MICROPROCESSOR-BASED INSTRUMENTATION FOR CONDUCTOR VIBRATION MEASUREMENT

Abstract

EHV and UHV overhead transmission lines are used for transmitting bulk power from generating end to distribu tion end. They pass through different types of terrains and experience changing environmental conditions. Due to their large span and height, winds induce high levels of vibration in them. These vibrations increase excessively in flat and open terrains, e.g., river or valley crossings. Mechanical vibrations of transmission line with low amplitude of the order of 25 nun (1 inch) peak to peak and frequency ranging from 5-100 Hz are called aeolian vibration or 'singing*, where as the vibrations of high amplitude, around 6m (20 ft) peak to peak, and low frequencies, 0.25-1.25 Hz, are termed as 'galloping' or 'dancing' of the conductor. Effect of aeo lian vibration is severe due to its persistence throughout the year. In the work reported here only aeolian vibrations are dealt with. The vibrations influence the conductor strands to such an extent that strand breakage may result, one after the other. This fatigue failure of the conductor is due to the alternating mechanical stresses produced in the conductor strands, specifically near the point of support. It is not physically possible to visualize the damage in the conductor from starting point of failures. But, if the break age is not detected at the early stages, the strands will break one after another resulting in the final breakdown of the

conductor. It will result in unexpected shut-down of the line causing inconvenience to the consumer and huge finan cial loss to the electricity company. If the nature and the causes of vibration failure are understood, then the level of the vibration can be controlled by properly placing the dampers, festoons or eveners and reinforce ment of the conductor at the clamps. The transmission line conductor can thus be saved from avoidable damages. Conti nuous measurement of vibrations can help in taking corre ctive measures before reaching the stage of conductor break down. Some experimental work has been done earlier using various types of transducer instrumentation systems for the measurement of overhead transmission line conductor vibra tions. The data is collected over a period of few days by keeping the recorder at the foot of the line tower and then the data is transfered to a mainframe computer for analy sis. This causes lot of inconvenience as the analysis is carried out on accumulated data of a past period. Further more, the processing of the data "after the recording period may lead to delay in taking corrective measures. In the present work, a microprocessor-based on-line instrumentation scheme is developed for the measurement of conductor vibration by using either a physical or a radio telemetry link between the transmitter (getting its input (v) from transducer and signal processing unit) and receiver (microprocessor-based instrumentation system). An experimental pseudo-line has been designed and fabricated in the laboratory for simulating the condi tions of conductor vibration similar to those experienced by lines in the field. A single span of ACSR conductor, with clamps at each end, is supported by vertical columns attached to a steel bed. Vibrations are simulated using a variable speed d.c. motor to push the conductor through an eccentric and sliding-shaft arrangement. A differential variable-area capacitive pick-up has been developed for sensing vibrations in the experi mental line. Capacitive pick-up has been chosen for the vibration measurement due to its outstanding features like simplicity of construction, light~vreight and requirement of negligible force from the measurand. It can withstand wide temperature variations and has high degree of stability and reproducibility. The pick-up provides variation in a differential capacitance in response to conductor displa cement. This in turn is converted into series of pulses of variable width through an external signal processing circuit. The change in the pulse width of pulse width modu lated (PV*d) signal is proportional to the displacement. The signal processing circuit is the integral part of the pick up unit. The main features of the circuit are a small number /

of components and a single 5-volt supply. The TIL level output provides direct compatibility to any digital system. Two types of telemetry schemes have been developed for the transmission of information from the experimental line to the processing and recording station : (i) Physical link, (ii) Radio link. In physical-link system, PWM signal is transmitted through a co-axial cable and is received at a short dist ance. The cable length is limited by the amount of signal attenuation and distortion. In practice, the use of a physical link (cable) is possible only in case of experi mental lines (erected in a laboratory or outdoor) and de-energised actual lines. As the physical link cannot be used with energi sed lines, and in inaccessible areas a telemetry scheme with radio link has been developed for such situations. With radio link, f.m. transmission has been employed for noise immunity. The frequency modulation of PWM signal is made at the central frequency of 108 MHz. The transmitter is powered with a 6-volt dry-cell battery and has a range of 650 m. At the remote station, the signal is received by a f.m. receiver sharply tuned to a frequency of 108 MHz. The output signal of the receiver is fed to a signal condi tioning circuit.

The information (pulse width) thus received is converted into a proportional count in binary code with the help of a high frequency clock, gating circuit and an up/down counter, interfaced with 8085 microprocessor system. The microprocessor first linearizes the incoming data (count versus displacement) on the basis of leastsquare curve-fitting principle and then uses linear inter polation for the evaluation of displacement. Finally, values of displacement are stored and recorded either conti nuously or periodically, as desired. After a predecided interval of time, the microprocessor evaluates and displays the peak-to-peak amplitude and frequency of vibration over the interval. Both the schemes using physical link and radio link have been extensively tested for their performance and the results of the tests found satisfactory. A mathematical model of the transmission line has been developed for the study of conductor vibration. A fourth-order differential equation is obtained assuming mass and stiffness to be constant all along the span, the tension ? is taken to be constant and not affected by the vibrations and the clamped extremities are assumed rigidly fixed. The partial differential equation for transverse vibrations has been solved by a numerical method using central diff erence approximation. ??

From the results of experimentation using the developed instrumentation scheme and the analytical results obtained from the mathematical modelling, variation of vibration amplitude with time are plotted for different excitation frequencies at supposidly the point of maximum stresses (i.e. 89 mm from the clamp). Plots between peakto- peak and excitation frequency are also obtained. An FFT program has also been written to evaluate the frequency spectrum. Using this program, frequency spectra of vibration displacement at different excitation frequencies have been worked out. A comparison of amplitude-time plots, frequency response "curves and frequency spectra shows a close agree ment between the results of experimentation and mathematical analysis. The results, at the outset, confirms the validity of the instrumentation scheme developed. It is also noted that the inclusion of the damping term in the mathematical model of the line brings the results of mathematical analysis still closer to the results of experimentation. *-i The instrumentation scheme developed not only measures displacement, but evaluates other vibration parameters of the interest, viz. velocity, peak-to-peak displacement, strain, frequency and number of excursions made by the conductor.. The data acquired by the microprocessor is analysed for eval uation of parameters immediately after the data acquision. With some minor modifications, the scheme can meet (ix) special requirements, such as mounting on conductor clamp, powering from the line, linking to a mainframe computer etc. Similarly, the experimental pseudo-line made here and the mathematical model developed by the author can find many additional uses (as elabo rated in last chapter).