This work presents an implementation and evaluation of an alternative approach for describing exchange of mass, momentum, and energy in diesel spray computational fluid dynamics (CFD) simulations using discrete droplet modeling (DDM). During the calculation, each parcel in the domain is surrounded by a spherical volume of ambient gas and interacts first with it instead of interacting directly with the cell volume hosting the parcel. In this way, the interaction volume is independent of the mesh and can be located in more than one cell. This model was implemented using the Open-FOAM CFD opensource C++ library. It was developed with the aim to reduce grid dependencies related to spray-grid mutual orientation and to the choice of the injector nozzle position with respect to the cell hosting it. All the submodel constants were set to match experimental data of a chosen baseline case in nonreactant vaporizing conditions. Then the new approach predictions were first compared to standard DDM on moving the injector position within the hosting cell and later on varying ambient density and injection pressure of fuel. Also, a study of the dependency of the results on the spray-grid mutual orientation was carried out. High-speed imaging and Rayleigh-scattering measurements taken from the engine combustion department (ECN) web database were used to assess numerical results: a good accuracy in the predictions of liquid and vapor spray penetration as well as axial and radial mixture fraction profiles, can be simultaneously achieved on varying thermophysical and geometrical settings. If applied to engine calculations, then the reduced dependency on the nozzle position becomes appreciable when injector with multiple nozzles are used.