Ventilation Is an Envelope Problem Before It Is a Plant Problem
Mechanical Ventilation with Heat Recovery (MVHR) is among the most frequently specified components of a low-energy home, and among the most frequently misapplied. The technology is straightforward in concept: two fans move a balanced volume of air, one extracting warm, humid and stale air from the kitchen, bathrooms and utility spaces, the other supplying filtered outdoor air to the living spaces and bedrooms. Between them sits a heat exchanger core in which the outgoing air passes its sensible heat to the incoming air without the two streams mixing. A well-designed unit may recover up to 90% of the sensible heat in the extract stream, so that a home can be ventilated continuously at a rate sufficient for indoor air quality while the associated heat loss is reduced to a small fraction of the unmitigated value.
The claim embedded in that description is conditional. It presumes that the air being ventilated is the air that passes through the machine. In a building whose fabric is itself leaky, that presumption fails: a substantial share of the home's air exchange takes place through gaps in the structure, entirely outside the heat exchanger, and no amount of efficiency in the core can recover heat from an airflow that never reaches it. This audit characterises the physical relationship between air permeability and background infiltration, quantifies the thermodynamics of the heat exchanger and the effect of leakage on the real-world recovery rate, examines the aerodynamic penalties that govern specific fan power (SFP), and outlines the realities of the remedial air-sealing that must precede mechanical plant specification in a retrofit.
Infiltration Physics and the Tipping Point
Air permeability is the standard metric for the leakiness of a building envelope. The score, denoted q50, is the volume of air leaking through the envelope per hour, per square metre of envelope area, when the building is held at a pressure difference of 50 Pascals relative to the outdoors by a calibrated fan, in a blower door test. It is expressed in m3/(h·m2) at 50 Pa. A related metric, n50, expresses the same leakage as air changes per hour at 50 Pa, which is the leakage volume divided by the volume of the building. The two are connected by the geometry of the building: n50 = q50 × Aenv / V, where Aenv is the envelope area and V is the internal volume. For a typical UK house the ratio of envelope area to volume is of the order of 1.0 per metre, so that n50 is numerically close to q50.
A blower door test is conducted at 50 Pa because the pressure is large enough to overwhelm the variable wind and stack pressures that act on a building and so to give a repeatable measurement. Under normal conditions, however, the pressure differences driving air through the envelope are far smaller, typically of the order of 1 to 5 Pa, being produced by the combination of the stack effect (warm air rising and escaping at high level, drawing cold air in at low level) and wind. Because the flow through a crack follows a power law with an exponent of between 0.5 and 1.0, and commonly about 0.65, the leakage at 4 Pa is a small fraction of that at 50 Pa. The widely used rule of thumb relating the two is that the average background infiltration rate in air changes per hour is approximately n50 divided by 20, with a divisor of 15 to 25 depending on the building height, exposure, shielding and the distribution of the leaks. It is a screening approximation rather than a precise prediction, but it is sufficient to establish the order of magnitude of the problem.
A Worked Example
Consider a three-bedroom semi-detached house with an internal volume of 250 m3 and an envelope area of 250 m2, so that n50 equals q50. Applying the n50 / 20 rule, an unrefurbished home measuring q50 of 10 has a natural infiltration rate of 0.5 air changes per hour, which is 125 m3/h or approximately 35 L/s. At a q50 of 5 the figure is 0.25 air changes per hour, or 62.5 m3/h (17.4 L/s). At a q50 of 3 it is 37.5 m3/h (10.4 L/s), and at a q50 of 1 it is only 12.5 m3/h (3.5 L/s).
These values must be compared with the airflow that the MVHR unit is designed to deliver. Approved Document F of the Building Regulations (England) sets a whole-dwelling ventilation rate of 21 L/s for a three-bedroom home, with a minimum of 0.3 L/s per square metre of internal floor area, and requires extract rates in the kitchen of 13 L/s, in a bathroom of 8 L/s, in a utility room of 8 L/s and in a WC of 6 L/s. The design flow is the larger of the whole-dwelling rate and the sum of the extract rates, so 21 L/s, equal to 75.6 m3/h (the conversion being 3.6 cubic metres per hour per litre per second), is a reasonable design flow for this example house. This is the figure by which the mechanical system is sized, and it is the figure against which the infiltration must be measured.
At q50 of 3 the uncontrolled infiltration is approximately half of the mechanical design flow. At q50 of 5 it is more than four-fifths. At q50 of 10 it is well over one and a half times the mechanical flow. These ratios define the tipping point. Because a balanced mechanical system and infiltration add approximately arithmetically, the total air exchange rate in a leaky home is the sum of both, and the fraction of that total that passes through the heat exchanger falls rapidly as the leakage rises. The commonly cited guidance that MVHR should be considered only below q50 of around 3 to 5, and ideally below 1 to 3, follows from this arithmetic, as does the Passivhaus requirement of n50 not exceeding 0.6 air changes per hour. By contrast, unrefurbished UK dwellings of pre-1980s construction commonly test between 10 and 15 m3/(h·m2), and many are higher.
The Thermodynamic Heat Exchanger Balance
The heat exchanger in an MVHR unit is typically a counter-flow plate core in which the extract and supply streams pass in opposite directions on either side of thin plates of aluminium, polymer or paper, or a counter-flow arrangement combined with a cross-flow entry section. Its performance is described by the temperature effectiveness, ε, defined for a balanced system, in which the two streams have equal heat capacity rates, as ε = (Tsupply − Toutdoor) / (Textract − Toutdoor). The supply temperature delivered to the house is therefore Tsupply = Toutdoor + ε × (Textract − Toutdoor). For an ideal counter-flow exchanger with balanced flow, the effectiveness is related to the number of transfer units by ε = NTU / (1 + NTU), where NTU = UA / (m-dot × c), U being the overall heat transfer coefficient and A the exchange area, which shows why high effectiveness requires a large exchange area and low velocities, and why the effectiveness falls as the flow through a given core is increased.
The heat rate recovered is Qrec = m-dot × c × ε × (Textract − Toutdoor), which for air, with a volumetric heat capacity of approximately 0.33 Wh per cubic metre per kelvin, is 0.33 × q × ε × ΔT where q is in m3/h. For the 75.6 m3/h example, an effectiveness of 0.9 and a 15 K temperature difference, the heat recovered is approximately 337 W, and the unrecovered portion of the mechanical ventilation loss is 37 W.
How Leakage Degrades the Real-World Recovery
In a leaky envelope, two separate mechanisms act on the system. The first is a dilution of the benefit. The balanced fans fix the flow through the core, but the air that infiltrates is additional to it, and its heat loss is borne in full. The effective recovery relative to the total ventilation heat loss is therefore εsys = ε × qmech / (qmech + qinf). With ε = 0.9 and the example values, the effective recovery is approximately 77% at q50 of 1, 60% at 3, 49% at 5, 34% at 10, and 26% at 15, so that a unit advertised at 90% effectiveness is delivering half or less of that in the draughty home.
The second mechanism is a depression of the exchanger's driving temperature difference, which is what is meant by the exfiltration of warm air starving the exchanger. Mass conservation for the house requires that the supply flow plus the infiltration equals the extract flow plus the exfiltration. Warm air that is supplied by the unit is lost through leaks at high level, driven by the stack effect, before it can reach the extract points, and cold infiltration entering at low level mixes with the room air and lowers the temperature of the air drawn into the extract terminals. Because Tsupply is a function of Textract, a reduction in the extract temperature reduces the supply temperature one for one, scaled by ε. With an outdoor temperature of 0 degrees Celsius, an extract temperature of 20 degrees Celsius and ε of 0.9, the supply temperature is 18 degrees Celsius; if the draughty envelope reduces the extract temperature to 17 degrees Celsius, the supply temperature falls to approximately 15.3 degrees Celsius, and the occupants experience cooler supply air and compensate with higher heating output. Where wind or stack pressures also unbalance the effective supply and extract flows at the terminals, the departure from the commissioned balance reduces the effectiveness further, since the formula above holds only for equal heat capacity rates.
It follows that in a leaky building, MVHR can increase rather than decrease the total ventilation heat loss relative to doing nothing. In the example house at q50 of 5, with a degree-hour total for the UK heating season of the order of 52,000 kelvin-hours, the infiltration alone represents a loss of approximately 20.6 W/K (62.5 m3/h at 0.33 Wh per cubic metre per kelvin), or roughly 1,070 kWh per year. Adding a running MVHR unit adds the unrecovered portion of the mechanical flow, approximately 2.5 W/K or 130 kWh, giving a total of approximately 1,200 kWh, which is higher than the baseline without the machine, in addition to the electricity consumed by the fans. In a very airtight home at q50 of 1, by contrast, the infiltration is 4.1 W/K and the unit adds the same 2.5 W/K, giving a total of 6.6 W/K or approximately 350 kWh, compared with approximately 1,300 kWh if the full ventilation rate were supplied without heat recovery. The same unit therefore delivers a substantial saving in a tight home and a net loss in a leaky one. The benefits for indoor air quality, such as filtered supply air and controlled distribution, are real in both cases but are distinct from the energy benefit.
| Building Air Permeability Score | Infiltration Ingress Profile | MVHR Thermal Effectiveness (effective, whole-house) | Dominant Ventilation Energy Vector | Net Carbon/Financial Return Profile |
|---|---|---|---|---|
| Below 1 m3/(h·m2) at 50 Pa (Passivhaus class) | Negligible: approx. 3.5 L/s or less; air entry is almost wholly controlled by the unit | Approx. 77% or above | Mechanical supply and extract; fan electricity is the main residual cost | Strong: the largest heat saving per unit of fan energy; best case for payback |
| 1 to 3 m3/(h·m2) | Low: up to approx. 10 L/s, around half of the mechanical flow | Approx. 60% to 77% | Mechanical flow remains dominant; moderate infiltration loss | Good: positive return, sensitive to SFP and filter maintenance |
| 3 to 5 m3/(h·m2) (the tipping zone) | Moderate: approx. 10 to 17 L/s, comparable to the mechanical flow | Approx. 49% to 60% | Shared between mechanical flow and infiltration | Marginal: savings largely offset by fan energy, particularly when heat is supplied by a heat pump |
| 5 to 10 m3/(h·m2) | High: approx. 17 to 35 L/s, up to one and a half times the mechanical flow | Approx. 34% to 49% | Uncontrolled infiltration dominates | Weak to negative: total ventilation heat loss may exceed that of the unventilated baseline, plus fan electricity |
| Above 10 m3/(h·m2) (typical unrefurbished stock) | Severe: over 35 L/s; draughts, floor voids and chimney leakage dominate | Approx. 26% or below | Uncontrolled infiltration, almost entirely | Negative: unjustified until the fabric is air-sealed |
The figures are indicative and assume additive combination of balanced mechanical flow and infiltration, an envelope-to-volume ratio of 1.0 per metre and the n50 / 20 rule. The real infiltration depends upon the location of the leaks, the height of the building, its exposure and the wind. The structure of the result, however, is robust: the return on MVHR is controlled by the ratio of mechanical flow to infiltration, and that ratio is set by the fabric.
Specific Fan Power and System Penalties
Specific fan power is the electrical power consumed by the fans per unit of air moved, expressed in W/(L/s). It is the principal measure of the electrical overhead of the system. The fan power is the product of the volumetric flow and the total pressure rise, divided by the overall efficiency of the fan, motor and drive: P = q × Δp / η. Dividing by the flow gives the SFP directly as Δp / η, with the pressure in Pascals and the flow in cubic metres per second, which produces a value in watts per cubic metre per second that can be divided by 1,000 to give W/(L/s). A total system pressure drop of 300 Pa, summed across the supply and extract sides, and an overall efficiency of 45%, gives an SFP of approximately 0.67 W/(L/s). Well-designed domestic systems commonly achieve between 0.5 and 1.0 W/(L/s) in commissioned operation; compliance guidance and certification schemes set limits in this range, and these should be checked against the current Approved Documents and, where relevant, the Passivhaus Institute criteria.
Every pascal of system resistance is therefore directly converted into electricity. At 21 L/s and 0.67 W/(L/s) the units draw approximately 14 W, which over 8,760 hours amounts to approximately 123 kWh per year. If the SFP rises to 1.5 W/(L/s) because of an unfavourable duct network, the figure becomes approximately 31 W and 276 kWh per year, more than doubling the electrical penalty. Where the heat is supplied by a heat pump with a seasonal coefficient of performance of 3, the 1,170 kWh of thermal saving available in the airtight example house corresponds to only approximately 390 kWh of electricity, so that the fan penalty of 120 to 280 kWh can consume between a third and two thirds of the net benefit. This is the quantitative answer to the question of whether an air-source heat pump removes the need for MVHR: the two are independent requirements, since the heat pump supplies heat and the ventilation system supplies air quality, but the lower cost and carbon of heat from a heat pump reduces the value of heat recovery and magnifies the penalty of a poorly performing fan system.
Aerodynamic Resistance in the Duct Network
The pressure drop in a duct is governed by the Darcy-Weisbach relationship, in which the loss per unit length is proportional to the square of the air velocity and to a friction factor that depends on the surface roughness and the Reynolds number. Semi-rigid ducting, which has a smooth internal bore, typically presents a friction factor well below that of flexible, corrugated ducting, whose internal ridges add roughness and turbulence. Corrugated flexible duct can present a friction loss two to three times that of smooth semi-rigid ducting at the same velocity, and when it is installed sagging, compressed, or kinked, the resistance increases dramatically. Each bend adds a local loss defined by a loss coefficient K, such that Δp = K × ρ v2 / 2, with sharp 90 degree bends having K values several times those of long-radius bends or two 45 degree bends. Terminal valves, silencers, plenum boxes and external grilles add further losses, as do the intake and exhaust ducts when they are long or restricted by insect mesh.
Filters constitute a further resistance that varies over time. A clean filter may contribute a few tens of Pascals, but as particulate loading increases, the pressure drop across the filter can double or triple. A constant-volume fan will respond by increasing its speed to maintain the flow, which increases the electrical consumption, and, if the unit is constant-speed, the airflow will fall below the commissioned rate. Because for a system with a resistance that varies with the square of the flow the fan power varies with the cube of the flow, an increase in flow of 20% to compensate for a perceived shortfall raises the power by approximately 73%. It follows that a poorly designed network, an unserviced filter, an undersized duct or a speed boost that is applied to overcome the resistance can erase a large part of the thermal saving.
The design rules that follow are well established. Duct velocities should be limited to approximately 3 m/s in the main runs and 1.5 to 2 m/s in the branches serving terminals, which controls both the pressure loss and the aerodynamic noise. Runs should be as short and straight as possible, with the minimum number of bends and with long-radius fittings. Rigid or semi-rigid smooth-bore ducting should be used in preference to flexible duct, flexible sections should be limited to short, straight connections, and the units should be located so that insulated ducting between the unit and the external wall is short. Filters should be inspected at least annually, with replacement at the manufacturer's recommended interval, and commissioning should record the flow rates, the pressures and the measured power of each unit.
Retrofit Realities and Remedial Air-Sealing
The conclusion that follows from the preceding sections is that, in a retrofit, the order of works is decisive. Air-sealing must precede the specification of the ventilation plant, since the airtightness achieved determines whether MVHR is appropriate at all, and, if so, how the unit and ducts should be sized. A unit specified for a home on the assumption of a q50 of 3 and subsequently installed in a home that tests at 8 will not deliver its modelled performance.
The process begins with a diagnostic blower door test, conducted to the relevant technical standard and in both pressurisation and depressurisation modes, to establish the baseline. With the building under test, smoke pencils, thermal imaging and anemometers identify the leakage paths. In legacy dwellings these are consistently found in the same locations: the junctions between the floor and the walls, including the voids behind skirting boards on suspended timber floors; the joist ends and wall plates at the eaves and at intermediate floors; service penetrations for pipes and cables; recessed lighting and sockets in the hollow dry-lined walls; loft hatches; chimneys and flues; window and door reveals and their frames; and the cavities behind plasterboard fixed with dabs of adhesive, where air can circulate freely and exit at the upper edge.
Techniques and Their Practical Friction
Solid and cavity masonry walls require a continuous air barrier on the warm side. The traditional and effective measure is a parge coat, a thin layer of sand and cement, lime or gypsum-based render applied directly to the masonry and run into the junctions with the floors, windows and ceilings, which seals the porous brick and the mortar joints behind the finishes. On a retrofit this means stripping the finishes, applying the coat and then reinstating the plaster and decoration, with considerable disruption to occupants. Where the interior is to be retained, the alternative is to seal from the outside with a render system or to treat the wall in conjunction with an internal or external insulation programme, so that the air barrier is installed as part of a coordinated build-up.
At the joist ends where timber floors penetrate masonry, the problem is the discontinuity between the floor structure and the wall. Techniques include the application of flexible sealants and tapes at the junction between the floor deck and the wall, the installation of airtight membranes lapped across the joist ends, and the sealing of the gaps behind skirting boards, which can account for a significant proportion of the total leakage in a house with suspended timber floors. Where the suspended floor void is ventilated, as is required to protect the timbers, the sealing must be applied on the room side of the floor, never by blocking the sub-floor ventilation.
Airtightness membranes and tapes are used around windows and doors, where the junction between the frame and the masonry reveal is a persistent path. A pre-compressed sealing tape or an airtight membrane lapped from the frame to the reveal, and then plastered over, creates a durable seal, and a sealing grommet or collar should be used at each pipe and cable penetration. Loft hatches require a compression gasket and a positive latch, chimneys that are no longer in use should be sealed with a ventilated cap or a balloon, and the dry-lined walls require airtight electrical boxes and sealed junctions at the perimeter. Dry-lining adhesive dabs should be supplemented with a continuous ribbon of gypsum adhesive around the perimeter in order to prevent air circulation behind the board.
Each of these measures carries a cost in labour, disruption and risk. In occupied homes, access is limited by furniture and finishes. Success is dependent on the quality of workmanship, since a single unsealed junction can offset the gains of the others, and the test must be repeated after each stage so that the effect of each measure is verified. The return on the effort follows a law of diminishing returns: the first measures, aimed at the largest leaks, bring the greatest reduction, while the last few units of q50 require disproportionate effort. For this reason the target should be set at the design stage, by reference to the ventilation strategy and the economics, rather than pursued indefinitely.
Moisture, Combustion Safety and the Duty to Ventilate
Improving airtightness reduces the natural air change rate, and so increases the risk of condensation, mould and the accumulation of indoor pollutants if adequate purpose-provided ventilation is not installed. In dwellings with open flues or fuel-burning appliances, the reduction in air supply can create a hazard from spillage of combustion products, and the appliances should be assessed by a competent person before work begins. Solid-walled, uninsulated or damp buildings that rely on air movement to dry their fabric need a moisture risk assessment before the envelope is sealed. These requirements are embedded in current retrofit standards, such as PAS 2035, which requires a ventilation assessment for any retrofit that affects airtightness, and in Approved Document F. They are the reason that the choice of ventilation strategy, which can include MVHR, decentralised or demand-controlled mechanical extract ventilation, or passive stack ventilation, should be made as part of a whole-house assessment rather than by default.
Specification Guidance for Practitioners
The sequence that follows from this analysis is as follows. First, measure the baseline air permeability with a blower door test, and estimate the infiltration with the n50 / 20 rule and the actual geometry of the dwelling. Second, calculate the required ventilation flows in accordance with Approved Document F and compare them with the estimated infiltration. Third, carry out a targeted air-sealing programme to bring the dwelling into the range where balanced mechanical ventilation delivers a positive return, treating q50 of 3 to 5 as the upper limit for consideration and a value below 3 as the working target. Fourth, only then should the MVHR unit be specified and sized to the corrected flow, with the duct network designed for the lowest practicable pressure drop and a target SFP, verified at commissioning, of no more than 0.7 to 1.0 W/(L/s). Fifth, commissioning should include the measurement of flow balance, the setting of boost rates, and a maintenance plan that includes filter replacement.
Where the airtightness cannot be improved to the necessary level, because of the construction, the heritage status or the cost, the correct conclusion is not to install an MVHR unit regardless, but to specify a simpler ventilation strategy that is matched to the actual condition of the fabric. MVHR is a highly effective measure in a building that is tight enough to benefit from it, and an unhelpful addition in a building that is not. Treating airtightness as a precondition rather than as an afterthought is the single most reliable way to ensure that the modelled thermodynamic and indoor air quality benefits are realised in service.