Layer | Fill | Outline |
---|
Map layers
Theme | Visible | Selectable | Appearance | Zoom Range (now: 0) |
---|
Fill | Stroke |
---|---|
Collaborating Authors
Results
Abstract A Probabilistic/mechanistic modeling was carried out to develop a predictive model for fully developed slug length distribution in horizontal pipeline. Statistical analysis suggested the appropriateness of a Log-Normal model over an Inverse Gaussian model for predicting slug length distribution. A total of 64 data sets were used to empirically correlate the Log-Normal model. Two empirical relationships for mean slug length and slug length standard deviation were developed. A statistical analysis revealed that, in addition to pipe diameter and mixture velocity, volumetric flow rate of the liquid film in the bubble region and the momentum exchange between the slug body and the liquid film are strongly correlated to mean slug length at a 5% significance level. The slug length standard deviation was found to have a significant correlation with film liquid holdup and momentum exchange. A model validation study demonstrated the capability of the probabilistic/mechanistic model to reproduce the experimental data with a satisfactory match. The match was improved when the developed correlations were tuned using the statistical confidence intervals of their coefficients. Introduction The significance of the slug length distribution in pipelines is mainly related to the design of downstream separation facilities. Although knowledge of average slug length in pipelines is important for pipeline design and deliverability calculations, it cannot be used to design separation facilities where more detailed flow characteristics are needed. The maximum slug length is the most crucial flow characteristic for proper separator or slug catcher design. Bernicot et al. proposed a fractal dimension approach to model the slug length distribution in a horizontal pipeline in which global stochastic behavior of slug flow time series is explored. Using experimental data sets, the self-similarity coefficients (H) were estimated for each data set and used to obtain a pre-assigned fractal dimension. The fractal dimension and the Inverse Gaussian distribution coefficients were used to describe a slug flow probabilistic model. Dhulesia et al. investigated slug characteristics, length and velocity in three multiphase field pipelines using a unique set of instrumentation. The data show that it is important to model slug length because of its strong relation to pressure drop. A qualitative relationship between slug length and liquid holdup in the slug body was observed from the field data. However, slug liquid holdup was not included in the available slug length correlations which are based on experimental data. A uni-modal distribution was found for slug length distribution in a horizontal pipeline; however, an unexplained bi-modal slug length distribution was observed in one of the horizontal pipelines. On the other hand, a bi-modal distribution was observed in another pipeline due to terrain induced slugging. An Inverse Gaussian distribution was a better fit to the acquired data than a Log-Normal distribution. Brill et al. developed a correlation from Prudhoe Bay data and a TUFFP experimental data bank to predict the mean slug length in horizontal pipelines, assuming the slug length follows a Log-Normal distribution. The data used include flow in 1.5-in., 2-in., 4-in., 7-in. and 16-in. diameter pipes with different fluids, including live oil and gas, kerosene and air, and water and air. Independent variables in the correlation are mixture velocity and pipe diameter. This correlation was modified by Norris using part of an expanded data set that included data from a 24-in. diameter flowline. In the Norris modified correlation, the mixture velocity term is excluded, which was found to be negligible. Brill et al. also presented a statistical based correlation to predict the maximum slug length in a pipeline as a function of the Log-Normal distribution parameters for a certain slug length distribution. This correlation is based only on the Prudhoe Bay pipeline data, which specifically gives the longest possible slug in 1000 observed slugs. Gopal and Jepson developed a mechanistic model to predict mean slug length and investigated the effect of liquid phase properties on the mean slug length. Their mechanistic model uses the Froude number of both the liquid film and the slug body. They found that the Froude number for the liquid film is always greater than unity, decreases in the slug-mixing zone, and tends to increase toward the tail of the slug where it reaches unity. This criterion was used to determine the slug length.
- Research Report > New Finding (1.00)
- Research Report > Experimental Study (1.00)
- Production and Well Operations > Well & Reservoir Surveillance and Monitoring > Production logging (1.00)
- Facilities Design, Construction and Operation > Pipelines, Flowlines and Risers > Pipeline transient behavior (1.00)
- Data Science & Engineering Analytics > Information Management and Systems (1.00)
ABSTRACT Two options are available for separating the gas liquid mixture at the exit of a two phase flow pipeline operating under slug flow conditions. These are a traditional vessel type separator and a finger storage type slug catcher. Use of a vessel separator is usually due to space limitations that exist, for example, on offshore platforms. Fingerstorage slug catchers are the obvious choice for long, large diameter pipes, especially those which undergo pigging. They are more cost effective and more simple to construct and operate. In the past, sizing of finger storage slug catchers were based primarily on experience andrules of thumb. Not surprisingly, most of the existing slug catchers have been oversized. Withthe recent trend of using a subsea compact finger storage slug catcher upstream of the platform riser, the need for more accurate design methods is even more crucial. This paper presents a new, innovative approach for the prediction of the required dimensions of slug catcher fingers. The approach is based on the effect of the finger pipe diameter and inclination angle on the transition boundary between slug flow and stratified flow. Predictionof the slug characteristics at the slug catcherinlet, under normal flow or pigging conditions, are incorporated. Based on the new approach, the required length and optimal downward inclination angle of the fingers can be determined. The new approach has been used to design a finger storage slug catcher for actual field conditions. The effect of operational conditions, e.g. pipeline diameter and finger inclination angle on the required slug catcher dimensions isdemonstrated. INTRODUCTION Pipelines can be operated under two phase flow conditions for several reasons. In hostile environments such as arctic and offshore fields, oil and gas are often transported in a singlepipeline to reduce the construction cost. In natural gas transportation, due to pressure andtemperature drop during flow in the pipeline, condensation causing two phase flow may occur. Depending upon the operating conditions, normal or terrain induced slug flow, may develop. Also, artificial slugs, possibly the largest ones, can be created during the removal of accumulated liquid by a pigging (sphering) operation in gas pipelines. It is a common practice to install a slug catcher to accommodate liquid slugs at the exit of a pipeline. A slug catcher can serve as both a separator and as temporary storage. There are several unconventional slug catcher types, such as a project slug catcher, a self supporting fluid separator, and a flexible subsea slug catcher. However, the vessel and finger storage types of slug catchers are the most widely used in the petroleum industry. Use of traditional vessel type separators as slug catchers are mainly dictated by space limitation and relatively small slug sizes. A number of studies have been conducted to design such catchers4 . In references 4, 5, and 6 the acceleration of the slugs during their production into the catcher and resultant loadson bends, fittings and slug catcher internals have been investigated for a specific slug catcher.
- North America > United States > Alaska > North Slope Basin > Prudhoe Bay Field (0.99)
- Europe > Norway > North Sea > Northern North Sea > North Viking Graben > PL 054 > Block 31/6 > Troll Field > Sognefjord Formation (0.99)
- Europe > Norway > North Sea > Northern North Sea > North Viking Graben > PL 054 > Block 31/6 > Troll Field > Heather Formation (0.99)
- (10 more...)
Abstract A mathematical model has been developed for a gas-oil separator system, and a computer program has been written to simulate the behavior of the system. The separator is assumed to operate under slug flow conditions at the end of a two-phase flow pipeline. The liquid level and the pressure of the vessel are controlled by pressure of the vessel are controlled by proportional integral controllers. proportional integral controllers. An optimization program has been written based on a linear model of the separator system. The program applies linear control theory to optimize program applies linear control theory to optimize the gains and reset times of the controllers. The optimization and simulation programs can be used to design the smallest possible separator that will handle given gas and liquid flow rates in slug flow production. A sample design case is given. Introduction A slug catcher is a gas-liquid separator that performs primary separation of gas and liquid, performs primary separation of gas and liquid, and is usually located at the end of a large diameter pipeline and upstream from any processing facilities. The most common flow processing facilities. The most common flow pattern received by a slug catcher is slug flow. pattern received by a slug catcher is slug flow. Slug flow is characterized by alternate flow of gas and liquid resulting in a wide fluctuation in liquid input rate into the slug catcher. A slug catcher should provide more steady liquid flow into the liquid discharge line. Very few studies have been conducted for simulating the behavior of vessel type slug catchers. A summary of previous work was presented by Giozza. The following modifications presented by Giozza. The following modifications and improvements of the previous studies are implemented.A mechanistic slug flow model is developed. The randomness in the slug length distribution is taken into account. The effect of maximum expected slug length was considered when determining the size of the vessel. A material balance between the output and input rates is ensured. The delays in the pneumatic transmission lines and the control valves are considered. Liquid level fluctuations in the vessel are tolerated in order to have smaller fluctuations in the liquid discharge rate. Stability analysis of the slug catcher system is performed and based on this analysis, the pressure and liquid level controller parameters are optimized. P. 549
Abstract In a major oilfield in Saudi Arabia, detailed Information on individual well performance is required for operating and planning purposes. Complete design data can only be obtained planning purposes. Complete design data can only be obtained through the addition of extensive facilities for periodic testing of all production in the field. Three different test pipeline configurations were considered. Variables included pipe diameter, pipeline length, oil production rate, gas/liquid ratio, and water cut. Lengths up to approximately 10 miles (16.1 km) were considered. Three types of analyses were undertaken. The first analysis included steady state pressure losses In not only the horizontal segments but also the vertical risers and downcomers at the platforms. The second set of calculations involved transient simulations for both increasing and decreasing flow rates. Finally, slug length and period predictions for both normal and "severe" slugging were predictions for both normal and "severe" slugging were performed. Severe slugging occurs only at very low no"' rates performed. Severe slugging occurs only at very low no"' rates and negative pipeline inclination angles. Pressure drop calculations based on the Beggs and Brill correlation, with a rough pipe friction factor, showed that 10 in. (25.4 cm) pipelines were adequate for most cases. Transient simulation results indicated that, in general, less than two hours were required to reestablish steady state flow for both increasing and decreasing rates. Normally occurring slugs were found to vary in average length from about 200 to 600 ft (61 to 183m) with a maximum possible slug length of approximately 2500 ft (762 m). Liquid slug lengths for severe slugging varied from about 100 to 800 ft (30. 5 to 244 depending on the pipeline length and riser height. Introduction In a major offshore oilfield in Saudi Arabia, detailed information on individual well performance is required for operating and planning purposes. The desired information can only be obtained through the addition of extensive facilities for periodic testing of all wells in the field. Three different possible test pipeline configurations were considered and are designated as alternatives A, B, and C. The various alternatives would have test stations located either on production platforms or on separator platforms (GOSPs). Some test stations could be located platforms (GOSPs). Some test stations could be located onshore with others located offshore. Each of the alternatives also has a variation in pipe diameter, pipeline length, production rates, GLRs and water cuts. Table 1 indicates the variables and the number of each variable to be considered for calculations. Because of the large number of data to be considered, it was necessary to develop the coding system described in Table 2. Nominal pipeline diameters of 8, 10 and 12 in. (20.32, 25. 40 and 30. 48 cm) were used for all calculations. Schedule 40 pipe was assumed with internal diameters of 7.981, 10.020 and 11.938 in. (20.27, 25.45 and 30.32 cm), respectively. Average lengths vary from 13,000 to 29,000 ft (3962.4 to 8839.2 m) with maximum pipeline lengths of 22,000 to 50,000 ft (6705.6 to 15240.0 m). Production rates to be considered were 2,000 and 25,000 STBL/D (318. 0 and 3974. 7 m3/d) and an average production rate of 8,000 STBL/D (1271.9 m3/d). Three types of two-phase flow calculations were performed for each alternative. The first calculations were performed for each alternative. The first calculations were steady state pressure losses for the various lengths, diameters, flow rates, gas-liquid ratios, and water cuts. The second calculations performed were transient simulations of flow behavior when inlet flow rate changes occurred. These simulations were necessary for predicting the length of time required to reestablish the steady state flow conditions required to obtain meaningful well test data. The final type of calculations performed were to estimate anticipated liquid slug length and other slug characteristics. For all three types of calculations, at least two methods were used to confirm results. Following sections describe these calculations and general results. Appendix A contains figures that Illustrate the four pipeline configurations considered for alternatives A, B and C. P. 607
- North America > United States (0.68)
- Asia > Middle East > Saudi Arabia (0.44)