moving compressed gas cyclically back and forth between 2 dist tanks 2

[jfgaltenco]

HI there,
This is my first post here but I have not found anything relevant to my problem.
I have to work on modernizing a process that involve moving potentially explosive gas.
My real life experience with compressor as of today is ZERO, so please bear with me. I know I have a lot to learn here.

I have 2 pressurized tanks same volume [1m3 each] separated by ~10000' [ not my choice this is a given I have no control on it].
Initially one tank is at 4000psi absolute the other at ~28 psi absolute. temperature is ambiant
The process I am working on requires the gas to be transferred back and forth continuously between the 2 tanks.
each transfer must be completed in ~2hour 2h30 max [ there is cost tradeoff involved in selecting exact duration of this transfer] [ so average volumetric flow rate is relatively low]

Constraints are
because of safety issues, suction pressure can NEVER fall below atmospheric.
because of the length and size [3/4"] of the line between the tanks, a relatively important friction induced pressure losses toward the end of each cycle exists.

I need an energy efficient compressor package that will perform this function [24/7 continuous operation]. it is not difficult to figure out how to do it in principle.

what I THINK I need is a package :
multi stage compressors 4 or 5 stages with interstage cooling, not sure at this point, each stage is individually short circuited by a check valve so that when inlet pressure exceed outlet's foreach individual stage, one does not compress gas that dont need to be compressed. [ energy efficiency is important here].
to be clear initially the gas flow naturally between the tanks until the 2 tanks are at the same pressure then one check valve close and only one stage is used.. when single stage capacity is not enough because of the decrease in the evacuated tank pressure.. second stage come online through the closing of the second check valve.. and so forth]

and then Variable capacity control is required to LOWER the inlet capacity toward the end of a transfer cycle to decrease pressure losses in the gas line between the evacuated tank and the compressor in order to achieve required evacuation while suction pressure remains above 1atm absolute. the decrease in capacity.
well I am pretty sure this problem must have been already adressed in this form or something close to it many time before me, but I am not able to find a vendor that has already done some packaging to adress this kind of problem or even a reference or a publication..

SO if somebody here could point me in the right direction this will be very much appreciated.

Thanks a lot in advance...

JF

 

[BigInch]

OK say Tank "A" has 28 psia, tank "B" has 4000 psia.
It's not clear at which point you want to reverse the process?

Provided the connecting line is large enough, presumedly you can equalize the two tanks to 2000 psia +/- within some appropriate portion of the time allowed, presumedly without the compressor, BUT you didn't exactly say what you want to do next. Right now tank "A" has 2000 psia, more or less, and tank "B" has the same +/- 2000 psia. Do you want to pump tank "A" up to 4000 psia while you reduce tank "B" to 28 psia, then reverse, OR do you want to begin a reverse process now and pump tank "B" back up to 4000 psia while reducing tank "A" back to 28 psia.

And let me get this straight. The connecting line is ten thousand feet (10,000) of 3/4" (0.75") diameter? That sounds "iffy".

BTW the friction losses are there at all times any gas is moving. A portion of the total pressure difference between tanks is due to friction, how much is dependent on how fast the gas is moving through the 3/4" diameter pipe. The remainder of the pressure difference between tanks can (at least for some small time step) can be considered as a static pressure difference.

And lastly, what gas. Molecular weight of the gas, amongst other things, makes a difference. What maximum temperature is the gas allowed to reach during the compression process. Will you have cooling equipment in place to keep the gas below that temperature during the next and subsequent uses of the compressor in the succeding cycles?

What would you be doing, if you knew that you could not fail?

 

[jfgaltenco]

OK sorry for not being clear enough let me try to clarify:
Initially,
Tank A is at 4000psi and B at 28 PSI
One need to move gas from A to B until
A is at 28 PSI and B 4000 PSI
Once this is achieved, gas flow is reversed until
A is at 4000 PSI again and B at 28PSI.... .. and so forth.

Regarding your comment about the line dimensions. I fully agree with you but there are other constraints that drive these dimensions.

To be clear, my first choice was to move these tank closer to each other, my second was to get bigger lines.
None of that make sense in this process context [site safety consideration + interconnection with other devices.]

Regarding the friction losses Of course they are here all the time and we keep track of them [ system is horizontal so no gravity involved] and we are willing to pay the HP price for them.

The only moment where they are really tricky to deal with [ not a HP issue,] is toward the end of a cycle and only because suction pressure CANNOT go sub atmospheric so in order to reach 28psi at tank A level the pressure drop through the 5000' line from A to compressor cannot be larger than ~14psi which can only be achieved through a gas velocity reduction in the line and accordingly a capacity reduction for the compression station.
BTW process is isothermal temperature being controlled otherwise, so temperature is not involved here]
what happen during the compression process, temperature for e.g, is not important as long as the compression is energy efficient [ we dont want to waste BHP when this can be avoided...].

Gas is a complex dry hazardous mixture hydrocarbon based with sg<1.
A solution that will work for dry pure methane could be adapted to our process.
Thanks
JF

 

[BigInch]

All properties similar to methane?

What would you be doing, if you knew that you could not fail?

 

[BigInch]

So you want to draw down pressure until you almost begin vaporizing water at 1 atmosphere pressure... or something.

My experience is that gas compression isn't so isothermal. Ambient gas temp is? And a max allowed temperature is needed, or by default will probably be around 250F max at compressor stage discharges. Final stage discharge maximum temp? Will that be cooled before entering the 5000 ft of discharge piping from compressor to tank.

Probably looking at something like a relatively small hi pressure recip, say of a type suitable for scuba tank filling, provided the pressure drop of the 1" pipe can be accomodated with required flow, has acceptable noise levels and at all inlet/outlet pressures.

The controlling parameter is probably going to be low flows when pipe differential pressures are low and gas tempeatures high. A 15 psi differential on 5000 feet of pipe may not give any real flowrate to work with, whether you like it or not, hi temperatures not helping. And it is difficult to move any useful amount of gas when pressures get to 1 bar and temperatures are high. The low pressure side may have to be raised. So what is the highest pressure you can have in the suction tank and still get this process to be a success. In other words we would like to see the highest possible minimum operating pressure used in the suction line (to maximize gas transfer in the suction line and facilitate completing the task within the allowed time.) Must you go as low as 28/14 in the suction side?

What would you be doing, if you knew that you could not fail?

 

[zdas04]

There are a couple of boundary conditions missing here:
1. The compressor has to physically be somewhere. If it is at one of the tanks, then you are pulling down a very long, very small suction line half the time (low pressure pipe has a much higher specific pressure drop than high pressure pipe). If it is in the middle then the small suction line is shorter, and the same length all the time. If it can be staged into 4-5 sites the overall performance improves. Remember that a recip wants suction pressure to stay +/- 5% of design pressure and a centrifugal wants to stay +/-2% of design suction (you are not going to get a compressor that works very well with 1999 psig suction and is still working well at +1 psig suction).
2. There is nothing about compression that is isothermal. If you can make that assumption then I can only conclude that this is a homework problem.
3. If it is a homework problem then I'm not wasting any more time on it.

Please convince us that this exercise is a real-world, business application. Maybe by telling us why you need to endlessly compress and expand this poor gas that never did anything to you.

David

 

[jfgaltenco]

HI guys thaks for all these comments.
here are a few answers

1 It is NOT a homework problem but a REAL LIFE situation. If you read my question above, I was /am NOT asking anybody to solve it for free but I am asking If somebody is aware of a comparable problem having already been worked through practically.
Again the concept is not difficult but it will require a carefully optimized packaging and component selection to be efficient.
At the end of the day I will hire a profesional to do this work.and at this time I am examining my options for that.
The one I am usually working with is aware that it is a special problem and that some prior experience could save a lot of time/money in development but if this does not exist we will have to start it from scratch together.

The highest possible pressure on the suction side is 28psia.

So this is really the extent of my question:
Does any body here knows or recognize in the case as described above a case that have been already worked through or not..?

If yes I will contact the professional that did the work and/or I will read whatever material is available.
I dont expect any easy solution coming out of an internet forum even this one to which I am new and that appear to be pretty good. But if I was proven wrong it will be all for the best.

The compressor is located in the middle of the tanks so the total length is 10000' made of 2 pieces of 5000' line and the compression station in the middle

Thanks,
JF

 

[jfgaltenco]

regarding isothermal process
I fully agree with you David and I think that my statement was not that the process was naturally isothermal but that "temperature was controlled otherwise". In fact, the gas system is part of a bigger plant and the layout is such that the tank and the line can be assumed in quasi static thermal equilibrium for the leading thermal effect is heat transfer through the wall with a time scale much faster than any other involved mechanism.
I did not emphasis it because I believe it makes no difference on the problem to solve. So isothermal assumption is a practically good working assumption here as far as I am concerned.
As I mentioned in my previous post also the compression station can and I believe will be cooled, but this cooling function has to be included in the package to be developed.

JF

 

[BigInch]

The thermodynamics will affect the compression and the discharge temperature going into the 5000 ft of pipe will be high, if not cooled at interstages and on final discharge. What temperature do you need there. 250 F won't do you much good going into that 1" line, so an aftercooler will be required too.

Problem I foresee is that at 14-28 psia I kind of doubt you will get any flow down 5000 ft of 1" pipe even at ambient temperatures. Higher pressure, lowest temperature and larger diameter will surely help you accomplish your goal, if it is possible at all. If I were doing this, I would model it using Stoner but I would have to first set up the tanks constraints by converting one into a variable pressure & flow source and the other into a variable pressure and flow sink. Stoner won't model a gas pressure tank directly; needs a Q = f(P).

You could do some quasi steady state in Excel, but you'd have to be very careful when deciding the duration of each time step.

What would you be doing, if you knew that you could not fail?

 

[zdas04]

"The highest possible pressure on the compressor suction is 28 psia? What happens when you turn the process around and the "suction" tank is at 4000 psia? The more you say, the less sense this makes. Could you describe what this juggling act is trying to accomplish?

David

 

[jfgaltenco]

the highest possible pressure in the tank on the suction side is 28 psi [ this is how much we must empty it]
The pressure at compressor inlet is what it is -function of pressure losses along the line from A to compressor but cannot be below atmospheric.

We already have a full computer model of the tank + line. and a conceptual design of an ideal compression station is not difficult to perform either and we have done it WITHOUT going into the "detail" that include compressors stage selection bottle sizes inter stage cooling requirements check valves... and so forth

the question I am asking here is about knowledge of existing practical work/experience that could be efficiently applied to the compression station effective design and construction, understanding that the specifics are:

1- move gas from Tank A to B with the least amount of work done, while pressure on A decrease from 4000 to 28 psia while pressure on B increase from 28 to 4000 psia. so VERY VARIABLE Inlet and discharge pressure through the cycle.

2- will efficiently reduce capacity when suction pressure come close to atmospheric to finish "emptying" the tank down to 28 psi without creating a sub atmospheric pressure point at the compressor inlet.

JF

 

[BigInch]

Come on. What's difficult about 10,000 feet of 1" pipeline, other than moving stuff through it. Get a VSD and a PID controller and rig compressor speed for suction pressure control, keeping suction pressure at or above Ps + 10%. As suction pressure decreases, speed will decrease and flow will decrease as required to keep the set pressure. Discharge pressure will also decrease with speed. The only trick here is actually moving enough flow within the time allowed. If you've got the model, you know all there is to know about that. The tech is old hat. Older than my hat. You don't have to do anything for the first 2000 psi of change, except let the tanks equalize. The compressor must pump one tank down from 2000 to 28 while pumping the other up from 2000 to 4000. Go down to the local SCUBA store that can fill tanks and see one in action. The only difference will be the 5000 feet of pipe not hanging off each end.

What would you be doing, if you knew that you could not fail?

 

[Compositepro]

What a bizarre process. The pipe probably contains more volume than your pressure vessels. Given that, the only way this could work is for the compressor to be half-way between for symmetry, or have a compressor at each end.

 

[BigInch]

I understood that the compressor is in the middle, but a compressor at each end, pumping from the tank, might not be a bad idea either. In any case, its a lot of 1" pipe at low pressure and, even if cooled, that'll be like blowing out a BBQ pit through a straw.

there's about 1.25 m3 in the pipe.

What would you be doing, if you knew that you could not fail?