Upload
others
View
5
Download
0
Embed Size (px)
Citation preview
Oil and Gas Spill and PipelineOil and Gas Spill and Pipeline Condition Assessment UsingCondition Assessment Using
Remote SensingRemote Sensing
Using New Tools forUsing New Tools for Situational AwarenessSituational Awareness
By: William E. RoperBy: William E. RoperGeorge Mason UniversityGeorge Mason University
SubijoySubijoy DuttaDuttaS&M Engineering ServicesS&M Engineering Services
Presentation OutlinePresentation Outline
�� IntroductionIntroduction�� Pipeline MonitoringPipeline Monitoring�� HyperspectralHyperspectral Oil Spill CharacterizationOil Spill Characterization�� ConclusionsConclusions
VisualizationVisualization
�� Assists the understanding of dataAssists the understanding of data�� Able to represent temporal changesAble to represent temporal changes�� A more challenging integrationA more challenging integration
requirementrequirement�� New software and hardwareNew software and hardware
developments are in this directiondevelopments are in this direction
Pipeline Monitoring andPipeline Monitoring and Condition AssessmentCondition Assessment
�� Imagery products inImagery products in multiple resolutionsmultiple resolutions and characteristicsand characteristics
�� Integration of dataIntegration of data sourcessources
�� Visualization ProductsVisualization Products�� Tailored products forTailored products for
the decision makerthe decision maker
• Mechanical Damage is #1 Pipeline Hazard
• Mechanical Damage Related to Encroachment – 29% of incidents and 20% of fatalities – Incident Distribution
• 72% Class 1 (rural land use) • 11% Class 2 • 15% Class 3 • 2% Class 4 (high density land use)
Motivation for Advanced Detection of 3rdMotivation for Advanced Detection of 3rd Party EncroachmentParty Encroachment
Satellite Monitoring for Pipeline AssetSatellite Monitoring for Pipeline Asset Safety and Security AssessmentSafety and Security Assessment
�� To develop and deliver a practical,To develop and deliver a practical, reliable, and economical means ofreliable, and economical means of monitoring pipeline assets usingmonitoring pipeline assets using earth observation data in twoearth observation data in two fundamental areas of pipeline safetyfundamental areas of pipeline safety
�� Third party encroachmentThird party encroachment�� Ground motionGround motion
Encroachment Monitoring
Encroachment Monitoring
Encroachment Monitoring
4 metre Multispectral
(colour) image
Encroachment Monitoring
4 metre Multispectral
(colour) image
RADARSAT
Overlay
1 metre
overlay
Targets
Encroachment Monitoring Combined Radar & Optical
8-9metre
panchromatic
Detected
Computerized change detection analysis
encroachment event
Notice Alarm Alert
Geo-referenced encroachment event
Field personnel are notified
Encroachment Monitoring Concept Service
Encroachment Event
Time Sequence Acquisitions
Satellite detected
Satellite Monitoring
88.0%Twice per day
78% to 93%70.0%Once per day
50% to 70%40.0%Twice per week
32% to 55%20.0%Once per week
5.0%Once per month
2.0%Once per 3 months
1.0%Once per 6 months
0.4%Once per year
Probability of Detection (%)
With Satellite Sensor Systems
Probability of Detection (%)
With Aerial Sensor Systems
Frequency of Imagery
Collection
Optical Data: Integration withOptical Data: Integration with GISGIS
Hyperspectral SensingHyperspectral Sensing
Single Pixel
Spectral Bands
Spatial Pixels
Flight Line
Wavelength
Inte
nsity
Pixel Spectrum
Single Sensor Frame Series of Sensor Frames
GREEN
Manmade and NaturalManmade and Natural ReflectancesReflectances50
40
30
20
10
0
0.4 0.6 0.7 0.8 1.3
Artificial turf Asphalt
Fallow field
Sandy loamy Soil
Concrete
Clear water
Wavelength (micrometers)
Grass
Visible 0.5
BLUE GREEN RED
Near IR
R E F L E C T A N C E
High
Low
Detected Large and Small GasDetected Large and Small Gas LeaksLeaks
PutuxentPutuxent River Oil Spill StudyRiver Oil Spill Study AreaArea
Illustration of Data Collected withIllustration of Data Collected with the AISAthe AISA HyperspectralHyperspectral SystemSystem
Supervised Classification of theSupervised Classification of the Image Data using ENVIImage Data using ENVI
ChallengesChallenges
�� Methods and authorities for improvedMethods and authorities for improved data sharingdata sharing
�� Disciplinary differences betweenDisciplinary differences between developers and usersdevelopers and users
�� Multi sensor data integrationMulti sensor data integration�� Interdisciplinary approach to needsInterdisciplinary approach to needs
development and productdevelopment and product requirementsrequirements
Research DirectionsResearch Directions
�� Information Integration and VisualizationInformation Integration and Visualization�� Expand Applications StudiesExpand Applications Studies�� Development of fuzzy classification systemsDevelopment of fuzzy classification systems�� Applications of neural networksApplications of neural networks�� Echelon analysis methodsEchelon analysis methods�� Data fusion and data miningData fusion and data mining �� Integrated Scenario ModelingIntegrated Scenario Modeling
For More InformationFor More InformationContact:Contact:
William Roper George Mason University Ph:703 993-1648 Email: [email protected]