In this article, I’ll show you how I used an ordinary stove element to convert my inefficient natural gas hot water tank to an ultra efficient electric tank on a timer, decreasing my energy consumption for water heating by about 80% and saving us about $175 per year (more when you consider that we were also able to cancel our natural gas account as a result of this conversion, saving an additional $140 per year in “basic charges” that the gas company charges regardless of the amount of gas consumed). Warning!!! This article describes modifications to a natural gas appliance, as well as custom 120V AC electrical wiring. Attempting to duplicate the modifications described here could result in injury and/or death and/or significant damage to your home. Do try this at home (where else would you?)… but do so at your own risk.
Why are electric water tanks more efficient than gas ones?
Put simply, most gas hot water tanks put more heat up the chimney than they put into the water. A gas hot water tank is effectively an insulated tank with a hollow tube running through its center. Hot exhaust gases from the burner at the bottom of the tank pass through the tube. The tube is designed for good heat transfer to the water around it but much of the heat goes right through the tank, up the chimney. Feel the chimney while the burner is on and it will likely be too hot to touch. A tremendous amount of heat is lost up the chimney whenever the burner is on. However, that isn’t the worst of it. Most gas hot water tanks in use today have continuous pilot flames and open flues. This means that even when the burner is off (which is most of the time), the heat transfer happens in reverse. The hot water in the tank warms up the air in the tube (recall the tube is designed for maximum heat transfer). The air rises up into the flue. The result is a continuous flow of warm air through the tube, extracting heat from the tank 24 hours a day and dumping it out the chimney. Feel the chimney above your gas hot water tank any time that the burner is off, and you will find that it is still quite warm to the touch due to all the hot air rising from the tank below.
Experiments on my own hot water tank showed that simply capping the tube at the top of the tank by placing a block of wood over it reduced the rate of heat loss from the tank to around 60% of it’s normal rate (ie 40% of the energy loss is up the chimney). If your gas hot water tank is located inside a living space, then I have even more bad news for you. In addition to carrying away heat from the tank, this continuous flow of air up the chimney carries away heat from your home. Air from within your home (that you’ve already heated up to room temperature) is drawn into the bottom of the tank, passes through the tank and is expelled out the chimney. The removal of this heated air from your home causes cold air from outside to be drawn in through all the cracks and openings in your home’s building envelope. It’s the same effect as having a small window open continuously. Capping your gas hot water tank (and plugging the chimney) will prevent this loss in addition to reducing heat loss from the tank itself. However, the only way you can get away with capping your gas hot water tank is by turning off the gas.
Ads by Google
Why not just buy an electric tank?
I could have purchased a new electric hot water tank, but that would have been expensive and wasteful since there is nothing particularly wrong with my gas hot water tank. In addition, installing a commercial electric hot water tank would require rewiring my home to supply a 240V circuit to the location of the water tank. I wanted to be able to run my tank off a 120V circuit and put it on a timer. Timers for 240V high current circuits cost hundreds of dollars. Timers for 120V circuits cost less than $20. On pondering the issue for some time, I realized there was a very simple conversion I could do myself with parts I already had or could get for free.
A 120V circuit doesn’t have enough power for a hot water tank… does it?
This is a common misconception. Commercial hot water tanks, after all, have elements requiring up to 10 kW of power while a typical 120V circuit is only capable of driving around 1.5 kW. The reason commercial tanks draw so much power is “recovery time”. Hot water tanks are typically designed for a recovery time of 1 hour. In other words, a typical hot water tank requires enough power that you can empty the entire tank every hour and still have hot water. Hot water tanks typically hold over 100 litres of water. I don’t know about you, but my household rarely requires 200 litres of hot water in a single day, let alone in a two hour period. If you can tolerate a longer recovery period, then you can easily reduce the power draw to something that a 120V, 15A circuit can supply.
One little compromise makes a huge difference
It is important to recognize that you don’t NEED hot water available all day all the time. In my home, showers are about the only thing we require hot water for. We wash our laundry in cold water and our dish washer has it’s own heating element. If you can tolerate restricting your hot water use to a particular time of day (morning for example), there is no reason to maintain a tank full of hot water 24 hours a day just dissipating it’s heat to the surroundings. If you do want hot water available at any time, you can still achieve that with the conversion I’ll describe, but you’ll see a larger reduction in energy consumption if you are willing to run your tank once per day on a timer, heating it up and turning it off just before you use it. Then as you extract hot water from the top of the tank, it will be replaced by cold water at the bottom of the tank. You can extract up to 1 full tank of hot water which will generally suffice for several showers each morning. The tank will then sit full of cold water most of the day, so it won’t be dissipating ANY heat to the surroundings. In fact, it may even absorb heat from the surroundings if the incoming water temperature is lower than the outside air temperature.
How much power does it take?
If you restrict your hot water use to one time of day (and less than one full tank of water), you clearly don’t need a 1 hour recovery period. You could, in theory, allow up to 24 hours for your tank to heat back up. Tests on my own tank, showed that it takes as little as 150W of continuous power to heat a cold tank from 15°C to 45°C in 24 hours. However, as stated above you can reduce energy consumption by allowing your tank to sit cold all day, heating it up as quickly as possible just before you need it. If 150W will heat a tank up in 24 hours, then 300W will heat it up in 12 hours, 600W will heat it up in 6 hours, 1200W will heat it up in 3 hours … you get the idea. Because I wanted to plug my tank into an existing 120V circuit, I did not want to draw so much power that I’d be likely to trip a breaker if I plugged something else in while the hot water tank was on (even though that’s pretty unlikely since I run the tank only at night). I found that a 520W heating element was satisfactory. Typically we don’t use a full tank of hot water per day so the tank is never completely cold and I’ve found it only takes about 4 hours per night to get the water temperature to around 45°C at point of use each morning. By running our tank on a timer between 3 am and 7 am, we have enough hot water for showers, and we only use the equivalent of about 87W of continuous power (520W times 4/24 since we only run it 4 hours our of every 24). This is in mid summer… I imagine we will need more power in the winter when the garage and the incoming water are colder.
How do you do the conversion?
Converting a gas hot water tank to electric is easier than you might think. If I had to do it again, it would take me only about an hour. The original equipment is not modified in any way either, so it’s easy to switch back later if you want to. Warning!!! Be sure to turn off your natural gas completely before performing this conversion. The process described could generate electrical sparks which could cause an explosion if any natural gas is present.
First, find a suitable stove element. You can often pick up a used stove element from any appliance recycling or salvage yard for free, or you can buy one. You want an element that will produce 500 to 1000 Watts of power on 120V AC. Most stove elements are rated for 240V AC. If you know basic electronics, then you will recall that if you halve the voltage across a resistive element, you will get only 1/4 of the power output. So a stove element rated for 2000 to 4000 Watts at 240V will give you 500 to 1000 Watts at 120V. The best method for selecting an element is to bring an ohm meter and measure it’s resistance. Power is equal to V2/R where V=120. Therefore you want a resistance R of 14 to 28 ohms. The resistance will increase a little when the element heats up but not too significantly. Note also that although you will be running the element for much longer periods than it would normally be run on a stove top, it will still last a long time because you will be running it at only 1/4 of its rated power. That said, you might want to pick up a few elements since they may break during the shaping process.
Shape the heating element to fit inside the tube through the water tank. I used a couple pipes as forms to bend the heating element into a “coil inside a coil” shape so that the two ends of the element were on the same end of the coil. See photos for a clearer explanation. Be sure that the terminal ends of the element are well away from the outside circumference of the coil so they cannot contact the inside walls of the tube through the hot water tank that the coil will eventually be lowered into. Test fit the element in the tank. First, turn off the gas, including the pilot flame. Then disconnect the chimney from above the tank. Pushing the chimney aside you should have access to the top of the tube running through the tank. Reach inside and pull out a spiral shaped length of metal. This is designed to improve heat transfer from the exhaust gases to the tank, but it is not needed for the electric conversion so it can be removed to make space for the electric element. Slide the coiled element into the tube. Ideally, it should slide in freely, making only light contact with the walls of the tube. Reshape the coil as necessary for best fit.
Attach wires, electrical cord and cap. I used some high temperature wire from the same salvaged stove I got the element from. This wire has a special insulation rated for high temperatures and also has some connectors for attaching to the element. The high temperature wiring should only be needed in close proximity to the element. About 2 ft of wire should be all that is required but if you have more, you might as well use it. The high temperature wire can then be soldered (and heat shrink tubing applied) to an electrical cord with plug. Before soldering, pass the ends of the electrical cord through a drilled hole in a metal cap. A large jar lid with the seal removed makes a good cap. WARNING!!! Do not use a wooden cap as shown in the photos. This was a temporary improvisation. It will work under normal operating conditions but could be a fire hazard if the element is ever left on continuously (intentionally or otherwise). Using non-flammable materials is just plain better when dealing with any heating appliance. The cap will act as both support for the element, hanging by the wires, and also to prevent air flow through the tank. Tie a knot in the cord so that the element hangs near the bottom of the tank without contacting the burner.
Install the heating element. Ensure that the bare ends of the element will not contact the tube walls or each other. Insert the element in the tube aligning the cap so the cord is centered above the tube and the block covers the tube end completely.
If you have a power meter, plug it into an outlet and plug the cord into it so you can see how much power the element is drawing. Leave the power on for a half hour or so, keeping close watch on the tank. You can lift the wood block occasionally and look into the tube, checking the wiring for any melting and smelling around for anything burning. Be careful when moving the cap from the top of the tube to look inside as hot air will rise from the tube.
If all goes well (no melting wires or the smell of anything burning), plug the cord into a timer. Set the timer to heat the element for roughly 6 hours, turning it off just before you generally use your hot water. You want the element to be turned off while you are using hot water so that you will not be wasting energy heating up the cold water that is entering the tank. If you find that the water is not hot enough after 6 hours of heating, then adjust the timer to start heating earlier. Likewise, if you find that the water is too hot after 6 hours of heating, then adjust the timer to start heating later. The system can take a day or two to reach equilibrium after you make an adjustment, so try not to make adjustments too large or too frequently. If you used a 520W element like I did, you’ll probably find that a few hours of heating per shower is plenty. For example, 4 hours seems to be about right for our two-shower-per day household in the summer. We’ve found that heating the tank only once in the early morning results in “high heat” for showers tapering off to “medium heat” by the evening. If you want high heat available all the time, then you can set your timer to heat a couple hours at a time at a few different times during the day, however, this will consume more energy since the average continuous temperature of the tank will be higher. You’ll likely need to adjust the timer a few times during the year to adjust for the temperature of your incoming water which will vary seasonally.
Lastly, you should post a warning or disconnect the gas lines completely. If someone were to turn the gas back on and light the pilot flame without noticing that you modified the tank (unlikely but possible) it could result in gas and/or carbon monoxide being released into your home (also unlikely but possible). The warning message shown in the image below was one I posted for a few days while testing to see if the 520W element was adequate. After that I disconnected the gas lines from the tank completely.
There are a few important safety considerations when performing the described conversion. First, if the timer beaks or if someone were to plug the element in without a timer, it will run continuously and eventually, it could heat the water tank to boiling temperatures which could cause scalding and in the worst case, could increase the tank pressure to the point of explosion. There SHOULD be a T&P (temperature and pressure) relief valve on the tank which will open, expelling hot water in the event that the temperature or pressure exceeds acceptable levels.
Warning! Whether you convert your tank or not, you should ensure your T&P (temperature and pressure) relief valve is functioning properly.
On my tank, this will expel water safely into a pipe that leads outside my home. Your tank may vary. It is possible, however unlikely, that water could be sprayed onto the element which could cause an electrical hazard. The tank could also leak internally causing an electrical hazard. Also, it is possible that the bare terminals of the element could contact the inside of the water tank, also causing an electrical hazard.
Warning! To reduce the electrical hazards described above, use a GFCI outlet or GFCI adapter. You should also add a note/label to your electrical cord indicating that it should only be plugged into a GFCI outlet to reduce the electrical hazard.
Thermal fuses are relatively cheap and can be wired in series with the element to reduce the risk of overheating. Cover the thermal fuse with heat shrink to prevent any danger of shorting out the leads and attach it adjacent to the outside of the hot water tank (under the existing tank insulation). Thermal fuses can be purchased from Digikey for about $1.30. Aim for about 80°C maximum working temperature of the fuse. There will be some internal heating due to the electrical current through the fuse, so it should trigger at 70 to 75°C. This type of fuse is non-resettable. Once you exceed the cut-off temperature the fuse will be permanently open circuited and will need to be replaced. If you’d prefer a resettable solution, you can use a thermal switch (also called a thermostat or temperature regulator). These open at a particular trigger temperature and close again at a temperature around 10°C below the trigger temperature. Thermal switches usually have a flat surface which is intended to contact the item whose temperature you wish to regulate (in this case, the exterior surface of the hot water tank). They are about $7 ea at Digikey. Whichever method you choose, cover exposed wire or terminals with heat shrink to reduce the danger of short circuits as much as possible. Also make sure you don’t exceed the current rating. The thermal fuse and switch linked above are rated for a maximum of 10A.
Warning! Install a thermal fuse or thermal switch to prevent overheating if the element is inadvertently left on continuously or plugged into an outlet with no timer.
In the absolute worst case that you don’t install a thermal fuse or thermal switch and your T&P valve isn’t functioning (or someone capped it because it was dripping) and you connect the heating element to continuous power (or your timer is faulty), you might be surprised to know what might happen. The temperature and pressure in the tank may rise until the tank explodes. Depending on the mode of failure of the tank, it can act like a rocket, literally destroying your home and launching itself several hundred feet into the air. This has been demonstrated on a couple episodes of Mythbusters. This myth, unfortunately, is entirely true.
Huh? Don’t you need to install a thermostat?
When I first thought about converting my tank, I intended to add a thermostat so that the tank would not overheat and would be at a consistent temperature each morning. However, during testing I found it wasn’t necessary because our hot water use follows such a regular pattern. Without a thermostat, the temperature of your hot water on any given morning may vary a bit depending how much hot water was used the previous day, but that is easily compensated for by adjusting the amount of cold water in the mix. In the worst case that no hot water is used for several days, the average tank temperature will increase higher than normal, but the change isn’t significant. There simply isn’t enough power going into the tank to overheat it (unless your timer breaks, or someone inadvertently plugs it in without a timer as discussed in the safety concerns section above). So my advice is not to complicate things. If you can restrict your hot water use to a single time of day and your hot water use follows a reasonably regular pattern then you don’t need a thermostat. If your hot water use does not follow a regular pattern, it’s entirely possible to make your own thermostat. For an installation at a friends house I converted a purchased programmable thermostat originally designed for controlling baseboard heaters and it worked great, offering complete programmability of both time and temperature without need of a separate timer. Unfortunately, converting a programmable thermostat for hot water tank use is beyond the scope of this article. I may add it as a separate article in the future.
What’s the best water temperature to aim for?
There is a hot debate (pun intended) about what is an appropriate temperature for domestic hot water heaters. Some recommend water temperatures under 50°C to prevent scalding (especially important if you have children). Others recommend a temperature of 60°C or higher to kill bacteria. The primary bacterial concern is Legionella bacteria which can cause Legionnaire’s Disease. The Centers for Disease Control and Prevention estimates 8,000 to 18,000 Americans contract the disease annually (about 1 out of every 20,000 people) with 5 to 30% of cases being fatal. High risk groups are the elderly, smokers, the immuno-compromised and those with chronic respiratory illnesses. According to this article, it’s not necessary that water temperature be maintained at 60°C to kill the bacteria. It’s only necessary that the water temperature be raised to 60°C at least once per day. A great way to accomplish this with minimal energy consumption is – you guessed it – by running a tank on a timer just as I’ve described above. Commercial electric tanks have heating elements located near their midsections leaving some water at the bottom of the tank that will be cooler than 60°C even when the tank is set to 60°C. Thus Legionella bacteria are almost always found in electric tanks rather than gas or oil ones. The converted hot water tank I’ve described, however, should not suffer from this deficiency since the heating element can be lowered to the very bottom of the tank. I personally aim for a temperature of only 45°C at the point of use for morning showers. It’s up to you whether you want to run your tank up to 60°C or reduce your energy consumption and scalding risk by targeting a lower temperature.
Ads by Google
How much can you save?
My gas statements from last summer showed that we consumed about 1.2 GJ of gas per month in the summer and about 2 GJ per month in the winter. That is a summer equivalent of about 11 kWh per day. This summer the converted tank is consuming about 0.52 kW x 4 hours = 2.08 kWh per day or about an 80% reduction. Our gas consumption for one full year was about 20 GJ or 5500 kWh. At $12.78 per GJ, that amounts to about $250 per year. Assuming the 80% reduction holds true year round, I expect to consume about 1100 kWh of electricity per year. At $0.07 per kWh that amounts to about $77. Therefore our savings are about $175 per year.
Aside: Our recent switch from gas to electric space heating already saved us about $355 per year. The hot water tank conversion has brought our savings up to about $530 per year. The icing on the cake is that we no longer use ANY natural gas so we were able to cancel our gas account, saving an additional $140 per year in “basic charges” that our gas company bills no matter how much gas we consume. That brings our total savings to around $670 per year by switching from gas to electric.
How does this compare to tankless “on demand” systems?
In theory, electric tankless water heaters approach 100% efficiency. In other words they put most of the energy into the water and dissipate hardly any to the surroundings. In practice they still dissipate heat from the pipes themselves and a preheating tank is typically installed in series to reduce the peak electrical load required to raise the water to the final temperature. The preheating tank is maintained at a constant temperature and dissipates heat just like any other hot water tank, reducing the efficiency of tankless systems. Despite the fact that tankless gas water heaters don’t require a preheating tank, they are not as efficient as electric systems due to our old friend, the chimney.
For the sake of argument let’s compare our converted tank on a timer to a 100% efficient “ideal” water heater. The specific heat of water is 4.186 kJ/l/°C. Our cold water temperature is about 15°C and our hot water temperature is about 45°C. We use about 40 litres of hot water per day for showers (yes, I measured it). Assuming 100% efficiency, the energy required to heat 40 litres of water by 30°C is simply 4.186*40*30 = 5023.2 kJ. The average power required in kW is simply that energy divided by the time (one day) in seconds or 5023.2/(24*60*60) = 0.058 kW. So a 100% efficient water heating system would require 58 W of continuous power to meet our hot water demand. Using our electric system on a timer our average power consumption (calculated earlier) is about 87 W. Thus our efficiency is 58/87 = 67%.
I think that’s pretty good. It is probably about equivalent to an electric on demand system with a preheating tank. To satisfy my curiosity I’ll also calculate the efficiency of the previous natural gas setup. Recall that prior to converting our tank our hot water tank was consuming about 1.2 GJ (1200000 kJ) per month in natural gas. The power is simply that energy divided by the time (one month) in seconds or 1200000/(30*24*60*60) = 0.463 kW. That’s 463 W. So our water heating efficiency prior to converting from gas to electric was 58/463 = 12.5%. This is typical of a natural gas water heater with an open flue and a continuous pilot flame (the most common type currently in residential use), and yes, it is pathetic.
Note that the efficiency of a hot water tank increases as hot water usage increases. This may not seem intuitive at first, but think about it. If you don’t use any hot water, your hot water tank efficiency must be 0% since it’s consuming energy but not producing any useful hot water. The efficiency numbers I calculated above are based on our own hot water usage of about 40 litres per day. If we used more hot water than that, the efficiency calculated would be higher. Most hot water tanks have an Energy Factor or EF rating. This value is supposed to represent the efficiency of the hot water tank assuming “typical” hot water usage. I’m not sure what the typical usage is that manufactures assume to determine their EF values but I suspect it’s in the ballpark of 2 full tanks per day. Given that gas hot water tanks tend to have EF ratings around 0.6 and electric hot water tanks tend to have EF ratings around 0.9, it’s pretty clear that our hot water usage of 40 litres per day must be MUCH less than the “typical” usage that EF values are based on. I suppose it is in the manufacturer’s best interest to assume greater usage than what is truly typical since it makes their products appear more efficient than they actually are. There are not many consumers who recognize there is a difference between “Energy Factor” and “Efficiency”, or that that actual efficiency depends on usage.
What about adding solar water heating?
Solar water heating is one of the easiest and least costly ways of extracting useful power from sunlight and has one of the shortest pay back periods of any solar technology. Solar water heating was commonplace in cities throughout North America (something like 20-30% of households had them) before cheap gas and electricity arrived on the scene. Now that gas and electricity are not so cheap anymore, there has been a resurgence of solar water heater installations. Luckily, an electric hot water tank on a timer is a perfect complement to a solar thermal collector. Heat only flows from hot to cold, so if your solar thermal collector is only heating up to 40°C because it’s an overcast day, you won’t be able to transfer any of that heat to a hot water tank that’s being maintained at 45°C. To get around this issue, solar water heating systems typically require installation of a preheating tank. The solar collector is used to heat the preheating tank which then delivers warm water to the regular hot water tank.
One way to avoid the expense and hassle of installing a preheating tank is – you guessed it – by using an electric tank on a timer. Then, instead of a preheating tank you have a “preheating period” which is simply the time that the heating element is turned off. If you shower in the morning, your water tank will be relatively cool during the day and a solar collector can be used to heat the tank directly. This will in turn reduce the amount of electrical energy required each morning to bring the tank up to temperature. You can imagine that if solar energy can bring the temperature of the tank up to 40°C, then you won’t need to power the heating element for very long the next morning to bring the tank up to 45°C. However, a thermostat would be a good idea to take best advantage of solar heating. Without one, the temperature of the water each morning will depend heavily on the amount of sunshine the previous day.
Ads by Google
What else can you do to reduce the energy consumption of a hot water tank?
There are several other things I’m planning to do as priorities and time permit:
- I could achieve a higher efficiency by installing a 1000W or 1500W heating element and reducing the heating time accordingly. Effectively this would allow the tank to spend more time at a lower temperature which would result in less heat dissipation. A simple way to do this would be to put two 520W heating elements in parallel to make a 1040W heating element. This would also reduce the recovery period (useful when we have guests and many people want to shower around the same time)
- My tank still dissipates heat to it’s surroundings whenever the inside water temperature is higher than the outside air temperature. It’s a law of thermodynamics that the power dissipated is proportional to the temperature difference so the power loss is greater in the winter when the air is colder (recall my tank is in an unheated garage). There’s no way to stop the dissipation of heat completely, but I could reduce it significantly by adding more insulation around the outside of the tank. There are hot water tank jackets marketed expressly for this purpose but they only add a couple inches of insulation. According to my calculations, the optimum insulation thickness is MUCH more than that. Update. I did this. It was easy and it was wildly successful. For more information see Super insulate your hot water tank.
- I mentioned earlier that our hot water tank is located in an unheated garage. That means that the heat dissipated by the tank is wasted. Therefore I plan to move the tank inside our home. There’s a small storage closet under our stairs that would hide it nicely. Then during our heating season (which is about 7 months out of the year) the heat dissipated from the tank will not be wasted. It will heat our home just as effectively as the electric space heaters we’re already using.
- When I turn on the hot water in our shower, about 1.75 litres of cold water is expelled from the shower head before any hot water arrives. This is because the hot water tank is currently far from the point of use. At first glance you might think that this is just a waste of water rather than a waste of energy – it’s only cold water after all – but what it really means is that after any hot water use, 1.75 litres of hot water remains in our pipes and simply dissipates heat to the surroundings. Our showers each morning only total about 40 litres of water so the 1.75 litres in the pipe represents about 4% of our hot water usage. During our heating season it’s not that important since the water pipe dissipates heat into our home where we want it, but in the summer the dissipated heat is simply wasted. Things could be improved by moving the tank closer to our point of use. Note that insulating your hot water pipes (as is a common recommendation) actually does very little good. Even an insulated hot water pipe will easily dissipate all its heat to the surroundings in a matter of hours because its surface area to volume ratio is so high. Fortunately the planned new location under our stairs will reduce the distance to the point of use to about 25% of the current distance. Another solution is to use smaller diameter pipe as long as you have plenty of water pressure.
- We typically use only 40-50 litres of hot water per day, but our tank holds over 150 litres. Hot water tanks are generally sized to meet peak demand rather than typical demand which makes sense since who wants to have visitors suffer through a cold shower. However, larger tanks dissipate more heat because they have a larger surface area. One solution to this issue is to install two tanks in series, one sized to meet typical demand, and another sized so that the total volume of the two tanks will meet peak demand. One tank would be heated on a timer to provide for normal hot water use. The other could be turned on manually when needed (ex when you have house guests). I may consider installing another, much smaller tank inside the home as described above, but keeping the current big tank in series. Perhaps about one week out of the year we might have house guest and need to heat up the big tank for the extra capacity. Since our garage is not insulated, I would have to heat the big tank in the winter but only enough to prevent freezing.
118 comments on “Convert your gas hot water tank to electric”
On the subject of reducing total amount of water heated by replacing oyur standard tank with two smaller tanks, it might be fairly feasible to install essentially something similiar to the little bob that hands in a toilet. Adding a manual switch of course that can be turned off (immersing the bob and taking no notice of its position) to simply stop the fill of the tank at a certain point. Though considerations of water to air loss of heat and perhaps the type of heater, and whether or not the heater can act if there is air or equivalent vacuum inside it. However, assuming no problems with the preceding a simple bob connected to the fill aparatus may solve the extra water to heat to problem straight away on the cheap and in theory reduce your current expenditures by a logical half or thereabouts in normal use scenarios.
That’s an interesting idea, but I don’t know how it could be implemented practically, and I have my doubts whether it would be effective. The power consumption of a hot water tank is related more to it’s surface area and insulation thickness than to the volume of water inside. A tank full of air maintained at normal operating temperature would likely require almost as much power as the same tank full of water maintained at normal operating temperature.
Also, a hot water tank is a closed, pressurized system with only a water inlet and a water outlet. If you want to switch from “full tank mode” to “half tank mode” using a system as you described, you’d need a way of injecting air into the tank and ensuring that it stayed there rather than simply being expelled through the outlet pipe. There is also a danger of explosion since air is compressible and will expand violently if the tank ruptures, unlike water which will simply leak.
I appreciate the comment, though.
I am interested to try this conversion however i would like to verify a few things if you don’t mind first i have no idea if the tube of the tank where you put the element inside is full of water or not because i’m just curious if it is full of water and the element will submerge on it (with the bare end) would it be creating a short cicuit? Thanks! and i appreciate your response and many thanks for all informative and very helpful articles you published and may the Lord bless you and your family.
The tube does not have any water in it. It is essentially a hole through the entire tank, open to the air at both the top and the bottom. The heating element is not in contact with anything except air. Some parts of the element might lightly touch the tube walls, but the surface of the heating element is not electrically conductive so that doesn’t matter. The element heats up the air in the tube. The heat is transfered from the air to the tank. You may think that is no as efficient compared to an element immersed directly in the water, but it actually has no effect on efficiency whatsoever. Since I plugged and insulated the top of the tube, the heat in the air has no where to go but into the tank. The only difference is that an element in air must heat to a higher temperature to transfer the same amount of heat as an immersion element. As long as the element can take the heat (and stove elements can) it is not a problem.
I’ve just scanned this, but you have covered a lot of topics. Very interesting.
Have you considered adding (building) a heat recovery unit for your sewage/drain line? A great deal of the heat disposed of in the shower waste water could be recovered to pre-heat cold water entering the tank.
I have seen a number of copper models, but am considering simply wrapping PEX tightly around 8 feet of my waste water line to preheat the cold water entering my tank.
I’ve looked at heat exchangers like you described but I determined there was not a big enough payback to justify the effort/expense for me, given how little hot water I use. I have many other projects to consider with a greater return (in both $ and reduced carbon footprint) and while it’s true that every little bit counts, I try to be very conscious of the fact that everything I spend time on has an opportunity cost.
PS. During the heating season (8 months of the year where I live), you can achieve much the same effect simply by leaving hot water in the tub to cool down to room temperature (heating up your home in the process) before you drain it.
Very interesting project. Hypothetically, what do you think would happen if it were left on for a few days straight with no water usage and no timer. How hot could the water get? Where will the overheat valve spay the water? Will the water create a electrocution hazard? Also, what temperatures are reached on the piece of wood? What if someone tries a 1200 watt element? The idea of using any flammable parts for the cap or the insulation really makes me cringe.
I really really think you should advise people to use both a GFCI power source and all non-flammable parts.
Great leading questions. The water reaches about 25Â°C above ambient with the 550W element running 4-5 hours per day, I would therefore conclude that the water could easily reach boiling temperatures if the element were left on continuously with no timer. With a 1200W element the danger would be even worse.
Therefore a thermal shutoff is in order.
On my tank the thermal expansion valve releases water into a pipe that goes into a drain pipe that empties the water outside the house. There is no danger of water reaching the element or any electrical wiring.
I never measured the temperature of the block of wood but using my well calibrated thumb (I know I can touch 50Â°C indefinitely and 75Â°C for about 1 second before it’s too hot) I would estimate the temperature was around 50Â°C during operation (about 5 to 10Â°C hotter than the water). Recall the element is about 3ft below the cap. However, the wood was a temporary improvisation and has been replaced with a sheet metal cap.
Agreed on the GFCI power source. I will update the article to add these suggestions. Thanks again.
I did a similar conversion to my water heater many years ago when i forgot to pay my gas bill.
I placed a 75watt lightbulb at the bottom of the water heater where the gasFlame is. I turned
off the gas valve just for safety, even though gas was OFF anyway. I also plugged the chimney
of the gaswaterheater like you did. I discovered the same thing concerning RECOVERY time.
It took about 20 hours for the tank to re-heat for showers. ONE THING I also discovered is
the INSULATION on the wires tended to BAKE… and eventually harden and crack and this could
be a hazard of your system also. ESPECIALLY since your wire is at the TOP! I suggest putting the
wire at the BOTTOM of the tank because heat rises. All that heat is going to RISE and overheat
your wires and cause hardening and cracking and baking of your wire! Your idea of adding the
TIMER and a bigger heater is GENIOUS ! I had to add many blankets of insulation around the
tank to allow the 75watts to heat sufficiently for 20 hours recoverytime. Any wire used should
be HIGH TEMPERATURE insulated, like those found on an old CLOTHES IRON. and should be
inspected regularly to be sure the insulation has not deteriorated causing safety hazard. A Note taped
to the water heater is important stateing “this is a modified water heater…etc etc ” full
description of what was done. … to inform next homeowner to REMOVE chimney blocks
if they RE-install a gas heater. Lots of safety hazards. Build at YOUR OWN RISK.
Can you convert it back to gas fairly easily. Or do you have to put in a new hot water heater.
Conversion back to gas takes 5-10 minutes. You just pull out the heating element, re-insert the original spiral shaped metal piece in the tube, re-attach the chimney, and re-light the pilot flame. Then figure out a way to deal with the guilt, knowing you’re consuming 5 times more energy than necessary. 😉
i have an electric water heater , not gas.
will i save something if i connect 110volts to the lower heating element?
I know im not rob .. but i think i can answer your question jappy, if your still around
Converting a typical 220V hot water heater, would save you money, yes .. but remeber electric water heaters dont have an opening at the top (ie chimney) .. and doesn’t have a opening at the bottom that you could use for a similar use
In addition, your already heating with electric :S ..
Where you would save money is if you installed a timer, to turn on the water heater only at the times you require it .. but a 220V timer is fairly expensive,
so jappy the easy answer is no .. your stuck with what you have .. correct me if im wrong ..
Yes. With an electric tank you’re already saving a lot over most gas hot water tanks with open flues and continuous pilot lights (the most common kind for residential use). But if you switch a 220V tank to 110V you won’t gain anything unless you also install a timer and/or insulate the tank better.
The only benefit to switching from 220V to 110V is that it will then be much cheaper to install a timer. The drawback is that you can’t get as much power out of a typical 110V circuit, so the recovery time (time to heat up an entire tank of cold water) is longer, but this is only important if you use more than a full tank of hot water each day.
We have a solar system for DHW (no standby) and on the third or fourth rare cloudy day the showers get shorter or skipped. It is a nice conservation feedback system.
hello would this work with a propane water tank?
I’m not familiar with propane hot water tanks but I assume they would be virtually identical to natural gas ones with the exception of the burner itself. If so, this technique should work just as well.
Try this, itâ€™s easier. After my shower one morning I turned the control on my gas water heater from On to Pilot. This keeps the main burner from running. The next day about an hour before my shower I turned the control back to On – But the burner didnâ€™t light?
Iâ€™m thinking now that my gas control is messed up, it made no sense, the burner hasnâ€™t run for over 24 hours so I know the tank has to be cold.
But it wasnâ€™t, the water was at full temp and the thermostat was keeping the burner off.
So I turned the control back to the pilot position and left it there. Every day I have a nice hot shower with water that is being heated only by the pilot flame!!
Yes, that is easier, but you’re still wsting a lot of heat up the chimney. Before converting my tank to electric I ran it for about a year on just the pilot light and I measured heat loss of around 40% up the chimney. See the section “Why are electric water tanks more efficient than gas ones?” in the article above for more details. If your gas hot water tank is located inside your building envelope, the draft generated by a warm hot water tank also draws warm air from within your home up the chimney resulting in significant heat loss from your home (you probably lose more heat from your home than you put into the tank). To reduce these losses, I tried restricting the air flow through my tank (running on pilot only you don’t need much air flow to remove the exhaust). I posted an article about that project but enough readers expressed concerns about carbon monoxide poisoning that I took it down (a ridiculous concern really since a single pilot flame is unlikely to produce enough carbon monoxide to do any harm, even if exhausted directly into your home). One issue I did find when I restricted the air flow was that the exhaust condensed on the inside walls of the hot water tank causing rust. I assume this would have resulted in premature tank failure, but I don’t know for certain since I stopped doing that and converted the tank to electric instead (better to get rid of the chimney entirely).
Rob: You’re my hero today, this website is fabulous, thank you very much.
You said: “I posted an article about that project but enough readers expressed concerns about carbon monoxide poisoning that I took it down (a ridiculous concern really since a single pilot flame is unlikely to produce enough carbon monoxide to do any harm, even if exhausted directly into your home).”
You are correct. Consider that old gas ranges and stoves had a full time pilot light.
In any case, if you have a gas hot water heater you should have a carbon monoxide detector nearby.
Thanks very much for this very informative article.
I bought a house in December which has natual gas heating and hotwater heater. Here in the desert of Arizona, very little heating is required, even during the coldest days of winter. Just the two of us, we never use a full tank of hot water each day and the gas furnace seldom comes on, but I’m paying the minimum $12.00 p/month gas bill. Plus, we spend approximately 4 months p/year away from the home, but we still have to pay the minimum gas payment.
For the last 3.5 months I’ve pondered how to convert the gas hotwater heater to electric. Now you have solved my delima.
The question I have is: would a common electric water heater element work just as well as the stove top element pictured in your article?
Thanks again, and keep up the great work.
Unfortunately electric water heater elements are designed for immersion. If they are used in air, they will almost certainly overheat (ie self destruct) since they do not have enough surface area to transfer the heat in the absence of water. Water has 24 times higher thermal conductivity than air so it does not require much contact area to transfer a lot of heat. A stove element on the other hand is designed to be operated in air. Even that might overheat in the confined space inside a gas water heater, if not for the fact that I’m operating at only 1/4 of its rated power output (due to operating on 110V AC instead of 220V AC). Good luck with your conversion.
Very good article! . Just to complement your article in how to increase efficiency, you can buy second hand electric kettle 1500w for $10, they are 110v.
Why wouldn’t you simply install the heating element where the burner is? Further how would you go about setting a thermostat? I wonder because i rent my home and I would like to convert without unduly damaging or stressing the existing system
You could install the heating element in place of the burner, but it is not an easy job to remove the existing burner and replace it with an electric one. By comparison it is quite an easy job to bend an element to the desired shape and drop it down the center tube, leaving the existing burner in place. It also means you can easily switch back to gas if you decide to. Since writing this article I have installed a programmable thermostat ordinarily used for controlling electric baseboard heaters. Inside the thermostat there is a thermistor (a temperature sensitive resistor) used for measuring the air temperature. Usually the resistance is 10k at 25 degrees C. To get the thermostat to operate in the desired range for a hot water tank, this thermistor needs to be replaced with a different thermistor that has the same resistance (10k in my example) at about 55 degrees C instead of 25 degrees C. The replacement thermistor also needs to be mounted in contact with the hot water tank (under the insulation) and connected to the programmable thermostat by wires. Thermistors are available from Digikey (www.digikey.com) for under $1. You use the programmable thermostat just like you would for a baseboard heater, except the real temperature will be 30 degrees C higher than what the device indicates (ie if you want 55 degree C water, program the thermostat for 25 degrees C). When I have time I’ll write an article showing how to do everything I described above (so many projects, so little time).
Thank You Rob,
I tried this and it worked GREAT! I did use the stove wire inside the unit and a heavier duty extension cord wire for the outside. The gas company can keep their ever increasing convenience charges. I will be using your method. I Will Try your electric heating method this winter. A smaller footprint and less $ for the monopolies. Thank You Again, Clint
P.S. Looking forward to the update! Total cost of conversion O$ Old extension cord and free burner and wire from recycle yard.
Can I hire a local electrician to do a job like this?
I like your idea.
I doubt it. Most electricians would be capable of doing the job, but they might be unwilling to accept the liability. It would be a very unconventional request.
Great article! Well, I just finished shaping my coil and wiring up my device only to find out that there is a “ridge” at the top of the tube which holds the length of metal in the tube. This “ridge” significantly decrease the diameter of the hole making it very difficult to reshape the coil for it to fit. This is a problem for me. Furthermore, the length of metal hanging from this ridge seems to be permanently attached to it and I don’t think there is a way to remove it. Have you experienced this and know of a remedy? Much thanks — jack
I haven’t encountered that before. I have converted a few tanks and in all cases the “ridge” was a removable ring. I guess one alternative is to insert the element at the bottom but I assume that is probably equally inaccessible. Sorry I can’t help more. If you do find a solution, please post it.
Thanks Rob–I was actually able to remove the ring with a crow bar and some gentle force. I dropped the coil inside and plugged it in. I waited a few moments then noticed a little smoke then unplugged it and pulled it out. I smelled a little burning. Any idea what this could be? The wires are high heat quality and everything is taped up beautifully. Nothing appeared burned when I pulled out the coil.
Upon further testing the coil outside the tank, I did smell burning again. I used solid copper 600 volt insulated wire to connect at the terminals and wrapped with layered electrical tape. I then connected the copper wires to an extension cord. Did I do something wrong that could cause this burning? Thank you again!
If you’re not tripping any breakers, then it’s unlikely there is a short circuit so you could just plug it in for a while in a well ventilated area (place it on a metal tray or hang it from something) and see what starts smoking. That will be the quickest way to find out what’s wrong.
Hi Rob…update to my project. The burning was likely caused by the electrical tape I used. I went ahead and inserted the unit in my tank and it’s been working now perfectly for 5 hours! I turned on my water and presto! It’s hot!! Rob, I can’t thank you enough for taking the time to share this creative genius with me and the world. Kudos to you!
Great article on super insulating water heaters. And a very interesting conversion of a gas water heater to low wattage electric. I recently did a conversion of my 240 volt/3000 watt 60 gallon water heater to 120 volt/750 watt. I’ll post some pictures and explanations at flatteningthecurve.
Waterheatertimer.org is a great resource on everything to do with water heaters.
All the best,
Thank you for all your valuable information and responses. If only people got together this way more often to address common challenges…
What could be other alternative “Heating Elements” that may be already available in our homes just in case. Thanks again and more power to you all.
I have done another conversion using an oven element. In this case rather than bending it to a spring-like coil, it’s easier to make a long element. The wattage is a bit higher (around 750W) for slightly faster recovery time. There is also a suitable element in most dishwashers, though I haven’t tried one. Note that on average, the power consumption of the heating element in my hot water tank is under 100W, so you could even use a single 100W light bulb as a heating element. The recovery time would be almost 1 full day though, so you would have to constrain your hot water use to once per day and live with luke-warm water most of the time.
Would it be a problem if most of the coil of the heating element touched the tube? Should the coil be “barely” touching or not at all in order for it to heat the air better? I noticed the coil I made was slightly bigger in diameter that it was touching the tube.
Thanks for your help.
Hi Robert, I don’t think it matters too much to the heat transfer whether it contacts the inner walls of the tank or not. However, over time it may wear away the wall or the element as it rubs during thermal expansion/contraction. I don’t know for certain if that will happen within the lifetime of the tank, but it is probably best if the element can slide freely inside the tube. It should not matter if it lightly contacts the wall (ie you don’t need to ensure there is an air gap all the way around).
Reading on this site has been an absolute delight, I can’t tell you how much of a breath of fresh crisp mountain air it has been. I’m giddy with excitement to try out some of your implementations as I may soon be a homeowner. It’s late so I can’t outline it now but I’ve got a wild solar hot water idea I’ll post soon and in the right project comments. My reason for writing right now is to call into question numbers 4 and 5 under “What else can you do to reduce the energy consumption of a hot water tank?” I am hesitant to say you are contradicting yourself because up until now you’ve been so deadly accurate with everything else I’ve read on your site.
I’ve already told at least 5 people about this site and have been raving only mildly so as not to completely scare them off with my enthusiasm.
Thanks for your vote of confidence. I looked over points 4 and 5, but I don’t see the contradiction. Can you elaborate?
4. “Even an insulated hot water pipe will easily dissipate all its heat to the surroundings in a matter of hours because its surface area to volume ratio is so high. ”
5. “However, larger tanks dissipate more heat because they have a larger surface area.”
Two smaller tanks would have a greater surface area to volume.
You are correct. Two small tanks would dissipate more heat than one large tank, all other factors being equal… ie same temperature, same insulation thickness, same total volume. But in point 5, the temperatures are not equal. I’m suggesting heating just one of the small tanks. The other would only be turned on during times of peak demand (ex when you have house guests). I edited the article to make that clearer. Thanks for the feedback.
Your conversion is awesome and is exactly what I’ve been looking for.
Is it possible to use an induction cooker element, for greater efficiency?
I don’t have any experience with induction elements to know whether they would work or not, but I can say with certainty that they would not be any more efficient. Any energy released from any kind of element inside the tank has no where to go but into the tank. Power in = power out. Therefore, the efficiency of conversion of electrical power to heat is 100% regardless of the type of element. Improving the heat transfer (ie reducing the thermal resistance) between the element and the water will not affect the efficiency. It will only affect the operating temperature of the element. For example, immersion elements such as those used in commercial electric hot water heaters have great heat transfer since they are in direct contact with the water. However, they are no more efficient than the stove element I used. For a given power input, the stove element will just run hotter.
Thanks for your feedback. If I understand this correctly, the element can be any shape, and goes to the bottom of the tube, the high temperature wire goes through the cover at the top, then connects to the extension cord?
There is no outlet near the WH cabinet, and on other forum they said a breaker should be within a short distance of the WH. Is that right? The breaker box here is outside, on the other side of the house, which seems sufficient to me. But it looks like I’ll need to put in a new outlet next to the cabinet. Should it be inside the cabinet, or would inside the adjacent cabinet be better (other side of the wall, same room, easy access). Do you think it should have a gfci? I was going to put those everywhere, then realized they all drain phantom electricity, so I’m not as in favor of them any more. I have also been wondering about putting the coil in through the bottom.
I’m going to use the element as you have suggested, but still wonder about induction, as a 500 watt induction coil can melt a lump of metal within 1 or 2 minutes, heating much hotter than an electric stove element. Perhaps a 140 watt induction coil would be just right for a WH, and use less electricity than an electric stove element. Also the coil would be cool inside of the tube, and only heat the metal that is holding the water, therefore not have any wearing effect on the tube.
Well I just realized the outlet has to be in the same cabinet to plug in.
Any ideas or suggestions about this are much appreciated. Thanks
thanx for the great idea,my gas was 326 dollars this month and tommorrow they are shutting my gas off,its just to costly,i have my element and wire,can you use the thermostat switch off the stove,to regulate the heat also,i do have a timer too,so after work im converting mine,thanx again for such a great money saving idea, dave
Most stoves do not actually have a thermostat (they don’t measure element temperature). They just turn the element on and off at a fixed duty cycle. The result is that the element puts out relatively constant power at a particular setting. Using a timer to limit the total energy input into the tank per day is safer. Using a modified programmable thermostat (one originally designed for electric baseboard heaters) is even safer and that is what I do now but it is a bit more work since you have to wire in a different thermistor to the thermostat. One day I will post how to do it.
Thanks for a fantastic and informative article.
My gas line is rusting out and needs replacing now, but it will cost many thousands.
And I hate those gas connection fees (not usage)
I really want to convert to all electric.
We have 240V AC (Australia).
Hopefully the element won’t run too hot.
I was thinking, could you use two elements in series, with the connection between the two elements at the bottom of the tank ? (not sure how to connect and insulate though).
Then it would be a lot easier to bend the element into a one way spiral, and screw one into the other.
I would prefer to have higher resistance and lower power.
Any thoughts on this ?
Power = voltage*current and with twice the resistance you will get half the current and thus half the power output. So yes, having two elements in series is a good way to reduce the power output, but as you say, it may be a challenge to connect the two elements and insulate the connection. Good luck, and be careful.
OK thanks Rob.
Just lately I have been able to get some better elements that I don’t need to join.
Your earlier tip about finding elements in dishwashers was a good one, since I had a useless old dishwasher in the shed.
Also got some oven elements, which may be good since less bending is required to straighten them.
(I hope to use them straight with just one bend at one end)
But I have also been able to scrounge a small electric water tank heater.
I will plumb this in a different location, and bulk insulate as you have suggested on this website.
Then I will have a emergency backup tank.
I guess a proper immersion heater would be slightly more efficient, than having the heating element in the air (in the centre column of a gas tank).
Actually, a “proper immersion heater” would not be any more efficient. Whether you heat the air column or heat the water directly, all the energy you put into the element is transferred into the water. It has nowhere else to go. All that changes is the thermal resistance between the element and the water. That means that for a given power input, say 100W, an element in the air column will run hotter than an element in the water. But in either case, putting 100W of electrical power into the element will put 100W of heat into the water so the efficiency is exactly the same.
The main advantage of immersion heating elements is that they can be much smaller to transfer the same power without overheating and melting.
Hi Rob, what about using a toaster oven element, along with it’s temperature control? Seems like that could give you temperature regulation fairly easily by reusing the toaster’s thermostat, and the temperature knob.
That should work, but depending on the toaster oven it may not actually have a thermostat. Some simply turn the element on and off at a variable duty cycle which would have essentially the same effect as plugging the element into a timer and running for a fixed amount of time each day. The best solution is to use a programmable thermostat designed for baseboard heaters. They are about $40 and you have to replace the thermistor to get to the right temperature range for hot water. I did this but haven’t posted the instructions yet… so many projects… so little time.
im a single mom…in my 30’S and ive been wondering how to convert my gas water heater to electric…i googled and came across your article..very smart..would you do this for me if i pay you?pls email me.thanks.
loll… i just thought about a fish tank heater..will it work??
Brilliant!!! A question though. Instead of unwinding the heating coil of the top element, how about using the oven element and bending it so that its length rests against the flue pipe? Any thoughts?
I have used oven elements to convert other hot water tanks. They work great and are a little easier to work with, bending them into a single long oval shape. I do not recommend putting the element in contact with the walls of the flue pipe though. Repeated thermal expansion and contraction of the element may cause rubbing which, over a period of years, may wear away the element (resulting in a short circuit) or the wall (resulting in a water leak). I recommend suspending the element in mid air inside the pipe. It may be non-intuitive, but it is important to recognize that there is no loss of efficiency by doing this. It takes exactly 1kWh of electrical energy to put 1kWh of heat into the tank, regardless of whether that heat goes directly from element to wall, or from element to air to wall. With either method, there is simply nowhere else for the heat to go but into the tank. The only difference is that an element suspended in air will get hotter due to the higher thermal resistance of the heating path.
Very interesting article however I wonder about one point. I believe the cost of electricity was calculated at 7 cents per killwatt hr. NOTE:- I live in Ontario near Niagara falls. The utility company says the cost of electricity is less than ten cents, how ever the true cost ( Total bill divided by number of hours consumed ) is above TWENTY CENTS. This FACT makes all advertized usage cost inacurate. In my opinion this is as unethical as allowing retailers to advertize a product for $100.00 knowing the customer must pay $115.00 to take it out of the store. One comment ref an aquarium heater, yes it could be made to work however if you consider the necessary modifications it is impractical.
Hi Don. My cost estimate of 7 cents was based on my total cost divided by my total kWh consumed. My cost has gone up to just under 10 cents/kWh since I wrote the article. Note that I heat my home with electricity too. If I only heated my water with electricity the cost per kWh would easily be double since the basic monthly fee (irrespective of consumption) would make up more than half my bill.
I submitted this before but for some reaon it didn’t appear on your site.?; I’ve used a porcelain light socket with 200W clear bulb and was wondering if you think it would be as efficient as the elements? It sits on top of the flue hole so would heat from the top down no doubt, but if it was, say, 12 hours on and 12 off, in a 33 gallon (imp) tank, would that compare with your calculations for a 750W element for example? (perhaps the link to pictures negated my sumbission so I will refrain from doing that this time.) Thanks for reply, your site was a true inspiration and much appreciated.
For a given tank, it is only time at temperature that determines waste heat loss. So whether one size element is more efficient than another really depends how you are controlling the element. If you are using a timer and heating up your tank once per day and then consuming exactly one tank of hot water replacing it with cold water, then a more powerful element will be much more efficient simply because the tank will be spending less time at temperature, thus dissipating less heat to the surroundings. For example suppose with a 200W element over 12 hours you put 1200 Wh into the water and 1200 Wh of waste heat into the surroundings. That would be 50% efficient. With a 400W element you would still put 1200 Wh into the water but only 600Wh into the surroundings since it would be hot for only 6 hours. That would be 66% efficient. An 800W element would put 1200 Wh into the water and only 300 Wh into the surroundings since it would be hot for only 3 hours. That would be 80% efficient… etc on up to 10kW or so where you essentially have a hot water on demand system.
However, in practice people rarely use their hot water tanks that way. If instead you are targeting a constant temperature setpoint and keeping your tank hot throughout the day, then a lower wattage element will be more efficient since that will result in less time spent at high temperatures.
When thinking about hot water tank efficiency, just think about “time at temperature”. The less time spent at high temperatures, the less heat you are wasting to the surroundings.
My apology for unnecessarilly submitting more than one comment/question. Thank you for your (typically) courteous reply. With a meager 150W bulb, it doesn’t seem very efficient, but then again not much energy can be wasted since I don’t use 33 (imp) gallons of hot water per day. That said I don’t know how many gallons/litres a 5-10 minute shower might use with a low-volume shower head.
But so far, many thanks to you, I’ve enjoyed just that immensely, (having curtailed what I regard as a gas price fixing monopoly long ago). Initially I let the light bulb heat the tank for 24 hours, but now experimenting with timer settings of 4.5 hours at equal intervals i.e. 9 hours total during a 24 hour period to see if that will satisfy my single person needs. Would still like to know if a 200W light bulb, for example, would be more efficient on this sort of schedule, than an oven element etc., but the light bulb is almost shameful symplicity, inspired by you 🙂
You’re very welcome. In terms of efficiency of a 200W light bulb vs a large stove element, think of it this way.
Operating a 200W element for 5 hours, twice a day will consume the same amount of energy (and thus cost the same) as operating a 1000W element for 1 hour twice a day. The difference will be that the 1000W element will result in greater water temperature fluctuation. The water will be hotter by the time the 1000W element is switched off because not as much energy is dissipated from the tank in the 1 hour heating time. However, the water will also be colder by the time the element is switched on since more energy is dissipated from the tank in the 11 hours the element is off.
If you care about the average temperature (ie if you want hot water available at all times), then the 1000W element offers no efficiency advantage over a 200W element. But if you care about peak temperature (ie if you only want hot water at a particular time of day, and don’t mind having cooler water for the rest of the day) then the larger element offers a significant advantage. It can achieve a higher peak temperature for less energy consumption.
Still testing with a timer, but so far a 200W bulb operating for, say, 15 hours per day seems satisfactory for my needs. The bulb is a simple clear type and this method of heating water is so utterly simple I’m ecstatic about it. The water heater is in an unheated utility room, so considering your calculations I will leave it uninsulated, although I have not detected any benefit towards room heating. An optimum 50+ temperature is indicated out of the tap before putting demands on it, but of course the volume at temperature varies depending on useage.
Hi again; would be interested to hear how you would compare the heating energy from a CFL (twist type) bulb rated at “42W = 150W”? Is it actually putting out 150W energy wise, or is that just a suggested comparison in lumines.?
“42W = 150W” simply means the CFL bulb consumes only 42W of electrical power but has the same brightness as an incandescent bulb that consumes 150W of electrical power. The amount of energy that any bulb puts out (light plus heat) is always exactly equal to the amount of electrical energy consumed. That is the law of conservation of energy.
That’s what I though Rob, thanks, I guess there’s ‘no free lunch’. Still testing the 200W long life incandescent clear bulb about 14 hours per day, with minor scheduling adjustments, and it seems to be ideal for my needs. All things considered I can’t see any advantage to using stove elements. Will now consider some of your other thoughtful projects.
At first I was skeptical, because of the air-gap from flue to tank, use of stove/oven components, etc.
But it works better than I could have hoped.
I had a bunch of new oven bake elements so I used one, bent in a long U shape, instead of the stove-burner unit you (originally) used.
I used a Range-Kleen bake element, with the STRAIGHT-blade terminals, but think now that the #7501, with the 90 deg.ANGLE-terminals might have worked better in this application. (But not sure I had one of those, so…)
Then I read the rest of the posts. (directions, we don’t need no stinkin’ directions)
I had also added a 100 watt light bulb,(or two depending on size) down in the burner portion (pilot light area)
150w, or small flood/spot lamps??? I just used the normal 40/60/75/100 lookig bulbs. (what I had)
I’m hoping this will have same effect as pilot light warming during summer months.
I hooked it up last night, gave it about 6 hrs., and water was hot as gas heat (I had the thermostat at like 140 F, before)
My daughter took a 5-10(or more) min. shower this morning (6am), and I just got out at 8(am). (another 5-10 min.)
So, 6 hours initial heat up,then 2 showers in 2 hours. And I’m sure it would have done more/quicker. (what?)
All on 110 volts, with about 20 mins. labor. Basement is about 75F, OUTSIDE temp is about 95F last few days.
But I’d (we,sorry) been taking COLD-water showers (no heat) for the past few days, and this seems to work as good as the gas (natural) did.
I just unplugged the heating element, but left the light bulbs on for now. I don’t have a timer yet.
MIGHT not need one if the PILOT-LIGHT/bulb thing pans out. (’til it gets cold, anyway)
I didn’t have any smoke at start-up (or yet, knock-on-head), but I used the crimp-on terminals, with-out tape.
Mine is in an out-of-the way place, with no chance of anybody, or anyTHING, messing with it.
I’m going to try add a tin-can cover (or SOMEthing) Mine is now still open and I’m loosing a LOT of heat that way.
I just now noticed the name of the web-site, and read the story behind it.
Makes me even more proud and appreciative.
I haven’t got a chance to check out the rest of the site.
Keep up the good work! I’ll let you know of any up-dates at this end.
I WILL TRY,
Brian D Nelson
There’s a guy here (KC) on CL, (not me) that does this for $150
I’m Assuming he does it this way, I didn’t call him.
I’m thinking if using a round dishwasher element (have one) down below instead of the light bulbs.
Any comments, questions, experiences, concerns, etc. ?
I WILL TRY,
Brian D Nelson
OK, Quick(?) update
YES, definitely hotter than the 140F I previously had set on the gas!
I unplugged the heating element about 8:30am yesterday, and about 2 hours later, I went in the kitchen and was running the hot water and almost scalded my hands. Was like, WTF?, then remembered, duh, no more gas.
SO, YES, Please use caution where children and/or elderly are involved.
I removed the gas burner and mounted a dishwasher heating element in its place.
Burner comes right out – 3 nuts (2 small, 1 larger gas lines) Push it back into the tank to get the lines clear
Dishwasher element was harder to bend, though. And it didn’t want to go in at first. I had to bend inside tank.
I don’t see why another oven element or even a stove-top burner (coil) wouldn’t work, if it would fit in the opening. The smaller stove top burner is about the same size and shape as the gas burner I took out.
D-W element seems to work as well, or better, than the oven element in the flue.
Looking into some sort of combination of the two. (series? parallel? 2 individual circuits? Not sure yet)
Definitely need a good 110v timer.
Tin-can cover at top of flue seems to help hold the heat, too. I put the hole for the wiring in the side, instead of having a hole in the top. And blocked the vent pipe to roof.
A little more fine tuning and I think I’ll be ready for insulation (Hot-water heater, then house)
I’m also thinking of a passive solar/radiant pre-heater in the attic. Run the lines down the old vent pipe
Not sure PCVC (or what) lines like a radiator in the rafters, a tank in a fake chimney, a combination of the 2, or what yet. Still thinking. But I’m not above stealing a good idea from ANYbody!
Thanks again!! Use less, waste less, want less.
I WILL TRY,
Brian D Nelson
I just re-read my last post and noticed something.
When I said, “I had to bend inside of tank”, I meant I had to bend the ELEMENT, while I was feeding it into the opening left where burner was removed.
I did not bend, fold, cut, mutilate, etc. the tank itself in any way.
Just to clarify.
I WILL TRY,
Our gas line allegedly began leaking somewhere in the back yard so the gas company shut off the feed. Not having the $4000 to run a new line from the meter to the house we were stuck showering at family’s houses. Then I read your post and showed it to the wife… She was entirely against my idea to implement a “suicide shower” and pretty much against your idea as well. So instead of a stove element I said I would use a light bulb, after all its merely an oversized easy bake oven right? I pulled out the twisted metal section from the water heater flue and dropped in a ceramic bulb holder from a brood heater with a 150 watt heavy duty bulb. 8 hours later we had 40 gallons of reasonably warm water. After 3 family members showered we ran out. 24 hours later hot not just warm water for everyone to have a controlled shower. This was fine until we had house guests so in anticipation I Bought 1.25 gpm shower heads and tried a 300 watt bulb. Holy cow. I know it isn’t the greatest thing ever but to me it was. No more 8 minute showers. On this setup we were able to have 3 showers in the morning and 2 baths, later that afternoon another 2 showers and a bath. Heating the tank from ambient to hot is less then 7 hours.
Here’s my setup: I bought a brood heater with a 600 watt rated ceramic fixture for $11 and 300 watt bulb for about 4$ at the local big box. I unscrewed and removed the metal “lampshade” and ran the end of the cord thru one of those red bricks with the holes in it and tied it in a knot at about 3 feet. So the bulb and heat rated ceramic receptacle drop about 3 feet or so down the flue and cannot go any further because the brick with the holes acts as a stop and the fireproof cap.
The bulbs are about 3″ wide and the flue in my heater is 3.5″ so it’s a nice fit without touching, and since the brick has 3 holes I can look down to makes sure that the bulb isn’t touching. This setup has served me and my family for almost a year now. I have had to buy 2 bulbs during this time but it is a small price to pay.
Sand does not conduct electricity and has 6 to 10 times the thermal conductivity of air.
Thermal Conductivity – k – W/(m.K) Air: 0.024 Dry sand: 0.15 – 0.25
You could streach out the heating element to one just one loop like a long U and fill the inner tube with just enough sand to cover the element. The benifit would be more efficent heat transfer to the water in the lower portion of your tank, and a longer lasting element do to less heat buildup on the element itself. Down side is converting back to gas requires removing the sand.
Hi Keith. Interesting idea. I am confident it would work and would result in a much lower element temperature as you say. However, although element life might be extended, the difficulty of replacing the element if/when it does fail would be increased significantly.
By the way, my original element is still functioning fine after 5 years with no noticeable deterioration so lifetime does not seem to be an issue. I suspect that when running a 220V element on 110V, the temperature is already much lower than the rated temperature of the element.
Just wanted to jump in to say thank you. I converted my house to electric a couple years ago and put in a 4 gallon electric water tank. Worked fine for me but then my daughter returned from college and wanted longer showers. I contemplate buying a larger heater but the remembered my old 30 gallon gas water heater was just sitting there. A quick search turned up this page. I am so glad. My sister just got a new set of kitchen appliances and gave me her old stove. I decided to use the bottom stove element. Heated it up a bit and bent it straight with a double loop at the bottom. Wired it up and sealed the connections with high temp RTV wrapped in fiberglass. Inserted it in almost to the bottom and stuffed some more fiberglass pulled from the oven to seal the top. Amazing. My daughter can shower as long as she wants and recovery time is only a little longer than when it was gas. About 2 to 3 hours for all 30 gallons. After running it for several weeks on a timer to make sure everything was okay I have just ordered a $17 temperature controller on Amazon. I plan to attach the sensor to the outside of the inner tank. That way if I sell my house I can convert it back to gas quickly and easily. No real changes to the existing heater. I can’t thank you enough. The main thing I would emphasize to others is make certain the electrical connections are well insulated with high temp rtv or there is a high temp fiberglass tape available on amazon for about $12. I was going to get it but my cheapness took over. But I didn’t skimp on safety and tested every step of the way including ohm out the element after bending and checking to make sure the inner wire didn’t short out to outside metal. And of course a GFCI is always a good idea.
I meant to say I used the bottom oven element. Not stove element.
Hey Rob, First off, thanks for this, i’ve been thinking about trying something like this since the winter when our oil fired furnace/hw heat died, and the subsequent propane stand alone “free” used HW heater also decided to give us multiple issues…
(family of 5 taking stove-top boiled “bucket baths” hasn’t been much fun…)
Coincidently, We were also was “lucky” enough to have our electric oven decide it was “its time” a few mod ago … (been one of the years..)
this post was just the thing i needed to give me the courage to give it a go (I have a Tesla style mind with a Jed Clampet education…i generally end up figuring these types of things out so long as the ADD doesn’t kick in, lol…
Anyhow, I’m about halfway through it , just had a few questions ….
1. would hooking a transformer to the element to increase the voltage gain anything?
2. the interior of the chute is very sooty, we got as much out as we could, but Im worried it may be very smoky, i was thinking of lining it would a length of Aluminium Dryer vent (formed around a pipe) and sticking the element through that (i went with a long U since i ended up using the biggest element (2500 watts btw)
and finally , i saw a neat video a kid posted on you tube where he made a free space heater out of an old coffee can, some bolts, and copper springs (i believe he shorted out the 2 car batteries he had wired in series to it) it was charged by solar i believe) could this possibly be viable here? or is nichrome different?
thanks in advance for your (or anyone else’s) reply – Jonny
What thrill to see article like this. I learned the rocket stove stuff from Paul Wheaton at Permies.com and richsoil.com. I also learned to insulate the refigerators and it is making a difference too.
I do have one question though- I am off grid solar, and I plan to combine my solar water panels with the house for hot water. But I also have a cool possibility with my Outback controller- it has an auxiliary port that looks like a telephone port. And we can run this to a heating element to dissipate any extra power that may be coming in. We have 2000 watts of solar panels currently and this is a possibility. I already bought 12 VDC elements in preparation for this. Instead of placing the element down the flu, is it possible I can drill into the side of the tank and place the element in like a regular electric? It has threads ready to go to do this.
We also want to use these elements to heat the livestock troughs in the winter with their own solar panels.
I wouldn’t advise mounting an element by drilling through the side of a pressure vessel like a hot water tank. On electric hot water tanks, the wall is reinforced where the elements mount to it. If you simply drill through the wall of a gas hot water tank you may cause stress concentrations that could cause the tank to rupture. Don’t underestimate the forces experienced by the walls of a hot water tank. If your water pressure is 100 psi, and your tank is 20 inches in diameter by 40 inches high, the walls will experience a tensile load exceeding 40,000 lbs. In the case of a water trough, it’s no problem because it is not pressurized. Worst case is you may have a leak, but the trough isn’t going to explode ;-).
I’m wondering, I literally just plugged mine in, how long does it take you to heat water from room temp? And also, I put my element in from the bottom, like a couple mentioned, but heat doesn’t seem to get to the top of the tube unless I take the cap off, is there a need for a small hole to allow air flow or should it be sealed? Because my cap is staying cool to the touch and it’s been about an hour
First of all, thanks for this great article!
I have successfully prepared everything and am a little anxious about a few of the safety issues/potential issues before I proceed.
First of all, when I first lowered my modified range element into the center space, the coil barely touching the inside wall of the cylinder as described, it began to smoke when I plugged it in for a test run. The smoke eventually dissipated so I believe it was just the buildup inside that space from the gas operation days.
But I want to make sure this is normal. Again, it doesn’t seem to be smoking after about 2 to 4 minutes, but the heater is indoors enclosed in a built enclosure by the landlord, in the kitchen. That enclosure has a vent at the top side, and of course there is the metal chimney pipe for the gas heater.
So I’m thinking it would be a good idea to leave that chimney functional, allowing any strange fumes to escape, but also so the enclosure where the water heater is located doesn’t get too hot! So instead of capping it, seems like allowing it breathe would be smarter? Thoughts?
Also, how critical is it to install a thermal fuse? If I have a timer that works well, is this really a major concern? Where exactly does this go?
Thanks in advance for your thoughts on all this!
It’s normal for the element to smoke a little bit or smell a little bit for the first several minutes of use, much like you might experience when plugging in a new stove element on a stove for the first time. The surfaces in nearest to the element may not be perfectly clean and and dirt/grime/etc will take a few minutes to burn off. The problem should not persist or recur after the first 2-3 minutes.
If you don’t cap the chimney, you’ll have an extremely inefficient waterheater. Much of the heat from the element will simply heat up the air which will rise and be exhausted through the chimney without transferring any heat to the tank.
The thermal fuse is kind of like a seat belt. You probably won’t ever need it, but if you do you’ll be glad you put it on. Using a timer with no thermostat or thermal fuse is like having a “boil” setting on your hot water tank. An unsuspecting user could always set the timer to run continuously which will eventually bring the water in the tank to boil.
The thermal fuse goes somewhere in series with the element. It is just like a switch. When the temperature gets too high, the switch opens. When the temperature drops, the switch closes again.
What’s the best way to bend the element? Or if I go the 200 watt bulb direction what kind of cord should I use
The element can be bent by hand with a little effort. If you have something round and solid to wrap it around (ex a rolling pin, a piece of PVC or ABS pipe, a coffee can, etc). I used a piece of 3″ diameter PVC tubing for the outside coil and a piece of steel pipe about 1.5″ in diameter for the inside coil. It definitely helps to have a shop vice. Good luck.
Hi Rob loved the article has really came in handy as my gas was shut off due to a gas line break in the area. Used the same method you suggested but instead of using a coil I made a metal cap and drilled a hole in the top and attached a 1500w heat gun. Just turned it on about an hour ago but already have got water in my sinks going to let it sit awhile longer for showers. What do you think of the heat gun idea since it is heating the air a little more evenly because it is forced air? Thanks again!
Thanks again 5yrs now and still going strong!
Regarding using a heat gun, I would be very concerned about safety. Heat guns are not designed for continuous operation and I would think that the lifetime will be short. Also, even if it lasts, it will be less efficient because you are blowing hot air through the center tube rather than heating air within the tube (ie the heat escapes rather than being contained and forced into the water tank).
Using a stove element at the bottom of the tube, natural convection will heat the air in the tube relatively evenly (no forced air required). But that is a moot point since there is no benefit to heating the air in the tube evenly. No matter how unevenly you heat the air, the heat has nowhere to go except into the tank. You could fill the top half of the tube with insulation and it would work just as efficiently (though the element would reach a higher temperature).
Thanks for the update. Glad to hear it’s still working for you. My installation is now 7 years and still going strong on the original element and water tank.
Love the info!!!!
I’m about to install water heater in new construction. Currently using “solar heat” via about 400′ 3/4″ black pipe on Texas roof. Would like to build hybrid system, maybe convection (thermo-siphoning similar to the one you built) but use a water heater as storage tank. I was hoping to use electricity as a backup.
Now, long story short, if you had a clean slate. Would you use a new electric or spend the extra for a new gas unit and convert as you did? If you choose electric, any advise on models to look for? Any advise you have on using 220v heater as 110v.
Look forward to hearing your advise. Thanks in advance..
Good questions. If I were building new, I would go with a new electric and run it on 220V. There is an advantage to being able to heat up the water quickly. The disadvantages are that timers for 220V are more expensive and you may not have 220V available close by if you are converting a gas hot water tank (not a problem for new construction). I would install an after-market 220V programmable thermostat/timer. I have not seen these designed for hot water heaters (though I haven’t looked recently), but they are common for electric base-board heaters (do a search for “programmable thermostat baseboard”. They are usually around $100 and I have seen suitable ones for as low as $40. There’s a thermistor (temperature dependent resistor) in these thermostats that can be replaced with a different thermistor to shift the measurement temperature range up 20 degrees C which you will probably need to do since the default temperature range likely only goes up to 35 to 40 degrees C. After the conversion, you just need to remember that when the thermostat says 30 degrees C, it really means 50 degrees C. Sorry I can’t give more details on the conversion… I’m consider doing a separate post about it.
I would also super insulate the tank as I have already described in another post.
I believe the elements for an electric water heater can be threaded into the one of the bungs(?)fittings common to all water heaters. This would allow the use of an off the shelf item.
i just wanted to say thank you for the great info here! i had been looking for a way to switch my gas heater to electric. i used a ceramic plug from a brooding lamp and a 200 watt bulb, and have HOT water again! first hot shower since summertime!
It’s 120/240VAC, not 110/220VAC. It’s the year 2016 and I still see ye old folk write it this way! I also highly doubt you live in Japan which actually uses 110VAC.
Hi Robert. Point taken. Article updated. Historically 110 V, 115 V, 117 V and 120 V have all been used in North America and it’s quite common for people to refer to all of them nominally as 110. But you are quite correct that the present standard is 120 V and has been for a long time.
Thank you for you great article
I’m currently having a natural gas hot water tank and is not functioning, the lease water tank company requires to open the ceiling to change the venting ABS pipe. After reading your article, I want to change my water tank to electric, i want to know if I use 240V electric hot water tank from the lease water company, how much I would save?
Hello Rob. The idea is interesting but I would like to make some comments please:
1) Air is an insulation medium. Its thermal conductivity at atmospheric pressure and 25 Celsius is just 0.024 W/mK. Water value instead at same conditions is 0.58 W/mK or about 25 times better conductor (plus it provide strong end effective convective currents in its medium that transfers the heat). This means the heating stove element heats the air first, than the metal and the water in a slow and inefficient heat transfer process… therefore, there will be significant heat losses from the tank to the exterior if it is not very well insulated (so your plugging feature at the end of the chimney tube has to be 100% sealed and of very good insulating thermal properties). Your surface area is also reduced compared to a gas heater that also heats the bottom of the tank. Also even the heat losses to the copper wiring (excellent conductor) that holds the stove resistance will be a source of heat losses to the exterior (it is also subject to higher oxidation associated with the heat). So keeping the heat generated by the resistance inside the chimney space without incurring significant heat losses to the exterior is critical. Have you quantified these losses? So I would assume that this heater has significant more heat losses to the exterior than a convectional electrical water heater tank… (even at the bottom of the tank once it is heated it is not directly insulated and eventually heat will “escape” from the bottom that has no insulation)…
2) The other question is the lifetime of the air stove resistance. In normal conditions it operates very hot but in your configuration it will operate at even hotter temperatures since it will be completely surrounded by hot air. I would imagine that its lifetime will be reduced…
I can see the benefits of this idea as a band-aid solution as a temporary arrangement. I attribute this as the main merit of it. Can you comment about the topics I mentioned above?
I’d be happy to comment on the topics you mentioned.
1. The air definitely creates additional thermal resistance between the heating element and the water tank. It’s less than you might expect since the heat is transfered primarily by convection rather than conduction, but that’s a moot point because additional thermal resistance has no affect on the speed of heating. The speed of heating depends only on the power. It doesn’t matter through what resistance the power is transfered to reach the water. The effect of the additional resistance is just that the element will reach a higher temperature than it would if it were immersed directly in the water (T_element = T_water + Power*Resistance). I agree it makes sense to seal and insulate the bottom of the tank and the plug at the top of the tube. I’ve gone a step further than that and insulated the whole tank with several inches of rock wool. The copper wires are an insignificant source of heat loss compared to the copper inlet and outlet pipes. These pipes are in direct thermal contact with the water tank and have a much greater cross sectional area than the wires. Inlet and outlet pipes are a source of heat loss regardless of what kind of tank you have, gas or electric.
Without adding external insulation a gas hot water tank converted as described would be less efficient than a standard electric hot water tank, but still much more efficient than the original gas hot water tank. After sealing the top and bottom and adding several inches of insulation around the tank I doubt there is much difference in efficiency between a converted gas hot water tank and a standard electric tank with the same additional insulation. But there is a big difference in cost if you’re starting with a pre-installed gas hot water tank. I was able to convert my existing tank including adding the extra insulation for much less than the cost of a new electric tank, let alone the cost of installing a new tank or wiring a new 240V circuit to provide the power. For new installations it definitely makes more sense to use a standard electric hot water tank and add extra insulation.
2. The electric element in this conversion actually runs much cooler than its design temperature since it is operating at half its design voltage (240V element running on 120V). That means it operates at 1/4 of the power it was designed for. It doesn’t even get red hot like it would on a stove. It experiences more frequent heating and cooling in the water heater than it would as a stove element though, so it’s not immediately obvious how the lifetime should be affected. I was not sure if this would turn out to be just a band-aid solution when I installed it, but it has been operating flawlessly for 8 years now which is permanent enough for me. I inspected it last year while checking on my sacrificial anode. The element and wires looked just like the day I installed them.
Not to sound to harsh but you have just created one scary tank. No thermostat?? So the fist time you get stuck away from home for a few days the tanks overheats, boils and the T&P valve spews if your lucky. If unlucky it explodes!. You used wood as a cap the potential for the wires getting hot and catching the wood on fire is another concern. My gas water heater costs me about $6 a month to run (one half an MCF per month) gas usage in the summer with a gas cook stove. An electric one would cost me about $40 a month. Maybe your gas is much higher or electric way cheaper than mine.
I’m not sure if you read the article. All your points are addressed. There is no thermostat but there is a thermal fuse in series with the element to protect against overheating. I believe I emphasized the importance of that with sufficient warnings about the potential for explosion if the tank overheats and the T&P valve fails. The wood cap was a photo from a quick proof of concept. I included a clear warning in the article to “not use a wooden cap as shown in the photos”.
I also did a cost analysis in the article comparing my measured gas usage before and my measured electrical consumption after. Prices have changed since I did this project (8 years ago) but where I live they still favour the electric conversion.
The reason your pipes should be insulated is not so much that the water will remain hot in them until the next usage – that as you note would not be particularly effective – but so that the hot water is at a higher temperature at the shower head, for example, and so you mix in more cold water – thus decreasing your hot water usage
Why not just keep the temperature controller dial from the original electric stove such as from a small electric cooktops? It should be easy to retain it, preset it to 120F to 160F and lock it in place so as to avoid fire hazard rather than putting in without that kind of governance, right?
We have 200 litre gas HWS system which is placed outside the house. In our place the temperature in winter rarely drops below 10°C. The power voltage in the grid is 240 V. I also have an 1350 W almost new oven element in the garage. We are just now getting 3.3 kW solar power. Would you please advise whether the 1350W element is too big or should I try to get one with lower Watts? How long would it take to heat 200 litres of water to 45°C? How long will the water stay warm? Say if the solar power turns off at 5PM, would the water be still warm for morning showers for two people? Otherwise we do not really need much warm water.
Also would it be OK to use a heat gun replacement element instead of an oven element? It would fit nicely into the heat exchanger tube without any bending.
And one more question. Where would the best place be for the thermostat sensor? I assume that it should control the temperature of the water, not of the air in the tube. Maybe the top of the tank?
hey rob , i put a 500watt charcoal starter where the burner would be an dtested it after three hours only acouple seconds of hot water .. any thoughts on what i could do with what i have ? … maybe let it sit longer
This was a great read. Hope you come back thanks Rob. It’s funny how it changed through the years some people don’t read. And want to contradict things that you so clearly you covered. I think it’s this new age of of trolls. Any way erythromycin helpful. I’m going to try a light bulb first. ??
I was talked into getting an on demand gas hot water heater with electric fan last year, but now in BC they are talking about everyone using electricity instead of natural gas within 12 years. Can I get it converted to electricity only? (It cost about $5000.00 all together with tax, etc. already) I should have just gotten a regular hot water tank, but he made this type sound so good, plus I got a rebate from Fortis BC of $500.00. Hmmm? I think I was railroaded, but it’s too late now.
67% efficiency for an electric water heater is terrible. Look up the ratings for real electric heaters. They are all over 90% because the elements are actually in the water. Your electric-air-metal-water chain is also super slow. A super-heated pocket of air in the middle of the tank is not the same as a real electric water heater. I bet you can find a good working used unit in your local classifieds for <$100 any day of the week.
You claim backwards compatible with gas as an advantage. But if it really works out better, why go back? This is just crippling the efficiency. The real advantages of your hacked unit is that you skip the real plumbing work, skip the real electrical work and you can use a cheap 110V timer instead of finding a 240V one. So yes, it's all good for proving the concept, but now it's time for the real implementation.
I think you may be interpreting EF ratings of 0.90, which are common for electric hot water tanks, as meaning that they are 90% efficient. That is not the case. “Energy Factor” or EF rating is not the same as “efficiency”. The 67% I measured is real efficiency which is the amount of energy that reaches the point of use (as hot water) divided by the amount of energy put into the tank (as electricity). This will vary significantly depending on daily water consumption. If you consume no hot water then even a tank with an EF rating of 0.90 will have 0% efficiency. The tank still consumes energy just to maintain temperature, but none of it reaches the point of use.
The 67% efficiency I measured was based on my actual hot water consumption of only about 40 litres per day. Typical household hot water use is around 125 to 250 litres per day. If I were consuming water at that rate, my efficiency would be much higher.
If hot water usage is less than “typical” you will get much lower efficiency than the EF rating. If you have a hot water tank with EF rating of 0.90 I suggest you take the measurements and do the math outlined in the article to determine your real efficiency. I think you’ll find your efficiency is much less than 90% unless you are consuming a full tank of hot water per day or more.
EF ratings are really not that useful except as a basis for making relative comparisons between similar sized tanks. Much more useful would be if manufacturers specified the thermal conductance of a tank (in Watts per degree C). From that spec, one could determine the baseline power consumption and cost that the tank will have just to maintain a certain temperature above ambient, assuming no water consumption.
As I’ve mentioned in response to other comments, the element being in air vs water has no effect whatsoever on efficiency or rate of heating. The heat has nowhere else to go but into the water, and the rate of heating is related only to the electrical power (in Watts) put into the element. The additional thermal resistance of the air pocket simply means the element will heat up to a higher temperature than it would if it were immersed directly in the water.
You said “If it really works out better, why go back?”. I agree. The issue for most people is they may not have the confidence it will really work without trying. Isn’t it an advantage to be able to do so without making any irreversible modifications?
My proof of concept is still working well 10 years later and I’m very happy with it. I’m confident there’s no net benefit to switching to a “real implementation”.
4 yrs ago I inherited a house and a crack was found on the gas line that was in the backyard so It was my responsibility to have it repaired, no way I could afford the $2800 cost so I’ve been heating water on the stove…
For the past 4 years which is a new level of suck. I found your article and was able to build this and very inexpensively.
Friday night was the first time in over 4 years I had a hot shower in my home with hot water that was NOT heated on the stove.
This might not be a big deal to some
But it was huge to me. You helped me out more than you know. Thank you
I’m really trying to do this….But I cant bend the burner into shape. I was using all my strength, it barely moves. maybe the one I got was a little thicker than what was used here…maybe I would have better luck with the oven element…anyone have any suggestions?
I simply used a 1500 watt heat gun blowing down from the top. It takes two hours to get to perfect 130 degrees.
Wouldn’t the heat gun finally burn out? They are not really built for extended use.. plus you are losing some because the air is actually blowing out the bottom where the burner is still hot. What did you use to check the temperature?
A lot of people are commenting on the air gap being an inefficient transfer medium. I haven’t seen a gas water tank tube before, so I may be out to lunch, but if the bottom of the tube is plugged why not use a regular hot water tank heater and fill the tube with oil?
I’ll repeat an excerpt from my last comment above:
As I’ve mentioned in response to other comments, the element being in air vs water has no effect whatsoever on efficiency or rate of heating. The heat has nowhere else to go but into the water, and the rate of heating is related only to the electrical power (in Watts) put into the element. The additional thermal resistance of the air pocket simply means the element will heat up to a higher temperature than it would if it were immersed directly in the water.
Why mess with the extra challenge (and risk of leaks) of filling the tube with oil or any other fluid when it will have no benefit whatsoever? The only possible benefit would be increased element lifetime since the element would operate at a lower temperature. But even in air it is already operating at a much lower temperature than it was designed for as it’s operating at 1/4 the design power. I’ve had no issues going on 12 years now operating on the same element.