Isa-tr84.00.09 File

The industry’s answer then was a shrug. The answer today, after TRITON, PIPEDREAM, and a dozen state-sponsored near-misses, is: catastrophe . For decades, functional safety engineers operated under a sacred pact: A safety system (SIS) must be fail-safe, deterministic, and isolated. If you pulled the logic solver’s plug, the valves went to their safe position. If a sensor failed, the system defaulted to shutdown. Safety was about physics, random hardware failures, and reliability.

Cybersecurity wasn’t part of the equation. Why? Because the assumption was that safety networks were air-gapped, proprietary, and obscure. No hacker would bother with a Beckhoff controller or a Triconex when they could go after corporate payroll. isa-tr84.00.09

A SIL 3 loop (one failure in 10,000 years) is mathematically robust against random hardware failures—but completely blind to a single malicious write command over Modbus TCP. TR84.00.09 introduced the concept of for security, arguing that a safety function can only claim its SIL if the supporting cybersecurity controls maintain the integrity of the logic, data, and timing. The industry’s answer then was a shrug