Forecast of Operational Sea State in Malaysian Waters

Currently, the wave height forecast for daily operation and support activities of PETRONAS are done based on estimation method. This method is found to be inaccurate. The requirement of improving the base model for wave height are critically needed to prevent the over estimation and for cost optimization. The objective of this research is mainly to develop forecast model for Malaysian waters by studying the metocean historical data. By having this forecast model, offshore operation cost optimization can be achieved. The metocean parameter that will be focused in this study is significant wave height. The study area will be Dulang (PMO), Erb West (SBO) and Tukau (SKO) platform which located in South China Sea extended from Peninsular Malaysia, Sabah and Sarawak. In order to forecast the wave height based on the historical data, Auto-regressive Integrated Moving Average (ARIMA) method will be applied. The best ARIMA model has been selected by using Bayesian Information Criterion (BIC) value. The parameters of the estimates (p,d,q) was found to be consistent for the three regions. The operation areas can be represented by one ARIMA model. In addition, it was found that the wave time series was weakly stationary and hence no differencing is needed. ARIMA model is reliable for short-term forecast and can be used for operational by Petronas Carigali Sdn. Bhd.