Please use this identifier to cite or link to this item:
http://hdl.handle.net/10266/5994
Title: | Design of humidity sensor for condition monitoring of power transformer |
Authors: | Kumar, Shailesh |
Supervisor: | Raina, Kuldeep Kumar Islam, Tarikul |
Keywords: | Transformer;Condition monitor;Moisture;Humidity sensors;Modelling;Breather;Moisture controller |
Issue Date: | 13-Aug-2020 |
Abstract: | The life of the transformer highly depends upon the insulation quality of the transformer oil. Moisture is one of the main causes for the degradation of the insulation quality of the oil. It decreases the dielectric strength and accelerates the aging of the insulation system of the transformer. Content of the moisture in insulation is measured in terms of parts per million (ppm). There are two main sources of water contamination of the oil in high voltage transformers such as (i) moisture ingress through cellulose paper used for transformer winding, (ii) moisture ingress by atmospheric humidity through transformer breather. In addition, a fresh oil sample always contains certain amount of moisture. The water contamination rate through the breather greatly depends upon the dehydrating capability of the silica gel present in the breather. The humidity level in the breather varies between 10% RH to 90% RH depending upon the condition of silica – gel. At present, there is no online method for condition monitoring of the silica gel of the breather. To avoid the need of frequent replacement of silica gel and for condition-based maintenance of the breather, a humidity sensor based auto regenerative dehydrating breather may be an advanced solution. Further, for protecting the high voltage transformer from the dangerous effect of excess moisture, there is a need of a humidity sensor, which directly detects moisture in ppm level (3-100 ppm) in the transformer (T/F) oil. Moisture measurement in ppm level is challenging, and a costly affair. The environment of the transformer is chemically harsh and there is a change in ambient temperature. Metal oxide thin film sensor is stable robust material which may be suitable for moisture measurement in oil and in breather. There are various technologies used for moisture measurement from ppm to ppb level. Popular technologies are Karl fisher titration, chilled mirror hygrometer, cavity ring down spectroscopy (CRDS), electrolytic, and polymer/oxide thin film sensors. Most of the technologies except thin film sensors are costlier and suitable for offline measurement only, require skilled manpower, have long measurement time and are not suitable for field application. A low-cost portable moisture measurement system can be developed using solid state capacitive sensor. Electrolytic sensor has limited life span, and the polymer sensor has low thermal stability. There is a limited work for direct measurement of moisture in ppm level using thin film sensor. Therefore, main objectives of the present thesis are (i) development of a humidity sensor-based system for online condition monitoring and controlling moisture of silica gel of v the breather (ii) online moisture measurement of transformer oil in ppm level using a parallel plate capacitive metal oxide thin film sensor and modelling its long-term drift. To accomplish the objectives (i) initially, a mathematical model for estimating the moisture content in silica-gel, and the degradation in the performance of the silica-gel inside a breather is developed, (ii) the anodized aluminum oxide parallel plate capacitive sensors and determination of response parameters of the sensors for humidity measurement in the range of 1-97% in breather have been designed and fabricated (iii) a microcontroller based interfacing circuit for auto refreshing of silica gel has been designed and implemented(iv) a thin film capacitive sensor for direct moisture measurement in transformer oil in ppm has been designed and fabricated and finally (v) the modelling the long-term drift due to aging of the ppm moisture sensor has been carried out to estimate the drift due to morphological change. The developed breather model analyses the saturation condition of the silica-gel. The simulation results of the model closely match the sensor based experimental results. The fabricated anodized aluminium oxide sensors with controlled pore morphology measure the humidity of the breather in the range of 1-97% RH with sensitivities vary from 0.8 pF/%RH to 15 pF/%RH. The automatic control of moisture using heater via moisture sensor in the breather (0.5 kg, used for 150 KVA distribution transformers) increases the refreshing efficiency by 43% and reduces the need of regular replacement of silica-gel. The sol-gel thin film humidity sensor fabricated to measure moisture directly in the oil can measure in the range of 3-100 ppm with the maximum and minimum sensitivity of 1.8 pF/ppm and 0.8 pF/ppm respectively. Effort is also made to measure moisture using anodized alumina capacitive sensor in the range of 180 ppm to 800 ppm. The developed long-term drift model estimates the variation in pore morphology due to the lateral seepage of the moisture in the pore cell of the sensor. Change in pore morphology leads to the variation of the sensitivity of the sensor. It is found that the maximum deviation in the experimental and model response is 2.5%. |
URI: | http://hdl.handle.net/10266/5994 |
Appears in Collections: | Doctoral Theses@EIED |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
thesis_shailesh.pdf | 5.52 MB | Adobe PDF | View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.