Perpustakaan Digital ITB

Arps Decline Curve is the most well known method for forecasting production decline data. Although it is simple to use but it brings uncertainty as its parameter is difficult to predict. We can easily meet historical matching condition using any parameters of exponential/ hyperbolic/ harmonic. However, this tends to bring uncertainty from its forecast. Within this thesis, we are going to use several statistical methods as comparison to Arps Decline Curve. There are two statistical methods being used. The first one is time series statistical method. The second one is geostatistical method. Since production decline data can be seen as series which are separated in equal time, we can naturally conclude that time series statistical method can be used for forecasting. However, geostatistical method is used for spatial data which are separated by distance-not exactly equal distance. Therefore, we have to see production decline data as 'spatial' time data which are fortunately to be equal time distance. Hence, we can also use geostatistical method for forecasting production decline data. A number of interesting results will be presented and validated using data that are not used in mathematical development of models. All results for each production decline data, either hypothetical or field data, would be collected and presented in the same figure to get understanding of the most fit forecasting method.