Some Observations on Data Management in Shell

1
Some Observations on Data Management in Shell
By Ax Hesterman and Tom Gulman SIEP EPT-AM
1
2
Outline

• Integration
• Business - the maturing world of assets and
  processes
• Technology - the virtual world of E P software
• Business drivers
• Understanding the work flow
• Technology drivers
• Understanding the information flow
• State of Play
• Current
• Vision
• Implementation Experience

2
3
EP realities
3
4
Integration vision
Analogy flying an aircraft - by hand - by
computer (shift over time and during flight)
5
5
Business drivers

• Increase profits and market share
• Reduce cycle time and cost
• Improve safety and environmental impact

Business response

• Target setting
• Asset management
• Process management

6
6
Generic business model
9
7
Technology drivers

• Object orientation - component software
• Networks - WWW-Internet / Intranets
• Lower (hardware) cost
• Data management tools

Business response

• Business reflected in software (virtual
  processes)
• Improved accessibility and sharing
• More data of higher quality

11
8
Business model in softwareEP virtual processes
12
9
Data transfer between applications"virtual
reality"
Analogy - European railway systems
13
10
Developing trend
Integration backbone A
Integration backbone B
Integration backbone C
14
11
Option 1
15
12
Option 2
16
13
Option 3
17
14
Reality?
18
15
Effort of building data bases

• Depends on span of data base over business area
• Increasing complexity - increasing cost
• If span of data coverage is small
• Several data bases needed
• Multiple interfaces needed
• Data management complex (e.g. naming conventions)
• If span of data coverage is large
• Complex data model
• Difficult to construct
• Difficult to adapt
• Agreement across business more difficult

19
16
Requirements

• Understanding of own business processes
• Work flow determines information flow
• Flexibility of data bases, applications and links
• To mimic the business in software (virtual
  processes)
• Understanding of information needs
• Standards for data storage (data model), data
  exchange (links) and nomenclature (naming
  conventions)
• Tools for data modelling and data mapping
• Data management guidelines
• Tools procedures to support data management
  (e.g. security, sharing, quality checking (e.g.
  inconsistent names))

20
17
Current view of market

• Multiple, but reducing, number of vendors
• Development of mega-applications by the major
  vendors
• Development of vendor applications and associated
  (project) data bases"INTEGRATION BACKBONES" (
  infrastructure)
• "Pearls" from small vendors and own developments
  by oil companies
• Infrastructure issues clashes
• Data connectivity
• GUI
• Data integration and application integration not
  aligned

21
18
Long term vision

• Supported Vendor-neutral plug-and-play, based
  on industry standards
• Industry standards
• De jure POSC - STEP
• De facto Vendors
• Unknown Unexpected
• Shell as member of POSC
• Will shift focus more to application of standards
• Away from development of standards
• Will work more with vendors
• Emphasise implementation of standards in products
• Not supported Vendor suite integration, just
  to reduce complexity of take-over products

22
19
Short term

• Long term view as guidance
• Integration, as far as possible, NOW
• Based on current commercial products
• Prioritisation for future developments
• Business processes
• Information needs
• CRITERIA for evaluating vendor backbones
• Strategic partnership(s)
• Vendor neutral
• Not necessarily with the traditional vendors

24
20
Implementation Experiences

• Centrally supported implementation of IBM's PDS
  environment
• Initial focus field development phase
• Moving towards production optimisation
• Loose confederation of Operating Units
  implementing Landmark's Openworks
• covers the acquisition phase
• Centrally supported development of a corporate
  data warehouse for EP technical data using
  Epicentre
• Some individual companies doing own thing
• One based on Finder
• One development using PPDM data model
• One data repository development since 1989 that
  continues to expand

21
Observations (1)

• One size fits none
• Each operating unit chooses development best
  suited for its environment, size, maturity, and
  future prospects
• No solution covers the entire EP life cycle today
• Multiple implementations a reality in many
  operating units
• More integration activity at the early part of
  the life cycle than later
• GG area is better covered than operations, well
  optimisation
• SAP is another intergration backbone
• Will expand into many service functions
• SAP will become a source of data
• e.g. procurement, materials, cost data needed for
  economic analysis
• Existing systems will need to feed SAP

22
Observations (2)

• Evolution, not revolution, is the order of the
  day
• Integration technology deployment has to link
  with working practices
• Existing projects are just a step in a longer
  process
• Integration projects are evolving as business
  processes evolve
• The target environment is years and several
  generations away
• Data integration and applications integration
  need a push to converge
• We need proactive oil company involvement
• Integration not the same for suppliers and OilCos
• Oil companies want flexibility to choose best of
  class applications and be able to swap out one
  for another
• Suppliers focus on their own stable of databases
  and applications
• Suite lock-in is real

23
Recommendations

• Be an active supporter of industry standards
• Shell supports POSC STEP Data access
  (LightSIP) Data exchange (SMDTI) Interoperab
  ility Business Objects Application
  connectivity (OpenSpirit)
• Get your data management strategy agreed at the
  highest levels
• Have a stated policy in your company
• Get it accepted within the rank and file
• Link support of standards to real business needs
• Use commercial products, but challenge suppliers
• Clarify what your supplier means by "open"
• Measure supplier by his conformance to your
  definition
• Test openness whenever possible
• Challenge suppliers to adopt and conform to
  standards as early as is feasible