THE TERMS, which must be distinguished precisely: HAZARD — something with the POTENTIAL to cause harm. An unguarded opening in a floor. RISK — the LIKELIHOOD that the hazard causes harm, combined with the SEVERITY if it does. INCIDENT — an unplanned event. A NEAR MISS is an incident that caused no injury, and it carries the same information about the system as one that did. ACCIDENT — an incident that resulted in harm. ── THE ACCIDENT TRIANGLE ─────────────────────────────────── Frank Bird's 1969 study of 1.75 million reported incidents found the ratio: 1 serious or major injury 10 minor injuries 30 property-damage incidents 600 NEAR MISSES with no damage at all (Heinrich's earlier 1931 study gave 1 : 29 : 300.) THE MANAGERIAL CONSEQUENCE IS THE POINT OF THE TRIANGLE: BENEATH EVERY FATALITY LIE HUNDREDS OF UNREPORTED WARNINGS. A site with no reported near misses is not safe — IT IS A SITE THAT IS NOT REPORTING. This is why mature safety systems actively encourage near-miss reporting and treat a rising near-miss count as a good sign rather than a bad one, which is counterintuitive and correct. ── HAZARDS IN CONSTRUCTION — "THE FATAL FOUR" ────────────── In construction fatality statistics worldwide, four causes dominate: 1. FALLS FROM HEIGHT — the single largest killer 2. STRUCK BY falling or moving objects 3. CAUGHT IN OR BETWEEN — trench collapse, machinery 4. ELECTROCUTION THE FULL RANGE: PHYSICAL — falls, falling objects, machinery, vehicles, EXCAVATION COLLAPSE, scaffolding failure, formwork failure, confined spaces, noise, vibration, heat and cold ELECTRICAL — contact with live conductors, overhead lines, temporary site supplies CHEMICAL — cement dermatitis, solvents, paints, ASBESTOS in demolition, silica dust BIOLOGICAL — contaminated ground, sanitation ERGONOMIC — manual handling, repetitive strain, posture PSYCHOSOCIAL — long hours, fatigue, stress, isolation from family for migrant workers NOTE THAT OCCUPATIONAL HEALTH IS SLOWER AND LARGER THAN OCCUPATIONAL SAFETY: silica dust and asbestos kill more construction workers over time than falls do, but the deaths occur decades later and are not attributed on site. THIS IS WHY HEALTH SURVEILLANCE MATTERS AND WHY IT IS NEGLECTED. ── THE HIERARCHY OF CONTROL ──────────────────────────────── THE SINGLE MOST IMPORTANT FRAMEWORK IN THE TOPIC, and it is ordered by EFFECTIVENESS, not by convenience: 1. ELIMINATION — remove the hazard entirely. PREFABRICATE AT GROUND LEVEL so no one works at height. THE ONLY CONTROL THAT CANNOT FAIL. 2. SUBSTITUTION — replace with something less hazardous. A water-based solvent for a flammable one. 3. ENGINEERING CONTROLS — isolate people from the hazard. Guardrails, trench shoring, machine guarding, ventilation, edge protection. THESE WORK WITHOUT REQUIRING ANYONE TO REMEMBER ANYTHING, which is why they rank above the two below. 4. ADMINISTRATIVE CONTROLS — change how people work. Permits to work, training, signage, rotation, method statements. DEPENDENT ON COMPLIANCE, therefore weaker. 5. PERSONAL PROTECTIVE EQUIPMENT (PPE) — helmets, harnesses, boots, eye and hearing protection. THE LAST RESORT, AND THE WEAKEST, because it protects only the person wearing it, only if worn, only if correctly fitted, and it does nothing to prevent the event. THE ERROR THAT DEFINES POOR SAFETY PRACTICE IS STARTING AT LEVEL 5. Issuing helmets is visible, cheap and immediate; designing the work so nobody is beneath a load is none of those things and is enormously more effective. A SITE WHOSE ENTIRE SAFETY PROGRAMME IS PPE HAS NOT MANAGED ITS HAZARDS AT ALL. ── SAFETY IN DESIGN ──────────────────────────────────────── The most powerful intervention available to an engineer, and it is exercised in the design office rather than on site: · design components to be lifted and fixed without working at height · design for safe MAINTENANCE ACCESS over the whole life — the person who cleans the gutter in year twenty · specify prefabrication to move work from an uncontrolled site to a controlled factory · design for safe DEMOLITION at end of life THIS CONNECTS TO THE LIFE CYCLE TOPIC'S CROSSING CURVES: THE ABILITY TO INFLUENCE SAFETY IS HIGHEST AT DESIGN AND LOWEST ON SITE, which is exactly where most safety effort is currently spent.

── THE SAFETY MANAGEMENT SYSTEM ──────────────────────────── ISO 45001 is the international standard. Its elements: · a POLICY signed at the top, because safety culture is set by what senior management demonstrably cares about · HAZARD IDENTIFICATION AND RISK ASSESSMENT, documented · a JOB SAFETY ANALYSIS or METHOD STATEMENT for each significant activity · PERMIT-TO-WORK systems for high-risk work: confined spaces, hot work, live electrical work, excavation · TRAINING AND INDUCTION — nobody on site untrained, including visitors · TOOLBOX TALKS — short, frequent, task-specific briefings · PPE provision, maintenance and enforcement · EMERGENCY PREPAREDNESS: first aid, rescue plans (including RESCUE FROM A HARNESS, which is a distinct and often forgotten hazard), fire, evacuation · INCIDENT REPORTING AND INVESTIGATION · INSPECTION AND AUDIT · CONSULTATION with workers, who know the hazards best · REVIEW AND CONTINUAL IMPROVEMENT ── MEASUREMENT ───────────────────────────────────────────── LAGGING INDICATORS — measure what has already happened: ACCIDENT FREQUENCY RATE = (lost-time injuries × 1,000,000) / man-hours worked Example: 7 lost-time injuries in 1,400,000 man-hours = 5.0 per million hours SEVERITY RATE = (days lost × 1,000,000) / man-hours worked Example: 210 days lost in the same period = 150 Fatality rate; incidence rate. LEADING INDICATORS — measure what predicts future outcomes: near misses reported, inspections completed, training hours, toolbox talks held, hazards closed out, percentage of work with a valid method statement. THE ARGUMENT FOR LEADING INDICATORS: LAGGING INDICATORS MEASURE FAILURES THAT HAVE ALREADY OCCURRED, AND ON A SMALL SITE THEY ARE STATISTICALLY MEANINGLESS — zero accidents in a year may mean excellent management or a small sample. Worse, rewarding a low accident rate creates a direct incentive to UNDER-REPORT, which destroys the information the system depends on. Leading indicators cannot be gamed in the same way because they measure activity that is either done or not. ── THE ECONOMIC ARGUMENT ─────────────────────────────────── Safety is often presented as a cost. THE ACCOUNTING SAYS OTHERWISE ONCE THE INDIRECT COSTS ARE COUNTED: DIRECT: medical treatment, compensation, insurance premiums INDIRECT, and typically SEVERAL TIMES the direct cost: lost production, work stoppage and investigation time, replacement and retraining, damaged plant, legal costs, delay and liquidated damages, reputational harm and exclusion from future tendering, loss of morale THE ICEBERG IMAGE IS APT: the visible direct cost is the small part above the water. BUT THE ARGUMENT SHOULD NOT REST THERE. THE PRIMARY REASON FOR SAFETY IS NOT ECONOMIC. A worker's life is not a cost line to be optimised, and an engineer who defends safety solely on financial grounds has conceded that the case would collapse if the arithmetic changed. ── THE NEPALI CONTEXT ────────────────────────────────────── LEGAL BASIS: the LABOUR ACT 2074 (2017) and Labour Rules place a duty on employers to provide a safe workplace, require safety committees in larger workplaces, and provide for compensation. The Social Security Act provides for an employment injury scheme. Contract conditions on public works impose HSE requirements on contractors. THE PRACTICAL SITUATION, stated honestly: · a large INFORMAL SECTOR outside effective regulation · widespread absence of basic fall protection on small building sites · limited inspection capacity · under-reporting of injuries, so statistics understate the problem · migrant workers, who often work with the least protection and the least recourse THE ENGINEER'S POSITION IS UNAMBIGUOUS DESPITE ALL OF THIS: the professional duty to public and worker safety does not depend on whether an inspector is likely to visit. A SUPERVISING ENGINEER WHO ACCEPTS UNSAFE WORK HAS FAILED PROFESSIONALLY WHETHER OR NOT ANYONE IS INJURED — because the failure is in permitting the exposure, and whether it results in harm is a matter of luck rather than of conduct.