Beer and Fluid Dynamics

[ZenBiker990]

I have a "friend" who has a "brewery". He has a pressure relief valve on the fermentation tank that pops open the drain at 100 psi. All was good until the tank that the finished beer drained into went bad and he installed a new tank. Nice, shiny, stainless steel, BUT (and there is ALWAYS a "BUT") the new tank also has a pressure valve in it, but it doesn't open until 200 psi. This second valve is in line with the valve on the fermentation tank. My buddy is all bent because his fermentation tank is only rated for 250 psi, and he believes adding the new 200 psi valve in line with the 100 psi valve makes 300 psi potential pressure, potentially rupturing the fermentation tank, and he is ready to spend big bucks to modify the valves. I told him "no, the operating pressure on the fermentation tank will only be 200 lbs maximum; the whole system will open when the resistance of the highest rated valve is reached - the pressures don't add all other things being equal." HELP US SAVE THE BEER!!!!! (If I'm right, I get free beer for life, so anyone on my side...well, you know; )

 

[Maggot]

Drink liquor and you wont have that problem.

 

[ZenBiker990]

not helping, Howard...

 

[Nik H]

I think you are correct...been a while since I had to do these kind of problems but the valves should be entirely independent.

Think about it this way; when the system first comes up, the first valve is closed until the pressure reaches 100 PSI. The second valve knows or sees no pressure. The first valve opens at 100 PSI and feeds the next run. The pressure valve at the 2nd segment will not open until 200 PSI is reached. The pressure at the first valve is still 200 PSI not 300 PSI. The second valve will open and let things function properly.

Your friend must be thinking about this like resistors in series but it doesn't work that way.

This being said, the company that sold him the new tank must have applications engineers that can walk him through it. My undergrad degree in physics is a bit rusty but I think I am right.

 

[...]

I think you are correct...been a while since I had to do these kind of problems but the valves should be entirely independent.

Think about it this way; when the system first comes up, the first valve is closed until the pressure reaches 100 PSI. The second valve knows or sees no pressure. The first valve opens at 100 PSI and feeds the next run. The pressure valve at the 2nd segment will not open until 200 PSI is reached. The pressure at the first valve is still 200 PSI not 300 PSI. The second valve will open and let things function properly.

Your friend must be thinking about this like resistors in series but it doesn't work that way.

This being said, the company that sold him the new tank must have applications engineers that can walk him through it. My undergrad degree in physics is a bit rusty but I think I am right.

as an engineer, you are correct.


also, the tank is not likely to burst at 300 psi if it has a rating of 250.......its unsafe for sure....but that tank probably has a burst rating of at least double or tripple the working pressure

 

[Nik H]

as an engineer, you are correct.


also, the tank is not likely to burst at 300 psi if it has a rating of 250.......its unsafe for sure....but that tank probably has a burst rating of at least double or tripple the working pressure

Electrical engineer (wireless communication and electro-magnetics) but a physics/math major in my undergrad. Nowadays, I am nothing as I am a VP of engineering so I do nothing useful anymore.

I appreciate the vote of confidence

 

[ZenBiker990]

Thanks guys. You have no idea. I'll post the results of this revelation. Lager or ale?

 

[ZenBiker990]

I think you are correct...been a while since I had to do these kind of problems but the valves should be entirely independent.

Think about it this way; when the system first comes up, the first valve is closed until the pressure reaches 100 PSI. The second valve knows or sees no pressure. The first valve opens at 100 PSI and feeds the next run. The pressure valve at the 2nd segment will not open until 200 PSI is reached. The pressure at the first valve is still 200 PSI not 300 PSI. The second valve will open and let things function properly.

Your friend must be thinking about this like resistors in series but it doesn't work that way.

This being said, the company that sold him the new tank must have applications engineers that can walk him through it.

That was my explanation as well. The tanks are expensive, and yeah, the burst ratings should be nominally higher than rated, but he isn't willing to risk going against rated values. As far as the new tank mfgr., they were the ones who told him the pressures were additive despite similar valve designs.

 

[1J04]

Thanks guys. You have no idea. I'll post the results of this revelation. Lager or ale?

Both. Definitely both!

 

[Nik H]

Thanks guys. You have no idea. I'll post the results of this revelation. Lager or ale?

I hope it works out...don't drink beer...only gin and bourbon and rye and scotch...LOL

 

[Veer_G]

Electrical engineer (wireless communication and electro-magnetics) but a physics/math major in my undergrad. Nowadays, I am nothing as I am a VP of engineering so I do nothing useful anymore.

I appreciate the vote of confidence

You need to go ask Magnum Boogers.

 

[Nik H]

You need to go ask Magnum Boogers.

Magnum Boogers???

Don't know who that is Veer...

I can ask my son as he is a chemical engineer that does plant design for Rockwell. I sent him an email but he is 25 and is likely out drinking today

 

[Veer_G]

Magnum Boogers???

Don't know who that is Veer...

I can ask my son as he is a chemical engineer that does plant design for Rockwell. I sent him an email but he is 25 and is likely out drinking today

http://www.magnusboner.com/

 

[Nik H]

http://www.magnusboner.com/

Practical experience wins the day...I will read it.

 

[ZenBiker990]

http://www.magnusboner.com/

Wow...all that copper...the moonshiner in my DNA twitched when that page loaded...singlin's...doubled and twisted...makes me think about a new career path...

add/edit: a new career in beer making, that is, because unlicensed liquor production is illegal. Just to be clear. Nonetheless, thinking of Nik, Maggot, and all the other spirit drinkers out there:

 

[Veer_G]

Practical experience wins the day...I will read it.

I don't know that he covers it specifically, but he might know. Just beware, it can be painful.

<iframe width="560" height="315" src="https://www.youtube.com/embed/PnAbRktI5Jk" frameborder="0" allowfullscreen></iframe>

 

[bogeybrown]

As mentioned, he "should" be fine, but for him to CYA he should speak to the supplier, via email with pictures etc and get a determination from them. Having worked in a commercial brewery a while back I got to see first hand how some of the companies will happily sell you several $100k worth of shiny stainless goodies but leave you SOL when it breaks claiming the fault of the user.

During active fermentation when pressure is building rapidly we just used a blowoff hose sitting in sanitizer to keep from having to rely on a PRV and have it blow shit all over the brewery. The PRV is a last resort, as tanks have pressure gauges and manual bleed off valves to keep from getting anywhere near an unsafe pressure.

I'll if I can upload a brief video or two of fermentation blowoff. Nope, can't figure out posting a vid from my phone.

Edit: from re-reading your original post it's unclear if his issue is with a fermenter or the Bright Tank (where finished beer resides before packaging). If it's the bright then any meaningful pressure comes from him adding it himself with CO2 to reach his final carbonation level. Unless he forgets he's got gas running to it (which can happen) the vessel won't reach a pressure he hasn't applied himself.

 

[diverdon]

I agree with everything written by the other Hiders above with only one possible exception. The exception is so preposterous that it is almost nonsense, but as the representive of his supplier told him that they were additive we should look at it.

Suppose one tank is completly contained inside the other, then the outer tank could be pressurized to its max say 200 psi, then that 200 psi becomes the enviorment for the inner tank, so to the inner tank relief valve the 200 psi links like zero so the inner tank could be pressurized to the 200 psi of its enviorment plus the additional 100 psi its relief valve will hold, or 300 psi when compared to the outside enviorment.

If we then suppost that the outer tank was all liquid and no gas, and then suppose that the inner tank had a gas over the liquid, then if the inner tank relief valve opened the outer tank wouls be at 300 psi until its relief valve released the pressure.

 

[Nik H]

The pressure from the 200 PSI tank would be force on the outside of the inner tank. It's 100 PSI pressure will be pushing in the other direction (inner surface) so I still think it would be OK; if I am reading you right.

 

[diverdon]

Unless the outer tank were at max pressure and the inner tank pressure release activated then the pressure of the inner tank would be added to the pressure of the outer tank. Of course the outer tank pressure relief would begin to vent but there could be a brief interval where the pressure was higher.

I can see no reason to set up the tanks in this manner and I doubt that they are, yet given that all we have is an imprecise description of how the tanks are being used it is difficult to be sure how they relate to each other. If they are simple connected by a pipe and one has a relief set for 100psi with the other being set for 200psi then both tanks will have the same pressure and the 100 psi relief will release if the pressure gets to 100 psi. The only way you would ever need the 200 psi release is if the 100 psi relief malfunctions or if there is a valve between the tanks.

 

[Nik H]

http://www.magnusboner.com/

I will read this a few times. Lots of excellent information and written well.

Thanks Veer

 

[Nik H]

Unless the outer tank were at max pressure and the inner tank pressure release activated then the pressure of the inner tank would be added to the pressure of the outer tank. Of course the outer tank pressure relief would begin to vent but there could be a brief interval where the pressure was higher.

OK...I will think about that scenario. It sounds right but...

Likely?

As you said...more for fun than practicality

 

[jrassy]

ALE

 

[ZenBiker990]

Diver, I think you right if we assume frictionless surfaces, that space is an Absolute Vacuum, and that Bill didn't inhale

 

[Sean the Nailer]

Throw this out there, into the 'stew' of the discussion. These relief valves. Are they measuring a pressure differential between upstream and downstream, OR are they measuring a differential between internal and atmospheric?

If the answer is atmospheric, then tank A could be prevented from venting until 200 psi. Depending on how much pressure is in tank B. If tank A needs to release, and there's no pressure in tank B, then everything flows as it should.

Whereas, if there is pressure in tank B, and tank A has to vent, then it would be a balance of the differentials between tank A and tank B.

On the contrary, if the relief valves were measuring upstream/downstream differential, then tank A "could" attain a pressure of 300 psi.

Pre-coffee math, ain't it fun?

 

[Mike Casselton]

^^ This, if A vents into B to atmosphere.

I was thinking the below statement until I remembered some things from my old job.
"Why not change B tank to a 100 psi relief valve and get rid of the worry.
Have both valves calibrated."

I don't think this will work

Also, if tank B has any built up pressure, the back pressure will increase the force required to open tank A relief valve.

Look at the design of a fill valve/ pilot valve system on an aircraft.
Liquid pressure opens the fill valve at 65 psi. As the tank fills and begins lifting the float, back pressure in the lines increases and puts pressure on the back side of the fill valve diaphragm. Flow slows as the float rises and increases back pressure. Once the float rises to its full travel, 65 psi is applied to the back side of the diaphragm closing it with the aid of the internal spring pressure. Flow into the tank stops.

If we use the above scenario, any internal pressure in tank B will increase the force required to open the relief valve for tank A.
Therefore, tank B should be atmosphere, or change A to a lower set point to compensate for any back pressure from B.

Just my useless non - engineer opinion fellas

 

[bogeybrown]

I couldn't even hope to keep up with the math involved in double-walled pressure vessels, but to help some of y'all who know the science here's some basics on brewery vessels:

Mash Tun (where hot water mixes with brewing grains) may be double-walled, but usually just for insulation purposes
Boil Kettle may be double walled with steam jackets for heating
Hot Liquor tank (for storing large quantities of on-demand hot water) may be double jacketed with steam jackets
Bright Tank (for finished beer) and Fermenters may be double-walled to utilize Glycol cooling jackets to set/maintain proper temps

I say "may" in those setups because there is ENORMOUS variability in brewing equipment, with some being direct-fire electric or gas burner and others being steam. Similarly with the cold vessels some breweries use Glycol jackets while others use single-walled vessels inside temp-controlled rooms/coolers to adjust the temp of the tanks. You'll see old dairy equipment, salvaged vessels from other industries, plastic fermenters, whatever.

What is not common practice is the transfer of fluids between vessels under high-pressure situations. Transfer between vessels is accomplished through centrifugal (or similar) pumps and the receiving tanks are usually manually vented where necessary during fluid transfer to allow the liquid to displace the gas inside.

High pressure steam and CO2 are the "normal" high-pressure systems and the kettle isn't a pressure vessel in itself as it has an open vent stack.

All that being said, the area where huge pressure can develop rapidly is in a fermenter during active fermentation. In that vessel, the liquid will be near the top of the vessel and actively producing CO2 as a by-product of fermentation. Because of the huge amounts of gas being produced in such a small air space it gets big quick, which is why they are manually vented to allow blow-off. If that vessel is double-walled the outer jacket will contain Glycol rather than another pressurized gas. Same thing with a Bright Tank taking high-pressure CO2 for final carbonation: you can forget you've got the gas on and blow a PRV, but the outer vessel will be Glycol and not gas.

Hope that confuses shit further

 

[Mike Casselton]

So, the outer tank (if it has one) is not part of the pressure vessel (liquid or fermentation tank).
If that is so, then the PRV should be specified to the max user pressure, not nominal or max operating pressure of the tank.
If I'm understanding you correctly, the tank is manually vented during fermentation and also during Co2 injection and transfer.
The only purpose for the PRV is if the manual vent it's left closed during the procedures?
If so, then using what the builder specifies should be well within the safe operating limits of the tank.
I would think an inspection on the PRV would be in order if used with potential contaminates (mash, hops, etc) to ensure its not blocked or held open. Do the inspection every drain/fill cycle.

Does the tank also have a negative PRV to prevent collapse during transfer in case the manual vent is left closed?
Or, is there some type of cross venting between tanks that can be transferred?

I would love to see a schematic of the system

 

[bogeybrown]

So, the outer tank (if it has one) is not part of the pressure vessel (liquid or fermentation tank).
If that is so, then the PRV should be specified to the max user pressure, not nominal or max operating pressure of the tank.
If I'm understanding you correctly, the tank is manually vented during fermentation and also during Co2 injection and transfer.
The only purpose for the PRV is if the manual vent it's left closed during the procedures?
If so, then using what the builder specifies should be well within the safe operating limits of the tank.
I would think an inspection on the PRV would be in order if used with potential contaminates (mash, hops, etc) to ensure its not blocked or held open. Do the inspection every drain/fill cycle.

Does the tank also have a negative PRV to prevent collapse during transfer in case the manual vent is left closed?
Or, is there some type of cross venting between tanks that can be transferred?

I would love to see a schematic of the system

Keep in mind that I'm speaking from the experience of working on one system, learning from a guy who worked on over 20 different brewing systems in the US and EU:
When transferring to the fermenter the vessel is manually vented to allow the liquid to displace the gas in the tank. If the tank isn't vented you'd be building huge pressures by displacement and later through fermentation gases. I wouldn't want to rely on a PRV during fermentation because of all the solids and foam being burbled at the top of the liquid that could potentially foul a valve. We also allowed the fermenter to blow off during vigorous fermentation to prevent pressure spikes. Once fermentation had slowed down we could seal the tank and just vent excess pressure manually if needed.
During CO2 injection to the finished beer in the Bright Tank the vessel is sealed because the intent is to force the CO2 into solution. You can have an oops there if you forget you're running gas to the tank. However, both of those vessels (if double walled) would have Glycol circulating in the outer jacket rather that pressurized air. The true pressure-vessel is the inside sealed tank where the liquid beer is contained.

Hugely short story long, I still recommend that the OP's buddy get a determination in writing from the supplier. The brewing world was strange to me: If you buy $300k worth of "start-up" equipment from me, with the almost definite future need of more product as you increase production, I'm definitely going to take your call if you need my help. A lot of these fucking guys will kiss your ass to get the sale and then totally blow you off after installation, pissing away continued customer loyalty. Hell, some of these suppliers will blow you off when you send a RFQ. It must be nice to keep selling expensive equipment with horrible CS.

 

[ZenBiker990]

The valves we're talking about here are differential with respect to downstream pressure; tank "A" valve dumps int drain pipe that attaches directly to tank "B" and the second valve. Pressure in the drain system and catch tank is slightly below atmospheric. Pressure relief valves that vent tank pressure to the atmosphere are not being considered in the discussion for sake of simplicity; it's alll about the drain valves and lines.

 

[Veer_G]

It must be nice to keep selling expensive equipment with horrible CS.

I know shit all about brewing, but that sounds just like H&K.

 

[bogeybrown]

I know shit all about brewing, but that sounds just like H&K.

Ain't that the damn truth, which is why I only own a few despite my lifelong love of them.

 

[Sean the Nailer]

The valves we're talking about here are differential with respect to downstream pressure; tank "A" valve dumps int drain pipe that attaches directly to tank "B" and the second valve.

Then, due to this statement alone, this implies to me that the THEORETICAL maximum pressure that tank A can attain, is 300 psi. This would happen/take place when the pressure is 'maxed' on both tanks.

If the pressure on the 'hot side' of a valve is 99 psi and the 'vent side' is 0 (14.7 psi Absolute) and the valve is set to unseat at 100 psi, then you have the capability of restraining 1 more psi. BUT,

If the 'vent side' is 199 psi, then the 'hot side' can attain up to 300 psi due to the fact that there needs to be a 100 psi DIFFERENTIAL in order for the valve to unseat.

So, in short order, the theoretical shows that chaos 'could' happen. My question for you (and him) is this:

How often (and/or likely) does his brewing process contain/create the likelyhood where both of the tanks will be pressurized, and by that I mean both pressurized to the extreme limit of the design safety factors?

I'm not a beerologist, but I do make my own wine. Any racking/transferring that happens here involves a full primary into an empty secondary/fermenter/carboy. So I'm a little unclear on THAT part of the process you are alluding to. But then we can also talk about the likelyhood of other events occuring, such as a butterfly flapping its wings in a garden in Japan, to an asteroid entering Earths orbit and taking out the continent, to palestinian commando's rappeling down the side of your building and crashing in through the windows....... the list of movie highlight "potentials" can go on for eternity. Surely I jest, and I don't know if Shirley is your middle name, but you see where I'm going, right?

Hope this helps!

 

[ZenBiker990]

Hate to say this, but the free beer is off. Here's why, and it took me a while to really say "OK, yeah, I get that.": the valve from the fermentation tank is a differential pressure valve, meaning it opens when the pressure difference on either side of the valve exceeds its rating. The run between the fermentation tank and the holding tank becomes pressurized up to the rating of the second valve, but as it pressurizes, conditions dictate that the first valve has to overcome the pressure in the run between the valves before it opens; in other words, the fermentation tank valve has to overcome a linear increase in resistance to opening. This process continues until the line reaches a pressure allowing the second valve to open. Thus, the total pressure on the fermentation tank is the sum of the ratings of the two valves. Apparently this has been bench tested, and it offends my sense of logic, but that sense has led me astray before, so there ya go.

I need a beer.

 

[2ndamendfan]

I disagree. I dont think the tank pressuure will get over 205psi if both valves are accurately calibrated. Tank pressure hits 100. Valve opens. And fills the cavity between both valves to 100psi. Then valve closes. It happens again with a 100psi difference from inlet to outlet which would be 200psi. At that point the second valve opens due to 200psi difference. That then drops pressure in the pipe between the 2 valves which would keep the 100psi valve open. If the 200psi didn't open then yea the first valve would have to see 300 psi to open. Once the pipe in between was at 200. But at 200 the 2nd valve should open draining that pipe and thus releaving pressure. I don't know about super accurate valves. But i have never seen one that opened at 100psi and closed at 99psi. Most i know of don't close until pressure as dropped about 50%. . As such the 2nd valve would close at about 100psi, leaving that in the pipe. Which actually may leave 1st valve open. If it does close it would open again at 200 psi starting the whole process again. Unless I missed where the 2nd valve doesn't vent to atmosphere (open tank).

 

[Fitz.]

Mechanical engineer. I design machinery and deal with process tanks and valving.

I don't know brewing so take my comments for crap!

Forgive me if I have missed something that has already been covered.
Some general comments and some summary. Some restating for my thought process as well.


1) At the pressures your talking about these are pressure vessels. There should be a label indicating certified tank pressure. Caution in going from 100 psi to 200 psi on the original tank. Also, wouldn't be a bad idea to check component ratings.
2) Based on your comment as I understand. The new PRV is rated for 200 PSI and has 0-100 psi from the other tank on the relief side? This would be 200-300 psi before the valve seat lifted. That's unusual variable safety feature. Are you sure the new tank is plumbed correctly? Between tanks is typically a process valve, not a relief valve. Again, I don't know brewing.
3) To clarify the 100 PSI PRV on the original tank vents to atmosphere, while the 200 PSI PRV on the new tank vents between tanks? That would mean that original tank is part of 200 psi safety circuit on the second tank. Volumes are probably sufficient for safety but I've never experience this. You can go low pressure to high pressure on a relief but high pressure to low pressure would be unusual. Is this the only PRV on the tank? Do you have a P&ID drawing of the new tank?
4) PRV's I use have an actuator for lifting the valve seat during CIP function, flood tank and flow through

Again, I don't know brewing so take my comments for crap.

Fitz.

 

[ZenBiker990]

I disagree. I dont think the tank pressuure will get over 205psi if both valves are accurately calibrated. Tank pressure hits 100. Valve opens. And fills the cavity between both valves to 100psi. Then valve closes. It happens again with a 100psi difference from inlet to outlet which would be 200psi. At that point the second valve opens due to 200psi difference. That then drops pressure in the pipe between the 2 valves which would keep the 100psi valve open. If the 200psi didn't open then yea the first valve would have to see 300 psi to open. Once the pipe in between was at 200. But at 200 the 2nd valve should open draining that pipe and thus releaving pressure. I don't know about super accurate valves. But i have never seen one that opened at 100psi and closed at 99psi. Most i know of don't close until pressure as dropped about 50%. . As such the 2nd valve would close at about 100psi, leaving that in the pipe. Which actually may leave 1st valve open. If it does close it would open again at 200 psi starting the whole process again. Unless I missed where the 2nd valve doesn't vent to atmosphere (open tank).

For the purposes here, no atmospheric venting. You and Sean both give eloquent examples. Upon reflection, I believe some extra information is needed. First, the particulars of beer brewing can be ignored; this is strictly a question of pressure within the described system. Second, perversely enough in an age of precise digital control, the valves in question are controlled soley by springs. An adjustable mechanism (screw agaist fulcrum) allows the valves to be calibrated to "open" (definition? Just craaaack open @ rated pressure!) Springs are flexible pieces of metal annealed and tempered to have an increased modulus of elasticity, and exhibit an exponential response to increasing stress. I didn't consider this at first. Additionally, I realized there is another missing variable in this discussion: flow rate, which I'm guessing will be a function of pressure rise. The degree of valve opening will be a combination of flow AND pressure. Considering those nuggets, now let's look at the second valve: rated to open @200 psi, with line pressure hits 200 psi; the valve strains against the spring, opening juuust enough to maintain a line pressure juuust below 200 psi, then closes.. Meanwhile, upstream, the first valve is seeing "downrange" pressures juuust approaching 200 psi. Depending on how each valve responds to the flow/pressure gradient, the first valve will see what Sean calls "hotside" pressures approaching 300 psi, the exact value of which depends on the overlapping behavior of two totally mechanical and independent valve mechanisms. Ironically. some electricity and some solenoids, pressure sensors and some kind of logic circuit would solve this problem. The error I made was not considering the manner in which the valves open; I assumed they popped wide open, merrily ushering freshly brewed beer along its way until enough had passed, at which time each valve would snap closed and go back to the sports page and stale coffee. In truth, like most unsupervised non-skilled tasks, the valves function at low flow/efficiency, maintaining system pressures approaching, but not quite, the sum of the two valve ratings. Dangerously (you see this coming), if the fermentation tank sees a sudden spike in pressure because of a "sweet spot" in the fermentation cycle or increase in tank pressure, a system already on the edge of rated safe values can suddenly see a dangerous surge, especially if the connecting pipe can't allow the flow volume required to safely lower pressures. This is what I failed to consider before, and it highlights my biggest character flaw: the assumption that everyhing somebody else designs is optimally designed, ie, "how I would do it". In this case it led me to a completely erroneous, and dangerous, conclusion. Still can't believe this is the best it gets. On the bright side, it turns out that out intelligent control with system monitoring is what my buddy decided was the best option.