Schoppa, Wade (Shell International Exploration and Production Inc.) | Zabaras, George J (Shell International Exploration and Production Inc.) | Menon, Raghu (Shell India Markets Pvt. Ltd.) | Wicks, Moye (Shell International Exploration and Production Inc.)
An experimental program was carried out to support the application of riser-base gas injection for deepwater production risers. The experimental equipment consisted of a 40 ft-tall, 11?? ID vertical riser connected to a 25 ft-long flowline section of the same inside diameter. The motivation for this work was the significant uncertainty that exists as to the behavior of vertical gas-liquid flow in large diameter pipes. Scarce published information exists for such flows and no known records of flow regime observations are available from the published literature.
Riser-base gas-lift has been identified by various deepwater project teams as a key operability element that can address three major project objectives:
a)flow assurance requirements necessitate the evacuation of the riser during an extended shut-down by using injection gas. As a result of riser evacuation, the hydrostatic pressure at the base of the riser is reduced to a level below the pressure required for hydrate formation at the ambient seawater temperature of between 37 and 40 F, b)production economics for many deepwater projects relies on riser-base gas-lift as an artificial lift method and, c)riser-base gas injection will be necessary for alleviating severe slugging problems in deepwater flowline-riser systems.
In order to verify the ability of gas-lift to fulfill the stated objectives for large diameter risers, a variety of air-water flow tests were carried out in the 11?? transparent riser system built at Shell's Houston Technology Center. The flow tests included transient and steady-state gas-liquid flow and flow regime visualization. Measured riser pressure drop data were used to compare with 1-dimensional flow model predictions. Serious model limitations were found as the models failed to accurately capture "churn flow?? effects that appeared to dominate over the range of conditions studied. Steady-state pressure drops were underpredicted by more than 25%. Riser evacuation predictions indicated continued liquid removal whereas the data showed that liquid is expelled as an initial slug with no additional carry over.
To overcome the shortcomings of 1-dimensional models, rigorous analyses using Computational Fluid Dynamics (CFD) simulations were performed. A full 3-dimensional CFD model of the flowline-riser system was used in FLUENT with the volume of fluid (VOF) method to predict both steady-state and transient behavior. In all cases simulated, CFD results compared well with the experimental data.
The goal of this work is to develop an improved model of CO2 bubble rise through porous media in the deep subsurface. Under the geologic carbon sequestration (GCS) conditions of interest, a rising parcel of CO2 will be subject to at least three dynamic forces: 1) buoyant forces; 2) surface tension forces; and 3) shear drag forces. To fully characterize these, this work involved several experimental measurements focused on the second and third forces in particular. To better understand the effect of shear drag forces, the viscosity of brines was explored under bubbly flow scenarios to understand the rheological conditions that might impact leakage. To better understand the role of surface tension forces on influencing flow, contact angle measurements were carried out for a range of relevant mineral, brine, CO2 combinations. Predicting leakage from geologic carbon sequestration sites is difficult because of the large length scales that are involved and because of the complex geophysics and geochemistry that a rising parcel of CO2 will be subject to as it travels to the surface. To better understand how quickly and where a parcel of CO2 is likely to escape, better modeling tools are needed. These tools must be based on experimental results collected for GCS-relevant conditions. The results of the brine viscosity work suggest that under vapor liquid equilibrium (VLE) conditions CO2-brine mixtures will exhibit complex viscoelastic behavior. This is because CO2 bubbles in the matrix will respond to the varying levels of shear that will exist in the porous media to resist flow. Similarly, the contact angle measurements suggest that CO2 is less wetting of some common minerals and clays that prevail near GCS sites. The experimental results described here will be used to describe an enhanced model of CO2 vertical flow through the subsurface. At smaller scales, this enhanced model could help explain preferential flow pathways and potential hysteresis that could influence leakage from GCS sites. At larger scales, the results of this work could contribute to more accurate prediction tools for managing the risk associated with GCS.
Keywords: Carbon sequestration, leakage/seepage, flow through porous media, bubble flow, contact angle.
Foamy-oil viscosity is a controversial topic among researchers regarding what happens to the oil viscosity when the solution gas starts coming out of solution because of decreasing pressure and the released gas starts migrating with the oil in the form of dispersed gas bubbles. For conventional oils, below the true bubblepoint pressure, the oil viscosity increases as the gas freely evolves from the oil. For foamy oils, it has been suggested that the apparent oil viscosity remains relatively constant or perhaps declines slightly between the true bubblepoint and a characteristic lower pressure, called pseudobubblepoint, which is the pressure at which the gas starts separating from the oil. Below this pressure, the viscosity increases, reaching the dead-oil value at atmospheric pressure. However, it is a well-known fact in dispersion rheology that the viscosity of dispersion is higher than the viscosity of the continuous phase. Therefore, the concept of foamy-oil viscosity being lower than the oil viscosity is counterintuitive. It is likely that the apparent viscosity for flow of foamy oil in porous media is not the true dispersion viscosity because of the size of dispersed bubbles being comparable to the pore sizes.
This study investigates this issue by measuring the foamy-oil viscosity under varied conditions. The effect of several parameters, such as flow rate, gas volume fraction, and type of viscometer employed, on foamy-oil viscosity was evaluated experimentally. Three different viscosity-measurement techniques, including Cambridge falling-needle viscometer, capillary tube, and a slimtube packed with sand, were used to measure the apparent viscosity of gas-in-oil dispersions. The results show that the type of measuring device used has a significant effect. The results obtained with Cambridge falling-needle viscometer correlate better with the observed behaviour in the sand-packed slimtube than the capillary viscometer results. Overall, the apparent viscosity of foamy oil was found to be similar to live-oil viscosity for a range of gas volume fractions.
Foamy oil viscosity is a controversial topic among researchers as to what happens to the apparent oil viscosity when the dispersed gas bubbles start migrating with the oil. For conventional oils, below the true bubble point pressure, the oil viscosity increases as the gas freely evolves from the oil. For foamy oils, it has been suggested that the apparent viscosity of gas-in-oil dispersion remains relatively constant, or perhaps declines slightly, between the true bubble point and a characteristic lower pressure, called pseudo bubble point, which is the pressure at which the gas starts separating from the oil. Below this pressure, the viscosity increases, reaching the dead oil value at atmospheric pressure. However, it is a well known fact in dispersion rheology that the viscosity of dispersion is higher than the viscosity of the continuous phase. Therefore, the concept of foamy oil viscosity being lower than the oil viscosity is counterintuitive. The major difference here is the extreme viscosity of the base liquid phase for foamy oil and how this interacts with the gas phase in a porous medium. The reported results appear to be very oil specific in this area, and are also a very strong function of how rapidly pressure is depleted in a given system. It is also likely that the apparent viscosity for flow of foamy oil in porous media is not the true dispersion viscosity due to the size of dispersed bubbles being comparable to the pore sizes.
This study aims to investigate this issue by measuring the foamy oil viscosity under varied conditions. The effect of several parameters, such as shear/flow rate, foaminess and gas volume fraction on foamy oil viscosity was experimentally evaluated.
A slim tube packed with sand was used to measure the apparent viscosity of gas-in-oil dispersions. The results show that the apparent viscosity of foamy oils is considerably higher than the live oil viscosity. . The results obtained also showed that the presence of an added foaming agent had only a minor impact on the apparent viscosity of foamy oil, especially at higher expansion factors. Also, at the same expansion factor, the apparent viscosity was higher at higher flow rate. Overall, the presence of a foaming agent resulted in enhanced dispersed flow of gas, as evidenced from the size of bubbles being produced and the observed pressure fluctuations
Head deterioration observed in electrical submersible pumps (ESPs) under two-phase flow is mild until a sudden performance breakdown is observed in the pump head curve at a certain volumetric gas fraction. This critical condition is termed surging. Consequently, the head that the pump generates with two-phase flow depends on whether the stages operate under conditions before (mild performance deterioration) or after (severe performance deterioration) the surging point.
The surging, for engineering purposes, can be predicted by published correlations, but the lack of a theoretical basis is a limiting factor for their application. Mechanistic models seem to be the proper alternative. However, the poor understanding of the physical mechanism that causes the surging hinders the development of such mechanistic models. This paper reviews some of these correlations and mechanistic models by comparing the correlation predictions against experimental data acquired in a closed loop with water and air using a commercial 24-stage ESP. The data cover a wide range of volumetric gas fraction, rotational speeds, and intake pressures. As a consequence of this analysis, a new correlation has been formulated. This correlation predicts the initiation of the surging as a function of rotational speed and fluid properties.
This paper provides an analysis of a widely-used pressure-volume-temperature (PVT) parameter, demonstrating common errors made in its use and the lack of detailed descriptions in common reservoir engineering texts. Use of the Y-function is a valuable tool for increasing the accuracy of bubble point measurements, and it has other important uses, but its detailed application is poorly understood, leading to unnecessary errors in one of the most fundamental PVT properties.
Best practices have been developed and are presented, to encourage proper application of the Y-function, together with detailed recommendations for setting up Excel spreadsheets to help identify the most precise bubble point, and optimize Y-function values.
More extensive use of the Y-function to obtain bubble points for down hole or surface oil samples is recommended as it could improve quality control and selection of more representative samples.
Increasing trend in the transportation of unprocessed reservoir fluid from assets located in remote locations and the drive towards maximising existing facilities and optimising production has encouraged many deepwater operators in the use of long subsea tiebacks.
However, the main challenges associated with unprocessed multiphase fluid flow in long subsea pipelines relate to constantly changing patterns and the likelihood of the suspended particles to settle out of flow and deposit in the transfer pipeline causing partial or complete blockage. Flow pattern transition is therefore a critical factor that must be accounted for in any particle transport models. However, current particle transport models do not account for these critical factors; besides the experimental data on which these models were validated are limited. Therefore they are often unreliable when subjected to varying operating conditions.
In this paper preliminary minimum transport velocity (MTV) models developed for rolling and suspended particles under different flow patterns are presented. The concept of particle velocity profile provided the basis on which the models were developed. Further work is planned to acquire large experimental data for the purpose of validating these proposed models.
The problem has been largely attributed to insufficient flow velocity among other parameters required to keep the solids in suspension and prevent them from depositing in the pipe. Additional complexities are introduced because of different flow regimes that occur within the pipe flow depending on the gas and liquid flow rates.
The development of minimum transport velocity models for suspension and rolling based on the concept of particle velocity profiles is a significant breakthrough in particle transport in multiphase flow. This has the potential to solve problems of pipe & equipment sizing, risk of sand deposition & bed formation, elimination of costs of sand unloading, downtime and generally improve sand management strategies.
Pressure drop and stability characteristics of the flow in subsea flowlines are important to the proper operation of offshore platforms, and they depend intimately on the flow regimes that occur. Air-oil-water three phase flows in simultaneous horizontal and vertical pipe orientation are presented. Fast-sampling (250 Hz) gamma densitometer units were installed at the top of the 50.8mm diameter, 11m high vertical riser and horizontally near the riser base in the Cranfield University multiphase flow test facility. Gamma radiation attenuation data were collected from the caesium-137 radioisotope-based densitometer for a range of air-oil- water flow mixtures spanning the facility's delivery range. Examination of the gamma densitometer signal response revealed the presence of quasi-periodic waveforms in the time-varying multiphase flow densities. The probability mass function (PMF) characteristics predict the flow in the horizontal flow loop as bubbly, slug and wave flows while the vertical riser flow regimes were bubble, slug and churn. Flow pattern maps were then developed based on these PMF plots for both horizontal and vertical pipe orientation. The effect of upstream conditions on the vertical riser flow behaviour was also investigated via two different air inlet configurations: (i) upstream flowline mixing and (ii) riser base injection. No significant difference exist in flow behaviour at low superficial air-liquid velocities for both configurations, but at higher superficial air-liquid velocities, the intermittent flow behaviour due to hydrodynamic slugging in flowline influences the riser flow pattern characteristics, thus controlling the riser dynamics.
Keywords: gamma densitometry, multiphase flow, flow regime, flowline-riser, signal analysis.
The fully automated fluid level measurement tool was developed recently. The paper describes the technical features of the tool and shows via case study the results of the field tests on various Electrical Submersible Pumps - ESPs and sucker rod pumps - SRP, running with and without Variable Speed Drive - VSD.
The unique feature of this system is its fully automated and purely electronic functioning. The measuring device is enclosed, mounted on the casing valve, has a pressure rating of 350 bar (5000psi) and works with zero emissions on the environment (no outlet of casing gas).
Compared with a conventional downhole pressure sensor, mounted on an ESP, the system is insensitive to high well fluid temperatures and simple to maintain due to its easy access on surface location. It additionally has a sampling rate of down to one measurement per minute. The measured fluid level data can be transmitted via SCADA system.
The measurement tool enables to run a pump in a more safe way. It can be used to avoid pump off conditions and the resulting equipment damage. It can also be used to control a VSD to keep the fluid level in a well at a specific depth to avoid bottomhole flow conditions below the bubble point pressure. Due to the availability of online fluid level data, all kind of pumps (e.g. ESP, Sucker Rod, PCP, Jet Pump) can be operated safely at more aggressive production rates in order to mobilize more oil.
Furthermore the vision and corresponding thoughts of acoustic well diagnosis, also a feature of the tool which is currently under investigation, are presented.
The low frequency spectral characteristics of signals emitted from marine airgun arrays are constrained by the source ghost, the response of the air-guns, and the air-gun array design. Both the ghost and the response of air-guns are related to the tow depth(s). This paper analyses the competing issues that must be addressed when trying to improve the low frequency output of air-gun arrays.