RiskBased Pipeline Route Optimization for the Persian Gulf

Risk-Based Pipeline Route Optimization for the Persian Gulf Region Scott Byron Exxon. Mobil Exploration Co. & Master’s Candidate, Penn State University MGIS Program 2010 ESRI Petroleum Users Group Conference February 23, 2010

Presentation Overview • Project overview • Analysis methodology • Results & Conclusion Courtesy: NASA 2010 ESRI PUG Conference

Project Overview 2010 ESRI PUG Conference

Project Objectives • Proof of concept • Create risk maps by Identifying & weighting analysis risk factors • Test system against existing pipelines Qatar – Abu Dhabi – Oman Pipeline Courtesy: Oil & Gas Journal 2010 ESRI PUG Conference

Why Focus on Risk? The Persian Gulf region topography & bathymetry • New pipelines will encounter high-risk environments • People can easily perceive risks • Risk management improves project performance • Must apply internal safety & risk management policies ~90 m (Approx. depth of Persian Gulf) -1, 788 m 2010 ESRI PUG Conference 0 m 2, 863 m Key Map

Constraints The Persian Gulf region topography & bathymetry • Availability of experts • No marine examples • Reliance on public data • Expectation that low risk path will equate to shortest, lowest cost path ~90 m (Approx. depth of Persian Gulf) -1, 788 m 2010 ESRI PUG Conference 0 m 2, 863 m Key Map

Analysis Methodology 2010 ESRI PUG Conference

Risk Categories Risk Factor Identification Construction 10 risks • 36 risk factors identified • 19 – terrestrial, 11 – marine, 6 - both • Divided into 4 categories Operation 13 risks Socio-Economic 7 risks Environmental 6 risks 2010 ESRI PUG Conference

Formal Risk Analysis Risk Scores • Probability of Occurrence & Potential Impact scores combined for Risk Score • High priority risks have Risk Score >= 3. 0 Risk Count • High priority risks used in weighting methods = High Priority Risks 2010 ESRI PUG Conference

Weighting Methodology Simple Weighted Index • Factors evaluated independently Increasing risk • Evaluated 21 risks 2010 ESRI PUG Conference

Weighting Methodology Pair-Based Comparison Overview • Evaluating Brown & Peterson method, Analytical Hierarchy Process • Factors compared to each other Factor A Factor C > Factor B < Factor A < Factor B • Data collected via survey • Fewer risks evaluated Factor A 2010 ESRI PUG Conference > Factor B > Factor C

Weighting Methodology Brown & Peterson Method • Captures selection frequency • Is row factor riskier than column factor? • No factor categories • Simple calculation of weights Factor ‘s column sum = Weight Maximum selection # 2010 ESRI PUG Conference Frequency matrix for 10 respondents

Weighting Methodology Analytical Hierarchy Process • Captures Degree of importance (How much riskier? ) on scale of 1 – 7 (similar – significantly riskier) • Weights categories • Uses matrix similar to BP method • Weights categories • Complex weight calculation 2010 ESRI PUG Conference Analysis Problem Category 1 Category 2 Factor A Factor C Factor B Factor D

Weighting Methodology Analytical Hierarchy Process Analysis Problem Category 1 Weight = 1. 00 Factor A Weight = 4. 28 Steps 1. 2. 3. 4. 5. 6. Sum each column Divide all cell values by column sum Sun each row Divide all row sums by smallest row sum Divide by combined row sums Multiply by Category weights, repeat steps 4 & 5 2010 ESRI PUG Conference Factor B Weight = 1. 00 Category 2 Weight = 1. 5 Factor C Weight = 6. 1 Factor D Weight = 4. 52

Results 2010 ESRI PUG Conference

Top High Priority Risks Seismic Low Landslide High risk areas 2010 ESRI PUG Conference Steep Slope (>5%) High Low High Crossings River, road, rail or pipe Commercial Shipping Low High Reefs & Spill Windows Spill response window (2 day) Coral reefs

Final Risk Surfaces Simple Weighted Index 0 • • Brown & Peterson 0. 424 Land riskier than ocean due to higher number of land risks Zagros has highest risk 2010 ESRI PUG Conference 0 Analytical Hierarchy Process 0. 466 0 0. 514 • Less ocean – land distinction • Further loss of ocean – land distinction • Zagros is riskiest • Exaggerated risk variation

Validation The IGAT 4 pipeline, Iran IGAT 4 Pipeline • Land oil pipeline • Crosses mountains, active earthquake area • ~$1 billion Key Map 2010 ESRI PUG Conference

Validation Alignments on BP Risk Map IGAT 4 Pipeline • SWI shortest, has greater slope variability • BP, AHP have best match • Key risks: seismic, slope, landslide, crossings Key Map 2010 ESRI PUG Conference

Validation IGAT 4 Pipeline 2010 ESRI PUG Conference

Validation The Dolphin Pipeline, Qatar - UAE Dolphin Pipeline • Marine gas pipeline • Crosses highly developed area • $3. 5 billion Key Map 2010 ESRI PUG Conference

Validation Alignments on SWI Risk Map Dolphin Pipeline • SWI has best match • BP, AHP similar, but route southward • Key risks: commercial shipping, pipeline crossings, coral reefs & spill windows Key Map 2010 ESRI PUG Conference

Validation Dolphin Pipeline 2010 ESRI PUG Conference

Conclusion 2010 ESRI PUG Conference

Learnings • Risk approach is valid • Lowest risk path: pair-based • Shortest path: SWI • Data quality impacts results • Finer resolution needed • Reevaluate some factors for coverage and weight 2010 ESRI PUG Conference

Questions? Dubai, UAE 2010 ESRI PUG Conference

Appendix 2010 ESRI PUG Conference

List of Risk Factors 2010 ESRI PUG Conference

List of Risk Factors (Con’t) 2010 ESRI PUG Conference

Resources Applying AHP in Weighting GIS Model Criteria. Retrieved from http: //www. innovativegis. com/basis/Supplements/BM_Sep_03/T 39_3_AHPsupplem ent. htm Brown T. C. , & Peterson G. L. (2009). An enquiry into the method of paired comparison: Reliability; scaling; and thurstone's law of comparative judgment. USDA Forest Service - General Technical Report RMRS-GTR. (216 RMRS-GTR), 1 -98. http: //www. fs. fed. us/rm/pubs/rmrs_gtr 216. pdf Longley, P. A. , Goodchild, M. F. , Ma. Guire, D. J. & Rhind, D. W. (2005). Geographic Information Systems and Science, 2 nd Ed. . London: John Wiley& Sons, Ltd. . 2010 ESRI PUG Conference

Resources (Con’t) Montemurro D. , Barnett S. , & Gale T. (1998). GIS-based process helps Trans. Canada select best route for expansion line. Oil and Gas Journal. 96(25), 6371. Saaty T. L. , Vargas L. G. , and Dellmann K. 2003. The allocation of intangible resources: The analytic hierarchy process and linear programming. Socio-Economic Planning Sciences. 37 (3): 169 -184. Pipelines Braestrup, Mikael W. Design and Installation of Marine Pipelines. . Blackwell Publishing. Online version available at: http: //knovel. com. ezaccess. libraries. psu. edu/web/portal/browse/display? _EXT_KN OVEL_DISPLAY_bookid=1384&Vertical. ID=0 2010 ESRI PUG Conference

Resources (Con’t) Guo, Boyun; Song, Shanhong; Chacko, Jacob; Ghalambor, Ali Offshore Pipelines. Elsevier. Online version available at: http: //knovel. com. ezaccess. libraries. psu. edu/web/portal/browse/display? _EXT_KN OVEL_DISPLAY_bookid=1258&Vertical. ID=0 Data Global Seismic Hazard Program. (2009 August 27). Middle East Hazard Map. Retrieved from: http: //www. seismo. ethz. ch/GSHAP/ National Center for Ecological Analysis and Synthesis. (2009, August 2). Commercial Activity (Shipping) Dataset. Retrieved from: http: //www. nceas. ucsb. edu/Global. Marine/impacts 2010 ESRI PUG Conference

Resources (Con’t) Nation. Master. com. (2009, September 22). Oil Consumption (per Capita) (Most Recent) by Country. Retrieved from: http: //www. nationmaster. com/graph/ene_oil_con_percap-energy-oil-consumptionper-capita Tupper, M. , Tewfik, A. , Tan, M. K. , Tan, S. L. , The, L. H. , Radius, N. J. , & Abdullah, S. (2009, September 2) Reef. Base: A Global Information System on Coral Reefs [Online]. Retrieved from: http: //www. reefbase. org United Nations Environmental Program. (2009, August 18). Global Environment Outlook (GEO) Data Portal. Retrieved from: http: //geodata. grid. unep. ch/ United States Geological Survey. (2009, September 2). Earthquake Hazards Program: Epicenters Database. Retrieved from: http: //neic. usgs. gov/neis/epic_rect. html 2010 ESRI PUG Conference

Resources (Con’t) World Database on Protected Areas. (2009, August 5). Annual Release 2009 Dataset. Retrieved from: http: //www. wdpa. org/Default. aspx 2010 ESRI PUG Conference