Optimization of pipelaying laybarge and stinger configuration

SLayOpti is desktop application which automatically finds optimal laybarge and stinger configuration for the chosen critical section at the pipeline route. Pipeline stress computation is obtained with Saipem’s in-house application NLay. SLayOpti application has been developed in C++ language and GUI is based on Windows forms and Microsoft .NET Framework.

SLayOpti relies on external data source (Fix database module) to get all needed information and data to perform laying analysis or optimization. Fix database module is XML (eXtended Markup Language) formatted database file of Saipem’s laybarges and stingers, created for using barge and stinger data in SLayOpti. Database is enhanced and expanded with important data concerning analysis and optimization. XML database file is easily editable with any text or XML editor, which makes easy to add or edit barges or stingers. Laybarge data and associated stingers data read from the Fix database are presented in graphical user interface in several places: number of laybarge supports, number of stingers supports, supports elevation limits and allowable reactions forces, support type (fixed, discrete, continuous), range of available nominal tension, other data not directly visible in GUI (laybarge and stinger geometry, supports position data, buoyancy tanks data, etc.). SLayOpti can be used as GUI (Graphical user interface) to configure, organize and run NLay structural analysis for multiple sections. Results are obtained from NLay output file and can be displayed in table or plot views. Preprocessing and postprocesing of NLay files and execution of NLay is done automatically.

Static laying analysis optimization will have a goal to find optimal combination of variables that define S-lay method laying configuration. Complete S-Lay laying optimization process is defined by sets of constant (external) and variable parameters. Constant (external) optimization parameters are specific for particular pipeline section and define optimization environment as a set of variable stages during laying process. In this group are listed only parameters that present NLay solver input parameters that influence analysis results and are derived from pipeline and concrete coating properties (concrete coating properties are included with and described by pipe weights and equivalent Young modulus): Water depth, Pipe OD, Pipe wall thickness, Pipe weight in air, Submerged pipe weight, Pipe steel yield stress (SMYS), Pipe equivalent Young modulus (pipe steel with concrete coating). Variable optimization parameters are parameters which values should be designated during laying analysis. Those parameters are classified as follows: Laybarge variables (Buoyancy tanks spacing, Tensioner force, Supports configuration, Trim angle, Draft), Stinger variables for every attached stinger component (Rollers configuration, Stinger angle, Floating stinger ballast plan).

Optimization requirements are divided into two groups according to necessity of their fulfillment after the optimization process:
1. Mandatory optimization requirements - compliance to those requirements is obligatory and they will serve as a basis for definition of optimization constraints functions and whole optimization process
2. Additional optimization requirements - optimization process should aspire to fulfill those requirements to the maximum possible extent. Majority of those requirements could present the basis for definition of optimization objective function. Optimization according to these requirements will be configurable and at least one additional optimization requirement should be selected in order to define optimization objective function. If more than one of additional optimization requirements is selected, then the optimization problem will be multiobjective and in that case for every requirement (goal) there will be a goal weighting factor representing relative importance of particular goal compared with other goals (requirements).

Pipeline laying section configuration optimization is handled by multiobjective constrained genetic algorithm extended with optimization procedure that mimics engineering practice. Multiple optimization objectives are transparently and flexibly configurable via flexible analysis criteria definition. It is possible to flexible include and exclude optimization variables and custom define discretization steps for each optimization variable. User does not have to be acquainted with nonlinear optimization and genetic algorithm.

Copyright 2013 - Department of Fluid Mechanics and Computational Engineering