Scrubbing Vacuum Tower Vent Gas/Safety Concerns of Air Intrusion

[travitch]

Like many refiners, we currently vent the non condensibles from the vacuum tower eductor train to a fired heater. My understanding is that the EPA will be mandating in the future that we scrub this of H2S prior to burning it. Therefore, we need some motive pressure to accomplish this. I had initially thought to slip this stream into the FCC Wet Gas Compressor to take advantage of the existing amine scrubbing system. The problem that I see is that should we develop a leak of sufficient size on the vacuum system we would be sending enough air down the line to be in the explosive regime even though we would know it quickly due to loss of vacuum on the tower. I am leaning to an O2 monitor on the line, response time of the analyzer for line sizing, and a shutoff/diversion valve to address this problem. I am wondering what other solutions have been employed as I suspect others may have already addressed this problem.

 

[gelpino]

It's also possible to think about a liquid ejector in wich the liquid can be for example the amine. I know some firm that product this kind of ejectors in different sizes and with different feed liquids.

Best regards

 

[pmureiko]

An oxygen detector will not serve as a leak detector on your tower. Composition is independent of flow. 100 lb/hr and 1000 lb/hr of air leakage will both have a composition of 21% O2 (neglecting equilibrium species). Composition is a function of temperature, pressure, and number of components.

You could use the O2 detector on the amine system to check that it is operating below the maximum safe oxygen content. Even then absorbing H2S in the scrubber will change the composition and you need to check if the gas would enter the flammable region.

 

[lastone]

I think you are confused pmureiko. An O2 analyzer in the line will be measuring total O2%. If the total flow of vacuum tower vent gas is 1000 lb/hr and the air leak is 100 lb/hr, then the %O2 would not be 21%. It would be something more like 4-5%. An O2 analyzer would, and does serve as leak detectors on vacuum systems all over the place. Although you usually realize if you have a big vacuum leak in other ways first.

I think gelpino has a good idea. Something like a venturi scrubber with recirculated amine as the scrubbing medium. Or you just continue with your original idea, but you are still going to have to re-compress the vacuum tower vent gas to get it over to the wet-gas compressor. I am assuming you have steam jets on the vacuum tower. If you have vacuum pumps then maybe you could get into the wet gas compressor. You probably would want some type of reliable interlock in the system should you ever lose tower vacuum though.

 

[pmureiko]

Lastone you are confused. Oxygen analyzers measure compostion and not flow rates. To measure flow you need a flow meter. Oxygen analyzers are used to measure if a system is operating with a safe composition. Changes in the composition can be used to infer air leakage, but you must be careful because this is not always the case.

If the non-condensible is air, then EQUILIBRIUM hydrocarbons (HC) will be present. The vent composition is independent of air leakage rate. More air leaking into the system means that more HC is in vent gas. The composition stays constant. So an oxygen analyzer is unable to detect a change in leak rate.

You appear to be referring to a situation with the non-condensibles consisting of air and process gases. In this case composition varies with air leakage rate, because of dilution from the base load of process gas. It will be hard to infer changes in leak rates when the process gas load is more than 3 times air rate because there is so much dilution.

 

[lastone]

I don't think I am confused pmureiko. I do know that the effluent gas off of a refinery vacuum tower is primarily uncondensed hydrocarbon. The %O2 will be the amount of air that is leaking into the vacuum jets or the vacuum tower. If a larger leak develops the %O2 would increase. Every furnace in the world tries to control to a set percentage of O2 in the vent stack. It is never 21%, it is usually around 10%. If the excess air increases or decreases the %O2 changes accordingly. O2 analyzers are used on Reformer regenerations. You can tell the extent of the regen based on small changes in %O2. Maybe you are thinking that the gas coming off of a vacuum jet condenser in a Crude Vacuum Tower is all air, but this is definetely not the case. We all realize that if the gas was 100% air to begin with, an O2 analyzer would always show 21% O2, but this is not what travitch is asking about. Needless to say, I don't think we have helped travitch at all.

 

[travitch]

Thanks for the info everyone. I was hoping someone had actually done this and could share what they had specifically done.

I looked into the liquid eductor a while back and don't think I can get the discharge pressure I would need.

I am going to use the O2 analyzer and send it to the FCC wet gas compressor. Basically, I know the max gas rate by the last eductor stage. So %O2 from an analyser tells me what the O2 potential flow could be and knowing the wet gas compressor absorber off gas typical flow and composition then I can determine a max O2 upstream before LEL at the highest concentration of non condensibles downstream and divert the flow so I don't ruin the compressor or something else downstream, if you know what I mean.

I don't know but the stack gas on my fired heaters is about 3 to 5% O2. 10% seems high and costly. Though you need to check design info from the heater vendor.

 

[pmureiko]

Many applications have been listed in this thread for oxygen analyzers. In each case the reading is interpreted specifically for that instance and generalizations are misleading. Here I cautioned travitch of the false sense of safety obtained from misusing a device that is only capable of measuring composition. Unfortunately I have seen several instances of equipment damage where oxygen detectors have been misued, but fortunately no injuries or fatalities.

 

[gelpino]

to travitch
Try also to contact

http://www.hijet.com...it's

possible they have good news for you....but i'm not sure