Galvanic Corrosion

I am a controls engineer, but the company I work for builds ss manifolds for process skids and use a lot of ss transmitters. We use a di-electric union if we go to copper, but I've never seen one for ss to carbon steel, never had a customer require one, and certainly have never installed one for ss to carbon steel. So, this does not say it is not necessary, but does imply that one may not be necessary here. Google dielectric union for ss to carbon steel and see what you can find. I couldn't find one.

Galvanic corrosion potential for most stainless steels against carbon steel is in the range of 0.3-0.4 V. This is considered to be likely to corrode the weldolet. That potential is probably calculated at standard conditions, probably considerably warmer than your chilled water loop (meaning that you should experience a lower rate of corrosion).

Usually, chilled water loops contain treament additives that should inhibit the corrosion reaction by shutting down the anode (clogs the anode to the extent that no further ion transport is possible). The potential of such an anode will become polarized to a more cathodic potential at that point. In chilled water loops, it is not unusual to find corrosion rates lower than 0.1 mil/yr (0.0025 mm/yr).

The current path between galvanic pairs is through the metal. Therefore, if you use an insulating thread sealant (as Teflon) the current path is also minimized, but only if the threads do not cut through the teflon during installation. Keep moisture off the outside surface of the threaded joint, as this can be a major source of uncontrolled corrosion in threaded fittings.

The general experience thus far has been that this is never more than a minimal problem. If still concerned, you have the option to install an isolating union, as others suggested.