Normal power is the everyday utility-derived electricity that energizes the vast majority of a healthcare facility — everything that is not on the Essential Electrical System. How the service enters the building, how it is distributed through the medium- and low-voltage network, and how loads are organized and estimated are foundational decisions that shape capital cost, operational resilience, expandability, and the facility's ability to keep treating patients while the rest of the building rides through routine events.
This article covers the normal (utility) side of the electrical system: incoming service and metering, the medium-voltage and low-voltage distribution architecture, transformer and switchgear strategy, load classification and estimation, demand and diversity, voltage and power-quality considerations, and how normal power hands off to the emergency systems. The structure of the Essential Electrical System itself (life-safety, critical, and equipment branches), generators, automatic transfer switches, UPS, and grounding/bonding are covered by sibling Articles and are referenced here only at the boundary.
In a hospital, the electrical system is deliberately split into two worlds:
The separation matters because the two are designed, sized, tested, and maintained to different standards. Normal power is governed primarily by the National Electrical Code (NFPA 70) general articles plus NEC Article 517 (Health Care Facilities) where it interfaces with patient-care areas. The EES adds the layered requirements of NFPA 99 (Health Care Facilities Code) and NFPA 110 (Emergency and Standby Power Systems).
A practical consequence: every load in the building is classified early as normal, equipment branch (delayed-automatic standby), or critical/life-safety (fast automatic). That classification — not the wattage — determines which side of the system the load lives on and therefore how much generator, transfer, and battery capacity the project must buy. Getting normal-vs-essential allocation right is one of the highest-leverage cost-and-resilience decisions in healthcare electrical design.
Even though normal power has no code-mandated outage survival, owners increasingly choose to make large portions of it optional standby — connecting HVAC, elevators, imaging, kitchen, and other "comfort and operations" loads to generation beyond the code minimum so the facility can keep functioning, not merely keep patients alive, during an extended outage. Those choices are made on the normal-power side and are discussed under load strategy below.
The service entrance is where utility reliability and facility resilience first meet. Key decisions:
Number and routing of utility feeds. Larger acute-care facilities frequently negotiate two utility services, ideally from separate substations or separate sections of the grid, routed into the site along physically diverse paths. A genuinely redundant pair (separate substation, separate right-of-way) meaningfully reduces the probability that a single utility event de-energizes the whole campus. A "second feed" from the same substation bus offers far less protection and should not be marketed internally as redundancy. Note that a redundant utility service does not substitute for the code-required on-site EES generation — utility feeds are not a "source" recognized for the EES under NFPA 99/110; they reduce how often the generators must run, not whether they are required.
Service voltage and configuration. Small facilities and clinics may take low-voltage service (e.g., 480Y/277 V or 208Y/120 V) directly. Mid-size and large hospitals almost always take medium-voltage service (commonly in the 5–35 kV class, with 12.47 kV and 13.8 kV typical in the U.S.) and own a campus medium-voltage distribution system with multiple unit substations. Taking service at medium voltage reduces feeder copper, lowers line losses over campus distances, and gives the owner control of the step-down transformers and the resulting redundancy topology.
Metering and utility interface. The service includes the utility metering (often current-transformer metered at medium voltage for large customers), the point of common coupling, and any utility-required protective relaying. Where the facility intends on-site generation that may parallel with the grid, microgrid behavior, demand-response participation, or net export, those plans must be coordinated with the utility's interconnection requirements early — a chapter taken up by the electrical resilience/microgrid sibling Article.
A useful framing for owners is a reliability tier target for the service: single radial feed (lowest), single feed with sectionalizing, dual feed with automatic throwover, and dual-substation feeds with primary-selective or secondary-selective distribution (highest). The tier should match the facility's clinical acuity and the cost of downtime, and it should be stated as a basis-of-design assumption rather than discovered during construction.
On a campus or large single building taking medium-voltage service, the MV distribution topology is the backbone that determines how a single fault propagates. Common architectures, in increasing resilience and cost:
| Architecture | How it works | Resilience | Typical fit |
|---|---|---|---|
| Radial | One feeder per load; a fault upstream drops everything downstream | Lowest | Small sites, clinics, non-critical buildings |
| Loop / open-loop (sectionalized) | Feeders form a loop with normally-open tie; a faulted segment is isolated and the rest is re-fed | Moderate; brief outage during switching | Mid-size campuses |
| Primary-selective | Each unit substation has two primary feeders; on loss of one, the substation transfers to the other | High; transfer at the substation primary | Acute-care campuses |
| Secondary-selective | Paired transformers with a normally-open low-voltage tie breaker; loss of one transformer/feeder is covered by the partner through the tie | High; N+1 at the substation | Critical buildings, central plants |
Many hospital campuses combine schemes — for example, a primary-selective MV loop feeding secondary-selective unit substations — so that no single transformer, feeder, or switch outage de-energizes a critical department. The MV system also defines where the facility's owned distribution begins and the utility's ends, and it concentrates the highest available fault energy, which drives arc-flash and equipment-rating decisions throughout.