inert gas closed loop process issues

[Runaround]

Looking for assistance with closed loop inert gas process system blower requirements. In an open system discharging to atmosphere the calculated system pressure drop (PD) would be "X". Now,
same system in a closed loop arrangement returning the process gas to the system inlet, we have an actual PD of approx. 2X with regen blower maxed out. The excessive PD in the closed loop introduces air leak issues into the system along with excessive energy consumption and most of all, unstable blower performance. I have found that with insufficient gas volume within the closed loop system, the blower will operate off of its design curve (beyond designed PD). When you add volume to the system by adding an inline vessel or tank you will lower the PD and get a performance from the blower as if it were in a standard open system. It is as if the blower is starved for air.

I want to show that we can eliminate the high (excessive) PD problem by adding gas volume to the system and I want to understand how to calculate how much volume would be required for a blower to perform as if or almost as if it were in an open system. Any ideas or experience sharing would be greatly appreciated. This is the second time dealing with the same trouble in a closed loop system.
Thanks
Mike

 

[rotw]

there is a relationship between pressure drop and gas density. The density is function of your discharge pressure.
A higher discharge pressure would increase your discharge gas density and pressure drop would also increase.
So some considerations:

- Irrespective of the pressure in/out conditions of the blower, ROUGHLY your machine needs to function at a certain actual m3/h - you could say its (+/-) the design point. Keep this in mind and keep also in mind that nominal point of your machine as a target.
- The higher the suction pressure, the higher the discharge via compressor ratio (unchanged) and the higher of course the delta P accross your machine. At the end the delta P across your machine is typically the pressure drop of the closed loop.

As an attempt:

My assumption is that when you switch from open loop to closed loop, your system resistance increases. That shifts your operating point to the left toward surge and because gas is routed back to suction, also increases suction pressure. Increase of system resistance increases the compressor ratio. Which then increases the suction pressure and ultimately the DP. So what you are trying to propose if I am correct, is to increase mass flow (by injecting more "kg" in the closed loop) as a way to "re-shift" the operating point of the compressor towards the right as a compensation. In this case, I think you should look at the change in am3/h from when compressor operates at the left (consecutive of switching from open to closed loop) to reaching the nominal point am3/h. Translate that m3/h difference into mass flow considering the operating pressure (just a guess, intuitively I think you should consider the pressures you want to restore, i.e. the open loop pressures), of course MW is unchanged.

 

[LittleInch]

You really need to sketch this up as it's not making a lot of sense at the moment.

"the calculated system pressure drop (PD) would be "X". " What system drop, from where to where is the PD being measured
What is the open loop inlet pressure into the regen blower

What is the difference between the two systems?

You can see all this but we can't.

Sounds to me like the velocity in the closed loop goes up, but the description isn't very clear. A diagram is worth a thousand words.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[Runaround]

I'll try to explain further. I attached a rough sketch and blower curve.

The expectation would be for the blower to operate at roughly 1000 - 1200 mm WG. Actual data taken from Clean duct between filter and blower is 2700-2900 mm WG, 250 M3/HR, motor draw 10 amps.

In my tangle with another closed loop system, the OEM at last minute increased a blower size without proper consideration. The centrifugal fan was performing to the left of the peak on the curve. I calculated all of the internal volume in the system in cubic feet. Then, measured air flow / internal volume ( cfm/cuft). From that, determined how much internal volume would be needed to achieve a desired cfm. Finally, this new required internal volume minus the current system volume is the amount of internal volume that needed to be added to the system to get the fan to performed where it belonged. The fan was starving for air. I modified the system that added an additional cyclone and duct that added the required volume. The fan then performed at a lower SP , the correct HP and air flow per the curve. In addition to the fan lining out, all of the components within the system performed properly as well.

I think this may be the case with this system,but I don't know regen blowers and I can't explain why the fix applied to the other system worked.

I really appreciate your feedback. It is very difficult to find the experience to trouble shoot this type of system.
Mike

 

[Runaround]

Sorry,
Here is the system sketch.
Mike

 

[LittleInch]

Still only about a quarter of the information needed.

You don't say what pressure drop across the blower is. By the look of it if somehow you know your flow is 250, then you're off the end of the curve for a 091610 type working at 60 Htz.

Are all these WG's positive pressure? (Is the 100-1200WG a positive number compared to atmospheric or negative - makes a big difference.)
what is happening to temperature?

No idea what the volume of this thing is or how does it pressurize / replace any air lost? No idea what is happening inside the "glove box"

The issue though is that the blower curves are designed for either atmospheric air entering the blower exiting at your desired pressure X. With too little volume in the system the closed loop is unstable and somehow feeding on itself or as you say starving itself. A bigger volume simply makes the system work better as there is more volume to use as a buffer storage.




Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[Runaround]

Thanks much LittleInch,
Right, so the total system volume is 4.1 M3. Temp is controlled. Inside the isolator is a material feeder (far right) followed by in line air classifier followed by cyclone. Cyclone air discharge goes to filter. The blower draws a vacuum on the system and returns the gas back to the isolator. The system is using nitrogen as the gas. The system is completely purged of oxygen.

The flow is measured with a pitot tube. I do not know the drop across the blower. We are reading our gas flow with the pitot tube and plotting on the curve to determine SP. Our data does not align with what should be happening.

Thanks Again,
Mike

 

[LittleInch]

OK, I'm finally getting there.

Re-reading your posts my understanding is that you had an "open" system with presumably an open inlet. This means that at either end of the system you have a fixed pressure value (atmospheric). It also means that in operation, you have a higher pressure and higher density than atmospheric air downstream of the blower, slowly reducing to atmospheric pressure. For the sorts of pressure you have if the blower is producing 300mbar, the volume is decreased by 30% or density increased by 30%. This means that there is a greater mass in the system than exists when the blower is off.

Now your closed loop system doesn't have these fixed pressure locations and the "zero point", where pressure equals atmospheric pressure can move around the system depending on pressure losses etc. If you don't pressurise the system or maintain a fixed pressure at the inlet to the blower then your total mass in the system when working will be lower than it was in the open system. Because parts of your system are now operating in "vacuum" mode that they weren't before when all parts of the system would be above atmospheric pressure, this means the density reduces, the velocity increases and as pressure drop is often a velocity squared rule, the pressure drop can increase a lot.

Simply adding more volume may help, but in reality this is probably just lowering the differential pressure.

Solution for me is to either pre pressurise your entire loop to say 2000mm WG or set a pressure of atmospheric pressure at the inlet to the blower and add N2 as required to maintain that pressure.

In summary, if, when you start the blower, your closed loop is only at atmospheric pressure, your system just doesn't have enough mass to operate in the same way as it does in a once through system. Therefore you need to add more mass into your closed system to replicate the once through system and its pressure drop and flowrate

Does that make sense?
Let us know.

LI

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[Runaround]

LI,
Thanks, I'm working on it. I'll get back to you.
Mike