DNVGL-RP-O501 Edition August 2015

Transcription

1 RECOMMENDED PRACTICE DNVGL-RP-O501 Edition August 2015 Managing sand roduction and erosion The electronic df version of this docuent found through htt:// is the officially binding version. The docuents are available free of charge in PDF forat.

2 FOREWORD DNV GL recoended ractices contain sound engineering ractice and guidance. August 2015 Any coents ay be sent by e-ail to This service docuent has been reared based on available knowledge, technology and/or inforation at the tie of issuance of this docuent. The use of this docuent by others than DNV GL is at the user's sole risk. DNV GL does not accet any liability or resonsibility for loss or daages resulting fro any use of this docuent.

3 CHANGES CURRENT General This docuent suersedes DNV-RP-O501, Noveber Text affected by the ain changes in this edition is highlighted in red colour. However, if the changes involve a whole chater, section or sub-section, norally only the title will be in red colour. On 12 Seteber 2013, DNV and GL erged to for DNV GL Grou. On 25 Noveber 2013 Det Norske Veritas AS becae the 100% shareholder of Geranischer Lloyd SE, the arent coany of the GL Grou, and on 27 Noveber 2013 Det Norske Veritas AS, coany registration nuber , changed its nae to. For further inforation, see Any reference in this docuent to Det Norske Veritas AS, Det Norske Veritas, DNV, GL, Geranischer Lloyd SE, GL Grou or any other legal entity nae or trading nae resently owned by the DNV GL Grou shall therefore also be considered a reference to. Changes current Main changes August 2015 The current 2015 revision includes the following ain udates: a) Docuent title has been changed fro Erosive Wear in Piing Systes to Managing sand roduction and erosion. b) Outline and list of considerations for a sand anageent strategy. c) New guidance on erosion odel for flexible ies with interlock carcass. d) New erosion odels for choke valves. e) Guidance on coutational fluid dynaics (CFD) erosion odelling. f) Erosion odel validation cases. Editorial corrections In addition to the above stated ain changes, editorial corrections ay have been ade. Acknowledgeent The current docuent is develoed in co-oeration with a large nuber of ajor oil and gas oerators. DNV GL is grateful for the financial suort to research and develoent and for being allowed to aly results fro rojects with these oerators to establish this industry guideline. Recoended ractice DNVGL-RP-O501 Edition August 2015 Page 3

4 CONTENTS CHANGES CURRENT... 3 Sec.1 General Introduction Alication of this docuent Reference to codes and standards Abbreviations Definitions Verbal fors Bibliograhic references...8 Sec.2 Sand anageent strategy General Consequences of sand roduction Sand roduction otential Philosohy for acceting and anaging sand roduction Goals and success factors Preises and accetance criteria Tolerable erosion Sand handling caacity Risk assessent Class of erosive service Risk assessent and ranking Safeguards Strategy ileentation Training requireents Status reorting and eriodic revision Status reorting Revision of strategy...16 Sec.3 Fundaentals of article erosion General List of sybols Indexes Erosive agents Non-erosive agents Characterisation of erosive wear Erosion resonse odel...21 Sec.4 Eirical odels for sand article erosion General Alication and liitations Geoetry correction factors Model inut araeters Erosion resonse odel Bulk roerties Sand content...28 Contents Recoended ractice DNVGL-RP-O501 Edition August 2015 Page 4

5 4.5 Sooth and straight ies Welded joint Pie bends Blinded tee Reducers Intrusive erosion robes Flexible ies with interlock carcass Production chokes General Choke selection Oeration Insection and condition onitoring Erosion odel for choke gallery...40 Sec.5 Model araeters for other erosive agents Sec.6 Software odel A. A Safeguards - anaging sand roduction and erosion A. B Material erosion testing A. C Coutational fluid dynaics erosion siulation A. D Production chokes A. E Model validation Contents Recoended ractice DNVGL-RP-O501 Edition August 2015 Page 5

6 SECTION 1 GENERAL 1.1 Introduction This recoended ractice is develoed for the oil and gas industry to rovide guidance on how to safely and cost effectively anage the consequences of sand roduced fro the oil and gas reservoirs through roduction wells, flowlines and rocessing facilities. The ultiate goal of this docuent is to assist revention of incidents related to sand that ay cause har to eole, environent or assets and without causing unnecessary restrictions to roduction erforance. This docuent was first develoed and issued by DNV in 1996 and has since then only been subject to inor adjustents. The current revision includes ore of the background aterial for the erosion resonse odels and further guidance on develoent, ileentation and follow u of a high-level sand anageent strategy. Objective of this docuent is to rovide guidance on how to safely and cost effectively anage the consequences of sand roduction and erosion through the different stages of design and oeration of oil and gas roduction facilities. 1.2 Alication of this docuent Sec.2 of this docuent rovides guidance on develoent, ileentation and follow u of a field sand anageent strategy. This section is riarily intended for oerating coanies, but should also serve as a reference docuent for engineering coanies in different stages of design, fabrication and construction. Sec.3 to [4.12] of this docuent rovides eirical odels for rediction of article erosion in standard iework coonents. The odels offer a ore secific ethod for diensioning of iework and coonents exosed to erosive wear coared to the erosional velocity aroach secified in API-RP-14E or NORSOK P-100. The erosion odels ay be used to deonstrate coliance between syste design, tolerable erosion and sand load either secified in design basis or exerienced in oeration. Suorting aterial relevant for the understanding and transarency of this recoended ractice is included in aendices. 1.3 Reference to codes and standards Docuent code API-6A API-RP-14E API 17J API 17B ISO 13703:2000 NORSOK P-100 NACE Standard MR DNV-OS-F101 Title Secification for wellhead and christas tree equient Recoended ractice for design and installation of offshore roduction latfor iing systes Secification for unbonded flexible ie Recoended ractice for flexible ie Petroleu and natural gas industries - Design and installation of iing systes on offshore roduction latfors Process systes Sulhide Stress Cracking Resistant Materials for Oil field Equient Subarine Pieline Systes Recoended ractice DNVGL-RP-O501 Edition August 2015 Page 6

7 1.4 Abbreviations Abbreviation ALARP ASR CRA DNV DNV GL EOR (IOR) ESP GLR GOR GRP HDPE IOR (EOR) MBR PCS PSD SI WC Descrition as low as reasonable ossible accetable sand rate corrosion resistant alloy Det Norske Veritas Det Norske Veritas Geranischer Lloyds enhanced oil recovery electrical suberged u gas liquid ratio gas oil ratio glass fibre reinforced lastic high density olyethylene increased oil recovery iniu bending radius ie class secification article size distribution secial ite water cut 1.5 Definitions Ter C-steel corrosion drolet erosion erosion erosion-corrosion low alloyed steel aterial degradation ixture velocity oil & gas iing syste stainless steel steel carcass suerficial velocity Definition steels containing less than 1.65% anganese, 0.69% silicon and 0.60% coer loss of aterial or loss of aterial integrity due to cheical or electro-cheical reaction with surrounding environent loss of aterial or loss of aterial integrity due to drolet iact on the aterial surface loss of aterial or loss of aterial integrity due to solid article iact on the aterial surface synergetic effect of erosion and corrosion steel containing agnesia, silicon and coer in quantities greater than those for C-steel and/ or other alloying eleents The total content of alloying eleents shall not exceed 5%. loss of aterial or loss of aterial integrity due to cheical or electrocheical reaction with surrounding environent, or erosive wear resulting fro article and drolet iingeent equal to the su of the suerficial velocities for all hases content in ie ay be either oil or gas includes ies for transortation of fluids and associated ie bends, joints, valves and chokes The general ter covers tubing, flow lines for transortation of rocessed and un-rocessed hydrocarbons. steels alloyed with ore than 12% Cr (weight) inner steel interlock layer used in flexible ies for transortation of hydrocarbon fluids fluid velocities of one hase in iing as if no other fluid hase were resent the ie 1.6 Verbal fors Ter shall should ay Definition verbal for used to indicate requireents strictly to be followed in order to confor to this docuent verbal for used to indicate that aong several ossibilities one is recoended as articularly suitable, without entioning or excluding others, or that a certain course of action is referred but not necessarily required verbal for used to indicate course of action erissible within the liits of the docuent Recoended ractice DNVGL-RP-O501 Edition August 2015 Page 7

8 1.7 Bibliograhic references /1/ dewaard C and Millias DE. Carbonic Acid Corrosion of Steel, Corrosion, Vol. 31, No 5, /2/ dewaard C, Lotz U and Millias DE. Predictive Model for CO2 Corrosion Engineering in Wet Natural Pielines, Corrosion, Vol 47, no 12, /3/ dewaard C and Lotz U. Prediction of CO2 Corrosion of Carbon Steel, NACE Corrosion 93, Paer no 69, /4/ Ikeda A, Ueda M, Vera J, Viloria A, Moralez JL. Effects of Flow Velocity of 13Cr, Suer 13Cr and Dulex Stainless Steels. /5/ Raask E. Tube erosion by ash iaction, Wear No. 13, 1969 /6/ Tilly GP. Erosion caused by Iact of Solid Particles, Treatise of Material Science and Technology, Volue 13, /7/ Finnie I. Erosion of Surfaces by Solid Particles, Wear /8/ Haugen K, Kvernvold O, Ronold A and Sandberg R. Sand Erosion of Wear Resistant Materials, 8th International Conference on Erosion by Liquid and Solid Iact, Cabridge /9/ Lindhei T. Erosion Perforance of Glass Fibre Reinforced Plastics (GRP). /10/ Hansen JS. Relative Erosion Resistance of Several Materials, ASTM STP 664, /11/ Kvernvold O and Sandberg R. Production Rate Liits in Two-hase Flow Systes Sand Erosion in Piing Systes, DNV Reort No , /12/ Huser A. Sand erosion in Tee bends. Develoent of correlation forula DNV Reort No , /13/ Oka YI, Okaara K, Yoshida T. Practical estiation of erosion daage caused by solid article iact. Part 1: Effects of iact araeters on a redictive equation, Wear 259 (2005) /14/ Oka YI, Yoshida T. Practical estiation of erosion daage caused by solid article iact. Part 2: Mechanical roerties of aterials directly associated with erosion daage, Wear 259 (2005) /15/ Barton NA. TÜV-NEL-Research reort 115. Erosion in elbows in hydrocarbon roduction systes: Review docuent, 2003 /16/ IJzerans S, Helgaker JF. Large Scale Erosion Testing of a Flexible Flowline. AOG Conference, 12 th of March 2015 /17/ Kvernvold O, Torbergsen LE, Eriksen R (DNV) and Kjørholt H (Statoil). New Strategy for Sand Manageent to safely Irove Production Perforance Dee Offshore Technology Conference; USA Noveber 2002 /18/ Selfridge F, Munday M (DNV), Kvernvold O (DNV), Gordon B (ConocoPhillis). Safely Iroving Production Perforance through Iroved Sand Manageent, SPE-83979, 2003 /19/ Eiliani CN, Lejon K (Statoil), Lindén M, Engene J, Kvernvold O (DNV), Packan C (Roxar), Clarke D (Coron), Haugsdal T (Claon). Iroved Sand Manageent Strategy: Testing of Sand Monitors under Controlled Conditions, SPE , 2011 /20/ Lejon K, Ree AB, Woster RB, Kvernvold O, Torbergsen L. Sand Production Manageent for Snorre B Subsea Develoent Lessons learned and actions taken, DOT, 2007 Recoended ractice DNVGL-RP-O501 Edition August 2015 Page 8

9 SECTION 2 SAND MANAGEMENT STRATEGY 2.1 General The urose of this chater is to describe an industry ractice alicable to subsea, toside and onshore oil & gas roduction facilities. In addition to the general guidance rovided in this docuent the basis for a sand anageent strategy needs also to consider local regulations, coany secific requireents and guidelines, field secific syste layout and oerational exerience. The current chater rovides guidance on how to establish, docuent and follow u a strategy for anaging sand roduction, considering the following key toics: sand roduction otential consequences of sand roduction hilosohy for acceting and anaging sand roduction goals and success factors reises and accetance criteria risk assessent safeguards strategy ileentation training requireents roduction otiisation condition onitoring and insection status reorting and revision of strategy. As a first rincile the sand anageent strategy shall address all systes that are considered likely to be exosed to solids roduced fro well. The ustrea and downstrea battery liits for the strategy shall therefore be roerly defined. Both assive and active eans to control and onitor sand roduction and the associated consequences should be considered. 2.2 Consequences of sand roduction Sand roduction can have significant consequences for both the roduction and the assets. Key failure odes are related to erosion, sand accuulation, lugging or containation by sand. For the ajority of oil & gas fields, sand fro the reservoir foration is an inevitable by-roduct. Monitoring and controlling sand roduction is iortant for the following reasons: Sand ay cause daage to well coonents such as sand screens, tubing, down-hole safety valves or electrical subersed us (ESPs). Sand ay cause erosion in iing syste and coonents that if undetected ay lead to loss of containent. Sand ay cause erosion in blow-down systes during ESD deressurization or inadvertent routing of roduction to knock-out dru. Particular focus should be on flow restrictions (valves or orifice) and iediate iing downstrea due to otentially high gas velocities resulting fro ressure let-down. Low wall thickness characteristic for flare systes should be acknowledged. Sand ay accuulate in wellbore if wells are roduced below lifting rate for sand, ultiately leading to sanding-in and loss of well. Accuulation of sand in roduction lines ay affect corrosion rates, cause usets during igging oerations or increased ressure resistance during oeration. Accuulation of sand in searators ay cause reduced searation efficiency and carry-over of sand to downstrea systes that are not designed for or have little tolerance for sand. Instruentation ay be influenced, otentially affecting safety critical systes for shut-down or rocess control. Challenges with sand volue handling caacity and reoval of sand fro the rocess ay cause usets or unlanned shut-downs. Recoended ractice DNVGL-RP-O501 Edition August 2015 Page 9

10 Sand roduction ay influence roduced water quality in a negative direction. Overboard disosal of roduced water containing sand ay cause erosion to iework and coonents or reduced well injectivity if re-injected. Accuulation of sand in rocess- or safety critical valves ay iair valve erforance due to blockage or increased friction. Sand ay daage rotating equient such as us (and coressors). Particularly for non-corrosion resistant aterials, even a oderate sand erosion otential ay affect flow accelerated corrosion (FAC). 2.3 Sand roduction otential The otential for sand articles being released and transorted fro foration to wellbore is deterined by a nuber of colex factors requiring exert evaluation by reservoir geologists and coletion engineers. The sand otential is norally assessed based on knowledge available early in the field concet develoent; however the inforation available at this oint is often fairly liited and associated with significant uncertainty. It is not the urose of this docuent to rovide guidance on how to assess sand roduction otential, rather to rovide a sufficient descrition of the key araeters relevant for anaging sand roduction in field oeration. Different forations have different echanical strength. Rock strength is characterised in ters of level of consolidation which describes how well sand grains are ceented together. Rock-strength is norally deterined based on core-sale testing, and the corresonding sand otential is assessed for the field life considering the lanned recovery strategy. It should be acknowledged that core sales norally taken fro exloration wells are not necessarily reresentative for all subsequent roduction wells and therefore associated with uncertainty. The true sand otential will also be a function of the actual reservoir recovery strategy during field life. A reduction in reservoir ore ressure will increase the load fro the overburden on the rock foration, which in turn will increase the otential for rock failure and foration of sand grains. This exlains why sand roduction is often exerienced to increase in tail end- and low ressure roduction. Water increases the obility of sand to well bore due to reduced surface tension between sand and water coared to sand and oil/gas and otentially increased drag on the sand grains. Onset of sand roduction is therefore often found to coincide with onset of water roduction. Raid transients in well oeration ay also affect the near well bore zone in a negative direction causing increase in sand roduction and should therefore to the extent ossible be avoided. 2.4 Philosohy for acceting and anaging sand roduction Adoting a hilosohy of zero sand roduction will in any cases ut significant restrictions on the field roduction otential or cause reature abandonent of wells. A sand anageent strategy should therefore be based on a cobination of iniising sand roduction by eans of sand control where it is coercially viable and racticable to do so and anaging with the consequences of sand roduction exerienced in oeration. Allowing for a certain sand roduction that can be safely and effectively anaged ay also significantly increase the field roduction otential. Managing sand roduction eans allowing for sand roduction fro individual wells and through co-ingled streas deending on consequences for integrity and/or availability of the facilities. This enables otiisation of roduction fro different wells and sub-systes, hence reventing unnecessary restrictions without coroising on safety and reliability of the syste. For a given cobination of field design and oerating conditions the accetable sand rate (ASR) will be liited by the following two ain factors: accetable rate of consution of erosion allowance for roduction iing and coonents sand volue handling caacity in the rocess, cleaning and disosal syste. A rerequisite for this hilosohy is that a sufficient syste for continuous onitoring of oerating Recoended ractice DNVGL-RP-O501 Edition August 2015 Page 10

11 conditions and sand roduction cobined with facilities for handling of sand in the rocess syste is in lace. Adoting an ASR hilosohy should be subject to a risk assessent where the required safeguards for anaging the consequences are established and followed u, ref [2.7]. The strategy for how to anage sand should be outlined early in the field develoent to ensure aroriate sizing and selection of equient and instruentation for onitoring, controlling and handling of sand roduction. The strategy should reflect both ethods to control and onitor sand roduction and aroriate safeguards to control the risk. The strategy should also be aligned with the overall Asset Integrity Manageent Syste for the roduction facilities. 2.5 Goals and success factors The overall goals for a sand anageent strategy should be defined and validated on regular intervals, aiing for: no loss of roduct to environent due to failures caused by erosion no erosion daages causing unlanned shut-down or aintenance/reair/relaceents liited (accetable) rocess usets due to sand roduction, accuulation, cleaning and disosal quality of disosed roduced water/sand in coliance with oerator and authority requireents axiized roduction otential (no unnecessary restrictions due to sand). The following key factors should be considered for a successful ileentation and adherence to the strategy: High level of knowledge within the field organisations related to the consequences of sand roduction. Clearly defined roles and resonsibilities related to follow u of safeguards. Corehensible steering criteria related to allowable sand roduction for the individual wells and systes. Systes in lace for onitoring and reorting the effects of cobined oerating conditions and sand roduction. Confired corresondence between systes for onitoring the effects of sand roduction (erosion and deosition) and results fro insection and aintenance caaigns building confidence in the strategy. 2.6 Preises and accetance criteria The strategy shall be based on the secific syste design, field develoent strategy and oeration. The reises for the strategy should be established rior to erforing the risk assessent (ref. [2.7]) and should as a iniu consider the following: reservoir conditions and lanned recovery strategy foration strength, exected sand otential and sand control sand characterisation, article size distribution (PSD) and fraction of erosive agents general field layout: considering subsurface, subsea and toside or onshore facilities - liited ustrea by the lower coletion of the roduction wells and downstrea by the battery liits where the rocessed strea can be considered free fro solids secifications for iing syste, anifolds, flowlines and coonents with resect to geoetrical layout, sizing and tolerable erosion rocess conditions sand handling caacity in the rocess syste, considering ethods for sand reoval, cleaning and disosal safe enveloe for sand transort, considering roduction wells, flowlines and iework field service life, also considering otential lans for life tie extension (IOR/EOR) ileented safeguards related to design, rocedures and instruentation for follow u of sand roduction and its consequences oerational exerience Recoended ractice DNVGL-RP-O501 Edition August 2015 Page 11

12 lanned future odifications, considering new tie-ins or change in rocess conditions field organisation and resonsibilities coany secific requireents local regulations Tolerable erosion An absolute zero tolerance for erosion is difficult to relate to fro a ractical ersective, hence a iniu tolerance for erosion needs to be secified. An accetable erosion rate (e.g. /yr) needs to consider the reaining target service life of the syste and the colexity and cost of syste reair or relaceent. For steel iework tolerable erosion should be identified with reference to one of the following otions: Allow for a iniu erosion allowance of 0.5 (1/50 ) with reference to tyical accuracy of handheld UT equient for wall thickness easureents. Erosion allowance according to ie class secification (PCS). Accetable utilisation of corrosion allowance or CRA cladding noted in PCS. To what extent the corrosion allowance can be utilised as erosion allowance needs to consider the level of corrosive service. For coonents with internal CRA cladding, the erosion allowance ay be taken as a ercentage of the cladding thickness. Erosion allowance identified based on iniu wall thickness requireent according to secified syste ressure rating (ie stress analysis ay be required for this otion). De-rating of the syste to increase tolerable erosion should only be considered as a last otion, and should be subject to a thorough assessent also considering future oeration of the syste. For secial ites the erosion allowance should be advised by vendor, considering otential effects on functionality, erforance or containent. E.g. for flow eters erosion ay affect eter calibration, for choke valves erosion ay affect controllability and for cyclone units erosion ay affect searation erforance Sand handling caacity To iniise rocess usets and to reduce erosion otential, sand should be searated fro the rocess strea as early in the rocess as ossible. In ost cases this will be the inlet searator(s) for the rocess train. The standard aroach is to let the sand searate with the liquid hase by gravity and accuulate in the botto art of the searator, ref. Figure 2-1. Accetable sand accuulation (before reoval is required) deends on a nuber of colex factors such as searation efficiency, ethod of reoval and sand carryover to downstrea rocess systes. In any cases carry-over of saller articles to downstrea systes cannot be fully avoided and needs to be anaged, and the consequences need to be assessed. General requireents to systes for reoval of sand fro rocess vessels are given in /NORSOK P100; section /. For rocess vessels where sand needs to be reoved anually, the sand handling caacity should be liited to an accetable sand build u between lanned intervals for anual reoval. Manual reoval will in ost cases require rocess shut-down having significant iact on lant availability. Fro oerational exerience the total sand volue accuulation that can be acceted is tyically in the order of deending on vessel size, configuration and oeration. Given a tyical interval of 3 years between ajor aintenance (etying of the searator) this eans that a axiu sand roduction of aroxiately 1 3 /year can be tolerated. This is equivalent to around 2 tons of water saturated sand. For rocess vessels with fully or artially autoated systes for sand reoval, the sand volue handling caacity ay be deterined fro the accetable volue of sand build-u and frequency of sand reoval. Intervals between sand reovals tyically range fro a few days to several weeks, deending on actual sand roduction. During the sand reoval sequence (flushing) the quality of roduced water ay be affected in a negative way. In service, the interval between sand reovals should be otiised based on exerience and ay vary over the field life. A artially or fully autoated syste for sand reoval significantly increases the sand volue handling caacity. Fro oerational exerience sand loads in the order of /yr has safely been handled for a single latfor (searator). Recoended ractice DNVGL-RP-O501 Edition August 2015 Page 12

13 Where ractically and econoically viable to do so, installation of wellhead or in-line de-sanders ay significantly reduce sand loading for inlet searators, thus reducing negative consequences associated with sand reoval and rocess usets. Figure 2-1 Tyical sand deosition in horizontal rocess vessel with conventional sand reoval syste 2.7 Risk assessent As basis for the sand anageent strategy a risk assessent should be erfored to identify both threats and oortunities associated with sandy service oeration. The urose of the risk assessent is to ensure that sufficient, effective and anageable safeguards are in lace in order to be reared for, detect and handle sand roduction. In the early concet and design hase the risk assessent ay be liited to a high level assessent considering sand and erosion otential and the requireents to sand handling caacity. The iortance of an early outline of a sand anageent strategy relates to decision of: need for sand control sizing of flowlines, iework and coonents instruentation for onitoring sand roduction and erosion systes for reoval, cleaning and disosal of sand fro rocess systes. For the oerations hase the risk assessent should be erfored by a dedicated sand anageent tea involving reresentatives fro relevant discilines in the oerating organisation. It should be ehasized that a successful ileentation of a sand anageent strategy requires a cross-disciline aroach Class of erosive service Ranking of the erosion otential for iework ay be erfored with reference to calculated bulk flow velocities, considering that the flow velocity deterines the order of agnitude for erosion. The class of erosive service for iework rovides a sile indication of whether a syste is suscetible to erosion. In reality, the exact size and geoetrical shae of the iework coonents in cobination with sand article size and fluid roerties such as density and viscosity will influence the actual erosion otential er aount of sand. Bulk flow velocities ay be available in the roduction control syste or be calculated based on a silified black oil odel ([4.4.2]) with reference to allocated or easured flow rates, oerating ressures and teeratures. The erosion classes defined in Table 2-1 ay be alied to identify whether a given oerating condition for a given iework syste ay be suscetible to erosion. The relative erosion otential referring to erosion class (1) deonstrates the order of agnitude increase in erosion otential as function of increase in bulk flow velocity. Recoended ractice DNVGL-RP-O501 Edition August 2015 Page 13

14 When increasing the bulk flow velocity fro erosion class (1) to erosion class (3) the exected erosion for a given sand load (kg) increases by a factor of 100. Siilarly if the velocity is increased to erosion class (6) the erosion otential increases by a factor of aroxiately In other words the aount of sand required to cause siilar erosion daage in a iework syste oerated in erosion class (1) is in the order of 5000 ties higher than for a syste oerated in erosion class (6). Table 2-1 Class of erosive service Erosion Class Piework Bulk Flow Velocity V (/s) Definition Relative erosion otential 1) Descrition Extreely high erosion otential 5000 Syste needs to be oerated close to sand free. Safeguards to onitor erosion should be in lace and closely onitored Very high erosion otential 1500 Tolerable sand roduction liited by risk of erosion High erosion otential 500 Tolerable sand roduction will in ost cases be dictated by erosion rather than sand handling caacity Mediu erosion otential 100 Tolerable sand roduction ay be liited both by erosion and sand handling caacity Low erosion otential Extreely low erosion otential 1) Relative erosion otential is given for the average velocity in each velocity interval 1 A large aount of sand is required to cause erosion. The accetable sand load will in ost cases be liited by the sand handling caacity in the rocess syste Effects of lain erosion, i.e. not considering any cobined effects of flow accelerated corrosion, can norally be neglected for realistic sand loads Risk assessent and ranking The risk assessent and -ranking should be erfored based on a breakdown of the syste into subsystes that is considered ractical with reference to oerator s organisation and resonsibilities of different discilines. A tyical syste breakdown is suggested below as a starting oint: lower well coletion, considering sand screens, tubing or other inflow control coonents down hole safety valve down hole artificial lift systes (gas lift, ESP) XT and valves iework and coonents between XT and anifold roduction choke instruentation and etering systes anifolds flowlines and risers boosting systes for unrocessed flow iework between anifold and first rocess vessel rocess vessels (searators) and internals roduced water syste fro searators, iework and coonents roduced oil syste fro rocess vessels, iework and coonents blow down systes and equalisation anifolds water injection systes, coonents Recoended ractice DNVGL-RP-O501 Edition August 2015 Page 14

15 safety critical instruentation. For each of the sub-systes, the risk associated with sand roduction should be assessed on a qualitative basis with reference to the risk categories suggested in the table below. Table 2-2 Risk ranking Risk High Mediu Low Definition Risk associated with sand roduction in conflict with accetance criteria or non-coliance with standards or regulations. Syste cannot be oerated without odifications to syste or oerating rocedures, or with additional safeguards Risk accetable. Additional onitoring or safeguards shall be evaluated by oerator according to ALARP rincile. Low risk with current oerational rocedures and safeguards The risk assessent should address: critical syste functions consequence of each failure ode related to sand accetance criteria related to containent, function, availability and erforance safeguards risk reducing easures risk level with and without active safeguards. The risk associated with sand roduction should be ade visible both with and without safeguards to ehasize the iortance of following u both active and assive safeguards. 2.8 Safeguards Possible safeguards to control sand roduction and anage its consequences within accetable liits are listed below and further elaborated in A.A: echanical sand control draw-down control otiisation of artificial lift echanis eriodic testing of safety critical valves continuous onitoring and control of sand roduction sot-check saling of sand roduction (e.g. routing of well to test searator, sand tras) erosion odelling continuous onitoring of erosion (e.g. erosion robes) onitoring by calculations; control of flow velocities onitoring choke condition and oeration onitoring sand-build-u in rocess vessels wellhead or in-line de-sanders reoval of sand fro rocess vessels onitoring of flow velocities in rocessed liquid streas igging (echanical or hydraulic) of infield flowlines online NDT insection. It should be acknowledged that soe of the safeguards are inherent to syste design (assive) and that others require continuous follow u (active). The relevance, feasibility and effectiveness of the individual safeguard needs to be evaluated on a case to case basis as art of the risk assessent. Recoended ractice DNVGL-RP-O501 Edition August 2015 Page 15

16 2.9 Strategy ileentation The sand anageent strategy should include a sand anageent anual that rovides a descrition of activities and resonsibilities related to follow u of critical safeguards identified in the risk assessent. Definition of the erson/grou resonsible for each activity should be as sile and secific as ossible to avoid diffusion of resonsibility. The sand anageent anual should serve as a ractical guideline in daily oeration. Sufficient guidance should be included to ensure that activities are correctly and effectively executed, reorted and counicated. Where racticable to do so, reference shall be ade to relevant oerating rocedures Training requireents Requireents to ersonnel training should be decided with reference to the activities and resonsibilities dedicated to secific ersonnel as described in the sand anageent anual. A general descrition of the secific field sand otential and strategy for anaging sand roduction should be included as art of this training Status reorting and eriodic revision Status reorting A eriodic sand anageent status reort should be established tyically on a 12 onth cycle. The overall objective of the status reort is to identify, counicate and execute necessary actions to adjust the strategy or its alication: confir that the objective of the sand anageent strategy is et for the reorting eriod rovide inut to insection lanning ensure that odification to the roduction syste or oerating conditions relevant for the next eriod are identified and ileented in the sand anageent strategy. rovide basis for roduction otiisation considering any liitations iosed by sand roduction cature any incidents/failures related to sand roduction over the last eriod of oeration Revision of strategy The sand anageent strategy should be subject to eriodic audit/review, e.g. on a 12 onth cycle. Related to the revious roduction eriod the following questions (check list) should be addressed and relevant actions identified if answered confirative: erosion in iing systes or coonents has resulted in loss of containent (external leakage) erosion has resulted in excessive consution of erosion allowance identified fro insection erosion has led to unlanned relaceent of coonents observed unaccetable rocess usets due to (jetting), cleaning or disosal of sand observed significant and frequent non-coliance with roduced water quality observed non-coliance with accetable oil in disosed sand systes for sand onitoring not calibrated according to lan significant roduction otential is restricted due to sand roduction other identified issues not covered by the above. Related to the next 12 onth eriod the following questions (check list) should be addressed and relevant actions identified if answered confirative: develoent of and tie-in of new roduction facilities odifications to existing roduction facilities; iing, valves, instruentation and chokes will be ileented for next eriod rocess conditions will be significantly changed for next eriod (Pressure, WC, GOR, ) identified down hole sand control failures or significant increase in sand roduction Recoended ractice DNVGL-RP-O501 Edition August 2015 Page 16

17 odification of accetance criteria, e.g. de-rating of systes increasing erosion allowance or iroved sand handling caacity odifications to established rocedures for sand onitoring. The sand anageent strategy should be udated based on the outut fro the audit/review as found required. Recoended ractice DNVGL-RP-O501 Edition August 2015 Page 17

18 SECTION 3 FUNDAMENTALS OF PARTICLE EROSION 3.1 General The selection of aterials and diensioning of ies are erfored in order to obtain necessary strength, caacity and service life to coe with the roduction conditions. Material degradation due to corrosion, erosion and/or erosion-corrosion, ay gradually affect the integrity of the iing syste. Material degradation will generally deend on the roduction characteristics for the syste; i.e. roduction rates, ressure and teerature, and the resence of corrosive coonents and erosive solid articles. The degradation ay also be strongly deendant on the ie aterial. Material degradation can in ost cases not be fully avoided, but the aterial ay be allowed to degrade to a certain extent and in a controlled anner. By roer diensioning, selection of suitable aterials, use of inhibitors or other corrosion/erosion reducing easures and/or by alication of corrosion/erosion allowance, a syste which fulfils the requireents can generally be achieved. Selection of such easures ay, however, be associated with high cost. The current section describes the fundaental theory for lain article iact erosion, roviding the basis for the eirical erosion odels in Sec.4 and detailed CFD erosion siulations described in A.C of this docuent List of sybols Sybol Descrition Unit A diensionless araeter grou [-] A t area exosed to erosion [ 2 ] A ie cross sectional area of ie [ 2 ] A ratio area ratio between cross sectional area in reducer [-] b function of Re [-] C 1 odel/geoetry factor [-] C 2 article size correction factor [-] C unit unit conversion factor (/s ~ /year) [-] c function of Re [-] D inner ie diaeter [] d article diaeter [] d,c critical article diaeter [] E actual surface thickness loss [] E actual aterial loss rate [kg/s] E, relative aterial loss rate [kg/kg] E L actual surface thickness loss rate [/s] E L, relative surface thickness loss [/kg] E L,y annual surface thickness loss [/year] E L,easured easured surface thickness loss [/year] F(α) function characterising ductility of aterial [-] G corrections function for article diaeter [-] h height of weld reinforceent [] K aterial erosion constant [(/s)-n] k g aterial constant ass flow of gas in ie [kg/s] l ass flow of liquid in ie [kg/s] M ass of sand [kg] ass rate of sand [kg/s] Recoended ractice DNVGL-RP-O501 Edition August 2015 Page 18

19 Sybol Descrition Unit total ass rate of fluids [kg/s] R radius of curvature given as Nuber of Internal Pie Diaeters. Reference of radius of [-] curvature is centreline of ie Re Reynolds nuber [-] U article iact velocity (equal to ixture fluid velocity) [/s] V 1, V 2 fluid velocity in cross section 1 and 2 of reducer [/s] V s g suerficial velocity of gas hase in iing [/s] V s l suerficial velocity of liquid hase in iing [/s] V ixture fluid velocity in iing [/s] α article iact angle [rad] β = ρ /ρ density ratio between article and fluid [-] γ = d /D ratio of article diaeter to geoetrical diaeter [-] μ g viscosity of gas hase [kg/s] μ l viscosity of liquid hase [kg/s] μ viscosity of fluid ixture [kg/s] n velocity exonent [-] ρ g density of gas hase [kg/ 3 ] ρ l density of liquid hase [kg/ 3 ] ρ density of fluid ixture [kg/ 3 ] ρ density of article [kg/ 3 ] ρ t density of target aterial [kg/ 3 ] Indexes 1, 2 - cross section 1 and 2. c - critical g - gas i - index L - length l - liquid - ixture - article s - suerficial t - target aterial 0 - standard conditions Recoended ractice DNVGL-RP-O501 Edition August 2015 Page 19

20 3.1.3 Erosive agents Erosion odels given in the current docuent are derived and validated riarily for quartz sand; however the odels ay also be alied to rovide estiates of erosion otential for other erosive agents. The ost coon erosive agents causing article erosion in oil and gas systes for which the erosion odels given in the current docuent ay be alied are described below. For other erosive agents than quartz sand, guidance is given in Sec.5. Sand Sand roduced fro oil & gas forations ay vary in ters of size, quartz content, shae and sharness. All these factors affect how erosive the sand will be to an exosed surface. In ters of size, sand is defined as articles in the range µ [W.C. Krubein & L.L. Sloss]. Particle erosion is governed by the sand quartz fraction. On the Mohs hardness scale quartz has a hardness of 7, aking it erosive to a wide range of construction aterials. Particles of size less than 62.5 µ are classified as fines. Erosion odels described in the current docuent are liited to article sizes above 20 µ. Fines are norally less erosive than sand both due to article size and quartz content. Particle sizes liited to fines can be achieved with various ethods for down-hole sand control. Conservatively the erosive character of fines can be assued as for sand. Particles larger than 2000 µ are classified as gravel. Barite / Calcite Proants Barite and Calcite used for weighting of drilling, coletion and kill fluids. On the Mohs hardness scale these aterials have hardness in the order of 3 aking it significantly less erosive to steel coared to quartz articles. Characteristic article size when used as weight aterial is in the order of 20 µ. Mixed in high density liquid fluids the sall article sizes will also suress erosion. For noral circulation rates the erosion otential is exected to be low, however under certain conditions a cobination of high velocities and large aounts of weight aterial ay cause erosion. Proants are custoised article distributions either based on treated sand or ade artificially. Proants are used e.g. for hydraulic fracking oerations or for acking of sand screens. In soe cases roants ay be roduced back to the roduction facilities and cause siilar concerns as sand roduced fro the foration. For steel, roants ay be equally erosive as quartz sand. Fro erosion testing variations are observed between intact and crushed roants Non-erosive agents Non-erosive solids are defined as solids having a sufficiently low hardness or size not to cause erosion to a aterial either as a result of surface ductile deforation or fatigue. Particles with hardness on Mohs scale less than 3 can norally be considered non-erosive to steel, see table below. Sodiu Chloride (salt) Sodiu Chloride (2.5 on the Mohs scale) used as weight aterial for brine can be considered non-erosive to steel for bulk flow velocities less than 100 /s. Clay/Silt Clay articles (2-2.5 on the Mohs scale) can be considered non-erosive to steel for bulk flow velocities less than 100 /s Recoended ractice DNVGL-RP-O501 Edition August 2015 Page 20

21 3.2 Characterisation of erosive wear Erosion resonse odel The ters erosive wear and erosion are in the resent docuent defined as aterial loss resulting fro iact of solids/sand articles on the aterial surface, ref. Figure 3-1. Figure 3-1 Notation - article iact velocity and angle Erosive wear can be estiated fro the following exression /8/, /13/, /14/, rovided iact velocities and angles are known for the articles iacting on the target surface, ref. Figure 3-1: n E, = K U F( α) E = K U F(α ) n Relative aterial loss [kg/kg] (3.1) Actual aterial loss rate [kg/s] (3.2) The function F(α) characterises the ductility of the target aterial. Steel grades are generally regarded as ductile, while cerets like tungsten carbide with a etallic binder hase are defined as brittle. Coatings ay be ductile or brittle deending on cheical coosition and deosition ethod. Metallic or hard coatings are alied by theral sraying, overlay or galvanic lating ethods. Soft coatings are defined as olyeric/eoxy coatings and are ductile in nature. Ceraics are generally characterised as brittle aterials. The article iact velocity (U ) is linked to the flow velocity either through secific correlations accounting for sli velocity or calculated directly by a article drag odel. The aterial coefficients (K) and (n) are derived fro testing for a given cobination of aterial and erosive agent. Coefficients for selected ductile and brittle aterials are given in Table 3-1. Angle deendency for ductile and brittle aterials are given in the exressions below and illustrated grahically in Figure 3-2. Ductile aterials exerience axiu erosion for iact angles in the range 20 to 50. Brittle aterials exerience axiu erosion at 90 (π/2) iact angle and erosion is gradually reduced for saller angles. The linear F(α) for brittle aterials shall be considered an aroxiation, however sufficient for ost ractical alications. Ductile aterials: F ( α ) ductile 2 k [ Sin ( α ) + B( Sin ( α ) Sin ( α ))] [ 1 ex( α )] = A C F ( α ) [ 0,1] for α 0, π 2 (3.3) A B C k Recoended ractice DNVGL-RP-O501 Edition August 2015 Page 21

22 Brittle aterials: ( α ) brittle α = 2 π F ; ( α ) [ 0,1] F for α 0, π 2 (3.4) Figure 3-2 Characteristic article iact angle deendency for erosion of ductile (steel) and brittle aterials Erosion otential for a given article iact characteristic is directly roortional to the aount of articles exosed to the surface, indeendent of tie. This is valid as long as the article concentration in the fluid is sufficiently low (< 500 V) 1) to avoid any significant article-article interaction. 1) For erosion testing the sand concentration should conservatively be ket below 100 V (100 c3 of sand er 1 3 of fluid). Within the tyical oerating teerature for oil and gas facilities (<200 C), variation in lain article erosion as function of teerature can norally be neglected. For velocities lower than 100 /s, the difference in erosion resistance for relevant steel grades is generally within ±25%. This also alies to Ni-based alloys tyically used in iing systes. Recoended values for the ost coonly used aterials and an erosive agent equivalent to quartz sand are given in Table 3-1. The araeters are deterined according to test rocedure given in A.B. Correction factors for other relevant erosive agents than quartz sand are given in Table 5-2. Recoended ractice DNVGL-RP-O501 Edition August 2015 Page 22

23 Table 3-1 Material roerties Erosive agent: Quartz sand/sei round/angular shae Material ρ t (kg/ 3 ) K (/s) -n n Angle deendency Steel grades Carbon steel 7800 Dulex 7850 SS E Ductile Inconel 8440 Alternative aterials GRP/Eoxy E GRP/Vinyl Ester E HDPE E Ductile Aluiniu E Brittle aterials DC-05: Tungsten Carbide E CS-10: Tungsten Carbide E CR-37: Tungsten Carbide E % Al2O3: Aluiniu Oxide E % Al2O3: Aluiniu Oxide E PSZ Ceraic: Zirconia E ZrO2-Y3 Ceraic: Zirconia E Brittle SiC: Silicon Carbide E Si3N4: Silicon Nitride E TiB2: Titaniu Diboride E B4C: Boron Carbide E SiSiC: Ceraic Silicon Carbide E Recoended ractice DNVGL-RP-O501 Edition August 2015 Page 23

24 SECTION 4 EMPIRICAL MODELS FOR SAND PARTICLE EROSION 4.1 General The following sections rovide eirical odels for estiation of erosive wear in tyical steel iing coonents. The odels and recoendations have been worked out based on exeriental investigations available in literature, dedicated erosion tests and exerience and odels available within DNV GL. The ajority of exeriental validation data have been obtained at low/oderate ressures in saller diaeter test facilities. Extraolation to higher ressure conditions and larger diaeter ies has been erfored based on detailed CFD erosion siulations and also by coarison with field exerience. A selection of controlled validation cases is included in A.E. 4.2 Alication and liitations The odels described in the following sections only address lain erosion, i.e. do not account for otential cobined effects of corrosion and article erosion, flow accelerated corrosion (FAC) or other echaniss such as drolet erosion or cavitation. Additional wear due to these effects needs to be considered searately. Alication of the erosion odels should be liited to quartz sand as the erosive agent, unless otherwise secified. Guidance on how the odels ay be used for other erosive agents is given in Sec.5 of this docuent. The erosion odels are develoed for what is considered realistic article concentrations, i.e. concentrations sufficiently low to assue arginal article-article interaction. Hence, the odels ay be conservative for article concentrations above tyically 500 V. All odels are based on ixture fluid roerties. For single hase fluids (liquid or gas) the single hase roerties shall be alied. For ultihase flow the ixture roerties shall be established based on the suerficial velocities and single hase roerties according to recoendations given in the current docuent. The odels are based on the assution of a iniu straight ustrea ie section corresonding to 10 ie diaeters. For colex iework, an aroriate geoetry correction factor according to [4.3] shall be alied. Alication of the odels should be liited to the range secified for inut araeters in Table 4-1. Table 4-1 Liitations to odel araeters Paraeter Unit Lower bound Uer bound Particle diaeter Particle ass density kg/ Pie inner diaeter Radius of bend (Nuber of ie inner diaeters) Pie aterial ass density kg/ Suerficial liquid velocity /s 0 50 Suerficial gas velocity /s Liquid density kg/ Gas density kg/ Liquid viscosity kg/s 1.0E E-02 Gas viscosity kg/s 1.0E E-04 Particle concentration V Recoended ractice DNVGL-RP-O501 Edition August 2015 Page 24

25 4.3 Geoetry correction factors The eirical erosion odels for article iact erosion given in the current docuent are based on the assution of a straight ustrea ie section >10 D. When the distance of straight iing between coonents is less than 10 D, a geoetry correction factor (GF) should be ultilied with the odel rediction. Recoended geoetry factors established fro a series of CFD erosion siulations are given in the table below. The geoetry factors are given as guidance only. For geoetrically colex iework and coonents the current eirical erosion odels ay be used to rovide an initial estiate of the erosion otential, however the need for erforing a ore detailed CFD erosion analysis (ref. A.C) should be considered on a case to case basis. Table 4-2 Geoetry factors Exale: For two ie bends in the sae lane, saced less than 10 diaeters (10 x D), a geoetry factor of GF=2 should be alied. I.e. if the erosion rate calculated with the bend odel given in the current docuent is (E L1 ), the exected erosion rate in the second elbow (E L2 ) shall be estiated as: E L2 =GF E L1 =2 E L1. Recoended ractice DNVGL-RP-O501 Edition August 2015 Page 25

26 4.4 Model inut araeters Erosion resonse odel Based on the fundaental erosion resonse odel linking aterial loss fro a surface to article iact characteristics, the aterial loss and wall thickness reduction can be calculated according to the following exressions: n E, = K U F( α) E E E = E, L, t E, = ρ A L = EL, t Relative aterial loss [kg /kg] (4.1) Actual aterial loss rate [kg/s] (4.2) Relative surface thickness loss [/kg] (4.3) Actual surface thickness loss rate [/s] (4.4) E =, E L M Actual surface thickness loss [] (4.5) Material araeters are listed in Table 3-1. For the eirical odels described in the following sections the characteristic iact angles, iact velocities and target area are secified as function of coonent geoetry, flow condition and article roerties. In the calculation rocedure, these effects are accounted for by eirical odel/geoetry factors. The odel/geoetry factors account for ossible ultile iacts of single articles, concentration of sand articles due to coonent geoetry and odel uncertainty Bulk roerties Particle iact velocity (U ) shall -if not otherwise secified- be deterined by the following rocedure: U = V = V + V V V s l s g s l 4 l = ρ π D l g 2 4 g = 2 ρ π D s g Characteristic article iact velocity [/s] (4.6) Suerficial liquid velocity [/s] (4.7) Suerficial gas velocity [/s] (4.8) Physical roerties of the fluid are described as ixture roerties and are deterined by the following exressions: ρ μ ρ V + ρ V = V + V l l s g g s l s g s μ V + μ V = V + V μ υ = ρ l l s g g s l s g s Mixture density [kg/ 3 ] (4.9) Dynaic ixture viscosity [kg/s] (4.10) Kineatic ixture viscosity [ 2 /s] (4.11) Recoended ractice DNVGL-RP-O501 Edition August 2015 Page 26

27 Unless other PVT odels are available to establish the ixture flow roerties, these ay be derived fro the following black oil forulation with reference to the following inut araeters: Table 4-3 Black Oil odel araeters Constants Sybol Value Unit Pressure at standard condition P bara Teerature at standard condition T K Universal gas constant R 8314 J/kgK Fluid roerties Oil density at standard condition ρ o Inut (default 800) kg/ 3 Water density at standard condition ρ w Inut (default 1000) kg/ 3 Gas Molecular Weight MW Inut (default 20) kg/kol Gas coressibility factor Z Inut (default 0.9) - Conditions Pressure P Inut bara Teerature T Inut K Water cut WC Inut S 3 /S 3 Gas Oil Ratio GOR Inut S 3 /S 3 Oil rate at standard condition Q o Inut S 3 /d Water rate at standard condition Q w Alternative inut S 3 /d Gas rate at standard condition Q g Alternative inut S 3 /d Pie cross section diaeter D Inut Definition of Water Cut (WC) and Gas Oil Ratio (GOR): Qw WC = Qo + Q Q GOR = Q g o w Calculation of ixture velocity: Q A ie = Q o Q V = A WC P o T Z GOR 1 WC P T π 2 = 4 D ie Water Cut [S 3 /S 3 ] (4.12) Gas Oil Ratio [S 3 /S 3 ] (4.13) Actual flow rate [ 3 /s] (4.14) Pie cross section area [ 2 ] (4.15) Mixture velocity [/s] (4.16) Recoended ractice DNVGL-RP-O501 Edition August 2015 Page 27

28 Calculation of ixture density: ρ = Q WC P0 MW 5 ρo + ρw + GOR 10 1 WC R T 0 = WC P0 T Z 1+ + GOR 1 WC P T 0 Calculation of ixture viscosity: WC P0 T Z μo + μw + GOR μ g 1 WC P T0 μ = WC P0 T Z 1+ + GOR 1 WC P T Sand content Mixture density [kg/ 3 ] (4.17) Mixture viscosity [kg/s] (4.18) If the article concentration is given as 'art er illion values' (), the resulting sand flow rate can be calculated according to the following relations. For article concentrations given on ass basis (W): = W 10 6 Mass rate of articles [kg/s] (4.19) For given on volue basis (V): = ρ V 10 6 ρ Mass rate of articles [kg/s] (4.20) Note that V will change with ressure and teerature. Care ust therefore be taken to relate the V to the secific conditions. The article loading ay alternatively be secified as gras er second (g/s) or ton er year (ton/yr). Considering that article erosion is directly roortional to sand loading, the referred aroach is to estiate the erosion rate (E L, [/kg]) and ultily this with a secified sand load to obtain actual erosion. Generally, sand content in the range 1-50 W is exerienced in well streas ustrea of the first stage searators. Tyical sand article sizes are exerienced to be in the range µ if no sand exclusion techniques are alied. When sand exclusion systes are alied, tyical article sizes range fro µ. 4.5 Sooth and straight ies Sand erosion in sooth and straight ies is generally low and in ost cases not the liiting factor with resect to erosion risk for a iing syste. The ain reason for the low erosion otential is related to generally low article iact angles. Figure 4-1 Sooth and straight ies Recoended ractice DNVGL-RP-O501 Edition August 2015 Page 28

29 Erosion rate for sooth straight steel ies under turbulent flow conditions can be estiated fro the following eirical exression: E L, = U D E, = L y U D Relative surface thickness loss [/ton] (4.21) Annual surface thickness loss [/year] (4.22) The odel is established for vertical ies but ay conservatively also be used for horizontal ies on the condition that fluid velocities are sufficient to diserse the sand in the bulk fluid 1). It should be noted that the eirical correlation is indeendent of the fluid density, viscosity and article size. 1) For fluid velocities where any significant erosion ay occur, articles are likely to be disersed in the bulk flow 4.6 Welded joint Particle erosion in welded joints with internal reinforceent is based on the initial weld geoetry. The geoetrical change of the weld reinforceent when eroded is not taken into account. Figure 4-2 Erosion on flow facing (left) and downstrea side of weld joint Erosion of the flow facing side of welded joints (ref. Figure 4-2) is based on the following 5 ste calculation rocedure: 1) Estiate the article iact angle, α, between the weld and the flow direction. If the angle is unknown, the conservative value α = 60 should be used. 2) Establish the value of the function F(α) by using the iact angle found in ste 1 and Figure ) Calculate the corrected ie cross sectional area: 2 π D Aie A t = = 4 sin( α) sin( α) Area exosed to erosion [ 2 ] (4.23) Unit conversion factor to convert erosion rate fro /s /year: C unit = = Conversion factor (4.24) 4) Calculate the article size and fluid density correction factor C2: 6 10 d 30 ( ρ ) 6 10 d 30 ( ρ ) 1/ 2 1/ 2 < 1, C ( ρ ) d 2 = 1/ 2, C 2 = 1 Particle correction factor [ - ] (4.25) Recoended ractice DNVGL-RP-O501 Edition August 2015 Page 29

30 5) The axiu erosion rate (/year) of the weld is then found fro the following exression: E α n L, y = K F( ) U C2 ρt Aie K=2.0E-09, n=2.6 sin( α) C unit Flow facing art of weld Annual surface thickness loss [/year] (4.26) It should be noted that erosion on the flow facing art of the weld will result in rounding and soothing of the weld, generally not affecting the integrity of the ie. Erosion of the weld will therefore norally not be a liiting factor with resect to diensioning or oeration of the ie. Maxiu erosion in the steel ie downstrea a weld is found to be larger than for the sooth art of the ie. This is due to the turbulence caused by vortex shedding on the downstrea side of the weld. With reference to the height of the weld h [] the erosion rate can be estiated by using the following eirical exression: E L, y = ( h) U D Downstrea side of weld (4.27) Annual surface thickness loss [/year] 4.7 Pie bends Particle erosion in ie bends can be estiated with the following rocedure: 1) Calculate the characteristic iact angle, Note: Radius of curvature (R) is given as the Nuber of Pie Diaeters: α = arctan( 1 ) 2 R Characteristic iact angle [rad] (4.28) 2) Calculate the diensionless araeter grous A and β: 2 ρ tan( α) U D ReD tanα A = = ρ μ β ρ β = ρ Diensionless grou [ - ] (4.29) 3) Use the diensionless grou, A, fro ste 2 in the following equation in order to obtain the relative critical article diaeter: d, c D ρ = γ c = ρ = [ 1.88 ln( A) 6.04] β [ 1.88 ln( A) 6.04] 0.1, 1, γ < 0.1 γ > 0.1 γ 0 c c c (4.30) d γ = D, where d () is the average article size 4) Calculate the article size correction function G by using the critical article diaeter found in ste 3: γ γ c G = 1, γ < γ, γ γ c c Particle size correction function [ - ] (4.31) Recoended ractice DNVGL-RP-O501 Edition August 2015 Page 30

31 5) Calculate the characteristic ie bend area exosed to erosion: 2 π D Aie A t = = 4 sin( α) sin( α) Particle iact area [ 2 ] (4.32) 6) Deterine the value of the function F(α) by using the angle, α, found in ste 1. The value of F(α) is in the range [0, 1], ref. 0. 7) The odel geoetry factor (C 1 ) is set equal to: C 1 = 2.5 8) The odel/geoetry factor accounts for ultile iact of the sand articles, concentration of articles at the outer art of the bend and odel uncertainty. 9) The following unit conversion factor ust be used (/s to /year): C unit = = Conversion factor [ - ] (4.33) 10) Maxiu erosion in the ie bend is found by the following exressions: E E E K F ( α ) U n 6 L, = G C1 GF 10 ρt At K F( α ) U n L, y = G C1 ρt At K F ( α ) U ρ A GF n = t t C G C1 GF M 10 3 unit Relative surface thickness loss [/ton] (4.34) Annual surface thickness loss [/year] (4.35) Actual surface thickness loss [] (4.36) The geoetry factor (GF) shall be selected according to [4.3]. If no inforation is available on the colexity of the iing isoetric a geoetry factor of GF=2 should be used. Calculation exale: A 4 steel (D = 0.1 ) iework syste is oerated with a suerficial liquid velocity of 5 /s and a suerficial gas velocity of 10 /s. The roerties of the liquid hase is characterised by ρ l =800kg/ 3 and μ l = 1.0E-03 kg/s and the gas hase is characterised by ρ g =100kg/ 3 and μ g = 1.5E-05 kg/s.the iework consists of elbows with radius of curvature R = 1.5 X D saced ore than 10 x D. Particles are characterised by sei angular quartz sand with a d 50 article size of 250 μ. The annual sand load for the syste is estiated to 0.1 ton. For the iework elbows the erosion rate is fro the elbow odel calculated to /ton. I.e. ore than 7 ton of sand is required to cause 0.1 erosion at this oerating condition. For the exected annual sand load of 0.1 ton, this corresonds to an estiated wall thickness reduction of /year. 4.8 Blinded tee Particle erosion in blinded Tee can be estiated with the following rocedure: 1) Calculate the following non-diensional araeters: d γ = D ρ β = ρ Re D V D = υ Ratio of article size to ie diaeter [ - ] Ratio of article to fluid density [ - ] (4.37) (4.38) Reynolds nuber [ - ] (4.39) Recoended ractice DNVGL-RP-O501 Edition August 2015 Page 31

32 For β < 40 d, c γ c = D 0.14 = β Noralised critical article diaeter [ - ] (4.40) 19 ln(re c = 0, 3 For b 40 D, ) C 1 = β 0. 3 γ < γ γ γ c c (4.41) Model factor [ - ] (4.42) Re D b = ln γ c β = b , ln(red ) c = ( ( C 1 = 10. β ) ), γ < γ γ γ c c Model factor [ - ] (4.43) 2) Calculate the article size correction factor: γ G = γ c c Particle size correction factor [ - ] (4.44) 3) Calculate the characteristic area exosed to erosion: A 2 π t = D 4 Characteristic article iact area [ 2 ] (4.45) 4) The following unit conversion factor ust be used (/s to /year): C unit = = (4.46) 5) Maxiu erosion in the blinded tee is found by the following exressions: E E K U n 6 L, = G C1 GF 10 ρt At K U n L, y = G C1 ρt At GF C n K U 3 E = G C1 GF M 10 ρ A t t unit Relative surface thickness loss [/ton] (4.47) Annual surface thickness loss [/year] (4.48) Actual surface thickness loss [] (4.49) Geoetry factor (GF) shall be selected according to [4.3]. If no inforation is available on the colexity of the iing isoetric a geoetry factor of GF = 2 should be used. Recoended ractice DNVGL-RP-O501 Edition August 2015 Page 32

33 4.9 Reducers The taered section of reducers are exosed to erosion due to change of flow direction cobined with flow acceleration. The figure below indicates location of erosion (red) and notation for the odel araeters. The odel is considered valid for reducer angles in the range [10, 80 ]. Figure 4-3 Scheatic, flow reducer Particle erosion in reducers ay be estiated according to the following 7 ste calculation rocedure: 1) Establish angle deendency F(α) fro secified correlations (ductile) and a as defined in Figure ) Calculate the area exosed to article iact (directly hit by the articles): A t π 2 = ( D 1 D 4 sin α 2 2 ) Characteristic article iact area [ 2 ] (4.50) 3) Calculate the ratio between area exosed to article iact and the area before the contraction: A ratio = 2 1 D2 D1 Area asect ratio [ - ] (4.51) 4) The article iact velocity is set equal to the fluid velocity after the contraction: U = V 1 2, 2 = V,1 ) 2 ( D D Characteristic article velocity [/s] (4.52) 5) Calculate the article size and fluid density correction factor C 2 : ( ρ ) d ( ρ ) d 1/ 2 1/ 2 < 1, C 1, C 2 = d 30 ( ρ ) 2 = 1/ 2 Particle size correction factor [ - ] (4.53) 6) Conversion factor (/s to /year): C unit = = Unit conversion factor [ - ] (4.54) 7) Maxiu erosion in the contraction is then found by the following exressions: E E K F( α) U n 6 L, = Aratio C2 GF 10 ρt At K F( α) U n L, y = Aratio C2 ρt At GF C unit Relative surface thickness loss [/ton] (4.55) Annual surface thickness loss [/year] (4.56) Recoended ractice DNVGL-RP-O501 Edition August 2015 Page 33

34 n K F( α) U 3 E = Aratio C2 GF M 10 ρ A t t Geoetry factor (GF) shall be selected according to [4-3]. If no inforation is available on the colexity of the iing isoetric a geoetry factor of GF=2 should be used Intrusive erosion robes Actual surface thickness loss [] (4.57) Intrusive erosion robes are extensively alied both subsea and toside for continuous onitoring and control of iework wear /19/. The robe erosion eleents are norally ade of aterials with erosive roerties siilar to the iework. The odel should be liited to robe surface angles between 10 and 90. Tyical robe angels are 45 ± 15. It is ilicitly assued that the articles are hoogenously distributed over the ie cross section. It should be noted that deending on the location and orientation of the robe and the effects of ustrea iework this ay not always be the case. Figure 4-4 Intrusive erosion robe Probe erosion is estiated by the following stes: 1) Calculate reresentative article iact area: A t π = D Sin( α) Equivalent article iact area for hoogenously distributed articles [ 2 ] (4.58) 2) Calculate F(a) fro correlation curve, or conservatively set F(a)=1, ref. Figure 3-2 3) Calculate article correction factor C2: 6 10 d For < 1, 1/ 2 30 ( ) ρ 6 10 d For 1, 1/ 2 30 ( ) ρ C 6 10 d 30 ( ρ ) 2 = 1/ 2 C 2 = 1 (4.59) 4) Conversion factor (/s to /year): C unit = = Unit conversion factor [ - ] (4.60) Recoended ractice DNVGL-RP-O501 Edition August 2015 Page 34

35 5) Calculate the robe erosion rate: n K U ( ) F α E L, = C 2 10 At ρt d correction E L, y / kg n K U F( α) = A ρ t t / kg C 2 d correction 6 C unit Relative surface thickness loss [/ton] (4.61) Annual surface thickness loss [/year] (4.62) In any cases it is of interest to use the real-tie easured erosion rate fro the robe to assess the real-tie aount of solids roduced. I.e. the real tie sand roduction can be deterined fro the easured erosion rate E L, easured (/ year) and the equation above: E = E L,easured L, M C unit ; C unit = Mass rate of sand [kg/s] (4.63) It should be noted that this aroach ay involve uncertainty, articularly at low bulk flow velocities. Orientation of the ie in which the erosion robe is installed also needs to be considered with resect to distribution of sand over ie cross section. Due to this uncertainty the aroach should not be used when bulk flow velocity (V ) is less than 5 /s Flexible ies with interlock carcass Flexible ies norally consist of a ultilayer coosite structure with an internal interlocked steel carcass to revent ie collase and to rotect the olyer ressure sheet fro echanical or abrasive daage. With reference to API 17J (secification for unbonded flexible ie), the anufacturer shall deonstrate with tests -or analytical data based on tests- that the carcass has sufficient erosion resistance to eet the design requireents for the secified service life and service conditions 1). 1) For carcass erosion test reference is given to API 17B, section Tolerable erosion to the interlock carcass should be liited with reference to risk of carcass collase, unlocking of the carcass (tensile load) and otential for direct exosure or collase of the olyer containent barrier following a carcass failure. Based on industry best ractice the tolerable erosion to the interlock carcass should be liited to axiu 10 to 30% of the carcass steel late thickness for the secified service life of the ie. Considering a characteristic late thickness of 1, the erosion allowance will tyically be 0.1 to 0.3 2). 2) Tolerable erosion norally less than for rigid steel ies The erosion otential should be assessed for the art of the ie exected to have the worst cobination of curvature (bending radius) and oerating conditions 3). 3) Conservatively the iniu bending radius (MBR) as er design ay be used Figure 4-5 Interlock carcass structure (left), tyical carcass erosion daage (right) Recoended ractice DNVGL-RP-O501 Edition August 2015 Page 35

36 Curved sections of a flexible ie will exerience erosion coarable to an equivalent rigid ie bend. Effects of rougher surface structure of the interlocked carcass coared to rigid iing are exerienced to have oderate effects on the average surface erosion. Fro controlled erosion tests /16/ on interlock carcass and fro detailed CFD erosion siulations, localised hot sot erosion is exected on the leading edge of the carcass, ref. figure above. The erosion odel for ie elbows secified in the current docuent is alicable to flexible ies u to a radius of curvature of 50 ties the internal diaeter of the carcass. When alying the ie elbow erosion odel given in the current docuent to flexible ies, the following guidance is rovided: a) The internal diaeter of the ie should be taken as the iniu internal diaeter of the interlock carcass. b) The radius of curvature should be taken as the iniu bending radius in oeration, also considering dynaic behaviour. c) A iniu geoetry factor of GF=2 for the leading edge of the interlock carcass should be alied on to of the elbow odel, ref. figure above 4). d) Tolerable erosion of the interlock carcass should be liited to 10% of the carcass late thickness unless otherwise secified by anufacturer. 4) The geoetry factor ay vary deending on carcass geoetrical layout. This ay be visualised with a detailed CFD odel. A geoetry factor of GF=2 ilies that the erosion rate calculated with the elbow odel for a flexible ie should be ultilied with a factor of 2. The above aroach is considered sufficient to docuent coliance with the requireents secified in API 17J related to lain erosive wear of the carcass. Additional degradation due to corrosion needs to be considered searately as found relevant. Further details on the distributed erosion on the interlock carcass ay be obtained fro detailed CFD erosion siulations, ref. exale below. Figure 4-6 Exale of interlock carcass CFD erosion contours, showing higher erosion on leading edge in the extrados of the ie 4.12 Production chokes General Production chokes generally stand out as the coonents in oil & gas roduction systes that are ost suscetible to erosion, which is also reflected in the statistics on erosion failures. This is riarily due to the otential for high flow velocities created by the ressure let-down across the choke. In addition to the risk associated with erosion, roduction chokes are also suscetible to lugging in case of high sand concentrations or articles larger than the assage through the throttling art of the choke. This ay cause both oerational robles and accelerated erosion. A brief descrition of tyical choke designs is included in A.D. Recoended ractice DNVGL-RP-O501 Edition August 2015 Page 36

37 Figure 4-7 Scheatic; angle style choke relative to adjacent iing Choke selection Selection of choke for a given sandy service alications should consider the following: Aroriate choke body sizing relative to ustrea iing to ensure that the choke body does not stand out as a significantly weaker link with resect to erosion otential. Choke orientation to reduce risk of lugging by sand and other articles or debris (deending on choke design) Aroriate sizing of flow caacity (C V ) to revent oerating the choke at unfavourable conditions that ay shorten the service life of the choke tri coonents. An aroriate tri size ay reduce the risk of loss of containent. A given choke body should be able to accoodation a sufficient range in tri sizes to fit the variation in oerating conditions during field life. Miniu accetable flow ath through choke (i.e. orts or channels) shall be evaluated with reference to exected size of articles that ay be roduced fro the well (deending on lower well coletion). Traing of erosive articles in the choke ay cause erosive wear significantly above what is exected for the observed sand load due to article re-circulation. Miniu assage should be evaluated according to standard bridging theory, i.e. diension of iniu assage should be at least 3 ties the exected size of articles. Tolerable erosion to the choke body should be established in co-oeration with the choke vendor s recoendations, considering aterial selection and any additional internal CRA cladding. Any tri arts that ay be exosed to high article iact velocities should be of erosion resistant aterial. Erosion resistance and resistance to brittle fracturing should be qualified fro testing. Certain choke tri designs ay cause high risk of erosion in the choke outlet. This is a articular concern for tri designs that generate distinct jet streas in the choke outlet. In soe cases a certain length of the choke outlet needs to be rotected by an extended wear resistant outlet bean or sleeve. Brittle fracturing of wear resistant tri coonents ay cause a sudden increase in flow to downstrea systes. The caacity of the downstrea syste -including the ressure rotection syste- needs to consider a choke collase scenario. The risk of choke collase varies between different choke designs (ref. A.D). A hilosohy for well clean-us needs to be established, i.e. will the roduction choke be exosed to clean-u oerations or will clean-u be erfored with a dedicated tri/choke. Potential effects of exected debris, articles and sand during the clean-u oeration on choke integrity and erforance should be evaluated. Potential for reverse flow through the choke (e.g. alternating injectors/roducers, clean-u wells and injection wells). It should be acknowledged that reverse flow in cobination with sand roduction ay be a significant threat to the integrity of the choke. Such oerations should always be subject to a risk assessent. Recoended ractice DNVGL-RP-O501 Edition August 2015 Page 37

38 Oeration To reduce erosive wear of the choke tri coonents, oerating the choke at less than tyically 20% of the axiu choke caacity (C V ) 1) in cobination with confired sand roduction should be avoided to revent reduced service life of the tri. For ost choke designs this is riarily a controllability issue and not a safety concern. However, for soe choke designs that generate distinct jet streas or swirl in the choke outlet under throttled condition, low choke oening ay also be a significant safety concern. Siilarly, biased flow caused by artial lugging or excessive wear of the throttling echanis can be an issue for any choke design. 1) Due account shall be taken for uncertainty in actuator osition When the roduction forecast ilies that the choke will be oerated at a choke oening less than tyically 20% with confired sand roduction for a long tie eriod (>1 year) it should be considered to teorarily install a choke tri with a saller C V. Figure 4-8 below rovides a general guidance on iniu choke oening for lug/cage and cage/sleeve tyes of chokes, referring to the throttling flow velocity in the choke and sand loading. The table is intended as an exale and should be adjusted according to recoendations fro a secific choke vendor. Recoended choke oerating range (% of axiu tri Cv) as function of choke throttling velocity (Vc) and sand load Oil field Gas field Sand load (g/s) Sand load (ton/yr) Vc= 0-50 /s Vc= /s Vc= (/s) Vc= (/s) < % - 100% 20%-100% 20%-100% 20%-100% 5% - 100% 10% - 100% 20%-100% 20%-100% 5% - 100% 5% - 100% 10% - 100% 10% - 100% Figure 4-8 Exale: Guidance on iniu choke oening for cage chokes Insection and condition onitoring Due to the colex shae and variation in aterial roerties for the different arts of the choke, insection of the choke body using conventional UT ethods is in ost cases not feasible or in best case associated with high uncertainty. Insection will therefore norally require slitting for visual access through flanges or bonnet. For a roerly sized choke body, insection of the adjacent iework (e.g. UT) should norally be a sufficient safeguard for controlling risk of erosion. For an undersized choke body this ay not be the case. For certain choke configurations it is considered feasible to detect erosive wear by onitoring changes in flow characteristics. This requires a sufficiently accurate odel for calculating the theoretical C V of the choke at the given oerating conditions. The calculated value is then coared to the actual choke setting. For ultihase oerating conditions in cobination with single stage tri designs this has roven to be difficult. For ultistage or labyrinth cage designs where internal wear of the cage gives a significant effect on the flow caacity of the choke the aroach is considered feasible. Interretation of any changes in C V that is erceived to be due to erosion needs to consider the accuracy of the theoretical C V odel and should be based on trend curves for a sufficient tie eriod. Recoended ractice DNVGL-RP-O501 Edition August 2015 Page 38

39 An alternative ethod for detecting erosive wear in the choke tri is to assess the internal leak through the choke at closed choke osition and then coare this with an as-new condition. Erosion to the choke tri coonents will in ost cases reveal itself as an increase in internal leakage. The test ay be erfored as art of a lanned well shut-in or as searate oerations if severe tri erosion is susected. Continued oeration with confired severe choke tri erosion should be subject to a risk assessent. Tyical cobinations of abnoral choke resonse and root causes linked to sand roduction given in Table 4-4 ay be used as background for aking qualified decisions on required corrective actions. Table 4-4 Guidance for RCA of chokes Sandy service oeration of roduction chokes - tyical cobinations of failure odes and root-cause Cause of failure Exerienced failure a) Choke cannot be run to fully oen b) Choke cannot be run to closed osition c) Flow through choke at closed osition d) Lower flow through choke than calculated for given ressure dro e) Higher flow through choke than calculated for given ressure dro Actuator failure echanical failure Cage/nozzle collase Erosion to choke body or flange Blockage of sleeve / lug due to foreign object stuck in cage (e.g. during or after iroer well clean-u) Erosion to cage ort holes x x x x x x x x x Erosion to lug nose / sleeve nose / seat Plugging of cage /nozzle Plugging of ressure balance chaber Brittle failure of lug nose or cage Daage by sand to balance seal or dynaic sealing surfaces x (x) x x x x x x x x f) Loss of containent x x Too low aterial fracture toughness, too high actuator load, iact of heavy object Too sall choke body, high bulk rates, low ustrea ressure, high sand load Too sall orts in cage Low choke oening, high ressure dro, low SG, oor TC quality, high sand load Low choke oening, high ressure dro, low SG, oor TC quality, high sand load Cage ort holes too sall High sand load or high otential for sand to travel u the ressure balance holes Vibration, theral shocking, insufficient fracture toughness Root cause Recoended ractice DNVGL-RP-O501 Edition August 2015 Page 39

40 Erosion odel for choke gallery Establishing erosion odels that are alicable to a wide range of choke configurations is challenging due to their relatively colex geoetrical layout coared to other iing coonents such as elbows and reducers. Different choke vendors also have secific design features both related to geoetrical layout and aterial selection that ay affect the erforance in sandy service oeration. In ost cases a detailed CFD erosion siulation is the referred otion for deterining erosion otential. The current odels are therefore liited to what is considered generic for a wide range of choke designs and relevant fro a ersective of choke selection and oeration. Erosion in the choke gallery for an angle style choke design is a otential concern articularly in the case where the choke body is sized significantly saller than the adjacent iing or the choke is designed with a narrow gallery between body and cage. The difference between a wide and narrow gallery is shown in the figure below. Figure 4-9 Generic choke gallery; wide gallery (left), narrow gallery (right); gallery highlighted with red line The choke gallery erosion odel is based on the erosion odel for a standard bend described in [4.7] with inut araeters according to Table 4-5. Table 4-5 Choke gallery erosion ode Paraeter R () D () H () A g ( 2 ) Inut Radius of curvature should be taken as the radius of the choke gallery Diaeter shall be taken as the ga between the cage and choke body Height (effective) of gallery The effective gallery area shall be taken as the 2 x H x D U (/s) Velocity shall be taken as: ¾ x Actual flow rate ( 3 /s) / A g ( 2 ) C 1 (-) C 1 shall be taken as 1.25 D R H The odel is validated towards a selection of reresentative CFD siulations and a liited set of full scale erosion tests. The odel shall therefore not be considered to be ore accurate than a factor of ±3 on the redicted erosion rate. Recoended ractice DNVGL-RP-O501 Edition August 2015 Page 40

41 SECTION 5 MODEL PARAMETERS FOR OTHER EROSIVE AGENTS Oeration of oil and gas facilities also involves other erosive articles than tyical quartz sand. Different tyes of solid aterials are used as weight aterials for drilling, coletion and kill fluids. Related to well coletions and fracturing, different tyes of artificially ade articles are used. For certain oerations or where an artificial gravel ack is set in cobination with a screen failure, roants ay be roduced back to surface. Physical roerties of calcite, barite and roants relevant fro the ersective of erosion are given in Table 5-1 below. Table 5-1 Proerties of other erosive agents than quartz sand Erosive agent Hardness (Moh) Mass density (kg/ 3 ) Shae Alication Quartz sand Sei angular - angular Fro reservoir Calcite Barite Proants Sherical - Sei angular (chalky) Sherical - Sei angular (chalky) Sherical / shar (crushed) Weight aterial Weight aterial Screen coletions/ fracturing Based on aterial erosion testing, using steel as target aterial, guidance on relative erosion of the different erosive agents coared to quartz sand is given in Table 5-2. Table 5-2 Relative erosion for other articles than quartz sand Erosive agent Calcite, Barite Proants (uncrushed) Proants (crushed) Metal chis Guidance: Relative erosion Erosion of steel by tyical barite and calcite articles can for ractical alications be aroxiated to be at least a factor 50 lower than erosion by tyical quartz sand. With reference to Table 3-1, the aterial constant shall for calcite and barite read: K=4.0E-11. The recoendation is based on aterial erosion testing using barite, considering a range in article iact velocity fro /s at iact angle of 30 degrees. Erosion of steel can be aroxiated as equal to quartz sand. With reference to Table 3-1 the aterial constant shall for uncrushed roants read: K=2.0E-09. Erosion of steel can be aroxiated as three ties higher than quartz sand. With reference to Table 3-1 the aterial constant shall for crushed roants read: K=6.0E-09. Associated with well clean-u oerations a saller aount of etal debris fro well erforation ay be entrained in the fluid. Erosion of steel by etal chis ay conservatively be assued as for quartz sand. Recoended ractice DNVGL-RP-O501 Edition August 2015 Page 41

42 SECTION 6 SOFTWARE MODEL Erosion odels described in the current docuent are incororated in a software alication that is coercially available and licenced through DNV GL Software. The alication enables a sile and tie efficient assessent of erosion otential for standard iing coonents as a function of oerating condition and sand loading. Recoended ractice DNVGL-RP-O501 Edition August 2015 Page 42

43 APPENDIX A SAFEGUARDS - MANAGING SAND PRODUCTION AND EROSION A.1 General The current aendix rovides a descrition of otential safeguards for controlling and anaging the consequences of sand roduction and erosion. The safeguards are included both for educational uroses and to serve as a reference for a sand anageent strategy risk assessent. A.2 Sand control Different ethods are alied to liit size and aount of sand fro entering the wellbore. Selection of the otiu solution is a balancing act between effective sand control and increased cost of well coletion together with otentially negative effects on well roductivity. Various down-hole sand screen (filter) solutions are available and tailored to secific field or well conditions and reservoir sand article size distributions (PSD). The noral aroach is to select a screen design that retains articles above a certain size, allowing saller articles to enter into the wellbore. Preventing articles fro entering the well bore ay increase the risk of lugging and reduced roductivity (PI). Sand screens ay be stand-alone solutions or screens backed e.g. with a gravel ack. Other solutions for sand exclusion are oriented erforated/slotted liners utilising the difference in vertical and horizontal foration strength. Various ethods for cheical sand consolidation in oeration have been atteted, however with oderate success. With reference to statistics, sand screen failures during well service life should norally be exected. To iniise risk of screen failure the differential ressure across the screen needs to be controlled to revent article erosion and breach of screen. Partial lugging of the screen is difficult to control or onitor and ay lead to reduced inflow area and hence hot sots for erosion. In case of screen failure the well ay in soe cases be re-coleted, however the standard aroach will be not to reair/relace the screen. Hence, in ost cases, sand roduction after screen failure needs to be anaged and controlled by other eans. For coletions with down-hole sand screens the required resonse tie (e.g. shut-in of well) before critical erosion occurs to the syste should be established. This erosion assessent should consider characteristic article sizes relevant for the unfiltered foration sand including any gravel/roants in the well. It should be evaluated whether the ethods available to detect sand screen failure will cature a sand screen failure within the required resonse tie. A screen failure will norally lead to both increased sand load and roduction of larger articles. A.3 Draw-down control Controlling draw-down is a ossible safeguard for liiting the aount off sand transorted with the reservoir fluid fro foration to well bore. Controlling draw-down will also reduce differential ressure across sand screen coletions, hence reducing the risk of screen failure. The criteria for axiu allowable draw-down should be based on exert judgeent by reservoir and roduction engineers and should be revised according to changes in reservoir conditions. Controlling draw-down directly affects the well roduction erforance; hence a coroise between well roduction rate and tolerable sand roduction is necessary. A.4 Otiisation of lift echanis Otiisation of artificial lift echaniss such as gas lift or ESP in cobination with choke oeration is a otential safeguard for controlling sand roduction and erosion. It should also be acknowledged that different concerns are associated with gas lift as coared to ESP related to sand roduction. ESPs ay exerience reduced service life in sandy service. Injection of lift gas increases the bulk flow rate, hence also the bulk flow velocity in the syste downstrea the injection oint. Systes oerated with artificial lift echaniss are often associated with a low rocess ressure. Particularly for gas lift oerated systes the effects of gas exansion ay result in a high class of erosive service. Recoended ractice DNVGL-RP-O501 Edition August 2015 Page 43

44 A.5 Periodic valve testing According to regulatory requireents, safety critical valves shall be subject to function and internal leak testing at secified intervals. Periodic valve testing ay be a sufficient safeguard to detect abnoral valve resonse due to any valve clogging or erosion, but needs to be considered on a case to case basis. A.6 Continuous onitoring of sand roduction Continuous (online) easureent of sand roduction is a otential safeguard for ensuring that the secified liitations in sand handling caacity are not exceeded. Cobined with aroriate erosion odels, online sand easureents ay be used to continuously kee erosion within safe liits. The ost coonly used technology is based on sensors correlating the distinct acoustic signal created by sand iacting on the ie wall to sand concentration in the rocess flow. With roer calibration the systes are able not only to distinguish between sand and no sand, but also to rovide a quantitative sand rate (g/s) that can be resented in the control roo. Systes are available both for subsea and toside alications. Constraints on detection liits and recision are a function of flow condition and article characteristics and need to be considered on a case to case basis. The acoustic signal fro the sand articles is norally reduced with increased density and viscosity of the fluid and with reduced flow velocity and article size. Sand roduction ay alternatively be indirectly quantified fro intrusive erosion robes by alying an aroriate erosion odel, ref. [4.10]. The ethod is considered alicable for bulk flow velocities above 5 /s and requires roer laceent and orientation of the erosion robe. A.7 Sot-check onitoring of sand roduction Sot-check easureents based on direct saling of sand roduction fro toside streas in a dedicated sand tra or by routing the well through the test searator is a otential safeguard to detect onset or changes in sand roduction. Based on changes in article size distribution a direct saling technique ay also reveal a sand screen failure (increase in article size). It should be acknowledged that sand roduction ay vary between sales. Sales fro rocessed streas will hel detect carry-over of articles to systes downstrea of the searator. An increase in article content in downstrea systes ay indicate solids build-u in the searator. A.8 Continuous onitoring of erosion Intrusive erosion robes based on direct easureent of aterial loss is a otential safeguard to directly onitor and control iework erosion within accetable liits. The location and orientation of the robe is essential for whether the robe signal constitutes a reresentative estiate of iework erosion. Cobined with erosion odels the outut fro an erosion robe ay be used to asses condition of adjacent iework. The robe eleent aterial should referably be siilar to iework aterial with resect to corrosive and erosive roerties. For toside alications the erosion robes can norally be relaced by a relatively sile oeration, and it can therefore be allowed for sensor eleents with shorter service life (erosion) and higher sensitivity. For subsea alications the standard ractice is to select erosion robe eleents that are aligned with the tolerable erosion in the iework which it is suosed to onitor for the target service life of the roduction syste. Current ractice is to use erosion robes with a axiu range u to 1 aterial loss. A.9 Erosion odelling Erosion odelling is a otential safeguard for deterining the erosion otential for a given syste design, oerating condition and sand load. In certain cases the erosion otential for a given syste design and oerating condition can be neglected for realistic sand loads, hence the requireents to sand and erosion onitoring ay be evaluated accordingly. In the oosite end of the scale a given syste design and oeration ay tolerate arginal sand before erosion becoes critical. Hence, erosion odelling is a valuable tool to correlate a given syste design and oerating condition to deterine tolerable sand load. Eirical erosion odels for standard iework and coonents are rovided in this docuent. For colex geoetrical layouts or when the exact location of the erosion attack is iortant Coutational Fluid Dynaics (CFD) erosion siulations ay be the referred otion. Recoended ractice DNVGL-RP-O501 Edition August 2015 Page 44

45 A.10 Monitor bulk flow velocities Considering the strong deendency between erosion otential and the flow velocity converting the oerating rocess conditions into bulk flow velocity is considered crucial. Monitoring the bulk flow velocity at secific locations in the rocess syste will enable the oerator to deterine the class of erosive service. Unless other ethods are available in the data acquisition and rocessing syste the sile black oil odel rovided in [4.4.2] ay be alied. A.11 Monitoring and controlling choke erosion Service life of roduction chokes is significantly affected by sand roduction due to the effects of both tri and body erosion. Perforance in sandy service oeration deends strongly on choke design. Reference is ade to descrition of state-of-the-art choke designs given in A.D. The riary safeguard to control erosion is to use tri coonents of erosion resistant aterial, even when the exected sand roduction is low. Oversizing of the choke tri (C V ) should be avoided as this ay lead to oeration at very low choke oening, hence increasing the risk of tri and choke outlet erosion. For roduction chokes exected to be oerated at low oening with sand roduction for a significant tie eriod, the otion of installing a choke tri with reduced C V should be evaluated. Choke tri erosion ay lead to both increased flow caacity and reduced sealing erforance in closed osition. Perforing an internal leak test for choke valves as art of the eriodic well barrier testing ay be utilised to assess choke tri condition and to identify onset or changes in sand roduction. A.12 Monitor sand build-u in searators Monitoring sand build-u in rocess vessels (e.g. inlet searator) is a otential safeguard to liit carry-over of articles to downstrea systes and detect sand roduction when no other eans of sand onitoring are installed. The liquid strea outlets fro searators norally extend a certain distance into the tank, hence allowing for a certain sand build-u. However, at a certain oint the sand level becoes too high and articles are carried over to the liquid outlets. Areas of sand build-u will due to lack of exosure to the warer roduction strea aear as colder on the external vessel surface. Different ethods for easuring the teerature variations on the outer shell have roven to be a sile and effective way to assess sand build-u. Feasibility of this otion needs to be considered on a case by case basis, articularly considering the vessel theral insulation. A.13 In-line de-sanders In certain cases installation of in-line de-sanders ay be an effective ethod for reducing the aount of sand into the downstrea rocess syste. De-sanders hysically reove sand fro the unrocessed strea and ay be installed in individual well lines or in a gathering line ustrea e.g. inlet searator. Proerly designed to the flow and rocess condition, de-sanders ay be highly efficient for sand reoval. Constraints associated with weight, sace as well as installation and oerational cost need to be considered. A.14 Use of cheicals to irove efficiency of sand searation Sand traed in eulsions ay cause reduced searation efficiency and increased otential for carry-over of articles with the liquid hases to downstrea systes. Potential effects of sand ay be considered in the otiisation of eulsion breakers. A.15 Reoval of sand fro rocess vessels Requireents to systes for reoval of sand fro rocess vessels need to be evaluated deending on exected or actual sand roduction. For rocess vessels not equied with autoatic systes for sand reoval, significant downtie ust be foreseen to anually reove sand. Inlet rocess vessels should as a iniu be reared for a retrofit sand reoval syste. Recoended ractice DNVGL-RP-O501 Edition August 2015 Page 45

46 A.16 Hydraulic igging In cases where the oerating condition of a flowline is outside the enveloe for safe transort of solids; (low roduction rates/velocities), teorary increase in flow velocities, either by increasing flow rate or changing oerating conditions ay be an effective safeguard to identify any solids build u or to obilise and reove solids fro the line. Tyical sand transort regies are illustrated in the figure below. In cases where sand accuulation ay be critical for the flowline, aroriate sand transort odels should be used to address the risk or to deterine conditions required for hydraulic sand reoval. Figure A-1 Sand transort regies A.17 Mechanical igging Mechanical igging ay be an effective ethod for reoval of saller aount of solids fro flowlines, however, the risk of stuck ig should be accounted for when designing the ig train. A.18 Monitoring flow velocities in rocessed liquid streas Liquid rocess streas are often associated with oderate flow velocities. To revent solids build-u in liquid lines, the liquid velocity should exceed 1 /s Ref. ISO In case of ultihase flow a ore sohisticated analysis is required. Models which can be used to assess the otential for sand accuulation in ultihase ielines are, however, available. A.19 Online NDT Different ethods are coercially available for online onitoring of iework wall thickness. Online wall thickness onitoring will norally be liited to one or a few locations on the iework and should for that reason not be considered a substitute for a conventional insection rograe. There are also constraints on ie diaeter, wall thickness and roxiity to flanges and inline equient. Any online techniques for onitoring wall thickness should be at a location where the axiu erosion is exected based on erosion odelling. Failing to do so ay give a false iression that the syste is not exeriencing erosion. A.20 Insection Insection of iework and coonents should be considered as a second line of defence. Hence the consequences of sand roduction should be anaged through other safeguards than insection alone. Feedback fro the insection rograe is however essential for establishing confidence in the sand anageent strategy. Also, actual erosion in coonents not accessible for insection can be estiated using a cobination of insection results fro adjacent iing and suitable erosion odels. A.21 ASR test An Accetable Sand Rate (ASR) test ay be erfored to test whether a roduction well can oerate on a higher roduction rate and still be within the steering criteria for accetable sand roduction. The test is norally erfored as art of the eriodic well test rogra. Recoended ractice DNVGL-RP-O501 Edition August 2015 Page 46

47 A rerequisite for erforing an ASR test is that a sufficiently accurate syste for onitoring sand roduction during the test (and subsequent oeration) is in lace. A tyical flow diagra for execution of an ASR test is shown in the figure below. For a secific field alication the flow diagra should be detailed in the well secific rocedures alternatively included as art of the well test rocedures. Figure A-2 Scheatic illustration of an ASR test Recoended ractice DNVGL-RP-O501 Edition August 2015 Page 47