he transformation of German energy system is taking place primarily in the distribution grid and is manifesting itself in various dimensions: More renewable energy sources are being added and there is an increased electrification on the consumption side. Additionally, part of the infrastructure itself faces a change regarding the increasing use of measuring points in various equipment and devices. All in all, developments are increasing the complexity of distribution networks and the requirement to implement the simulation of todays and future network load.
The paper describes an approach to generate and adjust parameters of a load flow calculation to fit the measurement in a real medium-voltage area. The simulation results indicate the future grid load and can be clustered on the base of deployed patterns.
The simulative reproduction of an existing network and its load require the use of available data sources. Public sources, anonymized data of the grid operator as well as already existing measurement data can be integrated. Some of these data are fragmentary or processed by systematic errors e.g. non-consideration of own consumption of PV rooftop plants and requires an adjustment. Despite the integration of various data sources, some relevant parameters are not available. This includes in particular the reactive power behaviour of the decentralised components. A sensitivity analysis is part of the method and considers the parameters which have to be changed in an interactive process to reduce the deviation between simulation and measurement. The applied parameter fitting allows the simulation of a real distribution network with a grid behavior close to measurement data.
Using the calibrate simulation as a starting point, various scenarios regarding the penetration of decentral components are integrated and the future grid load can be determined. The occurring grid congestions are examined by the use of a k-means cluster analysis to identify four typical patterns of grid congestion and their occurrence.
The processing of PV own consumption data results in an increase in consumption by all households in the grid area of approx. 0.2%. Despite the low value, this adjustment has a greater impact, especially in the local context, and can be even higher in other grid areas. By fitting the reactive power parameters, the quadratic deviation between simulation and measurement data series could be reduced by 98 %. For the cluster analysis, the previous correlation analysis shows that six characteristics are suitable to describe the grid congestion: flexibility demand, residual load, number of congestions, hour of the day, time step, PV generation and global radiation. Three of four resulting clusters, representing 98 % of all values, are caused by PV – even in scenarios with high penetration of electric loads. Therefore, the renewable generation remains the dominating factor for grid load.
South African electricity consumers are increasingly considering self-supply options as the price of grid supply increases and that of distributed technologies like solar PV and battery storage significantly reduces. This opens up the market and begins to enable the installation of Small-Scale Embedded Generation (SSEG). In South Africa, SSEG includes generators under 1 MW connected to distribution networks. With the bulk of SSEG falling within municipal networks, municipal licensees have been distributing electricity for many years but few have capacity, processes and policies in place to enable the structured connection and management of prospective SSEG adopters. The CSIR is implementing a work program focused on the municipal grid impacts of SSEG integration. Three municipalities have been partnered with as part of a pilot phase to provide support on assessing grid impacts of SSEG. This includes an initial gap analysis process looking at business processes for SSEG, personnel and skills capacity, availability of data and application of the most appropriate tools. This forms the basis for customised support for each municipality. This paper focusses on the outcomes of this gap analysis as well as learnings and recommendations, enabling further sharing for other South African municipalities and international distributor licensees, regulators and national institutions in jurisdictions where similar challenges may be experienced.
An economic evaluation tool for solar power self-consumption system with battery storage was implemented. This tool calculates output power of solar system and battery storage system to evaluate reduced purchased electric power by installing self-consumption system. System control methods can also be evaluated with this tool to confirm the value it claims. As a result, indexes like Return of Investment and self-consumption rate can be calculated according to system conditions and system control candidates. It is expected that a balanced system design can be obtained, in the phase of system planning, by using this tool to address concerns of equipment investor and solution provider. In this paper, several evaluation results were presented as use cases of this tool, where a self-consumption system with system control methods was tested for a system of 250 kW load consumption.
Since more and more large conventional power plants, based on direct coupled synchronous generators (SGs), will be phased out in the future, frequency stability must be guaranteed. Consequently, grid-forming inverters are needed to ensure the system stability of the future grids. In the joint research project “VerbundnetzStabil”, conducted by Fraunhofer ISE and its partners, the stability of a grid system with a high penetration of inverter-based renewable energy sources (RES) is investigated. This paper aims to develop a droop control concept of grid-forming inverters that can stabilize the system under all future grid scenarios (e.g. grid systems can be split into sub-grids with up to 100% RES and high-power transmission). The droop control structure is presented in both synchronous and stationary frames. Moreover, the droop control scheme is modelled and simulated in DIgSILENT PowerFactory with different test cases.
An microgrids basically involves distributed generations (DGs) and loads that can operate autonomously in islanded mode or collaboratively work with the main grid in grid-connected mode. As a result of the increasing microgrids level, the future grid's characteristics, including voltage and frequency will be predominately controlled by the inverters, which are generally classified into two basic types: grid-feeding inverter, as a current source, and grid-forming inverter, as a voltage source. The objective of grid-forming control for voltage source inverters is to ensure the stability of the voltage and frequency of the grid. However, due to the lack of physical inertia, those inverters cannot stabilize the grid during disturbances. To address this issue, grid-forming inverters are integrated with droop control to mimic synchronous generators' behavior in a conventional large-scale network . In this circumstance, the grid-forming inverters will have abilities to regulate voltage and frequency according to active power and reactive power demands. Also, the output current of the inverters should be limited in order to prevent damage to all of the components in case of grid faults.
In this paper, droop control schemes have been reviewed and evaluated throughout various test cases in PowerFactory. The controllers enable synchronization to an existing network or other voltage sources as well as load sharing between different generators in islanded mode. At the lowest control level, voltage regulation is implemented, which has a special current limiting function to limit the output current in the event of errors such as short circuits or overload situations. With droop control, grid-forming inverters are proved to be able to replace the properties of the discontinued synchronous generators.
Simulation shows that PI controllers on synchronous frame and PR controllers on stationary frame have similar performances under fundamental changes of the grid. The PR technique requires no transformations and less computation, that makes the controller schemes more simple and easier for implementation. However, the current method applied for the PR controller has limitation of eliminating ripples on the current magnitude. Therefore, a better method for the PR controller should be studied in future work. Whereas, the presented current limitation method based on the vector amplitude limitation shows an excellent performance for PI controller in synchronous frame.
Battery Energy Storage Systems (BESS) are more and more competitive due to their increasing performances and decreasing costs. Although certain battery storage technologies may be mature and reliable from a technological perspective, with further cost reductions expected, the economic concern of battery systems is still a major barrier to be overcome before BESS can be fully utilised as a mainstream storage solution in the energy sector. Since the investment costs for deploying BESS are significant, one of the most crucial issues is to optimally size the battery system to balance the trade-off between using BESS to improve energy system performance and to achieve profitable investment. Determining the optimal BESS size for a specific application is a complex task because it relies on many factors, depending on the application itself, on the technical characteristics of the battery system and on the business model framework.
This paper describes a generic simulation-based analytical method which has been developed to determine the BESS optimal size by taking into account both the application and the storage performance over its lifetime. Its implementation and the associated results are presented for two different BESS use cases: a smoothing and peak shaving application for PV injection and an off-grid hybrid microgrid case. In order to provide a better understanding of the most influencing drivers to consider during a BESS sizing procedure, several sensitivity analyses have been carried out on these two illustrative cases. The use of comparative scenarios led to quantify the degree of impact on optimal sizing results of several factors among the following topics: control strategy, forecast quality, degradation of battery performance due to ageing, precision of technical modelling.
The paper finally illustrates how BESS designers can take advantage of these study results at the stage of modelling or collecting data for sizing purposes. The conclusions should help them to:
Ringolab is a grid forming demonstrator to be built by Ingeteam and installed in the RTE network on September 2020 within the framework of the Work Package 3 (WP 3) of the EU-funded OSMOSE project. The main objective of the WP3 is to show the technical feasibility and economical viability of providing grid-forming function with commercially available energy storage systems (ESS). For this purpose, 2 utility-scale demonstrators are considered: 1) an existing 720 kVA/560 kWh Lithium-Titanate (LTO) battery installed at the EPFL campus already used for experimental activities [1], and 2) a one-Megawatt grid-forming inverter backed up by an hybrid energy storage system (HESS) composed of a 500kW - 1 hour lithium-ion battery and 1MW - 10 seconds ultra-capacitances. This paper focuses on the latter and it aims to share the experience gained on this technology during the factory acceptance test (FAT) that will be conducted at the end of July 2020 in the Ingeteam Power laboratory facilities in Zamudio, Spain.
The demonstrator is fully described in our previous work [2] where we showed, through simulations, stable association of the grid-forming control proposed in the MIGRATE project [3] with different DC side power sharing and energy management strategies. Facilitated by an active parallel hybrid topology, the proposed control makes the ultra-capacitances deliver the power during to the fast transients induced by the grid forming function smoothing the battery power output. Moreover, we have added a negative sequence (NS) control loop, including a NS threshold virtual impedance (TVI), which has proven to be effective in limiting overcurrents under unbalanced conditions including asymmetrical faults [4]. The ratio between NS and PS current during unbalance events depends on control settings.
A preliminary performance assessment of the demonstrator has already brought some light on the needed updates in the Connection Network Codes (CNC) to make regulation and grid forming behaviour compatible. Indeed, requirements for converter-interfaced storage plants are likely be aligned with other CNCs, such as Requirements for Generators (RfG) and High Voltage Direct Current (HVDC) [5], in which grid-support specifications have been conceived for grid following converter. In this paper we will present experimental results on the power in the loop test bench to provide recommendations and highlight remaining challenges requiring more research effort.
The FAT tests will bring knowledge on reference tracking and disturbance rejection, including faults, phase jumps and steady state grid frequency deviation, but also on synchronisation and behaviour under permanent unbalanced and distorted conditions (5th and 7th harmonics)
Abstract
This article presents the contribution of the renewable energies to the energy mix in the network transmission of Senegal, western Africa. The focus will be on the capacity of this network to absorb a photovoltaic generation to determine the maximal penetration rate. The approach has been applied to the electricity network of SENELEC (National Company of Electricity of Senegal), which presents good opportunities to host photovoltaic energy production.
Motivation
Within the last few years the installed energy power of renewable energy – mainly photovoltaic energy – has been increase to approximately 20 % of the installed capacity. Senegal is still heading an increase of energy demand, so that further increase of power generation is needed as well as grid expansion. Further increase of photovoltaic energy and other fluctuating energy such as wind in large scale will cause instabilities and limits to the national grid. On the grid integration workshop in 2018 – Stockholm – the need of further storages as balancing power has already discussed.
Results
This presentation will focus on simulations, which investigate grid restriction due to electrical and physical constraints of the national power grid of Senegal (SENELEC). Based on these results, grid simulations have been carried out to investigate the impact of the expansion of photovoltaic energy production. Different allocations for addition photovoltaic parks have been investigated as well as additional storages. Simulating various scenarios, grid transmission was optimized to allow a maximum of photovoltaic energy rate.
Approach
The network transmission and its various components were modeled using the PowerFactory software. Various characteristic parameters of the stability of the grid were measured and allowed to work out the maximum rate in photovoltaic energy. The simulations allow to determine the functional limits of the electricity grid under the given conditions. The result will be shown in the full paper. Further simulations and a comparison to empirical data of the SENELEC are planned.
Literature (extract)
Scenarios about the development of energy supply in Senegal with a focus on wind energy, Masterthesis, Jonne Finn Bibis, HAW Hamburg, Aug. 2017
Data analysis and simulation about the benefit of energy storage with existent photoaoltaic parks to the Senegalese elecricity grid, Master Project Tobias Schröder, HAW Hamburg, july 2018
Investigation of balancing power demand in the roll-out scenario of solar energy in Senegal, Birgit Koeppen, Anna Usbeck, et al., et al., Balancing Power Demand , 8th international workshop on the integration of solar power into power systems, Stockholm, Sweden
Determining the maximal penetration rate in photovoltaic power: case of SENELEC network in Senegal, Maguette Sarr et al., IC Smart-Grid, 2020
Optimization of the penetration rate in photovaoltaic power: case of the Senegalese electricity network, Maugette Sarr,Lamime Thiaw et al., 2020
Off-grid and decentralized Hybrid Renewable Electricity Systems (OHRES) offer a promising solution for addressing global energy poverty as they provide sustainable and reliable energy, and are adaptable to many geographical contexts. Building scalable solutions to alleviate global energy poverty requires a deeper understanding of OHRES and the factors for their success. This study focuses on the development and optimization of the Hybrid Off-grid and decentralized systems Techno-Economic Model (HOTEM), used to understand the technical and economic factors affecting the feasibility, sustainability, and reliability of OHRES. This model development is part of a comprehensive study methodology that aims to increase the level of understanding, acceptance, and deployment of such systems on a global level. The paper also covers the OHRES technical layout including an integrated 48V DC Hybrid lead-acid & lithium ion Battery Storage System (HBSS).
HOTEM is an OHRES sizing and techno-economic assessment tool. The model can perform feasibility analysis for electrical energy production, in order to obtain the best configuration and sizing of the hybrid system at the off-grid or decentralized location. The development of HOTEM is based on the on-ground experience gained from the deployment of two OHRES in highly contrasting case studies in Uganda and Canada, which allows HOTEM to be a practical tool, reflecting the practical market needs. The presented model includes optimized technical and economical sub-models, covering the technical modeling of Solar PV, Wind, Genset, battery storage, and load profiling. Besides, the calculation of the dynamic economic indicator, using detailed economical modeling through the newly integrated Weighted Average Cost of Capital (WACC) model. HOTEM performance is analyzed using the OHRES case study in Canada and benchmarked against an existing commercial software, showing that HOTEM can be successfully used in the planning and optimization of off-grid hybrid systems.
Australia is the global renewable energy pathfinder. Australia is deploying new renewable energy 10 times faster per capita than the global average and 4 times faster per capita than in Europe, China, Japan or the USA. Data from the International Renewable Energy Agency IRENA shows that Australia has the:
It is instructive to examine facts on the ground in Australia, where theory of integration of PV/wind gives way to actual practice.In Australia, the economic advantage of PV and wind over fossil fuels is compelling.
Renewable electricity (mostly PV and wind with support from existing hydro) has now passed gas and brown coal in the National Electricity Market (NEM) and on current trends will pass black coal in 2023.
The Australian experience is highly replicable in other countries. Australia is a pathfinder for the three quarters of humanity who live in the sunbelt (lower than 35 degrees of latitude) where there is high and consistent solar insolation and no cold winters. This is where most of the growth in population, energy use and Greenhouse emissions are occurring. Australia’s rapid deployment of PV and wind is an important reference point for other countries to also adopt this path and thereby reduce electricity costs and mitigate large amounts of future emissions.
Australia’s overall Greenhouse emissions are falling. Australia (popultion = 25 million) is tracking towards deep emissions reductions through the deployment of 6 Gigawatts per year of new PV and wind.
Deployment of more solar PV and wind is correlated with reducing electricity prices. PV and wind allow Australia to reduce emissions AND prices.
Balancing 50-100% variable PV and wind is straightforward using off-the-shelf techniques: stronger long distance transmission (to smooth out local weather), pumped hydro storage (without new dams on rivers), battery storage and demand management. Additional systems of each is being deployed at Gigawatt scale to manage Australia’s rapidly increasing PV and wind.
A global survey of off-river pumped hydro energy storage found 100 times more sites than required to support a 100% renewable global electricity system (23 million GWh). In combination with batteries and demand management, this solves the problem of low-cost storage to balance 50-100% renewable electricity.
Using batteries as energy storage is a fundamental practice to promote renewable energy generation and reduce fossil energy consumptions to mitigate greenhouse gas emissions. By integrating grid-connected photovoltaics (PV) and Battery Energy Storage System (BESS) into a local energy system, the renewable energy shares and self-consumption rate of the system can be significantly increased. Especially for high power energy users, e.g., industrial plants, energy storage even provides the possibility of peak shaving to reduce the cost of electricity and network utilization. This paper proposes a general workflow to simulate and optimize the PV-BESS system used for peak shaving in an industrial energy system, which consists of a fabrication plant for electronic devices, together with an attached office building, a battery storage system, and a PV system. The paper presents an innovative method to model the electric load of the system, considering both electric and thermal demand. The electric demand is a time-series model from the actual measurement of production-consumption of various production machines and electric devices in the office. The thermal demand contains the electricity required from the heat pump for cooling and heating. The state-space model describing the thermal behavior of the office building is applied to the calculation. Based on the first simulation results, an energy management strategy is developed to control the proposed system. The management strategy is implemented in a lifetime simulation.
Based on the simulation, the optimization problem considering the economic and environmental impact has been solved to find the optimal configurations, which minimize the annual cost of the system. The Levelized Cost Of Storage (LCOS) is integrated into the objective function. Local feed-in tariffs and environmental policies are also considered in the optimization. The simulation results are assessed in both technical and economic prospects. The optimization result provides an optimal capacity configuration of the PV and BESS system. By applying a suitable size of the PV/BESS system in the proposed industrial plant, it is possible to improve both the economic profitability and the ecological welfare by reducing CO2 emissions.
Fast frequency response (FFR) of Energy Storage System (ESS) is regarded as effective way to mitigate the system frequency deviation induced by fluctuation of power generation from the increased renewable energy resources and the impact of reduced system inertia caused by decreased synchronous generators. However, existing standards which refer to technical requirements for frequency measurements seem not to be specific for the frequency measurement used for plant power output control. This paper proposes improved frequency measurement logic based on phase locked loop (PLL), and shows the results of benchmarking testing against existing power meter product complying with IEC61557-12. Also, FFR performance of ESS with either has been tested in Power Hardware-in-the-Loop (PHIL) simulation. This PHIL composes of the actual Power Plant Controller (PPC), the actual inverter machine and the real time simulator. In PHIL experiments, several system disturbance situations which are likely to happen in actual power systems have been considered as test cases, and measurement accuracy and latency of each frequency measurement logic is evaluated. Also, plant power output control performance of ESS based on predefined FFR curve has been experimentally confirmed by closed loop power control simulation with PHIL environment. In conclusion, this paper reveals the effects of frequency measurement performance on FFR performance of ESS, and proposes considerations for frequency measurement method used for FFR.
This study reports the electricity market analysis with expansion of variable renewable energy by using zonal pricing-based market model in Japan. The study constructed the electric generation and transmission model by using “PROMOD” which is a fundamental electric market simulation solution with nodal and zonal price forecasting. PROMOD performs a security constrained unit commitment and economic dispatch that is co-optimized with operating reserve requirements. The data of power plants, transmission and hourly demand are quoted from public information provided by transmission system operators and the Organization for Cross-regional Coordination of Transmission Operators, Japan. For the validation of the model the marginal cost pricing simulated by PROMOD and day-ahead market price of Japan Electric Power Exchange were compared. In light of the Japanese energy policy outlook for 2030, two scenarios were compared: the government's target scenario with 22–24% renewables penetration (64 GW solar and 10 GW wind), and a scenario with higher renewables penetration (100 GW solar and 36 GW wind). The results showed that Japanese day-ahead market price will be depressed by 1647 Yen/MWh due to the change of energy mix and renewable energy expansion on governmental target. The result also showed that day-ahead market price will be depressed by 779 Yen/MWh additionally due to the further renewable energy expansion based on higher renewables penetration scenario.
Driven by the energy transition towards renewable generation, more and more storage systems will be installed into distribution grids for grid balancing tasks like frequency or ramp rate control, energy delivery on schedule etc. Those units are operated in grid following mode (GFL) to fulfill present grid standards.
In the meantime, many rural communities provide their landscape for renewable energy installations with negligible monetary profit in many cases. During bad weather conditions (thunderstorm, heavy rain, or snowfall) such communities often suffer from local blackouts due to interrupted main power lines, even if enough local renewable energy generation potential would be available at the same time.
Operating the discussed grid connected storages in grid forming mode (GFM), provides the chance to keep power up, even in case the main power line is disconnected.
In Bordesholm (Germany) such a system has been commissioned in 2019. The system design and operational results are described in this paper.
Poor electric power quality, grid failures and short circuits affect the supply of loads and cause high costs in industry and society. Uninterruptible power supplies decouple these faults and enable fault-free operation. The paper shows that UPS applications can be enabled with parallel grid forming inverters. The decoupling qualities are examined in the event of a fault regarding voltage quality and transition time.
The current European power system is currently undergoing a major transformation, as climate goals mandate the rapid replacement by fossil fuel powered plants by renewable power sources, such as solar energy. Consequently, the grid is facing a decline in inertia, which was commonly inherently provided by fossil plants. Against this background, the potential of grid-forming inverters that provide the capability to inherently behave as a voltage source with inert capability to mitigate rapid changes in voltage and frequency in the wake of a fault by providing inertia themselves, is currently studied. However, such inverters, actively controlling voltage at their point of connection, might be more vulnerable to externally induced voltage imbalance than their grid-supporting counterparts without any further measures. On the other side, this type of control might possess the ability to mitigate local voltage imbalance at the cost of increased inverter strain. This area of conflict is investigated more closely in this paper.
Firstly, the effects of voltage imbalance on a simple grid forming inverter model that is not equipped to handle or mitigate imbalance is depicted qualitatively and evaluated on a quantitative basis in a demonstration setup. It is shown that voltage imbalance causes oscillating power output that might be amplified by the control if it does not account for negative sequence components.
Based on the findings, negative sequence components are actively filtered in the control, thus allowing the inverter to operate in a stable fashion on the grid. It is shown that the inverter has the ability to improve voltage quality at its point of coupling when operated in this manner while strain on the power oscillation is reduced.
Thirdly, the control is further modified, allowing the inverter to completely eliminate power oscillation at the cost of not improving imbalance in the grid.
The paper further discusses the implications of this degree of freedom in operating strategy on the grid operation. It is analyzed whether it is possible to mandate for grid forming inverters to actively mitigate voltage imbalance if excess capacity is available while allowing them to reduce imbalance mitigation when operating close to their power limits.
Curtailment of generation in PV installations is necessary when the grid connection capacity is exceeded and the increase in voltage threatens its safety.
The article discusses a low voltage network, consisting of three lines with uniform distribution of load and PV generation.
The originality of the research carried out is due to the fact that they cover the entire year, and all calculations and simulations were made for all 35,040 quarters. The variability of the PV generation and network load, as well as the voltage variation on the side of 15 kV were taken into account.
The conducted analyzes show that if the installed power of the PV installation does not exceed the peak power of the demand, voltage increases above the critical value practically do not occur. An increase in the installed power to a value equal to twice the peak power causes the voltage exceeding the permissible value (440 V) even for 10% of quarters. This value is unacceptable to recipients and dangerous to the reliability of the network.
Four types of possible solutions to reduce generation in PV installations were considered for the critical quarter:
The natural reduction consists in the limitations occurring as a result of the operation of overvoltage protections of inverters. Their action is uncoordinated and spontaneous. If the voltage exceeds a critical value, the installation is shut down and then reconnected. Natural reduction cannot be accepted by network operators as a way to increase its capacity
The quasi-proportional reduction consists in the individual adjustment of the power generated in each PV installation to the voltage at the point of its connection, according to the linear characteristic P(U). As the control systems are not coordinated with each other, the final voltage profile may be stochastic.
Solidarity reduction consists in reducing the generation equally in all PV installations, so that the voltage at the end of the line does not exceed the critical value. The inverters respond to the signal sent by the external control system
Sectional reduction consists in reducing the generation in installations for which the voltage exceeds a critical value. There is no reduction for installations at the beginning of the line. The inverters of parts of the installations respond to the signal sent by the external control system.
The article provides an analytical description of the presented reduction methods and their comparison. The results of the calculations lead to the conclusion that although in some quarters the reduction in generation may be very deep (e.g. 50%), on an annual basis it may be kept significantly below 10% (in relation to the total available energy production of PV).
Positive results are achieved by combining the generation curtailment with another method of increasing the network capacity - the use of Q(U) characteristics, a slight increase in the conductor cross-section, tap control. Then the amount of reduction is symbolic (1% -2%).
In this paper, a fault detection algorithm for photovoltaic systems based on artificial neural networks (ANN) is proposed. Numerous literatures can be found on the topic of PV fault detection through the implementation of artificial intelligence. The novel part of this research is the successful development, deployment and validation of a fault detection PV system using radial basis function (RBF), requiring only two parameters as the input to the ANN (solar irradiance and output power). The results obtained through the testing of the developed ANN on a PV installation of 2.2 kW capacity, provided an accuracy of 97.9%.
Photovoltaic (PV) hot-spots is a reliability problem in PV modules, where a cell or group of cells heats up significantly, dissipating rather than producing power, and resulting in a loss and further degradation for the PV modules’ performance. Therefore, in this article, we present the development of a novel machine learning-based (ML) tool to diagnose early-stage PV hot-spots. To achieve the best-fit ML structure, we compared four distinct machine learning classifiers, including decision tree (DT), support vector machine (SVM), K-nearest neighbour (KNN), and the discriminant classifiers (DC). Results confirm that the DC classifiers attains the best detection accuracy of 98%, while the least detection accuracy of 84% was observed for the decision tree. Furthermore, the examined four classifiers were also compared in terms of their performance using the confusion matrix and the receiver operating characteristics (ROC).
Purpose: This paper presents the development and implementation process of a service-oriented, economically viable business model that deals with the use of electricity from photovoltaic systems that are no longer state-subsidized. The research is funded within the context of the innovative research project “Citizen Energy Transition”, subsidized by the Ministry of Science, Research and the Arts Baden-Württemberg (funding code Kap. 1403 Titelgruppe75)
Design/Methodology/Approach: First the catalyst for the innovative business model is investigated. Then the development and implementation process for the business model is explained and the business model itself.
Findings: The end of the guaranteed feed-in tariff for the first photovoltaic systems leads to the necessity for new business models in the energy sector.
Originality/Value: In design thinking workshops and ongoing feedback loops all stakeholders were involved in the development process. After refining and iteratively adapting the business model, it can be adjusted for further requirements due to regulative influences, customer needs, and ecological possibilities