The Paraguaná Refining Center, the union of the Amuay and Cardón refineries on Venezuela's Paraguaná peninsula and historically among the largest refining complexes in the world, has undergone a reliability collapse without precedent in the modern downstream industry, culminating in events such as the 2012 Amuay vapor cloud explosion — more than fifty fatalities, damage exceeding one billion US dollars, and an origin traced to a hydrocarbon leak following a pump-related containment failure (Sajid et al., 2021). The aim of this article is to explain and quantify how the systematic absence of preventive maintenance of pump and compressor sys- tems translates into the degradation of refining reliability. The research applies a mixed method- ological approach: a structured synthesis of the failure-mode and condition-monitoring literature for refinery rotating equipment; a documentary reconstruction of the maintenance-deprivation context of Venezuelan refining after 2013; and a model-based reliability simulation of a repre- sentative Paraguaná-scale rotating-equipment population, whose parameters are anchored to published failure data — bearing-dominated pump failure distributions of approximately 38%, Weibull characteristic lives of 510–657 operating days, and valve-dominated compressor failure structures of approximately 36% (Silva et al., 2023; Lv et al., 2021). The original contribution of the article is the Cumulative Maintenance Deficit Index (CMDI) — a criticality-weighted measure of unexecuted scheduled maintenance accumulated over time — together with an empirically fitted exponential degradation law, λ(t) = λ0·exp(κ·CMDI), which on the simulated dataset yields κ = 2.14 (95% CI: 1.87–2.41) with R2 = 0.93, implying that each 0.32 increment of maintenance deficit approximately doubles the failure intensity of the rotating-equipment population. Key findings: reliability decline under maintenance deprivation is strongly nonlinear; the failure-mode structure shifts from wear-out of monitored components toward systemic lubrication, sealing, and auxiliary failures; and workforce emigration acts as an independent degradation channel that spare-parts availability alone cannot offset (Garcia Zea, 2020; Gonyora & Ventura-Medina, 2024). All simulation outputs are explicitly model-generated and are intended as a transferable analytical template to be repopulated with plant data.