Notes on side-by-side radiant heating study

ESP-r provides facilities to define block shaped radiant sensors
within a room and to calculate how much that sensor views the other
surfaces in the room. This information is used in the results
analysis module to generate radiant asymmetry values.

To demonstrate the surface view-factor calculations and the definition
of radiant sensors lets look at a simulation model for a private
room in a hospital.  Two competing designs of radiant heating pannels
in room were under consideration. One design was claimed to be less
costly to install (rectangular unit at the facade) and the other
design was claimed to provide better comfort for the doctor and
patient (radiant heat better distributed in the room and a lower
temperature required).

The model was designed as a classic side-by-side virtual experiment.
The resolution of the model was dictated by the thermophysical
characteristics of the entities within the rooms and possible
interactions. The design was also dictated by the need to complete
the model while the client was available to supply details and so
the \fIextent\fR of the initial model was carefully considered.

The critical performance metrics were the comfort for a tall doctor
standing at the window and for the patient in the bed as well as
the frequency the pannels were on and how well they were able to
control temperatures within the room on moderately cold days.

In both rooms the bed was represented as internal mass, the radiant
pannels were represented via a thin-zone approach and the ceiling
plenum above and below the rooms were represented as rooms. Diversity
was included in the schedules for each zone and in particular the
heat output of clinical lighting during the doctor's visit was
included.

The adjacent wards were assumed to be at the same temperature as
the rooms being investigated.  The temperature of the upper ceiling
void is critical because heat leakage from the pannel was likely
to result in a warmer void. This approach would allow changes in
the thickness of upper insulation to be studied. The lower ceiling
void was controlled to the same temperature as the upper ceiling
void in order to form a representative boundary condition in the
patients room.

The thin-zone approach treats a heat generation device explicitly
as a zone rather than a system component. Briefly, a zone is created
which represents the shape of the radiant pannel and the lower
surface is metal and the upper surface is an insulated metal pannel.
High heat transfer coefficients are set within the thin zone so
that any heat injected into the air will be transferred to the
surfaces.   This approach was used because setting up system
components would have required additional time and the potentially
simplier injection into a ceiling construction could not be used
because it was located between two zones (a rather irritating and
persistent limitation).  For purposes of this assessment the approach
was seen to represent the expected pannel temperatures and response
time while the explicit shape of the pannel surfaces worked well
in the subsequent detailed comfort assessments.

The left patient room includes an "L" shaped heating pannel 400mm
wide and the right patient room uses a rectangular pannel 600mm
wide at the facade. The pannels were controlled via a multi-sensor
controller which regulated the room temperature to 21\(de C based
on allowing the thin-zone to rise to 74\(de C.  To reflect the
response time of the heating circuit the simulation timestep was
set to one minute. After composing the model calibration assessments
were run to tune the heat injection so that it matched the expected
pannel surface temperatures.

Surface long-wave view-factors were computed for the patient rooms.
It was not considered necessary to compute view-factors within the
radiant pannel zones (the high heat transfer at the surfaces would
tend to limit temperature differences).

Overall the model was up and running during the visit to the clients
offices and the predictions and fine tuning were carried out with
feedback from the client.  Assessments indicated that both designs
resulted in substantially similar comfort levels for the doctor as
well as the patient. There was a slightly increased risk of radiant
assymmetry discomfort for the case of the 600mm wide pannel design.
It was also clear, however that the 600mm wide pannel was ON more
often and tended to work at a higher temperature than the 400mm
wide pannel approach.
