CHEN20012 Fundamentals of Chemical Engineering Major Assignment Questions| UoM

CHEN20012 Assignment Questions

HYSYS is a widely used application for simulating sequences of Chemical Engineering unit operations to represent an industrial plant. But it is only one of about a dozen packages used in industrial applications (it’s the most common of all packages, but might only have 1/3rd of the total market share); it’s the most general and will be helpful in a range of future subjects (Digitalization in the Process Industries, Heat and Momentum Transfer, Safety and Sustainability Case Studies, Reactor Engineering, Process Engineering and Design Project).  

But as HYSYS is only one of many different packages, rather than trying to develop students into HYSYS experts, we want to focus more on an introduction to simulating chemical processes, with HYSYS as the worked example. This shapes the intended learning outcomes for the HYSYS modules and this assignment:  

• Develop an understanding of how simulation packages develop computational solutions  
• Understand that simulation packages function as advanced calculators and are only as good as the input data and people who operate them (fluid packages and user inputs)
• Be able to complete basic simulations in HYSYS as a case study.   

Summary  

Ethylene oxide is a widely used chemical: it’s an intermediate in the production of ethylene glycol (used in synthetic fibres or as antifreeze), it can be used as an industrial disinfectant, and its high flammability range means it’s also used in explosives. In this assignment, we will look at the design of a reactor oxidising ethylene to ethylene oxide in the presence of oxygen (a highly exothermic reaction).   

PART A: Analytic vs computational solutions  

In Part A, we’d like to understand how HYSYS solves reaction kinetics and how this differs from the analytic solutions we do in class. So we will take an over-simplified reaction set (that can be solved both analytically and numerically) and compare the 2 results to understand their differences. In the interest of full transparency, I just need to point out here that this isn’t a ‘real’ reaction set: in Part B we will use a real reaction set (that actually models the competing reactions in the epoxidation reaction and its real temperature dependence). The reaction kinetics used in Part A are a made-up simplification to help us understand HYSYS.

Initially, we will assume only one reaction takes place (an epoxidation reaction) with a stoichiometry of:  

C2H4 + ½ O2 → C2H4O    

We will assume the reaction kinetics can be given as a reversible reaction that is first order in the forward direction, concerning ethylene, and first order in the reverse direction, concerning ethylene oxide:     

We will assume our reaction takes place isobarically and isothermally at 1 atm and 30oC respectively, with an equimolar ratio of ethylene and oxygen and no ethylene oxide present in the reactor feed or nitrogen. This is a single-pass reactor (no recycle loop or purification of the product).  

Under these conditions, the rate constants are: 

kF = 0.0101 sec-1 
kR = 0.004 sec-1 

(Just note these reaction conditions are low, and the rate constants are made up here.)  

The reaction takes place in a PFR with a volume of 50 m3 and the inlet gas flow rate of 1.1 m3/sec. We will determine the concentration profile of ethylene in the PFR (concentration vs time).  

• Use the reaction stoichiometry and rate equation above to derive an analytic solution to the rate equation (ie using calculus and integrating the rate equation). Once you have solved the rate equation, plot your solution as concentration of ethylene (y-axis) against time (x-axis). You can put this in as a formula in Excel and create data points (helpful for part 2). You do not need to type out your calculations formally; you can include a copy of handwritten calculations in an appendix. Include an explanation or calculation for how you determined the initial concentration of ethylene.  
• Use HYSYS to simulate the ethylene concentration profile (ethylene concentration vs time); HYSYS will not do this by default (it can output molar flow rate or mole fraction), so you may need to export the data to Excel, modify and replot it. HYSYS will also not plot time by default (it will plot length position from the iterative calculation). But you will be able to convert this to time using the volumetric flow rate (think about this carefully if you are allowing for changes in volumetric flow rate through the system).  Clearly state any relevant assumptions invoked in replotting this data and in your HYSYS simulation (there are many different ways to calculate this, so make sure you say it explicitly). You may or may not need to assume the volumetric flow rate is constant (you should state this clearly). Assuming a constant volumetric flow rate will make your solution considerably simpler, and many people may select to do this.  
• You should include one plot of the concentration profile, but with both solutions on the same plot. The 2 results will be similar but will not match exactly, because they have different in-built assumptions. So you should look for deviations between the 2 plots and discuss why you think these have arisen. Neither result uses experimental data: they use the same starting reaction kinetics, but invoke different solution methodologies. So you should comment on the reason for any discrepancy between the 2 results and state which you think is likely to be closer to reality (and most importantly, why). You won’t be able to invoke all the same assumptions in the 2 solutions, but you will be able to think through and discuss how the assumptions will change the result (for example, will changing an assumption push the 2 solutions closer together or further apart- you can 

ΔCE segment = rE, segment x Δtsegment  

Where: 

Δtsegment:   Residence time per segment 
segment:   Volumetric flow rate of gas into the segment 
VReactor:   Total volume of the reactor 
Nsegments:   Number of segments the reactor is divided into (for the step-wise solution) 
CE segment:   Concentration of ethylene in the segment 

rE segment:   Rate of change of ethylene (wrt to time) in the segment  

So, for example, the concentration of ethylene leaving segment 1 and entering segment 2 becomes: CEthylene segment 2 = Cethylene segment 1 – rE Δtsegment 

The concentration of other gas species can be calculated in the same way (based on the reaction stoichiometry). 

The concentration of ethylene (and other relevant gas species) in segment 2 can again be substituted into the rate equation. A new volumetric flow rate of the gas can be calculated (based on the composition and inlet conditions) to determine a new residence time in segment 2. And so the change in concentration across segment 2 can be calculated, and so on for all 20 segments.  

This process is equivalent to the discrete form of integration (ie summation) used by computers to do a computational integral. As the number of segments approaches infinity, the solution approaches a continuous integral.

Part B: Reactor System Simulation

We now move on to a comprehensive solution of the real reaction kinetics under the conditions of the ethylene oxide reactor (for normal operations). The intention here is to study the reaction kinetics in detail to optimise the reactor performance for the next stage, which would be a mechanical design of the unit for a real plant.  

How the reactor performance is optimised (ie what metrics you choose to use) will be up to you. But you should not include any detailed costing information. Instead, you will be looking at operational considerations: conversion, yield, selectivity or overall energy demand or very general cost issues like equipment number or size of units (which roughly scales to cost). You will be trying to design your reactor (operating temperature, pressure etc) for optimal performance, and try to understand the competing considerations that affect this selection.   

Basic Process Information:

• Pure ethylene is supplied at 20 kmol/hr at 20 oC and atmospheric pressure.  
• Oxygen is purified from air in a cryogenic distillation column in an upstream process1 and is available at a purity of 95 mol% oxygen (balance nitrogen). The oxygen stream (oxygen and nitrogen) is supplied at a rate of 169 kmol/hr and 25oC.
• The gas mixture is sent to a reactor packed with silver catalyst in a long tube. So the reactor may be modelled as a plug flow reactor with a single tube. The reactor has a void fraction of 0.5 and a total volume of 0.025 m3 (VR). Select a reasonable reactor length, aspect ratio. A pressure drop of 10 kPa across the reactor will be an appropriate initial guess, but you can also estimate this in HYSYS. Further reaction kinetic details are provided below.
• The reactor operates isothermally, typically at temperatures between 170oC – 430oC and pressures between 3 MPa and 6 MPa.
• After the reactor, it is cooled to 25 °C and sent to a water stripper. In the stripper, the reactor outlet gases are contacted with water. The high pressure and low temperature favour mass transfer of ethylene oxide into the water; it is highly soluble in the water, but other gases are not. 250 kmol/hr of water is fed into the stripper, available at 25oC. Further details about the stripper splits are provided below.
• The water stream leaves the bottom of the stripper, rich with ethylene oxide and continues for further processing (which you do not need to simulate) 
  Unabsorbed reactant and product gases from the top of the stripper are recycled to the reactor inlet. To prevent the build-up of impurities (such as nitrogen and product gases), 5% of the gases recycled from the stripper are purged from the system. (Note: your simulation will include a recycle loop solver.)
• You may assume the pressure drop across any heat exchangers (or heaters, if simulating half a HEX) is 5 kPa.   

Reactor: 

You may model your reactor as a Plug Flow Reactor (PFR). Two competing reactions take place in the reactor: ethylene epoxidation (to produce ethylene oxide) and ethylene combustion (to produce carbon dioxide and water). The stoichiometry of the 2 reactions is:  

C2H4 + ½ O2 → C2H4O      (1) Ethylene Epoxidation

• Determine a metric to quantify the reactor performance, then use this parameter to optimise the reactor temperature and pressure, as well as the stripper pressure, within the range provided. Part of the exercise is for you to exercise independent critical thinking to determine what parameter(s) should be used to quantify reactor performance (and then optimised).  

Hint: You’ll need to think about this. You could think about things like conversion, yield or selectivity, and then also think about per-pass or overall. When you decide on a performance parameter, you’ll also need to think about how to define it (eg using a spreadsheet in HYSYS or just calculating it manually).

• If you choose to use case studies to optimise the process, make sure you include these plots in your report. Screenshots of case studies in HYSYS are usually unreadable- it is preferable to export the case study data and replot it in Excel. You can just cut and paste it from the HYSYS case study table. 
•  If you are using MyUniApps, note that you cannot save files to the desktop (it will be removed when you log out). You must save files within your student folder, or just email them to yourself for safety. 

Report

Please provide a short technical report to justify your design decisions in the HYSYS simulation. There is no specific word limit, but it should be about 8-12 pages as a maximum (if you are significantly exceeding this range, it implies you have a problem with your report layout). This limit excludes appendices, including things like scanned hand calculations.  

Your report should include the following key points:  

A title page, an executive summary (3-4 sentences) explaining your aims (what you are trying to do) and your key findings and then a contents page. The executive summary and contents should cover parts A and B of the report. (2 marks)   

Part A  

Your response to the 3 prompts in Part A and your discussion of why the 2 solution methodologies may differ. Don’t just list reasons for the discrepancy, or state differences within the assumptions of the 2 models (this can be pulled easily from ChatGPT). Instead, you should try to demonstrate your thesis with simple modifications to the 2 simulations: can you make small changes to the assumptions in either of them, that may help prove your thesis (and will help demonstrate your understanding of the discrepancy)? (9 marks)  

Part B

 Justification of fluid package selection. The fluid package selection must be correct, but more importantly, in grading, you will be marked heavily on the logic used to justify your fluid package selection. So you should think about ways you can check or validate this within HYSYS. (1 mark)  

The system Process Flow Diagram (PFD, screenshot from HYSYS) and justifications for the layout selected. There are multiple ways to organise your recycling loop, with ancillary heat exchanges and compressors. But not all these arrangements are equivalent; some configurations have significantly larger capital costs (equipment size) and operating costs (equipment duty). So you should think carefully about equipment ordering and explain the logical advantages of your layout. You do not need to do the costing of individual equipment items (this is beyond scope), but you can make common-sense statements (for example, a larger heat exchanger will cost more for the same duty and pressure rating). You may choose to compare different PFD layouts to illustrate the advantages of your layout. (9 marks)  

 Your selected operating pressure and temperature of the reactor. As with point 5, the marking rubric is heavily weighted towards critical thinking. For example, you may want to support your argument with case studies from HYSYS; poor reports will often just insert case studies with no explanation, but a good report will tend to use the case study results as evidence to support a recommendation. You will need to think about what parameter(s) can be used to help select optimal reactor conditions and how this can be presented clearly and concisely to support your case in the report. For example, for ease of control, reactors often don’t operate at a maximum.