This paper develops a structural and combinatorial interpretation of Goldbach partitions that complements the standard prime-based counting method g(N). Insteadof scanning for primes, the approach decomposes Goldbach’s function into three components: (1) potential residue-compatible pairs, (2) pairs eliminated (“blocked”) bycomposite occupancy in the 6k ± 1 columns, and (3) the surviving realized Goldbachpairs.Central to this framework are three composite-counting functions F0(N),F1(N),F5(N),which measure composite density in specific residue classes modulo 6. These functionsallow g(N) to be interpreted as:g(N) = Potential Pairs(N) − Bad Pairs(N),where bad pairs are deletions caused by composite interference in residue-structuredcolumns. This yields a transparent “pair deletion” model of Goldbach partitions thatexplains why the number of prime pairs grows approximately linearly and why composite blocking alone cannot eliminate all candidate pairs within any sufficiently largeprimorial segment.To support this framework, the paper introduces the Primorial Calendar, a conceptual structure for visualizing composite blocking, prime accessibility, and residue-classbehavior across iterative primorial ranges. Empirical graphs for the functions F0,F1,F5demonstrate stable patterns that reinforce the decomposition. The presentation aimsto clarify the internal mechanisms behind Goldbach pair formation, offering an explanatory narrative and heuristic structure rather than a proof.