Building Services & Facilities: A Practical Guide

A hermetic refrigerant motor-compressor — the kind found in most HVAC equipment — is sized under NEC Article 440, not the general motor rules of Article 430, because a sealed compressor motor behaves differently enough (particularly around starting current and thermal protection) that the Code gives it its own dedicated sizing method. Article 440's overcurrent protection sizing formula (percentages of rated-load current, rounded to standard breaker sizes with both a floor and a ceiling) looks superficially similar to Article 430's motor rules but isn't identical, and using the wrong article's formula for an HVAC compressor circuit can produce an incorrect result even though both articles are nominally about 'motor' overcurrent protection. When an HVAC unit combines a compressor with a condenser fan on one feeder, the combination-load sizing principle is the same one covered in the Motors & Drives guide — only the largest motor gets the full margin, everything else on the same feeder is added at its own rated current.

The lumen method's core relationship — required fixture count = (target illuminance × area) / (luminaire output × utilization factor × maintenance factor) — makes clear that floor area is only one of several inputs, and not even the dominant one for most design decisions. Target illuminance varies enormously by space type (a technical drawing office needs meaningfully more lux than a corridor, and an outdoor car park needs a small fraction of either), so two rooms of identical area can need very different fixture counts purely because of what the space is used for. Utilization factor (how much emitted light actually reaches the working plane, shaped by room proportions and surface reflectance) and maintenance factor (expected light loss from lamp depreciation and dirt accumulation before the next service) then further adjust that base relationship — which is why a lighting design pulled from 'square footage times a rule of thumb' rather than the actual target illuminance for that specific space type is a common and easy source of under- or over-lighting.

Sizing a shared feeder for multiple elevators raises the same underlying question covered elsewhere in this suite (Power Quality's maximum demand, Solar/EV's charge-point diversity): not every connected device draws its full rated current simultaneously, so is some reduction below the full connected load justified? For elevators specifically, NEC Table 620.14's demand factor is a straightforward table lookup based purely on the number of elevators sharing the feeder — no active load-management system or real-time control is required to apply it, unlike the EV charging diversity case covered in the Solar, EV & Renewables guide. That's because elevator dispatch behavior is inherently, structurally non-simultaneous by design (a dispatch algorithm rarely, if ever, runs every car in a bank at full duty at the exact same instant), so the code table reflects a statistical property of how elevator groups actually operate rather than requiring an active system to enforce it.

Power Usage Effectiveness (PUE) — total facility energy divided by IT equipment energy — is a simple ratio, but a bare PUE number without its measurement period and conditions stated is an incomplete figure: PUE typically varies with outdoor conditions (affecting cooling load), IT utilization level, and the specific period measured, so a single instantaneous reading and a well-documented annual average can tell meaningfully different stories about the same facility. It's also worth being precise about what's official and what isn't: ISO/IEC 30134-2 standardizes the PUE metric and its measurement methodology, but the commonly cited efficiency bands ('world-class,' 'efficient,' 'moderate,' and so on) come from The Green Grid's widely referenced but informal industry white paper, not a formally mandated ISO/IEC classification — useful context, but not an official grade.

Comparing a lower-upfront-cost option against a higher-upfront, lower-operating-cost option (standard vs. premium-efficiency equipment is the classic example) is a decision that shows up across several categories in this suite — VFD retrofits, power factor correction, generator fuel choices — and life-cycle cost (LCC) is the general tool for answering it rigorously. Simple payback period (how long until the extra upfront cost is recovered from operating savings) is intuitive but incomplete: it ignores the time value of money entirely, and it ignores every cash flow that happens after the payback point, including the fact that the cheaper-to-run option keeps saving money for the rest of the analysis period. A discounted-cash-flow LCC comparison — including present-worth of all future operating costs, any salvage value, and typically an escalation rate for energy costs — captures the whole picture, and can occasionally favor a different option than a naive payback comparison would, particularly when the analysis period extends well beyond the payback point.

Sizing a trace-heating circuit genuinely starts as a heat-transfer problem, not an electrical one: steady-state radial conduction through a pipe's insulation layer (a function of pipe diameter, insulation thickness and thermal conductivity, and the temperature difference being maintained against the coldest expected ambient) determines how much heating power per meter of pipe is actually required, with a design margin applied on top for a safety factor. Only once that required W/m figure is established does the calculation become electrical — sizing the circuit's power, current, breaker rating, and (importantly, and easy to overlook on long runs) voltage drop over the heating cable's own resistance. A trace-heating design that looks adequate on heater output and breaker sizing alone can still fail on voltage drop specifically because that check depends on a completely different input (the heating cable's own resistance per meter) that's easy to leave unset or unchecked.

An enclosure sized generously enough to physically contain all its installed equipment can still be thermally undersized, because natural-convection cooling capacity and equipment heat loss are governed by entirely different things — cooling capacity scales with the enclosure's external surface area and the allowable temperature rise above ambient, while heat loss scales with whatever components happen to be installed inside, regardless of how much spare physical room they leave. A busy, densely packed panel in a physically spacious enclosure can still exceed natural-convection capacity well before the enclosure looks 'full,' which is exactly why enclosure cooling is checked as an explicit heat-balance calculation rather than assumed to follow automatically from adequate physical fit.