Grid North Is Not True North
Distance is not the only thing a projection quietly bends. Point a compass, or better, compute a bearing from two GNSS positions, and compare it against the grid bearing shown on a plan for the same line. The two directions do not point the same way either, and the gap between them has a name of its own.
A projection's grid north, the direction its north-south grid lines actually point, is not the same direction as true north, the direction to the geographic pole. The angle between them, called convergence, is zero along a projection's central meridian and grows steadily east or west of it.
Why Straight Lines on a Grid Cannot Stay True
A projection's grid lines run perfectly straight and perfectly parallel to each other, by design, since that is exactly what makes a flat, subtractable coordinate system useful in the first place. Lines of true north on the actual curved Earth are not parallel to each other at all, they converge toward the pole. Forcing a straight grid onto a converging reality means the grid's own north-south lines can only point exactly at true north along one line, the central meridian. Move east or west of it, and grid north tilts slightly away from true north, more so the farther out a project sits.
Convergence is not a defect in the projection, any more than grid scale factor is. It is the direct, unavoidable price of representing a converging, curved reality with a set of perfectly straight, parallel lines.
A Bearing Problem, Not a Distance Problem
Convergence does not stretch or shrink a distance the way scale factor does. It rotates a direction. A grid bearing pulled from a plan, and a bearing computed directly from two GNSS positions relative to true north, can disagree by several arcminutes at a typical distance from a zone's central meridian, an amount easy to dismiss on a single short line and easy to compound into a real heading error across a long one, a survey traverse, or a pipeline alignment that runs for miles.
Convergence and grid scale factor are not the same correction applied to two different quantities. Scale factor answers how far apart two points are on the grid versus on the ground. Convergence answers which direction the grid itself is actually pointing at a given location, a completely separate question with its own separate correction.
Where This Shows Up in Practice
Any bearing pulled directly off a plan and compared against a compass, a solar observation, or a GNSS-derived azimuth is implicitly comparing a grid direction against something closer to a true one. Long, narrow projects, pipeline corridors and transmission lines especially, are the most exposed, since they often run a considerable east-west distance from a zone's central meridian, exactly where convergence grows largest.
A pipeline alignment computed from GNSS-derived bearings is compared against a plan's stated grid bearings, and the two disagree by a small, fairly consistent angle along the entire route. Does that point to a bearing measurement error? Not necessarily. A consistent angular offset that holds steady across a long, roughly east-west route is a recognizable signature of unaccounted convergence, not of any individual bad observation.
Convergence is the angle between grid north and true north, zero at a projection's central meridian and growing with distance from it. It rotates directions rather than stretching distances, a separate correction from grid scale factor, and it matters most for bearings on long, east-west-running projects.
Distance and direction both bend under a projection. The next distortion is not geometric at all. It is a units problem, quietly built into the very numbers a project has been trusting the whole time.