Eng-Tips is the largest engineering community on the Internet
Intelligent Work Forums for Engineering Professionals
How to Start a Chemical Process Design 4
- Thread starter louvreChE
- Start date
- Thread starter
- #4
I would like to thank you for your advice epoisses & IR stuff but unfortunately what I need is the deliverables that I would be needing to start the design. Like for example I need to produce an esterification plant. Should I start by doing the Chemical chain with all the oil structures on it to produce the finish product? Should I make the process description? How? please help.
-
3
- #5
bchoate
What you seem to be asking for is an answer(s) to a rather large question. There are systematic ways to go about process design but most really good process engineers have developed skills along the way that facilitate the process but perhaps are not easily described. I am going to sketch for you a 10,000 ft view of the process of developing a chemical process. You need to know that not am I a chemical process engineer with more than 20 years in the chemicals and fine chemicals industries, I am also a chemist with many years experience in new products development and process analytical work. My experiences include Quality Engineering for Energetics for defense industries and Six Sigma coordination for the same.
The first step of process design is to define what it is you wish to accomplish. You mentioned an esterification plant. What ester do you propose to produce. Research how that product can be produced and is produced if produced commercially. A process description is a good place to begin. You can leave some things undefined or posed as questions to answer. What is the chemistry to be carried out in this process? You may have more than one route to this product. Write down the chemical equations. Esters can be produced by direct esterification of an organic acid by an alcohol in the presence of a suitable catalyst or esters can be produced by transesterification of another ester or by a Tischenko condensation of two aldehydes, or by a Cannizarro reaction, or by a variety of other processes. Your process description might be the proposal to evaluate the production of the desired product by two different methods. Describe what feedstocks are needed (including catalysts), what assay is needed for each, and potential sources of the raw materials.
Once you have the what and the chemistry defined, work on defining what engineering operations might be required. An example might be receipt, storage, and feed equipment for raw materials, a reactor (CSTR, etc.) for the reaction, purification equipment for the product, equipment for any recyles and waste streams, and bulk storage and loading equipment for finished product. At this point you should be able to draw a block diagram of the process. The process description should contain a preliminary block flow diagram that will be built into a full process.
You can spend more time researching how some of the unit operations might be accomplished in your process. For a simple esterification plant producing a product such as butyl butyrate you might need bulk storage for n-butanol and n-butyric acid, a CSTR for a reactor, and distillation columns to purify the product. Distillation could include low boiler removal, high boiler removal, and product purification to a sales spec. There could be a recycle of n-butanol and disposal of waste products as well.
You need to spend more time studying what potential byproducts can be made in the process and in collecting physical properties of all the compounds in the process. Now you begin to develop a more detailed process flow diagram. Bulk storage tanks for raw materials and products need to include purge/vent systems for blanketing and emissions from the storage tanks. The reactor should include any operations to heat and/or cool the reactor or any of its streams and the pumps to feed, circulate, or remove product from the reactor. Distillation columns should include condensers, reboilers, and process tanks as well as pumps. Put together a more detailed process flow diagram.
Next you need to describe how each unit operation will function in the process. This can be data from literature sources, bench scale or larger work, and from process modeling of the process sections with Aspen or Chemcad or other package. This is where the property and VLE data you collected is put to use. You may begin to appreciate other operations that may be needed. In n-butyl butyrate it may be possible to decant some water from the reactor product and to make an ester/hiboiler split by taking low boilers and the water/ester azeotrope overhead in a distillation column. Low boilers might be separated from the ester in a second distillation column as their water azeotropes with an ester cut as a base vapor takeoff. The required separations, including all known byproducts, have to be defined from some source. At this point one can produce a fairly detailed process flow diagram with a wealth of information on it.
At some point one needs to do a theoretical balance around the process assuming 100% stochiometric yields and the process capacity. This could be done at the block diagram stage.
Once the unit operations have been defined, a detailed mass balance including recyles and waste streams is to be done to show reactor yields and losses across unit operations. This will establish mass flow rates through the process and raw material requirements. An energy balance has to be done next.
I will stop for now and complete this later if it appears to be what you are looking for.
What you seem to be asking for is an answer(s) to a rather large question. There are systematic ways to go about process design but most really good process engineers have developed skills along the way that facilitate the process but perhaps are not easily described. I am going to sketch for you a 10,000 ft view of the process of developing a chemical process. You need to know that not am I a chemical process engineer with more than 20 years in the chemicals and fine chemicals industries, I am also a chemist with many years experience in new products development and process analytical work. My experiences include Quality Engineering for Energetics for defense industries and Six Sigma coordination for the same.
The first step of process design is to define what it is you wish to accomplish. You mentioned an esterification plant. What ester do you propose to produce. Research how that product can be produced and is produced if produced commercially. A process description is a good place to begin. You can leave some things undefined or posed as questions to answer. What is the chemistry to be carried out in this process? You may have more than one route to this product. Write down the chemical equations. Esters can be produced by direct esterification of an organic acid by an alcohol in the presence of a suitable catalyst or esters can be produced by transesterification of another ester or by a Tischenko condensation of two aldehydes, or by a Cannizarro reaction, or by a variety of other processes. Your process description might be the proposal to evaluate the production of the desired product by two different methods. Describe what feedstocks are needed (including catalysts), what assay is needed for each, and potential sources of the raw materials.
Once you have the what and the chemistry defined, work on defining what engineering operations might be required. An example might be receipt, storage, and feed equipment for raw materials, a reactor (CSTR, etc.) for the reaction, purification equipment for the product, equipment for any recyles and waste streams, and bulk storage and loading equipment for finished product. At this point you should be able to draw a block diagram of the process. The process description should contain a preliminary block flow diagram that will be built into a full process.
You can spend more time researching how some of the unit operations might be accomplished in your process. For a simple esterification plant producing a product such as butyl butyrate you might need bulk storage for n-butanol and n-butyric acid, a CSTR for a reactor, and distillation columns to purify the product. Distillation could include low boiler removal, high boiler removal, and product purification to a sales spec. There could be a recycle of n-butanol and disposal of waste products as well.
You need to spend more time studying what potential byproducts can be made in the process and in collecting physical properties of all the compounds in the process. Now you begin to develop a more detailed process flow diagram. Bulk storage tanks for raw materials and products need to include purge/vent systems for blanketing and emissions from the storage tanks. The reactor should include any operations to heat and/or cool the reactor or any of its streams and the pumps to feed, circulate, or remove product from the reactor. Distillation columns should include condensers, reboilers, and process tanks as well as pumps. Put together a more detailed process flow diagram.
Next you need to describe how each unit operation will function in the process. This can be data from literature sources, bench scale or larger work, and from process modeling of the process sections with Aspen or Chemcad or other package. This is where the property and VLE data you collected is put to use. You may begin to appreciate other operations that may be needed. In n-butyl butyrate it may be possible to decant some water from the reactor product and to make an ester/hiboiler split by taking low boilers and the water/ester azeotrope overhead in a distillation column. Low boilers might be separated from the ester in a second distillation column as their water azeotropes with an ester cut as a base vapor takeoff. The required separations, including all known byproducts, have to be defined from some source. At this point one can produce a fairly detailed process flow diagram with a wealth of information on it.
At some point one needs to do a theoretical balance around the process assuming 100% stochiometric yields and the process capacity. This could be done at the block diagram stage.
Once the unit operations have been defined, a detailed mass balance including recyles and waste streams is to be done to show reactor yields and losses across unit operations. This will establish mass flow rates through the process and raw material requirements. An energy balance has to be done next.
I will stop for now and complete this later if it appears to be what you are looking for.
- Thread starter
- #6
Thanks for the advice bchoate. You are actually going through the process of answering my question clearly but where does the HAZOP study begins? Please feel free to explain further if you still have something to say since I'm learning from your point of views & experiences.
bchoate
You seem to have some parts of the design process in mind since you mention a hazop study. A hazop study is done when you have the process defined well enough for the study to be meaningful and to contribute to the design process.
Let's pause a moment and cover something that you know before starting this process; the site where such a plant could or will be built. Is this a grassroots effort? or an extension of an existing plant? or a conversion of an existing plant? The answer will influence several things in the process design.
Let's consider that you have a process description, property data for chemicals involved, process chemistry, block flow diagram, semi-detailed process flow diagram, theoretical mass balance, unit operations modeling, and a site. Now a detailed mass and energy balance based on the process flow diagram, the reaction yields, and the unit operations engineering yields has to be built. Process conditions (pressure, temperatures, flows, etc.) should be noted. The mass balance should be by component.
List what utilities will be required for the process. Electricity for pumps, instruments, etc. What will be used to heat or cool process streams. Closed loop cooling water system, river water, steam, hot oil, etc. Instrument air and compressed air. Nitrogen for blanketing. Where will these utilities come from? tie-in to existing facilities? new facilities?
The process diagram needs to be broken into sections for P&ID's. Moderately detailed P&ID's should be developed for each section. Items from the following considerations will be added to the P&ID's.
ENVIRONMENTAL: Will this plant require a permit? Will it be a new permit or a modification of an existing permit? What will be the disposition of vapors from process vents? process wastes?
SAFETY: Where will relief valves or rupture disks be required? What will be the bases for sizing the devices? Platforms, ladders, and process flooring must be built to standards. Are there other safety issues?
MATERIALS OF CONSTRUCTION: what materials are required for equipment, piping, utilities, etc. You will need to determine where corrosion can occur and what materials are appropriate for each service. If organic acids are to be used in an esterification plant, one should know that dilute amounts are very corrosive on 316 SS. Typically the service may require something like Sandvik 2205.
INSULATION: Where will insulation be used? what type. One needs to be concerned about under insulation corrosion also particularly with mineral wool or Calsil.
INSTRUMENTS: What type of control system will be used with this plant? DCS? Single loop controllers? PLC? What instruments will be required with each unit operation for level, flow, temperature, pressure, etc.
Pumps: what types and sizes of pumps will be needed for each application. Consider mechanical reliability here in the sealing and flush applications.
Equipment Sizing: You should begin to estimate the size of reactors, distillation columns, heat exchangers and other equipment that will be needed in the process.
Information from these items should be included on the P&ID's that you are compiling.
When the P&ID's are as well defined as one can make them, a hazop study can be done. The question to be asked is, "what are the effects of deviations from design intent" If a heat exchanger loses coolant flow, what is the effect and are there any process alarms or actions needed for this possibility.
You are also ready to begin the process of estimating the cost of the process. One can use the module/factor approach and arrive at a budget estimate. It has been my experience that cost estimates, once given, are never forgotten. It has been my practice to try to produce a 10% estimate if possible. You have a wealth of detail at hand by now. Equipment estimates can be obtained from vendors. Knowing the site and the equipment layout, you can begin to estimate lengths of pipe, conduit, etc. The Richardson system is a method that allows one to make a semi-detailed capital estimate of the components, labor, and indirect costs associated with a project.
I know there are many other details that I have not spoken of. You will need to incorporate them as you come across them. Engineering standards and approved vendor lists for your employer have to be followed. Try to bring some improvement to as many of the unit operations as you can.
Bill Choate
You seem to have some parts of the design process in mind since you mention a hazop study. A hazop study is done when you have the process defined well enough for the study to be meaningful and to contribute to the design process.
Let's pause a moment and cover something that you know before starting this process; the site where such a plant could or will be built. Is this a grassroots effort? or an extension of an existing plant? or a conversion of an existing plant? The answer will influence several things in the process design.
Let's consider that you have a process description, property data for chemicals involved, process chemistry, block flow diagram, semi-detailed process flow diagram, theoretical mass balance, unit operations modeling, and a site. Now a detailed mass and energy balance based on the process flow diagram, the reaction yields, and the unit operations engineering yields has to be built. Process conditions (pressure, temperatures, flows, etc.) should be noted. The mass balance should be by component.
List what utilities will be required for the process. Electricity for pumps, instruments, etc. What will be used to heat or cool process streams. Closed loop cooling water system, river water, steam, hot oil, etc. Instrument air and compressed air. Nitrogen for blanketing. Where will these utilities come from? tie-in to existing facilities? new facilities?
The process diagram needs to be broken into sections for P&ID's. Moderately detailed P&ID's should be developed for each section. Items from the following considerations will be added to the P&ID's.
ENVIRONMENTAL: Will this plant require a permit? Will it be a new permit or a modification of an existing permit? What will be the disposition of vapors from process vents? process wastes?
SAFETY: Where will relief valves or rupture disks be required? What will be the bases for sizing the devices? Platforms, ladders, and process flooring must be built to standards. Are there other safety issues?
MATERIALS OF CONSTRUCTION: what materials are required for equipment, piping, utilities, etc. You will need to determine where corrosion can occur and what materials are appropriate for each service. If organic acids are to be used in an esterification plant, one should know that dilute amounts are very corrosive on 316 SS. Typically the service may require something like Sandvik 2205.
INSULATION: Where will insulation be used? what type. One needs to be concerned about under insulation corrosion also particularly with mineral wool or Calsil.
INSTRUMENTS: What type of control system will be used with this plant? DCS? Single loop controllers? PLC? What instruments will be required with each unit operation for level, flow, temperature, pressure, etc.
Pumps: what types and sizes of pumps will be needed for each application. Consider mechanical reliability here in the sealing and flush applications.
Equipment Sizing: You should begin to estimate the size of reactors, distillation columns, heat exchangers and other equipment that will be needed in the process.
Information from these items should be included on the P&ID's that you are compiling.
When the P&ID's are as well defined as one can make them, a hazop study can be done. The question to be asked is, "what are the effects of deviations from design intent" If a heat exchanger loses coolant flow, what is the effect and are there any process alarms or actions needed for this possibility.
You are also ready to begin the process of estimating the cost of the process. One can use the module/factor approach and arrive at a budget estimate. It has been my experience that cost estimates, once given, are never forgotten. It has been my practice to try to produce a 10% estimate if possible. You have a wealth of detail at hand by now. Equipment estimates can be obtained from vendors. Knowing the site and the equipment layout, you can begin to estimate lengths of pipe, conduit, etc. The Richardson system is a method that allows one to make a semi-detailed capital estimate of the components, labor, and indirect costs associated with a project.
I know there are many other details that I have not spoken of. You will need to incorporate them as you come across them. Engineering standards and approved vendor lists for your employer have to be followed. Try to bring some improvement to as many of the unit operations as you can.
Bill Choate
-
1
- #8
LouvreChE, I'm surprised to hear you say that you always have a PFD that was complete when you started. In my experience a PFD is rarely, if ever, complete. It evolves, as more information becomes available - either before or after commissioning and operation. I would be surprised if there were no changes on any of the previous projects you were working on. So most likely, you already have some experience of this.
Your best option is to look to see if there are any similar plants operating or constructed by your company - or if you are working for a contractor, by the client company. Try to understand each piece of kit to determine what it is doing. Then decide if it's needed for your process.
If you are starting from scratch, and none of this information is available, then you need to sit down with the development chemist(s) and determine exactly what he/she does in the lab and why. It's important to understand what he/she finds easy to do and what is difficult. This isn't necessarily the same on the plant. Some things are easy in the lab, but very difficult on the plant (i.e. solids handling), whereas others are easier on plant than in the lab.
Then define a very basic flowscheme. What you are trying to do is copy what is happening in the lab on a larger scale. Don't try to size individual items of equipment, yet, and don't try to add instrumentation - unless it will help you understand how it will work.
Once you have a basic outline you can start adding details.
As epoisses says, the most important thing is to keep discussing it. Show your outline to the chemist(s) and get feedback on it. They (or you) may have forgotten something. It's very important to work as a team. They will most likely be proud of what they have done, and seeing what looks like a plant based on what they have done will give them a buzz. Take advantage of their enthusiasm!
Talk to others as well, people with experience of the chemicals or of the process (of developing a chemical process from scratch). With their help, you can start adding equipment sizes and the detail you need as described by bchoate above.
Don't forget, that most companies have several decision gates in any project. Don't get into too much detail before you are certain that the project is going ahead to the next stage. Doing a basic outline first will enable you (or someone else in your company) to come up with a rough cost, and, equally important, how long you think it will take to build. Feeding this back to the decision makers will give you an early indication as to whether this project is really going to happen or not. This is very important - it can save you a lot of time. Keep the dialogue going. Things change in the commercial world very, very quickly. It's no good designing the perfect plant if it's decided that the plant is no longer needed.
Doing a HAZOP is almost the last thing you do in the design, and is often only done once the money for the project has been approved. Don't worry about it until the very latter stages of the project. However, safety should be paramount in your design. In a perfect design, the HAZOP would be a rubber stamping exercise - your design should aim to make it so. Thinking about it, and designing around it at this very early stage makes it very much easier (and safer!) than trying to shoehorn changes in later on.
It may seem daunting at the minute, but personally, I find this the most satisfying and rewarding work I have done. Being given this project is a vote of confidence in your abilities by your management. Good Luck!
Your best option is to look to see if there are any similar plants operating or constructed by your company - or if you are working for a contractor, by the client company. Try to understand each piece of kit to determine what it is doing. Then decide if it's needed for your process.
If you are starting from scratch, and none of this information is available, then you need to sit down with the development chemist(s) and determine exactly what he/she does in the lab and why. It's important to understand what he/she finds easy to do and what is difficult. This isn't necessarily the same on the plant. Some things are easy in the lab, but very difficult on the plant (i.e. solids handling), whereas others are easier on plant than in the lab.
Then define a very basic flowscheme. What you are trying to do is copy what is happening in the lab on a larger scale. Don't try to size individual items of equipment, yet, and don't try to add instrumentation - unless it will help you understand how it will work.
Once you have a basic outline you can start adding details.
As epoisses says, the most important thing is to keep discussing it. Show your outline to the chemist(s) and get feedback on it. They (or you) may have forgotten something. It's very important to work as a team. They will most likely be proud of what they have done, and seeing what looks like a plant based on what they have done will give them a buzz. Take advantage of their enthusiasm!
Talk to others as well, people with experience of the chemicals or of the process (of developing a chemical process from scratch). With their help, you can start adding equipment sizes and the detail you need as described by bchoate above.
Don't forget, that most companies have several decision gates in any project. Don't get into too much detail before you are certain that the project is going ahead to the next stage. Doing a basic outline first will enable you (or someone else in your company) to come up with a rough cost, and, equally important, how long you think it will take to build. Feeding this back to the decision makers will give you an early indication as to whether this project is really going to happen or not. This is very important - it can save you a lot of time. Keep the dialogue going. Things change in the commercial world very, very quickly. It's no good designing the perfect plant if it's decided that the plant is no longer needed.
Doing a HAZOP is almost the last thing you do in the design, and is often only done once the money for the project has been approved. Don't worry about it until the very latter stages of the project. However, safety should be paramount in your design. In a perfect design, the HAZOP would be a rubber stamping exercise - your design should aim to make it so. Thinking about it, and designing around it at this very early stage makes it very much easier (and safer!) than trying to shoehorn changes in later on.
It may seem daunting at the minute, but personally, I find this the most satisfying and rewarding work I have done. Being given this project is a vote of confidence in your abilities by your management. Good Luck!
- Status
Similar threads
- Locked
- Question