Hydrate Management of Deadlegs in Oil and Gas Production Systems

Sep 20, 2017 - Oil and gas production systems comprise complex pipe networks in subsea, topside, and onshore facilities and require fit-for-purpose hy...
0 downloads 6 Views 5MB Size
Subscriber access provided by UNIV OF ESSEX

Article

Hydrate Management of Deadlegs in Oil and Gas Production Systems – Background and Development of Experimental Systems Keijo J. Kinnari, Kjell Magne Askvik, Xiaoyun Li, Torstein Austvik, Xianwei Zhang, Jeong-Hoon Sa, Bo Ram Lee, and Amadeu K. Sum Energy Fuels, Just Accepted Manuscript • DOI: 10.1021/acs.energyfuels.7b02051 • Publication Date (Web): 20 Sep 2017 Downloaded from http://pubs.acs.org on October 2, 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 of Deadlegs in Oil and Gas Production Systems – Background and Development of Experimental Systems Keijo J. Kinnari1, Kjell M. Askvik*2, Xiaoyun Li3, Torstein Austvik3, Xianwei Zhang4, Jeong-Hoon Sa4, Bo Ram Lee4 and Amadeu K. Sum*4 1

Statoil ASA, N-4035 Stavanger – NORWAY 2

3

4

Statoil ASA, N-5020 Bergen – NORWAY

Statoil ASA, N-7005 Trondheim – NORWAY

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

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

In the oil and gas industry deadlegs – pipe sections without through-flow – often pose hydrate control challenges to gas and oil production systems. The hydrate challenges, if not properly managed, can cause severe consequences in terms of safety and cost for oil/gas production. This paper provides an overview of deadlegs in oil and gas production systems with some examples of typical challenges faced in the oil industry. Two different types of vertical deadleg experimental systems have been developed to acquire a better understanding of hydrate risks in gas-dominated deadlegs. These systems offer valuable quantitative information of hydrate deposit, such as thickness, porosity, morphology, growth rate, distribution, temperature profile, and amount of water and/or gas consumed as a factor of time in the deadleg system.

Keywords: gas hydrate, flow assurance, hydrate deposition, deadlegs, hydrate management

ACS Paragon Plus Environment

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 Oil and gas production systems comprise complex pipe networks in subsea, topside and onshore facilities and require fit-for-purpose hydrate management strategies.1 Normal consequences of the system complexity are the presence of pipe sections that have no through-flow, so-called deadlegs.2, 3 These can exist in a large number of different constellations and can comprise both process pipes as well as instrument lines. These pipes, especially for gas dominated systems, are commonly introducing flow assurance challenges mostly related to hydrate restrictions and in some degree also to other solids deposition like wax. When these pipe sections are part of a safety system, such restrictions can have detrimental consequences. For example, a hydrate plug in a pressure safety valve (PSV) system would hinder emergency depressurization of a process tank that could result in over-pressurization and finally a rupture in some parts of the system; as such, the consequences could be catastrophic. Additionally, hydrate plugs in less safety critical parts of the production system could also result in large production loss and costly remediation solutions. Therefore, it is important to design production systems in a way that eliminates or minimizes safety risks and reduces the probability for flow assurance related problems in deadlegs. Despite of these challenges, very few studies or notes are publicly available on this topic,2, 4-8 so the understanding of hydrate formation in deadlegs is quite limited. This paper (i) introduces deadlegs in the oil and gas industry and typical challenges as specifically encountered in Statoil’s operations providing insight into design practices and operational solutions, and (ii) describes the development of gas dominated vertical deadleg experimental systems used for a quantitative research of the problem.

ACS Paragon Plus Environment

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

Definition of Deadleg in the Oil Industry A deadleg is typically defined as a pipe section that does not have any through-flow or the flow is very small and unintentional as a result of leakage. The pipe diameter of deadlegs can vary from a few millimeters in impulse/transmitter lines to tens of inches in production and export lines. Deadlegs can be categorized as short deadlegs with a length to inner diameter ratio (L/ID) below 100 and long deadlegs with L/ID above 100 (see Figure 1). The L/ID value of 100 is arbitrarily chosen only to differentiate between deadlegs in connection to subsea production templates and process facilities in subsea or topside/onshore and deadlegs formed by production flowlines not in use for a limited time. Short deadlegs can often be designed to an optimal length to mitigate hydrate related problems. Insulation can be used to increase the allowed length. L/ID for long deadlegs vary most often from about few hundred to larger than 10,000. The lower value range covers typically short jumpers and spools connecting wells or templates to main flowline or tie-in manifold. The upper range covers typically long flowlines connected to a production network. Long deadlegs require normally isolation valves to create acceptable L/ID. In cases where this is not possible, chemical injection can be used to mitigate the risk. Use of heating or heat tracing can eliminate hydrate risk.

ACS Paragon Plus Environment

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

Figure 1. Simplified categorization of the deadlegs with length/inner diameter ratio. Deadlegs shorter than 100 in L/ID ratio are classified as short deadlegs and larger than 100 as long deadlegs. Hydrate Challenges in Deadlegs Long deadlegs (L/ID > 100) connected to a live system pose different types of challenges compared to short deadlegs. As shown schematically in Figure 2, the long deadlegs will normally have several low and high spots, so (i) water would over time gradually fill the low points, (ii) pressure fluctuation in the main production lines provides adequate mixing to enhance hydrate growth, and moreover, (iii) water condensing on the cold pipe wall in the gas filled sections close to the warm fluid source will result in hydrate deposition. These mechanisms can over time result in clogging of the whole pipe cross section. If left untreated, the hydrate plug in the deadleg will grow more compact as more water will be converted into hydrates inside the restriction. The plug can also continue to grow in size as more water will be transported to the plug location.

ACS Paragon Plus Environment

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

Figure 2. Illustration of possible hydrate challenges in deadlegs. To avoid hydrate formation in long deadlegs, the normal design practice is to have isolation valves to isolate such deadlegs from the main system. However, due to cost issues, such valves are easily a prey to cost cut and thus often omitted or simplified in design, creating hydrate challenges. Actuated valves are therefore frequently avoided, and “manual” valves are chosen instead requiring a ROV. This increases the risk of the valves being left open for a prolonged time resulting in substantial hydrate growth. The time for a hydrate plug to be formed in such systems depends on the system design and the fluid system. Plugging would typically take several weeks or months for large pipe diameters (e.g., 10 in. and above), while smaller pipes (e.g., 6 in.) can be blocked in few weeks or even shorter time at favorable hydrate forming conditions. Hydrate plugging risks also exist in short deadlegs. Figure 3 shows various geometries of short deadlegs, covering production branches in a production manifold with a periodically closed branch, short branches for future tie-ins, pipe sections isolated with an isolation valve, flare lines, PSV lines, chemical injection lines that are used only occasionally, instrument lines and many more. The illustration characterizes the importance of two critical parameters: the length and the inner diameter of a deadleg. The arrangement of a deadleg also plays an important role. Obviously, the direction of flow in the main pipe and the angle of a deadleg can be crucial to the temperature profile and the mixing

ACS Paragon Plus Environment

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

in a deadleg. The characteristics of the fluid itself also affect the hydrate deposition growth. The flowing fluid system can be either a gas system or a multiphase system containing gas saturated with water. In the past, it was believed that a self-draining deadleg would be free of hydrate troubles, however, this belief is the opposite of what is actually observed and thus an improved understanding and design of deadlegs is required, especially in regards to hydrate formation.

Figure 3. Illustration of typical geometries of short deadlegs. If a deadleg is not properly designed or the operational guidelines are poorly implemented, hydrate deposition may occur in the deadleg. It is therefore important to design these parts of production systems in a way that hydrate growth can be avoided or minimized. For example, hydrate avoidance in new PSV systems is normally achieved by heat tracing both the pipe and the PSV. Existing facilities, however, do not always have such design. If the time window, the amount, and the location of hydrate presence cannot be predicted, no sufficient solution can be provided and the pipe may finally be blocked with hydrate. The presence of hydrates in deadlegs in safety critical pipes can pose a serious safety threat; for example, hydrates could impede a required depressurization attempt in a critical operation scenario.

ACS Paragon Plus Environment

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

Hydrate formation is more common in subsea systems than in topsides. The hydrate growth can be more complex in a system with a complex geometry or with a small gas leakage. Increased complexity of subsea production system has increased the awareness of this challenge and a large amount of work has been put forth to optimize designs to eliminate or reduce hydrate risks to an acceptable level. Some examples of hydrate challenges in deadlegs from Statoil’s operations are provided in the following section to illustrate the large variety of problems encountered in operations.

Examples of Hydrate Challenges in Deadlegs The following three examples illustrate hydrate growth in subsea chemical lines, subsea tie-in element and impulse line. High-Pressure and High-Temperature (HPHT) Field The first example is from a HPHT field with a template design that includes High-Integrity Pressure Protection System (HIPPS). As shown in Figure 4, this design requires special arrangements for injections of mono-ethylene glycol (MEG), a chemical commonly used for hydrate control. There are two small chemical lines between the two HIPPS valves. Due to the design complexity, the valves are located several meters from the connection point on the header. The larger pipe ID is 2 in. MEG injection line and the smaller pipe is a 5/8-in. ID annulus service line. Both lines are used intermittently, thus forming deadlegs during normal production. It was found early in the operation of the system that these pipes easily clog with hydrates. The 2-in. MEG line got plugged within 1–2 weeks and the smaller annulus service line within 3–5 days. The fluid temperature in the header was well above 80 ºC. A revised hydrate control strategy was needed to prevent clogging of these pipes with hydrates, and this was achieved by regularly injecting small amounts of MEG. Based on these operational experiences, a hypothesis was

ACS Paragon Plus Environment

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

defined to determine the plugging rate of deadlegs with different diameters. The hypothesis was that the plugging time ratio of two pipes is proportional to their inner diameter ratio squared, that is, t1/t2 ∝ (D1/D2)2. This hypothesis is purely empirical and the pipe length is not taken into consideration. However, it has provided a good enough basis for risk assessment of other pipes with larger diameters, and subsequently, has been successfully applied in a semi-quantitative analysis and guidelines for operations of other fields.

Figure 4. Part of a subsea template with sections of two deadlegs: (a) 2-in. MEG injection line and (b) 5/8-in. annulus service line. Subsea Tie-in This example is from a subsea wye riser base connecting two pipelines to a common flexible riser (shown in Figure 5). Pipeline A was shut down for several months and valve V was closed while production from line B was continued. The fluid system in B was a gas condensate system with low water content. The arrival temperature on the floater was well above the hydrate equilibrium temperature (HET). When the production from A was attempted, a hydrate restriction

ACS Paragon Plus Environment

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

was identified somewhere between the well (a few kilometers away) and the wye at the riser base. However, there was not enough information available to determine if the plug was in the wye structure, in the valve V or in the upstream part of the pipeline. As the location of the hydrate plug was not known it was not acceptable to remove the hydrate blockage by depressurizing the pipeline. A possible projectile during the dissociation of hydrates could have resulted in a serious scenario as the wye was directly connected to a flexible riser.

Figure 5. Measurements of average density of pipeline contents to locate hydrate restriction. (a) Illustration of the subsea wye riser base module and different positions. (b) Gamma-ray measurement results. The hypothesis for the hydrate restriction formation was that water was continuously condensing from the warm gas transported up into the deadleg. The condensing water on the cold surface could then turn into hydrate and gradually block the whole pipe. The worst-case assumption was to have a hydrate plug in a location around 6, 7 or 8 (see Figure 5a). If the remaining volume between the valve and the blockage were gas filled, there would be a large chance for a projectile during the depressurization process.

ACS Paragon Plus Environment

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

To know the location of the restriction (also possibly several hydrate plugs), a mobile Tracerco’s single source gamma-ray tool was used (at that time the Discovery tomography tool was not yet available)9.The tool provided an excellent quick method for identifying the plug location. The test result in Figure 5b suggested that there was not any hydrate blockage upstream the valve. This was confirmed by changing the pressure in this section of the pipeline and measuring the gas density by the gamma tool. It was quickly found that density was changing with the pressure and thus no plug existed in that part of the system. The plot in Figure 5b shows the density distribution at the different locations marked with 1 to 9. The low densities at location 9 and 8 correspond to the gas density at the operating pressure. The density then gradually increases from location 7 to 2. Location 1 provides the gas density at a higher pressure upstream of the valve. Densities at location 4, 3, and 2 are very close to 1 g/cm3 indicating that this part is water filled or filled with water and hydrates. Because the gamma method measures the average density, it is impossible to differentiate between liquid water and water bound in hydrates. If any hydrate existed, the pores would also be water filled due to the high-density value. It is difficult to conclude if the density differences among locations 2, 3 and 4 are real or an artifact as a result of the pipe property changes due to the presence of the bend. At last, it was concluded that the pipe is liquid filled (or filled with liquid and hydrates) up to location 4. Locations 5, 6 and 7 have intermediate densities suggesting that these parts of the pipe contain hydrates and location 5 contains the most. The intermediate density is unlikely to be caused by free liquid as these locations are above the top bottom of the pipe and the free water would have flowed back towards the riser base. The conclusion from the gamma measurement was that it was safe to depressurize the pipe, as there was not a gas pocket behind the restriction to cause a

ACS Paragon Plus Environment

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

projectile. The depressurization was carried out successfully and the hydrates were easily removed without any complications. Impulse Lines Small impulse lines pose a special tendency towards plugging with hydrates both in process facilities and subsea systems. One example of such an incident was observed in a Venturi meter element as shown in Figure 6. This was built for one of Statoil’s subsea field developments. The original design did not include any insulation of the impulse lines needed for pressure measurements. During an onshore test at Statoil’s test laboratory, hydrates formed in the impulse lines and clogged them even in cases with fluid temperatures in the main line around 80 ºC. A quick solution was to mold aluminum epoxy around the impulse lines to improve the heat conductivity, which promotes the heat transfer from the main line. CFD (computational fluid dynamics) calculations also confirmed that the unit would stay at a sufficiently high temperature. The unit has worked as planned for several years since its installation.

Figure 6. Subsea Venturi meter element: (a) the original design of the unit. (b) the unit after molding aluminum epoxy.

ACS Paragon Plus Environment

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

Design Guidelines for Systems with Deadlegs As the presence of hydrates in deadlegs can have a negative impact on production or be a safety risk, it is important to avoid hydrate formation or to minimize it to an acceptable level. In the past, Statoil has carried out CFD analysis and based on these results developed some general guidelines. Updated guidelines are in development based on the recent studies at Colorado School of Mines (CSM), CFD simulations and on the increased field experience and know-how within Statoil. The guidelines are moving towards being defined with respect to the flowline temperature and other process parameters relating to heat and mass transport. The establishment of such guidelines is the ultimate goal of the research program. While previous guidelines were formulated with respect to full plugging of the deadleg, it is understood that hydrates form wherever the temperature drops below the HET. Hydrates may therefore form at deadleg walls at lower L/ID ratios than those given in the earlier criteria, although hydrate plugs are not possible at such positions. This implies that the criteria may have to be stricter for safety critical dead legs, or indeed any presence of hydrates should be unacceptable. From experience, it has been observed that horizontal deadlegs connected to production pipes are not prone to hydrate plugging (instrument lines may on the other hand have problems). For example, production headers are normally horizontal and there has not been any report of hydrate plug incident in the deadleg section of a header even in cases where the producing wells are closed to the pipeline inlet, as shown in Figure 7. Statoil has no special requirement to insulate production headers to reduce the hydrate plugging risk. In gas injection systems, the risk may be quite different; If the temperature is below the water dew point, hydrate growth can become critical. Wells further down the production header having a branch forming a deadleg may experience plugging with hydrates if the deadleg is too long. It is therefore recommended to have an isolation

ACS Paragon Plus Environment

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

valve as close as possible to the production header to reduce hydrate deposit amount. In cases where the isolation valve cannot be operated or is missing, regular flushing with chemicals will be necessary.

Figure 7. Illustration of a typical 4-slot subsea template with headers and branch pipes. Well 1 is closed with an open branch valve forming a complex deadleg. Wells with closed branch valve on a given header form a short deadleg with a non-horizontal geometry. Horizontal deadlegs are formed upstream the branch connections of the wells flowing into a given header (marked purple).

Laboratory Studies of Deadleg Systems Statoil saw the need to better understand hydrate growth in deadlegs and especially short gasfilled deadlegs as defined in this paper. Hydrate restrictions in deadlegs can result in time consuming and costly mitigation and remediation operations. In addition, hydrate plugs in safety critical system parts could naturally have catastrophic consequences. Therefore, Statoil recently

ACS Paragon Plus Environment

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

launched a research project at CSM with the objective to understand hydrate growth in deadlegs and to develop design tools and guidelines to prevent/minimize hydrate deposition. New test systems were developed as part of the deadleg project. One of the main challenges was to have a test system that can provide as realistic results as possible over wide range of key parameters like water content at different temperature and pressure conditions. The geometry and dimensions of a deadleg design is naturally also of key importance. In order to meet these design requirements two different types of test systems were designed and built. The first system comprises of two relatively large test rigs with interchangeable pipes. The pipe sizes chosen for these rigs are 2-, 3- and 4-in. These rigs provide large flexibility and are run in parallel but with two different diameters. Conditions can be chosen independently for each rig. As the duration of the experiments for deadleg studies can inherently be long (several weeks), two rigs provide an excellent way to achieve larger amount of experimental data and to assess the repeatability of given test sets. The pipe sizes are in addition small enough to be cost effective in design and operation. The variation in diameters is large enough to allow development of a sufficient set of tools and knowledge basis combined with existing knowledge to have adequate extrapolation to larger pipes. This system will be described in some more details below to illustrate the capabilities of this versatile system. The description will cover only some of the basic principles in the design and methodology in running the experiments. Some examples will be provided for the interpretation of results and characterization of the depositions. Vertical Deadleg System Figure 8a shows the overall experimental system to study hydrate deposition in a deadleg of 2-, 3- and 4-in. pipes. The major components of the system include: main pipe piece, syringe pump, water reservoir, chillers, cameras, temperature and pressure sensors.

ACS Paragon Plus Environment

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

Figure 8. Illustration of the experimental setup for the 2-, 3-, and 4-in. deadleg systems. Temperature sensors, syringe pump, and camera systems are connected to a data acquisition system. Data acquisition system not shown. The main pipe piece of the system represents a vertical deadleg, which can be changed among 2-, 3-, and 4-in. (inner diameter) pipes. To have full control of the conditions in the pipe, including the hydrate deposit distribution, the pipe piece is designed to have five cooling sections (each 20 cm long), each with its own cooling-jacket for temperature control, and each connected to one chiller. The sections are numbered 1 to 5 from the top to the bottom. Each section is installed with resistance temperature detectors (RTDs) for temperature measurements of the surface, off-wall, mid-center, and center locations. In addition, each section has windows, which allows for direct view of the pipe interior. All the three pipes have the same length of approximately 1.24 meter. A syringe pump (Teledyne 1000D from Teledyne Isco Inc.) is used as a gas reservoir to maintain the system at constant pressure. The pump has a 1000 ml capacity capable of controlling the

ACS Paragon Plus Environment

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

pressure from 0 to 120 bar. The pump has a built-in pressure transducer to monitor the pressure change. A RTD is added to the top of the pump to monitor the temperature of the pump, which is usually different from the reservoir or the pipe. The syringe pump volume change is used to calculate gas consumed for deposition or recovered from dissociation. A 6-liter water reservoir is used to provide water for hydrate deposition. Typically, the reservoir is half filled with water. The water level is monitored via a side water level gauge, from which the water consumed for hydrate deposition or recovered from dissociation can be determined. The reservoir uses a ceramic heater to control the temperature up to 100 ºC. A magnetic impeller is used to maintain a uniform temperature in the entire water reservoir. To be able to change pipes with different sizes, a 200 mm adapter is used in between the pipe and the reservoir. To relieve the cooling pressure of the chillers, a thermal insulator made of polyetheretherketone (PEEK) is in between the adapter and the reservoir. There are two RTDs in the reservoir to measure the temperature at the top (gas phase) and at the bottom (water). There is also a pressure transducer connected to the top on the reservoir. Because of the relatively high temperature, the pressure transducer reading is not affected by hydrate formation. Two visual monitoring systems are installed in the deadleg system: one through the top window (made of polycarbonate) to observe hydrate deposit formation/dissociation, and another close to the reservoir to monitor the water level change in the reservoir. The system consists of multiple temperature and pressure sensors. All of them are connected to a data acquisition system (Yokogawa® GX10) for monitoring and recording. The syringe pump and the cameras are connected to a computer with a LabView® based program to monitor and record related data.

ACS Paragon Plus Environment

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

The 2-, 3-, and 4-in. system is designed with a capability to study the effect of forced gas flow by using a gas blower, which can be connected through the top of the water reservoir. The nozzle of the blower can be placed near the bottom of the deadleg pipe, allowing control of the angle and rate of the gas flow to assess the impact of forced convection. Experimental System Capabilities Tests in the deadleg system have two stages: hydrate formation and hydrate dissociation. During the first stage, hydrate deposition occurs under controlled conditions for given period of time, which can range from a few days to several weeks (the longest test performed lasted nearly 90 days). In the second stage, the system undergoes controlled dissociation to collect information on the hydrate deposition. Table 1 lists the controlled and measured variables from a typical test in the deadleg system. Table 1. List of controlled and measured variables. Typical values

Controlled variables

Measured variables

Pressure

(bar)

30 – 100

Surface Temperature

(oC)

-5 – 15

Reservoir Temperature

(oC)

30 – 80 CH4, CH4/C2H6

Hydrate former Duration

(day)

1 – 90

Temperature profiles

(oC)

Figure 10

Gas consumption

(mol)

Water consumption

(mol)

Gas recovery

(mol)

Figure 12

Water recovery

(mol)

Figure 12

ACS Paragon Plus Environment

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

Plug position measurement

(cm)

Figure 13

Top camera monitoring

Figure 11

Borescope inspection

Figure 14

X-ray CT scan

Figure 15

Hydrate Deposition For the hydrate formation, water is injected to the reservoir, which provides the water source to form hydrates in the deadleg. This water evaporates and saturates the gas inside the pipe. The deadleg system is then pressurized with the hydrate former, which is a gas or gas mixture, e.g., methane or methane/ethane. The experiment starts by setting the pipe surface temperature and the reservoir temperature to the desired values. For example, the surface temperature can be set to 4 o

C to mimic a typical subsea temperature, or to 15 oC to mimic a typical top-side platform

temperature. The water reservoir can be adjusted between 30 oC (needs to be higher than the room temperature for proper control) and 80 oC to mimic different header temperatures. All the experiments in the 2-, 3-, and 4-in. pipes are performed at constant pressure controlled by the syringe pump. All the experiments with the 1-in. pipe are performed at constant volume. In the deposition period, the system is monitored for hydrate deposition from temperature profiles, gas consumption, water consumption, and visual inspection from side and top windows. Hydrate Dissociation At the end of the hydrate deposition period, one of the two methods can be used to dissociate the hydrates and collect information on the quantity and distribution of hydrates in the deadleg. Each method has a different purpose. The first method is to dissociate the hydrates under pressure, which is fast and is considered more accurate than the second method. In this method, the reservoir temperature is first lowered to 30 oC to reduce the water transfer and hydrate growth. The syringe

ACS Paragon Plus Environment

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

pump is stopped and after the system is cooled down, the pipe surface temperature is increased section by section from section 5 to section 1, causing the hydrates to dissociate. The surface temperature is usually increased to 2 oC higher than the HET. The system pressure and the reservoir water level are recorded during the dissociation to later calculate the amount of hydrate in each section. The second method for dissociation starts by freezing and depressurizing the system, which preserves the hydrate deposit distribution and possibly the morphology. The freezing is achieved by reducing the surface temperature to -5 oC. After depressurizing the system to atmospheric pressure, the top window is removed to closely inspect the deadleg for the hydrate deposition distribution and morphology. The visual inspection is done by both a regular commercial-grade camera and a borescope. Theoretically, hydrates are converted to ice as they dissociate at -5 oC, but the process is relatively slow, so the solid observed in the deadleg during the inspection is usually a mixture of hydrate and ice. Sampling of the deposit can also be done at this stage. Tools and containers, which are precooled using liquid nitrogen, are used to sample the deposit. Due to the hardness of the hydrate deposit, a stainless steel knife is used to cut the deposit. The cut sample is quickly put into the container and immersed into liquid nitrogen. The samples can be used to perform various analysis, such as X-ray computerized tomography (CT). After the visual inspection and sampling being completed, the pipe surface temperature is increased, section by section (from the bottom to the top), above the freezing point to dissociate all the solids. The water recovery is monitored and used to calculate the amount of hydrate. Figure 9 shows how the data collected from the experiments are used to extract quantitative and qualitative information on the hydrate deposition for a given test. The amount of gas and water

ACS Paragon Plus Environment

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

consumed and recovered in the experiment are measured, which, together with visual observations through the top window, can provide data to estimate the thickness, distribution and porosity of the hydrate deposit.

Figure 9. Work flow from deadleg experiments to estimate the amount and porosity of hydrates formed. Hydrate Deposition Characterization The following sections describe the different data collected to assess the amount and distribution of hydrates formed in the deadleg, including the plug location for cases when one is formed. Temperature Measurements Figure 10 shows the center temperature profiles of a typical test in the deadleg. For this particular test, the induction time for hydrate formation is approximately 5 hours, based on the visual observation through the top window. The time zero in Figure 10 is from the start of hydrate

ACS Paragon Plus Environment

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

formation. The center temperature of all sections significantly increases after initial hydrate formation. The temperature increase in all sections results from the insulating effect of the hydrate deposit,10 as well as the heat released from hydrate formation. The insulating effect causes greater resistance to heat removal through the pipe wall. The rapid temperature increase levels out after approximately 2 days. In Sections 3 to 5, the center temperature becomes stable after approximately 10 days. Interestingly, the center readings in Section 2 slowly and gradually increases during the entire experiment after 4 days. For this particular test, the center temperatures for Sections 2 to 5 are out of the hydrate stable region, meaning that hydrates deposition cannot fully fill the cross sectional area in these section. However, the center temperature profile in Section 1 varies and enters the hydrate stable region after about 6 days, indicating that hydrates are gradually filling up the entire pipe cross section that may eventually result in a hydrate plug.

Figure 10. Example for the center temperature profile in a 3-in. experiment. The shaded area in the plot correspond to conditions outside the hydrate stable region. The temperature of Sections 1 to 5 are colored black, red, blue, orange, and green, respectively.

ACS Paragon Plus Environment

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

4.3.2. Top Camera Observation Visual observation of the pipe interior during the formation stage is possible through the top window. Figure 11 shows the results from the same experiment as that of Figure 10. In the figure, the bright spot is the deposit. The black area around is caused by the deposit on the top window, which is a few centimeters higher than the plug. The black area at the center of the image indicates the void space and its disappearance indicates a complete plug formation. Based on the visual observation, it seems that it disappears after around 38 days. Gradually, condensed water droplets on the top window can be observed, such in the 8-day image in Figure 11a. The condensed water gradually turns into hydrate and covers the top window, making the corresponding area black on the image. Figure 11a shows that the hydrate deposit formed seems to be angularly uniform in any given cross section. The thickness of hydrate deposit in Section 1 can be measured using the thickness and length of the RTD as a reference. Note that the images are from the top window, and the camera can be focused to observe various sections, but unfortunately, except for Section 1, the images are blurry due to the large difference of gas density at different heights in the pipe (caused by the temperature gradients). From the visual observations, a possible hydrate plug around Sections 1 and 2 formed after approximately 38 days, which agrees with the extrapolation of the growth curve shown in Figure 11c. The image analysis method is shown in Figure 11b.

ACS Paragon Plus Environment

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

Figure 11. Visual observation and analysis from the top window in a 3-in. experiment. (a) Hydrate deposit observation. (b) Methodology for measuring the hydrate deposit thickness at Section 1. (c) Thickness measurement results. Water and Gas Recovery If the hydrate deposit is dissociated under pressure, both the gas and the water released from the deposit can be measured. Figure 12 shows the results from the same experiment as that of Figure 10. Figure 12a shows the syringe pump gas recovery during the dissociation. These data are used to evaluate the amount of hydrate dissociated in each section. The result for this particular test suggests that the most amount of hydrate formed in Section 1 (largest amount of water and gas recovered), and only a small amount of hydrates formed in the other sections. The gas recovery does not show a monotonic trend. One possible reason is that, the syringe pump volume changes

ACS Paragon Plus Environment

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

are not solely caused by the hydrate dissociation. The system temperature change may have some contribution to the volume changes, too. Without some proper adjustments, the results can be different from the actual gas release amount from hydrate dissociation. During the hydrate dissociation, the water level in the reservoir is also monitored. The water recovery is shown in Figure 12b. It can be found that the ratio of gas and water recovery from each section are not at a constant value. Apart from the uncertainties in the measurement, one possible reason is that the deposit may be porous with a certain water and gas volume fraction. The recovery data only show the total amount of the free water (or gas) and the amount from hydrate.

Figure 12. Recovery of a 3-in. experiment. (a) Syringe pump gas recovery during hydrate dissociation. (b) Water recovery during hydrate dissociation. Plug Position Measurement If for a given test, a hydrate plug is formed, the plug position can be measured through the top window via a laser distance measurement device (Bosch GLM 80, ±1.59 mm), even when the system is pressurized. The refraction within the window and the gas phase can cause some small errors, but the measurements are overall accurate. Another method for determining the plug

ACS Paragon Plus Environment

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

location is to remove top window after depressurization, and use a ruler to directly measure the plug distance from the top. The two methods have been effective and shown to be in good agreement. Figure 13 shows the measurement of the plug position in a 2-in. experiment. The black squares show the actual center RTD distances in Sections 1 to 3 from the top window, and the red triangles show the measured distances. The noticed differences can be caused by several reasons which alter the light path, including the top window and the gas density gradient. Moreover, it has been noticed that the laser may go into the deposit before it can be reflected, which can cause an error of a few centimeters. However, as shown in Figure 13, the linearity of the measurement is still good. The extrapolation can be used to satisfactorily measure the plug position. In the experiment of Figure 13, the plug position is measured to be 1.06 m from the top window, locating at the bottom of Section 4.

Figure 13. Plug distance measurement by a laser distance meter in a 2-in. experiment. The distance is between the measured subject to the top window. The actual distances of the center

ACS Paragon Plus Environment

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

RTD of Sections 1 to 3 are colored black, while the measured distances are colored red. The measured position of the plug is measured in green. Borescope Inspection For the tests that the system is frozen and depressurized, visual inspection can be performed. While a regular commercial-grade digital camera can be used to take images through the window, a borescope can be inserted into the system to inspect the hydrate deposit after removing the windows. Two kinds of cameras provide complementary images. A regular camera usually generates better quality images, but the positions and angles are limited. A borescope, on the other hand, is more flexible. Moreover, a borescope combined with a custom-made bore gauge can also be used to measure the hydrate distribution and thickness. A borescope can be inserted into the system through either the top of the side window spots. Through the sides, a borescope can point towards either the top or the bottom to capture images of both directions. If a borescope is pointing the bottom, the image is considered taken from the top side. Figure 14 shows a series of photos taken by a borescope from the top and the bottom of the pipe. In this test, a plug was observed.

Figure 14. Representative images from a hydrate plug formed in the deadleg in a 3-in. experiment. Images are obtained via borescope inspection after depressurization of the system.

ACS Paragon Plus Environment

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

Sampling and X-ray CT Scan Hydrate samples can be collected from the deadleg at the end of a test after freezing and depressurization. These samples, preserved in liquid nitrogen, can then be further analyzed to determine the solid content via X-ray computed tomography, as shown in Figure 15. To do the Xray CT scan of Figure 15, the samples are taken out of the liquid nitrogen and put into a test cell protected by dry ice. The device is an X-Radia 400 instrument (Carl Zeiss Microscopy). The Xray source is 150 keV/10W. The resolution is controlled to be approximately 10 µm and each measurement takes approximately 60 min. In the figure, the light and dark colored regions represent the highest and lowest density, respectively, and the grey region is in between. The X-ray provides a non-intrusive way to characterize the deposit and helps understand the deposit in many aspects. For example, Figure 15 shows that the solid sample is relatively porous with large pores inside. If setting a density criteria to differentiate the void space and the solid, the porosity can be estimated.

Figure 15. Representative images from X-ray CT scan for one hydrate sample collected from the deadleg in a 3-in. experiment. Images (a) to (c) are for different cross-sections (in terms of depth) of the same sample. The boundaries between the sample and the background are illustrated by

ACS Paragon Plus Environment

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

the black line. The lines do not necessarily indicate the actual deposit surface. The straight line is the side in contact with the pipe wall.

Conclusion Deadlegs are pipe sections with no throughflow, which therefore often encounter hydrate management challenges. Deadlegs include pipes of a wide range of lengths, diameters, and geometries. They, however, can be categorized based on the L/ID ratio. Deadlegs with L/ID < 100 are defined as short deadlegs. Particularly, gas-filled short deadlegs usually suffer from hydrate deposition. Three specific cases encountered by Statoil have been reviewed to illustrate the complexity, the challenges and the possible strategies for solution. Two experimental deadleg systems for different pipe sizes have been developed to obtain more knowledge at such conditions. These systems are able to offer valuable quantitative information, such as thickness of hydrate deposit, temperature profile, and amount of water and/or gas consumed as a factor of time in the deadleg system. The developed systems to study hydrate deposition in deadlegs are versatile and have unique capabilities for laboratory scale measurements that can be transferable to guidelines actually used in the hydrate management of deadlegs. AUTHOR INFORMATION Corresponding Author * Corresponding authors: [email protected] (KMA), [email protected] (AKS) Author Contributions The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. Funding Sources

ACS Paragon Plus Environment

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

Page 30 of 31

Funding for this project was provided by Statoil. ACKNOWLEDGMENT The authors wish to express their appreciation to Statoil and Åsgard Licence (Petoro AS, Statoil Petroleum AS, Eni Norge AS, Total E&P Norge AS and ExxonMobil Exploration and Production Norway AS) for funding this project and granting permission to publish this paper. ABBREVIATIONS CT, computed tomography; HET, hydrate equilibrium temperature; HIPPS, high integrity pressure protection system; HPHT, high pressure – high temperature; ID, inner diameter; L, length of a deadleg; MEG, mono ethylene glycol; PSV, pressure safety valve; RTD, resistance temperature detector; ROV, remotely operated vehicle. REFERENCES 1.

Kinnari, K.; Hundseid, J.; Li, X.; Askvik, K. M., Hydrate Management in Practice. J. Chem. Eng. Data 2015, 60, (2), 437-446.

2.

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

3.

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.

4.

Anderson, H. Computational Study of Heat Transfer in Subsea Deadlegs for Evaluation of Possible Hydrate Formation. MS Thesis, Telemark university College, Porsgrunn, Norway, 2007.

ACS Paragon Plus Environment

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

5.

Cai, J. J.; Abedinzadegan Abdi, M.; Hawboldt, K.; Sundgaard, C.; Haghighi, M., Hydrate Prevention in Subsea Oil Production Dead-Legs. In Offshore Technology Conference, Houston, Texas, USA, 2013.

6.

Hagemann, O. Cool Down of Subsea Dead-leg with a Cold Spot - Experimental and Numerical Heat Transfer Analysis. MS Thesis, University of Oslo, Oslo, Norway, 2014.

7.

Kanstad, S. K.; Nilsen, E. Heat Transport Dead Leg. 9,328,586, 2016.

8.

Sundt, E.-M. W. Numerical and Analytical Study of Steady State and Transient Heat Transfer in Liquid Filled Dead legs. MS Thesis, University of Stavanger, Stavanger, Norway, 2012.

9.

Tracerco

Pipeline

Blockage.

http://www.tracerco.com/subsea/flow-assurance/pipeline-

blockage (1 September), 10. 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.

ACS Paragon Plus Environment

31