FORECASTING METHODS APPLIED TO MATERIALS DEMAND OF ONSHORE DRILLING OIL WELLS: RIO GRANDE DO NORTE AND CEARA.
Time Series. Drilling Wells. Optimization. Stocks. Accuracy.
Brazil is the ninth largest oil producer in the world. The demand for present and future drilling wells is fundamental to maintain current production, as well for approval of future projects and purchases of high engineering levels equipment. The literature considers different techniques of accuracy in forecasting methods, (also known as "prediction error"), considering different methods for forecasting those demands. The objective of this work is to compare different techniques of accuracy errors, such as: Mean Average Percentage Error (MAPE), Mean Square Error (MSE) and Root Mean Square Error (RMSE); The forecasting method considered in this work are: Moving Averages (MM), Simple Exponential Smoothing (SES) and Holt-Winters (SHW), Auto-Regressive (ARIMA) and Croston models; In order to optimize the material management of oil and natural gas wells in Rio Grande do Norte and Ceará. For such, the analyzes will be via R software for the monthly consumption of these materials in Rio Grande do Norte and Ceará during the period between 2010 and 2018, referring to the sum of the consumption of 707 different materials. As a result, it is expected to compare the current demand forecast model with the methodologies established by the above-mentioned indicators and identify the most appropriate approaches.