Wolfram for Chemical Engineering | Things to Try

Make edits and run any piece of code by clicking inside the code and pressing  + .
Chemical Engineering Computation & Modeling. For everyone from students learning the fundamentals to practising engineers and research teams. A single workflow for thermophysical properties, material and energy balances, phase equilibrium, reaction engineering, transport phenomena and process dynamics—combining symbolic derivations, numerical simulation, units-aware calculations, interactive visualization and experimental data analysis.

Find properties of chemicals and fluids

Find the molecular formula, molar mass, normal boiling point and density of methanol:
In[]:=
ChemicalData["Methanol",#]&/@{"CompoundFormulaString","MolarMass","BoilingPoint","Density"}
Out[]=
CH4O,
32.042
g/mol
,
64.7
°C
,
791.
kg/
3
m

Typing “=[...]” lets you look up a compound from its common synonym using natural language:
In[]:=
=[thf]
Out[]=
tetrahydrofuran
Find the dynamic viscosity of air at 300 K and 0.2 MPa:
In[]:=
ThermodynamicData["Air","DynamicViscosity",{"Pressure"->Quantity[0.2,"Megapascals"],"Temperature"->Quantity[300,"Kelvins"]}]
Out[]=
0.0000185515
sPa
Plot the density of liquid water between 20°C and 90°C:
In[]:=
temperatures=Quantity[Range[20,90,5],"DegreesCelsius"];
In[]:=
densities=ThermodynamicData["Water","MassDensity",{"Temperature"->temperatures}];
In[]:=
ListLinePlot[Transpose[{temperatures,densities}],Frame->True,FrameLabel->{"temperature","density"},PlotMarkers->Automatic]
Out[]=

Calculate with units, formulas and dimensionless groups

Convert between different units:
In[]:=
UnitConvert[Quantity[3.5,"Megawatts"],"InternationalTableBritishThermalUnits"/"Hours"]
Out[]=
1.19425×
7
10
BTU
IT
/h
In[]:=
UnitConvert[Quantity[1.2,"Centipoise"],"Pascals""Seconds"]
Out[]=
0.0012
sPa
In[]:=
UnitConvert[Quantity[250,"Kilomoles"/"Hours"],"Moles"/"Seconds"]
Out[]=
625
9
mol/s
Retrieve the Reynolds-number equation and identify its variables and their dimensions:
In[]:=
With[{f={"ReynoldsNumber","DynamicViscosity"}},Column[{FormulaData[f],FormulaData[f,"QuantityVariableNames"],FormulaData[f,"QuantityVariableDimensions"]}]]
Out[]=
Re

l
v
ρ
η
{
Re
Reynolds number,
l
characteristic length,
v
characteristic speed,
η
dynamic viscosity,
ρ
mass density}
{
Re
Dimensionless,
l
{{LengthUnit,1}},
v
{{LengthUnit,1},{TimeUnit,-1}},
η
{{LengthUnit,-1},{MassUnit,1},{TimeUnit,-1}},
ρ
{{LengthUnit,-3},{MassUnit,1}}}
Calculate the Reynolds number for water flowing at 2 m/s through a 50 mm pipe:
In[]:=
density=Quantity[998,"Kilograms"/"Meters"^3];
In[]:=
velocity=Quantity[2,"Meters"/"Seconds"];
In[]:=
diameter=Quantity[50,"Millimeters"];
In[]:=
viscosity=Quantity[1,"Centipoise"];
In[]:=
reynoldsNumber=densityvelocitydiameter/viscosity
Out[]=
99800
Calculate the required heat-transfer area for a specified duty, overall coefficient and terminal temperature differences using the log-mean temperature difference (LMTD) method:
In[]:=
duty=Quantity[2.5,"Megawatts"];
In[]:=
overallCoefficient=Quantity[500,"Watts"/("Meters"^2"Kelvins")];
In[]:=
deltaT1=Quantity[70,"Kelvins"];
In[]:=
deltaT2=Quantity[30,"Kelvins"];
In[]:=
logMeanTemperatureDifference=(deltaT1-deltaT2)/Log[deltaT1/deltaT2];
In[]:=
area=UnitConvert[duty/(overallCoefficientlogMeanTemperatureDifference),"Meters"^2]
Out[]=
105.912
2
m

Solve a material balance and flash calculation

Here is a simple flash calculation to determine vapour and liquid fractions for a three-component feed.
Specify feed components and overall composition:
In[]:=
components={"Methane","Ethane","Propane"};
In[]:=
z={0.45,0.35,0.20};
Assume we know the vapour/liquid ratios for each component:
In[]:=
kValues={2.2,0.8,0.25};
Solve the Rachford-Rice equations for vapor fraction:
In[]:=
vaporFraction=vf/.FindRoot[Total[z(kValues-1)/(1+vf(kValues-1))]==0,{vf,0.5}];
Determine phase compositions:
In[]:=
x=z/(1+vaporFraction(kValues-1));
In[]:=
y=kValuesx;
Collect results:
In[]:=
Column[{Row[{"vapour fraction = ",vaporFraction}],Dataset@MapThread[<|"component"->#1,"feed z"->#2,"liquid x"->#3,"vapour y"->#4|>&,{components,z,x,y}]}]
Out[]=
vapour fraction = 0.534926
component
feed z
liquid x
vapour y
Methane
0.45
0.274071
0.602956
Ethane
0.35
0.391931
0.313545
Propane
0.2
0.333998
0.0834996
Visualize:
In[]:=
BarChart[{z,x,y},ChartLegends->{"feed","liquid","vapour"},ChartLabels->components,Frame->True,FrameLabel->{None,"mole fraction"}]
Out[]=
feed
liquid
vapour

Solve differential equations for a reaction system

Define equations relating concentrations of species
A
,
B
,
C
:
In[]:=
equations={ca'[t]==-k1ca[t],cb'[t]==k1ca[t]-k2cb[t],cc'[t]==k2cb[t]};
Specify initial conditions:
In[]:=
initConds={ca[0]==ca0,cb[0]==0,cc[0]==0};
Solve symbolically:
Make an interactive plot of the concentration profiles:
Find the natural modes of the reaction network:

Visualize heat or mass transfer

In chemical engineering it is often necessary to model diffusion phenomena. Here we simulate the evolution of concentration in a one-dimensional slab starting from an initial distribution and with both ends maintaining zero concentration.
Clear variables from previous example:
Define the diffusivity in the slab:
Solve the one-dimensional diffusion equation for concentration profile:
Visualize the profile over time as a series of snapshots:
Visualize diffusion over time as a density plot:

Fit models to experimental data

Start with a reaction-rate data set:
Carry out a nonlinear fit:
Inspect the parameters:
Compare fit and data:

Analyze real-world data

Here we demonstrate Wolfram Language’s powerful data science capabilities on a real-world dataset of industrial boiler time-series data.
Connect to the data source on Kaggle:
Inspect the files in the dataset:
Download the data and wrangle into a convenient form, with columns nicely labeled:
This dataset contains many columns corresponding to various sensors in the boiler:
We’ll initially consider several representative sensors:
Let’s visualize their behavior as a time series:
Characterize their distributions:
Now let’s investigate the correlation between the six upper furnace pressure sensors:
Finally, we can use anomaly detection to flag states which may signal an operational issue: