Guidelines For Hydraulic Desig Of Multiple Pipe Slug Catchers z4642

PETRONAS TECHNICAL STANDARDS DESIGN AND ENGINEERING PRACTICE

MANUAL (SM)

GUIDELINES FOR HYDRAULIC DESIGN OF MULTIPLE PIPE SLUG CATCHERS

PTS 20.056 DECEMBER 1984

PREFACE

PETRONAS Technical Standards (PTS) publications reflect the views, at the time of publication, of PETRONAS OPUs/Divisions. They are based on the experience acquired during the involvement with the design, construction, operation and maintenance of processing units and facilities. Where appropriate they are based on, or reference is made to, national and international standards and codes of practice. The objective is to set the recommended standard for good technical practice to be applied by PETRONAS' OPUs in oil and gas production facilities, refineries, gas processing plants, chemical plants, marketing facilities or any other such facility, and thereby to achieve maximum technical and economic benefit from standardisation. The information set forth in these publications is provided to s for their consideration and decision to implement. This is of particular importance where PTS may not cover every requirement or diversity of condition at each locality. The system of PTS is expected to be sufficiently flexible to allow individual operating units to adapt the information set forth in PTS to their own environment and requirements. When Contractors or Manufacturers/Suppliers use PTS they shall be solely responsible for the quality of work and the attainment of the required design and engineering standards. In particular, for those requirements not specifically covered, the Principal will expect them to follow those design and engineering practices which will achieve the same level of integrity as reflected in the PTS. If in doubt, the Contractor or Manufacturer/Supplier shall, without detracting from his own responsibility, consult the Principal or its technical advisor. The right to use PTS rests with three categories of s : 1) 2) 3)

PETRONAS and its s. Other parties who are authorised to use PTS subject to appropriate contractual arrangements. Contractors/subcontractors and Manufacturers/Suppliers under a contract with s referred to under 1) and 2) which requires that tenders for projects, materials supplied or - generally - work performed on behalf of the said s comply with the relevant standards.

Subject to any particular and conditions as may be set forth in specific agreements with s, PETRONAS disclaims any liability of whatsoever nature for any damage (including injury or death) suffered by any company or person whomsoever as a result of or in connection with the use, application or implementation of any PTS, combination of PTS or any part thereof. The benefit of this disclaimer shall inure in all respects to PETRONAS and/or any company d to PETRONAS that may issue PTS or require the use of PTS. Without prejudice to any specific in respect of confidentiality under relevant contractual arrangements, PTS shall not, without the prior written consent of PETRONAS, be disclosed by s to any company or person whomsoever and the PTS shall be used exclusively for the purpose they have been provided to the . They shall be returned after use, including any copies which shall only be made by s with the express prior written consent of PETRONAS. The copyright of PTS vests in PETRONAS. s shall arrange for PTS to be held in safe custody and PETRONAS may at any time require information satisfactory to PETRONAS in order to ascertain how s implement this requirement.

Code 500.20.800 (Production) - IN 73186 Code 500.30.720 (Natural Gas) - IN 73210 GUIDELINES FOR THE HYDRAULIC DESIGN OF MULTIPLEPIPE SLUG CATCHERS (January 1982 - December 1984) by A. Bos Approved by: R.V.A. Oliemans

SUMMARY Guidelines are presented for the hydraulic design of multiple-pipe slug catchers. These criteria are based on engineering practice with existing slug catchers and on laboratory model studies carried out at KSLA with a proper two-liquid system to simulate high-pressure gas and condensate. January 1985

CONTENTS 1.

INTRODUCTION

2.

SLUG CATCHERS GENERAL

3.

4.

2.1.

Necessity of a slug catcher

2.2.

Types of slug catcher

EXPERIENCE WITH MULTIPLE-PIPE SLUG CATCHERS 3.1.

Den Helder slug catcher

3.2.

St. Fergus slug catcher

3.3.

Bintulu slug catcher

3.4.

Bacton slug catcher

3.5.

Eemshaven slug catcher

3.6.

Summary of lessons learned from field experience and model studies

DESIGN GUIDELINES FOR MULTIPLE-PIPE SLUG CATCHERS 4.1.

Slug catcher size

4.2.

Slug catcher geometry 4.2.1.

Inlet section

4.2.2.

Bottle section 4.2.2.1. 4.2.2.2.

4.2.3.

Choice of primary and secondary bottles Length of the entrance section of the primary bottles required for settling of small droplets 4.2.2.4. The dual-slope approach 4.2.2.5. Equalizer system 4.2.2.6. Bottom header Gas outlet section 4.2.3.1. 4.2.3.2.

5.

Gas risers Gas outlet header(s) and gas outlet(s) 16

FINAL REMARK 1

TABLE

13

FIGURES

APPENDIX I

: The two-liquid test facility for the simulation of slug catchers (1 Table)

APPENDIX II

: The liquid storage capacity and height of a multiple-pipe slug catcher 1 Figure)

APPENDIX III

: Settling of mist in the separation section of the primary bottles (1 Figure)

APPENDIX IV : Simulation of bottle choking (Kelvin-Helmholtz instability) in the Amsterdam Tray Test Column multiple- pipe slug catcher (1 Figure) LIST OF SYMBOLS REFERENCES

GUIDELINES FOR THE HYDRAULIC DESIGN OF MULTIPLE-PIPE SLUG CATCHERS (January 1982- December 1984) 1.

INTRODUCTION At the end of gas pipelines operating in two-phase flow mode normally a gas-liquid separator, known as "slug catcher", is installed. The function of this facility is to separate the liquid from the gas before the gas enters the gas-treating facilities downstream of the pipe and to store the liquid. A widely applied slug catcher configuration is the multiple pipe concept. Throughout the years the Group has accumulated experience with multiple-pipe slug catchers based on: 1.

engineering practice with and measurements on existing slug catcher,

2.

model studies carried out in the Laboratory (KSLA).

This has resulted in the formulation of a number of guidelines for the design of a multiple-pipe slug catcher which are presented in this report. Here only the hydrodynamic aspects are considered. Based on this report a more general PETRONAS Technical Standard manual will be prepared by PETRONAS, Kuala Lumpur. The recommended design codes and material selection procedures will be subject of a separate PTS manual at present in preparation by PETRONAS.

2.

SLUG CATCHERS GENERAL

2.1.

Necessity of a slug catcher Trunk lines transporting natural gas often show the phenomenon of liquid formation, mainly due to retrograde condensation. Due to a slip in velocity generally occurring between the gas and liquid phases the amount of liquid in the pipeline steadily accumulates to an equilibrium level. The total amount of liquid present at any time in a long two-phase pipeline operating in steady state condition can be significant. When operating conditions are changed large volumes of liquid may emerge from the pipeline. This can be caused by a change in volume flow i.e. in velocity, or by running a sphere (or pig) through the line. The largest slug that can ever occur is the one caused by sphering. These occasionally very large volumes of (live) liquids must be handled and stored on- shore as they emerge from the line preferably without reducing velocity the flow in the pipeline. For this reason a slug catcher is always connected to a two-phase pipeline. A slug catcher consists essentially of two parts: -

a separator part, separating the liquid from the mixed stream arriving under normal (steady) flow conditions;

-

the storage part, receiving and storing the incoming liquid slug created by upset conditions (which also includes running a sphere through the pipeline).

When a more or less continuous slug of liquid arrives, the liquid displaces the gas present in the slug catcher thus guaranteeing a continuous supply of gas to the downstream facilities (compressor, treating plant, LNG plant). Gas lines operate generally at velocities of up to 12 m/s and large slugs will take only a matter of minutes to arrive. The holding capacity of the slug catcher therefore must be essentially as much as the volume of the largest slug. Although liquid carry-over must be limited, a slug catcher is not meant to replace a high-efficiency separator.

2.2.

Types of slug catcher Slug catchers can be broadly classified into the three following categories: 1.

vessel type,

2.

multiple-pipe type,

3.

parking loop type.

The geometry of the vessel type slug catcher could range from a simple knock-out vessel to a more sophisticated lay-out as has been designed by Stone & Webster for a non-Shell submarine pipeline (see Fig. 1). Since a vessel-type slug catcher is relatively short for a given volume this type is preferred in the case of limited plot sizes (e.g. offshore platforms). 3

When larger liquid volumes have to be accommodated, say more than 1000 m either the multiple-pipe or the parking loop type is preferred. In Fig. 2 a typical example of a multiple-pipe slug catcher is shown (existing slug catcher in Den Helder at the end of a 36" gas pipeline from the K-14 field in the North Sea). It consists of an entrance section where liquid/gas separation takes place and an array of parallel down-sloping bottles (of standard line pipe size) for liquid storage. An incoming liquid slug flows via the splitter into the inlet manifold and then via downcomers into and down the sloping bottles. As a consequence the gas present in the bottles is displaced and flows through the gas outlet risers mounted on the bottles to the gas header and the gas-treating plant. The liquid/gas exchange that occurs in the bottles guarantees an uninterrupted gas supply to the downstream facilities during liquid slug arrival, provided excessive liquid carry-over can be avoided. The advantage of a multiple-pipe slug catcher is that it is easy to operate since no flow control measures are required. A disadvantage is, however, the counter-current gas-liquid flow in the bottles which will promote liquid carry-over. The problem of counter current flow is avoided in the parking loop type slug catcher. This 2 concept has been introduced by Texas Eastern and is shown in Fig. 3. In this novel concept the separating and storage parts are virtually disconnected: it consists basically of a large separator with the liquid outlet connected to a long single. The incoming gas/liquid stream is separated in the vessel. When the liquid level rises rapidly, indicating that a slug rather than a gas-liquid mixture is arriving, the gas flow from the vessel is restricted forcing the liquid to flow into the pipeloop. In this loop a sphere is present to separate the liquid entering from the gas present. With the other end of the loop now opened to the downstream facilities the gas is driven out in a direct co-current manner. The stored (parked) liquid can be discharged as a single slug by using high pressure gas in case the location is at a booster compressor station as shown in Fig. 3 or the liquid can be discharged gradually to a downstream treating plant. This type slug catcher is specifically suited for offshore application where the separator can be placed on the platform and the loop on the seabottom. All valves and controls will be on the platform. But also in onshore applications space will be saved, particularly if the pipeloop is laid parallel to and in the same trench with the incoming pipeline. This concept, however, requires more sophisticated control, in contrast with the multiple-pipe slug catcher and may also require additional facilities to effect a gradual discharge of the liquid to a downstream treating plant. For a further development of this type of slug catcher far more studies are required. In the following sections of this report, of the types of slug catchers mentioned so far only the multiple-pipe type will be discussed in more detail.

3.

EXPERIENCE WITH MULTIPLE-PIPE SLUG CATCHERS A great number of multiple-pipe slug catchers are in operation throughout the world. They vary widely in geometry, mainly because of the many degrees of freedom possible in the design of this type of slug catcher. Up to now a unique design procedure is still lacking. In Table I a few representative slug catchers operated by Shell or in which Shell has an interest are listed, together with relevant information. For the respective layouts see the Figs. 2, 4, 5, 6 and 7. To the list is also added the slug catcher projected at the end of the F-3 pipeline (in Eemshaven). It has not been built as yet. However, relevant information is obtained from model experiments. The performance of these slug catchers and their specific features will be discussed in the following paragraphs.

3.1.

Den Helder slug catcher The Den Helder slug catcher in operation since 1975 has already been briefly discussed in Section 2.2 of this report. Its geometry and discussions are shown in Fig. 2. The slug catcher contains eight primary bottles (slope 1.5 %). Primary bottles have a dual function: 1.

gas/liquid separation,

2.

liquid storage.

The slug catcher was designed as a symmetrical slug catcher with 4 bottles located between the two inlets to the inlet header (2-4-2 configuration). However, this was changed into a 3-3-2 configuration in the construction phase. In its first 4 years of operation the slug catcher performed satisfactorily. However, in 1979 when the gas flow rate in the pipe was increased above 20 3 million Nm /d liquid carry-over started to occur. Since ultimately the gas flow rate in the pipeline 3 had to be increased up to 30 million Nm /d the slug catcher performance had to be improved. From neutron backscattering (NBS) measurements carried out on the slug catcher it was found that the liquid carry-over was caused by a maldistribution of the slug over the inlet header, leading to an uneven distribution of the liquid slug over the slug catcher bottles and hence to overloading of several bottles. Subsequent tests in a two-liquid perspex model of the Den Helder slug catcher 4 indicated that the maldistribution and hence the carry-over could be remedied by installing constrictions in the downcomers of the slug catcher. The set-up of the model studies with the two-liquid perspex model is described briefly in Appendix I. In view of the results of the model studies NAM installed in the slug catcher an insert which narrowed the portholes of the down-comers from 24 to 15 inch, This modification improved the performance of the slug catcher substantially. A second series of NBS measurements carried out on the Den Helder slug catcher revealed an even distribution of the slug over the inlet header 3 5 and bottles. Furthermore the onset-of-carry-over flow rate shifted from 20 to 34 million Nm /d . These values were in good agreement with the ones predicted by the model studies. Summarising, the experience gained with the Den Helder slug catcher highlights the following points: 1.

Avoid maldistribution of the slug over the bottles by a proper design of the inlet section (e.g. avoid asymmetry and select a proper diameter of the inlet header and downcomer portholes.

2.

The two-liquid model slug catcher is representative of practice and is therefore a useful aid for slug catcher design.

3.2.

St. Fergus slug catcher The St. Fergus slug catcher (Fig. 4) in operation since 1982 is located at the end of the 36" FLAGS pipeline which transports natural gas from the Brent field to Scotland. Fig. 4 shows that the slug catcher consists of 13 bottles (slope 0.4 %). The bottles 1 to 9 inclusive are primary bottles into which the liquid slug flows directly from the inlet header. The bottles 10 to 13 inclusive are secondary bottles. Contrary to the primary bottles the secondary bottles have a storage function only and are filled from the lower end via the primary bottles and bottom header. Other features of the slug catcher geometry are the double riser and equalizer system. The philosophy of using two risers rather than one riser per bottle is to lower the flow rate of the exiting gas in the risers and therefore the chance on liquid entrainment. The equalizer system is intended to equalize the pressure in the bottles. 3

The slug catcher is designed for gas flow rates up to 30 million Nm /d. However, up to now 3 (1984) the gas flow rate did not exceed 20 million Nm /d. In most cases hardly any liquid was removed from the pipeline by pigging since the pipeline pressure was too high to allow the development of two-phase flow. Up to now, for those occasions that liquid slugs were present, there have not been any liquid carry-over problems. It was feared that similar to the Den Helder slug catcher the slug distribution over the primary. bottles was not even, which could give rise to liquid carry-over at higher gas flow rates. To 6 this, laboratory studies were carried out with a two-liquid model of the slug catcher . The model studies led to the following main conclusions. 3

1.

The slug catcher in its present state can handle slugs up to a size of 1700 m at gas flow 3 rates up to 40 million Nm /d. This is well above the maximum flow rate of 30 million 3 Nm /d anticipated for the FLAGS pipeline.

2.

The incoming slug is unevenly distributed over the primary bottles in particular at low gas flow rates.

3.

If carry-over takes place it originates from the secondary bottles after the slug has entered the slug catcher. Improvement of the slug distribution over the primary bottles did not have a significant effect on the onset-of-carry-over-flow rate.

4.

After the slug had been received the condensate level was not at the same height: the closer the bottle to the (eccentrically located) gas outlet the higher the condensate level. In particular in the secondary bottles. This "manometer" effect is caused by the pressure drop in the equalizer and riser headers due to the gas stream through these headers. This effect is promoted by the eccentric location of the outlet and could give rise to a relatively early onset of liquid carry-over. The experience gained with the St. Fergus slug catcher shows: 1.

Use of secondary bottles is advantageous. Because a part of the incoming slug will flow into the secondary bottles, less gas is displaced countercurrently in the primary bottles and the carry-over originating from the primary bottles is suppressed.

2.

A central location of the gas outlet is recommended so as to minimize the manometer effect.

3.3.

Bintulu slug catcher The Bintulu slug catcher (Fig. 5) in operation since 1983 is located at the end of the 36" pipeline transporting natural gas from the Luconia fields offshore of Sarawak to the MLNG plant at Bintulu. A second parallel pipeline will be connected to the slug catcher in about 1990 when the 3 gas production of the Luconia fields is at its maximum of about 46 million Nm /d. 3

Up to now the gas flow rate has been lower than 10 million Nm /d and only very small slugs have been received in the slug catcher. No liquid carry-over has been observed. Fig. 5 shows that this slug catcher consists of 10 primary bottles at a slope of 0.75 %. In order to promote stratified flow, the downcomers are at an angle of 45 °. Furthermore, the slug catcher is equipped with a triple riser system and a double equalizer system. The slug catcher has been designed by SIPM and in its conceptual stage model tests at KSLA have been performed making use of the two-liquid test facility. In the model studies the conditions of the 1990 scenario were simulated: 3

Two pipelines are in operation. Maximum flow rate through each pipe is 23 million Nm /d. One 7 pipeline only at a time is sphered . The model studies showed: 1.

In the original conceptual design (in which no constrictions were used, neither in the downcomers nor in the gas outlet header exits) it was observed that the slug was unevenly distributed over the bottles and also a maldistribution of the gas flow in the gas outlet headers was observed, which was very sensitive to the location of the gas outlet. Due to this maldistribution liquid carry-over took place at a flow rate per pipeline 3 equivalent to 14.5 million Nm /d, which is far below the maximum flow rate of 23 million 3 Nm /d per pipeline expected in 1990.

2.

By the installation of properly sized constrictions in the downcomers and in the gas outlet header exits this maldistribution was suppressed and no longer liquid carry-over was 3 observed, not even at a flow rate per pipeline equivalent to 35 million Nm /d. Lessons to be learned from these model studies are:

3.4.

1.

Avoid maldistribution of the slug over the bottles by a proper design of the inlet section.

2.

Avoid complex riser systems since these can lead easily to gas flow maldistribution.

3.

The position of the gas outlet has a significant effect on the slug catcher performance.

Bacton slug catcher The Bacton slug catcher in operation since 1968 is located at the end of the two 30" gas pipelines which transport gas from the Leman Bank Fields in the North Sea to England. Up to 1973 only one pipeline was in operation. The original version of the slug catcher (designed by Eagleton Eng.) is shown in Fig. 6a. It is seen that the original slug catcher consisted of 6 bottles (2 primary and 4 secondary bottles at a slope of 0.75 %). The bottle part of the slug catcher is buried. A special feature of this slug catcher is the use of vertical tee gas/liquid separators in the downcomers. The philosophy of the use of these devices was to achieve an early gas/liquid separation in the downcomers. Gas should escape through the horizontal branch (see Fig. 6). In 3 the period 1968-1973 the gas flow rate was below 10 million Nm /d and no liquid carry-over of the slug catcher was observed. From 1973 onwards the slug catcher serves two pipelines. For that purpose it was split into two slug catchers with 3 bottles each (one secondary and two primary bottles) (see Fig. 6b). In practice, in particular slug catcher 2 in the split-up version had an unsatisfactory performance (50 3 vol % liquid carry-over at a flow rate of about 13 million Nm /d per pipeline!). It was suspected that the use of a vertical tee separator had a negative influence on the slug catcher performance. This has been confirmed by laboratory studies carried out with a perspex model and using as test media water and air 8,9.These studies showed the vertical tee is a poor gas/liquid separator. At high gas flow rates and a high liquid loading (>80 vol %) it even acts as a gas/liquid mixer by sucking gas from the horizontal branch into the vertical stream. In slug catcher 2 the effect of the vertical tee separators on the slug catcher performance has been eliminated by the installation of a blind plate in the separation header which isolates this header from the gas outlet. This modification gave some improvement of the slug catcher performance. To avoid liquid carry-over during the arrival of a pigged slug now the strategy has adopted to decrease the gas flow rate several hours before the expected slug arrival. In summary, both laboratory studies and field experience suggest that the use of a vertical tee separator as a gas/liquid separator should be avoided.

3.5.

The Eemshaven slug catcher In the near future (1986): a 24" two-phase pipeline will be built to transport the gas/liquid hydrocarbons produced in the F-3 block (North Sea) to Eemshaven. A slug catcher will be installed at the end of this pipeline. Fig. 7 shows SIPM's conceptual design of the slug catcher. It has a symmetrical lay-out consisting of four primary bottles and two sets of four secondary bottles. The downcomers make an angle of 45° with the horizontal and their port holes are equipped with constrictions to suppress an uneven distribution of the slug over the primary 3 3 bottles. The total liquid storage concept is 3600m of which 1000m is intended for permanent liquid storage serving as a buffer for the gas/liquid treating facilities located downstream of the slug catcher. The bottle part of this slug catcher will be buried. By introduction of a new design feature, the dual slope concept, the slug catcher could be made more efficient from a liquid storage point of view and therefore less expensive. The separation part of the primary bottles is at a slope of 2.5%, while their storage part and also the secondary bottles are at a slope of 0.4%. Compared to a single-slope slug catcher (bottle slope 1.5%) the bottles are shorter by some 30m but the storage capacity is still the same. This is shown in Fig. 8. An additional advantage of the dual-slope concept is, that the height of the slug catcher is reduced by 2.7m. A point of concern, however, is that this concept may have hydrodynamic consequences. For instance, a hydraulic instability (hydraulic jump) could occur at or downstream of the point where the slope changes from 2.5% to 0.4%. Such instability could promote liquid carry-over during filling of the slug catcher, in particular when the slug catcher is nearly full. To investigate this, laboratory studies have been carried out with a Perspex model, using the two- liquid test facilities. These studies showed that the change in bottle slope did not introduce a hydrodynamic jump, but merely a gradual increase in liquid hold-up. The liquid carry-over started at a flow rate equivalent to about 40 million Nm3 /d, which is well beyond the maximum of 3 10 million Nm /d which will probably be reached in 1993. Further tests showed that a reduction of the bottom header diameter from a diameter equivalent to 48 to 36 inches did not have a significant effect on the flow rate at which the onset of carryover is observed. The major outcome of these studies is that the dual-slope concept is feasible.

3.6.

Summary of lessons learned from field experience and model studies 1.

Avoid asymmetry in the slug catcher design. For instance the inlet manifold should be symmetrical and if one gas-outlet is used it should be at a central location.

2.

The geometry should be as simple as possible. For instance multiple riser systems should be avoided.

3.

Avoid maldistribution of the slug over the bottles by a proper design of the inlet manifold.

4.

Application of secondary bottles is useful in that they relieve the flow conditions in the primary bottles.

5.

A vertical tee separator in the downcomers should be avoided.

6.

The dual slope concept is feasible.

7.

The two-liquid model approach is representative of actual conditions and is therefore a useful aid for slug catcher design.

4.

DESIGN GUIDELINES FOR MULTIPLE-PIPE SLUG CATCHERS Based on the experience obtained with multiple-pipe slug catchers (encoming engineering practice, field tests and model studies) a number of guidelines for the design of multiple-pipe slug catchers are formulated.

4.1.

Slug catcher size The size of the slug catcher is directly related to the maximum of liquid volume, Volsc , it has to hold. The slug catcher should be able to intercept the maximum possible slug size emerging from the two-phase pipeline at any moment (Volint). If required it should also contain a buffer volume of condensate, Volbuffer, in order to guarantee the buffer liquid supply to liquid-treating facilities downstream of the slug catcher Volsc = Volint + Volbuffer

(1)

As far as the intercepting capacity of the slug catcher is concerned, it has been pointed out in the Introduction that the largest slug which could occur is a slug generated by sphering. This slug II size is estimated with the computer program TWOPHASE developed by KSLA. With this computer program it is possible to calculate, among other things, the two-phase flow regime, the liquid hold- up in and pressure drop over a two-phase pipeline under steady-state conditions. In principle the size of the sphered (or pigged) slug and therefore Volint is calculated from the following formula: Volint = Lpipe

(H L − λ )

(2)

where L denotes length, H L is the liquid hold-up averaged over the pipe length and λ the liquid volume fraction flowing in the pipe. Note that eq. (2) gives the maximum slug size since in general between two subsequent sphering runs not sufficient time is lapsing to allow the liquid hold-up in the pipeline to build up to its equilibrium level. Also the pig or sphere will not have a 100% efficiency. No general rule can be given for the determination of Volbuffer since it is largely determined by the characteristics of the liquid-treating facilities. The slug catcher is considered to be full (contains Volsc) when the liquid level in the primary bottles just reaches the bottle section immediately underneath the risers (assuming the same liquid level in all bottles). See also Fig.II.1 in Appendix II. Under these conditions there is no liquid hold-up underneath the risers of the primary bottles and the risk of liquid entrainment in the gas stream exiting the primary bottles is minimal. In Appendix II the relations are given to calculate Volsc as a function of bottle slope, length and diameter and number of primary and secondary bottles both for single and dual-slope slug catchers. 4.2.

Slug catcher geometry In general a multiple-pipe slug catcher could be divided into three sections. -

Inlet section Bottle section Gas outlet section.

The guidelines for the slug catcher design are given below by reviewing the slug catcher sectionwise.

4.2.1.

Inlet section In this section the distribution of the incoming liquid over the bottles takes place. Also here a start is made with the gas/liquid separation by promoting the occurrence of stratified two-phase flow. This section entails: -

end of the pipeline splitter(s) inlet header downcomers expanders.

The geometry of the inlet system should be symmetrical. If there are fewer than 4 downcomers no splitter is required. To avoid maldistribution of the slug over the bottles more than 8 downcomers should be avoided. Based on the experience with the Den Helder slug catcher the portholes of the downcomers should be at most 40% of the inner diameter of the inlet header to guarantee an even liquid distribution. A smaller porthole cross- section is achieved by either a constriction in the portholes or simply by a smaller downcomer diameter. If constrictions are used they should be located eccentrically, close to the wall of the downcomer (see Fig. 9). In this way a jetting effect which could lead to excessive mist/foam generation is suppressed because the liquid is guided along the wall. Also dirt accumulation upstream of the constriction is avoided. The option of using a constriction rather than decreasing the downcomer diameter is slightly preferred since because of the expansion downstream of the constriction the segregation of gas and liquid is promoted. This segregation could be even more promoted by the selection of a downcomer slope of 1 rather than using vertical downcomers. From the open literature it is known that an angle of 45° with the horizontal plane is optimal for the development of stratified flow 12 In existing slug catchers the diameter of the downcomers is smaller than that of the bottle (rather arbitrarily: Ddowncomer ≤ 2/3 Dbottle). The downcomer is connected to the bottle through an eccentric conical expander either with the flat side up (Den Helder) or the flat side down (St. Fergus, Bintulu). Due to the expansion a further gas/liquid separation will take place. There is a slight preference for the flat side down option because in that case no discontinuity in bottle slope takes place, which could upset the development of stratified liquid flow. 4.2.2.

Bottle section This section encomes: -

primary bottles

-

secondary bottles

-

equalizer system

-

bottom header

In the primary bottles the gas/liquid separation is completed and the liquid is stored. The secondary bottles have a storage function only. The equalizer system is meant to equalize the pressure of the bottles.

4.2.2.1. Choice of primary and secondary bottles The choice of the number of primary, npb, and secondary bottles, nsb , to be used in the multiplepipe slug catcher depends on several factors. 1.

Gas flow rate in the pipeline

2.

Required liquid storage volume of the slug catcher

3.

Plot size available (length available for bottles)

4.

Diameter of the bottles to be used

5.

Slope of the bottles

6.

Single or dual slope concept slug catcher?

It is attractive to keep the number of primary bottles as low as possible because of the following reasons: 1.

The slug catcher becomes less expensive because fewer fittings are used in the inlet and bottom header.

2.

There is less chance of a slug maldistribution over the primary bottles.

On the other hand the higher the gas flow rate in the pipe (which determines the velocity of a slug generated by pigging), the more primary bottles are required so as to avoid overloading of the bottles and therefore liquid carry-over. The maximum flow rate a primary bottle can accept without liquid carry-over is a function of the physical properties of the gas and liquid, the bottle diameter and slope and the amount of liquid which can flow from the lower end of the primary bottle to other bottles. For more information on this flow rate see Section 4.2.2.3. For design purposes the worst case should be considered, i.e. no liquid flow to other bottles from the lower end. Also a certain degree of maldistribution of the slug over the bottles should be taken into . It is recommended to assume that the most heavily loaded bottle receives 20% more than in the case of an even distribution (120/npb %). Furthermore (if npb > 1) npb should be even from a symmetry point of view. Once npb is determined, the total storage volume of the primary bottles is calculated with the equations presented in Appendix II. By subtracting this value from the required slug catcher storage volume, the volume of the secondary bottles is found. Subsequently, the number of secondary bottles is calculated. 4.2.2.2. Length of the entrance section of the primary bottles required for settling the small droplets Under steady-state operating conditions when no sphering is carried out a continuous gas stream with some liquid continually enters the slug catcher. It is divided via the inlet section over the bottles. Due to this a reduction of the gas velocity takes place and consequently small droplets entrained in the gas phase could settle. The first part of the primary bottles (between the conical expander and the riser system) should be long enough for a sufficient completion of this process. In Appendix III it has been made plausible that in a full-scale slug catcher with 36 inch bottles the time needed to settle droplets larger than 0.5 mm is in the order of 4 s. Since it is highly unlikely that the gas velocity in the first part of the bottle will exceed say 2 m/s, a length of say 8m in the case of a 36 inch bottle, or more general an entrance length of about 10 bottle diameters should be adequate for most settling purposes.

4.2.2.3. Effect of bottle slope on choking The bottles of the multiple-pipe slug catcher should slope downwardly. This will facilitate the liquid filling of the primary bottles by gravity and the flow of gas displaced by the incoming liquid to the gas outlet system. It is essential that during the slug catching operation the liquid in the primary bottles flows down as a stratified layer so as not to impede the ascending gas stream. Beyond a given liquid flow rate into the bottle which is among other things a function of the bottle slope, stratified flow can no longer be maintained and choking of the bottle will take place. One of the two following mechanisms could be responsible for this: 1.

2.

More liquid flows into the bottle than can be transported as a satisfied layer in the bottle itself. This will be the case when the bottle slope is very small and therefore the gravity drain is insufficient. In Appendix I of Ref. 3 a calculation model has been presented to predict the onset of choking according to this mechanism. 13 Kelvin-Helmholtz instability . When the relative velocity between the descending liquid stream and the ascending gas stream in the primary bottle exceeds a critical value, the liquid/gas interface in the bottle becomes unstable and excessive wave formation will take place. In bottles equipped with a riser this will occur predominantly near to the riser (see also Ref. 3). The Kelvin-Helmholtz instability will take place when the bottle slope is relatively steep.

Based on laboratory studies carried out at KSLA (simulation of a slug catcher bottle using the Tray Test Column facilities) a model has been developed to predict this type of instability (Ref. 14 and Appendix IV of this report). The two models have been combined to a general choking criterion. Both for Den Helder and St. Fergus slug catcher conditions (pressure 70 and 110 bar, resp.) the onset-of-choking flow rate is calculated as a function of the bottle slope. It is seen from Fig. 10 that in the case of the Den Helder slug catcher the bottle slope (1.5 %) is about the optimal 3 3 slope. Furthermore, the predicted onset flow (0.67 m /h) is in good agreement with the 0.70m /h measured in practice 5. As far as the St. Fergus slug catcher is concerned, its slope (0.4%) is well below the optimal value as predicted by the choking criterion (2.5 %). Based on the choking criterion and realizing that the optimum is rather flat it is recommended to take 1% as a minimum value for the bottle slope in multiple-pipe slug catchers. Fig. 10 suggests that from a choking point of view slopes up to 3% would be acceptable. However, large bottle slopes could lead to an unacceptably high elevation of the slug catcher. 4.2.2.4. The dual-slope approach The experience with the model of the Eemshaven slug catcher has shown that the dual-slope concept is feasible. The advantage of this approach is that a more efficient use is made of the liquid storage capacity of the bottles. A drawback is that the liquid hold-up underneath the risers of the secondary bottles is relatively high. In the Eemshaven case the slope of the storage part of the slug catcher is 0.4%. From a choking point of view this value is too low (see Fig. 10) but still acceptable for the Eemshaven slug catcher because this slug catcher will not operate under severe flow conditions (see Section 3.5). In general it is recommended to take as minimum slope for the secondary bottles and the storage part of the primary bottle a value of 1%. Fig.11 shows the effect of the slope of the separation section on the liquid storage capacity of the primary and secondary bottles when the slope of the secondary bottles and the storage part of the primary bottles is either 0.4 or 1%. It is seen that particularly in the case of a 0.4% slope a substantial increase of the storage capacity is obtained, a slope of the separation section larger than say 2.5% gives only a minor further improvement. Therefore a value of maximal 2.5% is recommended for the slope of the separation section. 4.2.2.5. Equalizer system Both in the St. Fergus and in the Bintulu slug catcher an equalizer system is used. It was found in the model tests that the effect of an equalizer system on the slug catcher performance is very sensitive to the slug catcher geometry and is not always beneficial. In the model studies carried out for the St. Fergus slug catcher, it was found that the equalizer system had a negative effect on the slug catcher performance. Gas flowing from the primary to the secondary bottles through the equalizer system caused liquid carry-over of the secondary bottles 6 . It is therefore recommended not to use an equalizer system.

4.2.2.6. Bottom header It is common practice to design the bottom header with the same diameter as the bottles. The model studies carried out for the Eemshaven slug catcher showed, however, that a smaller diameter (down to 75% of the bottle diameter) did not affect the slug catcher performance. A further reduction is not recommended because of the risk of blockage of the bottom header by sludge or dirt always present in the lower part of the slug catcher. In relation to possible dirt accumulation one should keep the header accessible for cleaning. To prevent gas carry-under during liquid drainage of the slug catcher it is recommended to have the bottom header below the lower end of the bottle. A possible geometry is given in Fig. 12, where the bottle slopes downwardly at an angle of 45° into the bottom header. 4.2.3.

Gas outlet section This section entails: -

risers

-

outlet header(s)

-

gas outlet(s).

4.2.3.1. Gas risers The gas risers must act as vertical separators. The lower the gas velocity the lower the load factor λ defined as

λ=

ρG v ρL − ρ G SG

and the smaller the liquid droplets which remain entrained in the riser gas stream. It is proposed to take for slug catcher application λ ≤ 0.20. It can be derived with the equation of Appendix III that droplets with a diameter larger than 2mm will settle from the stream through the riser (assumed liquid fraction = 0.1). This is acceptable if one takes into that the slug catcher is not meant to be a high-efficiency separator for small liquid droplets. This criterion will determine the maximum value for vSG. In the worst case we have the following scenario: 1.

The riser(s) is (are) located at the primary bottle which receives (120/npb)% share of the incoming slug (in case of an even slug distribution over the bottles the share would have been (120/npb)%).

2.

The slug is generated by pigging or sphering.

3.

There is no liquid flow from the heavily loaded bottle to the other bottles. This means a maximum gas flow through the riser(s). If the number of risers per bottle is denoted by m, the gas velocity in the riser should obey to the following relationship:

12 . Q npb pipeline ρ − ρG vSG = ≤ 0.2 L π ρG m D 2 riser 4 To keep the gas outlet as simple as possible and to avoid maldistribution of the gas flow in the riser system, preferably only one riser should be used per bottle. This could be achieved by taking npb and/or Driser as large as possible with as practical upper limit Driser =Db . The riser should have a minimum height to allow liquid entrained in the riser gas stream to settle. It is recommended to take the riser height at least equal to 5Driser. 4.2.3.2. Gas outlet header(s) and gas outlet(s) The diameter of the gas outlet header(s) and gas outlet(s) should not be taken too small. In the case of small diameters the pressure drop over the gas outlet system could become significant. The consequence of this is that after the slug has been received the liquid level in the bottles closest to the gas outlet will rise relative to the level in the other bottles as has been observed with the model of the St. Fergus slug catcher (manometer effect, see Section 3.2 of this report). It is advised to take Doutlet = Dgoh ≥ Driser The gas outlet should not be located eccentrically. In Fig. 13 a few recommended gas outlet configurations are given. The lay-out should be as nearly symmetrical as possible.

5.

FINAL REMARK This report contains the current state of knowledge on multiple-pipe slug catchers at KSLA. It is intended to be a document which has to be updated regularly. Readers and s of this report are encouraged to make comments suggestions and to supply additional experience which could lead to the upgrading of the design criteria outlined here.

TABLE I : REVIEW OF REPRESENTATIVE MULTIPLE-PIPE SLUG CATCHERS

*

Original 6-bottle slug catcher divided into to serve two pipelines.

**

Dual-slope concept (end of 1983).

***

Between brackets the capacity of the full-scale slug cacther following from field measurements.

+

Related to one slug catcher of the twin-concept

FIG. 1 : VESSEL SLUG CATCHER WITH SEPARATE SURGE DRUMS

FIG. 2 : THE GEOMETRY OF THE DEN HELDER SLUG CATCHER

FIG. 3 : “PARKING LOOP” TYPE SLUG CATCHER FOR LOCATION AT A BOOSTER COMPRESSOR STATION (ONLY VALVE REQUIRED FOR OPERATION OF THE SLUG CATCHER ARE INDICATED)

FIG. 4 : THE St. FERGUS SLUG CATCHER

FIG. 5 : THE BINTULU (SARAWAK) SLUG CATCHER (LENGTHS IN METRES, DIAMETERS IN INCHES)

FIG. 6 : THE BACTON SLUG CATCHER

FIG. 7 : CONCEPTUAL DESIGN OF THE EEMSHAVEN SLUG CATCHER

FIG. 8 : EFFECT OF INTRODUCTION OF DUAL-SLOPE CONCEPT FOR THE EEMSHAVEN SLUGCATCHER ON LIQUID STORAGE CAPACITY AND HEIGHT OF THE RISER FOOT OF THE PRIMARY BOTTLES

FIG. 9 : THE LOCATION OF THE DOWNCOMER RESTRICTIONS

FIG. 10 : THE ONSET-OF-CHOKING FLOW RATE FOR A 36” OD BOTTLE AS A FUNCTION OF THE BOTTLE SLOPE (NO LIQUID FLOW TO OTHER BOTTLES)

FIG. 11 : THE STORAGE CAPACITY OF THE PRIMARY AND SECONDARY BOTTLES IN A DUAL – SLOPE SLUG CATCHER BOTTLE IN A DUAL-SLOPE SLUG CATCHER DISTANCE BOTTOM HEADER –RISER 300 m SLOPE OF STORAGE SECTION EITHER 0.4 % OR 1 %

FIG. 12 : THE RECOMMENDED LOCATION OF THE BOTTOM HEADER RELATIVE TO THE BOTTLE

FIG. 13 : RECOMMENDED GAS OUTLET CONFIGURATIONS (VIEW FROM ABOVE)

APPENDIX I THE TWO-LIQUID TEST FACILITY FOR THE SIMULATION OF SLUG CATCHERS For the assessment of the performance of existing and conceptual slug catchers a two-liquid test facility has been in use at KSLA since 1980. Through a perspex model of the slug catcher under study, kerosene is circulated representing the gas phase in practice. A condensate slug is simulated by diverting the kerosine stream through a tank containing a 55%w ZnCl2 concentrate in water, thus forcing the concentrate into the slug catcher. The reason for the choice of a kerosene/ZnCl2 concentrate system rather than air/water for instance is that, as far as the density ratio is concerned, a much better similarity is obtained (see Table I.1). The mismatch as regards the viscosities is less important because both in the full-scale and model slug catcher (if the scale is not too small) all fluids flow turbulently. In the simulation studies Froude scaling is applied, i.e. the flow conditions in the model slug catcher are thought to be representative for those in the full scale slug catcher characterized by the same Froude number Fr = v SG

ρG (ρL − ρ G ) Dpipe g

where vSG is the superficial velocity of the light phase in the pipeline upstream of the slug catcher. For the 1:20 perspex model of the Den Helder slug catcher this implies for instance: vfull-scale = 11.6 * vmodel The experience with the Den Helder slug catcher has shown that the concept of Froude scaling is correct and also that the two-liquid modelling approach is justified.

TABLE I-1 : COMPARISON OF PHYSICAL PROPERTIES

APPENDIX II THE LIQUID STORAGE CAPACITY AND HEIGHT OF A MULTIPLE-PIPE SLUG CATCHER 1.

Storage capacity The following assumption are made: (1)

The slug catcher is filled to its maximum capacity when the liquid level in the primary bottles is at the level of the bottle slope transition (in the case of a dual-slope slug catcher) and just reaches the bottle section immediately underneath the riser(s) (see Fig. II.1)

(2)

The bottle slope(s) is (are) smaller than 5 %. Then cos θ ≈ 1 and sin θ ≈ tgθ

(3)

There are npb primary and nsb secondary bottles.

(4)

For the bottle part between the bottom header and risers a length of L is available. Volsc = Volbh + n1 Volpb +n2 Vol sb Volbh =

π 2 D 4 bh

Lbh

Single-slope slug catcher Volpb = Volsb =

π 2 D 4 b

Db   L −   2tgθ 

Dual-slope slug catcher primary bottle : Volpb =

 Db  L −  2tgθ 2  

π 2 D 4 b

secondary bottle : Volsb =

π 2 D 4 b

HL, rf ≈ If θ 2 >>θ1, H L, rf ≈ 1 and Volsb ≈ 2.

  Db 1 − HL, rf  L − 2tgθ 2  

(

)

tgθ 2 − tgθ 1 tgθ 2

π 2 DL 4 b

Height difference between lower end of bottle and riser foot Single-slope slug catcher ∆hpb = ∆hsb = L tg θ Dual-slope slug catcher

 

primary bottle

:

∆hpb = L tg θ1 + Db  1 −

secondary bottle

:

∆hsb = L tg θ1

tgθ1   tgθ 2 

FIG. II-1 : THE FILLING DEGREE OF THE BOTTLES IN A MULTIPLE-PIPE SLUG CATCHER FILLED TO ITS MAXIMUM CAPACITY

APPENDIX III SETTLING OF SMALL DROPLETS IN THE ENTRANCE SECTION OF THE PRIMARY BOTTLES For the calculation of the time required to settle small droplets in the entrance section of the primary bottles the following assumptions are made. (1)

The mist droplets are spherical and have all the same diameter.

(2)

Effect of turbulence on settling velocity is neglected. In other words, the vertical velocity fluctuations in the turbulent gas flow compensate each other. Then the settling velocity is equal to that in stagnant medium.

According to Ref. 16 the following formulae apply for the settling velocity of an isolated mist droplet If Re =

v s ρG d ηG

for Re < 1

(ρL − ρ G ) g d2 vs = 18ηG 3

for 1 < Re < 10

vs follows from 2

C wR e =

4 d3 ρ G (ρL − ρ G ) g 3 n2G

and Fig. II-41 in Ref. 15 3

for 10 < Re < 10

5

v s = 176 .

(ρL − ρ G ) g d ρG

For the calculation of a droplet in a swarm the relationship of Richardson and Zaki holds (See Also Ref. 16) (vs) swarm = vs (1 – H)

n

where n ranges from 4.56 to 2.39 for 0.1 ≤ Re ≤ 500. Furthermore the conservative assumption is made that the mist droplet has to travel a distance D, so ts =

D ( v s )swarm

Fig. III.1 shows for the Den Helder slug catcher tsettle as a function of the droplet diameter for several values of Hmist. Rather arbitrarily it is assumed that droplets larger than 0.5 mm should settle. From Fig.III.1 it is seen that this is achieved within about 4 s even when H is as high as 0.2.

FIG. III-1 : THE TIME REQUIRED TO SETTLE SMALL LIQ. DROPLETS AS A FUNCTION OF THE DROPLET DIAMETER IN THE FIRST PART OF A PRIMARY BOTTLE

APPENDIX IV SIMULATION OF BOTTLE CHOKING (KELVIN-HELMHOLTZ INSTABILITY) IN THE AMSTERDAM TRAY TEST COLUMN In 1977 at KSLA a test installation was built at the Amsterdam Tray Test Column (TTC) to study the twophase flow phenomena occurring during filling of a slug catcher bottle. Tests were performed in a model slug catcher bottle with an inner diameter of 30 cm and an inclination of 1.5 % using amongst others butane up to a pressure of 13 bar. In contrast to the real situation the test bottle in the TTC experiments had no riser. The liquid phase flowed into the bottle from the upper end and the gas from the lower end. It was observed that if the velocity of gas relative to the descending liquid stream exceeded a critical value, excessive wave formation took place and eventually "choking" of the pipe occurred. This "choking" effect is known as "slugging" or Kelvin-Helmholtz instability in the open literature (Wallis and Dobson 13). For more details see Ref.14. 13,15 14 Based on the TTC data supplemented with information from the literature , Darton et al. arrived at the following criterion for choking. Choking takes place, if

HL vL

ρL

g D(ρL − ρ G )

where

φ=

HL vL ρL (1 − HL ) v G

ρG

−

1 0.36 6.34 2 + 3.6 HL θ 2 (1 − HL )

>0

LIST OF SYMBOLS Cw

drag coefficient

d

droplet diameter

Fr

Froude number

H

hold-up

g

gravity constant

m/s

L

length

m

m

number of risers per bottle

n

number of bottles or exponent in Richardson-Zaki equation (Appendix III)

v

velocity

m/s

Vol

volume

m

Re

Reynolds number associated with particle settling

m

  ρG  = v SG  (ρL − ρ G ) D g   2

3

 v ρ d  Re = S G  ηG   Greek symbols n

dynamic viscosity

λ

load factor

 ρG    = v SG ρL − ρ G  

Ns/m

2

m/s

or volumetric fraction of liquid flow in two-phase flow θ

angle between bottle and horizontal plane

ρ

density

φ

flow parameter

kg/m

  HL vL ρL =   (1 − H )v ρ G   L G

3

Subscripts b

bottle

bh

bottom header

G

gas

int

intercept

L

liquid

goh

gas outlet header

pb

primary bottle

rf

foot of the riser

s

settle

S

superficial

sb

secondary bottle

1

referring to separation section of the slug catcher in the dual-slope concept

2

referring to storage section of the slug catcher in the dual-slope concept

Superscript -

average

REFERENCES

1.

A.R. Huntley and R.S. Silvester, Hydrodynamic analysis aids slug catcher design, Oil and Gas J. 81(1983)95, Sept. 19.

2.

A.E. Martin, Handling liquids in offshore gaslines gets new approach. Oil and Gas J. 79 (1981) 143-148.

3.

A. Bos, J.A. van Klaveren and R. Meerhoff, "Liquid carry-over at the NAM slug catcher in Den Helder". I. Problem analysis using the neutron back-scattering technique", AMGR.82.288.

4.

A. Bos and J.G. du Chatinier, "Liquid carry-over at the NAM slug catcher in Den Helder. II. Model studies", AMGR.82.335.

5.

A. Bos, C.A. Kok and J.G. du Chatinier, "Liquid carry-over at the NAM slug catcher in Den Helder. III. Solving of the problem", AMGR.83.266.

6.

A.Bos and J.G du Chatinier, “Analogue modelling of the St.Fergus slug catcher”, AMRG.84.134.

7.

A. Bos and J.G. du Chatinier, "Improvement of the Bintulu (Sarawak) slug catcher", AMGR.83.354.

8.

C.M. Verheul, "A model study of existing and new slug catcher configurations for natural gas pipelines from off-shore platforms", AMGR.0208.71.

9.

A. Hortulanus and P.E.M. Duyvesteyn, "Model study for the modification of the Bacton slug catcher", AMOR.0001.73.

10.

J.G. du Chatinier and A. Bos, "Engineering studies on liquid entrainment in slug catchers. Model tests for the design concept of NAM's F-3 slug catcher", AMRS.84.07, PR-1.

11.

N. Trompe, R.V.A. Oliemans and J.A. ten Hagen, "TWOPHASE", A computer program for the hydraulic design of horizontal and inclined pipelines with two-phase gas/liquid flow. guide", AMGR.79.391.

12.

H.D. Beggs and J.P. Brill, A study of two- phase flow in inclined pipes, J. Petr. Tech. 25 (1973) 607.

13.

G.B. Wallis and J.E. Dobson, The onset of slugging in horizontal stratified air -water flow, Int. J. Multiphase Flow 1 (1973)173.

14.

R.G. Darton and G. Lentz, "Amsterdam tray test column. Countercurrent two-phase flow in a near-horizontal pipe. Investigation of conditions in a simulated slug catcher storage bottle during filling", Amsterdam Tray Test Column Test Report 79, Layout 28,AMGR.83.265.

15.

E. Kordyban and T. Ranov, "Mechanism of slug formation in horizontal two-phase flow", J. Basic Engng. 92(1970)857.

16.

W.J. Beek and K.M.K. Muttzall, "Transport Phenomena", John Wiley & Sons Ltd.,London, 1975, p. 101 ff.