High Pressure Gas Liquid Separation - II

The objectives of the HIPGLS-II project have been to improve the industries fundamental understanding of high pressure gas-liquid separation and its capability to model such processes. A challenge has been that key physical properties of these systems are poorly described in the literature. Thus, high pressure experiments have been carried out for some binary and ternary hydrocarbon mixtures. This data has been used to develop correlations for the physical properties of the systems. A special feature of h igh pressure hydrocarbon systems is the marked drop of surface tension and the reduction of the density ratio between the phases in the system. This influences droplet size distributions, coalescence efficiencies, breakup probabilities and distributions a s well as entrainment and deposition of droplets from wall films in the scrubbers. Experimental investigations have therefore been undertaken in an experimental scrubber unit (ID 252 mm) under realistic process conditions. Effort was made to the effect of the scrubber inlet momentum on the scrubber unit's performance. High quality data was obtained for the holdup in the scrubber column as a function of operating pressure, gas-load factor (K-value), inlet momentum, volume fraction liquid and height inside the otherwise empty column. Two model frameworks have been developed in order to integrate droplet dynamics into commercial computational fluid dynamics software. The main challenge has been to faithfully model the flow and the evolution of the droplet si ze distribution through a separation unit. Two frameworks have been developed. The first was based on quadrature-method-of-moments (QMOM/SQMOM), while the second framework was based a Lagrangian elements method. This latter framework has proven to be more robust than QMOM/SQMOM, and has been used in the last part of the project to model experimental data. The frameworks developed are useful in enabling modeling of droplet size distributions in scrubber units and summarize the continued need for high quali ty experimental data. High quality data is needed in order to develop model closures needed in the framework. The main challenge that still remains in closing this gap is measurement techniques that yield high quality data under high pressure conditions. Målene med HiPGLS-II prosjektet har vært å forbedre industriens fundamentale forståelse av høytrykks gass-væske separasjon og deres mulighet til å modellere slike prosesser. En utfordring er at viktige fysiske egenskapene til systemene er dårlig bes krevet i litteraturen. Høytrykks eksperimenter er derfor blitt utført for noen binære og tertiære hydrokarbonblandinger. Disse dataene har igjen blitt brukt til å utvikle korrelasjoner for fysiske egenskaper til systemene. En fremtredende egenskap ved hyd rokarbonsystemer når trykket øker er et markert fall i overflatespenningen og tetthetsforskjellen mellom fasene i systemet. Dette influerer på størrelsesfordeling, effektiviteten av koalesens og oppbrytning av dråper i systemet, samt på oppfanging og avre nning av dråper fra veggfilmer i gass-skrubbere. Eksperimentelle undersøkelser er derfor blitt utført i en eksperimentell skrubberenhet (indre diameter 252 mm) under realistiske forhold. Hovedfokus har vært på effekten fra innløpsmomentet i skrubberen på skrubberenhetens effektivitet. Høykvalitetsdata er blitt oppnådd for "hold-up" i skrubberkolonnen som en funksjon av operasjonstrykket, gass-lade faktor (K-verdi), innløpsmoment, volumfraksjon væske og høyde inne i en ellers tom kolonne. To rammeverk for modellering er blitt utviklet i prosjektet for integrering av dråpedynamikk inn i kommersielle dataprogrammer for fluiddynamikk. Hovedutfordringen har ligget i en troverdig modellering av strømningsmønster og utviklingen av dråpestørrelsesfordelingen gjen nom en separasjonsenhet. To rammeverk er blitt utviklet, hvor den første ble basert på såkalt kvadratur-moment metoden (quadrature-method-of-moment) (QMOM/SQMOM). Det andre rammeverket ble basert på en Lagrange-element metode. Det siste rammeverket har vi st seg å være meget robust og er blitt brukt i siste del av prosjektet for å modellere eksperimentelle data. Utviklingen av rammeverk for modellering av dråpestørrelsesfordelingen i skrubberenheter har vist behovet for mere detaljerte høykvalitetsdata ved reelle forhold. Disse dataene trengs for å utvikle lukningsmodellene i rammeverket. Hovedutfordringen videre for å tette dette gapet er behovet for nye eksperimentelle måleteknikker som kan fremskaffe disse dataene.

Fundamental understanding of high and medium pressure separation of liquids from gas is currently limited. This is due to the fact that most experimental studies have been performed on model systems at low pressures. Thus they are unsuited to account for the behaviour of real systems at elevated pressures. Existing high pressure facilities elsewhere are not suited for detailed studies such as those performed in the preceeding HiPGaS/HIPGLS projects. The combination of small scale cells covering low to m edium pressure at NTNU/SINTEF combined with the intermediate scale separation rig at Statoil which can handle medium to high pressures offers a unique research combination into the detailed dynamics of droplet dynamics. A literature survey in HiPGLS [L. Patruno OTC 2010] showed that high pressure data mainly consisted of nitrogen/air-water observations. However, this data cannot reliably be used to guide separation equipment for hydrocarbon systems since it does not cover the realistic range of density and viscosity ratios as well as surface tension ranges of hydrocarbon systems. Thus there is a need for fundamental studies of fluid properties, droplet sizes and size distributions, entrainment and deposition mechanisms in intermediate to high pressure s. What physical and process parameters influence coalescense? What is the liquid dynamics inside a flooded meshpad look like? Can new guidlines be developed for NORSOK design rules of gas scrubbers? Can engineering correlations, point models and CFD tool s be improved with respect to design of separators and their internals? These are the questions this project sets out to find data and models to answer.