Pump to or From modcon

NYplumber

Hey HH,



Interesting question I thought of.

When piping a modcon pri-sec to a buffer tank, with a DHW connections being between the modcon and the buffer tank, should one pump into the boiler or out?



I prefer to pump out, however boiler i&o manuals illustrate to pump in.



State your opinion and reasoning behind it.

Steve Whitbeck

pumping

You follow the instruction manual - Or you pump towards the most resistance.

I usually pump toward the boiler because I use pumps with check valves and don't want to trap air inside the pump if it is installed on a vertical pipe coming up from the boiler. ( On the return the air will rise out of the pump to the main because the check valve is on the discharge side of the pump.

croydoncorgi

Do what they do / say

....bearing in mind that 'system' boilers (from Viessmann etc. ie. those with a pump built-in)  ALL have pumps on the Return side.

Jean-David Beyer

ALL have pumps on the Return side.

I read somewhere that they put the pumps on the return side of mod-con boilers that are piped primary-secondary so that the circulator does not experience the resistance of the heat exchanger at its input. I suppose they may mean the pressure drop.



But since, at least for my boiler, the point of no pressure change is not even in the primary (boiler) loop; it is in the secondary loop right before the circulators to the heating zones are pumping (away). Now perhaps where the two closely spaced Ts are in the primary loop could be considered the point of no pressure change for that loop. In that case, the circulator on the return to the boiler is pumping away from the point of no pressure change.

Gordy

Pressure switch

 If you pump INTO the mod/con it will be sure to activate the pressure switch since it will see the high pressure side of the circ. At least this is how I have come to understand the reasoning for the manufactures wanting it this way.



Gordy

NYplumber

good point

Thanks gordy for the reasoning. Makes sense.



The reason I ask is, pumping at the relief valve in taller residential housing with deep basements and radiant wall panels on the wall of the attic post a concern to me, gotto keep that wall rad pressurized.



The remedy i can think of is place the boiler on the first floor of the home.

Gordy

Throw this in the mix

    So according to manufactures reasoning for keeping the pressure switch happy (pumping into it). If you are pumping away from the PONPC in the primary loop (boiler loop) away from the boiler, and the PONPC is on the supply side in that loop of course. Then the pressure switch should never see less than your initial fill pressure correct? 



   With that being said is the manufactures requirement for the sake of someone putting the PONPC on the return side of the boiler, and the pump on the supply in the primary loop. Then the boiler would see a drop in the pressure differential maybe causing a fault in the pressure switch.





 I agree with what you are saying in that keeping the system pressure up to get to the top in certain situations, and worrying about popping the relief pumping into the boiler. Which is one reason why pumping away from the PONPC on the return side has its limitations.





Gordy

Jean-David Beyer

Pressure switch: none.

My W-M Ultra 3 has no pressure switch. So they cannot instruct that the boiler circulator be on the return side, pumping into the boiler to keep the non-existant pressure switch satisfied. Their controller acts as a low water detector, though not all inspectors accept that. I have an M&M probe type low water detector in addition.



The tridicator is tapped into the supply line after the heat exchanger, just before the pressure relief valve, and the pressure does not change if the boiler circulator is running or not. It does change if the indirect circulator runs by a coupla pounds. So my pressure is around 15 psi (cannot read the tiny dial all that well). But one zone is on the floor level, so no pressure other than the weight of the water is needed to get the water there. The other zone is floor-level baseboard perhaps 6 feet above the output of the boiler. So I do not need all that much pressure there.



Coulld the answer be, "There is no reason for it; it is just our policy"?

icesailor

Adequate Piping:

If the boiler and system piping is sized adequately and properly, and not undersized and over pumped, there should be no increase or decrease in system pressure. In fact, some don't really understand how centrifugal pumps work.

If you take a bucket and fill it with water, and spin it around while holding it in your hands and spinning in a circle, the water will stay in the bucket. Gravity will hold it there until you stop. If you hold the opening of the bucket, horizontal, all the water will run out. Gravity holds the water in the bucket. If you drill a hole in the bottom of the bucket, and spin around, the faster you spin, the faster the water will run out of the hole and with higher pressure. The impellor is the bucket. If you pump a 007 circulator into a 50' loop of pipe, and put a gauge on either side of the pump, you wouldn't see any rise on pressure. If you put it on a 50' loop of 1/2" pipe, you would see a rise in pressure from restriction. Gravity systems had no circulators, no restriction and no pumps. It depended on a 7.5 degree rise in water temperature to flow and the pipes were sized to give the GPM flow through the large volume in the pipe. Maybe you could find a PONC with a manometer but you would have to look hard. If you can find a PONC in a system today, that is affecting the system, it is either over-pumped, undersized, or both. The only true PONC is in the volute of the pump, somewhere.  

The freaking circulator has absolutely nothing to do with pushing water to a level that is above the highest heat emitter or vent in the system. That is the job of the Pressure Reducing/Fill Valve. If you are depending on the circulator pump to get the water up there, the system doesn't have enough pressure.

The only time that circulators are designed to add pressure to a system is with an ejector water pump. Ejector well pumps are circulators, circulating water.

Gordy

I follow you Ice

Then I don't,



"If the boiler and system piping is sized adequately and properly, and not undersized and over pumped, there should be no increase or decrease in system pressure.



  Your speaking of a perfect world in piping. Even gravity systems have friction loss. very minimal but its there. for that matter in any closed loop system. You would need a very accurate pressure gauge to measure it, but its there.





 "If you pump a 007 circulator into a 50' loop of pipe, and put a gauge on either side of the pump, you wouldn't see any rise on pressure. If you put it on a 50' loop of 1/2" pipe, you would see a rise in pressure from restriction."



 Again there is friction loss there for a pressure differential. how many radiant systems use 1/2" piping.







 "Gravity systems had no circulators, no restriction and no pumps. It depended on a 7.5 degree rise in water temperature to flow and the pipes were sized to give the GPM flow through the large volume in the pipe. "







 Gravity systems had friction loss. They harnessed the friction loss through their pipe sizing. Yes to provide proper flow, but that pipe size for the flow they wanted also restricted flow through friction loss of the piping so they did not get more flow then they wanted. But then temperature differential, and gravity played the role of the circulator.









 "The freaking circulator has absolutely nothing to do with pushing water to a level that is above the highest heat emitter or vent in the system. That is the job of the Pressure Reducing/Fill Valve. If you are depending on the circulator pump to get the water up there, the system doesn't have enough pressure."





 Agreed there is static or system pressure,and there is pressure differential at the circ.



 I guess I get the impression from your post that if there is pressure differential at the circ then th system piping is poorly designed. If so then why do pump manufactures have pump charts, and different sized pumps.





Gordy



  







  

Gordy

JDB

 You would need to install a pressure gauge at the discharge, and return of the circ to measure its pressure differential, or pipe teed in on both sides with valves on either side of the pressure gauge. Even if it were only 1 psi thats 2.31 feet of head. But you need to use a 30psi pressure gauge to really see the small difference.



Like so.   

http://www.heatinghelp.com/files/posts/10844/PumpCurves.pdf

Gordy

Mike Kusiak_2

Optimum pipe sizing

In a properly designed system there will be a definite pressure differential across the pump. The best size and head loss is a compromise between oversizing the pump and oversizing the pipe.



Bell and Gossett goes into a lot of detail in their engineering manuals as to properly size the piping in a system. Their calculations assume a pressure loss of 0.4 inch of head for each foot of pipe or fitting equivalents as being about optimum. So for a loop of 100" this would result in a head loss of 40" ( 3.3' ) or about 1.4 PSI . This resistance to flow insures a flow rate of less than 4 feet per second to minimize noise, yet still provides reasonable piping and operating cost.



The B&G engineering manual is available here and really goes into a lot of design examples as far as calculation of pump and pipe sizing.



http://www.heatinghelp.com/article/123/Bell-Gossett-Handbook-Second-Edition-1949/955/Section-II



There are also a lot of nice graphs and tables for calculation of piping head loss, etc.

NYplumber

static fill

" If you are depending on the circulator pump to get the water up there, the system doesn't have enough pressure."



Icesalor, maybe you didn't understand. We aren't using the pump to keep the water in the highest rad pressurized, we are using static fill. However, with a high static fill, pumping at the boiler, makes it more likely that the relief valve will pass its threshold.

Jean-David Beyer

Huh?

"You would need to install a pressure gauge at the discharge, and return of the circ to measure its pressure differential,"



If I wanted to measure the differential, I would get a suitable gauge and pipe it across the circulator. But what I thought I was talking about was the change in pressure from running the circulator at the input to the boiler instead of the output. And I did not mean the pressure forcing the water around the loop. I meant the pressure experienced by the pressure relief valve. And to do that, you would want the pressure gauge as close to the relief valve as possible, and it could measure either absolute pressure or gauge pressure; gauge pressure seems to make the most sense. It would not be the differential pressure across the loop that opens the relief valve; it would be the gauge pressure, because that is what the relief valve measures.



And in my boiler, that does have the pump pumping the return water into the boiler, the supply water exits the heat exchanger and goes out the top of the boiler. Just before it exits, there is a well with a pressure and temperature sensor in it. Just after it exits, it hits a T with a street elbow in it and the relief valve goes in there. There may be a foot of 1" pipe between the pressure sensor and the relief valve. So all water exiting the boiler goes past that temperature and pressure sensor and next stop is the relief valve.



The thing that interests me is the boiler circulator is a Taco 007. The indirect circulator is also an 007. When the boiler circulator is running, there seems to be no pressure drop to the relief valve, but when the indirect runs, there is a pressure drop. I am not sure why that is. It may not matter. The pressure never goes up when anything runs.

icesailor

Static Fill;

NYP,

More than one person I have met thinks that a circulator is needed to get the water up to the top floor if there is an air problem.

Unless you have the equivelent to 58' of head above the boiler, the relief valve shouldn't ever be blowing off.

On a two stowy building, with the boiler in the basement, I figure 20' from the top of the boiler to the second floor baseboard. That's 30' of extra head to play with.  

icesailor

Lengthening

Again there is friction loss there for a pressure differential. how many radiant systems use 1/2" piping.



What's the longest length of radiant circuit you have come across in a system that wasn't working and installed by a "Green Contractor(Carpenter)?



I was taught way back when at a seminar that Dan did for Emerson Swan for Taco and Heatway that you should keep radiant loops at 200' or under. I've seen many over 700' How much restriction in 700'+ of 1/2" radiant tube"

NYplumber

Point

Hey ice, point taken and i agree with you. It's simple math to figure out pressure needed. However, the home in question will have towel racks, as well as wall radiant. To remedy the problem I put thought into keeping the static fill closer to the threshold and pumping away, however, with consideration of the flow switch needing pumping in, will moving the boiler up to the first floor give me the benefit of added pressure high up without sacrificing the relief valve?

Gordy

Okay

   But that is not what you were saying in your post Ice. You Still have friction loss in 200' loops, and I won't go there on the PONPC.



   Yes there are many people who think that radiant is as easy as throwing a bunch of tubing down with a circ, and  dare I say a water heater. Wha La radiant heat right. I have seen a guy who used in ground sprinkler zone valves on his pole barn system with a 40 gallon water heater. Wondered why it did not work. Ended up with LP heaters, and says radiant is not worth a crap. HE was screwed the minute the concrete got hard. No insulation, excessive loops etc. 

   I did not post to argue with you , but it is incorrect what you were saying Ice. There IS friction loss in any hydronic system, and there IS a PONPC that you should pump away from.



  The best you can do is minimize the friction loss through PROPER piping strategy so you can keep the circ as small as needed to do the job.

  Pump away from the ponpc.



  Static fill pressure, and pressure differential are two different things.



 Keep piping friction losses to a minimum.

Gordy

icesailor

Circuit Setters:

If I wanted to measure the differential, I would get a suitable gauge and pipe it across the circulator. But what I thought I was talking about was the change in pressure from running the circulator at the input to the boiler instead of the output. And I did not mean the pressure forcing the water around the loop. I meant the pressure experienced by the pressure relief valve. And to do that, you would want the pressure gauge as close to the relief valve as possible, and it could measure either absolute pressure or gauge pressure; gauge pressure seems to make the most sense. It would not be the differential pressure across the loop that opens the relief valve; it would be the gauge pressure, because that is what the relief valve measures.



Have you ever seen "Circuit Setters" on a large commercial Hydronic Job?

It a device for measuring the differential pressure in a circuit. You take the pressure and the pump curve, and adjust the balancing valves so that the designed GPM flow is running through the circuit. When all circuits are "set", the system, the system is supposed to be in balance and flowing as the engineer designed it to be.

There are people who are trained and certified to measure flows and set the systems. Once properly set, guys like you and I are not supposed to "adjust" the balancing valves.

Gordy

JDB

" But what I thought I was talking about was the change in pressure from running the circulator at the input to the boiler instead of the output. And I did not mean the pressure forcing the water around the loop. I meant the pressure experienced by the pressure relief valve."



Differential pressure is the pressure forcing the water around the system created by the circulator. This is the pressure that the relief valve will experience if the circ is on the return side of the boiler because the discharge side of the circ has more pressure than the inlet side of the circ

" It would not be the differential pressure across the loop that opens the relief valve; it would be the gauge pressure, because that is what the relief valve measure."





 Yes JDB, but the the added pressure of the discharge at the circ adds to the systems static pressure.



Example:If your static pressure is at 15 psi. The circ is pumping away from the PONPC, and there is 11.5 feet of head in your piping then the discharge of the circ would be at 20psi, and the inlet at 15psi because you are pumping away from the PONPC which is 15psi static pressure.

This is why it could be a problem if your static fill is say20-25psi, and the pump kicks on now exerting its discharge pressure bumping it up to 25-30psi



 If your PONPC is on the supply side, and the circ is on the return with the same above numbers then you would see 15psi on the discharge side of the circ, and 10 psi on the inlet side of the circ. But then this would not be pumping away.



The other scenario if there is a pressure switch, or sensor on the boiler. Is if you are pumping out of the boiler away from the PONPC, static fill is 15 psi, and using same numbers your inlet pressure at the circ is 10psi and if the pressure switch/sensor is calibrated to open at 12psi then you have a problem.

This is the premise of the discussion



"The thing that interests me is the boiler circulator is a Taco 007. The indirect circulator is also an 007. When the boiler circulator is running, there seems to be no pressure drop to the relief valve, but when the indirect runs, there is a pressure drop. I am not sure why that is. It may not matter. The pressure never goes up when anything runs."



 Its possible that the indirect may have more pressure drop then your system piping does.









.





Gordy

icesailor

Restricted arteries.

 Static fill pressure, and pressure differential are two different things.

I know that.



Keep piping friction losses to a minimum.

Gordy

I always pipe to the larger size to keep restriction down. The longer the small pipe, the greater the drop on flow.

I don't disagree with PONC, I just never saw it until mini-boilers were under sized in the piping and over pumped in the system.

I've said that many of the scorched air systems I see suffer from emphysema. I guess I should say that these under piped systems have atherosclerosis

Jean-David Beyer

I bet we could resolve this in 5 minutes face-to-face.

But in typing this we seem to be talking past each other.



"Yes JDB, but the the added pressure of the discharge at the circ adds to the systems static pressure."



I do not think so. The static pressure of the system changes only with a leak (out or in) unless the temperature changes and expansion tank too small or misfunctioning. Now dynamic pressure will change all around a loop if a circulator is running.  If a circulator is running, there are two kinds of pressure we can talk about. The desired pressure that pumps the water around the loop is one, and that is surely maximum at the exit of the circulator and minimum at the entry to the circulator. To measure that pressure, you need either two pressure gauges, one on each side of the circulator, or one gauge (pointing the right way) across the circulator.



Now depending on the relationship of the positions of the circulator and the relief valve. it might change the gauge pressure noticed by the relief valve. The extreme cases are if the valve is immediately after the output of the circulator (when the relief valve is most likely to open) and immediately before the input to the circulator (when the relief valve is least likely to open). I could imagine (an unlikely case) where the relief valve is at the input of the circulator, and the pressure drop at that point  keeps the valve from opening where some point after the circulator is overpressured so bad as to explode.; I have a good imagination; I do not really think this would happen with residential size systems.

Paul48

to or from



"Install the boiler circulator only on the boiler return piping.

This ensures the pressure drop through the boiler

will not cause low pressure in the circulator intake". Direct quote from WM Ultra manual.

Gordy

JDB Read this

http://www.heatinghelp.com/files/posts/985/Multiple%20tank%20locations.pdf



And get back to me.



Gordy

Jean-David Beyer

Read this

I read something like that is Siggy's big book  (Chapter 7.3 of the current edition) a coupla years ago, but I will read it again as you sent it to me sometime today. As I recall, it made a lot of sense.

croydoncorgi

Pump kick

Further reasons why pumps go on the Return side:



- some furnaces need  a pressure 'kick' when the pump starts to confirm the pump is running.  Pump is powered from the furnace and the kick is needed during each ignition sequence,



- Some furnaces have auto air vents between the Return and the HX, or on the HX itself.  If you have the pump on the Flow side, there's a possibility that air gets drawn into the AAV if/when system static pressure is accidentally gets too low.  Bad Thing!

Jean-David Beyer

Agree? or Disagree?

I think we are saying what seem to be different things by the way we talk about things, but that fundamentally. we agree. Part of the confusion is that I am thinking in terms of my system, and that is confusing. I think I am beginning to see why.



I will simplify my system by ignoring my indirect hot water heater (to begin with) and by assuming I have only one heating zone. My system is piped primary-secondary. The boiler is the only thing in the primary loop except for the relief valve, a flow check valve, the supply-side T, the return-side T, and the circulator pumping back into the boiler. In particular, there is no expansion tank or air eliminator in that loop.



In the secondary loop we go from the supply T to the air eliminator with the expansion tank right beneath it. we continue to the circulator to the heating zone. The return from the heating zone goes to the return T and back to the supply T. Down from the return T is the circulator back to the boiler. Those two Ts are supposed to be no more than 12 inches apart according to the installation manual. Actually they are 5 inches apart. The primary loop, that includes the two Ts must be at least one inch but the contractor elected to use 1 1/4 inch because he said it worked better. I suspect he is right, at least because that makes the water go through the air eliminator more slowly (in the sense of feet per minute, not gallons per minute).



I think we mean the same thing by Point of No Pressure Change: the input to the expansion tank (that is in the secondary circuit). Now the secondary circuit I imagine we agree is piped correctly in that the secondary circulator comes right after the air eliminator, and thus the expansion tank. The piping there is 1 1/4 inch copper even though W-M say 1 inch is sufficient. So there should be relatively little pressure drop between the PONPC and the secondary circulator.



Also, there should be relatively little pressure drop between the closely spaced Ts and the air eliminator and PONPC, because that is only about a foot of 1 1/4 inch pipe.



So, as seen by the relief valve in the primary loop, the closely spaced Ts are the PONPC as far as the primary loop is concerned (even though, strictly speaking, there will be a little pressure change at the closely spaced Ts compared with the input to the expansion tank. The change will be only the amount of pressure drop in the secondary loop from the Ts to the expansion tank, about a foor of 1 1/4 inch copper tubing away.



More important in terms of operating a circulator in the primary loop opening the pressure relief valve, I believe (but do not know if you do) is where that circulator is in relationship to the PONPC of the primary loop. If I am a purist, there is no PONPC in the primary loop. But if you accept that the location of the two Ts is close enough, we can continue. (If you do not agree, I am not sure wat to do next.)



My primary loop circulator pumps from the return T into the boiler. At the output of the boiler is the pressure relief valve, so the pressure it sees is the pressure rise caused by the boiler circulator running less the pressure loss going through the boiler. The rest of the pressure loss in that loop is the loss through the Flow Check valve, the two Ts, and all the 90-degree elbows, open ball valves, etc.



I pretty much know the static pressure of my system. The Tridicator says about 15 psi (gauge too small to be really sure); the fill valve says 12 - 15 psi. Now this gauge needle does move. I used to watch it very closely when I ran my system with the make-up water turned off. (I do have a LWCO.) It is on now because the contractor hassles me when it is off. There are two things that move the needle. One is the outdoor temperature when the boiler is not firing: as it get colder the pressure drops about the thickness of the needle. This is not a leak, because it goes back up in the warmer weather. The other thing that moves the needle is when the indirect runs. The indirect is piped across the primary loop. The water goes from the supply (right after the pressure gauge) through the tank-within-a-tank) indirect, through a Taco 007-IFC back to the return side of the primary loop. This drops the pressure indicate by the gauge, and also by the relief valve, by about 2 PSI. I must assume that the flow through the indirect is greater than the flow through the primary loop. The differences are the resistance of the indirect vs the  resistance of the Ts, and the difference of the resistance of the IFC valve vs the real Flow Check valve. My bet is that the flow through the Indirect is higher.



Can you tell if we are disagreeing, or if we are using different words for saying the same thing?

Magnehelic

Apples and Oranges

PONPC is discussed only and mainly when deciding where to put an expansion tank.  These are normally in the building loop, and hydraulicly seperated from the boiler loop (which is the whole reason for primary/secondary right??) so the location of the pump on the boiler loop should be of no realtion to the PONPC.  That being said......always follow the manufacturer's instructions......but also find out WHY they want it that way or you'll just be one of those guys that does things for 30 years a certain way "because that's the way I've alsways done it and it works"  It is normally recommended to pump INTO your highest pressure drop (or friction loss), and it is recommended so that there is less (or no) chance of sucking a lower pressure elsewhere in the system.  What I love about this is that vendors overspecify pumps because they have no idea what kind of piping system is actually going to be installed (and the pumps are ALWAYS oversized eh??) and then the boiler manufacturers tell you to pump into the highest pressure drop because they know there is a high likelihood of the pump being oversized.  In a perfect world........you could do it either way.......especially on primary secondary boiler loops.  Be careful with indirects like Amtrol or Lochinvar where the pressure drops are pretty huge (usually a lot more than the common mod/con heat exhanger).  I use Triangle Tube indirects for two reasons.....1) they actually put a factory air bleeder on the boiler side top of the tank....and 2) they have 60% better recovery and a very low pressure drop.

Cheers!!

Don

NYplumber

Carry on

Carry on in the conversation. All this talk got me thinking, and with the help of a manufacture rep I figured my solution to the system at hand.





Thanks again guys (& gals[?]).

AlHeating

On the return !

1)  less risk of cavitation if the heat exchangers is ever stuck with sediments



2)  you should always push in the restriction element, not suck from it.  It ensures the turbulence of the flow, which makes the heat transfer better.



3)  many wet rotors on the market.  Water chilled.  Where is the water colder?  In the return!  It gives a better chance for the pump to cold down.



Cheers !!

Gordy

Don

Totally agree with all you have said. Funny though how TT solo has their built in circ on supply side of the HX, low pressure drop fire tube design HX maybe?  But then Knights fire tube design which does not have a built in cir has you pumping in to the HX with Pimary/secondary.





Gordy.

icesailor

Pumps:

Gordy,

I don't care how old thaat drawing is or who did it, it is just plain WRONG!!!!

Are you or the drawing trying to tell me that the pressure is highest before the fluid goes back into the pump? Not ever. If so, where does it happen where it goes to less than system? 6"? 1 foot? No, it happens in the pump, in the impellor. Inside the volute. And the circular curve, is split from the middle so that the split is over/in the pump. High going out, low coming back. And half way through the system, the pressure would be at system. The second PONC.

If you put pressure gauges spaced along the circuit, it would show that. It's not a perpetual motion machine. Water coming out of a hose will have less pressure than the inlet. The pump becomes the pressure machine. It becomes the source of pressure over its static state.

When the wind is blowing 10 MPH on the ice, and I sheet in the sail on my ice boat, nothing will happen. I have to run and push itto overcome the resistance of the ice on the skates. If I push it and run at 5 MPH. the sail will "see" 15 MPH of wind. It will accelerate up to 40+ MPH. If the wind slows down, I will slow down. If the wind completely stops, I will stop. The wind is the pump.  

Jean-David Beyer

Drawings.

"Are you or the drawing trying to tell me that the pressure is highest before the fluid goes back into the pump?"



I do not think anyone or any drawing, is trying to tell you that. Figure 6 of that drawing, and Figure 7-12 of Siggy's book, both show that the pressure is minimum at the pump intake. (It may be even lower at the very entry to the impellor inside the pump, but the diagrams do not show that level of detail.)



"And half way through the system, the pressure would be at system. The second PONC. If you put pressure gauges spaced along the circuit, it would show that."



If you accept that the entry to the expansion tank is a point of no pressure change, then the figures show that there will be a slight drop in pressure between it and the entry to the circulator. There will be a considerable rise in pressure (to the maximum). From there, pressure drops around the loop until it gets to the expansion tank again. Let us make up some numbers that might be indicated by pressure gauges distributed around the circuit. In Siggy's diagram he gives the following.



Static pressure of the system: 10 psi



With the circulator running:



Pressure at expansion tank: 10 psi

Pressure at pump entry: 9 psi

Pressure at pump exit: 18 psi (pump puts out 9 psi pressure at whatever flow rate we have)

Pressure about 10% of the way around the circuit: 17 psi

Pressure about 35% of the way around the circuit: 15 psi

Pressure about 60% of the way around the circuit: 13 psi

Pressure about 85% of the way around the circuit: 11 psi



So the pressure change all around the system is to raise the pressure above the static pressure everywhere except the section between the expansion tank entry and the circulator entry. So there is no second PONPC

Mike Kusiak_2

Two PONPCs

Ice, your description of the pressure relationships around the system would be correct if there was no expansion tank. The pump differential would be split equally between the inlet and outlet of the pump, and the first PONPC would be in the middle of the pump. The second PONPC would be at the halfway point of the loop.



But the tank fundamentally changes the pressure relationships. First I believe you agree that the tank pressure remains at the static fill pressure whether the pump is running or not. The tank pressure can't change because the pump doesn't add any water to the tank. No change in water level, no change in pressure. The only thing that can change the tank pressure is adding water by thermal expansion, which is not happening here.



So if you agree that the tank pressure cant change, what happens when you connect it directly to the inlet of the pump. By definition then the pump inlet pressure is the same as the tank static pressure and cant change. The pump differential also doesn't change so it has to add the total pump head to the fixed tank pressure found at the inlet. So , in effect the tank "anchors" the pump inlet pressure to the static fill pressure of the tank and the outlet pressure becomes static pressure plus pump differential .

Now when the pump runs, the whole loop is above static fill pressure, except right at the pump inlet  (PONPC) where it remains at the static pressure of the tank. Now there is no second PONPC, since the whole loop is above static pressure.



What happens if you connect the tank to the outlet of the pump? Now the outlet is "anchored" to the static fill pressure of the tank and cant change, so the pump differential has to totally subtract at the inlet, producing a pressure at the inlet of static pressure minus pump differential. So now the whole loop is below static pressure except at the outlet of the pump and tank (PONPC) where it remains at system pressure. Thats the problem with pumping toward the tank, it reduces the pressure of the loop to below the static fill pressure.



It took me quite a while to accept all this, but if you actually install gauges on a working system you can see that it really does work as described.

.

Unknown

Opinion from a Pump Guy

At the risk of stirring the pot, here's my take on this...



A lot of mod con boilers are using low mass, high pressure drop heat exchangers.  These definitely require minimum flows to stop boiling, satisfy low flow/no water safety devices and keep the HX from internally fowling up.



Typically (and I recommend this) there is usually a dedicated boiler pump (shunt pump) for each boiler, sized to overcome the head loss and BTU rating pf the boiler.  Similar to copper tube boilers, manufacturers suggest pumping into the boiler to insure proper flow through the boiler.  Yes, I know water is not compressible but it does take the path of least resistance.  And yes, this does not fit into the "pumping away" idea.



But, if the circ is only sized for the boiler head loss and flow requirements, then an additional primary loop circ will probably be required anyway (especially with multiple mod cons).  Yes, even with hydraulic separators in the system.  The primary loop circ and related stuff (expansion tanks, purgers etc) can be installed to "pump away".  In addition, the primary loop pump can be controlled to provide the right delta T or to reduce return water temperature to assist condensing.



As long as the shunt pump is not over sized I can't see how it would cause a relief valve to blow unless it was deadheaded.  If the system pressure is that close to the set pressure of the relief valve to cause it to blow when pumping into the boiler than the relief valve would have to be replaced with a higher pressure type (keeping in mind not to exceed the pressure rating of the HX).  Consult the boiler mgf!!!



One more point, turbulent flow is required for the geothermal guys.  Hydronic heating uses laminar flows - turbulent flow in radiant actually potentially can reduce the heat output of heat emmiters - this is why hydronic rule of thumb is 4 to 4.5 feet/sec water velocity.



Hope this helps guys.  For those that pump away from mod cons with no problems, if it ain't broken, don't fix it as the saying goes.  It will probably work either way...

AlHeating

Turbulent ... not laminar!

The goal of the 4 to 4.5 ft/sec rule is not to keep the flow laminar, it's to make sure to prevent NOISE and EROSION in the system.  No matter the pipe diameter, you're fully turbulent at 4.5 ft/sec (see calculation below).



In any heat emitter or pipe, turbulence is preferable because you want the flow to be turbulent enough to "scrub" the inner surface of the pipe and uniformize the water temperature in it.  With laminar flow, the water in the middle of the pipe will be warmer than the one that touch the exchange surface.  Added to that, the lower density of the warmer water in the middle of the pipe, AND the absence of roughness in there, will increase that phenomenon.

 

If you calculate the Reynolds number of a 5/8" PEX pipe, at 4.5 ft/sec, with a density of 988.1 kg/m3 (density at 120F) and a Cp of 559 (viscosity at 120F), you obtain a Reynold's number of 3.6 x 10^4.  Which is fully turbulent!



Try with this calculator.  It's magic since you can mix SI and imperial units without any problem : http://www.efunda.com/formulae/fluids/calc_reynolds.cfm#calc



Note from the engineering toolbox : Velocity should in general not exceed 5 feet per second to avoid damaging noise and wear and tear of pipes and fittings



Hope this help !

bob_46

Speed

I was taught to keep velocity between 2 to 4 Ft./Sec. to keep air entrained. If velocity drops below 2'/Sec. air may separate and collect in an undesirable location.

Gordy

In all fairness to NY Plumber

  This thread kind of went off on a tangent. Which is healthy, and educational.



  Would like to know from NY Plumber what type of boiler he is using in this install.

Gordy

Gordy

Confusing lines

 Ice I think you are confusing the lines representing Static pressure when system circ is off ( hashed line), and system static pressure plus circ differential pressure added to the system pressure (solid line) pumping away from PONPC.



 Mike has already explained this so thats all I have to say.



Gordy

NYplumber

Knight firetube

Is the boiler.