Welding Health and Safety by the Numbers
Chapter 10 of 10 · 10 min read
Health monitoring, training and records
Monitor welders' hearing and lungs where the risk or the law requires, train for competence, calculate injury rates and why small shops' rates jump, choose leading indicators, and review each year.
A welding health and safety programme can look good on paper and fail in practice. The way to know whether it works is to measure it: the exposures and noise levels of earlier chapters, the health of the people exposed, whether they have the training to work safely, and the injuries and near misses that still happen. Each measure has its traps. Health monitoring finds harm only after it has begun; training records show attendance, not competence; injury rates in a small shop swing wildly from year to year on one or two events. Used together, and honestly, they show where the programme stands and where it must go next.
This chapter covers health monitoring for welders, training for competence, the calculation of injury frequency rates and their limits in a small workforce, the leading indicators that show whether controls are in place before anyone is hurt, the records the programme keeps, and an annual review. It closes by bringing the book together.
Health monitoring
Health monitoring is the monitoring of a worker’s health to find whether their work is affecting it. It does not prevent harm, but it shows whether the controls are working and catches early effects while action can still help. Under the work health and safety regulations, health monitoring is required for workers using certain hazardous chemicals where there is a significant risk to their health, and for workers who must wear hearing protection because of noise above the exposure standard; the regulations and the regulator’s guidance say which substances and situations, and how often.
For welders, the common forms are hearing tests, called audiometric testing, which find noise-induced hearing loss early, at the high frequencies where it begins; lung function tests and respiratory questionnaires, which can show the effects of fume; and, for some metals, biological monitoring of blood or urine, as Chapter 3 described. Health monitoring is carried out or supervised by a registered medical practitioner with experience in it, with the worker’s informed participation, and its results are confidential to the worker and the doctor, with the business receiving the information it needs to act, such as whether controls need review, without the worker’s private medical details.
Acting on the results
A health monitoring result that shows an effect, such as a shift in a welder’s hearing between tests or a fall in lung function, is a signal about the controls as well as the person. The business reviews the controls for that worker’s tasks and for others doing similar work, and the doctor advises the worker on their own health. In the running case, the first round of hearing tests found early high-frequency loss in two grinders and one gouger, which led directly to the noise controls of Chapter 9 and to fit-checking their hearing protection; their next tests showed no further change.
Training for competence
Every worker needs the training to do their work safely: induction for new starters, task-specific training for welding, cutting, confined spaces, hot work permits and the equipment they use, and refreshers when things change. Training is effective only if it produces competence, the ability to do the task safely, not just attendance. The coordinator’s training records for each worker listed each task they were authorised for, the training and the assessment that showed competence, and the date of the next refresher. Welder Training by the Numbers covers how skills are taught and how they fade without practice, which applies to safety skills as much as welding.
Injury frequency rates
The commonest measure of safety performance is an injury frequency rate: the number of injuries of a given kind for each million hours worked. The lost time injury frequency rate counts injuries that cost at least a full shift away from work; the total recordable injury frequency rate adds injuries that needed medical treatment or restricted duties.
- Hours worked
- about 30 × 1,900 = 57,000 a year
- Lost time injuries
- 2 in one year, 0 in the next
- year one: 2 × 1,000,000 ÷ 57,000 ≈ 35
- year two: 0 × 1,000,000 ÷ 57,000 = 0
- a single injury moves the rate by about 17.5
The calculation shows why a small shop’s injury rate is a poor guide to its safety from year to year. With about 57,000 hours a year, each lost time injury adds about 17.5 to the rate, so the rate swings from zero to thirty-five and back on the chance timing of one or two events. A shop with no lost time injuries in a year has not necessarily become safe; one with two has not necessarily become dangerous. Rolling rates over three or more years smooth the swings, and comparing the numbers of injuries by type, as Chapter 1 did, shows more than the rate itself.
Leading indicators
Injury rates are lagging indicators: they measure harm after it has happened. Leading indicators measure whether the controls that prevent harm are in place and working, before anyone is hurt.
| What it measures | Example in the running case | |
|---|---|---|
| Lost time and recordable injuries | harm that has happened | injuries by type, three-year rolling rate |
| Exposure results | fume and noise exposures against the standards | share of tasks below half the standard |
| Extraction tests | whether systems deliver their design airflows | share of hoods passing their six-monthly test |
| Inspections and audits | whether controls are in place on the floor | monthly walk findings closed on time |
| Training | whether people are competent for their tasks | share of workers current for every authorised task |
| Near misses and hazard reports | problems found before harm | reports made and acted on each quarter |
| Health monitoring | early effects in exposed workers | hearing shifts and lung function changes found |
Leading indicators are more useful to a small shop than injury rates, because they change with the controls, not with chance. A shop that tests its extraction, measures its exposures, closes its inspection findings and keeps its training current will have fewer injuries and less disease over time, even if the injury rate of any one year does not show it.
Near misses
A near miss is an event that could have caused harm but did not: a cylinder that fell but hit no one, a fire caught by the fire watch, an arc struck near an unprotected bystander. Near misses happen far more often than injuries and have the same causes, so they are the best early warning a workplace has. Workers report them only if reporting is easy, safe and seen to lead to action. The coordinator made near-miss reporting a short card or a phone message, reviewed every report at the committee’s monthly meeting, and posted the actions taken, and reports rose from almost none to several a month, which the coordinator counted as a success.
Investigating incidents
When an injury or a serious near miss happens, the investigation looks for causes, not blame. Asking why the event happened, and then why each answer happened, usually leads past the person who was hurt to the conditions around them: an extraction arm too short to reach, a job planned without time for the permit, a voltage-reduction device disabled because it made starting difficult, a helmet that was uncomfortable enough to lift. Those conditions are what the business can change. The coordinator’s investigations were short, done within a day or two with the people involved, and ended with actions that had owners and dates, which the committee followed until they were closed.
Consultation and the committee
The health and safety committee, of welders, a supervisor and the coordinator, was the programme’s engine. It walked the shop monthly, reviewed measurements, incidents and near misses, proposed controls, and tested new equipment before it was bought. Its welders explained the programme to their workmates in a way no notice could, and brought back what was not working. The work health and safety laws provide for health and safety representatives and committees, and give workers the right to be consulted; the running case found that the committee’s practical knowledge made the controls better as well as more accepted.
Contractors and labour hire
Workers who are not employees, labour-hire welders, contractors and apprentices from training providers, are owed the same protection, and the businesses involved must consult and coordinate with each other on who does what. Labour-hire welders often arrive with their own habits and equipment and without the shop’s induction. The shop’s induction, training records and health monitoring applied to them as to employees, and its arrangements with the labour-hire firm said which business provided which equipment and monitoring, so that nobody fell between the two.
Costs and benefits
A welding health and safety programme costs money: extraction, screens, respirators, hearing protection, measurements, health monitoring, training and the coordinator’s time. Its benefits are partly visible, in fewer injuries, lower workers’ compensation premiums, less lost time and better retention of skilled welders, and partly invisible, in diseases that never develop. The running case’s owner found that the visible savings in its first three years, from fewer injuries and lost days and a lower premium, roughly matched the programme’s running costs, before counting anything for the diseases prevented.
Return to work
When a worker is injured, helping them return to suitable work as soon as it is medically safe benefits them and the business, and the workers’ compensation laws require employers to support it. Suitable duties, planned with the worker and their doctor, keep the worker connected to the shop while they recover. The coordinator kept a short list of lighter tasks in the shop, such as inspection, programming and stores work, that could be offered during recovery.
Records
The programme’s records include the risk register, exposure measurements and their conditions, extraction test results, health monitoring results held by the doctor and the business’s summaries, training and competence records, permits for hot work and confined spaces, incident and near-miss reports and investigations, and equipment inspection and testing records. Some must be kept for long periods: records of exposure to some hazardous substances and of health monitoring may need to be kept for decades, because the diseases they relate to appear late. The coordinator kept them in the shop’s electronic safety system, backed up and linked to each worker and task, and retained them for the periods the regulations require.
The annual review
Once a year, the coordinator and the committee reviewed the whole programme with the owner, using a short checklist.
- Avoid: Risk registerevery task and hazard current; changes in work reflected
- Avoid: Exposuresmeasurements due or overdue; results against standards; actions from high results
- Avoid: Engineering controlsextraction, screens, guarding and electrical equipment tested and maintained
- Avoid: Health monitoringdone as required; summary of findings and actions
- Avoid: Training and competenceall workers current; gaps closed
- Avoid: Incidents and near missestrends by type; investigations completed; actions closed
- Do: ? Next year's planpriorities ranked by risk, with owners, dates and measures
The review’s output was the next year’s plan: the hazards to work on, ranked by their measured or recorded risk, each with an owner, a date and a measure that would show whether it had worked, the same structure as the first year’s plan of Chapter 1.
Bringing it together
This book began with the hazards of welding and the risk process the law expects, and with the observation that the immediate hazards fill the injury record while the delayed ones may do the most harm. It went on to fume: what it is, how much each process makes, how exposure is measured and why one sample proves little, how extraction captures it and why distance is everything, and how respirators protect the wearer only when they fit and stay on. It covered the eyes and skin, where the early manuals’ advice on filters still holds and screens protect the bystanders who suffer most arc eye; electricity, where the same voltage drives a hundred times the current through a wet body; fire, gases and confined spaces, where a small gas leak can change the air in an hour; and noise, heat and strain, the slow hazards welders accept too readily.
The numbers throughout point the same way. Measure what can be measured; control at the source before relying on people and protective equipment; check that controls work by measuring again; and watch the leading indicators rather than waiting for injuries. A welding shop that does those things protects its people, now and decades from now.
A failure: the zero year
This is an illustration. A fabricator elsewhere had a year with no lost time injuries and celebrated it as proof that its safety programme was working, cutting its safety budget for the following year. Its fume extraction had not been tested for three years, it had never measured exposures, and its hearing tests had lapsed. The following year it had three lost time injuries, and a former welder’s claim for occupational lung disease arrived soon after.
The zero year had been chance. The fabricator’s new safety manager started reporting leading indicators to the board, beginning with extraction tests and exposure measurements, and stopped treating a single year’s injury rate as a verdict.
Exercise: rates and indicators
A shop of 18 people works about 1,850 hours each a year. It had 1, 0 and 2 lost time injuries in three successive years. Calculate each year’s lost time injury frequency rate and the three-year rate. What does the comparison suggest about how the shop should judge its performance?
The shop works about 18 × 1,850 = 33,300 hours a year. The yearly rates are 1 × 1,000,000 ÷ 33,300 ≈ 30, 0, and 2 × 1,000,000 ÷ 33,300 ≈ 60.
Over three years, 3 injuries in about 99,900 hours give a rate of about 3 × 1,000,000 ÷ 99,900 ≈ 30. The yearly rates swing from 0 to 60 on single injuries, so they say little about whether the shop is getting safer. The three-year rate is steadier, and leading indicators, such as exposure measurements, extraction tests, closed inspection findings and current training, would show far better whether its controls are in place.
