Pressure Regulator functions and problems

  • Thread starter Jet Boer
  • Start date
  • #1
13
0

Main Question or Discussion Point

I have recently been experimenting with pressure to generate a mass flow and discovered something really bewildering.

First let me explain the procedure.
I use large stainless steel tanks filled to about 3/4 with water so there is a large enough capacity. I then pressurize the system with nitrogen to about 30 Bar using a high pressure regulator. The pressure in the nitrogen tanks is about 150 Bar.
I also record the tank pressure with respect to time.

Now, when I plot the Data on a graph I see a pressure drop in the tank and it settles at about 25 Bar, which is not good for the constant mass flow I desire.

Question:
Why is it that the regulator is unable to keep the pressure more or less constant at 30 Bar? Why is there a drop in pressure?
 

Answers and Replies

  • #2
Averagesupernova
Science Advisor
Gold Member
3,603
607
My first guess would be a crappy regulator or running it too close to the edge of its spec.
 
  • #3
13
0
No, the regulator is well within its specified input and output range. We also thought that regulator was faulty, and tried another regulator, but the pressure drop still occurred.
 
  • #4
Q_Goest
Science Advisor
Homework Helper
Gold Member
2,974
39
Regulators can not maintain their set pressure exactly. The decrease in downstream pressure which occurs as flow rate increases is due to how the valve functions. All regulators function basically the same way. Some are spring loaded and some are gas loaded (called "dome loaded" regulators). Regardless of which type you have, the valves function the same, but for the sake of explanation, I'll explain it using the more common spring loaded regulator.

The spring acts against a diaphragm, pushing down on the diaphragm. This assembly also pushes down on the poppet that controls flow. The further down the spring pushes, the farther the diaphragm moves and the further open the poppet goes. The further the poppet opens, the more gas that can get through and the higher the flow rate through the valve. The gas pressure downstream of the valve is being 'measured' by the diaphragm with that pressure on the inner side and atmospheric pressure on the outer side. So the force on the assembly is balanced by the spring pushing down and the pressure pushing up on the diaphragm.

When you set the regulator at a certain pressure under a no flow condition, the pressure downstream is acting on this diaphragm and the spring is balancing that pressure by applying an equal force downward while the poppet is on the seat, stopping flow. As downstream pressure drops, force up on the diaphragm also drops and the spring extends and the valve opens a little bit. But a spring has a spring constant (k) so the valve will come to equilibrium with a slightly lower pressure downstream, balanced by the slightly lower spring force and with the valve slightly open. As pressure downstream decays even further, the valve has to open more and the spring has to extend more and the forces balance at this lower pressure and lower spring force.

So the spring constant (k) is what is affecting the balance on the diaphragm. If you have an infinitely small spring constant, there would be no decay in downstream pressure with an increase in flow rate. But there is no such thing as an infinite spring constant and real regulators experience a decay in the downstream pressure set point as flow rate increases. This phenomenon is sometimes called "droop" although I've heard manufacturers call it by different names. It's a function of the spring constant and diaphragm area.

If you need a regulator with minimal droop, try a dome loaded regulator. The effective spring constant of a large volume of gas is generally lower than a spring.
 
  • #5
Averagesupernova
Science Advisor
Gold Member
3,603
607
Am I to understand that you want to maintain a pressure of 30 bar while the nitrogen is passing through the regulator? A pressure regulator is a feedback device which Q_Goest explained fairly well. Some things you may want to consider: The regulator opens its own valve based on how close the output pressure is to the reference. The reference being the spring mentioned in the above post. Sounds easy enough right? A pressure regulator that has a significant amount of flow through it will have to open its valve more than the same regulator having minimal or no flow through it. What opens this valve more is a lower pressure on the diaphragm. There will always be some error. There will also be some repeatability error. From no flow to flow and then back to no flow, or anywhere in between. You can try increasing the physical size of the regulator. Just a shot in the dark since I don't know your flow rate. Believe me, I've been there and done that on a pressure transducer calibration setup.
 
  • #6
13
0
Q Goest & Averagesupernova, thanks for the reply, but I got to an answer as well.

Q Goest: Yes the "droop" is something that I will keep in mind, but if the effects are minimal, that won't be my major concern.


(Tell me if you think I am wrong)

I think the problem is that I approached the problem from the wrong angle. I only focused on the pressure and to keep it constant because if you can keep the pressure constant you can keep your mass flow of water constant. I was thinking more in terms of static properties and less in terms dynamic properties.

I should have been rather focusing on the volume flow of gas into the tank in relation to the volume flow of water out. After asking around at my university today, I discovered that manufactures provide graphs specific to the regulator that buy from. On those graphs they relate the volume flow to the output pressure (because flow only occurs with a pressure difference).

I will soon be doing an experiment with the volume that I calculated instead of the pressure.
 
  • #7
Q_Goest
Science Advisor
Homework Helper
Gold Member
2,974
39
I should have been rather focusing on the volume flow of gas into the tank in relation to the volume flow of water out. After asking around at my university today, I discovered that manufactures provide graphs specific to the regulator that buy from. On those graphs they relate the volume flow to the output pressure (because flow only occurs with a pressure difference).

I will soon be doing an experiment with the volume that I calculated instead of the pressure.
Yes, air displaces the water so for every dV of water displaced, you need to provide enough air to make up for it. That's a fairly simple calculation. For example, consider a tank with at atmospheric pressure and a 'hole' in the top open to atmosphere so that the top surface of the water is always at atmospheric pressure. As water flows out, air flows into your tank at constant pressure to maintain atmospheric pressure. There's no reason the pressure in the tank should necessarily drop if there's a source of air that is capable of maintaining the pressure.

Similarly, there's no reason to insist that the pressure in your tank MUST drop as water is expelled. The volume or mass of air that needs to go into the tank to maintain the pressure can be calculated by knowing the volume of water it must displace. So you still need to look at your source to see why the source can't maintain a constant pressure. The reason I see is your regulator may droop. An alternative reason is you have irreversible pressure drop in your piping between the 150 bar cylinder and the regulator or possibly between your regulator and your tank. If you don't know how to calculate that, we can examine those possibilities too.

Regardless, I suspect you have a 'droop' issue with the regulator. Take a look at the documentation for the flow rates for a typical regulator; they're on the second page of the file:
http://www.documentation.emersonprocess.com/groups/public/documents/data_sheets/d44221731x012.pdf

You'll notice that the X axis shows flow rate of nitrogen and the Y axis shows the pressure. Given an inlet pressure and static set pressure (static set pressure is pressure regulator is set to with no flow) the pressure decays as flow rate increases. This phenomenon is due to the droop I described above.

Note that an alternative way to perform your test is to have a hand valve downstream of your regulator which you open to start the test. But instead of having the regulator set at 30 bar, you set it higher so that the tank pressure remains constant at 30 bar throughout your test. You can figure out what that value is fairly easily simply by cranking down on the regulator during the test so you maintain 30 bar in the tank. Now shut the hand valve, fill the tank, and perform the test by opening the hand valve. You'll find when you shut the hand valve, the pressure immediately downstream of the regulator will be higher than 30 bar, probably closer to 35 bar.
 
  • #8
13
0
You'll notice that the X axis shows flow rate of nitrogen and the Y axis shows the pressure. Given an inlet pressure and static set pressure (static set pressure is pressure regulator is set to with no flow) the pressure decays as flow rate increases. This phenomenon is due to the droop I described above.
Thanks for explaining the graph, I have this graph of my regulator [/PLAIN] [Broken]
[URL="[PLAIN]http://www.igs.org.za/images/high-pressure/304_Graph.jpgl"],[/PLAIN] [Broken] but couldn't understand it properly. It is the first regulator range.

Just to get my thoughts strait and make sure that I understand the graph, because my graph is not set out as clearly as the one in the documentation you suggested. (According to the graph above) Ex.1: Say I want a flow rate of 944 liters/min at 50 Bar, does that mean that my static set pressure needs to be about 70 Bar.
 
Last edited by a moderator:
  • #9
  • #10
  • #11
13
0
Thanks.
 

Related Threads on Pressure Regulator functions and problems

Replies
7
Views
70K
  • Last Post
Replies
2
Views
1K
Replies
7
Views
1K
  • Last Post
Replies
6
Views
2K
  • Last Post
Replies
6
Views
1K
  • Last Post
Replies
2
Views
822
  • Last Post
Replies
0
Views
1K
  • Last Post
Replies
1
Views
929
Replies
15
Views
10K
Replies
2
Views
2K
Top