Tankless water heater flow

[Terry Webb]

I recently put in a new tankless hot water heater the flow was very very low 1.5 gallons per me minute when I turned on the tub the water was really low when I turned on the tub and the bathroom sink cut it in half I call the manufacture and they sent me a new motherboard water Flo still the same when I installed the hot heater I used PEX waterlines to copper and I used gator bite fittings to include the valves could they possibly have slow down the flow of water I'm running out of options. There are no crossovers the filter is perfectly clean and the cold waterside is perfect

 

[Dana]

Is there a model number or other information to share on this unit?

Based on the behavior it's probably a lower-wattage electric tankless, better suited for the incoming water temperatures of tropical climates than New Jersey's water temps.

These things are designed to throttle back the flow to regulate the temperature once it has reached maximum power, so when the shower is running and hot water is turned on at the sink the total flow of sink + shower drops back to whatever it was delivering to the shower.

 

[Sponsor]

 

[Terry Webb]

Is there a model number or other information to share on this unit?

Based on the behavior it's probably a lower-wattage electric tankless, better suited for the incoming water temperatures of tropical climates than New Jersey's water temps.

These things are designed to throttle back the flow to regulate the temperature once it has reached maximum power, so when the shower is running and hot water is turned on at the sink the total flow of sink + shower drops back to whatever it was delivering to the shower.

The hot water heater is a Takagi model H3JDVN. The manufacture told me yesterday that up I opened up the relief valve that I should be getting at least 4 gallons per minute the unit is designed to give a maximum of 6.6 gallons I understand the difference in flow when you have more than one faucet on but with just a tub running by itself I get 1.2 gallons of water

 

[Dana]

The T-H3J takes 160,000 BTU/hr in at the high fire, and ~150K out, which is enough to deliver a decent amount of flow. Even with 35F incoming water you should be able to muster 4 gpm of output at 105F shower temperatures, a bit less at 110F (tub-filling temp) or higher.

I assume you're testing the flow at the tub/shower with a bucket and stopwatch?

The anti-scald valves on shower mixers sometimes interact with gas tankless units due to the hot-side to cold side pressure imbalance created by the impedance of the heat exchanger in the tankless, which may be part of the problem. If you have a hot water tap that isn't on a mixer (say, the connection to a clothes washer), test the hot-only flow.

If it's an anti-scald valve interaction you can sometimes fix it by raising the programmed output temperature of the tankless so that it requires less flow (and less of a pressure drop) through the tankless. The max temp setting on the T-H3J is 140F, but if you set it to higher than 120F it's wise to put a thermostatic mixing valve on the output plumbing to mix it down to < 120F before the hot water distribution plumbing. Sometimes setting the temp too high results in too low of a flow in the tankless for it to fire up reliably with a low-flow showerhead- if that happens there may be some tweaking to do to find the right balance.

If 105-107F is acceptable for your other hot water uses, you could turn the temperature DOWN to that range, and adjust the stops on the shower mixer to let it run full-hot rather than always mixing in some cold. That too will sometimes fix anti-scald valve interactions.

 

[Terry Webb]

The hot water heater is a Takagi model H3jdvn. The manufacture told me yesterday that up I opened up the relief valve that I should be getting at least 4 gallons per minute the unit is designed to give a maximum of 6.6 gallons I understand the difference in flow when you have more than one faucet on but with just a tub running by itself I get 1.2 gallons of water

Also want to thank you for responding

The T-H3J takes 160,000 BTU/hr in at the high fire, and ~150K out, which is enough to deliver a decent amount of flow. Even with 35F incoming water you should be able to muster 4 gpm of output at 105F shower temperatures, a bit less at 110F (tub-filling temp) or higher.

I assume you're testing the flow at the tub/shower with a bucket and stopwatch?

The anti-scald valves on shower mixers sometimes interact with gas tankless units due to the hot-side to cold side pressure imbalance created by the impedance of the heat exchanger in the tankless, which may be part of the problem. If you have a hot water tap that isn't on a mixer (say, the connection to a clothes washer), test the hot-only flow.

If it's an anti-scald valve interaction you can sometimes fix it by raising the programmed output temperature of the tankless so that it requires less flow (and less of a pressure drop) through the tankless. The max temp setting on the T-H3J is 140F, but if you set it to higher than 120F it's wise to put a thermostatic mixing valve on the output plumbing to mix it down to < 120F before the hot water distribution plumbing. Sometimes setting the temp too high results in too low of a flow in the tankless for it to fire up reliably with a low-flow showerhead- if that happens there may be some tweaking to do to find the right balance.

If 105-107F is acceptable for your other hot water uses, you could turn the temperature DOWN to that range, and adjust the stops on the shower mixer to let it run full-hot rather than always mixing in some cold. That too will sometimes fix anti-scald valve interactions.

 

[Terry Webb]

My inflow temperature is 73 my water flow is 1.2 I have none of the other issues that you mentioned there are no screens in the hot water side where the tub you turn on the hot water for the tub I finally contacted the manufacture they're going to send me another one. Thank you

 

[Dana]

ted t

My inflow temperature is 73 my water flow is 1.2 I have none of the other issues that you mentioned there are no screens in the hot water side where the tub you turn on the hot water for the tub I finally contacted the manufacture they're going to send me another one. Thank you

How are you measuring the flow?

If it's somehow stuck at the minimum firing rate for mechanical or electronic control reasons, 1.2 gpm is about what you'd get. Assuming it's 1.2 gpm (~600 lbs/ hr) a the shower head with a showering temp of 105F and an incoming temp of 73F water you're looking at about 19,000 BTU/hr out of the tankless, which is the dead rock-bottom of it's firing range.

 

[Tony Welsh]

Many of these units have flow restrictors to establish an ideal recovery rate. If the water pressure is to high - it will run through heat exchanger to fast and enough heat / energy extraction is not achieved. ( unit will not heat the water efficiently)

A consideration also has to be made to establish the flow rate in the actual fixtures. In the United States it is law to regulate water flow in all plumbing fixtures to ensure water conservation. This is done by the factory installation of flow restricting aerators.These are sometimes removed for they plug easily and can result in very low flow rates at the faucets.
If these are left in place the water flow rates in the home are limited enough for most tank less units to keep up with the load. If they are removed or if your trying to source hot water from an older tub spout or shower head, the flow has little to no restriction and you will often lack hot water at those fixtures.
Water temperature is important but anything close to room temperature is great and will not be the issue. If your pulling the water from a well or where a water source will be cool, a storage tank will allow the water to rise to ambient temperatures in a few hours. This can resolve low intake temperature issues, the tank should not be insulated for obvious reasons.

Hope this helps- Too bad your not told all the variables before you purchase the product- right? These units are like many items in the plumbing world, only ideal if the conditions are meet. Very few homeowners fall into that catagory.

 

[Jadnashua]

FWIW, a tub filler is one of the few unrestricted plumbing devices legal in the USA. Vanity and kitchen faucets, along with showerheads, are included in the flow restrictions. Some tankless systems require a higher draw to turn on, but once on, may stay on with a lower flow. It can be frustrating, and requires a learning/accommodation that not everyone is willing to endure. Throw in the higher initial cost and the more frequent maintenance, and they may not be worth it over more conventional options. Your home, your choice. It's a bummer to learn after the fact, though.

 

[Walkintom]

I have a Takagi TH3JDV that has a similar issue.

Mine is about 5 years old and has been increasingly having issues with restricted flow only with circulated water for our hydronic heating. I'm seeing 1gpm or less circulation but if I remove the resistance of circulation by taking the return line loose and sending that down the drain my flow rate shoots right up. Also, if I power down the heater, such as when I am descaling, the circulation is fine. Since I am recirculating the inbound water is typically nice and warm.

I have descaled this thing for hours (2x annually for an hour, but 4 hours this time with no real visible garbage), replaced a circulating pump and checked each and every loop for obstructed flow and I am still not at the end of the search for a cure for this malady. I am beginning to suspect a bad valve or board somewhere in the heater.

Given how much this thing saves me heating in the winter a 5 year life isn't unacceptable but I did think it would perform longer before issues began cropping up.

What am I not looking at that I should be?

 

[Dana]

I have a Takagi TH3JDV that has a similar issue.

Mine is about 5 years old and has been increasingly having issues with restricted flow only with circulated water for our hydronic heating. I'm seeing 1gpm or less circulation but if I remove the resistance of circulation by taking the return line loose and sending that down the drain my flow rate shoots right up. Also, if I power down the heater, such as when I am descaling, the circulation is fine. Since I am recirculating the inbound water is typically nice and warm.

I have descaled this thing for hours (2x annually for an hour, but 4 hours this time with no real visible garbage), replaced a circulating pump and checked each and every loop for obstructed flow and I am still not at the end of the search for a cure for this malady. I am beginning to suspect a bad valve or board somewhere in the heater.

Given how much this thing saves me heating in the winter a 5 year life isn't unacceptable but I did think it would perform longer before issues began cropping up.

What am I not looking at that I should be?

Are you using it strictly as a boiler, with recirculating water that is never replaced, or is it an "open" radiant heating system, or...???

If it's simply recirculating the same water, with the water replaced no more than annually it shouldn't be scaling up at all. I believe that violates the Takagi warranty (but so what? ) With no new water, there are no new mineral deposits.

If it's flow only seems restricted when the flow is through the heating system it's possible/likely that something in the heating system loop is liming up, not the Takagi.

 

[Walkintom]

It's an open radiant system. So there's some exchange of liquid occurring. The well water here was pretty horrid so I had inline filters cleaning up the particulate but nothing in place to deal with dissolved minerals of which I know there is abundant iron.

The heater keeps up fine with flow when there is demand for water for something like a shower, dishwashing, etc. The flow just slows to a crawl when it's only doing recirculation through the heating loops. If I remove any return line and direct the water into a bucket or drain the flow is restored while the line is loose.

We did a major renovation to our home (cabin) this year and added on a basement. The heater, along with everything else here was removed from service for about 5 months while this occurred...

Prior to circulating the loops as I have reconnected and reconfigured as necessary for the new heater location I flushed each one. Got some rusty water but that was all. The inlet screen on the Takagi stays pretty clean.

After the update I'm running 8 loops on 7 thermostat zones. The longest loop is about 340 feet and the rest are all just under 300 apiece. My pump is a Taco 008 and yesterday during my vinegar descaling I entertained myself by adding the spare pump in parallel so I have 2 Taco 008 pumps now. After descale and 2nd pump I'm getting about 2.2 gpm recirculating. But I also got rumbling from the unit several times so...

I'm currently running another descale, this time with Mag-Erad. I have only ever used vinegar prior to now but decided that maybe vinegar wasn't doing the job.

 

[Dana]

If opening up the return path on any branch dramatically increases the flow it means any obstruction/restriction is somewhere between the return manifold and the tankless. It's NOT likely to be scale in the tankless itself- fresh water is flowing freely into the heat exchanger to make up for the water that's gushing out the return line. If the heat exchangers was limed up it wouldn't do that.

A Taco 008 seems like a monster of a pump to be using in this application, but if the impeller is eroded that could be why there's low gpm in circulation mode. For a fresh water system it has to be a bronze, stainless or plastic version of the pump- the standard iron circulator pumps will always fail.

 

[Walkintom]

Thanks for all the replies on this. It really helps.

I am running the stainless pump(s).

Finished the additional descale cycle with only one result - an old leaky fitting that had stopped on it's own started back up and I actually fixed it this time. It's amazing how much more willing I am to deal with things like that when they aren't in a dirt crawl space any more...

I have actually been accepting the Takagi's GPM data and maybe that means I'm an idiot. When doing the descaling I know that pump is moving about 6-7gpm from ejecting the liquid into a bucket and counting off seconds. But the heater is unplugged then so I don't get numbers to compare.

Every piece of plumbing between the manifold and the heater is new. The old piping was all 3/4 copper and I replaced that with 3/4 pex. There is a check valve, and yes, it is oriented correctly. Double and triple checked because at this point I would be happy with that sort of finding. There's very little plumbed on that side right now. Manifold, ball valve with tenp gauge, check valve, 3/4 Tee into cold line, 3/4 elbow x3 and you're back into the cold inlet. Other direction of the 3/4 Tee goes up to a check valve to keep the circulated water from heading to a cold faucet.

If I put a ball valve between the fresh water source and the tankless I could isolate the heating entirely by closing it. If I did that I could run the circulation pumps, open the cleaning ports on the heater and direct the flow in/out of a bucket but that wouldn't really let me measure it, just eyeball the flow.

Right now 2gpm is enough to keep the place warm but that won't be so in a couple of months so I should mebbe figure this out.

 

[Walkintom]

Update: I've played around with a few things and I've improved my understanding of what's going on with my unit a bit.

First, I moved my parallel Taco 008 pumps from being situated by the hot water outlet of the heater to over by the cold water inlet side. I'd seen a couple of comments here and there that 'you just do it that way on Takagis' with no further explanation BUT it got my attention and got me to thinking that the inlet water is going to be denser and easier to pump just because it's cooler and that'll probably also be easier on pump hardware in general, regardless of what the pump is rated to take, heat-wise. In short, I talked myself into making the move. It made no change to my flow rate. At all. But - now I know that made no difference.

So I tried valving off zones down to a single zone (I intended to run each zone exclusively in sequence and record flow rates, etc) and I got into a whole mess where my flow rate fell off and wouldn't come back when the zones were reopened until I power cycled everything. I don't know why and I'm not too anxious to repeat that experiment without understanding exactly what occurred.

Anyway, thinking about the things I know and the things that I know that I don't know got me to thinking about the things that I don't know that I don't know and I got to wondering about the zone length of my underslab loop and whether my contractor reported that to me correctly. I wasn't there to see it installed so all I know without doing some additional investigation is what I was told was the length of the loop. I've been going off the known longest loop being 340 feet since that's the longest I've installed but maybe I *do* have a loop longer than that and I need higher head to get flow rate. (I should have closed just the slab loop first thing and checked my result, apparently)

After I went down this mental rabbit hole a bit I decided to put my two pumps in series instead of parallel. Never done that before but my understanding is that doing so is additive to head pressure. I was even able to bolt flange to flange for a space saving fit. When I fired up everything after I made this change, my flow improved to over 4 gpm.

It's NOT what I think it should be, but it's a move in the right direction. I also finally isolated the head loss information for my unit and at 4 gpm it's about 10 feet. I've attached that. So - if I accept that data as absolute truth, I'm losing about 10 head on the Takagi at 4gpm and the other 18 are needed fro the loop (Assuming 2 taco 008's in series generate 28 head at 4gpm). I suspect I'm losing more than 10 feet of head pressure on the Takagi. As I got the improved flow rate, I also got some rumbling from the unit several times. My understanding is that this indicates some combustion issues. I'm going to pursue that over the next couple of weeks and see if curing my combustion issue (if that's it) also allows me to free up some flow. I suspect that the Takagi may be closing down the flow to improve heating performance but I don't know that's the case.

I'm better off than I was, for sure.

 

[Dana]

In most cases the right thing to do would be to isolate the Takagi's pumping rate from the radiation pumping rate with separate pumps, a "primary-secondary". Done that way you can run the Takagi at a fairly high delta-T with high efficiency without pumping the hell out of it, which takes a toll on flow sensors and even erodes internal plumbing. Even if they're capable of 4-6 gpm in peak water heating mode, it doesn't mean they tolerate that at a high duty cycle. But they CAN tolerate very high delta-Ts of 70-80F at 2 gpm at a high duty cycle.

But the open system heat/hot water system complicates that, since you can't just crank up the temp of the Takagi to get the high delta-T.

Try running some fuel-use math to come up with a reasonable estimate of the design heat load. A 2gpm flow is about 1000lbs/hr, a 4 gpm flow is about 2000lbs/hr. So if the design heat load comes out to say, 40,000 BTU/hr at your 99% outside design temp, at 2000lbs/hr (gpm) that's only a 40K/2K= 20F delta-T needed out of the tankless. At 1000lbs/hr (2 gpm) it would need a 40F delta-T out of the tankless.

If the radiant floor design only need 90F water on design day (pretty common with radiant slabs), with the tankless cranked up to 130F you'd be able to get there with a 2 gpm pump on the primary side (the tankless loop) and you can run whatever monster flow you think you need through the heating system. But that puts you in scald territory on the hot water. At 3gpm (1500lbs/hr) that's a delta-T of 40K/1.5K= 27F, so you can set the tankless to a more reasonable 117F out.

The pumping on the heating system side probably doesn't need to be more than 1gpm, but you can play with it independently.

To make a cheap hydraulic separator between the two sides, take some 1" or 1.25" copper tees and solder them to as short a stub as possible, so that there are fewer than 4 pipe diameters center to center. Use reducers or reducing tees to 3/4" for the primary side, and make sure the secondary side and primary side flows are in the same direction through the fat stub of pipe, not counter-flow. There may be some flow tweaking to do here so a ball valve on the primary loop to be able to raise/lower the flows & delta-Ts isn't a bad idea. But with a hydraulic separator in place the radiation flows can vary by quite a bit without affecting the primary flow through the tankless by very much.

In this diagram they've reversed the primary & secondary sides, which reflects the needs of a high boiler flow requirement and low radiation flow, but with a tankless as the "boiler" the low-flow side is the primary.

Here's a picture of a pre-manufactured version- fatter is generally better for flow separation on the stub: