Risk assessment of Maroun gas and oil pipelines due to land sliding hazard



based on D-InSAR technique



Mohammad Sharifikia, @ Associate professor, Tarbiat Modares University, Department of Remote Sensing-



Iran



Meral Poorhamzah, postgraduate in Remote Sensing, Tarbiat Modares University



Abstract

It is importance to note that Iranian oil company have to transfer this valuable enrage from one side to other side of

country passing form several ridge and valley prone with several natural hazard. This is because the natural sources

of oil and gas generally lied in south west part of Iran locally calling Manathegh Nafte Khize Jonoub (south oil field

area). This area is closed to one of most active geological zone of Iran (Zakrose) covering thousands of kilometer

from south east to north west. Supplying natural enrages to central port of country need to crossing from this zone

which is suffering with several difficulties as well as neutral hazard. Out of neutral hazards can found to excite in

this area, the landslide hazard is a main restriction for pipeline crossing over.

The present research is dale with radar interferometry techniques applying for risk assessment and mapping over the

oil and gas pipelines suffering to landslides hazard in the part of Central Zagros (Maroun-Esfahan). For this purpose,

two individual radar dataset in C (ASAR) and L (PALSAR) band with deferent time were collected. Furthermore,

the D-InSAR technique was applied for land surface movement and land displacement detection. The outcome map

was showed the maximum rate of land displacement in this region is about 7.4 cm uplifted and 3.9 cm subsidence

with duration of almost one year. this is due to shape of landslide over the area’s slop. Overlying the landslide map

with the pipeline crossing route shown at lies three active landslides over the Maroun-Esfahan gas and oil pipelines.

For investigation about this three landslide and damage estimation over the pipeline the field study has been done

for accuracy assessment and land movement rat measuring and evaluation. Which, successfully identified and

mapped 3 landslides were located across the pipeline and damage it. Furthermore, map surveying by DGPS in RTK

method over the one of landslide shown that sliding transfer 20 m with falling 10 m over the length of 45 m of gas

pipeline. moreover, the press of landslide made curvatures on straight pip hogging pipe 43 cm. continued this

landslide activation and more pressing in close further can make a fracture and pessimistic pipe expulsion. With can

a kind of disaster if the event be close to settlements are.

The outcome landslide map shown the active landslide points (small area) very well, but the main think need to

suffusion information about interred area. For this exigency have to convert points data map to area as prediction

hazard. For this proses and to understanding the amplitude of landslide hazard in area the information value model

was applied for hazard zonation and mapping. The landslide hazard map resulting from D-InSAR technique as

inventory map along with 8 data set maps namely, lito-logy, soil, land cover, lineaments, faults, roads, derange

pattern and slop, has been interred to model for zonation and hazard estimation over the area. Furthermore, this map

was reclass in 5 individual hazard and risk class from low to high risk. The hazard map analyses and calculation was

show about 20 percent of area study was marked as high and very high risk zone. This is mainly because of

morphological and lito-logical exclusivity of area resulting by active tectonics. Crooning and overlaying the

landslide hazard map with pipeline track has been shown 28.5 percent of line length crossing over the high and very

high risk zone, where the 52 percent was prone with low and very low risk zone. This mine that near 1/3 of mention

pipeline length suffering from hazardous area which can classified as high risk part of pipeline.

Interpreting the hazardous classes on the prediction map is an important concern in landslide prediction models. For

this purpose, the prediction-rate curve was generated using validation group of landslide locations to validate the

prediction map obtained. This rate curve explains how well the model and factors predict the landslide. Results from

the success-rate curve are very promising, since the 3% area predicted as the most hazardous, includes 42.35% of

the total area affected by landslides, and this value grows to 90%, when about 25% area of highest susceptibility is

considered. The prediction accuracy can be assessed qualitatively by calculation the area under cover. The total area



equal to one means perfect prediction accuracy. In this model ratio area was 0.633 that means the prediction

accuracy was 63.3%.

Keywords: Differential SAR Interferometry, PALSAR, ASAR, Landslide, Oil and Gas Pipeline risk