No abstract

Failure detection and diagnosis (FDD) is one of the condition-based maintenance methods. Implementing chiller FDD one could identify the failure and locate where the failure occurs. With the ability to detect and locate the failure, proper maintenance action could be taken to remedy the situation. In most cases, it could restore the performance and extend the useful life of an asset. One of the well-known FDD methods is a Simple Thermodynamic Model (STM). In this paper, a chiller in a commercial building in Jakarta is subjected to the implementation of STM model. The chiller was monitored before and after maintenance. The necessary operating parameters were recorded and processed then they were used to identify two parameters of the STM model by regressing the STM equation. Based on the result, the faults could be detected and located during four different measurements time; on the second month, the eight-month, the eleventh month, and twelfth month. The results show that the STM model is able to detect the faults and locate where the faults have occurred.