DSS042: Fundamental Burning Characteristics of Five Combustible Dusts

In this episode of the DustSafetyScience Podcast, we talk about the fundamental burning characteristics of five combustible dusts. After discussing dust deflagration in general we talk about the combustion process for Carbon, Plastic, Coal, Iron and Aluminum. We also discuss factors the further complicate determining deflagration rates and dust reactivity such as isolated verses group combustion and scale dependency of propagating flames.

What Does a Dust Deflagration Look Like?

Dust deflagration involves a dust cloud and ignition source. The cloud consists of individual dispersed particles while the ignition source, such as a welding arc, creates a flame kernel somewhere in the cloud. If the kernel grows large enough, it will self-propagate and reach the point where it will deflagrate on its own. The flame progresses and heats successive groups of particles. In some cases, situational factors will play a role, such as turbulence stirring up the dust, causing the flame to accelerate and absorb more oxygen.

Let’s take a closer look at dust created by carbon, plastics, coal, iron and aluminum. We chose these dusts to review because they increase in complexity and provide a wide overview of different burning characteristics.

Carbon

The only fuel in a carbon particle is the carbon itself. There is no gas onboard or any other reactive material, so the reaction happens at the surface of the carbon particles.

This surface reaction occurs between two phases: gas (oxygen in the air) and solid (carbon at the surface). It’s sometimes called a nusselt flame, which is named after German engineer William Nusselt, who did a lot of work on dimensional analysis and heat transfer between phases. The flame heats up the carbon particle and causes it to react at its surface. Then a flame develops on the particle surface from this heterogeneous reaction. If the flame can ignite the next particle in the cloud the flame will continue to propagate.

Plastic Dust

With plastic dust, there is no solid reaction. Instead, there is a stepwise process: the dust heats up, melts, and then evaporates, creating a flammable gas that reacts. This process is known as homogeneous combustion. It’s also called a volatile flame because there is only reaction in the gas phase: polyethylene vapour reacting with gas phase oxygen.

Most organics burn this way, and off-gassing is a safety issue. With a wood pellet mill dryer, volatiles are released from the heated particles. If the gas doesn’t immediately react, it can build up and cause a safety hazard. In a textile mill, the dryer that’s flocking material can off the gas and create a gas explosion. You can even have a hybrid mixture explosion where both the off gas and the unreleased dust react under certain conditions.

Coal Dust

Coal dust shows both surface reaction and gas phase reactions. In other words, there are both heterogeneous and homogeneous modes of combustion.

High-volatile coal burns a lot like plastics which heat up and off gas. This volatile gas reacts, standing off some distance from the individual particle, which then contributes to the flame front and propagation.

Low-volatile coal reacts more like carbon. The reaction will be on the surface, burning the oxygen in the air with the solid carbon. If you take a sample of coal dust after doing an explosion test on it and look at it under a microscope, you can determine the volatility. If it was highly volatile, you would see blow holes at the surface of the paricle. There’s a picture of this in the Dust Explosions textbook by Wolfgang Bartknecht.

Metallic Dust

Metallic dust has a different combustion mode, with some dusts being more reactive than others. Some of the gases are much more reactive than some of the organics.

Iron

Iron dust is often measured as a low-KST dust. However, it is important to keep in mind that explosions from this material can still be devastating. An example of this is the but in the Chemical Safety Board report on the Hoeganaes explosions and fires where three incidents over a one-year period killed five workers and injured three more.

The comparatively low reactivity of iron dust is due to its low flame temperature and higher evaporation temperature. Since the flame temperature is lower than that required to evaporate the material is burns are a solid through heterogenous combustion. As we discussed above heterogenous combustion is slower than gas-phase or homogeneous combustion.

Aluminum

Aluminum is at the other end of the spectrum. It is highly combustible, with a high flame temperature and lower evaporation temperature. When that flame propagates, it’s evaporating aluminum vapor, which then reacts. This aluminum vapour is so reactive that it only exists for a very brief moment in time before being consumed with oxygen in a homogenous reaction.

Other Important Considerations

In summary, there are different combustion modes for different particle types. Carbon only has a surface reaction. Plastics generally only have a gas phase reaction away from the surface. Coal might have both. Metallic dusts that have a lower flame temperature relative to the evaporation temperature won’t burn in a gas phase mode and they’ll generally burn slower while metallic dusts that have a high flame temperature compared to the evaporation temperature will burn homogeneously and have a very fast reaction.

Taking all this information into account, trying to predict how a flame front would propagate in a deflagration in an industry setting, like a dryer, silo, or dust collector is very difficult. We want to answer questions such as:

What’s the maximum destructive force?
How strong does a container need to be designed?
How fast does the suppression system need to activate?

There’s still a big gap in our understanding, and one of the goals of this podcast is to close that gap by introducing some of these fundamental concepts and, even more important, connecting university researchers with industry and connecting industry with application specialists and equipment specialists so we can all get talking and solving some of these problems.

Now let’s look at a couple of complicating factors, namely isolated vs. group combustion and scale dependency.

Isolated Combustion vs. Group Combustion

There are two combustion types: isolated and group.

With carbon, for example, reaction takes place at the particle surface, so combustion will always be isolated. However, in a homogeneous-reacting dust like plastics or coal, the flame can overlap with the next particle, resulting in multiple particles burning and creating a much thicker flame.

While there are theories available for the fundamental burning characteristics of a single particle of carbon, plastics, coal, iron and aluminum, the problem is that there are no theories to predict what that means for a deflagration propagating through a cloud.

A lot of it comes down to this isolated combustion versus group combustion as being a complicating factor. Another complicating factor is the role of turbulence and particle size distribution.

Numerical modeling could be a useful tool in this instance: Dr. Chris Cloney wrote his Ph.D. thesis about using computers to simulate these types of flames, and we’re almost at the point where we can apply these systems in relevant industrial scenarios, but as a worldwide community, we’re still trying trying to develop the capability to do so.

Scale Dependency

There are multiple scale-dependent deflagration characteristics. Even if we could estimate the flame’s rate of propagation from understanding single particle combustion, we would still find in some cases, that it is dependent on the size of the flame. This factor was discussed in Episode #39: Does Size Matter – Why is the Standard Dust Explosion Testing Chamber 20-Liters?.

Radiation is a frequently-discussed topic in this area. High-temperature flames like aluminum emit radiation and heat up the particles in front of them. The longer the flame travels, the more it heats the particles ahead up, until the scale of the flame is larger and it accelerates faster.

Radiation in aluminum is the only one that’s been studied in a level of detail that may have some sort of industry application, but a big gap still exists. This is important because radiation and other scale-dependent deflagration characteristics raise questions.

For example, If you test a piece of equipment at six inches in diameter, can you use that piece of equipment at 36 inches? If you have a piece of equipment that works on a 200-liter vessel, can it be used in a 2000-liter vessel? When the reactivity of that flame front is dependent on the scale you’re using, the answer is generally that we don’t know. So you have to test at these very large conditions in order to know that the equipment is going to be safe.

Some dusts are not scale-dependent in the rates of pressure rise, reactivity and flame propagation rate, but some are and some might be. This situation raises a lot of questions about industrial equipment testing. Can we apply what we learn on a smaller scale to these larger scales? Do we know that when a piece of safety equipment is activated, when it’s actually in use, is it going to create a safe result at the end of the day?

These are some of the challenges and other complicating factors involved in bridging the fundamental burning characteristics of dust through to industry application. This type of information is being discussed at scientific conferences, like the International Symposium on Hazards, Prevention and Mitigation of Industrial Explosions (ISHPMIE), which will be hosted in Germany in 2020, and the Combustion Symposium.

Conclusion

Protecting people and facilities from dust explosions, deflagrations, and flash fires requires an understanding of the basic science of combustible dust and how it applies in different scenarios. This knowledge can, hopefully, help the universities, researchers and educators connect and communicate more effectively with industry experts.

If you have questions about the contents of this or any other podcast episode, you can go to our ‘Questions from the Community’ page and submit a text message or video recording. We will then bring someone on to answer these questions in a future episode