DEVELOPMENT OF NATURAL GAS ENERGY MEASUREMENT SYSTEM BASED ON SOUND SPEED AND DILUENT GAS CONCENTRATION

• Main Author: Padang, Elohansen
• Format: Dissertations
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/61119
• Institution: Institut Teknologi Bandung

Natural gas is a gaseous fossil fuel whose main component is hydrocarbon compounds and contains a small amount of non-hydrocarbons. As a fuel, natural gas is a clean fuel that produces less greenhouse gases than other fossil fuels, such as petroleum and coal. Based on data from British Petroleum in 2019, natural gas supplies energy by 24% or 140.14 exajoules from 583.9 exajoules of world energy consumption. Likewise in Indonesia, in 2017 around 21% or 1.49 exajoules of the 7.08 exajoules of Indonesia's energy needs were supplied from natural gas sources. It is estimated that the demand for natural gas in the future will increase due to its superiority as a clean fuel. In the natural gas trading process, the parameter that is generally used is the calorific value, which is an indicator that determines the energy content of natural gas. However, the problem until now is that there is no measurement device that can measure natural gas energy directly and in real time. Energy measurement techniques that are still used today are measurements using a calorimeter and gas chromatography composition analysis. Both of these techniques provide very accurate measurements of natural gas energy. However, it has several limitations, such as the price of equipment and equipment maintenance which is relatively expensive, requires skilled operators to operate the equipment, and measurements cannot be done in real time. Meanwhile, the varying composition of natural gas due to different sources, upgrading and transportation of natural gas requires measurements of natural gas energy that can be measured in real-time so that the trading process takes place fairly. Therefore, it is necessary to develop an alternative method of measuring natural gas energy which is low cost but accurate, real-time, and the measurement results can be monitored remotely. In this research, an alternative method of measuring natural gas energy based on measuring the velocity of sound and diluent gas (carbon dioxide and nitrogen) has been developed. Measurement of the speed of sound is carried out using an ultrasonic sensor and the concentration of carbon dioxide using a non-dispersive infrared (NDIR) sensor. Meanwhile, nitrogen gas is measured indirectly using deep learning techniques, especially the Deep Neural Network (DNN). The measurement results of these three parameters are related to the heating value using the correlative method, a method based on the idea that the measurement of the physical or chemical properties of natural gas can be used to provide the information needed for the energy conversion of natural gas. The development of natural gas energy measurement tools begins with the development of a model for calculating natural gas energy. A model for calculating the calorific value of natural gas was developed based on a database of natural gas composition from previous research researchers. The correlation model between the speed of sound, diluent gas concentration and bulk modulus to the heating value of natural gas obtained can be expressed as: ????????= 9820,685+[?17,830+(0,042 ?0,170????????2?0,233????????????2)????]???????????????????? In addition, an indirect measurement model for nitrogen concentration in natural gas was developed based on the DNN algorithm. The best DNN model obtained is a model with 4 hidden layers with 24, 24, 24, 24 neurons in each layer, ReLU activation function, and Adam optimizer optimization function. Furthermore, the natural gas energy calculation model and the DNN model to measure the concentration of N2 are implanted into the Arduino Uno and ESP32 microcontrollers using the Arduino IDE software. These two microcontrollers are part of the data acquisition system contained in the prototype module for measuring natural gas energy that was developed in this study. In addition, Internet of Thing (IoT) technology for remote monitoring of measurement results is also implemented in the prototype of the energy measurement module. Testing on a laboratory scale of the proposed prototype natural gas energy measurement module shows that the measurement of the calorific value of natural gas is generally within the range of ± 0.5% of the actual calorific value. The results of this test indicate that the prototype of the energy measurement module is accurate enough to measure the heating value of natural gas. Based on the test results, it can be stated that the prototype of this energy module can be further developed as an alternative measurement method for natural gas energy that is low-cost, accurate, real-time, and can be monitored remotely.