DESIGN AND IMPLEMENTATION OF ULTRASONIC SIGNAL PROCESSING FOR WATER VOLUME ESTIMATION BASED ON CROSS CORRELATION

• Main Author: Anra Purnama, Adro
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/82512
• Institution: Institut Teknologi Bandung

Water is a basic human need, and in Indonesia, each region has a Regional Drinking Water Company (PDAM) responsible for supplying it. Non-Revenue Water (NRW) is a major issue for PDAMs in Indonesia. According to Rudy Chaerudin, Senior Regional Manager at PDAM Bandung, a good NRW standard is below 20%, but the current level in Indonesia is 35%. NRW can be divided into physical and non-physical categories. Physical losses are due to technical issues like leaks in the distribution network, while non-physical losses are due to factors like meter reading errors, unpaid bills, or water theft. To address NRW, accurate data is essential. Using unreliable data can lead to meaningless results. NRW is calculated by subtracting the total volume of official consumption from the system input volume. At PDAM Tirtawening in Bandung, each meter reader records data from about 3,000 addresses every month, leading to a high likelihood of human error. Access issues to meters also cause unrecorded readings, with bills estimated based on the highest past consumption. Ideally, NRW analysis should match the billing period with the system input volume timeframe. However, customer meters are not read on the same day each month and issues with mechanical meters can only be identified monthly at best. Implementing water meters that automatically transmit data could reduce NRW. The accuracy of water meters also plays a crucial role in the analysis of NRW. Most meters in Indonesia are mechanical with a 2% accuracy rate. More accurate meters improve NRW analysis. Therefore, this paper proposes the use of ultrasonic sensors with the cross- correlation method to calculate water consumption volume. The results obtained show that the highest error at low flow rates is 3.5%, while at normal flow rates, the highest error is 1.19%.