Help Troubleshoot Monoflo ModCon Retrofit

[Mike Drew]

Hello! Apologies in advance for the long post, and thank you so much for any help provided!

My home has 10 main floor radiators and 1 basement radiator, all cast iron and all fed by a monoflo system. We just had an old 1940's radiator (roughly similar to this one) replaced by a Navien NHB-80. The new unit has been in place for about 2 weeks, and based on a few tests (no real cold weather yet), things don't seem to be working as they should.

New vs Old system overview:

(for clarity that should probably say 1" steel to AND ALONG outer walls. The 1/2" is only from the diverter T's to the rads and back)

1.25" to 1" junction at the front of the house:


New Boiler Setup:


So my problem... The HWS is leaving the boiler at ~125F, and coming back in the pipes at ~90F. Which sounds fine, except the 125F water isn't actually getting into the monoflo system and radiators. Instead it seems "blocked" by the cold water return, and the result is ~90 to 95F water actually getting to the system pump and being sent on from there. Meanwhile, my 90F return is being mixed with the 125F HWS just before reaching the boiler return, resulting in a return temp at the boiler of ~115+.

124.6 supply, 92.6 to System Pump:


Left pipe = to System Pump, center = 89F CWR, Right = 123F HWS


126.3 HWS, 90.6 Past System Pump



CWR Mixing with HWS, net 118F at boiler


The not-cold cold water return issue is annoying, because I feel like I "should" be getting more condensing efficiency than that will provide. But the ~90F actual supply temp is a far bigger concern - I don't see any way my rads will keep up with a Minnesota winter with 90F water. I try to understand the operations of my house as best I can, and I've done a ton of reading over the last week or so, but this is all still way over my head. I have so many questions...

-Am I crazy, and this all somehow normal and correct?
-The installers reversed the direction of the supply and return flows (see the new vs old image above) - could that be responsible for the issues I'm seeing? At first I thought so, but after reading more I'm not sure.
-Why was the new system plumbed to allow supply + return mixing? (the previous system didn't mix at all, at least as far as I could tell). What need in a mod con does that address?
-Is this the proper approach, just poorly implemented, or is this not something that should be done in this system?
-Why does the cold water seem to be "winning" when mixing? (120F supply + just under 90F return leading to just over 90F to the pump suggests almost exclusively cold water is getting through at that junction...?)
-Is 1" pipe upstream of the 1.25" main monoflo line problematic?
-How do I go about determining the optimum setting (high/med/low) for each pump? The installer just set them both to medium and said "that should be good".

Incidentally, I've tried the pumps at all combinations and the above behavior stays essentially unchanged. The only exception is when I set them both to low. In that case the boiler seemed starved for water when starting up, made some not great sounds, the supply temp started to spike, and I quickly set the boiler pump back to medium.

Other details...
-My system tests were done over 2 mornings with outside temps of ~55F, inside temp of ~68F, and the thermostat set to ~85F.
-All rads in the house ARE getting warm, just to around the same ~90 to 95F I'm seeing leave the system pump.
-During the install 2 new basement radiators were added, each of which tie into the new copper via their own runs of PEX.
-The pumps installed are Grundfos 15-58s.
-My home is ~1,800 sq ft, with a DHL of around 40,000 BTUs (or ~55k at 1.4x DHL) as calculated based on winter gas usage per Dana's Out With The Old article.
-FYI the black tape on the copper pipe in the pictures was put there by me, needed it to get accurate temps from the IR camera.

I've done my best to detail the new vs old system in the above pictures, but let me know any important details I've left out. I definitely plan to have the installer back out to correct whatever issues exist, but I'd love to go into that conversation as informed as possible. At this point I'm not feeling great about their ability to get it "right" on their own.

Thank you again for making it this far, this forum has been invaluable for me up to this point!

 

[Dana]

-The installers reversed the direction of the supply and return flows (see the new vs old image above) - could that be responsible for the issues I'm seeing? At first I thought so, but after reading more I'm not sure.

That is THE primary problem. Monoflow systems are directional, and will not provide the same amount of flow in the rads if the flow is reversed. It will still induce some flow when running backwards, and may work well enough in some systems, but it's far from ideal. Hacks working on high temp would just bump the temp 10F or crank up the flow in the loop something to compensate for the lower radiator flow and call it "good" but that's not what you want to do with a condensing boiler.

-Why was the new system plumbed to allow supply + return mixing? (the previous system didn't mix at all, at least as far as I could tell). What need in a mod con does that address?

The Navien NHB series uses water-tube heat exchangers, which have a high pumping head, and maximum flow rate well below normal flow rates for a mono-flow tee system. The boiler loop (aka "primary") is pumped separately with it's own pump to guarantee it's minimum flow rated, and the radiation loop (aka "secondary") is pumped at a rate required for the system to deliver the heat properly. Where the loops intersect has to be done correctly to provide sufficient "hydraulic separation" that the pumping of the secondary doesn't affect the flow in the primary too much and conversely. That is often done with closely spaced tees or a separate purpose-made hydraulic separator component. Navien sells a pre-engineered closely spaced tee manifold that works adequately, and that appears to be what the installers used, directly below the boiler:

^^Navien's hydraulic separator manifold^^

With a primary/secondary configuration the boiler output is mixed directly with return water flow, lowering the temp of the water delivered to the radiation, lowering the delta-T at the boiler.

The old cast iron boiler had l0w pumping head cast iron heat exchangers that tolerate the higher system flow well, and would not need primary/secondary plumbing unless the return water temp was too low, causing corrosive condensation inside the boiler or flue that the heat exchangers or flue would not tolerate without damage. By contrast the heat exchanger materials in the Navien were specifically selected to tolerate the acidity of natural gas exhaust condensation in order to take advantage of the heat of vaporization recovered by condensing inside the boiler.

The efficiency of a cast iron boiler is also quite high while condensing, at least at first. But operating it at condensing temps rapidly erodes the turbulence inducing features of the heat exchanger plates and shortens the equipment life cycle dramatically.

 

[Sponsor]

 

[Dana]

Perhaps I was too quick on the trigger. (Apologies for not reading it completely & carefully!) If the flow directions in this image are correct it's a problem.

The returns from all radiation loops (including the new basement loop) need to be tied together before the water enter the hydraulic separator, and the supplies need to come after the hydraulic separator. I didn't fully understand the squiggly bits on the system diagram until looking at the picture above. All three loops, the left, right, and basement all need to be supplied from a common point in the flow, and return to a common point.

When there is only one radiation pump you may need to install flow balancing valves to ensure the correct amount of flow on each circuit. In high flow systems this is best done with a globe valve which are not cheap, but more precisely tunable and don't wear out the way doing it with ball valves can. In smaller systems ball valves will usually work "well enough", at a lower cost but may induce a bit more pumping head.

Also, relative to reversing the flow, if both tees to a radiator are mono-flow tees it doesn't much matter, but if it's only one it does. If both tees are monoflow it reverses the flow through the radiator, but the flow is the same. If only one of each pair is monoflow reversing the flow lowers the amount of flow through the radiator:

[edited to add]

I think my relative inexperience of modifying systems with mono-flow systems is showing- if the system was designed correctly the flow will be about the same in either direction. But with multiple loops there needs to be some type of flow balancing to ensure proper flow on all loops.

 

[Fitter30]

One big problem is the boiler primary loop the tees to the secondary look can be no more that 4 pipe diameters apart center to center. I like back to back with a slight gap between them for ease of soldering. Boiler and secondary loops have only one direction. Basement and house are on to separate feeds with one pump and one return?

 

[Dana]

One big problem is the boiler primary loop the tees to the secondary look can be no more that 4 pipe diameters apart center to center. I like back to back with a slight gap between them for ease of soldering. Boiler and secondary loops have only one direction. Basement and house are on to separate feeds with one pump and one return?

The hydraulic separator tees is a fully manufactured assembly built & promoted by Navien. I agree that it quite apparently violates the standard rules of thumb regarding the number of diameters of separation and that a better hydraulic separator would be beneficial in very high radiation flow systems.

Purge tees are inherently tightly spaced, and work pretty well as hydraulic separators (as long as the ball valve is fully open to induce minimal turbulance), and can speed up air purging after working on the system.

But the Navien-built manifold should be fine here.

A also agree with (the implied notion in the question regarding separate feeds & one pump) that running the basement loop and the rest of the house as a single zone may never achieve good temperature balance between the basement and upper floors, especially if it's not a walk-out basement with well insulated foundation walls. It almost always needs to be a separate zone to work well in both mid-winter and shoulder seasons. The changes in load with outdoor temperature is usually too different between a basement & fully above grade floors.

 

[Fitter30]

The hydraulic separator tees is a fully manufactured assembly built & promoted by Navien. I agree that it quite apparently violates the standard rules of thumb regarding the number of diameters of separation and that a better hydraulic separator would be beneficial in very high radiation flow systems.

Purge tees are inherently tightly spaced, and work pretty well as hydraulic separators (as long as the ball valve is fully open to induce minimal turbulance), and can speed up air purging after working on the system.

Monoflow tees feeding radiators above main piping one tee on the return. Radiator below the main two. By looking at his piping lay out without a 2 balance valves on basement and house loops. Need to see how basement is tied in might not have enough head on that loop pump.

But the Navien-built manifold should be fine here.

A also agree with (the implied notion in the question regarding separate feeds & one pump) that running the basement loop and the rest of the house as a single zone may never achieve good temperature balance between the basement and upper floors, especially if it's not a walk-out basement with well insulated foundation walls. It almost always needs to be a separate zone to work well in both mid-winter and shoulder seasons. The changes in load with outdoor temperature is usually too different between a basement & fully above grade floors.

 

[John Gayewski]

The tees should have arrows on them. Are the arrows pointing in the direction of the flow of your new system?

I'm surprised they gave your system a makeover and left the monoflow tees. They are very much going out of style since there is no way to properly balance or adjust that system, also they have to be very accurately used.

I'm not familiar with the exact pump you have. Is it adjustable? Have you tried pumping your way out of it?

 

[John Gayewski]

I'm also kind of wondering why this if your problem. Didn't you pay to have a system that ran as it should? A 30 degree delta t isn't what it should be.

 

[Mike Drew]

Thank you Dana and Fitter for your responses!

Dana...

The returns from all radiation loops (including the new basement loop) need to be tied together before the water enter the hydraulic separator, and the supplies need to come after the hydraulic separator. I didn't fully understand the squiggly bits on the system diagram until looking at the picture above. All three loops, the left, right, and basement all need to be supplied from a common point in the flow, and return to a common point.

I may not be following 100%, but I'm pretty sure the system is set up with 1 common supply and 1 common return, with all of the supply after the hydraulic separator. I've added 2 new pictures that hopefully give a better view.

In both pictures,
A = the output to the mono flow main which then splits to the left and right loop.
B = Basement rad 1
C = basement rad 2.
The basement radiators are independent, not in series.

Regarding the hydraulic separator, Dana you are correct, it is that Navien built separator that is installed.

For the 2 basement radiators, what is the advantage of a globe valve to control flow vs using the radiator control valves themselves?

Thank you!

 

[Mike Drew]

The tees should have arrows on them. Are the arrows pointing in the direction of the flow of your new system?

The T's are all wrapped in asbestos insulation, so I can't verify the directions of the arrows.

The existing pumps have 3 settings (L/M/H) but I've tried all combinations between the 2 pumps with no meaningful difference in results.

As to why is this my problem, you're right. I 100% plan on getting the installer back out to fix the problem(s), but I want to understand what all of the potential issues are as best I can. The hope is by educating myself I can make sure things are fixed correctly, and not via a "hack" solution as Dana referenced above.

Thanks!

 

[John Gayewski]

The T's are all wrapped in asbestos insulation, so I can't verify the directions of the arrows.

The existing pumps have 3 settings (L/M/H) but I've tried all combinations between the 2 pumps with no meaningful difference in results.

As to why is this my problem, you're right. I 100% plan on getting the installer back out to fix the problem(s), but I want to understand what all of the potential issues are as best I can. The hope is by educating myself I can make sure things are fixed correctly, and not via a "hack" solution as Dana referenced above.

Thanks!

You want a delta t between 10 and 20. That's a good indicator of how the system is running. Ideally the system would run non stop dialing in the temperature as it runs. The non effeciant part of a boiler is start up that's also when the pumps are junked up. A low level of heat modulating constant is where the system is at its best. This isn't likley until it gets cooler outside though. Keep in mind your system is designed (or at least should be) to keep up with the coldest part of the year. So it should run at full clip constantly when it gets to the coldest part of the year. This time of year you won't lose as much heat as fast so it'll cycles quite a bit, but your delta t should not be that large.

 

[Fitter30]

Valve off the basement radiators see if house heats. C radiator doubt it would heat at all being piped both in and out of the side outlet ( bull) of tees compared to A piped on the run of the tees. Are the basement radiators above or below the main where the bottom picture and there moniflow piped?

 

[Mike Drew]

Valve off the basement radiators see if house heats. C radiator doubt it would heat at all being piped both in and out of the side outlet ( bull) of tees compared to A piped on the run of the tees. Are the basement radiators above or below the main where the bottom picture and there moniflow piped?

Thanks Fitter. I’ll close off the basement loop and see what impact it has when I get home tonight.

The B and C radiators (new basement radiators) do both heat up. They reach essentially the same temp as the upstairs rads. The B and C radiators are below the monoflow main. They’re floor level in the basement, same level as the condensate pump below the boiler.

 

[Mike Drew]

Closed the valve to the basement loops tonight and cycled through the various pump settings. No noticeable change to the system. Water still only getting to the system pump at 90ish degrees

 

[Dana]

It's possible that your radiation pump isn't providing the necessary gpm for the system to operate correctly.

Are you measuring the delta-Ts directly at the boiler, or are you measuring them at the input & output of the hydraulic separator?

 

[Fitter30]

What boiler temp is the boiler set for and it's readout. Also gpm and supply temp 2' down stream from boiler on piping. Looking at the infrared pics raise set point to 150* as long return water is 130* or below its condensing.

 

[Mike Drew]

That is THE primary problem. Monoflow systems are directional, and will not provide the same amount of flow in the rads if the flow is reversed. It will still induce some flow when running backwards, and may work well enough in some systems, but it's far from ideal. Hacks working on high temp would just bump the temp 10F or crank up the flow in the loop something to compensate for the lower radiator flow and call it "good" but that's not what you want to do with a condensing boiler.

The Navien NHB series uses water-tube heat exchangers, which have a high pumping head, and maximum flow rate well below normal flow rates for a mono-flow tee system. The boiler loop (aka "primary") is pumped separately with it's own pump to guarantee it's minimum flow rated, and the radiation loop (aka "secondary") is pumped at a rate required for the system to deliver the heat properly. Where the loops intersect has to be done correctly to provide sufficient "hydraulic separation" that the pumping of the secondary doesn't affect the flow in the primary too much and conversely. That is often done with closely spaced tees or a separate purpose-made hydraulic separator component.

Dana, thanks again for your help here. After 2 separate additional visits the install folks said the system needed to be switched back to its original direction. With that done the pumps run MUCH more quietly now, and system performance is maybe marginally better. But it still feels like a fundamental issue hasn't been addressed.

The boiler set point is still at 130F, and the HWS is still reaching the T to the system pump at ~125F. But the water from that T to the pump is still only 90-105F. I understand the need and benefit for the primary + secondary set up, but in thinking more about it it seems like what has been plumbed in my case is primary + additional primary + secondary loops. In the below pic, Red = primary loop utilizing Navien's own manifold, Yellow = "additional primary" loop, and green = secondary loop out to the radiators (HWS exits the green line on the right, returns on the left).

None of the piping schematics I have seen online show this type of set up (at least that I can tell). Some radiant floor set-ups seem somewhat similar, but those are expressly designed for low water temps to the radiation.

I don't understand why the yellow line should connect between the 2 green lines. The distance between those 2 T's is also 10" center to center, which seems to violate the 4x pipe dimension rule of thumb I've seen everywhere online.

I'm convinced that the path of least resistance is the return water traveling vertically along the yellow loop and circulating right back in to the hot water supply, which is preventing any substantial amount of the actually hot (meaning 125F+) water from getting out to the radiators. I can feel with my hand a constantly changing mix of hot and cold water passing through the horizontal pipe to the system pump.

Wouldn't I see better temperatures to the radiators with either no connection between the green system loop OR a valve there allowing the connection to be partially closed OR at least more tightly spaced T's?

I don't want to come off as a jerk to the install company asking them to come out for a 4th time, especially if I'm way off base. But the system as plumbed feels highly inefficient, and I have serious concerns that the low supply temp will be able to keep up with our Minnesota winter once we start getting real cold weather.

Thanks again to all for any insight or guidance!

FYI the system without my scribbling...

 

[Fitter30]

Dana, thanks again for your help here. After 2 separate additional visits the install folks said the system needed to be switched back to its original direction. With that done the pumps run MUCH more quietly now, and system performance is maybe marginally better. But it still feels like a fundamental issue hasn't been addressed.

The boiler set point is still at 130F, and the HWS is still reaching the T to the system pump at ~125F. But the water from that T to the pump is still only 90-105F. I understand the need and benefit for the primary + secondary set up, but in thinking more about it it seems like what has been plumbed in my case is primary + additional primary + secondary loops. In the below pic, Red = primary loop utilizing Navien's own manifold, Yellow = "additional primary" loop, and green = secondary loop out to the radiators (HWS exits the green line on the right, returns on the left).
View attachment 78139

None of the piping schematics I have seen online show this type of set up (at least that I can tell). Some radiant floor set-ups seem somewhat similar, but those are expressly designed for low water temps to the radiation.

I don't understand why the yellow line should connect between the 2 green lines. The distance between those 2 T's is also 10" center to center, which seems to violate the 4x pipe dimension rule of thumb I've seen everywhere online.

I'm convinced that the path of least resistance is the return water traveling vertically along the yellow loop and circulating right back in to the hot water supply, which is preventing any substantial amount of the actually hot (meaning 125F+) water from getting out to the radiators. I can feel with my hand a constantly changing mix of hot and cold water passing through the horizontal pipe to the system pump.

Wouldn't I see better temperatures to the radiators with either no connection between the green system loop OR a valve there allowing the connection to be partially closed OR at least more tightly spaced T's?

I don't want to come off as a jerk to the install company asking them to come out for a 4th time, especially if I'm way off base. But the system as plumbed feels highly inefficient, and I have serious concerns that the low supply temp will be able to keep up with our Minnesota winter once we start getting real cold weather.

Thanks again to all for any insight or guidance!

FYI the system without my scribbling...
View attachment 78140

Yellow loop is the primary loop it needs its own pump. Boiler needs a pump which it has. Green lines in previous pic are the the red and blue lines in the ceiling?

 

[Mike Drew]

Yellow loop is the primary loop it needs its own pump. Boiler needs a pump which it has. Green lines in previous pic are the the red and blue lines in the ceiling?

Hi Fitter. Yes, green lines go to the ceiling and are the same that we’re labeled red and blue in the prior pictures. Only difference since those pictures is the direction has been reversed, so what was red is now the return, blue is supply, and all arrows should point the other way.

When you say Yellow is primary and needs its own pump… I don't quite follow. Are you saying I should have 3 pumps? Or that the pump on the green line effectively IS the primary pump.

Per this image from the Navien manual, flow should go from the manifold loop to the system pump + system. No mention of another primary loop. ¯\_(ツ)_/¯

 

[John Gayewski]

Generally switching a monoflo system to flow in reverse won't work well. The pressure differentials in a hydronic system are not high enough in short runs of piping to overcome the friction in reverse. The tees are designed specifically designed to flow one way, hence the name monoflo.

I don't know what they did and why. I think it would be easier if you drew a simplified schematic that shows the system flow as you understand it. I realize you tried a version of this by writing on the picture. By there's not enough context that I can say whether it should work or not. I'm guessing since it isn't working and you suspect it's wrong then it's wrong.