Newbie looking for (re)design guidance and troubleshooting

SparkyBeaver

I have a newly built ICF home (2yrs old) with hydronic heating in a concrete slab foundation. I acted as an owner builder and sub contracted much of the work and did some myself. I have an Electrical Engineering background and so a lot of DIY projects plus manage rental properties. Needless to say, I did my own electrical but had a plumber do all of the DHW plumbing. (I'm definitely not a plumber - often have leaks) My primary heat source is hydronic heating on the main level with a multi split heat pump for back up heat and A/C. The contractor who did my ICF work has built around 2000 homes and has much experience with hydronic heating. He did all my PEX layout (Uponor 1/2" PEX A) and built the hydro pumping assembly with three zones. Each zone is a simple 120V thermostat that switches power to an outlet that powers on the pump for that zone. I have two four circuit manifolds and one three circuit manifold. All circuits are less than 250'. (See attached floor plan layout) Each zone has a Grundfos Alpha 2 15-55 circulator pump. Originally, the system was an open loop system with an 80gal hybrid heat pump water heater for DHW with a Rheem propane tankless heater in parallel with bypass valve on the hybrid unit if it ran out of hot water. However, the return of all the heating circuits only went through the tankless unit. My contractor said he has an 80gal gas water heater that provides all of his DHW and floor heating and he's never ran out of hot water. Because the ICF homes are so well insulated, the water heater probably puts out enough BTU's for this to work. He suggested a tank unit, but I thought I was being smart by getting a tankless unit with variable output to save space in my mechanical room. (Up to 199k BTU, 9.5gal/min). The original design worked OK but not great. I don't think my contractor ever did an official design of the system, just came up with something based on his extensive experience. The initial design caused me to have to run both water heaters at the same temp (130 deg) and adjust mixing valves for DHW and floor heating. The mixing valves had limited range and max on the DHW only gave me 118 deg water that didn't last for more than two showers. After one year of this I felt it should be better. The first year, I left my heat pump running with lower setpoints than the floor heating, but believe it was doing most of the work last season.

I decided to separate the two systems and convert my hydronic heating to a closed system. The 80gal hybrid would remain at 130deg and be dedicated to DHW. The tankless unit would be powering the radiant heating and would allow me to set the temp anywhere needed. My target was 105-115 deg. I did some re-piping to match the new design (see attached). Tried firing it up and it wasn't working! That's when I started doing more research and found this forum. I realized how little I really knew about hydronic heating and how I should have done a few things different when designing the system. First, I realized I had air in my circuits after sitting for a year and never fully purged the manifolds. I purged each circuit individually. Surprising there wasn't much air in the system. I also learned that the cold make up water was full of oxygen bubbles. No wonder the system would stop running after a short time because it had air. The biggest thing I learned was that the tankless water heater had a really high resistance to make it fire. (14 PSI ~ 32' head) My circulator pumps could barely produce enough pressure to turn it on and my flow rates were barely above zero. I never had a proper system design done so I didn't know what the flow rates should be. I'm also realizing that I should be using a boiler instead of a tankless water heater. The heat output and flow aren't the main problem, it's the high head that's holding me back. I think most of the system and components are in good shape, but I'm looking for advice on how to get my system running without investing a ton of money to redo things. What flow rates should I be using for each circuit? I had a load calc done to size my heat pump and have added the BTU calcs for rooms that had numbers. What pressure should I be running the system at? I can add a pressure reducing valve on my make up water line coming in. Can I add a high head pump right before the tankless unit to increase flow and reduce the load on the circulators? I do have two Grundfos UPS 26-99's in series that I salvaged from my geothermal heat pump system after a house fire at my old house. Don't know if they were damaged, but these two produce a head of 48' on speed one. I respect the knowledge and experience of people on this forum and hoping for some advice. Anything would be appreciated. I've attached some photos and system design info. Feel free to shoot out any questions. Thanks.

hot_rod

A primary secondary piping would solve the issue with the heater flow and pressure drop.
Three options, closely spaced tees, a hydraulic separator, or a buffer tank, in order of cost.

In the space you have, a clever hydroncian could shoehorn in either a P/S or sep and sort out the concerns.

Jamie Hall

?hydroncian ? I love it, @hot_rod ! Spell check doesn't, but I do!

SparkyBeaver

Thanks for the input Hot_Rod. I had to research what a hydraulic separator does, but still not sure how I would install the options you mentioned. Any chance you could provide a red line markup on the system schematic I provided?

hot_rod

in a simple drawing, here is the option.
The sep brings a 4 in 1 concept, with helps in tight spaces.
Pipe from the tankless to sep, them header the pumps into the other side of the sep.

SparkyBeaver

Thanks for the image. That adds much clarity for a Newbie like me! Looks like I'll still need another circulator pump for the primary heating loop. Any idea how to size that?

GroundUp

You can keep your existing zone circulators. The "boiler circ" will need to be sized according to your BTU requirements and desired delta T which should have been part of the original design. A 26-99 is usually a good option, but two in series should most definitely suit the need if you already have them laying around anyway. Flow per loop typically ends up in that .5-.8 GPM neighborhood, again based on heat loss and delta T. With ICF it's usually best to keep the flow high and minimize the delta T, so that the floor has a slightly higher average surface temp.

hot_rod

The installation manual may have a chart like thus showing the pressure drop through the heater at various flow rates. I suspect one 26 99 will get you the flow you need 

so pipe from the tankless to the sep with at 26 99

SparkyBeaver

Good input on the primary/secondary lines. I think I could add a couple of closely spaced Tees without too much trouble as my supply and return lines are about 12" apart. (See marked up photo) I already have the air and sediment separators installed. I do have a question about tee mounting position though. Do the tees need to disrupt the flow path or can they be a straight shot through (horizontally) with the vertical connections tied together? Also how close is close here? Is 12" sufficient or do we need a 1" close nipple tying the two together?


Regarding the extra pump, I plan to install a 1" paddle flow switch on the secondary loop so when the zone circulators turn on, it switches the auxiliary pump to my water heater on in the primary loop.

GroundUp

That orientation is great, but you want to do it as close to the tankless as possible and you need to get rid of those flex hoses to minimize flow restriction- screwing a flange directly on the inlet port to the tankless would be ideal. Tees as close together as possible, 12" is too far. I like to put a valve between them for ease of purging (Webstone purge tees are awesome for this) but it's not a necessity.

Jamie Hall

Oddly enough -- and perhaps counterintuitively -- so long as your velocities aren't outrageous whether the primary flow is straight through the two Ts, or into the legs of both Ts -- or (and this is where it gets weird) one side one way and the other the other won't make much difference. The orientation will change the head loss, but not that much.

That said -- do it neatly. Nothing looks worse than a plumbing job which looks like the installer was high or drunk...

SparkyBeaver

I had a chance to implement some of the ideas over the weekend and wanted to report back on the progress.  Good news and bad.  Good news is that I got the pipes split into primary and secondary loops with close tees installed.  Also added a pressure reducer on the intake and have the system running at 15psi now.  It went pretty smoothly.  The existing circulators on the secondary side now flow much better and I have more than enough capacity for each Zone (~1.5 - 3GPM).  



Now for the bad news.  The two 26-99 pumps I salvaged from the house fire don't appear to work.  I tried replacing the internal caps but they still don't run when applying power.  I purchased a cheap UPS15-58 3 speed pump but it doesn't appear to move enough hot water through the primary loop to supply all the secondary demand.  I don't know the exact flow on the primary side, but think it's around 2GPM.  If I only have one zone calling for heat, the secondary temp is just several degrees below my water heater setpoint.  However, when two or three zones call for heat it really drops the temp of the secondary side down 25 degrees or more.  My initial thought is to get a bigger pump.  I need something that will move about 6GPM through the water heater.

I did leave it running overnight with one zone and it did a nice job of heating a single zone.  

The other issue I've got is finding a good 1" paddle flow switch.  I purchased a few cheap ones from Amazon that had quick delivery, however, none of them have worked correctly.  Two of them had faulty contacts that didn't switch when the flow switch was pressed.  The other one I got did switch when pressed, however, after it was installed, I adjusted the flow setting down to switch at 1GPM, but the switch never turned off after flow stopped.  


Any recommendations on the pump or flow switch?

Thanks 

RickDelta

@SparkyBeaver
I first saw this pipe distribution configuration up in new York and didn't know the advantages or why you would do it like this.

...... could you explain (laymen's) what's going on it this arraignment please?

Thanks!

GroundUp

Your pictured gauge is on the return side- is that what you were using to check the temp on the secondary side? The 26-99 is still a pretty good choice for your primary

SparkyBeaver

RickDelta said:

@SparkyBeaver
I first saw this pipe distribution configuration up in new York and didn't know the advantages or why you would do it like this.

...... could you explain (laymen's) what's going on it this arraignment please?

Thanks!

I didn't understand the closely spaced tees either and had to do some research. @Hot_Rod shared a diagram of a hydraulic separator that helped. I also found a good article from an Australian company that explained how the flows work. In my case the tankless water heater is causing a lot of head loss in the system, therefore my zone circulators couldn't meet the demand. By separating one big loop into two smaller ones, you're giving the zones a shortcut to bypass the water heater and reducing the head in the secondary loop. You do have to add another pump for the primary loop but that only needs to meet the head of that side of the system. In my case, it's my water heater plus about 5 physical feet. I think it's about 18 ft. of head on the primary side.

SparkyBeaver

GroundUp said:

Your pictured gauge is on the return side- is that what you were using to check the temp on the secondary side? The 26-99 is still a pretty good choice for your primary

Yes, the gauge in the previous photo was a newly installed gauge on return side of the secondary. I still have my existing gauge near the water heater intake that acts as my primary gauge. There's another one on the input of the zone circulators to monitor temps going out to each zone. It was helpful to see how much heat is going into the secondary side vs. circulating back around in the primary side based on flows.

RickDelta

SparkyBeaver said:

Good input on the primary/secondary lines. I think I could add a couple of closely spaced Tees without too much trouble as my supply and return lines are about 1

SparkyBeaver said:

RickDelta said:

@SparkyBeaver
I first saw this pipe distribution configuration up in new York and didn't know the advantages or why you would do it like this.

...... could you explain (laymen's) what's going on it this arraignment please?

Thanks!

I didn't understand the closely spaced tees either and had to do some research. @Hot_Rod shared a diagram of a hydraulic separator that helped. I also found a good article from an Australian company that explained how the flows work. In my case the tankless water heater is causing a lot of head loss in the system, therefore my zone circulators couldn't meet the demand. By separating one big loop into two smaller ones, you're giving the zones a shortcut to bypass the water heater and reducing the head in the secondary loop. You do have to add another pump for the primary loop but that only needs to meet the head of that side of the system. In my case, it's my water heater plus about 5 physical feet. I think it's about 18 ft. of head on the primary side.

...... the "T's" I saw were aprox. 12" apart from each other, so how do I calculate an effective heat exchange for this?
also, it may have had a circulator inbetween these two "T's" as well.

hcpatel78

@sprkyBeaver...
Order a Complete Design Assistance Manual from Uponor. ...Learn the hydronic design....Also John Siggy's Modern Hydronic Heating and Cooling For Residential and Light Commercial Buildings, 4th Edition ...
these are Great books for Hydronic Heating....

SparkyBeaver

Ok, I've made some more progress this week.  First, I abandoned the flow switch idea and implemented a relay control option the turns on power to my primary pump whenever any of the three zone pumps turns on.  It is working great!  Should have just done this from the beginning.

Second, I replaced my UPS 15-58 with an Alpha 15-58.  I like so many things about this little pump.  Variable Speed, Multiple Modes, energy efficient, and a nice display with Bluetooth connectivity to monitor things.  It shows my head varies from 8-19ft depending on which mode it's running in.  According to the pump curve it should be able to put out over 4GPM at 19' head.  I've tried all the modes with manual and autoadapt but can't seem to get it to put out more than 1GPM.  It's never used more than 25W, but docs show it maxes out at 37W.  How do I unlock the potential of this pump? I've got two zones running at 4GPM and my primary pump isn't keeping up.  I'm tempted to call Grundfos to see if they have an Engineer I can work with.  I feel like I'm missing something.  Is there a hidden menu on this pump to unlock the full potential?  I want to test full throttle!

hot_rod

19' is pretty much the max head on that circ. Is this pump a tankless heater. Most often you need a high head circ to get the full output from that type of heater. You mentioned you had sone 26-99s?

Those flex hoses are probably pretty restrictive also, adding to flow resistance issues

GGross

@SparkyBeaver
I believe if you use the grundfos go app you can adjust the pumps minimum flow rate, You'll still be limited by the pressure drop of the equipment/piping the pump is attached to, it can only do so much. I know it was at least easier to control the parameters with the app, not sure if it allows more changes than the onboard controls or not. These are still a little new to me so I only played around with one for a little while, still selling through our old alphas before we switch to these haha

hot_rod

Did you try all the modes, constant flow for example?

GroundUp

As you were told from the beginning, the flex hoses need to be eliminated and you need something similar to a 26-99. That Alpha is too small for what you're trying to do.

SparkyBeaver

GroundUp said:

As you were told from the beginning, the flex hoses need to be eliminated and you need something similar to a 26-99. That Alpha is too small for what you're trying to do.

I hesitate to remove this flexible water heater hose and add more elbows in the system. This isn't a corregated hose, it's a smooth bore PVC hose on the interior with a long sweep. My primary loop currently has (11) 90 deg elbows. (8) are exposed and (3) are hidden behind the wall. There's about 6' of vertical rise plus the tankless water heater. I've seen several manufacturers of longer radius PEX elbows (~2"). I've thought about replacing some of the standard close 90's with several of the long sweep or (2) 45 elbows. Would that make much difference? If you still think it's work replacing the braided SS hose, I can do that also. Just not sure it would have much impact?

SparkyBeaver

hot_rod said:

Did you try all the modes, constant flow for example?

I have tried all four modes available. Constant Curve, Constant Pressure, Proportional Pressure, and Constant Flow. Tried setting the Constant flow to 4.5GPM, but actual flow was only 0.8GPM. As mentioned before, the estimated head changes depending on the mode selected. Proportional Pressure seems to show the lowest. Around 7' and still producing about 0.8GPM. Constant Curve shows the head at 19'. Constant pressure shows head about 15'. I'm not sure if the pump is accurately calculating the head because it's only seeing the tankless unit in front of it. I've toyed with the idea of moving the pump to the tankless hot output with the hope that it would sense a pressure change depending on how many zones are running. My desired method would be Constant Pressure running at the top of the curve, but the Grundfos Go app only allows you to choose setpoints in the middle of the regions. Ideally, I'd love a mode that adjusts for pump speed based on water temperature. This Alpha pump does have a temp monitor. Would like the pump to use the initial temp as a reference. After running, if it sees the incoming temp start going up it could increase flow rate. If incoming temp drops, it should slow down. That way it's regulating return flows from the secondary loops.

I could get another 26-99 and probably get by with running on speed one. I do still have the two salvaged units from the house fire in my previous house. Neither seemed to power up when testing though.
I haven't tried pulling the apart to see if they could be rebuilt. However, a standard 26-99 wouldn't be adjusting flow based on anything in the secondary side so primary temps would be fluctuating based on secondary demand. I guess that would be OK, as the heater would have less rise demand once temps start creeping up. Just not the most efficient. My preferred option would be to install and Alpha 2 26-99 but those are pretty pricey.

GroundUp

The flex lines have a small bore and they have a tremendous pressure drop, as does your tankless water heater. These are the reason you can not get adequate flow in that loop with a small circulator. The flex can stay, but you NEED a larger circulator to overcome the pressure drop. That is not up for debate. There is no reason whatsoever to adjust flow based on water temp. If you're going to keep throwing high dollar parts at this, you probably should have just replaced it with a boiler from the beginning. No matter what you do with this system, the tankless is still the wrong tool for the job

GGross

Oh this is a 199 tankless water heater? I don't know of any that would take less than a grundfos 26-99.

SparkyBeaver

Ok talked to a local rep for Grundfos who was able to find a head loss chart for my Rheem tankless unit.  He confirmed I'll need a bigger pump.

It's about 35' of head for just over 5GPM.  I'm just going to return the little Alpha 15-58 (I'll miss you), bite the bullet and go for the Alpha 2 26-99.  This is still variable speed based on flow needs.  Thank you all for your input.  Will report back once installed.

FYI - don't trust the head calculator on the Alpha 15-58.  It's a nice feature, but was terribly inaccurate.

GroundUp

What's the obsession with the Alpha? The flow needs are constant and will never change. Aside from the $7 a year you'll save on electricity, there is no reason to spend the extra $250 on the Alpha

hot_rod

I would also skip the Alpha 26-99 it will not have any advantage over the 26-99 for pumping a tankless. You don’t have a signal to send it to modulate 

SparkyBeaver

Actually, the primary reason I selected the Alpha2 26-99 over the UPS 26-99 was head capability. According to my design flow estimates, I need about 5.5 - 6 GPM to meet demand if all three zones are calling for heat. I have several areas that were unaccounted for in my original heat loss study that was performed so I'm estimating those areas and will adjust flow based on actual performance. The UPS 26-99 has a max head of 29' on Speed 3.

If the previously posted head drop through the tankless is accurate, it looks like it produces 25' of head around 4.5 GPM. I think that's about where the curves would intersect and max out with that pump. It would probably be sufficient for up to two zones heating, but not all three at once.
The Alpha2 has a max head of 42' and I'm hoping will meet my demand needs of 5.5 GPM on Speed 2, but has a little more capacity if needs are closer to 6GPM. I ordered one of these units that should be arriving tomorrow so I can test it. I didn't want to spend the extra $ for the fun of it, but didn't have the confidence in the UPS 26-99's ability to meet my max flow requirements.


My actual flow needs will vary from about 1GPM (Zone 3 only) up to 5.5 GPM for all three zones. I'm hoping to test the constant pressure feature on the Alpha to only produce as much flow on the primary as needed in the secondary. If I have a pump that isn't able to regulate flow, I'd have to set the primary pump to meet the max demand, correct? In my case, I'd need something producing about 5.5 GPM on the primary side. This would be great for max demand on the secondary side, but less than ideal for only 1 or 2 zones running as it would cause excess heat flow in the primary that would be recirculating on that side. I posed the question earlier if the pump on the primary side is able to "sense" the pressure drop on the secondary side corresponding to how many zones are calling for heat. I don't have the knowledge or experience to know if this is possible between the primary/secondary sides or having a tankless water heater in between the pump and secondary sides. I had an idea of moving the primary pump to the output of the tankless rather than the input side thinking maybe that would allow it to better sense the secondary side needs. The other idea about a primary pump that could monitor water temperature and regulate flow to maintain constant input temperature could be another way to regulate flow rather than pressure. I'm not aware of any circulators that do this though. Any thoughts on these ideas? Again, I understand I'm still a newbie trying to understand the dynamics of hyrdaulics but I'm trying to learn. Some of my unconventional ideas may seem crazy, but I hope it's OK to ask these questions anyway. Thanks again for listening. Not trying to offend anyone here.

GroundUp

You will never need anywhere near 5.5 GPM through the unit- 3 is more than adequate for your heat load. But you already did what you wanted and it'll be okay.

SparkyBeaver

Here's the flow calc sheet I'm using. The numbers in white were actual from the heating load calc. Number's with yellow highlight weren't part of the original study, so I've came up with my own estimates for these. They are smaller areas that are less critical. I'm shooting for a 15 deg delta based on @GroundUp's recommendation to have a lower delta with higher flows for ICF houses. Please let me know if these calcs look incorrect.

GroundUp

The flow in the zones is completely irrelevant to the flow through the unit. 5.7 GPM w/ 15* delta in the zones is only 42,000 BTU. That tankless will easily create 42k at proper temps with less than 2 GPM.

hot_rod

In the primary loop, the loop through the boiler, nothing in the loop would tell the Alpha to modulate its flow. Set it on the fixed speed that gets you the btu output you need.

A variable speed circ on a boiler primary loop gets a signal from the boiler control to modulate. You don’t have that on a tankless water heater. A plain 26-99 is all you needed.

SparkyBeaver

I got my new pump installed on the primary side and it has plenty of flow capacity. However, I'm running into an issue with not enough heat getting over onto the secondary side. Main unit set at 114 deg and secondary input to circulators were only hitting 90 deg with two zones running (~5GPM on secondary). I bumped up the temp to 125 and it went up to maybe 95 deg. Barely any difference. I'm trying to understand the relationship between temperature differences on the two sides. My initial thinking was that if the flows in the primary matched the demand in the secondary side that all the heat would transfer from primary to secondary side. However, based on previous comments from @GroundUp and @Hot_Rod, that doesn't seem to be the case. Could it be an issue with pressure differential in the Tees? My tees are only 3.5" apart, which should meet the criteria for closely spaced on a 1" line. However, I noticed the pressure on the primary side drops down when I have the primary pump running at higher speeds. System pressure is set at 15psi. Although my gauge is not very accurate at lower numbers, it appears to drop down to 5-10psi. Could this pressure drop be creating more recirculating flow in the primary instead of going on to the secondary side when they are calling for heat? What measures can I take to test this and get more heat transfer to the secondary side? What can I do to balance this out? Also wondering if there is still any air in the primary side after changing out the pump? The air separator seemed to bleed off most of the air, but I still hear water "gurgling" when the primary pump is running faster. I expected this to be quieter, but may just not be used to running higher GPM's through the primary side.

New Pump


Tankless Temp / Primary Return Temp


Secondary Circulator input Temp/Pressure


Secondary Return Temps

hot_rod

There is always some temperature mixing going on in closely spaced tees

You can calculate that by knowing the flow rate through the boiler loop and through the heating zones


With multiple zone pumps that zone side will have varying flow rates

If the boiler loop and heat loops were exactly the same gpm, the boiler swt would be going to the system.  That is not common, with multiple pumps 

GroundUp

What he said^^^. You're grossly overpumped on your secondary side when everything is running so perhaps cutting back on that flow and/or increasing the primary flow will reduce the between-the-tees mixing and allow the temps to be more stable.

SparkyBeaver

Another update on my situation. I did some more research and re-piping. I modified the input below the pump to avoid the sharp 90's near the intake. I also added a flow meter on my primary loop because my new pump doesn't show flow rates (Sadly) Per your recommendations, I replaced the SS braided flexible hose on my output with a long sweep PEX pipe.




Also found some of Hot Rods videos online that were helpful. One thing I noticed is that the min distance before and after the close tees should be 8X/4X the pipe diameter. I'm pretty close to 4X on my output before a 90 Bend, but only about 3X on the input. I'm thinking about raising that up to see if I can get the input closer to 8x before an elbow.


Also replaced a few 90's with (2) 45's. Found some 45 deg 1" sharkbite fittings at Lowes on clearance for only $4.92/ea.

However, my biggest issue still seems to be restrictions in my primary loop. I have a Spirovent Jr that should be releasing air from the top of the primary loop. Also, realized that the Watts service valves on my tankless have check valves built into them. Only allows flow from cold to hot. I suspect it's also added significant pressure drop on my primary side. This is preventing me from fully purging air from my lines. I hooked up a sump pump with bucket of water to the output of the tankless service valve and my low point drain to flush water through the system. Was barely getting any flow without even going through the tankless. (~0.25-.5GPM) My goal this weekend is to replace the tankless service valves and get rid of the check valves.


Any other thoughts? More to come after the weekend.

hot_rod

I don't know that I have seen those tankless service valves with checks. You don't need them for your application.

The P/S is not picture perfect but is not a problem for the heater flow.

What is your indication of flow through the heater, with the larger circ?

Sump pumps typically don't develop a lot of pressure, they are more about gpm, high flow rate.

SparkyBeaver

Great News! My system is running pretty well now. I took the tankless valves off and sure enough found a check valve on the cold side only. I was able to pop that out. Following removal, I was able to purge air both directions in my primary loop. I also raised the top elbow to about 8X above the close tees and added a micro-bubbler at the peak. Fired up the system and it started running. Prior to this I was only getting about 0.25 GPM with the new pump. It was initially pumping about 1.5GPM and got up to 3.5GPM after about 5min. You could hear the sound of the pump change and air bleeding off for the first few minutes. Definitely got much quieter. I have the pump in Autoadapt mode and it varies from about 2GPM to 4GPM depending on how many zones are running. Initial heater setting was at 120deg, however, this cause about 4deg overshoot from my T-Stat settings. I've since lowered down the temp to 114deg and will monitor the amount of overshoot caused. Hoping for no more than 2deg. Will keep playing with it to fine tune the system, but it's definitely putting out plenty of heat now! Thanks again for everyone's input on this forum. It's a great place to be.