As it becomes more fiscally feasible to pursue deepwater drilling, there will likely be a renewed interest in the field of deepwater production. At depths that approach two miles (three kilometers) subsea, and pressures that approach 20 ksi (1380 bar), there are significant challenges that arise. These challenges include ensuring asset reliability in aggressive environments, minimal access for inspection, limited options for repairs, environmental safety considerations and federal oversight. Historically, these challenges have been met using engineering judgement in the absence of a well-defined industry standards. Although there’s no replacement for good engineering judgement, in recent years, the API 17 TR8 has been developed and published to address the nuances of deepwater production.
The API 17 TR8 is a technical report intend to guide the design of high-pressure high-temperature (HPHT) subsea equipment. Structural Integrity was contracted by DeepStar (CTR 12302) to provide an independent usage of the 17 TR8 to evaluate a representative deepwater component – a 20 ksi 5-inch tee and flange assembly shown in Figure 1.
When verifying the design of the flange and tee, three methodologies were used,
all of which are acceptable per the design requirements of the 17 TR8:
ASME VIII-2 Linear Elastic Methodology (VIII-2-LE)
ASME VIII-2 Elastic Plastic Methodology (VIII-2-EP)
ASME VIII-3 Elastic Plastic Methodology (VIII-3-EP)
All three methodologies indicated that the 20 ksi operating pressure was sufficient, however there were notable differences in the permissible externally applied loads. As an example, it was found that for the design of this component, VIII-2-LE allowed for the most axial load. There is an important implication here. Depending on the selected methodology, there are multiple allowable working conditions that are acceptable. This is where good engineering judgement comes into play. It’s important for the operator of equipment to understand the margins (factors of safety) on all operating components. It’s also important to understand if different methodologies are used on different components in the same system to determine margin.
Furthermore, the design verification of the flange and tee required a fatigue assessment based on the expected design cycles. As a worst-case scenario from the three methodologies, the maximum fatigue damage didn’t exceed 25% over a 25 year required operating life. If a system were implemented to quantify the cyclic loading in service, there is a possibility to extend the fatigue life because design conditions are typically more severe than operating conditions.
As an additional layer of understanding to the design life, the possibility of existing cracks in the “as-fabricated” state was considered. Based on manufacturing process, non-destructive examination (NDE) technique, and quality assurance protocols, a minimum detection threshold for a crack can be set. Based on this information, it was assumed that the largest possible undetected cracks exists in the tee, flange, pipe and bolts. Cyclic loading can propagate the initial assumed flaw to a critical dimension. Ideally, the operating life is longer than the expected design. Alternatively, the initial flaw size may be found to be unacceptable, implying that there is a possibility that a undetected crack grows to a critical crack size in less than the specified design life. In the latter case, a more refined NDE technique could be developed and used to identify smaller cracks thereby increasing the life estimate.
What constitutes an unacceptable flaw? For starters, an unacceptable flaw is the smallest flaw subjected to the largest expected load that will result in fast fracture. As discussed previously, the smallest flaw is characterized by the resolution of the NDE technique, but what about the largest expected load? The largest expected load can be defined by the survival loading conditions. The survival loading condition is an unplanned event with less than a 0.1% chance of occurring in the total design life of a component. Survival events do not result in failure, but can result in irreversible degradations that can greatly decrease the service life.
Based on a design margin of 1.05 (Global Plastic Collapse Design Margins from API 17 TR8), two survival loading conditions were defined: one with an axial load in the direction of the connected pipe, and the second with a bending load applied to the end of the pipe. In practice, there could also be scenarios where a combination of axial and bending loads could act simultaneously. Crack stability was evaluated at critical locations (shown in Figure 2) for both loading conditions. It was discovered that cracks located at the intersection of the bores in the tee, and cracks in the flange neck are the most “at-risk” when subject to the survival loads (see cracks 1, 2 and 6 in Figure 2).
Hydrostatic testing can also potentially influence the design life of a component. Based on the accumulation of compressive residual stresses, fatigue mean stress and crack growth driving stresses can be lower than the un-hydrostatically tested case. In this assembly, the locations that limit the fatigue life remain largely unaffected by the hydrostatic test, however there is a noticeable increase in fatigue life in locations near a pressurized surface. In a purely elastic evaluation, there is no benefit seen because residual stresses are not considered. Additionally, from a crack life perspective, it was noticed at the intersection of the two bores in the tee, the crack design life was marginally higher with higher pressure hydrostatic tests.
To date, only verification work has been performed on the 20 ksi 5 inch tee and flange assembly. To get a clear picture of this component, and to be in compliance with regulatory requirements, it is also necessary to validate (or test) the design. A failure mode effects and criticality assessment (FMECA) has been established to do just that. Included in the FMECA are potential failures, their causes, and what action can be done to mitigate them. Among others, material testing, NDE and strain gage measurements are proposed as failure mitigation techniques.
In the near future, the work performed for DeepStar CTR 12302 will be provided to the American Petroleum Institute (API) for public reference and use.