280 The Role of Smart Meters in P2P Energy Trading in the Low Voltage Grid Christina Sigl1, Alexander Faschingbauer1, Andreas Berl1, Jakub Geyer2, Rudolf Vohnout2, Miloš Prokýšek2 1. Institute for Applied Informatics, Deggendorf Institute of Technology, Freyung, Germany, email: {christina.sigl, alexander.faschingbauer, andreas.berl}@th-deg.de 2. Institute of Applied Informatics, University of South Bohemia, Ceske Budejovice, Czech Republic, email: {geyer, vohnout, prokysek}@prf.jcu.cz Abstract: There are different approaches for energy behavior. As such, to enable this system to be dynamic and trading shown in the paper. The peer-to-peer approach is flexible it needs respective agile components to be able to the most interesting one of them. Therefore, the provide production-consumption control. Thus, smart meters underlying subsystem is essential for fast and accurate (SM) are important in such a P2P network. SMs provide trading. Smart meters are one of the most important information about power consumption and distribution for parts within the infrastructure to provide high quality billing (longer time intervals as for grid monitoring). Maybe, information and services for smart contracting and also they can be used to detect problems in the power grid, too for controlling the grid. Therefore, smart meters have to [1]. Within a P2P network SM are very important. Therefore, fulfill dedicated requirements. Especially the lowest layer required technologies, protocols and data quality must be of smart metering – the smart meter itself – is considered analyzed regarding to P2P trading. Thus, a simplified here. A very basic view on data acquisition and a possible consideration of measurement accuracy is given and some architecture for a test bench are presented in this paper. gaps in the measurement chain (e.g. timing issues) are Keywords: smart grid, P2P, energy, trading, smart depicted. Additionally, a test bed, especially for analyzing meter. SMs acquisition accuracy, is introduced. Moreover, in P2P I. INTRODUCTION trading, business models and means of trustworthy evidence of the contract (for example smart contracting/block chain) Smart grids are based on smart measurement and control must exist. Hence, P2P systems, business models as well as of energy supply, transportation and demand. The intelligent energy pricing models are considered. control of all components can be seen as a virtual grid, where The Paper is organized as follows. In section II, related energy can be traded over short (low voltage grid connected work is given. Section III presents P2P energy trading models to a sub-station) and/or long distances (middle- and high and a comparison of energy prices between Germany and voltage grid). In both cases the complexity of the control and Czech Republic. Information, that SMs have to provide for trading system is very high. P2P trading, are shown in section IV. Required technologies and protocols are pointed out in section V. The significance of data quality, for P2P trading, is stated in section VI. A test bed setup for analyzing SMs is shown in section VII. Finally, the paper is concluded in section VIII. II. STATE OF THE ART P2P trading, smart contracting and micropayment are modern concepts, which are widely used nowadays. There are several pilot projects and examples dealing with this topic. Murkin et al. show an example platform for P2P energy trading using the block chain technology. This makes the trading process accurate, authenticated and secure [2]. Alvaro-Hermana et al. use electric vehicles for P2P energy Fig. 1. Grid schematic [1]. trading [3]. In general, there are different approaches of P2P Fig. 1 compares the traditional grid with a smart grid. In energy trading systems on that we will take a closer look in traditional grid architectures there are producers (utility) and this paper. consumers. Whereas in the smart grid renewable energy Matamoros et al. investigated P2P trading between two sources and storages are integrated and therefore a consumer micro grids, thus looking at central versus distributed control can take the role of a producer as well. [4]. Zhang et al. submit that communication and control In case of peer-to-peer (P2P) trading also both roles networks are very important for P2P energy trading. They (prosumer) are used by each participant, which changes their also show a future scenario of P2P energy trading [5]. For ACIT 2018, June 1-3, 2018, Ceske Budejovice, Czech Republic 281 control of the P2P trading system data form smart metering distribution costs and taxes. For bigger consumers like must be used. houses there is a system of high and low tariffs. P2P energy trading bases on timing, a reliable Fig. 2 shows pricing components of consumed energy communication layer and accurate metering. Marshall et al. (E.ON, distribution rate D01d, 2018, [12], for Germany see did some investigations/simulations about the impact of [13]) excluding fix costs (monthly fee for consumption point, accurate metering. Most commercial metering systems reserved input fee). measure net flow only in intervals of 30 seconds or 300 Distribution 30 minutes, which is a barrier for accurate accounting. Therefore, sub-second level timing is required. Economic 200 Consumption impacts through faster energy fluctuations, than the Renewable Energy Surcharge measurement intervals, lead to economic inefficiencies and 100 also possible inaccuracies through meter timing (the question Tax is stated, which time period is the best) [6]. Capodieci et al. 0 Other [7] present a hardware/software solution for energy trading CZ DE using agents which use only six time intervals per day for Fig. 2. Energy price components (excluding fix costs). trading. The hardware architecture consists of a real SM Scheduled flexible pricing which is connected to a SM gateway through a SM interface. A flexible pricing model is based on PXE trading system The data is sent to an energy trading platform. and uses intraday trades. The system is now usually operated Nonetheless, SM accuracy also depends on temperature manually or with a low level of automation. This kind of effects of SMs [8]. SM accuracy is determined considering trading system is a good base for further P2P business models ADC resolution, SINAD and THD [9]. Thus, we take a closer and more advanced solutions. The amount of traded energy is look on data quality that is required especially for P2P based on prediction of consumption and could be supported trading and to evaluate SMs a test environment is set up. by data gathering from smart buildings. For trading in PXE a III. P2P ENERGY TRADING license is required. Therefore, a mediator with license (usually an energy supplier) is an easier choice for end-users. In general, there are differences concerning the electricity market of the neighboring countries Czech Republic (CZ) and Germany. In CZ, the current electricity market can be divided into several levels or areas. At first, there is a market for trading energy between producers and suppliers operated by market operator (OTE [10]). There also exists Power Exchange Central Europe (PXE [11]). This market is powered by EEX and provides also services for end-users, especially bigger consumers as municipalities or SMEs. In Germany, there are two different business models for the electricity market. The first one is the traditional producer/consumer model. The second one can be called Fig. 3. Energy source matching [14]. prosumer model. It is based on own production and consumption. Every customer (private/company) is Linking energy production resources connected to the grid by a distribution network operator The system of linking/matching produced energy with (DSO) (e.g. Bayernwerk AG). The DSO is a direct customer consumers is based on flexible pricing and fluctuating power of the four transmission system operators (TSO) (e.g. production of renewable energy sources. The idea is to offer TENNET). Each customer has a contract with an electricity cheaper energy to consumers when there is a surplus of supply company (e.g. E.ON), which does billing. Electric energy and provide clear data about how the energy is energy is typically traded on the stock market. produced. Consumers can than prioritize from which energy In a smart grid environment, the control of consumer source and for what price they want to buy energy and move behavior and demand could be based on many things like some energy consumption tasks accordingly. This allows social influence or responsibility, but the price for energy is savings for consumers while it also helps balancing local the most important factor. Hence, the metering of energy production and consumption. consumption and production is the crucial part of all P2P P2P energy market trading systems (See TABLE 1 for overview). P2P energy market is based on a similar idea as the linking Current pricing model of energy production resources, the idea of balancing local Electricity for smaller consumers is in CZ delivered by energy production and consumption. However, in this case, energy suppliers, usually based on an end-user agreement. any consumer can also become producer (prosumer) and his For households, there are two common pricing models. For energy surplus is primary offered to other local consumers smaller consumers (mainly in high density areas with (usually for a better price). If the offer is accepted by another prefabricated houses) the price of energy is constant and local consumer, the transaction is realized (often using block calculated from several parts, mainly from the price of power, chain mechanisms). Any additional surplus (or if there is no consumer at the time) is bought by a distribution company according to the agreed pricelist. ACIT 2018, June 1-3, 2018, Ceske Budejovice, Czech Republic TABLE 1. P2P ENERGY TRADING PRINCIPLES OVERVIEW Scheduled flexible pricing Linking energy production P2P Energy market P2P Energy trading platform resources (with battery storages) Companies / Projects PRE Piclo, Vandebron, AmperMarket TransActive Grid, SonnenCommunity, Lichtblick PeerEnergyCloud Swarm Energy Objectives Dynamic pricing based on estimated Linking electricity demand and local Direct local energy trading Distributed energy trading with power electricity production energy resources reserves and grid balancing capabilities Peers Producers–Distributors–Consumers Producers–(Distributors)-Consumers Prosumers – Prosumers Prosumers – Prosumers Key Features • Price driven energy consumption • Local energy production profiles • Tokenization • P2P payments (block chain) estimates • User consumption visualization • P2P payments (block • User consumption prediction • Weather prediction • User energy resources preferences chain) • Weather prediction • Consumption prediction • Virtual grid Infrastructure Level Any Micro-grids / grid-cells Micro-grids / grid-cells Any Smart Meter / • Daily / weekly / monthly readings • Daily / weekly / monthly readings • Readings several times • Readings several times per hour Gateway Demands per hour • P2P market support (online • P2P market support communication) (online communication) • Gathering user data consumption (profile) Prosumers Control Manual / parametric consumption Manual / parametric prosumers Manual / parametric Dynamic control based on user adjustment adjustment prosumers adjustment profile, weather prediction and grid 282 (scheduled) (scheduled) (dynamic) (community) demands Benefits • Dynamic pricing for consumers • Local production and distribution • Local consumption and • Consumption and distribution • Load distribution more optimized to more optimized to production distribution more more optimized to production production • More transparent billing optimized to production • Price reduction for consumers information • Price reduction for buying from prosumers • Price reduction for adaptive consumers buying local • Higher selling price for prosumers consumers energy • Peaks shaving • Possible direct support of • Higher selling price for • Power reserves renewable energy sources prosumers • Grid balancing ACIT 2018, June 1-3, 2018, Ceske Budejovice, Czech Republic URLs • https://www.pre.cz • https://piclo.uk/ • https://lo3energy.com/ • https://sonnenbatterie.de/en/sonne • https://vandebron.nl (TransActive Grid) nbatterie • http://www.ampermarket.cz/ • http://www.peerenergyc • https://www.lichtblick.de/ loud.de • http://www.smartmpow er.com/ 283 P2P energy trading platform (with battery storages) V. TECHNOLOGY/PROTOCOLS A P2P energy trading platform consist of 3 key parts: • P2P energy market From the standardization point of view, SMs can be divided into two groups: • Virtual grid with distributed battery storages • Using proprietary communication protocols and security. • Grid operator/distributor cooperation Backward compatible with SM standard protocols such Surplus energy can be stored and used in moments of low as DLMS or PRIME Alliance. production (peak shaving) if a battery storage is added to the prosumer system. Based on the energy consumption profile • Using standardized metering devices, certified by DLMS (done by prediction), the stored energy can also be sold. With and/or PRIME Alliance. grid operator/distributor participation this can help balancing The first group is mainly intended for island or isolated even across multiple grid cells. smart grid installations, where no interactions with other smart grid domains are expected and usually leads to vendor IV. SMART METERS IN P2P TRADING lock-in situations. Most of the smart grid solutions use standardized (certified) interoperable equipment to be able to Especially in the P2P trading market a smart metering exchange information not only within the same grid solution is required, which is able to actively participate in (domain), but also outside, to other installations supporting the smart shared grid infrastructure. This system has to report the same communication languages. For P2P trading we need electricity consumption values and trends. Additionally, it only selected elements of xDLMS protocol structure protocol has to provide information for P2P trading to enable it. Such structure, avoiding those intend for energy distributor information is: purposes. • Current demand of power consumption by a household/SME VI. DATA QUALITY • Total power budget available for trading in the (micro) grid shared (sub-) infrastructure Data quality is one of the most important topics in the • Tariff and/or price per kWh of redundant power to be used context of P2P energy trading. All parts of the described • Price for using/renting the infrastructure (distribution costs) subsystem can affect data quality. This starts at the very low for such model layer, where the signals are taped from the grid lines, as it can • Length of the contract (for example one hour) and smart be seen in the upper part of Fig. 5. Also the timing of the contract block chain evidence whole system must be known. Some blocks have a non- • Amount of electricity units to be contracted deterministic timing, which can lead to problems if the • Guarantee power supply to be provided for at least fluctuation rate of supply and demand is higher than the data contacted time frame and for unit price agreed processing by the P2P trading layer. Fig. 5. Timing of Data Acquisition Process The accuracy of the measured values is another important topic. There are several parameters which affect accuracy and data quality: • Sensor type, attached to the grid lines • Temperature Fig. 4. Prosumer energy consumption profile • Data reduction (with photovoltaics and battery). • Communication protocols/gateways • Processing speed The role of SMs in such a situation is to provide control • Acquisition bandwidth (harmonics) over the implementation phase of the P2P (smart) contract. Fig. 6 shows a simulation of the measurement error, As such, the internal real time clock accuracy and resulting from lowering the acquisition bandwidth, especially synchronization should be not lower than 5 x 10-3 seconds. for reactive power in dependence of the number of harmonics Because of the nature of smart contracts, SMs used in the considered. smart contract enabled grid infrastructure (with tariff less The following assumptions were done: Line voltage Urms trading) have to be able to store enough historical values. = 230 V, line current Irms = 10 A, grid frequency f = 50 Hz, Therefore, SMs primary oriented on distributed sample rate Fs = 100 kHz. The signal of the line current is a infrastructures and based on tariff templates will not be sine wave superimposed with harmonics (damping by 100). suitable for this scenario. Smart metering solutions available The calculation is done 100 times, starting with adding on the market differ in several aspects. For our model, the 100 harmonics to the base signal and decrementing it after minimum accuracy level (we count only on transformer each calculation step. For each loop the relative error is connected SMs) must be 0.5 S (better 0.2 S) to be able to calculated, using the actual test signal in relation to the first provide a fair smart contract for parties involved even in low signal. power demand conditions. ACIT 2018, June 1-3, 2018, Ceske Budejovice, Czech Republic 284 VIII. CONCLUSION This paper is focused on differences between traditional and smart grids from the view of business and metering. We identified the important role of smart metering and its necessary parameters for different business models of a P2P trading systems. We also discussed the price of energy for end-users/consumers on Czech and German energy market, where the component of energy distribution was identified as the most important and its reduction due the innovative trading scenarios could be the crucial motivation factor for P2P trading spread. Finally, we introduced a schema of a Fig. 6. Measurement Error in Dependence of Considered Harmonics smart metering test bench, developed by Czech-Bayern cross- border laboratory of smart grid for testing different scenarios VII. SMART METER TEST SETUP of P2P consumption-production analyses and security tests. The whole data processing chain from grid to the control ACKNOWLEDGEMENT system must be well known for well-balanced P2P trading. Smart grid in rural areas and SMEs – INTERREG V Fig. 7 depicts a simple test setup for testing SMs concerning (no.144). their time delay. REFERENCES [1] A. Berl et al., "Virtual Energy Information Network: A Resilience Perspective," OVE e&i Elektrotechnik Infor- mationstechnik, vol. 130, no. 4-5, pp. 121–126, Springer, July 2013. [2] J. Murkin et al., Enabling peer-to-peer electricity trading. Fig. 7. Architecture of a SM test setup. [3] R. Alvaro-Hermana et al., "Peer to Peer Energy Trading with Electric Vehicles," IEEE Intelligent Transportation The grid signals are fed into a Software Defined Radio Systems Magazine, vol. 8, no. 3, pp. 33-44, 2016. (SDR) through a voltage/current tap. The type of the current [4] J. Matamoros, D. Gregoratti and M. Dohler, "Microgrids tap is a Hall sensor with higher frequency bandwidth energy trading in islanding mode," in 2012 IEEE Third compared to current transformers. Both signals (voltage and International Conference on Smart Grid Communica- current) are applied to the SDR through filters, amplifiers, tions (SmartGridComm), 2012. attenuators and impedance matching circuitry. The SDR is [5] C. Zhang et al., "Review of Existing Peer-to-Peer directly connected to a computer for controlling and Energy Trading Projects," Energy Procedia, vol. 105, processing the incoming data. The used SDR provides pp. 2563-2568, May 2017. extremely high flexibility through the possibility of [6] L. Marshall et al., "Allocation Rules and Meter Timing reconfiguration. The SM is also connected to the grid lines Issues in Local Energy or Peer to Peer Networks". and considered as a black box, because the signal [7] N. Capodieci et al., "Smart Meter Aware Domestic conditioning and processing is not accessible from the Energy Trading Agents," Jan. 2011. outside. A digital interface allows only access on the [8] P. Wegierek and M. Konarski, "The temperature effect measurement results. Both acquisition systems are connected on measurement accuracy of the smart electricity to the same computer to get synchronized datasets, for easier meter," Poland, 2016. post processing and comparison. [9] DigitKey, Accurate Power Measurement in Smart Meters Part I. [10] "OTE," 2018. [Online]. Available: http://www.ote-cr.cz/. [11] "Power Exchange Central Europe," [Online]. Available: https://www.pxe.cz/. [Accessed 2018]. [12] "E.ON pricelist 2018," [Online]. Available: https://www.eon.cz/-a119343---WPHOMJe-/cenik- komplet-elektrina-i-k-1-11-2017-distribucni-uzemi-eon- Fig. 8. Test Setup. distribuce-2018-pdf. [Accessed 2018]. [13] D. J. Heidjann, "Electricity price composition (original Fig. 8 shows the test setup in detail. The device under test title: Strompreis Zusammensetzung," 2017. [Online]. (DUT) is a SM. DUT and SDR are both connected to a Available: https://www.stromauskunft.de/strompreise/ programmable pattern generator. Test signals, like a sine strompreis-zusammensetzung/. [Accessed 2018]. wave with a frequency of 50 Hz, superimposed by a variation [14] OpenUtility, "Piclo whitepaper," 2016. [Online]. of distortion signals of different speed, are applied. The [15] RiberaSolutions, "IoT and Smart Grid," [Online]. distortion signals frequency is varied until the SM can’t Available: www.riberasolutions.com/iot-and-smart-grid. detect any further changes. [Accessed 2018]. ACIT 2018, June 1-3, 2018, Ceske Budejovice, Czech Republic