Periodic Reporting for period 1 - GEiMS (GridEye State Estimation of Medium Voltage Distribution Systems)
Okres sprawozdawczy: 2018-09-01 do 2020-08-31
In GEiMS we developed a framework of State Estimation (SE) at Medium Voltage (MV) Distribution Systems (DS) of electricity. SE is a method for monitoring a power system operating and asset behaviors, in order to perform remedial actions in case of criticalities or plan for network upgrades. SE is performed by collecting synchronized measurements from various points in a grid and extracting a model (estimate) of the overall system status in as much detail (electrical variables values) as possible. The SE is expected to be executed in a Decentralized manner (DSE). The DSE framework will be integrated into GridEye Network Supervision, a fully operational monitoring and control ICT platform. GridEye Base Module (GBM) and the GridEye Network Supervision have been developed by DEPsys, a critically acclaimed Swiss technology company in the energy market field. DEPsys served as the Host Institution (HI) of this project.
SE at DS is important for society for a multitude of reasons. First of all, Renewable Energy Sources (RES) have been widely promoted and, thus, installed at DS. Despite their generally positive impact, most RES are characterized by volatility, which causes wear of infrastructure owned by the network operator to handle such volatility and power quality issues. Secondly, although Distributed Generation (DG) postpones investments in system reinforcement, it may not be a guaranteed outcome, while it is also unclear for how long and/or for which parts of a DS without specifically monitoring the effects of such deployments. Thirdly, employing SE at DS will increase system reliability, since it will improve handling of faults by locating them and responding to them more efficiently. Fourthly, DS reconfiguration will also be greatly favored thanks to SE at this level. The SE will facilitate the DS operator in monitoring the conditions which dictate the need to reconfigure a MV system. Lastly, if SE is active on a given DS, standards that limit the greater penetration of DG (stricter criteria calculated offline and prior to the installation of such units), may be applied in a softer manner. For example, voltage deviations averaged over specific time intervals may be contained through the online control of the DG units that cause them.
The main objectives associated with the desired characteristics of the framework are as following:
O1. Accurate calculation of MV phasors on distribution transformers (T/Fs) with low voltage (LV) side measurements by GBM devices.
O2. Realistic assessment of DSO metering infrastructure and aims of its use.
O3. Decentralized SE method for DS.
O4. Guidelines for the implementation of the SE method for DS.
O5. MV DS characteristics in the SE methodology.
O6. Performance of the proposed DSE of a DS on an actual system.
SE at DS is important for society for a multitude of reasons. First of all, Renewable Energy Sources (RES) have been widely promoted and, thus, installed at DS. Despite their generally positive impact, most RES are characterized by volatility, which causes wear of infrastructure owned by the network operator to handle such volatility and power quality issues. Secondly, although Distributed Generation (DG) postpones investments in system reinforcement, it may not be a guaranteed outcome, while it is also unclear for how long and/or for which parts of a DS without specifically monitoring the effects of such deployments. Thirdly, employing SE at DS will increase system reliability, since it will improve handling of faults by locating them and responding to them more efficiently. Fourthly, DS reconfiguration will also be greatly favored thanks to SE at this level. The SE will facilitate the DS operator in monitoring the conditions which dictate the need to reconfigure a MV system. Lastly, if SE is active on a given DS, standards that limit the greater penetration of DG (stricter criteria calculated offline and prior to the installation of such units), may be applied in a softer manner. For example, voltage deviations averaged over specific time intervals may be contained through the online control of the DG units that cause them.
The main objectives associated with the desired characteristics of the framework are as following:
O1. Accurate calculation of MV phasors on distribution transformers (T/Fs) with low voltage (LV) side measurements by GBM devices.
O2. Realistic assessment of DSO metering infrastructure and aims of its use.
O3. Decentralized SE method for DS.
O4. Guidelines for the implementation of the SE method for DS.
O5. MV DS characteristics in the SE methodology.
O6. Performance of the proposed DSE of a DS on an actual system.
The most substantial outcomes of this effort are the following:
• A digital twin of a distribution transformer (T/F) was developed. The digital twin uses real-time waveform measurements of the low-voltage (LV) side of the T/F and calculates with high accuracy the waveforms of the MV side of said T/F. This method does not require costly MV measurement devices or disruption of the grid to deploy it. It also enables the SE of the MV DS by limiting the direct MV measurement requirements.
• A survey of utilities and distribution system operators (DSOs) was conducted. The sheer numbers of DS feeders imply that the monitoring infrastructure to implement DSE will be immense and, thus, require prohibitive capital expenses. The survey shed light on what are the DSOs’ and utilities’ priorities and helped shape a DSE that handles these challenges.
• A DSE for MV distribution grids was developed. The DSE relies on a very limited extent of measurements, while it also avoids any requirements in pseudomeasurements extracted by historical data or forecasting tools. The proposed DSE estimates the operating status of the DS grid with high accuracy in voltage calculations and alerts for line congestion.
With regard to the exploitation and dissemination actions, we have published two conference papers, one journal paper, while one other of our planned conference papers was invited as a tutorial to the North American Synchrophasor Initiative (NASPI). We are also currently drafting an additional journal and one conference paper with our work on the SE tools. With regards to the exploitation of the results, the HI has already deployed the digital twin of distribution T/F for testing, and measures are taken to deploy for testing the DS SE framework. The HI and ER discuss the possibilities to continue their collaborations in the framework of exploitation of results specifically.
• A digital twin of a distribution transformer (T/F) was developed. The digital twin uses real-time waveform measurements of the low-voltage (LV) side of the T/F and calculates with high accuracy the waveforms of the MV side of said T/F. This method does not require costly MV measurement devices or disruption of the grid to deploy it. It also enables the SE of the MV DS by limiting the direct MV measurement requirements.
• A survey of utilities and distribution system operators (DSOs) was conducted. The sheer numbers of DS feeders imply that the monitoring infrastructure to implement DSE will be immense and, thus, require prohibitive capital expenses. The survey shed light on what are the DSOs’ and utilities’ priorities and helped shape a DSE that handles these challenges.
• A DSE for MV distribution grids was developed. The DSE relies on a very limited extent of measurements, while it also avoids any requirements in pseudomeasurements extracted by historical data or forecasting tools. The proposed DSE estimates the operating status of the DS grid with high accuracy in voltage calculations and alerts for line congestion.
With regard to the exploitation and dissemination actions, we have published two conference papers, one journal paper, while one other of our planned conference papers was invited as a tutorial to the North American Synchrophasor Initiative (NASPI). We are also currently drafting an additional journal and one conference paper with our work on the SE tools. With regards to the exploitation of the results, the HI has already deployed the digital twin of distribution T/F for testing, and measures are taken to deploy for testing the DS SE framework. The HI and ER discuss the possibilities to continue their collaborations in the framework of exploitation of results specifically.
With regards to the substantial outcomes of this action, we summarize the main aspects of progress beyond the state of the art and the plans forward.
With our work on the digital twin of a distribution system (DS) transformer (T/F) we contributed in the scoped field as follows:
- The waveform monitoring allows to determine power quality in real time, as it captures all harmonics content,
- Medium Voltage (MV) DS behavior under faults can be captured fully to alert the system operator and logged for further analysis,
- The MV-side waveforms outputted by the digital twin of the T/F are as accurate as the measurements of an instrument T/F (which are a particularly constly topology used to the day) on the MV side of the actual T/F,
- Technical personnel and system operator can assess immediately any remedial actions to system events,
- Installation of MV side measurement devices requires the MV network to be interrupted under fewer circumstances, thus, enabling a comparably seamless deployment.
From our work with the surveys of Distribution System Operators (DSOs) in Europe we concluded that:
• There is no monitoring for reliability or reinforcement purposes and, typically, the respective concerns are handled with reports and/or according to experience,
• DSOs have immense interests in reducing average outage duration for each customer served (the performance metric also known as SAIDI),
• Monitoring of DS is sought for to also accommodate distributed generation and storage,
• Many DSOs have used (to some extent) SCADA in monitoring distribution systems, which implies there are legacy effects from what is practiced in transmission systems,
• Gathering monitoring data and an easy-to-use interface are important features of the monitoring infrastructure DSOs plan to design and deploy,
• The low voltage grid specifically is very sparsely if not at all monitored, although most customers are equipped with advanced monitoring infrastructure in the form of smart meters.
Our proposal on the DS SE we achieve the following:
• It relies on a scarce measurement infrastructure, reducing, thus, the concerns for wide and costly monitoring device deployments,
• It is based on a linearized power flow formulation, to avoid processing requirements that contradict the aspirations for decentralized implementation.
• Pseudomeasurements are not required for the SE (as historical data that are needed to generate them, may not be considered readily available), thus, making it a self-contained method.
• The method answers the DSO expectations about congestion and power quality monitoring.
With our work on the digital twin of a distribution system (DS) transformer (T/F) we contributed in the scoped field as follows:
- The waveform monitoring allows to determine power quality in real time, as it captures all harmonics content,
- Medium Voltage (MV) DS behavior under faults can be captured fully to alert the system operator and logged for further analysis,
- The MV-side waveforms outputted by the digital twin of the T/F are as accurate as the measurements of an instrument T/F (which are a particularly constly topology used to the day) on the MV side of the actual T/F,
- Technical personnel and system operator can assess immediately any remedial actions to system events,
- Installation of MV side measurement devices requires the MV network to be interrupted under fewer circumstances, thus, enabling a comparably seamless deployment.
From our work with the surveys of Distribution System Operators (DSOs) in Europe we concluded that:
• There is no monitoring for reliability or reinforcement purposes and, typically, the respective concerns are handled with reports and/or according to experience,
• DSOs have immense interests in reducing average outage duration for each customer served (the performance metric also known as SAIDI),
• Monitoring of DS is sought for to also accommodate distributed generation and storage,
• Many DSOs have used (to some extent) SCADA in monitoring distribution systems, which implies there are legacy effects from what is practiced in transmission systems,
• Gathering monitoring data and an easy-to-use interface are important features of the monitoring infrastructure DSOs plan to design and deploy,
• The low voltage grid specifically is very sparsely if not at all monitored, although most customers are equipped with advanced monitoring infrastructure in the form of smart meters.
Our proposal on the DS SE we achieve the following:
• It relies on a scarce measurement infrastructure, reducing, thus, the concerns for wide and costly monitoring device deployments,
• It is based on a linearized power flow formulation, to avoid processing requirements that contradict the aspirations for decentralized implementation.
• Pseudomeasurements are not required for the SE (as historical data that are needed to generate them, may not be considered readily available), thus, making it a self-contained method.
• The method answers the DSO expectations about congestion and power quality monitoring.