.3 Optimal Segmentation
6.6.4 Balancing Group Problem
6.6.5 Price Cap and Price Function
7 Conclusion, Discussion and Outlook
Zusammenfassung
Durch die wachsende Nutzung von Systemen zur Produktion von erneuerbaren Energien und der zunehmenden Verbreitung von Elektrofahrzeugen stehen elektrische Verteilungssys- teme vor neuen Herausforderungen im Bezug auf das Engpassmanagement. Die vor- liegende Arbeit diskutiert die Flexibilisierung des Verbrauchs von Elektrofahrzeugen durch kontrollierte Ladevorgange als Losungsmoglichkeit bei kritischen Netzengpassen. Trotz erwiesener, positiver Auswirkungen durch die Nutzung steuerbarer Flexibilitaten, sind bisher wenig Anstrengungen in Richtung eines holistischen und leicht implementierbaren Systems getaatigt worden. Diese Thesis stellt einen neuartiger Ansatz zur Allokation lokaler Flexibilitaat von Marktteilnehmern durch Micro-Auktionen vor.
Ein funktionierender Software Prototyp simuliert eine virtuelle Markt- und Netzumge- bung. Jedes Fahrzeug handelt als unabhaangiger Agent durch die Abgabe eines Gebots- preis am lokalen Markt fuar Flexibilitaat auf eine 15-minuatige Ladeunterbrechungen. Die Gebotspreise der einzelnen Elektrofahrzeuge variieren basierend auf individuellen Risiko- faktoren und dem aktuellen Ladezustand. Der Verteilnetzbetreiber bewertet die aktuelle Auslastung und kontrahiert im Falle eines kritischen Netzzustandes positive Kapazitaaten durch Annahme von Geboten.
Die Auswertung zeigt, dass unabhaangig von der Anzahl an Elektrofahrzeugen, die unter- suchte Netztopologie unterhalb der maximalen Auslastung geregelt werden kann. Unter den gesetzten Annahmen kann somit die neunfache Menge an Ladeanfragen und Elektro- fahrzeugen gesteuert werden. Mit steigender Anzahl an Elektrofahrzeugen, konvergiert jedoch die durchschnittliche Ladungszunahme gegen null.
Durch die implementierten Intervalle von jeweils 15 Minuten werden die Elektrofahrzeuge sowohl in Gruppen zusammengefasst als auch deren Reaktionszeiten verzaogert. Sobald ein Elektrofahrzeug kontrahiert wird, befindet es sich fuar 15 Minuten in einem fixiertem Zus- tand. Die Zuordnung zu einer Gruppe aandert sich waahrend der gesamte Simulations-zeit nicht. Mit dem entwickelten Mechanismus laasst sich innerhalb der Simulation eine Reduk- tion der kritischen Netzzustaande um 49% erreichen. Quantitative Aussagen beschraanken sich auf die getroffenen Annahmen der Simulation, wohingegen qualitative Effekte in be- grenztem MaBe generalisierbar sind.
List of Figures
1 Three-node network with constrained capacity (Hirth et al. (2018))
2 Principally flow chart of local flexibility market (Wagler and Witzmann (2016))
3 One-line diagram of the European low voltage test feeder
4 Schematic price function
5 UMLdiagramofsystemlayout
6 Normal distribution of starting values
7 UML simulation procedure
8 Process pipeline
9 Consumption structure with dumb-charging
10 Consumption structure with controlled-charging
11 Compared consumption structure
12 Completed charging requests
13 Bid acceptance rate
14 Maximal bid-price
15 Total auction cost for DSO
16 ChargingpatternEV
17 Distribution of accepted bids over SoC
18 Normal distribution of starting values
19 Bid price and profit
20 Completed charging requests rate
21 Completed charging requests rate
22 Final SoC (avergage vs. minimal)
23 Highresolutiongraph:Latency
24 Simultaneous arrival time
25 Maximum deviation over 20 trails
26 SoC obtained versus charging time spent (Wang et al. (2016))
27 Schematic grid aggregation areas (VDE (2014))
28 Schematic representation of the ”corrected model”
List of Tables
1 TLP limits and thresholds
2 Random variables
3 Static variables
2 LISTINGS
4 Auction log of example EV
5 Summary time-series simulation
6 Overview robustness
Listings
1 Flexmarketclass
2 Car class
1 Executive Summary
With the rising adoption of Electric Vehicle (EV) technology and Renewable Energy Sources (RES), electric distribution grids are facing new challenges regarding congestion management. The present work steps into the topic of controlled charging mechanisms to reduce physical grid extension by utilizing flexible loads from EV. Although, existing research concludes a positive impact on congestion relief, less attention is given to a holistic and light system that is implementable under current circumstances. This thesis develops a novel system towards micro-auctions for local flexibility allocation amongst EVs to reduce grid congestion.
A functional software prototype simulates a virtual market and grid environment. Each EV acts as an autonomous agent submitting bids to the local flexibility market, offering 15-minute charging breaks. Based on individual risk preference and state-of-charge, bidprices vary amongst EVs. The Distribution Grid Operator (DSO) constantly asses grid status and contracts positive capacity during critical phases by accepting current bids.
It can be shown, that regardless of the penetration rate of EVs, the proposed model balances the tested grid topology below the maximum workload and within a predefined range. According to simulation assumptions, a ninefold increase of EVs can be accommodated with the proposed model. Although, with monotonically increasing penetration rate, average charge-increase converges to zero.
Due to the proposed intervals, EVs are grouped to continues batches with demandresponse latency. Once contracted, EVs remain charging or not-charging for 15 minutes. The assignment to certain batches does not change over simulation time. Based on the proposed request control mechanism, critical grid conditions can be reduced by 49%. Whereas quantitative results are limited to the proposed simulation assumptions, qualitative effects are generalizable to a certain extend.
2 Introduction
The power industry is undergoing a major paradigm shift. With the ongoing integration of RES and the forecasted mass-adoption of EV technology, traditional approaches to reinforce electrical power girds are limited to physical extension. As an integral part and point of access, the distribution grid in particular, is at the forefront of research and development 123. Whereas approaches towards the challenge of integration vary, the overarching objective is clear: Applying smart mechanisms and technologies to avoid physical grid extension while ensuring grid stability and availability 4. As these goals are causing each other, managing grid congestion and balancing consumption are the key activities that need to be transformed according to the new paradigm.
The Federal Network Agency in Germany (Bundesnetzagentur (BNetzA)) considers intelligent control mechanisms as an integral part of the investment strategy by DSO. In its annual report of 2018, the BNetzA additionally mentions the increasing planning complexity on a 10-year horizon. With a faster evolution of technological advancements, DSOs can no longer anticipate extension and maintenance strategies and need to utilize new approaches to comply with changing demand. 5
Past research has acknowledged challenges arising form RES and EV integration, but pointed out that these technologies might inherent a solution right away. More than 20 years ago, Kempton et al. (1997) identified bi-directional charging of EVs, so called Vehicle-to-Grid (V2G) technology as a possible tool for grid operators to perform decentralized congestion management 6. Also the balancing of exceeding RES energy on residential level, can be accomplished with the use of EVs 7. In the following decades, extensive research exposed various theoretical approaches towards the utilization of EV for this purpose 8 9 10. With the Traffic Light Concept (Ampelkonzept ) drafted by the BDEW an interaction framework was issued to make grid information more transparent and incentivize EV owners to actively participate in local congestion management. This framework allows the DSO to communicate current grid workload in the form of a yellow traffic light to EVs (and other market participants) that are, in turn payed for the responsive action 11 4. Although this concept paves the way for interaction between DSO and EVs, no specific market design to match supply and demand is prescribed.
2.1 Motivation and Problem Identification
However, V2G technology is barely applied and its implementation faces a set of problems ranging from physical limitations, legislative restrictions to potential customer satisfaction (see section 4.4.2). Considering the opening paragraph, the present work tries to bypass these problems by introducing a light system-design that complies with current limitations in regard to physical hardware as well as applicable law. This approach follows the idea of the Minimal Viable Product (MVP). Reducing complexity and rapidly push ideas to prototype are the core principles of lean startup 12. In cooperation with Paatz Scholz van der Laan GmbH, a German energy consultancy focusing on digitalization and liberalization of the energy industry, the author tries to implement expert knowlegede and customer perspective from the early start. Thanks to the vital exchange of industry knowledge and fruitful discussion with experts and researchers, many hypotheses were validated before framing the research and development question.
2.2 Research Question and Objectives
The present work introduces an auction-based marked design for the Traffic Light Concept to explore possibilities of local congestion management applying controlled charging of EVs. To validate the general applicability of such a mechanism, qualitative effects are subject of investigation rather than quantitative results due to the unpredictability of future EV penetration, evolving charging technology or extensive grid topologies. Nevertheless, quantitative results are presented with regard to stated assumptions for comprehension. In the course of this work, the below stated research questions and their respective outcome are a metric for evaluating potential advantages of the proposed model. Nevertheless, the objective of the present work is not limited to these questions, as the conceptual and architectural layout as well as exceeding findings allow for a broader understanding on the issue of congestion management.
1. Does the contraction of local flexibility from EVs via the proposed auctionbased marked design lead to congestion release on distribution grid level?
2. What generalizable effects arise from the proposed auction-based marked design?
3. What is the maximum EV penetration for the assumed grid topology and proposed auction-based marked design, with a minimum average chargeincrease of 15%?
2.3 Structure of the Thesis
Following a brief historical outline, chapter 3 introduces the foundation of electric power transmission and trade with a focus on Germany. The concept of energy and capacity markets is explained as well as procedures to settle contracts for energy delivery. Before elaborating on the basics of electric grid congestion and issues on balancing energy, auctioning as a tool to match supply and demand at energy markets is described. Subsequent, chapter 4 introduces the concept of local flexibility as a possible solution for challenges arising from EV and RES integration. The low demand flexibility from residential consumers describes the current shortcomings for an appropriate market design. With V2G and controlled charging, two viable approaches towards the integration of EVs are discussed before elaborating on local flexibility markets and the Traffic Light Concept. Section 4.7 completes the current chapter and sets the specific focus of the present work. Chapter 5 describes the proposed model and is split in two major sections. Chapter 5.2 introduces the underlaying assumptions, the context and dependencies whereas chapter 5.3 focuses on the practical implementation of the software and its mode of execution. Chapter 6 evaluates the simulation results based on various analytical methods. Effects of interest are highlighted in section 6.4 before limitations and problems to the proposed model are juxtaposed. Chapter 7 closes with the author's conclusion and an outlook on possible further research.
3 Theoretical Foundations
Electrical energy is an omnipresent commodity. Consumers expected electricity to be affordable and available without interruption. In the background, tremendous effort is spent every day, to fulfill this expectation 13. To ensure a ubiquitous and reliable energy supply, the German electric grid and its affiliated markets were developed, and evolved into a complex and abstract structure 14. In times of RES and decentralization, this commodity moves further into the focus of the public debate. To give the reader a glimpse of certain mechanisms, and to understand why the topic of this work is relevant, the following chapter tries to shed more light on this essential commodity.
3.1 Brief History of the German Energy Grid
With the electrification of Europe and Germany during the 19th and 20th century, an extensive electric power grid was established to supply consumers with the highly demanded electricity. During this time, the former German Empire issued concession to a hand full of energy producers and providers. These companies were exclusively allowed to build the grid infrastructure and sell energy what made them de facto monopolist in their respective area of service. In 1997, nearly 100 years later, the European Union decided to liberalise the market for electrical energy by forcing the existing companies to allow additional providers to use their grid infrastructure. More than 10 years later, in 2009, a last step towards a fully liberalized market was taken, by unbundling grid operations and power supply. The transmission grid was therefore transferred to four independent companies, the so called Transmission Grid Operator (TSO). In 2011, local utility providers followed and split off distribution grid operations into independent entities, the DSOs. 13 Since then, market mechanisms for wholesale energy apply independent from geographical location of supply and demand. The location of a generating unit and the point of consumption, is not reflected in the prices paid. A wind farm operator can produce energy at the North Sea and sell it at the energy exchange in Leipzig to an energy retailer located in Munich who in turn is supplying a consumer in Cologne. 15
On the other hand, the physical grid has requirements and limitations on every distribution level. As a result, grid operators have to monitor and balance the grid continuously, using different methods to achieve an equilibrium between production and consumption. With the integration of renewable energies, the anti-nuclear commitment and a likely adoption of electric vehicles, this task is becoming more complex. 16
3.2 Distribution and Transmission
The distribution of energy in Germany is divided according to voltage levels. The nationwide spanning transmission grid is operated at up to 380kV and supplies local distribution grids with energy from power stations located across the country. It has a total length of 35.000 km and connects the German electric grid to those of its neighbouring countries, thus enabling cross-border energy exchanges throughout Europe. Responsible for operating the transmission grid are the four TSOs. They divide the country in four Regelzonen or zones. 5
To connect households and residential consumers, the voltage is reduced and further distributed. The DSO maintains a medium voltage grid at 1-30kV for regional distribution as well as a low voltage grid at 230V or 400V for residential demand. There is a total of 4.000 medium voltage grids and about 500.000 low voltage grids operated by 888 DSOs in Germany. 15
3.3 Energy Markets
The liberalization and unbundling of energy markets and grid operators lead to a disconnection of the physical grid and virtual markets. These markets function as economic marketplaces to match supply and demand. The TSOs are responsible for administrating markets and guaranteeing access. Currently the most important markets are national and international markets hosted by the four German TSOs. The BNetzA monitors the markets mechanisms and utilizes incentive schemes to preserve grid stability, extension and maintains investments as well as a fair competition. 14
3.3.1 Energy-Only-Markets
At Energy-Only-Markets (EOM) 17, participants can purchase and sell energy either on long term agreements, over-the-counter (OTC), day-a-head contracts or intraday trades. The most substantial amount of energy is traded in this manner. Energy suppliers can hedge their portfolio against expected demands and gain revenues from precise predictions. All different sorts of energy sources are available, from brown coal, nuclear to RES like wind, solar and hydropower. To determine the current clearing price, the so called MeritOrder-Model is applied. In simplification, all primary energy sources are sorted ascending according to their marginal costs of generation. Starting from the very cheapest offer, the current total market demand will be satisfied by adding up until the market price is set at the last offer taken. All suppliers within this range are remunerated according this equilibrium price. 14
3.3.2 Capacity Markets
The limited storing ability of electrical energy and demand forecasting deviations make it necessary to retain a certain amount of capacity in case of exceeding demand. Demand can deviate positively or negatively from predictions and result in severe frequency changes (see 3.4.1). Thus, in contrast to EOM, capacity markets offer the possibility to trade and purchase the ability to produce or consume energy on demand. Power generators that can increase or decrease their production can offer this flexibility (see 4) at a capacity market to TSOs. So called Capacity Remuneration Mechanism (CRM) determine the price for retained capacity. The suppliers actually executing the offered capacity are remunerated additionally. Similar to the Merit-Order-Model implemented at EOMs, in case of grid imbalances, the suppliers are contracted ascending to the offered execution price. 17
3.3.3 Contractual Balancing
Due to the nature of electric energy, bilateral transactions are limited and TSOs operate so called Balancing Group (BG) to monitor contract fulfillment amongst market participants within their area of control. The administrator of a group is called Balance Responsible Party (BRP). This role is mostly executed by an energy provider. Within its BG, the BRP aggregates all contracted clients in the respective area of control. The BRP purchases energy at the EOM to supply its consumers according to their demand. As the name suggests, each BG has the same amount of positive feed-in energy and negative consumption to establish a constant equilibrium. In reality, this equilibrium is hardly achieved and balancing mechanism must be applied by the TSO. 14 16
3.3.4 Auctions in Competitive Electricity Pools
”As electricity is a completely homogeneous good and produced by a small number of firms, [..] power markets have become a major field of applied auction analysis.” 18
As Schwenen (2015) states, auctions are the preferred price- and matchmaking mechanisms in EOMs and capacity markets. Previously all auctions had been first-price auctions, therefore all bidders receive the same price for a homogeneous good like electricity. In simplification, sealed-bids are submitted before the auctioneer starts serving the highest bid first, giving them the number of units requested before serving the next until the good is finished 19. This mechanism describes the Merit-Order-Model introduced in section 3.3.1.
Nevertheless, several dysfunctions and cases of manipulation supported the implementation of a second auction design. The so called discriminatory or pay-as-bid auction rewards each bidder with the individual bid price. By today, it is not empirically proofen which design works more efficient for the case of energy markets and both frameworks are in use, with advantages for certain use-cases respectively. 20
3.4 The Problem of Balancing and Congestion
To understand the current challenges the German electric grid is facing and the necessity for innovative, decentralized technologies, it is crucial to define the problem-space carefully. The first subsection will clarify the two major physical grid threats before elaborating more on current mechanism and market designs.
3.4.1 Grid Congestions and Frequency Balance
Electrical grids differ from other distribution systems by the fact that power flows can not be steered and tracked throughout the system in the same way as a system of water pipes. Amongst others, two laws of physics describe the behaviour of electric power flows 21. First, Kirchhoff's laws states that the algebraic sum of currents in a network of conductors meeting at a point is zero. The sum of currents flowing into a node is therefore equal to the sum of currents flowing out of the node.
Abbildung in dieser Leseprobe nicht enthalten
Secondly, Ohm's law states that the current through a conductor between two points is directly proportional to the voltage across the two points. By applying changing voltage U on a node with constant impedance Z , the current I will proportionally change.
To put these two fundamentals of physics into action, a simplified example is given from Hirth et al. (2018). A triangle represents an electric grid. Edge A is connected to a power generator GA with power feed-in of 100 MW and a load of 100 MW is connected to edge C . All three sides (lines) of the triangle have equal impedance Z and electricity flows equally distributed according to Kirchhoff's law. The distribution factors are proportional to the inverse of the total impedance of a pathway 1a. Considering the same example with a capacity limit to line 3 of 60 MW and 100 MW of lines 1 and 2 respectively. With the same generation and demand structure, the grid is congested due to technical limits. This can lead to severe incidents and might cause protection devices to disconnect grid elements that again causes disturbance of frequency if generated power and load demand is imbalanced. 222324 In Germany, the transmission grid is balanced at a target value of 50Hz at all times 14. According to Hirth et al. (2018), feasible measures are complex to identify due to the interdependence of congestions, location and magnitude. 25 21
Abbildung in dieser Leseprobe nicht enthalten
Figure 1: Three-node network with constrained capacity (Hirth et al. (2018)).
3.4.2 Congestion Mechanism
In order to overcome congestion issues, the grid operator applies mechanisms according to three principles. (1) So called Network Options describe interventions on physical grid level according to §13 Abs. 1 Nr.1 Energiewirtschaftsgesetz (EnWG). Due to the principle of redundancy (n-1 criterion), the congested line of power flow, in simplification, can always be supported by a second line which connects the same generator and consumer but might have higher costs of transmission. In addition, the grid operator can cancel or delay consumption to release congestion. As an ultimate solution, physical grid expansion in frequently congested areas can be feasible. 21
The second principle (2), describes so called Market Options. This shifting of power generation is also referred to as Redispatch. According to §13 Abs. 1 Nr.2 EnWG, the grid operator is obliged to intervene in the current distribution of power generation across the country in case of a severe congestion. In simplification, if the transmission from the generation-intensive north to the high demanding south of Germany is congested, Redispatch measures switch off generation facilities in the north and ramp up capacity providers in the south. In this example, generation is shifted ”behind” the congested grid segment. Costs for Redispatch increased in 2017 and amount for e 901 million according to the Federal Network Agency. 5 21
(3) The third principle is handled by §13 Abs. 2 EnWG and §14 Erneuerbare-Energien- Gesetzt (EEG). As a last suitable option, the Feed-In Management allows grid operators to switch off RES in case of severe congestion. In Germany, this can be observed while wind turbines are switched off during windy but sunny days. Nevertheless, energy from RES is prioritized according to §14 EEG and only subject to curtailment if other principles fail. In this case, the facility operator is to be compensated. In 2017, power of 5,5 GWh was lost due to Feed-In Management on DSO and TSO level. Costs of e 609 million arose from this. 26 27
3.4.3 Balancing Mechanism
Due to forecasting errors of production and consumption patterns, imbalances between generation and demand have to be equalized. Balancing mechanisms are cascading from BG-level, to zonal-level, to the national level and ultimately to European level where supply and demand for capacity can be matched at capacity markets. First, the TSO tries to balance inequalities amongst BGs in its area of control. This virtual balance is called balancing energy (Ausgleichsenergie ) and will be payed by the BRP that caused in-parities. The cost for this balancing energy is by definition always higher than energy from EOM to not incentivize BRPs to shift sourcing preferences 16. If there is no equilibrium state reached within the TSO's zone, resources at the German control reserve market (Regelenergiemarkt ) are contracted 14. Depending on the contracting period and activation delay, this energy is categorised in primary (PLR), secondary (SLR) and tertiary control reserve (MLR) 28. In the meantime, the BRP will try to compensate losses and establish an equilibrium state within its BG by trading at intra-day EOMs or regulating feed-in and consumption that yields a lower cost.
4 Local Flexibility: A Novel Approach
The previous chapter explains the theoretical foundations of distribution and challenges arising from the adoption of RES and EVs. This section first introduces the current shift of responsibilities from TSO to DSO that now fuels research on new ways to guarantee stability, availability and affordability in times of volatile RES and power intensive EVs. Subsequently, the concept of local flexibility by EVs is discussed before an appropriate market design is illustrated.
4.1 The new importance of Distribution System Operators
Two major trends are currently sparking the discussion about the future of congestion and balancing mechanisms. Not that recent, the wide adoption of RES, even on residential level, made grid congestions happen to appear more frequently and severe 8 15 5. The medium and low voltage grid was originally designed based on the consuming loads connected. These were mostly constant and unidirectional and could be forecasted with low variance due to the repeating use by customers. This dramatically changed with the introduction of RES. Kinetic wind and solar energy depends on the presence of wind and sunlight respectively and, relative to conventional power plants, are complex to forecast. In 1990, the share of renewable energy sources was close to zero and congestions mainly occurred on transmission grids. By 2017, the installed capacity reached more than 110 GWh which translates into 52% of the total capacity in Germany 5. More than 90% of the RES facilities and 58% of all generation facilities in Germany are connected to the distribution grid. By 2035, this share will increase to 78%, making conventional production forecasts increasingly complex. 27
More recent, technological advancements gave rise to a wider adoption of EVs and joined the political debate on climate change. According to the BDEW, the federal government of Germany plans to have more than one million EVs on the road by 2020 and six million by 2030. This will impose tremendous stress on DSOs since a majority of these vehicles will be charged from the distribution grid with a substantial, accumulated charging power. 29
These figures illustrate the importance of electric distribution grids in the near future. The shift of responsibilities from TSO to DSO level, demand for increasing efforts in modernising the distribution grid and introducing new market mechanisms.
4.2 Definition of Flexibility
”Flexibility is defined as the modification of generation injection and/or consumption patterns, on an individual or aggregated level, often in reaction to an external signal, in order to provide a service within the energy system or maintain stable grid operation.” 30
Following this definition from the Union of the Energy Industry in Europe (EULECTRICS), flexibility in the energy sector is the consequence to the paradigmatic shift within the European Energy Markets from the principal of ”generation follows demand” towards a more decentralized and dynamic setup. To streamline the increasing amount of small, volatile and decentralized generation units with generally rising demands, the concept of flexibility offers a viable framework to shift supply and demand peaks to prevent grid congestions and imparity. 30
This concept is part of even bigger ambitions towards a smart grid infrastructure that sets its foundations on the ubiquity of advanced data structures and information exchange between a variety of entities. EVs represent a substantial part of the future market design. 15
4.3 The Problem of Low Demand Flexibility
The problem of grid congestion as well as balancing supply and demand that was described in section 3.4 as a rather technical disfunction, roots back to the unpredictable and inflexible consumption patterns on consumer side. Considering all consumers and connection nodes in an energy grid, current mechanisms are executed by a centralized authority to satisfy demand even during peak times. None involves the end consumer and his flexibility to adapt consumption according to current grid and market conditions. This goes back to the design of fixed, yearly supply-contracts for standard-load-profiles (SLP) and results in a high price-inelasticity in volatile demand periods. 31
Cramton et al. (2013) introduce the problem of low demand flexibility with the following statement:
”Suppose electricity markets did not suffer from demand-side flaws. In particular, suppose demand is sufficiently responsive to prices, such that the wholesale electricity market always clears. Then, the market would be perfectly reliable: If supply is scarce, the price would rise until there is enough voluntary load reduction to absorb the scarcity. Consumers would never suffer involuntary rationing.”
There is no price signal mechanism that would incentive consumers to reduce or increase demand accordingly 31. With the adoption of EVs, the daily demand during peak hours will increase and much likely lead to more congestion with no incentives to adapt behaviour 32 .
4.4 Flexibility from Electric Vehicles
Despite the current low demand flexibility and the risk of congestion due to increasing demand from EVs, the same technology implies great opportunity to release the distribution grid 11. In the past decade, research identified EV technology as a source of flexibility. Its ability to store power over a long period of time, controlled charging mechanisms and power release back to the grid, gave rise to an extensive amount of experiments. On the first sight, numbers are convincing. EVs are only utilized 4% of the time for transportation while being stationary parked for the rest. The battery is designed for frequent power fluctuations by its nature of roadway driving and cheap by the cost of capital per unit of power compared to large generators. 23
With a power grid unable to store energy and therefore constantly applying congestion and balancing mechanisms 14, this technology might tremendously change the state of energy transmission.
4.4.1 V2G
”The basic concept of V2G power is that EVs provide power to the grid while parked. The EV can be a batteryelectric vehicle, fuel cell vehicle, or a plug-in hybrid. Battery EVs can charge during low demand times and discharge when power is needed.” 23
According to Lund et al.(2015) there are various metrics defining flexibility ranging from the physical parameters of ramp magnitude, ramp frequency and response time to distinguishing flexibility power composition. V2G satisfies most of these criteria for providing grid ancillary services 33. In contrast to centralized capacity suppliers (see 3.4.3), local flexibility providers can release congestion in the distribution grid and even transmission grid with relatively small but aggregated interventions. During windy daytimes, when RES produce over-capacities, this energy can be stored in EV batteries to be consumed in peak hours. EVs must have three required elements 23: (1) a connection to the grid, (2) communication and control charging facility and (3) a bi-directional metering mechanism. An aggregator, an entity that virtually bundles more EVs together, which arbitrates between the needs of EVs and the grid, manages the process 34. The aggregator again has to fulfill three major tasks. (1) Data needs to be gathered about the state of charge of each EV as well as its capacity limits. (2) A charging and discharging schedule is derived from this data, to determine when to treat which vehicle and (3), the execution of this schedule must be forced and monitored 34. To market this capacity and turn it into revenues for the EV owner, the aggregator has several options and available capacity markets. Extensive research from the perspective of economics concludes with a wide range of different results for different markets. Kempton et al. (2005) project rather high net returns of 1700$ for participating in the regulation capacity market whereas Dalling et al. (2011) from the German Fraunhofer Institute only estimate positive returns of 180e per year from providing negative secondary regulation capacity 23 35. Naturally, providing bulk energy at EOMs does not represent a viable business case for V2G since no payments for withholding capacities are paid.
4.4.2 Current Limitations of V2G
Nevertheless, it can be questioned why V2G technology has not yet found its way to commercialisations apart from several small pilot projects (see for example: Nissan Leaf at Enervie). Literature suggests six major challanges to the integration of V2G 36 34 :
(1) Battery degeneration is still not fully understood and a higher frequency of charging cycles due to V2G can possible lead to costs that have to be offset by the reveunes. (2) Bidirectional charging is an advanced technical feature only two EV support by 2017 with little efforts from the industry to change this. (3) If capacity is contracted, the EV owner is left with less kilometer range which still has to satisfy the need for mobility. (4) There is no standard for charging facilities resulting in a fragmented market for viable business cases as well as (5) no standardized communication protocol for the highly complex procedure. (6) In addition, the regulatory framework needed is by far not in place and will take more time to be passed by legislative. Condensing the above, there are technical, regulatory and behavioural challenges to be tackled before V2G might be adopted in the future.
4.4.3 Controlled Charging
In contrast to a rather holistic V2G concept, controlled charging can be seen as a subset focusing on the issue of flexible demand side management. Several studies suggest that uncontrolled charging (dumb charging ) of an increasing amount of EVs will ultimately result in insufficient grid capacities and black outs during peak times 32. A controlled EV can therefore be considered as a flexible load that can have multiple positive effects for a smart controlled grid 37 38. (1) Load curve flattening, (2) frequency regulation and (3) voltage regulation. Wang et al. (2016) provide an extensive introduction to these aspects of controlled charging 39. The adoption of such a mechanism could ultimately accommodate a bigger amount of EVs in a particular grid segment that would have been forced to be extended 32. Additionally, when set against the current limitations to V2G, controlled charging is facing less challenges that would slow down an implementation. (1) Battery degeneration is a minor issue and only regarding the length of a single charging cycle. (2) There is no need for bidirectional charging capabilities, (3) the battery capacity can not fall below the initial state of charge and therefore not result in an unexpected condition for the EV owner. However problem (4) and (5) still represent a substantial barrier that has to be discussed in the following chapters. Regarding regulatory leeway (6), §14 EnWG describes the EV as a controllable consumer and is therefore subject to a set of rights and duties including reduced grid fees and external controllability. Nevertheless, this paragraph does not elaborate on particular remuneration mechanisms nor on possible market environments for these flexible loads.
4.4.4 Current Research to Controlled Charging
This sections points out current research to the concept of controlled charging on a more granular level.
First, the issue of (1) scheduling complexity arises. Literature proposes a range of different approaches to the problem. Alonso et al. (2014) try to solve allocation of charging capacity amongst an exceeding amount of demand via a heuristic algorithm to reduce complexity already during architectural design 40. Others propose a stochastic distributed algorithm and prove that an optimal equilibrium can be found 41. Substantial research has been conducted on multi-agent systems to simulate individual preferences of EV owners in a controlled environment 32 42. All have in common that according to their specific assumptions, an optimal solution can be found for the problem. In chapter 4.6.2, the present work will introduce a unique approach to this problem.
(2) Controllability of EV batteries implies a very delicate control of the supplied charging power 43. Nevertheless, with the currently available technology, discrete and constant charge control can be performed.
(3) As suggested by 39, the question on how to aggregate EVs before scheduling their capacity, a hierarchical approach seems to be more promising than centralizing or decentralizing coordination:
”Hierarchical coordination is regarded as a hybrid paradigms of both centralized and decentralized coordination. It commonly assumes the existence of an aggregator in a price-based mechanism, which operates as the intermediate between the smart grid and EV customers.”
Different concepts on how to compose an aggregator can be found in literature. A rather social community approach is given by 9. It illustrates a neighbourhood to be its own, independent aggregator that manages its capacity. The DENA (Deutsche Energie Agen- tur) positions more institutional entities as potential aggregators. Also energy suppliers are interested to function as intermediates 44. A very specific proposal is provided by the BNetzA, elaborating on possible aggregators that do not function as BRP within a BG and can act independent of these. This is particular of interest for the German market 45.
4.5 Local Flexibility Markets
Considering the shift of control responsibility from TSO to DSO level (4.1), the price inelasticity on residential level (4.3) and the inherent flexibility of EVs (4.4), local flexibility markets represent a possible market design to solve the respective challenges based on economic principles. At such a capacity market, local flexibility is offered to the affiliated DSO to prevent a critical grid condition for a market-based price. Naturally, these offers are small in terms of capacity and need to be aggregated. The Verband der Elektrotechnik, Elektronik und Informationstechnik (VDE) proposes the positioning of a local flexibility market ahead of national capacity and balancing offers. Therefore, in case of a local congestion, offers at the local flexibility market are evaluated and contracted punctually, before exceeding demand can be balanced at national markets. 3
Figure 2 illustrates the proposed procedure and the flexibility market ahead of centralized markets like the EEX energy exchange 46. According to the VDE, several aspects have to be considered before implementing such a market design: (1) Equal level of information, (2) a communication and monitoring framework and (3) secure authentication of market participants and activities. (1) and (2) will be discussed in the following section, as they are crucial for the proposed model, whereas possible approaches to (3) are hinted out in section 5.2.5.
[...]
-
Upload your own papers! Earn money and win an iPhone X. -
Upload your own papers! Earn money and win an iPhone X. -
Upload your own papers! Earn money and win an iPhone X. -
Upload your own papers! Earn money and win an iPhone X. -
Upload your own papers! Earn money and win an iPhone X. -
Upload your own papers! Earn money and win an iPhone X. -
Upload your own papers! Earn money and win an iPhone X. -
Upload your own papers! Earn money and win an iPhone X. -
Upload your own papers! Earn money and win an iPhone X. -
Upload your own papers! Earn money and win an iPhone X. -
Upload your own papers! Earn money and win an iPhone X. -
Upload your own papers! Earn money and win an iPhone X. -
Upload your own papers! Earn money and win an iPhone X. -
Upload your own papers! Earn money and win an iPhone X. -
Upload your own papers! Earn money and win an iPhone X.