Sizing Induction Furnace Water Cooling

Long story short - the water chiller for my 100 kW induction furnace hasn't survived storage and I need to replace it, but I want to check that the size is reasonable.

Original was 11 kW consumption 39 kW chilling - I can't get advice from the manufacturer so would like to find comparisons.

As I say it's a 100 kW driver (by CFEI of France), and the furnace body is an inverting Raydyne unit - driver, interconnecting cables and crucible unit are water cooled - relatively small crucible about 4 inches diameter.

I'd like to find ANY comparable to get a handle on the size of cooling unit used - bigger or smaller they can be pro-rata'd

Out of curiosity, are you planning for heat pump, forced convection (dry tower) or evaporative cooling (wet tower)?

The one that failed used a huge refrigeration unit like the one in the back of your fridge but ten times the size.

(Good to hear from you Mark, I hope that you are well in these troubled times)

The ASM Metals Handbook, Vol.15, Casting, suggests 10% heat loss for a channel furnace and 20% heat loss for a coreless (helical tube around the crucible) furnace.
Your original chiller was setup to handle a 39% heat loss. The reserve capacity may be needed to absorb additional losses from a mismatched coil. Or the refrigerant compressor was over sized to allow it to run on a duty cycle rather than continuously.

If it is a coil type furnace there are other constraints. The coil inside diameter, length, and maximum operating pressure determine the maximum flow rate through the plumbing. The coolant pump specification tag will show the flow rate at the design pressure.

The temperature rise for the coolant can be calculated based on the flow rate and maximum anticipated heat loss of 39 KW. The use of a chiller suggests that the system required a below room temperature coolant at the inlet to prevent boiling inside the copper coil.

Otherwise, a less expensive and more reliable alternative is a tube type heat exchanger for the coolant chilled using the municipal water supply. Some cities do not allow this type of use or have high sewage or water fees that make it too expensive.

The requirement for deionized water as the coolant will limit the choices for the recirculating pump and heat exchanger construction.

It may be possible to salvage the pump, heat exchanger, and other parts from the original unit for reuse. Or just replace the chiller's compressor. A 39 KW chiller will be less expensive to repair than it is to replace. The exception would be if you can locate a working unit in the industrial surplus market.

What happened to the original unit? Get a refrigeration guy to repair/recharge it.

Original is about 2000 year of manufacture, was stored from 2008 to a few months back, the Grunfos circulating pump had failed - I rebuilt it with expensive new ceramic seal. The original had a very odd varnish like tide mark where the water had dried out that stuck the seal up and destroyed it. We now think that the varnish like substance is the sealed compressor unit oil that has leaked from a dissolved heat exchanger that is immersed in the internal water tank resulting in the compressor seizing. The compressor internal motor is now stalled and drawing 90 amps per phase, 415 volts. if we could source a replacement sealed compressor unit which would be extremely expensive, the system is designed for a gas type that is now banned!

So the original is a right off I am advised

You can also end up with some weird reverse plating if you aren't using deionized water, at least with inductotherms. I can't imagine that is universal, because it would be stupid.

But I know we could ONLY use DI water in a 45 kw inductotherm.

Yes when I had this running 13 years ago I used de-ionised water laced with Glycol to prevent freezing in winter. In fact I still have a 205 litre drum of the coolant drained out when I moved 13 years ago. Frugal you see !

A few numbers:

The heat capacity of water is 1.0 BTU/lb-F

or
Cp= (1.0 BTU/lb-F)(.293 Watt-hr/BTU)(62.4 lb/ft**3)(.1337 ft**3/gal)(60 min/hr)= 147. Watt-min/F-gal

The 3/8" diameter copper tubing on the furnace can supply 5. gallons/min of cooling water

The temperature rise for the cooling water at 5. gal/min for a 39 KW power dissipation is

Delta F= (39 KW)/(5. gal/min)(.147 KW-min/F-gal) = 53 Deg F

If a tube type heat exchanger is supplied with city water at 50 Deg F. the furnace coolant loop will have a inlet temperature of roughly 70 Deg F.
The outlet temperature of the furnace coolant loop will be 70 + 53 = 123 Deg F.

If a air-water heat exchanger is used in your climate the coolant loop inlet temperature will be roughly 150 Deg F. (Assuming 50 Deg F inlet air and a large fan)
The outlet temperature of the furnace coolant loop will be 150 + 53 = 203 Deg F.

There is no risk of the coolant loop water boiling in either case.

The only reason to use a refrigeration system on the furnace is if you need to use the rejected heat to warm the shop during the winter.

If the furnace is only being run a few hours each day the least expensive solution is a 1000 gal plastic water storage tank supplying the tube heat exchanger.
The tank can be buried to remove heat quickly overnight. The 1000 gallon tank will increase in temperature by 48 Deg F if the furnace is run for 3 hours at the maximum heat load of 39 KW.

My suggestion is to remove the refrigeration compressor and freon heat exchanger from the water chiller and replace it either with a tube type heat exchanger or a radiator. Either part should be available from a junk yard.

Thanks for the figures Robert.

The furnace has 3/4” tubing
The driver and furnace loop needs a flow of 27 litres a min
Max loop temp is 40 C but I want to keep it lower than that to give the electronics an easier time
Max flow I can get to a heat exchanger cold side is 15 litre per min from mains water so I am investigating a bore hole that I happen to have.
A 1000 litre IBC will over heat in 27 minutes

You've got a farm. Dig a cooling pond.