7 replies to this topic

[Zauberberg]

Extremely rich acid gas from sweetening unit, recovered at 1.5bar abs, has to be recompressed in a multistage compressor system up to 100 bar before being exported outside of the plant battery limits. The stream is saturated with water and, as such, creates highly corrosive environment in downstream high pressure equipment. In order to aleviate this concern, dehydration of acid gas and reduction of water content down to 40mg/Sm3 is required.

According to my knowledge, TEG units are feasible within pressure range of 30-100bar which means that at least 3 stages of acid gas compression would require special and extremely expensive corrosion resistant alloys before dehydration could be performed.

Are there any recognized and proven technologies for dehydration of low-pressure acid gas (more than 50% H2S and CO2), and what are the possible drawbacks of these applications? The best solution would be if gas dehydration can be executed below 3 or 6 bar abs.

Thanks in advance,

[Art Montemayor]

Zauber:

No problema. I can dry the gas - down to ppm if need be, and the economics are available. I don't know the gas flow rate - and consequently the size of the facilities required in either option, but if the economics are in place, and the need for the gas is there, then I would look at using a pre-chiller to knock out the bulk of the water at around 3 to 5 ^oC. I would use a mechanical refrigeration cycle and heat exchangers. I would follow the bulk water removal with an adsorption unit (activated alumina) if the flow quantities are relatively large. If the flow rate is small, calcium chloride dessication can be applied. But I would rather look at a fixed bed adsorption unit - maybe 3 towers.

The product from an adsorption unit working at 90 to 100 bars could be in the ppms of residual water left in the product gas - probably negligible. Using adsorption below 30 bar pressure would have to be looked at from an economic return analysis. It all depends on how much money you can afford and what return on your investment you demand. The technology is no problem. It exists and is effective and efficient. The factors you haven't divulged are what will determine whether it is done or not: the project incentives and the minimum financial return acceptable.

[JoeWong]

May consider intermediate dehydration by adsorption to avoid mechanical refrigerant...

"Pumping" acid gas to 100 bar still remain a big challenge to process and mechanical... Please involve compressor vendor in early stage...

[Zauberberg]

Thanks for the replies.

By cooling the gas down to 3 or 5C, water content is reduced from 150,000 mg/m3 to 6,000 mg/m3. Mr Montemayor, why did you set refrigerated gas temperature to be 3 or 5C - are there any particular reasons for these values (activated alumina capacity, etc.)? As you can see from the chart below, water content gradually decreases as the temperature is reduced. I don't know if additional cost for refrigeration unit can overweight the cost of corrosion resistant alloy for acid gas compressor and auxiliary equipment?

I have turboexpander in this plant, with outlet temperature of 10C. My first idea was to use LTS vapor for cooling the acid gas down to 15-20C. This way, substantial amount of water would be knocked-out but without employing additional refrigeration unit. Turboexpander is already there, and it would be nice to utilize this low-temperature level for cooling the acid gas.

Can you tell me something more about activated alumina adsorption units, and where can I find references about these packages - inputs like minimum operating pressure, water content reduction, pressure drop of the system, etc.?

I received one very interesting answer from gas processing expert, saying that gas drying is not required in this case - since the gas will be undersaturated with water after compression and cooling. After 3rd stage compression, gas will be well undersaturated and there will be no possibility for water to condense - and this fact eliminates all corrosion concerns. I am quoting his post:

"First off, there is no need to dry the gas unless you just find it fun to do.

As the pressure approaches 1000 psi, the acid gas becomes hydroscopic and is undersaturated, therefore it is a noncorrosive mixture. In a paper presented at the GPA in 2007, an operating company spells out their whole design and compliance with NACE MR0172.

Quickly here's what you do. The compression cylinders are all carbon steel with ss304 valves and trim. The suction piping is all ss304. The discharge piping can be CS, but the interstage coolers will need to be SS304. After the last stage, all the piping can be CS with some exceptions that depend on the exact composition."

I already had these calculations with me, and the results are that water content is reduced from 150,000 mg/m3 (1st stage suction) to 5,000 mg/m3 (4th stage suction) - if I cool the discharge of each stage from 170C to 50C. But, the suction of each stage is saturated with water. Can you tell me what is your opinion about this very interesting issue?

Acid gas flowrate is 2.5MM Sm3/day, with more than 50% CO2. The rest is H2S and equilibrium water, with negligible amounts of hydrocarbons (C1-C4).

Greetings,


[attachment=809:WC.JPG]

[Art Montemayor]

Zauber:

I always set my minimum temperature on chillers at 5 to 8 ^oF above the freezing point of water (32 ^oF = 0 ^oC) in order to positively avoid any possibility of freezing out the water inside the chiller. The purpose and scope of the chiller is to maximize the physical condensation (& removal) of water vapor. As such, then, we want to come - as close as we practically can -to the freezing point of water. I have found that 5 ^oF is as close as I dare to come to the freezing point due to process variations, instrumentation limitations, and human accuracy. To allow the water to freeze to ice within the chiller often presents a major hazardous event. Of course, as you drop out more water before going into the adsorption unit you are also decreasing the amount of adsorbent required for ultimate water removal.

In your original post you wrote: “Extremely rich acid gas from sweetening unit, recovered at 1.5bar abs, has to be recompressed in a multistage compressor system up to 100 bar before being exported outside of the plant battery limits. The stream is saturated with water and, as such, creates highly corrosive environment in downstream high pressure equipment. In order to alleviate this concern, dehydration of acid gas and reduction of water content down to 40mg/Sm3 is required. You further added, “Are there any recognized and proven technologies for dehydration of low-pressure acid gas (more than 50% H₂S and CO₂), and what are the possible drawbacks of these applications?” My response was that I recommended you drop out the major portion of water with a pre-chiller, followed up by an adsorption unit. I still abide by that advice. You have now started to give us more basic data and a beginning of a scope of work. That’s good, but I didn’t know that before. If I knew of the availability of local refrigeration sources, I would have taken those into consideration. If you truly have a local refrigeration source and it doesn’t jeopardize the proper operation of both the source and the acid gas conditions, then by all means I would recommend you capitalize on the refrigeration source. But you haven’t told us what is the original temperature of the acid gas, so I don’t even have an idea of the relative attractiveness of pre-cooling the acid gas to only 20 ^oC. If this can significantly reduce the amount of water, then by all means, do it. You can always install a smaller mechanical refrigeration unit downstream that will essentially remove all liquid water at 38 ^oF. The adsorption unit is left to remove a minimum of water.

Gas adsorption units are well established in the industry and have been optimized for the past 50 years. A lot of our present hydrocarbon and petrochemical processing would not be possible today were it not for successful adsorption technology being used.

I don’t know who the “gas processing expert” is that you consulted with, but saying that gas drying is not required in this case is foolhardy – at best - since we don’t know what the total process is that you have in mind nor have you revealed your scope of work. The gas will be NOT BE “undersaturated with water after compression and cooling”. It will be TOTALL SATURATED AT THE FINAL COMPRESSED CONDITIONS. That is an engineering fact. I still don’t know your process and whether you want to compress humid gas or not. You did not state that as your problem. You only stated “The stream is saturated with water and, as such, creates highly corrosive environment in downstream high pressure equipment”.

My recommendation was to remove the majority of the water by cooling the compressed gas – not the un-compressed gas. “As you have stated, “the suction of each stage is saturated with water

Essentially, I agree with your “experts'” comment that normal compression will knock out the majority of the water in the gas – right after each of the intercoolers and the aftercooler. However, I don’t know what your conditions are or what your scope involves. If you are piping the compressor discharge with an ambient condition that is below the condensation temperature of the water content in the gas, then you are going to have problems. This is often one of the problems in pipelines that run between offshore platforms and land facilities. Hydrates can form within the line due to the cold sea water around the pipeline. So, in effect, it is all in your hands since the total scope of work and the basic data is all in your head.

[Zauberberg]

I think I understand why this guy is saying no dehydration is required for high-pressure acid gas injection. Based on my calculations for the given fluid composition, cricondenbar pressure of 3rd stage discharge gas (after it is cooled and part of water has been knocked-out as liquid) is about 76bar abs (4th stage discharge pressure is already above 80 bar abs). It means no condensation can take place, regardless of the fluid temperature. And I think this is the clue for high pressure injection.

However, there is a potential risk - in my opinion - during equipment blowdowns. Sudden pressure reduction can precipitate liquid water inside the system. Because of this fact, I am not sure if I want to play further with no-dehydration-concept.

I agree with your statements and conclusions - except for one: you said that fluid is saturated at all compression conditions. How can this be the case?

The option for solid adsorbent dehydration is the most likely, as you recommended. Based on feedback from different resources, employing TEG units for highly concentrated acid gas has always been considered as design and operational challenge - especially if complete Gas Plant needs to be designed for 99.9% availability. Can you tell me something more about minimum operating pressure of solid adsorbent units (activated alumina or molecular sieves), and what is the corresponding pressure drop, in general? Or should I approach process licensors directly for this information?

I am looking for saving materials of construction for the compressor circuit, but - obviously - if I dehydrate the gas at low pressure my drying unit is going to be much bigger in size. There should be an optimum somewhere in between.

Thank you for your time and quality feedback.
Best regards,

[Art Montemayor]

Zauber:

When you compress a low pressure gas that is saturated with water, the product will always result in a high pressure mixture that is ultimately saturated with water the moment you cool it and drop out the condensed, resultant liquid water in the knock out traps after each interstage cooler. What I am describing is, of course what normally happens in a reciprocating compressor. If you have intercooling between centrifugal stages of compression the same effect is reached. The condensed water is in equilibrium with the gas mixture from which it was condensed. That defines the gas mixture as 100% saturated – but at the higher pressure conditions under which it finds itself. Under these higher pressure conditions, the gas mixture obviously holds much less water vapor. And this is the “squeezing effect” that your expert is trying to take advantage of. In other words, as you compress your gas mixture to higher and higher pressures (with interstage cooling & condensate separation) the gas contains less and less water vapor at the higher pressures. But it is still SATURATED with water vapor (it can’t hold any more water at the conditions it finds itself under). I have taken advantage of this“squeezing effect” to liquefy air in air separation plants, liquefy CO₂, liquefy Nitrous Oxide, Carbon Monoxide, Oxygen, and Hydrogen. And I’ve done adsorption drying down to ppms with all gases. In the case of CO₂, and Nitrous Oxide I have employed the adsorbers usually at the 2nd or 3rd stage of compression. The reason for this is that these gases liquefy after the final compression since their critical temperature is relatively high (88 ^oF for CO₂ and 212 for H₂S, I think). I believe the same would apply to your mixture, is that right? I would suspect so.

The best and most intelligent way to attack an adsorption problem such as this one, in my opinion, is to deal directly with a pre-selected supplier of design and unit fabrication. There are many engineering companies in Europe and the USA that can easily design, build, and guarantee the expected results. They may even sell the process design package for fabrication elsewhere. The reason I say this is that the design selective sorptivity is the key to the entire package and an experienced fabricator has the information already developed (& perhaps proven). There should be no “experimenting” with an adsorption unit in today’s state-of-the-art. The next thing to preselect would be the type of adsorbent. I would select Activated Alumina as my adsorbent of choice – unless the experienced fabricator objected. I don’t think that will be the case in this application. Molecular Sieves are, in my opinion, an overkill for this application. They are also very expensive.

I have designed and fabricated my own adsorbers in several countries using only local fabrication and imported valves and electric heaters. These units worked under a rigorous 100% on-stream time, 24-hours a day, 7 days a week. And the product gas was always below the design dewpoint. That was years ago, so I know that you could easily carry out a gas dehydration project to success with a present-day adsorption unit. I believe the key is to locate and identify a credible, experienced and capable designer-supplier.

The minimum, optimum operating pressure for your application depends on what your incentives are and the costs attached to the various factors. An economic study may be required. It all depends on how your conditions are vested and your scope of work. I think I’ve mentioned this before. As a general rule, I believe you will find that your lowest credible pressure for drying would be 100 psig. If you dry your gas in between compression stages, then this might mean you dry at 200 – 300 psig. This would be a conventional platform for an adsorption unit. This is what I’m accustomed to.

A typical pressure drop that I’ve always designed for (& always stayed conservatively below) is 10 to 15 psig. I’ve operated CO₂ adsorbers with only 7 psig pressure drop across the system – and yielding 1 ppm vol. of water moisture (approximately -100 ^oF dewpoint).

The only problem I ever had with my adsorbers (in the 1960 -70’s) was the quality of the isolating valves. I finally found the proper type and make of valve that I later found out was also discovered by a lot of major international companies – Orbit valves. These valves successfully operated in the zone of great temperature and pressure differences across the valve plug – without leaking. This is a critical factor for an adsorption unit and I believe with today’s technology it no longer is a problem.

I hope this experience is of some help in getting you into a good and comfortable process design position.

[Zauberberg]

Thank you Mr Montemayor, this is exactly the information I needed.
Have a nice evening in Texas.


Regards,