speed up the start-up of airsepartion plant 4

[cryotechnic]

To reduce the costs and speed-up the start-up of one of our airseparation plants we are calculating if it's worth to pump in liquid once the plant is starting to make liquid.
Normally it takes us about 2 and a half day to produce enough liquid to get the first pure products.
Our idea is to start pumping liquid once the temperature of the installation is low enough (liquid temperature)
We are not sure yet in pumping it in de pressure collumn or in the lowpressure collumn.
The lowpressure collumn is a structured packed type column.

Does anyone have expirience with this?
Did anyone face problems with this?
What kind of alarms, protection are used for this?

Looking forward to answers and idea's

 

[wilmers]

We have experience in start-up an ASU plant daily, and acoording to our plant (produce LOX), the best is pumping LOX in the bottom of the low pressure column.
In the botton of the high pressure column we have a impure LOX, then is better pumping a pure product in the low pressure column, where you have pure LOX.

Maybe the only thing you have to be careful, is that the percent of oxygen in the column rise quickly, and if you use a waste stream to regenerate adsorbers you have to be very careful.

Wilmer

 

[cryotechnic]

Thanks Wilmers,
What I was thinking after reading your post. Do you pump in the LOX out of a tank or is it delivered with a truck? Do you have a special connection to pump in liquid?
In the case of our ASU, we don't have a tank in the vicinity. (plant is used for GOX and GAN production only, by use of internal compression)

In our case we have to make a connection for trucks to pump in the liquid. We are thinking about a pressure controll valve in the pipeline to control the pressure in the collumn during pumping.

you said, you start up the plant daily.... Is the plant switched off every day? No production during the night? No continues production?

Cryotechnic

 

[dcasto]

We have a nitrogen rejection unit at a gas plant. The speed up the process of starting up we bring in liquid nitrogen and charge the relux system with it and liquid methane the reboiler and refrigeration system. We went from a 6 day startup to a 1 day.

 

[wilmers]

ok cryotechnic:
Well, the way we inject the LOX in the column, is the same to pumping, we have a high pressure tank of LOX (about 200 psig)(that is the same if you have a high pressure truck). Then, when we need to inject LOX just we transfer LOX from the tank to the column, of course the pressure of the tank will decrease until 140-150 psi (depending the transfering ratio, I think the same will happen when you pump the LOX).

Definitely you have to make the necesary conections. And I believe is not necessary to have a control valve (you can regulate the pressure with a manual valve beetwen the truck an the column).

Yes, we have our plant stopped por 5 -hours period, then we start it again, the reasons are just economics. We save energy in this period (High costs of electricity in that period).

Wilmers

 

[NTCIINC]

Dear Cryotechnic,

Three things about speeding up the plant cooldown:

1. Start your plant at 75% air and product flow capacities. Do not try to start at 100% capacity or your plant will act as a yoyo.

2. Raise you upper column pressure (carefull with the safety settings) so that when liquid comes in contact with the condenser, your compressor will not surge. As you build a liquid level around the condenser, slowly reduce the upper column pressure towards its normal operating design. Watch closely your condenser level and also your air flow as it increases.

3. I am from the liquid nitrogen injection school. I do admire those who inject liquid oxygen but I find it costly and dangerous. LN2 is much cheaper than LOX, it is also much colder (LOX is around -280F, LIN is around -320 F at 15 Psia). The problem with LOX injection is that you may find oxygen everywhere in your cold box and out to the waste stream. This presents a risk if you consciously grease your revex valves every week or if you are in the heating cycle on your adsorbers. I expenrienced first hand how fast the temperature climbs in the adsorbers with 30% O2 in the waste. Luckily we caught it on time, no fire.

I agree that the best injection point is the sump of the upper column (The LP side of the condenser). After all, it is the condenser that makes the liquid...

Last word of wisdom: Liquid inject for BTU's, not for level. Inject a bit and let the plant do the rest. Inject a bit more etc. Do not try to build a level in your condenser by liquid injection, otherwise you risk freezing your turbine.

Hope this helps.

Nils

 

[cryotechnic]

thank you for your answer Nils, it was helpful.
Normally during the start-up our compressor is controlled by it's own pressure controller and his protonnic. So during this fase the bypass valve is openend and will be closed by the protonnic.
So once the liquid starts to build up and the pressure in the collumn sinks, the compressor is giving more air.

I don't get what you mean by: "Do not try to build a level in your condenser by liquid injection, otherwise you risk freezing your turbine"
Do mean that the whole column will be to cold wich results in a to low inlet temperature of the turbine, which causes liquid in the turbine? or what?

I was also thinking about what you've said in part 3. about the oxygen in the wastestream.
If you started with pumping in LOX, the whole column has to be allready on liquid temp.. So there will be allready liquid flowing downwards the column. I quess, once the GOX is at the wastestream outlet, it's washed out and will be impure.
Of course if you start pumping in LOX while there is not enough reflux liquid going down the column, the change of a high oxygen analyse in the wastestream will be greater.
I'm I right in this thought? or do I mis something?

Like you said, our reasong of pumping liquid is for BTU's. We want to try to speed up the start up by some hours. With a 20MW aircompressor, that's a lot of money to safe...
This plant is one with falling film evaporators and internal compression.
In an other plant, with a conventionel bath condensor, we have the possibilty to pump in LIN, that works well.

Cryotechnic

 

[NTCIINC]

Dear Cryotechnic,

Answering to your questions:

1. What was meant by "Do not try to build a level in your condenser by liquid injection, otherwise you risk freezing your turbine" is what you interpreted: if too much liquid nitrogen is injected at once, the heat balance is shifted in the main exchangers, resulting in liquid at the turbine inlet. In other words, when liquid assisting the plant at start-up, keep an eye on the exchanger cold-end temperatures.

2. "oxygen in the wastestream". I agree with you, especially with falling film vaporizers. The former arguments were meant for a liquid bath vaporizer, I share Wilmers' concern.

With a falling film vaporizer you may want to start early with LOX injection in order to maintain a minimum recirculating flow and avoid dry vaporization.

Correction on my previous posting: the best place to inject LN2 is at the top of the LP column.

Have you tried controlling the main air compressor from flow rather than pressure?

I also had a situation where my ASU was taking for ever to cooldown and make liquid on a warm start. After opening the purge at the sump of the MP column (lower column), we got rid of a "plug" of warm gas and liquid started building in the main condenser relatively fast. At that point we closed the purge.

I hope this helps.



NTCI, Inc.

http://www.ntciinc.com

 

[Carmela]

RE dcasto's mail

Didn't AP have a patent on this feature ?

See US6214258.

 

[asukid99]

LOX injection is typically avoided on new plants due to hydrocarbon safety issues. Typically LIN is injected at the top of the LP column so that it falls down into the sump of the main condenser (assuming stacked columns). As far as exactly where to connect the LIN injection that is difficult to determine not knowing your process. Most likely, you will probably not be able to justify draining perlite to install piping in your cold box so instead look for a manual liquid drain going to your snow blower. You can probably install some valving and tee into that line and run the tee to where a truck can fit his hose to it. You will probably need to use CGA pipe connections to match his hose. Don't forget to install thermal reliefs between valves. Piping should be 304 ANSI 150 and possibly insulated for personnel protection.

 

[cryotechnic]

Thanks asukid99,
We have skipped the idea of LOX allready, for the same reasons you are saying, Hydrocarbons.
This ASU is not one of a stacked type, 2 separate collumns with fallingfilm condensors.
Draining Perlite etc, that's sofar not an option, it will make the costs of the project too high, it's not worth anymore then.
Using a liquid drain pipe could be an idea...

Greetings,
Cryotechnic.

"Math is the ruler of your potential succes...."