Point of no pressure change - help me understand

28W

I'm just a homeowner, so I have no reason for needing to know this other than my own curiosity. I've read all sorts of explanations about why the expansion tank is the PONPC, but I can't wrap my head around it.



In a closed-loop system WITHOUT an expansion tank, the circulator has no problem creating a pressure differential and moving the water, right. So what is it about having an expansion tank attached to the pipe that creates the PONPC? Does the tank somehow "absorb" the pressure of the water coming into it?



If there is somebody out their both knowledgeable and patient enough to answer this question, I would appreciate it. If not, I guess I will just take it on faith . . . kind of like gravity.

JStar

Ponpc

The expansion tank is just a facade. The real reason is the compressed air inside of the tank. When a system is a closed loop, the water can neither expand nor contract unless heated or added manually. And something has to define what the pressure is. With just a loop of water, once the fill valve is closed, the pressure drops to 0 or whatever water weight there may be. With an expansion tank, once the fill valve is closed, the comprssed air still puts pressure on the water. Nothing in the system can affect that pressure in the tank (unless heated or added manually).

Jamie Hall

Good way to look at it

"Does the tank somehow "absorb" the pressure of the water coming into it?"



Not a bad way of looking at it... as Jstar said, the key isn't so much the tank, as it is the air trapped inside the tank.  Since air is much MUCH more compressible than water, as the system heats and cools (and the water expands and contracts) the air compensates -- but it doesn't take much of a pressure change to compensate for the volume of expansion.  If you had no air in there -- or no tank -- the pressure change would be huge (and probably burst a pipe or fitting somewhere).



This is also part of pumping away -- since the pressure in the tank is constant, or very nearly so, the pump will add to that pressure to push the water around the system (gradually losing pressure as it goes) until it comes back to the tank -- which is fixed.  If you pump towards the tank, then the output of the pump is at the fixed pressure and the pump has to suck the water around the system.  And pumps work much better pushing than sucking...

jumper

supposedly

JStar gives as good an explanation as any. Here's a complicated one. Pump raises pressure to the maximum in your circuit and from there on pressure decreases from flow resistance. But in the pipe going to expansion tank there is no flow so pressure doesn't change. Except for height. Like JStar says the PONOPC is a facade bladder tank salesmen use to make folks make the mistake of buying them. Bladder tanks have their uses but homeowners are better off the old fashioned way. Especially when the tank is the high point of your system. Aren't you glad you asked ?

28W

Pressure in the tank is constant

These explanations are helpful - thanks. So if I understand this, the air pressure in the tank is what keeps the water in the tank at a constant pressure. A circulator pumping toward the tank will not be able to increase the tank pressure because it will simply be compressing the air instead . . . yes? On the other hand, a circulator pumping away from the tank "doesn't care" because it is sucking in water that is flowing out of the tank, and creating higher pressure on the outlet side of the circulator.

SWEI

expansion tank

Vertical height and water weight fall out of the equation once the system is filled.



Water expands and contracts with temperature.  Without an expansion tank, the volume of the system remains constant, so pressure also increases and decreases with temperature.  The "spring" of compressed air behind the diaphragm in a pressure tank allows volume to change while keeping pressure constant.



A circulator in a closed system produces pressure at its outlet by "pushing" water across the impeller, which drops pressure at the inlet.  If the inlet is hydraulically "close" to the expansion tank, the inlet never sees negative pressure, preventing cavitation.

JStar

Tank

Not quite.



Water can never be pushed or pulled from the tank. That's why its pressure never changes. If you imagine the water moving in and out of the tank, where is it going to, or coming from? In a closed system completely filled with water, you cannot create a vacuum, or send water where water already exists.

SWEI

close

The air pressure on one side of the bladder is constant.  The water pressure on the other side of the bladder is equal and therefore also constant.  Any pressure added by a circulator is reduced by friction in pipes, valves and emitters.  The sum is ALWAYS zero, but since it started out at the expansion tank pressure, it returns to that same pressure.  By locating the expansion tank at the lowest pressure point in the system (the circulator inlet) the pressure at any given point in the system is always greater than that in the tank, and never negative.  This prevents both cavitation in pumps and air ingress.

RJ_4

drawing

This may help

Paul48

Expansion

Take the expansion out of the equation for now. Let's just take a closed loop of cold water. The expansion tank manufacturers say to adjust the air press in the tank to match system requirements. Let's say 12 psi  for arguement sake. You put 12 lbs of air in the tank, and pressurize the system with water to 12 psi. We have created no deflection of the diaphragm because we have 12 lbs on either side of it.It would seem the only area on no pressure change would be on the air side of the diaphragm, and would have no affect on the system.

bob_46

PONPC

The best source is probably Dan's book Pumping Away available at this site.

Read this link for a start.



http://documentlibrary.xylemappliedwater.com/wp-content/blogs.dir/22/files/2012/07/fhd-501a.pdf

bill_105

Classic!

Aren't you glad you asked??

28W

Great information, everybody!

Thanks so much for all the patient advice.  I'm just a guy who wants to understand how his heating system operates, and I really appreciate the information you have all provided.

I think I actually understand this PONPC concept now . . . at least at a very basic level.

icesailor

Two PONC's:

I see it differently.

There are two PONC's, one somewhere inside the pump housing in the impeller and the other, somewhere in the middle of the piping system. I see a closed loop pumped HW system as a Ferris wheel on its side, or a bicycle wheel on it's side. It doesn't take much energy to spin either one, whether it is horizontal of vertical. When the pump is off, the pressure is equal on any place of the wheel. If a motor is connected to the Ferris Wheel, with a motor and the motor is connected with a chain, and the motor is off, the chain is slack on the top and bottom. When the motor starts, the chain is slack on the bottom and tight on the top. A bicycle chain is the same, loose on the bottom and tight on the top when you pedal.

You can buy 1/8" tapped flanges for some circulators. If you put two gauges on the flanges, on the inlet and outlet of the pump, when the pump is off, the pressure will be equal on either side of the pump. If the pressure is at 12# in the system, and you turn on the pump, and the outlet pressure rises to 15#, the other side, inlet, should be 9#. That gives you 6# of pressure or 13.86 ' of head pressure. The only place that the pressure could be 12# is somewhere inside the pump. The impeller. Look on the pump curve chart and it will give you the GPM that the pump is pumping. Find the Delta T and it will give you how many BTU's are going into the system or loop.

Because the pressure going into the system with the pump running is 15#, and returning to the pump it is 9#, there has to be another place in the system where the pressure is equal to what is in the pump. Mother Nature approves of things in balance. If you go to the seashore at high tide, somewhere on the other side of the earth, it is high tide. Half way from there, it is low tide.

When the pump is off, the pressure is in balance. The pressure will never go into negative because of the outlet pressure overcoming the resistance of the piping system. Only when you put high head and high flow pumps into a system to try and overcome a badly designed system will the pump ever "see" a negative pressure.

Inside the pump, the impeller "sees" all kinds of different things. There are vacuums there around and inside the actual impeller where the heat of the fluid will cause cavitation or the creating of steam. But cavitation will occur whenever a fluid, a gas or liquid, will create cavitation from turbulence. The higher the flow rate, the higher the chance of cavitation.

The expansion/bladder tank must always be on the outlet side because it will always see the resting or static system pressure and also the pumping high side pressure.

Because my old late boss was a steamer by heart, we always put fills on the bottom of boilers where I still do today. We always put the circulators on the return with the circulators pumping into the boiler and pumping into the fill valve on the bottom. If you set the bladder tank to whatever you wanted at or above the fill valve pressure, you never got a negative fill pressure. And overfilling causing relief valve blowing.

And if you fill through the bottom of the boiler, you pushed hot water through the system when purging and you always knew when the water is around. When you fill through the supply extrol, you feed the system first and there is no way to know if the water really went around.

That's my story and I'm stickin' to it.

How come they are requiring primary pumps on Gianonni type HX's to be pumped INTO the HX and not the OUTLET where some want to put all circulators so they can "pump away"? So all the energy of the pump will be PUSHING into the HX and control negative spots where cavitation can set itself up.

Mark Eatherton

THE most mis-understood principle of hydronic heating...

Back when I was actively teaching at Red Rocks Community College, I would spend one whole 4 hour class explaining the principles of this theory to the students, and at the end of the night, there were STILL a whole lot more inquisitive, blank stares than there were light bulbs above the heads of the students. Those who got it were smiling, and those who didn't were frowning. Bottom line was, regardless of wether or not you understand the principles involved, just know that ALL pumps SHOULD pump away from the PONPC in order for the to produce POSITIVE pressure differentials.



I also told them that if they would follow this basic principle, that their lives, their customers lives and the lives of anyone connected to the heating system would be better. Seriously.



I've also had students come up to me and tell me that their grandfather, their father and themselves had always pumped towards to the PONPC and didn't have any issues. Really? You don't consider having to go back to the same job EVERY year to bleed air out of the system a "problem"? You have nothing better to do with your time?



And I've had these same hard core people come back to me at a later time in life, after having adopted the principle of pumping away, and tell me that I was correct. They switched out some older systems and hadn't been back to purge air since then, the customer quit complaining about strange system sounds, like Niagra Falls in the walls, etc.



Simply stated, centrifugal circulators/pumps are nothing more than pressure differential machines. They have one function in life, and that is to make a difference in pressure, which causes fluid to move through the system. If these pressure differential machines are only allowed to produce POSITIVE pressures, then micro bubbles are kept small due to positive pressurization, and are pushed back to the part of the system that traps and expels them or recovers them and stores them where they won't cause any problems (compression tanks.) If the circulator is pumping away from the expansion tank connection, then it has NO CHOICE but to present its pressure differential as all positive pressure, thereby keeping air in suspension issues at bay.



If the pump is pumping towards the PONPC, then it has no choice but to present its pressure differential as a NEGATIVE pressure, which pulls air out of suspension and causes micro bubbles to balloon to the size of golf balls, causing noise and air binding conditions.



If there is no PONPC (G.S. H.P. loop for example) half of it potential differential is presented in a positive form, and half in a negative form, for its whole differential potential.



In reality, all of these problems have to do with the pressure generation potential of the pump, and pressure drop within the system. If it is a low head pump and a low pressure drop design (converted gravity) the location of the PONPC is less critical. If it is a high head circulator (fairly common today in wet rotor pumps) and the pressure drop of the system is high (typical of radiant floor or home run systems) then it is problematic.



Will a system work if the expansion tank is not in the ideal location? Obviously, it WILL produce heat, but heat is just ONE component of comfort. If the system flow is noisy, and the consumer is hearing it and thinking about it, then they are not comfortable.



If you want a trouble free, maintenance (relatively) free system that is virtually silent in operation and delivers excellent comfort, then you MUST have all circulators/pumps pumping away from the PONPC.



And we didn't even get in to people trying to rectify air problems by throwing automatic air vents on the upper floors of an improperly pumped system and the problems associated with THAT....



Hopefully now you (and others reading this thread) know more about the PONPC than they really wanted to in a simplistic manner. I avoid Henry's Law, and Boyles Law and hopefully made my point ;-)



Just remember, regardless of wether or not you "get it", if they would follow this basic principle, that their lives, their customers lives and the lives of anyone connected to the heating system would be better for their effort.



Happy Educated New Year.



ME

Paul48

ME

How does a diaphragm type expansion tank compensate for a negative differential, when it can't, under all conditions. If some expansion has occured in the system, the tank becomes an alternate source of pressure, and could smooth out differential across the circulator.Cold, with equal pressure on either side of the diaphragm, there is no energy potential. Merely having a PONPC does not compensate for the need for a pressure change. The explainations are sometimes long, sometimes loud, but they all  seem to turn to "because I said so", at some point.

Mark Eatherton

Flexed diaphragm...

In order for the PONPC to be present, the diaphragm MUST be flexed such that there is some elasticity in the diaphragm. A dead diaphragm is the same as no PONPC. Hence the reason for the recommendation that fill pressure be slightly above the diaphragm pressure. In the real world, if the system has cooled below the temperature it was initialized at, the PONPC may not be where it should be, but as soon as the water starts heating and expands and flexes the diaphragm, the PONPC goes to where it is supposed to be and everything works as it is supposed to.



In reality, the PONPC is THAT place where the atmospheric pressure interfaces with the system fluid, be it a natural place (open expansion tank at the top of the system) or a man made atmosphere (diaphragmatic tank or compression tank).





Gil Carlson is the person who "discovered" the PONPC, and as he stated, he just "knew" it was there, and did research to confirm it and its influence on closed loop heating systems.



You can't make this stuff up....



It is a tough one to see in ones minds eye...



As for "Pumping Away" from the boiler, that is not whats important. Pumping away from the PONPC IS what's important. In Dan's book on pumping away, his presumption is that the expansion tank was historically connected to the boiler, hence the need to pump away from the boiler. There are some other issues associated with pumping towards a heat source, but the older low pressure drop boilers weren't as susceptible to those issues as todays boilers are.



Some people still don't get it...



Hope that helps.



ME

Paul48

Recommendations

That's my argument....The manufacturers recommend matching the air pressure to the required system pressure. It is a dead diaphragm, waiting for expansion.I understand the idea, and it makes sense. But for me, it doesn't work 100% of the time, and that's a problem.

Mark Eatherton

Chris... (Ice Sailor)

Oy, after all these years, it is evident that you still don't get it, and now you are spouting information on a public form that is flat plain ands simply WRONG, and WILL lead to trouble. Let me deconstruct your statements. You can think and say whatever you want to think and or say, but when you publicly tell people something that is completely off base, I feel a need to step in and correct you and hopefully keep other people from making a mistake that WILL cause problems.



You stated "There are two PONC's, one somewhere inside the pump housing in the impeller and the other, somewhere in the middle of the piping system." I won't argue with this statement, because I have seen some systems where there were literally two different expansion tanks on the system, one on top and one on the bottom. The pump was on the bottom, and it was obviously splitting its pressure differential half way between the two PONPC's. I could make the pump do what I wanted it to by isolating either one or both of the tanks and confirm the theory of the PONPC and its influence on pumps.



You stated "I see a closed loop pumped HW system as a Ferris wheel on its side, or a bicycle wheel on it's side. It doesn't take much energy to spin either one, whether it is horizontal of vertical. When the pump is off, the pressure is equal on any place of the wheel. If a motor is connected to the Ferris Wheel, with a motor and the motor is connected with a chain, and the motor is off, the chain is slack on the top and bottom. When the motor starts, the chain is slack on the bottom and tight on the top. A bicycle chain is the same, loose on the bottom and tight on the top when you pedal." I think you are confusing the explanation of a true pump, like an immersed deep well pump with a closed loop CIRCULATOR. In a true closed loop, the weight of the water being lifted is countered by the weight of the water falling via gravity on the other side of the system. Big difference between a "pump" which must overcome atmospheric pressure and a circulator, which is simply creating a pressure differential. The two should not be confused. Just because a circulator comes with a "feet of head" rating doesn't mean that it is intended to "lift" water. And in reality, I don't think it has anything to do with explaining the PONPC. You're just confusing the issue.



You stated "You can buy 1/8" tapped flanges for some circulators. If you put two gauges on the flanges, on the inlet and outlet of the pump, when the pump is off, the pressure will be equal on either side of the pump. If the pressure is at 12# in the system, and you turn on the pump, and the outlet pressure rises to 15#, the other side, inlet, should be 9#. That gives you 6# of pressure or 13.86 ' of head pressure. The only place that the pressure could be 12# is somewhere inside the pump. The impeller. Look on the pump curve chart and it will give you the GPM that the pump is pumping. Find the Delta T and it will give you how many BTU's are going into the system or loop."



Have you ever actually done this? I have, in a laboratory setting, and it proved the PONPC theory exactly as I have laid it out. If you are pumping towards the PONPC, the circulator presents its pressure differential as ALL NEGATIVE pressure. So in using your scenario, if the system static fill is 12 PSI, and the pump's pressure generation potential is 6 PSI, then the gauge on the outlet of the pump will read 12, and the inlet will read 6. (12 - 6 = 6) It subtracted its pressure potential from the static fill pressure.



Conversely, if your pump (same one) is pump away from the PONPC, the inlet gauge will read the static fill pressure (12 PSI) and the outlet gauge will read 18 PSI (12 + 6 =18). I have seen it with my own eyes, and in fact built a demonstrator for use in court to show jurors the effect of the PONPC and the pumps performance in relationship to the PONPC that caused the other side to drop that portion of their claim in court. I was actually able to change the PONPC onto either side of the pump, and show exactly what influence it had in real time... VERY graphic.



Your explanation of using the pressure differential converted to feet of head and then charting where on the performance curve the system is operating is a good one, based on the pumps overall pressure differential. The REAL problem has to do with what happens to air in suspension in the water. If it sees a drop in pressure (opening a bottle of seltzer water) the bubbles EXPLODE. If kept under pressure, they remain small, easily moved and removed or recovered. Hence the need to make certain that all pressure differential created by the pump/circulator is all positive, and the ONLY way to guarantee that is to insure that the pumps is pumping AWAY from the PONPC.



You stated, "Because the pressure going into the system with the pump running is 15#, and returning to the pump it is 9#, there has to be another place in the system where the pressure is equal to what is in the pump. Mother Nature approves of things in balance."



I won't argue with that, but it has nothing to do with the PONPC, its location with respect to the "pressure differential" machine, or pump or circulator, or whatever you want to call it. If the PONPC is at the boiler, and the pump is half way around the loop, the pump will split its pressure differential half way between the supply and the return. Does this mean there are two PONPC's in the system? No, it means that the pump is not in the right place, as close to the PONPC as is humanly possible, hence it presents the system with half positive pressure and half negative pressure. The further away from the PONPC you move the pump, the more the differential shifts between positive and negative pressure production. The overall pressure differential doesn't change, only the way it is "presented" to the system fluid.



You said "When the pump is off, the pressure is in balance. The pressure will never go into negative because of the outlet pressure overcoming the resistance of the piping system. Only when you put high head and high flow pumps into a system to try and overcome a badly designed system will the pump ever "see" a negative pressure."



This statement is partially correct, and you have recognized the results of improper pump selection, but I have seen systems of multiple stories that had had automatic air vents applied in an effort to get a handle on an air problem, without addressing the PONPC issue, and the auto air vent becomes a VACUUM breaker, and WILL dump air into the system when the high head circulator, pumping towards the PONPC turns on, compounding the air problem. Seen it too many times to recall...



You said "Inside the pump, the impeller "sees" all kinds of different things. There are vacuums there around and inside the actual impeller where the heat of the fluid will cause cavitation or the creating of steam. But cavitation will occur whenever a fluid, a gas or liquid, will create cavitation from turbulence. The higher the flow rate, the higher the chance of cavitation." You are correct for the most part, however, every pump manufacturer generally lists the minimum net positive suction head requirements for their pumps. This is to avoid those cavitation issues which you observed. If you ignore the NPSH requirements, you can expect problems and the pump will NOT perform to its performance curve. As John Siegenthaler joking said, NPSH, if not properly applied and understood, means "Not Pumping So Hot"... Details, details, details...



You said "The expansion/bladder tank must always be on the outlet side because it will always see the resting or static system pressure and also the pumping high side pressure." WRONG WRONG WRONG WRONG WRONG....



You are outside, standing alone, against every pump manufacturer if this is the theory you wish to adopt and promote. And if you want to do so, please feel free to do so, but you can expect me and others to jump on your case with this statement. Simply stated, this statement is W R O N G. No other way to say it. In order for a circulator to produce all positive pressure, and avoid issues of air in suspension, ALL pumps MUST pump away from the Point Of No Pressure Change, also known as the expansion tank connection, REGARDLESS of the type of expansion tank used (open closed, diaphragmatic, bladder type etc) If you WANT air problems, then do it your way. Guaranteed problematic.



You said "Because my old late boss was a steamer by heart, we always put fills on the bottom of boilers where I still do today. We always put the circulators on the return with the circulators pumping into the boiler and pumping into the fill valve on the bottom. If you set the bladder tank to whatever you wanted at or above the fill valve pressure, you never got a negative fill pressure. And overfilling causing relief valve blowing. "



The reason this works for you, is because the PRV fill valve is NOT seeing the negative pressure that the circulator is creating. If it could see it, it WOULD (and will and does in many cases) cause system pressure overflow problems. You have proven through this statement alone that the PONPC DOES exist, and works exactly as it is supposed to. The pump is pumping towards the PONPC, and hence can NOT influence the pressure at the PONCP, hence it remains stable at whatever pressure the system will filled to.



If the make up connection is connected to the inlet side of the pump, and the pump IS pumping towards the PONPC, then you WILL have a relief valve popping off on a regular basis.



Lastly, you stated "How come they are requiring primary pumps on Gianonni type HX's to be pumped INTO the HX and not the OUTLET where some want to put all circulators so they can "pump away"? So all the energy of the pump will be PUSHING into the HX and control negative spots where cavitation can set itself up."



I think you are getting confused by Mr Holohans book "Pumping Away". When Dan wrote that book, Gianonnis were just being developed. At the time the book was written, the expansion tank was connected to the boiler, where the temperature was the greatest, and the generated pressure was the lowest, hence the ideal spot for air to come out of solution. This is the reason he recommended pumping away from the boiler. It has to do with the expansion tank connection, not the heat source. The manufacturers WANT to see as much pressure as possible at their heat (Gianonni)exchangers to avoid possible steam flash occurring at the fluid to heat exchanger interface, and as we all know, water boils at higher temperatures with higher water pressures.



As I said before, you can do as you want, and believe in whatever theory satisfies your minds eye, but please don't tell people to put their expansion tanks connection at the outlet of the pumps. It is WRONG.



Respectfully,



ME

28W

I found this article very helpful

As a total amateur/homeowner, I found this article very helpful.  All the pros on this forum have probably read it already.

http://www.pmmag.com/Articles/Column/BNP_GUID_9-5-2006_A_10000000000000657694

Mark Eatherton

Its only a problem...

if it IS a problem.



So long as the circulator is NOT capable of generating pressure differentials that are in excess of the systems static fill pressure, then it won't be a problem, because as soon as the boiler fires, water starts expanding, the PONPC comes into "view" of the circulators, and everything falls into place. The PONPC doesn't stop the pump from generating a pressure differential, it only determines wether it is positive, negative or half and half. When the pump first turns on, water starts moving pretty much immediately. And the true PONPC is established shortly thereafter if not already established.



It (PONPC) is real, and its affects on circulators/pumps is well known and documented.



Trust me when I tell you that this has been field tested by hardened "Old School" hydronic's contractors who use to do it the other way, and have proven to themselves that it DOES in fact make a difference.



The primary reason that pumps ended up on the returns of boilers (old school) was because it was the most convenient point to ship the boiler pump mounted to keep it from bouncing around, and in the "good old days" when pumps first came out, they used a rag seal for the impeller shaft, and its life expectancy was much greater with lower fluid operating temperatures.



Some boiler manufacturers still show the pump on the return as an "alternate" location, just to keep these good ol' boys happy, but if questioned, they will tell you that it is best to place it so it pumps away from the PONPC.



Honestly, if the pump is a low head pump (like a series 100 pump) and the boiler is a low pressure drop boiler (like the good old cast iron boilers still out there) there is little to no potential for trouble in the field. It's when people start using high head circulators (like the gazillions of wet rotor circulators) and have system with high pressure drop (take your pick) that it really becomes a problem. This is why I get beat up by people who say "My grandpa, my dad and I have been doing it this way for over 100 years, and haven't had any problems..." Except for having to go back every year to bleed air from the system... which when they adopt this method, that problem goes away.



HEARING your heating system running is NOT in my definition of comfort...



HTH



ME

28W

Diaphragm

So having the system fill pressure a little bit higher than the tank (diaphragm) pressure creates a "spring" effect pushing down on the diaphragm?

Mark Eatherton

Yup...

It establishes THAT point where the atmospheric pressure has an influence on fluid pressure within the system. Only needs to be a couple of pounds difference. If diaphragm is charged to 12, then fill to 14. Or, lower diaphragm to 10.



If you have John Siegnthalers Hydronic Design Suite software, it tells you EXACTLY where the diaphragm needs to be air up to in order to guarantee proper operation, and its not always 12 PSI (SHOCK, GASP the HORROR...) ;-)



Something as simple as a large bubble of air trapped in the upper reaches of the system CAN and WILL confuse the system circulators. They can "see" both of these points where atmospheric pressure resides...



It's not Rocket Science, its WORSE. And its all dictated by Mother Nature... The laws of science simply try and explain it.



Our jobs are to either work with Mother Nature (PONPC for example), or work against her (counter her wishes for thermal equalibrium for example) to deliver comfort. He who understands and utilizes the control of Mean Radiant Temperature will get there (comfort) first...



ME

ChrisJ

Expansion tank?

Only thing I wish to add is,



Gentlemen, its a compression tank we are talking about here, not an expansion tank.



I started reading Dan's "Classic Hydronics" and what can I say, I love reading Dan's books.  Now I know what a compression tank is and I don't even own one.

JStar

Pressure

I feel a lot of pressure in this thread!



The only way to understand this concept is to test it. Or just listen to Mark.

28W

Interesting

I just looked at my boiler's pressure gauge for the first time.  That's not true . . . I've looked at it many times, but never paid any attention.  Now I notice that it is at exactly 14#.  I have no way of knowing for sure, but I'll bet the tank is at 12# and my installer did exactly what you mention.

bill_105

Again...

Aren't you glad you asked??

I swear these guys (Chris and Mark) are related to Tolstoy!!

Paul48

Search

for compression tanks leads you to wife-beater tee shirts.Apparently the manufacturers are equally confused, although the may wear wife-beaters. I don't know! Can you explain the clarification?

ChrisJ

expansion tank vs compression tank

Expansion tanks according to Dan's book and assuming I'm remembering it correctly are mounted in the attic and are vented.  They relied 100% on gravity.



Compression tanks have a bladder and an air charge on one side of the bladder.



I just noticed Mark also explained this in his responses above.

ChrisJ

Why a tank

My understanding, and keep in mind I have never touched a hot water heating system, is this.



When you change the temperature of water it either expands or shrinks.  If you filled the system with 200F water and let it cool to say 50F and had no way for air to get in, the pipes would implode.  If you filled it with 50F water and heated it to 200F and had no way to vent extra water, pipes would burst.



The tanks primary function is to make up for this and cars also need expansion tanks on their cooling systems, though they work differently.

The reason to have the tank on the input of the pump is when the pump fires up it temporarily pulls a small amount of water from the tank which creates pressure on the output side of the pump until the rest of the system catches up.  This stops the pump from creating a slight vacuum on the suction side and possibly pulling air in while giving a slight boost in pressure on the other side which helps push air out of the system.



Like I said, never touched a hot water system yet, but this is my understanding.

icesailor

PONC's

Mark,

Where in heavens name did I ever say that a PONC doesn't exist? I never did.

Words and definitions count. Where I see the term "Pumping Away", I think of the pump downstream and the force of the pump is into the system. As in, the circulator on the supply is before the extrol fitting rather than after it. If it is AFTER the fitting, it is pumping AWAY but the pressure on the extrol would be lower than if it was before it. If it is before the fitting, it is pumping INTO but is still "pumping away"?

When it goes into the main piping, regardless where it is located, there is friction loss from resistance to the piping. The longer and farther the pump has to move the fluid, the more the resistance comes and the pressure drops until finally, there is little or no pressure. But the pump adds pressure in the form of lesser pressure coming back. It keeps the pressure from ever going to less than atmospheric. As you said, and also said that a badly engineered pump will cause the system to go negative. There is absolutely no disagreement about that. Which brings up one of my points. I see people looking for pumps to "push" water up to 50' above the boiler because that is the fifth floor. You and I both know that the pump does not do that. Once filled by the pressure reducing fill valve, the system is a closed loop and it becomes the vertical Ferris wheel or bicycle wheel. And it doesn't take a pump that has the ability to push it up 50' like you would with a domestic water pump. Where the water coming out of the faucet at 50' is lost down the drain. The hydronic loop has the gift of the water going over the top (50') and adding weight to balance out the column on the other side.  I consider all of these "machines" as  Mini Atmospheres and Environments. Open gravity HW systems depend on the weight of the different water temperatures to achieve circulation. You don't need a pump to do it. A pump will speed up the process.

As far as the point of two PONC's, if you take a standard "Jet Pump" that has a nozzle and venturi to make it pump, and you stick the front end (Suction) into water that is 10' below the impeller, and another pipe on the outlet that goes 20' in the air, fill the pump with water and turn it on. Once the impeller/elector removes enough air from the suction pipe, atmospheric pressure pushes the water up the pipe into the pump. The water goes up the 20' pipe and overflows. Where is the PONC? There is only one, inside the pump, inside the impeller. It is not in the injector for a reason. Because it is open, there is no PONC between the suction side of the pump and the discharge because the atmosphere breaks it. If you connected the suction and the discharge together, it would be a closed loop. Somewhere in the between, there must be a PONC. Agreed?

When a circulator or water pump stops, everything becomes static. When it starts, it develops inertia by pushing whatever it needs to push around. Whatever is used to connect the driving force and the driven force goes back and forth. My example of the chain drive or bicycle chain. The same applies to drive belts. Or tree piece circulator spring couplers.

At my advanced age and early dementia, I can't remember all you said but I have to say that although I agree with everything you have said and say, you do not see things as I do.

You asked me about the circulator with the tapped flanges with gauges and had I ever done that. The answer is yes. I have an account that is P/S with circuit setters and gauges. After reading and not completely understanding all the IBR lessons on pumps and such in their 200 piping guide. it became clear as the fog lifting on how it worked when you put the differential gauges between the pump and system. How you can control and balance the flow through the circuit. The static pressure equals zero? The flowing differential pressure equals the head pressure? The pump curve head pressure equals how many GPM's are flowing through the circuit? The Delta T equals how many BTU's are going through and being delivered to the emitters?

But again, we seem to keep getting caught up on what is "negative pressure". What is that to you? I have never seen a compound pressure/vacuum gauge go to negative, less than zero in a heating system unless there was something seriously wrong with it. In a heating system, do we agree that "negative" is less than static? But still above zero?

Where "negative pressures come to be a problem is in higher temperature closed systems where the water can go over the boiling point. Like water boils at 244 degrees at 12# PSIG. Vapor/Vacuum steam systems work because the pressure goes negative and the water turns to steam at less than 212 degrees.

Your example of the gauge showing static with the other side showing 18# I would have to see. It could happen where you have a restriction in the system that allows the pump to push into the restriction but the flow back is equalized. 

On the same account, there was a 2000 gal. oil UST. It had pumps to pull the oil out of the tank. It never ran with much more than 3" of vacuum pressure because it had a big gauge that you could see. There was no filters on the primary pump side. I added two spin-ons to filter the oil. I have posted photos of what the filters looked like. The tank was removed and an above ground was installed by the low bid contractor. They ran 5/8 OD copper back and forth in 3" flex conduit. The lines are heavily kinked. The static pressure is one or 2, the level is now above the pumps. The restriction equals 15" and one filter canister is a vacuum chamber where only one filter half works. The installer says that this is how it is supposed to be. 15" of suction is normal for a recirculating oil pump. The return is 2". I have that gauged. Where's the restriction?

jumper

expansion is required

Room for thermal expansion is required. Bladder,vented, or conventional. Cushion and pressure are secondary if at all. I can argue that you should put the bladder tank at highest pressure (just past pump) because then expansion will increase pressure the least as a fraction of cold pressure. Think about that !



I'm old enough to remember when Amtrol began this bladder business. Then the supposed advantage was that you separate air from water. No more water logged tanks, ha ha, at least till the bladder rots.



The real issue is air removal. Here in southern California rich people build slab on grade with radiant. The so called specialists seem to not know what we knew sixty years ago and so they have to come back several times.



Another thing to think about. Gravity is cheaper and more reliable than pressure tank.

RJ_4

tanks

You have your compression/diaphram tanks all mixed up

Paul48

Ahhh

Sort of like circulator vs pump. Given the context, everyone knows what you are talking about.Taco will direct you to their pump curves for "00" series circulators. We talk about "Pumping Away", even though we are circulating. Just to avoid confusion.

Jamie Hall

I told Mark he'd lose people...

In a very real sense it is correct to say that the tank -- bladder, diaphragm, old fashioned air, open at the top of the house -- is there to allow for expansion and contraction as the temperature of the water changes.  As ChrsJ said.  Among others...



If it is considered solely on those grounds, it doesn't matter where you put it.  Anywhere on the system will do just fine.



However, that isn't the end of the story.



First, you want to make sure that there is no point in the system where the water pressure inside the system is less than the air pressure outside the system.  That point of lowest pressure will always -- always -- be at the intake of the pump.  It may, depending on the exact details of the pump's contstruction, be significantly less than the pressure in the intake pipe to the pump, oddly enough.  Second, that lowest pressure must be above the "vapour pressure" of the water at the corresponding temperature.  If it isn't, the water will actually boil (really).  This does two nasty things.  First, it is noisy.  Second, it can actually destroy the pump impellor in a remarkably short time.



Doesn't it just make sense, then, that you also put the tank at the pump input, to ensure that the pressure at the pump intake is above the pressure of the air outside the system?  And that the pressure everywhere else in the system is at higher than the pressure in the tank?

nicholas bonham-carter

Old gravity system

Checking out our one vacant building with a 1900 gravity hot water system (converted in 1955 to a pumped system). I noticed the pump is on the return side pumping into the boiler, and the supply side has the connection to the expansion tank (air-cushion, no bladder). So it is not pumping away.

If they had put the tank connection in front of the pump, any air bubbles would be trapped in the tank.

If I had to replace the boiler, would there be an absolute necessity to change the tank for a bladder type? I could just make sure the new setup is pumping away, and the tank would handle the air removal. The tank is drained and refilled every fall as part of the annual service.--NBC

icesailor

PONPC Agreements:

ME,

You have just agreed with every point I have made in this misunderstanding.

I have always set the diaphragm tank pressure higher than the system fill pressure. The only reason I put vents on emitters is if they can't be purged and there is no other way to get the air out. Other than that, if they are up at a high level above the system, and the caps aren't tight, and the fill valve stops maintaining pressure, they suck air. Float vents around boilers weep water and then the water turns to powder that stops the seat from working. If you read the instructions for diaphragm tanks, I think that 90% of them are undersized for possible conditions. Especially on cold start systems. I see cast iron Watts 1156F fill valves stuck closed and the system actually only has  5# pressure in it but vacuum pressure keeps sucking air in from high vents. Hit the fast fill and the pressure goes up to something. Most gauges I see on boilers are off their asses. So how do you know? ( what the actually pressure is)

Then, there are the systems that to save money, they use smaller pipes but go over the maximum circuit length. Why would you run a 1/2" ID, 200' circuit when there is 700' on the roll and I can run it out 700' and save and do it with 2 loops. Use a big huge high head circulator to overcome the resistance. And use 1/2" OD tube.

Many are clueless about flow. Put your hand on a big Airtherm cabinet heater. Turn the fan off. Put your hand on the outlet/return. It's hot. Turn the fan on low, feel it cool. Turn it to high. It might get cold. The flow didn't change, just the fan sucking more heat out of the fluid. The connections to an air handler are usually 1/2". But, because they are only 1/2", you don't run 1/2" pipe to it when it needs 3/4". and you run it right to the connection with the required pipe size. Growing up in Hollywood in the late 50's, early 60's, and being a motor-head, we had a saying about building fast cars. "If it don't go, chrome it". For the under piped heating system, "If it won't flow, put a bigger circulator in it".  Another, "It's hard to put in a circulator that's too big". 

That is the way and the ONLY way that you can pull negative pressures in a heating system on the inlet or suction side of a circulator. The air comes from the fact that the boiling point of the water drops so much. If you put a circulator on the supply side of a Gianonni type high resistance boiler, you WILL get air bubbles if it is a really high head circulator and isn't primary secondary piped. The secondary with the rocket powered high head circulator won't care.

If you get a high pressure differential pressure across a circulator, and the discharge is a high pressure (what is being pushed against the resistance), the system is piped badly and you are using a band aid (the pump) on an infection,the undersized system.

jumper

1900 system ?

Is the current tank still the high point ? Is it still open ?

Mark Eatherton

Not quite Chris J...

Water is incompressible. If it "takes" water FROM the expansion tank, it has to PUT it somewhere. IF there IS some place to put it, i.e. a bubble trapped in the upper portions of the system, then you have other problems to address. In a well design, properly commissioned system there will not be any air anywhere in the piping system, hence the system can not TAKE water from the expansion tank, because there is no place to PUT it.



As I previously stated, the pump is nothing more than a pressure differential machine. It does this through centrifugal force, slinging water to the outside of its impeller from the eye of the impeller. If you run the pump backwards, it WILL move water, but not at the same rate, and it will be VERY noisy...



Rocket science, I'm telling ya... Rocket science ;-)



ME

Jack

Dear Sir, Our Original Poster

I am not even going to attempt to get into this. This is the Third Rail of Hydronics...regardless of what Dan says;) Having just come on this evening and seen the time of original post and number of responses from guys I look to for advise, you get the award for the, The Question of the Year. Had you asked on 12/31 you would have gotten the Question of 2012 too! Well done!

Steve Thompson (Taco)

Pumps vs Circulators and Pressure vs Thermal Tanks

Sorry guys, couldn't resist...



Pumps boost pressure (aka submersible well pumps, pressure booster pumps, irrigation pumps, boiler feed pumps etc).



Circs (regardless of HP) provide flow to move the BTU train hopefully at the right speed to provide heat (delta T) in a comfortable and efficient manner by overcoming system friction loss in a closed loop system (whew, that was a mouthful).



Some/all (depending on the tank and circ size, air and system pressures etc) of the differential pressure a circ provides will be absorbed by the expansion tank instead of overcoming friction loss if the tank is installed after the pump.  This potentially can cause premature tank bladder failure, lack of flow (comfort) in some zones, boiler and circ cycling, difficulty removing air, excessive make-up water and premature relief valve failures, internal corrosion, just to name a few.



If in doubt think about the differences between a closed loop heating thermal expansion tank (absorbs thermal pressure change as the system heats and cools) and an open system potable water pressure tank (a pressure storage battery to reduce pump cycling).  Pump into the pressure tank with a pump and away from the thermal with a circ.



Yes, water is NOT compressible and yes, pumping away from the thermal expansion tank is, in my opinion, one of the few "has to be with no it depends" in hydronics.  And yes, some low mass boilers actually require the circs to be installed on the inlet (but PUMP AWAY FROM THE THERMAL TANK).



All the best for 2013 to all of you - hope to see in Dallas at the AHR/ASHRAE show and our Taco party!