I have an open radiant system with 7/8" PEX tubing in three zones, each with a low volume circulating pump:

Zone 1 is stapled between joists under floor with three circuits (600')
Zone 2 is in a newly poured concrete slab (300')
Zone 3 is upstairs, stapled between joists (200')

A Takagi TK-3 runs both my domestic hot water and radiant heating system.

We installed the radiant system ourselves, DIY style, as part of a major home renovation. We installed Zone 1 first, in the existing portion of our house, and ran it by itself for a year with no problems. Then we added a two story addition which included Zones 2 and 3. All three zones worked perfectly by themselves. However, we could not get all three zones to heat together. Zone 1 (the largest zone by volume) would "steal" hot water from the the other two zones. The other two zones would not heat, and I could actually put my hand on the supply and return lines and could feel they were cold while Zone 1 supply and return were hot / warm. If I manually closed Zone 1, I could instantly feel the supply lines to Zones 2 and 3 warm up.

The Radiant Floor Company tech suggested that there was not enough hot water supply to meet the demand of all three zones and, because the head pressure was lower, Zone 1 would steal from the others. However, our TK-3 should be able to produce enough hot water for our system so it was a mystery. After a year of manually turning each zone on / off - and wasting tremendous amount of energy - we finally figured out that our household pressure was too low for the Takagi which requires 50psi or above (our household pressure is at about 25psi coming from a community well)

We installed an inline booster pump. It is an "on demand" booster pump with a small priming tank that automatically kicks in when there is water flow or when the PSI drops below 30psi. At the same time we installed an expansion tank which was accidentally omitted from the initial installation because of confusion over our community well vs private well. Prior to installing the expansion tank our Takagi would turn on momentarily every time we flushed a toilet. Also, because our PEX tubing does not have an oxygen barrier, we have rust stains in our brand new toilets!!!! But, I digress...

Once we installed the booster pump all three zones worked together! However, our problems were not solved. Immediately after installing the booster pump and expansion tank, we were getting blazing hot water in our cold water domestic fixtures. After consulting with the Radiant Floor Company, we installed a check valve on our cold water inlet pipe which solved that problem but created a new one! Now, when we turn on the domestic hot water, it takes 1-2 minutes for the booster pump to kick in because it bleeds pressure from the radiant system before creating enough flow to trigger the pump. (The booster pump turns on instantly even with low flow from my cold water fixtures). That would not be a problem except, if the hot water is turned off before the pump kicks in, and the pressure remains below 50psi, our zones will no longer heat together!!!

The Radiant Floor Company has suggested we install a traditional tank style pressure boosting system but it is expensive and we are not convinced it would solve our problems because there are other, harder to explain, odd and intermittent problems with how water flows through our system (for example we get hot water flowing through the cold water side our domestic mixing valve). So, this is a long story to say that we have decided to switch from the open system to either a heat exchanger system or a closed system. I'm hoping to get some advice on which way to go.

Heat exchanger would separate domestic hot water from radiant heating system while still using one heat source, our Takagai TK-3. But there is some heat loss in the exchange. Parts would cost about $1335.

Closed system would mean that we simply purchase another heat source, either for our domestic hot water or for the radiant heating system. Run the systems independently.

Advice?

what water temp are you heating with?

Our Takagi is set at 131F

can you describe the construction quality of the areas in question? and is that a total of 1100 square feet, or is that footage linear feet of pex? if pex, please provide square footage for orientation. Finally, where are you located?

I'm in the San Bernardino Mountains of southern California, at 5300 feet elevation (we get real winters here with lots of snow). We have R13 insulation in walls of existing structure. R19 in addition. R19 in attic along with perforated radiant barrier. Floor in existing structure have double bubble radiant barrier. Slab is insulated with 2" rigid foam boards.

I gave linear feet of PEX but square footage is about the same, actually. 650 square feet in zone 1, 300 square feet in zone 2 and 250 square feet in zone 3.

I should mention that when all three zones are working, our house is toasty warm!

seems pretty unlikely that system pressure was your problem unless you had some really freaky intermittent pressure related air problem. Much more likely is that your circulators were "fighting" and the low head circuit won, so the suction of that pump prevented the others from flowing appropriately. If hot water went to one zone but not another, that's flow, not the firing or output of the takagi. If the takagi was not firing or didn't have the output, ALL the water would be colder than desired.

The booster pump is probably helping your system hydraulically, in other words it probably helped induce some flow . I would guess your piping isn't really appropriate for this kind of a setup and the booster but that you more or less accidentally got it to do some of your circulation pumping.

regardless, going closed is ALWAYS the right choice for any significant heating system... but depending on piping it may or may not solve this issue... depends how much of the issue is the huge pressure drop in the takagi. Note also in some parts of CA you will be expected to use a double wall heat exchanger if you do that route. that may escalate your cost.

I don't really like using on demands for heating, but since you already own it I'd probably do the heat exchange route. starting fresh, unless your DHW demand is very high, I'd probably have chosen a tank water heater with heat exchange.

Thanks, Rob. Agreed that the circulators were fighting and that the low head circuit won, but I'd like some clarification on the rest of what you've suggested.

The booster pump was added not to bring the system pressure up but to provide the pressure required for the Takagi to provide maximum flow, per their manual. By getting more flow from the Takagi we expected to solve the war of the circulators. It seems to have worked but as I understand you are saying that is possibly coincidental? Can you clarify for me?

Also, I don't think the booster pump can help with circulation pumping because it only turns on when a domestic hot water tap is open. But I'm curious what you mean when you say it is not appropriate for this kind of setup? Obviously, it isn't working with the radiant system because the radiant system inhibits flow from triggering the pump. Are you saying something else could be contributing the problem?

My biggest concern is your comment that going with a closed system might not solve our problem. I'd really like to know what issues we might face? The last thing we want to do is open yet another can of worms. Are we more likely to face problems with closed or with heat exchanger?

We don't have room for a tank style heater - our heating system is housed in an exterior "closet" cuz we don't have a finished basement like most homes back East...but if we go with a closed system I'm curious what heat source you would recommend? Use the TK-3 or something else?

Your problem was not insufficient flow through the takagi in heating mode, and if it was, a boost pump that doesn't run in heating mode couldn't solve it. At least, by what I get of your description of the problem, it couldn't have been insufficient flow.

Your problem was the flow was not going where it needed to go. again, if it were insufficient flow, all manifold legs would be cold, you wouldn't have some hot and some cold legs. Also, the booster pump could not help with flow since, as you note, it only runs when a fixture is open. so that pump should have nothing to do with your heating system or its operation.

so without knowing exactly what the factors causing the flow problem were, it's hard to guarantee that you won't see them again... *probably* not but it sounds like something is screwy with your piping. can you post a sketch of your piping? that would help.

Problems with a tankless water heater on a radiant floor? NO

When misapplying gas-fired equipment extra care must be taken. As Rob suggests, flow is an issue since the Takagi was engineered for domestic hot water service, which only remotely related to hydronic heating. The reason so many of these tankless jobs do not work up to expectations is the lack of appropriate hydronic design. Real hydronic people don't use them. The people you buy the equipment from should have the answers.

Naturally, the red flag should have gone up when an "open" combi-system was proposed at the onset. Since open water/space heating systems are not approved under any building, plumbing or mechanical code in N. America, the proper way to do it doesn't exist.

"so it was a mystery" defines much of what is available in the DIY radiant floor heating market. Your results are unfortunately typical. I would suggest a proper design using industry accepted practices. It should be enough to save money on DIY doing it the right way without the risk and long term hassle of using code defying design.

If others buy such systems they should have a heat load, near piping, CAD drawing and critical equipment specification before giving the online guys their credit card.

Though we rarely find a place of tankless water heaters (never as boilers), we do, on occasion, spec a true combi-boiler with the emphasis on the major load (usually heating) incorporating built-in weather sensitive controls and other features more appropriate to space heating. The advantages of a tankless are still available and properly engineered into the package instead of trying to make it work after the fact.

Another alarm bell is the proposition that 7/8" tube (non=-barrier as I recall) is appropriate for any residential radiant floor heating system in America. It is not by any accepted standard. If it is non-barrier pipe, keep in mind that you will have to isolate the pipe from any conventional (ferrous) hydronic heating components. If you have an open system we assume you have bronze pumps (three times the cost of conventional) and other like components compatible with potable water.

In the long term, a tankless water heater is not the short-cut (cheap) solution.

I think it pays to follow the code and conventional trade practice even (or more especially when) you are trying to avoid them. As you can see your system is far from simple, easy or cheap. I do not comment to belittle or berate, rather to warn and educate others.

Start with a proper heat load. If your supplier or installer can't produce it, move on. In this case, considering the low water pressure, an standard condensing boiler with indirect water heater would have been in the plan or perhaps a condensing water heater with a sub-assembly including an isolating SS heat exchanger. In no case (never non-barrier) PEX tubing stapled up below a sub-floor in radiant heating. This is "accepted" in some corners but not in ours.

I get these calls every week and it is quite disheartening. I feel your pain.

Actually, the Takagi itself requires 40psi or above for maximum flow. So simply by raising the pressure in the system, the Takagi does produce more flow. Once the pressure is raised, as long as the system remains closed, the pump does not need to keep running. The pressure stays above 40psi. That's actually why we think a closed system WILL work. There is definitely a cause / effect. When pressure is above 50psi all three zones have flow. When pressure is below 40psi zone 1 steals from the other two.

But I'm definitely not convinced that all our problems are solved because of the way that hot water wants to flow backwards into the cold water inlet, thus requiring the check valve. And the fact that hot water flows through the cold water side of our domestic mixing valve. Plus there are other wacky happenings. That's actually why we want to separate the systems. But want to be sure that either a heat exchanger or closed system will work for us so any further insight is appreciated!

I don't know how to post a sketch of my piping. Wish I could. This terrible diagram will have to suffice:

Diagram erased because it didn't display correctly!

Send a drawing direct to us.

Send the plan directly, I will try to help.

Hey there Badger,

I've read a lot of pros comments on various message board complaining about the DIY radiant systems that are sold to unsuspecting folks. Unfortunately, most of us hear of the problems after we have installed our systems so it feels kinda like punching someone when they are down to hear the lectures. I realize now, in retrospect, that there's a lot we could have done differently. Personally, the non barrier tubing is the thing that makes me the most upset! (see my comment on stains in my toilet. also copper patina stains on the white floor tiles in my shower) That said, we didn't choose the system because it was cheap. We chose it because we could install it ourselves and saving the labor is what we wanted to do. We didn't think we were compromising on quality of the components. We have done a major renovation, all up to code with only a few outside contractors when needed. Also, in my neck of the woods, it's hard to find radiant heating pros!

That said, when all three zones are running properly our system heats our house very well. Our Takagi seems to produce enough hot water as long as our domestic pressure is above 50psi. (if we always wait for the pump to kick on after opening a domestic tap our heat never fails - but we do have ongoing issues with our domestic hot water such as low pressure despite the booster pump!) Thus, again, we'd like to separate the systems! If we go to a closed system, I'd appreciate suggestions on a heat source other than the TK-3 if there is one? (but it has to be tankless). I'd also appreciate any advice on closed vs. heat exchanger systems.

Thanks!

Hey again, Badger!

Just saw your last posts. Will send diagram / photos your way.

Thanks!

the PSI only helps flow in DOMESTIC USAGE mode. it should have no effect whatsoever in heating mode and should not make any difference to the inoperable zones. If your standing PSI makes a difference in heating mode, then you've got some serious issue to deal with.. leaks, incoming air paths, bubbles impeding flow, hard to say but something like that would have to be the case.

You already own the tk-3, you might as well use it. switching to another tankless won't help you.

Also, the pipe not having an oxygen barrier, in the current configuration, is not a problem. you are bringing fresh water in constantly, an oxygen barrier would have been a waste of money for that because you are bringing in oxygen every time you open a tap. If your pumps are not cast iron and your expansion tank is a potable unit, you should be ok as long as you don't need glycol. all of that holds true in any system configuration you adopt from here on out.

Hey Rob. Thanks again for your insight. It's really appreciated. But it does make my head spin as I figured we had isolated the flow problem by raising pressure. I've inspected the PEX that is stapled under the house and the upstairs bedroom and found no leaks. I suppose we could have a leak in the slab but if we don't open a domestic fixture we don't lose pressure. Will test it again to be sure.

The cost to install the heat exchanger is about the same as the cost to add a second heat source. That's why I'm asking for reccomends on a heat source. But honestly, based on your comments I'm at a loss.

Pumps are stainless and expansion tank is potable. I'm not planning on replacing the tubing so I guess it's good that the oxygen barrier is irrelevant.

well, I can only guess without a diagram that whatever you did to put the boost pump in may have changed your piping in a useful way? the system PSI is the force that drives flow when you open a tap. but where there is no tap open, it doesn't move anything. the most it can do when the system is just recirculating is keep air in solution. but if you were at 25 PSI before, it's pretty hard to think you had an air solution problem...

the heat exchanger will have a much better lifecycle cost than a new heat source.

I understand you perfectly. I have the same issues within the trades, to which I belong. I come from a very rural area where codes are unheard of, let alone followed. I am particularly upset about gimmicks that take advantage of the uninformed, regardless of the buyers primary objective or motivation.

I am very pro-DIY - with reservations - gleaned from many years in the industry.

Waste and bad performance is what grinds my corn.

We can all use non-barrier tube, in fact most of who started in the 80s and 90s had nothing else, but today, as you can see, it limits your choices. The Takagi can work, perhaps best -dare I say - with an indirect or plate heat exchanger. You will have to add an expansion tank - potable water compliant.

I am glad the system meets the heat load without the benefit of plates. The systems that don't are expensive to fix.

One issue for you to look at is that there isa max flow through the takagi. The three pumps working together, and at a lower static pressure, could be way too high for the fluid to flow through the takagi with out a lot of cavitation in the pumps. It is quite possible that you will need to run a separate loop to a small tank, with a bronze pump, then run from the small tank to the heating loads with their own pumps. The t-stat on the tank will dictate when the pump turns on and heat the tank. The system is then hydraulically isolated (as much as you can) from the DHW flow without the use of a HX. Not ideal by any means, but I just fixed one like yours and this will work. I suspect that you do have some ferrous components somewhere. What are the pumps?

Rob and Badger - sounds like two votes (but not without serious reservations) for heat exchanger over 2nd heat source. and a good point re: lifecycle cost. i assume that is because the heat exchanger will work more or less forever but the heat source(s) need to be replaced.

installing the booster pump was very simple - it just goes on the main in line. here's a very simple diagram of our system:

water from main --> booster pump --> T splitting domestic cold / cold inlet to radiant (and also domestic hot water) --> radiant supply line to 3 zone manifold with 3x circulating pumps --> radiant tubing --> radiant return --> expansion tank --> Takagi --> pressure relief valve --> T splitting radiant mixing valve to radiant supply / domestic mixing valve to hot water fixtures

Mike - i don't totally understand what you are saying but i believe you are suggesting some form of "primary loop" system which we thought would be a solution but the Radiant Floor Company said would not help us. so we moved on to other options. but based on your success we may reconsider. any idea if raising our standing PSI would have an impact on this situation? i'd like to somehow connect the dots on why that seems to have helped us!

our circulating pumps are stainless Grundfos Model UP15. the only ferrous components that i know of are our household fixtures. mixing valves are honeywell am-1. expansion tank is amtrol proflo pftxt5. supply / return lines and connections are copper.