Abstract

This paper discusses the results of steady-state and transient flow simulators utilized in the design, operation, and troubleshooting of a complex multiphase pipeline network in a Saudi Arabian oilfield. The oilfield pipelines pass across a hilly terrain with sand dunes rising up to 200 meters, transporting wet crude from wells located on sabkhas (flat terrain) to three Gas Oil Separation Plants (GOSP). State-of-the-art transient multiphase flow model performance was also evaluated using pipeline slug flow tests measurements.

The results of various steady-state algorithms used for pressure drop and free-flow calculations in the Field's pipelines were found inaccurate and inconsistent. A comparison of these calculations and Field data demonstrates that they should not be used for predicting pressure drop, sizing multiphase pipelines, or designing production facilities over hilly terrain. Transient multiphase models can more accurately predict pipeline pressure drop, and optimize pipeline size. A transient multiphase model was used for simulating trunkline valves for slug flow control. Based on simulation results, plug valves were installed near trunkline outlets. The valves were found to be useful in mitigating effects of slugs into the receiving separators, thus eliminating the need for installing costly slug-catchers. Different free-flow alternatives, and crude pumping for transferring wet crude from satellite GOSP's to the Central Processing Facility (CPF) were examined. Crude pumping was found the most suitable and economical option due to large pipeline pressure drop resulting from slugs.

Transient multiphase flow continues to be highly complex. Its effects are hard to predict accurately. Pipeline pressure drop results from simulations using the latest commercially available flow models were about 20-30 % lower than Field measured values.

Introduction

Produced crude free-flows from wells through a network of flowlines and trunklines to three processing facilities located at the north (GOSP-1), south (GOSP-3) and the Central Processing Facility (CPF). The north and south satellite GOSP's have one-stage and the CPF has three-stage separation capacity. The three processing facilities receive crude into High Pressure Production Traps (HPPT) to separate gas from the oil at 400 psig. The separated crude from the satellite GOSP's is then pumped through 24-inch transfer lines to the Central Processing Facility (CPF) for Intermediate Pressure Production Trap (IPPT), and Low Pressure Production Trap (LPPT) gas separation at 150 psig and 70 psig respectively. The separated gas at the satellite GOSP's is free-flowed via 30-inch transfer-lines also to the CPF. The gas is dehydrated, dew point controlled, and then injected into the Field's gas cap. Four trunklines each transport crude from sabkhas production headers to the north GOSP-1 and CPF, and three trunklines to the south GOSP-3, for a total of eleven trunklines (Fig. 1). Each trunkline has a plug valve installed near the outlet to mitigate the slug effects on the receiving separator.

The terrain elevation changes in the Field were found to cause highly unstable slug flow regimes in the pipelines under certain flow conditions. Of particular concern to Saudi Aramco operations is the large pressure drop in the trunklines as a result of slug flow. It increases the wellhead back pressures, affects Inter-GOSP crude transportation, and limits low pressure producers. In addition, slug effects can upset the normal operation of the receiving separators, if not properly controlled. A slug flow test campaign was conducted, the objectives of the Field tests were as follows:

• Calibrate a transient flow simulator with actual Field test data to establish a benchmark. This benchmark will be used to correct simulator calculated values.

• Characterize slugs in Field's pipelines to enhance a new mechanistic transient flow model, aiming at pressure drop calculation accuracy improvement.