Étude expérimentale des écoulements diphasiques pétroliers en terrain accidenté

Experimental study of two-phase petroleum flow over rough ground

¹ Institut Français du Pétrole Département " Techniques de Production en Mer " Rueil-Malmaison
² Société Nationale Elf-Aquitaine (Production) Département " Exploitation " Paris
³ Compagnie Française des Pétroles Département " Exploitation " Paris

Abstract

Predicting two-phase behaviour in subhorizontal pipes has become critical for offshore oil and gas fields development, specially in great water depths or when the sea bed is uneven. Standard methods of calculating average pressure losses are lacking in precision and do not predict transient phenomena, e.g. liquid slugs length or separation of phases at low points in the lines. As a result, two-phase flow is generally ruled out as a method of transporting produced hydrocarbons to land, even though this process is very simple and requires only low investments. The objective of the "two-phase flow research group", associating CFP, IFP and SNEA (P) within the GERTH, is to improve methods of predicting flow regimes, spatial phases distributions and pressure losses, specially in inclined pipelines. After preliminary air-water experiments on a two inch test loop at the LMEF of PARIS-ORSAY, construction of a test loop working in industrial conditions was undertaken in 1977 at the SNEA (P) gas plant of BOUSSENS. Main characteristics of this 6" loop are: Length of the testline : 400 ft (260 ft in inclined position) Continuous variable slopes between - 10 % and + 10 % Average test pressure between 100 and 800 psi - Superficial gas flow velocity between 0 and 30 ft/s - Superficial liquid flow velocity between 0 and 10 ft/s. The loop is equiped with about 40 process instruments and 40 two-phase flow sensors such as slug detection probes or average liquid fraction probes. Instruments are monitored by a 3052A H.P. data acquisition system which controls measurements in real time and print out lists of processed results. First experiments were made at high GOR flows, in horizontal or slowly rising slopes, where the stratified flow pattern predominates. Results confirmed the strong influence of the inclination angle on pattern, liquid hold-up and pressure gradient, in particular at moderated gas flow rates. A critical gas velocity was observed at very high GOR in inclined flow : if the gas velocity is below this critical value, the liquid accumulates in the line, increasing drastically the pressure losses. The experimental data were compared to the results of the predictive models usually used in oil industry. All these methods do not take enough account of the slope influence, especially for high GOR at gas flow rates below the critical gas velocity. So, theoretical research is going on, in order to develop a more precise stratified flow model.