There is a quiet assumption behind most process safety work. Almost every standard hazard study assumes that the chemistry is understood. A HAZOP examines what happens when flow, temperature, pressure or level deviate from intent, but it takes the chemistry itself as a known quantity. That works well for a mature process running a familiar reaction, and far less well in fine chemicals, pharmaceuticals and active pharmaceutical ingredients, where the chemistry is often new, has only been run at laboratory or pilot scale, and has never been characterised for its thermal behaviour. A study cannot catch a hazard that no one has measured, and this is the gap where many of the most serious incidents in these industries are born.
The limitation of assuming the chemistry is known
In a bulk commodity plant a reaction may have been run for decades, its heat output and its failure behaviour known from long experience. In a specialty, pharma or API environment the situation is reversed. A route may have been invented recently, optimised for yield and purity in a flask by chemists focused on the product rather than the hazard, and then handed to engineers to scale up. The people who ran it at small scale may never have seen it misbehave, because a small flask sheds heat easily. When a HAZOP team reviews such a process, they can examine the deviations they can imagine, but the question of what the chemistry will do if it gets too hot is often unanswered, because the study is only as good as the chemistry knowledge it rests on and that knowledge is frequently thin.
The role of unknown chemistry
The hazards that hide in an uncharacterised route are not exotic. They are the ordinary behaviours of energetic chemistry that simply have not been looked for. An exothermic reaction releases heat, and if that heat is generated faster than it can be removed, the temperature climbs, which makes the reaction faster still, in the self accelerating spiral known as thermal runaway. Gas evolution can pressurise a vessel, unstable intermediates can accumulate and then decompose, and streams that are individually benign can be violently incompatible in the wrong order. None of this is visible on a flow sheet. Each hazard waits for a trigger, and the triggers are the events a plant experiences in normal life, a deviation, a scale up, a delay in an addition, a cooling failure.
Scale is what turns a laboratory curiosity into a plant hazard. A reaction in a flask has an enormous surface area relative to its volume, so heat escapes easily, while a plant reactor holding a thousand times the volume has a far smaller surface area for its size, so the same reaction that stayed cool in glass can retain its heat and run away in steel. The numbers that matter behave differently at scale. The heat of reaction is a property of the chemistry, but whether it causes a problem depends on the plant's ability to remove it. The adiabatic temperature rise, meaning how hot the batch would get if no heat were removed at all, can be startling once it is calculated for a full charge. The time to maximum rate, meaning how long a runaway would take to reach its most violent point, may be comfortably long in a small sample and dangerously short in a large one. And the onset temperature at which decomposition begins may sit only a little above the intended operating temperature, a margin that is not safe at all when a cooling failure lets the batch drift upward.
How you actually find it
The reassuring part of this story is that these hazards can be measured, and the tools to do so are well established. The work usually begins with screening, using a small sample to look for any sign of exothermic or gas generating behaviour. Differential scanning calorimetry, or DSC, uses milligram quantities to reveal where a material starts to release energy and how much, mapping the thermal hazards and the onset temperatures. Accelerating rate calorimetry, or ARC, studies a sample under near adiabatic conditions that mimic the poor heat loss of a large vessel, so it captures how a runaway would accelerate and the time to maximum rate. Reaction calorimetry, often carried out in an instrument such as the RC1, measures the heat released by the intended reaction under realistic conditions, which gives the heat of reaction and the adiabatic temperature rise directly and shows whether reagent is accumulating.
This data is then organised into a picture of how dangerous the process is and how much margin it has. The criticality classes described by Stoessel are a widely used way of doing this, ranking a process from one to five according to how the process temperature, the temperature the desired reaction could reach and the temperature at which decomposition takes over relate to one another. A low class describes a process that stays safe even under a cooling failure, while a high class describes one where a loss of control could reach decomposition and demands strong safeguards. The same measurements feed directly into the sizing of pressure relief, because you cannot size a relief system for a runaway whose heat and gas rates you have never measured. Together this evidence forms the basis of safety, the explicit statement of why the process is safe and what must be true for it to remain so, whether that rests on controlling the temperature, limiting the accumulation of reagent or providing adequate relief.
The financial implication of not knowing
It helps to frame this as a decision rather than a technicality. On one side sits the cost of characterising the chemistry, which is modest, known in advance, and buys a definite answer, since a programme of screening and calorimetry has a price that can be quoted. On the other side sits the cost of not knowing, which is not a single number but a distribution of outcomes, most of them mild and a few of them severe. The severe tail is where the argument lives. A thermal runaway that is not contained can destroy a batch and the reactor that held it, take out a campaign's worth of production time, bring regulatory action, and move insurance premiums or even the willingness of an insurer to cover the plant. It can also do lasting damage with customers, particularly the pharmaceutical and specialty customers who increasingly review the safety of their supply chain and do not wish to depend on a site that cannot explain its own reaction hazards.
Weighed this way, reaction hazard testing is one of the highest return investments available in process safety, since a known and bounded cost is spent to remove a small chance of a very large loss. Choosing not to characterise the chemistry is not choosing to save money. It is choosing to carry an unmeasured risk whose worst outcomes dwarf the price of measuring it.
From unknown risk to a managed basis of safety
The good news is that unknown chemistry does not have to stay unknown. The behaviour of a reaction under heat, under accumulation and under loss of control is a set of physical properties that can be measured, and once measured they can be designed around. A specialist can take a poorly characterised route, screen it, run the calorimetry, classify its criticality, size the relief and set out the basis of safety, so that what was an unquantified fear becomes a managed condition. If your process runs chemistry that has never been properly characterised, the most valuable next step is to have it understood, and a reactive chemistry specialist can turn that unknown into a managed basis of safety with you.