Hot Water by Solar Electric direct via MPPT ?

[bcroe]

Tricky to accomplish since morning and evening will always see a drop in current.
So what can I expect to happen if I power a resistive DC load on a non-MPPT CC when the current drops,
assuming I have matched the DC heater resistance (impedance) as you suggest?

Tgriff: As BRoe said, there are some pretty robust power relays but they are AC coil.

I have managed to use a few AC coil relays on DC. Using a power supply the relay can be expected
to pull in at a much lower voltage. I then boost that about 30% and call it the DC coil rating.

Certainly an MPPT control with an electrical on/off would be ideal, almost all require panel Vmp to be
greater than the load voltage at peak power. Without that, careful matching of components could
achieve decent efficiency but an on/off method is needed. Yes when sun intensity or angle change
you move away from MPPT. One point of view is, when power drops below 40% the efficiency does
not matter any more.

Here is an old school load line to see how a 12V vmp 20A imp performs with varying sun and different
loads. It can be scaled to your working numbers. My first attempt would be to set up for panel
voltage somewhat above MPPT at best sun, shifting toward and somewhat past MPPT with reduced
sun and current. Going way down, does MPPT still matter?

Another thing going on here, is to have multiple paralleled strings facing rising, possibly mid day, and
setting sun. This will tend to flatten out daily power as in the second curve, but costs many more
panels. It has the additional benefit of staying closer to your MPPT in good sun, AND bringing in
more power under somewhat dispersed (thin clouds).

Just what is the overall plan, is this totally stand alone? Is cold water OK when weeks go by with
very poor sun? We had 27 days in a row one Dec at 61084, no sun seen. Is the PoCo to be a
backup, with a complete rewiring? Bruce Roe

[psablo]

At least two companies I know of offer MPPT trackers for this purpose. (DC out to heating element) One of them also provides batteryless 120VAC for emergency use.

interesting, can you tell their names?

[psablo]

The efficient design will manage to feed the heaters without inverters. I see the first challenge as matching
the impedance of the heater to keeping the panels somewhere near MPPT. If there is plenty of sun keeping
the panel output current fairly constant, it may avoid an MPPT device. That leaves some series/parallel
combination of panels to match up to varied series/parallel heating elements. I doubt anything in the 12V
realm would be practical.

As for the control contact arc problem, there are devices to handle the switching, to be controlled by
the contact(s). We had some pretty hefty relays in phone offices, and now we have cheap FETs that
may easily be paralleled for this sort of thing, I have used 20. Inductive spikes may kill them if no
protection is used. Keep in mind, the panels can be shut down by shorting the output, maybe just
while a fragile switch is operated. Bruce Roe

Tricky to accomplish since morning and evening will always see a drop in current.
So what can I expect to happen if I power a resistive DC load on a non-MPPT CC when the current drops, assuming I have matched the DC heater resistance (impedance) as you suggest?

Tgriff: Have you considered going back to a 24VAC T-stat energizing the coil of a power relay which would switch your DC voltage?
As BRoe said, there are some pretty robust power relays but they are AC coil. I just don't believe those SSR's are going to last no matter what the rating says.
You may need one power relay per PV panel b/c the following are only rated up to 28VDC:
https://www.automationdirect.com/adc...ad-pr40_series)

As for Solar Thermal, it works great if the system is designed properly and maintained well; most systems I've seen are neither.
i would recommend a DrainBack system to avoid many headaches associated with high temp and high pressure.
Control is the tricky part. Resol controllers are very reliable.

[bcroe]

At least two companies I know of offer MPPT trackers for this purpose. (DC out
to heating element) One of them also provides batteryless 120VAC for emergency use.

If an MPPT control is to be used, it could also handle the job of turning power on and off. Bruce Roe

[jflorey2]

The efficient design will manage to feed the heaters without inverters. I see the first challenge as matching
the impedance of the heater to keeping the panels somewhere near MPPT.

At least two companies I know of offer MPPT trackers for this purpose. (DC out to heating element) One of them also provides batteryless 120VAC for emergency use.

[J.P.M.]

I know the reliability of using a PV solar system, but have read of problems with pumps & leaks in other systems. Does anybody here have personal experience living with one?

If you are referring to solar thermal flat plate DHW systems, yea - me. Designed my own and hired the installation by a licensed plumber. One leaky O ring on a flexible connector in 9 yrs. continuous operation. Lots of operational data. Annual solar DHW fraction ~ >0.95. Cost was $3,200 before tax credit. A nice toy that's probably paid for itself as well as a real learning/confirming tool. Most solar thermal systems don't do as well with respect to performance or reliability mostly due to neglect by owners. As a toy, mine gets a lot of attention.

[Tgriff]

Keeping the goal of heating water in mind, some comments FWIW:

1.) Make sure your heating elements are 5,500 W. Some/many residential units, in the U.S. anyway, are 4,500 W.

2.) Initial heating of 40 gal. of H2O + the tank thermal mass by 60 F. will take ~ 6 kWh of energy.

3.) Assuming a pretty standard tank with current insulation standards, and a 60 F. delta T., tank water temp. - to amb. temp. at the tank, I'd plan on the best (lowest) 24 hr. thermal loss from the tank to be something like ~ 1.2-1.4 kWh/day, and probably more unless the piping is designed to stop gravity induced fluid migration dead in its tracks. That translates to roughly 0.5 F./hr. temp. loss, again assuming the tank to amb. temp. diff. is ~ 60 F.

So, meeting that ~ 7.3 kWh/day = 2,665 kWh/yr. load. How much of that might be supplied by what you're planning is something like, as a maximum, whatever your PV array puts out annually multiplied by (1- impedance consideration %), less whatever other uses the PV output goes for.

4) I'd assume you'll still be using a fair amount of POCO power to supply most of the Hot water load.

5.) On a somewhat separate comment you made, I've got a solar thermal flat plate system on my roof. There are no roof penetrations for piping. Solar thermal for DW heating is still more efficient by ~ 2X than PV alone, but about as efficient as a PV/HPWH combination. Still, the solar thermal police won't nab you if you manage to forego roof penetrations.

I'd respectfully suggest you not believe everything you see or read or hear, particularly the technical equivalent of jackass videos as seen on the new idiots' bible (utube).

1. It has two elements: the upper is 3800 W and lower 5500 W, with a max of 5500 W (wired so only one element on at a time)

2 & 3. I didn't log water temperature or electrical performance for the few weeks I was testing it. I was just glad to have hot water at night and again in the morning - and it did that much well.

4. Actually, I used it for weeks without connecting the water heater to POCO - and it always had hot water for me (until the SSRs overheated). This whole discussion has me more interested in resuming where I took off ...

5. I know the reliability of using a PV solar system, but have read of problems with pumps & leaks in other systems. Does anybody here have personal experience living with one?

Also: I'll have to put in a plug for my YouTube video on a solar irrigation system: Solar Irrigation

[bcroe]

The efficient design will manage to feed the heaters without inverters. I see the first challenge as matching
the impedance of the heater to keeping the panels somewhere near MPPT. If there is plenty of sun keeping
the panel output current fairly constant, it may avoid an MPPT device. That leaves some series/parallel
combination of panels to match up to varied series/parallel heating elements. I doubt anything in the 12V
realm would be practical.

As for the control contact arc problem, there are devices to handle the switching, to be controlled by
the contact(s). We had some pretty hefty relays in phone offices, and now we have cheap FETs that
may easily be paralleled for this sort of thing, I have used 20. Inductive spikes may kill them if no
protection is used. Keep in mind, the panels can be shut down by shorting the output, maybe just
while a fragile switch is operated. Bruce Roe

[psablo]

Great discussion whether it'll work or not.

Tgriff - I appreciate your efforts at solving this because I am a solar installer and have access to as many PV panels as I want for free, but not batteries, inverters, etc.
I am not trying to heat water, but would like to heat my pump control house just enough so the pipes don't freeze. The pump house is well insulated and the pressure tank and pipes are within a second insulated space so I shouldn't need much wattage but it can get to -20F where I live.
There are plenty of 12VDC heaters (and some 24VDC) for automotive available, as well as this:

https://oemheaters.com/category/980/12-and-24-volt-heaters

But, I am concerned with arc-fusing as well since I would like to go direct from PV through cheap charge controller to resistive load. No thermostat, just a manual switch to disconnect in summer, and only heat when the sun's out (lots of sun here).

I couldn't find any other post discussing this so it's probably not done often (if ever).

PS, I used similar SSR's to build my rapid shut down relay box for my off-grid PV system and they lasted long enough to pass inspection but failed within a year.

[J.P.M.]

I considered using such an inverter, but the impedance wouldn't match. The other consideration is that I'd like to utilize as much solar power as possible - meaning that when the sun is not shining perpendicular to the solar panels, the power will drop off and the inverter will quit working, even though partial power is available. If the panels are connected directly, all of the DC power would be sent to the water.

Most water heaters are designed to heat a tank of water in roughly 30 minutes and to keep it hot all day long. My goal is to heat one tank of water in one day, as I won't use more than 40 gallons of hot water in a day. Most water heaters are well insulated and the water will stay hot even if not continually heated.

I'm driving it with 4 of the 230 Watt solar panels, or 920 watts or 96 VDC @ 9.5 A.

Each of the two elements in the water heater runs at around 5500 watts @ 240V. The heating element is basically a big resistor and impedance can be found like so:

P = V * I
V = I * R, therefore P = I^2 * R
or: R = P / i^2
In my case: I = 5500 / 240 v or 22.9 A and R = 5500 / 22.9^2 or 10.5 ohms

There's not enough power to drive it with a 240V inverter, as it demands 5500 watts

At 120V, it would still demand more power than available:

P = I * V or P or P = V^2 / R or P = 120^2 / 10.5 = 1371 Watts (more than my 920 Watt array at max power output)

It may work if I added another panel and would work better if I buffered the power with batteries and a charge controller, but then we're talking about lots more $$$.

Keeping the goal of heating water in mind, some comments FWIW:

1.) Make sure your heating elements are 5,500 W. Some/many residential units, in the U.S. anyway, are 4,500 W.

2.) Initial heating of 40 gal. of H2O + the tank thermal mass by 60 F. will take ~ 6 kWh of energy.

3.) Assuming a pretty standard tank with current insulation standards, and a 60 F. delta T., tank water temp. - to amb. temp. at the tank, I'd plan on the best (lowest) 24 hr. thermal loss from the tank to be something like ~ 1.2-1.4 kWh/day, and probably more unless the piping is designed to stop gravity induced fluid migration dead in its tracks. That translates to roughly 0.5 F./hr. temp. loss, again assuming the tank to amb. temp. diff. is ~ 60 F.

So, meeting that ~ 7.3 kWh/day = 2,665 kWh/yr. load. How much of that might be supplied by what you're planning is something like, as a maximum, whatever your PV array puts out annually multiplied by (1- impedance consideration %), less whatever other uses the PV output goes for.

4) I'd assume you'll still be using a fair amount of POCO power to supply most of the Hot water load.

5.) On a somewhat separate comment you made, I've got a solar thermal flat plate system on my roof. There are no roof penetrations for piping. Solar thermal for DW heating is still more efficient by ~ 2X than PV alone, but about as efficient as a PV/HPWH combination. Still, the solar thermal police won't nab you if you manage to forego roof penetrations.

I'd respectfully suggest you not believe everything you see or read or hear, particularly the technical equivalent of jackass videos as seen on the new idiots' bible (utube).

[PNPmacnab]

You shouldn't need to put more than 500W into a heating element with solar and 960W of panels is a very good match for that element. With that impedance difference you would always have sufficient dead time for an arc quench. I think you have come to the wrong place for any creative thinking on water heating with PV.

[Tgriff]

I considered using such an inverter, but the impedance wouldn't match. The other consideration is that I'd like to utilize as much solar power as possible - meaning that when the sun is not shining perpendicular to the solar panels, the power will drop off and the inverter will quit working, even though partial power is available. If the panels are connected directly, all of the DC power would be sent to the water.

Most water heaters are designed to heat a tank of water in roughly 30 minutes and to keep it hot all day long. My goal is to heat one tank of water in one day, as I won't use more than 40 gallons of hot water in a day. Most water heaters are well insulated and the water will stay hot even if not continually heated.

I'm driving it with 4 of the 230 Watt solar panels, or 920 watts or 96 VDC @ 9.5 A.

Each of the two elements in the water heater runs at around 5500 watts @ 240V. The heating element is basically a big resistor and impedance can be found like so:

P = V * I
V = I * R, therefore P = I^2 * R
or: R = P / i^2
In my case: I = 5500 / 240 v or 22.9 A and R = 5500 / 22.9^2 or 10.5 ohms

There's not enough power to drive it with a 240V inverter, as it demands 5500 watts

At 120V, it would still demand more power than available:

P = I * V or P or P = V^2 / R or P = 120^2 / 10.5 = 1371 Watts (more than my 920 Watt array at max power output)

It may work if I added another panel and would work better if I buffered the power with batteries and a charge controller, but then we're talking about lots more $$$.

[Mike90250]

The other alternative is to use a cheap mod-sine inverter, with remote control ON/OFF to power your heater elements.

Much safer than trying to kludge a Pulsed DC (all those OFF periods, are going to cut your heating capacity by 50%) to use a crappy AC thermostat contact.

[Tgriff]

Right. But:

1) most PWM controllers don't do that (they run slower)
2) if you are turning the power on and off anyway you don't need a separate switch in the thermostat to switch the current. Just use the thing that's already turning the power on and off.
3) a PWM controller isn't needed for a heating element anyway. You're not going to overpower it.

1. I'm not interested in using an "off the shelf" PWM controller, I'm just thinking of driving a SSR from an oscillator to pulse the DC so it will work on an unmodified AC water heater, where the thermostats will regulate the water temperature.
2. This will use the thermostats already inside the water heater.
3. But it DID overpower the water heater - it heated the water above the thermostat temperatures, causing the thermostat contacts to open, resulting in an arc that melted the contacts together. I was surprised at the tiny distance the contacts opened for switching 240VAC, no wonder DC arced across the contacts.

I was heating a 40 gallon water heater with two thermostats, controlling two heating elements with 240VAC. I used 4 solar panels, each with 24V @ 9.5A = 230, for a total of 920 watts. I used this setup as it roughly matched the impedance of one heating element (when operating normally, it powers only one element at a time). It heated slower than 240VAC, but over the course of a summer day, it collected more than enough energy to overheat the water.

[jflorey2]

Arc-fault checking isn't necessary if the supply power is chopped to near-zero several times / second.

Right. But:

1) most PWM controllers don't do that (they run slower)
2) if you are turning the power on and off anyway you don't need a separate switch in the thermostat to switch the current. Just use the thing that's already turning the power on and off.
3) a PWM controller isn't needed for a heating element anyway. You're not going to overpower it.