Facilities rarely outgrow their electrical systems in one visible step. They grow by accretion — a compressor here, a packaging line there, an office buildout, a row of EV chargers — and each addition passes because the breaker had room. Nobody recalculates the feeder. The system's margin quietly goes from comfortable to thin, and the first formal notice is a transformer failure or a drive that trips every August.
A load flow study is how that trajectory gets caught on paper instead.
What does a load flow study calculate?
The study solves your distribution system for normal operating conditions — the deliberate opposite of a short circuit study, which solves it for the worst milliseconds of its life. Into the model go the sources, transformer impedances and tap positions, conductor sizes and lengths, and the loads as they actually operate. Out of it come three families of numbers:
- Loading on every transformer and feeder, as a percentage of its rating;
- Voltage at every bus, from the service down to the last panel — the system-level version of the conductor-by-conductor question covered in Voltage Drop Explained;
- Power flow and losses through each element, including the reactive component that loads equipment without doing work.
The output reads like a physical of your distribution system: most elements healthy, a few flagged for observation, and occasionally one that needs intervention now.
Why is connected load the wrong number?
Panel schedules total connected load, and NEC Article 220 load calculations apply standardized demand factors for sizing new installations — deliberately conservative, as sizing rules should be. Neither is a statement of what the system actually carries on a working afternoon.
For an existing facility, the honest inputs are measured: utility interval data at the service, and metering or monitoring on the major feeders. A model loaded with real demand routinely tells a different story than the paper — panels that look full on the schedule while carrying a third of their rating, and one transformer nobody worried about running at nameplate every shift. Both findings matter: the first is capacity you did not know you had; the second is the failure you get to schedule instead of suffer.
What does it find before it breaks?
Transformers running past their comfort zone. Overloading a transformer does not trip anything — it just runs hotter, and insulation life falls fast with temperature. A unit quietly carrying its full nameplate through every shift is a failure with a date on it that nobody can read. The model shows loading against rating for every unit, under present load and under the scenarios you are planning.
Feeders with no headroom. The feeder sized for the original building, now carrying the original building plus fifteen years of additions. It shows up as marginal voltage at the far end and no room for the next project.
Voltage problems with a cause attached. The study does not just report low voltage at a bus — it shows where the drop accumulates, which is the difference between "upsize this one feeder" and "re-tap the transformer" as the fix.
Power factor loading you are paying for twice. Reactive current occupies transformer and conductor capacity and, under many utility tariffs, shows up on the bill. The model quantifies what correction would recover.
When is the study worth running?
Three occasions, in descending order of value:
- Before adding significant load. A new line, a chiller plant, an EV charging installation. The study answers whether the existing service, transformers, and feeders can carry it — before equipment is bought and concrete is poured. Conductor sizing is cheap on paper and expensive in conduit.
- During design. Load flow run at design time is how the system starts life with deliberate margin instead of accidental margin.
- When nobody knows where the margin went. If the one-line is fifteen years old and the facility has grown continuously, the study re-establishes the baseline every future decision needs.
Why run it with the other studies?
Because it is the same model. The impedances, conductors, and transformers that produce fault current in a short circuit analysis produce loading and voltage in a load flow — and the arc flash and coordination studies sit on the same foundation. Building the model once and running every analysis against it is why bundled studies cost meaningfully less than the sum of separate ones, and why our load flow and voltage drop service is most often delivered alongside the safety studies rather than alone.
The facilities that get the most from this analysis are the ones planning something — an expansion, an electrification project, a capacity question from corporate. If that is you, run the numbers before the purchase orders go out. The study is the cheap part of the project.