Professional Documents
Culture Documents
Applied Energy
journal homepage: www.elsevier.com/locate/apenergy
a r t i c l e i n f o a b s t r a c t
Article history: Thermally activated building systems (TABS) integrate the building structure as energy storage, and have
Received 22 August 2008 proofed to be energy efficient and economic viable for the heating and cooling of buildings. Although
Received in revised form 9 January 2009 TABS are increasingly used, in many cases control has remained an issue to be improved. In this paper,
Accepted 10 January 2009
a method is outlined allowing for automated control of TABS in intermittent operation with pulse width
Available online 15 February 2009
modulation (PWM). This method represents one part of a TABS control solution with automatic switching
between cooling and heating modes for variable comfort criteria which was published before. A first
Keywords:
pulse width modulation control solution is derived based on a simple 1st order model of TABS. Then a
Thermally activated building systems
Concrete core conditioning systems
second, even simpler solution is given that significantly reduces the tuning effort. Finally, the paper out-
Building control lines a pulse width modulation control procedure and gives two application examples of the PWM con-
Pulse width modulation trol carried out in a laboratory test room.
Intermittent operation Ó 2009 Elsevier Ltd. All rights reserved.
0306-2619/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.apenergy.2009.01.008
M. Gwerder et al. / Applied Energy 86 (2009) 1606–1616 1607
Nomenclature
2. Control concepts for TABS paper only the TABS zone control and in particular the intermittent
operation of the zone pumps is described in detail.
2.1. Conventional control concepts for TABS In Fig. 1, the overall control concept for a TABS plant with two
zones is shown. Both zone controllers are fed with the following
So far, only a few authors have reported on control of TABS, signals: outside air temperature #oa , supply water temperature
including Meierhans as one of the TABS pioneers [6], Olesen [7], #sw , room temperature #r (optional) and room temperature set-
Antonopoulos et al. [8] and Weitzmann [9] with a short overview. point range [#r;SpH ; #r;SpC ]. The zone controllers operate the zone
Tödtli et al. [4] evaluated existing control solutions and concluded pumps (on/off) and the zone heating and cooling valves. Together
that they mostly have the following properties: (a) they are based with the zone supply water temperature set-points #sw;SpH;PWM and
on an outside temperature compensated supply water tempera- #sw;SpC;PWM , these signals can also be used to set up a demand
ture control, the set-point of the supply water temperature being dependent control of the heat and cold generation and distribution
shifted with varying outside temperature according to the heating as indicated in Fig. 1.
curve without considering heat gains, and the cooling curve typi- In Fig. 2, the TABS zone control is shown in more detail. The
cally being a constant supply water temperature set-point based control is modular, i.e. it contains two or more functionality parts.
on the maximum load situation. (b) No feed-back variable from Mandatory parts are indicated by solid frames, optional parts by
the zones (return temperature, concrete core temperature or room dotted frames. If an optional part is not needed or required, only
temperature) is used for the control, thus banking on the self-reg- the feed-through indicated by dotted arrows in the corresponding
ulation effects of the concrete core conditioning system. (c) The control part function block has to be realized. The functionalities of
heat and cold generation are enabled or activated dependant on the four control parts are explained below.
the season and/or the outside temperature. (d) Free cooling is ac-
counted for in a heuristic way. 2.2.1. Room temperature feedback control
If one or several room temperatures #r are measured in the con-
2.2. New control concept trolled zone, a room temperature feedback control part can be
added to the zone control. Since TABS react slowly, only day-to-
In [5,10], a new control concept was presented that is one part day room temperature compensation is promising, an instant cor-
of an integrated process consisting of the tasks to design a TABS rection can not be achieved with TABS. In Fig. 2, the room temper-
(planning of HVAC system and of its control [11]), to commission ature controller adjusts the desired room temperature set-point
it and to optimize it during operation [12]. In the design phase of range [#r;SpH ; #r;SpC ] to a corrected set-point range [#r;SpH;FB ; #r;SpC;FB ]
a TABS, a base TABS zone control strategy is selected that only re- that is processed in the further control parts in order to maintain
flects the basic control action, e.g. the base control strategy ‘‘supply the room temperature within [#r;SpH ; #r;SpC ]. Advantages and disad-
water temperature control” (see Figs. 1 and 2, control with manda- vantages of room temperature control as well as a realization
tory parts only). In real operation, this base control strategy can be example are given in [10].
extended by optional control parts to improve comfort and/or en-
ergy efficiency depending on given requirements and installations 2.2.2. Outside air temperature compensated supply water temperature
[10] (see Fig. 2, optional control parts). The result is a modular con- control
trol concept for a TABS zone which – by configuration – can be This core control part determines a supply water temperature
adapted to different TABS plants. Besides the zones, also the heat set-point range [#sw;SpH ; #sw;SpC ] as a function of the mean outside
and cold generation and distribution have to be controlled. In this air temperature #oa of the last 24 h (moving time window) and
1608 M. Gwerder et al. / Applied Energy 86 (2009) 1606–1616
Fig. 1. The overall TABS control concept, including heat/cold generation and distribution system.
Fig. 2. Scheme of the modular zone control (mandatory parts with solid frame, optional parts with dotted frame).
the current room temperature set-point range [#r;SpH;FB ; #r;SpC;FB ]. For 2.2.3. Pulse width modulation
that purpose, the so called heating curve (HC) and cooling curve The control parts presented so far are based on continuous
(CC) are used. The heating and cooling limits serve to identify the operation of the zone pumps. A third control part (see Fig. 2) can
operating mode depending on #oa (see Fig. 2). In order to define ini- be added to operate the zone water circulation pump in intermit-
tial parameters of this control part, an unknown-but-bounded de- tent operation with pulse width modulation (PWM). Advantages
sign process can be followed [5,11]. Detailed information on this and disadvantages of intermittent operation control are given in
control part and its application can be found in [12]. [10].
M. Gwerder et al. / Applied Energy 86 (2009) 1606–1616 1609
!
2.2.4. Sequence control ~ þ Rl;f Þ
Rt ðR d#c ðtÞ ~ þ Rl;f
R
A standard sequence controller that controls the supply water Cs ¼ #c ðtÞ þ # ðtÞ
Rt þ R~ þ Rl;f dt ~ þ Rl;f sw
Rt þ R
temperature #sw (the according set-point range is [#sw;SpH;PWM ;
#sw;SpC;PWM ]) is acting on the heating and cooling valves. Depending Rt
þ ð#oa ðtÞ þ Rl;f q_ g ðtÞÞ ð1Þ
on operating mode and operating state (i.e. pump on/off com- ~ þ Rl;f
Rt þ R
mand) one or both sequences are disabled and the respective valve
We substitute two terms: the time constant of the model s1 and
is closed.
the perturbation #c1 ðtÞ.
!
3. A simple thermal model for TABS ~ þ Rl;f Þ
Rt ð R
s1 ¼ C s ~ þ Rl;f
ð2Þ
Rt þ R
The PWM control method introduced in Section 2.2 is derived
from a simple dynamic thermal model that reflects the thermal ~ þ Rl;f
R Rt
behavior of TABS. Therefore, this model is described first, followed #c1 ðtÞ ¼ #sw ðtÞ þ ð# ðtÞ þ Rl;f q_ g ðtÞÞ ð3Þ
~
Rt þ R þ Rl;f ~ þ Rl;f oa
Rt þ R
by the detailed derivation of the PWM control method.
For constant #oa ; #sw and q_ g , the solution of (1) is given by (4).
3.1. Modeling of the piping system The stationary solution is then given by #c1 .
st
For the purposes of design and performance simulations of #c ðtÞ ¼ #c1 þ ð#c ð0Þ #c1 Þ e 1 ð4Þ
TABS, a model was developed which depicts the heat transfer in
The first order differential equation for the slab core tempera-
the slab and to adjacent rooms [3]. This TABS model – under cer-
ture node of the model with zone pump deactivated can be
tain restrictions – allows reducing the 3-dim heat transfer in the
written:
slab to a 1-dim approach by establishing a correlation between
supply water temperature, core temperature (mean slab tempera- ~ þ Rl;f Þ d#c ðtÞ
C s ðR ¼ #c ðtÞ þ #oa ðtÞ þ Rl;f q_ g ðtÞ ð5Þ
ture in the plane of the piping system) and room temperature dt
(operative temperature). The main parameter to model the piping Again, we substitute two terms: the time constant of the model
system is the equivalent resistance Rt, in which the geometrical s0 and the perturbation #c0 ðtÞ:
characteristic, material parameters and the influence of the fluid
mass flow are summarized (Fig. 3). The dynamic thermal behavior s0 ¼ C s ðR~ þ Rl;f Þ ð6Þ
of the upper and lower parts of the slab can be modeled in an arbi-
trarily complex manner ranging from a simple 1-node-model as #c0 ðtÞ ¼ #oa ðtÞ þ Rl;f q_ g ðtÞ ð7Þ
used in this paper (see Section 3.2) to detailed multi-layer models.
For constant #oa and q_ g , the solution of (5) is given by (8). The
A detailed description of the TABS-resistances can be found either
stationary solution is then given by #c0 .
in [5] (Appendix 1) or [13] (Annex B).
st
#c ðtÞ ¼ #c0 þ ð#c ð0Þ #c0 Þ e 0 ð8Þ
3.2. Dynamic thermal model of TABS room
As thermal inertia plays a pivotal role in the behavior of TABS, a 4. Pulse width modulation (PWM) control
model used for TABS pulse width modulation should reflect in par-
ticular the thermal behavior of the activated thermal mass. Fig. 4 4.1. PWM calculations using 1st order model
shows the structure of a 1st-order resistance–capacitance model.
The slab core temperature node #c with thermal capacitance Cs is Here, the 1st order model from Section 3.2 is taken for PWM
linked to the supply water temperature #sw by the TABS-resistance calculations. A similar calculation procedure for a different applica-
Rt (only if zone pump is running) and to the operative room tem- tion can be found in [14].
perature #r by the resistance R. ~ The internal and solar heat gains Firstly, we examine the heating mode where heat is delivered to
q_ g affect the room node which is linked to the outside air temper- the thermally activated structure. In order to maintain a room tem-
ature #oa by a façade-resistance Rl,f (calculation of resistances see perature #r which is equal or higher than the lower room temper-
[3] or [5]). ature set-point #r;SpH also the core temperature must be kept above
The 1st-order differential equation for the slab core tempera-
ture node of the model with running zone pump can be written:
a certain level. This minimal core temperature #c;Min can be calcu- non-constant heat gains as can be seen in Section 5.3. For the out-
lated as (see Fig. 4): side air temperature disturbance, not the actual value is taken but
~ instead a typically very slowly moving average of the last 24 h
R ~ q_ g (moving time window), see Section 2.2. By considering these mea-
#c;Min ¼ #r;SpH þ ð#r;SpH #oa Þ R ð9Þ
Rl;f sures, the altered disturbances are approximately constant during
Considering a quasi-stationary operation with constant #oa , con- the PWM period and the use of Eqs. (12) and (14) in such a way is
stant #sw ¼ #sw;SpH;PWM , constant q_ g ¼ q_ g;elb and the zone pump adequate also in non-ideal operation as has been shown by simu-
switching on and off with constant switched-on phase Dt1 and lations [12].
constant switched-off phase Dt0, the core temperature has always
to be equal or higher than #c;Min . As can be seen from (10) at the 4.2. Simplified PWM calculations
beginning of a new switched-on phase (time Dt1 + Dt0) the core
temperature #c has to be equal to #c;Min . As an alternative to the method using a 1st order model, a sec-
Dt 1
ond approach for PWM calculation is derived here which is solely
s
#c ðDt 1 Þ ¼ #c1 þ ð#c;Min #c1 Þ e 1 based on an energy contemplation. The basic idea of this second
s0
Dt approach is that during the switch-on phase with PWM operation
#c ðDt 1 þ Dt 0 Þ ¼ #c0 þ ð#c ðDt 1 Þ #c0 Þ e 0 ¼ #c0 an equal amount of energy has to be delivered to the thermally
Dt 1 Dt
s s0 activated structure as would have been delivered during the time
þ ð#c1 þ ð#c;Min #c1 Þe 1 #c0 Þ e 0 ¼ #c;Min
span of an entire PWM cycle but with continuous pump operation.
ð10Þ Again, we consider first the heating mode where #c P #c;Min is re-
Combining (3), (7), and (10), the correlation (11) describing the quired in order to maintain thermal comfort in the room.
PWM operation is obtained: In continuous operation, the energy input Econt over one PWM
period DtH for constant #oa , constant #sw ¼ #sw;SpH and constant
~ þ Rl;f
R
#c;Min ¼ #c0 þ ð#sw;SpH;PWM #c0 Þ q_ g ¼ q_ g;elb is given by
~ þ Rl;f
Rt þ R Z Dt H
! ! 1 Dt H
~ þ Rl;f
R Rt Dt 1 Dt 0 Econt ¼ ð#sw;SpH #c;Min Þdt ¼ ð#sw;SpH #c;Min Þ
þ #c;Min # # e
s
1
s
e 0 Rt t¼0 Rt
~ þ Rl;f sw;SpH;PWM Rt þ R
Rt þ R ~ þ Rl;f c0
Dt H
ð11Þ ¼ ð# #r;SpH Þ ð15Þ
Rt þ R~ sw;SpH
Depending on what type of PWM control is desired, two out of In intermittent operation, the energy input EPWM over one PWM
the three quantities #sw;SpH;PWM , Dt0 and Dt1 can be defined and the period DtH for constant #oa , constant #sw ¼ #sw;SpH;PWM and constant
third one then can be calculated by applying (11). Doing so for gi- q_ g ¼ q_ g;elb can be calculated in a similar way. Since in PWM opera-
ven Dt0 and Dt1, we obtain (12) for #sw;SpH;PWM in the heating mode. tion, the pump is only partially running, the integration over time
~ þ Rl;f
Rt þ R only runs from pump switch-on to switch-off time, i.e. from 0 to
#sw;SpH;PWM ¼ Dt1,H:
~ þ Rl;f
R
2 Dt Dt
Dt
3 Z
0 1 0 Dt 1;H
#c;Min 1 e s0 e s1 #c0 1 e s0 1
6 Rt 7 EPWM ¼ ð#sw;SpH;PWM #c ðtÞÞdt
64 #c0 7
5 Rt t¼0
Dt
s0
Dt
s1 R ~
t þ R þ Rl;f Z Dt1;H
e 0 1e 1
1 h i
t
¼ #sw;SpH;PWM #c1 þ ð#c;Min #c1 Þ e s1 dt
ð12Þ Rt t¼0
Dt1;H s1
When following a similar procedure for the cooling mode, a cor- ¼ ð#sw;SpH;PWM #c1 Þ ð#c1 #c;Min Þ
Rt Rt
responding result can be derived. There a maximum core temper- Dt
s1;H
ature #c;Max can be calculated (see Fig. 4) in order to maintain a e 1 1 ð16Þ
room temperature which is equal or lower than the upper room
temperature set-point #r;SpC : In (16), the core temperature is taken from (4). For
~ Dt1;H =s1 << 1, the approximation eDt1;H =s1 1 Dt1;H =s1 can be
R ~ q_ g
#c;Max ¼ #r;SpC þ ð#r;SpC #oa Þ R ð13Þ used. By incorporating this approximation in (16) EPWM is derived
Rl;f as
By treating a quasi-stationary operation (constant switched-on
Dt1;H
phase Dt1 and constant switched-off phase Dt0, constant #oa , con- EPWM ¼ ð#sw;SpH;PWM #c;Min Þ ð17Þ
Rt
stant #sw ¼ #sw;SpC;PWM and constant q_ g ¼ q_ g;eub ), we obtain a result
similar to (12) for the cooling mode: As will be shown later this leads to a much simpler PWM con-
~ þ Rl;f trol solution.
Rt þ R
#sw;SpC;PWM ¼ When postulating that in both operating scenarios – PWM and
~
R þ Rl;f
2 3 continuous operation – the same amount of heating energy has to
Dt Dt Dt
0 1 0
#c;Max 1 e s0 e s1 #c0 1 e s0 be delivered to the thermally activated structure (EPWM = Econt), we
6 R 7
6
t
# 7 obtain:
4 Dt
s0
Dt
s1 R þ ~ þ Rl;f c0 5
R
t
e 0 1e 1
Dt1;H #sw;SpH #c;Min
¼ ð18Þ
ð14Þ Dt H #sw;SpH;PWM #c;Min
The deduced Eqs. (12) and (14) are perfectly valid only for ideal By using (15), an expression for #c;Min can be found:
quasi-stationary operation with constant disturbances. Since in
these equations, equivalent heat gain bounds are used that are con- Rt
#c;Min ¼ #sw;SpH ð#sw;SpH #r;SpH Þ ð19Þ
stant for at least 24 h (see [5]), the equations can also be used for ~
Rt þ R
M. Gwerder et al. / Applied Energy 86 (2009) 1606–1616 1611
Finally, by combining (18) and (19), we receive a PWM control sired (see also (26)). A similar result to (20) can be found for the
solution for the ratio between the switched-on phase Dt1,H and the cooling mode:
period DtH for given lower room temperature set-point #r;SpH , con-
Dt1;C 1
tinuous operation supply water temperature set-point #sw;SpH and ¼ ð21Þ
Dt C #sw;SpC #sw;SpC;PWM
(to be applied) PWM supply water temperature set-point 1 þ Rt
#sw;SpH;PWM : R þR
t
~ ð#r;SpC #sw;SpC Þ
Fig. 6. Comparison of pump switched-on phases for PWM control solutions based on 1st order model calculations (solid lines) and based on simplified calculations (dotted
lines).
1612 M. Gwerder et al. / Applied Energy 86 (2009) 1606–1616
with Eq. (16), we can derive such a solution based on the assump-
tion that the core temperature (#c ) devolution is equal to the case
with constant supply water temperature #sw :
R Dt 1
Dt1 Dt
1
t¼0
#sw ðtÞdt R Rþt R~ ð#sw;Sp #r;Sp Þ
t
¼ ð23Þ
Dt #sw;Sp R Rþt R~ ð#sw;Sp #r;Sp Þ
t
Eq. (23) can be applied in the heating, cooling and the undefined
mode. In reality, the assumptions made will not be met perfectly,
but – as simulation studies show – (23) can be applied in all typical
cases.
If #sw is outside the set-point range [#sw;SpH ; #sw;SpC ], the purge Table 2
phase is concluded and active heating or cooling is necessary Characteristics of laboratory test room and incorporated TABS.
5.1. Laboratory test room data 5.2. Applied heat gains and control parameters
The HVAC-Laboratory test room used is specifically designed to In the laboratory test room, heat gains were brought in exclu-
reproduce the thermal behavior of offices in modern commercial sively by heat dissipation devices (cooling load simulators, see
Fig. 8. HVAC-Laboratory test room equipped with TABS: external view from the laboratory hall (left) and view from the weather zone into the room (right).
1614 M. Gwerder et al. / Applied Energy 86 (2009) 1606–1616
Fig. 9. Heating/cooling curves and heating/cooling limits for test scenarios workdays (left) and weekend (right); here #oa represents the mean weather zone temperature of
the last 24 h).
Fig. 8 right). A repeating weekly sequence of heat gains was applied the defined comfort criteria (comfort band) can be met indepen-
to reflect the daily heat gain inputs in a typical office room, differ- dently of the chosen PWM period. In addition, with PWM periods
entiating between 5 workdays (maximal gains 20–30 W/m2) and 2 of 24 h, it is possible to place switched-on phases to a given time
weekend days (maximal gains 3–5 W/m2), cf. Fig. 10. of day interval and benefit from higher energy efficiency or lower
For both the workday and the weekend situation, an integrated energy costs.
design procedure as introduced in [10,11] was carried out. This at Room temperatures shown in the result figures (Figs. 10 and 11)
first led to the equivalent heat gain bounds (see Table 1 for equiv- are measured operative temperatures in the middle of the room at
alent heat gains of workdays), and finally to the heating and cool- 1.2 m height.
ing curves and limits shown in Fig. 9. Fig. 10 shows test results for case 1. During the first 3 days and
the last day shown, large heat gain inputs were present from 8 am
5.3. Test results to 6 pm (workday situation); during day 4 and 5 there were only
low heat gains (weekend situation). At the beginning of each
In Table 3, conditions and parameter values for two test cases PWM switched-on phase, a purge phase was initiated. The initial
are given. In the first test case, we considered a PWM in cooling purge phase was succeeded by a cooling phase (see Fig. 10, pump
mode with a PWM period of 24 h. The beginning of each axis, marked by pale areas) in the workday situation or was pro-
switched-on phase was set to be at 10 pm. Such a setting is energy longed in the weekend situation. Note that when the pump is
efficient when outside air temperatures are significantly lower switched off, supply and return temperature measurements are
during the night and cooling is provided by a cooling tower. In not meaningful for control.
the second test case, a PWM in heating mode with a PWM period Fig. 11 shows test results for case 2. During the first 2 days and
of 5 h was considered. the last 2 days shown, large heat gain inputs were present from
The choice of the PWM period is a trade-off between slightly 8 am to 6 pm (workday situation); during day 3 and 4 there were
worse room temperature control for long switched-off phases only low heat gains (weekend situation). At the beginning of each
(occurring when using long PWM periods) and a higher number PWM switched-on phase, a purge phase was initiated. Here, the
of (unproductive) purge phases with short PWM periods. For typi- initial purge phase was prolonged for all but six switched-on
cal concrete core-activated structures and PWM periods up to 24 h, phases in which heating activities were necessary (see Fig. 11,
resulting room temperature amplitudes from PWM are small and pump axis, marked by pale areas).
Fig. 10. PWM control test results for cooling mode with DtC = 24 h (case 1).
Fig. 11. PWM control test results for heating mode with DtH = 5 h (case 2).
1616 M. Gwerder et al. / Applied Energy 86 (2009) 1606–1616
switched-on phases can be shifted to times with high energy gen- References
eration efficiency (e.g. benefit from colder night time outdoor tem-
peratures). The quantification of energy reduction potentials using [1] Koschenz M, Dorer V. Interaction of an air system with concrete core
conditioning. Energy Build 1999;30(2):139–45.
PWM control is a topic of ongoing and future investigations. [2] Lehmann B, Dorer V, Koschenz M. Application range of thermally activated
Two PWM control solutions are outlined and compared in this building systems TABS. Energy Build 2007;39(5):593–8.
paper, one based on a 1st order thermal model of a room equipped [3] Koschenz M, Lehmann B. Thermoaktive Bauteilsysteme TABS. Duebendorf
(Switzerland): Empa; 2000. ISBN 3-905594-19-6 [in German].
with TABS, and a simplified PWM control solution. Simulation [4] Güntensperger W, Gwerder M, Haas A, Lehmann B, Renggli F, Tödtli J. Control
studies [12] show that both solutions are sufficiently accurate for of concrete core conditioning systems. In: Proceedings: 8th REHVA world
application in building control, the simplified solution having the congress Clima 2005, Lausanne (Switzerland).
[5] Gwerder M, Lehman B, Tödtli J, Dorer V, Renggli F. Control of thermally
advantage of only one control parameter to tune. That simplified
activated building systems TABS. Appl Energy 2008;85(7):565–81.
solution was additionally evaluated in tests performed at the Sie- [6] Meierhans R. Room air conditioning by means of overnight cooling of the
mens HVAC-Laboratory in Zug, Switzerland. For two of the test concrete ceiling. ASHRAE Trans 1996;102(1):693–7.
cases carried out, results are shown and analyzed in the paper. In [7] Olesen BW. Radiant floor heating in theory and practice. ASHRAE J
2002:19–26.
the laboratory test set-up, it has been confirmed that the simplified [8] Antonopoulos KA, Vrachopoulos M, Tzivanidis C. Experimental and theoretical
PWM control solution successfully works for all assessed test cases. studies of space cooling using ceiling-embedded piping. Appl Therm Eng
The PWM control module has proved to fit well into the newly 1997;17(4):351–67.
[9] Weitzmann P. Modelling building integrated heating and cooling systems. PhD
developed TABS control solution. Furthermore, the module is easy thesis. Danish Techn University. Rapport BYGDTU R-091 2004; ISBN 87-7877-
to engineer and commission which qualifies it for practical 155-2; 2004.
application. [10] Gwerder M, Tödtli J, Lehman B, Renggli F, Dorer V. Control of thermally
activated building systems. In: Proceedings: 9th REHVA world congress Clima
Further details of the PWM method and the entire newly 2007, Helsinki (Finland).
developed TABS control solution can be found in the design [11] Tödtli J, Gwerder M, Lehman B, Renggli F, Dorer V. Integrated design of
and commissioning handbook, published in the frame of our thermally activated building systems and of their control. In: Proceedings: 9th
REHVA world congress Clima 2007, Helsinki (Finland).
work [12]. [12] Tödtli J, Gwerder M, Lehmann B, Renggli F, Dorer V. TABS-Control: Steuerung
und Regelung von thermoaktiven Bauteilsystemen. Faktor Verlag Zurich
Acknowledgements (Switzerland); 2009. ISBN: 978-3-905711-05-9 [in German].
[13] Standard EN 15377-1. Heating systems in buildings – design of embedded
water based surface heating and cooling systems – part 1: determination of
The partial funding of the project by the Swiss Confederation’s the design heating and cooling capacity; 2008.
innovation promotion agency CTI and the input of the project [14] Tödtli J. Influence of thermal inertia on heating controller adaptation: an
steering committee are gratefully acknowledged. analytical investigation. In: Proceedings: 3rd REHVA world congress Clima
2000 (1993), London (England).