Chemical Change Control: 5 Evidence Gaps That Let New Hazards Enter Production

A chemical change is not controlled when the new product appears in the inventory. It is controlled when the organization can show, before use, what changed in the hazard, exposure, storage, process, emergency response, and worker understanding. The five evidence gaps below explain why a familiar approval workflow can still introduce an unmanaged chemical risk.

Why chemical change control fails before the new product arrives

A chemical change can create a new occupational exposure even when the replacement product performs the same production task. The evidence is incomplete when the review compares product names but does not compare hazard statements, quantities, routes of exposure, storage conditions, process temperature, waste streams, and emergency arrangements. OSHA's Hazard Communication Standard, 29 CFR 1910.1200, requires accessible hazard information, but access to a safety data sheet does not prove that the work system is ready.

A typical substitution looks harmless because procurement, engineering, and operations each see only one part of the decision. Procurement checks availability. Engineering checks compatibility. Operations checks whether the line can run. The risk appears between those views, where a different vapor pressure, incompatibility, cleaning method, or disposal route changes the exposure profile.

That gap matters because a chemical review is a decision about work design, not only a document review. A plant that approves the purchase before defining the control boundary has already allowed schedule pressure to shape the risk decision.

Andreza Araujo's work across multinational EHS environments is relevant here because chemical safety becomes reliable when engineering, operational discipline, and care for people are treated as one decision rather than three disconnected approvals.

Evidence gap 1: the safety data sheet is treated as the complete assessment

A safety data sheet describes the supplier's hazard information, but it does not describe how a specific site will receive, store, transfer, use, clean up, or dispose of the substance. The missing evidence is the connection between the SDS and the actual task, equipment, people, and conditions at the workplace.

Review the SDS against the proposed use, including concentration, temperature, pressure, application method, ventilation, container size, and foreseeable misuse. Section 8 can identify exposure controls and personal protection, yet those recommendations still require local verification. A respirator recommendation is not a respiratory-protection program, and a glove material listed by a supplier is not proof that the glove remains suitable for the task duration. Label quality matters as well, which is why the chemical-label review should test whether information remains visible and usable at the point of work.

The review should record which hazards are new, which controls already exist, and which controls must be designed or changed before the first delivery. If a risk owner cannot explain that chain in plain language, the chemical change is not ready for approval.

Use the existing SDS review checks as a starting point, then add task-specific evidence that the supplier document cannot provide.

Evidence gap 2: inventory records do not match the worksite

A chemical inventory is useful only when it matches what workers can actually find in the workplace. The evidence gap appears when the central register lists approved substances while secondary containers, decanted products, obsolete stock, and temporary storage remain outside the same control system.

Walk the route from receiving to point of use and compare every container with the proposed inventory record. Look for different trade names, concentration changes, unlabeled spray bottles, damaged labels, partially used drums, and chemicals held by contractors. The purpose is not to produce a perfect spreadsheet. It is to test whether the organization knows what people may encounter during normal work, abnormal work, and cleanup.

Quantity also changes the decision. A product stored in a small laboratory bottle presents a different planning problem from the same substance held in a process tank or a series of intermediate containers. The assessment should therefore capture maximum credible quantity, not only the average amount ordered each month. The 30-day chemical inventory audit provides a practical field sequence for testing whether the register matches the workplace.

OSHA's process safety management rule, 29 CFR 1910.119, provides a useful discipline for highly hazardous processes because process information, operating procedures, training, and management of change must connect. Even outside its direct scope, that logic exposes why a static inventory cannot substitute for operational evidence.

Evidence gap 3: exposure assessment starts after the process is installed

Exposure control should be evaluated before commissioning because the chosen process can determine whether workers are exposed during routine operation, sampling, maintenance, or cleaning. Waiting for complaints or monitoring results makes the first workforce the test population for a decision that should have been designed more carefully.

Map the task sequence and identify inhalation, skin, eye, ingestion, injection, and take-home pathways. Consider who opens containers, who connects hoses, who clears blockages, who samples the process, and who cleans spills. A job title is not an exposure assessment because the same role can contain several very different tasks.

Industrial hygiene evidence may include representative sampling, ventilation verification, enclosure performance, local exhaust testing, or a reasoned qualitative assessment when measurement is not yet possible. The important point is that the decision record should explain why the selected evidence is sufficient for the proposed work, including what will trigger reassessment.

For a new chemical process, the hierarchy of controls should shape the design before PPE becomes the headline control. Andreza Araujo makes this distinction central to her practical safety writing because protective equipment can support a barrier, but it cannot repair a process that releases an avoidable exposure into the work area. The hierarchy of controls gives leaders a clearer order for deciding whether the process should be redesigned before PPE is relied upon.

Evidence gap 4: emergency response is copied from the previous chemical

Emergency readiness is not transferable merely because two products are used in the same room. A chemical substitution can change first-aid priorities, firefighting media, incompatibilities, spill behavior, medical evaluation, waste handling, and the time available for a safe response.

Compare the new SDS with the existing emergency plan and ask what must change before the product arrives. Check eyewash and shower access, alarm assumptions, spill-kit contents, isolation points, fire protection compatibility, rescue boundaries, contractor instructions, and the information that responders will receive at the scene.

The strongest test is a short tabletop exercise based on the actual task. Ask a supervisor to describe the first five decisions after a hose failure or splash, then ask whether the required equipment, authority, and information are available without improvisation. If the response depends on a person remembering a detail that is absent from the work area, the change has not reached operational readiness.

ISO 45001:2018 expects organizations to prepare for and respond to emergencies, but the standard cannot supply the site's chemical-specific sequence. The local plan must show that the emergency controls fit the substance, the equipment, and the people who will respond.

Evidence gap 5: worker understanding is inferred from training completion

Training completion confirms attendance, not understanding. A chemical change is not ready when a learning record is complete; it is ready when the people who perform the work can identify the changed hazard, explain the critical controls, and act correctly when the task deviates from the plan.

Ask operators and maintenance workers what changed, what must never be mixed, which container is acceptable, what to do after a splash, and who can stop the job. Their answers should match the approved procedure without requiring them to repeat technical language from the SDS. If a worker can sign the training record but cannot identify the new control boundary, the evidence is weak.

Supervisors need a separate verification because they authorize work under changing conditions. They should know when the chemical change requires a permit, an updated job hazard analysis, a revised line break plan, or escalation to an engineer or industrial hygienist.

In Safety Culture: From Theory to Practice, Andreza Araujo emphasizes the distance between declared systems and daily decisions. That distance is visible when training says one thing while the storage cabinet, label, work instruction, and supervisor response say another.

How leaders should decide whether the change is ready

A chemical change should be approved only when the evidence shows that the new hazard is understood, the controls are designed, the emergency response is adapted, and the people responsible for the work can act on the decision. A signature can authorize a review, but it cannot replace the evidence that the review was complete.

Use a decision record that separates four outcomes. The first is approved, where all required controls are verified. The second is approved with conditions, where specific actions have owners, dates, and a clear boundary that prevents use before closure. The third is deferred, where the evidence is incomplete. The fourth is rejected, where the proposed change creates an exposure that the site cannot control acceptably.

Make the risk owner explicit. Procurement may initiate the change, engineering may define compatibility, EHS may assess exposure, and operations may own execution, but someone with authority must accept the residual risk or stop the decision. Shared responsibility without a named decision owner is an invitation for unresolved gaps to become normal.

During her PepsiCo South America tenure, where the accident ratio fell 50% in six months under a 180-day plan, Andreza Araujo's documented career context supports a practical lesson: measurable improvement depends on turning safety intent into operating decisions with visible ownership, not on adding another approval layer.

What to verify during the first use

The first use is a validation event because the workplace reveals conditions that documents cannot fully predict. A chemical change should be observed at the point of use, with attention to container handling, ventilation, connection points, labels, waste, housekeeping, and the decisions workers make when the process does not behave as expected.

Keep the verification focused on the evidence gaps identified during review. If the concern was vapor release, verify ventilation under operating conditions. If the concern was incompatibility, verify segregation and connection design. If the concern was manual transfer, observe the actual posture, reach, container stability, and spill response.

Record what was verified, what differed from the approval assumptions, and whether the difference changed the control decision. A finding that produces no change in the work, procedure, equipment, or supervision may be a note rather than a control improvement.

The first-use review should also define a recheck trigger. A new supplier, concentration, container, process temperature, ventilation arrangement, or cleaning method can reopen the chemical change even when the product name remains the same.

The management question is not whether the chemical is approved

The useful leadership question is whether the organization can prove that the changed chemical will be controlled in the work that people actually perform. That proof requires more than an SDS, more than a current inventory, and more than a completed training record.

Leaders should ask which evidence changed the decision, which control prevents the most credible exposure, who owns the remaining risk, and what field observation will confirm that the approval assumptions are true. Those questions make the review consequential because they connect paperwork to conditions that can harm people.

A chemical change process becomes credible when it can say no without becoming a barrier to responsible improvement. It protects production by preventing emergency improvisation, protects workers by exposing weak controls early, and protects leaders by making the residual-risk decision visible.

Andreza Araujo's broader safety message is consistent with that discipline. Safety is about coming home, which means a chemical substitution is complete only when the new work remains understandable, controllable, and survivable under pressure.

A new chemical should not enter production because the approval packet looks complete. It should enter only after the evidence shows that the task, controls, emergency response, and worker decisions still hold together.

Chemical change control is strongest when it treats evidence as a test of readiness rather than a collection of signatures. Closing the five gaps above gives leaders a defensible basis for approval and gives workers a safer system to operate.