Does Nuclear Grade Graphite burn?

The Union of Concerned Scientists's Ed Lyman never met a reactor he liked, despited his profession that he is not prejudiced against nuclear power in principle. Are Lymans concerns about nuclear safety sound? Or is Lyman trying to lead us off the deep end? Is Lyman trying to convince us that a safe reactor is not possible? Take for example the Pebble Bed Modular Reactor, a reactor that seemingly is safe. Unlike Japan's ill fated GE Mark 1 reactors if you shut down the coolant system of the PBMR, nothing bad happens. The PBMR is melt down proof. Now isn't that a safer reactor? "No way," Lyman tells us:

The PBMR has been promoted as a “meltdown-proof ” reactor that would be free of the safety concerns typical of today’s plants. However, while the PBMR does have some attractive safety features, several serious issues remain unresolved. Until they are, it is not possible to support claims that thePBMR design would be significantly safer overall than light-water reactors.

You see there Lyman is ready to rescue us from our nuclear safety illusions. What is wrong with the PBMR is simple,

A second unresolved safety issue concerns the reactor’s graphite coolant and fuel pebbles. When exposed to air, graphite burns at a temperature of 400°C, and the reaction can become self-sustaining at 550°C—well below the typical operating temperature of the PBMR. Graphite also burns in the presence of water. Thus extraordinary measures would be needed to prevent air and water from entering the core. Yet according to one expert, “air ingress cannot be eliminated by design.”

Rainer Moormann, a German reactor scientist argued that,

graphite burning caused by a huge air ingress may lead to massive fission product releases into the environment.

Genera Atomic says Lyman is wrong because nuclear grade graphite does not burn. It is often incorrectly assumed that the combustion behavior of graphite is similar to that of charcoal and coal.

Numerous tests and calculations have shown that it is virtually impossible to burn high-purity, nuclear-grade graphites. Graphite has been heated to white-hot temperatures (~1650°C) without incurring ignition or self-sustained combustion. After removing the heat source, the graphite cooled to room temperature. Unlike nuclear-grade graphite, charcoal and coal burn at rapid rates because:
* They contain high levels of impurities that catalyze the reaction.
* They are very porous, which provides a large internal surface area, resulting in more homogeneous oxidation.
* They generate volatile gases (e.g. methane), which react exothermically to increase temperatures.
* They form a porous ash, which allows oxygen to pass through, but reduces heat losses by conduction and radiation.
* They have lower thermal conductivity and specific heat than graphite.
In fact, because graphite is so resistant to oxidation, it has been identified as a fire extinguishing material for highly reactive metals.

The oxidation resistance and heat capacity of graphite serves to mitigate, not exacerbate, the radiological consequences of a hypothetical severe accident that allowed air into the reactor vessel. Similar conclusions were reached after detailed assessments of the Chernobyl event; graphite played little or no role in the progression or consequences of the accident. The red glow observed during the Chernobyl accident was the expected color of luminescence for graphite at 700°C and not a large-scale graphite fire, as some have incorrectly assumed.

Is this true? The New Scientist published a discussion of the General Atomic claim in its November 4. 1989 edition. The New Scientist investigation pointed out that the graphite in the Windscape fire was inpure, while the relatively pure graphite at Chernobyl contributed little to the that fire's heat. General Atomics in the past offered a demonstration to skeptics who wanted further convincing of their "Graphite does not burn," claim. A block of graphite would be brought out and heated to a red hot temperature. Then oxygen would be blow ovr the red hot graphite which would not catch fire. Needless to say Ed Lyman did not attend one of those demonstrations. The New Scientist did not entirely support the General Atomics Graphite does not burn claim, but the analysis came down on the side of a graphite does burn reluctantly, and is not very dangerous conclusion, pointing to Peter Kroeger's research for support.


Peter Kroeger of Brookhaven National Laboratory used a compluter simulation to check on General Atomic's claim. He found that if openings developed at two opposite ends of a graphite reactor containment structure, air could flow through the core, and graphite structures would burn some, but not very much, and certainly not enough to release radioactive materials embedded in the graphite. Kroeger remarked,
Air ingress into the primary loop requires prior depressurizatlon with significant subsequent air inflow. Scenarios that have been considered are, for Instance, a primary vessel leak such that during decay heat removal via a
main loop or an auxiliary loop, significant amounts of gas can be exchanged between the primary loop and the RB, while the operating loop forces the re- sulting gas mixture through the core [34]. (It may be hard to conceive signi- ficant air ingress and combustible gas discharge from a single break; butonly with such a large break or with several separate breaks and with simultaneous forced flow conditions can significant amounts of air be forced through the core.) Order of magnitude computations indicate that natural circulation can only result In about .1 to .3 kg/s of gas circulation through the core of a typical modular pebble bed reactor. The initial RB air Inventory of about 80 kg mol (even if none were lost during the Initial blowdown) can only cause the burning of about 400 kg of graphite. Thus, air Ingress consequences under natural circulation conditions appear to be less severe than those under the above forced cooldown scenarios.
Four hundred kilograms? That is less than a thousand pounds, hardly a roaring confligration.
Kroeger found that,
Separate code applications for air Ingress with auxiliary loop cooling [34,43,44] generally indicate that fuel temperatures are only raised slightly due to local burning, at most reaching 1200 C for a core with 1000 C design temperature. Thus, fuel failure from excessive temperature is not to be ex- pected. With auxiliary cooling the oxidation stops after 4 to 96 hrs, depend- ing on the assumed air ingress rate and the number of loops operatlij^. The maximum burn-off (averaged over a pebble) ranges from 100 to 350 mg/cm , which represents about 10 to 40% of the total exterior graphite coating of the fueled pebbles. (It should be noted that the higher values are obtained for extremely large assumed air ingress rates, which may not be realistic.)
A further review of the Lyman's (and Moormann's) claim that graphite fires an PBMR are serious nuclear safety issues, is the composition of the Pebbles of Pebble Bed Reactors. The Pebbles are complex manufactured objects. Each pebble contains an inner coat of silicon carbide a nonflamable material that is designed to contain radioactive fission products within the pebble. Any fire on the graphite surface of the pebble would be stopped by the SiC coat, and thus would not lead to a dangerous release of radioactive materials.
Needless to say, Ed Lyman forgot to mention any of Peter Kroeger's research, the General Atomic's argument, or other arguments that makes his simple "Graphite burns" statement less than a serious enditement of pebble bed reactor safety.
Even less so, does the "graphite burns" statement a serious safety objection to the use of graphite in the core of Molten Salt Reactors. It should be noted that the presence of liquid fluoride salts would be a serious inhibitor of any graphite fire, and in the event of salt drainage from a MSR core, a graphite fire would not be a safety issue, because both fission products and nuclear fuel would drain out of the core along with the coolant salt. Thus even if we reject the General Atomic's contention that Nuclear Graphite does not burn, the graphite burns objection does not appear to raise a serious concern about Molten Salt Reactor safety.