Reducing energy consumption and heat exchanger network

8 replies · 3K views · last reply
gfd43tg
Gold Member
Messages
949
Reaction score
48

Homework Statement


If you are allowed to make changes of column pressure and temperature, stream preheating or cooling temperature, please propose a process change that can significantly reduce the energy consumption at DTmin of 10°C. The modified process should maintain the product quality and production rate unchanged.
Aspen process flowsheet for acetone production from isopropyl alcohol, showing reactor R1, columns D1, T1, T2 and T4, heat exchangers and stream labels with flow rates and duties.
Utility Information

Low pressure steam (LP, 140°C) = 0.03 $/kWh = 240 $/kW-Yr

Medium pressure steam (MP, 170°C) = 0.05 $/kWh = 400 $/kW-Yr

High pressure steam (HP, 200°C) = 0.07 $/kWh = 560 $/kW-Yr

Hot Oil (HO, supply 450°C – return ≥ 300°C) = 0.15 $/kWh= 1200 $/kW-Yr

Electricity = 0.1 $/kWh = 800 $/kW-Yr

Cooling water (CW, 30°C) = 0.005 $/kWh = 40 $/kW-Yr

Chilled Water (ChW, 5°C) = 0.05 $/kWh = 400 $/kW-Yr

Refrigeration

(R-10, -10°C) = 0.1 $/kWh = 800 $/kW-Yr

(R-50, -50°C) = 0.3 $/kWh = 2400 $/kW-Yr

(R-100, -100°C) = 1 $/kWh = 8000 $/kW-Yr

(R-200, -200°C) = 5 $/kWh = 40000 $/kW-Yr

Homework Equations

The Attempt at a Solution


This is my converged aspen simulation of a process to produce acetone (AC) from isopropyl alcohol (IP). I used a heat exchanger simulation software and here are my streams
Table of process stream data with columns Stream, Name, TS and TT in °C, DH in kW, CP in kW/K, DT, HTC in kW/m^2K, and Cap Cost Class, listing hot and cold streams R1, C1, T2C, T2R, T3C, T3R, T4C, T4R, HP, MP, LP, CW, HO, and ChW.

Basically just inputting the heat duties of all the units. I just do a heat curve analysis and when the heat capacity changes I add a separate cooler, which is why there are blank rows in "Name", because the specific heat capacity is assumed to be constant only in one phase.

This is my composite curve
Composite curves plot of temperature in C against enthalpy in kW, with a red hot curve and a blue cold curve, axes running 0 to 350 C and 0 to 50000 kW.


This is my heat exchanger network
Process flow diagram of a heat exchanger network with horizontal stream lines labelled 1 to 14 and vertical red and blue lines joining numbered heat exchanger matches, with utility labels CW, HO and R on the margins.


I'm wondering what I could do to improve the efficiency of this process? Or at least places I could start looking and what I should look for.
 
Last edited:
Some new idea, I decided to add a heater before the reactor in order to preheat the IP, since I can use low pressure steam which is cheaper than hot oil for the reactor. However, for some reason the Aspen file will not converge when I add this heater. The amount of IP being fed is changed, even though I specify the amount. In the OP the mass flow rate is 19483 kg/hr, now its 21123 kg/hr. It is quite strange.

Aspen process flowsheet with stream mass flow rates in kg/hr and duty values (Q) for a reactor R1, heat exchangers and columns T1 and T2, with a legend for temperature, pressure, mass flow, duty, errors and warnings.

The Energy report from SPRINT
Text report of a pinch analysis: DTmin 10.0 C, minimum hot utility 22348.3 kW, minimum cold utility 12604.3 kW, process heat recovery 8729.70 kW, followed by a utility cost table.


This is my original energy report from the previous iteration
Text-based pinch analysis summary showing a minimum DTmin of 10 C, minimum hot utility 22219.9 kW, minimum cold utility 13554.9 kW and process heat recovery 7779.13 kW, with a table of hot and cold utility loads, energy and total cost in dollars per year.
 
Wouldn't trust any answer at all from a non-converged A+ simulation.
I have preciously little to go by, but my advice is to check the pressure drop in the feed preheater. T1 has trouble now (offgas flow disproportional) and then T2 crashes.
Good luck !
 
If you wish, I can send you the Aspen file, but in the mean time I will try your suggestion
 
Either that or increase the feed pressure.
Something else that came to mind: does the reactor do the same thing with a feed of 25 degrees as it does with a feed of 140 degrees ?
 
As far as I know there is no difference
 
If there are no temperature dependencies in the reactor, then it must be the pressure ...
 
Ah yes, I forgot that the pressure in the preheater should be 2 atm.
 
Okay, so I added a preheater before the reactor to IP, and a cooler for the AC.

Process flow diagram of a chemical plant with streams labelled with mass flow rates, temperatures and pressures, equipment including a reactor R1, columns T1, T2 and 3, and duties marked Q on heaters and coolers.


This is my grand composite curve, and I should somehow use it to improve my energy efficiency. I know the flat blue lines are the utilities that I need, but I am unsure how to translate this curve into my process so that I know where I should put heaters/coolers.

Balanced grand composite curves plot from Sprint pinch software, temperature in C against enthalpy in kW, with hot (red) and cold (blue) curves and DTmin of 20 C.


And here is my utilities cost (this one is higher because DTmin is 20 C), but I have improved it with what I have done so far

Text output from a pinch analysis tool showing DTmin 20.0 C, minimum hot utility 17449.7 kW, minimum cold utility 15624.0 kW, and process heat recovery 6677.03 kW, with a utilities cost table listing hot utility and cold utility energy costs and a total cost of 0.125116E+08 $/yr.
 
Last edited: