Storing heat in sand?

[Honuz]

My point is you have to operate the heater at a temperature it (and the sand) can handle.
If only raising to 100C maybe 200C, fixed power could be OK.
Raising sand to 600C, depending on conductivity, temperature of the element may rise as the drum of sand gets hotter, so tapering to a lower final wattage would allow faster heating earlier. Otherwise, a constant reduced power.

You can experiment, see if you wreck elements. My guess is they will get too hot if powered with 240V and buried in sand. It might come out encased in glass, and/or it may fracture.

The stove control may simply be a dimmer switch. That or a light dimmer would be a good way to test power vs. temperature.

To inject lot power in sand without overheating the vicinity of the resistance :

1 - Increase the number of resistances : my wife is a ceramiciste and there are like 3m/4m of twisted wire for a pretty small furnace

Wire look like this :

This pack .. cost 350€, but it heat ceramic to 1300°C, and they do not last long => higher temperatures also mean more expensive maintenance.
Firestone/firebrick are used in first layer in contact with the high temperature material.

2 - Decrease power

As an example i can send 3.6kW in my water tank without any problem (i use dead water), i got 2 x 3 resistor (600W each), i use solid state relays to switch them on depending on the available excess pv power, controlled with an arduino ( i got 8kWc of pv).

Switching to an other thought, heating a thermal mass (with sun power) the whole summer to use it in winter is not .. that usefull, periodicity is too important. It would be like going on a 15 days hike and taking all needed water for the trip while there are a lot of places to take water on the path. My system deliver 12kWh on a sunny winter day (35kWh in summer).

Slowly storing a lot of energy for late use will lead to a lot of loses, taking that as a quality, the easier solution would be to put the sand tank in the middle of the house with enough insulation to let him radiate what needed to keep the house at the right temperature for an average day of winter, just let it radiate.... add a simple sleeve over it to have a little control of the output.

 

[MyK3y]

My point is you have to operate the heater at a temperature it (and the sand) can handle.
If only raising to 100C maybe 200C, fixed power could be OK.
Raising sand to 600C, depending on conductivity, temperature of the element may rise as the drum of sand gets hotter, so tapering to a lower final wattage would allow faster heating earlier. Otherwise, a constant reduced power.

You can experiment, see if you wreck elements. My guess is they will get too hot if powered with 240V and buried in sand. It might come out encased in glass, and/or it may fracture.

The stove control may simply be a dimmer switch. That or a light dimmer would be a good way to test power vs. temperature.

I’m confused as to where melting sand, 240V, and conductivity came into play - it’s a fixed 64V and the sand has infinite resistance at 64V.

The object of the exercise is to convert every watt of electricity to heat and store that heat for use in warning air.

The panels vary in output as the sun climbs and will never exceed the fixed maximum parameters of the panels

Using the accurate solar irradiance data from NIWA I can calculate the optimum direction and angle of panels during the period I wish to heat - June, July, August, typically

By using a specific gauge of nichrome 80 at a specific length and with the set value for voltage, which never changes, I can create a coil or coils that most closely matches the maximum output of the panels at the time of year I need to use them

The very last thing I want to do is decrease load or add resistance.

Nichrome 80 will keep adding heat to sand right up to its melting point - 1400C. Short of a mechanical failure, only adding more volts will make it reach melting point and I have no way of increasing volts - that is a product of the number and size of discrete solar cells in my panels.

Defining my coil length so at peak amps at 64V it reaches a maximum temperature of 500C means it won’t exceed 500C - you can’t exceed that temperature due to the nature of the coil and maximum output of the panels - but you can increase the number of joules stored.

It’s possible to create a coil that has too much resistance, so you have untapped potential. It’s possible to have too little resistance, resulting in too high a temperature and too much current drain.

But, short of a massive change in the laws of physics nothing will make the sand fuse or the wire go over 500C or magically start conducting.

 

[Hedges]

I’m confused as to where melting sand, 240V, and conductivity came into play - it’s a fixed 64V and the sand has infinite resistance at 64V.

If you have a heating element that operates at appropriate temperature in air, then if you bury it in sand I think it will run hotter, possibly too hot.

So a 240V stove element, if buried in sand and powered with 240V, I think it might overheat.

If you make a custom heating element of course you can tailor it to the application.

Defining my coil length so at peak amps at 64V it reaches a maximum temperature of 500C means it won’t exceed 500C - you can’t exceed that temperature due to the nature of the coil and maximum output of the panels - but you can increase the number of joules stored.


But, short of a massive change in the laws of physics nothing will make ... the wire go over 500C

Given low TCR, constant voltage means constant power dissipation. The sand will rise in temperature. At some point, power will be lost to surrounding environment. It will reach equilibrium when power lost equals power dissipation of heating element. The element will be above environment temperature by PowerDissipation/ThermalConductivity. That temperature could be above 500C.

Only thermal resistance of insulation and thermal resistance of sand will determine final temperature.

If you put the element at center of an infinite volume of sand, there would be a temperature gradient due to thermal resistance, power input, and volume the heating element is distributed over.

I think there is a maximum wattage you want for a given design, so you can heat all the sand to desired temperature without having heating element too much hotter than extremities of the sand "battery". As you say, too low a resistance would result in too high a temperature.

I suggested an element powered by AC and dimmer switch to experiment, measure temperature rise and gradient.

 

[MyK3y]

While experimenting and determining what element to run at what voltage, AC makes it easy to modulate power levels.

When powering a resistive load, you can achieve MPPT with just switching and a capacitor, no inductor needed. There is a unit which does that for water heaters, I think that's what brum is referring to. Might be worthwhile if goal is to get "all of them" watts.

There seems to be a disconnect, somewhere.

While the panels aren’t operating at peak power, they are still generating electricity and that electricity is still at 64V and still heating the element and that heat is still being stored in the sand.

MPPT? It’s a panel directly wired to a resistive wire. Every erg coming out of that panel is directly converted to heat - it can’t do anything else.

It’s the simplest form of electrical circuit - a wire.

Things to consider are: carrying capacity of the sand; how best to harvest the heat; mechanical connections and resistance; length and diameter of feeder cable; length and diameter of the Nichrome 80; insulation and safety of the vessel;

 

[Hedges]

While the panels aren’t operating at peak power, they are still generating electricity and that electricity is still at 64V and still heating the element and that heat is still being stored in the sand.

MPPT? It’s a panel directly wired to a resistive wire. Every erg coming out of that panel is directly converted to heat - it can’t do anything else.

It’s the simplest form of electrical circuit - a wire.

For a given amount of sun, PV panels deliver peak power at a particular voltage. The light produces current, some of which leaks back through PV diode and some of which is delivered to your load.

If PV goes to a capacitor and you PWM the resistive load on/off, you can vary power delivered from capacitor to resistive load, drawing capacitor voltage down to MPPT of PV array.

This would be a potential future improvement, after initially designing resistive element to match PV array. But it can only serve to raise operating voltage, not lower it.

 

[Honuz]

For a given amount of sun, PV panels deliver peak power at a particular voltage. The light produces current, some of which leaks back through PV diode and some of which is delivered to your load.

If PV goes to a capacitor and you PWM the resistive load on/off, you can vary power delivered from capacitor to resistive load, drawing capacitor voltage down to MPPT of PV array.

This would be a potential future improvement, after initially designing resistive element to match PV array. But it can only serve to raise operating voltage, not lower it.

MMmm let's imagine i got a string with an Mppt connected to it, then in parralel of this MPPT i connect a variable load (through PWM) going to a resistive load.
Would the Mppt still "drive" the panel correctly to the MPPT ..? (i'm kind weak on electronic circuitry and their behavior)

 

[Hedges]

I think so, within limits. Although commercially available MPPT SCC behavior as programmed for batteries might get in the way, for instance automatically recognizing 12V, 24V, 48V battery voltage causing it to change its settings. A dumb one might even work.

If you had an MPPT circuit that didn't try to outsmart you, you could vary resistance of load (e.g. electric stove, turn on more/fewer burners) the MPPT would adjust to follow. Alternatively, if you had a continuously variable resistor (carbon pile battery tester or potentiometer), you could adjust resistance to maximize power harvested.

The typical Maximum Power Point Tracking Solar Charge Controller (MPPT SCC) might harvest maximum power from PV and deliver it to a load, but only within a certain range of voltage and current.

They are designed to act as a CV/CC (constant voltage/constant current) power supply for a period of time. They deliver up to their rated current (e.g. 30A) up to programmed battery voltage (e.g. 56V). But after a time "absorption", they drop to a lower voltage "float".

While doing so, they adjust current drawn from PV array up & down, always trying to draw the maximum W = V x A, therefore delivering maximum to the battery, or less if they have hit target current or voltage (e.g. 30A or 56V). If they reach that limit, they draw less power from PV array. The extra power from the sun is then dissipate in the array.

I was loosely using the term "MPPT" to mean a circuit which seeks to operate PV array at Vmp/Imp. For the sand heater, you would want to similarly deliver all available power from PV, and a supply delivering CV/CC output could do that. Maybe some SCC would accomplish that if set to suitable limits including a voltage high enough that it is never reached, so SCC doesn't drop to "float" voltage.

MPPT SCC are switching power supplies that contain an inductor, switched with pulses to act like a variable ratio transformer but for DC.

The neat thing about driving a resistive load is the power circuitry to achieve MPPT can be much simpler. Just a capacitor bank connected to PV array, and a transistor being switched on and off, at maybe a few kHz.

If you simply connect a PV array directly to resistive heating element, you could tune it (adjust element length for a particular resistance) so it was exactly at MPPT for one particular amount of illumination. Simple, and it would work reasonably well. But as amount of light varies the PV will not be at exactly Vmp/Imp so you won't get the full amount of power available. At half the light (half the available power), Vmp drops maybe 3%. But at 97% of the voltage, resistive element will draw 97% of the current, 94% of the power. PV can't deliver 97% of the current so voltage drops further, no longer Vmp/Imp. Half power would be delivered to resistor at 1/sqrt(2) = 0.71 times the voltage, but at that voltage PV array puts out maybe 85% to 90% of available power. You don't get all available power, lose a modest 10% to 15% I would estimate from graph below. If there are shadows on panels, further off optimum point.

A circuit that switches resistor on/off rapidly, while measuring computing power delivered and varying on/off timing to maximize power delivered, would harvest all available power. This complexity is not needed for initial design, which should work fine with resistance or array series/parallel configuration selected appropriately. But I think 10% or so additional power could be achieved with MPPT design.

 

[brum]

Here is an example for MPPT heater controller - https://projecthub.arduino.cc/steve...art-pv-mppt-solar-hot-water-controller-5264c7 . Details are pretty well explained.

 

[MyK3y]

You guys are still confused.

Panels in series wired directly to nichrome wire coil in a circuit. That's it. No other elements involved. I found a bunch of SSRs, so I will wire in a breaker just for safety.

So, further developments:

Just picked up aluminium tubing - 10M @ 50mm x 2mm - and redrew the diagram for the water jet. Cutting tonight. We've had a massive storm blew through for the last 6 or so days which really put a damper on things - I wasn't going to go out and wrestle with stuff in the high wind and rain when I could stay inside and read my book.

Nichrome arrives sometime this week, so I can start test wiring. I plan to make three elements in parallel, buried at three different levels in the sand. That allows me to run them hotter without the chance of burning out and are evenly heat the sand.

I was going to go through-hole with the baffle ducts, but have decided on flush-mount using exhaust gasket goo - good to 350C

46mm holes centred - the gap is to allow fuss-free routing of wiring, temp sender, etc.

 

[daklein]

Sorry I only got through first 4 pages of this interesting thread so far.

Skip sand or water, use white hot graphite. Resistance heating input, output thermal energy as infra red to liquid, and output electricity from the glowing graphite with special PV cells (cooled by liquid takes heat out too). Output controlled by mechanical shutters to stop the light getting out.... https://www.volts.wtf/p/why-electrifying-industrial-heat

I'm hooking up two 85 gal marathon water heaters as buffer tank for our home hydronic system. Air source heat pump split system provides heat, with upper tank elements from grid power as backup, and lower elements as excess solar dump load. House already had hydronic with Nat gas boiler, which I replaced last fall with the first Marathon tank and resistance elements were only heat last winter. Wood stove is primary home heat, but would also like to use that less.

That said, higher temp sand thermal storage is interesting.

 

[upnorthandpersonal]

Panels in series wired directly to nichrome wire coil in a circuit. That's it. No other elements involved.

As long as you can calculate the resistance required to match the panel's maximum output point, this works (think of it as impedance matching for maximum power transfer like one does with analogue telecom transmission lines). I've wired up solar panels in this kind of configuration as a test for some family, where we connected them to a heating element in a hot water buffer. You need to play around with series/parallel strings to find an appropriate match, or be able to change the resistance of the heating element (your nichrome wire coil) itself.

 

[daklein]

Geothermal heating would be perfect.
Move to Yellowstone, and live in complete peace and contentment.

Until....

 

[Warpspeed]

Until....

A good fuse should prevent any overload problem.

 

[rogerdw]

One of these came up for sale on marketplace recently ... a TSF32K ... basically a box of bricks interspersed with big oven type elements ... 4,571 watts with an additional 1,500 watt element. Rating 7 hrs ... 32kWh

Only wanted $100 but it weighs 176kg and it all seemed too hard so I let it go. ?

Anyway, seeing it made me think of this thread.

 

[OzSolar]

One of these came up for sale on marketplace recently ... a TSF32K ... basically a box of bricks interspersed with big oven type elements ... 4,571 watts with an additional 1,500 watt element. Rating 7 hrs ... 32kWh

Only wanted $100 but it weighs 176kg and it all seemed too hard so I let it go. ?

Anyway, seeing it made me think of this thread.

Very interesting. I don't think we have a similar sort of product available to us in the States but with the trend towards TOU rates across the country it might be time for one.

So an off the shelf product that stores 32 kWh worth of heat weighs almost 400lbs. Teasing that out for seasonal heat storage and calling that one day worth a of heat. A month would take ~30x's that much weight (and area) and a heating season would take about 100x's time that. (this is just back of the napkin math) Give or take that works out 40,000lbs to do that.

 

[Hedges]

I think the math shows it is reasonable.

TSF Turbo | Creda

www.credaheating.co.uk

"It retains up to 50% of its total stored heat after 17 hours, making it one of the most energy efficient storage heaters on the market today."

"

Charge Acceptance 7 hours

32.00

"

Units not quoted.

177 kg


Another model:

Creda TSRE050 Storage Heater 500W Slimline

WE WONT BE BEATEN ON PRICEWe regularly check our prices against our competitors, to ensure we can offer you the best deal possible.

www.electricpoint.com

"

• Heat output: 500W
• Input rating: 1020W
• Fan wattage: 0.5W
• Rated charge period: Up to 7 Hours
• Boost mode rating: 340W
• Maximum storage capacity: 7140Wh
• Storage element rating: 936/1020W

"
(According to a retailer)


What brick temperature do you guys come up with for it to do that?

Consider Silica,

Properties: Silica - Silicon Dioxide (SiO2)

Silica is one of the most abundant oxides in the earths crust. It exists in 3 crystalline forms as well as amorphous forms. It hasmany useful properties and is used in a range of applications such as silicon, elctronics, refractories, sand, glass making, building materials, investment casting etc.

www.azom.com

"

Specific Heat

680

730

J/kg.K

"



7140 Wh x 60 minutes/hour x 60 seconds/minute = 2.6 x 10^7 J
2.6 x 10^7 J / 730 J/kg.K / 177 kg = 200 degree K (or C) rise

Did I get that right, relatively modest temperature rise?
7.14 kWh, equivalent to operating 1800W heater continuously for 4 hours.

Back up the truck, and you could reasonably heat your house for a few months.

 

[Turnermjt]

Hi all,

i am new to this forum and based in the UK,

Wondered if someone could help me with some Maths please.

If I have an insulated container of sand approx 2.5 x 2.5 x 2.5 mtrs and over the summer months I believe I can heat it to 750c evenly using excess solar power.

If I then take the heat back out in winter by blowing air through a coiled tube In the sand.

How many hours/days would I get from that amount of sand at 750c if I was withdrawing 3 kws an hour??

Thanks Mark

 

[upnorthandpersonal]

- Mass of dry sand, about 1600 kg/m3
- Specific heat capacity is about 800 Joules per kg per degree delta
- 1kWh equals 3 600 000 joules

Your container: 15 cubic meter (rounded). Let's assume a delta of 750C - 50C = 700C

Energy stored: 15 x 700 x 1600 x 800 = 13 440 000 000 Joules, or 3733 kWh.

At a draw of 3kW, this will last 1244 hours, or about 51 days.

Important note: this assumes perfect insulation, which is not possible, thus a theoretical maximum. It will be very hard to insulate a box of that size at 700C. Remember that when insulating you need to keep the delta in mind between inside and outside the box. The higher the delta, the better the insulation has to be.

 

[Turnermjt]

thanks -much appreciated - if i extract the heat just at night, i should be around 80 days then realistically .

does anyone know how thick the ceramic wool would need to be to reduce the exterior of the steel container to 25c ?? i worry it is going to make it all too expensive.

how much bigger does the container of sand need to be to use the sand as an insulator??

thanks

 

[Warpspeed]

If I were attempting something like this myself, I think I would build a much scaled down working prototype to test and evaluate.
It would then be possible to make changes and improvements, and learn a lot in the process.

Whenever I build something that is highly experimental, afterwards I can always think of ways I could have made it better, simpler, cheaper.

Don't overlook using vacuum as an aid to insulation.
A perfect high vacuum is obviously not going to be possible or practical, but somewhat reduced air pressure in whatever type of insulation you finally decide upon may be beneficial.