another way of "looking" at PONPC

hot_rod

This piping arrangement is seen often in manufacturers manuals. Expansion tank in the distribution "primary" loop.

What can happen is the indirect tank calls, heat circ shuts down and is checked. The PONPC needs to be referenced thru all the components and piping between the indirecct tank and expansion tank connection.

Notice what happens to the pressure in various parts of the circuit. Exact numbers could be put to this if you know you boiler, tank coil, piping pressure drop.
So this could set up a condition (sub-atmospheric) where air can be pulled into various float type air vents in the system and possibly coax cavitation, especially when high head circs are used for high pressure drop boilers. It also hampers air removal, even with an excellent purger.

Notice, as Dan has explained for several decades now, "the only place in the system where the circulator cannot change the pressure", the PONPC point of no pressure change

DanHolohan

Thanks, H.R. I wrote about a similar situation in Primary/Secondary Pumping Made Easy. It involved a check valve and a three-way mixing valve between a primary- and secondary circuit. When the mixing valve went to full bypass the secondary pump could no long "see" the PONPC in the primary loop and it could cavitate. The solution was to use a three-way diverter valve on the return side of the secondary instead of the mixing valve on the supply side.

Like most things regarding P/S, I got this tip from Gil Carlson, who coincidentally died on this day in 1994. Thanks.

hot_rod

Yep, a few ways to "better" the piping and tank shown, here is another option, notice how the red dotted line stays at or above the static and the circulators delta P assures the boiler doesn't go into low pressure lock out.
Showing it in a graph format helps understand what is going on with the dynamic pressure.

R.I.P Gil

Peter_26

So the delta P adds just enough pressure to compensate for the drops until it matches the static pressure? So to achieve this the expansion tank must be where the water is cooler, which is the return right before the pump? Just trying to get a better understanding of the concept.

R.I.P Gil

HomerJSmith

hot_rod, how does the setup on Idronics14 page 44, Figure 7-21 fair with your placement of the Xtank? What about zoning with circulators rather than ZVs. Isn't a hydro sep equivalent to closely spaced tees?

This post has really got me thinking. That's a good thing.

hot_rod

HomerJSmith said:

hot_rod, how does the setup on Idronics14 page 44, Figure 7-21 fair with your placement of the Xtank? What about zoning with circulators rather than ZVs. Isn't a hydro sep equivalent to closely spaced tees?

This post has really got me thinking. That's a good thing.

Good eye, Homer.

A properly engineered separator is basically a super sized "closely spaced tees".
This example shows a Caleffi 2" pipe size sep with a 40 gpm flow rate. When a properly designed barrel diameter, or space between the close tees, flow velocity drops to near zero. Roughly, the barrel is 3 times the diameter of the pipe connections to the sep.

Notice less than 1/2 fps velocity from top to bottom with 40 gpm flow on one side. That space inside the barrel becomes know as the low velocity zone. As such it is an excellent place to remove air, dirt, and it assures PONPC to any of its branches.

So the schematic you noticed has 7-21, all 3 circs pumping away. Very little pressure drop between any circulator connection and the exp tank connection. It can never be zero pressure drop, even 1" of pipe has some pressure drop with flowing water.

Shown in fig 7-21 The system circ must have a check so the DHW circ cannot pump at the PONPC. So the DHW priority circ references it's PONPC on its suction side thru the indirect. The graphs I attached a few posts above show that concept, and how to calculate what that pressure drop could be.

If you built a P/S closely spaced tees for a 40 gpm flow, the space between the tees would probably be a 2" pipe. While that gets you hydraulic separation, greatly increasing that diameter between the tees drops velocity. That is how a sep becomes a 3 in 1 device
The engineering works out to the flow velocity between top to bottom in a sep being 1/9 of the flow into and out of it.

With a 40 gpm flow on one side, no flow goes out to the secondary. Until a circulator on that side runs, same a P/S closely spaced tees really. So the importance of the tees being closely spaced is critical to P/S working properly, assuring no flow to secondaries under design flow thru the primary. Any flow restriction between the tees encourages flow into the secondaries. Same issue when you see a "value engineered" hydraulic separator with a skinny center barrel dimension.

hot_rod

Peter_26 said:

So the delta P adds just enough pressure to compensate for the drops until it matches the static pressure? So to achieve this the expansion tank must be where the water is cooler, which is the return right before the pump? Just trying to get a better understanding of the concept.

R.I.P Gil

These early B&G examples cleared it up for me. Now in living color:)
Notice when pumping away from the tank the system sees that increase in pressure from the circulator. It shows up as a pressure increase at the circ discharge, then drops as it goes around the circuit due to flow resistance of the loop.

In the second example pumping at the tank, that pressure difference, delta P, mshows up on the suction, inlet side. Still workable, but positive pressure helps aids air removal.

#3 shows a worse case, where you pull a sub atmospheric condition. In this example we dropped static fill to show the condition. A medium or high head circ would do the same with a 10 psi fill. Under this condition air vents on that negative pressure locations would pull air in! Also at high temperatures, negative pressure you could boil, flash to steam.
Lastly all circulators like positive pressure at their suction. Small resi circs get by with a few psi. But at higher operating temperatures that inlet pressure needs to be higher. I think something around 4-5 psi at 190F, on small wet rotor circs.

Bottom line. Always pump away from the expansion tank connection, why tempt fate?

As odd as it looks a low pressure drop boiler could have the tank mounted at the return, circ at the boiler supply out, you would be still pumping away.

This graphic is intended to show the best places for air removal, but it shows that ∆P concept also. Highest temperature is best air removal, second best is lowest pressure point, at the very top of the piping or top radiator.

HomerJSmith

Thank you, hot_rod. I read it six times and am probably not done.

So, per, our earlier discussion on the PONPC being the primary circuit as emphasized by Gil, in Figure 7-21 the Xtank connected to the hydro sep would be the primary circuit (load circuit) making the labeled Secondary pump really the Primary pump? If so, you can see the confusion with those terms (primary/secondary).

hot_rod

HomerJSmith said:

Thank you, hot_rod. I read it six times and am probably not done. So, per, our earlier discussion on the PONPC being the primary circuit as emphasized by Gil, in Figure 7-21 the Xtank connected to the hydro sep would be the primary circuit (load circuit) making the labeled Secondary pump really the Primary pump? If so, you can see the confusion with those terms (primary/secondary).

Probably better to just label it distribution pump instead of secondary. I suppose they could all be considered primary pumps, as they all see the expansion tank in their circuit. 

If the tank was connected at the upper left hydro Sep branch, then which pump is considered primary?

Peter_26

Thanks hot_rod, I can grasp the concept of "PONPC" better now. I was also trying to make sense of how a ∆P circulator works and it went down hill after that .