Chemical Processing: Scaling Safety Safely

‘Sensible people will see trouble coming and avoid it, but an unthinking person will walk right into it and regret it later. ‘ (Proverbs 22:3).

Careful planning is nothing new. Although chemical development and manufacturing were not exactly commonplace (or even existed) in biblical times, the basics for forward planning were as essential for life then as they are for us all now. People often quote Benjamin Franklin: “By failing to prepare, you are preparing to fail”, but preparation without wisdom and appropriate application can also be just as ineffectual.

In today’s society, health and safety is often the butt of jokes, often seemingly due to over-zealous bureaucrats demanding over-the-top responses to what are perceived to be minor situations^[1]. However, without the correct application of health and safety measures and regulations we have seen how the consequences of not assessing risks or implementing findings correctly can be severe. This ranges from environmental damage^[2,3] and financial loss, to personal injury and, in the worst situations, death^[4,5].

In The Beginning…

So where should we start assessing chemical process safety, for either R&D or manufacture?

For any process, on any scale, the first step should be having a full and thorough understanding of it, 

• identifying all the chemicals involved (both the reagents and the products and by-products), 
• the conditions for each reaction step, 
• the equipment being used, and 
• the potential hazards associated with each step, including the physical, health, and environmental hazards of each chemical identified, along with any expected reactivity hazards (e.g. gas evolution, exotherms).

After hazard identification, the next step is protecting personnel from harm, which can be done using the risk assessment hierarchy of the control system^[6]:

• Elimination:

We may consider the hazards associated with certain chemicals/conditions too great of a risk. In this case, alternative synthetic routes using different materials, conditions or equipment are required. Examples would be avoiding “all-on-board” processes or avoiding equipment that would be unavailable on a large scale (e.g. very high pressure hydrogenations or microwave reactors). Similarly, we should aim to avoid a reaction that produces significant quantities of hazardous gases which needs to be controlled or removed. Control of this on a larger scale presents additional hazards.  

By identifying the chemical and physical hazards at the earliest opportunity, we have a greater chance of eliminating them from projects from the outset.

• Substitution:

While eliminating the hazard should be our primary objective, this is not always possible. In these cases, it would be necessary to look to complete the chemical process with less hazardous materials or processes. One example of a hazard we may wish to remove could be the skin sensitisation potential associated with an increasing number of chemicals (e.g. palladium catalysts^[7]), or process development to avoid hazardous conditions.  

• Engineering Controls:

The purpose of engineering controls is to reduce exposure by preventing hazards from coming into contact with workers, while still allowing them to do their jobs. The prime example of this would be fume cupboards. Some other simple examples of general changes that can be made include the use of heating blocks instead of oil baths, the use of machine guards on overhead stirrers, or the use of interlocked cabinets on equipment where hazardous radiation is generated.

• Administrative Controls:

Administrative controls are intended to change the way work is done or give workers more information. They include

• Procedures, such as standard operating procedures; 
• Training in safe systems of work or for the authorised use of specific items of equipment (e.g. hydrogenation equipment); 
• Warnings, such as signs and labels to notify both the workers and other nearby staff to hazards present in the area; 
• Health monitoring and active monitoring of the work environment; 
• Actions to be taken by emergency responders in the event of an incident. 

These controls are often used together with higher-level controls.

• Personal Protective Equipment:

This is probably the control measure most people would immediately think of for safety. However, this really is the last line of defence, so careful thought must go into ensuring workers not only have ready access to the correct PPE, but also that the PPE provided is compatible with the chemicals being handled, that it is well maintained, and training has been given for its use (e.g. appropriate face fit tests for RPE).

So, that’s it. We’re done. We’ve followed the steps, identified the hazards, ascertained appropriate control measures, and can look upon the assessment as a job well done.

Or can we?

Following the ‘Plan, Do, Check, Act’ approach for managing safety, we have merely completed the first step – we have put plans in place to mitigate against the identified risks, but these plans must then be correctly implemented, or else the assessment is nothing but a waste of time and paper, and this implementation requires the buy-in from everyone. Throughout the assessment, we must always consider whether the controls we are looking to implement are feasible. The controls identified should be: 

• Right for the hazards; 
• Appropriate, based on how likely injuries/illnesses are; 
• Consistent with any existing legislation or company policies; 
• Not too burdensome to workers; 
• Effective, reliable, and durable; 
• Readily available;
• Cost-effective in both the short- and long-term.

Once the appropriate controls have been implemented, and the chemical processing completed, we must ensure that we review the control measures for their effectiveness, reviewing any incidents and accidents associated with the process. Once we have reviewed the process safety we can then look to act on any deficiencies or unexpected events to prevent recurrence, not just for this process, but for others which may utilise similar reagents or processes.

While we have completed our safety assessment, the process still does not end there. Any change to the process from this point, either changes to the chemistry or conditions – or both – will introduce new hazards which must also be assessed. We must then go through the same safety assessment cycle for these hazards as any other previous iteration of the project.

Bigger is Better…

So, what new hazards may we encounter when we change scale?

Scaling a chemical process requires careful consideration of various factors, including reaction kinetics, heat transfer, mass transfer and equipment design. Our projects keep these factors in mind during development, as often 50L scale production is our end-goal. Throughout the development of any process, the gathering of data is essential to assist with the safe design of a scaled process. Communication is therefore key between scale-up and development chemists in this regard.

Some of the most important information to gather is regarding the quantities of gas and heat generated during the process. Understanding reaction mechanisms and stoichiometry will allow for the calculation of volumes of gases generated, so suitable controls such as appropriate venting on the vessel, or appropriate pressure relief systems can be incorporated into any equipment design. It is important to remember that gas evolution occurs just as frequently during a reaction work-up as during a reaction, so when scaling a process, it is important to allow for sufficient venting and headspace in vessels to tolerate this.

Understanding the stability of materials in the reaction conditions is essential. This includes any potential decomposition of reagents, intermediates or products which may also be exothermic, as increasing scale often results in longer times to heat or cool reactions, giving more time for materials to degrade. Conducting process stress tests at higher temperatures and longer times during development gives a far better insight into the process when it comes to increasing the scale.

Similarly, understanding the thermal profile of any reaction is critical to safely scaling a chemical process, especially identifying any expected heat generated from the reaction. In a development lab, one of the most commonly used cooling media for highly exothermic reactions is a dry ice/acetone bath. The high difference in temperature between the coolant and reaction and the large surface area per unit volume means that exotherms that are not noticeable on a small scale can present significant problems as the scale increases. Therefore, planning scalable and suitable conditions is paramount during development, and failing to achieve these would lead to problems such as thermal runaways during scale-up.

For any exothermic reaction, the amount of heat produced increases with the volume (V) of the reaction mixture, and the heat removed depends on the surface area (A) available for heat transfer. As the scale of a process increases, and therefore the ratio of surface area to volume (A:V) decreases, cooling efficiency also decreases, so it is important to replicate the largest scale conditions as early as possible. The change from a round-bottomed flask to a jacketed vessel of the same volume has little effect on the A:V ratio, but as soon as the scale increases, the A:V ratio itself decreases rapidly. But with a limit to how efficient the cooling supplied to a jacketed vessel can be, the heat generated can quickly dominate over any cooling resulting in a thermal runaway.

There is no fixed rule about the size of jump between the initial scale of a process and the final target scale, though many industries and businesses utilise a ten times scale as a rule of thumb. At Onyx, to have greater control over the scaling of processes we often aim for a lower scaling factor for reactions which are particularly exothermic. For example, scaling by two times rather than ten only requires a couple more experiments but the thermal data on these intermediate scales, coupled with the stress studies described above provide crucial data for making safety decisions. These intermediate scales must be sufficiently large to provide representative information for each increase in scale and then ultimately the final scale design. In this way, we minimise the risk of unexpected exothermic changes and identify where process heating or cooling may be affected, ultimately providing a sound basis of safety for the largest-scale reactions.

Having proven that efficient cooling is a key aspect when increasing scale, the maintenance of reaction cooling systems is clearly critical to avoid potential accidents. Many industrial disasters have come about due to the loss, removal, or failure of cooling systems or back-up systems including the disasters at Bhopal^[8] and the T2 laboratories in Florida^[9]. While these disasters were clearly at sites where production is far greater than the scales operated on in scale-up laboratories, the same risk is still present. Therefore, to mitigate against potential tragic situations, it is important that all critical systems are regularly checked and maintained.

At the End of the Day… 

When looking at the development of chemical processes, and the subsequent scaling of these processes, we can see that forward planning and the integration of safety are inherent in all of the steps from the initial concept of a process; through the retrosynthetic design of the chemical scheme and process development, then ultimately to the transference of the finalized synthesis to intermediate and production scales. Each stage of this process presents safety challenges which must be carefully considered, assessed and constantly reviewed before we ‘walk right into it and regret it’.

 

References:

^{[1] https://www.independent.co.uk/news/education/education-news/university-bans-throwing-mortarboards-at-graduation-and-tells-students-hats-can-be-photoshopped-in-a7035211.html, accessed on 15 March 2024}

^{[2]: https://www.newscientist.com/article/dn19425-the-eight-failures-that-caused-the-gulf-oil-spill/, accessed 07 May 2024}

^{[3]: https://world-nuclear.org/information-library/safety-and-security/safety-of-plants/chernobyl-accident.aspx, accessed 03 April 2024}

^{[4]: https://en.wikipedia.org/wiki/Rana_Plaza_collapse, accessed 07 May 2024}

^{[5]: https://www.csb.gov/bp-america-texas-city-refinery-explosion/, accessed 07 May 2024}

^{[6]: https://www.cdc.gov/niosh/hierarchy-of-controls/about/, accessed 18 Jul 2024}

^{[7]: Anonymous (2002) Environmental Health Criteria 226: Palladium. World Health Organisation, Geneva, Switzerland, http://www.inchem.org/documents/ehc/ehc/ehc226.htm, accessed on 27 March 2024}

^{[8]: Environ Health. 2005; 4: 6}

^{[9]: https://www.csb.gov/t2-laboratories-inc-reactive-chemical-explosion/, accessed on 03 April 2024}