Hydrate blockage detection, remediation and removal are successfully applied in a hydrate plug incident in the GOM. A hydrate plug occurred in mid-October 2002 in the Williams Field Services (WFS) owned and operated, 36-mile long, 10- inch gas export pipeline from ChevronTexaco's Genesis Green Canyon 205 (GC 205) platform (2,600 ft water depth) to the downstream Ship Shoal 354 platform (SS 354) platform (460 ft water depth). ChevronTexaco's Flow Assurance experts, ChevronTexaco's Genesis Asset Team, and WFS team worked together to detect, locate and remediate the hydrate plug in a safe and timely fashion while ensuring personnel safety, maintaining pipeline and riser integrity and minimizing loss of production. This paper shows how a concentrated team effort and deployment of the right tools such as transient models, hydrate formation and dissociation models, can be used to predict and alleviate a flow assurance problem, such as hydrates, which can occur in deepwater oil and gas production in a subsea system. Introduction The Genesis development1 was brought on stream in February 1, 1999. In mid-October 2002, following two major shutdowns due to strong hurricanes and one for WFS planned maintenance, gas flow into the dry gas export pipeline to SS 354 was obstructed and gas delivery to shore was halted. Analysis of the measured hydraulic data prior to and during the incident showed that the obstruction in the gas pipeline was due to the presence of hydrates corresponding to low spots in the pipeline. Operations personnel in New Orleans, working in conjunction with Flow Assurance personnel in Houston were able to locate the hydrate plug via hydraulic and hydrate models based on the measured pressure, temperature and gas flow rate during the incident. Hydrate formation caused by the presence of water and natural gas components at high pressures and low temperatures in a deep-water subsea pipeline can occur during shutdown or start-up of a pipeline as was the case with Genesis. Under normal flowing conditions, dehydrated gas flows through the uninsulated Genesis export gas pipeline. Although pressures in the line were high (~1,600 psig at inlet) and temperatures as low as 42 °F in the coldest sections of the pipeline, hydrate formation was not expected due to the low water content of gas from dehydration at GC 205. However, in the presence of sufficient water, pressures and temperatures of the flowing stream in the pipeline were conducive to hydrate formation. Since water content of gas entering the pipeline was routinely monitored below hydrate formation content while flowing, hydrates should not have formed in the pipeline during steady state flow. When hydrates did form in the pipeline, many factors may have contributed to that formation such as the multiple shutdowns within a few days of each other and the added complexity in the procedures of shutting the platform down and starting it back up. During these back to back platform shut downs and start-ups, if any water had entered into the pipeline due to non-optimal dehydrator performance, the potential for hydrate formation would have been be high. Overview Genesis is ChevronTexaco's first deepwater oil and natural gas drilling and production facility located in 2,600 ft of water in the Gulf of Mexico, 150 miles south of New Orleans. The field is operated by ChevronTexaco with 56.67% working interest, and partner ExxonMobil with 38.38% working interest and PetroFina Delaware, Incorporated, with 4.95% working interest.

Summary Slug flow is a flow pattern commonly encountered in offshore multiphase flowlines. It is characterized by an alternate flow of liquid slugs and gas pockets, resulting in an unsteady hydrodynamic behavior. All important design variables, such as slug length ad slug frequency, liquid holdup, and pressure drop, vary with time and this makes the prediction of slug flow characteristics both difficult and challenging. This paper reviews the state of the art methods in slug-character sizing and slug-volume predictions. In addition, history matching of measured slug flow data is performed using the OLGA transient simulator. This paper reviews the design factors that impact slug-tracking option under a process-control system. The slug-tracking option of the simulator is applied to predict the slug length and the slug volume during a field operation. This paper will also comment on the performance of common empirical slug-prediction correlations. Introduction There is an increasing need to develop more oil and gas design and development of economical offshore production fields in harsh offshore environments. The special problems associated with offshore operations pose new challenges in the systems. In an offshore environment, the transient effects during startup and shutdown of the production system become more pronounced. New transient simulation tools, such as OLGA and PLAC, are being used more in the design and modeling of these transient effects. Slug flow is one of the most common flow patterns. It is characterized by an unsteady, alternating flow of liquid slugs and gas pockets. Because of its highly complex nature, the prediction of slug length, slug frequency, and pressure drop by theoretical means is almost impossible. Because of its importance to oil and gas production operations, many studies have been carried out and empirical and mechanistic models have been developed. "Steady-state" slug flow can be classified as either hydrodynamic or terrain-induced slugging. "Transient" slugging can also occur in pipelines as a result of changing operating conditions, pigging, or during startup operations. Empirical correlations based on field or laboratory data have been developed to predict the transition to slug flow, the slug velocity, the slug length, the slug frequency, and the statistical distribution of slug lengths. Hydrodynamic slugging is the normal slugging pattern encountered in flowlines. Most of the empirical methods developed for predicting the transition to slug flow were developed for horizontal or near-horizontal pipes. The most often used slug-length prediction methods in the industry for large pipe diameter are the Brill1 correlation and its revisions, such as the Norris2 correlation and the Scott et al.3 correlation. Terrain-induced slugging is induced by low points in the flowline that may shrink or grow after the dip. It is more dynamic and less understood compared with hydrodyanmic slugging. Every flowline through hilly terrain has its own elevation profile, and therefore, has its own slugging characteristics. The worst kind of terrain-induced slugging is severe slugging, caused by an abrupt change from the horizontal to vertical flow directions. Severe slugging is frequently seen in risers. This usually occurs when both gas and liquid flow rates are relatively low.

Slug flow is a flow pattern commonly encountered in offshore Multiphase flow lines. It is characterized by an alternate flow of liquid slugs and gas pockets, resulting in an unsteady hydrodynamic behavior. All important design variables, such s slug length and slug frequency, liquid holdup, and pressure drop, vary with time and this makes the prediction of slug flow characteristics both difficult and challenging. This paper reviews the state of the art methods in slug catcher sizing and slug volume predictions. In addition, history matching of measured slug flow data is performed using the OLGA transient simulator. This paper reviews the design factors that impact slug catcher sizing during steady state, during transient, during pigging, and during operations under a process control system.. The slug tracking option of the OLGA simulator is applied to predict the slug length and the slug volume during a field operation. This paper will also comment on the performance of common empirical slug prediction correlations. INTRODUCTION There is an increasing need to develop more oil and gas design and development of economical offshore production fields in harsh offshore environments. The special problems associated with offshore operations pose new challenges in the systems. In an offshore environment the transient effects during start-up and shut-down of the production system become more pronounced. New transient simulation tools such as OLGA andPLAC are being used more in the design and modeling of these transient effects Slug flow is one of the most common flow patterns. It is characterized by an unsteady, alternating flow of liquid slugs and gas pockets. Due to its highly complex nature, the prediction of slug length, slug frequency and pressure drop by thematically means is almost impossible. Due to its impedance to oil and gas production operations, many studies have been camied out and empirical and mechanistic models have been developed. "Steady state" slug flow can be classified as either hydrodynamic or terming induced slugging. "Transient" slugging can also occur in pipelines as a result of changing operating conditions, pigging or duringstart-up opemtions. Empirical correlations based on field or labotary data have been developed to predict the transition to slug flow, the slug velocity, the slug length, the slug frequency, and the statistical distribution of slug lengths. Hydrodynamic slugging is the normal slugging pattern enounced in flowlines. Most of the empirical methods developed for predicting the tradition to slug flow were developed for horizontal or near horizontal pipes. The most often used slug length prediction methods in the industry for large pipe diameter are the rilll correlation and its revisions, such as the Norris2 correlation, and the Scott et. al.3 Correlation. Terrain induced slugging is induced by low points in the flowline which may shrink, or grow after the dip. It is more dynamic and less understood compared with hydrodynamic slugging. Every flowline through hilly termin has its own elevation profile, therefore has its own slugging characteristics. The worst kind of terrain induced slugging is severe slugging.

Dynamic simulation results of thermal and hydraulic performance of the Gemini deepwater subsea production system of three wells and two pipelines are presented. This paper shows how transient models such as OLGA are used to predict and alleviate the flow assurance problems associated with deepwater production of a gas condensate subsea system. The paper addresses the importance of flow modeling before and during production. The results of this study can be applied to the design of new deepwater production systems in order to maintain flow assurance, and to safely operate a subsea pipeline/riser system. Comparisons of predictions with the measured production data are presented. Simulation ChallengesSteady State and dynamic simulations were performed throughout the conception, design and operation of the Gemini