Page 335 - Handbook Of Multiphase Flow Assurance
P. 335

334                          10.  Research methods in flow assurance

                                                        polymer type
                              0


                             –20

                             –40
                           Eadsorption (kcal/mol)  –60



                             –80

                            –100


                            –120
                                                                                    std.dev.
                                                                                    std.dev.
                                                                                    ave
                            –140
                 FIG. 10.85  Strengths of adsorption for all simulated chemicals.



                 Guest molecules adsorption on a hydrate surface
                   In order to investigate the hypothesis of an adsorbed inhibitor hindering building blocks from
                 reaching the growing crystal surface, a separate set of MC simulations was performed. Water was
                 assumed to be abundant near the hydrate surface. Guest molecules adsorption on a hydrate sur-
                 face was modeled with a bare hydrate surface and with a hydrate surface having a preadsorbed
                 polymer. Position of the preadsorbed polymer was obtained from the polymer adsorption runs.
                                                                                        2
                   The guest molecule selected to be a guest molecule was prepared using Cerius , and small
                 partial charges were assigned to its atoms. Monte Carlo simulation for methane used the same
                 program as was used for polymer adsorption except the pivot moves were discarded com-
                 pletely, and rotation moves were attempted once in 10 cycles. Methane is a spherical molecule
                 and its rotations were almost always accepted so that acceptance was monitored only for trans-
                 lational moves. Methane molecule was atomically correct (e.g., had explicit carbon and four
                 hydrogens with non-zero Lennard-Jones diameters). It could interact both with the hydrate sur-
                 face and with the polymer fixed on surface. Only Lennard-Jones and electrostatic interactions
                 between methane and hydrate/polymer were calculated. Hydrogen bonding was neglected.

                 Validation of the methane adsorption model
                   In the hand-written code we used only the nonbonded interactions in DREIDING which
                 consist of van der Waals or dispersion (E vdw ), electrostatic (E Q ), and explicit hydrogen bond
                 (E hb )terms (Fig. 10.79):
                                                 E =  E vdw  + E + E .
                                                            Q
                                                                hb
                                                  nb
                   At one atmosphere the volume occupied by 1 mol of methane is 22.4 l. This translates to
                        3
                 37,191 Å  per methane molecule at 1 atm. The volume of the simulation box above the hydrate
                                 3
                 surface was 7597 Å  which is 5 times smaller than the volume per molecule of methane at one
                 atmosphere. This suggests that the pressure in the simulation can be estimated as 5 atm.
   330   331   332   333   334   335   336   337   338   339   340