Browse Prior Art Database

New physical model for the control of distillation columns in air separation units Disclosure Number: IPCOM000197372D
Publication Date: 2010-Jul-05
Document File: 50 page(s) / 3M

Publishing Venue

The Prior Art Database

This text was extracted from a Microsoft Word document.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 6% of the total text.

New physical model for the control of distillation columns in air separation units

Cryogenic distillation columns are operated at extremely low temperature to separate air into nitrogen, oxygen and argon in Air Separation Units (ASU). The major operating cost of air distillation is electricity. AL ASU electricity consumption represents 600 M€ each year. Therefore, small improvements in process control would lead to significant savings.

Current control technology is based on linear dynamic models and linear model predictive control. Until now this linear control technology had given satisfaction over a small range of production rates. Namely, it efficiently stabilizes steady-state. However, deregulation of the electric utility will lead to frequent changes in the cost of electricity. The production of air separation units will change as well and steady-state will not represent the most common range of an ASU. With these new operating conditions come new control needs. It should be noticed that the main part to control in an ASU is the distillation column. This invention proposes a new physical model for the control of distillation columns in ASU. This model is more accurate than most of the other existing ones and it can be used to implement nonlinear purity control.

The state-of-the-art of distillation column model is presented here. Many papers about distillation dynamic modeling have been published since 2001. As far as we know, nothing concerning global ASU optimization has been published yet.

It should be noticed that:

-      At the end of every model presentation, a conclusion recalls the hypotheses and the specificities.

-      In the remaining, vertical axis is oriented up to down.

1.1. Stage-by-stage model

Symbols used in the following.


Theoretical tray height



Tray efficiency



Liquid component molar fraction



Vapor component molar fraction



Liquid molar flow, Nm3/h



Vapor molar flow, Nm3/h



Flow rates velocity



Liquid molar hold-up



Vapor molar hold-up



Relative volatility



As index, relative to oxygen



As index, relative to argon



As index, relative to nitrogen



As index, relative to stage i









Saturation pressure








Distillation columns used to be composed of a superposition of trays. The liquid was dropping from tray to tray and the vapor was going up the column through little holes in the tray, which insured the liquid/vapor contact necessary to the distillation process. Columns are now made of packings. Liquid and vapor are in contact all along the column and the distillation is more efficient.


Figure 1 Packing


However, distillation simulation models have kept the shape of a tray column with 2 empiric parameters: theoretical tray height  and tray efficiency . For example if , then a 6 meters high column is made of 20 theoretical stages.


engineering group currently uses Aspen Custom Modeler to simulate an entire ASU dynamic behavior. The distillation part of the model relies on a stage-by-stage model. The...