Nugroho, Adi (2024) Peramalan Tekanan Operasional Pipa Salur Minyak Bumi untuk Pencegahan Oil Congealing Menggunakan Metode Hibrida. Masters thesis, Institut Teknologi Sepuluh Nopember.
![[thumbnail of 6032221129-Master_Thesis.pdf]](http://repository.its.ac.id/style/images/fileicons/text.png) |
Text
6032221129-Master_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 October 2026. Download (4MB) | Request a copy |
Abstract
Dalam industri minyak bumi, jaminan aliran fluida minyak melalui pipa salur merupakan aspek kritikal, dimana gangguan oil congealing atau pengentalan minyak menjadi semi-padat dapat berakibat fatal terhadap operasi produksi. Meski sistem monitoring real time aliran pipa saat ini telah tersedia, namun tindakan pencegahan seringkali terlambat, sehingga meningkatkan risiko terhentinya operasi. Penelitian ini mengusulkan peramalan tekanan operasi pipa sebagai indikator kondisi aliran minyak beberapa hari ke depan untuk mencegah terjadinya oil congealing, menggunakan model time series berdasarkan data historis. Data yang digunakan meliputi parameter tekanan operasi pipa dan data eksternal berupa suhu lingkungan dan curah hujan dari tiga segmen pipa yang diamati. Penelitian ini membangun dan memvalidasi model peramalan time series menggunakan algoritma ARIMA/ARIMAX, FFNN, dan metode hibrida ARIMA/ARIMAX-FFNN. Model ini kemudian dibandingkan berdasarkan akurasi menggunakan nilai RMSE dan MAPE. Hasil penelitian menunjukkan bahwa model FFNN dengan variabel eksogen adalah yang paling efektif dalam memprediksi tekanan operasi pipa, seperti dibuktikan dengan nilai RMSE sebesar 15.40 dan MAPE sebesar 14.73 pada salah satu segmen pipa, yang menunjukkan kinerja lebih baik dibandingkan dengan model ARIMA/ARIMAX dan metode hibrida. Penelitian ini memberikan kontribusi signifikan bagi praktisi industri minyak dan gas, menawarkan model peramalan time series ilmiah yang memungkinkan tim operasi mengambil tindakan preventif lebih cepat terhadap risiko oil congealing. Manfaat bagi perusahaan termasuk menghindari kerugian akibat terhentinya produksi dan pengoptimalan sumber daya dalam penanganan oil congealing. ===================================================================================================================================
In the petroleum industry, ensuring the flow of oil fluids through pipelines is a critical aspect, where disruptions like oil congealing or the thickening of oil into a semi-solid state can be fatal to production operations. Although real-time pipeline monitoring systems are currently available, preventative measures are often delayed, thereby increasing the risk of operational shutdowns. This research proposes forecasting pipeline operational pressure as an indicator of oil flow conditions for the next few days to prevent oil congealing, using a time series model based on historical data. The data used includes pipeline operational pressure parameters and external data such as environmental temperature and rainfall from three observed pipeline sections. This study develops and validates a time series forecasting model using ARIMA/ARIMAX, FFNN, and a hybrid ARIMA/ARIMAX-FFNN method. The models are then compared based on accuracy using RMSE and MAPE values. The results indicate that the FFNN model with exogenous variables is the most effective in predicting pipeline operational pressure, as evidenced by an RMSE value of 15.40 and a MAPE of 14.73 in one of the pipeline segments, showing better performance compared to the ARIMA/ARIMAX models and the hybrid method. This research provides significant contributions to practitioners in the oil and gas industry, offering a scientific time series forecasting model that enables operational teams to take faster preventive actions against the risk of oil congealing. Benefits to the company include avoiding losses due to production stoppages and optimizing resources in handling oil congealing.
Actions (login required)
 |
View Item |