Grid Resolution and the Cost Tradeoff

You Can Only See What the Cells Resolve

A grid represents the reservoir only as finely as its cells. A curved sand body, a thin shale baffle, or a sharp permeability contrast finer than a cell simply cannot be captured; the grid stair-steps a smooth shape into a blocky one and averages anything smaller than a cell away. The widget overlays a smooth sand body with its gridded version. Raise the resolution and the blocky approximation tightens onto the true curve.

Resolution Costs, and the Cost Compounds

Why not just make every cell tiny? Because the cell count, and the simulation cost, grow explosively. Halving the cell size in all three directions multiplies the cell count by eight, and the time to run a flow simulation grows faster still, because more cells also force smaller time-steps. A model fine enough to resolve every lamina would never finish. Resolution is a budget.

Spend Resolution Where Flow Needs It

Good grid design spends that budget where it changes the answer. Fine cells go where the flow is fast or the geometry is critical (near wells, across thin high-permeability streaks, around faults), while distant, quiet rock is left coarse. The static geological grid is often built fine to capture the geology, then upscaled to a coarser simulation grid, a step we devote a later chapter to. Knowing what to resolve, and what to coarsen, is much of the craft.