Oxidizability and explosibility of pure graphite powder

Full text of publication follows: While graphite is widely considered a heat-resistant material, e.g. able to screen metallic shielding from thermal damage, and graphite powder is used as a fire extinguisher agent where water or carbon dioxide should not, it still can react with air and - being carbon - give forth a significant amount of heat. Whether this makes it a hazard in operations such as dismantling nuclear reactors that contain hundreds of tons of graphite, including a small percentage of powder, is a question that has to be answered, considering that dismantling implies the use of such potential fire initiators as thermal cutters and electrical equipment. For this reason EDF commissioned the Centre National de Prevention et Protection (CNPP) to carry out explosibility tests on unirradiated, nuclear grade (i.e. with about 100 ppm of impurities) graphite powder. CNPP tests were so designed as to simulate realistic conditions that might result from a severe mishap during a dismantling operation, such as the crash of heavy equipment on graphite blocks coupled with the bruise of a high power electrical cable. EDF-CNPP tests complement others, done either in Italy most notably on irradiated graphite dust contaminated with various pollutants, or in the UK where the ability of settled graphite dust to propagate an initial gas explosion into an adjacent volume was assessed. EDF-CNPP tests comprise two steps. Step one was intended to produce a qualitative understanding of how nuclear grade graphite behaves while heated in air. In a first series of experiments graphite samples were heated up to 900 C during two and a half hours and their mass loss measured: it was found that while fine powder is wholly oxidised, coarser powder and chunks retained about two thirds of their initial mass. Oxidation kinetics, as assessed by oven temperature shoot-up, begins at 580 C and is quite low, compared with that of iron powder. In a second series of experiments a graphite piece electrically heated up to 1500 C during 2 min. showed no trace of alteration. In a third series graphite blocks, heated at 850 C, were expose several min. first to the flame of a blowtorch (temp. between 1600 C and 1800 C) with no visible effect, then to that of a thermal lance (temp. between 2000 C and 2400 C). The latter was able, helped by some mechanically performed erosion, to pierce a hole through the 1 cm thick block, while no other alteration or damage propagation showed. Graphite powder, also heated at 850 C, was submitted to a blowtorch flame that made it airborne, and then became a characteristic red from oxidation. On the other hand the contact of molten iron had no oxidation effect. When saturated with oil, deposited graphite powder at ambient temperature may be locally ignited by molten iron, though not propagate the flame. Airborne fine graphite powder, exposed to a blowtorch flame, gets partially oxidised, as evidenced by a moderate temperature rise. As a conclusion of this first step it was established that blocs of graphite simply cannot burn, while graphite powder with a fine enough size is able to do so if airborne and exposed to a high energy source. The reactivity of graphite powder is markedly weaker than that of iron powder and flour. Step two was devoted to studying the ability of an extended cloud, with between 150 and 500 gram of graphite powder per cubic metre, to smother a local explosion. As previous tests showed, even the largest amount of heat, such as the application of a flame may provide, cannot initiate an explosion in airborne graphite powder: the initiator has to deliver its energy in a very short period. From previous explosion tests in France and in the UK we derived that a spark of 4 MW power was required. We choose an electrical initiator, as more representative of potential mishaps in dismantling operations than a chemical or gas explosion. The location was the Les Renardieres test centre of EDF R and D, which allows generating extremely powerful electric sparks. The sparks were triggered at one end of a 5 m long plexiglas tunnel, alter it had been filled with the right concentration of graphite powder. Preliminary tests carried out by CNPP allowed to calibrate opacimeters so as to measure the cloud density. The graphite powder was of the finest size compatible with quick dispersion, as larger particulates are known to act as heat sinks and smaller ones tend to agglomerate, and ground shortly before the tests since it has been noted that graphite powder looses some reactivity with time. The propagation of the explosion within the tunnel was filmed with a fast shooting camera, while temperature and pressure were recorded. The results show that, while the graphite is brought locally to incandescence by the spark, this dies out in about 100 milliseconds without having been propagated to other parts of the dust. It is expected that the EDF-CNPP experiments contribute together with results gathered in other countries to a shared understanding of the low oxidizability and explosibility properties of nuclear grade graphite