It should be noted that there are also some special applications where an ejector may be fully steam jacketed but for reasons other than preventing the ejector from icing up. Sulfur pit ejectors, as an example, are fully steam jacketed because the sulfur in the process stream solidifies at 240 ºF and the entire ejector must be heated above that temperature to prevent the process from solidifying. In all instances, if the ejector is designed with a steam jacket, then that steam jacket is designed to be run continually. If there is any doubt as to whether or not a steam jacket is needed please consult with the ejector’s manufacturer.

The purpose of a steam jacket is to prevent the motive steam from forming into ice on the inside of the ejector. When ice forms on the inside of an ejector, that ice distorts the internal geometry negatively impacting the ejector’s performance. H2O does not need to condense from a gas(steam) into a liquid(water) before freezing into a solid(ice). There is a point on the curve where H2O in the form of gas(steam) will go through a process called desposition, where it transforms directly from a gas(steam) into a solid(ice). The point where it starts to have the ability to make this change is called the triple point. This point is directly related to the operating pressure and temperature, which happens to be a temperature of 32 ºF and a pressure of 4.59 mmHgA. See the chart below for reference.

The point at which one needs to worry about ice build-up is therefore directly related to the ejector’s operating pressure. Ejectors that have a design suction pressure approaching the triple point pressure of 4.59 mmHgA are normally constructed with a steam jacket. That jacket could be a fabricated jacket or coiled tubing, both of which apply heat to the outside of the ejector’s diffuser to prevent ice build-up.  In most systems only the first stage ejector or first couple of ejectors operate at a deep enough vacuum to start freezing, which is why not all of the ejectors in a system are steam jacketed.

Ejectors at low loads will pull a deeper vacuum than what they would at their design loads. It’s possible for an ejector that isn't normally going to be operating at a deep enough vacuum to build ice, to pull a deeper vacuum because of low loads and ice up. Increasing the ejector loading normally corrects this issue but if the user intends to regularly operate at these very low loading points, steam jacketing may be required.

How do I know if my ejector is freezing up? Taking a suction pressure and comparing it to the triple point pressure of 4.59 mmHgA can identify if you are at risk of icing. An external diffuser temperature can also be taken to determine if it’s below 32 ºF. If your external diffuser temperature is cooler than 32 ºF than the internal temperature is also cooler than this. If your diffuser is cooler than 32 ºF you may also start to freeze moisture in the air to the outside of the diffuser. The ejector behavior will also start to become erratic as it freezes up. Normally the ejector will build up ice over a period of time as the pressure degrades. Then the ejector will break performance and heat up causing that ice to melt off. At that point the ejector suction pressure will pull back down and the process repeats itself. This is normally seen as a pressure spike in the system that repeats itself on a cyclic basis.