Hydrate Management in Deadlegs: Effect of ... - ACS Publications

Subscriber access provided by University of Florida | Smathers Libraries

Article

Hydrate Management in Deadlegs: Effect of Header Temperature on Hydrate Deposition Xianwei Zhang, Bo Ram Lee, Jeong-Hoon Sa, Keijo J. Kinnari, Kjell Magne Askvik, Xiaoyun Li, and Amadeu K. Sum Energy Fuels, Just Accepted Manuscript • DOI: 10.1021/acs.energyfuels.7b02095 • Publication Date (Web): 24 Oct 2017 Downloaded from http://pubs.acs.org on October 25, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Energy & Fuels is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

Hydrate Management in Deadlegs: Effect of Header Temperature on Hydrate Deposition Xianwei Zhang,1† Bo Ram Lee,1,2† Jeong-Hoon Sa,1 Keijo J. Kinnari,3 Kjell M. Askvik,4* Xiaoyun Li,5 and Amadeu K. Sum1* 1

Hydrates Energy Innovation Laboratory, Chemical & Biological Engineering Department, Colorado School of Mines, Golden, CO 80401 – USA 2

Department of Chemical Engineering, Pohang University of Science & Technology, Cheongam-Ro, Nam-Gu, Pohang, Gyeongbuk, 37673–KOREA 3

Statoil ASA, N-4035 Stavanger – NORWAY 4

5

†

Statoil ASA, N-5020 Bergen – NORWAY

Statoil ASA, N-7005 Trondheim – NORWAY

Author contributions: X. Zhang and B.R. Lee contributed equally to this work.

ACS Paragon Plus Environment

1

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 31

ABSTRACT Deadlegs in oil and gas production systems often encounter hydrate plugs by deposition. Temperature is generally known to be an important variable in hydrate formation, but the effects in deadlegs are not exactly known. This study focuses on the effects of the header temperature on the hydrate deposition in gas-filled vertical deadlegs at constant wall temperature. All the experiments are conducted with a methane/ethane gas mixture at constant pressure. The pipe wall temperature is kept at constant, while considering different header temperatures. The tests show that the header temperature has a significant impact in the hydrate deposit growth rate and distribution in the deadleg. It is also found that the hydrate deposit can, in turn, change the temperature field inside the pipe. The header temperature or the pipe temperature field can be used to estimate the hydrate distribution in the deadleg. Under the right conditions, hydrates can form a restriction in the deadleg and its location is usually close to the boundary of a hydrate stable region. The location of the restriction can be correlated to the header temperature. At 80 °C, the location is estimated to be 15-18 ID and at 30 °C, the location is estimated to be 9-12 ID. The results of this study contribute to the understanding of the hydrate deposition mechanism in deadlegs.


2

Page 3 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

INTRODUCTION Gas hydrate is a crystalline structure formed by water and small guest molecules usually at low temperature and high pressure.1 In the oil and gas industry, hydrate is usually undesirable, as its occurrence in the pipeline, if not properly managed, can cause capital losses and raise safety concerns.2 Deadlegs, pipe sections which have no through flow, are common features in a subsea and topside production system and are often exposed to the risk of hydrate formation.3 Deadlegs are typically process lines and instrument lines, but they could also represent sections of the main flowline without flow. Deadlegs constitute of many pipe types with different sizes and geometries and can be categorized based on their diameter and length.4 Deadlegs are usually connected to a main flowline and are usually colder than the main flowline because the fluids in the deadleg are nearly stagnant. The fluids insides a deadleg can contain gases, liquids, or multiphases depending on the conditions. Depending on system conditions, deadlegs can face different challenges. For example, in oil and gas production systems, liquid-filled deadlegs have been reported to have corrosion issues if free water phase exists.5,

6

From practice, gas-filled

deadlegs are particularly prone to hydrate deposition and possibly hydrate blockages.4 The consequences of hydrate deposition blockage of these deadlegs can be critical. Potential usage of deadlegs can be chemical injection, depressurization, service line, and remediation to remove hydrate blockage. A hydrate blockage can prevent the usage of deadlegs in production. Several deadleg cases related to hydrate challenges have been reported by Kinnari et al.4 To date, despite the importance of hydrates formation in deadlegs, studies on this topic are scarce in the open literature. In industry, the current hydrate management strategies are based


3

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 31

on experience (rule-of-thumb) and are rarely reported. Most relevant studies to date focus on deadlegs without hydrates or hydrate deposition in the main flowline. A number of related studies have been done on liquid-filled deadlegs. Bloom investigated the mixing region in vertical downward deadlegs with isothermal water.7 The dimensionless mixing depth is correlated to the flow Reynolds number and the fluid kinematic viscosity. Habib et al. experimentally and numerically studied the velocity field of an isothermal oil/water flow and its separation in deadlegs with different geometries.5, 6 In their study on vertical downward deadlegs, L/ID = 3 was found to be a critical value after which the liquid velocity becomes negligible. Asteriadou et al. reported similar studies on non-isothermal water flow in vertical downward deadlegs with different inlet velocities.8 They found that the center temperature in the deadleg decreases exponentially after a threshold depth. These studies provide basic understandings of the velocity and temperature field inside liquid-filled deadlegs. Gas-filled deadlegs is prevalent within oil and gas pipelines. Anderson used CFD to evaluate the temperature field and then hydrate risks in a gas-filled deadleg.9 The study shows the effects of the header velocity and the possible natural convection on the velocity and temperature field. However, because the calculations did not include hydrate and there was no experimental data for verification, the study lacked sufficient data to suggest any design criteria or risk evaluation of deadlegs. Only one experimental study has been reported on the hydrate formation in a gas-filled deadleg, by Nazeri et al.10 In this study, a 3-inch-ID and 20-inch-long pipe is used for form hydrate with a temperature gradient from 20 °C water at the bottom to 4 °C at the top. The major


4

Page 5 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

conclusion from the study was the confirmation of hydrate deposition in the deadleg. However, there is no mention of the investigated variables or hydrate deposition mechanism. Hydrate deposition has been speculated to be a significant problem in real flowlines and was only recently confirmed in bench-scale rocking cells and flowloops.11-16 The research investigated hydrate deposition at various conditions, such as different flow patterns, but no stagnant flow conditions were covered. Two possible mechanisms of hydrate deposition have been identified: through water condensation and through particle attachment. The water condensation mechanism states that the water vapor in the warm bulk fluid in a flowline can condense on the cold pipe wall. Given enough subcooling and a supply of gas molecules, the condensed water can convert into hydrate. This mechanism is also the most likely in gas-filled deadlegs. As can be seen, the knowledge for hydrate formation in deadlegs is scarce, as such, there is a great need to better understand hydrate deposition in deadlegs so better hydrate management strategies can be developed to minimize the risk of hydrate blockage in oil and gas flowlines. This work represents a part of the first systematic study to quantify hydrate deposition in a deadleg. This paper is the first in a series that will discuss the effects of header temperature on hydrate deposition. The experimental apparatus mimics a gas-filled vertical straight uninsulated deadleg. The direct liquid cooling and the wall temperature are chosen to mimic the seafloor environment. Header temperature is observed to affect the temperature field, the hydrate distribution, and the hydrate growth rate in a deadleg. EXPERIMENTAL SECTION EXPERIMENTAL SETUP


5

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 31

One of the first challenges in this project was to design and build a deadleg system. Figure 1(a) shows the deadleg system used in the study, which was first introduced in detail elsewhere.5 The system includes: i) a vertical straight pipe to mimic a deadleg, ii) chillers and cooling jackets to control the pipe wall temperature, iii) a water reservoir to mimic a main flowline header as well as to provide a water source for hydrate formation, and iv) a syringe pump to maintain the system pressure. The pipe is 7.4 cm (3-inch) in inner diameter and 121 cm in length. The cooling part is 115 cm and is divided into five sections. Each section is 23.0 cm and the wall temperature in each section can be independently controlled. The distance of each section from section 5 is listed in Figure 1(b). The uncooled part is above section 1 and is insulated with plastic foam. The adapter connecting the pipe to the header is 30 cm and uninsulated. The temperature is measured by several intrusive resistance temperature detectors (RTDs) located at each section and the header. No temperature sensor was used in one of the tests (experiment 3) in order to evaluate the possible artifacts brought by the RTDs inside the system. There are at least two RTDs at each section: one 3.7 cm away from the pipe wall (center position), and another 2.0 cm away from the wall (off-wall position). The pressure of the system is monitored by sensors located in the syringe pump and the header. The header has a water level gauge to monitor the water consumption/recovery. A camera is mounted at the top of the pipe (sealed with a glass window) to monitor hydrate growth inside the pipe. The setup of the system allows for control of the header temperature, the wall temperature for each of the section in the pipe, and the system pressure.


6

Page 7 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

Figure 1. Schematic illustration of the deadleg system used for hydrate deposition studies. (a) Main components of the system include: vertical pipe, header (water reservoir), chillers, and syringe pump. (b) The relative distance of each section in terms of ID. EXPERIMENTAL PROCEDURE Each experiment has two stages: hydrate formation and hydrate dissociation. In the formation stage, approximately 2.0 liters of water is injected to the header. The deadleg system is vacuumed and pressurized with a methane/ethane (75/25 mol.%) gas mixture. The system is then left for 24 hours to reach an equilibrium state. The experiment starts by setting the wall temperature (Tw) to 4 °C (seafloor temperature) and the header temperature (Tr) to the desired value (fluid in the main flowline temperature). The syringe pump is then started to maintain the system pressure constant at 100 bar. All the temperature and pressure data are monitored and


7

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 31

logged by a data acquisition system. The cameras are used during the entire experiment to monitor the hydrate formation inside the pipe. In the dissociation stage, one of two methods is used: i) controlled hydrate dissociation under pressure or ii) borescope inspection after depressurization and controlled dissociation. In method i, Tr is first lowered to 30 °C to reduce the water transfer and hydrate growth. After the system cools down, Tw is increased section by section from section 5 (bottom) to section 1 (top) to 25 °C to dissociate the hydrate deposit. The system pressure and the header water level are recorded during the entire process so that the amount of hydrate in each section can be estimated. In method ii, Tw for all sections is first lowered to -5 °C and then the system is depressurized so that the top and side windows can be removed. Details of the depressurization is described in the Supporting Information. A borescope is next inserted into the system to visually and closely inspect the deposit, which is a combination of ice and hydrate. Because the temperature is below 0 °C, any hydrate dissociated will convert to ice, but because the dissociation process is relatively slow, it is expected that most of the observed solid is hydrate. The solid deposit is finally dissociated by increasing Tw section by section. The water recovered during the dissociation from each section is monitored to calculate the amount of hydrate and the hydrate properties. The conditions of the experiments reported in this study are listed in Table 1. All experiments are performed at 100 bar with methane/ethane (75/25 mol.%), with a hydrate equilibrium temperature (HET) of 18.9 °C from CSMGem v1.10, which is close to those predicted by Multiflash v6.1.25 (CPA Infochem model) and PVTSim Nova v3.1.117 (SRK Peneloux model) of 19.6 °C and 19.3 °C, respectively. Tw is maintained at 4 °C. The study


8

Page 9 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

focuses on the Tr. Experiments 1, 2, and 3 are at Tr = 80 °C and they correspond to experiments to confirm the repeatability, as well as to estimate the time evolution of hydrate deposition. The discussion about the repeatability is included in the Supporting Information. Experiments 4, 5, and 6 are at Tr = 60, 40 and 30 °C, respectively. The experiment duration of experiments 1, 4, and 5 were determined by the time we observed signs that the deposit had completely filled the pipe cross section. The experiment duration of experiment 6 is not determined at the start. The purpose was to test whether the system can be observed to approach a steady state. There was no clear indication that the system had reached steady state after 2006 hours (about 83 days), and as such, the experiment was stopped. Table 1. List of experimental conditions for tests performed to quantify hydrate deposition in the deadleg system.

Exp. #

Tr (°C)

Duration (hours/days)

Dissociation method

Borescope Inspection

1*

80

1004 / 41.8

i

No

2

80

474 / 19.2

ii

Yes

3

80

380 / 15.8

ii

Yes

4

60

860 / 35.8

i

No

5

40

887 / 37.0

ii

Yes

6 30 2006 / 83.6 ii Yes 4 *Some data for this experiment are first shown in our previous paper , including the center temperature profiles, the top window observation, and the water/gas recovery. RESULTS The temperature profiles in each section over the course of representative experiments are shown in Figure 2(b) (off-wall) and Figure 2(c) (center). The header and the pipe wall reach the


9

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 31

desired temperature within 1.5 hours. The header acts as the only heat source in the system. When Tr decreases, the temperature throughout the pipe decreases. The pipe wall is the only active cooling sink, thus low temperature is observed near the wall. In the initial period of the experiment, after the point when hydrate deposition initiates, for any point inside the pipe, the temperature is observed to increase, as in Figure 2(a). The increase is caused by the insulating effect of the hydrate deposit.17 Pure hydrate has a heat conductivity in the order of 0.1 W/m/K1, 18-20

, which is much smaller than the stainless pipe wall. The deposit layer thus reduces the heat

flux through the pipe and increases the inner gas bulk temperature.

Figure 2. Temperature profiles. (a) Center for the initial 20 hours. (b) Off-wall. (c) Center. Hydrate stable region is unshaded. The black, red, blue, and orange lines represent experiment 2 (Tr = 80 °C), experiment 4 (Tr = 60 °C), experiment 5 (Tr = 40 °C), and experiment 6 (Tr = 30 °C), respectively. At the relatively low and central positions, the temperature after the initial increase is usually higher than the HET, where hydrate cannot form. Temperature usually stays stable


10

Page 11 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

afterwards. At the relatively high and side positions, temperature is closer to the wall temperature, which is usually below the HET and hydrate can potentially form. Temperature usually decreases and gradually approaches Tw. The temperature profiles of the intermediate area are interesting as the temperature fluctuates after the initial increase. The temperature fluctuates when the temperature falls below the HET. It is also found that the sections that have temperature fluctuations are the section(s) with or just above where the most amount of hydrate forms. All these pattern changes are related to the thermodynamics and the kinetics of the hydrate deposit growth, which is discussed in the following paragraphs. Figure 3 shows the hydrate deposit growth in section 1 recorded by the camera at the top window. Different hydrate deposit growth rates can be observed in section 1 from Figure 3(a) to Figure 3(d) as the Tr decreases from 80 °C to 30 °C. The growth rate from the wall can be seen an decrease which is related to the temperature profile in Figure 2. Because the gas bulk inside the pipe is warm and the pipe wall (or hydrate deposit) is cold, water condenses on the surface. When the temperature falls below the HET, the condensed water can convert into hydrate. When Tr = 80 °C, water condensation on the top window is significant, forming a hydrate deposit on the window and blocking the view into the pipe in the later stages of the experiment; whereas when Tr = 60 °C or lower, only a slight amount of water condensation on the top window is observed. The observation of the deposit growth rates in Figure 3 shows the same trend as the observation of condensation rates. Hydrate growth is also observed on the RTDs, because the RTDs also provide a surface for condensation. The hydrate deposit growth on the RTDs also matches with the temperature profile changes, which provides another evidence for the relationship between the temperature and the deposition. When Tr = 80 °C (Figure 3(a)), the center of section 1 is so warm that only a


11

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 31

small amount of hydrate deposit forms on the center RTD. When Tr = 60, 40, and 30 °C, the center temperature falls lower than the HET, which promotes hydrate formation. However, there is less condensation in section 1 from Tr = 60 to 30 °C due to a decreasing axial temperature gradient in section 1 and thus less hydrate deposit. When Tr decreases, more hydrates can be expected closer to the header.

Figure 3. Images of hydrate deposit growth in section 1. (a) Experiment 2 with Tr = 80 °C; (b) Experiment 4 with Tr = 60 °C; (c) Experiment 5 with Tr = 40 °C; and (d) Experiment 6 with Tr = 30 °C. For each experiment, the images from left to right are taken at 0 h, 48 h, 176 h, and 480 h. Images from the top camera are analyzed by the image processing software, e.g., ImageJ, to obtain the hydrate deposit thickness growth in section 1 radial direction (Figure 4). Analysis methods and uncertainties are described in the Supporting Information. Measured hydrate deposit thickness for experiments 1, 2, and 3 show good agreement, demonstrating the reproducibility of the experiments. These results also suggest that the RTDs do not interfere with the hydrate deposition, at least when Tr = 80 °C. The measurement of the hydrate thickness for all experiments shows a relatively fast growth period at the beginning of the experiment (about the initial 120 hours), and then decreasing growth rates for several days. Interestingly, the growth rates (in terms of deposit thickness) seem to approach a limiting value. The results also show that


12

Page 13 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

when Tr is lowered (so does the center temperature), the deposit growth rate in section 1 also decreases, as seen from Figure 3.

Figure 4. Deposit thickness at the middle of section 1. Solid lines are drawn to facilitate visualization of the trend. The black, red, blue, green, orange, and wine symbols represent experiment 1 to 6, respectively. In experiments 2, 3, 5, and 6, a borescope is inserted from the side port after depressurization to closely study the morphology and the thickness distribution of the hydrate deposit. We note that there is no obvious change in section 1 during the depressurization, as the pipe is cooled to -5 °C. Other sections, however, cannot be observed, and changes to the hydrate deposit are still possible. For example, small pieces of the deposit can fall due to the gas release


13

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 31

from any hydrate dissociated. Because the density of ice and hydrate are slightly different, the solid deposit porosity or volume may change during the transition from hydrate to ice. The inserted borescope can capture images either pointing toward the top or bottom. Figure 5 shows representative images of experiment 5 (Tr = 40 °C). Figure 5(a) and Figure 5(b) are taken from sections 1 and 2 towards the bottom, while Figure 5(c) and Figure 5(d) are taken from sections 4 and 5 towards the top. Because there is a significant amount of hydrate deposit in section 3, it is impossible to use the borescope in that section. Figure 5(e) shows an illustration of the hydrate deposit throughout the pipe based on the images collected using the borescope. For experiment 5, the inspection into the pipe with the borescope shows a substantial amount of hydrate deposited forming a plug filling the whole cross-section. The distribution of the hydrate deposit is asymmetric. Figure 5(a) and Figure 5(b) show that the hydrate deposit narrows when it gets close to the plug and the narrowing is more pronounced closer to the plug. The hydrate deposit is observed to be angularly uniform at any given distance above the plug and the top of the plug is observed to be relatively flat. However, the hydrate piece in Figure 3(b) is a fallen piece during borescope inspection rather than a part of the plug, that is, a hydrate deposit on the center RTD of section 2, which can be seen until opening the system. As can be seen in Figure 2(a), there is no hydrate deposit on the RTD anymore. Figure 5(c) and Figure 5(d) show the plug from the bottom. For this experiment, the plug is located slightly above the RTDs in section 3. Unlike the area above the plug, the hydrate deposit below has a narrow opening profile, with respect to the axial distance to the plug, and it appears to be less angularly uniform. The plug in section 3 may only occur within a small range of height. In experiment 5, the axial thickness of the plug was not measured.


14

Page 15 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

Figure 5. Borescope inspection in experiment 5 (Tr = 40 °C). Images are taken (a) towards the bottom at the center of section 1, (b) towards the bottom at the center of section 2, (c) towards the top at the center of section 4, and (d) towards the top at the center of section 5. (e) Illustration of the hydrate deposit distribution (colored in white) in the deadleg. The thickness of the hydrate deposit at the middle of the sections can be estimated by using the RTDs as a reference. The method used is described in the Supporting Information. The thickness measurements via the borescope inspection are summarized in section d) of Table 2, which also contains data on the water recovery to estimate the porosity of the hydrate deposit. After the borescope inspection, the hydrate/ice deposit is dissociated section by section, from bottom to top. The data for the water recovery are plotted in Figure 6 and listed in section c) of Table 2. The details to calculate the water recovery is described in the Supporting Information. Because of the different time scales of each experiment, comparison of the results is focused on the distribution rather than the total recovery amount (even though time may affect


15

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 31

the distribution). The results show that the distribution of hydrates (water) is not uniform along the pipe. Typically, the most amount of water is collected in one section, with proportionally less water in the other sections. It is also observed that the section with most water shifts towards the bottom as Tr decreases, that is, most water is recovered from sections 1, 2, 3, and 4 for Tr = 80 °C, 60 °C, 40 °C, and 30 °C, respectively. Experiments 1, 2, and 3 are for the same experimental conditions, and all three show very similar distributions, with the most amount of water from section 1, a considerable amount of water from sections 2 and 3, and a negligible amount from sections 4 and 5. In experiment 4 (Tr = 60 °C), the water is more evenly distributed along the pipe, with the most amount of recovered water in section 2, but the differences among all sections are not as large as that in experiments 1, 2, and 3. In experiment 5 (Tr = 40 °C), the distribution is more or less uniform from section 1 to 4, with the most amount of water recovered in section 3, and significantly less in section 5. In experiment 6 (Tr = 30 °C), even though that particular experiment lasted 2006 hours, little water was recovered from sections 1 and 2, and most recovered from section 4, followed by sections 3 and 5.


16

Page 17 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

Figure 6. Water recovery from hydrate/ice deposit dissociation and center temperature at 24 h. Columns of each experiment from left to right represent sections 1 to 5 and are colored black, red, blue, orange, and green, respectively. Columns with ‘*’ refer to negligible amount of water recovered. The symbols and line represent the center temperature at 24 h of each experiment. The symbols from left to right represent the temperature of section 1 to section 5 and the header. No temperature line in experiment 3 because no RTD was used. DISCUSSION Based on the results, Figure 7 summarizes the understanding for hydrate deposition in deadlegs through conceptual pictures highlighting the key steps in the mechanism. At the beginning (Figure 7(a)), a temperature field is established due to the hot header and the cold wall. The temperature gradient induces a density gradient along the pipe causing the gas and water concentration to vary accordingly. The gas density gradient is considerably large. For example, the density of the methane/ethane (75/25 mol.%) mixture at the 100 bar is 75.78 kg/m3


17

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 31

at 80 °C, 102.0 kg/m3 at 30 °C, and 130.8 kg/m3 at 4 °C (calculated from Multiflash 6.0 SRK model). The saturation of water vapor only barely affects the density. The large density difference along the pipe induces natural convection (Figure 7(b)), carrying water vapor from the bottom to the top of the pipe. It is likely that the natural convection is so strong that the gas phase is well mixed. The characterization of the natural convection is important but it is beyond the scope of this paper. Some temperature measurement results for the natural convection using nitrogen are shown in the Supporting Information to better illustrate the mixing in the system. Because of the convection, the radial temperature gradient generally is relatively small in the bulk gas phase and the temperature may only decrease sharply at the boundary. For example, as shown in Figure 2, the temperature difference between the center and the off-wall RTD probes in section 5 of experiments 1 at 48 h is about 2.5 °C, while the difference between the center and the pipe wall is about 36 °C at the same time.


18

Page 19 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

Figure 7. Conceptual pictures of hydrate deposition in a deadleg. Water and hydrates are colored blue and white, respectively. The gas temperature is colored from pink (hot) to purple (cold). The hot header and the cold pipe wall established (a) a temperature gradient, induces (b) natural convection, and results in (c) water condensation on the wall. (d) Hydrate forms from the condensed water and covers the entire pipe wall. Given suitable conditions, (e) hydrate deposit grows and accumulates in the deadleg. As the vapor is saturated with water, the water condenses from the hot vapor to the cold surface (Figure 7(c)), and given that the surface is cold, the condensed water can convert to hydrate (Figure 7(d)). Hydrates are observed to firstly form at a random spot but spread to the entire pipe wall within usually 10 minutes. Given suitable conditions, hydrate deposit can grow and accumulate in the deadleg (Figure 7(e)). It should be emphasized that hydrates are not likely to form in the gas phase, but only from the condensed water on the surface. As such, the relevant hydrate equilibrium is for liquid-hydrate-vapor. During the hydrate deposition, water is transported from the header into deadleg, causing a water consumption in the header. The water (or hydrate deposit) distribution along a deadleg largely depends on the heat and mass transfer limitation for water transport into the pipe, water condensation on the wall, and hydrate deposition. A number of factors can affect the hydrate deposition, however, the temperature in the pipe is the most important in these experiments. The difference between Tr and Tw determines the temperature gradient in the system and therefore determines the magnitude of natural convection and the water mass transfer rate from the header to the deadleg. Thermodynamically, the temperature profile in the deadleg also determines the regions where hydrates can form. Kinetically, the subcooling, which is the difference between the actual


19

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 31

temperature and the HET, determines the rate of hydrate deposition, provided the availability of water and gas. The hydrate stable region, where the temperature equals or is less than the HET, can be estimated based on the measured temperature profile in the deadleg (Figure 8). Because hydrate will only form within the region, an estimation of the region at a given Tr and Tw is important. The estimation can suggest a design criteria (e.g., L/ID) of deadlegs to minimize the conditions for hydrate deposition. For example, if the hydrate stable region can be well controlled from the design, a complete plug may be possibly avoided.

Figure 8. Illustration of the temperature field in the axial and radial cross-sections at Tr = 80 °C and Tw = 4 °C. The blue color represents the hydrate stable region where the temperature is below the HET.


20

Page 21 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

Figure 8 shows an illustration of the hydrate stable region in a deadleg without hydrate. The region is marked with blue color. The boundary of the region is approximately parabola in the axial cross-section view based on the temperature measurement. The actual boundary depends on the Tr and Tw and may not be smooth. In the presence of the hydrate, the boundary will shift and narrow, but remain in a similar shape. The shape of the hydrate stable region generates the hydrate deposit of the similar shape, as visually observed. At the very bottom, the region exists only close to the wall. The bottom is a relative concept which changes with the boundary conditions. Even though the water condensation can be significant, the deposit growth is thermodynamically limited. As the height increases, the region expands and gradually fills the entire radial cross-section. In the hydrate stable region, hydrate deposition, gradually changes from heat transfer limited to mass transfer limited. The subcooling increases in the hydrate stable region with height but water becomes limited. Most water from the bottom condenses and converts into hydrate in the bottom sections. Such decrease in condensation is also partially related to the temperature because the water vapor pressure in the gas exponentially decreases with decreasing temperature. The heat and mass transfers reach the balance at a certain height and hydrate deposition can reach the fastest growth at that location. Figure 2 shows that after the initial increase in temperature, the temperature only slightly changes if it is higher than the HET. Interestingly, the temperature in the initial stage can be correlated to the water recovery distributions. The center temperature at 24 h are plotted with the water recovery results in Figure 6, which shows that the most water recovery in each experiment is from the section where the center temperature is slight below the HET. In fact, in experiments


21

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 31

4, 5, and 6, these are sections where the temperature of the entire section (not just the center temperature) is lower than the HET. In experiments 1 and 2, no section is entirely into the hydrate stable region. Most of the water only condenses and forms hydrate in section 1. Because the choice of the sections is arbitrary, a more precise statement is that the deposit distribution should increase continuously with an increasing height until it peaks at a certain location and decreases afterwards. Figure 6 suggests that the restriction, rather than the initial hydrate stable region boundary, locates at where the temperature initially (24 h) is a few degrees lower than the HET. The temperature field in the pipe affects the hydrate deposit growth and distribution. The hydrate deposit, in turn, changes the temperature field. The growing hydrate deposit narrows the pipe and changes the heat transfer. Specifically, the hydrate deposit restricts the gas convection from the bottom to the top of the pipe, causing the gas temperature in the upper sections to decrease over time. In the meantime, the hydrate deposit may also cause the gas temperature to increase in the lower sections as heat accumulates and is not transported well upwards the pipe. For example, for experiment 6, the hydrate restriction is determined to be within section 4, and all sections above section 4 have initially a temperature increase, and then a gradual decrease, especially at the off-wall positions (Figure 2). On the other hand, the temperature in section 5 increases and stabilizes. The temperature readings in section 4 noticeably fluctuate around the HET, suggesting that the hydrate deposit is continuously forming, melting, annealing, or possibly falling, as what is observed in the frost deposition.21 The hydrate distribution in the deadleg is obtained from visual observations (Figure 3 and Figure 5) and the water recovery during dissociation (Figure 6). The results from the three sources give consistent information for the hydrate distribution in the deadleg. This suggests that


22

Page 23 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

the water recovery results can be used to represent the amount of hydrate deposit. The amount of recovered water, however, is not necessarily equal to the amount of water that formed hydrate, because depending on the conditions, not all the condensed water may be converted into hydrate. Some water may condense on the hydrate deposit and fill the hydrate pores, resulting in a “wet” hydrate. With the current setup of the system, the actual conversion ratio of water to hydrate cannot be determined. The actual thickness at certain positions, in addition, can be measured from visual observations and the hydrate deposit morphology can be inspected. Assuming a water conversion ratio, a hydration number (5.94 from CSMGem) and a porosity value, the thickness of the deposit can be calculated, as shown in section e) of Table 2. Details of the calculation is described in the Supporting Information. In section e) of Table 2, the water conversion and the hydrate porosity are assumed to be 100% and 0%, respectively. The porosity assumption is conservative compared to the actual scenario, as to have a solid with zero porosity would mean that a single crystal was formed. The results show that the hydrate deposit thickness at the restriction can be at least around 1 cm. At this thickness, the effective pipe diameter would reduce from 7.4 cm to around 5.4 cm and the cross-section area would shrink by 47%. With a larger porosity, the pipe diameter can be reduced much more. Based on the borescope inspection and the water recovery, the porosity is estimated as shown in section f) in Table 2. During the experiment, if possible, the syringe pump is used to record the gas recovery. However, the gas recovery is not used in Table 2 because it is not available in all experiments. As seen, the porosity in most cases, at the experimental time scale, is around 50%. Based on other published studies,13, 15 the porosity can change over time. Initially, the porosity value could be high, and then gradually decrease as the hydrate deposit grows/anneals by filling the


23

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 31

pores. This may be particularly true after a hydrate restriction is formed in the deadleg, as the hydrate deposit essentially ceases to grow in volume, and anneals over time. Another distinction that needs to be made is the hydrate porosity for the pores filled with water or gas. A gas-filled pore would result in a permeable hydrate deposit, whereas a water-filled pore would have very low permeability. Table 2. List of water recovery and calculation results. Unavailable data are marked with ‘-’. 1 2 3 4 5 6 1004 474 380 860 887 2006 80 80 80 60 40 30 1 29.93 38.82 24.8 (c) 15.26 10.35 5.34 2 16.69 9.18 10.87 20.07 11.21 1.91 Water recovery (mol) 3 11.14 0 8.16 12.85 16.82 17.17 4 5.16 0 6.44 9.64 8.62 25.56 5 0.38 0 0 10.44 4.31 12.97 1 2.08 1.76 1.34 0.78 0.39 (d) 2 1.31 1.12 1.04 0.32 Measured thickness (cm) 3 0.61 0.2 2.86 0.4 4 0.22 1.11 2.2 5 0.06 1 1.23 1.74 0.98 0.56 0.37 0.19 (e) 2 0.81 0.42 0.50 1.00 0.52 0.08 Deposit thickness 3 0.52 0 0.37 0.60 0.82 0.83 a with zero porosity (cm) 4 0.23 0 0.29 0.44 0.39 1.35 5 0.02 0 0 0.48 0.19 0.61 (f) 1 0.446c 0.111 0.369 0.523 0.495 0.501 2 0.632 0.508 0.532c 0.459 0.731 Average porosity calculated c 3 0.586 0.410 from visual observationb 4 0.605 0.286 5 a Assumptions: 1. Calculation is based on the water recovery. 2. All recovered water is from hydrate dissociation. 3. Hydrate has zero porosity. 4. Hydrate deposit thickness of each section is uniform. b Assumptions: 1. Calculation is based on water recovery. 2. All recovered water is from hydrate dissociation. 3. Measured thickness from the borescope at the center of the section represents the average thickness of the entire section. 4. Porosity corresponds to an average value of the section. c Porosity is calculated by assuming the hydrate thickness is the radius (instead of using the thickness measurement from borescope,), i.e., assuming the deposit has plugged the entire section. (a) (b)

Experiment # Duration (hours) Tr (°C)

Sec.


24

Page 25 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

For practical design purposes, it is important to relate the environment variables to the hydrate deposit restriction location where most hydrates are expected to deposit. The results show that the hydrate restriction forms at the very top of the deadleg when Tr = 80 °C and moves downwards as Tr is lowered (Figure 9(a)). Assuming the restrictions are formed at the middle of each section, a possible correlation between Tr and the hydrate restriction location for a 3-inch deadleg can be identified as illustrated in Figure 9(b). When L/ID is smaller than the given values for a given Tr, the deadleg will be most likely not have a complete hydrate restriction. In fact, when the pipe becomes shorter, the temperature at the same height may be even higher due to less cooling and the relative hydrate amount should be even less. Therefore, the position of the hydrate restriction suggested is a conservative L/ID value. While these results indicate a simple promising correlation between Tr and the hydrate restriction location, it is too premature to state whether this correlation is valid in other systems. The correlation is possibly very system specific with respect to the tested conditions.

Figure 9. (a) Illustration of the hydrate deposit distribution at different Tr for a 3-inch deadleg. The deposit is colored white. (b) Estimation of the plug location at different Tr.


25

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 31

The present study on the effects of the header temperature has provided very valuable insight into the hydrate deposition process. On the other hand, the study also leaves open questions to be addressed in future studies to better understand the process. It will be necessary to continue to study the effects of wall boundary conditions (temperature/heat flux), pipe size, pipe geometry, etc. Knowledge of the effects of each parameter would help to quantify the heat, mass, and momentum transfer in the deadleg which may lead to correlations for the hydrate deposition. Experimentally the water and gas consumption need to be measured more accurately to better quantify the deposition kinetics. The results suggest that the water condensation can be crucial to the hydrate deposition and, as such, better characterization of the water condensation will be helpful. Moreover, CFD simulations should be attempted to model the deadleg as it can provide insightful understanding of the various transfer processes. One difficulty is to include the hydrate deposition kinetics into CFD simulations as no models yet exist for deposition, but approximations and simple correlations may be used as a first attempt to understand the coupling of heat/mass transfer and hydrate deposition.

CONCLUSION The deposit and the temperature mutually affects each other. Temperature is observed to significantly affect distribution and formation rate. When the Tr decreases, the temperature inside the pipe generally decreases. The hydrate stable region thus shifts towards the bottom of the pipe. Interestingly, the temperature profile in the early stage can be correlated with the deposit distribution. It seems that the most deposit will appear in an area just within the hydrate stable region and less deposit will form when the temperature is too low or too high. The growth rate at


26

Page 27 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

the top of the pipe is also monitored through the camera. It also confirms that the rate decreases with the temperature, when the temperature is below the HET. The deposit, in turn, gradually changes the temperature profiles. Depending on the amount of deposit, the hydrate stable region rapidly changes at the beginning and gradually changes afterwards. Initially the change can be contributed to the insulating effects of the deposit. In the late stage, the changes are caused by the pipe effective diameter reduction. The deposit forms a restriction in the deadleg. The restriction essentially divides the deadleg into two compartments. The temperature of the top sections usually decreases afterwards, while that of the bottom stabilizes and that of the middle fluctuates. The results are important to understand the deposition mechanism, but more experiments are needed to recommend practical hydrate management strategies.


27

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 28 of 31

ASSOCIATED CONTENT Supporting Information include additional details for: image analysis, depressurization procedure, deposit thickness estimation from hydrate dissociation, repeatability, and nitrogen temperature measurement.

AUTHOR INFORMATION Corresponding Authors *E-mail: [email protected] (KMA), [email protected] (AKS) ORCID Jeong-Hoon Sa: 0000-0002-8579-1643 Amadeu K. Sum: 0000-0003-1903-4537 Notes The authors declare no competing financial interest.

ACKNOWLEDGMENT The authors wish to express their appreciation to Statoil for funding this project and granting permission to publish this paper.


28

Page 29 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

REFERENCES 1.

Sloan, E. D., Jr; Carolyn, K., Clathrate Hydrates of Natural Gases. 3rd ed.; CRC Press:

2007. 2.

Sloan, E. D., Jr, Natural Gas Hydrates in Flow Assurance. Gulf Professional Publishing:

2010. 3.

Lunde, K.; Kinnari, K.; Labes-Carrier, C. Deadleg. US Patent 9,297,237, 2016.

4.

Kinnari, K. J.; Askvik, K. M.; Li, X.; Austvik, T.; Zhang, X.; Sa, J.-H.; Lee, B. R.; Sum,

A. K., Hydrate Management of Deadlegs in Oil and Gas Production Systems - Background and Development of Experimental Systems. Energy & Fuels Article ASAP. 5.

Habib, M. A.; Badr, H. M.; Said, S. A. M.; Hussaini, I.; Al-Bagawi, J. J., On the

Development of Deadleg Criterion. Journal of Fluids Engineering 2005, 127, (1), 124-135. 6.

Habib, M. A.; Said, S. A. M.; Badr, H. M.; Hussaini, I.; Al-Bagawi, J. J., Effect of

geometry on flow field and oil/water separation in vertical deadlegs. International Journal of Numerical Methods for Heat & Fluid Flow 2005, 15, (4), 348-362. 7.

Bloom, G. R. Turbulent hydraulic penetration from turbulent pipe flow into stagnant

columns. Washington State Univ., Pullman, Pullman, Washington, USA, 1978. 8.

Asteriadou, K.; Hasting, A. P. M.; Bird, M. R.; Melrose, J., Computational Fluid

Dynamics for the Prediction of Temperature Profiles and Hygienic Design in the Food Industry. Food and Bioproducts Processing 2006, 84, (2), 157-163. 9.

Anderson, H. Computational Study of Heat Transfer in Subsea Deadlegs for Evaluation

of Possible Hydrate Formation. MS Thesis, Telemark university College, Porsgrunn, Norway, 2007.


29

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

10.

Page 30 of 31

Nazeri, M.; Tohidi, B.; Chapoy, A., An Evaluation of Risk of Hydrate Formation at the

Top of a Pipeline. Oil and Gas Facilities 2014, 3, (2), 67-72. 11.

Creek, J. L., Efficient Hydrate Plug Prevention. Energy Fuels 2012, 26, (7), 4112-4116.

12.

Dorstewitz, F.; Mewes, D., The influence of heat transfer on the formation of hydrate

layers in pipes. International Journal of Heat and Mass Transfer 1994, 37, (14), 2131-2137. 13.

Grasso, G. A.; Sloan, E. D.; Koh, C. A.; Sum, A. K.; Creek, J. L.; Kusinski, G., Hydrate

Deposition Mechanisms on Pipe Walls. In Offshore Technology Conference, Houston, Texas, USA, 2014. 14.

Nicholas, J. W.; Koh, C. A.; Sloan, E. D.; Nuebling, L.; He, H.; Horn, B., Measuring

hydrate/ice deposition in a flow loop from dissolved water in live liquid condensate. AIChE J. 2009, 55, 1882-1888. 15.

Rao, I.; Koh, C. A.; Sloan, E. D.; Sum, A. K., Gas Hydrate Deposition on a Cold Surface

in Water-Saturated Gas Systems. Ind. Eng. Chem. Res. 2013, 52, (18), 6262-6269. 16.

Zhao, J.; Wang, B.; Sum, A. K., Dynamics of hydrate formation and deposition under

pseudo multiphase flow. AIChE J. 2017, 63, 4136-4146. 17.

Lingelem, M. N.; Majeed, A. I.; Stange, E., Industrial Experience in Evaluation of

Hydrate Formation, Inhibition, and Dissociation in Pipeline Design and Operation. Annals of the New York Academy of Sciences 1994, 715, 75-93. 18.

Zhao, J.; Wang, B.; Yang, L.; Cheng, C.; Song, Y., A novel apparatus for in situ

measurement of thermal conductivity of hydrate-bearing sediments. Review of Scientific Instruments 2015, 86, (8), 085110. 19.

Ross, R. G.; Andersson, P.; Bäckström, G., Unusual PT dependence of thermal

conductivity for a clathrate hydrate. Nature 1981, 290, 322-323.


30

Page 31 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

20.

Krivchikov, A. I.; Gorodilov, B. Y.; Korolyuk, O. A.; Manzhelii, V. G.; Conrad, H.;

Press, W., Thermal Conductivity of Methane-Hydrate. Journal of Low Temperature Physics 2005, 139, (5), 693-702. 21.

Hayashi, Y.; Aoki, A.; Adachi, S.; Hori, K., Study of Frost Properties Correlating with

Frost Formation Types. J. Heat Transfer 1977, 99, 239-245.


31

Hydrate Management in Deadlegs: Effect of ... - ACS Publications

Recommend Documents