Chemical engineering and the continuous process: from soda and ammonia to process safety

Most manufacturing makes things you can count: parts, assemblies, products. The chemical industry makes things you measure by the tonne or the litre, flowing through pipes, reactors and columns that run day and night. Its products are everywhere, though rarely seen: the fertiliser that grows food, the acids and alkalis used to make glass, paper, soap and metals, the plastics and fibres in everyday goods and the fuels that power transport.

Turning chemistry into industry required a new kind of engineering. Laboratory reactions had to be scaled up from flasks to vessels holding many tonnes, run continuously rather than in batches, controlled automatically and made safe despite high pressures, high temperatures and hazardous materials. The discipline that emerged, chemical engineering, developed general methods for designing and operating such processes. It also learned, sometimes through catastrophe, that a continuous process is a tightly coupled system in which small changes can have large consequences.

This article traces, in general terms, how the chemical industry developed from early alkali works to continuous processes such as ammonia synthesis, how chemical engineering became a profession, what major accidents taught about process safety, and the lessons for any business that runs continuous or hazardous operations. It is a general historical overview.

Alkali, acid and the first chemical industry

The early chemical industry grew around a few basic chemicals needed in large quantities. Soda ash, sodium carbonate, was essential for glass, soap and textiles. In the 1790s, the French chemist Nicolas Leblanc developed a process for making soda from salt, sulfuric acid, limestone and coal. The Leblanc process spread through Britain in the nineteenth century and created a large industry.

It also created severe pollution. The process released hydrogen chloride gas, which damaged crops and property around the works, and produced large heaps of foul-smelling waste. In 1863, Britain passed the Alkali Act, which required works to capture most of the acid gas and appointed an inspector to enforce the rule. It is often described as one of the first modern pollution control laws. Faced with the requirement, manufacturers captured the gas, which led them to new products made from it.

From the 1860s, the Belgian chemist Ernest Solvay developed a process for making soda from salt, ammonia and limestone that produced far less waste and could run more continuously. The Solvay process gradually displaced Leblanc’s. Sulfuric acid, the other foundation chemical, was made in large lead chambers and later by the contact process, in which sulfur dioxide is oxidised over a catalyst.

From batch to continuous

Many early chemical processes ran in batches: load the vessel, run the reaction, empty it and start again. Batch operation is flexible and suits small volumes or many products, but it wastes time in loading, heating, cooling and cleaning, and each batch can differ.

Continuous processes feed raw materials in and draw products out steadily, with every stage running at a stable condition. They suit large volumes of a single product. Once running steadily, a continuous process can be efficient and consistent, but it is harder to start up and shut down, it relies on instruments and control systems, and a disturbance in one stage quickly affects the others. Choosing between batch and continuous operation remains one of the basic decisions in process design.

Ammonia: scaling up a laboratory reaction

At the start of the twentieth century, farmers depended on natural sources of nitrogen fertiliser, such as guano and Chilean nitrate, and there were fears that supplies would not keep pace with growing populations. Nitrogen makes up most of the atmosphere, but turning it into a usable compound is difficult.

Around 1909, the German chemist Fritz Haber demonstrated in the laboratory that nitrogen and hydrogen could be combined into ammonia over a catalyst at high temperature and very high pressure, while recycling unreacted gas. The chemical company BASF assigned the engineer Carl Bosch to scale the process up. Bosch’s team had to solve formidable problems:

• Equipment that could hold hot, high-pressure hydrogen. Hydrogen attacked the steel of early reactors, making it brittle, and reactors burst. The team developed reactors with a protective inner lining and a strong outer shell designed to cope with the hydrogen that leaked through.
• A practical catalyst. Researchers tested thousands of formulations to find a cheap, robust catalyst based on iron.
• Large supplies of pure hydrogen and nitrogen, and compressors, heat exchangers and controls to run the plant continuously.

The first commercial plant began operating in 1913. The Haber-Bosch process became one of the most important industrial processes in history: synthetic nitrogen fertiliser supports a large share of the world’s food production. It also had a darker side, since ammonia is a starting point for explosives, and the process helped sustain Germany’s munitions production during the First World War. Haber and Bosch both received Nobel Prizes for their work.

The story is a classic example of scale-up: a laboratory success becomes an industrial process only through years of engineering on materials, equipment, catalysts and control.

Chemical engineering becomes a profession

Early chemical plants were designed by chemists and mechanical engineers working together. Towards the end of the nineteenth century, universities began offering courses in chemical engineering, and in the early twentieth century the idea of unit operations gave the field its organising framework. Rather than treating every process as unique, engineers recognised that most processes are combinations of a small number of standard operations: heat transfer, fluid flow, distillation, evaporation, filtration, drying, mixing, crushing and chemical reaction. Each operation could be studied, modelled and designed on general principles.

This approach made design faster and more reliable and allowed knowledge to transfer between industries. A distillation column in an oil refinery and one in an alcohol plant obey the same principles, and the same is true of heat exchangers, pumps and dryers. The the petroleum value chain article shows these operations at work in refining.

Instruments and automatic control

A continuous process must hold temperatures, pressures, flows and levels at their set points despite changing feed, weather and demand. Early plants relied on operators reading gauges and adjusting valves by hand. Through the twentieth century, pneumatic and then electronic feedback controllers automated much of this work, continuously comparing measurements with set points and adjusting valves. Control rooms concentrated information and control in one place, and from the 1970s, computer-based distributed control systems managed whole plants.

Automation brought consistency and efficiency, but also new risks. Operators can lose touch with the process when automation runs it most of the time, alarms can flood control rooms during upsets, and a failed instrument can mislead both people and controllers.

Lessons written in disaster

Several major accidents shaped how the industry manages hazards.

Flixborough, 1974

At Flixborough in England, a temporary pipe was installed to bypass a reactor removed for repair in a chemical plant. The bypass was designed and installed without proper engineering assessment. Some weeks later, it failed, releasing a large cloud of flammable cyclohexane, which exploded, killing 28 people and destroying the plant. The disaster made management of change a central principle: any modification, however temporary, must be assessed by competent people before it is made.

Bhopal, 1984

At Bhopal in India, a release of toxic methyl isocyanate gas from a pesticide plant killed thousands of people and injured many more. Investigations pointed to failed and disabled safety systems, poor maintenance, inadequate staffing and training, and housing built close to the plant. Bhopal remains the worst industrial disaster on record and a lasting reminder of the obligations of hazardous industries to surrounding communities.

Longford, 1998

In September 1998, an explosion and fire at the Esso gas plant at Longford in Victoria killed two workers and injured others, and cut Victoria’s main natural gas supply for about two weeks, affecting homes and businesses across the state. A royal commission found that a loss of warm oil flow had allowed a heat exchanger to become extremely cold and brittle, and that when warm oil was reintroduced, the vessel fractured. It found failures in hazard identification, training and the management of the plant’s operating knowledge. Longford is widely studied in Australia for the distinction it highlighted between personal safety, such as slips and trips, and process safety, the prevention of major hazard events, which had not received comparable attention.

Methods for hazardous processes

From these and other experiences, the process industries developed structured methods, including:

• Hazard and operability studies (HAZOP), developed in the British chemical industry in the 1960s and 1970s, in which a team systematically considers how deviations in flow, pressure, temperature and other parameters could arise and what their consequences would be.
• Management of change, requiring assessment and approval of modifications to equipment, procedures, materials and organisation.
• Layers of protection, combining design, control, alarms, safety systems, relief devices and emergency response so that no single failure leads to disaster.
• Major hazard facility regulation, under which, in Australia, operators of facilities storing large quantities of hazardous chemicals must identify hazards, assess risks and demonstrate their controls to a regulator.

The so far as is reasonably practicable article explains how Australian workplaces make and record safety risk decisions.

Lessons for engineers and businesses

Scale-up is engineering, not copying

The ammonia process worked in the laboratory long before it worked at industrial scale. Any process taken from a trial to production deserves engineering attention to materials, equipment, control and failure modes.

General principles travel

Unit operations allowed knowledge to move between industries. Businesses benefit from recognising that their processes share principles with others and learning from them.

Every change is a potential hazard

Flixborough began with a temporary fix. A simple management-of-change routine catches most dangerous modifications.

Process safety is different from personal safety

Low injury rates do not prove that major hazards are controlled. Businesses with hazardous processes need specific measures for them.

Regulation can drive innovation

The Alkali Act pushed manufacturers to capture and use acid gas. Environmental and safety requirements often lead to better processes.

Automation needs alert people

Control systems run processes, but people must understand them well enough to recognise when something is wrong.

A worked example

This is an illustrative example. The business and figures are hypothetical.

A small food manufacturer runs a continuous cooking and cooling line. During a weekend repair, a fitter installs a temporary flexible hose to bypass a faulty valve on the hot-water circuit so production can restart on Monday. The hose is not rated for the circuit’s temperature and pressure, and nobody records the change.

On Monday, the production supervisor notices the hose and, recalling the company’s new management-of-change procedure, stops the line. A quick review takes forty minutes: the maintenance planner confirms the hose’s rating, finds it unsuitable and sources a correctly rated hose from the store, the change is recorded with a date for permanent repair, and the operators are briefed. The line restarts with a delay of under an hour.

A failure of the original hose at 85 °C would have released scalding water near operators and stopped the line for a full shift. The supervisor’s decision cost less than an hour of production; the alternative risked injuries and a much longer stoppage.

How this applies to a small Australian business

• Use a simple management-of-change procedure for temporary and permanent modifications.
• Identify process hazards separately from everyday safety risks.
• Study how your process behaves during start-up, shutdown and upsets, not just steady running.
• Keep instruments calibrated and alarms meaningful.
• Train operators to understand the process, not just the buttons.
• Check regulatory obligations if you store hazardous chemicals, and seek specialist advice.

Common mistakes

• Treating temporary fixes as harmless.
• Judging safety by injury rates alone.
• Flooding operators with alarms that do not need action.
• Scaling up a trial process without engineering review.
• Letting operating knowledge leave with experienced staff instead of recording it in procedures and training.

Frequently asked questions

What is the difference between a batch and a continuous process? A batch process loads materials, processes them and empties the vessel before starting again. A continuous process feeds materials and removes products steadily. Batch suits variety and small volumes; continuous suits large volumes of one product.

Does a small business need a HAZOP? Not always. Simpler hazard studies may be enough for simple processes, but businesses handling flammable, toxic or high-pressure materials should seek specialist advice on the right level of study.

Questions to ask

• Which changes to our equipment or procedures were made without review?
• What are our major process hazards, and how do we know they are controlled?
• How does our process behave during start-up, shutdown and upsets?
• Do our operators understand why the process behaves as it does?
• Which of our trial processes are being scaled up without engineering review?

Bringing it together

The chemical industry moved from polluting batch processes such as Leblanc’s to cleaner, continuous processes like Solvay’s and, with Haber and Bosch, scaled a laboratory reaction into a process that now helps feed the world. Chemical engineering organised the field around unit operations and automatic control. Flixborough, Bhopal and Longford taught hard lessons about change, maintenance, communities and the difference between personal and process safety. For any business running continuous or hazardous operations, the lessons are to engineer scale-up, manage every change, treat process safety as its own discipline, keep operators engaged with the process and see regulation as a spur to better design. The process FMEA and control plans article offers a structured way to apply them.

---

Source: KEVOS editorial notes. Examples and figures in this article are illustrations. This article is a general historical overview and does not constitute safety or regulatory advice.