Health Monitoring of Gas Turbine Hot Sections Based on Standard Deviation and Skewness

Gas turbine hot sections, which work in a high temperature and pressure conditions, could be exposed to failure. The cause of failure on the hot sections are influenced by several factors, including fatigue, creep, oxidation, degradation of the coating on the turbine liner and blade, corrosion, erosion, and surface degradation. Therefore, early fault detection can detect the sign of fault occurrence as soon as possible, improve availability and arrest critical incidents of the plant. Monitoring the variation of mean low-pressure turbine exhaust gas temperature (EGT) is an effective early fault detection method. In this paper, mean LPT EGT of a gas turbine is monitored during six months by 16 thermocouples, which installed after low-pressure turbine. In order to analyses the temperature trend more accurately, we used two numerical methods of standard deviation and skewness. According to the analyses, the closer the standard deviation to zero with absolute positive or negative skewness means that the data collection is in a small temperature range with the least data dispersion and therefore the possibility of damage caused by these fluctuations has significantly reduced. The month of April had the lowest temperature fluctuation with a standard deviation of 0.79 and a skewness of -9.58. However, the highest fluctuations have been calculated in the three months of May, June and July respectively. In addition, to validate the analytical method, borescope inspection of hot sections, has done in three intervals of 2000, 4000 and 6000 hours, and the result was obtained and represented.