BASRI, WIDI HASTUTI HASAN (2023) PENINGKATAN AKURASI MODEL RESERVOIR DENGAN MENGGUNAKAN ASSISTED HISTORY MATCHING PADA LAPANGAN “WD”. Other thesis, UPN "Veteran" Yogyajarta.
![[thumbnail of ABSTRAK_113190014_WIDI_HAST.PDF]](http://eprints.upnyk.ac.id/style/images/fileicons/text.png) |
Text
ABSTRAK_113190014_WIDI_HAST.PDF Download (2MB) |
|
Text
COVER_113190014_WIDI_HASTUT.PDF Download (1MB) |
|
Text
DAFTAR_ISI.PDF Download (199kB) |
|
Text
DAFTAR_PUSTAKA_113190014_WI.PDF Download (2MB) |
|
Text
LEMBAR_PENGESAHAN_113190014.PDF Download (186kB) |
|
Text
SKRIPSI_FULL_TEXT_113190014.PDF Restricted to Repository staff only Download (6MB) |
Abstract
vi
RINGKASAN
PENINGKATAN AKURASI MODEL RESERVOIR DENGAN
MENGGUNAKAN ASSISTED HISTORY MATCHING PADA
LAPANGAN “WD”
Oleh
Widi Hastuti Hasan Basri
NIM: 113190014
(Program Studi Sarjana Teknik Perminyakan)
Lapangan "WD" memiliki enam sumur yang telah berproduksi hingga November 2021,
dengan nilai OOIP sebesar 40 MMSTB. Tantangan untuk proses matching pada lapangan
ini adalah adanya tiga lokasi patahan yang dapat mempengaruhi aliran fluida pada sumur-
sumur di sekitarnya. Penelitian ini merupakan hal baru yang menganalisa hasil dan proses
antara manual dan assisted history matching. Bertujuan untuk menambah pemahaman
mengenai assisted history matching, memodifikasi nilai-nilai dari uncertainty parameter
yang digunakan, menganalisa hasil mismatch data produksi setiap sumur dari metode
manual history matching & assisted history matching, menganalisa parameter paling
berpengaruh, dan lamanya proses matching dari metode history matching & assisted
history matching.
Analisis akan didasarkan pada perbedaan antara proses manual dan assisted history
matching. Dalam melakukan assisted history matching, terdapat dua proses, yaitu
experimental design dan optimization algorithm. Experimental design dilakukan dengan
algoritma latin hypercube dan tornado. Dilakukan proses optimasi dengan menggunakan
perbandingan antara response surface (proxy model) dan particle swarm optimization.
Metode terbaik dalam history matching adalah menggunakan AHM dengan algoritma
PSO (latin hypercube), memberikan hasil yang memuaskan untuk setiap sumur,
sementara manual history matching menghasilkan perbedaan yang cukup dekat antara Np
simulasi dan sejarah lapangan. Parameter yang paling mempengaruhi hasil matching pada
hasil manual dan assisted history matching dari kedua algoritma yang digunakan adalah
aquifer dan relative permeability. Lamanya waktu yang dibutuhkan untuk running manual
history matching selama 15 hingga 20 menit. Lamanya waktu yang dibutuhkan untuk
running algoritma dari assisted history matching satu varian memakan waktu 5 hingga 7
menit.
Kata kunci: Assisted History Matching, Manual History Matching, Experimental Design,
Optimization Algorithm.
vii
ABSTRACT
ENHANCING RESERVOIR MODEL ACCURACY USING ASSISTED
HISTORY MATCHING IN THE "WD” FIELD
By
Widi Hastuti Hasan Basri
NIM: 113190014
(Petroleum Engineering Undergraduated Program)
The "WD" field has six wells that have been producing until November 2021, with an
OOIP value of 40 MMSTB. The challenge for the matching process in this field is the
presence of three fault locations that can influence the fluid flow in the surrounding wells.
This research is a novel approach that analyzes the results and processes between manual
and assisted history matching. The aim is to enhance the understanding of assisted history
matching, modify the values of uncertainty parameters used, analyze the mismatch results
of production data for each well from both manual and assisted history matching
methods, analyze the most influential parameters, and compare the duration of the
matching process between the two methods.
The analysis will be based on the differences between the manual and assisted history
matching processes. In assisted history matching, there are two processes: experimental
design and optimization algorithm. Experimental design is carried out using the Latin
Hypercube and tornado algorithms. The optimization process is performed by comparing
the response surface (proxy model) with the particle swarm optimization.
The best method for history matching is using assisted history matching with the PSO
algorithm using Latin Hypercube, which provides satisfactory results for each well. On
the other hand, manual history matching shows a relatively close difference between
simulated Np and the field's historical data. The most influential parameters in the
matching results for both manual and assisted history matching from the two algorithms
used are aquifer and relative permeability. The runtime for manual history matching is
approximately 15 to 20 minutes while running the algorithm for assisted history matching
for one variant takes 5 to 7 minutes.
Keywords: Assisted History Matching, Manual History Matching, Experimental Design,
Optimization Algorithm
| Item Type: | Thesis (Other) |
|---|---|
| Uncontrolled Keywords: | Assisted History Matching, Manual History Matching, Experimental Design, Optimization Algorithm |
| Subjects: | T Technology > T Technology (General) |
| Divisions: | Faculty of Engineering, Science and Mathematics > School of Engineering Sciences |
| Depositing User: | Eko Yuli |
| Date Deposited: | 22 Aug 2023 05:03 |
| Last Modified: | 22 Aug 2023 06:46 |
| URI: | http://eprints.upnyk.ac.id/id/eprint/37055 |
Actions (login required)
 |
View Item |