Selecting circuit breakers by current, cables by ampacity, and transformers by capacity—most engineers get these basics right. But what truly causes headaches in distribution engineering are not these primary parameters, but the easily overlooked peripheral issues that don’t appear on product catalogue covers and are seldom mentioned by sales engineers. Yet when they go wrong, the consequences range from equipment malfunction to catastrophic failure and safety incidents. This article distills four of the most common and most overlooked engineering issues in distribution selection, helping you navigate around these hidden traps. The first issue is that voltage is not as simple as “380V” — there are hidden details within the voltage specification. “Three‑phase four‑wire, 380V” describes most industrial distribution, but in practice, voltage matching is far more nuanced. Rated voltage versus actual operating voltage: the nameplate value is a design figure, but real system voltage fluctuates—transformer output is typically 5% above nominal, while line ends under heavy load may drop to 360V; equipment rated 380V ±10% accepts 342‑418V, but if you have imported equipment rated 400V ±5% and your system runs at 430V under no‑load, the power supply may fail from overvoltage—always measure actual operating voltage rather than blindly trusting the drawing. Voltage sags are more dangerous than blackouts: many users focus on outage risk but ignore voltage dips where voltage drops to 70% or less for a few cycles and recovers; this “flicker” phenomenon often damages variable frequency drives, contactors, and PLCs more than actual blackouts, causing unexpected shutdowns and motor surges; if your process is sensitive to sags, consider ride‑through measures such as UPS, dynamic voltage restorers, or VFDs with low‑voltage ride‑through capability. The second issue is that IP protection ratings are not “higher is better” — the wrong choice is worse than none. The common misconception that IP54 is always superior to IP30 leads to sealed cabinets with poor heat dissipation; if the cabinet contains heat‑generating devices like VFDs or soft starters, IP54 can raise internal temperatures by more than 10°C over IP30, and every 10°C rise roughly halves electronic component life; if you don’t need dust or water protection, IP30 or IP40 may actually support longer equipment life. Also, IP rating alone doesn’t address condensation—if a room’s humidity exceeds 85%, even IP54 cabinets may suffer internal moisture problems because condensation forms from internal air moisture on cooler surfaces, unrelated to external splashing; in such cases, anti‑condensation heaters and slight positive‑pressure ventilation are required, not just a higher IP grade. The third issue is that grounding is not just “connect it to ground” — the grounding system type determines protection configuration. In TN systems, the transformer neutral is directly grounded and equipment protective earth connects to this neutral; a single‑phase ground fault produces a large fault current that quickly trips the breaker. In IT systems, the neutral is ungrounded or high‑impedance grounded; a single‑phase fault produces only a small current, allowing continued operation (typically limited to 2 hours for fault clearance). IT systems are used in applications with extreme continuity requirements—operating rooms, continuous steel production, and HVDC systems—and require insulation monitoring devices, with leakage protection adapted to the system type. Regarding grounding resistance, while lower is generally better, the question is “how low must it go” for compliance—different standards vary from below 1Ω to as high as 4Ω; achieving below 1Ω may require deep excavation, chemical treatment, or soil replacement at enormous cost, so target a value appropriate to the system and local code rather than obsessing over minimal resistance. Repeated grounding and equipotential bonding are two often‑overlooked concepts: repeated grounding means multiple earth connections on the neutral conductor to reduce shock risk if the neutral breaks; equipotential bonding connects all conductive metal parts (enclosures, pipes, structural steel) to the same ground potential to prevent dangerous voltage differences during faults—many designs address equipment grounding but neglect bonding, leading to equipment damage or shock hazards. The fourth issue is electromagnetic interference — the invisible ghost. The distribution system is both a transmitter and receiver of EMI, with VFDs, rectifiers, UPS, and switch‑mode supplies generating harmonics and high‑frequency noise that affects other equipment on the same system via conducted or radiated paths. Symptoms include intermittent PLC freezes, flickering touchscreens, high communication error rates, or jumping meter readings without any clear electrical fault; the most frustrating aspect is that symptoms are sporadic and often disappear when you bring the test equipment in. EMI has three elements: source, coupling path, and susceptible device—treat it by cutting any one. Common measures include installing EMC filters at VFD input and output; separating control and power cables by at least 30cm; using shielded cables with shield grounded at one end; and installing isolation transformers or power line filters at sensitive device inputs. If none of these work, more radical measures may be needed—such as separating “dirty” loads (VFDs, rectifiers) from “clean” loads (PLCs, instruments, communications) right from the transformer secondary with dedicated busbars or even separate transformers. Distribution selection is not a simple checklist of catalogue parameters. Verified actual voltages, rational IP choices, correct grounding system identification, and effective interference suppression—these invisible details often determine project success far more than the primary ratings. Investing extra time on these issues during design will pay off many times over during years of reliable operation.