Planar longrange correction affects surface tension in evaporation

[IsaNiTU]

Describe the bug
In a stationary evaporation scenario of methane, the surface tension depends on the thickness of the liquid phase. This seems to be caused by the planar lrc. The region sampling plugin is used to get the normal and tangential component of the pressure tensor. The difference of both is supposed to be zero in the phases and only non-zero in the interface region. In the liquid, this difference is larger than zero if the lrc is used, but equal to zero, if the lrc is not used.

To Reproduce
A scenario with input files is attached. For a liquid thickness of 150, the surface tension is 9.06 and if the thickness is doubled to 300, it raises to 10.22 which is clearly unnatural. Used branch is "surface_tension".
CH4_evap.zip

Expected behavior
Pressure difference is supposed to be zero such that surface tension is not affected by system size.

Screenshots

Build environment (please complete the following information):
gcc (Debian 10.2.1-6) 10.2.1 20210110

Additional context
The problem is not related to the evaporation. It also happens if a VLE is set up like this. The mere presence of the liquid reservoir for the mett deamon might be a problem for the lrc.