PODS Adaptation for PIM at Dominion

1
PODS Adaptation for PIM at Dominion

• Nancy Gleason
• Dominion Resources
• Sue Whiting
• Ten Sails Consulting

PODS Users Group Presentation 2006
2
Overview

• Implementation history
• Linear Reference Data Model
• Survey Data Model
• Architecture
• ILI Data Flow
• Lessons Learned
• Summary

PODS Users Group Presentation 2006
3
History

• Integration Common Linear Reference
• 2002 2003 Initial Implementation
• 2004 Adjustments First Official Risk Ranking
• 2005 New Applications Implemented
• Cost of manually realigning surveys
• Inconsistent format of surveys
• 2006 Continuous Improvement Efforts in Progress

PODS Users Group Presentation 2006
4
Linear Reference Model

• Dominion already had an existing GIS data model
• Develop a linear reference (LRS) model that had
  as minimal as possible impact to the existing
  data model.
• Develop a LRS that satisfied the needs of
  Pipeline Integrity Management (PIM) and could be
  extended if needed.
• PODS
• Open Industry Standard, not vendor specific.
• Active committees comprised of industry operators
  reviewing and providing feedback on the model.
• Made sense to Dominion Business/Technical Teams.

PODS Users Group Presentation 2006
5
Implementation at Dominion
Pipeline
Centerline (Comparable to Route)
Station Point
Event Range
Feature Table
Any Facility
Any Facility
Any Facility
PODS Users Group Presentation 2006
6
Survey Data Model
_CL fields have associations to code list
tables. For discussion purposes, those tables
have not been shown here.
PODS Users Group Presentation 2006
7
Architecture
ILI and CIS
PODS Users Group Presentation 2006
8
ILI Data Flow
PODS Users Group Presentation 2006
9
Data Flow

• Populate MS Access DB
• GIS based report generator to pre-populate the MS
  Access database (IDX db) with data related to the
  pipes to be surveyed.
• Pre-populated IDX db is sent to vendors.
• Survey is run and results are fed into the IDX
  db.
• Survey is completed and vendor returns the IDX
  db, now augmented with the results from the
  survey. IDX db contains validations to ensure
  vendor populated the tables correctly.

Access Exchange Database
ILI_Inspection_Range
GIS
ILI_ Data
Trace_Report
PODS Users Group Presentation 2006
10
Trace report
PODS Users Group Presentation 2006
11
ILI Inspection Range
PODS Users Group Presentation 2006
12
ILI Data
PODS Users Group Presentation 2006
13
Survey Data Returned from Vendor
Oracle Database Perf_Data_Mgmt Schema
Access Exchange Database
Raw_Vend_ILI_Inspect_Range
1 Record
ILI_Inspection_Range
Raw_Vend_ILI_ Data
ILI_ Data
Millions of records
GIS
Trace_Report
Many months may elapse
PODS Users Group Presentation 2006
14
Survey Integrate and Alignment

• Process performance data in compliance with rule
  DOT 192 subpart O, ASME B31.8s, NACE RP 0502
• Align and integrate performance data, in house.
• Data interpolation to do initial alignment and
  on-going alignment of performance data.
• Track anomalies throughout lifecycle
• Establish common business processes if practical.
• Where are we?
• ILI is implemented.
• CIS is in development.

PODS Users Group Presentation 2006
15
Align and Integrate with GIS
Oracle Database Perf_Data_Mgmt Schema
1
ILI_Inspection_Range
2
Fit Points Create if valid
Raw_Vend_ILI_ Inspect_Range
ILI_Data
Raw_Vend_ILI_ Data
Survey_Fit_Point
3
Survey_Table_Xref
Fit Points Projects Validate only
Rupture_Failure_Pressure
Many months may elapse
PODS Users Group Presentation 2006
16
Survey Alignment Tool

• Extensible to future survey types.
• Filtering
• View by fittings (applicable for GIS users),
  welds, anomalies, etc.
• Fit Point identification
• Automatic when possible
• Manual if further refinement is needed
• Validation
• Call out any fittings found in survey that are
  not present in the GIS
• Call out fittings that fall outside of tolerance
  (i.e. 50 difference between where the survey
  says a fitting resides and the GIS says the
  fitting resides

PODS Users Group Presentation 2006
17
Interpolation

• Xn (En Sn) (X - So)/(Eo So) Sn
• X Represents any point on the pipe
• Xn Denotes the new value of X after applying
  the adjustments specified by the transaction
  table
• So The Original Start value
• Eo The Original End value
• Sn The New Start value
• En The New End value

PODS Users Group Presentation 2006
18
Interpolation Example

• Survey has been aligned to the GIS valves giving
  us
• Valve1 odometer 1100, measure 100
• Valve 2 odometer 1200, measure 200
• Anomaly was found at odometer 1150 but what does
  that equate to in the LRS?
• Use interpolation formula
• Anomaly Measure
• (200 100) (1150 - 1100)/(1200 1100)
  100
• 150

Concept would be similar for CIS
PODS Users Group Presentation 2006
19
Nightly Interpolation

• With data residing in the GIS, Oracle and other
  repositories we needed a strategy to keep the
  measures up to date.
• WHY?
• Pipes, and their related appurtenances and
  fittings change.
• Data cleanup
• Actual physical changes in the field.
• These modifications result in changes to the
  measures within the GIS, which have to be passed
  to the enterprise repository

PODS Users Group Presentation 2006
20
How?

• A nightly synchronization from the GIS to the
  enterprise repository occurs.
• Data tied directly to a physical GIS object will
  automatically pick up the new measures. An
  example would be the inspection data.
• Data tied indirectly to a physical GIS object,
  such as the ILI survey data, are re-interpolated
  using an Oracle stored procedure.

PODS Users Group Presentation 2006
21
Nightly Interpolation Example

• Initial measures at time of survey were
• Valve1 0 Valve 2 200 and the Anomaly (X)
  has an measure of 150.
• Adjustments to pipe have added an additional 100
  to the line.
• Valves are automatically adjusted in the GIS
  based on the LRS.
• The anomaly is adjusted using the interpolation
  formula
• Anomaly New Measure
• (300 100) (150 - 0)/(200 0) 100
  
• 250

PODS Users Group Presentation 2006
22
Alignment Sheet Generator

• The data is aligned . Now we want to see it!
• Have all data in-house, why send out for sheet
  generation?
• Leverage the use of the existing in-house GIS
  application.
• Easy to use!
• End users are engineers, not GIS users. We
  didnt want the engineer to have to become a GIS
  expert.

PODS Users Group Presentation 2006
23
Alignment Sheets

• View alignment sheets and GIS data.
• Ability to run on entire pipeline or sections.
• Basic navigation capabilities of zoom in, out and
  panning.
• Filtering on the bands
• Drilldown capability. Make the data bands
  intelligent.
• Less reliance on paper. Use a more interactive
  environment to analyze the pipeline data

PODS Users Group Presentation 2006
24
Where does the data come from?
GIS - Map with any related ortho photography,
raster, etc.
Oracle - ILI Survey Data
3rd Party data, i.e. Soil
Oracle Threats and Assessments
GIS Pipe data, HCA, Population Classifications
PODS Users Group Presentation 2006
25
Lessons LearnedSummary
PODS Users Group Presentation 2006
26
Lessons learned from ILI

• Vendor meetings prior to using the new format.
• During development received input from vendor.
• Took several iterations with the vendor to get
  the survey submitted correctly they were
  testing and developing too.
• Access to Oracle issues
• Number vs. Auto
• Number values combined in a text field are not
  program friendly ex. CIS Interruptions cycle
  should be two numeric columns
• Data volume and performance.
• A common data model doesnt automatically equate
  to common business practices.
• Difference in precision for fitting survey data
  to the GIS data.
• Different level of software sophistication.
• Dont assume the vendor will recall the processes
  put into place each year.

PODS Users Group Presentation 2006
27
Summary

• PIM requires an integrated view of certain events
  and characteristics affecting the integrity of
  gas transmission pipe.
• Dominion achieved integration by implementing a
  common linear reference based on the PODS
  centerline concept and use of a GIS generated
  enterprise key to tie objects to the common
  linear reference measures.
• From an initial implementation focusing on
  integrating existing electronic data, Dominion
  has continued to expand data capture and
  integration to more fully meet the requirements
  of the PIM. The chosen architecture has been
  able to support the additional functionality.
• Use of the IDX standards for surveys has enabled
  Dominion to integrate this data dynamically at a
  one time implementation cost.

PODS Users Group Presentation 2006
28
Questions?
Thanks for coming!!!
Nancy_Gleason_at_Dom.com Sue.Whiting_at_Tensails.com
PODS Users Group Presentation 2006