THE DEVELOPMENT OF AN INDUSTRIAL INTERNET OF THINGS (IIOT) SYSTEM FOR A DATA CENTER WITH WATER-BASED COOLING

• Main Author: Hamid, Abdul
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/84325
• Institution: Institut Teknologi Bandung

This research aims to build an IIoT (Industrial Internet of Things) platform for a water-based cooling system in a tier 2 data center. The data center, measuring 3x4 square meters, houses 34 servers that need to be maintained at a temperature between 18°C-20°C. The cooling system consists of one chiller that supplies cooling water to two in-rack cooling systems within the data center, with each having an operating temperature range of 10°C-22°C and 18°C-22°C. Between the chiller and the data center, there is a 15-meter-long piping system. To monitor the performance of this system, 10 temperature sensors, 8 pressure sensors, 2 water flow sensors, 2 power meters, 2 centrifugal pump motors, 2 3 kW inverters, and a Programmable Logic Controller (PLC) have been installed. The communication module on the PLC uses RS485 and analog modules, while the communication module on the in-rack cooling uses RS485 via the Ethernet protocol. Data acquisition is carried out using Node-RED, which is connected to a local database. All observations of data and system performance can be done through an interface, making the entire system an IIoT platform for monitoring the data center's cooling system. Based on the data collected from this platform, static and dynamic performance analyses of the data center's cooling system have been conducted. The static performance analysis of the measurement system on the chiller was conducted using data reconciliation principles and gross error detection with the Least Square method. Measurement biases were found on four temperature sensors, each by 2.52°C, 2.57°C, 2.52°C, and 2.57°C. After correction, the accuracy improved with an average of 0.504. Meanwhile, for two flow sensors, after filtering, a residue of 1 l/s was obtained, and biases of 7.08 l/s and 18.87 l/s were detected, and after correction, the accuracy improved with MAE values of 0.045 and 0.015. The IIoT system that has been built features data filtration, correction, and real-time monitoring displayed through an HPHMI interface created based on ISA 101 standards. Keywords : Industrial Internet of Things, water-based cooling system, data reconciliation, real-time monitoring ?