Flexibility services in the electrical system Sébastien Mathieu March 02, 2016
S. Mathieu Flexibility services in the electrical system 2/50 Outline The European electrical system Flexibility in electrical system Flexibility in the day-ahead energy market the reserve market the distribution network Flexibility from heat pumps a price signal in the distribution network Conclusion
S. Mathieu Flexibility services in the electrical system 3/50 Producers, consumers and the market Electricity is a commodity which main roles are: Producers Trade Bilateral trades Consumers Trade Market operator The most common market is the day-ahead energy market.
S. Mathieu Flexibility services in the electrical system 4/50 Electricity retailers Small consumers do not buy their electricity individually but make use of the services of intermediation provided by retailers. Producers Big consumers Small consumers Market operator Retailers
S. Mathieu Flexibility services in the electrical system 5/50 The network Once traded, electricityneedstobeconveyed. Thephysicallinkis called the network. The network is divided into 2 layers: I The distribution network operated by the Distribution System Operator (DSO ). Low-voltage < 1kV & medium-voltage < 36kV I The transmission network operated by the Transmission System Operator (TSO ). High-voltage > 36kV
Figure: Example of baseline and realization. S. Mathieu Flexibility services in the electrical system 6/50 Imbalance One of the most important task of the TSO is to maintain the balance at every moment between production and consumption. This equilibrium is first ensured on the trades with the obligation for each participant to submit a planning of production and consumption called baseline. The baselines are provided by Balancing Responsible Parties (BRP ) which may regroup multiple actors. To access to the network, every user of the network needs a BRP.
Imbalance of the system: +1MW S. Mathieu Flexibility services in the electrical system 7/50 Balancing responsibilities In real-time, realizations may deviate from baselines. The TSO is responsible to compensate the overall imbalance. Due to the unbundling of the electrical system, the TSO does not own production or consumption assets and must contract and activates balancing services according to the needs of the system. BRP 2-1MW Balancing service 2MW BRP 1 +5MW Transmission Network 2MW BRP 3
S. Mathieu Flexibility services in the electrical system 8/50 Imbalance price The most expensive activated service defines the imbalance price which serves as a basis for the payment or compensation of BRPs in imbalance. Figure: Imbalance prices on June 1, 2015 of Elia, the Belgian TSO.
S. Mathieu Flexibility services in the electrical system 9/50 Outline The European electrical system Flexibility in electrical system Flexibility in the day-ahead energy market the reserve market the distribution network Flexibility from heat pumps a price signal in the distribution network Conclusion
S. Mathieu Flexibility services in the electrical system 10/50 Definition of the flexibility Flexibility in electrical system is the modification of the production or consumption in reaction to an activation signal to provide a service. e.g. a gas unit producing 25MW asked to produce 1MW more. The actor providing flexibility is called in this work a Flexibility Service Provider (FSP ). The needs for flexibility can be triggered by: I the loss of a production unit, I an unexpected over-production of a wind turbine, I a congestion in a line of the network, etc.
Changes of the electrical system The current electrical system is extremely reliable. Why looking at flexibility since everything is already working? The electrical system is changing from its traditional conception. % Renewable energies Production in distribution & Flexible thermal units Congestions and voltage issues & Control on the production % % % S. Mathieu Investments or flexibility % Flexibility of the consumption Flexibility services in the electrical system 11/50
S. Mathieu Flexibility services in the electrical system 12/50 Renewable energy curtailment Production units may change their output to provide flexibility. Flexibility Thermal unit Wind turbine % modulation % fuel - & modulation & fuel blades pitch To provide an upward modulation, a production unit runs below its maximum power. Power Installed capacity Maximum power Reserved capacity Production Time Figure: Wind turbine production plan with reserved capacity. Reserving capacity on a wind turbine is equivalent to throwing away this free energy.
S. Mathieu Flexibility services in the electrical system 13/50 Flexibility of the consumption A a large part of the consumption can be controlled. e.g. electrical heaters, air-conditioners, industrial consumption, etc. The control of loads gives rise to new challenges: I Developing solutions which considers the constraints of all processes behind the electrical consumption. e.g. getting flexibility from heat pumps. I Integrating the payback e ect coming from the modulation of energy consumption into flexibility services. If the consumption of a fridge decreases for 1h, the temperature inside the fridge rises. In the 2 sd hour, the fridge consumes more to restore the temperature. I Determining the cost to provide the flexibility.
S. Mathieu Flexibility services in the electrical system 14/50 Flexibility in distribution networks The DSO may resort to di erent options to obtain flexibility: I Using flexibility services in the distribution network. Producer / Consumer? FSP? DSO? BRP?? TSO I Changing the distribution network tari s to shift consumption.
S. Mathieu Flexibility services in the electrical system 15/50 Outline The European electrical system Flexibility in electrical system Flexibility in the day-ahead energy market the reserve market the distribution network Flexibility from heat pumps a price signal in the distribution network Conclusion
S. Mathieu Flexibility services in the electrical system 16/50 Outline The European electrical system Flexibility in electrical system Flexibility in the day-ahead energy market the reserve market the distribution network Flexibility from heat pumps a price signal in the distribution network Conclusion
S. Mathieu Flexibility services in the electrical system 17/50 Retailers and day-ahead energy market We study the system... Market operator Retailer Retailer Retailer......... using game theory, in particular congestion games. Each retailer with flexible consumption can choose its consumption in each hour to minimize the cost of buying its energy.
S. Mathieu Flexibility services in the electrical system 18/50 Shifting demand according to energy prices Changing the consumption profile... MW x t x t Price e e t Time... a ects the energy prices. O er curve x t x t Quantity
Namur S. Mathieu Flexibility services in the electrical system 19/50 Congestion game An example of congestion game is the car trip Liège-Bruxelles. The more care there is on the road, the more time it takes. At 7am,the time taken is identical for the two paths. This situation is a Nash equilibrium. Schematically we have Leuven Bxl Lg Cars
S. Mathieu Flexibility services in the electrical system 20/50 Retailers-market system as a congestion game The mapping of the retailers-market system as a congestion game is: Hour 1... Hour 24 Retailer Retailer Retailer......
S. Mathieu Flexibility services in the electrical system 21/50 Results We define and work on a particular regime of the game: the laminar flow, valid if retailers are big enough. Contributions If the Nash equilibrium is laminar, we obtain I a bound on the ratio maximum/minimum market price, I a bound on the ratio of the total system cost at the Nash equilibrium and the optimal one, I the price of flexibility, i.e. the cost of shifting energy away from a given hour.
S. Mathieu Flexibility services in the electrical system 22/50 Outline The European electrical system Flexibility in electrical system Flexibility in the day-ahead energy market the reserve market the distribution network Flexibility from heat pumps a price signal in the distribution network Conclusion
S. Mathieu Flexibility services in the electrical system 23/50 Impact of load flexibility on the reserve market What is the impact of a retailer controlling flexible consumption on the day-ahead energy market and on the secondary reserve market? Producers Balancing services Market operator Bid Reserve sizing Balance TSO Bid Retailers Balance Balancing services
S. Mathieu Flexibility services in the electrical system 24/50 Agent-based modeling Each actor is modeled individually in an agent-based system. Input Agent Output The behavior of the actors are given by optimization problems. Maximize revenues / minimize costs subject to agent s constraints. This kind of models allows to I model individually each actor and observe their reaction in the complete modeled system, I to highlight and understand unpredicted behaviors of the actors.
General view of the simulated system S. Mathieu Flexibility services in the electrical system 25/50
Results Drastic reduction of the reserve procurement costs even though the total cost increases above 4% of flexible consumption. Reserve services provided by the consumption are less e cient due to the payback e ect. S. Mathieu Flexibility services in the electrical system 26/50
S. Mathieu Flexibility services in the electrical system 27/50 Outline The European electrical system Flexibility in electrical system Flexibility in the day-ahead energy market the reserve market the distribution network Flexibility from heat pumps a price signal in the distribution network Conclusion
S. Mathieu Flexibility services in the electrical system 28/50 Active distribution network management Flexibility in the distribution network is mostly unused up to now and could solve local congestion problems. There is a need of a framework detailing the interactions needed to exchange flexibility. Producers Flexibility services DSO Flexibility services Retailers Flexibility services Imbalance TSO Imbalance
Timeline of the interaction between the agents Dynamic Ranges Global baselines Flexibility contracts Producers &Retailers TSO Baseline proposals Local baselines Flexibility o ers Settlement DSO Flexibility needs Flexibility activation requests Flexibility platform Time Figure: To each arrow corresponds an exchange of information whose source is given by the dot and the destination by the arrow head. S. Mathieu Flexibility services in the electrical system 29/50
S. Mathieu Flexibility services in the electrical system 30/50 DSIMA: an open-source testbed The testbed to evaluate interaction models is available as an open source code at the address http://www.montefiore.ulg.ac.be/~dsima Figure: Screenshot of the user interface.
S. Mathieu Flexibility services in the electrical system 31/50 Interaction models DSIMA allows comparing quantitatively di erent interaction models. In the thesis, six interaction models are studied. These interaction models are di erenciated by 1. their type of access contract to the distribution network, 2. their financial compensation of flexibility services. Table: Summary of the parameters of 4 of the studied interaction models. Interaction Access Financial model type compensation Model 1 unrestricted Model 2 restricted... Model 5 unrestricted bids & imbalance Model 6 dynamic
S. Mathieu Flexibility services in the electrical system 32/50 Results Results obtained on a 75 bus network for a representative 2025 year. Figure: Quantitative comparison of four interaction models. Protection costs reflect the cost of shedding of production or consumption due to problems in the management of the system.
S. Mathieu Flexibility services in the electrical system 33/50 Coordination problem Assume that the DSO predicts that the power flow will exceed the capacity of the line 3 by 1MW. To solve this issue, the DSO curtails a wind turbine by 1MW. At the same time, assume that the TSO asks a storage unit to inject 0.4MW. These activations lead to a remaining congestion of 0.4MW.
S. Mathieu Flexibility services in the electrical system 34/50 Outline The European electrical system Flexibility in electrical system Flexibility in the day-ahead energy market the reserve market the distribution network Flexibility from heat pumps a price signal in the distribution network Conclusion
S. Mathieu Flexibility services in the electrical system 35/50 Outline The European electrical system Flexibility in electrical system Flexibility in the day-ahead energy market the reserve market the distribution network Flexibility from heat pumps a price signal in the distribution network Conclusion
S. Mathieu Flexibility services in the electrical system 36/50 Flexibility from heat pumps Heat pump Consumer FSP This electrical energy consumed by the heat pump is converted into heat which is stored by the building and, depending on its thermal inertia, allows consumption to be shifted.
A flexibility service considering payback This flexibility service consists in a modulation in one quarter followed by a well defined and limited energy payback in the following hour. Modulation Payback Baseline Power (a) Modution signal Baseline (b) Modulation added to the baseline Time S. Mathieu Flexibility services in the electrical system 37/50
S. Mathieu Flexibility services in the electrical system 38/50 Optimization of the baseline Minimize energy cost subject to I thermal state transition model, I state constraints, I power limitations, I heat pump constraints. using an accurate thermal model of the building and the heat pump. R 4 Q t s, Q t g, Qt sol T t a R 1 T t m R 2 T t i R 3 T t f R 4 T t a C 1 C 2 C 3
S. Mathieu Flexibility services in the electrical system 39/50 Optimization of the modulation To obtain the maximum flexibility in quarter t, solve Maximize flexibility available in quarter t subject to I thermal state transition model, I state constraints, I power limitations, I heat pump constraints, I payback limited on k quarters. Accurately defining this payback is the key element that allows flexibility to be used in one quarter without risking creating other issues afterwards.
S. Mathieu Flexibility services in the electrical system 40/50 Results A modulation of 2.5 kw with a payback of 1 hour provided by space heating from a 4.3 kw heat pump. Domestic hot water counterbalances space heating to limit deviations during payback.
S. Mathieu Flexibility services in the electrical system 41/50 Outline The European electrical system Flexibility in electrical system Flexibility in the day-ahead energy market the reserve market the distribution network Flexibility from heat pumps a price signal in the distribution network Conclusion
Price signal in the distribution network The DSO may use the tari of distribution as a price signal to shift the consumption and alleviate problems in its network. Consumer FSP DSO This solution has the major advantage that the infrastructure is already in place with the system of o -peak and on-peak tari s. In Belgium, there are night-only meters that switch on electric boilers and heaters only during o -peak periods. S. Mathieu Flexibility services in the electrical system 42/50
S. Mathieu Flexibility services in the electrical system 43/50 Problem statement Find the best o -peak hours pattern by solving Minimize energy costs and photovoltaic curtailment costs subject to I distribution network model and constraints, I automatic tripping of solar panels, I a fixed number of o -peak hours, etc. Examples of o -peak patterns 0h 6h 12h 18h 21h 24h Classic Optimal
S. Mathieu Flexibility services in the electrical system 44/50 Results The test case is a sunny summer day on a 108 buses medium-voltage network with 5040 houses. 5.7% of houses are equipped with night-only meters and 30% with PV panels of 6 kw. This is therefore a good short-term solution to increase the amount of renewable energy produced in the distribution network.
S. Mathieu Flexibility services in the electrical system 45/50 Outline The European electrical system Flexibility in electrical system Flexibility in the day-ahead energy market the reserve market the distribution network Flexibility from heat pumps a price signal in the distribution network Conclusion
S. Mathieu Flexibility services in the electrical system 46/50 Summary This thesis assesses impacts of exchanging flexibility in the electrical system and analyzes the resulting complex interactions. The modeling techniques used to carry the analysis are I game theory, I agent-based modeling, I optimization. The impacts on di erent parts of the electrical system are presented: I the day-ahead energy market, I the secondary reserve, I the distribution system. Methods to obtain flexibility from the consumption are broached: I direct control of a portfolio of heat pumps, I dynamic pricing to control electric heaters and boilers.
S. Mathieu Flexibility services in the electrical system 47/50 Discussion This thesis strengthens the state of the art by formalizing the interactions needed to use flexibility in the European electrical system. This work could be continued along two major lines: I doing further academic researches: I more detailed models, I other modelization techniques, I compare investments in flexibility and in the network, I etc. I going from theory to practice.
S. Mathieu Flexibility services in the electrical system 48/50 Discussion This work could be continued along two major lines: I doing further academic researches: I going from theory to practice. Therealreadyare I many applied projects on flexibility, I an increasing number of flexible consumption devices, I algorithms able to coordinate these flexible loads. My opinion is that one important remaining step to perform is to complete the regulation of flexibility in the electrical system. Fortunately, many regulators of various countries of the European electrical system are currently writing this legislation.
S. Mathieu Flexibility services in the electrical system 49/50 Conclusion TSO DSO FSP BRP Producer Big consumer Small consumer Market operator Retailer
Further information in the thesis: S. Mathieu, Flexibility services in the electrical system, PhD thesis, University of Liège, 2016 and in the corresponding papers: I S. Mathieu, Q. Louveaux, D. Ernst, et al., Dsima: A testbed for the quantitative analysis of interaction models within distribution networks, Sustainable Energy, Grids and Networks, vol.5,pp.78 93,2016,issn: 2352-4677. doi: 10.1016/j.segan.2015.11.004 I E. Georges, B. Cornélusse, D. Ernst, et al., Direct control service from residential heat pumps aggregation with specified payback, in Proceedings of the 19th Power Systems Computation Conference (PSCC), IEEE, 2016 I Q. Louveaux and S. Mathieu, Electricity markets with flexible consumption as nonatomic congestion games, in Submitted, 2016 I L. Merciadri, S. Mathieu, D. Ernst, et al., Optimal assignment of o -peak hours to lower curtailments in the distribution network, in Innovative Smart Grid Technologies Europe (ISGT EUROPE), 2014 5th IEEE/PES, IEEE, 2014 I S. Mathieu, Q. Louveaux, D. Ernst, et al., A quantitative analysis of the e ect of flexible loads on reserve markets, in Proceedings of the 18th Power Systems Computation Conference (PSCC), IEEE, 2014. S. Mathieu Flexibility services in the electrical system 50/50