Hazard identification, feedback loops, risk tolerance

There are many lessons to be drawn from the explosion in Beirut last week. Perhaps the most important is that we, as a society, always have a choice: to learn from our mistakes, or to repeat them.

We can debate the root cause and find individuals and systems at fault — and we must do these things — but if we don’t make use of those learnings, and change our future behavior, we’ll have missed the point. As one commentator put it, the cause of the Beirut explosion was clear-cut: “Once again, society not learning from its own mistakes. For hundreds of years, there have been situations where too much ammonium nitrate has been kept in the same place at the same time.”

Before I go further, let’s remember that it is a luxury to discuss a tragedy that befell others. Beyond the dead and the injured, there are lost livelihoods, lost identities, and lost futures — none of which directly affects me nor, probably, you. So, in addition to abstract thinking, we can also take concrete action. For example, we can donate money to a group like Medecins sans Frontieres that is on the ground in Beirut, actively helping.

Hazard Identification

LBC The News Explained, Professor Hal explains what happened during Beirut explosion, August 5, 2020

One aspect of this disaster struck me as new and different from previous disasters, however: the excellent response of journalists and commentators to identify the facts, communicate the science of ammonium nitrate, and explain the meaning of all this information. (The quote above is from Hal Sosabowski, the Professor of Public Understanding of Science at the University of Brighton in the UK).

Those of us who work with ammonia or advocate for ammonia energy are keenly aware of a history of confusion between these two chemicals, intricately linked yet completely different. I have seen too many occasions where respected journalists, scientists, or policymakers simply treat the two chemicals as synonyms (Exhibit A: National Geographic, April 2013).

But this is changing. As a society, we are now doing a better job of understanding the differences between ammonia and ammonium nitrate. And this is crucial because we can only manage risks when we correctly identify which risks exist.

Despite its hazards, ammonia is gaining acceptance as a responsible solution for decarbonizing large sectors of society: a maritime fuel, a hydrogen carrier, a way to store and transport renewable energy. To a large extent, this is because ammonia is such a well known chemical with such a long history. We know the risks, we know how to manage them, we know how to regulate them.

Ammonia and Ammonium

The chemical leap from ammonia (NH3) to ammonium (NH4) is at the root of all life, quite literally. As a fertilizer, the nutrient nitrogen in ammonia cannot be taken up by plants until soil moisture, pH levels, and microbes convert it to ammonium and, later, to nitrite (NO2) and nitrate (NO3). The nitrogen cycle continues, up the food chain and down the waste streams, until ammonia hits the ground again.

Ammonium nitrate is a man-made fertilizer that contains the nutrient nitrogen in both its forms: ammonium and nitrate. The only similarities between ammonia (NH3) and ammonium nitrate (NH4NO3) are their shared supply chains and their shared fertilizer use.

Ammonium nitrate is explosive. Ammonia is caustic. Two very different hazards.

Ammonia is a “reducer,” meaning that it needs air or oxygen in order to burn (on top of this, its low flame speed makes combustion difficult). Ammonia will never detonate, only deflagrate.

Ammonium nitrate contains both a “reducer” (ammonium) and an “oxidizer” (nitrate), making it a “monofuel” that needs no air or any other source of oxygen. This is why ammonium nitrate can detonate.

There are other nitrogen-based chemicals that pose high risks: hydrazine (N2H4) is a rocket fuel; trinitrotoluene (C6H2[NO2]3CH3) is better known as TNT. Even nitrogen itself, completely inert, is suffocating. Every molecule has its own risks and must be managed appropriately. A good word for this is “respect.”

Feedback Loops

I said above that we know the risks of ammonia, and we know how to manage them. That statement is true but it invites disaster because it encourages complacency and discourages feedback. We know the risks of ammonium nitrate, and we know how to manage those too. Knowledge is useless if it isn’t applied, and it wasn’t applied in Beirut.

Seven years ago I gave a presentation on Communicating Risk, in which I categorized the causes of ammonia accidents (based on a presentation Making ammonia refrigeration plants safer by Andy Pearson of Star Refrigeration) as follows:

  • Operator not trained for the job,
  • System design / equipment selection flawed,
  • System design / equipment selection compliant but flawed,
  • Mechanical failure.

The first two categories are instances when regulations are not being followed (note: increased regulation will have no impact if the regulations are already not being followed). The fourth is also unlikely to be impacted by regulation. The “compliant but flawed” category is difficult to understand but it can be directly addressed by regulation, not necessarily by making new rules but by simplification of existing ones and reduction of ambiguity.

In response to the Beirut explosion, regulators around the world are now trying to respond appropriately. Here’s an illustration of the need to reduce ambiguity and complexity within the US:

It’s exactly that jumble of federal and state agencies regulating ammonium nitrate and long-standing concerns about a lack of coordination that has experts once again sounding the alarm. On the federal side, OSHA, DHS, EPA, BATF, the Agriculture Department and the little known Chemical Safety Board all have a piece of the ammonium nitrate oversight. Experts say that is a patchwork with dangerous gaps.

NPR All Things Considered, U.S. Is At Risk Of Beirut-Like Explosion, Experts Warn, August 12, 2020

As my 2013 presentation continued, I made the argument that we humans don’t tend to minimize risk. Instead, we optimize risk.

Each one of us defines our own comfortable level of risk (our risk tolerance) and we push our environment to that limit. To do this successfully, we need a correct assessment of our skill (ability to keep a poker face, to drive a car, to manage inventory of explosives). Next, we need feedback, constant feedback, so that we can adjust our behavior and keep the danger we face always just within our risk tolerance (this is called Risk Homeostasis Theory).

If it gets dark when we’re driving a car, we turn on headlights to maintain vision. If it starts raining, we drive slower because we know that brakes are less effective when wet. If we feel like we’re turning a corner too fast, we slow down until we feel in control again. Constant feedback, constantly integrated into the choices we make that define life and death.

Beirut is too much feedback. This accident should never have happened.

Instead, we — industry, regulators, media — should have been communicating “near misses” all they way, so that the unacceptable risks were known and understood. Nobody would intentionally do a welding job beside a stash of fireworks beside 2,750 tons of ammonium nitrate that had been stored for too many years in the heart of a densely populated city beside the nation’s grain silos. The layers of this tragedy are too deep.

Risk tolerance

The most effective aim of safety planning isn’t to mitigate dangerous behavior (a safety belt, a sprinkler system) nor to punish dangerous behavior (a speeding fine, a jail term). The most effective aim is to reduce our level of risk tolerance — that way, we avoid dangerous behavior.

As we move forward with ammonia energy, we bring with us a century of experience and respect that, today, makes ammonia one of the most highly regulated chemicals with one of the best safety records. But soon we will be entering new markets with new applications, and creating new risks, some of which we might not yet fully comprehend. Certainly, we’ll be bringing the chemical into new communities.

Communicate the risks. Communicate how they can be managed. Communicate why this is a good idea. Transparency and openness — access to information and explanation — will create the positive feedback loops necessary to reduce our level of risk tolerance.


Inline Feedbacks
View all comments
Joe Beach

I agree with a lot of what you are saying here, Trevor. I do have some concern about the notion that the aim of safety planning should be to reduce our level of risk tolerance, mainly because I have seen that taken too far. I wouldn’t want the response to the Beirut ammonium nitrate explosion to be a decision that ammonium nitrate can only be stored in 50 pound bags separately housed in 1 bag capacity concrete bunkers with a 100 foot separation. That would be incredibly low risk, but also incredibly expensive and logistically impossible. We need to have… Read more »

Wim Hamelink

I agree to most of it, however I am convinced that knowledge is the most important tool to prevent disasters like this one. Everyone who is working with ammoniumnitrate shall be trained into the hazards and properties including poduction, handling and storage. Everyone includes also maintenance and laboratory people and authorities.
I have given for years this kind of training and also made an AN safety standard. I found that a lot of misunderstanding about this subject exists.