Periodic Reporting for period 1 - RESPONDENT (Renewable Energy Sources Power FOrecasting and SyNchronisation for Smart GriD NEtworks MaNagemenT)
Berichtszeitraum: 2022-11-01 bis 2024-04-30
As sustainable and abundant renewable energy generation is highly weather-dependent and variable, RES produce significant fluctuations in the generated output power. These power fluctuations can cause an inherent integration difficulty in legacy power supply and distribution networks. Despite the advent of smart power grids, that offer more efficient and automated power distribution, metering, control and security, which contribute to solving the seamless integration of RES into the power network, issues such as synchronisation and supply/demand balancing still exist. Being able to predict the amount of power generated, according to local weather and environmental conditions contributes to the seamless integration of RES. Accurate power demand forecasting allow the power network operators to achieve efficient and balanced system operation.
The EU’s Copernicus EO Programme offers advanced capabilities for the estimation of the local weather/climate parameters, which affect the power generation performance of the climate-dependent RES. The same parameters also provide information on the consumers’ demands.
Furthermore, the seamless and efficient integration of RES into the smart grid requires the effective monitoring of the grid, which in turn rely on Phasor Measurement Units, whose function is to perform electric signal measurements with precise timing and synchronisation.
RESPONDENT addresses and mitigates these issues pertaining to seamless and efficient clean integration of RES into smart grid power networks, by providing power generation and demand forecasting, and smart grid timing and synchronisation.
The EU’s Copernicus EO Programme offers advanced capabilities for the estimation of the local weather/climate parameters, which affect the power generation performance of the climate-dependent RES. The same parameters also provide information on the consumers’ demands.
Furthermore, the seamless and efficient integration of RES into the smart grid requires the effective monitoring of the grid, which in turn rely on Phasor Measurement Units, whose function is to perform electric signal measurements with precise timing and synchronisation.
RESPONDENT addresses and mitigates these issues pertaining to seamless and efficient clean integration of RES into smart grid power networks, by providing power generation and demand forecasting, and smart grid timing and synchronisation.
The following presents work performed and main achievements per WP:
WP2 Use Cases, Requirements and Integration Planning:
The activities commenced with the identification of the potential key stakeholders that would be interested in the RESPONDENT project and its results, and who would be willing to interact with the project and its beneficiaries. Identified stakeholders were invited to a requirements elicitation workshop that was organised by the project in Spain in February 2023.
The three RESPONDENT modules were presented to the participants with related questionnaires, and the workshop was met with very constructive discussions. All the feedback was synthesised and translated into user requirements. Finally, representative use cases were compiled mapped to the ensuing piloting activities for demonstration and verification of the RESPONDENT outcomes.
Activities also incorporated the determination of the integration plan, which served as the basis upon which the RESPONDENT components would be integrated into the RESPONDENT Solution Suite. All components' interfaces and their interconnections were also defined, and the time plan for the actual integration to take place was determined.
WP3 RES Power Generation Forecasting:
The activities commenced with the specifications, design and architecture of the Power Generation Forecasting Module. The work package's main objectives were to develop two main forecasting models and algorithms, one related to weather forecasting and one related to RES power generation forecasting (PV parks), the latter incorporating the former.
The work included a detailed analysis of the climate-related parameters most suited to the forecasting algorithms and available COPERNICUS EO databases were studied, whereby the most relevant were selected, namely CAMS and ERA5. Furthemore, data from the in-situ weather stations installed at the Greek pilot site, was collected and analysed. The respective data sets from both the IoT-WS and COPERNICUS were correlated and fed into the weather algorithm, which was trained with said data.
For the power generation forecasting, the work included an extensive analysis of the most suitable algorithms to be used, and the development of the models. As the initially foreseen algorithm wasn't suited to provide the best forecasting, particularly in the case with cloud-cover, a hybrid-approach was adopted, and the PGF algorithm was developed based on this analysis.
Finally, the work also incorporated the development of the Power Generation Forecasting user dashboard, component of the PGF module itself, an interactive web interface.
WP4 Power Demand Forecasting:
The initial activities of the WP were to define the specifications, design and architecture of the Power Demand Forecasting module, which included the interfaces with the consumer's historical data providers, as well as with the weather forecasting data. Furthermore, the determination of the socio-economic data was established.
Additionally, a number of forecasting algorithms were investigated and analysed, and a pattern recognition algorithm was applied to the industrial and residential consumers, and a clustering to differentiate the consumer types was defined.
The algorithm was trained with data from the three clusters of consumer types, and improvements were made in regards to its accuracy.
With respect to the power consumption simulator and module, integration with the partners' data sources were finalised and a first set of data sent to the FINoT platform.
WP5 Smart Grid Galileo-Based Synchronisation and Monitoring:
The work commenced with an analysis of smart grid monitoring systems, and a comprehensive description of smart grid architecture was provided.
Furthermore, and algorithm to determine the optimal placement of the phasor measurement measurement units was developed and optimised specifically for ANELL's grid.
Additionally, and identification and compilation of the desired set of characteristics was conducted in relation to the PMU and Galileo receiver.
In order to test the data format necessary to represent the power grid topology, an initial data set from the Greek transmission grid was selected, as data from ANELL's power grid were being prepared. This activity incorporated the preparation of the messaging system from ANELL's grid, and integration with the FINoT platform for data exchange.
WP6 RESPONDENT Suite, Demonstration and Validation
In relation to the Greek pilot, the installation of the FINoT weather stations in the PV park in Artmida was conducted early on in order to obtain the weather-related data in a timely fashion. Furthermore, integration between the FINToT platform the PV park's inverters was also realised.
In relation to the Spanish pilot (PILOT 2), the Galileo antenna and selected PMU were purchase to be installed in ANELL's grid.
WP2 Use Cases, Requirements and Integration Planning:
The activities commenced with the identification of the potential key stakeholders that would be interested in the RESPONDENT project and its results, and who would be willing to interact with the project and its beneficiaries. Identified stakeholders were invited to a requirements elicitation workshop that was organised by the project in Spain in February 2023.
The three RESPONDENT modules were presented to the participants with related questionnaires, and the workshop was met with very constructive discussions. All the feedback was synthesised and translated into user requirements. Finally, representative use cases were compiled mapped to the ensuing piloting activities for demonstration and verification of the RESPONDENT outcomes.
Activities also incorporated the determination of the integration plan, which served as the basis upon which the RESPONDENT components would be integrated into the RESPONDENT Solution Suite. All components' interfaces and their interconnections were also defined, and the time plan for the actual integration to take place was determined.
WP3 RES Power Generation Forecasting:
The activities commenced with the specifications, design and architecture of the Power Generation Forecasting Module. The work package's main objectives were to develop two main forecasting models and algorithms, one related to weather forecasting and one related to RES power generation forecasting (PV parks), the latter incorporating the former.
The work included a detailed analysis of the climate-related parameters most suited to the forecasting algorithms and available COPERNICUS EO databases were studied, whereby the most relevant were selected, namely CAMS and ERA5. Furthemore, data from the in-situ weather stations installed at the Greek pilot site, was collected and analysed. The respective data sets from both the IoT-WS and COPERNICUS were correlated and fed into the weather algorithm, which was trained with said data.
For the power generation forecasting, the work included an extensive analysis of the most suitable algorithms to be used, and the development of the models. As the initially foreseen algorithm wasn't suited to provide the best forecasting, particularly in the case with cloud-cover, a hybrid-approach was adopted, and the PGF algorithm was developed based on this analysis.
Finally, the work also incorporated the development of the Power Generation Forecasting user dashboard, component of the PGF module itself, an interactive web interface.
WP4 Power Demand Forecasting:
The initial activities of the WP were to define the specifications, design and architecture of the Power Demand Forecasting module, which included the interfaces with the consumer's historical data providers, as well as with the weather forecasting data. Furthermore, the determination of the socio-economic data was established.
Additionally, a number of forecasting algorithms were investigated and analysed, and a pattern recognition algorithm was applied to the industrial and residential consumers, and a clustering to differentiate the consumer types was defined.
The algorithm was trained with data from the three clusters of consumer types, and improvements were made in regards to its accuracy.
With respect to the power consumption simulator and module, integration with the partners' data sources were finalised and a first set of data sent to the FINoT platform.
WP5 Smart Grid Galileo-Based Synchronisation and Monitoring:
The work commenced with an analysis of smart grid monitoring systems, and a comprehensive description of smart grid architecture was provided.
Furthermore, and algorithm to determine the optimal placement of the phasor measurement measurement units was developed and optimised specifically for ANELL's grid.
Additionally, and identification and compilation of the desired set of characteristics was conducted in relation to the PMU and Galileo receiver.
In order to test the data format necessary to represent the power grid topology, an initial data set from the Greek transmission grid was selected, as data from ANELL's power grid were being prepared. This activity incorporated the preparation of the messaging system from ANELL's grid, and integration with the FINoT platform for data exchange.
WP6 RESPONDENT Suite, Demonstration and Validation
In relation to the Greek pilot, the installation of the FINoT weather stations in the PV park in Artmida was conducted early on in order to obtain the weather-related data in a timely fashion. Furthermore, integration between the FINToT platform the PV park's inverters was also realised.
In relation to the Spanish pilot (PILOT 2), the Galileo antenna and selected PMU were purchase to be installed in ANELL's grid.
WP3:
- Weather forecasting algorithm incorporating correlated in-situ IoT WS and COPERNICUS EO data sets.
- Power forecasting algorithm based on a hybrid multiple-model combination approach, providing higher accuracy compared to the individual models.
WP4:
- Development of a rule-based classification for distinguishing different types of consumers. Rule-based classification algorithms offer several advantages, including interpretability, ease of understanding, and transparency. T
- Development of the Power Demand Forecasting models, specific for the different consumers.
- Development of a data-driven simulator for Power Demand, able to capture complex interactions, being adaptable to real new data, and easily scalable.
WP5:
- GNSS-based grid timing and synchronization enhanced with EGNSS (Galileo).
-
- Weather forecasting algorithm incorporating correlated in-situ IoT WS and COPERNICUS EO data sets.
- Power forecasting algorithm based on a hybrid multiple-model combination approach, providing higher accuracy compared to the individual models.
WP4:
- Development of a rule-based classification for distinguishing different types of consumers. Rule-based classification algorithms offer several advantages, including interpretability, ease of understanding, and transparency. T
- Development of the Power Demand Forecasting models, specific for the different consumers.
- Development of a data-driven simulator for Power Demand, able to capture complex interactions, being adaptable to real new data, and easily scalable.
WP5:
- GNSS-based grid timing and synchronization enhanced with EGNSS (Galileo).
-