Mar-2006

Proper interpretation of freezing and hydrate prediction results from process simulation

This paper focuses on the modelling of solid phase behaviour in systems that are frequently encountered in natural gas processing. The ability to perform accurate calculation of freezing or solids formation conditions in processes from dry ice, hydrates, and water ice is quite important.

Michael W Hlavinka and Vicente N Hernandez Bryan Research & Engineering, Inc
Dan McCartney Black & Veatch Energy

Article Summary

Although the primary focus in this work is on dry ice formation from carbon dioxide, analogies with hydrate formation are presented. A description of the phase equilibria at different conditions of temperature and pressure is included. The paper compares the predicted results from simulation with selected experimental data sets, and illustrates that accurate results are obtained over a wide variety of conditions. However, due to the complicated phase behaviour of these systems, improper interpretation of results, or incorrect use of the tools within the simulator is possible due to the multiplicity of incipient formation points. One fact that is not well known is that lowering the temperature may cause a solid that has formed to melt under certain conditions of pressure and composition. While recent work has been done to mitigate the incorrect application of these tools, knowledge of some of the different types of phase behaviour is generally desirable to understand and exploit the results. Phase diagrams are presented to aid in understanding the solid formation behaviour.

The accurate prediction of the formation conditions of solid species is commonly required in process simulation. In gas processing, the most common solid forming species are dry ice from carbon dioxide (CO2), and hydrates and ice from water. Dry ice formation can occur in cryogenic gas processing as in turboexpander plants where methane (CH4) is typically separated from heavier paraffinic natural gas components. In particular, the top trays and the associated draws and feeds of the demethaniser in these facilities are particularly susceptible to dry ice formation when CO2 is present due to the low temperature conditions present on these trays. Water ice can also form in natural gas processing. Dry ice and water ice form a virtually pure CO2 and water solid phase, respectively. Due to the increase in volume on freezing of ice and the associated negative slope of the melting point line on a pressure-temperature diagram for water, ice cannot exist as a pure phase above its triple point (32.018°F/0.01°C). Ice will never appear in a process that is warmer than the water triple point temperature. Carbon dioxide exhibits a decrease in volume on freezing, and consequently a positive slope for the melting point line on its pressure-temperature diagram. While it is possible for dry ice to form above the triple point temperature (-69.826°F/-56.57°C), dry ice normally forms below this temperature in gas processing.

Hydrates are a particular type of solid species known as a clathrate. Hydrates enclose a guest molecule in a solid water cage, and consequently do not precipitate as a pure solid compound. While hydrates are a significant problem for gas transport in pipelines, they can form any time a hydrate forming guest species and enough water are present, and the temperature and pressure are in the hydrate formation region, which can be above the freezing point of water. In gas pipelines, enough water is frequently present to form hydrates at ambient temperatures and pipeline pressures.

In this paper, we will demonstrate that accurate predictions of solid forming conditions across all regions are provided by ProMax, the general purpose process simulator by Bryan Research and Engineering, Inc. While ProMax generally offers improved predictions over its predecessor PROSIM), the predictions from PROSIM were also good. Recent papers [1, 2] have suggested that unreliable results for dry ice formation conditions were prevalent in several commercial simulators. Although this may appear to be true at first inspection based on the results of their presentation, in actuality the values are reliable when using ProMax and PROSIM for solid formation from a system comprised of vapour and liquid (including immiscible liquids) phases if the tools are applied in the context for which they are designed. The authors are not drawing any conclusions or inferences regarding the results or behaviour of the tools present in any other process simulation program. Unfortunately, misapplication of the results is possible if familiarity with the complex phase behaviour of these systems is not considered. We have attempted to revise the presentation within our newer ProMax software to prevent incorrect application of its solid formation calculations. However, even with these efforts in program revision to remove ambiguity, we feel that writing this paper is necessary to minimize the likelihood of inappropriate application of results. Without a fundamental understanding of the phase behaviour, inappropriate application is still possible.

The solid formation utilities available in ProMax and PROSIM are quite complete. Solid formation temperature can be calculated from stream overall or individual phase composition and pressure through a dedicated stream utility. Additionally, the incipient conditions are calculated by a separate phase envelope utility for tabular and graphical display. In order to prevent modelling a system that would operate in a solid region, the program checks for incipient formation conditions in every unit operation and stream in a simulation. This also includes every stage of a distillation column, as well as internal increments within heat exchangers. An appropriate warning indicating the location of the solid formation condition is issued if operation occurs within a user defined threshold.

While most of this paper focuses on CO2 dry ice formation in a binary mixture with methane, the same fundamental principles apply to dry ice, ice, and hydrate formation in more complex systems. Since similar phase behaviour is also present in these systems, the results from these simulations should be analysed in a similar manner.

Thermodynamic review
From thermodynamics, a system of N components at a given temperature and pressure is in equilibrium when the chemical potentials for each component in all phases are equal. Mathematically this can be expressed as:

Again, the CO2 chemical potential for the solid phase is a pure component chemical potential, while the chemical potentials of the liquid and vapour phases are mixture properties, dependent on phase composition.

The equations for chemical equilibrium (Equation 1) result from the equilibrium requirement of the minimization of the total free energy of the system. Any other phase composition combination other than the equilibrium composition will result in a higher free energy than the equilibrium composition. Additionally, a phase (vapour, liquid or solid) will only form (or disappear) if it lowers the free energy of the system, driving the system to a minimum free energy level. Therefore, dry ice only forms when its presence will lower the total free energy of the system.