co2 gas compression and dehydration

[ratash]

My gas stream is 98% co2, it has 0.46% c1, 1.1% c2, 0.06% h2s, 0.049% argon, 0.063% o2, 0.46% n2. It is the flue gas from amine stripper. I need to compress it to 2000 psig, the flowarte is about 1.8 MMscfd, and transfer it into a pipeline.
What method of dehydration do I need?
What type of compressor?
Could you please outline everything that needs to go into this package?
What critical issues need to be considered?

Thanks

 

[dcasto]

to get a more exact set up, we'd have to know your contract terms too. What is the specifications on the CO2 into the pipeline.

Generally, I'd use a 5 stage reciprocating compressor with a TEG dehydrator after the fourth stage.

Read the contracts it'll tell you where you will have problems.

 

[ratash]

Thanks.

Why not 4 stages?
Why TEG and not mole sieve dehydration?

This is very preliminary no sales contract, because 2 people sat at breakfast and came up with this idea. They have no knowldge of what will be involved. This is new to me too.

I have simulated it in HYSYS, by saturating it with water, then 4 stages of (recipe)compression (75% efficiency) with final delivery at 2000 psig and 120 F. Inlet gas is at 1.5 psig and 70 F. I knocked out most of the water but still have water going to the pipeline, which means I can have corrosion problems.

I need to know what are the main factors, which need to be considered in designing this system.

 

[dcasto]

At 4 stages you will exceed normal maximum operating temperaures for most most compressors.

The only reason to use mole sieve is if you need to have water in the 0 to 4 ppm range. Most Co2 sales contracts allow the water to be 7 pounds/mmscfd. The reason for the 7 lb/mmscfd, bceause thats a resonable level of water from a TEG unit.

You will not have corrosion problems if you do not remove the water in a 2000 psig line because the CO2 is above its criticle point and it is hydroscopic and will carry twice as much water in the vapor phase. water in the vapor phase is not corrosive.

 

[ratash]

You are absolutely correct, the compressor discharge temp is over 325 F with 4 stages. Thanks.

I saturated my co2 flue gas from the stripper and ended up with 224 lb/MMscf of water at 2000 psig and 120 F. As you mentioned it is all in the gas phase. So, does it mean after the last stage it can go to a C.S. pipe as long as it stays in critical region? I have a feeling this will go underground and it could get to 60 F and start condensing. In that case is the 7 lb/MMscf ok, when it comes to corrosion?

Is Natco the one for TEG DEH units?

Thanks for your help!

 

[zdas04]

I just did an analysis on a similar project and I have a couple of questions:
- What is the final disposition of the CO2?
- Does it really have to be dried for a process reasons or corrosion reasons?
- Have you looked at doing the first two stages with flooded screws instead of recips? The shape of the constant entropy lines on your mollier diagram really favor using screws to get to about 70 psig.
- How sure are you of your water number? At 2000 psig and 120F I get a number much closer to 20 lbm/MMCf than 224 lbm/MMCF
- Are you going to be handling this stuff as an SCF or a liquid?

If you're just trying to get rid of it then it makes more sense to use corrosion resistant pipe and don't dehydrate at all. It sounds to me like you are a long ways from the point where the TEG vs. Mole Seive decision is critical.

David

 

[maddocks]

Like dcasto, I would look at a glycol based system, however you may wish to consider the use of glycerol to reduce glycol losses. As well, you should look at which stage do dehydrate - after stage 3 or 4 is typical but look at the economics of tower size vs glycol rate.

I wouldn't be overly concerned with actual water content, instead, you should look at hydrate temperature, and how cold the line could possibly get during winter conditions.

regards,

Jim

 

[dcasto]

You stated the delivery pressure is 2000 psig, but what is the pressure at the end of the line, then you can look at dew point along the line...

A single recip is just fine. A srewcip sounds nice but at this low volume, not needed.

 

[ratash]

Dear maddocks and dcasto,

You are the best! Thanks.

 

[dcasto]

do we see stars in our future?

 

[maddocks]

I feel like a kid in grade 2 asking for a pink star....

 

[dcasto]

he's a newbie and I probabilly have met him before...