Barring those who have "smartened up" the liquid feed valves: At less than 30% load on a Recirc Arrangement, The Real Recirc Rate is going to be something like 10 or 12 to 1; presuming the original objective was 3 to 4:1 at design load.

At low Mass Flux The ratio between whats tolerable as a return main pressure loss and whats required to lift liquid gets smaller as temperatures reduce......That is: if a 12" main is sized for a temperature change of 2-deg at 0, the temperature penalty at Minus 40 is a whole lot larger for the same refrigeration effect.

At low loads: the smaller pipe takes all the flow, the larger being sealed at the trap. As the load increases, the pressure difference through the small pipe increases and the liquid rises in the larger riser...With a lot of temperature difference available the larger riser will eventually clear; that requires to throw much of its liquid to the higher level.

If the small pipe is tied to the top of the lower main: vapor will pass over the top of a considerable liquid level to reach the small pipe so liquid will accumulate in the upstream pipe. Once the load increases this accumulation will start travelling again. At a really low percentage of the design load, the smaller pipe will seal at its trap and the upstream pressure will increase until 1) the heat exchanger runs out of temperature difference or 2) the smaller trap gets cleared.

On the "in between": the bigger pipe successively accumulates; burps clear, flows for some interval depending on the real load, the difference in levels and the available temperature differences; the return main pressure difference almost disappears; the evaporator temperature drops; the "real" load increases; the cooled medium gets cold and the whole cycle starts over again.

With process freezers: This tends to happen at line breaks that is everybody goes to lunch and the freezer has no "warm" product load. When the line resumes, the product load can also be light until the freezer is back to (say) 30% belt coverage. Then both traps clear as the upstream pressure rises and the vessel at the other end needs to be able to accomodate the liquid...As well as the compressors have to be able to pick up on the higher CFM flow as the vapor from the freezer suddenly "looses" half its evaporating pressure. In worst cases, the compressors will start tracking the liquid accumulation and the changes in return pressure.