Repeat Hose Failure: Investigating a Near Miss That Keeps Recurring
At a plating shop, a transfer hose on a 20% sulfuric acid line failed at the crimped fitting during unloading. About 40 L escaped, but secondary containment caught it and no one was injured. The shift supervisor investigated over two days, interviewed only the day-shift operator who connected the hose, and concluded "employee error - hose not fully seated." Corrective actions were to retrain all operators and add a sign at the connection point. Three similar hose failures have occurred in 18 months, all at the same fitting type; the previous investigation after the second event recommended the same retraining. Maintenance records show the hoses are purchased from the lowest bidder and have a five-year shelf life; the failed hose was seven years old and stored outdoors. The investigation report does not include a timeline, physical evidence, or interviews with the unloading contractor. The site uses a simple five-why template and closes actions when training records are complete.
1. Critique the investigation and identify the causal factors it missed.
Read the worked answer
The investigation is not adequate for a repeat near miss. It relied on one interview, did not interview the contractor who may have connected or monitored the hose, did not preserve the failed fitting or hose, and produced no timeline. The conclusion "employee error" stops at the last barrier rather than explaining why the error was possible and why it recurred.
Missed causal factors include: hose age (seven years against a five-year shelf life), outdoor storage and UV or ozone degradation, procurement driven only by lowest bid, absence of incoming inspection or shelf-life tracking, fitting design that permits incomplete seating, lack of a breakaway or locking coupling, and no engineering control to prevent transfer under adverse conditions. The fact that three failures occurred at the same fitting type points to a design or procurement problem, not individual carelessness.
A proper causal analysis would use a timeline, physical evidence, interviews across shifts and contractors, and a logic tree or barrier analysis. It should distinguish immediate causes, contributing factors, and root causes such as procurement and storage practices. The report should also assess why the previous corrective action failed to prevent recurrence - retraining alone did not change the hardware, storage, or procurement system.
2. Propose corrective actions using the hierarchy of controls and specify effectiveness metrics.
Read the worked answer
Higher-order controls should come first. Elimination or substitution: evaluate whether the acid transfer can be hard-piped, whether a safer acid or lower concentration can be used, or whether the hose connection can be eliminated with a fixed transfer line. Engineering: install a locking camlock or breakaway coupling that cannot be left partially seated; provide a drip leg or dry-break connection; ensure secondary containment and ventilation; add an interlock or flow restriction until the connection is verified. Administrative: establish a hose inspection and shelf-life program, store hoses indoors or in UV-protective covers, specify material and pressure rating in procurement, require incoming inspection, and train operators and contractors on connection verification. PPE is the last line and should not be the primary fix.
The previous retraining should be recognized as ineffective because it did not address the age, storage, procurement, or fitting design. Corrective actions should be assigned owners and due dates and verified in the field.
Metrics should include leading indicators: percentage of hoses within shelf life, inspection completion rate, percentage of hoses stored indoors, contractor audit scores, and near-miss reporting rate. Lagging indicators: hose failures per 1,000 transfers, acid releases, and repeat findings. Track them monthly and investigate any recurrence as a failure of the corrective action, not just of the operator.
3. How should the site improve its incident investigation and lessons-learned program?
Read the worked answer
The site needs a defined investigation methodology and competence. Investigators should be trained in evidence preservation, interviewing, timeline construction, barrier analysis, and causal logic. The program should define investigation levels by potential consequence: a near miss with serious potential should receive the same depth as an injury, not a two-day supervisor review. Contractors and temporary workers should be included in interviews and action development.
The five-why template is insufficient when there are multiple contributing factors. A causal tree or barrier analysis should be used for significant events, with a review of prior similar events to detect trends. Reports should include physical evidence, interview notes, timelines, and a clear distinction between immediate and root causes. Actions should be tracked to completion with verification of effectiveness, not closure based on training records.
Lessons learned should be shared across shifts, sites, and contractors, and the site should monitor repeat events as a key performance indicator. A periodic review of investigation quality - for example, an audit of a sample of reports - would identify weak causal analysis and premature closure. Finally, the site should feed investigation findings into hazard identification and management of change so that repeat failures trigger hardware and system changes, not repeated retraining.