Endothermic Atmosphere Chemistry and Control
Allen Golden - Surface Combustion Inc.

What is "endothermic protective atmosphere"?

In simplest terms it is 40% Nitrogen, 40% Hydrogen and 20% Carbon Monoxide (when made from Natural Gas "Methane" and Air). It can also be generated from propane and other hydrocarbons, with slightly different resulting compositions.

It is the result of reacting one volume of Methane with one volume of air and results in seven volumes of Endothermic Atmosphere.

The reaction takes place in two stages. In stage one the reactants yield Water, Carbon dioxide, and Nitrogen along with excess methane. It is performed by heating the reactants to around 1850 - 1900F over and alundum media in the reaction tube.

The transformation is completed over a nickel catalyst in a second stage of the reaction tube which breaks up most of the water to form hydrogen gas and combines the water's oxygen with the additional methane to produce Carbon Monoxide and additional Hydrogen gas.

The actual final composition of the endothermic atmosphere is approximately: 37.9% Nitrogen, 40.6% Hydrogen, 20.7% Carbon Monoxide, with 0.1% Carbon Dioxide, 0.3% Water, and 0.4% Methane remaining in the stream.

Both the Dew Point (water content to temperature relationship) and Carbon Dioxide content can be used to relate to the carbon potential of the atmosphere, which is a major factor in determining case hardening effectiveness for ferrous alloys.

Modern "Endothermic Generators" include features to reduce maintenance, such as filter systems for pump lubrication (requiring only yearly change), air cooling (as an option vs. a water system), separate heaters for each reaction tube which allows "burn-out" (reduction of carbon deposit) of the catalyst of one tube while others continue to generate gas, and automated control of "Dew Point" via oxygen probes.

Surface Combustion recently participated in a study requiring "very tight" control of case depth through better control of the endothermic atmosphere. The goal was to reliably produce coupons with +/- 0.05% maximum deviation of carbon from the set point of the endothermic controller.

The study was conducted over a wide range of temperatures, and set points using batch "integral quench" furnaces at a Texas oil well bit manufacturing facility with carbon ranging from 0.2% up to 1.1%.

When controlled with only "oxygen probes" for dew point the deviation was much to great (+/- 0.2%). Therefore, and IR analysis system with feedback for control was added to the generator system. This system could acurately measure Carbon Monoxide, Carbon Dioxide, and Methane content of the gas.

It was found that using theoretical calculations and on-line Carbon Monoxide and Carbon Dioxide results from the analyzer, it was possible to obtain the +/-0.05% carbon accuracy over 90% of the time. (Final verification of the carbon was by measuring carbon in actual metal samples carburized during the gas generation periods using a LECO carbon analyser).

An interesting point of the study was that when IR Methane analysis was added to the control parameters, the accuracy of the control set point to LECO measurement actually went down, and the +/-0.05% parameter could only be held about 40% of the time. This would suggest that "methane content" should be monitored (as it serves as an excellent indicator of problems and maintenance cycle for the generator), but it should NOT be figured into the control parameters.