56 replies to this topic

[A_D_M_MII]

Hi everyone,

 

I am doing a simulation of a steam reforming plant, the objective of this simulation is to achieve the real operation data of controllers and indicators.

 

I have already simulated this plant with the design data but like everyone knows the reality is more complex, however, this is my question:

 

How can i simulate the decrease of the catalyst activity?, i mean i have simulated the reformer reactor like a PFR with the kinetic equations, but it is a theoretical and experimental equation that doesn't take into account other parameters like the catalyst deactivation. 

 

In the other hand, my principal problem is that flue gas that goes to the convection zone has more temperature than the real temperature that is given by the indicator.

 

I have simulated it like a combustion adiabatic reactor (100% Conv), the product is a current of flue gas with high temperature that gives the "radiant energy" by a cooler to the reformer (PFR), at the exit of this cooler is supposed to be the temperature indicator.

 

I also have the feed (kg/h), the composition (% mol), temperature, and pressure of the inlet currents of the combustion reactor (Furnace).

 

I hope i was clear.

 

Thank you all!

 

ADM 

 

[Pilesar]

A few comments for consideration... 

Consider using equilibrium reactors instead of PFR. Adjust the approach to equilibrium to tune for catalyst performance. 

Are you adding excess air to your combustion? Extra air lowers flue gas temperature and affects the convection section heat transfer. The heat transfer to the reaction is in the balance also. An inaccurate heat of reaction might be affecting your results.

Including the convection section in the model will provide additional balances that have to be matched with the flue gas flow and temperatures. Adding these additional balances will help data reconciliation efforts to identify misleading field measurements.

[SilverShaded]

I agree with Pilesar an approach to equilibrium apporach is fine.  In terms of catalyst deactivation why model it?  Some HPU's have run for 25 years without changing catalyst (mostly because they were running at low temperature) but these catalysts do not deactivate quickly.

Edited by SilverShaded, 15 September 2021 - 08:46 AM.

[A_D_M_MII]

 

A few comments for consideration... 

Consider using equilibrium reactors instead of PFR. Adjust the approach to equilibrium to tune for catalyst performance. 

Are you adding excess air to your combustion? Extra air lowers flue gas temperature and affects the convection section heat transfer. The heat transfer to the reaction is in the balance also. An inaccurate heat of reaction might be affecting your results.

Including the convection section in the model will provide additional balances that have to be matched with the flue gas flow and temperatures. Adding these additional balances will help data reconciliation efforts to identify misleading field measurements.

 

OK thank you!

 

1. But, excess air could be better because of the decreases of the initial flue gas temperature at the outlet of the adiabatic combustion reactor, so for the same thermodynamic of the reaction of the reformer, the heat which is taken by the reactor is the same, and the flue gas to convection zone will be less than with a reduction of the air flow.

 

2. Heat of the reaction is given by Hysys, i can do nothing, the composition is also correct, this is the fact the made me think about the deactivation of the catalyst, but it's difficult.

 

3. Yes i know what you mean, i could verify the heat balance with the rest of the process with the convection zone.

 

Thank you again!

[PingPong]

It's not clear to me how your model works, but I agree with the others that you best use ATE.

 

Note also that your firebox has a heat loss that impacts the flue gas temperature entering the convection bank.

Convection bank also has some heat loss.

 

I suggest you read this topic: https://www.cheresou...-reformer-unit/

 

and in this topic I explained how to verify whether lab analysis are correct or not: http://www.cheresour...ormer-catalyst/

 

EDIT: first topic link was wrong, replaced by correct one.

Edited by PingPong, 16 September 2021 - 01:52 PM.

[SilverShaded]

 

 

A few comments for consideration... 

Consider using equilibrium reactors instead of PFR. Adjust the approach to equilibrium to tune for catalyst performance. 

Are you adding excess air to your combustion? Extra air lowers flue gas temperature and affects the convection section heat transfer. The heat transfer to the reaction is in the balance also. An inaccurate heat of reaction might be affecting your results.

Including the convection section in the model will provide additional balances that have to be matched with the flue gas flow and temperatures. Adding these additional balances will help data reconciliation efforts to identify misleading field measurements.

 

OK thank you!

 

1. But, excess air could be better because of the decreases of the initial flue gas temperature at the outlet of the adiabatic combustion reactor, so for the same thermodynamic of the reaction of the reformer, the heat which is taken by the reactor is the same, and the flue gas to convection zone will be less than with a reduction of the air flow.

 

2. Heat of the reaction is given by Hysys, i can do nothing, the composition is also correct, this is the fact the made me think about the deactivation of the catalyst, but it's difficult.

 

3. Yes i know what you mean, i could verify the heat balance with the rest of the process with the convection zone.

 

Thank you again!

 

Check the heat of reaction carefully, some older versions of hysy were calculating it incorrectly in some models. e.g. Gibbs reactor gave a different value to conversion reactor for the same amount of conversion.
 

Edited by SilverShaded, 17 September 2021 - 03:38 AM.

[PingPong]

@ A_D_M_MII

 

Note that I edited my previous post: First topic link was wrong, replaced by correct one.

[daraj]

i have a question regarding efficiency of steam reformer to model the fuel side. If say the endothermic reactions in the reformer require 100MW of duty, what is the actual input energy needed from NG combustion? i.e what is the efficiency will you consider here? is it 90%?

 

will you burn NG equivalent of 110MW or 120MW?

 

what is the current state of the art efficiencies of reformers?

[Pilesar]

To model the fuel side, you need to quantify all the uses for the fuel. The reaction heat is just a part of the overall heat required in a steam reformer. Some more uses: heat the feed up to the reaction temperature, heat the water to the boiling temperature, boil the water, superheat the steam, heat the fuel to the combustion temperature, heat the air to the combustion temperature. All these steps require energy. Efficiency of the operation is another way of asking 'How much energy escapes out of the combustion flue stack unused?' 

[A_D_M_MII]

Hi everyone, 

 

I have other question, considering that my plant uses NG with a low quantity of C2 and C3 hidrocarbons, what is the limit of the temperature at the inlet of the pre-reformer, i have read about the sintering of the catalyst but i don't know what is the limit. I also know that a high temperature could happen carbon deposition but, what is the safety band or the upper limit of this temperature  at the inlet of pre-reformer?

 

If we forget this first question, i have another one:

 

On the other hand, could be possible that part of "partial reforming" (i.e C2H6 -> H2 + CH4) takes places in the reformer ?

 

Thank you so much for your help! 

 

ADM 

[SilverShaded]

Hi everyone, 

 

I have other question, considering that my plant uses NG with a low quantity of C2 and C3 hidrocarbons, what is the limit of the temperature at the inlet of the pre-reformer, i have read about the sintering of the catalyst but i don't know what is the limit. I also know that a high temperature could happen carbon deposition but, what is the safety band or the upper limit of this temperature  at the inlet of pre-reformer?

 

If we forget this first question, i have another one:

 

On the other hand, could be possible that part of "partial reforming" (i.e C2H6 -> H2 + CH4) takes places in the reformer ?

 

Thank you so much for your help! 

 

ADM 

 

It depends on the catalyst.  Some units have pre-reforming catalyst in part of the furnace tubes.
 

[daraj]

To model the fuel side, you need to quantify all the uses for the fuel. The reaction heat is just a part of the overall heat required in a steam reformer. Some more uses: heat the feed up to the reaction temperature, heat the water to the boiling temperature, boil the water, superheat the steam, heat the fuel to the combustion temperature, heat the air to the combustion temperature. All these steps require energy. Efficiency of the operation is another way of asking 'How much energy escapes out of the combustion flue stack unused?' 

Thanks Pilesar. 

 

what I found in general was that if i extract enough neergy from flue gas, such that they leave around 150-175C from the reformer, the general

maximum efficiency is close to 920-93% in current state of art reforers? is this right? i.e you can recover almost 92% of the input energy and only 8% is lost via flue gas(and maybe other losses like radiation leaks,insulation etc)

[SilverShaded]

 

To model the fuel side, you need to quantify all the uses for the fuel. The reaction heat is just a part of the overall heat required in a steam reformer. Some more uses: heat the feed up to the reaction temperature, heat the water to the boiling temperature, boil the water, superheat the steam, heat the fuel to the combustion temperature, heat the air to the combustion temperature. All these steps require energy. Efficiency of the operation is another way of asking 'How much energy escapes out of the combustion flue stack unused?' 

Thanks Pilesar. 

 

what I found in general was that if i extract enough neergy from flue gas, such that they leave around 150-175C from the reformer, the general

maximum efficiency is close to 920-93% in current state of art reforers? is this right? i.e you can recover almost 92% of the input energy and only 8% is lost via flue gas(and maybe other losses like radiation leaks,insulation etc)

 

Flue gas exit temperature is usually designed to be about 120C

[daraj]

Hi silver, for isothermal wter ga sshift reactors, cna they be modeled as RGIBBS with an approach to equilibrium?(Linde or haldoe topsoe isothermal shift(?)  what is the maximum CO conversion typically seen industrially and whatare te preferred steam : CO molar ratio at inlet of the reactor?

[daraj]

 

 

To model the fuel side, you need to quantify all the uses for the fuel. The reaction heat is just a part of the overall heat required in a steam reformer. Some more uses: heat the feed up to the reaction temperature, heat the water to the boiling temperature, boil the water, superheat the steam, heat the fuel to the combustion temperature, heat the air to the combustion temperature. All these steps require energy. Efficiency of the operation is another way of asking 'How much energy escapes out of the combustion flue stack unused?' 

Thanks Pilesar. 

 

what I found in general was that if i extract enough neergy from flue gas, such that they leave around 150-175C from the reformer, the general

maximum efficiency is close to 920-93% in current state of art reforers? is this right? i.e you can recover almost 92% of the input energy and only 8% is lost via flue gas(and maybe other losses like radiation leaks,insulation etc)

 

Flue gas exit temperature is usually designed to be about 120C

 

 

 

can i model the fuel combustion as an adiabatic RSTOIC? if so what will be the diffeernc ebetween exit gas temperature from this reactor vs the reformer exit? lets say reformer exit is 850C? should i have an exit of 1000C from my RSTOIC? iam talking about the temperature of the combustion gases generated. what is the delta T between this and reformer exit temp. in top-fired reformers? this will determine heat input needed to the reformer

[daraj]

pilesar, awaiting your responses. also for assumptions of pressure drops across SMR and shift reactors what are some reasoanable values?

[daraj]

also while ch4-water reaction is endothermic, the water ga sshift in SMR is endothermic. so if i were to model SMR as RGIBBS how wil i set an overall temperature approach? i know only methane slip at the exit of reformer and the temperature, do not know other compositions. should i keep changing approach to equilibrium until i get desired ethane dslip and at that targetted temperature? 

[daraj]

no responses for these?

[SilverShaded]

Either you have a real unit your tryiing to model in which case you have operating data or you don't, in which case your asking for details from a licensed process unit and in all likelihood providing that detailed information would violate NDA agreements with the licensor.

You will need to used some enginerring judgement or do some actual calculations.

[daraj]

silver and others, for calculating duty of the SMR(in order to calculate required fuel needed for SMR) should i calculate duty using simulator assuming NG and steam entering at what temperature in the RGIBBS reactor block? I believe the natural gas and steam are partially pre-heated by the flue gas before they enter the SMR reactor tubes. This pre-heating is also provided by the flue gas heat content(heat recovery). So that makes it a bit tricky to calculate the net heat duty which needs to be provided by combustion. any thoughts on how to do this?

[daraj]

problem iam facing is the flue gas from the combustion reactor(even whne NG and air are fed at near ambient condiitons) is quite hot at around 1900C or so, and even after giving up heat for SMR duty they stll continue to be around 1430C, which is much higher than what is needed. this should be close to 1050C ideally. Not sure how to adjust this. should i deduct any further heat to account for radiation losses? will that help?

[daraj]

Okay i reduced fuel gas proportioantely to bring the temperature down. can you help me with following?

 

1. what is the typical temperature of flue gas finally leaving to stack? around 150C?

 

2. in the model we assume that all heat producedby combustion is used in providing heat duties + heat leaving through flue gas. But there may be some losses also. what % of combustion duty is typically lost? how can i incorporate this in the model?

[SilverShaded]

Okay i reduced fuel gas proportioantely to bring the temperature down. can you help me with following?

 

1. what is the typical temperature of flue gas finally leaving to stack? around 150C?

 

2. in the model we assume that all heat producedby combustion is used in providing heat duties + heat leaving through flue gas. But there may be some losses also. what % of combustion duty is typically lost? how can i incorporate this in the model?

Flue gas exit temperature can be as low as 120C.  Theres normally no need to account for losses in the simulation.

[daraj]

silver, if reformer exit is 850C, then my flue gas temperature can be about 1050C AFTER it has provided the duty for the reforing reactions, right? what is the typical delta temperature range between reformer exit temp. and flue gas temperature after it has provided reformer duty?

is this what is alled the bridgewall temperature?

[daraj]

also is desulfurized NG used for reformer fuel, or pipeline undesulfurized NG as such?