Cooling Coil Selection: Rows, Circuiting, ADP and Carry-Over Limits

A heating coil sizing problem has one variable that matters — air-side dP versus capacity, as covered in our heating coil selection guide. A cooling coil adds two more axes: how dry the leaving air actually gets (apparatus dew point and bypass factor), and how fast air can move across a wet coil before it starts carrying condensate into the ductwork. Get either wrong and the datasheet duty is "met" on paper while the installed unit either misses its dehumidification target or wets the downstream ducting.

The four knobs

• Row count — same diminishing-returns pattern as heating coils: each added row buys progressively less capacity per unit of dP, but for cooling coils the row count also controls how close the leaving-air state gets to the coil's apparatus dew point.
• Fin pitch — tighter pitch adds surface area and lowers bypass factor, but it is also where condensate has to drain against airflow; too tight a pitch on a wet coil raises the carry-over risk at the same face velocity that would be safe on a dry coil.
• Circuiting — governs water-side velocity and pressure drop, not the psychrometric outcome, but it is still autosized off actual water flow per the coil duty — never a fixed number copied from a similar project.
• Face velocity — the one knob that trades directly against carry-over risk (see below), on top of its usual SFP consequences covered in our face velocity guide.

Apparatus dew point and bypass factor

The apparatus dew point (ADP) is the coil surface temperature the leaving air would reach if the coil had infinite rows and zero bypass — effectively the saturation point on the chilled-water side. No real coil gets there: a fraction of the air, the bypass factor (BF), passes through without full contact with the cooling surface, so the actual leaving-air state sits on the straight line between the entering-air state and the ADP, at a distance controlled by BF:

leaving state = ADP + BF × (entering state − ADP)

More rows and tighter fin pitch lower BF (more surface, more contact, closer to ADP); fewer rows or a coarser pitch raise it. This is the same physics the psychrometrics guide describes as the coil-leaving state point — a consultant checking a quote will ask whether the implied ADP is one the coil's actual row/fin-pitch combination can reach, not just whether the sensible + latent split totals correctly.

Carry-over: the limit that caps face velocity on wet coils

Every AHU casing sizing exercise picks a face velocity for SFP reasons. A cooling coil operating wet (condensing) adds a second, independent ceiling: past a certain velocity, air shear starts pulling condensate droplets off the fin surface and carrying them downstream into the duct or the next component — a real, verifiable operational failure, not a datasheet nuance. Manufacturers publish a maximum face velocity for wet operation per coil geometry (fin spacing, tube row layout); above it, an eliminator/droplet-separator section is required downstream, adding both casing length and air-side dP. Sizing a cooling coil purely for SFP and free-cooling face velocity, then discovering post-installation that half the year it runs wet above the carry-over limit, is a common source of coil-section water damage complaints.

Indicative row-count behaviour for a wet cooling coil

Standard fin-and-tube geometry, face velocity in the coil manufacturer's wet-operation range, sensible + latent duty:

Rows	Typical bypass factor	Relative latent capacity	Typical air-side dP
3	0.20-0.25	baseline	60-90 Pa
4	0.12-0.16	+25-35%	90-130 Pa
6	0.05-0.08	+15-20% over 4-row	140-190 Pa
8	0.02-0.04	+5-10% over 6-row	190-250 Pa

Past 6 rows, most of the added dP buys a lower bypass factor rather than meaningful extra capacity — useful when the spec genuinely needs a tight leaving humidity ratio (e.g. a hospital OT AHU), wasteful when it's a default carried over from a previous project.

Where the sizing loop breaks in practice

The failure mode is the same shape as on heating coils: row count, fin pitch, face velocity and circuiting get picked from four different reference points instead of solved together against the actual entering-air state and duty. On a cooling coil the consequence is worse than extra SFP — an ADP that was never achievable, or a face velocity above the wet-operation limit that nobody cross-checked against the manufacturer's carry-over data.

Where the tooling matters

AirSelect3D computes the actual bypass factor and ADP from real manufacturer coil geometry for the entering-air state and duty you specify, autosizes circuits per module the same way as on heating coils, and flags face velocity against the coil's published wet-operation limit — so the carry-over check happens during selection, not after a service call.

See ADP, bypass factor and carry-over checked live during coil selection →