A Method to Quantify Data Quality in Asset Health Indices Model

Asset Health-Indexing is a data-intensive process. Depending on the complexity of the asset and the number of failure modes included in the model, the number of condition indicators involved in the process can be in the range of tens to hundreds. Consequently, data quality becomes a critical factor in the model’s output accuracy. The issues include incomplete data, old data, incorrect data format, and data validity. Therefore, measuring the data quality level can benefit the decision-maker. This paper introduces a method to quantify data quality in the asset health indices (AHI) model. The technique considers six data quality parameters: completeness, consistency, uniqueness, validity, accuracy, and timeliness. The output is the data quality indices, shown in % for each data quality parameter. The paper provides the implementation of the method in a power transformer AHI model from a case study in PLN Indonesia. The paper suggests that the data quality indices become an additional flag in the AHI modeling to assist the decision-maker in justifying better.