On demand water heaters are a great way to save money on your domestic water heating costs......most of them are quite reliable.........cant say i would choose one over another having installed a few different ones......the biggest things you have to look out for is your gallon per minute usage and your temperature differential..that is your incoming water temp. verses your desired hot water temp. ............do some research first to find the one that is right for your part of the country ( colder inlet temp. in the winter )and for your expected water usage.

I've considered one of these too for a future wing/bathroom but mine would have to be electric. Our well water is frigid in the winter and down right cold in the summer so not sure if it would work good enough here.

I have a Rheem/Rudd/Paloma unit and I am very happy with it. Payback will be about 4.5 years.

Typically, I would not advise using an electric tankless. And as poorplmber noted above, be aware of your inlet temp and max flow rates. Use a thermometer and don't just guess. Make sure the unit that you are looking at qualifies for the tax credit. Also be aware of the cost of the vent. You must use special stainless steel vent and it is expensive. A long vent run can make an installation prohibitive.

Also, I don't know about all brands, but I would advise against Bosch units. They seem to have a history of trouble due to their design.

Just don't expect to save much money with a on demand heater.
How much do you think it will save and why?
Keeping water warm in a well insulated tank actually uses very little energy.
My total gas usage on a 40 gallon tank is about 7 therms average. That figure includes a shower per day, as well as about 2 dishwasher uses weekly. So figure half the therms are for actual usage while perhaps 3 to 4 therms go for keeping it warm. Your cost of gas may vary but here, that only costs about $3 to $5 per month to keep it warm 24/7
Whats the payback on a $1,000 heater if you are only spending $5 a month to keep water warm in the tank.
That advantage over a traditional tank heater is about the same as the interest you could earn.
OK.....say you save $5 per month and that the cost for the on demand unit is $500 more than a tank unit.
So after covering for lost interest on the $500 then you have about 35 you can put towards breaking even on the tankless unit. Instead of 4 to 5 years payback, you are talking about 15+ years payback. And by that time, you probably need to buy a new tankless.
So the savings may be very little if any.

PuzzleFan,

You should expect to save between 20-25% on your hot water costs. Some manufacturers are saying "as much as 50%" -- that is nonsense.

These heaters save money in two ways, they eliminate standby heat loss and the gas heaters are more efficient at converting the heat input to hot water than a conventional heater.

In my case, I was comparing my tankless unit to a powervent tank unit. I didn't want to vent my water heater through the brick chimney again.(Common, but not Kosher without a flue pipe) The purchase cost of the tankless was similar to the powervent.

Again, for me the payback vs a traditional unit was about 4.5 years including the tax break. (but I would not have purchased a traditional unit, it would have been a power vent that cost as much as the tankless unit)

Although I think he is trying too hard, Baxxter does have a point in that the savings aren't huge. If you have other places to put your money (such as insulation, new furnace, etc.) the payback may be higher.

I will also say that similar arguments were made against condensing furnaces. Natural gas prices have gone up considerably since then and I have never been sorry that I bought the hi-efficiency furnace.

The point is that you replace your major appliances once every 15 years. If you go with the appliances that are just borderline reasonable cost vs. energy savings, energy prices usually increase over time to make them a very good investment.

I agree with jake. For us, it is the difference of a tankless unit or a powervent. The tankless will be mounted outside saving money on venting. The price is also just about equal, due to the federal rebate.

The real cost savings come with the fact that a tankless is much more efficient at heating water. It is over 30% more efficient. On top of that I do not have to store hot water and keep it hot. We do have cold water in the winter, so we will be limited to about 5 gal/minute during that time. That should be more than plenty for us.

I also get the space in my utility/laundry room that the tanked unit would consume.

Another reason to consider a tankless is the length of pipe run from the HWH to the fixture. In fact, the only reason I have installed tankless heaters in the past is because the client was tired of waiting forever for the hot water to get hot.

In addition to energy getting more expensive, so is water. You can figure out a water payback as well, since you wouldn't be dumping X amount of water down the drain while you wait. When combining the energy and water savings you often get a considerably more favorable payback period. But there is also the value of convenience.

I agree that you may get more bang for the buck by spending your money elsewhere, including installing pipe insulation on your hot water piping. If that is possible, I'd try that first before going with a tankless, and see if that is satisfactory for you.

I love my tankless water heater, but I would caution anyone in a remodel situation to really examine the costs associated with retrofitting one into their home. In new construction, where you can put whatever you want, it makes the most sense and really isn't that cost prohibitive when planned for from day one. When you have to upgrade your gas line, vent size and material, and maybe factor in relocating the heater itself, then payback of intial startup costs is prohibitive. For electric, it involves upgrading the whole house electrical service, possibly from the pole to your home, and certainly an upgrade to your breaker box. The smallest units use approximately 80 amps on a dedicated service. For older homes with only a 100 amp service, you're talking 80% of the entire electrical service to the home. So you can see why even folks with 200 amp panels may need to upgrade as well.

There are definate plusses to a tankless water heater, such as endless hot water, but "faster" hot water isn't one of them. The length of time that it takes for the hot water to travel to the fixtures will remain the same, and all fixtures will have to purge the cold water in the lines before obtaining the hot water. Only relocating the heater closer to the end fixture will reduce the travel time of the hot water. Or you can install a recirculating pump that will defeat any energy savings you might have realized on switching to a tankless, and may even cost you more $$ to boot.

Tankless heaters are a great option. They're just not for everyone.

Most houses in Japan have tankless/ or on demand water heaters. Since water pipes run up outside the house and are not insulated, the water is the same temperature as the outside temperature (39F this morning). I am confused why people think cold climates can not use tankless water heaters. Also the water flows fast enough to fill my Japanese soaking tub quite quickly.

Cindyasahi - it's not that people here think those in cold climates cannot use tankless. It's that tankless has been promoted as a money-saver, which where I live at least, it is not. So the recommendations you are seeing on this Forum are more along the lines of don't get a tankless if you're doing it to save money but you live in a cold climate. I would like it very much to have your temps of 39 degrees in the morning - we've had many days so far this winter where morning temps have been below 0, and so incoming water temps are around 40. There needs to be a larger-sized gas line to be able to heat that water in a tankless. The installation of the gas line adds expense, as does additional venting since it is unlikely that the tankless would be installed outside as yours is. Furthermore, in Japan they have been using tankless for years, and you have many installers who have the expertise to install them. Here, the number of knowledgeable installers is fewer, and the risk of a poor installation is greater, made more so by the demands in a colder climate. So I hope you now understand why some do not recommend tankless in colder climates here - it's not that it won't work, it's that it is a quite expensive upfront and installation cost (compared to tank heaters), and it needs to be installed by a knowledgable plumber.

Interesting and thanks for your reply. We will be installing a water furnace this June in a rental unit in Minnesota and I am trying to convince the geo thermal installer (very experienced and very highly recommended) to do some checking on installing tankless water heater to work with the geo thermal. Thus when no one is renting (it will be a B and B but in a separate building) we do not need to keep the water heater on. Plus we are installing a Japanese soaking tub which will take ?? many gallons of water and the guests do not want to run out of hot water.
So any thoughts about the tankless working with geo thermal

How exactly is an on demand 30% more efficient ? All the gas models I have seen are around 85% efficient, the same as storage type heaters, to get into 90% it would need to be a condensing type with a drain and plastic flue and I haven't seen any of those.
If you replace a storage heater with a tankless expect the same exact delay in getting hot water along with the bonus of a cold slug of water when the unheated water that was sitting in the heater ( not the insulated piping ) finally reaches you.
You cannot use a recirculating system with a tankless without adding a small tank to the system which defeats the whole purpose.
Many people who go tankless end up using even more energy as family members end up taking longer showers, showers that ended previously when the tank ran out of hot water.
The only savings are from standby loss and that is very small and can be made smaller by insulating tanks and pipes.
Where tankless heaters are used in other countries they usually have a seperate T/L for the kitchen and for the bathroom, short runs, like a point of use heater. One can understand why they would be popular in Japan where space is a premium and a storage tank would take up too much room.
TBS, IMHO there are some legitimate uses for T/L such as a seperate H/W supply for a large tub.

HoundHandler,

It's not at all the way that you describe it. To begin with, look at the energy factor, not "efficiency". The term efficiency gets tossed around without a good definition. The Energy Factor takes into account actual usage. So, in the case of a water heater, it takes into account the standing heat loss-- or not.

A standard gas tank water heater has an energy factor of .58 -.62 The .62 is a "high efficiency" model. That's right off the rheem website if you want to look it up.

Most tankless have an energy factor of .82-.85 there is one that I know of that gets into the low .90s, but it's unusual. There's also one tankless that I know of, a Bosch unit, that's about .65. The point is that most are in the low to mid 80's but you need to check.

To go from an energy factor of .60 to .83 is a 38.3% gain... That's significant!

I don't know if you own a tankless heater, but I do. I have never experienced the "cold slug". In addition, I think the "taking longer showers" argument is bogus as well.

My tankless cost me right about $1000 with the vent kit - self installed. I didn't need to increase the gas line and I had a short, through the wall vent installation. The payback is about 4.5 years at current gas rates compared to a standard gas tank water heater. The thing is that I would have either needed a powervent heater ($1200) or line the brick flue ($??) if I had gone with a standard tank heater. The tankless was less than a powervent.

As I have stated in other threads, I would never advise someone to install an electric tankless except for point of use. There is no operational savings with electric tankless.

As to the cold weather issue, it takes a lot of BTU's for a 70 degree temp rise. (35 degrees to 105 degrees) A 200k BTU unit will heat about 5 gpm at that temp rise. 200-230k BTU is about the top end of the residential before you get into multiple units. So, in cold climates, 2 showers or a two showers and a bath sink (230k BTU unit) is about the best that you can do... and that's usually enough.

The gas line may or may not need to be increased. In many of the older homes, they were built gas supplies to deal with no insulation and inefficient furnaces. I have a 90% furnace and a 150k BTU water heater and the gas supply is fine.

That said, newer homes may not have the extra capacity. In any case, you really need to size the unit properly, determine if the gas line is sufficient and determine if the flue can be short (cheap). Don't go into it blindly... but don't listen to the FUD either. Just do the math.

cindyasahi,

I can't imagine how geothermal and tankless would work together. To heat 5 gallons per minute 70 degrees takes a huge amount of heat at one time. A geothermal unit wont do that. A geothermal unit provides less heat over a longer period of time.

Geothermal can be an appropriate solution, but a tankless seems appropriate for the intermittent use that you describe.

jakethewonderdog,
This topic has been covered extensively before, just do a search.
Heat loss during standby is greatly exagerated, and if you throw some extra insulation around the tank it's even less.
And if the tank is located within the house envelope, any small amount of heat radiating off the tank contributes to home heating, at least during the winter.
The real question is how efficiently the water is heated, and there is no difference.
In fact there are 90+ storage water heaters available.
If you really want to save money heating water, install solar.
Storage type water heaters are one of the best devices invented, no moving parts, extremely reliable, requires no electricity to run.
T/L W/H's have P/C boards to burn out, impellers to lime up and require electricity to run so if the power is out, no H/W, they are prone to freezing and bursting in cold weather, one breakdown and the high cost to fix it negates any possible savings over it's lifetime.
Because it is not possible to use a recirc system you will have long waits for H/W as costly cool water is wasted down the drain.
In short, T/L heaters are an expensive solution to a non-existant problem.

We built our custom home in 2006, and I spent many hours researching all manor of mechanical items, as I hope to never move again. I wanted the best mechanical/structural construction which are the things no one readily changes later. We built to a Canadian super energy efficient standard called R2000, plus geothermal heat.

I was enamored with tankless water heaters, having suffered with poor performance in our previous house, and wanted electric tankless. I decided I would pay the extra for 400 amp service, which I did, as here electricity is only a little more than natural gas. Then as the design fleshed out, I'd need two tankless heaters, at 100-120 amps each for our cold well water. Then the deciding factor was to minimize water usage since we're on a septic system, and new water protection bills were passed in our province, increasing homeowner's liability for failed systems. Once minimizing water consumption became the starting point, part 2 was a recirc system, so part 3 became a storage type heater.

Our water is very hard, 36 grains, so we needed a 70,000 grain softener. Soft water eats steel tanks, so my final decision was made when I discovered Marathon water heaters. They are polybutylene/fibreglass tanks, EF of .92, with only 5 degree heat loss in 24 hours. The plastic can't take the heat of a burning flame, so electric only.

Other steel water heaters specify standby losses of 60-100 watts heat loss. So 100 watt heat loss x24h x 30d per month x 6.5 cents/kwh here=$4.68 per month cost in heat loss. Plus taxes.

So, my initial enthusiasm was tempered with facts. If you have 6 teenagers showering twice a day, a large tankless gas heater is perfect. If you have a tiny house where each sq ft is precious, any tankless is perfect to save space. Otherwise, if you have a mechanical room, need/want a recirc system, want to save money on capital costs but accept $5 a month in extra operating costs (even $10 a month for a standard tank at higher US electricity prices), I think storage type water heating is still an excellent and cost effective idea.

houndhandler,

Jake here. I know that there is research posted in other threads, because I posted it.

The energy savings comes from three places in a tankless, listed in order of savings:

1. Lower flue temps- Tankless get the flue temps to near condensing, that's why they must have stainless vents. Tankless can do this because of the power venting, SS vent and modulating gas valve. The one tankless that had an EF in the .90's did go all the way to condensing... but it cost a lot more to accomplish that.

Standard tanks must keep the flue temps high enough to vent reliably, all the way to the roof, via barometric pressure. That's a lot of wasted heat.

2. Tankless have no standby losses through the flue. Standard heaters have a flue pipe that goes through the center of the heater that acts as a chimney, even when the heater is off.

3. No standby heat loss through the jacket.

4. No standing pilot.

Now, this isn't my opinion. They are facts. You may check my facts if you wish, but I haven't seen a standard gas tank water heater with an energy factor above the low .60's -- no mater how much insulation. If you know of one, please point it out to me.

While manufacturer's claims are exaggerated (on tank and tankless) The energy factor is the no BS way of getting to the real bottom line on energy use. BTW: There are no standard tank water heaters out there with a EF in the .90's If you think there is, please back up your claims with a link. There are some condensing tank units, but they cost 2x the cost of a tankless.

As to the heat loss and the envelope, since much of the heat loss is actually going up the flue, that argument doesn't hold water either.

So please, before you tell me I don't have my facts straight you should look the information up yourself. I have done it in several of the post already -- Show me a standard tank heater with an EF significantly higher than .62 Also, if I'm wrong about the EF of tankless being around .83-.85 say so.

The rest of your post about reliability is FUD, really. There was only one tankless unit that had significant reliability problems... and that was a Bosch. It used the water flow to generate the ignition spark (very lame) it also had an EF of ~.65 (also very lame). In very hard water situations, I would do some research on the impact on tankless. Other than that, your FUD is baseless.

As to the post above about the electric tankless, I completely agree... the EF difference between an electric tank and electric tankless is only about 3% if I remember from the research I did. That's not enough to justify the increases in electrical service except for special point of use applications.

Here is a question - why are you comparing against a "standard" water heater. Why don't you compare against another condensing water heater as a basis for comparison ? It's a total cost scenario and not whatever you want to compare against to make one technology look better over another.

Granted a condensing water heater costs more than even a tankless, but you also have to factor whether its a new build or existing build, how much water is being used and efficiencies involved. A condensing water heater's ef (or thermal efficiency - standby losses) make tankless water heaters look like inefficient pigs.

While I'd agree that a tankless has its places and is more efficient in certain scenario's, I disagree with a lot of their overall comparative statistics. It seems every tankless manufacturer uses a tanked water heater with an eF of 0.50 as a basis of comparison to announce their 33 to 50% gains in cost reduction. .

Overall, depending upon usage - it's really a wash. Do you want to spend more up front and try and recover that over 8 to 12 years to break even or spend less up front to pay more over time ? Then the next big kicker - if you use enough to break even sooner, the condensing unit's start to factor in much more heavily as they have a big jump in efficiencies as well.

This is all assuming that one doesn't have a boiler/hydronic heat as piggy backing that would make a lot more sense for HW generation. Solar is cheapest on a mtm basis, but the startup costs there force a 6-10 year payoff as well (depending upon usage).

Basically it isn't a foregone conclusion as either side makes it out to be. It's 6 or 1/2 dozen (maybe skewed a little one way or the other but not heavily).

Here is an interesting chart, did not cut & paste right but her it is.
Water Heater Type Efficiency (EF) Installed Cost 1 Yearly Energy Cost 2 Life (years) Total Cost (Over 13 Years) 3
Conventional gas storage
0.60
$850
$350
13
$5,394
High-efficiency gas storage
0.65
$1,025
$323
13
$5,220
Condensing gas storage
0.86
$2,000
$244
13
$5,170
Conventional oil-fired storage
0.55
$1,400
$654
8
$11,299
Minimum Efficiency electric storage
0.90
$750
$463
13
$6,769
High-eff. electric storage
0.95
$820
$439
13
$6,528
Demand gas (no pilot) 4
0.80
$1,600
$262
20
$5,008
Electric heat pump water heater
2.20
$1,660
$190
13
$4,125
Solar with electric back-up
1.20
$4,800
$175
20
$7,072

Over 13 years you will save $306.00 with a tankless gas over a tanked gas but this does not take into account higher water bills and sewer bills which are often tied to the water bill.
And since you are paying extra upfront for T/L and hope to recoup savings over time don't forget that you cannot hang onto your money and make some small interest on it.
And if you are borrowing the money to buy the more expensive T/L you will have to pay interest on that borrowed money, perhaps over a 30 year mortgage.
Also how long do you expect to stay in the house ? most people will move before 13 years, before any savings can be recouped.
Jake you say that a T/L is " more efficient at converting the heat input into hot water than conventional heaters " but this just isn't true, the efficiency % is the same, standby loss is what is saved.
The energy factor (EF) indicates a water heater's overall energy efficiency based on the amount of hot water produced per unit of fuel consumed over a typical day. This includes the following:

* Recovery efficiency how efficiently the heat from the energy source is transferred to the water
* Standby losses  the percentage of heat loss per hour from the stored water compared to the heat content of the water (water heaters with storage tanks)
* Cycling losses  the loss of heat as the water circulates through a water heater tank, and/or inlet and outlet pipes.
My point is that T/L may be slightly more efficient on paper but in the real world longer showers may use more energy and you cannot use a recirc and being that a T/L is like a bottleneck in the system flow rate is limited.

Look, houndhandler, I have explained exactly how the gas tankless achieves the savings. I spelled it out in the four points above.

The largest savings doesn't come from standby losses through the jacket, it comes from lower flue temps and not having the stanby losses through the center of the tank. As the electric tankless show, the standby loss just through the jacket is only about 3%.

I showed you energy factors to support my case. I have told you where you can find the energy factors for the various equipment. The energy factors provide measurable unbiased information, unlike marketing hype.

I have shown that an increase in the energy factor provides a measurable ~35% overall gain without cherry picking my numbers. The longer shower argument is bogus, but If they take longer showers then they are getting 35% more utility for the same money.

Now, I have consistently given people real information that they can then use to assess their own situation. I also have the benefit of having actually installed and used one in my own home. You, on the other hand, want to spread inaccurate information and FUD.

Gas tankless isn't right for every situation. It is particularly not good if you have long flue runs or have to upgrade the gas supply. It can be very good if the alternative is a powervent, direct vent, or if the tankless allows you to avoid some other cost-- such as lining a flue.

In my situation, the payback over a standard water heater, without any other considerations (such as the fact that I couldn't use a standard water heater without lining the flue) the payback was just under five years. The truth is that if I didn't go tankless, I would have gone with a powervent unit to avoid lining the masonry chimney. A powevent unit was more than the tankless.

My payback period did include a tax credit, but no utility credits. The payback will be shorter if gas prices increase in five years.

And why would you say that you can't use a recirculating pump on a tankless? Of course you can. I think it's a better idea to design the plumbing so that you don't have the water heater at one end of the house and the master bath at the other... but that's beside the point. If water is expensive or you place a high value on "instant" hot water... use a pump. I also recommend insulated pipes and a dedicated return.

I've suggested in other posts to combine a motion sensor in the room and a thermostat on the pipe. Inconspicuous 120 volt motion sensors that fit in a 2x3 box are cheap. Hey, you walk into the bathroom and the pump kicks on until the pipe is hot and then it goes off. It doesn't cost much, gives you the hot water you want, and very little in the way of standby losses. It might kick on some times when you don't need it.. . but no big deal. Other people just use a manual pushbutton switch that runs the pump for five minutes or so just when they want it.

To answer Anthem, the reason I don't compare tankless to condensing tank heaters is because the condensing tank heaters I've seen cost 2x or more than a tankless for an energy factor of .89. (or an 7.2% gain over a tankless)I can't imagine a scenario where that would ever make sense, Can you? Note that I wasn't cherry picking to make my point... most of the standard gas water heaters have an EF of 58-62 with .62 being the "high efficiency". I used the Polaris 50 gal condensing heater for comparison, If you know of condensing tank units less than $3000 give me a link to it. If you can think of a design situation where it makes sense to spend $3000 on a water heater, please tell me what it is.

Jakethewonderdog,
I am a plumbing/hvac contractor and have installed T/L heaters for customers and in fact have one in my own house.
T/L heaters convert gas to H/W at around 82-84 % same as a tank heater.
You are spouting EF or energy factor which includes standby loss and cycling loss.
Since you have never tried to hook up a recirc system to a T/L heater and I have let me explain to you why it won't work, been there, tried that, talked to tech support ad nauseum.
You assume that because the T/L heater instructions say you need for instance 1/2 GPM flow rate to activate the burner simply using a small recirc pump rated @ 1/2 GPM will activate the burner, wrong.
That's 1/2 GPM on an open system like when you open a tap.
A recirc is a closed system and will require signifigantly more GPM to activate that burner.
My system ended up using the T/L indirectly to heat a 50 gal. tank and the pump is pretty big, I could go under the house and check it out but it was big and costly.
I had to use the tank because I found out you cannot just recirc the water back thru the T/L heater, it just doesn't work, water gets too hot or doesn't come on at all. A buffer is needed.
Now you don't need a 50 gal heater, you can use a 10 gal but at the time I had just built a crazy shower with 5 showerheads and bodysprays and felt I needed the 50 gal head start to get a prolonged shower with all the sprays going. Turns out I'm a cheapskate and just use one showerhead !
For a real good detailed explanation look in the "Finehomebuilding" magazine Dec 07/Jan 08 issue, there is a pretty good article about just this, however the author of the article makes a few mistakes, for instance he tries to show a tempering valve hooked into the H/W recirc loop that just will not work.
Using a tank with a T/L heater with pumps and sensors amounts to a Rube Goldberg setup, I should know I have one.

Houndhandler is correct on the tankless and recirc setup. While certainly possible - its pretty difficult. He esssentially tackled it by using a holding pump.

"If you can think of a design situation where it makes sense to spend $3000 on a water heater, please tell me what it is."

Easy. Actually quite a few that I've been involved in the last couple months. All involve high volume

-Homeowner frequently has guests over and doesn't want hot water to run out when running 3 to 4 showers. Either very large storage, ganged/parallel tankless, ultra efficiency condensing gas. Went with ultra efficient gas(condensing) as they essentially got the best of both worlds - efficiency and volume and recirc to boot.

-homeowner has shower that can output over 12+gpm (whether you or I like them isnt relevant). Using tanked - can go thru a 100gal tank in less than 8min. Again stacked tankless or ultra-efficient gas tanked. Gas tanked wins as it was higher efficiency and could handle recirc.

Now this may not always be norm as these were both high-end installs, but anything requiring volume will tax traditional and tankless systems. Both of those were well over 3k (about 5)

Like I said before though - in regular use the differences are very close. Maybe a few hundred dollars either way over the lifespan depending upon how you count.

I would agree, very high demand situations (any time you are exceeding the capacity of the largest tankless unit and looking at multiple units) could make sense for a large condensing tank unit. Particularly in high-end installs.

On the other hand, you asked why tankless was being compared to a standard tank and not to a tank condensing unit. The reason is that tankless are on the edge of making sense financially in an average residential install. More so in some situations-- less so in others. Condensing tank units are appropriate for high-end, large capacity installations where performance, not payback, is important. Your average family of four probably isn't going to consider a Polaris that costs 2-3 times the cost of a tankless unit.

"You assume that because the T/L heater instructions say you need for instance 1/2 GPM flow rate to activate the burner simply using a small recirc pump rated @ 1/2 GPM will activate the burner, wrong.
That's 1/2 GPM on an open system like when you open a tap.
A recirc is a closed system and will require signifigantly more GPM to activate that burner."

Its just the opposite; a closed system requires less energy, thus less pump/circulator to activate the burner.

Take a look at circulator performance charts-Open system versus closed, you will see closed performance with same circ is better.

With regards to hooking up circ to on-demand heater, my Takagi TK-3 has recirc pump leads to connect to built right in and it works sweet!

The Takagi TK-3 recommends a pump that can deliver at least 2gpm at 15' head plus piping head losses. It also wants the pump to deliver a flow rate of less than 4 ft/sec or 5 gpm with 3/4 pipe. That's good information to know. I appreciate companies that take the guesswork out of an installation.

It looks like the pump connections on the TK-3 provide the temperature sensing for the water temp on the recirculating loop. That's pretty cool.

The TK-3 is a 200k BTU unit that sells for about $830 making it very comparable in price to the Rheem 74 (I own a Rheem 53). It has a 10 year warranty on the heat exchanger and 5 years on everything else. The Rheem has a 10 year on the exchanger and 3 years on everything else.

So, particularly if you are considering using a pump, look at the Takagi TK-3.

zl700,
Perhaps I do not have all my terminology correct, I'm just a plumber not a fluid engineer, let me try to explain it this way ;
When the T/L heater needs 1/2 GPM flow rate to activate the burner this is easily achieved by opening a faucet, there is no resistance, the water is trying to get out.
When you are trying to achieve 1/2 GPM flow rate by pump in a closed loop without any taps open the pump must work against the resistance of all the pipe friction and fittings, the T/L itself with all it's narrow passages, the water itself as it's pushed around in a circle and head if any.
This is the difference I was trying to show, an open pipe vs. a pump.
Now it seems that perhaps at least one T/L company has figure out a way to make a recirc work with T/L, perhaps a lot has changed in 10 or 15 years. I notice that all the T/L are now electrically operated with P/C boards that must have computer algorythms to solve some of these issues.
I found this on the web ;

The traditional way is to install a recirculation system is to install pump and a return loop back to the inlet side of the heater with a couple of flow check valves. This will work of course but you have to install pump with enough "head" capacity to get through the high head loss of the tankless. Higher head pumps also use more electricity to operate than their lower head counterparts. How big a deal is that? Well if a higher head pump uses $.02 more to operate per kWh then the other pump, over a 20-year operating period that calculates into the thousands of dollars over the life of the pump!

Another consideration is most tankless water heaters get their energy efficiency and their longevity by being "always off". Installing a recirculating system this way puts a lot more wear on the unit and may cut its warranty in many cases. Add the cold water sandwich to the equation and there is a better way.

One of the first things that installers dealt with was the cold water sandwich. This was managed by installing a very small storage tank (About 5 gallons) right after the tankless heater. This way, if a small slug of cold water got out of the tankless it would mix with the 5 gallons of hot water in the tank and be diluted to a point where no temperature change would be noticed. That fixes the cold water sandwich but still leaves the high head pump and the warranty issue to deal with.

At some point someone came up with the idea of turning that small storage tank into a 5 or 6 gallon 110V tank water heater. They are very small, insulated, and can plug into an outlet. Then they hooked up the recirc loop between the tankless heater and the small tank heater effectively by-passing the tankless, saving the warranty and keeping the energy efficiency of the unit. The small tank heater does not have to heat the water very much. The tank and recirculating system only have to use enough energy to maintain the loop temperature. This is usually only about a 5° rise. With insulation on the hot water piping, it is a very efficient solution! Instant hot water, as efficient as you can be with very low energy demands, and all the hot water you will ever need.

Yes hound, no doubt when a faucet is opened the rush starts and water take the path of least resistance. However on a closed system as the pressure in the system rises, circulator pump efficiencies increase. Notice I dont say pump because a pump moves the water from one place to another, a circulator circulates and wants the medium it discharges to return to its suction.

One of the best ways to show the difference is take a common performance spec from Grundfos, comparing the UPS15-58 circulator. This is a common circ pump on most residential systems. Lets compare closed loop, Hydronic versus open loop, solar, open storages, etc.

Closed system:
5 GPM @ 15 Head

Open system:
5 GPM @ 8.5 Head

You see, same pumps and the closed system application performs better. This is common knowledge in Hydronic systems and pressure plays an important part but even more important is air, or lack of it. In a closed system removing more air means easier, more efficient circulation.

Many years ago before I was afraid of legionnaires, I had a Bosch instantaneous heater for domestic hot water. On this heater I also had a bronze circ that was controlled and was used to heat some baseboard in my basement rec room. I learned the principals of pumps/circ very early when I discovered that I lost circulation to my heating baseboard every time someone opened more than a hot water faucet in the house. Seems I ended up with priority DHW without even trying, water takes the path of least resistance.

Yes, I was trying to compare using a pump to activate the T/L (CLOSED system) vs. using it the way it is intended, OPENing a tap.
When I hooked up my T/L the specs said it needed 1/2 GPM to activate the burner, so I figured I'll just take a little recirc pump rated @ 1/2 gpm, should do the trick, wrongo.

Understood.

Like a grundfos comfort pump, watts, or lang DHW pump, just not enough.