Sand Battery Experement

I am about to start with my sand battery experiment.

I plan on using it as a simple heat exchanger by blowing the heated air through metal pipe(s) into a room. Dimensions of the build will be 4x8x4 to start, bigger if needed.

I've decided to bury standard stove-top heating elements like the attached photo into the insulated battery. The individual element is rated 2600 watts. I'm thinking of putting them up 12 inches apart from each other for a total of 8 buried elements. Each element will be pulled apart so they form a cone shape and will heat more sand.

I need help calculating how many panels to attach to each element to continuously keep them heating during the day. I still need to figure out how hot I should get them & how I would do that.

Thanks!

i was watching a guy on youtube a little while back who was doing a lot of sand heat battery experiments....off-grid survival mike. maybe you can get some ideas there.

cool, will be following your experiment. Eventually I will be doing a water thermal battery experiment.

To control the heat and how warm they get you will likely need to use some temp probes and an Temp Controller to control the heat turning on and off.

I hope you have another use for all that sand later.

efficientPV said:

I hope you have another use for all that sand later.

Click to expand...

I live in Arizona's high desert! I'll just tip it over & it will blow away or stay.

jasonojordan said:

To control the heat and how warm they get you will likely need to use some temp probes and an Temp Controller to control the heat turning on and off.

Click to expand...

Good idea. Now that I recall, I think the YouTube battery theory guy talked about it.

Ceefiveceefive said:

I am about to start with my sand battery experiment.

I plan on using it as a simple heat exchanger by blowing the heated air through metal pipe(s) into a room. Dimensions of the build will be 4x8x4 to start, bigger if needed.

I've decided to bury standard stove-top heating elements like the attached photo into the insulated battery. The individual element is rated 2600 watts. I'm thinking of putting them up 12 inches apart from each other for a total of 8 buried elements. Each element will be pulled apart so they form a cone shape and will heat more sand.

I need help calculating how many panels to attach to each element to continuously keep them heating during the day. I still need to figure out how hot I should get them & how I would do that.

Thanks!

Click to expand...

I'm going through the process of building one right now.

But, some caveats: it's stand-alone from the rest of the house solar. No controller - just raw DC straight to the elements. The elements I'm looking at are the largest in diameter I can find, at 2400W/240V - they typically are between 25 and 45 ohms - smaller diameter are higher resistance. I have based calculations on 30ohm.

So, a 2400W element at 240V and with a resistance of 30 ohms runs at ~800C at full power.

Given my winter panels will be delivering between 350-530W/M2 ea for ~6.5 hours in mid winter at 36V, I am looking at ~300 watts/m2 from the element for a single panel but I'm planning two panels at 72V with an area of 3.85M2 so hopefully will get ~in excess of 1000W for around 5-6 hours.

My system is planned to be a ducted warm air implementation, from a 200L/44 Gal drum sand battery with a single element, insulated on all sides and heat-broken from the ground it stands on. Under the house, it will be protected from wind and weather and thermal losses kept to a minimum by locating it directly below the duct outlet, so at most a couple of metres of duct.

If successful, I will make another for the other end of the house.

Reasons behind everything are here: Sand battery

So water stores about 5x more heat per LB than sand. And dry sand is less dense than water. So you need a lot more space to store much heat with sand. You can buy a ready made hot water storage tank, already insulated -- its called an electric water heater. Those elements work AC and DC IIRC. You can buy any number of 12/24/48/ 120AC/240AC small pumps to move water around. Your heat exchanger is near to free. Visit a local pic and pull and a small automotive radiator from a used car does the trick.

Some ducting, which sounds easy for you, and you have a mini hydronic to air heating system.

Put some ethylene glycol in there (antifreeze) and you can even let it freeze safely, if that problem applies to your locale.

Rocks and sand don't transfer heat very well. And they don't move around on their own. Water does both. Water also integrates with a waste boiler system very well. Meaning: you build a barrel stove, wrap some loops of flexible copper around the barrel, have a diverter valve so that your flow pump can also be used to circulate water around your trash incinerator. Bingo, you have hot water on a cloudy day burning your trash.

All this stuff would be frowned upon in most urban environments. But it sounds like you are out in the open and free part of the world.

I don't know where you get your figures from - dry sand is 1600kg/m3, water is 1000kg/m3 - that is more dense. If sand was less dense than water, there would be no sand or beaches and making concrete would be really, really hard, no?.

Water can store 4000J/kg vs sand at 835J, but sand has more kg per L of space. so you end up with 1335J in the same amount of space as 1kg of water.

Water holds 5x more specific heat per kg but 3x specific heat/L, I showed (via link) the figures that show 1cuM of sand holding twice the kJ of water - because the water can only go to 100C in free air, vs >500C for sand. 500C is readily achievable with stove top elements, which typically run at 800-950C, depending on size and wattage.

You seem to be confusing free-air specific heat retention, which is of less interest as the vessel will be heavily insulated. The whole reason for a battery is to insulate it against uncontrolled thermal loss. The reason to use sand is because of its physical properties - it won't change state until you reach 1700C. Sand absorbing and releasing Joules at a higher transfer rate is an advantage in a battery, where you seem to think it's a negative. It would be a negative if you weren't insulating.

Or, you can go and tell the Finns they're doing it all wrong and need to convert their municipal sand batteries to water?

MyK3y said:

Or, you can go and tell the Finns they're doing it all wrong and need to convert their municipal sand batteries to water?

Click to expand...

Sand batteries are sort of experimental also in here, one(1) unit in use.
Whereas large 2-3m3 water tanks are rather common and have been in use at least last 40 years.
I know half a dozen relatives and friends houses that have such a system installed. Forum member upnorthandpersonal also has one.

cdherman said:

So water stores about 5x more heat per LB than sand. And dry sand is less dense than water. So you need a lot more space to store much heat with sand. You can buy a ready made hot water storage tank, already insulated -- its called an electric water heater. Those elements work AC and DC IIRC. You can buy any number of 12/24/48/ 120AC/240AC small pumps to move water around. Your heat exchanger is near to free. Visit a local pic and pull and a small automotive radiator from a used car does the trick.

Some ducting, which sounds easy for you, and you have a mini hydronic to air heating system.

Put some ethylene glycol in there (antifreeze) and you can even let it freeze safely, if that problem applies to your locale.

Rocks and sand don't transfer heat very well. And they don't move around on their own. Water does both. Water also integrates with a waste boiler system very well. Meaning: you build a barrel stove, wrap some loops of flexible copper around the barrel, have a diverter valve so that your flow pump can also be used to circulate water around your trash incinerator. Bingo, you have hot water on a cloudy day burning your trash.

All this stuff would be frowned upon in most urban environments. But it sounds like you are out in the open and free part of the world.

Click to expand...

Thanks for your input!

I am in a very rural area with lots of free sand on my property.

I am not interested in the water version, only the sand version.

MyK3y said:

I'm going through the process of building one right now.

But, some caveats: it's stand-alone from the rest of the house solar. No controller - just raw DC straight to the elements. The elements I'm looking at are the largest in diameter I can find, at 2400W/240V - they typically are between 25 and 45 ohms - smaller diameter are higher resistance. I have based calculations on 30ohm.

Click to expand...

My system will be stand-alone as well, completely separate from my main solar array.

MyK3y said:

So, a 2400W element at 240V and with a resistance of 30 ohms runs at ~800C at full power.

Given my winter panels will be delivering between 350-530W/M2 ea for ~6.5 hours in mid winter at 36V, I am looking at ~300 watts/m2 from the element for a single panel but I'm planning two panels at 72V with an area of 3.85M2 so hopefully will get ~in excess of 1000W for around 5-6 hours.

Click to expand...

This is what I need help with in my calculations.

I have a total of at least ten 240W & 250W panels to use for this project. I'm not sure if I need to combine them and how many to use per element? My elements are rated at 2600W each.

You might need to factor in the insulation factor. Composting systems have been considered for heat generation. Basically tubes are run through composting material to pull the heat away.

It does work but as the heat is pulled away from the compost surrounding the tubes it then acts as an insulator limiting the potential exchange from compost further away from the tubes.

MattiFin said:

Whereas large 2-3m3 water tanks are rather common and have been in use at least last 40 years.
I know half a dozen relatives and friends houses that have such a system installed. Forum member upnorthandpersonal also has one.

Click to expand...

Not here. Never seen such a thing.

This thread is specifically about sand storage - why has it been taken over by water heaters?

Ceefiveceefive said:

I have a total of at least ten 240W & 250W panels to use for this project. I'm not sure if I need to combine them and how many to use per element? My elements are rated at 2600W each.

Click to expand...

It's really just down to maths - you need to know your solar radiation specifics for your site, your panel orientation and generation, efficiency, cable budget, voltage, etc.

I started with a solar generation spreadsheet from NIWA - our government agency that covers weather and atmosphere - you put in your exact location and orientation and panel angle and it creates a power profile of watts/m2 of collector. I did that for every panel angle from 30deg to 75deg. The sweet spot for winter was 65 deg, for summer it was 40 deg. - another reason for wanting it separate from the house solar generation - this is specifically for winter so getting the angle optimised for winter is key.

Depending where you are in the world, you may have access to the same sort of resource

The highlighted section is 'cloudless W/m2' - so what you would expect to get in perfect conditions. It's averaged for the whole month - the spreadsheet covers a whole year, by hour per month. As you can see, 3W at 7am rising to 931W at noon and back to zero at 5pm. The figures just to the left are 'cumulative kWh/m2', so how much you can expect to generate on a particular day - 2.7kWh in May, but my panels are 3.75m2 for the pair so 10.125kWh of generation for each day in May at 72V. Just as an indication, for all-year optimisation, with panels at 40deg, generation in May would be 3kWh less per day - so worth getting right.



I already had the specs from Trina for the panels - m2 of collector, voltage, amperage, etc.

I am not using an inverter/controller - just raw DC. My intended elements are 3kW at 240V, but that will be significantly reduced at the ~72V DC my panels will operate at. And that number changes during the day - they are giving ~15V at 8am, rising to 36V at peak, so from 30V to 72V in series. An element that gives 2400W at 240V will be giving 360W at 36V - your sand reservoir and how much heat you want to pull from it will dictate how much power you put in.

At 240V and 10A a 2400W element will be at 24 ohm of resistance
At 36V and 10A the same element will be giving 360W at 3.6 ohm
At 72V and 10A the same element will be giving 720W at 7.2 ohm

Look at 'Off-Grid Survival Mike's videos - he is the origin for my setup. https://www.youtube.com/@off-gridsurvivalmike8120/videos

Also look at the other thread 'Storing Heat in Sand' https://diysolarforum.com/threads/storing-heat-in-sand.27147 for more detailed reasoning and discussion - I'm going to stop posting in this thread now.

MyK3y said:

I am not using an inverter/controller - just raw DC. My intended elements are 3kW at 240V, but that will be significantly reduced at the ~72V DC my panels will operate at. And that number changes during the day - they are giving ~15V at 8am, rising to 36V at peak, so from 30V to 72V in series. An element that gives 2400W at 240V will be giving 360W at 36V - your sand reservoir and how much heat you want to pull from it will dictate how much power you put in.

At 240V and 10A a 2400W element will be at 24 ohm of resistance
At 36V and 10A the same element will be giving 360W at 3.6 ohm
At 72V and 10A the same element will be giving 720W at 7.2 ohm

Click to expand...

Ideally, you will want to load the panels to close to their NOTC current at solar noon.

The resistance of a stove heating element will likely change some with temperature, but not this much. A 24 ohm element will likely still be close to 24 ohms when running at just 72 volts. The power will be much less than the rating though. When you cut the voltage in half, you only get 1/4 of the power. At just 72 volts, the power is way down.

72 volts / 24 ohms = just 3 amps. And 3 amps x 72 volts is only 216 watts. That will not take full advantage of the PV solar panels. If you run 4 of those solar panels in series, you get 144 volts. Into a 24 ohm load, that will try to pull 6 amps. The solar panels should easily be able to supply that. I think that would be a good place to start to get some idea of the heat available. 6 amps x 144 volts = 864 watts. That is only asking for 216 watts per solar panel with 4 in series. This is still well below the rating of the element, so it should not be in any danger of burning it out. You might be able to run it even a bit harder, but I would keep it to less than half of the rated power. Stove heating elements are normally run in open air, or in contact with a pan full of water to pull heat out of the element. Having the element buried in sand will pull some heat out, but it is then in an insulated casing, the temp could rise quite a bit higher than what it would do out in free air. Half power (1,200 watts) is probably still safe though.

In most areas, you are lucky to get a solid 3 or 4 sun hours of energy out of the solar panels. Using 3.5 x 864 watts means you might put a bit over 3 kilowatt hours of energy into the sand. Then blowing the air through it to make heat, it would produce similar heat to running a 1 kilowatt space heater for 3 hours.

Depending on the V/I curve on the panels, it might work better to parallel 2 of those stove heating elements. That puts you down to 12 ohms instead of 24 ohms. If the panels IMP current is 12 amps, you will get to the same 144 volts, but now at nearly double the peak power. 144 volts x 12 amps = 1,728 watts at peak solar noon power. This is asking for 432 watts per solar panel, so I think it will fall a bit below that in the real world. Are these 300 to 400 watt panels? What is the VMP and IMP? We should try to hit those as close as possible. Assuming the panels max out at only 400 watts, the voltage would fall a bit to 138 volts / 4 = 34.5 volts per panel and about 11.6 amps. Each stove element makes less heat, but you now have 2 of them pushing heat into the sand. The only problem with this setup is that the voltage will fall off a lot when the sun is not ideal. So the total energy per day might end up less. That is the problem with not using an MPPT tracker. You can only optimize a single point with a fixed resistor load.

JJJJ said:

You might need to factor in the insulation factor. Composting systems have been considered for heat generation. Basically tubes are run through composting material to pull the heat away.

It does work but as the heat is pulled away from the compost surrounding the tubes it then acts as an insulator limiting the potential exchange from compost further away from the tubes.

Click to expand...

I plan on heavy insulation!

GXMnow said:

In most areas, you are lucky to get a solid 3 or 4 sun hours of energy out of the solar panels.

Click to expand...

or 7-8 in winter, like we do.

MyK3y said:

or 7-8 in winter, like we do.

Click to expand...

No, not even close.
You may have 7-8 hours of daylight, but the sun is at such a low angle, the energy is much lower.

One "Sun Hour" is an amount of sunlight that will produce 100 watt hours from a 100 watt solar panel, but it may take 4 hours to get it. I am in a very sunny area, and in summer, we rarely hit 7 sun hours a day. And in the winter, it can fall under 3 sun hours even on a clear day. Add in clouds, and it falls to just 1 sun hour all day.

Look up "Solar irradiance at my location" and you should be able to find a chart or calculator to find this data. I have used this one and it is a pretty good estimate.


I typically get 80 to 90% of the estimate from this site each month. I think most of that error is that it does not allow exact panel angles, and I have some shading in both the early morning and the late evening. And my power reading is after the inverter which has some loss and also clips a little near noon on sunny but cold days.

For the whole year, my 4,800 watts of solar panels produce right about 8,000 kilowatt hours. That is an average of just over 4.5 sun hours a day.

Cloudless Wh per m2 at my location, middle of winter. To the left, cumulative kWh for the day per m2 of collector. Further left, average Wh per m2 over the month

I don't need your ad site - besides which, it's broken

GXMnow said:

No, not even close.
You may have 7-8 hours of daylight, but the sun is at such a low angle, the energy is much lower.

One "Sun Hour" is an amount of sunlight that will produce 100 watt hours from a 100 watt solar panel, but it may take 4 hours to get it. I am in a very sunny area, and in summer, we rarely hit 7 sun hours a day. And in the winter, it can fall under 3 sun hours even on a clear day. Add in clouds, and it falls to just 1 sun hour all day.

Look up "Solar irradiance at my location" and you should be able to find a chart or calculator to find this data. I have used this one and it is a pretty good estimate.

Solar Irradiance Calculator

www.solarelectricityhandbook.com

I typically get 80 to 90% of the estimate from this site each month. I think most of that error is that it does not allow exact panel angles, and I have some shading in both the early morning and the late evening. And my power reading is after the inverter which has some loss and also clips a little near noon on sunny but cold days.

For the whole year, my 4,800 watts of solar panels produce right about 8,000 kilowatt hours. That is an average of just over 4.5 sun hours a day.

Click to expand...

Here are my results with the panels flat. I however adjust them every month & I do not have shading.
My guess is that I will have at least 5-6 hours per day in the winter.

Ceefiveceefive said:

Here are my results with the panels flat. I however adjust them every month & I do not have shading.
My guess is that I will have at least 5-6 hours per day in the winter.

Click to expand...

I am 30 miles north of Los Angeles, so our numbers should be pretty close, but AZ typically has a bit less clouds, and you are a little further south than me, so you will do a little better. And adjusting the panel angle every month really helps. You also selected Due South, my main array is about 25 degrees west.

If you click on the "Adjusted Throughout the Year", it will give you the optimum for each month.

For my Los Angeles location, going due south, and adjusted each month, the prediction does go up to 4.85 for Dec., 4.8 for Jan., and 4.95 for Feb.
Doing the same for Chandler, AZ it is a bit better at 5.28 for Dec., and 5.29 for Jan. That has the panels way up at 48 degrees (42 from vertical). I did not expect it to go up that much. That is truly excellent winter production. To get that much, there has to be nothing around to create shadows. In winter here, I get shadows from all my neighbors trees with the low sun angle. I make enough power/energy in winter, so it is not a big deal for me, the 3 sun hours I get is enough. And I do think that web site is also a little optimistic. I am only hitting about 90% of what they predict. But still, those are good numbers.

Location and panel angle means a lot for energy production. Too many people seem to assume a solar panel just makes it's rated power whenever there is daylight, and that simply does not happen. For 90% of the year, my panels top out at under 80% of the STC rating. And you also need to accept the losses when the panels get hot. This will certainly be a factor in AZ. My panels exceed 140 degrees F (about 60C) on hot summer days here. And that has a big effect on the panel output voltage. My panels lose about 0.3% per degree C over 25. This is 35 degrees C hotter. 35 x .3 = 10.5% drop in output power. That makes my 300 watt panels into 268 watt panels just from the heat. Some newer panels can get the temp coefficient down to 0.22% per degree C, but they are more expensive panels.

MyK3y said:

Cloudless Wh per m2 at my location, middle of winter. To the left, cumulative kWh for the day per m2 of collector. Further left, average Wh per m2 over the month

I don't need your ad site - besides which, it's broken

Click to expand...

I took the numbers from your chart. If you add them up, you get a grand total of 6,056 watt hours per square meter. That would calculate to 6.056 sun hours. BUT... What surface was that measured on? I am guessing it is total from a tracked panel, always facing directly at the sun. Are you going to use a tracker? And even with a tracker, those numbers still seem very optimistic. Where is that data from? Is that a spread sheet available on the web? I would be curious to see what it produces for the Los Angeles area of California. How close are the numbers to what I am getting from real solar panels?

That site I linked is not an ad site, I have nothing to do with it, it is just one I found that seems to work fairly well. You can search the web and find another, there are many on the web. The calculator is not perfect and it does sometimes error out, especially if you try to change entries after the initial choices. All I do is click "home" at the top, then click "online calculators", and bottom left is the "irradiance calculator". Load the table and again and then re-enter the data. I did it just now for Aukland, NZ. I do not know how far that is from you as I don't know NZ very well. I guess I could go to a map and find the closest city to you, but I do not care enough at this point.

With the panels facing straight north, and the tilt angle at best for winter, the May results are coming in at just 3.64 sun hours. I have real panels, and I measure the real output, and this site is predicting a bit more than I get most of the time. A few days, when the air is cool, I am able to match it and even did beat it a few days, but on average, I get about 10% less than the prediction. Unless you have a tracker following the sun, I would not expect you to get over this 3.64 sun hours a day in May, unless you are much closer to the equator than Aukland, but there is not a lot more of NZ to the north from there. You might do better, but if you do, I would call that a bonus. And the 6 sun hours from your table seems too far off to be a good prediction.

I am not telling you this to say you are wrong or to make you mad. I am doing this to give you a better idea of what to expect in the real world. If you spend a lot of money on this project, expecting to get 6 sun hours in May, and you only get 4, you may be very upset. But if you expect just 3.64 sun hours, and you do get 4, then you are over producing. And that tends to make people a bit happier. If you go in expecting too much, you won't be happy. The calculator I linked is not the best or the only one around, and I don't care if you never go to the page. But when you do put up PV solar panels, and you end up getting only 60% of the energy you expected, don't be mad at me.