Adini, Anisa Gemelia (2021) Prediksi Rate of Penetration pada Sumur Bor Pertamina Menggunakan Jaringan Saraf Tiruan Propagasi Balik. Undergraduate thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
05211740000118_Undergraduate_Thesis.pdf - Accepted Version Download (2MB) | Preview |
Abstract
Pada saat ini, industri pengeboran sumur minyak bumi dan gas mengalami peningkatan kompleksitas. Dalam banyak kasus, biaya pengeboran dapat dikurangi dengan meningkatkan kece-patan pengeboran melalui upaya untuk memaksimalkan rate of penetration (ROP). ROP merupakan ukuran yang sangat pen-ting untuk mengefisienkan waktu pengeboran sumur. Dalam pengoperasiannya, ROP sangat bergantung pada berbagai va-riabel seperti kedalaman su-mur, kecepatan rotasi putar alat bor, dan parameter lainnya.
Tugas akhir ini bertujuan untuk membangun model prediksi ROP menggunakan jaringan saraf tiruan propagasi balik (JST-PB). Masukkan dari model prediksi ini berupa semua variabel yang diperlukan dalam pengeboran sumur bor Per-tamina yaitu tipe sumur, ukuran bit, kedalaman, beban mata bor, beban pengait, kecepatan rotasi, dan variabel pendukung lainnya.
Hasil uji coba menggunakan data pelatihan tiga set data su-mur bor Pertamina (sumur A, B, dan C) dan metode pene-lusuran kombinasi parameter arsitektur JST-PB berbasis grid mengha-silkan arsitektur JST-PB terbaik dengan nilai koefisien deteminasi (R2) dan RMSE berturut-turut sebesar 0,568 dan 2,856. Hasil uji coba penggunaan arsitektur JST-PB terbaik pada data tes ketiga sumur berturut-turut mem-berikan pasangan nilai R2 dan RMSE sebesar 0,704 dan 2,213 (untuk sumur bor A); 0,340 dan 6,166 (untuk sumur bor B); dan 0,023 dan 1,749 (untuk sumur bor C).
======================================================================================================
At present, the oil and gas well drilling industry is experiencing increasing complexity. In many cases, drilling costs can be reduced by increasing the drilling speed, i.e. maximizing the rate of penetration (ROP). ROP is a very important measure to streamline well drilling time. In its operation, the ROP is highly dependent on various variables such as the depth of the well, the rotational speed of the drill bit, and other parameters.
This final project aims to build an ROP prediction model using a back propagation neural network (BPNN). The input of this prediction model is in the form of all the variables needed in drilling Pertamina's boreholes, namely well type, bit size, depth, drill bit load, hook load, rotational speed, and other supporting variables.
Experimental results of using the Pertamina's A, B, and C dril-led wells and the use of grid search method in generating the combination of BPNN architecture parameters resulted in the best BPNN architecture with the coefficient of determination (R2) and RMSE scores of 0.568 and 2.856, respectively. The results of the trial using the best BPNN architecture on the test data of the three wells, respectively, gave a pair of R2 and RMSE scores of 0.704 and 2.213 for drilled well A; 0.340 and 6.166 for drilled well B; and 0.023 and 1.749 for drilled well C.
| Item Type: | Thesis (Undergraduate) |
|---|---|
| Uncontrolled Keywords: | prediksi, rate of penetration, jaringan saraf tiruan propagasi balik, pengeboran sumur minyak bumi dan gas. |
| Subjects: | T Technology > T Technology (General) > T57.5 Data Processing |
| Divisions: | Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Information System > 57201-(S1) Undergraduate Thesis |
| Depositing User: | Anisa Gemelia Adini |
| Date Deposited: | 01 Sep 2021 04:01 |
| Last Modified: | 11 Nov 2024 08:31 |
| URI: | http://repository.its.ac.id/id/eprint/91638 |
Actions (login required)
 |
View Item |