A Practical Guidance: Using Simulation for Efficient Column Operation How to save a few hundred thousand USD per year with the column optimization
Distillation is probably the most widely used separation process in the process industry; its applications ranging from the rectification of alcohol, which has been practiced since antiquity, to the fractionation of crude oil.
Because of their widespread use in the process industry, they also represent a huge consumer of energy and, therefore, a constant target of process optimization initiatives. Because of the required purity of products, distillation columns in petrochemical industry have the leading role as huge consumers of energy.
This article is providing a practical optimization guidance through the example taken from petrochemical industry. This approach, however, can be applied on practically all operating distillation columns in the industry with the goal to analyze current operating conditions and explore the opportunities for improvements.
Once a distillation column is in operation, its design including geometry, number of stages, duties of reboiler and condenser and general control strategy is defined. What can be reviewed and changed are operating conditions.
It is good to perform this type of analysis occasionally or after some changes in upstream units, to review optimal column conditions with the goal to use the minimum energy for defined product requirements. We will explore the steps to be done through optimization of reflux ratio of the propane/propylene column.
The following steps with description can be applied to any column:
1. Definition of the column
The subject of this analysis and optimization can be any column. Columns which are considered as good targets and can contribute with significant savings are columns with high reflux ratio, high utility costs, and high differential product top/bottom values. Definition of this example is given by propane propylene column and defined feed composition.
2. Formulation of the objective function
An objective function is a measure of profitability and includes factors as profit, cost, safety, energy, feed&products yields, product quality etc. For this example, the objective function is related to the profits of yields vs. costs of utilities needed for their production, given by the following equation:
3. Selection of appropriate tool to build a model
You will need to define an available and appropriate tool to perform this analysis.
We will use Chemstations Chemcad Simulator on this particular example. You can use any other process simulation tool such as:
CHEMCAD (Chemstations, Inc.)
HYSYS (Aspen Technology, Inc.)
SimSci Pro/II (Schneider Electric Software)
UniSim Design Suite (Honeywell)
DWSIM - open source simulator
or any other that is listed here.
4. Creating the flowsheet
A simple exercise of drawing a process flowsheet. You need 1 inlet and 2 outlet streams and a column. We suggest choosing a column type with the condenser and reboiler already included in the column symbol.
5. Definition of the column variables and parameters
This step includes a definition of components and a feed stream. Our example is propane/propylene column, therefore, those two components need to be selected in the component list.
|Feed Temperature, °C||25|
|Total Flow, kg/h||9205|
Definition of thermodynamics: we will use Soave-Redlich-Kwong. You can read more about the reasons for the selection of the thermodynamics:
To define the initial calculation, column pressure, a number of stages, type of condenser, top and bottom key component and initial reflux ratio need to be defined. Example parameters are shown in the Table below.
|Column pressure, MPa||1.8|
|Number of stages||241|
|Initial reflux ratio||10.1|
|Condenser duty, MJ/h||-33400|
|Reboiler duty, MJ/h||33200|
6. Column solving
This is a relatively simple calculation problem, so no convergence issues should occur in solving this column if you have defined all parameters as explained above. To initially solve the column, we suggest a definition of heat at the column bottom as defined in the table Initial parameters. Ones your column has converged, it is appropriate to define the bottom product characteristics: mole fraction of propylene in the bottom product to be 0.05.
7. Profitability analysis
Once a distillation column is in operation, the number of trays is fixed and very few degrees of freedom can be manipulated to minimize operating costs. The reflux ratio is frequently used to control the steady-state operating point and optimize column energy consumption.
The optimization of reflux ratio is particularly attractive for columns that operate with
- High reflux ratio
- High differential product values (between overhead and bottoms)
- High utility costs
- Low relative volatility
- Feed light key far from 50 percent
Propane-propylene column matches practically all of the mentioned assumptions.
We analyzed how the change in the column pressure is influencing the reflux ratio and energy consumption of the column. The goal is to keep the feed and the product compositions and flows constant and to be focused on the energy consumption.
Results are shown in the table below:
|Pressure, MPa||Reflux Flow, kg/h||Condenser Duty, MJ/h||Reboiler Duty, MJ/h|
The results are showing how pressure reduction in the column is influencing reflux flow reduction and reduction of energy consumption. The most significant saving is the reduction of the steam used to evaporate the bottom of the column.
Between the maximum and minimum analyzed pressure, there is a difference of close to 1500 MJ/h. This number converted to the cost of energy is between 300-350 thousand USD per year, dependant of the cost of the steam. It is a significant amount of money that shouldn't be neglected. The column operation should be regularly checked to make sure it is operating under optimal operating conditions.
The procedure and analysis presented in this article could and should be applied to any column and the results of analysis compared to the data obtained in the column operation. There is always a space for improvements which can be checked with the aim of a process simulation. By missing this improvement and operating the column away from the optimal point, you could easily end up with the energy cost a few hundred thousand higher that is necessary.