Solar water heater

[Andy999]

I’m wanting to put a couple of solar panels on to outbuilding roof and use output to heat immersion heater for domestic hot water ( I have boiler backup)
I don’t want a grid tied system - completely standalone
I want to keep the existing 3kW element as I don’t think it will come out of the cylinder!
Don’t want batteries- just want to feed the output via inverter direct to immersion
Is there an inverter that will do this?
Anyone done it?

 

[efficientPV]

What size tank and is it just one element. Your heating element is 20 ohms. That takes a lot of voltage to get any significant wattage. If you drive a heating element with half voltage, you get 1/4 rated watts. Four grid tie panels in a 120V array would give you about 700W.

 

[5275K_Dude]

My friend just asked me the same question. Solar panels for hot water heater element only. Not the first time it’s been done. Easy way is solar heating the bottom water heater element, utility on the top unless you do a transfer switch to shift it completely one way or another.

One other consideration is to add a tempering valve and set the water heater to say 140 or even higher. Even a single panel running all day at 400 watts can heat up a lot of water. And the tempering valve can nearly double your useable water.

Depends on number of individuals, how much you use, etc, but the tempering valve really makes the difference for me. I have a closed loop solar collector and turn off my water heater at the circuit breaker in March (Phoenix area) and back on around late October. Or when my wife “reminds” me she’s running low on hot water. My water heater regularly hits 150 and with that tempering valve, it’s nearly free (12V circulation pump that I haven’t switched to the solar circuit yet)

Let us know what you find out!

 

[robby]

Keep in mind that the thermostat in a normal water heater is not designed to work with DC.

 

[Daddy Tanuki]

Keep in mind that the thermostat in a normal water heater is not designed to work with DC.

true but their are several videos on how to wire a capacitor to remove the inrush spark to make them work. sorry I am too lazy to get the videos from youtube for the OP. but they are out there.

 

[robby]

true but their are several videos on how to wire a capacitor to remove the inrush spark to make them work. sorry I am too lazy to get the videos from youtube for the OP. but they are out there.

I can see how a capacitor can solve the problem on inrush when the contacts close but I see know way they can help when the contacts open with DC still on them and an Arc is formed.

 

[robby]

Anybody thinking of doing this just remember there are only two things that Prevent a Water Heater from exploding. The Thermostat and the Emergency pressure valve. I have actually seen an old hot water heater with a bad valve, so be real careful or this can happen.

 

[BartHenry]

disconnect all the wires from the bottom element ( most water heaters have 2). replace it with a DC rated element.
a quick google search and you will find what your looking for.

 

[efficientPV]

true but their are several videos on how to wire a capacitor to remove the inrush spark to make them work. sorry I am too lazy to get the videos from youtube for the OP. but they are out there.

Those people don't know very much. There are ways to protect the element if case the capacitor shorts and to allow full current when the contacts open, then limit current when they close again preventing contact burning.

 

[robby]

disconnect all the wires from the bottom element ( most water heaters have 2). replace it with a DC rated element.
a quick google search and you will find what your looking for.

Are you saying that ONE DC rated element will prevent an explosion? Because it will not.
You need some kind of Thermostat if the Heater is completely sealed. Heating water in a sealed container is a Bomb, you have to have the means to prevent it from going critical. Pressure relief valves do that but they are a last resort and they can possibly fail. The thermostat is the main control mechanism and HVDC will weld the contacts shuts on an AC thermostat.

 

[BartHenry]

Yes, you need a thermostat on the DC rated element, weather it is built in or not..

 

[Vigo]

I can see how a capacitor can solve the problem on inrush when the contacts close but I see know way they can help when the contacts open with DC still on them and an Arc is formed.

When i read this i immediately thought 'sure it would help!' but then i got ready to write something and realized i.. didn't have an explanation for that.. I was trying to relate it to 'points and condenser' on an old car which is a similar situation, but when i realized neither side was an electromagnetic coil i lost the thread and got all confused.

If you had the capacitor wired across both sides of the contacts, when closed it would have no charge, and when contacts open it would appear as a path of much less resistance in the moment the arc is forming? Given that the solar is a limited current source, i feel like it's possible if you put a large enough capacitor across the contacts it could 'eat' almost the whole current potential of the source during contact opening, dropping voltage down close to nothing in that instant, and it might be enough to stop an arc forming?

I feel like im missing something. Might be full of crap on this one.

 

[efficientPV]

You are correct in that the capacitor acts as a dead short initially. The voltage across the contacts quickly rises as the capacitor charges. These are mechanical contacts, and the hope is the gap will increase enough so an arc will not form. Some metal always vaporizes from resistance heating which also helps the arc formed. Generally, if the voltage remains under 30V the mechanical speed is fast enough to gain distance for an arc to extinguish. When the contacts close again, you will have the capacitor current added to the heater current which may be enough to weld the contacts. A diode could be used in parallel with a resistor. The diode will allow full current to the capacitor when the points open. When they close again, that resistor limits added contact current from the capacitor. Resistor only needs to be 200 ohms. A couple amp time delay fuse in series with the capacitor will insure the fuse blows should the capacitor short. Never use electrolytic capacitors.

 

[robby]

You are correct in that the capacitor acts as a dead short initially. The voltage across the contacts quickly rises as the capacitor charges. These are mechanical contacts, and the hope is the gap will increase enough so an arc will not form. Some metal always vaporizes from resistance heating which also helps the arc formed. Generally, if the voltage remains under 30V the mechanical speed is fast enough to gain distance for an arc to extinguish. When the contacts close again, you will have the capacitor current added to the heater current which may be enough to weld the contacts. A diode could be used in parallel with a resistor. The diode will allow full current to the capacitor when the points open. When they close again, that resistor limits added contact current from the capacitor. Resistor only needs to be 200 ohms. A couple amp time delay fuse in series with the capacitor will insure the fuse blows should the capacitor short. Never use electrolytic capacitors.

So let me get an idea of what you are saying.

I have 300VDC of PV at 10 Amps going to a thermal switch and then into a 3000W element.
At some point in time the contacts in the switch are going to open with that full 3000W of power on them.
How exactly is this circuit going to prevent a massive Arc?

 

[efficientPV]

I would never do it. That never stopped anyone on the internet from trying it. Doing things the right way is always cheaper and easier.
Solutions that work at 60V do not scale up.

 

[robby]

I would never do it. That never stopped anyone on the internet from trying it. Doing things the right way is always cheaper and easier.

Yep I have noticed that people always head for the cheaper way even when it is very dangerous.
I cannot see any method that is not going to be complicated that is going to solve this problem.
Yes you could use a proper HVDC relay and some circuitry to disconnect the power when thermal switch opens but that adds a layer of complexity to something that is designed to be dead simple because failure can mean your house will be blown apart. And yes this has happened and peoples houses have been destroyed.

Another fun one and keep in mind what happens to people who are in the house and don't get hit but are exposed to the super heated steam.
BTW anyone who did not see these episodes. The only thing they did was remove the Over Pressure relief valve on the top of the heater and then they just plugged them in with no thermostat.

 

[Vigo]

Well, I guess the easy way around it would be to switch it out for an actual DC element with a much lower resistance, so that you could wire your panels to a much lower voltage and use cheaper/safer switching and protection devices. Depending on how long your wire runs are it may end up more expensive in pounds of copper wire, but for a lower power setup it will probably be cheaper.

Regarding lower power, my limited experience so far is that 1100w out of one element does a pretty ok job on my water heater. So i would add that there's no law saying you need to get the full 3000-4500w that the existing elements are rated at, to have an acceptable hot water experience.

Also, i wouldn't try to convince anyone that a water heater is LESS dangerous than they think, because usually thinking the opposite makes us act more responsibly, but.. If your tank is basically 100% full of liquid which basically do not compress, it is not really storing 'explosive' energy and if it ruptures it will mostly go PUH and dump all its water out. But if you manage to heat the water past normal boiling temp before it ruptures, then AS the tank ruptures the water will undergo phase change and expand by a huge amount, which is what is creating all the 'thrust' of a flying water heater. ? So basically don't heat past boiling EVER, and your explosion potential even if the tank DID fail, should be pretty dang low. A tank rupturing at 180f would be pretty uneventful unless you were standing right next to it in the wrong kind of shoes.

 

[efficientPV]

When someone says they need 3,000W for an element my first response is how many elephants do you own. I don't get involved with those people. The real problem is if you tell someone they have to do five things, they will just do the three things they want. I get by with 500W.

 

[Daddy Tanuki]

Well, I guess the easy way around it would be to switch it out for an actual DC element with a much lower resistance, so that you could wire your panels to a much lower voltage and use cheaper/safer switching and protection devices. Depending on how long your wire runs are it may end up more expensive in pounds of copper wire, but for a lower power setup it will probably be cheaper.

Regarding lower power, my limited experience so far is that 1100w out of one element does a pretty ok job on my water heater. So i would add that there's no law saying you need to get the full 3000-4500w that the existing elements are rated at, to have an acceptable hot water experience.

Also, i wouldn't try to convince anyone that a water heater is LESS dangerous than they think, because usually thinking the opposite makes us act more responsibly, but.. If your tank is basically 100% full of liquid which basically do not compress, it is not really storing 'explosive' energy and if it ruptures it will mostly go PUH and dump all its water out. But if you manage to heat the water past normal boiling temp before it ruptures, then AS the tank ruptures the water will undergo phase change and expand by a huge amount, which is what is creating all the 'thrust' of a flying water heater. ? So basically don't heat past boiling EVER, and your explosion potential even if the tank DID fail, should be pretty dang low. A tank rupturing at 180f would be pretty uneventful unless you were standing right next to it in the wrong kind of shoes.

You do realize that takes all the “fun” out of it…

 

[charliefleet]

Watching with interest... i'm doing something similar as my rough calculations (based on October 2022's new tarrif of around 34p/kwkr) is a saving of approx. £500 per year just on the hot water cylinder alone.

Proposed system: 3 x 455w panels --> MPPT with no batteries --> inverter --> existing 3kw AC heating element connected in series with a new dedicated AC boiler thermostat. This will typically run at well below the 3kw rating, probably more like 700 - 1200w, depending on time of day/year.

Alternatively, there is an all-in-one solution, something like this: https://www.bimblesolar.com/elwa-PV-water-heater. But it's about £675 including VAT in the UK.

My only concern with the cheap option is running an MPPT with no battery attached... i know the water heater is purely a resistive load, but i would need to research how stable the connected system would be under fluctuating light levels. To some extent, it doesn't matter given it's just a heating element - nothing critical. But i just have a concern that batteries are there for a reason (at least in some MPPT/Inverter systems) - e.g. to give a reference voltage, smooth out rapid light transients, etc. But i really don't need the additional expense of a battery for this application.