The most typical situation is a size refuge, i.e. prey/hosts that are too large or too small to attack (or possibly too small to be useful) are not attacked, or are attacked less, than hosts that are the natural enemy (parasitoid/predator)'s preferred size.
Walde et al (1989) considered whether an observed spatial refuge pattern (less parasitism on the trunk and woody branches of trees) could be explained by a size refuge (i.e., hosts on the trunk and woody branches were smaller on average, so may have escaped predation due to their size)
Depending on what you mean by "situation" and "non-physical", other examples are
- inversely density-dependent attack rates: if hosts occur in patches of different population density, and if natural enemies are not disproportionately attracted to high-density patches, then hosts in higher-density patches will have lower per capita rates of parasitism (because the risk of parasitism is shared by all of the individuals in the patch) (Price 1988)
- gall wasps in galls with thicker walls are better protected from parasitoids (Ito and Hijii 2002)
Ito, M., and N. Hijii. “Factors Affecting Refuge from Parasitoid Attack in a Cynipid Wasp, Aphelonyx Glanduliferae.” Population Ecology 44, no. 1 (2002): 23–32. https://doi.org/10.1007/s101440200003.
Price, Peter W. “Inversely Density-Dependent Parasitism: The Role of Plant Refuges for Hosts.” Journal of Animal Ecology 57, no. 1 (1988): 89–96. https://doi.org/10.2307/4765.
Walde, Sandra J., Robert F. Luck, Dicky Sicki You, and William W. Murdoch. “A Refuge for Red Scale: The Role of Size-Selectivity by a Parasitoid Wasp.” Ecology 70, no. 6 (December 1989): 1700–1706. https://doi.org/10.2307/1938104.