16 replies to this topic

[bikrom]

Hi,

 

I have a problem with steam reformer in hydrogen generation unit. Feed to reformer is Natural gas and a small quantity of hydrogen rich off-gas from other units. These hydrogen rich off gas are treated for sulfur and chloride removal before being fed to reformer along with Natural gas and steam. Current plant load is 55 % of design only. S/C ratio is 3.9. (pretty high compared to required 3.5 at this plant load as per design)

 

I am attaching  LAB test results for reformer inlet and outlet gas compositions as image files.

 

Please note that CO in reformer outlet has increased from normal values of 12-13 mol% to 30 mol%. However, methane in reformer outlet is as low as earlier, indicating that methane conversion is fine

 

CH4 + H2O = CO + H2

 

 

it is the secondary water shift reaction which is not taking place well.

 

CO + H2O = CO2 + H2

 

And that puzzles me. Could anybody please throw some light on this?

 

with regards

Bikrom

Attached Files

•  Before problem.bmp   1.42MB   65 downloads
•  After problem.bmp   1.44MB   58 downloads

Edited by bikrom, 01 March 2014 - 10:01 AM.

[PingPong]

At first glance one notices the 1.59 mol% C6+ in the present natural gas. On a weight basis that is quite a lot. That may be a problem for the catalyst run length, however it seems then strange that the methanation equilibrium looks unaffected but the watergas shift is. I could imagine the opposite, but then, I am not a catalyst expert.

To enable others to advise, you should add following information:

What are heater outlet Pressure and Temperature in each occasion?

What are S/C ratios in each occasion? Or better: what are Natural Gas and Steam flowrates to the reformer?

What exactly is meant by Before and After? Before and After what? What happened in between? An upset? Or simply passing of time? Or is Before the original design (theoretical) situation, and After the actual present operation at 55 %?

Edited by PingPong, 01 March 2014 - 12:07 PM.

[bikrom]

I will gather the data and put here.

 

The problem just materialized one day. One day LAB test was OK. The next day it was not. That is before and after the problem. There is a high temperature shift reactor after the reformer, which brings down the abnormally high CO conc. to normal levels and hence no apparent change was faced in unit operation. It is only from the LAB test results that we know that there is a problem. There was no upset in between or any intentional or unintentional change in any unit parameter. Unit load also remained same.

 

However, there was a upset in the upstream unit where hydrogen rich off-gas is treated for sulfur and chloride removal and we are ready to accept the possibility of poisoning due to escape of sulfur/chloride. But again as you pointed out, one reaction perfectly ok the other totally not ok. This confuses me.

 

And yes, C6+ is really high since last few days too. Earlier it used to be less than 0.5 mol%. It actually is higher since the problem started. (exactly coincides). I hadn't noticed that though.

 

Anyway, I will gather data and post here. Thank you for your reply.

Edited by bikrom, 01 March 2014 - 04:54 PM.

[PingPong]

I had a closer look at the compositions according to the LAB tests, because they seemed not reliable to me.

 

For example: the amount of dry gas normally increases by a factor 3 to 4 between reformer inlet and outlet. As a result of that, the concentration of inerts such as N₂ should decrease by a factor 3 to 4. However in those LAB results it does not. Before it even increases. Not a big deal by itself as N₂ concentrations are small anyway, but it made me suspicious.

 

One can calculate from a Carbon balance (C atoms in = C atoms out) that Before the amount of dry syngas is 3.85 mol per mol NG feed. I assumed for the moment that C6+ is all C₆H₁₄

All O atoms in the CO and CO₂ product must have come from the reacted steam, except for any CO₂ and CO already present in the NG feed, so one can calculate that Before 1.35 mol Steam per mol NG must have reacted. However there is also a Hydrogen balance and from that one can calculate that only 0.99 mol Steam per mol NG must have reacted. That is a big difference, so the LAB analyses for Before are not very accurate.

 

Doing the same exercise on the After data gives even worse results:

2.70 mol Dry Syngas / mol NG, calculated from C-balance

1.29 mol Steam Reacted / mol NG, when calculated from O-balance

-0.39 mol Steam Reacted / mol NG, when calculated from H-balance (note the negative sign)

That is obviously not possible, so these LAB analyses cannot be right at all.

 

I am not saying that there may not be a problem in your unit, but with such lousy LAB test data it will be hard (or impossible) to determine anything with certainty. It could also be that the LAB is the real problem.

 

Certain people in your organisation will not want to hear that the LAB analyses must be wrong, so I suggest you try to calculate the above numbers yourself, and in doing so get a good feel for this kind of calculations, before you confront them.

Edited by PingPong, 02 March 2014 - 11:50 AM.

[rychurek]

I will gather the data and put here.
 
The problem just materialized one day. One day LAB test was OK. The next day it was not. That is before and after the problem. There is a high temperature shift reactor after the reformer, which brings down the abnormally high CO conc. to normal levels and hence no apparent change was faced in unit operation. It is only from the LAB test results that we know that there is a problem. There was no upset in between or any intentional or unintentional change in any unit parameter. Unit load also remained same.

Hi

Almost 2.5 CO increase in reformer outlet gas should influence in shift reactor equilibrium:
1. Higher CO concentration after shift reactor.
2. Because of shift reaction exothermic energy effect. Higher temperature rise across shift reactor.

[bikrom]

 

I will gather the data and put here.
 
The problem just materialized one day. One day LAB test was OK. The next day it was not. That is before and after the problem. There is a high temperature shift reactor after the reformer, which brings down the abnormally high CO conc. to normal levels and hence no apparent change was faced in unit operation. It is only from the LAB test results that we know that there is a problem. There was no upset in between or any intentional or unintentional change in any unit parameter. Unit load also remained same.

Hi

Almost 2.5 CO increase in reformer outlet gas should influence in shift reactor equilibrium:
1. Higher CO concentration after shift reactor.
2. Because of shift reaction exothermic energy effect. Higher temperature rise across shift reactor.

 

 

Hi rychurek,

 

You are right. 

 

1. Co concentration after HTS did increase from 1.3 to 4.1 mol%. But  we can manage 4.1. (and we came to know about it only from the LAB test)

 

2. Delta T across reactor also increased by 5 deg. C .Still not that great a deviation. (from original 61 deg. C to 66 deg. C)

 

That's why there was no significant apparent change in unit parameters.

[PingPong]

1. Co concentration after HTS did increase from 1.3 to 4.1 mol%. But  we can manage 4.1. (and we came to know about it only from the LAB test)

If those CO concentrations were determined by the same person and/or same apparatus as the syngas analyses then I would not put any value on it.

 

 

 

2. Delta T across reactor also increased by 5 deg. C .Still not that great a deviation. (from original 61 deg. C to 66 deg. C)

So according to the LAB analyses:

Before the CO content of the syngas decreased 11.8 % (from 13.06 % to 1.3 %) in the HTS resulting in a temperature rise of 61 ^oC

and After the CO content of the syngas decreased 25.9 % (from 30.02 % to 4.1 %) in the HTS resulting in a temperature rise of 66 ^oC

Does that look consistent? I don't think so.

 

Once again: the LAB analyses for the syngas are wrong, especially the After one which is complete rubbish.

With the technique that I described before anybody can easily determine whether a syngas analyses is correct or not.

 

An even simpler method to illustrate that the After syngas composition is rubbish is as follows:

From the After NG analyses one can simply determine that the H/C ratio = 3.81 by counting all H and all C atoms.

Doing the same with the After Syngas analyses gives H/C ratio = 3.09

As the amount of C-atoms does not change in the reformer, this means that the After Syngas contains less hydrogen atoms than the NG feed. In reality it must be much more due to the hydrogen from the reacted steam.

Conclusion: that syngas analysis is rubbish.

Edited by PingPong, 03 March 2014 - 04:46 AM.

[rychurek]

I agree with PingPong there should be Lab analysis mistake. HTS reactor could not convert higher content CO without higher steam to carbon ratio or/and higher HTS reactor inlet temperature. You wrote that was no change in product yield, but as Lab alalysis indicate CO was 4.1 so hydrogen production should be less and more unreacted staem(more cold and hot condensate- compare condensate valve openings)

[bikrom]

rychurek and Pingpong, thank you very much for your time. A lot of calculations have been done by you. I will now do them myself to learn as well as confirm. I am aware that our LAB data is not reliable (despite all the certifications they have), but cant really help it. We always take it with a pinch of salt. If they report 13 mol%, we understand that it could be anything between 10-15 %. But when they say 30, we give them some attention. And they are consistently giving this new set of test results. Coincidently, as Pingpong pointed out in his first post, this new phenomenon (if you will) coincides with the increase in C6+ content. 

 

I will do these calculations, and get back to you with any developments.

[bikrom]

I happen to have a model of the plant. I will check what it says about HTS reactor performance with change in inlet gas regarding delta T and all.

 

rychurek, when I said there was no change in unit parameters, I meant no strikingly apparent change. Hydrogen yield from PSA, which depends on a lot of factors including ambient temperature, easily varies here + -1%. There has been a reduction in yield. BUt since then we have changed so many things, it is difficult to say which caused which, e.g. S/C ratio has been increased from 3.9 to 4.4 now. 

 

I will get back with some calculations from my side. Thank you for your time.

Edited by bikrom, 17 January 2015 - 12:35 AM.

[mahesh6903]

Please send me exact flow sheet of hydrogen prodcution from naphtha

desulfurization of naphtha by two ways, first with co-mo catalyst and secondly with zno bed. then feed to reformer and through psa h2 purification and in first step after co-mo bed h2s to be removed with the help of stripper. Naphtha inlet s content 1000 ppm and after zno bed , 0.1 ppm., please help me. i am student.

[mahesh6903]

my mail id is mahesh6903@yahoo.com

[MJ94]

Do you have a PreReformer??
Preformer which operates at 480 'C majorly converts Ch4 to h2 via Shift COnversion , also methanation reaction which is reverse of SMR also takes place , both these leads to consumption of CO , if your Prereformer catalyst will deactivate , its load will shift to Reformer which operates at higher temperature 800 ' C cot where predominantly SMR rxn takes place which produces CO .

[Steve Ransome]

I think you may have a sampling issue. It is common to have reverse water gas shift, converting CO2 to CO in the sample lines. It is critical that the sample is cooled before being taken. Have you been out with the laboratory when they take the samples to see that they are doing things properly? All the calibration certificates in the world won't give you good analysis if you don't sample properly.

[tarafdar]

Hi Steve,

 

You say there is a possibility of reverse shift reaction in the sample line.Could you please explain how this may happen?

 

Reverse shift will be very helpful to control global warming & to produce many chemicals.

 

Regards

[MJ94]

I dont know about sample lines but by reverse shift reaction Steve means FT synthesis , in Reformers which have a HT and LT , Fe in the catalyst promotes FT synthesis , this is Reverse Shift reaction in Shift Converters.

 

And yes it is helpful technology known as GTL gas to liquid

[bikrom]

Thank you everybody. I had not checked this thread since long. To put original doubt to rest. As it turned out, there were serious issues with the LAB machinery and the tests were indeed very wrong. One fine day, after they sorted out their equipments, results became normal. Sorry for the trouble. But i learnt at least one thing, 'dont trust lab results'.