THE ELECTRICITY COUNCIL RESEA THE ELECTRICITY COUNCIL RESEARCH CENTRE Capenhurst
Chester
METHODS OF MEASURING VENTILL No charge- Other reports to follow i n due course when they are published LEAKAGE O F HOUSES externally.
by
D, J. Dickson Ktx % a n z f i l i n S l z t d
J.B. Siviour
SUMMARY
Two methods have been used a t ECRl
houses. The decay method permits s very simply. The equilibrium concl
tubes for tracer gas distribution and measurement. Both nitrous oxide an, , ,,-- - - _
tracer gases. Calibration procedures for the gas analysers and flowmeters
a r e described.
A pressurisation test is used for measuring house leakage, with a fan installed
in an outside door. The calibration procedures a r e given for flow measurement
and assessment d effective leakage area. The use of a smoke puffer to locate
leaks is also discussed.
This Memorandum is published a s part of the Electricity Councills Research Programme and any technical query on the contents or requests for permission to reproduce any part d it should be addressed to the Author.
April 1981
THE ELECTRICITY COUNCIL RESE.A
METHODS OF MEASURING VENTIL; LEAKAGE OF HOUSES
by
D. J. Dickson
SUMMARY
Two methads have been used at ECRf
houses. The decay method permits s
very simply. The equilibrium conci
tubes for tracer gas distribution and
THE ELECTRICITY COUNCIL RESEARCH CENTRE Capen hurst
Chester
~o charge. Other reports to follow 1n due course when they are published externally.
J.B. Siviour
measurement. Bath nitrous <»tide an", ...... ______ _
tracer gases. Calibration procedures for the gas analysers and flowmeters
are described.
A pressurisatien test is used for measuring house leakage, with a fan installed
in an outside door. The calibration procedures are given for flow measurement
and assessment of effective leakage area. The use af a smoke puffer to locate
leaks is also discussed.
This Memorandum is published as part of the Electricity Council's Research Programme and any technical query on the contents or requests for permission to reproduce any part of it shauld be addressed to the Author.
April 1981
-1-
CONTENTS
1. Introduction
Measurement of ventilation rate using tracer gases
2.1 The decay method
2.1.1 Procedure
2.2 The equilibrium concentration method . 2.2.1 Procedure
2.3 Gas Analyser calibration
2.3.1 Carbon dioxide analyser
2.3.2 Infrat-red gas analysers for N20
2.4 Flowmeter calibration
3. Houseleakagemeasurement
3.1 Procedure
3.2 Interpretation
3.3 Leak location
4. References
Table 1
L i s t of F i g u r e s
PAGE NO.
ECRC/M1419
CONTENTS PAGE NO.
1. Introduction 3
2. Measurement of ventilation rate using tracer gases 3
2. 1 The decay method 4
2.1.1 Procedure 4
2. 2 The equilibrium concentration method ... , 6
2.2.1 Procedure 7
2.3 Gas Analyser calibration 7
2.3.1 Carbon dioxide analyser 7
2.3.2 ,Infra.-red gas analysers for N20 7
2.4 Flowmeter calibration 8
3. Hoose leakage measurement 9
3.1 Precedure 9
3. 2 Interpretatian 10
3. 3 Leak location 11
4. References 11
Table 1 12
List of Figures 13
·-2-
1. INTRODUCTION
Ventilation is the process of replacing a i r already ,in the house-hy fresh a i r
from outside. In practice, the process i s one of dilution since a t least partial
mixing of the incoming fresh a i r with the a i r already 'in the house takes place
before this a i r is removed.
Methods used a t ECRC for measuring the ventilation rate in houses a r e
described.
The driving forces of ventilation a r e not discussed, only the ways in which the
ra te can be measured.
Under any given weather conditions, the ventilation rate will depend on how
leaky the house i s and also on where the leaks a r e situated. A technique is
described for measuring and locating house leakage.
2. MEASUREMENT O F VENTILATION RATE USING TRACER GASES
Since ventilation is simply the process d contaminant removal by dilution, the
change in concentration with time of a contaminant o r t racer gas can be used to
measure the ventilation rate.
The tracer gas would ideally have the same density a s air , be readily
obtainable, harmless and easily analysed. The choice of tracer gas has been
discussed by Hitchin & Wilson (1). NO n nltraus oxid have been u used a t ECRC. Both a r e 50% more dense than a i r but, at the concentrations
appear to be a problem.
up to concentrations of a t least 0.5% and relatively
cheap equipment is available for measuring the concentrat ion. The main
disadvantage is that it is already present in the atmosphere a t a slightly
variable concentration around 0.03% and also the presence of people exhaling
C02 disturbs the measurement and precludes occupation of the house during
tests.
s more convenient because tihere i s no background concceart~atton
in the atmosphere. Therefore much lower ccaacentrations mw Are used and it
fa permissible to sntsr the hotas@ during tebsts. However, the infra-red amlyseraa
required to measuf e nitrow oxide gas ~ m c e n t s d i o are fairly aqe~;aneive,
ECR.C/M1419
1. INTRODUCTION
Ventilatien is the process of replacing air already ,in the house-by fresh air
from outside. In practice, the process is one of dilution since at least partial
mixing of the inceming fresh air With the air already 'in the hoose takes place
befere this air is removed.
Methods used at ECRC for measuring the ventilation rate in houses are
described.
The driving forces of ventilation are not discussed, en1y the ways in which the
rate can be measured.
Under any given weather oenditiens, the ventilatien rate will depend en how
leaky the heuse is and also on where the leaks are situated~ A technique is
desoribed for measuring and locating heuse leakage.
2. MEASUREMENT OF VENTILATION RATE USING TRACER. GASES
Since ventilation is simply the process ef contaminant removal by dilution, the
ohange in oonoentration with time of a centaminant er traoer gas can be used te
measure the ventilation rate.
The tracer gas would ideally have the same density as air, be readily
obtainable, harmless and easily analysed. The ohoioe of traoer gas has been
discussed by Hitohin & Wilson (1). (Carbon~;Xide }ind\ nitrooJ'So~~ have been
used at ECRC. Beth are 50% more dense than air but, at the ooncentrations
used, ~does not appear to be a problem.
\§i'rbon dimtide)is ~l-up to oonoentrations of at least O. 5% and relatively
oheap eqUipment is available for measuring the oenoentration. The main
disadvantage is that it is already present in the atmosphere at a slightly
variable oonoentration around 0.03% and also the presenoe of people exhaling
CO2 disturbs the measurement and preoludes oooupation of the house during
tests.
~itrou~id;\S more oonvenient beoause there is no baokground ooncent:ration
i.n the atmosphere. Therefore muoh lower concentrations can be used and it
Is permissi.ble to enter the house during tests. However. the infra-red analysers
requJ,ll."EHl to measure nitrous a,dde gas concentration are fairly expensive.
-3-
are the decay method and the equilibrium
The decay method
The tracer gas is first and distributed the a t a which the analyser readily detect. The ra te change of
is then measured.
ra te
of V then the in tracer concentration in time interval dt is
the a i r e rate, in house per unit time dc -
the a i r rate is equal to
The gas concentration must be uniform throughout the
are
the decay method and the equilibrium
The decay method
The tracer gas is first and distributed the
at a which the analyser readily detect. The rate change of
is then measured.
rate (
of V then the in tracer concentration in time interval dt is
the air e rate, in house per unit time
dc
the air rate is equal to The gas concentration must be uniform throughout the
4-
In the ECRC test houses it was convenient to have the trace as supply tubing
permanently installed. The gas supply cylinder i s in an instrument space adjacent
to the house and six 6mm bore PVC supply tubes pass from there into six main
rooms. The tubes connect through individual taps to a \'@%+?d"I( man o so that selective
introduction of t racer gas is possible. o that
the individual pressure drops and therefore flows a r e equal. In the house, the
f ree ends of the tubes terminate at the mixing fans. Gas flow i s measured by a
rotameter between the gas cylinder regulator and the manifold.
gas distribution is by a long flexible tube
attached through a needle valve and a flowmeter to a gas cylinder in the hall.
By systematically walking round the house with the end of the tube and tiniing
the gas input to each room proportional to its volume, a uniform gas concentration m
can be achieved throughout the house. An advantage d this system is the ease with
which the gas cylinder and supply tubing can be completely removed from the
house thus reducing the risk of accidental leakage d gas into the house during
measurements.
The initial gas concentration aimed for i s typically or between 100 ppm
and 200 ppm of N20. With internal doors open and mixing fans running, a uniform
concentration d tracer gas is attained throughout the h o u s e G 0 to 15 minutes3
This is checked by individual room sampling.
For sampling six PVC tubes a r e used with inlets 1 metre above the floor in the &
centre of the main rooms (i. e. lounge, kitchen/diner, mid-stairs/hall, large front bedreom, back bedroom, landing). A l l these sampling tubes a r e the same
length to ensure equal sampling rates a t all points using one pump The six
sampling tubes lead into a mixing bax, each through a tap to permit individual
room sampling. A single tube leads from the mixing box to the gas analyser.
The C02 analyser uses 6mm bore sampling tubes and a sampling rate of
100 cc/minute. The, MIRAN N 0 analyser requires 12mm bore sampling tubes to 2 enable the required sampling flow rate of 20 litres/minute to be achieved. A
particulate filter, required by the MIRAN ana ly~e r , is situated in the mixing box,
The actput from the gas analyser goes to an X-t recorder. The C02 analyser
has a e u r r e ~ t output which is taken through. a micrmmmeter which converts it
to a vdtage signal and also permits zero sugs~sression. The tracer gas
ECRC/M1419
In the ECR C test heuses it was cenvenient te ha ve the trace s supply tubing
permanenUy installed. The gas supply cylinder is in an instrument space adjacent
t@ the h@use and six 6mm b0re PVC supply tubes pass frem there inte six main
rooms. The tubes connect through individual taps to a 'inU-(1~1, that selective
introductien @f tracer gas is possible. All six tubes are the same leng!:h so that . ....,
the individual pressure dreps and therefore flows are equal. In the heuse, the
free ends 0f the tubes terminate at the mixing fans. Gas flew is measured by a
retameter between the gas cylinder regulater and the manifold.
A satisfactery alternative metped of gas distribution is by ~ lang flexible tube
attached threugh a needle valve and a flawmeter to a gas cylinder in the hall.
By systematically walking reund the house with the end of the tube and timing the gas input to each room proportienal te its volume, a uniform gas concentratien
can be achieved throughout the house. An advantage of this system is the ease with
which the gas cylinder and supply tubing can be cempletely rem0ved from the
heuse thus reducing the risk ef accidental leakage ef gas inta the house during
measurements •
.:Ihe initial gas C0Ilpentration aimed fer is tYPically~o/~or between 100 ppm
and 200 ppm ef N20. With internal doors open and mixing fans running, a unif0rm
concentrati0n of tracer gas is attained throughout the house~ to 15 minutes;)
This is cheeked by individual room sampling.
For sampling six PVC tubes are used with inlets 1 metre abeve the· floor in the
centre of the main rooms (i. e. lounge, kitohen/diner, mid-stairs/hall, large
front bedroom/back bedroom, landing). All these sampling tubes are the same
length to ensure equal sampling rates at all peints using ene pump. The six
sampling tubes lead into a mixing bax, each threugh a tap te permit individual
room sampling. A single tube leads from the mixing bex te the gas analyser.
The CO2 analyser uses 6mm bere sampling tubes and a sampling rate of
100 cc/minute. The MIHAN N20 analyser requires 12mm bore sampUngtubes to
enable the required sampling flew rate ef 26 litres/minute to be achieved. A
particulate filter, required by the MIRAN analyser, is situated in the mixing boxo
The output from the gas analyser goes to an X-t recarder. The CO2 analyser
has a current output which is taken through a microammeter which converts it
to a voltage signal and also permits zero suppressien. The tracer gas
-5-
•
ECRC/M1419
ration is recorded for 0.5 to 3 hours, depending on the. ventilation rate.
,L inside and outside temperatures/ a r e also recorded.
ical C02 and N20 decay curves a r e shown in Figures 2 and 3. The a i r
change rate is obtained by replotting the decay curve on log-lin paper and , .. .
measuring the slope or by direct calculation of the slope of log (concentration)
versus time.
When using GO2, correction for the ambient concentration must be done before
plotting the results. w
2.2 The equilibrium concentration method
If t racer gas continuously flows into the house a t a constant known rate, and -.----- - - - mixes uniformly with the a i r in the house, the rate of loss d t racer gas in the -----I
ventilating a i r will eventually equal the rate of supply of tracer gas. A steady
equilibrium tracer gas concentration will then be obtained.
Let q = rate of supply of tracer gas
v = ventilating a i r flow
c = equilibrium concentration of tracer gas ,When the rates of supply and removal of t racer gas a r e equal:
If CO is used a s the tracer gas and is also present in ambient a i r at2 concentration c a '
q + v c = ( v + q ) c a
. . v = C - ca
T ~ h e advantagJof this method is that the ventilation rate can be continuouslv "
that the accuracy of the measurement depends on the absslute accuracy d the gas aaalyser and p a f law meter, w h e r s a ~
a constant peraenhge error in the @E: &alyeer calibgdltton dm$ B P ~ matter when
ueing the decay method,
ECRC/M1419
is recorded for G. 5 to 3 hours, depending Gn the ventilation rate.
and outslde are also recorded. ~~~~==~~~~~~
Typical CO2 and N20 decay curves are shown in Flgures 2 and 3. The air
change rate is obtained by replotting the decay curve on log-lin paper and
measuring the slope or by direct calculation. Gf the slope of log (concentration)
versus time.
When using CO2, correcth;m fer the ambient concentration must be done befere
plotting the results.
The equilibrium concentration method
If tracer gas continueusly flews into the house at a constant knOwn rate, and _ .. _---------mixes uniformly with the air in the house, the rate ef less 0f tracer gas in the -.------ventilating air will eventually equal the rate of supply of tracer gas. A steady
equilibrium tracer gas concentration will then be ebtained.
Let q = rate at supply of tracer gas v = ventllating air flew
c = equilibrium cencentratlon of tracer gas
When the rates of supply and removal of tracer gas are equal:
q = (v + q) c
. . V = q - go 1 n = - - = q (- - 1) ~ ..::J. o 0 C
If CO2 is used as the tracer ga~ and is also present in ambient air at
ooncentration 0a:
. . q + vc = (v + q) c a
_ 9 - gc v- , c-o a = q'- ~ (
1 - 0 ) C - ca
g c - Os
this method is that the ventilation rate can be continuously ~--~-,-=----------~ . menitored. The main disadvantage s that the accuracy rat the measurement
depends on the absolute accuracy of gas analyser gas flow meter, whereas
a constant percentage error in the ga.e analyser calibration dosl$ nat matter when UB decay method.
-6-
2.2.1 Procedure
The gas supply and sampling arrangement is similar to that described for the deay method except that provision is made for maintaining and measuring a
constant low flow rate af tracer gas into the house. The flow rate used in the
E CR C test houses i N20 measured on a rotameter with a
l Flostat1 to oontrol the flow rate. The flow is monitored by the pressure drop across a constriction in the supply line measured by an electrcmic micro-
manometer, the output being recorded on one channel of the X-t recorder used to record the gas concentration.
(since the time to reach equilibrium a t the operating flow rate will be several ours, it is preferable to begin by charging the house with t racer gas a s for
decay method to approximately the expected equilibrium concentration.
\ 2.3 Gas analyser calibration
Most gas analysers require to be checked a t lzero' and at least one other known gas concentration. Some analysers claim to require only a zero check.
The Hampden Gas-0-Mat C02 analyser is omparatively arid has been used successfully a t E CR C for decay and continuous measurements of ventilation
rate. It has been found that , more
frequently if the ambient temperature is varying.
l Carbosorbl in a glass U-tube i s used to obtain a i r f ree of C02 for setting the
zero. Commercially obtained cylinders of standard C02 in a i r mixtures a r e used to set the calibration span.
Measurement of the CO concentration in ambient a i r is necessary a t the beginning 2 and end of each run, and gives an additional check that the Glibration is approximately correct. Ambient a i r con ins about 0.03% CO 2' p. 3.2 ~ r a - r e a gas analysers for N
Two models have been used. The MIRAN 1 is an all-purpose instrument which requires selection of wavelength and slit width, and then must be calibrated a t those settings. The MIRAN 103 is much simpler to use since the 'factory
ECRC/M1419
2.2.1 Procedure
The gas supply and sampling arrangement is similar to that described for the
decay method except that provision is made for maintaining and measuring a
constant low flow rate en tracer gas into the house. The flow rate used in the
ECRC test houses i~ 50 cc/minute }f N20 measured on arotameter with a
'Flostat' to contrel the flew rate. The flew is menitored by the pressure drep .
across acenstriction in the supply line measured by an electranic micre-
manometer, the eutput being recerded on one channel of the X-t recerder used
te record the gas concentration.
Since the time to reach equilibrium at the operating flow rate will be several
hGUrs, it is preferable to begin by charging the house with tracer gas as fer the decay method to appreximately the expected equilibrium coocentratien.
Most gas analysers require to be checked at 'zere' and at least ene other known
gas cencentratien. Seme analysers claim to require ooly a zero check.
(2. 3. 1 carbon dioxide ;~alyserl
The Hampden Gas-O-Mat CO2 analyser is comparatively in5pep@iy~ arid has been
used successfully at E CR C for decay and continueus measurements of ventilation
rate. It has been feu:nd that ~ dailY.#.ere check and cp.libration is neces~ary, more
frequently if the ambient temperature is varying.
'Carbesorb' in a glass U-tube is used te ebtain air free of CO2 for setting the
zero. Commerciallyebtained cylinders of standard CO2 in air mixtures are used
to set the calibration span.
Measurement of the CO2 concentratien in ambient air is necessary at the beginning
and end of each run, and gives an additional check that the calibratien is
apprOXimately correct. Ambient air cen ins about 0.03% CO2•
~ 2. 3. 2 Infra-rea gas analysers for N2 NTwo models have been used. The MffiAN 1 is an all-purpose instrument which
requires selectien of wavelength and slit Width, and then must be calibrated at
these settings. The MIRAN 103 is much simpler to use since the Ifactory
-7-
ECR C/M1419
ated1 slit and scale appropriate to the gas in use a re simply inserted in
instrument. It is claimed that standard gas mixtures a r e not required for
the MIRAN 103. However, our tests showed inconsistencies in calibration and
subsequent investigation showed that the meter scale on our instrument was
wrong.
Figure 4 shows the method used to calibrate the MIRAN analysers. The inlet
and outlet of the analyser a r e short-circuited giving a closed loop into which
known volumes of gas a r e injected. The MIRAN 103 has an internal pump
which can be used to circulate the gas in the loop; an external pump is
necessary with the MIRAN 1.
Using a microlitre gas syringe, N20 is injected 50 IJ 1 at a time into the gas
loop through a rubber septum. A calibration trace a s in Figure 5 is obtained.
Blockage of the syringe needle is indicated by discontinuities in the step size
on the recorder trace. When this happens the loop must be flushed with N 0 2 free a i r and calibration repeated.
Calibration setting of these analysers has proved to be very stable and so a
monthly calibration check is sufficient. The zero setting i s also very stable but
since it can be checked very quickly it is checked daily.
v. 4 Flowmeter calibrqtio4 The equilibrium concentration method requires an absolute measurement of the
tracer gas flow into the house. The flow is measured by a rotameter with
10-150 cc/minute a i r scale. For a given rotameter it i s expected that Q 6- i s constant where Q is the flow rate and P the gas density (2). Thus when
using nitrous oxide the true flow should be 0.8 times the reading on the a i r scale.
This was not confirmed by calibration.
calibration of the rotameter was checked using the travelling soap film device
shown in Figure 6. A vertical graduated tube of volume 500 ml i s connected to
the outlet of the rotameter. By squeezing a small rubber bulb containing soap
solution a t the base of the tube, a soap film is formed which is driven up the tube
by the gas flow. The speed a t which the soap film travels is measured by stop-
watch.
ECRC/M1419
!lalibrated' slit and scale appropriate to the gas in use are simply inserted in
the instrument. It is claimed that standard gas mixtures are not required -for
the MmAN 103. However, our tests showed inconsistencies in calibration and
subsequent investigation showed .that the meter scale on our inst:t~ument was
wrong.
Figure 4 shows the method used to calibrate the MmAN analysers. The inlet
and outlet of the analyser are short-circuited giving a closed loop into which
known volumes of gas are injected. The MmAN 103 has an internal pump
which can be used to circulate the gas in the loop; an external pump is
necessary with the MIRAN 1.
Using a microlitre gas syringe, N20 is injected 50 J.l 1 at a time into the gas
loop through a rubber septum. A calibration trace as in Figure 5 is obtained.
Blockage of the syringe needle is indicated by discontinuities in the step size
on the recorder trace. When this happens the loop must be flushed with N20
free air and calibration repeated.
Calibration setting of these analysers has proved to be very stable and so a
monthly calibration check is sufficient. The zero setting is also very stable but
since it can be checked very quickly it is checked daily.
t 2.4 Flowmeter calibr~tioi\ The equilibrium concentration method requires an absolute measurement of the
tracer gas flow into the house. The flow is measured by a rotameter with
10-150 cc/minute air scale. For a given rotameter it is expected that Q rp is constant where Q is the flow rate and p the gas density (2). Thus when
using nitrous oxide the true flow should be 0.8 times the reading on the air scale.
This was not confirmed by calibration.
Catibration of the rotameter was checked using the travelling soap film device
shown in Figure 6. A vertical graduated tube of volume 500 ml is connected to
the outlet of the rotameter. By squeezing a small rubber bulb containing soap
solution at the base of the tube, a soap film is formed which is driven up the tube
by the gas flow. The speed at which the soap film travels is measured by stop-
watch.
-8-
The flow rate for N20 was found to be 1.12 times the reading on the a i r scale. A calibration check with a i r gave the true flow a s 1.04 times the indicated
value. The anomalous behaviour with N20 is not explained.
Ventilation takes place through intentional and fortuitous gaps in the house
envelope. The actual ventilation rate a t any time depends both on the leakheslfn
of the house and on the weather.
The house leakage can be evaluated by finding the pressure required to cauee
a i r t o flow out through all the leaks a t a measured rate.
The house is pressurised by a fan fitted to a substitute external door. The a i r
flow Q through the fan is measured. At equilibrium the a i r flow into the house
through the fan equals the total a i r leakage out of the house. If the pressure
difference between inside and outside the house is Ap then (3f V
where A is the effective leakage area
K and m a r e constants.
The value d m is given in the literature a s 0.5 for a
0.63 for window cracks.
3.1 Procedure
The apparatus is shown i A centrifugal fan set into an adjustable
plastic panel is fitted into the door recess. The remainder d the opening is
blocked off by a sheet of 25 mm thick ICI tPurlboardt or almilar material.
Obvious gaps a r e closed with adhesive tape. h 4 ~ts, '.
The fan has a gauze screen across the outlet to achieve a sufficiently uniform
a i r distribution for the flow to be reliably obtained from a single measurement
of a i r speed using a vane anemometer. Calibration of the flow was carried out
a t indicated a i r speeds of 3 m/s and 6 m/s by traversing a pitot-static tube
across the end of the duct at the intervals shown in Figure 8. The a i r speed was
recorded against distance on an X-Y recorder and thence the average speed in
each of the 24 equal rectangles could be found. When the vane anemometer
ECRC/M1419
The flow rate for N20 was found to be 1.12 times the reading on the air scale.
A calibration check with air gave the true flow as 1.04 times the indicated
value. The anomalous behaviour with N20 is not explained.
Is. HOUSE LEAKAGE MEASUREMENT]
Ventilation takes place through intentional and fortuitous gaps in the house
envelope. The actual ventilation rate at any time depends both on the leakiness
of the house and on the weather.
The house leakage can be evaluated by finding the pressure required to cause
air to flow out through all the leaks at a measured rate.
The house is pressurised by a fan fitted to a substitute external door. The air
flow Q through the fan is measured. At equilibrium the air flow into the house
through the fan equals the total air leakage out of the house. If the pressure
difference between inside and outside the house is LlP then (3)
where A is the effective leakage area
K and m are constants. I\l\i.". ~N.)\.!.'f- O~ The value of m is given in the literature as O. 5 for a' t.1:~~;'d; orifice and
O. 63 for window cracks.
3.1 Procedure
The apparatus is shown inG;-gur0A centrifugal fan set into an adjustable
plastic panel is fitted into the door recess. The remainder af the opening is
blocked off by a sheet of 25 mm thick ICI 'Purlboard' or similar material.
Obvious gaps are closed with adhesive tape. ~\ ~~ ~ The fan has a gauze screen across the outlet to achieve a sufficiently uniform
air distribution for the flow to be reliably obtained from a single measurement
of air speed using a vane anemometer. Calibration of the flow was carried out
at indicated air speeds of 3 mls and 6 mls by traversing a pitot-static tube
across the end of the duct at the intervals shown in Figure 8. The air speed was
recorded against distance on an X-Y recorder and thence the average speed in
each of the 24 equal rectangles could be found. When the vane anemometer
-9-
E CR C/M1419
3 an speed of 3 m/s the flow rate was 765 m /h i, e. 255 x a i r speed.
recorder.
between inside and outside is measured by an electronic
micromanometer, the output of which is connected to the other channel of the
X-Y recorder,
A motorised thyristor controller i s used to slowly increase and decrease the
fan speed s o that flow versus pressure difference is plotted directly on the
X-Y recorder. The fan speed is increased from minimum to maximum in 5
minutes; coincidence of the traces for increasing and decreasing fan speed indicates
that pressure equilibrium is achieved a t each flow rate. This can also be checked
by holding the fan speed steady a t any intermediate value.
In low wind conditions, very reproducible smooth traces a r e obtained a s in
3.2 Interpretation
The results of 40 measurements in 26 houses gave values ofw between 0.61
and 0.75 a s shown in Table 1. The mean value is m = 0.66.
The proportionality constant K was found by using calibrated 1
house leakage was reduced a s much a s possible by dosing an
internal doors except that of the W. C. The fan was installed in the f r m t door-
way. So that the leakage could be increased by known amounts, the W. C. window
was replaced by a panel with an 8 x 50 cm hole in it. Figure 10 shows the flow
versus pressure curves for the three conditions:
2 (1) Background leakage only ( = A m ) 2 (2) Background + 8 x 25 cm hole (= A + 0.02 m ) 2 (3) Background + 8 x 50 cm hole (= A + 0.04 m )
ECRC/M1419
eed of 3 mls the flow rate was 765 m3/h i. e. 255 x air speed.
The carres,anding figures at 6 mls were-1575 m3/h and 263 x air speed. The
mean value gives
Flaw (m 3 /h) ::: 259 x Vane anemameter reading (m/s)
The output fram the vane anemometer is cannected to ane cannel af an X-Y
recarder. ~\~ '{\\~~. ~ The pressure ~difl'erence between inside and outside is measured by anelectrenic
micromanometer, the output af which is cannected ta the ather channel af the
X-Y recarder.
A motorised thyristar cantroller is used to slc:)wly increase and decrease the
fan s,eed so that flow versus pressure difference is plotted directly an the
X-Y recorder. The fan speed is increased from minimumta maximum in 5
minutes; coincidence of the traces far increasing and decreasing fan speed indicates
that pressure equilibrium is achieved at each flow rate. This can also be checked
by holding the fan s,eed steady at any intermediate value.
In lew whid conditions, very re,roducible smoath traces are abtained as i:n
~gur~ 3. 2 Interpretation
The results af 40 measurements in 26 houses gave values ef-m between O. 61
and o. 75 as shawn in Table 1. The mean value is m = O. 66 •
The ,roportionality constant K was found by using calibrated lea.
house leakage was reduced as much as possible by closing and
internal doors except that of the· W. C. The fan was installed in the front daor-
way. Sa that the leakage could be increased by knewn amounts, the W. C. window
was replaced by a panel with an 8 x 50 cm hale in it. Figure-10 sh0ws the flow
versus ,res sure curves for the three conditions:
2 (1) Backgraund leakage only ( = Am) 2 (2) Background + 8 x 25 cm hale (= A + O. 02 m -) 2 (3) Background + 8 x 50 cm hole (= A + O. 04 m )
-10-
ECR C/M1419
These curves satisfy respectively:
Hence (la) 0.278 = K A
(3a) 0.596 = K (A + 0.04)
From (la) and (2a) K = 7.8
(la) and (3a) K = 7.95
(2a) and (3a) K = 8.1
giving a mean value of K = 8
We now have V = 8 A A go* 66 where V i s the vane anerndrmeter reading in metre/second. . From this family of curves, Figure 11, the effective leakage
area can be immediately obtained.
3.3 Leak location - When the house i s pressurised the places where the a i r is escaping can be
located by using the smoke puffer, Figure 12. Occasionally the leaks can be
seen more easily when the pressure inside I s reduced, by *versing the fan.
Alternatively, selected leakage paths &n be sealed with adhesive tape and their
contribution to the house leakage found by difference,
4. REFERENCES
(1) Hitchin, E. R. & A Review of Experimental Techniques for the Wilsc~n, C. B. Investigation $I Natural Ventilation in Buildings.
Building Seience, 2, 59-82, Pergamon (1967).
(2) Ower, E. & The Measurement of Air Flow. Pankhurst, R. C. Pergamon (1977), p. 296.
(3) IHVEGuide, SectionA4: AirInfiltration(1976)
•
These curves satisfy respectively:
(l)V = 0.245 8p 6.697 ~
(2) V= 0.490 8p 0.624 ~
(3) V = o. 740 8P 0.596 ~
Hence (1a) 0.278 = K A
(2a) 0.434 = K (A + 0.02)
(3a) O. 596 = K (A + 0. 04)
Frem (la) and (2a)
(1a) and (3a)
,(2a) and (3a)
K = 7.8
K = 7.95
K = 8.1
gi ving a mean value m K = 8
. ECRC/M1419
0.278 8p 0.66
0.434 b. 0.66 . . P .
0.596 8p 0.66
We new have V = 8 A 8·pO. 66 where V is the vane anemometer reading in
metre/secGnd;· FrGm this family Gf curves, Figure· ll,the effective leakage
area can be immediately ebtalned.
3. 3 Leak location I\JVvVv...IVVVV When the house is pressurised the places where the air i~ escaping can be
lecated by using the smoke puffer, Figure 12. Occasionally the· leaks Can be
seen more easily when the pressure inside is reduced, by reversing the fan. ..
Alternatively, selected leakage paths can be sealed with adhesive tape and their
contribution to the house leakage f(!)UIld by difference.
4.
(1)
(2)
(3)
REFERENCES
Hitchin, .E. R. &; Wilson, C. B.
OWer, E. &; Pankhurst; R. C.
AReview ~ E~xperimental TechniquesfGr the Investigation m Natural Ventuatlon in Buildings. Bu.ilding Science, !, 59-82, Pergamon (1967).
The Measurement of Air Flow. Pergamon.(1977), p.296.
IHVE Guide, Section A4: A ir Infiltration (1976) .
-11-
0) s 00 0 s 3 2 a3 a o T i 3 d b c.3 d * li2 (O
$ a ". (O w w ( O W W co o d d d 6 d o' d d 8 %
z h h h h h h
z 3 d 03 Q,
b (0 I
c s 8 60' o' e; d a
-4 W
c.3 m m (O d b b W b 2 z!
S3 Ek? (O k? (O cD 4 d c ; c ; Irj 0' Ci;, d d I4
-4
8 a 9 % a d M c l l M u ) 0 3 4 0 0 2 9 b m w b C c 8 $ g g z (0 C W b % 3 t o w 29 p d d d d d d d d o' d d do' 0) 0 1 4 . -'tk ;=! cn M b b b 0 c n * M d a(0 * d c
Cc S3 i ? w CD ( O w d m (O Zk?$ me0 w m
cj o'de; e; e; d & c ; ej do; do'
k cd d 0 m
0) a a 9 9 ; 7 * s g 0) s n 4 a B o
9) 0 Qak 3 g 8 Olk Vl. ur I3 4 3 : 4
U U U g 1 d 8 m W W cd h $ cd
4
& 0- (O- & 4 E E ! 8 d d d 4 a x .;: cf 6 s l a , * e X 2 $ 4 cd
9) Qa X a 0) a i Q & a S Q * a 3 2 $j 8 % E E I 8
x a x x x - 8 J-, 8 5 %J U 0 -
E
jj 8 m % B % g 8 3 F a" W F Q a m
% S d g 4 4 n( d B) * (O m X X
t-'
N
No.
of
Hou
ses
1 1 1 1 9 2 6 5
Hou
se
Tes
t H
ouse
No.
10,
EC
RC
Tes
t H
ouse
No.
10,p
art
Tes
t H
ouse
Ne.
16,
ECR
C
Tes
t H
ouse
No.
18,
EC
RC
Det
ache
d, K
emna
y
End
terr
ace,
Kem
nay
Beb
ingt
on S
olar
Hou
ses
Beb
ingt
on T
radi
tion
al
Tab
le 1
: V
alue
s Gi
exp
onen
t in
pres
sure
dep
ende
nce
fer
diff
eren
t hou
ses
Type
V
alue
of m
.for
M
ean
valu
e pm
G
im
Flow
Sem
i-de
tach
ed,
mod
ern
0.62
9 0.
631
0.62
3 0.
628
. Semi-deta~hed,
mod
ern
0.70
3 0.
712
D.7
05
0.72
3 )
0.70
0 re
oms
seal
ed o
ff
0.67
7 0.
633
0.71
5 0.
642
) 0.
747
0.75
0
~emi-detached,
mod
ern
. 0.6
77
0.68
0 'cr.
666
-
0.67
4 fr
ame
wal
l
. Sol
id w
all,
sem
i-de
t.
. 0.
660
0.64
4 0.
652
. Wel
l in
sula
ted,
woo
den
o. flB
9 0.
695
0.71
1 0.
721
) 0.
692
0.71
2 0.
643
J1.6
67
0.74
3 )
0.64
3
Wel
l in
sula
ted,
wee
den
0.67
1 0.
707
0.68
9
With
sol
ar w
all
0.66
2 0.
645
0.67
7 0.
629
) 0.
651
0.66
6 0.
626
)
As
abov
e w
ithou
t sol
ar
0.61
2 0.
617
0.67
4 0.
617
wal
l 0.
584
0.59
7 OV
ERA
LL M
EAN
0.
663
t:rJ ~ -3: t-'
oj::- t-'
\0
.w
FIGURES
1. Apparatus for ventilation rate measurement
2. CB2 decay trace
3. N20 decay trace
4. ~sfibratioa d gas analgsers
5. Calibration trace for gas analysers
6. Rotameter calibration
7. House leakage apparatus
8. Flew calibration of house leakage apparatus
9. House leakage curve
10. Flow versus pressure f 9r calibrated leaks
11. Family of leakage curves '
12. Smoke puffer
FIGURES
1. Apparatus for ventilation rate measurement
2. CO2 decay trace
3. N20 decay trace
4. Calibration af gas 'analysers
5.
6.
7.
8.
9.
10.
11.
12.
Calibration trace, for gas analysers
. Rotameter calibration
House leakage apparatus
, Flow caUbration of hoose leakage apparatus
Hoose leakage curve
Flow versus pressure f~r caUbrated leaks
Family of leakage curves·
Smoke' puffer
- 13 -
ECRC/M1419
box
Figure 1. Apparatus for measuring the ventilation rate
sample pOint
fan T
mixing box I I I I I
gas analyser
X-t recorder
Figure 1. Apparatus for measuring the ventilation rate
Figure 2.
•
ECRC/M1419
----gas concentration--i"
Figure 2. ~o; decay tr~
gas concentration----+
Figure 3.
ECRC/M1419
o
1-
t ~ .,..
AoLL Sl41AfC.. '.,tlTS *' FlIfoIj .Iff A41. , •• "s •
100
"""' •• ,,,5 C L.' "'J" -Itw..,. Mllfltt>l 103
t:ctee H./O.
r50 "00,,.,. ttlaD
f minl4ta ~~----~~------~----------~--------~ __ -4 ____ ~ ---------- gas concentration-". Figure 3. (N;o d~;; traC;)
syringe
Figure 4. Calibration of gas analysers
20Um
valve
/gos syringe
septum
MIRAN 103
H rMIRAN I I I t •
~igure 4. Calibration of gas analysers
ECRC/M1419
gas concentration ---+ Figure 5. Calibration trace for gas analyser
.--rr :..
J .
I \
!.
1 CAl.iJ ~n.'" ~Q
MtII/Io %.
S-Opl .,.,.. .10 10 JO ... .. ,. 70 ., 10 up
/7".'7' tno
------gas concentration----+
. Figure 5. Calibration trace for gas analyser
ECRC/M1419
stop watch
soap film
in rubber bulb
50 100 true f low (soap f i lm) cc /minute
Figure 6. Rotameter calibration
(() stop watch
soap film l'
~ ::J
soap soluti on in rubber buLb
~ 100 u u
"6 6 en 50 c::
1
rotameter to be calibrated
Cii=====+gas supply
50 true flow (soap film)
100 cc/minute
Figure 6. Rotameter calibration
ECRC/Ml4l9
Pressurisation panel viewed from inside
Fan unit viewed from outside
Figure 7, House leakage apparatus
Pressurisation panel viewed from inside
Fan unit viewed from outside
Figure 7. House leakage apparatus
ECRC/M1419
ECRC/Ml4l9
A .1. .2. - 3· 4 '. 5 - rills
5 B .A R C
- 2. ·.3 ·4 5
4 A ·B C
C 1 2 3 4 5
,0- 3 5
146-A .R C
4 3
2
1 2 -
k" 'B ,c A B C
1
B c Figure 8. Flow calibration'of house leakage apparatus
Inside t o outside pressure di f ference
Figure 9 . House leakage curve
till oS "G .. ~ ... .. ~
I I
mls
B
7
6
5
4
3
2
1
10 20 30 40 50
ECRC/M1419
I.y. Bas ,.~ WII" 1<..,..,.,"'''
./1.7'''' '1' I$'tIS"-~~ • .e.
Inside to outside presaure difference
Figure 9. House leakage curve
Inside to outside pressure difference
Figure 10. Flow versus pressure for calibrated leaks
DO c:I .... 1 • ... ... • w I
J
m/s 8
7
6
5
4
3
2
1
10 20 30 40 so Inside to outside pressure difference
Figure 10. Flow versus pressure for calibrated leaks
ECRC/M1419
60 N/rrf
Figure
Pressure difference
11. Family of leakage curves
m/s 10
.:l Equivalent leakage area metre--i' 0'" d~ a~ fS~ ~t Jf"
Air ~
flow c:: 9· ... m'1h
.. "tJ at ., ~
~ ., 41
2000 8a a ., .If
7
1500 6
5
1000 4
3
500 2
1
10 20 40 50 Pressure difference
Figure 11. Family of leakage curves
ECRC/M1419
o·r?
0" • o·
0.0 '2-
_ 0·01
60 N/m:J.
Figure 12. Smoke puffer
ECRC/M1419
Figure 12. Smoke puffer