Facilities & Plant Operations — Engineering Decision Guide

Size the compressor to the plant's real demand profile (CFM at the required pressure), not peak nameplate, and add storage (receiver tank) to absorb spikes. Use a rotary screw for continuous industrial duty, a reciprocating unit for intermittent/lower duty, and a variable-speed drive (VSD) compressor where demand swings widely. Set the pressure no higher than the highest tool actually needs — every extra 2 psi wastes ~1% energy.

Compressed air is often the most expensive utility in a plant per unit of useful work, and most of that cost is electricity, so sizing and pressure dominate the bill. Oversizing wastes energy at part load (a fixed-speed screw idling still draws power); undersizing starves tools. Storage lets a smaller compressor ride out short peaks instead of running big. Running system pressure higher than needed "to be safe" silently taxes every hour of operation and increases leak losses.

An oversized fixed-speed compressor burns energy idling. Too little storage makes pressure sag when a big tool kicks on. Cranking system pressure to mask a sizing or leak problem multiplies the energy cost. Wrong compressor type for the duty cycle wears out or runs inefficiently.

Use a chiller when you need coolant below ambient / tightly controlled temperature (lasers, spindles, molding, EDM). Use a cooling tower (evaporative) for large heat loads where the process tolerates near-wet-bulb temperatures and you can manage water treatment. Use a dry cooler / fluid cooler where water use, freezing, or contamination rule out a tower and above-ambient cooling is acceptable.

Cooling choice follows the temperature you must hit and the ambient you're rejecting heat into. A chiller uses a refrigeration cycle to push below ambient — necessary for precision and process-temp control, but energy-hungry. A cooling tower rejects heat by evaporation, reaching near the wet-bulb temperature very efficiently for big loads, at the cost of water, treatment, and Legionella management. A dry cooler just blows ambient air across a coil — simple and waterless but can only get near (above) dry-bulb ambient.

A dry cooler where sub-ambient temperature is required can't hit the setpoint. A cooling tower without water treatment scales, fouls, and risks Legionella. An oversized chiller where a tower or dry cooler would serve wastes energy. Ignoring freeze protection on outdoor loops in cold climates bursts them.

Capture contaminant at the source (hood/enclosure at the machine) rather than diluting the whole room, size the system for the required capture velocity and duct transport velocity, and pick the collector to the dust: cartridge/baghouse for fine dry dust, wet collector for combustible metal dust (aluminum, titanium, magnesium), and mist collector for coolant/oil mist. Follow combustible-dust (NFPA) requirements where they apply.

Source capture uses far less airflow (and energy) than room dilution and actually protects operators, because it grabs the contaminant before it disperses. Duct velocity has to stay high enough to keep particulate airborne or it settles and plugs the ducts (a fire/explosion risk with combustible dust). Collector type follows the hazard: combustible metal dusts can detonate in a dry collector, which is why they go to wet collectors and why NFPA combustible-dust rules govern the design.

Room dilution instead of source capture moves huge air volumes and still exposes operators. Too-low duct velocity settles dust and plugs (or, with combustible dust, creates an explosion hazard). A dry collector on aluminum/titanium dust is a deflagration risk. Ignoring NFPA on combustible dust is a life-safety and code failure.

Use LED high-bay/area fixtures as the default retrofit and new-build choice, target the illuminance (lux/foot-candles) the task needs — higher at inspection/assembly benches, lower in aisles/storage — and add controls (occupancy sensors, daylight dimming, zoning) to cut energy where light isn't needed. Mind color temperature and CRI where color judgment (inspection, paint) matters.

LED has displaced HID/fluorescent because it uses far less energy, lasts longer, switches instantly (enabling occupancy control that HID can't do), and holds color. Designing to task illuminance rather than a blanket level puts light where work happens and saves it elsewhere. CRI and color temperature matter specifically where operators judge color or fine detail — a low-CRI source hides defects a high-CRI one reveals.

Uniform over-lighting wastes energy; under-lighting task areas hurts quality and safety. HID where instant switching/occupancy control is needed can't deliver it. Low-CRI lighting at an inspection station lets defects slip through.

The compressor runs more than the work justifies, pressure sags at the far end of the plant, and tools underperform. Leaks are invisible and audible only up close — yet they often waste 20–30% of all compressed air generated.

Leaks at fittings, couplings, hoses, drains, and FRLs (they never self-heal and grow over time), undersized or long/restrictive piping causing pressure drop, and the common overreaction of raising system pressure to compensate — which increases both energy use and the leak rate.

1) Run a leak survey (ultrasonic detector) and fix leaks — the fastest energy payback in most plants. 2) Fix pressure drop at the source: adequately sized loop piping, clean filters, short drops. 3) Lower system pressure to the minimum the tools need once leaks and drop are fixed. 4) Add zone shutoff valves to isolate idle areas. 5) Make leak detection a scheduled program, not a one-time fix — leaks recur.

Nuisance breaker trips, overheating transformers and neutrals, VFD/drive faults, flickering lights, and electronics that misbehave or fail early — often traced to the plant's own electrical loads, not the utility.

Harmonics from nonlinear loads (VFDs, rectifiers, LED drivers) distorting the current and overheating neutrals/transformers, voltage sags when large motors start, low power factor drawing penalty charges, and undersized or unbalanced distribution.

1) Measure it — a power-quality survey identifies harmonics, sags, and power factor before you guess. 2) Mitigate harmonics at the source — line reactors/chokes on drives, harmonic filters, or multi-pulse/active-front-end drives. 3) Correct power factor with capacitor banks (with harmonic-safe designs). 4) Soft-start or VFD large motors to cut starting sag. 5) Balance and adequately size feeders and neutrals for the harmonic load.