Dumb heating element and water pump

[MisterSandals]

I'm just hoping I don't damage the panel in some way.

A resistive load pulling as much power as it can is exactly what an MPPT charge controller is designed to do.

Get a heating element that will handle the number of watts for the panels you intend to use. Like a 2000W element for 2000W of solar.

 

[AntronX]

Get 240V heater without thermostat. For example this one is 3kW 240V for $7. Calculate heater resistance: 240^2 / 3000 = 19.2 ohms. Calculate solar panel impedance: 28 Vmp / 7Amp = 4 ohms. 19.2 / 4 = 4.8 or ~5 panels needed in series to have their max power point match this heater element. That's it, theres nothing else you need. Use 15A 240V plug and socket as PV disconnect. No need for thermostat because 1000W of panels are unlikely to boil your radiant floor water loop.

 

[AntronX]

If you want 2 strings of panels in parallel and maintain impedance you would have to go with 4500w heater. You can do 7 series 2 parallel or 14 panels with that element for 2400w total PV heating power.

 

[fatjay]

Thanks for the info.

What if the panel impedance is lower than the heater? Such as times when the panels aren't operating at 100%? Early mornings, late afternoons, cloudy days?

 

[AntronX]

They still produce power but at less efficiency. They will produce max current they can but at mismatched voltage because at low solar irradiance their effective impedance will go up.

 

[AntronX]

Keep in mind that it takes 5.5kWh to heat your concrete slab by 1 degree. So 14 panels in winter weather will not be effective at all. You are better off just running 3kW 240V heater (since you have 20A circuit in the shop already) from grid power. It will take 230 kWh or 3 days to bring up your slab temp to 70F.

 

[fatjay]

I don't need 70, it would just be nice to not be working on a 20f slab, 40f is much warmer. I have a 110k kerosene bullet heater I use there now but it's loud and I go through $30 of kerosene in about 6 hours, to hit 50f. I want to take the chill out of things, not run around naked.

And at current grid rates, 230kwh is $43. I'm trying to reduce consumption as much as possible, adding a new thing on the grid is going in the wrong direction.

 

[AntronX]

Getting up to 40F will take 110 kWh and less energy per day to offset thermal loss to maintain that temp. Where are you located? If you are up north then solar will not help you keep your shop warm in winter. You really have to insulate first then measure energy required to maintain desired temp inside. Then you can decide which heating strategy will be most cost effective.

 

[AntronX]

Check if your power company offers time of use rates. It may be much cheaper to heat the shop during off-peak hours.

 

[Hedges]

Good to hear. I'm debating just buying a heating element for $13, applying random dc voltage, and seeing what happens. I've played with this stuff all my life, but never tried to use it for anything other than it's intended use. In my mind, a heating element is just a wire with resistance that heats up when amps are applied because current is resisted.

Do some math and design it before you buy it or try it.

Resistive heating elements work the same at 240VDC as at 240Vrms AC.
Thermostats and over-temperature safety switches do not; they will fail with high voltage DC. They could be used for "pilot duty" at 24VDC to control a relay or SSR rated for the DC voltage. But as long as you have a non-pressurized tank and element remains underwater (automatic fill valve?) no hazards. Use a suitable DC rated switch to shut off.

Look up the specs of the heating element, 120V or 240V. If it is rated 4800W at 240V, that is 4800W/240v = 20A. 240V/20A = 12 ohms.
With that, you can calculate how much current it would draw at Vmp of your PV string, and how many watts it would deliver.

Probably don't configure series string of PV panels such that Vmp exceeds the 240VAC rating of heating element, to avoid damage.
See if you can make an array 240 Vmp able to deliver around the 20A (if using 4800W element).

And at current grid rates, 230kwh is $43. I'm trying to reduce consumption as much as possible, adding a new thing on the grid is going in the wrong direction.

$0.19/kWh.
PV panels can be had for $0.005/kWh amortized over 20 years, give or take. So 6 months payback compared to using electric grid for heating, if mounting hardware and the rest was free and you wanted the heat year round. 1 year if heating 6 months. Depending on how well you scrounge materials, maybe a few years payback

 

[fatjay]

I'm in eastern PA, it definitely gets cold. I'm going to be building out a better system next summer, I was just hoping to get some relief this winter with a system i could reuse the components of when i build something bigger.

Around here 240w panels are $30-60. I have 18 sitting in the shop waiting on an application.

It'll be a closed system, I was going to find an old hot water tank and use that. Circulation pump off the top port pulling water from the system and pushing it into the tank where it would be warmed and then out to the the manifold. Tanks typically have 2 heating elements.

If memory serves the panels are 31Vmp and 37Voc. 240/31 is 7.7 panels or 7. 7 panels is peaking at 1680 watts for a 3kw element, so I should be able to get 2 strings of 7? Or 6 to be safe?

It's been a long time since my electronic principals course, volts and watts and amps are easy but ohms is resistance, heating elements have resistance, that's how they produce heat. But I'm not sure why the panels have resistance. Or is it their ability to overcome impedence?

If it is reduced efficency, wouldn't it be better to get a small mppt to convert that to a solid 48v output with varying amps and get a 48v dc heating element to match? Or won't it be much different?

 

[Hedges]

7s2p or 6s2p should work fine.
But it will only put out that power at Noon, or whatever orientation the panels. Other times, they get less than full sun.
Since you have 18, try 6s3p, oriented SE, S, SW. Maybe all oriented S and tilted for winter sun, one string rotated E and the other W on that axis to catch morning and afternoon sun.

 

[AntronX]

I was going to find an old hot water tank and use that.

Be careful not to contaminate your radiant floor loop with gunk from old water heater. You really don't need it. Just get a pipe, weld on threaded fittings on the sides, cap the end and weld on threaded adapter on top to fit heater element threads. A DIY tankless water heater.

It's been a long time since my electronic principals course, volts and watts and amps are easy but ohms is resistance, heating elements have resistance, that's how they produce heat. But I'm not sure why the panels have resistance. Or is it their ability to overcome impedence?

Panels have a sweet spot where they will make peak power provided that load impedance (resistance) is just right. MPPT controller acts as variable load impedance tuner to present panels with optimal load they need in order to transfer max power available. You can manually pick a load vs. panels combo to get them as close to their ideal load as reasonable. Since you have 18 panels you can try 6 in series and 3 strings in parallel and not worry too much about getting it perfect right now. During winter you will not get 1000w/m^2 sun anyway.

 

[efficientPV]

Not a bad idea though, i have metal buckets. what's the worst that can happen.

What is the best that can happen? Not much. You can stick two wires into a dead chicken and heat water, it doesn't prove anything. If you don't have numbers to backup what was done it is absolutely worthless. Heating water isn't rocket science, almost all the information is already available. A solar calculator will tell you what energy is available. With direct connect you will at best get half of that. Solar panels do not work well into fixed resistance. Heating is a function of squares and a small drop in current dramatically reduces the actual heating. That is the reason that the calculated ideal resistance can be almost doubled and still produce more daily heating. At least four heating elements are needed to sequence in as sun intensity increases and decreases thru the day. If you want an education, buy a power meter and connect it between the panels and the element. It will soon become apparent that direct connect is a pipe dream. Hydronic heating is total folly unless the floor can be brought up to temperature and maintained. The mass that has to be heated is too large to increase any significant amount. Go with radiant electric. Heat pump would be best. However, that is a major investment in solar equipment.

 

[MisterSandals]

Get 240V heater without thermostat. For example this one is 3kW 240V for $7. Calculate heater resistance: 240^2 / 3000 = 19.2 ohms. Calculate solar panel impedance: 28 Vmp / 7Amp = 4 ohms. 19.2 / 4 = 4.8 or ~5 panels needed in series to have their max power point match this heater element.

This is great. I learned something pretty darned interesting today.

What is the best that can happen? Not much.

It will soon become apparent that direct connect is a pipe dream.

So is it good to do a $10 heating element sanity check or not?

If you don't have numbers to backup what was done it is absolutely worthless. Heating water isn't rocket science, almost all the information is already available.

Are you saying @AntronX and @Hedges numbers are worthless or incorrect?

 

[efficientPV]

No, sticking a heating element in a bucket connected to a panel is stupid without getting any data.

A great heating element nobody knows about is the HE90240. This is a single phase 9KW 240V element for on demand water heaters which is 6.5 ohms. It is physically the same size as a standard 5500W found in many tanks.

 

[AntronX]

A great heating element nobody knows about is the HE90240. This is a single phase 9KW 240V element for on demand water heaters which is 6.5 ohms.

Its $40. But here is idea fop OP to use tankless water heater instead of tank. Just make sure to wire PV direct to heaters bypassing temp switches.

 

[SajgZi]

I know this thread is about dumb heating elements and water pumps for concrete heating but you should probably also consider mini split heat pumps. In heat mode, they run like an AC in reverse (cool outside, heat inside). They are commonly 3x more efficient per watt-hour than pure resistive heating. They are also not super expensive (you don't need the hyper heat stuff since it is for shop heating). Maybe something like this: https://www.pioneerminisplit.com/co...lit-air-conditioner-heat-pump-system-full-set. You'd use the mini split during typical operation and if it's super cold and you need to be in the shop, supplement with kerosene/propane (if it's super cold, they'll be running as basically pure resistive heating anyways).

 

[fpgt72]

Do you have anything for the floor heat now? Pumps, heater anything.

Reason I ask is if you could hook it up to a generator you already own, or could borrow you will get a real world estimate on just how long it is going to take your floor to get up to temp. Personally I think it is going to take longer then you think it will. The concrete is a great deal of mass, I don't think it is going to be an on, and 5 hours later it will be nice in there, I have a feeling the floor will still be ice cold to the touch. Naturally the farther from the heater the less you are going to feel it.

I really think a test is in order before you go too far down this road.

Folk I know with this system leave it on all the time. This way the water is not ice water coming back requiring more energy to get it up to temp. They never turn it off. I don't know what their power bills are however.

Just my two bits. And yes I agree with you a cold floor no matter how good your boots are will get to your feet sooner or later.

 

[fatjay]

Right now it's just the manifold capped off holding pressure. I can rig up a pump pretty easily to circulate water, just the main problem is making it warm.

You have a point, I should just experiment with ac power first. But it's not good for the long term. Even if a 1000w heating element is running 24/7 that's ~9a or 24kwh per day, for 30 days, at $0.19/kwh that's $136.8/mo. That's a lot of juice.

1000w is also fairly low, which is why I was thinking 2-3 strings of panels each being 1200w each that will blast it with heat. Like I said I am not looking for a balmy 70f, I just don't want my feet to be freezing on an ice cold concrete floor.