Lessons for Data-Driven Modelling from Harmonics in the Norwegian Grid
With the advancing integration of fluctuating renewables, a more dynamic demand-side, and a grid running closer to its operational limits, future power system operators require new tools to anticipate unwanted events. Advances in machine learning and availability of data suggest great potential in u...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-05-01
|
| Series: | Algorithms |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1999-4893/15/6/188 |
| _version_ | 1797490800086482944 |
|---|---|
| author | Volker Hoffmann Bendik Nybakk Torsæter Gjert Hovland Rosenlund Christian Andre Andresen |
| author_facet | Volker Hoffmann Bendik Nybakk Torsæter Gjert Hovland Rosenlund Christian Andre Andresen |
| author_sort | Volker Hoffmann |
| collection | DOAJ |
| description | With the advancing integration of fluctuating renewables, a more dynamic demand-side, and a grid running closer to its operational limits, future power system operators require new tools to anticipate unwanted events. Advances in machine learning and availability of data suggest great potential in using data-driven approaches, but these will only ever be as good as the data they are based on. To lay the ground-work for future data-driven modelling, we establish a baseline state by analysing the statistical distribution of voltage measurements from three sites in the Norwegian power grid (22, 66, and 300 kV). Measurements span four years, are line and phase voltages, are cycle-by-cycle, and include all (even and odd) harmonics up to the 96 order. They are based on four years of historical data from three <span style="font-variant: small-caps;">Elspec</span> Power Quality Analyzers (corresponding to one trillion samples), which we have extracted, processed, and analyzed. We find that: (i) the distribution of harmonics depends on phase and voltage level; (ii) there is little power beyond the 13 harmonic; (iii) there is temporal clumping of extreme values; and (iv) there is seasonality on different time-scales. For machine learning based modelling these findings suggest that: (i) models should be trained in two steps (first with data from all sites, then adapted to site-level); (ii) including harmonics beyond the 13 is unlikely to increase model performance, and that modelling should include features that (iii) encode the state of the grid, as well as (iv) seasonality. |
| first_indexed | 2024-03-10T00:38:15Z |
| format | Article |
| id | doaj.art-49506a23e1b4431c811b134093d2daaf |
| institution | Directory Open Access Journal |
| issn | 1999-4893 |
| language | English |
| last_indexed | 2024-03-10T00:38:15Z |
| publishDate | 2022-05-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Algorithms |
| spelling | doaj.art-49506a23e1b4431c811b134093d2daaf2023-11-23T15:13:02ZengMDPI AGAlgorithms1999-48932022-05-0115618810.3390/a15060188Lessons for Data-Driven Modelling from Harmonics in the Norwegian GridVolker Hoffmann0Bendik Nybakk Torsæter1Gjert Hovland Rosenlund2Christian Andre Andresen3Department of Sustainable Communication Technologies, SINTEF Digital, Forskningsveien 1, 0373 Oslo, NorwayDepartment of Energy System, SINTEF Energi AS, Sem Sælands vei 11, 7034 Trondheim, NorwayDepartment of Energy System, SINTEF Energi AS, Sem Sælands vei 11, 7034 Trondheim, NorwayDepartment of Energy System, SINTEF Energi AS, Sem Sælands vei 11, 7034 Trondheim, NorwayWith the advancing integration of fluctuating renewables, a more dynamic demand-side, and a grid running closer to its operational limits, future power system operators require new tools to anticipate unwanted events. Advances in machine learning and availability of data suggest great potential in using data-driven approaches, but these will only ever be as good as the data they are based on. To lay the ground-work for future data-driven modelling, we establish a baseline state by analysing the statistical distribution of voltage measurements from three sites in the Norwegian power grid (22, 66, and 300 kV). Measurements span four years, are line and phase voltages, are cycle-by-cycle, and include all (even and odd) harmonics up to the 96 order. They are based on four years of historical data from three <span style="font-variant: small-caps;">Elspec</span> Power Quality Analyzers (corresponding to one trillion samples), which we have extracted, processed, and analyzed. We find that: (i) the distribution of harmonics depends on phase and voltage level; (ii) there is little power beyond the 13 harmonic; (iii) there is temporal clumping of extreme values; and (iv) there is seasonality on different time-scales. For machine learning based modelling these findings suggest that: (i) models should be trained in two steps (first with data from all sites, then adapted to site-level); (ii) including harmonics beyond the 13 is unlikely to increase model performance, and that modelling should include features that (iii) encode the state of the grid, as well as (iv) seasonality.https://www.mdpi.com/1999-4893/15/6/188machine learningpower systemsharmonic distortionpower quality |
| spellingShingle | Volker Hoffmann Bendik Nybakk Torsæter Gjert Hovland Rosenlund Christian Andre Andresen Lessons for Data-Driven Modelling from Harmonics in the Norwegian Grid Algorithms machine learning power systems harmonic distortion power quality |
| title | Lessons for Data-Driven Modelling from Harmonics in the Norwegian Grid |
| title_full | Lessons for Data-Driven Modelling from Harmonics in the Norwegian Grid |
| title_fullStr | Lessons for Data-Driven Modelling from Harmonics in the Norwegian Grid |
| title_full_unstemmed | Lessons for Data-Driven Modelling from Harmonics in the Norwegian Grid |
| title_short | Lessons for Data-Driven Modelling from Harmonics in the Norwegian Grid |
| title_sort | lessons for data driven modelling from harmonics in the norwegian grid |
| topic | machine learning power systems harmonic distortion power quality |
| url | https://www.mdpi.com/1999-4893/15/6/188 |
| work_keys_str_mv | AT volkerhoffmann lessonsfordatadrivenmodellingfromharmonicsinthenorwegiangrid AT bendiknybakktorsæter lessonsfordatadrivenmodellingfromharmonicsinthenorwegiangrid AT gjerthovlandrosenlund lessonsfordatadrivenmodellingfromharmonicsinthenorwegiangrid AT christianandreandresen lessonsfordatadrivenmodellingfromharmonicsinthenorwegiangrid |