Balanced Ventilation and Heat Recovery

Setting the Number Both Columns Must Hit

Continuous whole-building ventilation gets one design figure, and both supply and exhaust are then obliged to hit it. In residential work that figure usually comes from ASHRAE Standard 62.2, Ventilation and Acceptable Indoor Air Quality in Residential Buildings, which combines a floor-area term with an occupancy term derived from bedroom count. Editions differ in how the terms are weighted and in what infiltration credit, if any, is permitted, so the adopting jurisdiction and its amendments decide the answer — not the standard in isolation. Canadian work is commonly governed by CSA F326, Residential Mechanical Ventilation Systems; work in England by Approved Document F, Ventilation. Commercial and institutional zones follow ASHRAE Standard 62.1, Ventilation for Acceptable Indoor Air Quality, with per-person and per-area components summed by zone and then corrected at the air handler.

Local exhaust is a separate obligation from the whole-building rate. Kitchens and bathrooms carry their own intermittent or continuous minimums, and whether continuous local exhaust may be credited against the whole-building total varies by code edition and by jurisdiction. Getting that credit wrong in either direction is expensive: claim it where it is not allowed and the system fails inspection, ignore it where it is allowed and the unit is oversized, noisier and harder to balance at low speed.

Fix the design flow before anything is ordered. That single number sets the branch schedule on both sides, the terminal count, the core selection and the acceptable duct velocities. Changing it after the ducts are hung means re-balancing every register in the house.

Sizing the Core to Carry Both Streams

Selecting a heat or energy recovery unit means selecting two fan curves, not a cabinet. Catalogue flow at zero external static is a marketing figure; the number that matters is delivered flow at the external static the installed duct system will actually impose, on both the supply and the exhaust side, which are rarely equal. Certified performance data — HVI-listed ratings, or ratings developed under AHRI Standard 1060, Performance Rating of Air-to-Air Exchangers for Energy Recovery Ventilation Equipment, and the laboratory method in ASHRAE Standard 84, Method of Testing Air-to-Air Heat/Energy Exchangers — gives flow, power and recovery effectiveness at stated conditions. Recovery effectiveness quoted at a flow the installation will never see is worth nothing on site.

Choosing between sensible-only and enthalpy transfer is a climate and moisture-load decision, not a preference. An HRV moves heat and leaves the moisture where it found it, which suits a cold winter climate with an indoor humidity surplus. An ERV returns a fraction of the moisture with the heat, which reduces winter dryness in very cold dry regions and reduces the latent load imported in a humid summer. What an ERV does not do is dehumidify: it moderates what crosses the envelope through the unit, and a house with a genuine latent problem still needs dedicated dehumidification.

Leave headroom deliberately. A unit that only reaches design flow on its top speed tap has nowhere to go once the outdoor filter loads, once a defrost cycle bites, or once the client adds a HEPA bank. Sitting near the middle of the available range at commissioning keeps both fans trimmable in the same direction for the life of the installation.

Ductwork as the Silent Debit

Two duct systems leaving one appliance are never symmetrical, and every asymmetry is a debit against the balance. Long flex runs, crushed elbows within a diameter of the unit's collars, undersized branches feeding distant bedrooms — each imposes static on one side only. Sizing to ACCA Manual D principles and fabricating to the SMACNA HVAC Duct Construction Standards keeps the two systems close enough that the balancing dampers have authority rather than acting as makeshift throttles for a design error.

Duct leakage does not merely lose air; it relocates it. A leaking supply trunk in an unconditioned attic delivers less to the rooms than the fan reports, so a unit balanced at the cabinet is unbalanced at the envelope. Seal and pressure-test both ducted systems, and treat the outdoor-air and exhaust ducts running through conditioned space as condensation risks — they carry near-outdoor temperatures and need continuous vapour-tight insulation, not a taped blanket that opens at the joints.

Velocity is the constraint that decides whether the occupants will run the system at all. Undersized trunks produce audible rush and register noise, the client drops to low speed permanently, and the designed ventilation rate quietly stops existing. Check velocities at the design flow rather than assuming a duct size, and reserve the highest velocities for short, straight, well-supported runs away from bedrooms.

Hood placement is part of the same balance. Intake and exhaust terminations too close together short-circuit, and the unit begins recovering heat from its own discharge while the building receives progressively staler air. Separation, prevailing wind direction, and distance from combustion vents, plumbing stacks and dryer terminations all belong on the drawing, not on the roofer's judgement.

Filters, Frost and the Slow Drift

Filter loading is inherently one-sided. The outdoor-air filter sees pollen, road dust and combustion soot; the return filter sees indoor lint. They load at different rates, so a system balanced perfectly in September is supply-short by spring. Matching filter media between the two streams helps only slightly — the real defence is a commissioning point set with enough fan reserve to absorb the drift, plus a replacement interval the client will actually keep.

Higher-efficiency filtration multiplies the effect. Adding a deep-pleat or HEPA bank on the supply side imposes a pressure penalty that grows steeply as the media loads, and that penalty applies to one fan only. Size the bank for face velocity, not just for rated efficiency, and confirm the unit can still deliver design flow with the filter at its terminal resistance rather than clean.

Frost control is the largest scheduled imbalance in the system. Below a manufacturer-specific outdoor temperature the exhaust stream begins depositing ice in the core, and the unit responds in one of a few ways: it throttles or stops the supply fan, it recirculates indoor air across the core, or it energises a preheater on the incoming stream. The first two run the building exhaust-dominant for the duration of every defrost cycle; only preheating keeps the ledger square. In a tight, cold-climate house with atmospheric combustion appliances, that distinction is a safety decision, not an efficiency one.

Defrost thresholds, cycle length and recovery behaviour vary widely between products and are set by the manufacturer's controls. Read the installation manual for the specific model on the job and record the defrost strategy in the commissioning file, because the next technician diagnosing a winter pressure complaint will otherwise chase a phantom duct leak.

Proving the Balance with Instruments

Balance is a measurement, not an intention. Flow measuring stations built into the unit are a starting indication; independent verification at the terminations or by duct traverse is what stands up. ASHRAE Standard 111, Measurement, Testing, Adjusting, and Balancing of Building HVAC Systems, and the procedural standards published by NEBB and AABC describe accepted methods — pitot traverse in straight duct, powered flow hood at grilles, and calibrated fan flow measurement at the unit's collars.

Sequence matters more than instrument choice. Set the unit to its highest normal operating speed and balance supply against exhaust there first, because that is the condition with the most fan authority and the least measurement uncertainty. Then drop to the continuous speed and re-verify; motor and control behaviour is not linear, and a unit square at boost can be several percent out at trickle. Many programs and specifications require the two streams within roughly ten percent of each other at every operating speed, and certification schemes such as the Passive House criteria for comfort ventilation are tighter still — confirm the tolerance the job is actually being held to before setting dampers.

Balance at the room level as well as at the cabinet. Total supply may match total exhaust while one bedroom receives half its design flow and the return path is choked. Measure every terminal, adjust the branch dampers, and re-check the totals afterwards, because branch adjustments interact.

Record the result in a form the next person can use: measured flow per terminal, total supply, total exhaust, external static on both sides, fan settings, damper positions marked and locked, filter type fitted, and defrost strategy. A commissioning sheet left in the unit turns a future service call into a comparison rather than an investigation.

Return Paths Inside the Envelope

Air supplied into a room must have somewhere to go, and closed doors are the most common reason a properly balanced system misbehaves. A bedroom with supply and no return path pressurises as soon as the door shuts; the excess leaves through whatever leak exists — usually the exterior wall — while the rest of the house goes correspondingly negative. Whole-house totals stay perfect and the occupants still complain.

Provide transfer paths deliberately: generous door undercuts, transfer grilles, or jump ducts sized for the room's supply flow. Program specifications commonly hold pressure across a closed interior door to a low single-digit pascal figure, and a manometer with a short length of tubing under the door confirms it in seconds.

Extract-point placement follows the same logic. Pulling exhaust from bathrooms, kitchens and utility rooms while supplying bedrooms and living spaces sets up a deliberate cascade from cleanest to wettest. Reversing that pattern, or exhausting from a room the supply never reaches, moves humidity and odour the wrong way through the plan regardless of how well the unit is balanced.