Real heat recovery with air handling units

C.-A. Roulet
a,*
, F.D. Heidt
b
, F. Foradini
c
, M.-C. Pibiri
a
a
Laboratoire d'Energie Solaire, et de Physique du Batiment, Ecole Polytechnique Federale, CH 1015 Lausanne, Switzerland
b
University of Siegen, Siegen, Germany
c
E4Tech, Av. Louis Ruchonnet 57, CH 1003 Lausanne, Switzerland
Abstract
More and more air handling units are equipped with heat recovery systems, with the aim of decreasing the energy use in buildings for
heating and cooling. The efciency of the heat recovery system is often used to calculate the energy saving. However, air-handling units do
not always function as planned. In particular, parasitic shortcuts and leakage may decrease dramatically the efciency of ventilation and
heat recovery. In addition, these units need electrical energy for fans, which may be more precious than saved heat. Measurements, using
tracer gas dilution technique have detected various malfunctions in several units.
This paper addresses real energy recovery with air handling units from a theoretical point of view and presents results of measurements
on 13 units. In the best three cases, the real, global heat recovery efciency was between 60 and 70% for units having a 80% nominal
efciency. In the three worst cases, the global efciency was less than 10%. For these cases, the heat recovery system uses more energy than
it saves. # 2001 Elsevier Science B.V. All rights reserved.
Keywords: Heat recovery; Air handling units; Efciency
1. Introduction
Ventilation in buildings especially in large buildings
and advanced low-energy and passive-solar houses is
becoming increasingly important for many reasons. One of
them is the excellent standard of thermal insulation, which
easily raises the contribution of ventilation losses
depending on the building's compactness and air change
rate to more than 50% of total thermal loss. Another
reason for the importance of ventilation is air-tightness of
buildings' envelopes, which avoids air inltration heat loss
but does not anymore provide sufcient ventilation. To cope
with ventilation requirements with regard to hygiene and
building physics, mechanical ventilation systems are of
increasing use. In order to reduce energy consumption,
ventilation systems with energy-efcient systems, recover-
ing heat from the exhaust air, are almost mandatory.
However, air-handling units may have parasitic shortcuts
and leakage [15], which can decrease dramatically the
efciency of ventilation and heat recovery. Moreover, leak-
age in a building's envelope allows warm air to escape
outdoors without passing through the heat recovery system.
In addition, these units use electrical energy for fans, which
may, in some cases, overpass the saved heat. The inuence
of these various phenomena on the real energy saving is
addressed in this paper.
2. Effect of leakages and shortcuts on heat recovery
2.1. Airflow rates, heat loss and heat recovery efficiency
Let us consider the air and heat ows in the unit sche-
matically represented in Fig. 1. Outdoor air o, enters the inlet
grille i, and is blown through the heat recovery system HR,
where it is either heated or cooled. Then, after subsequent
heating or cooling rs, it enters the supply duct s, to be
distributed into the ventilated space. As the envelope is not
perfectly airtight, the supply air is mixed with inltration air
inf, in the ventilated space. A part of the air may also be lost
by exltration (exf). The extract air x, passes through the
other part of the heat recovery system re, where it is either
cooled (if inlet air should be warmed up) or heated (if fresh
outdoor air should be pre-cooled). The air is then blown to
the outside through the exhaust duct e, to the atmosphere a.
If the exhaust and inlet grilles are not well situated, it is
possible that a part of this exhaust air re-enters the inlet
grille, resulting in an external recirculation rate R
e
. Leakage
through the heat recovery system may also result in an
Energy and Buildings 33 (2001) 495502
*
Corresponding author. Tel.: 41-21-6934557; fax: 41-21-6932722.
E-mail addresses: claude.roulet@epfl.ch (C.-A. Roulet),
heidt@physik.uni-siegen.de (F.D. Heidt), foradini@e4tech.com
(F. Foradini).
0378-7788/01/$ see front matter # 2001 Elsevier Science B.V. All rights reserved.
PII: S0 3 7 8 - 7 7 8 8 ( 0 0 ) 0 0 1 0 4 - 3
internal recirculation rate, from inlet to exhaust R
ie
, or from
extract to supply R
xs
.
In simplied methods to calculate heating (or cooling)
demand of buildings, ventilation heat loss, F
V
, is often
calculated by [6]:
F
V
= m(h
x
h
o
)(1 Z
G
) (1)
where m is the mass flow rate of outdoor air in kg/s, h
x
the
specific enthalpy of extract air, which is considered as
representative of the average indoor air, h
o
the specific
enthalpy of outdoor air, and Z
G
the global efficiency of
the heat recovery system.
This global efciency Z
G
, should consider the whole
system, consisting of the ventilated building and its ventila-
tion equipment. But, instead, often the nominal efciency of
the heat recovery unit itself e
HR
, is used. This efciency is
measured with balanced intake and exhaust airow rates
( m
re
= m
rs
) and is
e
HR
=
h
rs
h
i
h
x
h
o
=
h
x
h
re
h
x
h
o

y
x
y
re
y
x
y
o
(2)
where the signification of subscripts can be seen in Fig. 1,
and h are specific enthalpies of the air in J/kg. As a first
approximation, only sensible heat is considered, and the
temperatures at the same locations can be used. As shown
below, this replacement leads to optimistic results when the
air-handling unit has parasitic recirculation or when the
building has infiltration or exfiltration.
2.2. Global heat recovery efficiency
If there were no heat recovery, the heat loss of the building
F
L
, resulting from these airow rates is the sum of extract
heat ow and exltration heat loss, equal to the heat
necessary to bring outdoor air to indoor climate conditions:
F
L
= ( m
x
m
exf
)(h
x
h
o
) = ( m
s
m
inf
)(h
x
h
o
) (3)
Nomenclature
c heat capacity of air (J/(kg K))
C matrix of the tracer gas mass concentrations at
various locations
COP coefficient of performance of heat recovery
f
p
production factor, or primary energy used to
produce 1 kW h of electrical energy
f
r
part of the fan power recovered as heat in the
supply air
h specific enthalpy of the air (J/(kg K))
~
I vector containing the tracer gas injection rates
at various locations (kg/s)
m mass airflow rate (kg/s)

~ m vector of the air mass flowrates inthe ducts (kg/s)

Dp pressure drop
R recirculation ratio, i.e. part of the air that is
recirculated
R
e
recirculation ratio from exhaust to inlet through
outdoors
R
ie
recirculation ratio from inlet to exhaust through
heat recovery system
R
xs
recirculation ratio from extract to supply
through heat recovery system
SNES specific net energy saving per cubic meter of
supplied outdoor air (W h/m
3
)

V volume air flow rate

Greek letters
e
HR
nominal efficiency of the heat recovery unit
F
L
building ventilation heat loss without heat
recovery (W)
F
R
recovered heat flow rate (W)
F
v
ventilation heat loss (W)
g
inf
infiltration ratio, i.e part of the outdoor air that
enters the building by infiltration
g
exf
exfiltration ratio, i.e part of the outdoor air that
leaves the building by exfiltration
Z
fan
global efficiency of the fan, including the
motor efficiency
Z
G
global efficiency of the heat recovery system
Z
o
fresh air efficiency, i.e. the part of outdoor air
in the supply air
Z
re
air recovery efficiency, or part of the extract air
passing through the heat recovery unit
Z
x
extraction efficiency, or part of the air leaving
the ventilated volume, that is extracted through
the AHU
y temperature (K or 8C)
r density of air (kg/m
3
)
Subscripts
a atmosphere
e exhaust
exf exfiltration
HR heat recovery system
Fig. 1. The simplified network representing the air handling unit and
ducts. Arrows represent considered airflow rates.
i indoor
inf infiltration
o outdoor
re recovery unit, exhaust side
rs recovery unit, supply side
s supply duct
x extract
496 C.-A. Roulet et al. / Energy and Buildings 33 (2001) 495502
Recovered heat is
F
R
= m
re
(h
x
h
re
) = m
rs
(h
rs
h
i
) (4)
Since, in first approximation, all the heat taken from extract
air is given to supply air. The global heat recovery efficiency
of the system is then
Z
G
=
F
R
F
L
=
m
re
(h
x
h
re
)
( m
x
m
exf
)(h
x
h
o
)
=
m
re
( m
x
m
exf
)
e
HR
(5)
It can readily be seen that this global efficiency is not equal
to the nominal efficiency of the heat recovery system e
HR
.
An expression giving Z
G
as a function of the fresh airflow,
exfiltration, and recirculation rates can be derived from
Eq. (5), by taking account of mass conservation at the nodes
of the system (see Appendix A). This relation is
Z
G
= Z
x
Z
re
e
HR
(6)
where
Z
x
=
m
x
m
x
m
exf
(7)
is the extraction efficiency, i.e. that part of the air leaving the
ventilated volume which is extracted through the AHU, and
Z
re
= 1 R
xs
=
m
re
m
x
(8)
is the air recovery efficiency, or that part of the extract air
that passes through the heat recovery unit.
Looking to Eq. (7), it seems at rst glance that the global
heat recovery efciency depends only on extract and exl-
tration airow rates. However, the purpose of ventilation is
to provide fresh, outdoor air in the ventilated volume. Let us
see howEq. (7) is changed when the fresh airowrate is used
as reference.
Fresh air entering the AHU is m
o
. Because of external
recirculation, this air is mixed with exhaust air into the inlet
duct. Apart R
ie
of this mix is recirculated to the exhaust duct.
All the fresh, outdoor air that enters the building through the
AHU is found in m
rs
, which is, from the denition of R
ie
and
using Eq. (A.6) of Appendix A
m
rs
= (1 R
ie
) m
i
= (1 R
ie
)( m
o
R
e
m
e
) (9)
Since m
e
is no more fresh, the only part of m
rs
which is fresh
is m
o
(1 R
ie
). Therefore the total fresh airflow rate entering
the ventilated space in building is
m = m
o
(1 R
ie
) m
inf
(10)
which means that
m
o
(1 R
ie
) = m m
inf
(11)
replacing in Eq. (7) m
x
by its value given by Eq. (A.10), and
taking into account the above relation gives finally
Z
G
=
[1 g
exf
R
e
R
ie
(g
inf
g
exf
)[(1 R
xs
)
1 R
e
R
ie
(g
inf
g
exf
) g
exf
[R
e
R
xs
(1 R
e
)[
e
HR
(12)
where
g
inf
=
m
inf
m
and g
exf
=
m
exf
m
(13)
are, respectively the infiltration and exfiltration ratios. In
other terms, the extraction efficiency is
Z
x
=
1 g
exf
R
e
R
ie
(g
inf
g
exf
)
1 R
e
R
ie
(g
inf
g
exf
) g
exf
[R
e
R
xs
(1 R
e
)[
(14)
which depends on all parasitic airflow rates. When there
is no external recirculation (R
e
= 0), Eq. (12) simplifies
to
Z
G
=
(1 g
exf
)(1 R
xs
)
1 R
xs
g
exf
e
HR
(15)
and infiltration has no effect. In this case, exfiltration
through the envelope, and internal recirculation from extract
to supply ducts have the same effect, since both drive air
away from the heat recovery device. Eq. (12) is illustrated in
Fig. 2.
Global efciency Z
G
equals the effectiveness e
HR
only if
there is no exltration, and there is neither external- nor
extract-to-supply recirculation. Otherwise, Z
G
is smaller
than e
HR
.
The inlet to exhaust recirculation, as well as the inltra-
tion ratio have only a small effect on heat recovery ef-
ciency, but reduces the amount of fresh air supplied by the
unit to the ventilated space. In order to get the same amount
of fresh air, the supply airow rate should be increased.
Fresh air efciency can be dened by
Z
o
=
m m
inf
m
s
=
m
o
(1 R
ie
)
m
s
(16)
This recirculation obviously results in an increased con-
sumption of electric energy for the fans, which is approxi-
mately proportional to the cube of the airflow rate, without
delivering more fresh air. However, such parasitic recircula-
tion is often not noticed, and hence can lead to an undis-
covered reduction of indoor air quality.
Fig. 2. Global heat recovery efficiency in function of exfiltration ratio g
exf
and internal recirculation rate R
xs
. In this figure, e
HR
= 100, R
ie
= R
e
= 10
and g
inf
= 50%.
C.-A. Roulet et al. / Energy and Buildings 33 (2001) 495502 497
3. Electric energy consumption for fans
The electric power used by the fan to blow air in a
ductwork is
F
fan
=

V Dp
Z
fan
=
mDp
rZ
fan
(17)
where

V is the volume air flow rate, Dp the pressure drop
across the fan, equal to the pressure drop for the ductwork
ventilated by this fan, Z
fan
the global efficiency of the fan,
including the motor efficiency, r the density of air.
The pressure drop through a ductwork is proportional to
the square of the airow rate
Dp = K
m
2
r
(18)
where K is a constant (with dimension m
4
) depending on
the characteristics of the ductwork. Therefore
F
fan
= K
m
3
r
2
Z
fan
(19)
4. Specific net energy saving
A crucial issue is that HR-systems recover thermal
energy, but use electrical energy for the fans. As a useful
gure to deal with this fact we introduce the specic net
energy saving per cubic meter of supplied outdoor air (SNES
in W h/m
3
) averaged over a heating period, for which the
mean outdoor temperature is y
o
. This gure is calculated by
SNES = r
Z
G
F
L
F
fan
(f
r
f
p
)
m
(20)
where F
L
= mc(y
x
y
o
) is the ventilation heat loss, based
on average internal and external temperature during the
heating season, f
r
the part of the fan power recovered as
heat in the supply air. This factor f
r
is close to one for supply
fans and zero for exhaust fans, f
p
a production factor,
accounting for the fact that the production of 1 kW h of
electrical energy requires much more primary energy.
Only if SNES is positive, a net gain in thermal or primary
energy is achieved by the HR-system. Otherwise the system
even wastes energy.
Another interesting gure is the coefcient of perfor-
mance (COP), dened by the ratio of recovered heating
power and used electrical power
COP =
Z
G
F
L
f
r
F
fan
F
fan
(21)
This COP is defined without taking account of the produc-
tion factor f
p
, as it is usually the case for heat pumps.
5. Measurement methods
5.1. Airflow rates and leakages
The tracer gas dilution method is used since several years
for diagnosis of air handling units [79]. It has the advantage
of being applicable on any unit and to allow the detection
and quantication of unexpected airows, such as leaks or
shortcuts.
The technique is described in more detail elsewhere [5,9
11] gases are injected, most often at a constant owrate, at
carefully chosen locations in the air-handling unit. Experi-
ence has shown that most practical and efcient injection
locations are as indicated in Fig. 3.
Tracer gas concentrations are measured at various loca-
tions, in order to obtain enough equations from conservation
of airow and tracer gas ows to determine all required
airowrates. Conservation of air and tracer gas mass ows at
each node shown in Fig. 3 provides a system of linear
equations, assuming steady state and perfect mixing of
tracer gas at sampling locations [9]:
~
I = C

~ m (22)
where C is a matrix containing the tracer gas concentrations
measured at the various locations,
~
I the vector containing the
tracer gas injection rates,

~ m the vector of the air mass flow
rates to be determined.
Fig. 3. Schematics of an air handling unit showing main and secondary airflow paths (arrows), tracer gas injection (stars), and sampling points for
concentration measurements (C
i
). The equivalent network used to derive equations of system (22) is also shown.
498 C.-A. Roulet et al. / Energy and Buildings 33 (2001) 495502
Depending on the number of tracer gases injected and on
the number of sampling locations, this system may contain
more equations than unknowns. Several possibilities to take
advantage of this situation were tested. The most robust
method, providing the most accurate results, was found to
select a sub-set of N equations (N being the number of
airow rates to be determined), which provide the required
airow rates in the most direct way, that is using the smallest
number of input variables in each equation.
Noting the injection starting time of tracer gas and
recording the evolution of its concentration in the exhaust
duct versus time, allows the assessment of the mean age of
air and the air exchange efciency in the ventilated space
[9,12].
5.2. Measurement of fan power efficiency
The mechanical power delivered by a fan is the product of
the volume airow rate

V delivered by the fan, times the
pressure difference Dp across the fan. Airow rate is mea-
sured by tracer gas dilution technique, and pressure differ-
ence is easily measured with a differential manometer. The
electrical power consumed by the fan motor F
fan
, is mea-
sured with a wattmeter, and the fan efciency is
Z
fan
=

V Dp
F
fan
(23)
5.3. Efficiency of heat exchangers
The energy (or enthalpy) efciency of the heat recovery
system itself is the ratio of the enthalpy ow delivered to the
supply air over the enthalpy ow available in exhaust air. It
can be calculated using the rst term in Eq. (2).
Since airow rates are known from previous measure-
ments, assessment of temperature and moisture content of
air at three locations only allows the determination of the
enthalpy efciency. Care should be taken to measure tem-
perature and air humidity in well mixed zones, especially
downwind the exchanger, and to check that that the air is
neither heated, cooled, dried or humidied between the heat
recovery system and the measurement location.
Much simpler to assess and often used in practice, is the
temperature efciency, telling how well the temperature is
recovered. This efciency is simply calculated from tem-
perature measurements upwind and downwind the heat
exchanger in both supply and exhaust channels, using the
third term in Eq. (2).
6. Results
6.1. Measured airflow rates
Thirty air handling units located in 14 buildings were
measured during several measurement campaigns design
values and measured recirculation rates are compared in
Fig. 4 left. These are seldom equal. Note that 11 units out of
23 designed without recirculation have shown sig-
nicant, but unexpected recirculation. These leaks strongly
reduce the indoor air quality by mixing return air to supply
air.
In principle, supply and exhaust airow rates are either
balanced, or put the building under a slight pressure differ-
ence against outdoor conditions. When the envelope is not
airtight, and when the difference between supply and
exhaust airow rates is too large, air may be leaking through
the buildings' envelope. This has not much inuence on
indoor air quality, but may strongly decrease the efciency
of heat recovery. Fig. 4 right shows design and measured
extract efciencies, i.e. the part of the indoor air that is
extracted through the heat recovery unit. As shown in this
gure, more than 50% of the supplied air is lost that way in
one third of the audited units.
6.2. Fan efficiency
The global efciency of fans was measured in several
AHU's. It is the ratio of the mechanical power of the air
moved by the fan and the electrical power used by the fan.
Important is to use measured quantities, since design values
are not always achieved, as shown in Fig. 5.
Fig. 6 shows that the global fan efciency grows, as
expected, with the fan measured power. There are, however,
Fig. 4. Comparison of design and measured recirculation rate (left) and exfiltration ratios (right) in 30 units. Triangles are related to air handling units
mentioned in this paper.
C.-A. Roulet et al. / Energy and Buildings 33 (2001) 495502 499
signicant differences between fans within the same power
class. For example, fans with 350550 W have efciency
between 14 and 32%, the best using the smallest power. In
the 13001700 W class, the efciency also varies between
15 and 38%, and so on. Even in the 30004000 W class,
there are variations from 55 to 70%.
6.3. Global efficiency
Airow rates and heat exchanger efciencies were mea-
sured in 10 large units located at the EPFL, and three small,
wall-mounted room ventilation units, measured at the Uni-
versity of Siegen. The main characteristics of these units are
summarised in Table 1.
Recirculation ratios and efciencies measured in these
units are given in Table 2. In this table, the SNES and COP
are calculated with 16 K indooroutdoor average tempera-
ture difference during 210 days, a recovery factor for fans,
f
r
= 0X5 (taking account that here are two fans in these units,
one of them in the supply duct) and a production factor
f
p
= 3X55, which is the average for low-voltage electricity in
Europe according to Frichtknecht et al. [13]. Note that a
common value used in Germany for f
p
is 2.8. French and
Dutch regulations give smaller values 2.58 and 2.56, respec-
tively.
Major leakages have been observed in several buildings.
In three of them, inltration represents a signicant part of
the outdoor air, and in four of them, most of the air leaves the
building through the envelope instead of passing the heat
recovery unit. Signicant internal recirculation is observed
in the three small units, and external recirculation above
20% is measured in three large units. These leakages sig-
nicantly affect heat recovery efciencies, which drop from
nominal values between 50 and 90% down to actual values
ranging between 5 and 68%. On the average, the heat
recovery effectiveness e
HR
is 70%, but the global, real
efciency is only 43%. In the best case, an 80% heat
recovery effectiveness is reduced by 15% down to a 68%
real efciency.
Specic net thermal energy savings (SNES) can be very
small or even negative. In the best case, it reaches 2.7
W h/m
3
, corresponding to 8 K average temperature increase
of fresh air. It should be also noticed that the COP might
be much smaller than expected, as it is often the case for
air-to-air heat pumps. A COP less than 2.5 indicates that
Fig. 5. Measured electrical power of fan motors compared with design
values.
Fig. 6. Global fan efficiency as function of measured power.
Table 1
Measured airflow rates with experimental uncertainty band (when available), total and specific fan power in audited units
Unit Airflow rates (m
3
/h) Fan power
Outdoor air Supply air Extract air Exhaust air W W h/m
3
B30 1900 100 2070 70 1790 40 1600 200 990 0.27
TP 2530 80 2900 200 1860 50 1500 200 850 0.19
BH 2380 70 2480 70 1930 40 1830 50 1800 0.42
CS 2200 300 3400 100 3240 90 2000 2000 1800 0.33
E1 5000 200 5400 100 6000 700 5500 700 3710 0.34
E2 15000 2000 16400 700 11000 1000 10000 3000 11800 0.45
E12 11000 400 11600 200 10000 300 9500 900 8180 0.39
E13 16000 1000 17400 700 13400 600 12000 2000 9760 0.33
E14 9000 1000 10000 2000 1970 90 1000 3000 3800 0.35
E15 14300 600 16200 400 3420 70 1000 1000 7970 0.45
HA 25 36 34 24 13 0.22
HB 42 75 74 41 27 0.24
HC 74 87 87 74 32 0.20
500 C.-A. Roulet et al. / Energy and Buildings 33 (2001) 495502
the heat recovery is less efcient than heating the air with a
gas boiler with 75% efciency [14].
Best net energy saving in large units (E12 and E13 in
Tables 1 and 2) is 80 00090 000 kW h per winter season,
but another unit (E15) actually spills as much energy. Small
units, (HA, HB and HC) save between 80 (HB) and
350 kW h (HC) during an entire season. From energetic
and economic aspects, only such ventilation units are dis-
advantageous and hard to recommend.
Note that these results are obtained when the heat recov-
ery is functioning. Yearly average efciency may even be
smaller, due to reduced operation time [15].
7. Conclusions
Heat recovery from extract air is often installed in
advanced low energy buildings in order to ensure efcient
ventilation at low energy cost. However, global efciency of
heat recovery depends signicantly on air inltration and
exltration, which should be minimised during the heating
period. Internal and external recirculation also decrease the
efciency of the heat recovery units. Moreover, electrical
energy for fans has to be used in order to supply fresh air and
to recover thermal energy from exhaust air.
Characteristic gures for the evaluation of ventilation
units with heat recovery have been dened and measured
using the tracer gas dilution method. The most important of
them are fresh air efciency, global efciency of heat
recovery and specic net energy savings.
For several examined ventilation units energetic savings
were small or even negative. Even if best technical perfor-
mance is assumed (airtight building, e
HR
= 90%, specic
fan power equal to 0.27 W h/m
3
), the economic viability of
small ventilation units remains questionable. This, however,
does not affect the other qualities of ventilation systems such
as steady supply of fresh air with low concentrations of
contaminants.
Measurements on a series of air handling units clearly
show that, specications of these units should be greatly
improved, in order to ensure that design airow rates are
achieved, and that no parasitic airow's such as leaks or
shortcuts do occur. The measured units were located in
Switzerland and Germany, and more measurements should
be performed to get similar information in other countries.
Acknowledgements
The authors thank the Ministry of Science and Technol-
ogy of North Rhine-Westphalia, Germany, for supporting
investigations within the project ``Labor fur bauphysika-
lische Messungen und Materialuntersuchungen'' of the AG
Solar NRW under No. 253 134 96, as well as the Swiss
Federal Ofce of Energy, supporting the MEDITA project
(contract 19358) and the Swiss Federal Ofce of Education
and Science, sponsoring the Swiss participation to European
Contracts JOU2-CT920022 and JOR3-CT97-0171. They
also thank Pascal Cretton for careful checking of the equa-
tions. Dr. Heidt is also indebted to his former co-worker
Dipl.-Phys. T. Fischer for his valuable help in the compila-
tion and evaluation of the experimental data presented in this
paper.
Appendix A. Derivation of Eq. (6)
We have mentioned the following recirculation rates
External R
e
=
m
i
m
o
m
e
=
m
e
m
a
m
e
(A.1)
Inlet to exhaust R
ie
=
m
i
m
rs
m
i
=
m
e
m
re
m
i
(A.2)
Extract to supply R
xs
=
m
s
m
rs
m
x
=
m
x
m
re
m
x
(A.3)
Table 2
Outdoor air efficiency Z
o
, exfiltration and infiltration ratios g
exf
and g
inf,
external and internal recirculation rates R
e
, R
ie
and R
xs
, heat recovery effectiveness
e
HR,
global heat recovery efficiency Z
G
, specific net energy saving SNES in W h/m
3
, and coefficient of performance COP, of audited air handling units.
Unit Z
o
(%) g
exf
(%) g
inf
(%) R
e
(%) R
xs
(%) R
ie
(%) Z
x
(%) e
HR
(%) Z
G
(%) SNES COP
B30 97 16 0 6 7 0 86 70 56 1.55 6.5
TP 92 47 9 20 5 0 59 70 39 1.35 8.0
BH 100 29 7 0 5 0 72 90 62 1.18 5.2
CS 68 77 76 55 1 0 31 30 9 0.05 3.3
E1 98 8 17 0 7 0 92 80 69 1.92 6.7
E2 97 43 8 0 6 0 61 90 52 0.69 4.5
E12 100 14 0 4 2 0 87 80 68 1.45 5.5
E13 97 25 0 0 0 0 77 70 54 1.17 5.5
E14 95 97 49 0 0 0 10 50 5 0.37 1.8
E15 93 91 18 100 6 0 18 50 8 0.92 1.5
HA 74 8 0 0 33 0 94 63 40 1.37 6.2
HB 57 2 0 0 44 4 99 80 44 2.21 6.8
HC 68 0 0 0 39 25 100 90 55 2.69 8.2
C.-A. Roulet et al. / Energy and Buildings 33 (2001) 495502 501
The mass flow balance for the whole building is
m
a
m
exf
= m
o
m
inf
(A.4)
Combining this equation with the definition of the external
recirculation rate, we get
m
e
=
1
1 R
e
( m
o
m
inf
m
exf
) (A.5)
Then, writing the mass flow rate balance at node (see
Fig. 1), we get
m
i
= m
o
R
e
m
e
=
m
o
R
e
( m
inf
m
exf
)
1 R
e
(A.6)
The mass balance at node gives
m
rs
= (1 R
ie
) m
i
=
(1 R
ie
)
(1 R
e
)
[ m
o
R
e
( m
inf
m
enf
)[
(A.7)
From mass balances at nodes and
m
s
= m
rs
R
xs
m
x
(A.8)
and
m
s
= m
x
m
exf
m
inf
(A.9)
we get
m
x
=
1
1 R
xs
[ m
rs
m
inf
m
exf
[
=
m
o
(1 R
ie
) (1 R
e
R
ie
)( m
inf
m
exf
)
(1 R
xs
)(1 R
e
)
(A.10)
Mass balance at node gives
m
re
= m
x
(1 R
xs
) (A.11)
Therefore
Z
G
=
m
x
(1 R
xs
)
m
x
m
exf
e
HR
= Z
x
Z
re
e
HR
(A.12)
References
[1] A.R. Hanlo, Use of tracer gas to determine leakage in domestic heat
recovery units (HRV), in: Proceedings of the AIVC 12th Conference
on Air Movement and Ventilation Control within Buildings, Ottawa,
Canada, 1991.
[2] T. Fischer, F.D. Heidt, Testing the ventilation efficiency of ventilation
units with tracer gas methods, in: Proceedings of the Second Inter-
national Conference on Buildings and the Environment, Paris, 1997.
[3] F.D. Heidt, T. Fischer, A. Thiemann, Energetische Beurteilung
dezentraler Raumluftungsgerate mit Warmeruckgewinnung Meth-
oden und Beispiele, Heizung-Luftung/Klima-Haustechnik 49 (11)
(1998) 5262.
[4] C.-A. Roulet, F. Foradini, P. Cretton, in: Proceedings of the Healthy
Buildings'94 on Use of Tracer Gas for Diagnostic of Ventilation
Systems, Budapest, 1994.
[5] C.-A. Roulet, F. Foradini, L. Deschamps, in: Proceedings of the
Indoor Air'99 on Measurement of Air Flow Rates and Ventilation
Efficiency in Air Handling Units, Edinburgh, 1999.
[6] EN 832, Thermal performance of buildings, Calculation of energy
use for heating, Residential buildings, CEN, Brussels, 1999.
[7] K.H. Presser, R. Becker, Tracing air flows with laughing gas.
Measuring of air flow in air conditioning plants by means of the
tracer gas method (Mit Lachgas dem Luftstrom auf der Spur
Luftstrommessung in Raumlufttechnischen Anlagen mit Hilfe der
Spurgasmethode), Heiz Luft Haustech 39 (1) (1988) 714.
[8] S.B. Riffat, Airflow rate through a heat exchanger coil: comparison
of measurements made with a rotating vane-anemometer and a tracer-
gas technique, Appl. Energy (1990).
[9] C.-A. Roulet, L. Vandaele, Airflow patterns within buildings
measurement techniques, AIVC technical note, Vol. 34, 1991,
Bracknell, Berkshire RG124AH, GB: AIVC 265.
[10] F.D. Heidt, R. Rabenstein, MULTI-CAT fur Luftungskenngroen,
KI-Luft-und Kaltetechnik 30 (3) (1994) 119124.
[11] C.-A. Roulet, et al., in: Proceedings of the CISBAT'99 on
Measurement of Air Flow Rates and Ventilation Efficiency in Air
Handling Units, LESO-PB, Lausanne, 1999.
[12] M. Sandberg, M. Sjoberg, The use of moments for assessing air
quality in ventilated rooms, Building Environ. 18 (1984) 181197.
[13] R. Frischknecht, et al., Oekoinventare fur Energiesysteme, Zurich,
ETHZ, 1994.
[14] P. Ruyssevelt, Ventilation and heat recovery in superinsulated houses.
In: Proceedings of the UK-ISES Conference on Ventilation and Heat
Recovery in Superinsulated Houses, 1987, pp. 5767.
[15] M.K. Drost, Air to air heat exchanger performance, Energy and
Buildings 19 (1993) 215220.
502 C.-A. Roulet et al. / Energy and Buildings 33 (2001) 495502