Mechanical state estimation of transmission line sag using tilt sensors

The electrical transmission infrastructure in the U.S. has witnessed severe disruption due to natural events and accidents. There is also the potential for such disruptions due to malicious attacks. For the security of the nation's grid, any disturbance, whether of mechanical or electrical in origin, is identified by the electrical performance of the system. To ensure safe and reliable operation of the power system, detection of mechanical damage to the line at the preliminary stage is indispensable. It is important to distinguish between outages due to electrical faults and by mechanical disruption of line components with the aid of mechanical sensors and analytical methods. This work deals with the online monitoring of the transmission line sag through mechanical state estimation. The measurements of tower tilt, temperature and the current for a multi-span dead-end tilted line section are transmitted to the control center. Measurements of the tension and of the inclination angle of conductor with the horizontal at the suspension points for each span are synthesized using geometric transformation (GT) for the mechanical sag estimation.;Mid-point sag is determined by GT using data of the line section geometry and tower tilt angle for different combinations of tower tilt in the line section. Sag-tension results of GT are compared with that of a more accurate finite element based software package Power Line Systems-Computer Aided Design and Drafting (PLS-CADD). The results of GT prove to be accurate when compared with PLS-CADD.;The state equations obtained with the use of GT, and the measurements of tower tilt, conductor inclination, tension, temperature and current measurements are used in a mechanical state estimation algorithm to determine the sag levels. Mechanical sag estimation is quite accurate when compared to PLS-CADD and demonstrates true mechanical health of the transmission line. The sag estimates are compared with the prescribed sag levels to generate appropriate warning alarms in case of the significant changes or violations.;Bad data detection and identification using the largest normalized residual test is carried out in weighted least squares state estimation algorithm. This test successfully detects and identifies the bad measurements in any of the spans.