Process Engineering

Safety – Reactor thermal runaway

System:

The reaction considered in this example is normally carried out isothermally at 175 °C and about 35 bar. The ambient temperature of the cooling water in the heat exchanger is 25 °C. By adjusting the coolant rate, the reactor temperature could be maintained at 175 °C. At the maximum coolant rate, the ambient temperature is 25 °C throughout the heat exchanger.

Simulation:

A batch process with exothermal reaction needs external cooling. To increase production rate, it is important to understand the coupled dynamic effects of cooling capacity and reaction kinetics. In this example, simulations of the failure of the cooling system with different filling levels of the reactor are analyzed.

Results:

With dynamic system simulation we can observe the effects of temporarily cooling failures. For an increased initial filling level and despite the cooling system at its full capacity, the necessary heat cannot be removed from the reactor. Here, the increased initial filling level leads to temperature runaway and explosion of the reactor.

‍

PtX - Optimal thermal management of methanation process

System

In this example a cooling system is designed for optimal dynamic operation of a methanation reactor. The reactor is modelled with a heterogenic 1D-model considering detailed heat transfer, reaction kinetics and pressure drop to enable the analysis of temperature hotspots during start-up and load changes. The model is adjusted to work well within our dynamic optimization software MUSCOD.

Optimization:

The ideal cooling temperature and mass flow rate is determined with respect to optimal methane yield during reactor start-up as well as positive and negative load changes.

Results:

As part of the design of the cooling system, co-current flow is determined as the most favorable for the reactor under consideration. In addition to the optimal cooling temperature profile for safe and fastest possible run through of the load profile, the optimization results also reveal potentials for reducing the start-up time by increasing the coolant mass flow. The increased load flexibility by shortening the start-up process can potentially reduce the costs of upstream buffer storage of hydrogen.

High performance distillation column

Efficient parallelization of flash calculations leads to significant speed improvement on multi-core machines. The effect of parallelization is demonstrated at example of the distillation of a 6-component mixture on a 112-stages column. A step change on the feed composition affects the composition of the head product as shown in the graphic below. The distillation column is modelled including varying pressure and vapor hold-up using a U,V-flash approach.

The resulting system of differential equation is solved by state-of-the-art variable step size solvers. Multi-species multi-phase fluid property calculations are handled by Multiflash™, an industry approved advanced thermodynamics package with very efficient internal numerical solvers.

System:

The impact of parallelization on computational simulation time is demonstrated by example of the distillation of a 6-component mixture on a 112-stages column. The distillation column is modelled including varying pressure and vapor hold-up using a U,V-flash approach. The pressure is controlled by the condenser heat stream and the liquid level with the bottom flow rate.

Simulation:

Starting from a uniform distribution of components on all stages the transient behavior the system can be analyzed. Additionally, a step change on the feed composition affects the composition of the head product as shown in the graphic below.

The resulting system of differential equation is solved by state-of-the-art variable step size solvers. Multi-species multi-phase fluid property calculations are handled by Multiflash™, an industry approved advanced thermodynamics package with very efficient internal numerical solvers.

Result:

Efficient parallelization of flash calculations leads to significant speed improvement on multi-core machines: We achieve a 14.6 speed-up on a 16-core machine!