Hazardous Liquid Pipelines Leak Detection Methodologies Pipe Tech Americas Summit 2013
Dec 05, 2015
Hazardous Liquid Pipelines
Leak Detection Methodologies
Pipe Tech Americas Summit 2013
US Pipeline Infrastructure
Pipeline Integrity Management Components
PIMProgram
In-line Inspection (pigging)
• $1,000-5,000/km
• Once every 5-7 years
• Not well suited for gas lines
• 30% of liquid pipelines un-
piggable
Direct Assessment (digging)
• >$50,000/location
• Method of last resort
• Applied to un-piggable
pipelines
Hydrostatic Testing
• >$10,000’s/km
• Pipeline is out of service
• Water disposal issues
Leak Detection (ground)
• >$100/km
• Once every 1-2 years
• Private land access issues,
low quality control
Leak Detection (aerial)
• $50 – 70/km for once a year
• <$15/km if done with monthly
visual flights
• Guaranteed pipeline coverage
Aerial Patrolling
• $10-30/km
• Usually performed monthly
• Visual/photographic assessment
of pipeline ROW condition
Note: Not a complete list
Leak Detection (continuous)
• Uses SCADA and CPM
systems
• Typically leak >1% of flow
• False-alarms problematic
Hazardous Liquid Leak Detection Systems
• “Internal” continuous:• Leveraging sensors used for pipeline operation
(SCADA)
• Computer models sometimes used to interpret
measurements (CPM)
• Mass balance, pressure, flow
• External continuous or periodic:• Requires installation of new sensors
• Most technologies infer a leak from a secondary
physical effect
• Acoustic
• HC sensing using fiber optic
• Temperature sensing using fiber optic
• Liquid sensing cables
• Vapour sensing tubes, sensors
• Optical cameras
HL Leak Detection: Continuous SCADA/CPM
• SCADA systems are widely deployed and the first
line of defense at detecting large leaks and
ruptures
• Contrary to expectations, <50% effective at
detecting leaks (2012 PHMSA Leak Detection Study
– DTPH56-11-D)
SCADA/CPM: HL Transmission Pipelines
• Average release: 29,000 gallons
• Largest release: 843,000 gallons
• SCADA system in place: 77% (6% not installed)
• CPM system in place: 44% (39% non installed)
SCADA/CPM: HL Transmission Pipelines
Initial Identifier, All Releases
SCADA/CPM: HL Transmission Pipelines
Initial Identifier, Above Average Releases
SCADA/CPM: HL Gathering Pipelines
• Largest release: 8,400 gallons
• SCADA system in place: 23% (41% not installed)
• CPM system in place: 5% (59% not installed)
SCADA/CPM: HL Gathering Pipelines
Initial Identifier, Above Average Releases
SCADA/CPM: NG Trans/Gathering Pipelines
• Largest release: 615,000 MSCF
• Average release (from pipe body): 30,000 MSCF
• SCADA system in place: 89% (10% not installed)
• CPM not typically deployed
Initial Identifier, All Releases
SCADA/CPM: NG Trans/Gathering Pipelines
Initial Identifier, Above Average Releases
SCADA/CPM Summary
SCADA
CPM
All, Detected
Above Average, Detected
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
HL Transmission
HL Gathering
NG Trans/Gathering
HL Leak Detection: Acoustic
• Array of sensors (microphones)
installed onto or close to the
pipeline
• Used within smart pigs or balls• Can accurately pinpoint a leak location
• Relatively limited coverage area (few miles)
HL Leak Detection: Fiber Optic Hydrocarbon Sensing
• Detecting change in refractive
index of fiber mounted near
pipeline
• Used as handheld probes or
continuous installation• Can accurately pinpoint a leak location
• Relatively limited coverage area (few miles)
• Expensive to retrofit legacy systems
HL Leak Detection: Fiber Optic Temperature Sensing
• Detecting change in refractive
index of fiber due to temperature
changes• Can accurately pinpoint a leak location
• Relatively limited coverage area (few miles)
• Expensive to retrofit legacy systems
HL Leak Detection: Liquid Sensing Cables
• Electrical cables buried
underneath pipeline
• Liquid leak will change impedance
between two wires• Cheaper than fiber optic cables
• Deployed over shorter distances and less
accurate than fiber optic
• Impacted by other liquids (water)
• Limited coverage area (few miles)
• Expensive to retrofit legacy systems
HL Leak Detection: Vapour Sensing Tubes
• Small diameter perforated tube
buried underneath pipeline
• Air gas samples are drawn from
the tube and analyzed for
hydrocarbons• Deployed over very short distances
• No pinpointing ability
• Expensive to retrofit legacy systems
HL Leak Detection: Optical Sensing
• Remote Sensing • Forward Looking Infrared (FLIR)
• Hyperspectral
• Differential Absorption Lidar (DIAL)
• Gas Filter Correlation Radiometer (GFCR)
Remote Sensing of Hydrocarbons
• Remote Sensing: Measuring properties of an object from
a distance, without physically sampling the object, by
detection of the interaction(s) of the object with an
electromagnetic (EM) field• Do not have to be in the plume of the leaked gas
• Can detect in inaccessible areas (“over-the-fence”)
Remote Sensing: Hydrocarbon Absorption
Forward Looking Infrared (FLIR)
Forward Looking Infrared (FLIR)
Differential Absorption LIDAR (DIAL)
Differential Absorption LIDAR (DIAL)
Gas Filter Correlation Radiometry (GFCR)
May – Aug 2000 CO distribution Real-time ammonia plume
Gas Filter Correlation Radiometry (GFCR)
33
realSensTM Technology - Platform Instrument
• Canadian Space Program derived technology
• $53 million in research and development
(Synodon $15 M to date, $1.5 M annually)
• remote hydrocarbon detection from 300
meters altitude
Leak point
30 mi/hr wind
Gas Leak Image – Controlled Leak
Gas Leak Image – Facility Leak
Liquid Leak C5+
Liquid Leak C5+
Remote Sensing: Hydrocarbon Interference
Controlled Flight Tests
• Field Testing Results • Pentane: < 0.04 lpm
• Gasoline: < 0.2 lpm
• Condensate: < 0.15 lpm
Condensate Evaporation Rate
Figure 4-1: Volume of a 4’×8’×3” pool of condensate (226.5 L) over a 200 minute period. Ten chemical
components were modelled to be evaporating, comprising 66.5% (by volume) of condensate.
Gasoline Evaporation Rate
Figure 3-1: Volume of a 4’×8’×3” pool of gasoline (226.5 L) over a 200 minute period. Fourteen chemical
components were modelled to be evaporating, comprising 62.9% (by volume) of gasoline.
Gasoline Evaporation Rate
Figure 3-2: Volume of 14 chemical components of gasoline (plus ethanol) over a 200
minute period.
realSensTM HL Leak Detection
• realSensTM could detect leaks that are < 1 lpm of
hydrocarbons with 5% light ends
• For a 100,000 bpd line, this would correspond to 0.01% of
the flow or 10 bpd
• A typical SCADA/CPM system would detect a 1% leak
roughly 50% of the time
HL Leak Detection: Conclusions
• SCADA/CPM are line of first defense but not 100%
reliable
• Best Leak Detection Solution: continual
SCADA/CPM monitoring with periodic over-line
surveys and external sensors in critical HCA
Contact
Synodon Inc.6916 Roper Road
Edmonton, AB, T6B 3H9
Tel: 780.468.9568
www.synodon.com
Adrian [email protected]
Tel: 780.468.9568