Initial Question:
We are having an issue with one of our furnaces that is completely stumping both me and my maintenance department.

We are experiencing extremely high dew-point levels – our analyzer faults out at +69°F (20°C) in the heat chamber, and we are finding high oxygen levels (up to 0.52%) in the cooling chamber. We have probed the furnace with a flammable gas detector and have not been able to find a leak. If we drop the mix out of the furnace and have only the process nitrogen used during the purge cycle in the furnace, both the dew point and oxygen readings are good throughout the furnace.

We had this issue on another furnace and found a fitting in the incoming mix line that was cracked. Once this was replaced, the issue was gone. We thought that we had a similar issue with this furnace but have been unable to determine a leak. We have plugged every section of the plumbing and pressurized the system for several hours with no pressure drop. We are unable to pick up anything with the flammable gas detector and have tried a bubble test on every joint without finding anything.

The furnace in question is a small mesh-belt unit, operating at 1800°F (980°C) using an atmosphere with a 75% hydrogen, 25% nitrogen mixture and a total of 325 cfh (9.2 m3/h) of atmosphere flow. The front door opening is 6 inches (152 mm) wide x 3 inches (76 mm) high. The rear opening is covered by fiber curtains.

The muffle and most of the flowmeters and valves are all relatively new (6 months or less), and the cooling chamber has been pressurized and held its pressure for a significant period of time. We are not experiencing any issues with the surrounding furnaces.

Right now we are scratching our heads and pulling out our hair trying to figure out anything else to check. I'm desperately hoping that there is something glaringly obvious that you would be able to point out to us. I would greatly appreciate any advice/wisdom you could offer!

Answer:
It is my pleasure to try to help. I am quite familiar with the type of furnace you describe. Based on my interpretation of your needs, the first mission is to determine if you have a water or air leak in the furnace. There is a simple test to determine what is going on. The procedure is as follows:

1. Gather relatively thin – less than ¼ inch (6 mm) – and clean samples made of mild steel, copper and (if available) a piece of 304 stainless steel. A short length of copper tubing, sanded to produce a shiny surface, can be used for the copper sample.

2. The furnace (inside) temperature should be lowered (or raised) to 1800-1850°F (980-1010°C) and the belt speed adjusted to ensure dwell time in the high heat and cooling chambers of at least 20 minutes. (Caution: Pure copper melts at 1981°F (1083°C), so care must be taken not to approach this temperature. You know that the temperature inside a muffle furnace will be different than the temperature outsid,e so exercise appropriate caution.)

Results of the test are interpreted as follows: Steel parts will discolor (oxidize) in an atmosphere contaminated with either air or water. Copper will emerge shiny if water is present but will be discolored if air is present – usually with black spots or streaks. Stainless steel will be gray if air or a very small water leak exists and various shades of green if the leak is large.

When running the copper/steel test, I like to include stainless steel to determine the "severity" of the problem. As you may know, the formation of chromium oxide can be from either an air or water source, but the "richness" of the green coloration is usually a clue to the degree of the problem. The steel and copper test will tell you whether it is air or water. When stainless comes out green, I consider it a severe or major problem.

Remember too that hydrogen loves to attract oxygen and form water. If enough air is entering into the muffle, it will create water in the high-heat chamber and could be responsible for the readings you’re getting. Run the steel-copper-stainless test, and let me know what you find.

It might be helpful to know what your "normal" dew-point level in the high-heat chamber is as well as the oxygen (ppm or percentage) in the cooling chambers (as a function of distance from the front door). I like to see 10-20 ppm oxygen levels in the cooling chamber away from the exit end, certainly no higher than about 50 ppm.

More to follow …