AmericanChemicalSociety.com

Natural gas supplies ~23% of U.S. energy needs. For off-shore wells, treatment near the wellhead is critical to prevent clathrates from forming and plugging the pipeline as gas flows to the mainland. The raw gas is normally treated by adding chemicals or reducing its dew point, but standard processing equipment is often large and requires manned platform operation. An alternative approach is to use supersonic natural gas separators that (1) cool the gas in a supersonic expansion to induce droplet formation and growth, (2) separate the droplets from the gas, and, (3) recompress the dried gas using a diffuser to minimize pressure losses. These separators are smaller than traditional process equipment, have no moving parts, and require no chemicals. Thus, they are suited for both off-shore and sub-sea applications.

 As these devices are adopted, critical questions remain regarding droplet growth in these complex vapor mixtures. The goal of this work is to improve our fundamental understanding of water - hydrocarbon droplet growth at cooling rates comparable of those found in the novel separators. In particular, hydrocarbons can inhibit water condensation because they wet the water surface while water cannot wet the hydrocarbon surface. Since removing water is critical to preventing clathrate formation, understanding co-condensation in this highly non-ideal system is vital to the success of supersonic separator technology and efficient natural gas recovery.

Our experimental apparatus includes a series of supersonic nozzles with cooling rates that match the supersonic separators. The characterization methods include pressure trace measurements, Fourier transform infra-red (FTIR) spectroscopy measurements of the gas and liquid phases, and in situ small angle x-ray scattering (SAXS) or neutron (SANS) measurements to directly follow the growing droplets. To date we have completed an extensive series of pressure and SAXS measurements that investigate 3 different starting concentrations for the pure components, nonane and D₂O, as well as the 9 corresponding binary systems. Figure 1(a) illustrates the position resolved average droplet sizes of aerosols formed during condensation of pure D₂O, pure nonane and co-condensation of D₂O – nonane. Under the conditions used in these experiments, pure nonane nucleation occurs upstream of pure D₂O nucleation, at a temperature that is about 3 K warmer. The droplets formed when D₂O is present are always smaller than the pure nonane droplets, while the number density of the droplets formed via co-condensation appears higher than for the pure nonane droplets (not illustrated). This observation is surprising because under the temperatures reached in the experiment, the formation of additional particles via D₂O nucleation is unlikely. Furthermore, the non-ideality of this system suggests that binary nucleation should not be an efficient particle production pathway.  By using D₂O we can also conduct the complementary small angle neutron scattering (SANS) measurements to better observe the distribution of condensed “water” within the aerosol. Using D₂O also makes it somewhat easier to follow “water” condensation with FTIR spectroscopy.

We have also completed FTIR measurements to follow condensation of the pure components and better understand how the presence of the hydrocarbon inhibits water removal. One challenging aspect of the work is that although the vapor and liquid regions of the spectrum are well separated for D₂O, they are not for nonane. Nevertheless, by combining the FTIR and SAXS data we have developed a robust method to deconvolve the measured alkane spectra into contributions from the vapor and the liquid. Figure 1(b) illustrates the spectra corresponding to the vapor and the liquid, as well as the fit to a spectrum that comprises 60% vapor and 40% liquid. The current challenge is to measure and deconvolve the spectra measured during co-condensation of water and nonane and combine these results with the SAXS measurements to develop a more comprehensive picture of droplet growth in this complex system.

Figure 1(a): Position resolved average droplet sizes of aerosols formed during condensation of pure D2O, pure nonane and co-condensation of D2O – nonane under conditions where nonane condensation begins at somewhat higher temperatures than D2O condensation.  (b): Nonane spectra corresponding to pure vapor and pure liquid can be combined to yield a good fit to the measured nonane spectrum.