Ammonia gas detection system 3

[spencer1515]

Hello! I am an engineering intern who is working at a natural gas power plant this summer. One of my first projects is centered around finding a detection and alarm system for the plant's ammonia pits. The two aqueous ammonia tanks are each in their own rectangular concrete pits next to one another. The pits are roughly 12 feet in width, 30 feet in length, and 10 feet deep with the storage tanks poking out. There is a catwalk above the tanks. I need three sensors and an alarm system to let us know if there is a leak and to continually monitor the area. One sensor in one pit, one sensor in the other, and one on the catwalk. I would like to do a good job on this but I do not really know where to start. I was not given that much information on this and I think I have free reign to come up with something. It would be ideal to 1)continuously monitor the area, 2)alert to a leak in area and 3)connect back to the control room's DCS when the alarm does go off.

I have researched the types of sensors and I think an infared one would be best. It needs to be something that can be installed and then not have to worry about it. It should be durable since it will be outside and it should last a long time ideally. Something that does not have to be calibrated regularly. I am unsure whether I will need a controller. In the event of a leak, having the controller near the tanks' area does not make sense. Is a controller necessary?

If anyone has any advice whatsoever or knows someone who has done similar projects please reach out and let me know. I would really appreciate it.

Sincerely,
An Intern Who Wants To Do Well On Project #1

 

[pierreick]

Hello,
You should contact vendors to support your query . In the past (Soda ash plant) I was using equipment supplied by Draeger and Oldham for similar application.
One important thing is the calibration which needs to be performed regularly , at least once a year . Vendors provide this kind of service and deliver certificate.
Probably good to check with your HSE manager the best monitoring system for this application , probably a separate system dedicated to safety .
my experience.
Pierre

 

[spencer1515]

Thanks Pierre. I have contacted a few vendors already, Draeger being one. I am looking to hear back from them today.

I was wondering about the range I should be considering as different types of sensors monitor different levels. Obviously, if there was a major leak and the ppm was high we would need to be notified, but I also probably need a system for continuous monitoring at a much lower ppm. OSHA says 50 ppm, so I would have an alarm for that level and then just have a continuous monitoring system in place too. I do not want the alarm to trip every time a little ammonia comes from the vents.

I am curious about the fact that the tanks are outside, albeit in a recessed pit area, but with the wind fluctuating I do not know the best spot to house the sensors in order to detect them best. Anyway, I'll pose these questions to the vendors I talk to. Any other advice you have would be much appreciated. Thanks for your time.

Spencer

 

[Compositepro]

FYI, ammonia is lighter than air.

 

[shvet]

FYI, ammonia is lighter than air.

Understood molecular weight of NH3 is less than that of air. But as per my experience with leakages of gaseous NH3 and spillages of liquid NH3 on a real facility it is heavier than air. As an example:

https://www.youtube.com/watch?v=sNkdAs1e7Cw

https://www.youtube.com/watch?v=qIi4_Poo2HY

https://www.aria.developpement-durable.gouv.fr/wp-content/files_mf/FD_717_jonova_1989_ang.pdf

Are you sure? Please comment or give some links.

 

[spencer1515]

Yes, my initial thought was to house the sensors at the top of the pits as it should rise. Also, I have seen where water vapor in the air mixes and brings a plume of that mix lower to the ground such as shvet's videos; thus, I am confused once again. Most likely I will house sensors right in the middle of the two pits by the tanks and then one on the catwalk above. I tend to over think things, and I am probably doing that with this project, but I want to do it well and make sure it is done right. Thanks for the responses everyone!

 

[Compositepro]

The boiling point of ammonia is -28F so when liquid mixes with air, the mixture will be denser than ambient air for a while due to the very cold temperature. The clouds in the video are water vapor condensed from the air, so you are only seeing where the air has gotten cold enough to form a fog. Ammonia vapor is invisible.

 

[spencer1515]

Yes the invisible ammonia vapor is in those low level clouds for a bit, right? If this was to happen in the pit I have described it should be obvious not to go near it. Ammonia vapor itself will rise and go over my sensor housed at the top of the pit. That is the idea at least.

 

[Compositepro]

Yes, it is certainly in the clouds but it is also rising above the clouds and cannot be seen.

 

[EdStainless]

We had a similar application (albeit a different gas). We set a caution trip very low, but wrote the logic to require those low level trips over a long time period before raising the alarm. We did find some very minor transient leaks this way.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[spencer1515]

That's a good idea. I am new to all of this, but when I do hear back from these vendors I plan to tell them the dimensions of the pit that house the tanks, the tank's contents, level, and capacity, the two main functions of the system (continuous monitoring and leakage detection), and I'll mention what you, EdStainless, said as well. What other questions will they have for me? I want to be prepared for when I hear back. Thanks everyone.

 

[pierreick]

Hi,
You may consider to have additional sensor(s) , calibrated at an higher set point to ensure the intervention team will be wearing the appropriate PPE and use the correct breathing equipment ( SCBA vs cartridge mask) . To be discussed with your management team.
Pierre

 

[EdStainless]

Maybe O2 sensors in the pits. If they are four sided they are likely considered confined spaces, so lots of rules about entry.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[shvet]

Sorry for english, I have rare practice.

1/ First deal with specific density. Specific density of vapors of boiling NH3 is 17*273/(273-33)*29=0,67. As you can see NH3 is lighter than air and temperature of air&NH3 has a little effect. It means NH3 vapors shall move upward and accumulate under roof. No doubts.

2/ NH3 vapors are invisible (areas with high concentration of NH3 were slightly visible as a change of refraction). I recall a couple of incidents with leakages of gaseous NH3 at ~ ambient temperature I participated. First was a ground release and ground area downwind was exposed to plume. Second was an aboveground release (~4m) and a little ground area downwind was exposed to plume. In both cases we had no means to find out a shape of plume but we had no reasons to do so as only exposed zone was important. It means NH3 vapors at least can move downward or slightly downward.

3/ Vapors of boiling liquid NH3 are visible. I recall an incident with spillage of little amount of NH3 (<1m3) I participated. Vapors of a boiling spill were visible and invisible plume exposed rather large ground area downwind. In all cases above a sense of smell was our detector. The same I recall story from colleagues about release of several cubic meters of liquid NH3 when clearly visible front of cloud had propagated rapidly at several hundreds meters (as on 2nd video above) and further a huge area had been exposed to a toxic cloud.

4/ A chemical plant I had been working on produced many chemicals including those extremely toxic. But I recall risk assessment documents where only toxic release of NH3 was shown (so called "toxic wave"). As I recall that "toxic wave" had several kilometers radius and therefore had the largest area affected among other possible toxic releases.

5/ You can find that at Jonave 1989 toxic cloud exposed area ~400 km2 and was observed as far as 35 km away from source (see a link above). Would it be so large if vapors of NH3 behaves as an ordinary light gas?

The only thought I tried to write down is that as per my experience NH3 dispersion is not a routine case. May anyone share a worth links to a study of behaviour/modelling of NH3 leakages/releases?

 

[Compositepro]

https://www.tandfonline.com/doi/pdf/10.1080/00022470.1982.10465371

The following deductions may be made.(a) If air, wet or dry, is mixed with pure ammonia vapor, the mixture is always buoyant.(b) There is a critical value for the initial liquid fraction F,between 4% and 8% of the total airborne mass of ammonia for air at 20° C such that if F is less than this value then the cloud will remain less dense than the sur-rounding air, regardless of the humidity.(c) There is a second value of F, between 16% and 20%, such that, if the fraction exceeds this value, the cloud will become denser-than-air at low dilution and will remain so throughout the subsequent dilution process regardless of the humidity of the air.(d) Between these two critical values of F the density relative to air will be determined by the humidity of the air as well as by the dilution, lower values of humidity being associated with higher cloud densities.

https://www.google.com/search?client=firefox-b-1-d&q=ammonia+spill+dispersion+pattern

 

[iconprocon]

Hello, I know about some liquid sensors but I think Levelpro LDS-YN Industrial Series is the best for your job. You will find their product here: [URL unfurl="true"]https://iconprocon.com/product/ld-leak-detection-sensor/[/url]