It may have been a quiet hurricane season, but as Keith Leung discovers, there are still plenty of challenges facing offshore energy in the Gulf of Mexico.

Despite the widely-held expectation in the insurance community of another disaster-filled year, the hurricane forecasts for 2006 have been adjusted downwards since August, with analysts citing factors from sea level pressure to Saharan dust as reasons for a relatively benign season so far.

Last year, hurricane losses shocked the insurance community and raised many questions about the apparent inadequacies of the risk models, with catastrophe-modelling professionals receiving their share of the blame. In response to this, the models were altered and rewritten, and much more attention was paid to offshore energy, particularly in the Gulf of Mexico.

But serious issues remain. Despite the ongoing improvements and understanding of science in the models, the underlying exposure data has seen very little improvement. Given the unpredictability of nature's behaviour, using flawed exposure data seriously jeopardises the process of risk management for operators and insurers alike. According to AM Best, management needs to be "acutely aware of the specific issues that relate to their own geographic exposure and have the ability to properly manage those risks with accurate data".

Numerous issues

The issue of inaccurate or incomplete exposure data is not unique to offshore energy. Similar problems, for example, are experienced in property insurance, although street addresses, construction characteristics and specific occupancy types of individual risks are increasingly being used to analyse and quantify property risks. The challenges facing the offshore energy market, however, are somewhat less straightforward.

What information does an insurer need to fully understand the risks associated with an offshore platform in the Gulf of Mexico? The information is not dissimilar to that required by property insurers, essentially covering platform location, value and physical characteristics such as the year of construction and platform type.

In theory, obtaining accurate exposure information for platforms in the Gulf of Mexico should not be difficult. The number of platforms and mobile units is relatively small - something in the region of 5,500 - and accurate platform location can be easily obtained through third parties such as the Mineral Management Services of the USA and some commercial providers. These data providers produce location information in terms of latitude, longitude and block, updated on a regular basis, and supply details such as year of construction and water depth.

Despite this, the Gulf of Mexico still appears to suffer from the lack of good exposure data. Although precise location information for each platform and mobile unit is readily available, it appears to be rarely significant in underwriting decisions. Most offshore energy risks present location data at a block level (normally representing a 5km by 5km area). In worst case scenarios, underwriters might only be able to use a coarser way of recording location eg at the protraction level (an area measuring up to 200km across). This is almost certain to produce estimates of limited worth when calculating the proximity of a platform or mobile unit to the anticipated path of a storm.

For example, in the 2006 Lloyd's realistic disaster scenario (RDS) calculations for Scenario 17 (offshore), the prescribed damage ratios for physical damage are 25% and 10% of sums insured for locations within ten miles and those between ten miles and 25 miles of the RDS storm track respectively. The difference in the two damage ratios is significant and the lack of precision in the location could therefore potentially influence (either over-estimating or under-estimating) the calculated ultimate physical loss.

Sometimes only the platform or complex name is used to represent the location of a risk. Generally the risk capturing process entails significant data entry so insurers often end up with ungeocoded locations, ie there are no specific details of latitude or longitude. This is of particular concern if the insured values for such locations are high or if there is a high concentration of values.

Multiple ownership

But location data is only part of the problem. Another issue surrounds data regarding asset values or sums insured. Platforms in the Gulf of Mexico region are often owned by more than one party. The main owner of a platform is its designated operator, and any platform can be leased to numerous leaseholders, each owning a percentage interest in the platform, which are often traded amongst leaseholders with some frequency.

Individual owners of a platform will usually arrange separate insurance for their interest. This makes sense when it comes to business interruption or liability coverage in view of each leaseholder's needs. However, the same logic surely does not apply when insurance against physical damage is considered.

When assessing physical damage to a platform it would be helpful if the asset was regarded in its entirety, on the 100% basis. The fact that different platform owners utilise different valuation criteria and arrange their own cover without any consultation with one another means it is not possible to know if the platform is fully insured for physical damage. Indeed, it is not uncommon to see submissions from various owners on the same platform using different 100% asset values. A situation may arise where an owner of a platform with a smaller interest can obtain insurance with a higher physical damage limit than another owner who has a larger interest in the same platform. This could ultimately lead to issues on consistency and fairness of pricing.

Multiple ownerships can lead to another issue that relates to the application of policy limit, excess and deductibles. If the terms of the policy are to be applied on the "assured interest" basis, the limit, excess and deductibles are designed specifically for the individual owner. If the terms are to be applied on the "100%" basis, this means the terms applicable would have to be netted down to take into account the ownership share of the platform by the insured.

The consequences are twofold. First, the leaseholders' interest in the platform varies in percentage terms. In other words, netting down the policy limit, excess and deductibles accurately for all platforms covered under a policy may not be possible. The second problem is a situation whereby the basis upon which the policy terms are to be applied is inconsistent across the terms, eg the limit is applied on the "100%" basis while the excess is applied on the "insured interest" basis.

Another problem relates to the complexity of policy coverage. A single platform can be covered for physical damage, control of well, loss of profit, removal of debris, seepage and pollution, third-party liability, business interruption, contingent business interruption and so on. Each of these could have a different geographical coverage. For instance, while the limit for control of well is often applied on a block level, the one for physical damage can be applied on the whole policy. This creates the complication that losses have to be calculated separately for different coverages under the same policy before the total loss for a given platform owner can be estimated. This can be further complicated by the existence of such arrangements as Oil Insurance Limited (OIL), which typically covers the losses for all coverages (except business interruption) in excess of $250m, for example.

Simple solution

So how can the insurance community address these issues? In essence, it boils down to good communication among the many parties involved in the offshore energy value chain.

For example, if you look at the chain of communication for a typical offshore platform, first there could be multiple owners of the asset and second, multiple energy brokers could be involved in placing the various portions of risk in the market. Often, more than one insurer is involved in underwriting each insurance contract. The flow of information then continues depending on whether reinsurance is arranged by those insurers and if retrocession is utilised by the reinsurers.

Although there are numerous chains of communication, arguably the most important step is to ensure the quality of data of the original source - the platform owners. Failure at this point will have a negative impact on the quality of any analysis all the way down through the chain. Monitoring aggregates, the validity of catastrophe modelling and RDS analysis become highly questionable if the original data is accurate or inconsistent.

Currently there is no standard format for offshore risk data. The quality of data received by insurers fluctuates immensely and the variations are vast across cedants and brokers. Such fluctuation in quality damages the confidence in the insurance market as well as those providing capital to support this risk class, leading to an unstable marketplace. Having established standards for the provision of risk data in terms of the platform location and total value would significantly improve the quality of analysis, giving all interested parties a better understanding of their risks.

For a standard of information to emerge, there needs to be ownership of the processes. This seems to be one of the areas where brokers can add a tremendous amount of value. Brokers have always been pivotal in facilitating process and coverage improvements within the industry. Of course, this would not be effective without the involvement of underwriters, as platform owners also need to see the benefit of providing detailed platform information to their insurers.

Indeed, the achievements of Lloyd's brokers and underwriters in the offshore energy market have recently been recognised and honoured in the US by the Offshore Energy Center Hall of Fame. The foresight of the individuals in this field in the 1950s are being called on now more than ever.

It is crucial not to underestimate the scale of change required or the enormous frictional costs currently generated. The offshore risk business is complex from all perspectives. But there are areas where the current uncertainties can be significantly reduced and understanding of the risk can be significantly enhanced through attention to basic risk data.

- Keith Leung is an associate at JLT Re


- OIL shareholders agree windstorm proposal extension

- Aon Re: Hurricanes dampening insurance profits

- Catlin: Coastal dwellers should pay for the risk.