10 replies to this topic

[Kakashi-01]

(7a and 7b) and (12a and 12b) are streams that undergo a phase change. Stream 7 is broken up into two dummy streams, one representing the stream as it is cooled to its dew point and the other representing the stream as it is further cooled to its desired temperature. Same goes to stream 12. Streams 17 and 18 represent the streams that pass through the column reboiler and condenser. I didnt break down those streams because their temperature change while they phase change is minimal

 

I adjusted the temperatures of the hot and cold streams, rank each temperature in descending order and calculated the enthalpy of each temperature interval. An enthalpy cascade was designed. In the initial pass, no heat was added from outside. The residual enthalpy was always positive indicating no process pinch. However, looking at the composite curve there is a psuedo-pinch temperature at 142.4996427 C.  The design for the heat exchanger network should start from the psuedo-pinch and move away from it.  

 

Above the psuedo-pinch, it must follow that the number of hot streams be less than the number of cold streams and CPH<=CPC to pair hot and cold streams. But, why do I have to split the streams? Looking at the data all the hot streams can satisfy the cold streams above the pseudo-pinch to the no utility end. Namely, stream 21 can be matched with cold streams 5,10,and 19 and stream 7a can be matched with stream 17 in the reboiler satisfying their duties and not violate the minimum temperature difference.

Attached Files

•  StreamData.JPG   95.48KB   0 downloads
•  TCC.JPG   53.42KB   0 downloads

Edited by Dumpmeadrenaline, 21 February 2025 - 10:36 PM.

[Pilesar]

With pinch analysis, the flow rate, location, and phases of the streams matter when judging whether a match is practical. For example, you propose using a 870C stream to reboil a 146C stream. I highly doubt this would be done. I don't know your process, but here are a few concerns I would investigate:

1) the hot stream is far away from the reboiler.

2) the materials of construction and exchanger design would be super expensive.

3) the reboiler would be in the undesirable film boiling regime.

4) the high surface temperature in the reboiler would cause degradation of the process stream, forming unwanted byproducts.

Energy savings is not the only consideration... capital cost and constructability should also be considered.

You are the one doing network analysis - I don't understand what you mean by 'why do I have to...'

[Kakashi-01]

With pinch analysis, the flow rate, location, and phases of the streams matter when judging whether a match is practical. For example, you propose using a 870C stream to reboil a 146C stream. I highly doubt this would be done. I don't know your process, but here are a few concerns I would investigate:

1) the hot stream is far away from the reboiler.

2) the materials of construction and exchanger design would be super expensive.

3) the reboiler would be in the undesirable film boiling regime.

4) the high surface temperature in the reboiler would cause degradation of the process stream, forming unwanted byproducts.

Energy savings is not the only consideration... capital cost and constructability should also be considered.

You are the one doing network analysis - I don't understand what you mean by 'why do I have to...'

Thank you, I appreciate your insights and help! The primary objective of my study is to reduce operating costs, and capital costs are not considered.

I have attached the process flow diagram for methanol synthesis by steam reforming of natural gas. In this process, low-pressure steam is used in the reboiler, while high-pressure steam is used to preheat the syngas feed to the methanol reactor (130–250°C).

I have already optimized the process by adjusting operating variables such as the steam-to-methane ratio, system pressure, and distillation column pressure while considering constraints like reactor volume and temperature. Now, I am conducting a pinch analysis to further reduce utility usage and operating costs.

I understand that after identifying the pinch or pseudo-pinch, stream matching must follow two key criteria:

1. The number of hot streams must be less than or equal to the number of cold streams.
2. The heat capacity flow rate of the hot stream (CPH) should be less than or equal to that of the cold stream (CPC) to ensure proper pairing.

However, in my case, the number of hot streams exceeds the number of cold streams. I would need to split the cold streams to balance the numbers.

 

Why is splitting necessary when the available heat in the hot streams is sufficient to satisfy the demand of the cold streams while still meeting the minimum temperature difference requirement? 

Attached Files

•  1740200838742.png   149.4KB   0 downloads
•  1740200864005.png   107.27KB   0 downloads
•  1740200891986.png   36.46KB   0 downloads

[Pilesar]

This looks like a useful exercise for your learning, unlike some other academic problems that have been reviewed in this forum. Pinch analysis is something you almost have to immerse yourself in... it is not a casual 5-minute solution. Your composite curve looks to me like it is incomplete. There is a significant hot stream at the unnumbered distillation column overhead into the condenser that I do not see listed. I think the reboiler feed stream is the unmarked '17' stream. You may be better served by putting all your streams in the list at the beginning. Your initial post suggested that you did not include all streams because the temperature change was minimal during phase change. Those streams are important! The grand composite will show the duty required versus temperature. The duty value is where the energy savings matters. The composite curve and the pinch can look much different depending on the streams included. As to splitting streams, I will suggest that large temperature differences between hot-cold streams are more valuable and you don't want to waste temperature driving force unnecessarily. Your observation that you can get a good solution more easily without going through all the analysis steps may be correct, but I think the goal is to get the theoretical optimum solution. You provided a lot of information, but there is not a lot I can do further to help. Your questions reflect a good engineering mindset. I wish you success.

[breizh]

Hi,

Consider this resource to support your work.

Good luck,

Breizh

Attached Files

•  Pinch Technology - Basics for the Beginners.pdf   1.52MB   15 downloads

[Kakashi-01]

Thank you @pilesar and @breizh for your feedback and link. I wanted to clarify a couple of points:

1. The hot stream at the unnumbered distillation column overhead is actually included in the stream data as stream 18.

2. When I mentioned that I did not break down all streams due to minimal temperature change during phase change, I was referring to the streams that pass through the column reboiler and condenser. For those, I assumed a single heat capacity since their temperature change during phase change is minimal. However, for streams 7 and 12, I created dummy streams to account for the phase change more accurately.

I also wanted to ask for your thoughts on an alternative approach. Given the practical limitations you mentioned regarding using a high-temperature stream like 7a to reboil a much lower-temperature stream, could we instead use the cooling water proposed in the process flow diagram to bring the temperature down from 850°C to 130°C to make the required LP steam on-site, rather than use purchase LP steam?

[Pilesar]

 could we instead use the cooling water proposed in the process flow diagram to bring the temperature down from 850°C to 130°C to make the required LP steam on-site

Cooling water should not be used at this high temperature. If water is used for cooling here it must be very pure. As water temperature increases, any minerals in the water become less soluble. Excessive scale tends to form on heat transfer surfaces above 60 C. As your goal is energy savings, you would use this hot stream to generate steam or increase the temperature of process streams. Is the 850 C stream a fired heater flue gas? Typically, there are coils in the convection section of fired heaters to recover heat. Often there are coils for several purposes including high pressure steam superheat, boiler feedwater heating, furnace process feed heating, furnace fuel heating, and possibly combustion air heating. Produced steam can be used to drive rotating equipment, generate electricity, and transfer heat. If you generate steam, you will need steam drum, water purification equipment, steam control system, auxiliary boiler and more. I expect methanol synthesis process flow diagrams and material balances should be findable if you research them.

[Kakashi-01]

the 850°C stream is not fired heater flue gas but rather syngas at the outlet of the steam reformer (R-401). All combustible species in Stream 15 are completely oxidized in the burner, and the heat (Qin) needed to drive the steam reforming reactions is provided by this combustion process. The high-temperature flue gas leaving B-401 (Stream 21) is then used to preheat the reactor feed (Stream 5) to 850°C in E-401.

The available utilities in the problem include boiler feedwater (BFW), low-pressure (LP) steam, medium-pressure (MP) steam, high-pressure (HP) steam, and cooling water. From what I’ve read, the convection section of the reformer includes three coils, boiler feed water preheat, steam  generation and steam superheat convection coils.

 

The reaction products pass through the tubes, exchanging heat with boiler water which naturally circulates through the shell from the High Pressure Steam Drum above to produce a saturated liquid/vapor mixture. The steam-water mixture rises back to the HP Steam Drum. Saturated steam exits the top of the HP Steam Drum and passes through the superheat coil in the hottest zone of the convection section. The superheated steam is split: Reactant Steam: A portion is mixed with natural gas feed for the reforming process and the remaining is exported.

 

To improve energy efficiency, I was thinking to implement an alternative configuration in my simulation based on the above where BFW is used instead of cooling water. A pump would pressurize the BFW to the HP steam pressure (41 bar). The 850 C stream could then generate saturated steam at 41 bar. If the simulation results show excess steam production beyond process requirements, the surplus could either be exported or expanded through a turbine to a lower pressure to meet LP steam demands in the reboiler of the column.

Edited by Kakashi-01, 25 February 2025 - 03:42 PM.

[Pilesar]

It is common to have the first quench after the reactor generate steam. Usually, this is fed by high pressure water from the steam drum. This heat exchanger is tricky to design. Specialized exchangers are used but still leaks are not uncommon, resulting in emergency plant shutdown. Because of the elevation of the steam drum, steam vapor is suppressed in the exchanger. When the heated water returns to the drum, a portion of the hot water flashes due to the lower pressure in the drum. Steam drums must be at even higher elevation in high pressure systems because steam at high pressure is closer in density to the liquid water. To reduce energy, you will generate steam to reduce the amount of steam utility imported and even possibly become a net exporter of steam. It may even make since to combust more fuel to improve the steam generation. Typically you want to generate steam at high pressure and high superheat for use in turbines. The turbine can be designed for steam effluent at lower pressure levels as needed in the plant steam balance. Every parameter you change affects so many others! Closing the heat and energy balance is difficult even for experienced plant designers.

[Kakashi-01]

Thank you so much. I’m finding the complexity of these systems quite overwhelming, but I’m trying to ensure my simulation is as realistic as possible.

 

I pressurized the boiler feed water (BFW) at 120°C and 21 bar (as given in the problem statement) to 40.35 bar, matching the pressure of the  saturated steam (40 bar) mixed with natural gas as feed to the steam reformer. Instead of using cooling water, I used BFW to recover heat from the steam reformer product gases (syngas). For heat exchangers, I set pressure drops as follows (also as given in the problem statement):

Shell side: 0.1 bar

Tube side: 0.35 bar 
 

Since the heat duty is fixed to cool the steam reformer outlet from 850°C to 130°C, I needed to specify either the outlet temperature or vapor fraction of the pressurized BFW. I chose to set the vapor fraction to 1, as the steam mixed with natural gas is saturated. This is why I didnt include a steam drum in my simulation though I understand that it would be required in reality.

 

The simulation yielded 2859.5 kmol of saturated steam at 40 bar, while the HP steam required for the reforming reactions is 1774.7 kmol/h, leaving a surplus of 1084.8 kmol/h.

 

To utilize this excess steam, I am considering two options:

 

Exporting it for a credit of $0.006/kg (Any steam produced in the process that is not consumed within the process may be exported for a credit of $0.006/kg).
 

Take advantage of its lower specific heat capacity and superheat it using flue gas for use in turbines, improving overall energy efficiency

The flue gas leaving B-401 (Stream 21) is already used for:

 

E-401: Preheating the reformer feed to 850°C

E-407: Heating combustion air

E-408: Generating MP steam
 

I tried matching the excess saturated steam to a superheater but was unsure what outlet temperature to set for the superheated steam. For maximum heat recovery from the flue gas, if a 10°C minimum temperature approach is used, the steam outlet temperature would be set to 645°C, resulting in an estimated flue gas outlet temperature of ~523°C.
Does the superheat level (645°C) seem reasonable for typical turbine applications?

Attached Files

•  Superheated steam.png   364.95KB   0 downloads
•  Canvas.png   528.7KB   0 downloads

[Pilesar]

A few comments... The water in the quench exchanger will be at a higher pressure than the steam pressure due to the elevation of the drum. Assuming complete boiling in the quench exchanger may be okay for energy balance, but it is wrong and will cause many problems if you have to size an exchanger. These exchangers are designed so there is no vaporization as much as possible. Bubbles on the heat transfer surface have much less ability to remove heat from the metal and so the metal gets too hot in a local region causing failure. The superheat seems too high. If you sell the steam, you can use the available heat to make more steam instead of superheating it so much. If I were doing your project I think I would use the artificial intelligence tools now available on the internet. I asked one free Ai engine 'what is typical superheat for 40 bar steam' and got the answer below. Good enough? Accurate? That is for you to investigate!

 The typical superheat for steam at 40 bar (4,000 kPa) can vary depending on the specific application and system design. However, for industrial steam systems, the typical superheat is usually in the range of 10°C to 50°C above the saturation temperature at that pressure.

At 40 bar, the saturation temperature of steam (the temperature at which water boils at 40 bar) is around 250°C. So, the superheated steam temperature could range from 260°C to 300°C in typical industrial applications.