Nitration, Hydrogenation, and the Hazard Review That Comes First

There is a category of chemistry that is entirely routine at the bench and entirely unforgiving at scale. Nitration is the standard example. Hydrogenation is another. Both are run every day, at volume, by people who are not taking a gamble, and both have a history of incidents at plants where the chemistry was well understood and the physics was not.

The difference between those two situations is a hazard review done before the batch is scheduled rather than a procedure written after it.

The four numbers

A thermal hazard assessment comes down to a small number of quantities and the margins between them.

The heat of reaction. How much energy the transformation releases per mole. At bench scale a flask sheds it into the room. At fifty litres it has to be removed deliberately.

The adiabatic temperature rise. Where the batch ends up if cooling is lost entirely and nothing is removed. This is a calculation, not a measurement, and it is the single most useful number in the assessment because it describes the worst case honestly.

The onset temperature of decomposition. The point at which something in the vessel begins to break down and release energy of its own, generally faster than the reaction that got you there.

The margin between them. If the adiabatic rise takes the batch past the decomposition onset, the process has a credible path to a runaway and it must be redesigned rather than merely procedurally controlled. Dilution, a slower feed, a lower operating temperature, or a different solvent are all real answers. Careful operation is not.

What makes these numbers necessary at scale and optional at the bench is geometry. Scale a vessel a thousandfold in volume and its surface area grows only a hundredfold. Heat generation follows volume; heat removal follows the wall area available to take it away. The same reaction becomes progressively harder to cool as it gets bigger, and there is no operating discipline that reverses that ratio.

Nitration: accumulation is the hazard

The instinct about nitration is that the exotherm is the danger. The exotherm is the mechanism. The hazard is accumulation.

If nitrating agent is fed to a batch that is not consuming it at the rate it arrives, unreacted reagent builds up in the vessel. The reaction appears calm, because little is happening. What has been created is a quantity of stored energy waiting for whatever eventually initiates it: a temperature rise, the point at which the batch finally begins to react, or the restoration of agitation that had stopped without anyone registering it.

The scenario that recurs in incident reports is not a violent reaction observed and mishandled. It is a quiet vessel that had accumulated an inventory nobody was tracking, followed by all of it reacting at once.

That drives the controls, which are unglamorous and effective:

• Feed rate governed by heat removal, not by schedule. The addition should be limited by what the jacket can take away, which usually makes it considerably longer than the bench procedure suggests.
• Interlocked feed. Addition stops automatically on loss of agitation, loss of cooling, or a temperature excursion. Stopped agitation with continuing feed is the classic path into an accumulation.
• Temperature monitored where it matters, including the possibility of a hot region near the feed point that the bulk measurement does not see.
• Confirmation of the conversion during the run rather than assumption, so that accumulation is visible while it is small.

Hydrogenation: three hazards sharing a vessel

A hydrogenation puts a flammable gas, a pyrophoric catalyst, and an exothermic reaction into a pressure vessel simultaneously. Each has its own controls and they interact.

Hydrogen has a wide flammable range in air and ignites on very little energy, which makes the vessel's inerting sequence part of the chemistry rather than a formality. Purging before hydrogen is admitted, and purging again before the vessel is opened, is what keeps a flammable mixture from ever forming.

The catalyst, typically palladium or nickel on carbon, is pyrophoric when dry and loaded with hydrogen. It is handled wet for that reason, and the filtration at the end of the batch, when a spent catalyst cake can be sitting in air, is the step that deserves the most care rather than the least. A cake that has been allowed to dry has ignited more than one filter.

The exotherm is governed by the hydrogen uptake rate, which means the reaction can be throttled by controlling gas supply. That is a genuine advantage over a liquid addition, and it is only an advantage if the uptake is actually being monitored.

Sitting behind all three is pressure relief that has been sized for the credible worst case, which is not the operating pressure but what the vessel would see if the reaction ran away while the gas supply was open.

Gases you generate without intending to

Some hazards do not come from the reaction being run. They come from what happens when it meets something else.

Thioglycolate chemistry is a straightforward example, and one we work with directly. Potassium and sodium thioglycolate solutions evolve hydrogen sulphide on contact with acid. Nothing about that is surprising as chemistry. It becomes an incident when an acidic waste and a thioglycolate waste meet in the same collection drum, in a room with no gas detection, because the two processes that generated them were assessed separately.

H2S is acutely toxic and, at the concentrations that matter, it deadens the sense of smell rather than warning through it. Segregated waste collection is not tidiness in this context. It is the control.

The general point is that a hazard review scoped to the reactor misses the hazards that live between operations: what is quenched into what, what shares a drain, what is stored next to what, and what a spill would produce if it reached the wrong thing.

What the review produces

A hazard review is finished when it has produced things a plant can act on, not a document that records that thinking occurred.

That means the thermal data and the margin to decomposition, written down. The feed rate the cooling can support. The interlocks that stop the process automatically and what each responds to. The relief sizing basis. The handling procedure for each material by class, including waste. And an explicit statement of what has to happen if the batch deviates, decided while everyone is calm rather than during the event.

Our laboratory and pilot facility run to written procedures for this class of work, with a hazard review before any unfamiliar chemistry is run at scale and segregated collection for chemical waste. The operations are certified to ISO 45001:2018 for occupational health and safety. We do not hold GMP or GLP status and do not present the discipline as something it is not.

If you have a route with an exotherm you are unsure of, the thermal data is worth having before the campaign is booked. It is the cheapest part of the programme and the only part that cannot be recovered later.