Valhall Corrosion Events and Management · 2009. 12. 1. · duetos cale buildu p Ongoing Biocide...
Transcript of Valhall Corrosion Events and Management · 2009. 12. 1. · duetos cale buildu p Ongoing Biocide...
Valhall Corrosion Events and ManagementLearning Pack
Eldar�Larsen,��Head�of�Operations,�BP�Norway.
INTERNAL�DISTRIBUTION�ONLY 2
Valhall�Corrosion Events - Summary
Summary
• Hydrocarbon�leak�HIPO�on�6th�March�2009�caused�by�Microbial�Induced�Corrosion�(MIC)
− resulted�in�crude�and�gas�leak,�First�Aid�Incident,�emergency�plant�shutdown�&�muster.�
• Valhall�eventually�shutdown�for�10�weeks�for�corrosion�inspection�and�repairs.
− during�which�three�further�incidences�of�MIC�and�44�anomalies�identified.
Major Lessons
• Behaviours of MIC were misunderstood
− Areas�of�high�flow�rate�were�previously�thought�unlikely�to�experience�MIC,�this�assumption�failed�to�identify�internal�pipe�work�scaling�can�provide�environment�for�MIC�activity.
• Inadequate application of hazard Recognition, risk awareness and control of work.�
− The�initial�inspection�routine�was�inappropriate�and�directly�led�to�exposure�of�the�individual�and�the�resultant�FAI.�
• 15�specific�improvement�opportunities�identified�as�a�result�of�the�incident�and�plant�investigations�(7�organisational,�5�work�process�and�3�performance�management).
INTERNAL�DISTRIBUTION�ONLY 3
Microbial Induced Corrosion
What�is�MIC?
• MIC�is�Microbial�Induced�Corrosion.
• MIC�of�carbon�steel�typically�manifests�itself�as�localised�corrosion�or�pitting�attack�following�the�development�of�a�surface�biofilm.
• The�anaerobic�environments�of�oil�and�water�transport�pipework�often�support�the�growth�of�biofilms,�which�almost�invariably�contain�sulphate-reducing�bacteria�(SRB),�a�major�cause�of�MIC.�
• SRB�containing�biofilms generate�H2S.
• H2S�results�in�formation�of�Iron�sulphide�within�the�biofilm.�
• Iron�sulphides�are�cathodic�to�bare�steel�and�can�have�the�effect�of�greatly�increasing�corrosion�at�anodic�sites.
Why�does�MIC�occur?
• MIC�can�occur�in�all�systems�that�contain�water�with�SRB’s.
• Prediction�of�corrosion�rates�attributed�to�MIC�is�unreliable�
− uncertainty�of�the�onset�of�localized�corrosion/pitting�,whether it�proceeds�at�a�constant�rate.
• Parameters�that�can�influence�MIC�include�pH,�temperature,�nutrition,�velocity�of�fluids,��presence�of�deposits.
• MIC�is�more�likely�to�occur�at�locations�with�low�fluid�velocity such�as�in�dead�legs,�under�deposits,�scale.
• All�metals�are�susceptible�to�MIC.�Carbon�steels�have�generally�been�found�to�be�most�susceptible.
• MIC�mitigation�relies�on�effective�biocide�dosing�regimes.
INTERNAL�DISTRIBUTION�ONLY 4
Reference - Valhall Timeline
1982
Valhall�start�
production,
Incl.�Jet�
washing
separators�
using sea-
water.�By�
mid-’90’s�
using mainly
produced
water
1998
New�Metering�
Skid�installed.�
Small�bore�
pipe�work�not�
commissioned�
in�Workmate
1999RBI�Analysis�performed�–MIC�internal�corrosion�not�sufficiently�
identified�as�a�threat.
2000
Inspection�
scheme�
constructed�
which�
identified�
external�
inspection�of�
Liquids�
metering�
package�as�a�
whole
2004-2006
Chemical�
cleaning�of�
fast�loop�to�
aid�
calibration�
due�to�scale�
build�up
Ongoing
Biocide�
treatment�of�
Sulphate
Reducing�
Bacteria�
(SRB)�on�
Produced�
Water�(PW)�
system�
ongoing�and�
working
2007
SRB’s in�2nd�
stage�separator�
0-460/ml
Sept 2007
Cor.�Mngt.�Health�
Check:�Valhall�
Prod.�Water�
Sulfate�Reducing�
Bacteria�(SRB)�
contamination�as�
a�high�risk�and�
recommends�a�
microbial�audit�by�
microbial�
specialist
06.Mar.2009
Leak�in�1”
pipe�fast�
loop
HIPO
02.Apr.2009
Upstream�
skimmed�oil�
tank
17.Apr.2009
Leak�in�6”
prod.�water�
pipe
INTERNAL�DISTRIBUTION�ONLY 5
Reference - Valhall Timeline
Events and response:
6th�March�2009�”Metering�fast�loop” (F&G�→ ESD).
• Limited�awareness�of�MIC,�deemed�low�probability.�Limited�spot�checking�by�X-ray�outside�the�
fast�loop.�No�SD�justified�by�residual�corrosion�risk.�No�additional�resources.
2nd�April�2009�“T-337�pressure�increase” (No�ESD).
• Start�up�after�a�compressor�incident�increased�flare�backpressure�which�burst�a�MIC�in�the�
produced�water�system.�
• 15�days�(6�+�9)�shutdown�looking�for�MIC�with�Eddy�current�and�X-ray�w.�crew�of�3.
17th�April�2009�“6” produced�water�pipe” (No�ESD).
• Restart�attempt�failed�when�a�MIC�burst�in�the�prod.�water�system.�Had�been�inspected�in�very�
close�vicinity.�Increasing�realization�of�the�criticality�of�adequate�inspection�rigour.
• 10�days�shutdown,�100%�inspection�of�prod.�water�system�by�Eddy current�and�X-ray,�extended�
risk�based�approach.�Increased�offshore�resources.
27th�April�2009�“Condensate�return�to�prod.�water�system”.�(F&G�→ ESD).
• Pipe�burst�during�start�up�attempt.�Contained�highly�volatile�condensate.�
• Realized�technology�gap.�Changed�to�SLOFEC�scanning�successfully.�
• 30+�team�offshore.�Task�force�onshore.�True�100%�inspection�beyond�produced�water.�RIB�
revision.
1st�June�2009�“Successful�restart”.
• 10�weeks�of�production�loss�and�serious�HSE�incidents.
INTERNAL�DISTRIBUTION�ONLY 6
Valhall Corrosion Event (March 6th 2009)
Summary of first event.
• A�‘seep’ or�sweating�leak�observed�coming�from�an�insulated�1” carbon�steel�pipe�was�prepared�for�inspection�following�removal�of�the�insulation.
• The�pipe�was�on�the�crude�export�metering�skid�and�contained�a�mixture�of�crude�oil�and�gas�at�25�bar�pressure.
• On�removal�of�insulation,�the�leak�was�observed�not�to�originate�from�the�flange�as�originally�expected.
• An�inspector�examining�the�pipe�initially�used�a�blade�to�scrape�off�paint�at�the�leak�site.
• During�a�later�visit,�the�inspector�touched�the�leak�site�with�his�finger�and�was�sprayed�in�the�face�with�oil�under�pressure.�
• During�hydrocarbon�release,�the�hole�enlarged�to�10mm�diameter.
• Valhall�went�to�full�Emergency�Shutdown�and�muster.�
Hydrocarbon leak on metering skid – First Aid Incident – Emergency Plant Shutdown
INTERNAL�DISTRIBUTION�ONLY 7
Inherent Design
Plant Layout
Control, Alarm &
Shutdown system
•Inspection (RBI) and Corrosion Management
practices failed to identify and mitigate the internal corrosion threat/risk that led to the failure•RBI assessment not updated to reflect changes to operating
and/or process conditions
Maintenance
& Inspection
Learning from
the Past
•Actions from 2007/2008 Health Checks to improve Biocide treatment
and to conduct Microbial survey not completed
•Force Inspection procedures inconsistent with requirements of HSE
Directive for visual inspection
Operations
Procedures
•Inspection Engineer
delegated inspection task to NDT Technician.
Effective
Supervision /
Leadership
Training &
Competency
�Failure to follow CoW
and RBI procedures
Work Control
•Inadequate Hazard Recognition and Risk Awareness – job treated as ‘routine’ and Level 2 work permit.
•CoW - Failure to isolate energy source prior to work•CoW - Failure to comply with requirements for ‘Visual Inspection’
Support to Next
of Kin & Injured
Relief and
Blowdown
system
Audit & Self
Regulation
•No apparent MoC for chemical treatment of Fast Loop line.•1” line not identified on risk-based inspection schedule
Management
of Change
Active & Passive
Fire Protection
Escape /
Access
Rescue &
Recovery
HAZARD
Hydrocarbon
Inventory at
25bar in Fast
Loop line
Investigation &
Lessons Learned
HAZARD
REALISATION
Loss of containment
First Aid injury & HiPo
Hazard Barrier Diagram
Heirarchy of control – Bias towards hardware/inherent safety & reducing the scope for human error – multi barrier defence
•Four Point Check
not used.
INTERNAL�DISTRIBUTION�ONLY 8
Valhall Corrosion Event (April 2nd 2009)Summary of second event.
• A�leak�in�the�Produced�Water�System�followed�
a�gas�compression�upset.�Area�isolated�&�
upstream�vessel�drained.
• 25mm�hole�found�in�drain�line.
• Areas�of�low�flow�and�dead-legs�were�identified�
and�inspected.�6�further�MIC�incidents�
observed.�
• Investigation�initiated�to�ensure�appropriate�
analysis�and�response�prior�to�restart.
• Management�verification�of�process�mitigations�
and�corrosion�mapping�&�repair�effort.
Targeted inspected discovers further incidence of MIC
INTERNAL�DISTRIBUTION�ONLY 9
Valhall MIC Response The third event: Defining the Scope of Work
• Produced�Water�identified�as�being�the�main�
host�of�Sulphate Reducing�Bacterias (SRB’s)
• Considered�potential�contamination�of�other�
systems�vs�HC�leak�risks.
• Scope�agreed�by�Ops,�Process�TA,�Corrosion�
Mgmt�TA,�Inspection�&�CM�Contractor.
• Produced�Water�System�was�100%�inspected.
• Risk�Based�”Hotspot” Execution�– Water�wetted�
surfaces�eg Low�spots,�Sumps,�Deadlegs etc.
− Waste�injection�on�drilling�platform�(DP).
− Well�systems,�confined�to�slurry�lines.
− Crude�Oil�processing�including�the�
metering�system.
− Closed�Drains�System.
Oil Return from PW Flash Drum to Oil Return tank
Orientation of defect; 9 o’clock position on vertical
pipe
Unpredicted nature and location of corrosion required a new approach to
immediate reaction inspection.
INTERNAL�DISTRIBUTION�ONLY 10
• Leak�in�6”produced�water�high�flow�pipework
• A�8�mm�hole�was�discovered�during�preparation�for�plant�restart�following�inspection�program�triggered�two�gas�detectors�&�ESD.
• Critical Factor�1�
− Microbial�induced�Corrosion�(MiC)�mechanism�in�the�produced�water�system
• Critical�Factor�2
− Inspection�(RBI)�and�Corrosion�Management�practices�failed�to�identify�the�corrosion�risk�in�high�level,�high�flow�pipework
Targeted inspected discovers further incidence of MIC – in high flow pipework
Valhall�Corrosion�Event.Summary�of�forth�event.
INTERNAL�DISTRIBUTION�ONLY 11
Valhall MIC project – Inspection
• Extensive�inspection�effort�required�to�execute�
identified�scope�– 30+strong�team.
• SLOFEC�technology�used�
• Priority�1�scope�Identified�44�anomalies
− 23�MIC�Related
− 21�Non-MIC-related
• 19�spools�replaced,�5�temporarily�repaired�&�20�
scheduled�for�ongoing�inspection�following�start-up
• Assurance�processes�followed�by�No�go/Go�
verification�reviews
• Priority�2�inspection�scope�identified�for�post�start�up�
inspection.
No.�10�– T-337�Flange�corrosion
Following fresh focussed risk assessment an intense and prioritised
inspection programme was initiated.
INTERNAL�DISTRIBUTION�ONLY 12
Lessons Learnt
Direct cause
• Cross�contamination�of�oil�production�system�with�SRB�via�the�produced�water�
system.
• Inadequate�awareness,�inspection�and�mitigation�as�analysis�did�not�identify�oil�
system�as�at�risk�from�SRB�corrosion.
• Insufficient�chemical�injection�(such�as�biocide�and�corrosion�inhibitor)�and�
sampling�and�analysis�to�control�microbial�activity.
• As�a�result�MIC�occurred�within�oil�systems.
Underlying causes
• Improvement�opportunities�identified�within:
− Organization (7�subcategories�with�recommendations)
− Work Processes (5�subcategories�with�recommendations)
− Performance Management (3�subcategories�with�recommendations)