7 replies to this topic

[Khepa]

Dear all,

         Thank you in advance for going through my query.

 

I have joined a company who are  responsible for etablishing a Natural gas processing plant of 60MMSCFD using silica gel.

Though There is a consortium with a reknowned company but the process implementation has to be done by us. I am facing a dilemma regarding the accuracy and efficiency of the Desiccant tower switching issue.

 

Please do correct me and guide me if I am wrong; The main objective is to dehydrate the gas and then regenerate the tower again using heated gas. The process description is given below.

 

1. Our project is 3 tower based

2. One will be in adsorption while one will be in cooling and the last one will be in regeneration stage and that's how the cycle revolves.

3. In adsorbtion the gas enters from top of the tower but for regeneration (heating) and cooling the gas enters at the bottom.

4. All this switching will be done on the valve skid where 18 solenoid valves gonna be in action

 

If the switching was done instaneously there would have been no issues,but the problem is as per vendor of the silica gel the time for adsorbtion is the highest and for heating it is 84 seconds less and for cooling it is 168 seconds less than adsorbtion.

 

My question is if I wait for the adsorbtion to complete and run the heating and cooling process as it is won't it be harmful for the silica gel properties. Also each process has its own pressure setting; adsorbtion 78 bar, Heating 80 bar and cooling 76 bar.

 

I would really appreciate if someone kindly and considerably shares his/her experience regarding this issue. Whether to adjust the process such way that all the three process consumes almost the same amount of time or switch the valve in any other way which will allow and satisfy all the process requirement including the equalization of the pressure?
 

Apologies in advance if you find my query novice, which I am . With all due respect I just don't want things to get done but to understand how they are done and why particularly that way. As I have just started my carrer as a process engineer I hope that my respected seniors and experienced members here will rectify me and  guide me to become a better engineer.

Thank you & best regards,

Mohaimin Sunny

[ankur2061]

Hi,

 

Are you aware of "Time Delay Relays" for automatic on-off valves which allow staggering of valve opening or closure. A time delay relay provides a delayed a delayed on/off actionwith an adjustable time delay. Talk to your instrumentation engineer and also refer the link below:

 

https://books.google... valves&f=false

 

Regards,

Ankur.

[manojkaila]

Dear Khepa,

 

I have near to same system ( Dryer with moleculer sieve ) experience. I would like to share following things.

 

1 You must have some ideal time after regeneration. Because initially system efficiency is more as time pass efficiency will reduce & your cycle time for adsorption will increase (it must depends on outlet moisture).

2 I worked with MOV based system with logic generated in DCS and which works well. We can adjust the timer as per the requirement. 

3 For inlet from the TOP use some good distributor and avoid direct impact for which I have very bad experience.

4 For pressure equalization you can put small (2") bypass to main Valve to avoid impact of sudden pressure change.

 

 

Regards 

[Bobby Strain]

It is usual to buy the system as a package. The vendor will provide a PLC for cycle control. Best not to mess with it.

 

Bobby

[ColinR33]

We design and build these sorts of dehydration packages.  On a three tower system, two towers are typically in adsorption while the 3rd tower is in regeneration or standby.  The cycle times are staggered to ensure no interruption in dehydrated gas flow and take into account the adsorption, heating, regeneration, cooldown and standby time as well as the switching times of the valves.   These are dependent on the gas composition, water content, dry gas spec, flow rate, available regeneration gas (typically a dry gas slip stream) and selected dessicant. As Bobby states, the PLC with all this programmed in can be supplied with the package, or loop narratives and cause/effect diagrams with the switching sequence can be provided for 3rd party programming of the overall facility DCS.

 

Cheers,

Colin

Edited by ColinR33, 11 December 2015 - 12:47 PM.

[Reliability Eng]

I think there is some missing info in your inquiry, I am totally agree with the explanation that given by Colin. The three beds should be organized to be two beds in adsorption mode and one in regeneration mode. what I would like to add is before doing any modification in the cycles timer you should conduct a breakthrough test to evaluate the bed performance. I am sure there is a moisture online analyzers that telling you how the bed is performing. by following the online analyzer and make the trend you can obviously see your bed performance.

 

based on the analyzer trend you can modify the cycle time either by extend or reduce the adsorption time. you have to take care when you are modifying this timers as it is very critical and can cause moisture breakthrough and trip the plant.

 

Please send us more details about the online analyzers and their locations in order to give you clear feedback.

[ColinR33]

Very good point, if you want to start changing cycle times on an existing unit then yes, you definitely have to do breakthrough testing.

[Khepa]

Dear All,

       Thank you for your kind and considerate reply. I also apologize for such a late reply.  I understood the perspective the members tried to impose and personally thanks to manoj for sharing his field experience.

 

If possible I will definitely share how we finally tackle and accomplish the situation

 

Thank you all again, Take care