Session: 04-02-03 Rigid Risers III

Paper Number: 79044

79044 - Cost-Effective Optimal Solutions for Steel Catenary Riser Using Artificial Neural Network 

Free hanging steel catenary riser, SCR, has been regarded over decades as a cost-effective and simple system for riser solution in deepwater and ultra-deepwater by National oil companies NOCs, international oil companies IOCs and independents players. However, the free-hanging SCR comes with its owns challenges, which can result in system failures if not properly evaluated during design stages. The three prominent challenges facing the free-hanging SCR usage are high increased self-weight, yielding failure, and high fatigue damage especially at the touchdown point and hang-off point. Fatigue damage is the failure resulting from stress accumulation during the riser design life. The involvement of new techniques has over the years improved the design method for these systems. two of such innovative contribution is the inclusion of optimization algorithm and artificial neural network, ANN, algorithm. The aim of this research is to automate the process of riser design, with the implementation of an optimization process as a key resource for initiating and finalizing the acceptance of a reliable riser system. The method involves the combined use of ANN and genetic algorithms GA in determining the optimum solution of the SCR. The genetic algorithm was integrated into the process because of its capability in handling a variety of complex nonlinear optimization problems in order to ascertain the global optimum solution, and its capacity to self-moderate the number of calls to the fitness function and the constraint so as to reduce the computation time. While the ANN method is employed for its accuracy in the prediction of nonlinear finite element analysis FEA response. The method has shown to be promising due to its time save and less computational cost when compared to integrating GA and finite element analysis FE model. This method was illustrated using a prospective 10-inch SCR to be installed in a 2000 meter deepwater offshore field off the coast of the oil-rich Niger Delta region, Nigeria. The obtained results are in agreement with the GA-FEA method and show a 90.83 percent reduction in computational cost.

KEYWORDS: finite element model FEA, artificial neural network ANN, genetic algorithm GA optimization, steel catenary riser SCR

Presenting Author: Joshua Abam Newcastle University

Authors:

Joshua Abam Newcastle University
Yongchang Pu Newcastle University
Zhiqiang Hu Newcastle University