Materials Qualification for Bolting Applications NAS Committee on

Materials Qualification for Bolting Applications NAS Committee on Connector Reliability for Offshore Oil & Gas Operations Oliver Moghissi 10 April 2017 1 DNV GL © 2017 10 April 2017 SAFER, SMARTER, GREENER

Bolting Background § Subsea bolted connections can be critical to system integrity – Wellheads, Xmas trees, Flanges, Structural connections, etc. § Specs & standards – Usually adequate but not uniform – Non-conformance § Bolting material performance – Strength, Corrosion, Galling, Hydrogen susceptibility § Bolting manufacture & quality management – Varied & not fully known because of traceability § Few identified failures – Is it just statistics of crack distribution? – Overlapping/Aggregate uncertainties – Significant consequence – Weak reporting of incidents (failure or not) when not mandated 2 DNV GL © 2017 10 April 2017

$Qualification vs Fitness for Service § Significant # are brittle fracture (primarily H related)$

Commonly Used Materials § Low Alloy Steels both ferritic and martensitic (up to 125 ksi) – Typically limited to 105 ksi subsea § Monel (e. g. , K 500) – Used both topside and subsea both by O&G and Navy (with known failures) § Stainless Steels (ASTM A 286) – 300/400 series (susceptible to localized corrosion & HE) § Duplex – 2507 (HE susceptible) § Nickel Based Alloys – 718/725 (HE susceptible) – C 276/625/686 ‘Alloying up’ is not always the answer 4 DNV GL © 2017 10 April 2017

What Are We Testing? § What are technical causes for failures? – Material properties – e. g. , Hardness criteria (e. g. , >34 HRC); non-homogeneous in bolt – Excess hydrogen – Before service (i. e. , and not baked out) – In-service (e. g. , CP, coatings) – Environment – e. g. , crevice, biofilm, contact with internal fluids – Load & design characteristics – Bolt installation – In service load profile (e. g. , strain rates) 5 DNV GL © 2017 10 April 2017

Why are We Testing? § Ultimately, we are assessing a material’s fitness for service § Material qualification is normally testing according to a standard and comparing to a pass/fail criterion – A single criterion represents many service conditions (usually hardness) – Prescriptive over Performance – A safety-factor or other conservatism is usually built in – In most cases, it is costly – In some cases, it compromises safety – Leads to exception requests – Can failures occur despite meeting every existing materials spec? 6 DNV GL © 2017 10 April 2017

Industry Standards § Current standards specify basic material properties and do not directly address inservice performance – ASTM A 193 ”Alloy steel and stainless steel bolting materials for high temperature service” (ferritic steels Grade B 7, B 7 M) – ASTM A 320 ”Alloy steel bolting materials for low temperature service” (ferritic steels Grade L 7, L 7 M) – ASTM A 354 ”Quenched and tempered alloy steel bolts, studs and other externally threaded fasteners” – ISO 898 -1 ”Mechanical properties of fasteners made of carbon steel and alloy steel – Part 1: Bolts, screws and studs – API 20 E - Alloy and Carbon Steel Bolting for Use in the Petroleum and Natural Gas Industries – API 20 F – Corrosion Resistant Bolting for Use in the Petroleum and Natural Gas Industries 7 DNV GL © 2017 10 April 2017

Example – Low Alloy Steel Microstructures in API 20 E § Requirements not explicitly tied to in-service damage modes § Requirements include – Processing (e. g. , cast, forged, continuous Cast (not allowed for BSL-3)) – Limits on banding, porosity, segregation – Wrought Microstructure is desired 8 DNV GL © 2017 10 April 2017

Qualification Testing – Hardness § Hardness measurements used as proxy for hydrogen susceptibility § Requirements for steels vary from 34 – 38 HRC among standards and company specs for subsea service – Hardness measurements required on center of bolt – Properties vary across the bolt (especially for rolled threads) – Are acceptance criteria already conservative to account for this? – Or do we measure highest hardness? – Do we have different specs for different bolts? 9 DNV GL © 2017 10 April 2017 *B. Craig

Qualification for Hydrogen Content § H content measured using ASTM F 1113 – H ingress during manufacture (e. g. , plating) and typically baked out § Hydrogen ingress can also occur in-service (e. g. , CP) 10 DNV GL © 2017 10 April 2017

Standard ASTM Qualification for H Embrittlement § ASTM F 1624 ‘Measurement of Hydrogen Embrittlement Threshold in Steel by the Incremental Step Loading Technique 1 – Step loaded method to identify threshold stresses for fastener crack initiation in environment (e. g. , primarily seawater + CP) 11 DNV GL © 2017 10 April 2017

Load Rate and Profile affects Fracture Toughness Measurement § Slow continuous rising displacement test (modified ASTM E 1820) shows low toughness § Step loading (ASTM 1624) is sensitive to hold time, even at same effective loading rate § Which represents service conditions? § Which is conservative or non-conservative? 12 DNV GL © 2017 10 April 2017

Simulating Service Conditions – HPHT Example § Typical service conditions involve some load changes and long periods of constant loads § Test methods must simulate loading profiles to represent service conditions 13 DNV GL © 2017 10 April 2017

Factors for Bolt Performance have Distributions § Distributions come from – Measurement uncertainties – Non-homogeneous structure and environment – Stochastic processes § Examples – Material properties – Hydrogen in-service (e. g. , CP, coatings) – Environment (e. g. , geometry/chemistry) – In service load profile (e. g. , strain rates) Hydrogen – Hydrogen before service Hardness 14 DNV GL © 2017 10 April 2017

Simple BN example 15 DNV GL © 2017 10 April 2017

Conclusions § Current state – Majority of fasteners perform as intended – Exceptional failures need to be addressed – True failure rates not known § Currently being addressed – Existing specs & standards for materials qualification adequate most of the time – Lack uniformity – Conservative most of the time and non-conservative some of the time – Non-conformance is possibly deficiency § Future – Step-change improvement realized by performing materials qualification through fitness-for -service lens – Improved understanding of failures will require better failure analysis and reporting – Probabilistic component of performance requires understanding of distributions and how they aggregate 16 DNV GL © 2017 10 April 2017