Connecting solar panels to hot water heater

oil pan 4

> @mike95490 said:
> oil pan 4 said:
>
>
> How about a Hot water heat pump?
>
> Use half the power for the same amount of hot water.
>
>
>
>
> Those units are starting to exhibit field failures and  they are non-repairable.

I have an add on heat pump unit that can be installed on any electric water heater. Any hvac shop can repair it.
Those package unit hybrid heat pump hot water heaters are junk, especially the GE units.

NANOcontrol

It is absolutely necessary to have some form of energy storage. In a simplistic explanation a panel can produce a maximum of 10A. A heating element is sized to match that at power point. The best sun of the day the switching circuit is fully on and all that current goes to the panel.  In lesser sun, the panel can produce 5A and that is still at the same power point voltage.  A capacitor bank is connected to the panel.  When the switching device turns on it produces 5A from the panel and 5A from the capacitor bank. The heating element still sees 10A, but for a shorter time period. The same holds true when the panel only produces 1A. The 1A goes into the capacitor for a long time.  Then the switcher turns on for an even shorter time. 9A from the capacitor bank and 1A from the panels. There are variations in execution of the method, but they are all basically the same.  Anything else is just not efficient. Just like batteries, the capacitor bank has to be able to handle the current or it overheats and eventually fails.

FET have transition heating. Doing it 100 times a second produces very little heat. 50,000 times a second adds up. The trade off is lower switching requires a larger capacitor bank and this is really advantageous for neophytes building these systems.

With enough panels water can be heated with a dead chicken between them and a resistive element. It won't be efficient. Nothing I've heard so far is even remotely efficient.

jonr

Here is a better link for a device that works efficiently.

http://techluck.com/

NANOcontrol

But, you can't parallel it with an existing charge controller.

MarkP

My main concern is how to avoid burning up contacts with 100vdc.  First I have bought SSRs that should allow me to control the raw array power to the water heater using the 12vdc 750ma max AUX signal from the charge controller.  In addition I will rout the power directly from the SSRs to the elements, bypassing the thermostats.  I will put the thermostats in the control loop so they will only have to switch a few ma at 12vdc.  They should be able to do that just fine.

mike95490

When you are using a SSR, be sure you follow the recommendations for applying Protection (snubbing) the  circuit the SSR is in.  Even with simple restive heating elements, there can be high amp/volt surge that can fry the SSR.   And be sure of your sources, many SSR's have been cloned with cheap counterfeits that don't last a couple weeks .

jonr

> But, you can't parallel it with an existing charge controller.

I have no knowledge either way, but what makes you say that?   I can think of various ways that two MPPT controllers could work well together in a primary/secondary mode.


> many SSR's have been cloned with cheap counterfeits 

I second this. I think most on ebay are fake.

NANOcontrol

There is a manual reset over temp control usually on the upper element only.  I highly recommend that stay in the heater element circuit to disconnect all power in the event of a problem.  The lower control can be replaced with this type shoul you reach maximum current capacity of this reset. You never know when a SSRY will short out. A mechanical contact should last a couple times. You can tell us if the SSRY work with this unknown duty cycle and frequency.

MarkP

Oooog...  I ordered the SSRs on eBay.  Supposed to be two Crydom D4D12Ls.

bill von novak

NANOcontrol said:

It is absolutely necessary to have some form of energy storage. 

Not really  All you need to do is match the impedance.  There are lots of ways to do that; an MPPT controller, a dedicated controller (like the Sunflux) or more 'creative' methods.  You don't need to store energy (other than in the hot water heater tank, of course.)

NANOcontrol

You said exactly the same thing I did, a MPPT controller stores energy in capacitors.  Some have discussed using diversion outputs with a solid state switch.  That is better than just not using available energy, but it is not efficient without capacitors to store the energy. The PV hot water systems I build are much like the Sunflux.  I like how they say up to 96% efficient.  Only the Shadow knows. If you find someone more creative than me, let me know.

BB.

Technically, a typical (switch mode "buck type" power supply) stores the energy in inductors (current flow sort of like momentum).

The capacitors are just there for "filtering" noise (if needed).

https://en.wikipedia.org/wiki/Switched-mode_power_supply

You can build a buck mode converter without capacitors--You cannot build one without an inductor.

There are power supply designs that use capacitors for intermediate energy storage.

There are various Capacitor Based power supply circuits... Manly used for voltage doubling (such as a typical 120/240 VAC input to DC power supply--Voltage doubling for 120 VAC input, just rectification for 240 VAC input):

https://en.wikipedia.org/wiki/Voltage_doubler#Delon_circuit

-Bill

mike95490

NANOcontrol said:

You said exactly the same thing I did, a MPPT controller stores energy in capacitors. ....

No, they use inductors.

But I'm reading apples and oranges.   Storing energy as hot water vs storing in batteries ?

bill von novak

NANOcontrol said:

You said exactly the same thing I did, a MPPT controller stores energy in capacitors.  Some have discussed using diversion outputs with a solid state switch.  That is better than just not using available energy, but it is not efficient without capacitors to store the energy. The PV hot water systems I build are much like the Sunflux.  I like how they say up to 96% efficient.  Only the Shadow knows. If you find someone more creative than me, let me know.

Again, no.

You don't need capacitors at all in an MPPT converter; a continuous mode converter uses capacitors only for filtering and EMI suppression.  You could easily do a specific continuous-mode MPPT converter for this application with no capacitors whatsoever.

Why don't you need storage?  Because what you need is an impedance converter, not energy storage.  (The water provides that.)  You could do it with relays and 2 heating elements, wired to give you 4 different impedances.  But that's only 4 steps, rather than the infinite number of impedance steps you can get with a continuous mode converter.

At the end of the day, all you need to do for a PV hot water heater is to deliver current to an element.  Impedance matching gives you the maximum possible power (V*I) at all solar fluxes.

Dave Angelini

All of these  (PV)  hot water systems over the decades never seem to live up to their promise in an offgrid home with 3 or 4 people living there. They get into trouble with shorting and burning things up in winter. Often they end up buying a propane tank heater or a on-demand heater because they need the PV for batteries.

A simple domestic hot water collector, a small pump, timer, and a propane tank heater will run for decades if drained during the cold weather. In really cold climates a more complex drain back will also work. For offgrid  reliability, simplicity, and multiple sources are really nice concepts to live by. Less is more

jonr

While it will certainly work, you don't have to use a buck converter (with an expensive inductor) to modulate power supplied to a heating element and to hold the panel at Vmp.  PWM + input capacitance can suffice.  

If you don't use input capacitance in either design, the panel will not be held at Vmp (the voltage will rapidly vary above and (probably) below Vmp as switching occurs .  It won't be a MPPT design.

PS: I get paid to design DC-DC converters.

MarkP

I can see solar PV hot water not providing enough hot water but electric resistance water heating is as simple and trouble free as it gets.  The problem is when people forget that it is truly an opportunity load and don't plan for back-up heating.  Running out of hot water mid-shower is likely to make you think "This doesn't work" but it should never be expected to work as the base load for heating water.

NANOcontrol

I played around with a few cheap grid tie inverters. They have a habit of blowing the input capacitors.  You should see the full wave power line ripple on the panel power. It is about 25%.  Most people are lucky and don't even know the capacitors have blown from high ripple currents. Knowing the input power and the output power doesn't always give the real efficiency in solar even though it does with line operated devices.

bill von novak

MarkP said:

I can see solar PV hot water not providing enough hot water but electric resistance water heating is as simple and trouble free as it gets.  The problem is when people forget that it is truly an opportunity load and don't plan for back-up heating.  Running out of hot water mid-shower is likely to make you think "This doesn't work" but it should never be expected to work as the base load for heating water.

If it's really going to be a way to use opportunity power, then having a separate preheat tank (sized for a morning's hot water usage) can work well.  That way the "main" water heater is still there, using whatever energy source it usually uses.

NANOcontrol

Expecting solar by any method to supply all hot water isn't very cost effective. Sure, there are some places where a barrel painted black on the roof will work. That will not be an option for most people. Supplemental PV heating is a very exciting idea. I bought 520W of panels for $240.  That is something that can be done any day by anyone. That few of panels can easily find a place on any home and running a wire is nothing. Putting that power into a water heater is far more cost effective than a heat pump. As supplemental, 100% of the potential power will go into heating. The tank will be at near temperature and it doesn't take much to  raise the water temp significantly. This reduces grid consumption and higher temperatures get you thru  a lot of the night. Direct solar is DOA for most homeowners except in rare cases.

MarkP

I still plan to work a solar thermal panel into the system if I can find one cheap but will not attempt to install "enough" solar thermal.  If I did then there would be many times that the system was sitting there cooking itself.  An undersized thermal panel will make the water warm while operating at modest temps where it is most efficient.  The PV will take it the rest of the way up to usable temperatures, will be very controllable, and losses from the tank at the higher temps are much more controllable than from a flat plate collector.  Everything will be doing the job it does best..  I bought the water heater for 25 cents on the dollar so it would have been crazy not to go solar PV to hot water using power would have been thrown away.

bill von novak

NANOcontrol said:

Expecting solar by any method to supply all hot water isn't very cost effective. Sure, there are some places where a barrel painted black on the roof will work. That will not be an option for most people. Supplemental PV heating is a very exciting idea. I bought 520W of panels for $240.  That is something that can be done any day by anyone. That few of panels can easily find a place on any home and running a wire is nothing. Putting that power into a water heater is far more cost effective than a heat pump. As supplemental, 100% of the potential power will go into heating. 

Just as a note, yes, it's a neat idea.  But:

1) You can't skimp on mounting or wiring just because it's not grid tied.  You still need to safely mount the panels and run the wiring.  A wiring fault can start a fire on a water heating system as easily as it can on a grid tied system; it must be wired properly to be safe, and you need the knowledge and experience to do it correctly.  It cannot be done "any day by anyone."

2) You also need a thermal cutout (i.e. a thermostat) for safety.  You cannot rely on the PT valve, or the low power of the solar input, to protect you against boilover and scalding.

3) You need a means of impedance matching if you want "100% of the potential power to go into heating."  There are a few off the shelf solutions for this, but they aren't cheap.

Dave Angelini

I would add it is more important offgrid to exceed safety/code as there will be few to help you.

 It always happens at night in winter when few will be around to help you