New experimental techniques for validation of marine discharge models

Abstract

A new three-dimensional Laser-Induced Fluorescence (3DLIF) system was applied to study the mixing of a produced water outfall discharge typical of that used in offshore production operations. The outfall consists of a pipeline terminating in a multiport diffuser discharging into water about 12 m deep. Scaled laboratory experiments were done using conditions based on measured oceanographic conditions near the site. The current was perpendicular to the diffuser, and the receiving water was linearly stratified. The model was undistorted and the jet densimetric Froude number was the same as in the prototype. This scaling results in correct modeling of the source volume, momentum, and buoyancy fluxes, and equality of other line source dimensionless parameters. The experiments were conducted in a constant refractive-index environment which enabled detailed three-dimensional concentration measurements in the near field mixing zone to be obtained despite the density variations in the flow field. The mixing processes were found to be complex. The jets discharging upstream were quickly swept downstream, where they began merging with the downstream jets. Lateral mixing caused the concentration profiles to quickly become laterally homogeneous. Initially, dilution increased rapidly with distance from the diffuser, but farther away the rate of increase of dilution slowed as the turbulence became affected by the ambient density stratification. The dilution eventually becomes approximately constant; the location where this occurs defines the end of the near field, where mixing is primarily due to turbulence and other processes induced by the discharge itself. The end of the near field is marked by collapse of the self-induced turbulence under the influence of the ambient stratification. The results were compared to predictions of commonly used mathematical models of marine mixing zones: RSB, UM3, and CORMIX. The best predictions of the near field dilution and the length of the near field were by RSB. UM3 gave good predictions of the near field dilution, but CORMIX overestimated the dilution and rise height substantially. It is believed that this overestimation by CORMIX is due to its approximation of the diffuser as a row of vertically discharging jets of equivalent total momentum flux; this is clearly a poor approximation to the diffuser conditions modeled here. 3DLIF is an exciting new technique that can give unprecedented insight into the complex hydrodynamics and processes occurring in buoyancy-modified mixing zones and to improvement of near field mathematical models. Experiments on other discharge configurations are continuing.