Sand Battery Prototype Help!!!!!!

[Ceefiveceefive]

Greetings!

I am going to assemble my larger sand battery I posted about here: https://diysolarforum.com/threads/sand-battery-experement.67368/

I have made a few slight changes & have purchased a large metal barrel measuring 6' tall and a 4' diameter. Before I start assembling this large outdoor sand battery, I want to do a smaller indoor prototype with a few deviations.

This video is basically what I want to do:

Instead of using a plastic bucket and a coffee can for containers, I will use a small metal trashcan inside of a larger one.

I will use 2 elements stretched out (see pics) evenly in the small trashcan along with aluminum bars throughout to help transfer heat. The elements are rated 2100 watts each.

Questions:

1. How should I attach the solar panels safely to the element connections? (See pic)
Is it safe to attach PV wire straight to the connectors?

2. Is it advisable to attach the elements in series or parallel?

3. What is a good estimate of how many panels I will need to attach to the element connection? I have a variety of panels including 240 , 250, & 270 watts.

Thanks! If clarification is needed, let me know!

 

[Hedges]

Heating element contacts probably get too hot for PV wires.

Inside heating appliances you will see wire insulated with asbestos, fiberglass, silicone, something able to take the temperature.

I suggest lengths of such wire in between, and keep PV wires outside the insulated area.

Also ground the metal enclosure, so it doesn't become a shock hazard if circuit shorts to it.

Elements and panels, connect each series/parallel as needed to get resistance which will draw Imp at around Vmp.

Start out well below rated voltage of element. Surrounded by sand, it may have much greater temperature rise than it was designed for.
A view port allowing you to see it's glow could help you gauge whether it's temperature is OK.

 

[Warpspeed]

Most efficient (and simple) way to connect solar panels to any heating element is to keep the solar panels operating at very close to VMP no matter how much or how little solar there is.

One way to do that, is to connect a very large electrolytic capacitor bank directly across a suitable high voltage series string of solar panels, so that the capacitors charge up to a voltage slightly higher than VMP.
When that voltage is reached, switch the heating load across both capacitors and solar panels with a mosfet to discharge the capacitor bank down to a voltage slightly lower than VMP.

The capacitor bank then charges back up and the cycle repeats. How fast it cycles depends on how much solar there is and the capacitance of the electrolytics.

When the sky is gloomy grey, the capacitor bank may take a comparatively long time to charge up, but it will eventually discharge into the heating load, and the system will cycle much more slowly, but it still puts the maximum available power into the heater in short sharp pulses.

With full solar in a clear blue summer sky, capacitor charging will be much faster, and it still puts all the available power into the heater.

This works very well, because the voltage across the solar panels always stays fairly close to the maximum power point.
In effect its a slow motion low frequency MPPT controller, the fact that it pulses on and off makes no real difference to heat transfer.
It very effectively precisely matches the power source to the heating load, and its very simple to build and to adjust.

Several people now have these systems successfully running, used mostly as hot water pre heaters, but it would work equally well heating up something else. People are pleasantly surprised at how much electrical heating energy can be harvested on very cloudy days.
This definitely works.

Just connecting solar panels up directly to a random heating element will not work anywhere nearly as well.

 

[rogerdw]

I posted a link in another thread to a "box of bricks" storage heater and thought a picture from the manual might be worthwhile here in regard to element placement etc.

Can also see the full manual

 

[jfet]

Most efficient (and simple) way to connect solar panels to any heating element is to keep the solar panels operating at very close to VMP no matter how much or how little solar there is.

One way to do that, is to connect a very large electrolytic capacitor bank directly across a suitable high voltage series string of solar panels, so that the capacitors charge up to a voltage slightly higher than VMP.
When that voltage is reached, switch the heating load across both capacitors and solar panels with a mosfet to discharge the capacitor bank down to a voltage slightly lower than VMP.

The capacitor bank then charges back up and the cycle repeats. How fast it cycles depends on how much solar there is and the capacitance of the electrolytics.

When the sky is gloomy grey, the capacitor bank may take a comparatively long time to charge up, but it will eventually discharge into the heating load, and the system will cycle much more slowly, but it still puts the maximum available power into the heater in short sharp pulses.

With full solar in a clear blue summer sky, capacitor charging will be much faster, and it still puts all the available power into the heater.

This works very well, because the voltage across the solar panels always stays fairly close to the maximum power point.
In effect its a slow motion low frequency MPPT controller, the fact that it pulses on and off makes no real difference to heat transfer.
It very effectively precisely matches the power source to the heating load, and its very simple to build and to adjust.

Several people now have these systems successfully running, used mostly as hot water pre heaters, but it would work equally well heating up something else. People are pleasantly surprised at how much electrical heating energy can be harvested on very cloudy days.
This definitely works.

Just connecting solar panels up directly to a random heating element will not work anywhere nearly as well.

nice idea! how do you know what the VMP is?

 

[Warpspeed]

The VMP will be shown on the ratings plate of the solar panel.
Its not really that critical, the power falls away quite slowly either side of maximum.

 

[Hedges]

More slowly toward lower voltage, faster toward Voc.
And Vmp trends lower with less illumination.
so aim low.

 

[Warpspeed]

This is how I did it using a 555 timer chip:
Sorry for the very low resolution schematic, not sure what happened there.

 

[Warpspeed]

Ah !
found it on another hard drive, had to convert .BMP to .GIF file before I could post it.

Suppose you had seven solar panels connected in series, each with a rating plate VMP of 31v, that would give a theoretical VMP total of 217v
powering a 230v rated heating load.

As Hedges says^, we ideally need to have more operating range on the low side than on the high side.
So we decide to set our on trip point at 228v (11 volts above VMP) and our off trip point at 184v (33 volts below VMP)

R1 and R2 set the two trip voltages.
For example if R1 = 180k, then the load switches off at 184v. (180k at 1ma = 180v + 4v)
If R2 - 220K, then the heating load switches back on at 228v. (220k at 1mA = 220v + 8v)

All this is not in the least bit critical, the nearest standard resistor values will be close enough, and work fine for whatever trip voltages you decide.
But 44v of hysteresis (20%) for a 217v VMP system, might not be an unreasonable starting point for some further experimentation.