Tankless Water Heater Hazard

Unfortunately that's not all that uncommon of problem to have happen regardless of brand or model. Some are better than others but still pipecorsosion and mineral buildup will affect most of those types of water heaters at some point.

Steam burns are lawsuit bait.

A double pole relay fed through that thermal cut-off would prevent it.

A word to the wise....

"During our long-term testing, an indicator on the tankless model warned of scale buildup. We paid $334 for special valves and a plumber to flush out the water heater with vinegar. Many industry pros recommend that tankless models be serviced once a year by a qualified technician. Calcium buildup can decrease efficiency, restrict water flow, and damage tankless models. Experts suggest installing a water softener if your water hardness is above 11 grains per gallon. Ignoring this advice can shorten your warranty."

http://www.consumerreports.org/cro/appliances/heating-cooling-and-air/water-heaters/tankless-water-heaters/overview/tankless-water-heaters-ov.htm

Thanks. I am now aware of that. I didn't learn about it until after I installed ours. I intend to install the service kit and mineral reduction equipment when I replace this unit.

A full water softener system is in the neighborhood of $2000. (I could replace the whole water heater every couple years for less money!) 3M "Aqua Pure" is a mineral treatment cartridge system for $62 on Amazon.

YOu are definitely making me appreciate my 'less efficient tank type heater' that always has hot enough water in the flow rate I need and that will run maintenance free for a decade or two.

So a $2000 water treatment system plus a few hundred a year on servicing and consumables plus a few hundred more a year on tankless heater maintenance...

Lets see now that works out to roughly ~$5000 a decade to operate plus purchase costs which would buy me about 15 years worth of electricity for my cheap low efficiency water heater.

Sorry just not seeing how a tankless water heater is more efficient and cost effective in the long run once everything else related to it is factored in.

Me too. I was intrigued by tankless water heaters for a while, especially raising 4 kids in the house (teenager showers ) we kept that gas fired 40 gallon unit humming along. But we purchased the house in '96 and the nameplate indicates the water heater was manufactured in '91. Still going strong but we are probably on borrowed time with that water heater that almost 25 years old. Our municipal water has some hardness as evidenced by a modest buildup of minerals in the supply lines but not too bad. I do make a point of flushing the crud out of the bottom of the heater about once a year. I usually get about a quarter cup of sediment in the five gallon bucket when I run a hose into it from the drain valve.

I figured that raising an active family never lets hot water sit around for very long so eventually I talked myself out of it. Plus I like the fact that when the power goes out, our natural gas fired hot water heater still makes hot showers.

Yea my last water heater was propane fired and it made it about 30 years before it rusted out from the outside in.

After that I went to electric for the convenience and to get rid of the second of the last gas fired appliance in the house.

As of two weeks ago all the original propane lines got taken out and now I have a single line feeding directly from the outside main regulator to the furnace and nowhere else which eliminated about 75 feet of iron pipe and ~20 assorted fittings of which to my amazement after 40+ years of service about half of them were only hand tight at best when I was expecting them to be rusted solid.

I guess the pipe dope kept you from smelling the ethyl mercaptan from those less than tight pipe joints. Given the price of propane, I don't blame you for trying to reduce your consumption of it. How does your price of propane compare on a BTU or joule basis to electricity costs out your way? I understand that heat pumps are about useless in your neck of the woods in the winter months.

We have natural gas from the utility that is hard to beat on a joule basis. A solar thermal system for hot water and/or space heating would require decades to realize the ROI. I like the idea of 'free energy' but the equipment to collect it is anything but.

"I guess the pipe dope kept you from smelling the ethyl mercaptan from those less than tight pipe joints."

I'm guessing that TCM's "hand tight" is what we would call "wrench tight," since he gave 'hand tight' and 'rusted solid' as the only conditions of the piping.

And if the joints kept the ethyl mercaptan in, it also kept the propane in, so they were 'tight enough.'

Hand tight as in grab a elbow or Tee joint bare handed and turn it or what my 7 year old could turn with a common vice grips. Way too loose even for a low pressure gas line by my standards.

I had figured that after 40+ years of being in service that everything would be rusted solid since the piping looked pretty rusty on the outside. Was I ever surprised and a bit mad when I put the 18" pipe wrench on roughly horizontal to some fittings and it's own weight was enough to get them to turn.

All of the new line is doped and torqued to 40 - 50 Ft/LBs like it should be.

Two comments:

1) I stand corrected on my interpretation of your description of 'hand tight.'

2) 40+ years at 'hand tight' and the house hadn't blown up from a leak, that is some abso[*BLEEP*]ing FANTASTIC pipe dope. Wish they made it that good these days.

Yea I had a few thoughts for whomever was on the crew back when this place was built.

Guessing someone just doped things up put them together hand tight and never came back to do the final torque down before the next piece went on. Most of the fittings were good but it was the fact that there were several random ones that were that loose just makes me think it was done in a hurry of was done by more than one person.

A far as the dope goes that stuff was dry as an old stone.

"A far as the dope goes that stuff was dry as an old stone."

Dry, but apparently still 'air-tight.'

There is,in fact a pipe dope like that.

LocTite makes it.It is an anaerobic adhesive type pipe sealant.

I don't recall the name,but you can put pipe together with very light force,and after a few minutes, it is permanent.

It can be taken apart,but much effort is required.

This was developed with the fire protection sprinkler systems in mind, when hundreds of joints are made throughout a building,and repairs to leaky joints would be very expensive.

I have used it on loose hydraulic threads,and even at 3000psi no leaks.

Not sure if it is rated for potable water,though.

I will try to dig out the name if interested.

My bulk tank fill a few weeks ago was 95 cents a gallon and our electrical rate is ~12 cents a KWH.

About the only time I run propane as the primary heating fuel is in the fall and spring on these days where it's not worth the effort to load up the boiler in the evenings with wood or keep it running 24 hours a day on used oil for only an hour or so of heating at night.

My main reason for switching things over to electric has been more to do with AE power and my futile efforts to keep a grid tied wind generator working.

I just can't get any decent blades online that hold up to our winds and weather. Fiberglass weather sofast and either tear open at the seams or rip out at the roots and steel blades stress crack due to turbulence buffeting. The only blade type that has ever lasted here is the good old fashioned all wood hand carved ones and I have just not had the time to spend or the right tools to carve a set.

Next time read the warranty before you call the manufacturer. It clearly states that damage from corrosion isn't covered, something I'm sure that their tech support is taught to always blame for any failures, and is almost impossible to be certain of without an internal inspection.

According to the wiring diagram there are two or three sealed heating chambers, each of which has a double pole thermal safety cutout. The control of the elements takes place via triacs on a pcboard which monitors the output temperature and fires the triacs accordingly.

It does seem odd that the only overpressure protection is the thermal cutouts, but when you consider that it has a very small volume of heated water you have to conclude that they are also relying on the volume of water contained by the supply piping to absorb any overpressure from a stuck thermostat.

Consider yourself lucky that you didn't get scalded, and replace the unit with one that has better protection. If you want to push for damages you'll need a lawyer that specializes in product liability litigation.

You must be looking at a different model. This is an AE9.5. There is one heating chamber that both the elements are in. It is a copper pipe that curls around the perimeter of the enclosure. It's about 5/8" diameter and les than three feet long. There are two single pole microswitches operated by a mechanical flow actuator. There are no triacs and no circuit board.

There is no thermostat, just a thermal cut-off that you manually reset when it goes off(at 194F according to the manual). But that didn't happen because the cutoff and flow switches control only one leg of power. The other ties right to the other end of the elements.

I don't think pressure really built up that much because it would have backed up into the supply pipe.

The tech support guy knew immediately what happened when I explained the symptoms, so he's heard this problem before.

Ironic you should post this today. This morning I posted an answer to a tank electric hot water heater question, concerning this same problem and how to cure it on a tank unit. Refer to "Re: Hot Water Heater Question" post #20.

Not only will this happen due to calcification but also water rich in iron or rusty water. They both cause premature, very premature, cracking of the element electrical insulation. As you described the problem creates an electrical path from the element through the more conductive than ordinary water to the ground thus having no switch (thermostat) on this errant circuit.

For a tank heater a way to prevent or at least slow this down is to drain the major portion of a 5 gallon bucket from the bottom drain valve each month. This helps remove the precipitated calcification and other conductive solids. It also helps remove any solids such as sand that may have been brought in with the water.

For either a tank or a tankless heater, if it is rust that is causing the problem, put a particulate filter, 5 micron or smaller, on the inlet pipe to the heater. This will remove solids including rust from the water keeping it out of the heater. Depending upon your use, replace the filter cartridge at least every month. If you get one with a clear plastic housing you will be shocked (no pun intended) at how quickly the filter cartridge gets dirty. I would recommend putting this on your heater whether or not it currently needs it.

Hope this helps you a little. Ironically when I posted this on the other blog it was OT'd. I hope it is of more help or interest to you than it was the OT'er.

Good Luck, Old Salt

Install an earth leakage circuit breaker (has different names depending on where you live). If it detects a difference of about 30mA between current in Active and Neutral, it trips. A particular case is an earth short like you had. It would detect that there was Active current but no Neutral current (because Neutral current was flowing thru the Earth circuit). Instant off!

In America that's spelled GFCI for Ground Field Circuit Interrupter.

There's a nicely detailed article on it on Wikipedia (which I consulted because I wanted to make sure i was spelling it right, my hands wanted to type GCFI for some reason).

Do you have an expansion tank on the hot water heater?

It will temporarily absorb excessive pressure when the water gets hot,or overheats.

Some local codes require this if you are on public water supply,because a check valve is required at the entry to the house to prevent possible back flow.

If you are on a privately owned well, there is normally not a check valve required at the tank,so the storage tank itself absorbs the pressure.

I tried a tankless heater,and the problem was regulating the temperature to a desired level.

The water left in the exchanger would overheat when the flow stopped,and the next time you used it,it would be much too hot for a few seconds.

This unit had 3 heater elements enclosed in individual copper tubes as heat exchangers.

They were staged according to demand and switched by triacs,and there was a magnetic flow switch to detect flow.

Looked like a good system on paper.

I even tried a PID controller and mixing valve on the system,but it was still very hard to control due to residual heat in the copper tubes which comprised the exchanger.

I finally admitted defeat,and installed a regular hot water heater.

With the modern insulation of the tanks, they do not lose much heat.

I have been on vacation for a week,with the heater turned off,and the water was still hot when I returned before turning unit back on.

I don't have an expansion tank. Wouldn't the excess pressure just back up into the supply pipe? Even if there is a check valve from the street, I would think the house and yard plumbing could absorb a few cubic inches of expansion.

I'm installing tankless service ports. They are fitttings on either side of the heater that allow a pump to be attached that pumps mineral cleaner through the unit.

I've also noticed the temperature rise briefly when restarting flow. I'm putting a temperature mixer on the outlet. It senses the temperature of the hot side and throttles it down to allow more cold to be added.

And with all this extra crap installed, I'm going to have to move the whole thing under the house!

CAVEAT: Comply with your local codes regardless of my less than expert opinion.

If you are on city water, there is probably a check valve to prevent backfeeding the municipal supply.

A tankless heater probably does not require an expansion tank like a regular hot water heater fed from a municipal supply. Reason being is that the volume of water being heated is much smaller, unlike the change in volume that occurs when you heat 40 gallons of water from 60 degF to 140 deg F.

I found this calculator on the Watts web site.

http://www.watts.com/pages/support/sizing_DET.asp

If you heat 2 gallons of water from 60 degF to 140 degF, the acceptance volume is about 0.03 gallons. That corresponds to a volumetric increase of about 1/2 cup. If you have a 40 gallon hot water tank, when you heat that the same amount, the volume increase is about 0.6 gallons. The system pressure can rise dramatically as the volume starts stretching (and stressing) everything in your plumbing system.

In a previous residence of mine fed by the city of Worcester at over 120 PSI, I had a regulator in the basement that dropped the pressure down to 45 psi and instead of an expansion tank, I had a relief valve that emptied into the wash sink in the basement set to 55 psi so that when expansion occurred, it would just dump a little water 'overboard'. The relief valve would also save the plumbing if the regulator got stuck or started to fail open.

P.S. In that house, I put one of my outdoor hose faucets on the 'high side' of the regulator. That made for quick washing and rinsing of my car.

"P.S. In that house, I put one of my outdoor hose faucets on the 'high side' of the regulator. That made for quick washing and rinsing of my car."

That must have worked nicely back then, but these days you can pick up a cheap ($10USD or so) 'power washing wand' which will take 'low pressure' residential water and, through the magic of a Venturi nozzle, transform it into a high pressure stream or fan of water, depending on the tip you use on it.

Those wands are also hood for shooing away pigeons who are trying to monopolize your bird feeders, then simply flap up on top of the garage when you get near, with the intent of resuming the bir feeder raid once you've left. City pigeons are like little gang-bangers, no respect for property rights, or for their fellow birds.

Tank-less heaters must have a built in expansion system for the heating, but not for the hot water, you cannot see it, but its there. Its a large dish or round shaped unit with a thick rubber membrane to absorb the volume changes.

Mine is inside the water heating system, which does hot water and house heating, but they can also be external to it, something like this one:-

...which is actually for a system where a hot water check valve is installed, but it is the same type of unit as this is. It simply allows expansion and contraction. Water changes its volume quite a lot with varying temperature....

Best of luck.

Brave Sir Robin,

Your analysis is typically darn good. However, this time I think you have not thought this one through in your typically thorough fashion.

.

There is no reason that two similar failures , one in a tank less and one in a tanked waterheater would heat the water to the same temperature in an accident such as this. While a lower volume of water will expand less in absolute terms, if heated to the same temp, in an uncontrolled addition of heat such as this it will see a much faster rise in temperature for the same heat input and thus a much greater temperature and pressure rise, all other things being equal.

.

There much less time before a tank less waterheater begins to resemble a pressure cooker in a situation like this than for a tanked water heater. While a tankless system does have higher total surface area to volume ratio than a system with a large tank, the temperature rise for a two gallon tank occurs about 20 times faster than in a forty gallon tank system.

.

While it is nice that a relief valve backs up the regulator, relief valves are not immune to failure.

Whether or not the relief valve opens, the smaller volume also increases the possibilities of a fire should there be a fault in the electrical system such that a circuit breaker is delayed for some arbitrary time.

If the tank-less heater is at the point of use, steam burns are also more likely.

.

Are the elements on typical tank-less and and tanked system usually similar in output?

.

The marketing people for tank less systems should be ashamed for promoting as one of the benefits the reduced chance of flooding if this is a common tank less issue.

With respect to your "relief valves are not immune to failure" fortunately they are a fail safe device that will only result in a leakage.

USA codes mandate relief valves on all standard water heaters (that is one with a downspout on the side of the water tank in the diagram showing the pressure expansion tank); but, I cannot speak as to the quick heat unit requirements.

Wow! Relief valves are immune to sticking shut! Incredible! Truly unbelievable. No really, I'm being serious (and absolutely literal) this truly is unbelievable.

.

The preceding ridicule is directed at your attempted condescension 'with respect' to (my?) relief valves. Yes I am aware of the ubiquitous nature of pressure relief valves on water heater. Yes I am aware thuis is code.

I am also aware that ubiquity and being a code requirement does not bestow magical powers.against corrosions, fouling, improper use, inadvertent damage, line blockage, or manufacturing defects.

.

Yes a relief valve failing to operate properly will be rare, but as it turns out water heater failures are not exceedingly rare. The chance of both occurring is not out of the question.

What is out of the question; the existence of your magically perfect pressure relief valves.

That was not intended as a condesending response; but, as a rebuttal to a loosely made statement about the failure of relief valves to open.

In that respect, the last 20 years of my career were spent in new product development, testing and ultimately engineering design supervision of high pressure relief valves for ASME Section VIII Pressure Vessel Code service; of which, there are thousands of that type of valve installed in chemical and petroleum process plant services around the world.

Thankfully, during that entire period I was never called on to address the issue of a valve failing to open and provide the critical protection for which it was installed.

Alternatively, on occasion, I was called on to address an issue of a valve prematurely opening, leaking or exhibiting a blowdown longer than the customer preferred.

So, while I would never claim that a relief valve could absolutely not fail to open due to one of those issues you stated, I have a pretty solid basis for my response.

As far as anecdotal evidence.goes, yours is in the top of its class. It does provide a little insight and I can tell you I learned someone can work with valves for 20 years and never have one stick shut.

.

Valves do occassionally stick shut.

Drain lines may become packed with mud and create a robust seal.

Sometimes a rare manufacturing defect escapes notice.

People are prone to error and can botch even something as simple as plumbing a water heater.

Atypical Joe Schmo does not do any preventative maintenance or even cycle valves not in regular use. Even if these valves were sticking closed as frequently as elements fail, it would go mostly unnoticed...."the water heater needs to be replaced." "Okay replace it" No one is.going to test the valve to see if it opens too slowly or too late or for.evidence it had been stuck.

.

The point is, for these valves, because these are not refaced or repacked or otherwise rebuilt after being in service, you are not going to have personal experience for every mode of failure possible for these valves, not even necessarily all the more likely modes of failure. Moreover, the valve might not be the.soirce of the problem, as in the case of a plugged drain

Experiencing something first hand does bring certainly about its existence. Having experience something does not disprove its existece.

.

Believe what you will.

The main point of all my posts is that , based upon the successful percentage basis of relief valves for more than a century, it is far better to have a relief valve than not to have one.

I also included in my post that I would never claim that there is absolutely no chance of an opening failure. In that respect, as opposed to relief valves, control, shutoff etc. valve malfunctions do regularly occur due to a multiple number of factors. That is exactly why relief valve codes require self actuating relief valves as a last stage of defense in protecting pressure systems. Additionally, it is also the reason that the design, material, certification and installation requirements of relief valves are held to a much higher standard than most other in-service valves.

Even for a water heater tank, if one takes an opportunity to inspect the installation of its relief valve it will be noted that this valve is specifically located high on the tank remote from the lower tank region of contaminant collection and high temperature calcification around the heating elements. As for corrosion, the valve body is bronze with stainless steel internals, all being well suited for corrosion resistance in domestic water systems.

As to my basis being anecdotal, ie observational, in the real world all evidence of reliability of every product, from the FDA product approval of prescriptions to the FAA certification of aircraft and every product in between is based base upon testing, which is observational. Additionally, the ultimate efficacy and reliability indexes for all products are based upon their observed history of performance and safety.

Even in the worlds of physics and astronomy, no matter how strong the mathematical evidence, no theory is considered to be absolutely confirmed until it or a predicted associated action is observed.

Of course, that is the problem with anecdotal evidence; many people fail to recognise the distinction between anecdotal evidence and rigorous empirical data.

The FDA and FAA and many other TLAs eschew anecdotal evidence for good reason. Their decisions are based on data gathered in ways that is.careful to exclude the problems inherent in anecdote evidence.

.

Relying on anecdotal evidence (instead of the far more tedious methods required for legitimate empiral evidence) leaves the door open wide for introduction of a veritable who's who of cognitive biases and logical fallacies.

.

So, yes, real world validation through testing is very important. Far too important to subject to heuristic distortions and cherry picking recall.

In this case I wasn't really addressing the failure mode but the OPs question about the requirement for an external expansion tank for normal operations. All the water heaters I have seen have a thermal/pressure (T&P) relief valve for extremis conditions versus normal heating and cooling. Usually those valves will open at 180 degF and I don't recall the over-pressure relief rating. I think it might be around 125 psi. On the Watts website they call out 210 deg F and 150 psi but the hang tags on the ones I've seen are lower rating.

Well, as for marketing people, I work in engineering so you know my view on that.

Have a good one, Cheers !!

Oh, that makes much more sense, both as an answer and as coming from you. Got it.

Many houses use tank-less heaters here in Germany, but we also often have thermostatic water mixers/taps, which handle any pressure (someone flushing a toilet for example) and temperature changes by keeping the water at a required temperature, by mixing the hot and the cold, fully automatic and really fast....

I would not have anything else...

As anything else is either dangerous or uncomfortable, or both, one shortly after the other!!

It sounds like a good reason to have a RCD, which would have surely tripped with the fault you are describing (if I understood your explanation properly!).

I have added them to every house I have owned and had them added in several rented ones too....

Although it does of course usually save people and animals from being electrocuted when something goes wrong!!! It would have prevented scalding steam in your case I believe.....think about it!!

By the way, it does sound like a corrosion problem to me, maybe combined with hard water....

Although their product is seemingly badly designed (none of us will ever buy one!!), any short to ground should cause an immediate drop of power to that circuit at least!! Usually the whole house with a single phase supply....

Best of luck!!

I have 5 of them installed, with three being replaced, two under warranty. The problems are legion, the worst being something similar when the high temp switch tripped (a rare event), and, while I was standing on the step ladder in front the unit in my basement, I called to my assistant to flip the breaker back on after I had manually reengaged the switch button. All seemed to be OK for a few seconds while my assistant went to the nearest tap to check water temperature, but just before the tap was opened I heard a rumble in the unit and the internal compression joint blew apart on the cold size spraying water (fortunately cold) all over me and the Bosch Powerstar electrics, causing the panel breaker to trip.

Not surprisingly, the Bosch tech was in disbelief, but sent me a new unit, which, by now, I am very experienced at removing and re-installing. Scale build up (mostly at the hottest point in the unit's coil pattern) will frustrate the thermal safety switch function. An added safety feature needs to be a high pressure shunt (as with tank heaters) on the unit that bleeds at a marginally higher pressure above the peak utility or private system pressure to a secure hot water drain or small drainable receiving tank. The high pressure shunt should have an interlocked toggle-to-off switch that is in electrical series with the high temperature switch that has failed and is still closed. Consumer info should make it clear that if this secondary safety function (toggle-to-off) occurs, the entire unit requires replacement.

But don't get me started on the temperamental flow switch, circuit board failures due to overheating and the little temperature rheostat or whatever dial setting. I have a sixth unit (30 amp, 240 vac) that is the oldest (12 years) and was built by the British company Bosch bought; it has a more sensitive thermal switch which requires a manual reset about twice a year, but it keeps on plugging along.

Around 20 years ago, I was visiting in Chile. The home I was staying at had an instant tankless water heater. Not being familar with these types of heaters, I stepped into the shower and turned on the hot water fawcet. I was immediately hit by a blast of boiling hot water that took some skin off my stomach. I'm guessing the temperature was above 180° F. I guess it was a second degree burn and took about a month to heal. The burned area was very painful and is still visible today. I should have turned the cold water on first, but didn't want to be greeted by cold shivering water. I have been thinking about getting a whole house tankless water heater, but afraid of the burn potential.

Some of those are showerhead units. I saw them in Costa Rica. Screwed on to the shower pipe with about three inches of wire hanging off in mid air with wirenuts. It's common to get shocked from them. Needless to say, they are not UL approved nor are they available in the US. If you can bring them back yourself though, they are great for an outside shower if you install them correctly and use them carefully.

....and install RCDs on all phases?

(I know that US & the UK usually only get one phase per house, but here we have all three coming into a house or any building for that matter!!)

(RCD = GFCI in USA)

I did install that in the new kitchen unit. I also have one for when I replace the bathroom one. But I'm not sure it's essential.

Under what circumstance would a water heater be dangerous that this would prevent? I can hypothesize and imagine an insulation break inside the heating chamber that would send current through the water. Neither the new one or the old one call for it in the installation instructions, and I've not heard of it being required in electric code.

My old tankless DID short out halfway down the chamber. It shorted to ground and overheated the water. That current was not enough to trip the breaker, but it did overheat the water and cause a scalding hazard. A RCD/GFI would have detected that short to ground and shut it off.

My new one has temperature overload that shuts off BOTH legs of the power. The worse that could happen is a short AND malfunctioning TO switch. Hmmm.

Its also there to safeguard the humans and their pets from being electrocuted by any contact by them to a phase...not necessarily just electrical water heaters.

Though, as you mentioned, they do trip rather well on water heating element leaks to ground....

Years ago, I had a kettle that occasionally did it!! Annoying!! Till I identified the problem and replaced the element....

Sometimes identifying the reason for the trip takes some good discovery work!!

By the way, you mentioned correctly that "RCD = GFCI in USA", but the main Wiki page for such devices is still called "RCD".

Check here:-

http://en.wikipedia.org/wiki/Residual-current_device

It notes that "GFCI" redirects here, as do GFI, or ALCI and even RCBO.....

So they seemingly consider "RCD" to be the "main" name world wide, I have no idea why!

Plus I do believe Wiki is an American website too......interestingly!!

These heaters have always struck me as dangerous, in spite of what they claim. To heat water in real time, you have to be supplying quite a bit of power. Water has a high heat capacity, and for some reason it is not there to absorb the energy and the safety cutoff fails, the situation could turn out badly.

OP Follow-Up Thread and with Buyer's Guidance:

http://cr4.globalspec.com/thread/104860/Tankless-water-heaters