Tankless limitations/ workarounds?

rbphhc

We've had a Rinnai RU199iP hooked up here for a few weeks now, and the benefits and shortcomings are becoming clear:

Benefits:
- it definitely provides endless hot water (but not limitless) once it gets going;
- it's not running when we don't need it, which is nice;
- it takes up way less space than the old tank, and is quieter.

Shortcomings:
- it takes even longer to get up to temp at the faucet than the old tank, b/c additional few seconds delay launching vent and flame while more cold is running into hot pipes;
- noticeable pressure drop at the hot faucet compared to the tank, b/c restrictions through heat exchanger + possible internal restrictions on flow so the tankless can keep up;

Since we have the benefit of a nice new indirect (NTI S20W) to work into the system, we're looking for ways to use the indirect to make up for the shortcomings of the tankless. Such as:
- tempering/ preheating incoming cold to reduce Delta T;
- using indirect as 20gal buffer, so tankless has a 20gal 'head start' to get going;
- definitely recirculating so it's always hot at the tap.

We've come up with 3 options for how to plumb the indirect to make the most of the tankless, and the least of its shortcomings. Would appreciate anyone's 2¢ (see dwgs):

Option #1:
Run CWS straight to indirect, run DHW straight from indirect to house. Circulate primary pump through tankless to keep indirect hot. Recirculate DHW to/from indirect tank via separate small pump. Basically conventional boiler piping.
Advantages:
- full pressure of CWS feeding DHW - no pressure restrictions through tankless;
- tankless has 20 gal of hot buffer 'ahead' of it, giving ample time get up and running;
- preheated water to tankless allows rapid reheating and tank recovery.
- recirc simple/unrestricted - could run one of those cheapo rigs on a timer;
Disadvantages:
- feeding the tankless hot intake water (Delta of only 10 to 20°f) might make it unhappy. Short-cycling? Increased scale? Not sure what optimal Delta T is for this unit.

Option #2: CWS directly into tankless, but trigger indirect circulation any time DHW runs. That way, tankless is preheated by indirect water, and flow relaxes through unit. DHW conventional off of tankless only. Recirc conventional to/from indirect.
Advantages:
- tankless gets warm intake water but not hot, more volume through, maybe happier?;
- indirect loop gets 'head start' on recovery; no wait for temp drop (we know it's coming);
- having primary circ running w/ DHW might help pressure at fixtures?
- primary loop keeps running after DHW stops, to restore indirect temp - single long burn, maybe tankless happier? Recovery shorter.
Disadvantages:
- no DHW buffer from indirect (although 'indirectly,' b/c storage heat is used to temper incoming CWS, increasing throughput);
- still 2 pumps necessary;
- recirc is still separate system.

Option #3:
Run CWS into tankless, but *mix* with recirc water to preheat. Possibly tee recirc line into CWS via Monoflow? That way we'd get passive recirc anytime DHW is running. Unit would still have full pressure from CWS line. (Might have to oversize Monoflow?) Rest is conventional: run hot out of tankless to indirect, and DHW out of indirect, w/ primary circ pump through tankless to maintain indirect temp. Normal recirc on separate small pump through indirect.
Advantages:
- preheated CWS allows more flow through tankless;
- 'free' recirc anytime DHW is running;
Disadvantages:
- still need 2 pumps, one for primary loop, one for 'offline' recirc;
- unsure how Monoflow might affect CWS pressure, especially w/ multiple fixtures open?

Alternative #3: instead of Monoflow, size up recirc pump and trigger anytime tankless runs to push recirc into CWS intake for tankless. *MIGHT* be able to use single pump in this scenario, if we consider indirect tank + recirc line as ONE buffer to maintain @ temp.

Theses ideas may be stupid and/or crazy. I'm not a plumber, just a lowly GC. Property owner only wants to use 'her guy,' who is semi-retired (her friend of 30+ years). He's willing to do the wrenching if I engineer/ buy/ prep everything. Hence the mess above.

Thanks for any thoughts!

PS
I included a pic of the inside of the NTI S20W b/c I thought it was cool. (Also wanted to inspect pre-install in case any damage, and wanted to see exactly where pipes terminated. The low 'diamond' on is the cold intake, midway up is the recirc line, top is the hot out, and the coil is the boiler I/O (although in this case will be used to heat fluid for in-floor radiant).

Larry Weingarten

Hi, I’m thinking demand controlled recirculation could fix the problems without an indirect. You don’t want to preheat water to the tankless as it may not be able to modulate down enough. Look into gothotwater.com and see if that could be a good fit.
Yours, Larry

hot_rod

Maybe you should just install a tank type DHW, instead of some Frankenstein work around:)

Why try to make a tankless be something it wasn't designed to be with a bunch of expensive components?

rbphhc

Larry Weingarten said:

You don’t want to preheat water to the tankless as it may not be able to modulate down enough.

Look into gothotwater.com and see if that could be a good fit.
Yours, Larry

Thanks Larry!

Interesting stuff. Basically a Grundfos UP15-76 with proprietary controls, and you push a button when you're about to use the hot water? Smart.

Re: preheating:
We are in VT where ground water temps are close to 40°f this time of year. These units are designed to function normally with incoming temps up to 70°f.
Not to mention their own recirculating options that must take incoming water at 110°f or so. So they must have the ability to modulate quite a bit.

I'm curious what the *ideal* operating parameters are for these units. What's the best Delta T for long life, reduced scale, and maximum output?

Maybe it's an 80°f Delta. But if it's more like 40°, I'm cool with mixing in a little recirc or indirect water to approximate Florida.

rbphhc

hot_rod said:

Maybe you should just install a tank type DHW, instead of some Frankenstein work around:)

Why try to make a tankless be something it wasn't designed to be with a bunch of expensive components?

Here's why:

Property owner wanted clear floor space for an addition they are planning. Which dictated tankless wall-mounted only.

They also wanted reliable, so I've been trying to go low-tech. Combis had mixed reviews on reliability, and they seemed like pretty finicky, complex machines. Tankless only has one function, fewer moving parts, and the indirect has zero.

So we are committed to the tankless + indirect Frankenstein. They are already purchased and on site, and the tankless is installed and running well.

But I'm trying to maximize the system here, for reliability and performance, learning from how the tankless is working so far. There's room for improvement.

So. Can we deliver more optimal incoming water temps to the tankless?
Can we simplify indirect circulation to possibly include recirc?
Can we increase pressure at the faucets without losing capacity?

I'm pretty sure there's a way to make the most of this combination of devices to get the best system for the house. Hence the spitballing and brainstorming. Thanks for all the help!

rbphhc

Does anyone like the Monoflow idea? I thought that was maybe a good one.

rbphhc

Larry Weingarten said:

You don’t want to preheat water to the tankless as it may not be able to modulate down enough.

The Rinnai manual shows diminishing returns below a 40°f Delta T. So if we want 130°f out, we want 90°f in. No?
Seems like mixing with indirect or recirc might be a simple way to approach that. But, like it says, I will 'contact Rinnai.'

Larry Weingarten

Hi, I imagined the indirect feeding the tankless, so if the indirect stays at 90F or less you're probably good. About getting even temp and flow at taps, look into lower flow Neoperl aerators and shower heads. They will give even flow over a wide range of pressures.

Yours, Larry

rbphhc

Larry Weingarten said:

Hi, I imagined the indirect feeding the tankless, so if the indirect stays at 90F or less you're probably good.

In Options 2 and 3 above, I have the CWS at the tankless *tempered* by water from the recirc loop or indirect. Not feeding it directly. So it's a mixture of cold and warm that I hope would get us near 70-90°f. My goal is to eliminate any flow restriction or 'slow down' due to high DT.

Any thoughts on either idea? 1/ Monoflow tee into CWS to introduce recirc water passively (and get 'free' recirc); or 2) sync indirect circ pump w/ tankless so they always run together. This would mix indirect H2O w/ incoming cold.

Thanks for any thoughts!

rbphhc

Here's one example of tankless with indirect to provide DHW + radiant heat. This is similar to my proposed setup. (Note: tankless fed off *bottom* of the indirect for warm water in, not hot.)

Jamie Hall

One thing to remember with a tankless is that the "restriction" is not actually a restriction on flow rate, but on temperature rise. Some may restrict the flow rate to maintain the set outlet temperature, however. They also have a minimum flow rate, below which they won't fire at all. That's usually pretty small, but it is a consideration.

Since you have no need to separate the various hot water flows, I don't think I would take advantage of the indirect's heat exchanger capability at all. I'd just treat it as a nice buffer tank, and use the tankless with a pump as the heat source for the buffer tank. Pipe the recirculation line -- and its pump -- in the usual way. Fire up the circulating pump for the tankless with an aquastat on the indirect.

Larry Weingarten

Hi @rbphhc , I like the simplicity of Jamie's suggestion. About trusting a mixing valve to perform nicely, long term, that's not been my experience. Water where I am can be quite hard, and mixing valves all seem to fail sooner or later. Also, with the temperature differential needed to activate the valve, it's like driving a car with a very loose steering wheel. I prefer systems that are so simple you can troubleshoot them in your sleep.

Yours, Larry

rbphhc

Jamie Hall said:

One thing to remember with a tankless is that the "restriction" is not actually a restriction on flow rate, but on temperature rise. Some may restrict the flow rate to maintain the set outlet temperature, however. They also have a minimum flow rate, below which they won't fire at all. That's usually pretty small, but it is a consideration.

Since you have no need to separate the various hot water flows, I don't think I would take advantage of the indirect's heat exchanger capability at all. I'd just treat it as a nice buffer tank, and use the tankless with a pump as the heat source for the buffer tank. Pipe the recirculation line -- and its pump -- in the usual way. Fire up the circulating pump for the tankless with an aquastat on the indirect.

Thanks Jamie! Understood on flow restriction. We are in VT, water temps @ 40°f, so DT is pretty high. At a Delta of 80 or 90°f, the Rinnai can only move 4 to 5gpm max. Hence the desire to preheat a little.

The HX coil in the indirect is for a small in-floor radiant load. I didn't want to complicate the question with that info, but there will be some additional draw on that 'buffer.'

Other than that, I plan to do exactly as you suggest. Primary pump will circulate indirect tank through the tankless. Recirc will be on its own separate circuit through the indirect.

Only tweak I may experiment with is preheating incoming a bit to lower that delta. Any suggestions on how to do that welcome. Thanks!

rbphhc

Larry Weingarten said:

Hi @rbphhc , I like the simplicity of Jamie's suggestion. About trusting a mixing valve to perform nicely, long term, that's not been my experience. Water where I am can be quite hard, and mixing valves all seem to fail sooner or later. Also, with the temperature differential needed to activate the valve, it's like driving a car with a very loose steering wheel. I prefer systems that are so simple you can troubleshoot them in your sleep.

Yours, Larry

Thanks Larry! I very much agree with you and Jamie on simplicity. (Also with Einstein: As simple as possible, but no simpler!)

My preference is for simpler fittings - Monoflows over mixing valves, etc. A nice globe valve can be low-tech but high-performance.

There's still the question of preheating for DHW. A delta of 80 or 90 just seems too high to me. Especially when there's that nice fat tank of heat sitting right nearby.

I continue to think that recirc or indirect flow could be diverted into tankless CWS to raise incoming temps a bit without getting too complicated...

rbphhc

OK, so I spoke to Taco and Rinnai directly. Upshot:

- Yes, the Taco 006e3 is adequate to circulate the buffer tank. At 3gpm the pressure loss through the tankless is @ 10ft of head. The Taco will do twice that on max setting. So plenty of pump, especially only recovering 20 gal.

- The Rinnai likes to see at least a 20°f differential. It prefers more. Shooting for 40°f should be optimal (in at 90°f, out at 130°f).

- YES, they do recommend diverting recirc water into the intake supply. They just don't recommend exceeding that 110° input ceiling temp. I am attaching their own schematic for maintaining direct HW storage off the RU199k tankless (note: red arrow from recirc loop to tankless pump).

So I am gonna go ahead w/ the 006e3 and option #2 above. Which is basically Jamie's recommendation, and similar to Rinnai's schematic. I will mess around with different ideas for pre-heating the tankless intake. Any suggestions appreciated. Thanks!

hot_rod

Vermont is one of 4 states that require mixing valves on all DHW sources. If this is a customers residence, best have one both for code compliance and liability.

No question they need maintenance depending on water hardness. Valve around it knowing this.

To get much output from the coil inside, you may need to maintain the tank at 140 or higher. That coil is designed for 180 f or higher SWT for indirect use. Pulling heat out, the opposite way will take some delta T, depending on what you expect for btu/ hr on the radiant? The only way to increase that output would be raising tank temperature, since your flow is limited by the tankless circulator.

rbphhc

hot_rod said:

Vermont is one of 4 states that require mixing valves on all DHW sources. If this is a customers residence, best have one both for code compliance and liability.

No question they need maintenance depending on water hardness. Valve around it knowing this.

To get much output from the coil inside, you may need to maintain the tank at 140 or higher. That coil is designed for 180 f or higher SWT for indirect use. Pulling heat out, the opposite way will take some delta T, depending on what you expect for btu/ hr on the radiant? The only way to increase that output would be raising tank temperature, since your flow is limited by the tankless circulator.

Thank you, Rod! As usual, extremely helpful.

140° is the ceiling out of the RU199, so that's the system limit. I was planning to ramp up to that if necessary, once the radiant is hooked up.

Running 125° now for DHW, and it's definitely not too hot. We do have one mixing valve ready to go, mare can be gotten if needed.

This property has never had a plumbing inspection, and probably never will. But certainly if we're running 140°, we'll need to temper.

I think I only need 90° for floor temps. Is 120 or 125 in the indirect not enough to keep the floor happy?

And you're right, we will also be limited by the flow of the primary pump. But at 10ft of head, the Taco 006e3 can move 6gpm (maxed out). At 6gpm, the Rinnai can raise water 65°f. Tank is only 20gal, so with no further draw, tank recovery from 70°f to 130°f should happen in 3 to 4 minutes. Call it 10 min. Is that optimistic?

Thanks again, everybody! Super helpful here.

rbphhc

hot_rod said:

To get much output from the coil inside, you may need to maintain the tank at 140 or higher. That coil is designed for 180 f or higher SWT for indirect use.

NTI claims high heat transfer rates for this indirect. It certainly is a big, long coil (see above). Under standard operation, w/ boiler water 200°f @ 15gpm, they say 212 gph output is possible continuously, @ 140° raised from 50° input. That's over 3 gpm at a 90° increase.

My conditions are nothing like this, of course. We've got a 20 gal tank at 140° or less (possibly a lot less, if showers and dishes), with cool water coming into the coil at maybe 70° that we want to raise to 90°.

But all we're trying to get is maybe a 20° increase from 50° hotter water, at only maybe 2 gpm. I was concerned abt overheating floor, hence the mixing valve.

And I'm thinking the indirect may work more efficiently in reverse? The coil holds a small amount of cool water, surrounded by a LOT of hot water. Like putting ice cubes in tea. They melt pretty quick.

The normal way, there's a small amount of hot water trying to heat a lot of cold water. Like Wim Hof in the arctic ocean. That seems inherently like more work.

Anyway, we'll find out. Bottom line, this house ran fine on two 35k btu HWTs. Pretty sure we should be able to get sufficient heat out of more than twice the BTUs.

hot_rod

A bit of a catch 22 with mix valve. Depending on the brand and model, the hot to the valve needs to be 25 degrees hotter than the mixed outlet. So the tankless or tank would need to be 145.
The 520 caleffi is stable with an 18 degree differential. So the tank would need to be 138f.

I think you already experienced the main difference between tank and tankless, the dump load. Tanks don’t need large burners as they act more like a battery. So 35k tanks in 40 or 50 cover the majority of homes in the US. Millions are sold every year.

Keep a 50 gallon tank at 140, mix the outlet to 120 you get more like 65 gallons of tank.

The delay to produce is the biggest complaint. With combi boilers you can maintain boiler temperature. So that takes a good 3 minutes or more out of the wait time that a tankless needs to rev up.

You need to focus on the couple most important functions that you want from this package. As you are discovering you will not get everything you desire from a tankless regardless of the tank interface.

Rinnai actually builds some tankless mounted onto tanks to “fix” some of these short comings in commercial applications where you just can’t transfer enough btus instantly without a 400k burner or stacking 6 tankless, which is the common hotel setup, a bank of tankless to cover morning and afternoon loads.

rbphhc

hot_rod said:

A bit of a catch 22 with mix valve. Depending on the brand and model, the hot to the valve needs to be 25 degrees hotter than the mixed outlet. So the tankless or tank would need to be 145.
The 520 caleffi is stable with an 18 degree differential. So the tank would need to be 138f.

I think you already experienced the main difference between tank and tankless, the dump load. Tanks don’t need large burners as they act more like a battery. So 35k tanks in 40 or 50 cover the majority of homes in the US. Millions are sold every year.

Keep a 50 gallon tank at 140, mix the outlet to 120 you get more like 65 gallons of tank.

The delay to produce is the biggest complaint. With combi boilers you can maintain boiler temperature. So that takes a good 3 minutes or more out of the wait time that a tankless needs to rev up.

You need to focus on the couple most important functions that you want from this package. As you are discovering you will not get everything you desire from a tankless regardless of the tank interface.

Rinnai actually builds some tankless mounted onto tanks to “fix” some of these short comings in commercial applications where you just can’t transfer enough btus instantly without a 400k burner or stacking 6 tankless, which is the common hotel setup, a bank of tankless to cover morning and afternoon loads.

OK, so bottom line: plumb it conventionally, correct?

- DHW off top of indirect;
- make up CWS comes in through tankless, HOT to indirect;
- tankless pulls off bottom of indirect to circ, preheated but not hot;
- Recirc RTN into indirect, simple/ separate/ small pump;

Basically we are building a piggy-back Rinnai. A tank with a tankless.

Only problem I see with this approach is not having enough Delta T in circ mode at the tankless. I.e. if we are maintaining 140°f in the indirect, low limit has to be 120°f. Otherwise DHW too cold. But that's *barely* the minimum differential the tankless wants to see. They actually recommend 25 or 30° - maybe pulling off the very bottom will accomplish that - maybe it's 125 to DHW and 115° to tankless when sensor reads 120°...

Hey, Rod, we did look seriously at the Polaris tanks you suggested months ago, and I thought they were perfect, but owner balked on size and cost.

Ah well. Onward!

Jamie Hall

You do have a potential problem with the delta T for the tankless, and i honestly don't see a good way around that. Perhaps if you got a little more complicated and arranged the piping so the makeup water cold feed went into the cold water pipe from the indirect to the pump, it would at least ensure that when you were drawing hot water for, for instance, a shower, at least some of it would go through the tankless... giving you more delta T for the tankless.

Frankly, whatever you come up with is going to be a kludge, since you are using the tankless in application for which it was never intended. Some fiddling may get it to work... one major thought: when you do get it to work, assuming you do, document in writing very thoroughly exactly what was tried and didn't work, and exactly how it is set up to work, and attach the documentation securely somewhere on the whole rig -- so someone coming after you doesn't have to go through the whole process again.

hot_rod

rbphhc said:

hot_rod said:

A bit of a catch 22 with mix valve. Depending on the brand and model, the hot to the valve needs to be 25 degrees hotter than the mixed outlet. So the tankless or tank would need to be 145.
The 520 caleffi is stable with an 18 degree differential. So the tank would need to be 138f.

I think you already experienced the main difference between tank and tankless, the dump load. Tanks don’t need large burners as they act more like a battery. So 35k tanks in 40 or 50 cover the majority of homes in the US. Millions are sold every year.

Keep a 50 gallon tank at 140, mix the outlet to 120 you get more like 65 gallons of tank.

The delay to produce is the biggest complaint. With combi boilers you can maintain boiler temperature. So that takes a good 3 minutes or more out of the wait time that a tankless needs to rev up.

You need to focus on the couple most important functions that you want from this package. As you are discovering you will not get everything you desire from a tankless regardless of the tank interface.

Rinnai actually builds some tankless mounted onto tanks to “fix” some of these short comings in commercial applications where you just can’t transfer enough btus instantly without a 400k burner or stacking 6 tankless, which is the common hotel setup, a bank of tankless to cover morning and afternoon loads.

OK, so bottom line: plumb it conventionally, correct?

- DHW off top of indirect;
- make up CWS comes in through tankless, HOT to indirect;
- tankless pulls off bottom of indirect to circ, preheated but not hot;
- Recirc RTN into indirect, simple/ separate/ small pump;

Basically we are building a piggy-back Rinnai. A tank with a tankless.

Only problem I see with this approach is not having enough Delta T in circ mode at the tankless. I.e. if we are maintaining 140°f in the indirect, low limit has to be 120°f. Otherwise DHW too cold. But that's *barely* the minimum differential the tankless wants to see. They actually recommend 25 or 30° - maybe pulling off the very bottom will accomplish that - maybe it's 125 to DHW and 115° to tankless when sensor reads 120°...

Hey, Rod, we did look seriously at the Polaris tanks you suggested months ago, and I thought they were perfect, but owner balked on size and cost.

Ah well. Onward!

Is this homebuilt, one of a kind actually cheaper than an off the shelf Polaris, or HTP tank?

I admire the persistence but you are taking a simple task of heating water and making it overly complicated in my mind

rbphhc

Jamie Hall said:

You do have a potential problem with the delta T for the tankless, and i honestly don't see a good way around that. Perhaps if you got a little more complicated and arranged the piping so the makeup water cold feed went into the cold water pipe from the indirect to the pump, it would at least ensure that when you were drawing hot water for, for instance, a shower, at least some of it would go through the tankless... giving you more delta T for the tankless.

Frankly, whatever you come up with is going to be a kludge, since you are using the tankless in application for which it was never intended. Some fiddling may get it to work... one major thought: when you do get it to work, assuming you do, document in writing very thoroughly exactly what was tried and didn't work, and exactly how it is set up to work, and attach the documentation securely somewhere on the whole rig -- so someone coming after you doesn't have to go through the whole process again.

Thanks Jamie, I appreciate the confirmation.

The weird thing is, even the thing tankless was DESIGNED for doesn't work that well. I mean - triggered by water flow means there's at least several seconds of cold water flowing through before the thing gets up to temp; and even then there's a ceiling on temp rise and flow rate. Not to mention the exorbitant gas use.

So, to me, there's just a device that works this particular way. It has a flame and a heat exchanger, and it's triggered by flow. Let's figure out how to use THAT thing best. An indirect or buffer tank seems like a good start.

Re: Delta T: Rinnai said optimal differential would be 40°. So *if* we want to maintain the indirect at 140°, the tankless wants to see 100°f.

In DHW mode, that means temper the incoming 45° w/ something hot.

In Circ mode, that means temper the incoming 115° w/ something cold.

Sure sounds like a mixing valve to me. Only problem is, it's an open system on DHW, and closed on Circ. So it's difficult to mix them.

UNLESS I let CWS into the indirect. Then we are using the tankless exclusively for circulating. Tank does all the instant stuff.

THEN we could rely on natural *stratification.* Set the delta at 20°f, but in reality we might see 110° at the bottom feeding the tankless. Even less if there is significant draw, like DHW and radiant at the same time.

And that's the system that Rinnai themselves have drawn up (above).

ROGER on documenting. Will certainly write up, especially if I can make this all work WELL.

BTW, any recommendations on water treatment/ descaler pre-tankless? The supplier didn't have one, and I want to make sure we are conditioning somewhat before this thing gets caked up.

rbphhc

hot_rod said:

Is this homebuilt, one of a kind actually cheaper than an off the shelf Polaris, or HTP tank?

I admire the persistence but you are taking a simple task of heating water and making it overly complicated in my mind

I agree. But working within parameters I don't control: building owner, plumber, and future goals. Trying to do the best I can within those, and come up with a system that does both DHW and radiant effectively, reliably, and transparently.

rbphhc

OK, so this is a direct translation of the Rinnai schematic to my application.

DHW mode:
- open fixture pulls HOT off the top of the indirect;
- make-up cold enters the system, forced through the tankless;
- HOT enters the IND at R position, mid-tank (it's cold for a few seconds);
- this continues until fixture shut.

CIRC mode:
- IND tank hits low limit, calls for heat;
- circ pump kicks on and forces *IND tank bottom* water through tankless;
- HOT flows into IND at R position until tank recovered.

Recirc mode:
- sensor or timer kicks on secondary pump;
- pump pulls HOT of top of IND and pushes in at R position until recovered.

Problems:
- Inlet COLD still too cold in DHW mode;
- Inlet COLD still too hot in CIRC mode.

Possible solutions too COLD:
1/ trigger primary circulator to run whenever tankless runs; this would force incoming C to mix with bottom-of-IND water - 50 + 120 = 85?
2/ tee Recirc water into CWS pre-tankless with Monoflow - this would draw warm into the cold inlet stream passively;

Possible solutions too HOT:
1/ set lower limit out of INDirect; tank stratification should keep top of tank at @120 even when bottom hits 100; set limit for 110? That's a 30° Delta.

Thoughts? Thanks!

hot_rod

Isn't your recirc and tankless circulator in series? Are they both the same model? Head doubles if both run at the same time? You could end up with some high velocities in the DHW piping. This graph shows that relation ship.
Ideally you would plot a system curve, blue line on this graph to determine the OP, operating point.

Idronics 16 shows you how to develop and plot a system curve.

The pressure drop is the entire circuit, not just the tankless coil.


If its copper you may get erosion, hot, hard water is very aggressive to copper tube.

If it is pex, you are limited to temperature you can run continuously with hot, chlorinated water. If you have any pex in the DHW loop look for the chlorine rating on the tube.

You're building a complex hydronic loop, really, disguised as a recirc loop

The tank bypass line also bypasses the 3 way valve? its not clear which is H, C and mix ports. If so, that is where you could get the high temperature into the loop.

The mix valve needs checks on H & C ports. Some models include checks.

rbphhc

hot_rod said:

You're building a complex hydronic loop, really, disguised as a recirc loop

Crap, I hope not! Couple lines drawn wrong above, thanks for catching.

Here's a simplified dwg of the 3 modes - does this still look crazy?

(The recirc-return was supposed to tee into the R line, not the Cold. Although Rinnai does have recirc coming into the cold, and yes their pumps are in series!)

hot_rod

Is a homeowner insisting you make this concept work? Got a plan B?

I think the tank and tankless can work ok. Complication sets in when you add recirc, a mixing valve and the need to pull radiant off a bottom coil in the tank at a required temperature.

Yet another potential issue, when you recirc thru a mixing valve the need to have a specific piping.
Basically what happens after a period of non use, no cold coming into the C port of the mix valve. The cold port sees warm water hitting the C side, from the recirc loop, and the valve cannot control temperature, no cold is entering to mix down the 140. It’s called temperature creep. So someone opens a faucet in the morning and the hot water will be whatever the tank temperature is, until the C port gets hit with 50 degree water so the valve gets its mind back.

There is a piping fix that involves two balancing valves. So the goal is to move just enough gpm thru the recirc loop to overcome the temperature drop. The rest of the flow goes back to the tank

Another fix is to have point of use mixing valves at ever HW outlet

With your system needing flow thru the tankless, and different flow rates depending on one or two circulators running, I not sure how you will tame the beast? If you knew the flow rate, a PIC valve would be a solution, an automatic pressure independent balancing valve

A motorized mixing valve like the legiomix could be a solution also as it can close either port 100%

Download Idronics 11 for some more info on recirc and tankless with preheated water supply.

A few pages from that issue here.

Cody does a good, quick job of explaining temperature creep, find this at the Caleffi you tube channel

rbphhc

hot_rod said:

Is a homeowner insisting you make this concept work? Got a plan B?

I think the tank and tankless can work ok. Complication sets in when you add recirc, a mixing valve and the need to pull radiant off a bottom coil in the tank at a required temperature.

Plan B is the original Plan A:
Tankless functions just as a tankless - CWS in, DHW out.
Indirect just does radiant, with a single circ pump.
Everything is at 120, except floor at 90.
No mixing valves.
Recirc just simple 120 to/from tank.

Should I go back to that?
One downside: that initial slug of cold off the tankless.
Is there any way to divert that until running hot?
Another downside: floor temps limited by that 120°.

Tankless functions fine right now as simple tankless, EXCEPT slow to get going. Recirc would fix most of that.

hot_rod

As I see it you will not check all the boxes on your wish list. Systems don’t always work like the lines on the drawing show.

So pick the most important.

Eliminating the cold sandwich effect? Not waiting the ramp up time?
A dump load beyond what the tankless gpm can provide?
Recirculating for instant hot at every faucet?
Dual purpose DHW and radiant?
Complexity and cost?
Serviceability?
Proving a point🥹

If your name is on this system, liability pass through to future owners, as far as codes and safety. How’s your liability insurance capped? Scald/ soft tissue burn lawsuits can run into the millions, attorneys love them. SAD!

rbphhc

hot_rod said:

As I see it you will not check all the boxes on your wish list. Systems don’t always work like the lines on the drawing show.

So pick the most important.

Eliminating the cold sandwich effect? Not waiting the ramp up time?
A dump load beyond what the tankless gpm can provide?
Recirculating for instant hot at every faucet?
Dual purpose DHW and radiant?
Complexity and cost?
Serviceability?
Proving a point🥹

If your name is on this system, liability pass through to future owners, as far as codes and safety. How’s your liability insurance capped? Scald/ soft tissue burn lawsuits can run into the millions, attorneys love them. SAD!

Definitely dual purpose DHW and radiant.
Beyond that, it would be nice to have hot when you open the tap. Without unexpected cold a minute later.

If I limit the system to 120, I don't have to worry about mixing valves or lawsuits. Floor might be a little cooler, but we can live with that. I think that might be the sweet spot.

rbphhc

MikeAmann said:

For the recirc, what about using a very small pump and a restrictor in that line (say 1/8" dia) operated by a timer to run for xx minutes every so often? Keep it simple!

Definitely planning very simple recirc, tank-to-tank, small pump, timer. Thx!

john123

Gosh, I'm so confused. Can someone first explain the downside of simply running the dhw from the tankless into the indirect and then running the further heated dhw from indirect into the house system. Even if there was a huge pressure drop through the tankless, could you fix that with closely spaced T's (and the tankless on a "secondary" loop including the tankless)? no valves, no pumps no complications?

hot_rod

If the tankless is able to handle the recirc, this should work. Is there a way to make the tankless operate with an aquastat call, or just when it sees flow? Some how the recirc would need to run always to maintain the tank temperature, and the instant hot function. What happens if the pump times out all night? Tank would drop 8-10 degrees perhaps.

Maybe pull up the wiring schematic on the Rinnai that is factory built with the tank connected.see how the control logic works.

HX life expectancy could be shorter if you recirc 24/7 they are small thin tubes intended for intermittent flows. That is the big difference between a boiler and tankless, but you know that by now.

Other than not being legal, in Vermont, I think this will work. It’s an easy enough piping to try and see.

rbphhc

john123 said:

Gosh, I'm so confused. Can someone first explain the downside of simply running the dhw from the tankless into the indirect and then running the further heated dhw from indirect into the house system.

That was my original design, way back in the paleolithic era, a couple months ago. Plan was that DHW usage would recharge the indirect. No pumps, no recirc.

But there's a radiant in-floor loop running through the indirect coil. Only a single floor, but there's draw on that indirect unrelated to DHW use. I.e. the radiant usage could cool the indirect between DHW usage (at least in winter).

So my next modification was a brazed plate HX between the in and out of the indirect (attached). That would assure hot coming out of the indirect (scavenged from tankless feed input). Problem with that is, even less hot entering the indirect. Same problem of radiant draw without recovery.

Hence the circ pump.

THEN, after installing the tankless, noticed how long it takes to get DHW at the faucets. Same long run of cold pipe, PLUS the tankless ramp-up time. Way too much water waste.

Hence the recirc pump.

I think Rod and Jamie and Larry (and Rinnai) are all right: do it simple, conventional, standard. Which means two loops, two pumps, DHW off the indirect off the tankless.

Which is pretty close to your suggestion (and my original plan) - the tankless is still pushing the DHW through the indirect. But we added pumps to keep the floor and faucets happy.

I am going to start the system at 120° and see how that works - temps going to the floor, recovery time for the tank. If I need to raise temps significantly, there's a Taco 5120-2 standing by. That may or may not cut it.

But I think at this point there's a reliable design. We're gonna build it and see how it performs. My hunch is, we'll have to top it out at 140. And then we'll need the mixing valve, and we'll have issues with recirc running too hot b/c no CWS.

But one problem at a time. Let's run this at 120-ish and see what happens. Thanks!

rbphhc

hot_rod said:

If the tankless is able to handle the recirc, this should work. Is there a way to make the tankless operate with an aquastat call, or just when it sees flow? Some how the recirc would need to run always to maintain the tank temperature, and the instant hot function. What happens if the pump times out all night? Tank would drop 8-10 degrees perhaps.

Maybe pull up the wiring schematic on the Rinnai that is factory built with the tank connected.see how the control logic works.

HX life expectancy could be shorter if you recirc 24/7 they are small thin tubes intended for intermittent flows. That is the big difference between a boiler and tankless, but you know that by now.

Other than not being legal, in Vermont, I think this will work. It’s an easy enough piping to try and see.

Hey Rod! So you're saying one big loop, one pump, circulate often, that'll maintain indirect and HW lines?

Is that different from my original Alternative #3 above? "Size up recirc pump and trigger anytime tankless runs to push recirc into CWS intake for tankless. *MIGHT* be able to use single pump in this scenario, if we consider indirect tank + recirc line as ONE buffer to maintain @ temp."

rbphhc

One workaround for too-high-recirc temps: pull the recirc loop off the BOTTOM of the indirect. Splice it into the hot line. If the top is at 140, I'm expecting the bottom to be @ 120. That way we can bypass mixing valve.

Hot_water_fan

I hope you bill for this time 

rbphhc

Hot_water_fan said:

I hope you bill for this time 

I was until last week. Now I'm just in it for myself. (As Rod said: Prove a point!)

rbphhc

Btw, Taco claims the 5120 only needs a 10° difference between hot and mix. Anyone know if that's true?

(also, I don't know who wrote this for them, but 3*f is not 3.4*c.)

john123

Suppose you ran the output from the tankless into a stand-alone hot water tank (either electric or gas/propane) and then ran the output from the tank into the house dhw system, what problems might you face? possible low pressure at the taps? If so could you fix it with a primary/secondary loop and CST's? Do tankless hot water heaters usually have supplementary circulator pumps?