Recording a Startup That Takes Two Days
Emerson's AMS 6500 ATG 5.0 adds continuous unlimited data transmission into AMS Machine Works 2.4, so a turbomachine's startup and shutdown can be captured end to end rather than in samples.
Emerson has released AMS 6500 ATG version 5.0 and AMS Machine Works version 2.4. The significant change is continuous, unlimited transmission of machine data from the protection system into the analytics platform, so that startup and shutdown events lasting hours or days can be captured end to end. The protection system also gains a web-browser-native programmable dashboard supporting up to five pages with as many as 25 widgets each, including trending displays, Boolean alarm screens, live measurement fields and machine component visualisations, plus Modbus-independent connectivity. Ben Swisher, vice president and general manager of reliability solutions, is quoted.
Why that matters requires knowing what a machinery protection system is for. Under API 670 it is a safety function: it watches shaft vibration, position and temperature on a turbine, compressor or large pump, and it trips the machine when a threshold is crossed. Its job is to be reliable and boring, and historically the condition-monitoring system that does diagnosis has been a separate box tapped off the same transducers, because the protection path could not be burdened with data collection.
The diagnostic value, though, is concentrated in exactly the events a sampled system misses. A machine at steady state gives up very little information; almost everything interesting happens during a run-up or a coast-down, when the rotor passes through its critical speeds and the response reveals unbalance, rubs, cracks, bearing condition and misalignment in a way no steady-state reading can. A run-up on a large turbomachine can take many hours, and a system that captures periodic snapshots will reliably miss the transient that mattered. Continuous capture across the whole event is what allows a proper Bode or waterfall plot, and that is the instrument a rotating-machinery specialist actually uses.
Two practical points follow. Machines that start and stop rarely — which is most large turbomachinery — offer very few of these opportunities, so a missed run-up is a missed year, and that asymmetry is the whole argument for the upgrade. And the Modbus-independent connectivity is worth reading as a maintainability change rather than a feature: Modbus register maps between a protection rack and a monitoring server are a well-known source of silent mismatch, where a register moves after a firmware update and a value keeps being read from the wrong address. Removing that layer removes a class of fault that is hard to notice and easy to distrust the whole system over.