Stability Analysis of AGC in the Norwegian Energy System. Ingvar Andreassen Master's Thesis.
abstract   bibtex   
In the last recent years it has been observed increasing power system instability in the Nordic power system, this is observed by an increasing incidents of frequency deviations. The theory of power system control and actual control of the Norwegian power system was investigated, and a simplified model was made for stability analysis on the Norwegian power system to find probable causes for system instability and better parameter settings. The theory of the main contributing factors in power system stability was studied, such as Automatic Generation Control (AGC), Load Frequency Control (LFC) and turbine controllers with droop. The implementation of power frequency control in Norway and the Nordic power system was investigated and described in this report. The functional description of an AGC system from ABB used by Statkraft at an actual operations centre was studied, in addition to how this system was used and implemented. An actual AGC control area with hydropower units was then used as a basis for making a dynamic model. The model was made using the Modelica programming language together with the computer tool Dymola. Components from the Hydro Plant Library (HPL) from MODELON AB were used in making models of the hydropower units and the Nordic power system. One detailed plant with HPL water way components was made, while the other hydropower units in the area were simplified to ease the computational burden in simulations. The model contains a total of seven hydropower units where six is controlled by an AGC. The turbine controllers were implemented with and without frequency dead band. An AGC system model was made realistically including low pass filters and non linear functions such as sampling, dead bands and rate limiters. The AGC PI controller were tuned, and then tested and compared to the recorded real response. This showed similar system dynamics, although the model responded slightly faster. Simulations showed that a slow response of the AGC could be an advantage, as it gave both a minimized Area Control Error (ACE) and better stability. It was also observed small oscillations in steady state of the system, which was partly caused by a dead band zone filter in the AGC system. Load change tests were also performed on the model, where a sudden large drop in grid frequency occurs. The turbine controllers with dead band on the frequency measurement caused a poorer primary frequency control in the system, as expected. The tests also showed that the ACE regulation of the AGC controller model cancelled the primary control action from the turbine controllers, which caused a larger grid frequency deviation. However, the model needs to be investigated for the verification of this last result.
@thesis{IngvarAndreassen2011,
  title = {Stability {{Analysis}} of {{AGC}} in the {{Norwegian Energy System}}},
  author = {{Ingvar Andreassen}},
  date = {2011-06-03},
  institution = {{Telemark University College}},
  abstract = {In the last recent years it has been observed increasing power system instability in the Nordic power system, this is observed by an increasing incidents of frequency deviations. The theory of power system control and actual control of the Norwegian power system was investigated, and a simplified model was made for stability analysis on the Norwegian power system to find probable causes for system instability and better parameter settings. The theory of the main contributing factors in power system stability was studied, such as Automatic Generation Control (AGC), Load Frequency Control (LFC) and turbine controllers with droop. The implementation of power frequency control in Norway and the Nordic power system was investigated and described in this report. The functional description of an AGC system from ABB used by Statkraft at an actual operations centre was studied, in addition to how this system was used and implemented. An actual AGC control area with hydropower units was then used as a basis for making a dynamic model. The model was made using the Modelica programming language together with the computer tool Dymola. Components from the Hydro Plant Library (HPL) from MODELON AB were used in making models of the hydropower units and the Nordic power system. One detailed plant with HPL water way components was made, while the other hydropower units in the area were simplified to ease the computational burden in simulations. The model contains a total of seven hydropower units where six is controlled by an AGC. The turbine controllers were implemented with and without frequency dead band. An AGC system model was made realistically including low pass filters and non linear functions such as sampling, dead bands and rate limiters. The AGC PI controller were tuned, and then tested and compared to the recorded real response. This showed similar system dynamics, although the model responded slightly faster. Simulations showed that a slow response of the AGC could be an advantage, as it gave both a minimized Area Control Error (ACE) and better stability. It was also observed small oscillations in steady state of the system, which was partly caused by a dead band zone filter in the AGC system. Load change tests were also performed on the model, where a sudden large drop in grid frequency occurs. The turbine controllers with dead band on the frequency measurement caused a poorer primary frequency control in the system, as expected. The tests also showed that the ACE regulation of the AGC controller model cancelled the primary control action from the turbine controllers, which caused a larger grid frequency deviation. However, the model needs to be investigated for the verification of this last result.},
  annotation = {Master's Thesis},
  type = {Master's Thesis}
}

Downloads: 0