PDA

View Full Version : Ammonia and CO2 reactor



gwapa
03-01-2007, 03:38 PM
Dear Friends

With new labor laws in the working areas , many costumer had asked to pipe all the discharge safety valves to a water tank in order the eliminate the leaks to the working areas and also to the atmosphere.

However as everybody knows, when the water is saturated with ammonia, one has to get the water off the vessel and drains it to the sewerage. In the sewerage, part of the ammonia separate from the water and leave out anywhere, which is a trouble:o

I have heard about the use of Carbon Dioxide (CO2) to combine with Ammonia to produce a salt named Ammonia Carbonate . This salt is like a paste and can be handling “easily”:confused:

My questions are :
what is the residence time required to the CO2 and Ammonia to form the salt?
It is fast or it takes long?.
How could the reactor be made?

Please let me now any reference, literature o experience

Thanks
Gwapa

US Iceman
03-01-2007, 04:32 PM
Hi gwapa,

The use of water dispersion tanks is more complicated than you have listed.

First there are some application problems with the relief valves. The calculation of the relief header sizing has been complicated by a change in the ASHRAE Standard 15 (if you follow this as a safety standard).

According to the revision you have to account for the back pressure imposed on the outlet of the relief header. With the relief header submerged in a tank of water the static head (due to the height of the water in the tank) must also be considered as part of the imposed back pressure.

When ammonia flows into the water, a chemical reaction occurs. This reaction is exothermic, which means the reaction generates heat.

Several other potential problems are:

water getting back into the refrigeration system
water can freeze, so freeze protection of the tank will be needed if the ambient drops below 32F (0C).
Allowing the ammonia/water from the tank to flow into the sewer an add operational problems to the sewer digesters, since ammonia is a fertilizer also. This depends on the type of water treatment system used by the sewer system.
The ammonia/water might also be considered a hazardous waste since it could also have oil in it.Water dispersion tanks can and do work, but they are not the answer in my opinion.

A flare system would be more environmentally responsible and present less operational impacts, but these are expensive also.

This is the first time I have heard of anyone wanting to use CO2 for this purpose.

From some of the stories I have heard about ammonia and CO2 reactions the salt is relatively hard and not very easy to remove.

I am going to look forward to learning more about this as others post to here.

This is an interesting question...;)

NH3LVR
03-01-2007, 05:04 PM
A flare system would be more environmentally responsible and present less operational impacts, but these are expensive also.
;)

US Iceman, I am unfamiliar with the term flare system:confused: .
Could you enlighten me?:)

Josip
03-01-2007, 05:13 PM
Hi, Gwapa :)

Dear Friends


With new labor laws in the working areas , many costumer had asked to pipe all the discharge safety valves to a water tank in order the eliminate the leaks to the working areas and also to the atmosphere. However as everybody knows, when the water is saturated with ammonia, one has to get the water off the vessel and drains it to the sewerage. In the sewerage, part of the ammonia separate from the water and leave out anywhere, which is a trouble:o


I have heard about the use of Carbon Dioxide (CO2) to combine with Ammonia to produce a salt named Ammonia Carbonate . This salt is like a paste and can be handling “easily”:confused:

My questions are :
what is the residence time required to the CO2 and Ammonia to form the salt?
It is fast or it takes long?.
How could the reactor be made?

Please let me now any reference, literature o experience

Thanks
Gwapa



I want to suggest to connect all safety valves discharge lines into the lowest pressure vessel (some safety valves are not affected with back pressure) and then from that vessel one safety valve discharge pipe into water tank. That tank can be emptied by some company capable to remove ammonia from that water.

For example 1 kg of water can absorb up to 0,5 kg of liquid ammonia or about 700 liters of ammonia gas.

I'm not sure but I was informed in Germany some plants have this solution. That depend on pressure and safety regulations what I do not know for Germany, but some other guys here can say more about;)

I'm not sure about possibility to install another chemical plant within ammonia refrigeration plan :eek: . I believe there is something more then just mix CO2 with gaseous ammonia.

Best regards, Josip :)

nh3wizard
03-01-2007, 05:22 PM
If I remember correctly, California has something similar in place as far as piping the reliefs into a water vessel, its been a long time ago, might want to search the EPA site for the state.

NH3LVR
03-01-2007, 06:05 PM
I want to suggest to connect all safety valves discharge lines into the lowest pressure vessel (some safety valves are not affected with back pressure) and then from that vessel one safety valve discharge pipe into water tank. That tank can be emptied by some company capable to remove ammonia from that water.

Josip;
I am not as familiar with the Codes as I should be. I mainly do service/repair/troubleshooting now, and can use that as an excuse. It has become so complicated and applied more stringently in the past few years. Of course the codes in the US are different than other Countries, so knowledge of one Code does not qualify one to comment on another.

However the basic idea here is that we have to increase the required Relief Valve piping one size, then go into a header, which is up sized as each new relief is piped in. (This is not the way the Engineer who stamps the plans does it) This ends up being a rather large header in most cases.
We are about to commence a job of this nature. I have not seen the drawings yet, however it will probably mean at least two 4 (10CM)Inch headers going to the tank, plus some smaller ones.

This would most likely cause a problem connecting back to another vessel. At the very least it would require cutting into and modifying a existing vessel.
The lower pressure vessel would also require a equal sized outlet to the tank.

Please do not think I am criticizing your ideas, just discussion. I have just never seen this done in the manner you describe.

Josip
03-01-2007, 08:59 PM
Hi, NH3LRV :)


However the basic idea here is that we have to increase the required Relief Valve piping one size, then go into a header, which is up sized as each new relief is piped in.

You are right. We all do the same (mistake) because, if open then only one relief valve open in one/same time-thus it is enough to make header the same size as required by the biggest relief valve. I do not know for such case where more relief valves act simultaneously, but maybe I am not right.


This would most likely cause a problem connecting back to another vessel. At the very least it would require cutting into and modifying a existing vessel.
The lower pressure vessel would also require a equal sized outlet to the tank.

Here I was thinking about design of new plant to consider this execution. For existing plants still is possible to install one separate vessel (without refrigerant and pressure in) just for safety purpose ;)


Please do not think I am criticizing your ideas, just discussion. I have just never seen this done in the manner you describe.

Me either:) , as I wrote I've been told about, but I believe that is possible, because there are safety valves working independent of back pressure.

My colleague told me to look on:

http://www.danfoss.com/Products/Categories/Categories.htm?segment=RA&category=http%3a%2f%2fwww.ra.danfoss.com%2fra%2fProducts%2fProductCatalogue.asp%3fNavigation%3dHideO nAllPages%26Footer%3dHideonallpages%26Division%3dRC%26HL%3d2%26HLID%3d542%26AppID%3d%7bc2a95dac-5e01-47df-92cd-5972d837cf1a%7d
http://www.herl.de/herl/english/start_e.htm

not found anything about, but still thinking that is not bad idea....;) at least for discussion


Best regards, Josip :)

US Iceman
03-01-2007, 09:14 PM
Hi NH3LVR,

A flare system is a simple gas burner that burns off the chemical admitted into the flare stack. In these case, the ammonia vapor from the relief header(s) would flow into the flare stack and get burned off.

I believe they use cast iron as a catalyst and the flame to disassociate the ammonia back into hydrogen and nitrogen.

These are used in gas plants to flare off the vapors they do not need or want I believe.

Here is a link for a quick review.
http://www.mrw-tech.com/pia_enclosed_flares.html

US Iceman
03-01-2007, 09:24 PM
However the basic idea here is that we have to increase the required Relief Valve piping one size, then go into a header, which is up sized as each new relief is piped in. (This is not the way the Engineer who stamps the plans does it) This ends up being a rather large header in most cases.


The old ASHRAE Standard allowed something like this in concept. The new method that was adopted several years ago is very involved (a lot of number crunching).

One thing I would caution you about is doing something different than the plans/design sealed by the project engineer.

With his design and stamp, he is the responsible party. If you change the design, then you are responsible. Good intentions can sometimes be the wrong thing to have.

If your company thinks there is something that needs to be done differently, then you and the engineer should discuss this, before it gets installed.

Simply going up one pipe diameter may not be sufficient as the actual length of the relief header is based on total equivalent length. This includes all of the pipe fittings too, which can add a substantial equivalent length to the lineal feet of pipe.

NH3LVR
03-01-2007, 09:38 PM
Josip;
I do agree that is extremely unlikely that all the relief valves would blow at the same time. In fact I would be surprised if that ever has happened. However I could be incorrect in that assumption.

However, in my part of the US are required to install systems to those standards by the authorities. If the local people do not have the expertise to make sure the system is in compliance they bring in a consultant. The consultant will make sure everything is by the book, and I mean everything.
And there will be a list of things to repair/replace/mark etc.

I pass no judgment on the Authorities in these cases. They do not want to take responsibility for an accident by weakening the standards, and I do not blame them.

I had looked before at the Danfoss and Herl Valves. The Danfoss are only available in small sizes. The Herl valves are available in somewhat larger sizes, although it might be difficult to get them past our local inspectors without a design review. I suspect even though the opening is not back pressure dependent the standard tables would not apply because they are not relieving to Atmosphere.

I would certainly appreciate anyone who has done this sort of thing before joining in.

US Iceman
03-01-2007, 09:41 PM
Hi Josip,

You raise a good point about the relief valves. Some are dependent on back-pressure, while others are not.

Those that are not dependent I think are much more expensive and typically used in chemical plants. I think these are pilot operated and they compensate for the back-pressure.

The normal relief valves we use in refrigeration are dependent on the back-pressure though.

In the new ASHRAE Standard, the relief headers have to be sized with this in mind. It involves the calculation of the pressure loss in header during a relief event. Any flow through the pipe has friction loss, so in essence this increases the back-pressure on the relief valve.

If I'm not mistaken the relief event is considered when all of the valves opening simultaneously, which to me is unlikely, but extremely conservative.

To further compound something already difficult it is my impression you cannot/or should pipe relief headers together where different set pressures occur. Each header for a specific set pressure should be installed independently of another relief header with a different (lower/higher) set pressure.



I'm not sure about possibility to install another chemical plant within ammonia refrigeration plan :eek: . I believe there is something more then just mix CO2 with gaseous ammonia.


I too think there is more to this than meets the eye. I'm more than a little nervous about introducing CO2 into a relief header. If the CO2 ever gets into the ammonia (back into the system) it will be a big mess.

It might work, but I think there are a lot of things to consider....



Returning the relief outlets back into the system makes a lot of sense, since we would like to contain the ammonia, instead of releasing it. And there are rules for that also.:D

This is one of those topics that is confusing to solve.

NH3LVR
03-01-2007, 10:18 PM
USIceman;
Please be assured that I was not recommending a method of sizing Relief Valve lines.
"This is not the way that the Engineer who stamps the plans does it" was meant as a disclaimer, so no one would think that I was recommending such. The only proper way to do this is to follow the procedures.
We used to do this, but the days of installing such things by rule of thumb are over. The explanation was for discussion purposes of general ideas only.

US Iceman
03-01-2007, 10:37 PM
NH3LVR,

Ooops, sorry about that. I tend to jump on those sort of comments quickly.

No offense meant.:o

gwapa
04-01-2007, 12:56 AM
Dear friends
The main ploblem (Focus) is how to get rid the ammonia off the working area.
The capacity of a safety valve is a function of the difference of pressure between the back a discharge pressure. So if we increas the discharge pressure in any way we have to check de capacity of the safety valves according with the new Delta P in order the vessel be protected. If you install a second safety valve in series down stream you again have to check all the valves and all the systems( vessels and heat exanger).Of couse I do not agree with install down stream a second safety valve.
In a food procesing plant sometimes one can have in the main receiver 3000 kg of ammonia and also in the traps and separators tank. In case a safety valve open what can a human do? I know your answer......run and faster
The idea to burn the ammonia sounds good. I think that we have to mixture with some air to have a mixture of 30% of ammonia and burn it.
thanks for your good ideas

US Iceman
04-01-2007, 01:38 AM
Hi gwapa,

We have been busy discussing your question as you see.:D



If you install a second safety valve in series down stream you again have to check all the valves and all the systems( vessels and heat exanger).


I did not think this was a legal procedure for unfired pressure vessels. Have you seen this done before?



The idea to burn the ammonia sounds good. I think that we have to mixture with some air to have a mixture of 30% of ammonia and burn it.


From the data I have:
The upper flammability limit is 25-28%
The lower flammability limit is 15-16%
and an ignition temperature of 1204F (651C)

I believe the flare stacks have a powered fan to draw in the required volume of air to support the combustion process. I also seem to remember the flare should be on continuously in the event of a release of ammonia. So, there are some operating costs too.

Josip
04-01-2007, 01:10 PM
Hi, Gwapa :)


The main ploblem (Focus) is how to get rid the ammonia off the working area.

How much ammonia you have there?:confused: Why so many leaks? Are you producing ammonia? As I remember you make a beer:D


So if we increas the discharge pressure in any way we have to check de capacity of the safety valves according with the new Delta P in order the vessel be protected.

I'm :confused: why to increase discharge pressure? Which vessel? You must have your safety valve matching size and pressure of your vessel. On the vessel you have a plate with max. working pressure and your safety valve must open at that pressure or before.


If you install a second safety valve in series down stream you again have to check all the valves and all the systems( vessels and heat exanger).Of couse I do not agree with install down stream a second safety valve.

You cannot install safety valves in series:eek: except if you want to have a bomb.


In a food procesing plant sometimes one can have in the main receiver 3000 kg of ammonia and also in the traps and separators tank. In case a safety valve open what can a human do?

For that reason safety valve must be installed in gas area of your vessel (top side) thus to reduce pressure of gas and eventually release that gas into air/water. It is forbidden to install it in liquid line-there we have overflow valves, bypass valves internal execution i.e. to release that pressure back into closed system.


The idea to burn the ammonia sounds good. I think that we have to mixture with some air to have a mixture of 30% of ammonia and burn it.

I must disagree with you. That is only possible and applicable for chemical plants where you have constant waste and you have to disintegrate it with flame.

We have some refrigeration plants (not chemical) with more than 40000 kg of ammonia and no smell there. I think it is better to start with leaks repairing;)

If you have a good beer I can come and help about;)


Best regards, Josip :)

gwapa
05-01-2007, 12:04 AM
Thanks Josip/ US Iceman for your very kind advises

I do not talk about normal leaks. I am talking about a method to get rid the ammonia from safety valves off the working area. That is, just in case a safety valve opens.

Once I received a proposal for a system to purge the noncondensables gases from an ammonia system. In that design the purged “air” was discharged in a pot with water with an atmosphere of CO2. We did not buy the system and I could not know the detail of operations. I am wandering if one can make a big system to fight against a discharge of a safety valve .

Remember what the project is. The costumer wants to connect all the outlet of the safety valves to a manifold. The manifold will be designed with a diameter big enough to handle the discharge of the safety valves. Also the manifold should introduce minimal pressure drop. Each safety valve would be change to a dual model with shut off valve (keeping one of them always open). Also we are recommending installing a rupture disk with sensor to detect which one of the valves is leaking or opened .Now the question is: what can we do with the ammonia discharged in case it happed?
Normally the first idea is to discharge it to a vessel with water. And the second question is : where will we discharge the solution of ammonia –water?.

US Iceman point out that the mixture is not very good for either the water treatment plant nor sewages

Could we use CO2 to neutralize the ammonia water mixture?
How fast the reaction with the CO2 is?
How will be the shape of the reactor?

These are some question which I wish to post

Thanks to every body

Andy P
05-01-2007, 12:16 AM
I have heard about the use of Carbon Dioxide (CO2) to combine with Ammonia to produce a salt named Ammonia Carbonate . This salt is like a paste and can be handling “easily”:confused:

My questions are :
what is the residence time required to the CO2 and Ammonia to form the salt?
It is fast or it takes long?.
How could the reactor be made?

Please let me now any reference, literature o experience

Thanks
Gwapa
Hi all

back to Gwapa's original question - we did some testing a year ago (and I still haven't written up the formal report :( ) on mixing CO2 and ammonia to reduce venting from ammonia relief valves. There are two possible compounds produced. They are quite similar and both relatively non-toxic, but different in detail. If the ammonia and CO2 are mixed in a pressure vessel (without air) then ammonium carbamate will form. This is hard white crystals that will block everything and are also very corrosive to steel - not acceptable. If the ammonia is bubbled through water at the same time as CO2 then ammonium carbonate is formed. This is a white precipitate (solid forms in the water) and relatively non-toxic. The compound is also known as Hartshorn (and was originally made (allegedly) from powdered reindeer antlers!) and is used in baking Scandinavian biscuits. Unfortunately it needs a huge amount of CO2 to react with the ammonia - 2 molecules for every one of ammonia, so 88kg of CO2 for every 17kg of ammonia. Also the resulting liquid still smells strongly of ammonia and would need to be treated as hazardous waste. We concluded that the experiment taught us a lot, but mainly that this had not been a good idea!

cheers

Andy P

gwapa
07-01-2007, 05:50 PM
Andy P
It was a very good test .Congratulations. May I have a copy of your final report ?
What is the density of the resultating liquid?
Could one pump it away?
What will happen if we pump the liquid to an hermetic vessel? The free Co2 or free ammonia will make an internal pressure?
Best regards
Gwapa