Identification of Optimum Closure Relationships for a Mechanistic Model Using a Data Set From a Low-Liquid Loading Subsea Pipeline
| Autor: | Pablo Adames, David Roullier, Eduardo Pereyra, Cem Sarica, Mack Shippen |
|---|---|
| Rok vydání: | 2017 |
| Předmět: |
Petroleum engineering
business.industry Pipeline (computing) Point model Flow assurance 02 engineering and technology Structural engineering 010502 geochemistry & geophysics 01 natural sciences Data set Identification (information) 020401 chemical engineering Closure (computer programming) Wet gas 0204 chemical engineering business Geology 0105 earth and related environmental sciences Subsea |
| Zdroj: | Day 1 Mon, October 09, 2017. |
| DOI: | 10.2118/187327-ms |
| Popis: | Multiphase flow mechanistic models are macroscopic conservation equations (mass and momentum) supported by several empirical closure relations. There are several empirical corelationsavailable for each closure relationship. The success of mechanistic models largely depends on the employed closure relationships. Although, ideally, the closure relationships are desired to be applicable for all operating conditions and fluids;however, in reality, it is difficult to find such universal closures due to empiricism involved in the development of the closureships. Until the development of universal closure relationships, another viable approach is to find different set of closure relationships for identifiably different flow conditions such as low liquid loading flow or high viscosity liquids. In this study, we present a procedure to test the performance of multiple closure relationships against available databases. As an initial application of the procedure, we have tested different combination of closure relationships in the Unified Model (Zhang et al., 2003) and identified the combinations of closures that produce the best model performance for low liquid loading conditions using the field data set consisting of the nineteen field measurements from Frigg to St. Fergus gas and condensates export line. The results of the 37,620 simulations were analyzed. The results showed that the original set of closure relationships used in Zhang et al. (2003) was not optimal for this particular set of data. Instead, a different set of closure relationships is proposed to improve the model's performance for these specific field cases. Total absolute average pressure drop deviation was reduced form 11% to 10%. Total absolute average liquid content deviation was reduced from 66% to 1%. Using the proposed procedure, with the availability of additional large-scale field data, the set of closure relationships can be further refined, and even the elusive goal of finding universal closure laws would be possible. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|