Hello, new here and need some help.



I am trying to replace a method of heating a graphite rod. Currently we use a very inefficient and expensive open air propane heating and I would like to try and replace it with an electrical heating system.


I have a graphite rod of about 25kg at room temperature and would like to heat it up to 1000 centigrade. The quicker the better.


I need to get about 18 MW into the thing (about 5kWh) to reach about 1000 centigrade.




So far the only reasonable idea I have had is to buy/build a furnace with some nichrome wire (or similar) and heat the thing up slowly over a few hours.


I don't think induction heating would be suitable or easy to get a rod that size to 1000 centigrade.
Arc heating would be difficult to get that much energy in there in any suitable time plus potentially dangerous.




I remembered today as a kid I used to short batteries with wires and they would glow hot very quickly. Would this be possible with the graphite rod? Just pass a huge current though it?


Now my electronics isn't very strong but I studied physics so know a little and can read/learn about it.


The “problem” I see is that the graphite rod is low resistance and hence will not heat up easy as power dissipated is current squared times by resistance. Now I would like 25kWh of power into the rod to heat it up in about 12minutes. The resistance is tiny. Some 0.001 ohms. That means I will need about 5,000 amps which is HUGE.


I take it I would need a large drop down voltage transformer and it should be possible.






Any ideas about this or another method or comments/suggestions more than welcomed.


Ideally I would short a 100 volt line with this rod. 100,000 amps through it. 10MW power. Should take 2 seconds to heat to 1000 centigrade. How cool would it be to draw 10MW even if only for 2 seconds but something downstream is likely to blow or melt before the graphite rod reaches 1000 centigrade.


So is there any easy way to heat this rod up quickly? Would a bunch of car batteries do ok (don't know where I can find a drop down transformer that can handle 25kw)? would a bank of capacitors work?

 

Have you considered induction heating?

John

 

See if you can find a surplus submarine battery. Lots of volts and amps in one of those.

McMaster-Carr used to carry an oxy-gasoline torch that would cut 150 lb rail in less than one minute.

 

You're kind of in a conundrum here. You want to heat the graphite up as quickly as possible, but yet, you want to restrict power flow into the rod.

The rod will require (1000-25)*0.79*25*1000J or ~19.25MJ of energy assuming a 100% efficient process. This doesn't mean that you need 19.25MW to heat this up, although it would if you needed it to be heated within a second.

You need to create requirements of how long you are willing to wait to heat this graphite up. The resistance isn't a big factor, since it is magnitudes larger than copper. You can use a constant current or voltage source to heat the copper, although I would stick to constant current because the voltage needed would be low and the temperature will have an effect on the graphite's resistance.

So, you're needing to decide on how much time you're willing to take to heat up your graphic. Depending on the dimensions, you might need an excessive amount of current, but that should cause a limitation on time alone. Clearly, you would want to heat it up as soon as possible because of the power lost through heat transfer, thus making an inefficient process.

What are the dimensions of the graphite? You should apply the current so that it takes the longest path possible. If it isn't uniform, you may have a problem with hotspots.

Steve

 

Something just occurred to me that might apply for your application. I happened to be there when an engineer tried something called a Halogen torch. I believe it was around 1KW (5KW?) in a single lamp, it would scorch paint off the wall. As it happened he was trying this on a ceramic substrate, I watched this sucker flex like paper before it shattered into a thousand pieces. I also remember something like this being tried out as an alternative to microwave cooking (maybe a suppliment). Definately an alternative to search on.

 

They use graphite rods in arc furnaces, yes? And for carbon arc welding? Are you opposed to heating two rods at once? "Rubbing sticks together?"

 

You're kind of in a conundrum here. You want to heat the graphite up as quickly as possible, but yet, you want to restrict power flow into the rod.

The rod will require (1000-25)*0.79*25*1000J or ~19.25MJ of energy assuming a 100% efficient process. This doesn't mean that you need 19.25MW to heat this up, although it would if you needed it to be heated within a second.

You need to create requirements of how long you are willing to wait to heat this graphite up. The resistance isn't a big factor, since it is magnitudes larger than copper. You can use a constant current or voltage source to heat the copper, although I would stick to constant current because the voltage needed would be low and the temperature will have an effect on the graphite's resistance.

So, you're needing to decide on how much time you're willing to take to heat up your graphic. Depending on the dimensions, you might need an excessive amount of current, but that should cause a limitation on time alone. Clearly, you would want to heat it up as soon as possible because of the power lost through heat transfer, thus making an inefficient process.

What are the dimensions of the graphite? You should apply the current so that it takes the longest path possible. If it isn't uniform, you may have a problem with hotspots.

Steve

thanks for the post steve, one of my concerns was that the current would take the path of least resistance and i think graphites resistance goes down as temp goes up so i may end up with one really hot streak where the current is flowing.

that would probably shatter the rod. no harm in trying though.


constant current instead of constant voltage is also something to consider.

 

Something just occurred to me that might apply for your application. I happened to be there when an engineer tried something called a Halogen torch. I believe it was around 1KW (5KW?) in a single lamp, it would scorch paint off the wall. As it happened he was trying this on a ceramic substrate, I watched this sucker flex like paper before it shattered into a thousand pieces. I also remember something like this being tried out as an alternative to microwave cooking (maybe a suppliment). Definately an alternative to search on.

thanks, looking for halogen torch on google i came across this

http://www.noblelight.net/products/infrared/infrared_industrial_heaters/shortwave_water-cooled.asp

seems perfect for what i require.

 

This is outside my area of expertise, but it would seem to me that one of the main problems is that the rod will radiate energy when heated to this temperature. Basically the black-body radiation is going to represent a power loss. Wouldn't you expect to need an insulated chamber, like a kiln?

 

This is outside my area of expertise, but it would seem to me that one of the main problems is that the rod will radiate energy when heated to this temperature. Basically the black-body radiation is going to represent a power loss. Wouldn't you expect to need an insulated chamber, like a kiln?

Yes it will radiate a lot of heat but the hope is heating it quickly hence reducing the amount of energy lost.


An alternative is a large insulated furnace heating the thing up slowly over a few hours.


Problem with slow heating is that the graphite burns away (carbonizes) so fast heating has the advantage that the graphite doesn't burn away so much.

 

thanks for the post steve, one of my concerns was that the current would take the path of least resistance and i think graphites resistance goes down as temp goes up so i may end up with one really hot streak where the current is flowing.

that would probably shatter the rod. no harm in trying though.

constant current instead of constant voltage is also something to consider.

You're welcome!
I think the current profile would need to be experimentally derived. Isn't graphite a decent thermal conductor? You would need to hold off of the power a bit to account for thermal lag, but it should even out okay. I'm not sure what kind of quality of process you need, since I don't know the application.

Maybe induction heating, as John mentioned, may be a good option. The halogen idea is okay, but I don't like the fact that it would be heating from the outside-in. You would need a minimum of two of these. Also, these would probably need to be replaced often.

Steve

 

Yes it will radiate a lot of heat but the hope is heating it quickly hence reducing the amount of energy lost.

Would not black-body radiation represent an instantaneous power loss? No matter how fast you heat the object, it will radiate electromagnetic power according to it's temperature. The final temperature will be based on the equilibrium of radiation output and electrical energy input, assuming thermal conduction/convection effects are neglected.

A good example of this is a common incandescent lightbulb. So your approach needs to provide enough power to allow steady state blackbody radiation at 1000 C. You should be able to estimate this power level. Then, you can calculate the voltage and current needed to achieve this. Assuming these are achievable levels, you will then have one very powerful infrared light bulb.

For comparison; the principle of the kiln is that the surrounding air and chamber allow thermal conduction to counteract the blackbody radiation. Or, perhaps a better way to say it, is the blackbody radiation is recaptured to heat the surroundings and help slowly raise the temperature.

 

steveb,

Great call! I didn't even think of the radiation-mode component, because it is usually not as significant as conduction or convection. Especially in a process that is this quick.

Graphite turns out to be an excellent emitter of radiation, which will make it extremely difficult to heat up without insulating (reflect a better word??) it well. Check out this information and calculations:

http://www.egglescliffe.org.uk/physics/astronomy/blackbody/bbody.html

Steve

 

Graphite turns out to be an excellent emitter of radiation, which will make it extremely difficult to heat up without insulating (reflect a better word??) it well. Check out this information and calculations:

http://www.egglescliffe.org.uk/physics/astronomy/blackbody/bbody.html

Steve

Steve,

This is a nice reference and provides the power formula directly. I wasn't looking forward to digging out my old solid state physics book.

Cells,

Assuming we are thinking about this correctly, the power formula Steve gave should provide you a reasonable estimate of the minimum required power. You should have the details for your graphite sample; surface area and resistance. You can then estimate the electrical power and current to heat this directly with current. I ran some quick estimates and it appears your original comments and concerns were on target. Power would approach 100,000 W and you would need very high current of the order of 10000 Amps and very little voltage of the order of 10 V.

I suppose this is possible to achieve, but it won't be easy, and will be expensive. I'm almost certain you will have to implement parallel IGBTs to do this (if you want DC and/or controlled output). Hopefully, there is someone here that can guide you about the details of implementing such high current drivers. If you are interested, I may be able to get some part numbers for IGBTs from a friend that works with high power systems like this.

 

Industrially, graphite is commonly used as a susceptor for induction heating. A "susceptor" is used to indirectly heat material which is not itself suitable to induction heating. The susceptor absorbs the energy from the induction work coil and convects or conducts the heat into the workpiece. Temperatures as high as 3K C are attainable in practice.

Cells, what are the dimensions of your 55 pound hunk of graphite?

 

Have you considered induction heating?

John

Yep...definitely the way to go. At the UCLA plasma physics lab, we created the highest man-made temperature (outside of a nuclear reaction!) using an inductively coupled R.F. plasma. (55,000 degrees C!)

Your job will be much easier, as inductive heating apparatus is commerically available

eric

 

Steve,

This is a nice reference and provides the power formula directly. I wasn't looking forward to digging out my old solid state physics book.

Cells,

Assuming we are thinking about this correctly, the power formula Steve gave should provide you a reasonable estimate of the minimum required power. You should have the details for your graphite sample; surface area and resistance. You can then estimate the electrical power and current to heat this directly with current. I ran some quick estimates and it appears your original comments and concerns were on target. Power would approach 100,000 W and you would need very high current of the order of 10000 Amps and very little voltage of the order of 10 V.

I suppose this is possible to achieve, but it won't be easy, and will be expensive. I'm almost certain you will have to implement parallel IGBTs to do this (if you want DC and/or controlled output). Hopefully, there is someone here that can guide you about the details of implementing such high current drivers. If you are interested, I may be able to get some part numbers for IGBTs from a friend that works with high power systems like this.

Yep I know about black body radiation and knew about STEFAN'S LAW ect.


Not really a problem if you want to rapidly heat something but if you want to maintain a temperature then you would lose a lot of power due to the power of 4 relationship.



Just a quick update, I have installed some thermocouples onto the rod and measured a typical 2 hour propane heat (we currently use open oxy-propane where some 5% of the energy goes into the rod and some 95% wasted into the atmosphere). Was able to figure out that in the first 25-30minutes the rod reached a steady state temp of about 800 centigrade and the other 90min where really wasted. So we have changed standard practise to heat the rod for only 30-40minutes instead of 2hours which is a massive saving already (someone must of picked 2hours min time years and years ago, such as waste. Looking at some £3m (~$5m) wasted over 10 years by this simple error. Oh well, better late than never.




I think the original idea of just blasting a massive current through the thing isn't feasible especially considering I would need a BIGG transformer to ramp the 440 volts on site down to some 10 volts.




I am now leaning towards using a simple Nichrome heating system to heat the rod up. The rod and heating elements enclosed in an insulated housing.




I am also working on a easy robust igniter and I think nichrome would work with that too. Currently our guys leave “pilot” flames on all the time which they use to light large “hand scarfers” which are hand held propane burners. The problem with the “pilot flame” is that they are HUGEEE. Think of a flame about 30-50cm high and 10-20cm wide, probably costing more than £10k ($15k) per year per one and there are probably a dozen of them.


Any ideas on simple “pilot flame”. I was thinking nichrome wire with a foot push switch. The step on it and it heats the wire up to 1000 and they use that as their pilot “flame” to light their hand held propane burners. Could be done easy I think and work off simple AA batteries.


Any other ideas?

 

Cells, what are the dimensions of your 55 pound hunk of graphite?

Its a tube about 700mm high and 200mm diameter. Inside is hollow about 100mm so the wall of this graphite tube is some 50mm thick.

 

Thanks for all the help so far, this “project” was to see if we could cut down on the various gases used in the factory. Propane was the largest one costing some £1.6m (~$2.5m) per year and heating this graphite part was the largest chunk at £0.6m so it was the first one I tried to tackle.


Some £0.4m pa was saved just by figuring out that we don't need to heat the dam thing for 2hours and 30min would do.


An electrical heating method would save a further £0.18m pa and using an electric “pilot light” would save some £0.12m pa


its surprising the amount of waste we had/have.


Thanks again.

 

Any ideas on simple “pilot flame”. I was thinking nichrome wire with a foot push switch. The step on it and it heats the wire up to 1000 and they use that as their pilot “flame” to light their hand held propane burners. Could be done easy I think and work off simple AA batteries.[/FONT][/COLOR]

Pilotless furnaces are extremely common in the US. Some use a heated wire, others use a spark. The spark can be like an automotive or other HT ignition or from a piezo device. I would simply hack one of the piezo lighters.

John