Tag: functional safety

Machine Guarding Compliance for Australian Manufacturers: A Practical Guide

Guarding Is Where Machine Safety Becomes Visible

Walk through any production plant and the guards tell you a great deal. Guards that fit, close properly and are in place suggest a site that takes machinery seriously. Guards propped open, cable tied back, missing fasteners or stored in a corner tell a different story, usually about a machine that jams and a guard that gets in the way of clearing it.

This guide explains what Australian law and standards require of machine guarding, the types of guard and where each belongs, why guards get defeated and a point that catches many manufacturers out: what happens to your legal position when you modify a machine.

What the Law Requires

In Queensland the requirements sit in the Work Health and Safety Regulation 2011 and are explained in the Managing the risks of plant in the workplace Code of Practice 2021. Section 208 of the regulation deals specifically with guarding and sets an order of preference for the type of guard to be used.

A permanently fixed physical barrier where access to the guarded area is not needed during operation, maintenance or cleaning.

An interlocked physical barrier where access is needed. It prevents access while the area is hazardous and allows it when the risk has been removed.

A physical barrier that can only be altered or removed using a tool where neither of the first two is reasonably practicable.

A presence sensing safeguarding system where a physical barrier is not reasonably practicable. This is equipment such as a light curtain that detects a person and stops the hazard before they reach it.

The regulation also sets requirements for the guard itself. It must be of solid construction and securely mounted so as to resist impact or shock. It must make bypassing or disabling the guarding as difficult as is reasonably practicable. It must not create a risk in itself. It must be properly maintained. And where a guard can be removed for maintenance or cleaning, the plant must not be able to be restarted until the guard is replaced.

The Standards Behind the Law

The detailed engineering guidance is in the AS/NZS 4024 Safety of Machinery series. It is a large series, and a handful of parts do most of the work for guarding.

AS/NZS 4024.1100 is the application guide and the place to start.

AS/NZS 4024.1201 covers risk assessment and risk reduction, which determines what guarding a machine needs.

AS/NZS 4024.1601 covers the design and construction of fixed and movable guards.

AS/NZS 4024.1602 covers interlocking devices associated with guards.

AS 4024.1603 covers prevention of unexpected start-up.

AS/NZS 4024.1604 covers emergency stop principles.

AS/NZS 4024.1801 gives safety distances to prevent upper and lower limbs reaching danger zones.

AS/NZS 4024.1803 gives minimum gaps to avoid crushing parts of the body.

Where a guard depends on a control system, as an interlock or a light curtain does, the reliability of that system matters as much as the barrier. ISO 13849 is the standard used to determine and achieve the performance level a safety function requires.

Choosing the Right Guard

The order of preference in the regulation maps directly to how often people need to get in.

If nobody needs access during normal running, cleaning or routine maintenance, fix the guard permanently.

If people need regular access, to clear a jam, change a reel or clean, use an interlocked guard. Opening it stops the hazard, and the machine cannot run until it is closed.

If access is occasional and an interlock is not practicable, a guard that needs a tool to remove is acceptable, on the understanding that removing it is a maintenance task carried out under isolation.

Presence sensing devices suit openings where product has to pass or where a physical barrier would make the machine unworkable. They demand careful design. The device has to be far enough from the hazard that the machine stops before a person can reach it, which means the stopping time has to be measured.

The reach distances and gap sizes in the 4024 series exist because a guard with an opening is only safe if a hand or finger passing through it cannot reach the hazard. A mesh guard with apertures too large for its distance from a nip point looks compliant and is not.

Why Guards Get Defeated

Operators rarely defeat a guard out of carelessness. They do it because the guard makes the job harder than it needs to be. The common reasons are consistent across plants.

The machine jams often, and opening an interlocked guard triggers a slow restart. The guard blocks the view of something the operator needs to see. Cleaning requires removing a guard that takes ten minutes and a spanner. The interlock switch is unreliable and causes false stops. The guard was added after the machine was built and does not fit the way the machine is actually used.

The regulation’s requirement that bypassing be made as difficult as reasonably practicable is one answer: coded interlock switches in place of simple ones, fasteners that stay with the guard, and switches positioned where they cannot be reached with a spare actuator. The better answer is to remove the reason. Fix the jam. Add a viewing window. Design the guard to open without tools where cleaning is frequent, with an interlock. A guard that suits the task does not get defeated.

Guarding in Food and Beverage Plants

Hygiene and safety can pull in different directions. A fully enclosed guard is safe and hard to clean. An open design is easy to clean and exposes the hazard. Good guarding for food plants resolves this by design: stainless construction, sloped surfaces that shed water, no ledges or hollow sections where product collects, hinged or lift off panels with interlocks for frequent cleaning access, and interlock devices rated for washdown.

Conveyors deserve particular attention. Nip points at transfers, tail drums and return rollers are among the most common sources of injury in packaging halls, and they are often left unguarded because they are low, numerous and awkward.

Modifying a Machine Changes Your Position

This is the point most often missed. Manufacturing plants modify machines constantly: a guard is cut back to fit a new conveyor, a section is extended, a machine is speeded up, two machines are joined, an imported machine is adapted to local product.

Under work health and safety law, designers and manufacturers of plant carry specific duties to ensure it is safe. The Queensland code of practice makes clear that when plant is altered, the person making the alteration takes on the duties of a designer or manufacturer for that alteration. A plant that modifies its own machine, or engages someone to do it, has stepped into that role.

The practical consequences are these. A modification needs a risk assessment before it is made, not after. Guarding and safety functions affected by the change need to be redesigned and validated. The work should be documented with drawings, the assessment and test records. And the original manufacturer’s compliance no longer covers the altered parts.

The same thinking applies to second hand and imported machinery. Suppliers have duties, but a machine built for another market may not meet Australian expectations for guarding, and the plant that installs it should assess it before it runs.

A Guarding Audit in Seven Checks

Walk each machine and ask. Is every hazard that a person can reach guarded? Is each guard the right type for how often access is needed? Are guards intact, secured and free of damage? Do interlocks work, and does the machine stay stopped while the guard is open? Can any guard be bypassed easily, and is there evidence that it has been? Are openings small enough for their distance from the hazard? Has the machine been modified since its guarding was last assessed?

Record what you find with photographs, rank the findings by severity and deal with the worst first. Then look for the pattern. If the same guard is defeated on every shift, the fix is to the machine and not to the people.

Where BevTech Fits

BevTech is a family owned engineering business in Richlands, Brisbane, with more than 25 years in beverage and FMCG manufacturing. Guarding work usually needs three disciplines at once, and we have all three in house. Our fabricators and machine shop design and build guards in stainless steel to suit washdown environments. Our industrial electricians install guard interlocks, light curtains, safety relays and emergency stop circuits. Our TUV certified functional safety engineers assess the risk, determine the performance each safety function needs and validate the result.

If you have machines whose guarding has not kept pace with the changes made to them, use the contact form at bevtech.com.au/contact.

This article is a general guide current as at October 2026 and is not legal advice. Refer to the code of practice and Workplace Health and Safety Queensland for the requirements that apply to your plant.

October 7, 2026
Risk Matrix Explained: How to Use One for Plant and Machinery

A Simple Grid With a Lot Riding on It

The risk matrix is probably the most widely used safety tool in Australian industry. It appears in risk assessments, permits, contractor paperwork, incident reports and board papers. It is also widely misused, and a matrix applied badly can leave a plant feeling safer on paper while nothing on the floor has changed.

This guide explains what a risk matrix is, how to score one sensibly, where it fits in Queensland’s approach to managing risk and why machinery needs more than a coloured square.

What a Risk Matrix Is

A risk matrix is a grid that combines two judgements: how likely it is that something will cause harm, and how bad the harm would be. Each is given a rating, the two are combined, and the result is a risk level, usually shown as a colour from green to red.

The most common layout is five by five. Likelihood runs from rare to almost certain and consequence from minor to catastrophic. If each is scored from 1 to 5 and the two are multiplied, the result falls between 1 and 25. Organisations then set bands, such as low, medium, high and extreme, and attach a required response to each.

The purpose is to rank. A matrix helps a site decide which risks to deal with first and who needs to approve living with what remains.

Where It Fits in Queensland’s Rules

Queensland’s guide to this subject is the How to manage work health and safety risks Code of Practice 2021. It sets out a four step process: identify hazards, assess risks if necessary, control risks, and review hazards and control measures to ensure they are working.

Two things in that code surprise people. First, the words if necessary. The code says a risk assessment may not be necessary if you already know the risk and how to control it effectively. Many hazards have well established and accepted controls, and the expectation is that you simply apply them. Second, the code does not prescribe any particular matrix. It asks you to consider how severe the harm could be and how likely it is to occur, and it notes that risk increases as severity and likelihood increase.

So a risk matrix is a tool some organisations choose to use. It is not the legal requirement. The requirement is to eliminate risks so far as is reasonably practicable and, where that is not possible, to minimise them. Some activities do carry a mandatory risk assessment, including entry to confined spaces and live electrical work.

Scoring Likelihood

The Queensland code describes five levels of likelihood: certain, very likely, possible, unlikely and rare. It asks you to consider how often people are exposed to the hazard and for how long, how effective existing controls are, whether changes in the organisation could increase the chance of harm, the working environment, and how people actually behave, including when they are tired or rushed.

The common mistake is to score likelihood from memory: it has never happened here, so it must be rare. A plant that has run a machine for ten years without an injury may have been careful, or lucky. Look at exposure. A task done forty times a shift by every operator is a very different proposition from one done twice a year by a fitter.

Scoring Consequence

Consequence should be the most severe outcome that could reasonably result, not the most common one. Most contacts with a moving conveyor end in a fright. The one that matters ends in an amputation.

The code’s questions are useful here. Could the hazard cause death, serious injury or illness, or only minor injury? How many people could be affected? Could one event lead to others? If the honest answer includes a fatality or a permanent injury, score it that way, however unlikely it seems.

A Worked Example

Take a labelling machine where operators reach in to clear jams. The guard door has an interlock, but it has been unreliable and operators have learned that the machine can be jogged with the door open.

Likelihood. Jams happen several times a shift and every operator clears them. With a defeated interlock, contact with moving parts is likely.

Consequence. Crush or entanglement injuries to the hand, potentially permanent.

Rating. High, on any reasonable matrix.

Now the important part. The matrix has told you this is urgent. It has not told you what to do. That comes from the hierarchy of control.

The Hierarchy of Control Is Where the Work Happens

The code ranks control measures from most to least effective. Elimination comes first. Then substitution, isolation and engineering controls. Then administrative controls such as procedures and training. Personal protective equipment comes last.

For the labeller, eliminating the jam at its source might mean correcting a label feed problem so that nobody needs to reach in at all. An engineering control would be a reliable interlock that cannot be defeated, with a jog function that works only under safe conditions. A sign and a toolbox talk are administrative controls, and on their own they sit near the bottom of the list.

A risk assessment that scores a hazard, adds a procedure and rescores it as low has usually skipped the levels that matter. The code is direct about cost as well. It is one of the factors in deciding what is reasonably practicable, but it cannot be used as a reason for doing nothing.

Five Ways Risk Matrices Go Wrong

The score becomes the goal. Teams argue about whether something is a 9 or a 12 instead of what to do about it.

Residual risk is adjusted to fit. A control is written down, the likelihood is dropped two levels and the square turns green, whether or not the control is in place and working.

Severe but unlikely events are averaged away. A fatality risk scored as rare can land in the same band as a frequent minor injury. They are not the same problem and should not be managed the same way.

Different people give different answers. Without agreed definitions, two assessors will score the same hazard two bands apart.

The assessment is never revisited. The code expects controls to be reviewed when they prove ineffective, before changes to the workplace and when new hazards are found. A matrix filled in at installation says little about a machine that has been modified three times since.

Why Machinery Needs More Than a Matrix

A general risk matrix is a reasonable first pass for a machine. It is not enough to design a safety system. Machine safety work in Australia follows the AS/NZS 4024 Safety of Machinery series, in which AS/NZS 4024.1201 deals with risk assessment and risk reduction.

When a control measure depends on a control system, such as an interlock, a light curtain or an emergency stop, the question changes from how risky is this to how reliable must this safety function be. That is answered with a structured method that considers the severity of injury, how often people are exposed and whether the hazard can be avoided, and it produces a required performance level for the safety function. The circuit then has to be designed and validated to achieve it.

This is the point where a plant risk assessment hands over to functional safety engineering. It is also where modifications cause trouble, because changing a machine can change the risk and the performance its safety functions need.

Making Your Matrix Useful

Define every likelihood and consequence level in words people can apply, with examples from your own site. Assess with the people who do the job, because they know how it is really done. Record the controls that actually exist, not the ones in the manual. Treat any credible fatality or permanent injury as needing attention regardless of the score. Check that controls are in place before rescoring. Keep the record, since the code notes that records demonstrate what you have done to comply. And set a trigger for review: any modification, any incident and any change to how the task is done.

Where BevTech Fits

BevTech is a family owned engineering business in Richlands, Brisbane, with more than 25 years in beverage and FMCG manufacturing. Much of our work begins where a risk assessment ends: designing and installing the engineering controls it calls for. Our TUV certified functional safety engineers and industrial electricians upgrade guard interlocks, light curtains, safety relays and emergency stop circuits, and our machine shop makes the guards and change parts that remove the need to reach in.

If a risk assessment on your site has produced actions that need engineering, use the contact form at bevtech.com.au/contact.

This article is a general guide current as at October 2026 and is not legal advice. Refer to the code of practice and Workplace Health and Safety Queensland for the requirements that apply to your workplace.

October 6, 2026
The Cost of Non Compliance: Why Functional Safety Engineering Matters

Non compliance with machine safety requirements is often discussed purely in terms of regulatory exposure, the possibility of a fine following an inspection or incident. This framing significantly understates the real cost structure facing FMCG manufacturers, where safety non compliance carries financial and operational consequences across several layers, many of which materialise long before any regulator becomes involved.

Beyond the regulatory fine

A safety incident, even a relatively minor one, triggers a chain of consequences that extends well past any direct regulatory penalty. The affected machine or line is typically taken out of service during incident investigation, creating production loss independent of any fine. Workers compensation claims and associated insurance premium increases follow most incidents involving injury, regardless of severity. And for FMCG manufacturers supplying major retail chains or beverage brand customers, a documented safety incident can trigger a supplier safety audit or review, since many large retail and brand customers now actively monitor the safety performance of their supply chain as part of their own risk management, meaning a safety failure at a supplier site can create commercial consequences entirely separate from the regulatory process.

The insurance dimension

Insurers increasingly factor documented safety system quality into premium calculations for manufacturing risk, and a plant with thorough, current risk assessments and properly engineered safety systems is generally better positioned in insurance negotiations than a plant relying on informal or outdated safety documentation. Beyond premium impact, in the event of a significant incident, the quality of a plant’s documented safety due diligence, including formal risk assessments and evidence of functional safety engineering input, can materially affect how an insurance claim or liability matter is resolved, since demonstrated good faith effort to properly assess and address known hazards is viewed very differently from a plant that had no structured safety assessment process at all.

What functional safety engineering actually adds

The gap between surface level compliance and genuine functional safety engineering is significant, and it is where much of the real risk reduction value lies. Surface level compliance might confirm that guards are present and emergency stops are installed. Functional safety engineering goes considerably further, calculating the required safety integrity level for each hazard based on severity, frequency of exposure, and possibility of avoidance, and then verifying that the actual safety circuit design, including the specific components used, their failure modes, and the overall system architecture, genuinely achieves that required integrity level rather than simply appearing adequate on visual inspection.

This distinction matters enormously in practice. A safety relay or interlocking system that looks correctly installed can still fail to achieve the required safety integrity level if the underlying circuit design, component selection, or redundancy does not match what the actual assessed risk demands. TUV certification reflects formal training in this calculation and verification methodology, which is why functional safety engineers bring a level of rigour that general electrical competency, while valuable, does not automatically include.

The cumulative risk of incremental compromise

Non compliance rarely arrives as a single deliberate decision. It typically accumulates through a series of individually reasonable feeling compromises, a guard modified during an urgent breakdown repair and never properly restored, a safety interlock temporarily bypassed to keep a line running during a busy period and never formally closed out, a machine integrated with new equipment without the interlocking logic being fully updated to reflect the new combined hazard profile. Each of these decisions, made under production pressure, can feel justified in the moment, but together they create a safety integrity gap that a structured functional safety review will typically surface in detail, often to the surprise of plant management who were not aware how much had quietly drifted from the original design intent.

Closing the gap proactively

The cost of proactively closing safety integrity gaps through structured functional safety engineering is almost always lower than the layered cost of non compliance materialising through an actual incident, insurance impact, or customer audit failure. For procurement and operations leaders, this argues for treating functional safety engineering investment as core risk management rather than discretionary compliance spend, and for engaging genuinely qualified functional safety expertise rather than relying solely on general electrical or mechanical competency to assess and design safety critical systems.

Qualified functional safety expertise

Bevtech Engineering and Automation’s team includes TUV functional safety engineers who provide genuine safety integrity level assessment and verified safety circuit design for food and beverage manufacturing equipment, going beyond surface level compliance to close real safety gaps. To discuss a functional safety review for your facility, contact Bevtech on +61 400 881 321 or admin@bevtech.com.au, or visit 25 Silvio St, Richlands QLD.

June 20, 2026

Get Your FMCG & Beverage Plant Shutdown Planning Checklist

We wont spam you :) or share your details.
We will send you an email with the PDF checklist.