A reliability study of a Deepwater Vertical

Xmas Tree with attention to XT retrieval

rate

Oda Ingeborg
Stendebakken

Marine Technology
Submission date: June 2014
Supervisor: Jan Erik Vinnem, IMT
Co-supervisor: Endre Willmann, GE Oil & Gas

Norwegian University of Science and Technology

Department of Marine Technology
Faculty of Engineering Science and Technology
Department of Marine Technology

MASTER THESIS 2014

for

M.Sc. student Oda I. Giske Stendebakken

Department of Marine Technology

RELIABILITY STUDY OF A DEEPWATER VERTICAL XMAS TREE WITH

ATTENTION TO XMAS TREE RETRIEVAL RATE
Pålitelighetsstudie av dypvannsvertikalt tre med fokus på trekkerate av treet

Thousands of subsea Xmas trees have been installed since first introduced subsea in the
1950’s. Now, there are a number of subsea tree variations, including vertical, horizontal, dual
bore, mono bore, TFL (Through Flow Line), drill-thru horizontal, vertical with tubing head
spool, mudline vertical and mudline horizontal trees.

The primary function is control of flow, usually hydrocarbons from the well, but also
injection of gas or water to maintain reservoir pressure, or injection of lift gas to assist the
flow of hydrocarbons. A tree often provides numerous additional functions including
chemical injection, monitoring (such as valve positions, pressure, temperature, corrosion,
erosion, sand detection, flow rate, flow composition, etc.) and well intervention means. Tree
complexity and functionality has increased over the last few decades.

In relation to subsea development projects, reliability and availability performance targets are
normally part of the contractual requirements. As a general requirement, the subsea
contractors are also responsible for optimizing the system design, in a life cycle perspective,
taking account of various aspects including production availability, installation/intervention
risks as well as ability to support reservoir management operations.

In order to optimize the tree system design/configuration with respect to reliability and
availability performance, there is a need for proper reliability models that are able to
differentiate between different options and variants.

  i  
  
 
There are two main approaches to predicting the reliability of subsea Xmas trees:

1. Bottom-up: Reliability modeling of system, component by component, using

component reliability data available in industry recognized sources such as the
OREDA Handbooks.
2. Top-down: High-level reliability data for systems, captured from actual field
experience/operations, typically reflecting the reliability performance as seen from
the operators point of view.

The two approaches results in totally different results. For instance, an attempt to model a
conventional tree using component data from OREDA may typically predict an “Xmas tree
retrieval rate” in the order of once per 15-20 Xmas tree years, as opposed to once per 150-200
years expected or experienced by the operators/customers.

The idea is to use actual in-field performance data to calibrate the ‘bottom-up models’, such
that the top-level predictions are in accordance with field experience. Although GE Oil & Gas
have a good understanding of the overall reliability and availability performance as seen from
the customers, they do not really have a well-documented understanding of actual field
experience for their Xmas trees (number of trees installed, years in operation, number of
failures, corresponding repair activity etc.).

The main objective is to study a specific Xmas Tree (XT) system and to estimate the retrieval
rate due to tree failures based on commercially available reliability data. Further this shall be
compared to high level experience data presented. This is to initiate the process to alleviate
the gap seen between generic calculations of the tree retrieval rate compared to known field
experience.

The master thesis shall cover the following tasks:

1. Literature study: The candidate shall perform a literature survey and, on the basis of
this survey, describe:
a) Main types of Xmas trees, and main Xmas tree sub-systems and components
b) Describe the most important differentiating factors with respect to reliability
and availability performance of Xmas Tree Systems (including
installation/intervention issued and downhole operations).
c) Essential terms, definitions and industry standards for performing
probabilistic analysis of subsea tree systems, and describe main
methodologies
d) Relevant reliability data source(s), with emphasize on limitations and
applicability in relation to the current topic.
2. Component-level FMECA of a selected Xmas tree configuration (excluding control
module).
a) Develop an understanding of main components with essential functional
requirements and criticality and effect resulting from functional failures.
3. Probabilistic Reliability Analysis of the selected Xmas Tree System
a) Bottom-up approach:
i. Based on reliability data sources, develop a Xmas Tree specific
database containing component reliability data for Xmas Tree
components/items.
ii. Use this to develop a first pass reliability model for the Xmas Tree
System.

  ii  
 b) Top-down approach:
i. View statistics on the high level reliability and availability
performance of subsea trees.
ii. Seek literature for relevant comparative information.
iii. Use this information to develop a ‘simplified’ top down reliability
model for the selected Xmas Tree configuration.
c) Comparative assessment: Derived from a comparative assessment of results,
recommend a baseline set of component reliability data, additional model
parameters, modifications factors and other refinements as required for
calibration of the bottom-up model

Outlook: We are currently struggling to obtain Reliability Targets for the next generation of
subsea trees and associated sub components. We believe this should be driven by a Systems
top-down approach rather than assigning arbitrary values at the component level (bottom-up
approach).

In the process of identifying Design Practices for 2014, the thesis work will be seen in
relation to this. This will include methodologies for collection of field statistics on subsea
trees.

The main objective is to establish a standard benchmark model (simple probabilistic rather
than time-based) and associated process (Design Practice) for the allocation of component
reliability targets based on goals for overall system reliability and availability.

The thesis must be written like a research report, with an abstract, conclusions, contents list,
reference list, etc.

During preparation of the thesis it is important that the candidate emphasizes easily
understood and well-written text. For ease of reading, the thesis should contain adequate
references at appropriate places to related text, tables and figures. On evaluation, a lot of
weight is put on thorough preparation of results, their clear presentation in the form of tables
and/or graphs, and on comprehensive discussion.

The thesis is to be handed in electronically.

Thesis supervisors:

Prof. Jan-Erik Vinnem, NTNU

Endre Willmann, GE Oil & Gas

Deadline: 10th June 2014

  iii  
PREFACE  
This work comprises my master thesis for the Department of Marine Technology at
the Norwegian University of Science and Technology, NTNU, spring 2014. The
master thesis account for 30 credits in the last semester of the final year, and
completes a Master of Science degree in Marine Technology, within the
specialization Marine Operations and Maintenance. The master thesis has been
executed in collaboration with GE Oil & Gas.

The main purpose of this thesis is to estimate the retrieval rate of a specific tree
system due to tree failures based on commercially available data and further compare
the results to experiential data presented. This is to initiate a process alleviating the
gap seen between generic calculations of the tree retrieval rate compared to known
field experience.

As I did not have any significant knowledge about XTs, a part of the thesis was to
gain knowledge about different XT systems, their function and reliability issues
thereof. The work has been awarding and exciting. It has been especially interesting
to work with an actual problem and to get insight into the reliability engineering
industry.

I would like to give my most genuine thanks to my responsible supervisor at NTNU,

Jan-Erik Vinnem, for valuable help, input and for always being online, even when he
was not campus. I would also like to give my most sincere thanks for my supervisor
in GE Oil & Gas, Endre Willmann, for taking the time and for this give valuable
insight, feedback and patience throughout the thesis.

Finally, I would like to thank Oline Giske Stendebakken for moral support and input
on my writing.

Trondheim
10th
June
2014

__________________________
Oda Ingeborg Stendebakken  

  I  
EXECUTIVE  SUMMARY  
The  first  subsea  XT  was  installed  in  the  Gulf  of  Mexico  in  1961.  Since  then,  the  
XTs  are  an  essential  part  of  the  subsea  fields.  Now  there  are  a  number  of  XT  
variations  with  technology  modified  to  fit  each  unique  well.  Tree  complexity  and  
functionality  has  increased  over  the  last  few  decades.  

The  focus  of  this  thesis  is  towards  the  subsea  Xmas  Tree  (XT)  system.  The  XT  is  
recognized  by  the  industry  as  an  overall  reliable  configuration,  but  as  in  all  
development  projects,  it  is  a  constant  battle  to  optimize  the  design  in  a  life  cycle  
perspective  taking  account  of  several  aspects  such  as  safety,  availability,  
maintainability  and  reliability.  The  ambient  seabed  conditions  and  continuously  
increasing  intervention  cost  require  a  higher  standard  on  the  equipment  and  
keep  pushing  the  technology  development.    

There  are  two  main  approaches  for  predicting  reliability  of  tress:  

• Bottom-­‐up:   Reliability   modeling   of   system,   component   by   component,  

using  component  reliability  data  available  in  industry  recognized  sources,  
such  as  the  OREDA  handbook.  
• Top-­‐down:  High-­‐level  reliability  data  for  systems,  captured  from  actual  
field  experience/operations,  typically  reflecting  the  reliability  
performance  as  seen  from  the  operators  point  of  view.  

The  main  objective  is  to  study  a  specific  XT  system  and  to  estimate  the  retrieval  
rate  due  to  tree  failures  based  on  commercially  available  reliability  data.  Further  
this  has  been  compared  to  high-­‐level  experience  data  presented.  This  is  to  
initiate  the  process  to  alleviate  the  gap  seen  between  generic  calculations  of  the  
tree  retrieval  rate  compared  to  known  field  experience.  The  XT  configuration  
chosen  to  evaluate,  is  the  Deepwater  Vertical  Xmas  Tree  (DVXT).  A  generalized  
case  was  constructed  as  the  DVXT  with  help  from  Endre  Willmann,  the  
supervisor  in  GE  Oil  &  Gas.    

The  scope  is  limited  in  this  context  to  the  DVXT  system  and  systems  that  
influence  the  DVXT  system  in  terms  of  tree  retrieval  rate  and  downtime  due  to  
failures  in  the  tree  system.  Therefore,  the  subsea  control  systems  with  associated  
monitoring  equipment  are  excluded  from  the  analysis.    

To  assess  the  DVXT  system,  a  reliability  analysis  is  performed.  The  reliability  
analysis  is  achieved  in  the  following  steps  with  proven  methods  from  the  
reliability  engineering  discipline:  

1. FMECA/Failure  analysis  
2. RBD/Reliability  analysis  

  III  
  
 
A  component-­‐level  FMECA  is  conducted  to  develop  an  understanding  of  main  
components  with  essential  functional  requirements,  criticality  and  effect  
resulting  from  functional  failure.  The  results  from  the  reliability  analysis,  
conducted  as  a  bottom-­‐up  approach,  indicate  a  retrieval  rate  of  the  production  
tree  near  22  years.    

Through  the  performed  reliability  analysis,  the  DVXT  system  has  confirmed  its  
reputation  as  a  reliable  configuration  with  high  operating  reliability  and  
associated  low  risk.  Nonetheless,  several  assumptions  have  been  made.  The  
focus  of  this  thesis  is  not  at  the  absolute  result,  but  to  illustrate  a  reliability  issue  
experienced  by  GE  Oil  &  Gas  in  the  calculation  of  reliability  based  on  generic  
reliability  data  versus  field  experience  data.  The  OREDA-­‐2009  Handbook  is  
deliberately  used  as  a  sole  source  for  raw  data  to  illustrate  this  issue,  as  the  
handbook  is  known  to  give  conservative  results  when  calculations  is  performed  
purely  based  on  it.  However,  it  can  well  be  seen  as  desirable  that  calculations  are  
more  cautious  than  a  real  situation,  but  a  natural  question  here  is  to  what  
extend.  

It  is  shown  a  significant  gap  between  the  bottom-­‐up  approach  and  the  field  
experience  data  presented.  Based  on  the  field  experience  collected  it  is  indicated  
a  MTTF  for  XT  retrieval  on  the  top-­‐down  approach  between  100  to  200  years.  
This  implies  a  factor  of  5  to  10  between  the  bottom-­‐up  and  the  top-­‐down  
approach.  It  should  be  noted  that  the  estimates  for  retrieval  rate  is  meant  for  a  
population  of  trees  in  operation.  A  tree  will  obviously  not  be  able  to  operate  for  
200  years.  

Further,  it  is  indicated  by  GE  Oil  &  Gas  that  a  Pareto-­‐rule  seem  to  apply  when  
deciding  if  failures  require  heavy  workover  such  as  XT  retrieval  or  light  
intervention  means  such  as  ROV  remedial  actions  upon  repair.  Applied  to  the  
sensitivity  case  to  the  bottom-­‐up  approach,  assuming  that  in  fact  80%  of  XT  
critical  failures  can  be  restored  by  light  intervention  means,  the  total  MTBF  of  15  
years  predicted  for  XT  critical  failures  then  results  in  an  XT  retrieval  rate  of  75  
years.  This  is  closer  to  the  expectations  indicated  by  the  top-­‐down  approach,  but  
still  not  close  to  the  levels  indicated  by  recent  field  experience.  

This  indicate  that  the  bottom-­‐up  model  should  be  calibrated  with  input  data  that  
is  based  on  experience  data  rather  than  solely  based  on  generic  to  alleviate  some  
of  the  distance  between  the  two  approaches.  This  can  be  performed  in  shape  of  
additional  model  parameters,  modification  factors  or  other  refinements.  The  
solution  to  this  is  however  not  presented  in  this  thesis.    The  Pareto-­‐rule  can  be  
utilized  in  order  to  calibrate  this  gap,  if  shown  applicable  based  on  
comprehensive  historic  data.  This  has  to  be  studied  further  thoroughly.  

  IV  
SAMMENDRAG  
Det  første  juletreet  på  havbunnen  ble  installert  i  Mexicogolfen  i  1961.  Siden  da  
har  treet  vært  en  essensiell  del  av  havbunnsfelt.  I  dag  eksisterer  det  en  rekke  
varianter  av  trær  med  teknologi  modifisert  til  å  passe  hvert  unike  felt.  
Kompleksitet  og  funksjonalitet  til  trærne  har  økt  de  siste  tiårene.    

Fokuset  i  denne  masteroppgaven  er  rettet  mot  juletresystemet  på  havbunnen.  

Treet  er  anerkjent  av  industrien  som  en  pålitelig  konfigurasjon,  men  som  i  alle  
utviklingsprosjekter  er  det  en  konstant  kamp  å  optimalisere  designet  i  et  
livsløpsperspektiv  hvor  det  blir  tatt  hensyn  til  aspekter  som  sikkerhet,  
tilgjengelighet,  vedlikehold  og  pålitelighet.  Omgivelsene  på  havbunnen  og  de  
stadig  økende  kostandene  for  intervensjon  krever  en  høyere  standard  på  utstyr,  
og  fortsetter  å  drive  teknologiutviklingen.  .    

Det  er  to  hovedtilnærminger  for  å  forutsi  påliteligheten  til  et  tre:    

• ”Bottom-­‐up”:  Pålitelighetsmodellering  av  systemet,  komponent  for  

komponent,  basert  på  pålitelighetsdata  på  komponentnivå  tilgjengelige  
kilder  anerkjent  av  bransjen,  slik  som  OREDA-­‐2009  håndboken.    
• ”Top-­‐down”:  Pålitelighetsdata  for  systemer  på  et  overordnet  nivå,  tatt  fra  
felterfaring/drift.  Disse  dataene  reflekterer  vanligvis  pålitelighetskrav  
sett  fra  operatørenes  synspunkt.    

Hovedmålet  med  oppgaven  er  å  studere  et  bestemt  tre  og  estimere  trekkerate  
for  dette  treet  på  grunn  av  feil  i  treet.  Trekkerate  vil  si  hvor  ofte  en  skal  påregne  
å  erstatte  et  tre;  ta  det  opp  og  sette  inn  et  nytt  tre.  Beregningene  er  basert  på  
kommersielt  tilgjengelige  pålitelighetsdata.  Videre  skal  dette  sammenlignes  med  
erfaringsdata  fra  oljefelt  presentert  i  oppgaven.  Dette  blir  gjort  fordi  GE  Oil  &  Gas  
har  erfart  en  signifikant  avstand  mellom  beregnet  trekkerate  og  den  reelle  
trekkeraten  i  felt.  Hensikten  med  denne  masteroppgaven  er  å  sette  i  gang  
prosessen  med  å  minske  gapet  mellom  generiske  beregninger  og    presentert  
felterfaring.  Den  konfigurasjonen  som  ble  valgt  som  case  er  et  dypvannsvertikalt  
tre.  En  generell  case  av  det  vertikale  dypvannstreet  ble  konstruert  ved  hjelp  av  
Endre  Willmann,  veileder  i  GE  Oil  &  Gas.    

Omfanget  av  oppgaven  er  begrenset  i  denne  sammenheng  til  det  vertikale  
dypvannstresystemet  og  systemer  som  påvirker  treet  i  form  av  trekkerate  og  
nedetid  på  grunn  av  svikt  i  selve  treet.  Derfor  er  kontrollsystemet  på  havbunnen  
med  tilhørende  overvåkningsutstyr  for  trykk  og  temperatur  ekskludert  fra  
analysen.  

En  pålitelighetsanalyse  er  utført  for  å  vurdere  treet.  Analysen  er  utført  i  følgende  
trinn:  

1. Feilanalyse  (med  feil,  modus,  effekt  og  kritikalitetsanalyse  (FMECA))  

  V  
  
 
2. Pålitelighetsanalyse  (med  pålitelighetsblokkdiagram)  

Feilanalysen  er  utført  med  en  feil,  modus,  effekt  og  kritikalitetsanalyse  på  
komponentnivå.  Denne  er  gjennomført  for  å  utvikle  en  forståelse  av  
hovedkomponenter  med  essensielle  funksjonelle  krav  og  kritikalitet,  og  videre  
virkning  som  følge  av  funksjonssvikt.  Pålitelighetsanalysen  er  gjennomført  med  
rådata  fra  den  kommersielt  tilgjengelige  OREDA  håndboken.  Resultatene  fra  
denne  analysen  predikterer  en  trekkerate  på  treet  til  å  være  lik  22  år.    

Gjennom  de  utførte  pålitelighetsanalysene  har  treet  bekreftet  sitt  omdømme  

som  en  pålitelig  konfigurasjon  med  høy  driftssikkerhet  og  en  tilhørende  lav  
risiko.  Fokuset  på  denne  oppgaven  er  ikke  rettet  mot  et  absolutt  resultat  på  
pålitelighet  av  treet,  men  er  ment  å  illustrere  det  GE  Oil  &  Gas  oppfatter  som  et  
pålitelighetsnøyaktighetsproblem  ved  beregning  av  pålitelighet  basert  på  
generiske  pålitelighetstall  kontra  erfaringsdata.  OREDA-­‐2009  håndboken  er  
bevisst  brukt  som  eneste  kilde  til  rådata  for  å  illustrere  dette  problemet,  siden  
håndboken  er  kjent  for  å  gi  konservative  resultater  når  beregninger  er  utført  
utelukkende  basert  på  den.  Det  er  naturlig  og  kan  godt  sees  som  ønskelig  at  
beregninger  er  mer  forbeholdne  enn  real  situasjonen,  men  i  denne  konteksten  er  
det  et  naturlig  spørsmål  å  stille  seg  i  hvilken  grad.  

Det  er  påvist  et  betydelig  gap  mellom  den  beregnete  og  den  erfaringsbaserte  
trekkeraten  på  treet.  Basert  på  samlet  felterfaring  presentert,  er  det  indikert  et  
estimat  på  tid  til  feil  for  trekkerate  mellom  100  til  200  år.  Dette  innebærer  en  
faktor  på  5  til  10  mellom  den  beregnete  og  den  erfaringsbaserte  tilnærmingen.  
Det  bør  bemerkes  at  anslagene  for  trekkerate  er  ment  for  en  populasjon  av  trær  i  
drift.  Et  tre  vil  selvsagt  ikke  kunne  operere  i  200  år.    

Videre  er  det  gitt  av  GE  Oil  &  Gas  at  en  Pareto-­‐regel  tilsynelatende  gjelder  når  en  
bestemmer  om  en  potensiell  feil  krever  tung  brønnoverhaling  i  form  av  å  trekke  
treet  eller  en  lettere  intervensjon  med    fjernstyrt  maskineri  (ROV)  på  stedet.  Hvis  
Pareto-­‐regelen  gjelder,  vil  den  beregnede  totale  tid  til  feil  på  det  beregnede  treet  
lik  15  år  resultere  i  at  80  %  av  kritiske  feil  på  treet  kan  repareres  ved  lettere  
intervensjon,  mens  de  resterende  20  %  vil  kreve  trekking  av  treet.  Dette  vil  da  
reflektere  en  trekkerate  på  treet  lik  75  år.  Dette  er  nærmere  forventingene  fra  de  
erfaringsbaserte  tallene  presentert,  men  fortsatt  ikke  i  nærheten  av  nivåene  
angitt  av  nyere  felterfaring.    

Dette  tyder  på  at  konvensjonelle  data  bør  kalibreres  med  erfaringsdata  for  å  
minske  avstanden  mellom  de  to  tilnærmingene.  Dette  kan  utføres  i  form  av  
modellparametere,  modifiseringsfaktorer  eller  andre  forbedringer.  Resultatene  i  
denne  oppgaven  gir  ingen  klare  svare  på  hva  slike  modellparametere  bør  være.  
Pareto-­‐regelen  kan  muligens  benyttes  for  å  oppnå  mer  realistiske  beregninger,  
dersom  en  slik  omregning  finner  støtte  i  omfattende  historiske  data.  Dette  
kreves  det  i  så  fall  grundig  videre  forskning  på.    

  VI  
LIST  OF  ABBREVIATIONS  
XT   Xmas  Tree  
DVXT   Deepwater  Vertical  Xmas  Tree  
FMECA   Failure  Mode,  Effect  and  Criticality  Analysis  
DHSV   Downhole  Safety  Valve  
MTTF   Mean  Time  To  Failure  
RBD   Reliability  Block  Diagram  
FMEA   Failure  Mode  and  Effect  Analysis  
FTA   Fault  Tree  Analysis  
MV   Master  Valve  
TH   Tubing  Hanger  
SCSSV   Surface-­‐Controlled  Subsurface  Valve  
ROV   Remotely  Operated  Vehicle  
SCM   Subsea  Control  Module  
RAM   Reliability,  Availability  and  Maintainability  
fpmh   Failure  per  million  hours  
PWV   Production  Wing  Valve  
HXT   Horizontal  Xmas  Tree  
VXT   Vertical  Xmas  Tree  
ISO   International  Organization  for  Standardization  
BOP   Blow  Out  Preventer  
LRP   Lower  Riser  Package  
EDP   Emergency  Disconnect  Package  
THS   Tubing  Head  Spool  
FCM   Flow  Control  Module  
WGFM   Wet  Gas  Flow  Meter  
MPFM   Multiphase  Flow  Meter  
SCMMB   Subsea  Control  Module  Mounted  Base  
IWOCS   Installation  and  Workover  Control  System  
PGB   Production  Guide  Base  
THRT   Tubing  Hanger  Running  Tool  
MEG   Mono  Ethylene  Glycol  
API   American  Petroleum  Institute  
FAT   Factory  Acceptance  Test  
EFAT   Extended  Factory  Acceptance  Test  
SRT   Site  Receipt  Test  
PTT   Pressure  Temperature  Transmitter  
PCV   Production  Choke  Valve  
PMV   Production  Master  Valve  
AMV   Annulus  Master  Valve  
AWV   Annulus  Wing  Valve  
CCV   Chemical  Control  Valve  
AIV   Annulus  Isolation  Valve  
AVV   Annulus  Vent  Valve  
XOV   Crossover  Valve  

  VII  
  
 
MIV   Mono  Ethylene  Glycol  Isolation  Valve  
CIV   Chemical  Isolation  Valve  
WV   Wing  Valve  
PSV   Production  Swab  Valve  
ASV   Annulus  Swab  Valve  
CV   Check  Valve  
IV   Isolation  Valve  
WOCS   Workover  Control  System  
MODU   Mobile  Drilling  Unit  
LIV   Light  Intervention  Vessel  
ROVSV   Remotely  Operated  Vehicle  Service  Vessel  
MSV   Multipurpose  Service  Vessel  
O   Productivity  impact  
E   Environment  impact  
S   Safety  for  human  life  and  health  
WH   Wellhead  
LDHI   Low  Dosage  Hydrate  Inhibitor  
SI   Scale  Inhibitor  
HPMEG   High  Pressure  Mono  Ethylene  Glycol  
 

  VIII  
TABLE  OF  CONTEXT  
PREFACE  ................................................................................................................................  I  
EXECUTIVE  SUMMARY  ...................................................................................................  III  
SAMMENDRAG  ....................................................................................................................  V  
LIST  OF  ABBREVIATIONS  .............................................................................................  VII  
TABLE  OF  CONTEXT  ........................................................................................................  IX  
LIST  OF  FIGURES  .............................................................................................................  XII  
LIST  OF  TABLES  .............................................................................................................  XIII  
1   INTRODUCTION  ...........................................................................................................  1  
1.1   BACKGROUND  ............................................................................................................................  1  
1.2   OBJECTIVES  ................................................................................................................................  2  
1.3   SCOPE  AND  LIMITATIONS  .........................................................................................................  3  
1.4   RESEARCH  APPROACH  ..............................................................................................................  3  
2   THEORY,  METHOD  AND  LITERATURE  REVIEW  ................................................  5  
2.1   RELIABILITY  METHOD  ..............................................................................................................  5  
2.1.1   Failure  Mode  and  Effect  Criticality  Analysis  ........................................................  5  
2.1.2   Reliability  Block  Diagram  ............................................................................................  6  
2.2   RELIABILITY  DATA  ...................................................................................................................  6  
2.2.1   Qualification  and  Application  of  Reliability  Data  .............................................  6  
2.2.2   Failure  Rate  ........................................................................................................................  7  
2.2.3   OREDA  Offshore  Reliability  Data  Handbook  ........................................................  8  
3   OVERVIEW  OF  SUBSEA  XMAS  TREE  SYSTEMS  AND  STATISTICAL  REVIEW    
3.1   INDUSTRY  REQUIREMENTS  ...................................................................................................  11  
3.2   ACCIDENT  AND  STATISTICAL  REVIEW  ................................................................................  12  
3.2.1   Accident  Review  .............................................................................................................  12  
3.2.2   Statistical  Review  ..........................................................................................................  14  
3.2.3   Review  of  XT  Field  Data  Performed  in  1999  .....................................................  15  
3.2.4   Review  of  XT  Field  Data  performed  in  2014  .....................................................  17  
3.3   XMAS  TREE  FUNCTIONAL  REQUIREMENTS  ........................................................................  19  
3.4   TYPES  AND  CONFIGURATIONS  OF  XMAS  TREES  .................................................................  19  
3.4.1   Vertical  Xmas  Tree  .......................................................................................................  19  
3.4.2   Horizontal  Xmas  Tree  .................................................................................................  20  
3.4.3   Comparison  of  Horizontal  and  Vertical  Xmas  Trees  .....................................  22  
3.4.4   Selection  Criteria  ...........................................................................................................  23  
3.5   MAIN  COMPONENTS  OF  A  XMAS  TREE  SYSTEM  .................................................................  24  
3.5.1   General  Components  of  a  XT  System  ....................................................................  24  
3.5.2   Tubing  Hanger  ...............................................................................................................  24  
3.5.3   Wellhead  Connectors  ...................................................................................................  26  

  IX  
  
 
3.5.4   Valves  .................................................................................................................................  27  
3.5.5   Flow  Control  Module  ...................................................................................................  28  
3.5.6   Main  Components  That  Vary  Between  HXTs  and  VXTs  ................................  28  
3.5.7   Xmas  Tree-­‐Mounted  Controls  ..................................................................................  29  
3.6   XMAS  TREE  INSTALLATION  AND  SERVICE  CONDITIONS  ...................................................  29  
3.6.1   Xmas  Tree  Installation  ...............................................................................................  29  
3.6.2   Service  Conditions  .........................................................................................................  31  
3.7   XMAS  TREE  DESIGN  AND  ANALYSIS  ....................................................................................  31  
3.8   TEST  PROGRAM  FOR  XMAS  TREES  ......................................................................................  32  
4   CASE  STUDY:  DEEPWATER  VERTICAL  XMAS  TREE  ......................................  33  
4.1   DESCRIPTION  OF  A  DEEPWATER  VERTICAL  XMAS  TREE  .................................................  33  
4.2   BASE  CASE  DEFINITION  ........................................................................................................  34  
4.3   BOUNDARY  DEFINITION  ........................................................................................................  35  
4.4   KEY  SYSTEM  ELEMENTS  AND  CHARACTERISTICS  ..............................................................  36  
4.4.1   General  ..............................................................................................................................  36  
4.4.2   Tubing  Head  Spool  .......................................................................................................  37  
4.4.3   Tubing  Hanger  ...............................................................................................................  38  
4.4.4   Production  Xmas  Tree  ................................................................................................  39  
4.4.5   Flow  Control  Module  ...................................................................................................  41  
4.4.6   Xmas  Tree  Installation  and  Workover  Control  System  ................................  42  
4.5   MAIN  ASSUMPTIONS  AND  LIMITATIONS  .............................................................................  42  
4.5.1   Analysis  Level  ..................................................................................................................  42  
4.5.2   Operational  Phases  ......................................................................................................  42  
4.5.3   Analysis  Assumptions  ..................................................................................................  42  
5   RELIABILITY  ANALYSIS  APPROACH  APPLIED  TO  CASE  ..............................  44  
5.1   XT  FAILURES  ..........................................................................................................................  44  
5.2   WELL  INTERVENTION  MEANS  .............................................................................................  45  
5.2.1   Heavy  workover  .............................................................................................................  45  
5.2.2   Light  Intervention  .........................................................................................................  46  
5.3   FAILURE  CRITICALITY  CLASSIFICATION  ..............................................................................  46  
5.4   FAILURE  MODE  EFFECT  AND  CRITICALITY  ANALYSIS  ......................................................  47  
5.4.1   Risk  Evaluation  of  the  Components  ......................................................................  49  
5.4.2   Identification  of  Component  Failures  ..................................................................  51  
5.5   RELIABILITY  BLOCK  DIAGRAM  .............................................................................................  54  
6   RESULTS  ......................................................................................................................  57  
6.1   FMECA/FAILURE  ANALYSIS  ...............................................................................................  57  
6.1.1   Failure  Criticality  Assessment  of  Components  and  Sub-­‐Systems  .............  57  
6.1.2   FMECA  ................................................................................................................................  59  
6.2   RBD/RELIABILITY  ANALYSIS  ..............................................................................................  64  
6.3   COMPARISON  OF  BOTTOM-­‐UP  AND  TOP-­‐DOWN  RESULTS  ..............................................  66  
7   DISCUSSION  ................................................................................................................  69  

  X  
 7.1   DISCUSSION  OF  RESULTS  .......................................................................................................  69  
7.2   DISCUSSION  OF  THE  FRAMEWORK  USED  TO  OBTAIN  THE  RESULTS  ...............................  70  
7.2.1   Simplification  of  the  System  .....................................................................................  72  
7.2.2   Simplifications  and  Weaknesses  of  the  FMECA  ...............................................  73  
7.2.3   Simplifications  and  Weaknesses  of  the  RBD  .....................................................  73  
8   CONCLUSION  AND  FURTHER  WORK  ..................................................................  75  
8.1   SUMMARY  AND  CONCLUSION  ................................................................................................  75  
8.2   FURTHER  WORK  ....................................................................................................................  76  
BIBLIOGRAPHY  ................................................................................................................  78  
APPENDICES  .....................................................................................................................  79  
A.   RELIABILITY  DATA  ................................................................................................  81  
B.   MOBILIZATION  AND  REPAIR  TIME  ...................................................................  85  
C.   FAILURE  MODE  AND  EFFECT  CRITICALITY  ANALYSIS  ................................  87  
D.   RELIABILITY  BLOCK  DIAGRAMS  .....................................................................  113  
D.1   PRODUCTION  XMAS  TREE  ..................................................................................................  113  
D.2   FLOW  CONTROL  MODULE  ..................................................................................................  116  
   

  XI  
  
 

LIST  OF  FIGURES  

 

FIGURE  1  THE  BATHTUB  (LIFE)  CURVE  (RAUSAND  &  HØYLAND,  2004)  .......................................................................  7  
FIGURE  2  GENERAL  VXT  CONFIGURATION  ((ENI),  ET  AL.,  2012)  ...............................................................................  20  
FIGURE  3  GENERAL  HXT  CONFIGURATION  ((ENI),  ET  AL.,  2012)  ..............................................................................  21  
FIGURE  4  MONOBORE  AND  DUAL  BORE  TUBING  HANGER  (BAI  &  BAI,  2012)  .............................................................  25  
FIGURE  6  H4-­‐CONNECTOR  (BAI  &  BAI,  2012)  .................................................................................................................  27  
FIGURE  7:  DIFFERENCE  BETWEEN  HXT  AND  VXT  (RED  DOTS  ILLUSTRATE  THE  LOCATION  OF  VALVES)  (BAI  &  BAI,  
2012)  ...........................................................................................................................................................................  28  
FIGURE  8  VXT  COMPLETED  ON  A  THS  (COURTESY  OF  GE  OIL  &  GAS)  .............................................................................  33  
FIGURE  9:  SCHEMATIC  DESCRIPTION  OF  A  VXT  COMPLETED  ON  A  THS  (DUALBORE  CONFIGURATION  TO  THE  LEFT  
AND  MONOBORE  TO  THE  RIGHT)  (ISO  13628-­‐4,  2010)  ....................................................................................  34  
FIGURE  10  DUAL  BORE  CONFIGURATION  TH  INSTALLED  WITHIN  THE  THS  ((ENI),  ET  AL.,  2012)  .......................  39  
FIGURE  11  ILLUSTRATION  OF  INTERVENTION  MEANS  FOR  THE  PRODUCTION  XT  ......................................................  64  
FIGURE  12  FAILURE  DISTRIBUTION  IN  THE  PRODUCTION  XT  .........................................................................................  65  
FIGURE  13  FAILURE  DISTRIBUTION  IN  THE  FCM  .............................................................................................................  66  
 

   

  XII  
LIST  OF  TABLES  
TABLE  1  BLOWOUT  DURING  PRODUCTION  (1980  -­‐  2007)  (MOLNES,  2012)  ...........................................................  13  
TABLE  2:  KEY  TECHNICAL  DESIGN  DATA  .............................................................................................................................  36  
TABLE  3  FAILURE  CRITICALITY  CLASSIFICATION  (PROVIDED  BY  GE  OIL  &  GAS)  .......................................................  47  
TABLE  4  SEVERITY  MATRIX  (PROVIDED  BY  GE  OIL  &  GAS)  ...........................................................................................  49  
TABLE  5  PROBABILITY  MATRIX  (PROVIDED  BY  GE  OIL  &  GAS)  ....................................................................................  50  
TABLE  6  DEFINITION  OF  RISK  CRITICALITY  LEVEL  (PROVIDED  BY  GE  OIL  &  GAS)  ...................................................  50  
TABLE  7  FAILURE  MODES  IN  THE  COMPONENT-­‐LEVEL  FMECA  ...................................................................................  51  
TABLE  8  FUNCTIONAL  CRITICALITY  ASSESSMENT  –  GENERAL  FUNCTIONS  .................................................................  57  
TABLE  9  CRITICALITY  ASSUMPTIONS  FOR  MAIN  COMPONENTS  .....................................................................................  58  
TABLE  10  CRITICALITY  ASSUMPTIONS  FOR  MAIN  VALVES  .............................................................................................  58  
TABLE  11  TOTAL  RISK  ..........................................................................................................................................................  59  
TABLE  12  EFFECTS  ON  OPERATIONAL  RISK  ......................................................................................................................  60  
TABLE  13  ENVIRONMENTAL  RISK  .......................................................................................................................................  61  
TABLE  14  EFFECTS  ON  HUMAN  RISK  ..................................................................................................................................  62  
TABLE  15  EXPOSED  COMPONENTS  FOR  FURTHER  EVALUATION  ....................................................................................  63  
TABLE  16  RELIABILITY  OF  THE  PRODUCTION  XT  .............................................................................................................  64  
TABLE  17  RELIABILITY  OF  THE  FCM  ..................................................................................................................................  65  
TABLE  18  FAILURE  DATA  FOR  THE  WELLHEAD  SYSTEM  .................................................................................................  81  
TABLE  19  FAILURE  DATA  FOR  THE  CONNECTORS  .............................................................................................................  81  
TABLE  20  FAILURE  DATA  THE  TUBING  HEAD  SPOOL  FRAME  AND  FLOWLOOPS  ..........................................................  81  
TABLE  21  FAILURE  DATA  FOR  THE  TUBING  HANGER  ......................................................................................................  82  
TABLE  22  FAILURE  DATA  FOR  TREE  CAP  ...........................................................................................................................  82  
TABLE  23  FAILURE  DATA  FOR  THE  PRODUCTION  XT  ......................................................................................................  83  
TABLE  24  FAILURE  DATA  FOR  THE  MAIN  VALVES  ...........................................................................................................  83  
TABLE  25  FAILURE  DATA  FOR  THE  CHOKE  VALVES  .........................................................................................................  84  
TABLE  26  INTERVENTION  VESSELS  WITH  MOBILIZATION  DATA  ...................................................................................  85  
TABLE  27  REPAIR  TIME  ........................................................................................................................................................  86  

  XIII  
     Introduction  

 
1 INTRODUCTION  
 

1.1 BACKGROUND  
As  offshore  technologies  have  advanced,  more  and  more  of  the  operations  
previously  performed  on  the  surface  are  moved  down  to  the  seabed.  Today’s  
subsea  technology  encircles  a  wide  range  of  subsea  components,  i.e.  XTs,  
manifolds,  risers,  templates,  flowlines,  ROVs,  hydraulic  and  electric  power  
systems,  control  systems,  fluid  pumping,  reinjection  and  separation.    

The  first  subsea  XT  was  installed  in  the  Gulf  of  Mexico  in  1961.  Since  then,  the  
XTs  are  an  essential  part  of  the  subsea  fields.  Now  there  are  a  number  of  XT  
variations  with  technology  modified  to  fit  each  unique  well.  Tree  complexity  and  
functionality  has  increased  over  the  last  few  decades.  

The  XT  is  recognized  by  the  industry  as  an  overall  reliable  configuration.  With  
the  demand  for  production  of  hydrocarbons  from  deeper  water  with  higher  
pressures  and  temperatures,  the  XTs  meet  additional  design  constraints.  These  
keep  pushing  the  technology  to  evolve  to  meet  the  challenges  coming  both  now  
and  in  the  future.  The  ambient  seabed  conditions  and  continuously  increasing  
intervention  cost  require  a  higher  standard  on  the  equipment  and  keep  pushing  
the  technology  development.    

As  in  all  development  projects,  it  is  a  constant  battle  to  optimize  the  system  
design  in  a  life  cycle  perspective  taking  account  of  various  aspects  such  as  safety,  
availability,  maintainability  and  reliability.  In  subsea  development  projects,  
reliability  and  availability  performance  targets  are  normally  part  of  the  
contractual  requirements.    

Generally,  unplanned  stoppage  of  equipment  result  in  high  equipment  

downtime,  high  cost  of  repair,  extensive  repair  time  and  high  penalty  associated  
with  loss  of  production.    

In  order  to  optimize  the  tree  system  design  with  respect  to  safety,  availability,  
maintainability  and  reliability  performance,  there  is  a  need  for  proper  reliability  
models  that  are  able  to  differentiate  between  different  options  and  variants.    

There  are  two  main  approaches  for  predicting  reliability  of  XTs:  

• Bottom-­‐up:   Reliability   modeling   of   system,   component   by   component,  

using  component  reliability  data  available  in  industry  recognized  sources,  
such  as  the  OREDA  handbook.  

  1  
Chapter  1      
 
• Top-­‐down:  High-­‐level  reliability  data  for  systems,  captured  from  actual  
field  experience/operations,  typically  reflecting  the  reliability  
performance  as  seen  from  the  operators  point  of  view.  

The  two  approaches  results  in  totally  different  results.  For  instance,  an  attempt  
to  model  a  conventional  tree  using  component  data  from  OREDA  may  typically  
predict  a  retrieval  rate  of  XTs  in  the  order  of  once  per  15-­‐20  years,  as  opposed  to  
once  per  150-­‐200  years  estimated  MTBF  based  on  experience  implied  by  the  
operators/customers.    

The  idea  is  that  the  bottom-­‐up  model  should  be  calibrated  with  input  data  that  is  
experiential  rather  than  generic  to  alleviate  some  of  the  distance  between  the  
two  approaches.  Although  GE  Oil  &  Gas  have  a  good  understanding  of  the  overall  
reliability  and  availability  performance  as  seen  from  the  customers,  there  is  an  
unrealized  potential  regarding  a  structured  data-­‐collecting  and  organizing  tool  of  
actual  field  experience  for  their  Xmas  trees  (number  of  trees  installed,  years  in  
operation,  number  of  failures,  corresponding  repair  activity  etc.).  In  its  absence,  
the  generic  and  therefore  misleading  retrieval  rate  are  emphasized  to  a  higher  
degree  than  one  might  wish.  

1.2 OBJECTIVES  
In  this  master  thesis,  the  main  objective  is  to  study  a  specific  Xmas  Tree  (XT)  
system  and  to  estimate  the  retrieval  rate  due  to  tree  failures  based  on  
commercially  available  reliability  data.  Further  this  shall  be  compared  to  high  
level  experience  data  presented.  This  is  to  initiate  the  process  to  alleviate  the  gap  
seen  between  generic  calculations  of  the  tree  retrieval  rate  compared  to  known  
field  experience.  The  elected  tree  system  to  study  is  the  Deepwater  Vertical  Xmas  
Tree  (DVXT).  

More  specifically,  the  objectives  are  too:  

1) Perform  a  literature  survey  and,  on  the  basis  of  this  survey,  describe:  
a) Main  types  of  Xmas  trees,  and  main  Xmas  tree  sub-­‐systems  and  
components  
b) Describe  the  most  important  differentiating  factors  with  respect  to  
reliability  and  availability  performance  of  Xmas  Tree  Systems  (including  
installation/intervention  issued  and  downhole  operations).  
c) Essential  terms,  definitions  and  industry  standards  for  performing  
probabilistic  analysis  of  subsea  tree  systems,  and  describe  main  
methodologies  
d) Relevant  reliability  data  sources,  with  emphasize  on  limitations  and  
applicability  in  relation  to  the  current  topic.  
 
2) Conduct  a  component-­‐level  FMECA  of  a  DVXT  configuration  (excluding  
control  module,  see  Chapter  1.3).  

  2  
     Introduction  

a) Develop  an  understanding  of  main  components  with  essential  functional  

requirements  and  criticality  and  effect  resulting  from  functional  failures  
 
3) Conduct  a  Probabilistic  Reliability  Analysis  of  the  Deepwater  Vertical  Xmas  
Tree  System  
a) Bottom-­‐up  approach:  
i) Based  on  reliability  data  sources,  develop  an  Xmas  Tree  specific  
database  containing  component  reliability  data  for  Xmas  Tree  
components/items.  
ii) Use  this  to  develop  a  first  pass  reliability  model  for  the  Xmas  Tree  
System  
b) Top-­‐down  approach:  
i) View  statistics  on  the  high  level  reliability  and  availability  
performance  of  subsea  trees.  
ii) Seek  literature  for  relevant  comparative  information.  
 
4) Comparative  assessment:  Derived  from  a  comparative  assessment  of  results,  
recommend  a  baseline  set  of  component  reliability  data,  additional  model  
parameters,  modifications  factors  and  other  refinements  as  required  for  
calibration  of  the  bottom-­‐up  model  

1.3 SCOPE  AND  LIMITATIONS  

The  scope  is  limited  in  this  context  to  the  XT  system  and  systems  that  influence  
the  XT  system  in  terms  of  XT  retrieval  rate  and  downtime  due  to  XT  failures.  
Therefor,  the  control  systems  with  associated  monitoring  equipment  are  
excluded  from  the  analysis  conducted.    

1.4 RESEARCH  APPROACH  

The  work  has  mainly  consisted  of  two  activities;    

1. Acquiring  through  relevant  theory  with  regard  to  reliability  data  

(commercially  available  data  and  field  experience  data),  applicable  
standards,  reliability  theory  and  XT  accidents.  
2. Gain  detailed  knowledge  about  the  XT  system  and  variations  within  to  be  
able  to  perform  a  proper  reliability  assessment  followed  by  analysis  of  the  
system  through  proven  methods  from  the  reliability  engineering  
discipline.    

A  component-­‐level  FMECA  have  been  conducted  to  evaluate  failure  modes  

relevant  for  the  Deepwater  Vertical  Xmas  Tree  (DVXT)  in  order  to  estimate  the  
Mean  Time  To  Failure  (MTTF)  through  Reliability  Block  Diagram’s  (RBD’s).  The  
focus  is  directed  to  the  retrieval  rate  of  trees,  found  through  the  RBD’s.  This  
retrieval  rate  is  compared  to  field  experience  to  initiate  the  process  to  find  a  
sound  strategy  to  alleviate  the  distance  between  the  bottom-­‐up  and  top-­‐down  
approach.    

  3  
     Theory,  Method  and  Literature  Review  

2 THEORY,  METHOD  AND  LITERATURE  REVIEW  

In  this  chapter  relevant  theory,  method  and  literature  review  are  presented.  The  
literature  review  executed  on  tree  variations  and  its  components  are  presented  
in  Chapter  3.    

2.1 RELIABILITY  METHOD  

2.1.1 FAILURE  MODE  AND  EFFECT  CRITICALITY  ANALYSIS    

A  Failure  Mode  and  Effect  Analysis  (FMEA)  is  normally  the  starting  point  of  a  
systems  reliability  study  and  is  used  as  a  systematic  technique  to  identify  and  
evaluate  potential  failure  modes  in  a  system  and  the  further  effects  these  failures  
may  have  on  the  system.  An  FMEA  becomes  a  Failure  Mode  and  Effect  Criticality  
Analysis  (FMECA)  if  the  failure  modes  are  ranked  in  terms  of  criticality.  A  FMECA  
is  usually  carried  out  during  the  design  phase  to  identify  possible  weaknesses  so  
that  corrections  and  potential  extra  barriers  might  be  added  at  a  relatively  early  
stage  in  a  project.  FMECA  is  also  used  for  maintenance  planning  and  as  a  basis  
for  more  detailed  reliability  analysis.  

The  main  objective  of  a  FMECA  is  to  identify  areas  where  improvements  are  
needed  to  meet  safety  and  reliability  requirements.  This  is  achieved  through  
systematic  assessment  of  the  likelihood  that  faults  might  occur  and  the  severity  
of  the  potential  consequences.  Main  elements  of  the  system  and  their  
relationship,  such  as  series,  redundancy  and  the  like,  shall  be  identified  as  part  of  
the  analysis.  Based  on  the  output  form  the  FMECA,  effects  on  system  availability  
and  maintenance  planning  may  be  established.  A  FMECA  can  provide  a  basis  for  
a  more  detailed,  tailored  reliability  analysis.    

Largely,  a  FMECA  can  be  divided  into  several  levels  of  accuracy:  The  following  
breakdown  structure  is  used  by  GE  Oil  &  Gas  for  subsea  application:  

i. Level  1:  System-­‐level  FMECA  with  project/application  specific  

conditions  and  requirements  applied.  
ii. Level  2:  Subsystem-­‐level  FMECA,  also  with  project/application  
specific  conditions  and  requirements.    
iii. Level  3:  Component-­‐level  FMECA;  with  generic  component  
conditions  and  functional  requirements  applied.    

A  FMECA  can  be  conducted  both  using  bottom-­‐up  and  top-­‐down  approach.  In  a  
bottom-­‐up  approach  one  evaluates  the  system  by  identifying  all  potential  failure  
modes  on  a  component  level  and  precede  upwards  in  the  hierarchy.  This  is  
distinct  from  the  top-­‐down  approach,  where  the  analysis  is  carried  out  by  
splitting  the  system  into  a  number  of  subsystems  and  then  identifying  possible  

  5