RISK ENGINEERING MANAGING RISK INTELLIGENTLY OMAN RISK ENGINEERING SEMINAR 13 OCTOBER 2015 Adrian Louis BEng CEng MIChemE Risk Engineer Dubai - Marsh Emirates
Agenda Introduction Risk Management Framework Estimated Maximum Losses Conclusion MARSH 14 October 2015 1
Section 1 INTRODUCTION 14 October 2015 2
About Marsh s Global Energy Practice Vertically integrated global practice 60+ key locations 13 global energy hubs 2000+ clients globally Managing over US$2.5 billion premium annually 350+ energy professionals around the world 30 specialist energy risk engineers globally, mainly chemical engineers MARSH 14 October 2015 3
What Do We Do? Some Examples Risk Appraisal Reviews; Construction and Operational Risks Design / New Projects Risk Reviews Sabotage & Terrorism Risks Reviews Risk Benchmarking Studies Third Party Liability Reviews Thought Leadership 100 Largest Losses Loss Database and Learning from Losses Seminars Loss Control Newsletter Position Papers Specific Bespoke Client Training and Reviews Process Safety Management Hazard & Operability with Insurance Risks Fire Preparedness and Emergency Systems MARSH 4
Section 2 RISK MANAGEMENT FRAMEWORK 14 October 2015 5
Risk Management Framework MARSH 6
The Risk Management Framework Risk Management Cycle Identification Risk harvesting, using multiple parties and stakeholders MARSH 14 October 2015 7
The Risk Management Framework Typical Risk Landscape MARSH 8
The Risk Management Framework Risk Management Cycle Identification Risk harvesting, using multiple parties and stakeholders Measurement Audit, Monitoring, Sampling, Inspection, Checking, Identification MARSH 14 October 2015 9
Risk Map; e.g. for Downstream O&G and Petrochem Companies Unmitigated Material Risks Frequency 5 4 Customer concentration High skilled employee retention Capital investment (pipes, M&A) Legend Size of bubble implies volatility Red color implies critical risks 3 Contract disputes Supplier & sourcing Increased Regulations Auto/fleet liability Operating risks Health & Safety Permitting & Approvals Refining Margins/ Commodity prices 2 1 GHG legislation Cyber security Terrorism Interest rate Tech. evolution/ Stranded assets Contingent BI Railcar risk Construction Catastrophic Property Damage 0 0 1 2 3 4 5 MARSH Severity / Risk bearing Capacity 10
The Risk Management Framework Risk Management Cycle Identification Risk harvesting, using multiple parties and stakeholders Measurement Audit, Monitoring, Sampling, Inspection, Checking, Identification Feedback Analysis, Trends, Evaluation, Actions External Influence Laws, Industry Standards, Stake Holder Pressure, Public Concern, Company Image MARSH 14 October 2015 11
Corporate Risk Management Structure Identification What are the dangers? Measurement How big are the potential losses? Analysis and Solutions What can be done about the risks? Risk Avoidance Risk relative to reward is too high Risk Mitigation Not economical to transfer Risk Mitigation then Transfer Economical to transfer after mitigation Risk Transfer Economical to transfer risk immediately Exit Risk Area Organisational Solutions Enhance management processes to better manage risk Risk Management and Mitigation Financing Solutions Strategy Improvement People Improvement Process Improvement Systems Improvement Insurance Financial (capital markets) Hybrid Structure Engineering Data Collection Market Management Capacity Providers Compliance & Regulation Claims MARSH 12
Typical Risks Faced By Downstream O&G and Petrochem Companies Type of Risk Sub-type of Risks Typical mitigation measures Operational Risks Financial Risks Regulatory/ Strategic risks MARSH 1. Operating risks (fires, explosions, spills etc..) 2. Catastrophic Property damage and business interruption risk 3. Retention of highly skilled personnel 4. Construction cost risk 5. Terrorism risk 6. Employee and Contractor Health & Safety risk 7. Supplier and sourcing risks 8. Cyber security risk 9. Auto/Fleet Liability Risk 10. Contingent Business interruption risk 11. Railcar risks 1. Refining Margins and Commodity Prices Risk 2. Debt financing rate risk 3. Capital investment risk (new projects, M&A) 4. Customer concentration risk 5. Contract disputes risk 1. Permitting and approval risk (delays, moratorium or suspension) 2. Increased regulatory requirements (e.g. additional testing of equipment, contingency plans etc.) 3. GHG and climate change regulation risk 4. Broader technology evolution/stranded asset risk Operational excellence HSE excellence Supply chain and vendor management Business Continuity & Resilience practices Talent management Property & Casualty Insurance Other Insurance Commodity/financial derivatives/swaps Internal controls/risk Committee Investment committee E&O/D&O insurance Industry wide policy advocacy Industry coalitions Compliance training Internal controls/risk committee 13
The Risk Management Framework Risk Management Cycle Implementation & Performance Measurement Training, Supervision, Selection, Manning Interpretation Procedures, Methods, Job Description, Responsibility Identification Risk harvesting, using multiple parties and stakeholders Measurement Audit, Monitoring, Sampling, Inspection, Checking, Identification Policy Making Policy Statements, Corporate Goals, Standards Feedback Analysis, Trends, Evaluation, Actions External Influence Laws, Industry Standards, Stake Holder Pressure, Public Concern, Company Image MARSH 14 October 2015 14
Risk Map; e.g. for Downstream O&G and Petrochem Companies Unmitigated Material Risks Frequency 5 4 Customer concentration High skilled employee retention Capital investment (pipes, M&A) Legend Size of bubble implies volatility Red color implies critical risks 3 Contract disputes Supplier & sourcing Increased Regulations Auto/fleet liability Operating risks Health & Safety Permitting & Approvals Refining Margins/ Commodity prices 2 1 GHG legislation Cyber security Terrorism Interest rate Tech. evolution/ Stranded assets Contingent BI Railcar risk Construction Catastrophic Property Damage 0 0 1 2 3 4 5 MARSH Severity / Risk bearing Capacity 15
Risk Map For Downstream O&G and Petrochem Companies Post operational measures and Insurance solutions Frequency 5 Legend Size of bubble implies volatility Red color implies critical risks 4 3 2 1 Customer concentration GHG legislation Terrorism Cyber security Contract disputes Supplier & sourcing Interest rate Tech. evolution/ Stranded assets Employee retention Auto/fleet liability Railcar risk Increased Regulations Construction Contingent BI Operating risks Health & Safety Catastrophic Property Damage Capital investment risk (new pipes, M&A) Permitting & Approvals Refining margins/ commodity prices 0 0 1 2 3 4 5 MARSH Severity / Risk bearing Capacity 16
Section 3 INSURANCE AND ESTIMATED MAXIMUM LOSS 14 October 2015 17
Insurance Principles Insurance is a pool of premiums which are collected from Insured s and from which claims are paid form of mutual Insurance is there to provide cover for potential losses and to place the Insured into the same position as he was before the loss An Insured must act as if he was not insured Insurance is there to provide back up in case of accident If there is a shortfall in this pool, then premiums have to increase If one Insured causes more than his fair share of that shortfall, then he will asked to pay a higher disproportionate premium MARSH 18
What affects Insurance Pricing? Available market capacity Technical (market) rates for given industry; e.g. medical vs. automobile EML and Risks aggregation; especially for large industrial parks Natural Catastrophe load; Japan vs. KSA Deductibles / Retentions Claims and Loss History Asset Valuation; accuracy Gross Profit; for Business Interruption purposes MARSH 14 October 2015 19
Property Damage Why determine the EML? Why not purchase on Full Property Value There is no holy grail to suggest buying to a loss load (i.e. EML determination) or to full property value (FPV). The loss should be examined via risk modelling exercise to determine the strategy Traditionally FPV was the common place for insurance purchase however FPV may not yield overall Economic Cost of Risk (ECoR), because: There are unused premium allocation: Cover is purchased on assets which has a very low probability for loss occurrence It reduces overall market capacity; increasing the cost of purchase There is lack of competition: Only the big insurers will have access to large sums of capital There is an inability to obtain full cover for given asset value, e.g. a US 10 billion refinery Buying to FPV maybe justifiable where the loss can destroy the whole asset in its entirety For the majority of energy assets, the default is to buy to a loss load / EML, as the alternative might be a dollar exchange exercise MARSH 14 October 2015 20
Property Damage Estimated Maximum Loss The EML is suggested as the worst case scenario for a given asset Assets can be Onshore / offshore Operational In construction Upstream Downstream Cross country They may exist in harsh operating conditions, for example: Natural conditions; at sea, underwater, at altitudes Operating temperatures and pressures Ambient conditions; freezing to scorching MARSH 21
Property Damage Estimated Maximum Loss There is agreement the EML has the following characteristics: Expressed as a monetary value Considers a single event Event has a low probability of occurrence, although the actual likelihood has not been quantified Event is based on the Industry loss experience Derivation of the EML is by: Identifying and defining potential scenarios and loss mechanisms The loss needs to be modelled The effects of the loss needs to be turned into a monetary value MARSH October 14, 2015
Property Damage Estimated Maximum Loss The EML is a function of probability, within an event return period of 100 to 1000 years (10-3 to 10-4 ) That doesn t mean that very low probability events have not occurred; e.g. World Trade Centre, Buncefield, Tianjin, but these are very rare Within the Energy sector, out of the top 10 largest losses, 8 of these are due to vapour cloud explosions (VCE). Probability of event of a VCE is around 2 x 10-4, seems low, however, There are around 670 refineries worldwide, This gives a return period of a VCE of around 7.5 years for the refinery population alone, This drops to around 4 years for both refineries and petrochemical facilities MARSH 23
Property Damage Loss Scenarios Different types of asset can yield different loss scenarios: Fire & Explosion events VCE Marsh SLAM Jet Fire DNV PHAST Vessel disintegration SLAM Tank fire TOOL PHAST Pool / spill / plant fire PHAST BLEVE PHAST Building fire Review using TM Machinery Failure / Breakdown Review using TM Natural Catastrophe Review using TM Storm / Hurricane / Flood / Earthquake Toxic Release / Pollution Control of well / blowout Sabotage and Terrorism PHAST SLAM + TM SLAM MARSH 24
Property Damage Sedgwick Loss Assessment Model (SLAM) Fundamental basis is that vapour cloud explosions are congestion and confinement dependent with a stoichiometric mix TNT-equivalence basis is also used within the insurance markets to determine VCE type EMLs: Compiled from WW2 data on how detonations behave Does not take into account congestion / confinement characteristics Need to keep fiddling the yield factor to match observed overpressures Does not work well with hydrocarbon volumes of less than 5 Mt and more than 40 Mt MARSH 25
Property Damage Sedgwick Loss Assessment Model (SLAM) SLAM is an empirical model run using Shell s CAM (Congestion Assessment Method), developed in 90s based on large scale experiments Identified that congestion and confinement can influence explosion overpressures Suggests that the mechanism of a VCE is primarily deflagration as opposed to detonation Detonations have high peak pressures with quick decay as opposed to deflagration with lower peak pressures however damaging subsequent overpressures due to formation of vortices when in contact with structures SLAM has been compared against actual losses: Flixborough, 1974 Norco 1988 La Mede, 1992 Ju aymah, 1987 Lake Maracaibo, 1993 MARSH 26
SLAM performance against Industry losses SLAM MARSH R.F. Barton, 2nd Int. Conf. on Loss Prevention & Safety, Bahrain, Page 361, 1995 27
Property Damage Estimated Maximum Losses Fundamental characteristics which can affect the PD-EML Scenario VCE Explosion (S&T) Building Fire Tool SLAM SLAM Technical Method Information required Asset value by area Plot plan; showing congestion and layout Inventory Volume Inventory Composition Inventory Pressure and Temperature Asset value by area Plot plan Payload; - Bag (25 kg) - Car (450 kg) - Truck (4,500 kg) Asset value by area Construction Materials Building Fire Load Fire Protection Systems Fire Risk Areas MARSH 28
Property Damage Sedgwick Loss Assessment Model (SLAM) 26 Refinery Tank Farm North 21 Main Utilities (Pkg 5B) 2 3 4 1 HP HDSLP HDSNHDT CDU/VDU 5 5 HPU HPU 4 10 NHDT 1 CDU/VDU 3 LP HDS 2 HP HDS 19 17 PX 15 Cat Ref 14 Fuel 16 BEU 13 10 12 SGU ARU 11 SWS SRU 7 3 4 DHC 8 MHC 6 5 FCCU 23 2 SHU 24 11 12 13 UGP 9 8 7 9 Alkyl 18 1 18 6 18 DCU / Merox 25 Coke handling Pkg 5C Refinery Tank Farm South 20 Flare Fire sys & plant water 0 500m MARSH 29
Property Damage Sedgwick Loss Assessment Model (SLAM) 26 0.70 bar 0.35 bar Refinery Tank Farm North 21 0.20 bar 0.10 bar 0.05 bar Main Utilities (Pkg 5B) 2 3 4 1 HP HDSLP HDSNHDT CDU/VDU 5 5 HPU HPU 4 10 NHDT 1 CDU/VDU 3 LP HDS 2 HP HDS 19 17 PX 15 Cat Ref 14 Fuel 16 BEU 13 10 12 SGU ARU 11 SWS SRU 7 3 4 DHC 8 MHC 6 5 FCCU 23 2 SHU 24 11 12 13 UGP 9 8 7 9 Alkyl 18 1 18 6 18 DCU / Merox 25 Coke handling Pkg 5C Refinery Tank Farm South 20 Flare Fire sys & plant water MARSH 30 0 500m
Business Interruption Losses affecting the financial performance of a site post an event can be insured as Business Interruption or Contingent BI Income from sales Costs that reduce with throughput e.g. feedstock / utilities Variable cost reduction Fixed costs that reduce e.g. contractual obligations Fixed cost reduction Revenue Insurable Gross Profit Includes costs that continue, net profit and debt servicing Gross Profit MARSH 31
Business Interruption Dependency Schematic MARSH 32
Section # CONCLUSION 14 October 2015 33
Conclusion Risk Engineering Managing Risk Intelligently Risk Engineering represents an innovative element of risk management, beyond that of just insurance requirement A critical success factor is to establish a robust risk management framework, and this by a cycle of: Identification Measurement Prioritisation Analysis Mitigation Implementation Performance Measurement Each type of facility can yield a different estimated maximum loss scenario There are genuine premium savings in buying to a loss load instead of full property value Risk Engineering is just there to get the right insurance coverage in place If you have further questions post this session, get in touch with your friendly Risk Engineer or your Marsh contact THANK YOU MARSH 34