Currently there is some confusion about the validation of the datafile, wells and the eclbase (concept), which should be improved. In the future we might want to split all reservoir simulator specific into a new section of the configuration file.

Some background:

To balance between the reservoir specific tooling in Everest and its general purpose capabilities we need a sane way to detect whether this is indeed a reservoir setup or not 🤷 Currently that is done by checking for the datafile key in the configuration. This decision can probably be documented and communicated better... It can also be discussed whether is should be changed.

Even an empty datafile will today result in all field keywords being internalized. And if wells are added in the wells section they will be internalized as well. But to have some kind of "toggle" for activating the reservoir stuff makes a lot of sense to me at least...

Reply:

If Everest is given an empty datafile Everest will not fail it will only internalize the default keys, or at least try to do it and fail if also there is is no wells section defined.

For now, would it be acceptable either:

To assume that if a wells section is present in the config file we are dealing with a reservoir setup.
Have an actual flag in the config file that needs to be toggled to internalize any keys.
Keep things as they are now and just document it better, even though currently there is a Note in the Everest documentation that explains that without the datafile there will be not well or group specific summary keys internalized