Use of LLM/ChatGPT and NeqSim

[EvenSol]

In this thread we can discuss how input to ChatGPT/language models can be formulated to generate useful python code with minimal need for manual corrections.

[EvenSol]

AI tool:
ChatGPT-4

Date:
04.11.2023

Input to chatGPT
make a python script using neqsim that takes a feed stream with at 50C and 50bara of 50% methane and 30% n-decane and 20% water . Set the flow rate to a total flow rate 20000 kg/hr. Separate it in a three phase separator. The gas from the separator is sent to a compressor that compresses the gas to 100 bar. Set the adiabatic efficiency of the compressor to 75%. Calculate the compressor power.

ChatGPT genrerated code:

import neqsim
from neqsim.thermo import fluid, phaseenvelope
from neqsim.process import clearProcess, stream, separator, compressor, runProcess

# Define the feed stream
feed = fluid('srk')  # Using the Soave-Redlich-Kwong equation of state
feed.addComponent("methane", 50.0, "mol%")  # Adding components by mol%
feed.addComponent("n-decane", 30.0, "mol%")
feed.addComponent("water", 20.0, "mol%")
feed.setTemperature(50.0, "C")
feed.setPressure(50.0, "bara")
feed.setTotalFlowRate(20000, "kg/hr")  # Set total mass flow rate

# Create a stream
feedStream = stream(feed)

# Three-phase separator
threePhaseSeparator = separator(feedStream, "Three phase separator")

# The gas from the separator
gasFromSeparator = stream(threePhaseSeparator.getGasOutStream())

# Compressor
compressor1 = compressor(gasFromSeparator)
compressor1.setOutletPressure(100.0)
compressor1.setIsentropicEfficiency(75.0)

# Run the process
clearProcess()
feedStream.run()
threePhaseSeparator.run()
gasFromSeparator.run()
compressor1.run()

# Retrieve results
compressorPower = compressor1.getPower()  # Compressor power in kW

print(f"The power requirement of the compressor is: {compressorPower:.2f} kW")

Corrections needed to make the script run

1. Change component name of n-decane to nC10
2. Calculated power has unit Watt.
3. Use 3 phase separator (separator3phase)

Corrected script that runs:

import neqsim
from neqsim.thermo import fluid, phaseenvelope
from neqsim.process import clearProcess, stream, separator3phase, compressor, runProcess

# Define the feed stream
feed = fluid('srk')  # Using the Soave-Redlich-Kwong equation of state
feed.addComponent("methane", 50.0, "mol")  # Adding components by mol%
feed.addComponent("nC10", 30.0, "mol")
feed.addComponent("water", 20.0, "mol")
feed.setTemperature(50.0, "C")
feed.setPressure(50.0, "bara")
feed.setTotalFlowRate(20000, "kg/hr")  # Set total mass flow rate

# Create a stream
feedStream = stream(feed)

# Three-phase separator
threePhaseSeparator = separator3phase(feedStream, "Three phase separator")

# The gas from the separator
gasFromSeparator = stream(threePhaseSeparator.getGasOutStream())

# Compressor
compressor1 = compressor(gasFromSeparator)
compressor1.setOutletPressure(100.0)
compressor1.setIsentropicEfficiency(75.0)

# Run the process
clearProcess()
feedStream.run()
threePhaseSeparator.run()
gasFromSeparator.run()
compressor1.run()

# Retrieve results
compressorPower = compressor1.getPower()  # Compressor power in W

print(f"The power requirement of the compressor is: {compressorPower:.2f} W")

Output:
The power requirement of the compressor is: 1099.11 W