The Engineer Who Signed What He Didn't Check
You are a senior engineer on a building services package. A junior engineer, under deadline pressure, stamped a commissioning certificate confirming that fire dampers had been inspected and tested. In fact, sampling records show only 40% of dampers were tested; the rest were 'assumed compliant' because they were from the same batch. The junior engineer told you he was 'told to get it signed' by the package manager. The certificate has already gone to the client and the building is due to be occupied in three weeks. The package manager is a long-standing colleague and has asked you to 'let it ride - the batch is fine, we all know it.'
1. What are the ethical issues, and what must you do first?
Read the worked answer
This is a false certification of a life-safety system. The issues are:
- The certificate is untrue on its face. It asserts inspection and testing that did not occur for 60% of dampers. 'Same batch' is an assumption, not a test, and batch conformity does not establish installation correctness - dampers fail on install, not just manufacture.
- Pressure to certify without evidence is a professional-conduct breach, and the instruction came from the package manager. The junior engineer is junior and was directed; responsibility does not sit with him alone.
- The building is three weeks from occupation. Fire dampers are a life-safety element; a false certificate creates a risk to occupants and exposes the firm and the client.
- The request to 'let it ride' is the moment the problem becomes yours. Agreeing would make you a party to it.
First action: do not conceal or delay. Immediately notify your line manager and the package manager in writing that the certificate is not supported by records and must be withdrawn or corrected. Do not wait for the three-week deadline to force the issue - the shorter the runway, the more dangerous silence becomes.
2. The package manager says withdrawing the certificate will delay occupation and cost the client money. How do you proceed?
Read the worked answer
Delay and cost are real, but they do not change what the certificate says. A false certificate is not a schedule tool.
Proceed in this order:
- Correct the record. Notify the client in writing that the certificate was issued in error and is withdrawn pending completion of testing. Own it plainly - do not blame the junior engineer. The firm issued it.
- Test the remaining 60%. This is the only route to a true certificate. Mobilize the resource now; it is the fastest path to a defensible position. Do not substitute a desk review for physical testing on a life-safety system.
- Quantify the real delay and give the client a recovery plan. Three weeks may be enough to test and re-certify if action starts immediately - which is an argument for acting today, not deferring.
- Escalate if overruled. If the package manager or anyone else blocks correction, escalate up the line and, if necessary, to the relevant professional body. Document each step.
- Address the cause. The instruction to 'get it signed' is a systemic problem; raise it so the same pressure does not produce the same result on the next package.
The colleague relationship is not a factor in the decision. It is a factor in how you communicate it - directly and privately first, but the correction happens regardless.