data modelling and witsml

59
1 Use this layout for text on top of a horizontally striped picture. Visualising WITSML using Entity-Relationship Models Christopher Bradley Chief Development Officer ECIM 2012 Haugesund, Norway

Upload: christopher-bradley

Post on 04-Dec-2014

1.464 views

Category:

Technology


16 download

DESCRIPTION

Visualising WITSML XML Data Structures in Data Models. ECIM E&P conference, Haugesund Norway, September 2012

TRANSCRIPT

Page 1: Data Modelling and WITSML

1

Use this layout for text on top of a horizontally striped picture.

Visualising WITSML using Entity-Relationship Models

Christopher BradleyChief Development OfficerECIM 2012

Haugesund, Norway

Page 2: Data Modelling and WITSML

2

Use this layout for a title with a vertically striped picture.

PresenterChris BradleyChief Development [email protected]+44 1225 475000

My blog: Information Management, Life & Petrolhttp://infomanagementlifeandpetrol.blogspot.com

@InfoRacer

Page 3: Data Modelling and WITSML

3

Introductions

Christopher Bradley

Chief Development Officer

Chris has spent 32 years in the Information management field, working for leading organisations

in Data Management Strategy, Master Data Management,

Metadata Management, Data Warehouse and Business Intelligence.

Graduating in 1979 Chris worked for the MoD(Navy), Volvo, Thorn EMI (as Head of Information

Management), Readers Digest Inc (as European CIO), and Coopers and Lybrand Management

Consultancy where he established and ran the International Data Management practice.

Chris heads IPL’s Business Consultancy practice and is advising several Energy,

Pharmaceutical, Finance and Government clients on Business Process and Information Asset

Management.

Chris is a member of the MPO, Director of DAMA UK and holds the CDMP Master certification.

He co-authored

“Data Modelling For The Business – A Handbook for aligning the business with IT using high-level data models”

.

Chris is a columnist and frequent contributor to industry publications. He authors an

experts channel on the influential BeyeNETWORK, is a recognised thought-leader in Information

Management and regular key speaker at major International Information Management

conferences.

[email protected]+44 1225 475000

Blog: Information Management, Life & Petrolhttp://infomanagementlifeandpetrol.blogspot.com

@InfoRacer

Page 4: Data Modelling and WITSML

4

Use this layout for text on top of a vertically striped picture.

Who is IPL?Trusted, independent consulting & solutions co

• 30 year track record• 300 staff, £28m+ turnover• High-stakes, business & mission critical contexts• Consistently exceed expectations

Business Consulting Division• Information Management

• IM Strategy• Information Security & Assurance• Data Governance• Information Exploitation• Master Data Management• Information Architecture• Business Intelligence

• .......turning Information into a strategic asset• Enterprise Architecture• Business Process Management• Programme Management

Business Sponsor

IPL Consulting Clients

Page 5: Data Modelling and WITSML

5

Page 6: Data Modelling and WITSML

6

Page 7: Data Modelling and WITSML

7

Use this layout for a title with a horizontally striped picture.

I

Introduction & Agenda

Page 8: Data Modelling and WITSML

8

AgendaIntroductionsWhat is WITSML?How is WITSML defined?Making sense of WITSMLFrom XML to ERD: TechniqueUsage ScenariosThe Wider Landscape

Page 9: Data Modelling and WITSML

9

What is WITSML?Energistics is industry recognised body that provides a neutral environment for the development of common data exchange standardsWellsite Information Transfer Standard Markup Language is one such standard

XML-based markup languageFor the transfer of wellsite information

Page 10: Data Modelling and WITSML

10

Supported Data Objects 1.4.1

Page 11: Data Modelling and WITSML

11

Example: WITSML Wellbore TrajectoryThis message describes the trajectory of a wellbore (a unique, oriented path from the bottom of a drilled borehole to the surface of the Earth)StructuredBut a bit cryptic…

dTimTrajStartaziVertSect

For uninitiated, difficult see overall organisation… and it’s difficult to see the context.

XML Message

Q. Would you show this to a business person?

Page 12: Data Modelling and WITSML

12

Page 13: Data Modelling and WITSML

13

How is WITSML Defined?http://w3.energistics.org/schema/WITSML_v1.4.1.1_Data_Schema/witsml_v1.4.1.1_data/doc/witsml_schema_overview.htm

1. List of WITSML Data Objects

Page 14: Data Modelling and WITSML

14

How is WITSML Defined?

Click

http://w3.energistics.org/schema/WITSML_v1.4.1.1_Data_Schema/witsml_v1.4.1.1_data/doc/witsml_schema_overview.htm

1. List of WITSML Data Objects

Page 15: Data Modelling and WITSML

15

How is WITSML Defined?l The tree reflects

the hierarchical structure.Each XSD defines a type of XML element.

2. Tree of Element Types

(for a Data Object)

Page 16: Data Modelling and WITSML

16

How is WITSML Defined?l The tree reflects

the hierarchical structure.Each XSD defines a type of XML element.

Click

2. Tree of Element Types

(for a Data Object)

Page 17: Data Modelling and WITSML

17

How is WITSML Defined?Each XSD defines a type of element including

AttributesNested Elements

3. XSD Element Definition

Page 18: Data Modelling and WITSML

18

How is WITSML Defined?Each XSD defines a type of element including

AttributesNested Elements

Click

3. XSD Element Definition

Page 19: Data Modelling and WITSML

19

How is WITSML Defined?There can be many levels of nestingThe schema for a single data object spans many files

3a. XSD Element Definition

(Nested Element)

Q. Would you show this to a business person?

Page 20: Data Modelling and WITSML

20

How is WITSML Defined?There can be many levels of nestingThe schema for a single data object spans many files

3a. XSD Element Definition

(Nested Element)

Q. Would you show this to a business person?

Page 21: Data Modelling and WITSML

21

Page 22: Data Modelling and WITSML

22

Look at the real world example

Page 23: Data Modelling and WITSML

23

What is the problem?

Page 24: Data Modelling and WITSML

24

XML message fundamentalsXML is hierarchicalER models are relationalER models represent real world data

Page 25: Data Modelling and WITSML

25

XML implementation of ER modelAn XML schema generated from this model must choose one “parent”We could choose BOOK as the root, in which case WRITER would become a child of BOOK AUTHORSHIPWe could choose WRITER as the root, in which case BOOK would become a child of BOOK AUTHORSHIP

Page 26: Data Modelling and WITSML

26

Book

Constraints

Book ISBN code

Amazon URL

Book name

Category

Publication date

Publisher

Recommended price

Book Authorship

Constraints

Agreement id

Book ISBN code

Writer id

Writer

Royalty %

Draft delivery date

Profile delivery dateConstraints

Writer id

Writer name

Specialism

Affiliation

XML implementation of ER modelBook

Book Authorship Writer

Page 27: Data Modelling and WITSML

27

WITSML Logical Model Objectives

Digestible for business usersMeaningful names for entities and attributesAppropriate level of detail – hide “noise”Show appropriate logical relationships or business rules

Not just the tree structure

Easy to review

Page 28: Data Modelling and WITSML

28

WITSML Logical Model ObjectivesPrecise for IT users

Accurate reflection of the WITSML standardTraceable to the WITSML standard

DetailedDistinguish between physical and logical constructs

Normalised, but showing hierarchical structure of messagesRigorous, formal analysis and design processPrecise meaning of terms and symbols and rulesDefinitions support rigour

Page 29: Data Modelling and WITSML

29

WITSML Logical Model ObjectivesIncorporate within Enterprise Data Model (EDM)

Map objects to the relevant layer(s) in EDMLink enterprise level data assets through to WITSML Objects…more on this later

Baseline for data requirements analysis and data modelling efforts undertaken at the project levelReduce the time taken for impact analysisMinimise reworkPromote reuse.

Page 30: Data Modelling and WITSML

30

Making sense of WITSML

1. Submodel: WITSML Data

Objects

1. List of WITSML Data Objects

Page 31: Data Modelling and WITSML

31

Making sense of WITSML

1. Submodel: WITSML Data

Objects

1. List of WITSML Data Objects

Page 32: Data Modelling and WITSML

32

Making sense of WITSML

2. Submodel for a Data Object

2. Tree of Element Types

(for a Data Object)

Page 33: Data Modelling and WITSML

33

Making sense of WITSML

2. Submodel for a Data Object

2. Tree of Element Types

(for a Data Object)

Shows the tree structure (inside the box labelled “Data Object: Trajectory”)

Also shows the context (outside the box labelled “Data Object: Trajectory”)

Page 34: Data Modelling and WITSML

34

Making sense of WITSML

3. Entities

3. XSD Element Definition

Page 35: Data Modelling and WITSML

35

Making sense of WITSML

3. Entities

3. XSD Element Definition

Page 36: Data Modelling and WITSML

36

Making sense of WITSML

Attribute DefinitionAzimuth: Azimuth used for vertical section plot/computations.

Entity DefinitionTrajectory: A set of Trajectory Stations that describes the path of a section of a wellbore or of the entire wellbore.

Page 37: Data Modelling and WITSML

37

Can we turn XSDs into E/R models automatically? Yes and No!

Yes: Tools such as E/R Studio and PowerDesigner can create models by inspecting XSDs. They can…

Identify entities, attributes, and data typesImport definitions from <xsd:documentation> nodesInfer relationships based on nesting of element types.

But the human touch is needed too!

Page 38: Data Modelling and WITSML

38

Can we turn XSDs into E/R models automatically?

No: Manual effort is needed to…Create logical namesIdentify implied relationshipsNormalise / denormaliseClassify into subject areas and map to conceptual modelsLayout diagrams

Page 39: Data Modelling and WITSML

39

XML versus E/R StructuresXML

Hierarchical - tree structure. Each entity has just one parent.Used for transfer of data.Shared data appears multiple times in multiple messages.

E/R StructuresRelational - network structure.Each entity can have many parents.Used for storage and maintenance of data.Shared data typically appears just once.

Page 40: Data Modelling and WITSML

40

From XML to ERD: Techniques

Page 41: Data Modelling and WITSML

41

Using the Logical ModelScenario 1: Impact analysisProposal to allow multiple fluids to be specified in the schema for the cementJob object

What is the overall organisation of things the object describes?What is the impact on the business rules?Do the definitions still reflect the essence of the object that are impacted?

Page 42: Data Modelling and WITSML

42

Current

Proposed

Page 43: Data Modelling and WITSML

43

Page 44: Data Modelling and WITSML

44

Relationships

Page 45: Data Modelling and WITSML

45

Entity DefinitionCement Pump Schedule: Records the elapsed time, fluid rate and other pump related properties for the Cement Stage.

Definitions

Page 46: Data Modelling and WITSML

46

So…Is the Cement Pump Schedule for the Cement Stage? orFor each Cementing Fluid?…does relaxing the constraint require the definition of the Cement Pump Schedule to be revised?Remember, definitions add rigor to models!

Page 47: Data Modelling and WITSML

47

So What?Reduces the time taken for impact analysisDefinitions and business rules highlight issues/questionsInformed response to proposalBetter quality model/WITSML

Page 48: Data Modelling and WITSML

48

Using the Logical ModelScenario 2: Fit with existing application architectureRequirements to integrate data supplied in XML messages with existing systems

Inc. reporting, data warehouse

Or when assessing suitability of a system/application with business data requirements, e.g. SiteCom

Page 49: Data Modelling and WITSML

49

Existing Architecture ExampleRequirement to report on planned and actual wellbore trajectory

Page 50: Data Modelling and WITSML

50

Identify candidate target objectsEntity DefinitionA set of Trajectory Stations that describe the path of a section of a wellbore or of the entire wellbore.

Page 51: Data Modelling and WITSML

51

Identify candidate target objectsEntity DefinitionA set of Trajectory Stations that describe the path of a section of a wellbore or of the entire wellbore.

Page 52: Data Modelling and WITSML

52

BenefitsCan be used to highlight:

Fit and Integration challenges early in projectGaps and redundancyData element sourcing issues – data type and sizeHelp estimate development effort

Overlap points to reuseGaps require development – understand size and complexity

Page 53: Data Modelling and WITSML

53

The Wider LandscapeWITSML doesn’t exist in a vacuum!

Page 54: Data Modelling and WITSML

54

Enterprise Data Model

Conceptual Domain Model

ApplicationLogical Data Model

Physical Data Model

Described in more detail by

Generates schema of

Described in more detail by

Domain of an Enterprise data concept

Within subject area/domain

Reverse engineered into

Implemented in Reverse engineered into

Physical IT System

Implementation focus (Low)

(High)

(High)

(Low)

Communication focus

Data Model Levels” Models

Page 55: Data Modelling and WITSML

55

Use this layout for a title with a horizontally striped picture.

Summary

Page 56: Data Modelling and WITSML

56

In SummaryRepresenting WITSML through data models:

Easy to review by business and technical alike – ‘a picture paints a thousand words’Facilitates a shared understanding of conceptsRigorous, formal analysis and design processReduce the time taken for impact analysisMinimise reworkPromote reuse

Page 57: Data Modelling and WITSML

57

Further information:Articles including:

• Seven deadly sins of data modelling• The IT Credibility Crunch• Information Management Deficiency Syndrome• Modelling is not just for DBMS’s• Data mining - where’s my hard hat?• Master data mix-ups• Drowning in spreadsheets• Why bother with a semantic layer?• Business Intelligence in a cold climate• Data Management is everybody's business

Download from:http://bc.ipl.com/

Page 58: Data Modelling and WITSML

58

Contact details

Chris BradleyBusiness Consulting [email protected]+44 1225 475000

My blog: Information Management, Life & Petrolhttp://infomanagementlifeandpetrol.blogspot.com

@InfoRacer

Page 59: Data Modelling and WITSML

59

Questions?

business knowledgetechnical depth

engagement modelexcellence

experience

satisfactionvaluedeliveryqualityrelationship