Calculating Losses In DC Induction Furnace

I suggest that you monitor the power wherever you can.

Checking the output from the rectoformers would be a good place to start. It may be that they have different losses and thus for the same furnace output there will be different input energy. Remember to monitor both AC and DC since the rectifiers might still have ripple to the DC furnace coils.

You might also have iron losses in the laminates of the cores of any coil sets in the system, either from laminate thickness, material or air gaps from aging processes.

Finally, could be insulation of the furnace space allowing energy loss.

Thankyou for the reply !

Monitoring and calculating the energy losses is the main aspect of this project, i would surely see where i can calculate the losses in terms of energy..i have the data with me of one week for vtg,current,energymeter readings,temp,unit consumption per kg

can u suggest me which methods i can use to calculate the losses with these readings??

As simple as kWhr (ampsxvolts at uniform intervals) or energy meter reading consumed during each tonne of product, that will give you relative efficiency, actual efficiency is theoretical kWhr/tonne compared to actual, though maybe your loads are in kilograms, at 8,000 amperes per furnace.

Your process engineer will know theoretical values for your product, assuming all energy goes into the carbon.

The camera may help you improve efficiency, but having a good baseline of your current install is required to know if any improvements suggested by the camera are justified by the savings.

Agreed! ... and it does not necessarily require a $5k FLIR camera. I've solved many problems (and saved 10s of k$) with images from my FLIR One camera (≈$250) attached to my iPhone.

Thankyou for the reply !

yes it sounds like a nice idea to use a thermal imagery but 2 of 4 furnaces had their insulation replaced just an year ago still there are discrepancies and losses in those furnaces

Hello,

Just because insulation was replaced doesn't mean it was done well. Were there reading taken after that refurbishment/renewal to confirm the effectiveness of the outcome.

In elation to your response to my initial response, wherever you can measure and compare the results for the various furnaces, it's the comparison that will give you the possible areas of improvement.

Just because the insulation was replaced it doesn't mean that it was replaced correctly. You might even find out from FLIR imagery that some other paths not covered by insulation are cooling the inefficient furnaces. Use the FLIR imager to identify where there are heating and cooling differences between each of your furnaces and then analyze why those differences exist. Then and only then can you systematically work on optimizing these furnaces.

You should also understand that your original post contains a contradiction. An induction heater works by AC current producing large amounts of eddy currents in a magnetically coupled conductor. That AC frequency will be far higher in frequency than your mains power grid. A DC furnace is a purely resistive process that has little to no inductance.

Thermal imaging camera is very costly even for a small flir one camera for android but i have given the suggestion u provided to my superior and he said he will arrange the camera..i will check and update as soon as possible

As I indicated in post #3, my "FLIR One" camera for iPhone only cost $250, three years ago. I presume the one for Android has a similar price. If $250 is considered expensive, then these must be very small furnaces!

Using the camera to find a poor connection on one of our furnaces has easily saved that much in a single day, without even considering the cost of a rejected furnace load. The cost of that load is commonly tens of thousands of dollars.

Here's a sample image, showing a bad connection:

Lots of very good answers (suggestions) On top of these I would run an ultra sound scan this may detect electrical issues such as corona and arcing. Along with ultra sound a simple infrared thermometer usually a good one is under $100.00 may identify some hot spots that are heat losses. I would do this several times during a heat and with these tests mark the location of each scan for future reference and trending.

The loss is the input power minus the output power. The input power is the applied voltage times the current drawn by the load (for each furnace separately). The output power is the energy measured divided by time. Just make sure you have the units correct (i.e. KWA for 1 hour / 1 hour).

Good start, but you need to be sure that the work done is the same for each furnace. There must be a theoretical heat value for graphitization, ie btu per pound in English units, so your furnaces consume more than this value to do the job. This will help tell you if any of your instruments are out of calibration, and give you a check on load variation.

<...was assigned...can anyone...>

The person that made the assignment is the best source of guidance.

<...monitored them for 2 months...>

Clearly that person is not in the least bit curious about the time spent and the value of the outcome! One wonders what that really means...

I am confused. What is a DC induction furnace? Do you mean an induction furnace which is supplied DC to an oscillator to produce high frequency for actual furnace? What is a rectoformer? Is this a transformer and rectifier set that supplies the DC? If both of these guesses are correct describe the oscillator. That may be the source of your different losses. Leaking capacitor, inter turn leakage in inductor?

Thank you everyone for the response and i would like to update on the current scenario regarding the furnace.

The insulation of the furnace was changed based on the temperature rating and the total of 0.6 KW of energy saving was observed.

The Insulation was changed based on the heat losses and the plant head was not happy as i was not able to provide an appropriate solution for the electrical losses happening in the power circuit of the furnace.

I have attached the layout of the furnace power circuit if anyone wants to look into it.

I would appreciate if some one would suggest me possible ways i can curb the power losses in the circuit.

i have also added some data from furnace for reference.

Link for the data image - https://ibb.co/wzs0CtX

Link for the power circuit image - https://ibb.co/qrdpfvV

https://ibb.co/b75wCCN

I can post few links of the description of the instruments provided by the manufacturer of the transformer if that might help.

LINK: https://ibb.co/kyGNKRV

https://ibb.co/CVr42YY

https://ibb.co/QQty08K

After reading the text in those documents, and re-studying the main circuit diagram, I now see that the system can be supplied with or without the variable autotransformer. That is NOT obvious on first observation of the diagram (at least it wasn't to me). I assume your installation does have the control feature.

Like Redfred, I thank you, but even more for posting the links. Thery are definitely more legible than the CR4 images.

These transformer and other symbols are not the ones I'm accustomed to, and it's not at all clear whether there are magnetic connections between the main transformer and the booster and/or regulator.

The fact that those latter two exist, indicates that you will have a variable Power Factor, depending on some kinds of settings or conditions. The Power Factor can have a huge impact on energy measurement. Check the capabilities and settings of your energy meter.

I also agree with post#13. Induction furnaces use high frequency AC to induce the heating. DC (zero frequency) produces zero induction heat. Can you provide brand and model of furnace?

Thank you for responding dkwarner.

Firstly I would like to apologize for not making the furnace type clear.

The furnace used her is a DC induction furnace which is used for the process of graphitization and not the induction furnace used in steel industry,the furnace is heated by the means of graphite elements itself as no other element is able to withstand the high temperature which is between 2100 to 3500 degree/Celsius.

So u can say furnace is made on the site for the process of graphitization.

I am attaching the link for the picture of the furnace for reference.

LINK : https://ibb.co/fNSgwTD

https://ibb.co/Hpq0B5T

https://ibb.co/z4stzdb

https://ibb.co/SdTthBY

Ok. In U.S. terminology, at least, that is NOT an induction furnace. Those graphite elements are very large resistors, and it is a DC Furnace.

I presume the process of graphitization refers to the conversion of some other form of carbon into graphite.

The furnaces I'm most familiar with are vacuum brazing/heat treating furnaces. At least one of ours does use graphite heating elements, but all of ours use AC, not DC for the heating current. At the moment, I see no reason for converting the AC to DC in your furnaces, and that conversion does waste some energy.

On the other hand, since your furnaces are powered by transformers, they will inherently have an inductive power factor. Adding appropriate capacitors can significantly raise your power factor, and thereby significantly reduce your energy cost.

Do some searching for Power Factor Correction Capacitors, here on CR4 and elsewhere.

Thank you for the reply dkwarner !

I am based in INDIA and there are only two majority Graphite Industries in India and ours is one of them, i agree we are technologically behind in the graphite production as the setup is atleast 10 to 12 years old in few of the plants here.

So as for the power factor is concerned APFC has been installed on the incoming supply from the power supplier to maintain the power factor unity.

By suggesting to add the capacitor, do u mean adding the capacitor after the dc conversion and before it is given to the furnace ? Or before it is given to the rectoformer?

HA! 10-12 years isn't that old! At least a couple of our furnaces are over 30 years old. They HAVE been upgraded and maintained...

I have limited knowledge of APFC, but I understand that it is more efficient if installed separately at each furnace/device. The capacitors I referred to are the ones in one or more APFC units. To my knowledge, there do not exist capacitors capable of smoothing 9kA at 12V or even 4kA at 24V.

I haven't yet succeeded in convincing the leaders of our plant to invest in APFC

Do you have any idea why your system uses DC on the heating elements? With 6 diodes effectively operating on 6 phases, your ripple will be pretty small, but definitely present.

Also i would appreciate if u tell me if there are furnace available in market, which uses AC to heat graphite elements which may achieve temperature upto 3100 degree celsius. As u said they are being used in US ,as i would love to suggest them to the management as a possible upgradation on current furnace which i agree lose lots of energy in conversion.

Unfortunately for you, all of my experience with large furnaces is with vacuum furnaces. In a vacuum furnace, almost all the air is removed before starting the heat cycle, so the heating elements don't oxidize much. For what it's worth, ours are made by Ipsen. https://www.ipsenusa.com/

Yes, the furnace used in our plants are also vaccum furnaces, as the graphite electrodes are to be prevented from being oxidised to increase their productivity.

And please provide me some time [like it is 3 am right now please give me few hours] to give u detailed answer regarding why DC is preferred in our plant over AC for heating.

Of course! It is 1:30 PM here, and currently snowing hard! Get some sleep!

the only reason someone has come up for the DC is its properties and its advantages over AC for heating purpose

"the only reason someone has come up for the DC is its properties and its advantages over AC for heating purpose"

Which properties and advantages are those?

If you were heating a fine wire, as in a small incandescent lamp, there could be some noticeable flicker (brightening with each half-cycle peak of the AC and dimming near the zero crossover), but with these massive graphite heaters, the thermal mass will take hundreds of cycles to heat up or cool down noticeably.

The only possibility I can think of is if the varying magnetic field of the heater current were to influence some other component of the furnace or its load. I know very little about the graphitization process. I'm assuming you are referring to graphitization in the sense: "Graphitization can be defined as the structural change from highly disordered or defective carbon atom structures towards a perfect 3D crystal", rather than graphitization of steel, which is normally a bad thing.

If so, I can conceive of AC magnetic fields interfering with the crystallization process, but I can also conceive of the AC fields helping to jiggle atoms into the crystal structure.

Does your furnace have 1 electrode (we would call it an element or resistor) or several? If one element the advantage of DC is to equalize load on all 3 phases. If multiple elements one could put some on each phase. I have not seen graphite elements (1500 degree C is higher than we go) but we often use silicon carbide elements similarly.

A problem we have (and I would expect same with graphite) is that as elements age their resistance increases. This results in different power characteristics.

It has two elements one for positive and one for negative

I suspect we have a translation problem here. In our terminology, a heating element is whatever object acts as a resistor to convert the electrical energy into heat energy. Most three-phase furnaces have those heating elements in sets of multiples of three, so an equal number of heating elements gets power from each phase.

As I indicated in an earlier post, but you have neither confirmed nor denied, I now suspect that your single rod being graphitized IS itself the only heating element.

That would be really efficient, as long s you have a good method to connect the electrical source to the rod, because there would be no need to transfer heat from other heating elements to the furnace rod.

If that it true, then what you meant here is that your furnace has two contacts or connections to that single rod

Yes! The power factor is likely the root of this puzzle. I suspect the numbers posted earlier are actually the apparent power and not the real power drawn by this system. It is also very important to know where in the power distribution these power readings are taken.

As Rixter mentioned in your other query, a more advanced rectifier topology (like a Vienna Rectifier) can both perform power factor correction and output power regulation. This is not a simple redesign that can be handled over a public engineering forum. A redesigned supply by an outside firm may be your only option. They might be able to reuse your original transformer but this can easily extend your downtime at the same time.

However, a cost versus benefit discussion can be started here. Assuming that this is a power factor problem and that you are being billed for apparent power consumption, the difference column in your chart is the most you can hope for by improving the power factor. Will this be worth the upgrade costs?

The readings were taken on the load side, energy meter reading is from the transformer and there is control panel by which the current and voltage is adjusted manually according to the load,it controls the current and voltage given to the furnace for heating and it is increased or decreased based on the diameter of the graphite rod and the reading is noted by the operator every hour i have attached the link to the sheet of measurement

Link : https://ibb.co/5K2WPrZ

There are many kinds of energy meter; some take power factor into account, and some don't. We currently use a Dent Elite Pro, which does take power factor into account for up to four phases.

Interesting! I gather the rod you are graphitizing IS itself the heater. Some of our furnaces use graphite bars as the heating elements, but our loads are always metals.

yes but how m i suppose to measure DC power factor,i assume u are telling power factor before conversion to DC??

See post #33

I think I just realized WHY you need DC! Since your load is your (single) heating element, you must supply the current at the ends of the rod. That means there must be diodes to prevent feedback from one AC phase into another.

We always have heating elements in multiples of three, so there is effectively a separate heating element for each phase pair.

The furnace is operated in on DC supply and the energy is supplied by rectoformer so i assume power factor should not be an issue in DC.

Forgive me if i am not able to get any point you are trying to make.

NO! The rectoformer is powered by AC, and is a highly inductive load for the AC. It is there that the power factor comes in to account.

You are right that there is no power factor in DC, but your company isn't paying for DC; it is paying for the AC coming in to the rectoformer. Depending on how your electric utility charges, there is probably a large penalty for low power factor.

You need to be measuring the energy at the lines feeding into each rectoformer, if you want to compare one unit to another. Since you say you have a single APFC unit for the whole plant, The only measurement that will be valid regarding energy cost is a power measurement on all 3 or 4 lines leading into that AFPC unit. Again, that measurement must be taken with an energy meter that accounts for power factor.

While taking those measurements, pay attention to phase balance. Any significant difference between the power in any of the three phases indicates a problem.

There is a combined APFC of 500 KVAR for all the four rectoformers..i checked with engineer's and they told that a separate APFC is being used for this plant which is as said above