2016_EJRNL_PP_MOHAMMAD_ALI_AHMADI_11.pdf
Terbatas  Suharsiyah
» Gedung UPT Perpustakaan
Terbatas  Suharsiyah
» Gedung UPT Perpustakaan
The importance of the flow patterns through petroleum production wells proved for upstream
experts to provide robust production schemes based on the knowledge about flow behavior. To
provide accurate flow pattern distribution through production wells, accurate prediction/representation of bottom hole pressure (BHP) for determining pressure drop from bottom to surface play
important and vital role. Nevertheless enormous efforts have been made to develop mechanistic
approach, most of the mechanistic and conventional models or correlations unable to estimate or
represent the BHP with high accuracy and low uncertainty. To defeat the mentioned hurdle and
monitor BHP in vertical multiphase flow through petroleum production wells, inventive intelligent
based solution like as least square support vector machine (LSSVM) method was utilized. The
evolved first-break approach is examined by applying precise real field data illustrated in open
previous surveys. Thanks to the statistical criteria gained from the outcomes obtained from LSSVM
approach, the proposed least support vector machine (LSSVM) model has high integrity and performance. Moreover, very low relative deviation between the model estimations and the relevant
actual BHP data is figured out to be less than 6%. The output gained from LSSVM model are closed
the BHP while other mechanistic models fails to predict BHP through petroleum production wells.
Provided solutions of this study explicated that implies of LSSVM in monitoring bottom-hole
pressure can indicate more accurate monitoring of the referred target which can lead to robust
design with high level of reliability for oil and gas production operation facilities.