Catastrophe modeling improves risk management

Jan. 1, 2008
Gulf of Mexico operators and insurers need better tools to understand hurricane risk in order to avoid losses from storms such as those in 2004 and 2005, or worse.
Offshore-tailored design yields best protection

Richard L. Clinton, EQECAT Inc.

Gulf of Mexico operators and insurers need better tools to understand hurricane risk in order to avoid losses from storms such as those in 2004 and 2005, or worse. A clear understanding of risk leads to viable decisions for platform design, operations, and maintenance. With decades of experience in risk assessment of offshore platforms, ABS Consulting engineers have designed a product to enable the industry to protect its assets, while maximizing the rate of return on the investments. The product is called EQECAT Gulf of Mexico US Offshore Energy Model, developed specifically for the offshore oil and gas industry.

Hurricane damage

Hurricanes Katrina and Rita devastated the Gulf’s oil and gas infrastructure. Katrina blasted through the GoM with sustained winds at 160 mph, destroying 47 platforms and severely damaging 20 others. Less than 30 days later, Rita made landfall near the Texas-Louisiana border, destroying 66 platforms in its path. These storms also damaged over 650 pipelines, which contributed to delayed production recovery. API estimated the damage to offshore production facilities and pipelines at $18-$31 billion. Others have estimated the damage at $15 billion, with about two-third physical damage and one-third business interruption (BI) and operator extra expenses (OEE). BI losses were mainly due to shut-in production from damaged pipelines; OEE losses were primarily driven by well control and debris removal.

About 75% of the Gulf of Mexico’s platforms were in the path of hurricanes Katrina and Rita. Sources: MMS, US Department of the Interior, and EQECAT Inc.
Click here to enlarge image

These losses could have been significantly worse. ABS Consulting’s EQECAT Gulf of Mexico US Offshore Energy Model, part of EQECAT’s WORLDCAT enterprise catastrophe management software suite, includes a worst case scenario – a Camille-intensity hurricane that tracks through the heart of the Gulf’s production facilities causing more than $65 billion in damage. Losses include platform damage, debris removal, extra expenses, well plugging and abandonment, and lost production from damaged platforms, pipelines, and onshore terminals (shut-in BI-Contingent Business Interruption [CBI] loss). An event of this magnitude impacts oil and gas supplies, which could cause a spike in commodity prices and reduce production for a number of months. The financial implications of an event like this are significant. Adequately planning for such an occurrence is critical.

Loss assessment

Operators use sophisticated financial modeling to optimize return on capital over the life expectancy of their assets. Decisions are made on platform configuration where the trade-off is cost versus performance. Underlying the financial modeling are assumptions regarding hazard intensity probabilities and asset vulnerability. If either the hazard or the structural performance assumptions are wrong, the financial decision will be flawed. Misunderstanding the risk can impact the risk transfer decisions, which can result in inadequate mitigation. These are all areas where catastrophe modeling can improve the process.

Pipelines are vulnerable to damage from mudslides in areas with heavy silt and steep slopes. Source: EQECAT Inc.
Click here to enlarge image

The EQECAT model enables the most comprehensive loss estimates for offshore assets. The model considers not only physical damage to platforms, but also OEE, well control, debris removal, BI, and CBI losses. The model simulates thousands of hypothetical hurricane events, accounting for waves and current, wind, tidal surge, and mudslides to estimate platform and pipeline damage. The damage functions have been validated with data from hurricanes Ivan, Katrina, and Rita. One innovative aspect of the model is its ability to quantify the risk of disruption of product delivery to onshore facilities due to pipeline damage. This is possible through a network analysis that considers pipeline connectivity, redundancy, and the impact of wave scouring and mudslides to determine residual pipeline capacity after a storm. The program estimates the initial loss of oil and gas product delivery capacity, and also takes into account the time needed to restore full delivery capability.

The paths of some of the major hurricanes in the Gulf from 1950-2006. Source: EQECAT Inc.
Click here to enlarge image

After the storms in 2004 and 2005, offshore energy facility insurance capacity was greatly reduced and pricing substantially increased, making insurance cost prohibitive for many operators with BI and CBI. “Insurers are trying to make up for their losses by raising premiums,” says Aon Corp. as reported by Bloomberg. “Coverage for wind damage to offshore facilities costs three or four times as much as before Katrina. The amount of coverage offered has dropped about 70%.” This shortage of capacity created in some instances a crisis for operators needing insurance to finance costly production facilities. The largest operators are in position to self-insure, but many others need the protection.

Offshore-tailored model

Better understanding of risk by both insurers and operators will help to rationalize the current market, leading to better risk transfer decisions. Improved risk modeling tools will help insurers to regain confidence in pricing risk for offshore energy exposure. The problem with many of the models is they are deterministic in nature. These models give only crude damage and loss estimates, with little differentiation of risk. The problem with most currently available probabilistic models is that they were designed largely for estimating wind losses for onshore commercial and residential properties. They have been adapted to cover the Gulf area, but lack the specialization needed to properly model the agents of damage or the vulnerability of these highly specialized assets. Some models lack adequate financial representation of the types of losses that can occur and how insurance policies would handle these losses.

The EQECAT Gulf of Mexico US Offshore Energy Model uses a network analysis method to model platform to onshore connectivity, pipeline damage, residual capacity, and restoration time. Source: EQECAT Inc.
Click here to enlarge image

Quantifying risks for the offshore energy market is different and more challenging than risk assessment for onshore property, which is why it is critical to use a model developed specifically for this risk. Most onshore hurricane damage is generated by wind. Offshore platform damage is caused primarily by waves with some topside damage from wind. Pipeline and wellhead damage is caused by wave and current action, and sometimes by errant rigs dragging anchors. Specialized vulnerability functions must be used that reflect changes and improvements in platform design standards. It is also critical to account for water depth, air gap, age, and configuration of a platform to enable platform damage estimates.

A significant component of loss is associated with interruption of oil and gas production or shut in capacity. A major element of loss in ability to deliver product to onshore facilities is pipeline reliability. In some cases, product can be delivered through redundant routes. A sophisticated “network” analysis is required to understand what the residual capacity of the pipeline network will be after a severe storm, which platforms are most vulnerable to reduced delivery, and how long it will take to restore capacity to pre-storm levels. In addition, insurance policy terms and coverage types are different and more complex than land-based counterparts, making insured loss more difficult to calculate.

Many of these issues have been addressed by improvements to recently released offshore energy models. Some have integrated wave models for platforms and pipelines. Some have incorporated the necessary pipeline network modeling to properly quantify business interruption risk. Vulnerability functions have been developed to differentiate risk between older facilities and newer designs that mitigate much of the wave risk with larger air gaps. Recent claims data have enabled improvements to model damage prediction capability. As the new models are adopted, users will become more comfortable with their ability to predict losses consistent with recent loss experience. Operators will have a better tool to make platform design choices and a clearer understanding of both risk and financial returns.

All parties in the risk chain – operators, insurers, reinsurers, and the capital markets, through the use of sophisticated catastrophe modeling – will improve their understanding of risk and its potential financial consequences on their company, which will lead to better decisions on how to manage and/or insure this risk.