Life estimation of ceramic insulators in costal areas using predictive model based on weather conditions

Insulators are the foremost vital part of the electrical power distribution system, because it determined the system?s insulation level. It is important for us to know the factors that cause damage to insulators that is why we first study from various research and experiments that happened in the pa

Project Title

Life estimation of ceramic insulators in costal areas using predictive model based on weather conditions

Project Area of Specialization

Electrical/Electronic Engineering

Project Summary

Insulators are the foremost vital part of the electrical power distribution system, because it determined the system’s insulation level. It is important for us to know the factors that cause damage to insulators that is why we first study from various research and experiments that happened in the past which explains how different types of contamination on ceramic insulators cause great damage to the distribution systems reliability and also more power loss.

As the range of distribution voltage increases, the pollution severity of the site becomes the most important factor in determining the insulation level of the system.

The major reason behind the polluted ceramic insulators are their surface because it tends to be wet by water, and it results in creating continuous pollution layers over the surface of ceramic insulators in wet conditions.

Continuous pollution layers are generated resulting in leakage currents which result in arc bands being formed. Such arcs are formed in these dry regions because of an increase in the strength of the electric field. Because of emissions and contamination, flashover for insulators and substations nearby highways is much troublesome, which are routinely salted, thus revealing the insulators to dust and contamination gathered.

Flashover on polluted insulators poses serious threats to the reliability of the distribution system. Also the contamination which causes damage to insulators leads to system outages and power loss. Due to this leakage current flow through damage area of insulator causes loss of power.

An experimental chamber will be used to perform artificial aging on the distribution insulators. An electrical test bench will be made on which an aging sample collected from an experimental chamber will install and measure the leakage current. A predictive model using machine learning approach will be design which tell us how long our insulator can survive on the basis of data collected.

Project Objectives

Following are our project objectives;

• Perform aging analysis on ceramic insulators in Experimental Chamber by using artificial environmental effects.
• To create 5 different artificial environmental effects in experimental chamber including temperature, rain, dust and fog effect.
• To check the behavior of weather conditions of coastal areas (like Temperature, Raining, Fog, Dust etc.)  On the ceramic insulator.
• A Test bench will be created for calculating the leakage current flowing through the insulator.
•  Using Machine Learning we will design a predictive model which helps to estimate the life of insulator.

Project Implementation Method

• Our primary purpose is to perform aging analysis on ceramic insulators and for that we required an Experimental Chamber in which we can create artificial environmental effects.
• Artificial environmental effects for our experimental chamber that includes temperature, rain, dust and fog effect.
• After performing the environmental effects, sample will be collected from the chamber and will be mounted on test bench to test the leakage current and factors pertaining to predict the health of insulator.
• After the acquisition of relevant data from testing bench the preprocessing to remove the outliers and error point need to be done. Furthermore, more suitable feature set can be investigated.
• Extraction of suitable features for machine learning Algorithm.
• Selection of machine learning algorithm. For example neural network, support vector machine, KNN (K nearest neighbors), Naive Bayes etc. We will select one of them.
• Specifying a subset of data to be used for training the model.
• Validate the predictive model accuracy on validation not used for training or testing.
• Use the model for testing if satisfied with prediction accuracy.

Benefits of the Project

• Experimental Chamber:

 Our experimental chamber is useful for performing those tests which include environmental effect testing on Ceramic insulator, by utilizing this experimental chamber one can also perform testing on polymer insulators.

• Electrical test bench:

In our electrical test bench all instruments are calibrated and located in one place, which are used to measure the leakage current flowing through insulators by applying voltages and then vary current.

• Machine Learning:

As our project is research based we can easily determine the Data that we required by using machine learning algorithms. We don’t need to train a model with complete data, we can specifying a subset of data to be used for training the model.

After specifying a subset of data to be used for training the model, the proper life estimation of ceramic insulator can be determined and with replacement of insulators at known time, we can overcome the system outages and power losses.

We can also overcome the financial loss of power equipment’s due to insulator damage.

Technical Details of Final Deliverable

Our final deliverable will include an experimental chamber made of 5mm thick acrylic sheet with dimensions of 2x2x2 ft and An Electrical Test bench made of thick sheet with dimensions of 1.5x1.5x1.5ft.

In Experimental Chamber we have used DHT22 sensor. In this Heat and cold effect system the input value of required temperature can be set by us using a TFT touchscreen, this heat and cold effect system is a closed loop system. We also used Flow sensor to measure the flow rate of water, by this flow rate we can set the rain rate value in mm/hr, Rain rate can be change by changing the speed of pump which is done by using BTS7960 motor driver. The rain rate input value can also be given through TFT touchscreen. It can control rainfall's intensity and duration.  PH sensor is also used to measure the PH value of the water. We also placed a 500W halogen light on top of the chamber which is facing downward, it is for boosting the heat effect in chamber. The intensity of halogen light can be controlled by dimmer. .Ultrasonic mist maker fogger is also placed inside of the chamber for creating Fog effect.

In Electrical Test bench we will use following components;

• Power supply of following rating;

(Power rating = 11KV

Input Voltage = 110-220VAC

Output current = 0-50A

Output Voltage= 220VAC)

• Iron clamper (use to hold the ceramic insulator).
• Steel Wire (3m, 80A).
• Insulation Pole (1.5feet).
• Ammeter.
• C.T (100A).

In Machine learning there are many programming platforms for example C++, PYTHON, JAVA etc. We will use C++ programming language to apply the algorithm on Data collected from electrical test bench.

Final Deliverable of the Project

HW/SW integrated system

Core Industry

Others

Other Industries

Others

Core Technology

Others

Other Technologies

Others

Sustainable Development Goals

Industry, Innovation and Infrastructure

Required Resources

If you need this project, please contact me on contact@adikhanofficial.com