Upstream Research

Slugging in a Subsea Compact Separation System

Ed GraveFractionation & Separation Advisor Houston, TX USA

MCE Deepwater DevelopmentApril 8 & 9, 2014Madrid, Spain

Upstream Research

ExxonMobil Subsea Compact Separation SystemExxonMobil Upstream Research Company (URC) designed and is testing a compact separation system for application in 3000m water depth and internal pressures up to 690 barEM qualification philosophy is to qualify for a wide range instead of specific field conditions to reduce timeline of applicationRobust, flexible to inlet fluids, and scalable

API Gravity:  19˚‐38˚Oil Rate: 60 kBPD/trainGas Rate: 1250 – 4000 Sm3/dayWater Cut: 0‐90%Slug Size: 5m3

Currently being qualified at ProlabNL(3) Crudes with API 19˚, 28˚ & 38˚Scaled to 10‐15 kBPDMethane Gas at 45 barsGas Rate: 33‐497 Am3/hrWater cuts 10‐70%Slug Tests: 0.2 to 0.6m3 Upstream Research

2

Simplified Schematic of ExxonMobil Subsea Compact Separation System

being tested at ProLabNL.

ExxonMobil’s Subsea Compact Separation System

OilURC Inlet Gas/Liquid Slug‐Catcher

URC Liquid/Liquid Pipe Separator and Level Controls

Inlet

Gas

Water

ASCOM Inline De‐sander

Cyclonic Inline Gas Polisher

Reject

Fast‐Acting MokveldSubsea Control Valve

Liquid Slug Generator

FMC‐CDS IEC tested at FMC separately

Aker CEC™

OiWMonitors

3

ASCOM Hiper™ MixedFlow 2‐stage de‐oiling Hydrocyclones and Canty Oil/Water Monitors

Slugging in a Compact Separation SystemCompact Inline Separation devices do not perform well during slugging conditions

Very small residence time and buffer volumeNormally optimized for one phaseSensitive to change in fluid density (GVF) and flow rates…inlet momentum changes and potential collapse of the swirlHomogenous flow is desired

This results in lower separation performance with undesirable liquid entrainment and gas carryunderSlugging effects can propagate through other downstream equipmentSlugging effects can be minimized with sophisticated slug dampening control systems in combination with a flow conditioning device

Deliquidizer - Courtesy of FMC Technologies, Inc.

Ref: Hannisdal, et. al. 2012 OTC 23223

4

Slugging in a Compact Separation SystemSubsea Compact Three Phase Separation System Studied the effects of hydrodynamic slugsStudied different size slugs and frequency to determine:

Effectiveness of ExxonMobil’s Subsea Slug Catcher in buffering slugsThrough CFDs air/water tests model fluid high pressure tests studied a number of proprietary designs. Final design more effective for liquid dominated servicesMinimize liquid carryover to the ASCOM Monoline on the gas outletReduce dramatic liquid rate fluctuations downstreamMeasure the liquid outlet GVF to downstream equipmentLiquid level control & hold-up

ASCOM MonolineLiquid CarryoverMeasure gas carryunder from bootLiquid level control

5

Slugging in a Compact Separation SystemStudied different size slugs and frequency to determine:

ExxonMobil’s Pipe Separator The effect on oil-water separationChange on interface height and control. Effect on emulsion stability due to increased GVF and rate changesMeasure oil outlet GVFMeasure OIW & WIOLiquid level control & hold-up

Dynamic simulation based on a potential field case

Optimized control systemsIP around novel control system

Never the intention to validate dynamic model based on flow loop data

Due to pressure balance around a close loop systemNo oil/water pumps on dischargePreferred control system could not be implemented 6

Slugging in a Compact Separation System

Slug Generator at ProlabNL

7

Slug GeneratorAllowed for controlled slug size feeding the system of various sizes: 0.1 to 0.6m3

Number and frequency could be controlledSplit flow possible with steady state flow and liquid slug volume

ExxonMobil’s Subsea Slug CatcherMain function to separate the gas and liquid phasesTemporary storage of liquids to buffer downstream equipment and maintain steady state conditionsIt did not consider the increase in flow rate tolerance of the downstream equipment as this was a test loopLimitations of a closed flow loop

0

70

140

210

280

350

420

0

20

40

60

80

100

120

11:57:20 AM 11:59:20 AM 12:01:20 PM 12:03:20 PM 12:05:20 PM 12:07:20 PM

Level (mm)

Flow

 (m3/h)

Time (hh:mm)

Level and flow vs Time

Flow Slug Catcher outFlow Oil outLevel Slug GeneratorLevel Slug CatcherFlow water out

0

70

140

210

280

350

420

0

20

40

60

80

100

120

12:37:50 PM 12:39:50 PM 12:41:50 PM 12:43:50 PM 12:45:50 PM 12:47:50 PM

Level (mm)

Flow

 (m3/h)

Time (hh:mm)

Level and flow vs Time

Flow Slug Catcher out

Flow Oil out

Level Slug Generator

Level Slug Catcher

Flow water out

Slugging in a Compact Separation SystemResults – Slug Catcher Medium Crude

API 28˚ 50˚C 20%WC – 0.4m3 Slug Auto Control

Slug ReleasedSlug Released

Fluid out of SC feed to PSFluid out of SC feed to PS

Oil out of PSOil out of PS• Slug catcher caught the slug effectively. • However, control valve released the slug

too quickly• Manual control of water outlet valve did

not help

API 28˚ 50˚C 20%WC – 0.4m3 Slug Manual Control

Manually opened water valveManually opened water valveRef: run 221M50SLmRef: run 221M50SLm

Ref: run 221M50SLaRef: run 221M50SLa Fluid out of slug generatorFluid out of slug generator

8 All data generated by or on behalf of ExxonMobil

Slugging in a Compact Separation SystemResults – Monoline Medium Crude

0

70

140

210

280

350

24

26

28

30

32

34

36

12:37:50 PM 12:39:50 PM 12:41:50 PM 12:43:50 PM 12:45:50 PM 12:47:50 PM

Level (mm) a

nd Gas flow

 (m

3/h)

Flow

 (m3/h)

Time (hh:mm)

Level and flow vs Time

Purgeflow MonoLine

Level Slug Catcher

Gas flow System in

Gasflow MonoLine out

Slug ReleasedSlug ReleasedGas rate leaving boot through purge lineGas rate into monoline decreased during sluggingGas rate leaving boot through purge lineGas rate into monoline decreased during slugging

0.01

0.1

1

10

0

20

40

60

80

100

120

12:37:50 PM 12:39:50 PM 12:41:50 PM 12:43:50 PM 12:45:50 PM 12:47:50 PM GVF

 (%), Ve

locity (m

/s) a

nd Residen

ce time 

(min)

Flow

 (m3/h) and

 Level (%

)

Time (hh:mm)

Level and flow vs Time

Flow Slug Catcher out

Filling Slug Catcher

GVF Slug Catcher out

Velocity Gas

Velocity Liquid

Residence time

GVF increases due under carry from additional turbulenceGVF increases due under carry from additional turbulence

Results – Slug Catcher Medium Crude• Liquid carryover rate to the Monoline during slug

trials were <2-3%LVF

• Decreased gas flow rate into Monoline due to liquid slug

• GVF in liquid leaving slug catcher remained fairly constant ~1%

Ref: run 221M50SLaRef: run 221M50SLa

Courtesy of Ascom Advanced Separation Company

ASCOM Monoline

9 All data generated by or on behalf of ExxonMobil

Slugging in a Compact Separation SystemResults – Pipe Separator Medium Crude

12:37:50 PM 12:39:50 PM 12:41:50 PM 12:43:50 PM 12:45:50 PM 12:47:50 PM

Den

sity (k

g/m3)

Time (hh:mm)

Density Tracerco vs TimeSlug ReleasedSlug Released

API 28˚ 50˚C 20%WC – 0.4m3 Slug Manual Control

Interface level increased as the slug was releasedInterface level increased as the slug was released

Slug ReleasedSlug Released

• Interface level remains fairly constant with some effect on oil-water separation quality. WIO suddenly increased

• Note there’s a process delay and a smoothing function with the TRACERCO ProfilerTM

Interface level at steady state conditions prior to sluggingInterface level at steady state conditions prior to slugging

Ref: run 221M50SLaRef: run 221M50SLa

10 All data generated by or on behalf of ExxonMobil

8:43:10 PM 8:47:09 PM 8:51:10 PM 8:55:09 PM 8:59:09 PM 9:03:10 PM

Flow

 (m3/hr)

Time (hh:mm)

Flow vs Time

Water inOil inSlug Generator & Catcher inPipeSep inOil outWater out

Slugging in a Compact Separation SystemResults – Slug Catcher Heavy Crude

API 19 70˚C 20%WC – 0.4m3 Slug Manual Control

Changed parameters on control valve to limit %openChanged parameters on control valve to limit %openSlug ReleasedSlug Released

• With this crude slate modified parameters on oil outlet control valve

• Slug Catcher now holding and buffering flow better

• Interface level in pipe separator remains constant

• Temporary change in oil-water separation

Results – Pipe Separator Heavy Crude

Slug ReleasedSlug Released

Ref: run 221H70 DMSLRef: run 221H70 DMSL

11

All data generated by or on behalf of ExxonMobil

Slugging in a Compact Separation SystemResults – Slug Catcher Light Crude; Multiple slugs

API 38 50˚C 70%WC – 0.6m3 Triple Slugs

0

90

180

270

360

450

0

50

100

150

200

250

11:35 11:37 11:39 11:41 11:43 11:45 11:47 11:49 11:51

Level (mm)

Flow

 (m3/h)

Time (hh:mm)

Level and flow vs Time

Flow Slug Catcher outFlow Oil outLevel Slug GeneratorLevel Slug CatcherFlow water out

0.01

0.1

1

10

0

40

80

120

11:35 11:37 11:39 11:41 11:43 11:45 11:47 11:49 11:51

GVF

 (%), Ve

locity (m

/s) a

nd 

Reside

nce tim

e (m

in)

Flow

 (m3/h) and

 Level (%

)

Time (hh:mm)

Level and flow vs Time

Flow Slug Catcher outFilling Slug CatcherGVF Slug Catcher outVelocity GasVelocity LiquidResidence time

Slugs ReleasedSlugs Released Flow and level of slug catcher the same…different scalesFlow and level of slug catcher the same…different scales

• Very large slugs of 0.6m3

• Lighter crudes lower overall separation efficiency due to higher GOR

• Separation efficiency about 98%• Liquid carried over to monoline

GVF out of slug catcher peaked at ~0.3%GVF out of slug catcher peaked at ~0.3%Ref: run 710aL50SLRef: run 710aL50SL

12 All data generated by or on behalf of ExxonMobil

Slugging in a Compact Separation SystemResults – Monoline; Multiple Slugs

Liquid entered the slug catcher into monolineNote liquid density…lower due to gas carryunderLiquid entered the slug catcher into monolineNote liquid density…lower due to gas carryunder

Gas leaving from boot stops as it’s mostly liquid with gas carryunder. Insufficient time to degas inside the boot

Gas leaving from boot stops as it’s mostly liquid with gas carryunder. Insufficient time to degas inside the boot

• Monoline effectively captures the liquid carryover with an efficiency of 99.97% for this run

• Due to sudden increase in liquid, boot did not have sufficient time to degas. This gas carryunder had insignificant effect on the pipe separator due to overall small liquid volume.

0

200

400

600

800

1000

1200

0

5

10

15

20

25

30

11:35 11:37 11:39 11:41 11:43 11:45 11:47 11:49 11:51 Level (mm) a

nd Gas flow

 (m3/h)

Flow

 (m3/h)

Time (hh:mm)

Level and flow vs Time

Flow monoline boot out.

Level MonoLine boot

Density Boot out

0

50

100

150

200

250

0

5

10

15

20

11:35 11:37 11:39 11:41 11:43 11:45 11:47 11:49 11:51 Level (mm) a

nd Gas flow

 (m3/h)

Flow

 (m3/h)

Time (hh:mm)

Level and flow vs Time

Purgeflow MonoLine

Level Slug Catcher

Gas flow System in

Gasflow MonoLine out

API 38˚ 50˚C 70%WC – 0.6m3 Triple Slugs

Ref: run 710aL50SLRef: run 710aL50SL

13 All data generated by or on behalf of ExxonMobil

Slugging in a Compact Separation SystemResults – Pipe Separator Light Crude; Multiple slugs

11:35 11:37 11:39 11:41 11:43 11:45 11:47 11:49 11:51

Den

sity (k

g/m3)

Time (hh:mm)Density Tracerco vs Time

Slug ReleasedSlug Released

API 38˚ 50˚C 70%WC – 0.6m3 Triple Slugs

Note: Interface level higher at 70% WC than at 20%

Note: Interface level higher at 70% WC than at 20%

• Very little effect downstream into the pipe separator on interface control

• WIO temporary increased

Ref: run 710aL50SLRef: run 710aL50SL

14 All data generated by or on behalf of ExxonMobil

Slugging in a Compact Separation SystemConcluding Remarks

Slugging effects in a compact separation system can be dampened and controlled with a properly designed flow conditioning systemAttention to control system is a MUST!!Pressure drop around each piece of equipment needs to be considered to balance flows during upset conditionsRecycling streams to diminish slugging effects or increase turndown capabilities not necessaryClean 3-phase separation that meets target performance was achievedCompact separation system much more complex than traditional gravity separation vesselValidation of dynamic simulation not possible in a closed test flow loopHigh pressure testing provided valuable dynamic and separation performance data to be incorporated into the dynamic simulation.Fast acting control valves are not requiredCompact separation systems are less forgiving to reservoir model prediction errors

15

Thank You!! Questions?

Upstream Research

Thank You! Questions?

Ed GraveFractionation & Separation Advisor edward.j.grave@exxonmobil.com1‐713‐289‐4770Houston, TX USA

BACK‐UP SLIDES

ProlabNL Simplified Flow Loop Schematic

18

ExxonMobil URC Subsea Compact Separation System

Test Unit