Acid Gas Injection Pipeline 2

[AggieCHEN04]

I have a mixture of acid gas that is (by mass) 73% CO2, 26% H2S, and about 0.2% H2O with trace hydrocarbons. The acid gas is compressed to 1100 psig and then flows to the injection well through a buried carbon steel pipeline.

My concern is that of an aqueous phase may condense one the gas hits the colder pipeline and cause localized corrosion. Should I be worried? Are there any good epoxy coatings for this type of application? I had originally planned on dehydrating the gas, but I have run into some budgetary and time constraints.

 

[elguero]

What is the gas temperature? How much can the composition vary? You say the pipe is buried. What is it buried in, and how deep? There's a difference between being buried in Brazilian swamp and Alaska permafrost.

 

[SJones]

If condensation is proven to be an issue, it will be unlikely that an epoxy coating will be of much help as corrosion could be rapid at pinholes. What is the flow regime? The following may be of use:

http://www.nace.org/nacestore/assets/ConferencePapers/1996/96026.pdf

Steve Jones
Materials & Corrosion Engineer

http://www.pdo.co.om/pdo/

 

[AggieCHEN04]

What is the gas temperature?

The gas is assumed to come out of the air cooler at 120F after the final stage of compression. Actual operating conditions will probably vary somewhat. According to my simulations the acid gas would actually be fully condensed at about 100F at the design pressure. At slightly lower pressure and aqueous phase would form between 120 and 100.

How much can the composition vary?

quite a bit actually. Any where from 90% CO2 to the design conditions.

You say the pipe is buried. What is it buried in, and how deep? There's a difference between being buried in Brazilian swamp and Alaska permafrost.

East Texas probably a couple of feet deep.

What is the flow regime?

Fully turbulent. Somewhere around 300,000 to 500,000 Reynolds number if liquid and 1.3 million if gas.

 

[JLSeagull]

I would be worried. On an offshore project we had acid gas that was lower CO2 and higher H2S. We diluted the acid gas with fuel gas to keep the H2S below about 18%. I suspect that the metalurgist worked with the chemical engineers to determine the maximum H2S level for use with carbon steel. The gas was used for reinjection and also could be further processed onshore.

 

[AggieCHEN04]

Hmm....I just realized I posted this in the wrong forum.

 

[elguero]

I sent a message about three days ago, but do not see it. Here goes again.

Even though water is present as a small percentage to be mostly in the form of a liquid -- droplets and surface film, probably. Carbon steel will be attacked by the acids formed by gases disolved in the water.

Carbon steel is not acceptable. Don't put any faith in epoxy coated pipe.

Use a nonferous alloy. A nickel alloy might be best, but I can't say for sure. You must investigate. Buena suerte.

 

[maddocks]

I believe that carbon steel is acceptable as long as you verify that no aqueous phase will form within the pipeline. The current understanding is that you require an aqueous liquid phase for the corrosion mechanism to start to work.

 

[AggieCHEN04]

It's kind of difficult to charactarize the acid gas as an aquious or hydrocarbon phase when it's initially condensing since the water condenses first and the acid gas and hydrocarbons condense within a narrow temperature range. Well, so far it's up and running for a few weeks and hasn't sprung any leaks yet.

 

[SJones]

When is the first inspection planned and by what method?

Steve Jones
Materials & Corrosion Engineer

http://www.pdo.co.om/pdo/

 

[AggieCHEN04]

I work for the engineering firm. The client will determine the scheduling for inspection.

 

[elguero]

I repeat what I stated above. According to the given conditions, water will be present as a liquid. This will lead to corrosion. Don't ever install anything when you know it will fail.

That is the whole point of engineering, isn't it. Do the calculations; then if a bad condition is revealed by the calculations, change the design.

Read Engineering Design for Process Facilities, McGraw-Hill (1993).

 

[AggieCHEN04]

It isn't certain that the material will fail. After doing some research I have found similar acid gas injection systems that use carbon steel pipelines:

http://www.tartaneng.com/download/Acid Gas Injection in Western Canada.pdf

..particularly the ones that inject acid gas with formation water.


I have also read a few papers on the subject. There is no real consensus on the appropriate material for pipeline construction. Generally, if the acid gas is compressed to the dense phase or has a water content of less than 0.2%, carbon steel is acceptable since there isn't sufficient free water to form an aqueous phase. In this case it is borderline, but probably will not form. Currently we are looking into chilling the acid gas into the dense phase to save horsepower going into the well, so that should remove the potential for corrosion as well.

 

[elguero]

Buy a quart bottle of soda water at your local market. Unscrew the cap. Put in a half dozen steel nails. Put the cap back on, to keep the CO2 from escaping. Check once a week for two months. You will notice a dark brown sediment of fine powder after a while. After two months, check the nails. You will note deep pits in the nails. After a year or so, the nails will start falling apart. To spur things along more rapidly, add a pinch of table salt.

 

[maddocks]

This reply is simply not even relevant to the discussion. We're talking about undersaturated acid gas without an aqueous phase. Industry experts have discussed this extensively and the general concensus is that a corrosion mechanism will not be active without an aqueous phase. If anybody has data to show otherwise, fill us in. There are a number of acid gas injection schemes using carbon steel piping - we're unaware of any that have experienced high corrosion rates when there is no liquid water phase present.

 

[SJones]

In the initial posting, the originator expresses their concern regarding the possibility of water condensation. None of the above replies have alleviated that concern. The originator does not allude to any provision for corrosion monitoring or chemical mitigation facilities. The apparent approach is: 'leave it to the operator to sort out'. I don't know if it's just me, but the engineering aproach, as described above, is a little disconcerting.

Steve Jones
Materials & Corrosion Engineer

http://www.pdo.co.om/pdo/

 

[AggieCHEN04]

The apparent approach is: 'leave it to the operator to sort out'

The statement that is paraphrased here is specifically in response to the operators mechanical integrity program. The operating compnany is free to go to whomever they choose to perform radiography, ultrasound or whatever they want to after the facility is operational. We are only able to make sugestions for the future operation and mantanence of the facility. These issues have been addressed, but I don't feel comfortable explaining every detail of said project.

 

[maddocks]

My calc's show that the 0.20% water is equivalent to about 95 lb/MMSCF of water content and that you will be undersaturated to roughly 16 Deg F. I'll check this a bit more just from curiosity and run it past our hydrate guru (John Carroll) to get his thoughts in this as well. I still don't see a problem with your approach. I would ensure that exposed parts of the pipeline (ie, headers, risers, etc.) are traced and insulated to prevent them from getting colder than 20-25 DegF in wintertime. Not sure where you're at on your project, but you may want to consider select chem as well as a weld procedure that will give you a low residual hardness.

Jim

 

[elguero]

Water content in the gas is higher than its dew point. It will condense. Here is the data:



psia psia
lbs lb-moles mole-fr Pp degF Pv
______ _________ ________ _______ _______ ________
CO2 | 73 | 1.659 | .682 | 760 || 100 | 0.943 |
H2S | 26 | 0.763 | .314 | 350 || 110 | 1.28 |
CO2 | 0.2 | 0.0111 | .0046 | 5.13 || 120 | 1.69 |
Sum = 2.4331

degF Pv
_____ _____
100 | .943
110 | 1.28
120 | 1.69



Stated water content is 0.2 lbs. This is a mole fraction of 0.0046. Multiply this times 1115 psia to obtain partial pressure of 5.13 psia. This pressure is compared to vapor pressure, Pv, (the last column of the table) at 100, 110, and 120 degF. For example at 120 degF the vapor pressure is 1.69 psia. Therefore, water will condense until its concentration is: mole-fr = .0046*(1.69/5.13) = .00152, or a gas composition of 73% CO2, 26% H2S, .066% H2O.

 

[Bala0610]

Hi

From the originator's response on conditions prevailing, it is clear that the gas is compressed, cooled and sent for re-injection. He does mention cooler performance. Under the circumstances, it would be logical to assume the gas to be water saturated (is the KOD having water after the final cooling?) and any further cooling can and will drop out water. Rest follows...

We have an application where we have a separate dewpointing unit to dewpoint the gas - and the gas is heated up to near ambient to give a 15 deg C superheat or so. We dont expect any cooling (ambient conditions, length of pipe etc).

I would be very concerned also (especially if the length of pipeline is large) about upset / startup/shutdown conditions.

I can suggest one thing though...leave some of the superheat in the compressed gas to avoid condensation...if acceptable. Or better still, use some of the compression heat (in another exchanger) to superheat the final gas after the cooling / knock out. The degree of superheat will have to be decided based on the ground conditions (ambient, length of pipe etc). These methods should also be used along with something else to address the upset / startup / shut down conditions.

Hope this helps.