Advances in modeling methods support moorings and risers

According to the International Energy Agency “World Energy Outlook 2013,” in 1990 the contribution from deepwater (defined as in excess of 400 m [1,312 ft]) exploration was 60 Mbbl/d. In 2012, this rose to about 270 Mbbl/d (6% of all worldwide crude). The projections for Brazil are that by 2035, 11% of conventional crude will be produced from deepwater fields. This rapid demand for deepwater field development presents an opportunity for large returns on investment but at the same time high technical and economic risk as the industry ventures into new territories.

Publications at the 2014 Offshore Technology Conference on the survey of past failures, pre-emptive replacements and reported degradations for mooring systems of floating production units show a persistence of high annual rate of mooring line failure (per year, per facility) over the last 15 years, with more than a third of failure events occurring between initial installation and the third year of operation.

Mooring incidents occur due to poor mooring integrity management such as a lack of in-service inspection and/or structural monitoring. Additionally, line failures occur due to lack of design analysis and testing, especially when innovation is introduced to the design process. DNV GL, in collaboration with Cambridge and Strathclyde universities in the U.K., has developed more innovative means of modeling ultradeepwater mooring and risers.

Ultradeepwater model-scale tank-testing

Model testing is primarily needed as proof of concept. It confirms the assessment of hydrodynamic loads on the vessel and can be used to verify assumptions. It is used to validate/calibrate numerical models (loads, motions and line tensions) and can reveal unexpected and/or highly nonlinear phenomena such as wave steepness, slamming, run-up and vortex-induced vibrations. It is particularly important for assets with high strategic importance and when innovation such as new technology, vessel concept or territory is introduced. The challenges with ultradeep water are that numerical time domain models have long simulation times due to the large number of line elements. Subsequently, design optimization is limited, and tank testing of the complete system at conventional model scales is not possible due to depth limitations in wave tank basins.

FIGURE 1. The hybrid verification approach uses an equivalent numerical model at reduced depth, which then uses a conventional model scale for tank testing followed by a model-the-model stage to calibrate the numerical model and extrapolate to full water depth. (Source: DNV GL – Oil and Gas)

This issue was addressed through the DeepStar and Verideep joint industry projects (JIPs), where line equivalence (or truncation) was developed.
The hybrid verification approach (Figure 1) uses an equivalent numerical model at reduced depth, which then uses a conventional model scale for tank testing followed by a model-the-model stage to calibrate the numerical model and extrapolate to full water depth. This work has proven model scale equivalence for certain systems such as taut leg moorings. The method for setting up the equivalent mooring system is primarily based on a mooring/riser system that replicates the static characteristics of the full depth system using an optimizer to preserve the line geometric properties. This is referred to as a passive-equivalence model, which generally leads to model test data that poorly reflect line dynamics (tension and damping).

With increased water depth and number of moorings and risers, coupling between vessel motions and line dynamics will increase. This renders the importance of developing truncation approaches for ultradeepwater moorings and risers that reflect not only the quasi-static performance in full scale but also the dynamic performance as closely as possible.

Work at Cambridge University has developed a methodology that drives the truncation based on the expected mooring line dynamics rather than the static response. The equivalent numerical model is set up at reduced depth by modeling the upper sections of each line in detail, and these terminate to an approximate analytical model that represents the rest of the line.

The analytical model is a simple spring-damper model that could be confidently replicated in a test basin. In such an approach, the truncation point (where to “cut” the line) is key, and this is based on a minimum truncation length criterion below which the transverse response of the line is inertia-dominated. Figure 2 shows some results of mooring line tension for an example case study with both truncation schemes applied (passive equivalence and localized truncation, Cambridge approach).

Technical innovation only covers one aspect for successful ultradeepwater model tests. As identified in DeepStar, further effort is required to provide guidance for when to test and what type of truncation is suitable, dealing with uncertainty in the two-step process and to improve model test schedule efficacy. To this end, DNV GL is looking to launch the DeepTest JIP that aims to develop a complete set of model test guidelines that the offshore oil and gas sector can work to when designing floating production systems in ultradeep water.

Mooring damping

In related work at Strathclyde University, the effect of line dynamics on mooring system damping is studied in a general way using numerical analysis. It shows that for a turret-moored FPSO unit, wave frequency motions will dramatically increase the mooring system’s low-frequency damping. This is because energy dissipation, which is mainly caused by fluid drag opposing the transverse motion of the line, is greatest close to the seabed for a semi-taut or catenary mooring configuration due to the geometric coupling of axial and transverse vibrations. These dynamics are highly sensitive to the period and amplitude of the wave frequency motion at the top.

FIGURE 2. Numerical analysis of a four-line spread moored semisubmersible in violent storm conditions showed that the system response (motions and mooring static/dynamic tensions) of the localized truncation model matches the response of the system in full depth. In contrast, the truncation model based on line equivalence has a poor estimation of line dynamics. (Source: DNV GL – Oil and Gas)

This work also emphasizes the importance of having a reliable estimation of the drag coefficient (Cd) for chain located at depth. Operationally, this means having accurate predictions of marine growth through monitoring and comparing the observed drag diameter with values assumed in the analysis to ensure there is sufficient margin to allow uncertainty in the marine growth prediction methods.

Strathclyde has studied the sensitivity of Cd using coupled analysis of a turret-moored FPSO vessel in hurricane conditions and revealed that an increase in Cd by a factor of two reduced vessel motions by about 40% but increased total tension by up to 20%. This forms the motivation for the second theme of research at Strathclyde—using computational fluid dynamics (CFD) to estimate Cd for chains.

Flow past a smooth cylinder is first considered to validate the CFD model against published experimental data. The CFD model of the chain consists of one full link and two half-links modeled using periodic boundary conditions, and the domain is defined with no slip walls, inflow/outflow boundaries and symmetry of lateral sides.

For the case of steady flow past the chain links, the analysis showed that flow direction does not impact the estimation of Cd and that DNV-RP-C205 recommended values provide a good average over a large range of Reynolds number (Re ~ 104 to 106). For the case of oscillating chain in still water, the analysis showed that Cd can vary up to 30% depending on the Kuelegan Carpenter number for flow with low Reynolds number. Such a variation in Cd will impact the calculation of extreme tension and therefore provides motivation for future work to analyze Cd values for chain in oscillating flow.

As industry engages in operations in increasing water depths, it becomes more important for operators to fully understand riser and mooring system design limitations.

Technology developments are progressing at a rapid pace to adapt to the new environments encountered in ultradeepwater locations. It is therefore important that modeling methods, both numerical and physical, keep pace to assess the reliability of the floating production system with confidence at the design phase.