No abstract

Over the years, underwater inspections of analytically “fatigued” tubular joints have mostly concluded with lack of weld cracks or indication thereof. Life extension of platforms, however, continues to honor fatigue results by considering all “fatigued” joints as severed in the end-of-life models. With many severed joints, linear design assessments seldom work, typically triggering nonlinear pushovers, with fewer qualified contractors and software, higher running times, leading to risk assessments, fit-for-risk conclusions, and mitigation measures such as downgrading platform manning and additional underwater inspections. All these struggles continue today while underwater inspections continue to negate analytical fatigue results. This paper proposes two approaches to reconcile this conflict. The first is based on use of underwater inspections of “fatigued” joints to reset their lives, and also to regress an update for similar uninspected yet “fatigued” joints. The second approach is based on leveraging the statistics of the underlying S-N curves to suggest scenarios of severed joints. In the first approach, only one end-of-life extension model is retained albeit with fewer “fatigued” joints. In the second approach, several end-of-life extension models are developed albeit each has fewer severed joints consistent with S-N statistics. In both cases, the conflict between end-of-life models and underwater inspections is addressed, linear design checks become more successful, nonlinear pushovers are mostly avoided as are the ensuing risk assessments and mitigation measures. The two approaches were developed internally in ExxonMobil and successfully applied to its offshore fleets worldwide. The two methods with example applications are presented in this paper for industry consideration.

The kernel underwater inspection interval of offshore steel piled jackets in API 2SIM is 5 years. Thousands of steel piled jackets worldwide use this value to set Level II underwater inspection frequencies. However, the genesis of the 5 year interval is uncertain. It could be either based on the damage found in early platform inspections or it could have more logically followed from the ship hull Drydocking survey practice. It is worth noting, however, that the later does not directly apply to jackets which are completely different structural systems. In any case, the uncertainty in the 5 year origin, plus the lack of anomalies in most underwater inspections beg the question: are we over inspecting offshore platforms? To answer this question, we commenced a quest to establish an inspection interval methodology from fundamentals of jacket response to environmental loading and resulting reliability. A reliability-based inspection interval (RLII) methodology was born. It is based on the use of platform annual reliability degradation due to corrosion, fatigue, and potential mechanical damage. First, nonlinear pushovers of platform at onset and as expected at a future point is determined; the future model captures severed joints based on fatigue analysis and any expected corrosion damage. The lateral capacities are then mapped to a platform base shear hazard curve to estimate the corresponding probability of failure. The difference in failure probability at onset and in the future, factored by platform life, determine annual reliability degradation. With probability of failure mapped to a Normal distribution, a minimum target reliability index and a corresponding factor of safety selected, a reliability-based inspection interval can then be established. Furthermore, a cap on inspection intervals can be recommended to cover general uncertainty in finding platform damage including corrosion. In this paper, we summarize and publicize the novel RLII approach, and then demonstrate its application to example jackets. This methodology has been successfully used internally to establish inspection intervals for ExxonMobil fleets of steel piled jackets.

A comprehensive structural measurement program was carried out on Exxon's Harmony and Heritage platforms during their trans-Pacific tow and subsequent launch in 1989. The objectives of this program were to 1) provide the towmaster with wind and barge motion data during the tows, 2) collect jacket/barge response data to help define post-tow jacket inspection plans to account for actual tow conditions, and 3) collect a comprehensive set of oceanographic and jacket/barge response data for use in calibrating Exxon's tow and launch analysis procedures. Metocean conditions, jacket/barge motions, and loads in a broad range of structural members were measured throughout both tows. This paper summarizes how the measurements were performed, as well as, how the data was analyzed and used to accomplish the program objectives. INTRODUCTION In 1989 the Harmony and Heritage jackets were towed from Ulsan, Korea to the Santa Barbara Channel, offshore California. These jackets are the largest of all Exxon's platforms, standing in water depths of 1200 ft and 1075 ft, respectively. Harmony departed Korea aboard Heerema's 851 foot long launch barge on May 18, 1989 and completed the 5840 nautical mile voyage in 30 days at an average speed of 8.2 knots. After a successful launch, the barge returned to Korea for the Heritage towout on August 29, 1989. Following the same rhumbline course as Harmony, Heritage arrived offshore California 36 days later. Because of the great tow distance and the potential for severe weather (29 ft significant design seastate), much of the strength and fatigue design of the jackets was dedicated by tow conditions. In all, about 13% more steel was added to the jackets to accommodate the trans-Pacific tow, compared to a local west coast tow(1). With the potential for severe weather and the need for confidence that the jackets were fit for in-place service after completing the tow, Exxon chose to use a monitoring system for both tows. This paper describes the tow instrumentation program which addressed the tow monitoring needs as well as additional research objectives. A brief discussion of measurements conducted during jacket launch is also provided. TOW INSTRUMENTATION PROGRAM Objectives The tow instrumentation program was defined by two operational objectives and one research objective. The first operational objective was to provide the towmaster with quantitative data that could be used to decide if a change in barge heading or speed was required during severe weather. In this way, excessive jacket/barge motions could be prevented, reducing the risk of jacket overstress or fatigue damage. Referred to herein as "motion monitoring", this application of tow instrumentation has been used extensively in the past and is generally accepted as a valuable asset for large open ocean tows(2,3). The second operational objective was to collect jacket/barge response data that could be used to modify post-tow inspection plans to account for actual tow conditions. The importance of this objective was accentuated by the desire to expedite the Harmony launch so that the Heritage jacket tow could occur before encroaching too far into typhoon season.