CHILLED BEAMS IN HEATING: DESIGN CRITERIA AND CASE STUDY

Maija Virta1 and Risto Kosonen1

Halton Oy, Haltonintie 1-3, 47400 Kausala, Finland

e-mail: maija.virta@halton.com

Keywords: chilled beams, ventilation, heating, thermal comfort

ABSTRACT

The chilled beam system is suitable for spaces with a high cooling requirement, low
humidity load and relatively small ventilation requirement. Although the chilled beam
system is quite popular and well known, there is still limited experience in using active
chilled beams for primary heating. The main objective of this paper is to give design
criteria for active chilled beam system and present results of a case study, where active
chilled beams were used for heating. The study is carried out in laboratory conditions.
The results show that the occupied zone conditions are good and the temperature
gradient is low.

BACKGROUND
The design of the heating system begins by defining the required heating capacity. In traditional
heating systems the design is often based on high safety margins when heat losses are calculated. When a
chilled beam system is used for heating, proper system operation cannot be achieved by oversizing the
heating system. In a new office building 30 –45 W/m2,floor of heating capacity is typically enough.

The required heating capacity is strongly dependent on the thermal properties of the building and the
outdoor climate conditions. Figure 1 shows the required heating capacity as a function of the average
thermal conductance and the outdoor temperature. In this case, the room dimensions are 2.8 m (room
width), 4.5 m (room length) and 3.0 m (room height), giving a the floor area of 12.6 m2 and a total
external wall area of 8.4 m2. In this case, the glazing ratio of the total outdoor wall area is fixed to 50 %.

In the European case, there is a double-glazed window (U-value 3 W/m2K) and the U-value of the wall
is 0.45 W/m2K. In the Scandinavian case, there is a triple-glazed window (U-value 2 W/m2K) and the U-
value of the wall is 0.28 W/m2K. Figure 1 also shows an example, which incurs high heat losses: this case
is like a European one, except for having a single-glazed window.
 In addition, a case representing low heating demand is shown. Here, the heat losses are simply fixed 50
% lower than the Scandinavian case. Based on this calculation, the required heating capacity of 500 W
per room-module is enough for the normal case in various European climate and design conditions.

900 High losses G=27.1 W/m2K

800 European G=14.5 W/m2K

700
Scandinavia G=9.6 W/m2K
Heating demand (W) 600
Low energy G=4.8 W/m2K
500

400

300

200

100

0
-30 -20 -10 0 10 20
Outdoor temperature oC

Figure 1. The required heating capacity as a function of thermal conductance and external temperature for a case
room-module (room area 12.6 m2 and external wall area 8.4 m2).

The basic guideline for thermal comfort is laid down in the ISO 7730 Standard (1990). The major part
of this report concerns thermal comfort in terms of the temperature gradient in the room space and the
room air velocities. ISO 7330 recommends the following:
• Air velocity (to avoid draughts):
• < 0.15 m/s (winter)
• < 0.25 m/s (summer)
• Vertical air temperature difference:
• < 3°C from foot to head when sitting, 0.1 m to 1.1 m, or standing, 0.1 m to 1.7 m.

With the traditional heating system, the maximum gradient is typically 3 – 5 °C. Of course, the
existing conditions considerably affects a lot for the gradient, e.g. heat losses, positioning of the heating
device, surface temperatures, and the design temperature levels of system. All in all, the installation has a
consequential effect on the gradient: it is more difficult to keep the gradient low when the installation
height of the heat emitter is high.

HEATING DESIGN CRITERIA WITH ACTIVE CHILLED BEAMS

The active chilled beam is one of the most popular air-conditioning systems in Europe. With chilled
beams, it is possible to create high-quality indoor climate conditions, including thermal comfort and a low
noise level within reasonable life-cycle costs. Chilled beam systems are dedicated outdoor air systems to
be applied primarily in spaces where internal humidity loads are moderate. They can also be used for
heating.

Active chilled beams are connected to both the ventilation supply air ductwork, and the chilled water
system. When desired, hot water can be used in this system for heating. The main air-handling unit
supplies primary air into the various rooms through the chilled beam. Primary air supply induces room air
to be recirculated through the heat exchanger of the chilled beam. In order to cool or heat the room either
cold (14-18°C) or warm (30-45°C) water is cycled through the heat exchanger. Recirculated room air and
the primary air are mixed prior to diffusion in the space. Room temperature is controlled by the water
flow rate through the heat exchanger.
Figure 2. Typical exposed installation of active chilled beams in office environment.

If the heating inlet water temperature of a chilled beam is higher than 45°C (linear output of an active
beam is higher than a. 150 W/m) in a typical installation, secondary air is often too warm to mix properly
with the room air. The relatively low temperature gradient in the space raises the air temperature near the
floor thus maintaining comfortable thermal conditions, as well as ensuring the energy efficiency of the
system (by decreasing the short circuit and thus the exhaust air temperature).

Figure 3. Typical example of an active chilled beam system operation principle in heating mode.

This has been confirmed in previous papers based on laboratory measurements, where both the heating
capacity and the temperature gradient of the room were analyzed. The studied inlet water temperature
levels were 40 °C, 60 °C and 70 °C.
 3

2,5
Room height (m) 40 deg.C
2 60 deg.C
74 deg.C
1,5

0,5

0
0 2 4 6 8 10
Temperature difference dT (C)

Figure 4. Temperature gradient in the space with different inlet water temperatures.

The data in Fig. 4 shows the effect of the inlet water temperature level on the temperature gradient.
With relatively high temperatures, the gradient in the room rises: the temperature level of 74 °C leads to
far too high gradient 8 – 10 °C in the room. 40 °C inlet water temperature gives much lower gradient in
the room of under 3 °C. Increasing the supply airflow rate decreases the temperature gradient but in this
measured case, using 22 l/s instead of 13 l/s reduced the gradient only about 1 °C. Therefore, the key
factor is to control the inlet water temperature at a reasonably low level. With the temperature level of 40
°C, it is possible to reach 500 W, which is in most cases enough.

The mixing is also dependent on the window size and surface temperature. The higher and colder the
window, the colder the air falling down to the floor, and the temperature gradient between secondary air
and room air becomes higher. For this reason using beams for heating is recommended when the heat
transmission of the windows is moderate (e.g. surface temperature is higher than 14°C and height is not
more than 1.5 m).

The heating capacity of active beams is dependent on the primary airflow rate. This is why ventilation
must be operating when heating is required. When an office room is occupied the internal heat sources
normally reduce the required heating output and the temperature gradient stays at an acceptable level.

The control principle in heating case is very important. It can be on-off, time proportional on-off or
modulating. The selection of the control system is dependent on the system design. In most cases all the
above mentioned control principles provide reliable operation of the system. In cases where the difference
between the design and normal operating conditions is large (e.g. the design is based on a much higher
heating capacity level than the real level), time proportional on−off or modulating controls are
recommended. Otherwise the mixed air temperature of chilled beam becomes too quickly too warm, and
there is no sufficient mixing in the space and therefore the occupied zone becomes too cold.

In some cases the operation of heating system can be improved by installing the room air temperature
measurement at higher level in the space. In that case the heating mode is controlled based on higher
room air temperature in heating mode and in cooling mode with fully mixed situation, the positioning of
temperature sensor is irrelevant.
 CASE STUDY OF EXPOSED CHILLED BEAMS IN HEATING

The main idea of the conducted case study measurement was to study a chilled beam system in heating
mode with correctly calculated heat losses (35 W/m2) without overdesign and relatively low inlet water
temperature (35 OC). The office room is also relatively high (3.6m) and chilled beams are installed at 800
mm under the soffit. This measurement was so-called “early morning” case without any internal loads.

The measurements were carried out in a laboratory measurement room with room length of 6.5 m,
width of 4.4 m and height of 3.6 m. There was cold window (surface temperature of 15 OC and height
2m) in the room. This project was designed at London weather conditions with double-glazed window.
Two open types of chilled beams with active length of 4.6 m and width of 0.3 m were installed exposed at
height of 2.8 m. The beam installation can be seen in Fig. 5.

Figure 5. The measurement room sizes, measurement grid and chilled beam location during the measurement.
 The comfort conditions were measured in 25 different points throughout the space at 6 different
heights (0.1m, 0.2m, 0.6m, 1.1m, 1.5m and 1.8m). The air velocity, air temperature and turbulence
intensity was measured in each measurement point and from these results the draught rate was calculated
using following formula:

DR = (34 − Ta )(v − 0,05) 0,62 (0,37 * v * Turb + 3,14) (1)

Where:

DR = draught rate (%)

Ta = air temperature (OC)
v = air velocity (m/s)
Turb = turbulence intensity (%)

Table 1. gives all the input data values. The design inlet water temperature was 35 OC and with 0.04
kg/s water flow rate, the chilled beam heating output was 1425 W. The primary air was slightly colder
than room air in order to generate free cooling in the space when all internal heat loads are effective and
therefore cooling is needed instead of heating. At room air temperature of 21.3 OC the 18 OC primary air
has a cooling effect of 487 W. Therefore the net heating capacity to compensate heat loads was 937 W,
corresponding heating capacity per floor area of 33 W/m2.
Table 1. Input values of a chilled beam case study.

o
TR Room Temperature 21,3 C
qvS air Supply Air Flow Rate 116,0 l/s
qvS air/m Supply Air Flow Rate /m 25,8 l/s/m
qvE air Exhaust Air Flow Rate 116,0 l/s
o
TS Supply Air Temperature 17,8 C
o
TE Exhaust Air Temperature 21,7 C
o
TW Window Surface Temperature 14,6 C
o
TW1 Inlet Water Temperature 35,0 C
o
TW2 Outlet Water Temperature 26,5 C
qmW Total water Flow Rate 0,040 kg/s
PW Total capacity of water 1425 W
PA Capacity of air -487 W
PT Total Capacity 937 W
T R -average (T W1 andT W2 ) o
dT -9,45 C

The chilled beam heating operates fine also in this kind of installation. The achieved temperature
gradient is low enough and velocities are very low. Results of comfort measurement are shown in Table
2. In all measured points the air velocity is under 0.18 m/s and majority of them is under 0.15 m/s.
Calculated draught rates in all points are under 15% (except 3 points at 1.8m height).

Temperature gradient in the occupied zone is less than 2 OC and is below the comfort criteria of ISO
7730 Standard (<3OC between 0.1m and 1.7m). This means that the airflow from chilled beam is well
mixed with room air in the whole occupied zone and therefore the ventilation efficiency is good. In
addition it also shows relatively good energy efficiency of system with exhaust air temperature of 21.7OC
even the space was 800 mm higher than the installation height of chilled beams.

With internal loads (lights, persons and computers all together 55 W/m2), the room air temperature
increases a. 2OC if water flow rate is not reduced. This means that even the maximum heating output is
defined with room air temperature of 21OC, in practice the personal control of +2OC can be achieved.
Table 2. Comfort measurement results and calculated draught rates in various measurement points.

1 2 3 4 5
Heigh v Ta Turb. DR v Ta Turb. DR v Ta Turb. DR v Ta Turb. DR v Ta Turb. DR
o o o o o
t (m) (m/s) ( C) (%) % (m/s) ( C) (%) % (m/s) ( C) (%) % (m/s) ( C) (%) % (m/s) ( C) (%) %
1,80 0,05 21,4 55 - 0,05 21,4 51 - 0,03 21,6 56 - 0,04 21,6 45 - 0,03 21,6 42 -
1,50 0,07 21,3 41 5 0,05 21,3 47 - 0,04 21,4 47 - 0,04 21,4 35 - 0,03 21,4 44 -
1,10 0,06 21,1 48 3 0,04 21,1 39 - 0,04 21,1 38 - 0,03 21,1 39 - 0,02 21,2 38 -
0,60 0,03 20,6 42 - 0,03 20,6 37 - 0,03 20,6 31 - 0,04 20,6 48 - 0,03 20,7 35 -
0,20 0,02 20,3 35 - 0,02 20,4 40 - 0,03 20,5 34 - 0,03 20,4 53 - 0,02 20,5 44 -
0,10 0,02 20,1 25 - 0,03 20,3 27 - 0,05 20,3 36 - 0,04 20,3 36 - 0,03 20,3 34 -

6 7 8 9 10
Heigh v Ta Turb. DR v Ta Turb. DR v Ta Turb. DR v Ta Turb. DR v Ta Turb. DR
o o o o o
t (m) (m/s) ( C) (%) % (m/s) ( C) (%) % (m/s) ( C) (%) % (m/s) ( C) (%) % (m/s) ( C) (%) %
1,80 0,09 21,7 49 8 0,13 21,8 44 13 0,18 22,0 44 21 0,09 21,8 46 8 0,07 21,8 49 5
1,50 0,10 21,6 48 10 0,09 21,6 40 8 0,14 21,9 50 16 0,10 21,6 49 10 0,04 21,6 47 -
1,10 0,10 21,4 44 9 0,06 21,3 40 3 0,06 21,5 75 3 0,06 21,2 70 3 0,02 21,2 40 -
0,60 0,06 20,8 64 3 0,04 20,4 35 - 0,04 20,5 36 - 0,03 20,5 34 - 0,05 20,6 34 -
0,20 0,03 20,4 59 - 0,02 20,3 44 - 0,02 20,4 32 - 0,02 20,4 43 - 0,03 20,4 33 -
0,10 0,03 20,2 36 - 0,02 20,2 27 - 0,03 20,3 34 - 0,03 20,3 32 - 0,03 20,3 33 -

11 12 13 14 15
Heigh v Ta Turb. DR v Ta Turb. DR v Ta Turb. DR v Ta Turb. DR v Ta Turb. DR
o o o o o
t (m) (m/s) ( C) (%) % (m/s) ( C) (%) % (m/s) ( C) (%) % (m/s) ( C) (%) % (m/s) ( C) (%) %
1,80 0,07 21,8 37 4 0,12 21,9 31 11 0,15 22,1 52 17 0,10 21,9 47 9 0,05 21,8 51 -
1,50 0,07 21,6 42 5 0,08 21,7 43 6 0,11 21,8 47 11 0,07 21,6 52 5 0,05 21,6 40 -
1,10 0,08 21,3 43 6 0,05 21,3 54 - 0,06 21,2 53 3 0,03 21,1 38 - 0,03 21,3 54 -
0,60 0,04 20,7 57 - 0,03 20,6 42 - 0,03 20,5 34 - 0,03 20,5 36 - 0,04 20,6 50 -
0,20 0,02 20,4 47 - 0,02 20,5 41 - 0,02 20,4 34 - 0,02 20,4 31 - 0,03 20,4 52 -
0,10 0,03 20,2 34 - 0,03 20,3 28 - 0,02 20,3 24 - 0,02 20,3 23 - 0,03 20,3 51 -

16 17 18 19 20
Heigh v Ta Turb. DR v Ta Turb. DR v Ta Turb. DR v Ta Turb. DR v Ta Turb. DR
o o o o o
t (m) (m/s) ( C) (%) % (m/s) ( C) (%) % (m/s) ( C) (%) % (m/s) ( C) (%) % (m/s) ( C) (%) %
1,80 0,09 21,8 50 8 0,10 21,9 46 9 0,13 22,0 59 15 0,08 21,6 58 7 0,05 21,6 35 -
1,50 0,10 21,7 43 9 0,08 21,6 46 6 0,08 21,7 63 7 0,04 21,3 54 - 0,05 21,4 36 -
1,10 0,10 21,5 38 9 0,04 21,1 46 - 0,04 21,1 47 - 0,03 20,9 30 - 0,05 21,0 22 -
0,60 0,04 20,6 56 - 0,02 20,5 41 - 0,03 20,4 40 - 0,03 20,3 41 - 0,04 20,3 37 -
0,20 0,02 20,3 34 - 0,02 20,3 28 - 0,03 20,3 36 - 0,03 20,3 48 - 0,06 20,3 31 3
0,10 0,02 20,1 24 - 0,03 20,2 29 - 0,03 20,1 30 - 0,03 20,2 44 - 0,05 20,2 44 -

21 22 23 24 25
Heigh v Ta Turb. DR v Ta Turb. DR v Ta Turb. DR v Ta Turb. DR v Ta Turb. DR
o o o o o
t (m) (m/s) ( C) (%) % (m/s) ( C) (%) % (m/s) ( C) (%) % (m/s) ( C) (%) % (m/s) ( C) (%) %
1,80 0,06 21,4 46 3 0,06 21,4 49 3 0,05 21,3 46 - 0,03 21,4 42 - 0,04 21,4 52 -
1,50 0,08 21,3 43 6 0,06 21,2 52 3 0,04 21,1 43 - 0,04 21,2 38 - 0,04 21,1 47 -
1,10 0,08 21,1 45 7 0,05 20,9 66 - 0,04 20,8 41 - 0,04 20,8 40 - 0,04 20,9 34 -
0,60 0,04 20,5 53 - 0,03 20,5 42 - 0,03 20,5 49 - 0,03 20,4 46 - 0,03 20,5 40 -
0,20 0,04 20,4 41 - 0,05 20,5 45 - 0,06 20,5 38 3 0,08 20,4 30 6 0,08 20,4 36 7
0,10 0,04 20,2 52 - 0,10 20,3 30 9 0,11 20,3 25 10 0,10 20,3 25 9 0,12 20,3 27 11

Temperature gradients with and without internal loads

with max heating capacity

2,00
3
8
1,50 13
18
Height (m)

23
1,00 3 with internal loads
8 with internal loads
13 with internal loads
0,50
18 with internal loads
23 with internal loads
0,00
20,0 21,0 22,0 23,0 24,0 25,0
Temperature

Figure 6. Temperature gradients are lower than 2 OC throughout the whole floor area. Measurement point 3 is
nearest from the window and 23 is close to the door. All the points are in the middle of the room halfway between
the chilled beams. With internal loads the room air temperature is a. 2 OC warmer, if water flow rate is not reduced.
 CONCLUSIONS
Integration of the heating mode in the active chilled beam system is a challenging task. The main
problem has been to control the temperature gradient in the space and to creating thermal comfort
conditions in an energy-efficient way. On the other hand, meeting the heating capacity requirement with
chilled beam is not a problem in most cases.

The key factor is the temperature level of the inlet water. Lower the inlet water temperature is, lower is
also the temperature gradient in the space. Based on the measurements, a temperature level of 40 °C or
below, gives a reasonable gradient for the room space. For design practice, this also means the possibility
to use low temperature level heating sources.

During working hours, there is normally enough or even too much of a heat loads to keep indoor
temperature within acceptable comfort levels. This means that this worst case exists in the beginning of
the working day when the external temperature is also near design conditions.

In the case study of exposed chilled beams in heating mode, the right design values (inlet water
temperature of 35 OC) with relatively low heating output (33 W/m2,floor) demonstrated excellent thermal
comfort and good energy efficiency. It was also showing, that earlier given design guidelines are a good
basis for design of active chilled beams in heating.

LITERATURE

Kosonen, R., Horttanainen, P, Dunlop, G. Integration of heating mode into ventilated cooled beam.
Proceedings of Roomvent 2000.

Virta, M., Butler, D., Gräslund, J., Hogeling, J., Kristiansen, E. Reinikainen, M., Svensson, G. Chilled
Beam Application Guidebook, Rehva Guidebook n:o 5.

International Standard ISO 7730, Moderate thermal environments- Determination of the PMV and PPD
indices and specification of the conditions for thermal comfort, 1990.