Chemical research and manufacturing rest on proven safety principles developed through decades of incident analysis. Whether in a lab, pilot plant, or full-scale facility, chemists define failure scenarios, test protection layers, and define when a process is safe.
Digitalization puts that certainty under pressure. It introduces a new class of failure that sits outside of traditional chemical risk models but directly affects their conclusions. Chemists must consider the inherent safety not only of their physical chemical processes but also of the digital systems that monitor and control them.
Cyber intrusion can alter control logic, disrupt communications, and change configurations remotely without any visible sign at the equipment level. Chemists now face a greater responsibility to understand how digital factors influence the safety decisions they make.
A fundamental shift in risk assessment for chemical facilities
From a safety perspective, the origin of an event matters less than the performance of the protection layer. In a connected plant, that performance depends partly on cybersecurity, governance, and verification. This is a dependency most chemical facilities have not yet incorporated into their hazard analysis—for example, where a safety function depends on a networked signal or shared system.
For chemists moving their work from lab to plant, the credibility of digital systems matters more than it might appear. The validated process handed to operations assumes that the control environment behaves as designed. In a networked plant, cybersecurity risks mean that integrity can be hard to guarantee.
The publication of edition 2 of the International Electrotechnical Commission (IEC) 61511 standard recognized this with a new requirement for security risk assessments of safety instrumented systems (SISs), clearly stating that the availability and integrity of safety functions form part of functional safety. Yet many facilities have missed this change, underestimate it, or treat it only as a compliance checkbox.
Extending the lens of safety processes to digital pathways
The industry cannot remain stuck in its old safety processes. Hazard and operability (HAZOP) exercises help facilities identify the ways in which process faults could occur and the potential impact if they do. Consider a HAZOP exercise that addresses the loss of a critical sensor that sends a shutdown signal, thus preventing a shutdown when a shutdown is needed. Engineers design redundancy and proof testing to address the direct mechanics of the fault, but what if the shutdown signal itself is delayed, altered, or blocked due to network issues? This falls outside traditional HAZOP boundaries.
Practical SIS security assessments are relatively straightforward. They begin by defining boundaries around critical safety functions and their digital connections, mapping any dependencies they might have, and then evaluating potential scenarios where their availability or integrity could be compromised. In turn, engineers should select cybersecurity controls consistent with the IEC 62443 standard, which defines technical and procedural requirements for securing operational technology and industrial automation and control systems. These should then be integrated into the facility’s existing management of change processes.
What chemical and other industrial plants are missing
Field studies at chemical and other industrial facilities consistently show a limited separation between process-control and safety networks, shared credentials across stations, legacy systems that are maintained but not hardened, permanently enabled remote access, and backup procedures that are rarely tested.
None of these areas automatically create incidents, but in shared digital environments, a compromise in one area can affect multiple protection layers simultaneously. Devices which appear independent on a diagram of piping and instrumentation may actually share a common digital pathway, connecting them together in a way which the diagram does not acknowledge.
These gaps stem from incremental system evolution. Engineering teams inherit architectures that grew organically as connectivity, remote support, and vendor tooling were added over years. Each addition introduced dependencies that standard safety reviews were not designed to evaluate. For those handing off apparently validated work, this inherited complexity may be invisible.
Rethinking hazard analysis at chemical facilities
Conceptually, the structure of digital hazard analysis will likely feel familiar. Chemical engineers already excel at examining process deviations. HAZOP reviews should extend to cover network interruption, delayed transmission affecting shutdown timing, unauthorized parameter changes, and loss of synchronization between redundant systems. Reviews of layers of protection analysis, which evaluate whether safety layers reduce hazards to an acceptable level, should be similarly altered to encompass digital risks.
Few facilities have rehearsed what happens when confidence in the control environment itself is lost, so it is important to develop controlled shutdown procedures for when system integrity cannot be confirmed. These should include pre-agreed criteria for switching to manual operation or initiating shutdown, along with facility-wide incident-response drills.
Making cybersecurity practical for chemical plants
Digital safety integration does not dismantle existing risk-evaluation frameworks; it extends them to account for the risk presented by connected systems. Network segmentation, which reduces device connectivity to only required communications, is akin to preventing cross-contamination in a process environment.
Denrich Sananda Credit:
Courtesy of Denrich Sananda
For plant engineers and information technology teams, practical steps include maintaining that segmentation, strengthening the governance of remote access, and formally including digital configuration changes in the management of change processes. For process-safety teams, proof testing already validates hardware reliability assumptions, but it absolutely must be extended to cover digital integrity, ensuring that data, logic, and communications function in real operating conditions.
Chemists do not need to become cybersecurity specialists, but they do need to ask the right questions and work with those administering the control environment to ensure that plant processes behave as expected.
Denrich Sananda is managing partner and senior consultant at Arista Cyber.
Views expressed are those of the author and not necessarily those of C&EN or ACS.
Do you have a story you want to share with the chemistry community? Send your idea or an 800-word submission to [email protected].
