This is new construction, and I have been thinking I would go tankless for my domestic hot water. There is no gas - house will be all electric.

However I've been reading conflicting information about electric tankless, its efficiency and return of investment. I understand they draw a lot of amps, so accommodation has to be made in the circuit panel for that.

House will have 3 full baths, and the usual appliances (dishwasher, washing machine), but we're just a family of three.

Winters are very mild, but another consideration is the incoming water temp in Louisiana in the summer. The mains are so close to the surface of the ground that the water comes in hot already. How hot, I cannot accurately say, but hot/warm enough that you can wash your hands using the cold faucet and you wouldn't comfortably drink the water, but not hot enough to shower with. Seems like tankless would allow me to take advantage of warm water coming from the mains. But I also want to be sure that I can set the thing so that you can take a hot shower even in the summer, without the unit kicking off because it thinks the incoming water is hot enough already.

I'll need more than one tankless unit as well as increased amperage, but it's new construction. Would the savings in energy be enough to warrant electric tankless, or would I be better off to go with a conventional electric storage unit? Any opinions, or experience?

What about propane?

I edited an article in Fine Homebuilding by a master plumber/builder, Michael Chandler from Chapel Hill, NC on getting the most of a tankless water heater. He is against electric tankless units but all for gas-fired units. Propane is what he installs on most houses he builds. Frequently he couples a tankless to a small tank and a solar water system to provide heat for water ans space (radiant floors).

Also, a heat pump water heater might be a good choice for you as they cool the air from the room they're in.
Read a lot about heating water at www.greenbuildingadvisor.com/green-basics/water-heating

Hope that helps,
Dan

Ancillary reason to go for an electric hot water heater (heat pump or conventional) is that it can be renewably powered at any point in the future. Gas or propane units will always run off fossil fuels.

Burning gas or propane can often produce less CO2 than electric power because of no transmission losses, no coal powered electricity at source. But if PV magically gets cheaper soon (been hearing that for a long time now...) you then have the option of converting over.

Can be low first cost, also.

It really depends on your location. My town is close to several Nuclear plants as well as several hydro stations, so electricity is cheap and clean. Also, in my location, propane is more expensive per BTU than electricity, but natural gas(if you have it) is even less per BTU. One advantage of a tank heater is you could use off peak electricity to heat the water at night and just store the water in the tank until you are ready to use it during the day.

On a similar note, what is the best R value tank heater avalible? I think it is silly that most of the water heaters I can find have less R value than the walls of my house.

I think Marathon makes the best insulated tanks, although if it’s an electric tank you can add insulation on the outside without any worry.

I also wanted to point out there are losses in pumping or moving NG, less than trucking propane and the price reflects that, but there are losses involved. Not sure energy wise how it compares to moving electricity.

If you're thinking that you're somehow "greener" 'cuz your local generating capacity is one energy source vs. another, think again:

The N.American grid is ONE MACHINE, all generators feed it, all loads retrieve from it, without distinguishing where that power came from.  No load is greener than any other based on location from a carbon point of view- a grid load is a grid load, period.

Check that- in fact, loads placed closer to COAL PLANTS are slightly greener, 'cuz it lowers the distribution losses from a heavy carbon-emitting generator.  Hydro or nuke power transmission losses are already carbon-free and lossy exports along 1000 miles and 25 transformers can be incurred without additional carbon burden, even if very lossy along the way.  (It's a subtle distinction in the aggregate, but TRUE!)  Wanna be greener- place your load within one transformer of the coal-generator.  Unless you're generating & using your own carbon-free power and exporting any excess to the grid, your load is just like any other load, sipping off the aggregated grid.  Plopping the load next to a grid-hydro plant doesn't spare the world any carbon (but it probably saves you money! :-) )

The best R-value in electric tanks are solar-storage tanks designed for higher temperature storage.  Cheapo electric & gas tank heaters are about R4-R6, the better ones about R8-R12, but there are a few electric tanks out there with R20 (eg. BRADFORD WHITE M250T6DS). 

OTOH, most existing houses in the US (2x4 studs 16" o.c., with R-11 batting, typical installation) run about R7-R8.  Even typical 1980s type 2x6 construction with R19 batting only have clear-wall values around R15-R16.  The reason it's generally not cost-effective to go higher than R12 (or R20) on tank heaters is that the delta-Ts are relatively well defined: Ambient air of 65-70F, water temp of 120-130F for a fixed delta-T in the 50-60F range rarely more.  Losses from the tanks are also generally into conditioned or semi-conditioned space, which is (kinda) OK for 3/4 of the year in much of the US, since it's lowers the space heating load.  The surface area of a tank (relative to the exterior surface area of your house) is also very small.  If you add width to the tank for more insulation it fits in fewer places, making it an even harder sell for very marginal decreases in operating cost.   Solar storage tanks can handle storage temps in the 180F range, and are insulated to be able to do so. Most solar storage tanks come rigged with (or are designed to be retrofitted with) electric heating elements as backup heat.  But they're also designed for high stratification- you need about 1.5-2x the storage capacity to get the same useable volume out as an electric or NG fired tank.)  Unless you're in a superinsulated house (R25 SIPS or better) your whole-wall & clear wall R-values are probably lower than that Bradford-White electric tank.

On gas fired tanks there are necessary gaps in the insulation (under the burner, f'rinstance) and have high standby loss from flue convection.  You could triple the insulation value, but the standby losses won't go down very much, since even at R4 insulation levels the majority of the standby loss is flue convection, not conduction through the tank wall into the adjascent space.  Electric tanks have FAR far better standby characteristics at almost any insulation level.  (It's the gross source-fuel inefficiency that makes electric water heating less desirable.)

Going with off-peak electricity doesn't make it significantly greener from a carbon-footprint point of view either, even though it makes it cheaper.  The base-load generators are the same coal/gas/hydro/nukes, peak generators are often gas or oil-fired, but the net effect of the grid-source mix isn't much affected. (Remember, the base-load generators are already WORSE than natural gas-fired generation on average.)

Have you thought about checking with the local utility company? Many offer really great deals on high efficiency tank type water heaters. A local utility in our area gives them away. I talked to my HVAC guy and he does not recommend any type of tankless. He says that most people that install them end up installing some sort of buffer tank with them. There are many inherent problems with tankless water heaters that you just don't hear about. Another thing with your utility is that they may offer you a much better rate if you install a tank type water heater if you allow them to put a timer on it so that it only heats during off-peak times. You can also get a very efficient Marathon water heater that has virtually no heat loss. I have read that many areas of Europe are moving away from tankless water heaters because of the demand they put on the local utility at certain times of the day, especially when everybody is taking a shower in the morning and evening.

The peak loads to the very local grid represented by electric on-demands is HUGE, but not dramatically different than a fleet of tank-electrics on the only slightly wider (say city-wide instead of neighborhood or city-block) grid. 

The only way tanks can reduce peak loading is if the tanks get locked-out of drawing from the grid during peak-load conditions.  Smart-grids (and smart high-load appliances like water heaters, heat pumps, air conditioners, and clothes dryers) can smooth out peak load management considerably.

(Coming to a grid near you... sooner than later, hopefully! :-) )

Much of Europe uses NG-fired (not electric) point-of-use tankless water heaters with reasonable efficiency and utility.  Point of use heaters tend to have fewer "issues" than the oversized "whole house" forced draft gas-fired tankless systems more familiar to US plumbers.  These aren't unmanageable beasts though, and DO provide a significant efficiency boost over NG tank heaters in many/most applications (I'm undeterred by the critics.) 

I'm not a big fan of any electric resistive-element heating (for water, space heating, etc), but as modulated finish-heat to low-temperature solar or desuperheater preheating, electric on-demands definitely have a place in greening-up DHW.  A high efficiency electric tank heater (or on-demand) with a 0.98EF is still only 27% source-fuel efficient, which is abyssmal compared to any legal gas-burner.   But as the final 10-20% finish heat bringing low-temp solar up to temp it starts to look pretty good! (Especially since by running low temps the solar runs far more efficiently, so it can be sized somewhat smaller, etc.)

I think you have a misconception, the N-Americian grid isn't just one machine, it is many large machines with many large zones that can be connected together if the seperate companies running them want them to be. I seriously doubt that TVA is buying any power from other utilities in my zone when they have that beast of a Reactor working less than a mile from my house. (Side note-I'm a Power Electrical Engineer.)

The walls of my house are about R-20, 2*6 + rigid foam, and my next house will far exceed that. Becuse the tank is so small, adding extra Rs to the tank is much less expensive than extra R-s to a house. It wouldn't cost much to get one of those tanks wrapped in a foot of foam, but I can see your point about it being bulky. And even though 3/4 of the year it is "helping", I would much rather heat my house with a more efficent source. It's the same logic as clf's vs incandecents, even though the incandecenst are heating your house 3/4 of the year.

Off peak electricity is actually "greener", becuse in the US, we have many many megawatts burning off as waste heat at night because most of the base load generators can't be turned off at night. Even though the base generators still aren't "green" in most locations, using the energy for something constructive is still greener than wasting it to heat. Thats why electric charge at night cars are so attractive, even though they are less efficent at turning inital energy into motion.

Point taken on the more nuanced model of the grid interconnection, but as long as it's interconnected, whether one's buying from a remote generator or not, it's still basically one machine. As long as the local generation meets or exceeds the local local load, the transmission losses are also limited to just the local grid (and it still saves more carbon to park the load next to a coal burner than to a nuke, eh. ;-) )

Off peak load dumping is a major problem with using nukes as base generators (France is practically paying Spain & Italy to TAKE it all night long so as not to have to throttle back their nukes), and of the large base-generator model in general. In heating-dominated Netherlands & Denmark are doing pretty well with micro & mini gas-fired cogens (getting over 85% fuel-utilization) and other small sources supporting 35-40% of the grid load even without smart grids (with anticipation of much greater things to come.) But the carbon footprint of heating with off-peak isn't really ALL that green as compared to much higher end use efficiency or using a much more flexible burner control, with no need to dump heat in the first place. (But I understand completely how it's greener to use power otherwise wasted by the nuke sending more heat into the heat dump.)

With R20 batting and some insulating sheathing your house just MIGHT squeak ahead of the high-R electric tanks, even if the national average is far below that number. But if you crunch the numbers, the total power/heat loss from an R20 tank is miniscule compared to that dissipated by incandescent lighting vs. high efficiency lighting, or as a fraction of the total power burned in the tank elements. (Bottom of the line R4-insulated tanks are pretty eggregious though, eh? Especially in places with high air-conditioning loads.) A typical cheapo 50 gallon tank with a 0.90EF uses ~4800kwh/year, which means over 365 days it "wasted" 480kwh in standby losses (most of it from jacket losses, fixable by increased R-value), or 1.3kwh/day. It doesn't take all that many 75W-100W incandescents running 3hrs/day to beat that level of waste- I see a lot of "kilowatt kitchens" with a dozen or more R30s blazing away, even here in the land of $0.18/kwh residential power rates. They've wasted the same amount of power in the kitchen between the time dinner got started and dessert was served than the bottom-of-the-barrel electric tank heater wasted in 24hours.

The typical "high efficiency" electric tank only cuts that by half. How much do we want to spend on saving that last 1/4kwh/day at the tank? Given that the distribution losses within the house are typically 15% or more, isn't the focus more rightly placed on standards for hot water distribution layout (and possibly pipe insulation)? Wouldn't putting multiple point-of-use 0.90EF mini-tanks around be more efficient than running the entire DHW system from a central 0.95EF (or the mythical 0.999EF) high efficiency central tank? (Methinks, it is...) There's more to making the system efficient than making the tank less lossy, and it's the system efficiency that is what determines the true level of power waste. The shorter the distribution pipe the better, especially where end use is a preponderance of frequent short to very short draws (lavatory sinks, etc.). The distribution losses from a shower or bath can be pretty low even at a distance, since the distribution pipe is filled very few times/day compared to hand-washing situations.

EF is a lousy test- it doesn't address low-volume use standby or short-draw issues very well, but I have a good hunch that even at low volume use, dramatically reducing distribution losses with point of use tanks does a lot more good than just raising the efficiency of the central tank. See:

http://www.nrel.gov/docs/fy03osti/32922.pdf

http://aceee.org/conf/08whforum/presentations/1a_davis.pdf <<
But back at the tank, it takes a heat pump to get anywhere near the same source-efficiency as a bottom-of the line barely-legal NG gas heater (let alone a really good one.) The grid as whole needs a whole lot of greening-up before that won't be true-in-general. (If we're ascribing local grid sources to the carbon footprint of the load, imagine the difference between an electric vs. gas water heater in coal-heavy UT or WY!?! But even in nuclear/hydro TN there is a big coal component.)

The electric automotive fleet issue works on several levels, not just off peak loading. The "smart garage" concept to be able to use the grid-attached car batteries for peak load support has some merit, but it'll take a large infrastructure investment to get there- not holding my breath on that one.

Jelly,

It is true, they do take a lot of amps but they are efficient. The cost of operation is based on your price per KW/h. If you are in the .03 to .06 range - I would seriously consider electric tankless over convential tank type (savings ,I think, is 30% to 50%).

Go to this link for excellent techs and specs (this company has been very helpful):http://www.houseneeds.com/shop/HeatingProducts/WaterHeating/stiebeleltron/stiebeleltronspecs.htm

If propane is an option, get a tankless unit that is approved for the 30% federal tax credit. Same company (houseneeds.com) has that info also.

Hope this clears the air a little

yoda, I've read those efficiency claims, but I've also read quite a bit that indicates the claims aren't quite accurate. Add to that the extra initial cost of the tankless units and some of their functional problems (I lived with a tankless unit for ten years and suffered cold showers every summer), as well as the increased amperage and I'm not so sure the payback is there.

But having said that, I really think you should write your posts in another format... "True, it is true, they do take much amps but, efficient, are they. Yeesssssss." ;)

Distribution losses can exceed the difference in as-used performance between a tankless & tank ELECTRIC hw heater. Reducing the distribution losses by careful placement of the tank (or even additional point-of-use minitanks) can be a bigger net efficiency boost for less money than a single centralized tankless with a mega-amp service.

As always, its the system, not the heater, that determines the ultimate efficiency. The difference between a 0.99 EF tankless and a 0.95 EF (or even a 0.90 EF) tank is all too easily defeated by lengths of plumbing.

Electric Tankless vs. Conventional Hot Water - neither. You want geothermal, and not just an opportunistic de-superheater.

If I had an extra 40k I might consider geo. But I don't, so...

True!

The proof:   http://www.nora-oilheat.org/site20/uploads/FullReportBrookhavenEfficiencyTest.pdf

But note, getting the boiler correctly sized is critical to performance- look at the differences between 2x & 3x oversizing in the Brookhaven study:  The summertime "water heating only" mode drops below even standalone tank efficiencies, even if the annual efficiency beats a standalone tank. (Oversizing 3x or more, meaning the boiler can deliver 3x+ the heat required even on the coldest hours of the heating season, is all too common.  This is why the Brookhaven study looked at those configurations, not just the rare idealized perfectly sized boiler.) It's the SYSTEM, not the boiler's efficiency ratings, that ultimately determines what your efficiency is.  Oversizing is universally bad for efficiency.

One way to buy a significant portion of the steady state efficiencies back on oversized boiler (and even gain a significant shoulder-season efficiency edge with correctly-sized boilers) is to use a "reverse indirect" hot water heater plumbed and controlled as a buffer tank for the heating system rather than a seperate zone. 

eg:  http://www.ergomax.com/New-Tanks.htm  or http://www.thermo2000.com/pdf/en-US/manu/turbomax.pdf

By making the tank, not the boiler the center of the system and keeping the boiler "agnostic" of the zone calls, only serving the tank a it's master, the thermal mass of the buffer tank creates a guaranteed minimum burn length- the system can never "short cycle" and lose efficiency, even on micro-zoned systems.  (In the tankless heater scenario, their efficiency falls off a cliff in the sub-1-gallon handwashing draws due to short-cycling effects, even though their average efficiency still beats a standalone tank every time.) For a pretty good discussion  on how/why buffer tanks improve effiency (and why it's a good idea, even with low-mass modulating boilers) see this:

http://www.patkelco.com/uploads/files/4bf77cff8bc8411bb00c67e962d0ef7d.pdf

This discussion may be couched in large-facility system terms, but the math works for micro-level residential sized systems too- only the decimal point shifts- the percentage savings remains the same.

As houses get tighter and average heating loads drop, the instantaneous domestic hot water load can easily become several times the peak heating load (and an order of magnitude higher than the shoulder-season average), and boilers with indirect hot water get sized for the DHW load instead of the the heating load.  In those systems setting up the hot water heater as a separate zone leads to lower system efficiency, since the boiler is now oversized for the heating system.   Going to a reverse-indirect buffered topology lets the mass of a buffer tank keep the heating system efficiency up despite being oversized for the load.

The only system in the Brookhaven review that was remotely like a reverse-indirect-as-buffer system was #2, which used a sophisticated set of controls to guarantee minimum burns, using an indirect zone as a heat dump when the minimum burn length exceeded the heating system demand.  With a low-mass boiler the buffer-centric approach can achieve similar results to the low-mass steel boiler/indirect system #2 with less complexity.  Even medium-mass systems with cast iron boilers will see significant gains in annual efficiency, but will still suffer summertime efficiency losses if they're oversized.  Sizing the burner to the load is critical.

BTW: "...just 1° F per hour at 180° F...." standby loss on a 40 gallon indirect adds up to ~8K BTUs/day, or ~3,000,000 BTUs/year (which becomes ~35 therms of natural gas use, or ~25 gallons of heating oil at typical "pretty good" system efficiencies).  If you can reduce the standby temp to 120-130F and still have enough hot water for your needs you can cut that heat loss in half.  Alternatively (or additionally), adding R20 of insulation to the indirect will cut it by half or more, and is usually cost-effective as a DIY project.  Insulating the near-boiler plumbing to the tank to at least R6 is also usually cost-effective.  If your total fuel use is 1000 therms/year, that's a very cheap ~2% system-efficiency gain.  If you use 500 therms that's a ~3-4% efficiency gain.

Also:  Circulators can and ARE adapted for use with gas-fired tankless heaters, (your point #4), but it can cut into efficiency significantly if it's set up to maintain a minimum temperature at the remote sink, even if the pipes are insulated. (It generates an efficiency robbing short-cycle on the burner.)

But of course, none of this has anything to do with Jelly's all-electric system in a cooling-dominated climate. :-)

The frequency of boiler cycles is determined by the hysteresis in the aquastat, not the storage temperature.  (It takes just as many BTUs to go from 175F to 185F for a nominal 180F setting as it does to go from 115-125F for a nominal 120F.  It's the same delta-T, the same number of lbs. of water.)  If it's being maintained by a mod-con, any time the  the temperature of the return loop to the boiler from the heat exchanger is under 122F you get a significant condensing-mode performance boost that is NEVER there if you're maintaining a 180F temp.  But even with cast-iron boilers, lowering the return water temps to the boiler results in a ~3% fuel savings for every 10F you lower the temp, a result of improved combustion efficiency.  (Sustained burns of under 130F return water can be an issue for NG fired cast-iron boilers though, since it increases flue condensation/corrosion. Below 125F there can be a hazard of destructinve heat-exchanger condensation as well, but there are many solutions to those problems.)

If the controls for the indirect are more complex, and allow a much bigger hysteresis, yes, you'll get longer burns in standby-maintenance mode, but if you're using some hot water daily, almost all burns will be true recovery burns, not maintenance burns.  Most sophisticated controls can set a minimum burn time (many mod-cons may have that function incorporated into the boiler controls anyway. Short cycling losses aren't likely to ever become an issue with 40 gallons of thermal mass to work against.