3 replies to this topic

[A_D_M_MII]

Hi everyone,

 

I am doing a simulation of a real steam reforming process, i have some data of sensors and indicators and a problem has just shown up. (It is a academic project)

 

• I have simulated the combination of furnance and the steam reforming reactor by this way:
• A combustion reactor, which i know the inlets (T, P and composition), these inlets are: Natural gas, pre-heated air from the outside of the plant and PSA purge.
• A reformer reactor (FPR reactor with the kinetics equations os methane steam reforming).
• A cooler that transfers the heat flow to the FPR from the product exhausted gas of combustion.

+ i have supposed about 3,5% heat losses in combustiion reactor from the total heat transfer to reformer.

 

So, i know the temperature of the exhausted gas product of the combustion at the end of the radiant zone, i.e. the temperature after it has transfered the neccesary heat flow to the reformer. I also know that this current has about 3% v/v in oxigen, so i have thought that there is a incomplete combustion.

 

The problem is that the final temperature in my simulation is 1200ºC, but, i have to say that i have simulated it like a 100% conversion reactor (combustion). The temperature gauge indicates that this current may be 1020ºC.

 

How could i modelate it (Combustion)? I dont know what could be happening inside the furnace.

should i add the incomplete combustion reaction? or maybe i should put less than a 100% conversion in the combustion.

 

I don't have idea what could fit better whit the real process.

 

Thank you so much!

 

ADM 

 

 

[Pilesar]

The oxygen in the flue gas is not incomplete combustion. Excess air is used at the burners to guarantee complete combustion of the fuel. Adding in enough air so that you have the right flue gas composition will add a lot of volume to your flue gas. The 3% oxygen by volume may be misleading. Try reducing the excess air until you match the temperature out of the radiant section to see its effect in the simulation. If the oxygen content of your flue gas is according to the measurement you supplied, consider whether it may be caused by air leakage into the flue gas downstream of the combustion zone. You are also needing to balance the reformer reactor duty with the combustion duty so check the reformer duty. If your simulation is correctly working and it were a real plant, the operators would reduce the firing rate of the burners to reduce the flue gas temperature.

Edited by Pilesar, 29 June 2021 - 06:44 AM.

[A_D_M_MII]

Hi thank you Pilesar!!

 

I know what you mean, but, i also have the mass flow of the inlet air (so i can't move it much), and, the initial and final temperature of the reformer inlet and outlet, so the heat flow which is necessary to complete the reaction is " forced ".

 

The only data i have not is the air composition, but, it is usually the same and i don't believe that a variation from 21% to 20% in oxygen, for example, can make a big effect...

Maybe i should model the combustion reactor in another way instead of a 100% conversion reactor, but, i have not idea  .

 

Thank you again,

 

ADM 

 

 

The oxygen in the flue gas is not incomplete combustion. Excess air is used at the burners to guarantee complete combustion of the fuel. Adding in enough air so that you have the right flue gas composition will add a lot of volume to your flue gas. The 3% oxygen by volume may be misleading. Try reducing the excess air until you match the temperature out of the radiant section to see its effect in the simulation. If the oxygen content of your flue gas is according to the measurement you supplied, consider whether it may be caused by air leakage into the flue gas downstream of the combustion zone. You are also needing to balance the reformer reactor duty with the combustion duty so check the reformer duty. If your simulation is correctly working and it were a real plant, the operators would reduce the firing rate of the burners to reduce the flue gas temperature.

[Pilesar]

Add the convection section into your model and include the stack temperature measurement. The duties and temperatures of the convection section coils will also need to be balanced against the flue gas rate and duty. There is enough redundancy available to allow you to identify which data is suspect. Data reconciliation requires tuning. Measurements may very well be imprecise... especially for something like air flow that doesn't really matter to anyone except the person trying to simulate the process. Operators generally only care if the flow is going up, down, or steady. Is your combustion model sensitive to thermo method change? I would use an equation of state like PR. Complete combustion of the fuel is reasonable. Air includes some water vapor also and the air oxygen content should be known. There is negative pressure in the furnace so air leakage bypassing the air feed meter is also a real phenomenon.