In e-mails and in these fora, ALL CAPITALS is taken as shouting, and there is no need to shout; we can hear you very well!

Is your question about whether it is feasible to supply an electric arc furnace from Diesel generators as an isolated load, meaning not connected to any other loads, or are you asking whether such a furnace can be supplied via an existing system supplied exclusively by Diesel-engine-driven generators?

Providing more information yields better answers!

CR

"As iron sharpens iron, so one person sharpens another." [Proverbs 27:17, NIV]

Dial it down first, will ya?

Muthu
www.edison.co.in

The answer is MAYBE.

Arc furnaces can be very noisy, and some generators don't like that.

Could you expand on that, cranky? From what I know of electric arc furnaces, [a] starting one is not far removed from creating a dead short, [b] they draw significant amounts of lagging reactive current and therefore have a very poor power factor, necessitating voltage regulators capable of handling such severe and abrupt duty, and [c] they generate great honking amounts of scatter harmonics that have to be addressed with things like smoothing reactors, filters, and/or similar "flicker fighters" . . .

I've gathered that much of the newer generator technology is highly susceptible to performance degradation and/or reduced service life where these conditions exist, whereas older technology not only performs better but can often be rendered relatively immune to such conditions.

CR

"As iron sharpens iron, so one person sharpens another." [Proverbs 27:17, NIV]

If you look at the transformers on a modern arc furnace they look like something out of the 19030's, and for good reason.
CR you are correct, there is a ton of electrical noise to soak up.
The voltage and current swings during the initial stages of melting are huge.
That is why these usually have dedicated sub-stations, the rest of the mill gets power from another source.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

I was involved in a number of rental projects where the primary load was an arc furnace. It was a pretty ugly experience and it required a good bit of work to make it successful for the time required for the utility to finish up their substation work.

As I remember everything had to be way over-sized, one job I still have a few details on,

4 groups of 10 2MW generators feeding utility supplied 40MVA transformers each,
To get first unit on line for each group we had to do a dead gen/dead bus gen breaker close and ramp the generator up, then the other 9 units could be paralleled to make the group "ready".

When first two groups up and running some auxiliary plant equipment was brought online, not sure what it was since the plant was about 2 miles away, but I do know whatever that equipment was it was very tolerant to a low voltage condition.

Third and then fourth step up transformers were energized by the two groups already online, then the remaining generators were brought online.

With all four groups of generators online, the arc furnace start sequence came on, for that "first start" we had to have a second under/over voltage settings group active to allow for the large swings, once arc was established and stable we could slowly bring units offline and plant would bring on other aux loads. We would usually end up with 12-15 units remaining online at about 40-50% kW and about 80-90% kVA load. Technically we were off the reactive capability curve but the factory engineers felt we had enough "extra" in those tail ends at the time.

Every Monday morning at about 2am we did the start, at midnight on Friday they did a stop sequence that took about 4 hours.

On that particular job we failed 12 AVR's and 2 generator ends over a 2 month period. At that time our 2MW portable units had a reputation of being very rugged and reliable, it was one of the worst rental jobs we ever had, but we ended up doing 7 or 8 of them over the next couple of years.

As I understood it these were fairly small furnaces, mostly dealing with scrap recycling in the USA midwest.

I seriously doubt that the current model 2MW portable power units by several companies would be able to deal with the harmonics, low power factor and load instability we saw on the units we used at the time, at least in my opinion.

Hope that helps, MikeL.

Hey elec777, I'm guessing this post is a follow-up to https://www.eng-tips.com/viewthread.cfm?qid=448233 and that you didn't have a lot of success the first go-round . . . true?

CR

"As iron sharpens iron, so one person sharpens another." [Proverbs 27:17, NIV]

Mike; Where the transformers selected with higher than normal %imp? Would higher $imp transformers make life any easier for the gen-sets?

Bill
--------------------
"Why not the best?"
Jimmy Carter

Bill,

The utility who was our customer drove most of the engineering and selection of the transformers.

On the first project we did with them we used transformers more closely matched in size to the gen group output, 25MVA, but the instability was severe and there were some other issues that while initially the utility threw it at us saying our units weren't capable enough, all the sudden new transformers showed up, got connected and we got in service without too much more trouble.

It was on that first job we figured out we had to go to a way better protective relay, the engine control panel protections just weren't up to the task, so we installed SR489's on each unit, with two setpoint groups, also implemented stator and bearing temp protections due to the low power factors. We also had to retrofit the units with Basler SSR AVR's as the early version digital regulators died off pretty quick.

What we learned from those jobs led to development of a "utility grade" module, wasn't a perfect match for everyone's need, but it did live thru some really crappy projects. On the second job the utility tried adding some cap banks, but the control of them was poor, and in some cases came on too late or early and added to our grief. Likely nowadays there are some way better controls or power electronics that may help with some of the problems we saw, but when we did those jobs we were pretty limited. Those jobs ran from 1997 to 2005.

MikeL.

Thanks Mike.

Bill
--------------------
"Why not the best?"
Jimmy Carter

What happened to elec777? Post & ghost?

CR

"As iron sharpens iron, so one person sharpens another." [Proverbs 27:17, NIV]

My understanding is arc furnaces require supply from a system with substantial short circuit power.

what I understand from the answers:
1- most of the colleagues are not in favor of an isolated system of diesel generators.
2- am thinking about turbines, I mean isolated combined cycle power plant has 2.5 power capacity
more than the EAF.
3-the turbines will be equipped with a highly responsive speed and efficient governing system(wadward)
4- step up transformers will be with low impedances
5- today we have the supper capacitors and control systems, means that we can solve the big need for reactive power.
6- I agree that the more the short circuit power of the source, the more the power source capable of handling the fluctuating loads.
thank you

dears,
I found this paper gives good ideas regarding the case we are discussing.

• https://files.engineering.com/getfile.aspx?folder=a1ab033c-4938-4e4c-9235-3

How much short circuit current do you need?
A higher impedance transformer will limit the short circuit current to lower levels and cause less disruption to the supply grid.
When the load is a short circuit there is not much energy being transferred to the load. Most of the energy is lost in the transformer windings.

Bill
--------------------
"Why not the best?"
Jimmy Carter

The big advantage that taking supply from a large grid has over islanded generation is that the inertia of the large system is very much greater than anything you can contrive using islanded generation. The energy stored in the rotating masses can handle the wildly fluctuating load of the EAF without the machine governors swinging in a similar fashion.

You might consider supporting local generation by operating in parallel with the utility - this arrangement used to be commonplace at aluminium smelters and the like: the load was met locally, but the grid was there to handle the load swings which exceeded the capability of the local plant.

THANK YOU SO MUCH SCOTTYUK,
(You might consider supporting local generation by operating in parallel with THE utility)THIS IS WHAT I
PERSONALLY PRACTICED IN A PLANT THAT HAS 20MW STEAM TURBINE FOR LOCAL GENERATION WORKING IN PARALLEL WITH THE GRID, BY ADJUSTING THE LOAD SHARE CONTROL OF THE TURBINE, I COULD DERIVE 3MW FROM THE GRID AND THE REST FROM THE TURBINE TO SUPPORT A LOAD OF 17MW AND ALL WENT WELL.
NOW, THE QUESTION IS :
CAN I APPLY THE SAME ABOVE MENTIONED SCENARIO TO FEED THE EAF, SO THE THE PP AND GRID TOGETHER CAN HANDLE
THE FLUCTUATING LOAD OF THE EAF?

You should be able to operate in parallel, especially if the grid connection is large relative to the local generation. You'd run your machines in MW control mode, and likely set the AVR to keep the machine neutral on the grid from a reactive point of view. The grid will meet the fluctuating load. I would suggest you'd need some impedance to de-couple your generators from the EAF, i.e. a GSU transformer onto the transmission system, and a separate transformer from the transmission system to the EAF, but you need to get a proper load study done - not rely on opinion from the internet!

Whether you can sell this idea to the transmission system operator is a different question, I can't see the TSO being especially happy supplying a violently unstable load and even less happy if they receive little revenue stream from accepting that load because you're running internal generation.

Scotty. How do you feel about transformer impedance for this application?
Some seem to be suggesting a low impedance to provide short circuit current.
I feel that you may not need fault levels of current and that a higher impedance transformer will supply adequate current while absorbing a larger part of the violent load swings.
Does an EAF really need high fault levels of current?

Quote (Scotty)

I can't see the TSO being especially happy supplying a violently unstable load and even less happy if they receive little revenue stream from accepting that load because you're running internal generation.

I can see the TSO addressing this with onerous demand charges. Possibly instantaneous rather than a 15 minute average.
Again higher impedance transformers will help if they are suitable for AEF service.

Bill
--------------------
"Why not the best?"
Jimmy Carter

Hi Bill,

Straying a little outside my normal world, but the TSO's priority will be to minimise the disturbance on their system, which means having sufficient transformer impedance to decouple the EAF from the transmission substation. With (I'm guessing a little here) 15% - 20% impedance you're looking at a substantial fault current on the LV side and if there's a fairly direct connection to the transmission system then the EAF transformer impedance is going to be the dominant factor in determining the fault current.

Hopefully someone like prc will have a bit more knowledge of EAF transformers themselves, but my understanding isn't so much that the fault level is abnormally high as the transformer and ancillaries like the tapchanger are designed to handle sustained and repetitive fault-level currents on a more-or-less continuous basis.

Thanks Scotty.

Bill
--------------------
"Why not the best?"
Jimmy Carter