The Nancy Grace Roman Space Telescope has a planned duration of five years. Is this a hard limit due to expendables on the spacecraft; it is an infrared telescope so presumably will have a coolant, also it will be station keeping at L2 so there will be propellant issues, or is it a budget limitation that may be extended later?
What limits the duration of the Roman Space Telescope mission?
Murphy's law: "Anything that can go wrong will go wrong."
Sensors can stop working correctly, thrusters can fail off, moving parts (e.g., reaction wheels) can get sticky, avionics (computers and communications systems) can get fried, and so on. All kinds of things can and eventually do go wrong with spacecraft.
For a spacecraft to have an expected life of five years, each of the critical components must have an expected life significantly greater than five years. The planned mission duration of a spacecraft that is still under construction dictates the kinds of equipment that should be used and how much redundancy needs to be built in.
The planned mission duration is a metric for success. As spacecraft designers do not want missions to fail, they typically design and build spacecraft to remain healthy throughout the planned mission duration. That means the spacecraft most likely will remain operational well past the planned mission duration. As an extreme example, the rovers NASA sends to Mars often have a planned mission duration of only 90 days. There are so many things can might go wrong that were not foreseen that a rover that lasts only 90 days is deemed a success. The most recent set of Mars rovers have continued to operate for years.
The Roman Space Telescope (formerly WFIRST) will have one very hard limit on its useable life, the consumables (e.g., propellant and pressurant) needed for orbit maintenance. The consumables have been sized for a mission life of ten years rather than five.
Maybe some overkill for this question, but here is full mission design report of 2013:
The design had evolved since than, the latest I could find is 2016 presentation:
Compared with Spitzer telescope NGR will have no evaporating cooler, but closed-type Brighton cooler. (Would not need ever this, but spy telescope assembly was designed for ambient temperature 280 K, and the sensor should be at 100 K).
From experience of other space missions probably the most vulnerable parts are gyroscopes. Other subsystems usually degrade slowly and predictably.