1

This paper was published in Journal of Cultural Her itage, vol. 11, issue 2

(2010), 205 – 219.

Original publication is avaible at: doi:10.1016/j.culher.2009.02.008

An advanced church heating system favourable to art works:

a contribution to European standardisation

Dario Camuffoa*, Emanuela Pagana, Sirkka Rissanenb, Łukasz Brataszc,

Roman Kozłowskic, Marco Camuffod, Antonio della Vallea

a National Research Council, Institute of Atmospheric Sciences and Climate,

Corso Stati Uniti 4, 35127 Padova, Italy b Finnish Institute of Occupational Health, Aapistie 1, FIN-90220 Oulu, Finland

c Institute of Catalysis and Surface Chemistry, Polish Academy of Sciences,

ul. Niezapominajek 8, 30-239 Kraków, Poland. d BOCA Associate Architects Studio, via S. Mattia 12, 35121, Padova, Italy

*Corresponding author. Tel: +39-049-8295902, E-mail address: d.camuffo@isac.cnr.it

Keywords: Church heating; Preservation of cultural property; Warm-air heating;

Pew heating; Heating carpets; Localized heating; Thermal comfort; European standardization

2

Abstract

The European project FRIENDLY-HEATING: comfortable to people and compatible with

conservation of artworks preserved in churches addressed the problems caused by the

continuous or intermittent heating of historic churches, which disturbs the microclimatic

conditions to which the building and the artworks preserved inside have acclimatised. As

thermal comfort and the preservation of artworks often conflict with each other, a balance

between the two needs is necessary. The proposed heating strategy is to provide a small

amount of heat directly to people in the pew area while leaving the conditions in the church,

as a whole, undisturbed. This novel heating system is based on some low-temperature radiant

emitters mounted in a pew to provide a desirable distribution of heat to the feet, legs and

hands of people occupying it. Due to little heat dispersion, this novel system not only

significantly reduces the risk of mechanical stress in wooden artworks and panel or canvas

paintings, fresco soiling and cyclic dissolution-recrystallization of soluble salts in the

masonry, but is energy-efficient. The detailed environmental monitoring was conducted in the

church of Santa Maria Maddalena in Rocca Pietore, Italy over a 3-year period to verify the

performance of the novel heating system in comparison to the warm-air system that was

active earlier in the church. The methodology and results of this comprehensive and

multidisciplinary study were included in three draft standards of the European Committee for

Standardisation intended for use in the study and control of environments of cultural heritage

objects.

3

1. Introduction

A great deal of artistic wealth has survived for centuries in unheated buildings, if only

dampness and related mould growth were avoided. Thick walls are a typical feature of many

historic buildings, and help to smooth out the daily cycles of air temperature (TA) and relative

humidity (RH) and attenuate the seasonal ones, creating a natural microclimate favourable for

the preservation of many artworks. The natural microclimate can be significantly altered after

heating is introduced, especially in the case of churches, due to the necessity to raise the

temperature to an acceptable level in a short time and at low cost. Different heating systems

and regimes are used, and various aspects of church heating have been extensively reviewed

[1-5]. In many cases, the comfort requirements have had dramatic consequences on artworks

preserved in churches, mostly managed by people who fail to grasp the implications of

heating on the preservation of artworks.

This paper will concentrate on the everyday problems generated by heating for the thermal

comfort of the congregation, i.e. typical church heating operated a few times a week during

services or, on less frequent occasions, operated continually during the cold season in

churches used on a daily basis. The paper will deliberately leave out the exceptional case of

conservation heating, operated to stabilise RH on an acceptable level throughout the whole

year by adding or withholding heat. Conservation heating is especially efficient in preventing

dampness in buildings during the cold, rainy seasons.

Continuous heating is sometimes considered a favourable strategy which avoids cyclical

changes in temperature and humidity. In cold climates, however, even modest heating can

cause excessive RH drops due to a very limited amount of water vapour contained in the cold

air. For instance, when the external TA is -10 °C and the internal TA is +10 °C, the indoor RH

drops to extremely low levels of 10-20% for outdoor RH of 50-100%, respectively. Low RH

levels are damaging to artworks. For example, they can cause the excessive shrinkage of

wood leading to its irreversible stretching and eventual cracking. Rapid heating for one or

two hours is sometimes claimed to be a good strategy as it is considered too short to damage

artworks. However, sharp variations in T cause sharp variations in RH and both can be

dangerous. Therefore, clarification of safe heating strategies, based on sound science, is

needed.

4

The European Commission funded project FRIENDLY-HEATING (FH), implemented between

2002 – 2005, started with analysing the main risks for preservation, pointing out the pros and

cons for each heating methodology and using many case studies [1]. Then research was

undertaken to develop a heating system which would provide direct confined heat just to

people sitting without dispersing too much heat in the church as a whole reducing thus the

variability in T and RH in the vicinity of artworks.

The design and assessment of the novel system developed were based on extensive laboratory

tests, numerical simulations using Computational Fluid Dynamics (CFD) [6] and direct

measurements in two churches. Wooden artwork response [7-9] was monitored as well as the

origin, distribution and transport of atmospheric pollutants [10-13]. This paper reports the

results of comprehensive field monitoring which was carried out during a three-year period

and concerned microclimate, heat diffusion and control, internal air motions and building

response. Also, the thermal comfort of churchgoers was monitored in a number of

independent ways in order to establish an objective methodology and assess a compromise

between artwork preservation and thermal comfort. The final verification included a thorough

control of the above environmental variables in the area where people are and around

artworks, as well as the artwork response

2. Churches selected for the study

Two churches with very cold indoor climate were selected in the Italian Alps to rigorously

test the compatibility between preservation needs and thermal comfort for the novel heating

system.

The first church was Santa Maria Maddalena in Rocca Pietore, situated at 1,143 metres above

mean sea level in a small valley shielded from the winter sunshine, close to a glacier, with a

daily minimum temperature between –10 and –20 °C. The church was built in the 15th

century from local stone, and features one-metre-thick walls, and a nave with a chapel on

either side which were added in the nineteenth century. The building is relatively small and

measures 25 m long. The nave is 8 m wide and 9 m tall, while the two side chapels are square,

with each side measuring 4.5 m.

The church has various types of artworks, which was useful for the completeness of the

research programme. These included wooden altarpieces, paintings on canvas and panels,

choir stalls, a decorated organ-loft with modern organs, and frescoes from the fifteenth

5

century. The magnificent altarpiece sculptured and painted by Ruprecht Potsch, dated 1518, is

particularly valuable. It was cleaned and restored around 10 years ago. Originally, the

altarpiece was used as recommended in 1523 by A. Stoss:

“Usually, the altarpiece wings should be kept closed, and only opened for important events. It

should be cleaned twice a year. It should be softly lit to avoid any blackening from smoke:

two small candles of pure wax are enough; additional candles should be placed far away”

[14].

The closed wings kept the internal microclimate particularly constant and protected the

altarpiece against unnecessary smoke and dust deposition. The original use of soft candlelight

reduced the generation of smoke and radiant heat from the flame. Stoss thus suggested a

preservation-oriented management with limited regard to public enjoyment, and almost

permanent self-protection. At present, the use is oriented to the enhancement of the public’s

enjoyment by keeping the wings permanently open, which makes sculptures exposed to

environmental hazards and, consequently, increased risk to preservation, especially when the

polychrome wood was blasted with warm air blown by the previous heating system, or when

spot lamps light the altarpiece.

At the time of the last restoration, the church was provided with warm-air (WA) heating,

planned for occasional use, mainly once or twice a week, for around 100 minutes of

operation. Two grilles, one in the side chapel and the other in the nave (featuring blown-air

velocities of 2.7 and 0.4 m/s, respectively), supplied warm air (70 – 80 °C) inside. After the

WA heating was installed, some cracks increased in width, disfiguring the faces of the figures

of St. Mary Magdalene and St. Catherine in the central part of the altarpiece. In addition, the

warm air blown inside generated strong turbulence near the ceiling and the walls. The high air

speed, elevated TA and thermal gradients close to the walls were responsible for accelerating

the deposition of smoke and dust particles via aerodynamic capture, Brownian motions,

thermophoresis and electrophoresis. The surface blackening was worsened by downdraughts

of cold air which were formed on contact with cold surfaces. Around 10 years ago the

Superintendence to Artistic and Historical Cultural Heritage asked for accurate environmental

monitoring to understand the causes of the rapid deterioration of the artworks just restored.

The study identified the WA heating as the cause [15], but the congregation preferred to stay

with the inadequate heating system rather than to return to the unheated church. The solution

required a heating system that was at the same time comfortable to people and safe for

6

artwork preservation, and when the FH project was launched, the choice of this church was

natural, after it had been recognized as a very difficult problem needing a general solution.

The second case study was S. Stefano di Cadore, a church erected in the fourteenth century

also situated in the Italian Alps. The church used infrared emitters heated by gas combustion

and then a WA heating; however, none of these systems was able to create a thermally

comfortable environment.

Very careful monitoring was carried out in both churches to provide a holistic representation

of what was happening in natural conditions and when the existing WA heating and the novel

FH prototype were operating. The problems encountered in both churches were similar, as

well as the results. For the sake of brevity, this paper will concentrate on the church of Santa

Maria Maddalena in Rocca Pietore only.

3. The novel heating methodology of the EC Friendly-Heating project

Previously, the church had a heating system based on a central inflow of warm air, i.e.

focused on pre-heating the entire church space to provide a thermally comfortable

environment before people entered for a service, and then keeping this temperature level for

the necessary time. This was a difficult task, because the warm air remains buoyant in the

upper part of the church, and a huge amount of heat is needed before some benefit is brought

to people sitting in the pews. Air mixing by fans could have reduced the thermal layering but

would have increased noise.

The FH methodology changed the heating principle: it leaves the church in its natural, historic

microclimate and focuses on the localised heating of people in the pew area when they are in

the church. The adopted solution is, however, different from well-known pew heating

systems, generally based on only one high-temperature heating element placed below or

between the pews, or on the emission of mildly warm air from the floor or footboard [1,3].

Computer modelling, physical simulations, laboratory work and tests with the prototypes in

the church, conducted during the first year of the project, indicated that low-temperature

radiant heaters, operating at 40 to 70 °C in the far infrared (dark emitters) provide the best

solution to keep heat localized. The emitters used in the project comprised heating foils of

varying sizes and temperatures. The foils are made of an electrically heated layer of graphite

granules deposited on fibreglass and sealed between two plastic foils. When the electric

power is on, the graphite warms up, the resistance of the granules increases with T and

7

reduces the current intensity. Consequently, the maximum temperature is self-regulated

(Fig.1) at levels specifically selected for various parts of the body - in our case between 40

and 70 °C. This self-regulation provides a natural cut-out for the system and eliminates the

risk of ignition or burning skin. A thermostat is added for further fine regulation and safety.

The heating foils are enclosed in a metal frame, with a stainless-steel grid in front of the foil,

to protect it against mechanical damage by sharp objects or vandalism. The grid has a mesh of

0.8 mm to stop umbrella tips. However, as the absorptivity of stainless steel is 6%, i.e. very

low, the IR radiation passing through the grid is partly scattered and almost not absorbed. A

further advantage is that the grid reaches an equilibrium temperature (e.g. between 30 and 50

°C) absolutely safe to prevent skin or cloth burning. On the back of each foil, IR mirroring

and thermally insulating layers are placed, to avoid back heat dispersion, i.e. energy waste and

potential damage to the pews.

In the case study, the pews had no cultural value, and the emitters were simply fixed below

the seats and the kneeling pads with four small screws. When this is not possible in the case

of pews of historic value, a metal frame should be used, leaving pews untouched. There is

thermal insulation between the emitters and the pews, and the wooden pews are heated less

than due to the contact with sitting or kneeling persons, e.g. 10 °C below the seat and the

kneeling pad, 4 °C in the vicinity of the hand warmer. This has been considered a reasonable

reference for the allowable heating of wood.

The efficiency and performance of the heating foil emitters were tested in the laboratory by

measuring the two-dimensional distribution of the heat that reaches a vertical plane placed at

various distances, i.e. 10, 20,… 100 cm, from the foil. The measurement was done with the

help of a black strip, totally absorbed radiation, placed in front of the emitter. A regular mesh

was drawn on the strip to mark the points at which T was measured with the IR thermometer.

The measurements provided a 3D distribution of the heat emitted from the source. Once the

prototype of the novel system was built and installed in the church, vertical T profiles

representative of the situation for a person sitting, kneeling or standing between the pews

were monitored with the vertical black strips (described later). The monitoring has provided

information about the efficiency of the heating-foil emitters, their best location in the pews,

and the distribution of heat in the area where people are.

The typical metabolic activity of a human being in a church was considered to be for a seated,

quiet person: 1.0 MET; for a standing, relaxed person: 1.2 MET; for singing: 1.5 MET; for

8

walking: 2.0 MET [16]. One MET is equivalent to metabolic rate consuming 1 kilocalorie per

kilogram of body weight per hour, also expressed as 58.2 Wm-2, for a person sitting at rest.

Due to low physical activity and thus low metabolic heat production, the feet, legs and hands

were identified as parts of the human body that needed more external heat in a cold church.

The upper body was usually adequately insulated by clothing and needed less external heat.

Individual heaters with a diverse range of size, temperature and location were provided,

accordingly. A number of low-temperature emitters were considered for each pew and

strategically placed to satisfy the warmth requests of the specific body parts. An effort was

made to balance the heat distribution on feet, tibia, calf etc. and to assess the heat transferred

to people and the small amount escaping vertically (Fig.2). It was found out that feet were

most comfortably warmed when placed beneath the kneeling pad, however skirts could leave

unprotected the lower part of legs which might need heating; in contrast, the central and upper

part of the body should be adequately protected by overcoats and needed little heating.

Each pew was provided with three emitters: an under-seat emitter, an under-kneeler emitter

and a hand warmer (Fig.3). The under-seat emitter (55 °C), in which the heating-foil was

rolled into half-cylinder, was placed below the bench, to emit radiation in the front and back

direction, i.e. on the calf of the persons sitting in this bench, and on the tibia of the person

sitting in the back bench, respectively, as well as vertically, on the calf of the person, when

kneeling. A further heating element (65 °C) was placed under the kneeler pad to warm the

feet. Should feet warm too much, the occupier of the bench could withdraw them, reducing

the heat supply. A hand-warmer (60 °C), or trunk warmer when kneeling, was placed on the

back of the front pew. This element is useful in the case of severe cold, as in the Alps, but is

not necessary in milder regions.

A strip made of a purely insulating layer was added on the seat to provide a more comfortable

sensation when sitting. The insulating strip quickly reaches the body temperature and

provides comfort, without consumption of external energy. Heated pillows may be used to

supply heat to the body, but such a solution was not considered in this project to avoid any

possible health risk to sitting people.

An alternative, very comfortable solution would be to provide pews with back heaters.

However, they would change the appearance of the pews with empty backs. Heating glass

panes, made of a very resistant tempered glass, with a transparent sub-micrometric layer of

sputtered metal oxides inside, were tried for the purpose. They are heated up by the Joule

9

effect when electric power is supplied. A thermostat maintains the temperature at the desired

level of 40 °C and a second thermostat guarantees extra safety in the case of failure of the first

thermostat. They provide heat by IR radiation or direct contact with the back or the hands.

The glass panes are fully transparent, non-reflecting, dirt-resistant and self-supporting, i.e. no

frame is needed, so that they are almost invisible. Obviously, the hand warmer is no longer

needed in cases where the back heater is installed, while the under-kneeler pad for feet and the

under-seat emitter for legs are still useful.

In the FH prototypes installed in the two Alpine churches, the pews are subdivided into a

number of groups, which are powered according to the needs. When the church is crowded,

all pews are operated; however, when only a few churchgoers are in, only the occupied part of

the pews is heated, with the advantage of reducing energy consumption.

For the area of the altar and the choir, a heating carpet and additional remote quartz tube IR

emitters were used. A heating carpet consists of a heating foil or a heating wire placed

between an insulating layer on the bottom to avoid heat dispersion to the floor and a carpet-

like layer on the top. The top layer should protect the heating foil against mechanical damage

by sharp objects (e.g. spike heels), fire, water etc. It is fixed with Velcro fasteners and can be

easily exchanged so various colours and patterns can be used. The surface temperature is in

the range 20-25 °C [3]. The temperature is comfortable for the feet but insufficient for the rest

of the body, so that overhead IR radiant heaters are a valuable addition. In the warm season,

the heating carpet can be removed.

A possible addition, popular in some parts of Europe, is a heating foil (30 °C) placed below

the altar cloth. The thickness of the heating foil is less than 1 mm, so that it is invisible below

the altar cloth. The celebrant has the possibility of heating his hands and of receiving some IR

radiation. However, not all celebrants accept technical devices like heating foil or a

microphone placed on the altar, which has a highly symbolic meaning.

Short-wave infrared heaters (e.g. halogen quartz heaters) and medium-wave infrared heaters

(e.g. ceramic, carbon or quartz tubes) are capable of quickly heating surfaces reached by

radiation with low operating costs [1,3,12,13]. The former, similar to halogen lamps, are less

suitable due to the annoying glare and some UV emission. The latter, operating at lower

temperatures of red without glare, convert nearly all electrical energy into heat, which is

directly used to heat people. In both systems some energy is unavoidably lost due to

conduction or convection. New, technically advanced models of glare-free radiant sources,

10

which can be positioned horizontally or vertically, were tested in the laboratory for the altar

area. Two such heaters were symmetrically placed on the walls, on the right and the left side

of the altar area, and were found to raise the equivalent temperature by a maximum of 6 °C,

which was sufficient, as an addition to a heating carpet, for the celebrant and his assistants.

4. Monitoring the outdoor and indoor microclimate, as well as the building and

artwork response

Environmental monitoring was performed in order to understand the natural microclimate in

the church in relation to the external conditions, as well as to evaluate the perturbations

generated by the WA heating and the novel FH prototype when they were operating. The

indoor and outdoor data were automatically sampled every 15 minutes and transmitted every

night by GMS network to CNR-ISAC in Padua. Real-time control of the proper operation of

the station was possible, as was quick intervention in case of malfunctioning.

4.1. Outdoors

TA, RH, wind and net radiation were automatically sampled. Humidity mixing ratio, i.e. the

mass of water vapour to the mass of dry air (MR) and dew point (DP) were calculated from

TA and RH readings.

4.2. Indoors

The automatic monitoring included:

• vertical profiles of TA, RH and TS measured near or on the altarpiece, the entrance and

the organ, with sensors located at five levels (0.7, 2.5, 3.5, 4.5, 6 m);

• continuous remote IR monitoring of TS of the ceiling, the two side walls, the floor and

the pew area, which was done with an array of six high-precision Everest radiometers (+0.2

°C) installed on the organ choir;

• additional continuous monitoring of TA and RH, performed on the tie beam across the

nave, in the altarpiece of the left side chapel and between the pews;

• special monitoring surveys were carried out in the church, the pew area, near the heat

emitters and the artworks to study the impact of the existent WA heating on the artworks and

11

to verify to what extent the heat remained confined to the pew area after the novel FH

prototype was installed;

- horizontal cross sections of TA and RH, inside the church, including the area

both occupied and not occupied by people. Sampling was done on a horizontal plane

at the level of 1 m above the floor. MR was calculated from the direct observations of

TA and RH,

- vertical cross-sections along the nave (from floor to ceiling) of TA, RH, MR

from vertical profiles obtained with sensors fixed to a 10 m fishing rod (vertical

resolution 1 m);

• vertical profiles of the equilibrium T derived from a balance between radiation, TA

and convection on black strips (vertical resolution 10 cm), in the pew area from floor to a

man’s height (180 cm);

• surface and emission T of the heating elements in the pews to test safety concerning

skin burning and material ignition;

• pew TS to test the impact of the heat emitters on the pews;

• TS of paintings, altarpieces, floor, walls, ceiling, windows, doors in order to know the

baseline values and to test the impact of heaters. Detailed mapping of TS was performed with

a remote IR sensor and a laser pointer, following a regular grid with 30 sampling points

uniformly distributed;

• indoor air motions and fluctuations (draughts) using sonic anemometry (Campbell

Scientific).

Laboratory tests were carried out to understand performance and efficiency of heat emitters,

i.e. the heating foils under the seats and the kneeling pads, the hand-warmer, the heating

carpet and the quartz tubes.

The temperature measurements, the calibration and the accuracy of the sensors followed

the UNI 11120 standard [17]:

• air and walls were continuously monitored with 30 Pt100 temperature sensors

(uncertainty +0.1 °C)

• TS of altarpieces, paintings etc. were monitored with a quasi-contact thermometer by

Finger to avoid any risk of damaging the artwork surfaces by direct contact with the sensor.

The instrument is provided with a parabolic mirror that converges on the pyrometric sensor

12

the total radiation emitted from the surface and at the same time provides a shield against

extraneous radiation from other bodies. Thanks to the parabolic mirror, the sensor is reached

by direct and reflected IR radiation emitted by the artwork surface as it were inside a

blackbody cavity. The measurement is therefore independent of the original surface

emissivity that is virtually brought to 1, and reaches a high precision (+0.2 °C). This

methodology has proven to be easy and accurate, and at the same time guarantees the highest

safety level to artworks, untouched by the instrument. Therefore, it has been adopted by CEN

in a new draft standard for temperature measurements in the field of artwork preservation

[18].

• remote spot measurements of TS were done with an IR Raytek thermometer. This was

a specially improved instrument that in addition to the surface emissivity compensation was

provided with further electronic compensation for the different temperature levels between the

pyrometric sensor and the target surface. This compensation was highly beneficial and gave

excellent results at calibration (+0.1 °C in the range -40° to +200 °C).

• vertical profiles of the equilibrium temperature derived from a balance between

radiation, TA and convection could be in principle measured by an array of well-known globe-

thermometers. However, the globe-thermometers cannot be used in this case because they

have a very slow response (20-30 min) when compared with the duration of T cycles during a

service for example and would act as a low-pass filter. In addition, their spherical shape was

not suitable to the aims of the measurements, especially to assessing the flux of heat across a

plane surface at different distances from the emitters. Vertical black strips were used as a

target reference. They had a low-thermal capacity and were able to quickly reach equilibrium

between incoming radiation, TA and convection. The strips were made of a non-woven textile

behaving as a blackbody, i.e. absorbing almost totally the short- and long-wave radiation up

to 25 µm (the absorptivity was in the range 0.95-0.98). The absorption was assessed by

controlling the transmission and reflection levels of both visible and IR radiation through or

from the black strips in the laboratory. The response time of the black strips was assessed

from a continuous monitoring of their equilibrium temperature using the Rayteck pyrometric

transducer and was in the order of 1 to 2 min, which was appropriate for monitoring the

dynamic changes during the heating and cooling phases in the church. From the practical

point of view, and as a very crude approximation, the equilibrium temperature measured by

the black strips represents the effective temperature reached by clothing, if we ignore the

13

additional heat supplied by the body. The black strips were 180 cm high, as a standing person,

10 cm wide and 0.1 cm thick, supported by a frame. The measuring points were marked every

10 cm to get a high resolution vertical profile. The black strips provide objective, useful

information on the efficiency of the tested heaters, the distribution and attenuation of the

radiant heat in the space, which allows assessing the optimum distance between the heat

source (emitter) and the receptor (e.g. feet, legs, hands). Due to the application potential and

several innovative aspects of this methodology in the field of environmental assessment of

cultural heritage preservation, it has been adopted in the draft CEN/TC346 standard

prEN15758 [18].

Relative humidity sensors were of capacitive type; their calibration and accuracy followed the

UNI 11131 standard [19] i.e. uncertainty +0.2 % in the range 30 to 95%. Psychrometric

measurements were done only when possible, i.e. with no frost on the wet bulb, with a

Technoel electronic psychrometer provided with an internal aneroid capsule for atmospheric

pressure detection and electronic compensation for it. This correction was relevant in the case

of sampling in the mountains.

5. The prevailing indoor climate

The climatic patterns are evaluated by plotting the daily mean values of indoor and outdoor

TA, emphasizing general, long-term tendencies (Fig.4a). In the warm season, the massive

building and soil accumulate heat that is given back in the cold season. The monthly average

difference is largest in winter, when the church is 7 °C warmer than the outdoor air. In terms

of daily averages, the church is 5 to 10 °C warmer.

During the whole calendar year, the indoor MR (Fig.4c) followed external variations, with

slightly greater moisture content by 1 g/kg on average, and a daily scatter in the range of 4

g/kg. This can be explained by several factors: though most of the rainwater absorbed by

walls evaporates outside, a part of it migrates and is given away inside; moisture is also

released by the congregation (breathing and transpiring) or transported inside on rainy days.

In summer, the indoor/outdoor concentrations are almost the same due to better ventilation.

The indoor RH range is smaller than the outdoor RH range (Fig.4b) due to the buffering effect

of the building and the wooden structures, i.e. the organ-loft, choir stalls, pews, confessionals,

doors and altarpieces. The natural indoor RH variations generally lie between 50 and 80%;

however, strong departures occur when the WA heating is operated.

14

6. The old warm-air heating system and its negative effects on artworks

6.1. Floor

The floor temperature remains unchanged.

6.2. Impact on the ceiling and frescoes

At each operation of the old WA system, the increase in TS of the upper part of the church

was 5 to 10 °C, half of that observed in the air, due to relatively high thermal diffusivity of the

masonry (Fig.5). However, this increase was very large compared to the natural TS variability

in the building which did not exceed the maximum value of 1 °C during clear nights, when

the ceiling cooled due to the infrared loss. The TS distribution was not uniform but reflected

the internal flow of the blown air: two maxima corresponded to the air inlet grilles; the

additional one in the mid-vault, reflected the trajectory of the blown air from the side chapel.

6.3. Vertical T and RH profiles

At each operation of the WA heating, the air stratified aloft forming a cushion of warm, dry

air in the upper part of the church, with e.g. TA sharply passing from 0 to 20 °C (Fig.6) and

RH dropping from 70% to 20%. The abrupt TA change augmented with height, reaching a

maximum under the vault, and almost zero near the floor. This left the congregation with cold

feet. The WA heating system produced a sharp vertical gradient, with most heat dispersed

aloft, with little advantage to people’s comfort and very low heating efficiency. The

temperature of the frescoes on the vault rose by some 10° C generating heating-cooling

cycles.

6.4. Simultaneous evaporation and condensation on artefacts characterised by different

thermal conductivity

It is well known that, when the church is crowded, the moisture released by people raises

both the moisture content (expressed as humidity mixing ratio [MR]) and the dew point (DP).

When WA heating is operating, TS of walls and frescoes rises at a rate slower than DP. When

DP>TS, water is absorbed by condensation in the masonry which gives rise to dampness in

cold areas (e.g. the side chapel). Therefore, two effects are simultaneously observed when the

15

WA heating is operating: the RH drops and all artworks which have a low thermal inertia (e.g.

paintings on canvas with TS close to TA) desiccate; at the same time the DP increases and

artefacts with high thermal inertia (TS<DP<TA) suffer from dampness. Dryness or dampness

occurring simultaneously in artefacts characterized by a different thermal conductivity have

already been observed [1,20] and are typical of dynamic heating regimes.

6.5. Simultaneous evaporation and moisture condensation on masonry at different levels from

the floor

An interesting new mechanism was observed when the church was empty and the WA

heating was operating. In the upper part of the church, an increase in TA generated a drop in

RH which forced evaporation from the masonry. The evaporation determined a modest

increase in the moisture content of 1-2 g/kg, homogeneously distributed within the whole

church volume (Fig.7). The effective mixing of the air in the church can be explained by the

moisture evaporation from the plasters (frescoed on the ceiling) which are colder than the air.

The moisture enters an internal boundary layer (IBL), i.e. a thin air layer adherent to the wall

surface, which has an intermediate temperature between the cold wall and the warm air

indoors. The denser IBL sinks flowing along the walls and collects all the evaporating water.

In this way, a continuous down-draught air transport mixes the excess moisture with the air,

which finally leads to its fully homogeneous distribution in the church. The mechanism has

been documented with profiles of T and MR, and with direct measurements of air motions

(sonic anemometry) along the walls and in the church interior.

The vertical RH profile in the air and at the interface air-plaster is controlled by the vertical

profiles of TA and TS, respectively, the MR being homogeneously distributed. The RH profile

reaches a minimum at the top of the nave, and a maximum at floor level. The small DP rise

(for the increase in MR), combined with the low TA and TS values near the floor, may lead to

high RH levels and even to surface condensation, favouring microbiological colonisation on

the lower part of walls and the floor. This floor-level condensation is fed by ceiling-level

evaporation, which is forced by the WA heating, and has a visual appearance that might be

mistaken for a slight moisture capillary rise. The damage area is visible in the left side of the

chapel, where salt driven deterioration of the plaster occurs in the wall just above the floor.

The environmental monitoring showed that the plaster temperature there often drops below

DP.

16

6.6. Impact on the altarpiece

Conflicting ideas exist about the deterioration mechanisms and the impact of the WA heating

on wooden artworks. It is sometimes believed optimistically that wood has such a slow

dimensional response that any short-term T and RH change is unable to generate damage. It is

argued in addition, that if a wooden object has been exposed in the past to a specific, large T

and RH fluctuation, it may be expected that similar fluctuations will not cause any further

damage if repeated. Under such assumptions, a heating event over a short time of a

celebration is not considered dangerous. However, it can be argued less optimistically that

wood’s internal damage can be cumulative and invisible microfractures can precede the

visible damage (e.g. a crack) which appears only after the internal structure has progressively

weakened or the structure has accumulated sufficient stress, in both cases with dramatic

consequences.

The direct on-site monitoring of mechanical damage in the elements of the main altarpiece

using acoustic emission (AE) has confirmed that each WA heating event gave rise to

characteristic bursts of AE activity, a proof of an on-going climate-induced damage to the

altarpiece [9]. Furthermore, the threshold in the magnitude of the RH variations, above which

damage of the historic wood appears, was found not to be a discrete value. The AE events are

recorded much below that threshold as fracturing occurs in locally weakened areas in wood.

Therefore, even a short heating episode for a service can be dangerous, especially when it is

repeated before the wood has completely relaxed from the previous strain.

7. The novel Friendly-Heating prototype: controlling the dispersion of heat inside the

church and its effect on artworks and pews

7.1. Floor

The floor temperature remains unchanged except the deliberately heated areas, i.e. the

wooden board directly irradiated by the IR emitters and the heating carpet in the altar area.

7.2. Impact on the ceiling and the frescoes

Below the vault, TA increased by 0.5 °C after a one-hour operation of the novel FH heating

system. Under normal operating conditions, the heating rate was of the order of 0.5 °C/h,

comparable with the natural thermal cycle in the warm season. The temperature of the

17

frescoes and the ceiling showed no detectable changes, because during the heating the ceiling

temperature in some parts increased by a few tenths of °C, and in others decreased, the

changes being overshadowed by the natural climate variability in the room. The maximum

increase observed was 0.5 °C (Fig.8): the same as generated by spot lamp lighting, or natural

daily cycles. This means that this heating system does not generate dissolution-

recrystallization cycles of the soluble salts present in the frescoes and masonry.

7.3. Vertical T and RH profiles

The vertical profiles from floor to ceiling were determined with 10 cm resolution in the first

180 cm, and then with one-metre resolution. The pew area had the required increase in

temperature, especially at the feet and leg level, but almost ceased at the two-metre level.

Micromapping on horizontal and vertical cross sections of the church (Fig.9) showed that the

heat remains contained in the areas occupied by people and the dispersion outside this area is

negligible. The continuous T and RH monitoring at various levels near the altarpiece and

entrance showed very small fluctuations at each FH operation, not greater than the natural

variability of the historic microclimate.

A comparison was made between the daily T and RH cycles when FH or the old WA heating

systems were used. The distribution of the cycles shows two distinct clouds of scattered data

(Fig.10). The cloud with the largest cycles corresponds to the WA heating and indicates high

risk of damage to vulnerable artefacts. The cloud with the smallest cycles corresponds to

much safer conditions and is generated by the combined effect of the natural microclimate

variability and the perturbation generated by the novel FH system. The daily cycles in T and

RH on days when FH was operated did not much exceed those with no heating at all, i.e. they

do not exceed the natural variability of the historic microclimate.

7.4. Simultaneous evaporation and moisture condensation

None observed.

7.5. Impact on the altarpiece

When the novel FH heating system was operating, the heating was only evident in the pew

area, but almost not outside it. The temperature was at its maximum within the pews, where it

reached the level required for human comfort. Above the pews, temperature decreased and the

18

effect of the novel heating system virtually disappeared at head height for a tall, standing

person. Further, the monitored elements of the altarpiece showed a very limited response, not

exceeding the natural variability due to the historic microclimate, as already mentioned.

In the 3-year period, the average T variation recorded below the ceiling and near the altar,

when FH was operating, was about 0.5 °C and the extreme variation was 2.5 °C.

Consequently, RH dropped gently only, the maximum departure being 5% (Fig.11).

7.6 Impact on pews

A general problem with pew heating is that emitters are installed on pews and might damage

them because of T cycles. The problem is particularly relevant in the case of historic pews. To

verify the impact of the FH system on pews, the pew temperature was measured at different

points. In all cases the most heated part was the surface where the heating element was

placed. The under-seat gets 12 °C warmer, the seat 3 °C, the kneeling pad 3 °C, the internal

part of the upper element containing the hand warmer 6 °C but no temperature increase was

recorded on its top surface where the forearms rest. If necessary, these values can be reduced

with better insulation; however, they all are smaller than those generated by people sitting (30

°C) or kneeling and therefore, are of no risk to the pews.

8. Draughts, comfort and the European standardization

8.1 Negative effects of draughts: wall blackening and thermal discomfort

Draughts are frequently a problem in churches, especially tall churches, which have a heat

source near the floor (e.g. underfloor heating, pew heating, radiant heaters warming up the

floor, convective heaters), use systems based on the forced inflow of warm air, or simply have

cold windows, ceiling or walls. Draught activity grows when an internal heat source increases

the temperature contrast between air, ceiling and walls, enhancing potential instabilities

determined by building features.

During this project, draughts were measured in the natural, basic condition of the unheated

empty church, as well as with the WA heating and with the operating prototype of the novel

heating system. Measurements were carried out at different locations within the church, e.g.

near the walls, in the pew area occupied by people, in the central part of the nave, near the

altar and in the two side chapels. The natural draughts were attenuated in the central part of

19

the nave, far from the walls; the absolute minimum (0.5 cm/s) was below the organ loft where

the wall extension was reduced to half the height of the nave and the second minimum was in

the altar area, far from the disturbance brought about by the two side chapels. Draughts were

stronger (2 ÷ 2.5 cm/s) near the walls, as expected, and in places where the nave opens to the

two side chapels added in the 19th century. The walls of the two chapels are thinner and

therefore colder than those of the nave. The wall temperature is similar in both chapels as

thickness and thermal conductivity of their walls are the same, and both are shielded from

sunshine for the whole winter period. However, the southern Trinity’s chapel, having a wide

part of the walls covered with wooden confessionals, showed smaller draught activity

compared to the northern Virgin’s chapel, which has a larger extension of bare walls, finished

with traditional plaster and a tempera coating. The situation would improve if the walls of the

Virgin’s chapel were painted with a thermal insulating coating, which would prevent the air

from direct contact with the cold underlying structure [21].

Even in the absence of heating, when the building envelope is colder than the indoor air, the

air entering into contact with the walls becomes cooler and denser, and sinks, forming a

dynamic IBL. The IBL accounts for the down draught activity and grows in thickness as the

air flows down along the wall. The intensity of this turbulent flow is controlled by the

difference between the air and wall temperatures and has two negative effects:

Draughts increase the deposition rate of suspended particles, i.e. wall blackening [20].

Temperature gradients enhance two main deposition mechanisms. The first is thermophoresis,

in which suspended particles undergo incessant collisions with air molecules from all sides;

however molecular impacts are more energetic from the warmer side. The consequence is that

particles gain momentum from warmer air and move towards the colder wall, where

ultimately, they arrive and stick. The second deposition mechanism is aerodynamic

impaction. When air flows along the wall, in this case due to decreasing buoyancy after

contact with the colder surface, the wall roughness induces perturbations to the air stream,

especially when it passes from the laminar to the turbulent regime. When the inertia of the

airborne particles dominates over the viscous drag of the air stream, the trajectory of the

particle deviates and may deposit on the wall surface. In addition, two other deposition

mechanisms increase their efficiency with temperature: Brownian deposition, because the

random motion of fine particles becomes more intense with higher energy of impinging air

20

molecules, and electrostatic capture because warmer air means also lower RH and a reduced

dissipation of electric charges.

Draught, from a human thermal comfort point of view, is defined as the unwanted local

cooling of the body caused by air movement. Draught sensation depends on the air speed, air

temperature, turbulence intensity, human activity and clothing. In a cool church, a draught

may cause unpleasant sensation especially for the head and legs where the body is not

covered with clothing. Moreover, the thermal state of a person affects her or his draught

sensation i.e. people feeling cool complain of draughts more than those feeling neutral or

warm [22].

Discussions concerning thermal comfort are not commonly found in the literature of cultural

property preservation. A common opinion is that heating for people’s comfort is dangerous

for artworks, i.e. thermal comfort and preservation are based on conflicting requirements.

However, a thorough analysis of the problem is crucial to finding a balance between the two

opposite needs. Obviously, adequate clothing is a very important aspect, and some regulations

exist for required clothing insulation [23]. However, churchgoers are often lightly dressed and

expect comfortable temperatures in churches rather than adequately prepare themselves for a

cold environment.

8.2 Assessing thermal comfort

The thermal comfort was assessed on the ground of three different kinds of information:

• a questionnaire for churchgoers in order to understand their subjective sensation;

• monitoring of air and black-strip temperatures, air velocity and turbulence in the

environment of the church;

• measurements of skin temperature, recorded at several parts of the body with the

thermistor probes.

The first method, i.e. the questionnaire, brought widely scattered answers on people’s thermal

sensations, ranging from ‘warm’ to ‘cool’. The answers were influenced by highly subjective

factors, like expectations, personal habits and requests and were considered scarcely

representative.

The second method was based on environmental monitoring and calculations.

21

Thermal levels before and during the operation of the heating systems were measured with

black strips placed in the middle of each row of pews to represent the effective temperature

felt by a standing person on different parts of the body. This was the case of a single person in

a pew. In the case of a usual congregation density, each person receives and supplies infrared

radiation, which improves the comfort. The situations for the FH and the previous WA

heating system are schematically shown in Fig. 12. The WA heating was characterized by

thermal stratification: with the warm head region (TA increasing by 5-10 °C) and cold feet

region (almost no increase in TA). This inverse temperature profile, i.e. cold feet and warm

head, does not represent a really comfortable solution. The profiles for FH indicate that the

comfort level is mainly attained in the area where people are. Even in severe conditions when

the indoor temperature in the unheated interior is close to 0 °C, the occupied area is mild on

average, with warm feet region (TA 18-25 °C), moderately cool leg region (TA 10-15 °C), and

cool head region (TA 6-8 °C). A large vertical TA difference between the head and ankles

may cause discomfort. The allowable temperature difference is 3°C, which applies to

situations where the temperature increases with height from the floor (i.e. the head is warmer

than the feet) [16].

As described above, convective air movements and turbulence were always present, but the

air velocity increased when the heating was operating. The average vertical velocity of

convection at shoulder level for FH was in the range 10 to 30 cm/s; almost twice the values

for the WA heating. The air motions were experienced as a draught sensation to the head and

face. Thus, the air velocity during the operation of FH exceeded, at least occasionally, the

desired value of 15 to 20 cm/s for sedentary situation according to ISO 7730 [24].

Draught Rating (DR), which corresponds to the percentage of people predicted to be

dissatisfied due to draught, was calculated according to ISO 7730 [24] and ASHRAE 55 [16]

using the following formula:

DR = (34-TA) ·(<v>-0.05)0.62 · (0.37 <v> Tu + 3.14) (%)

where the variables are: <TA> (°C) average air temperature, <v> (m/s) average air velocity ,

<Tu> (%) average turbulence intensity defined as <Tu> = Std Dev/<v>, where Std Dev is the

standard deviation of the v vector, sampled at 4 Hz frequency. DR is valid for sedentary,

thermally neutral persons dressed in normal indoor clothing and at an TA range from 20 to 26

°C.

22

For working sites, the requirements for maximum allowable air speed is based on 20%

dissatisfied [16]. This is, however, not realistic for a church which is colder and visited for

short periods by people dressed for outdoors. For the WA heating, DR was about 30% and for

FH in the range of 28 – 70%. As the requirement was originally defined for working sites

which are characterized by mild temperature (i.e. 20 < T < 26 °C) and in which people stay

longer, it is logical to expect that the DR range acceptable in a church should be much wider.

Furthermore, DR 20% is based on the greatest sensitivity of draught (the head region with

airflow from behind) and, therefore, may be conservative for other locations of the body and

other directions of airflow. In conclusion, this index is not convenient to provide an absolute

judgment about thermal comfort for conditions different from those of a standard office.

However, it seems possible to use this index for relative evaluations, i.e. to compare thermal

sensation levels achieved by different configurations of heating systems, in order to establish

which of them is more comfortable, and which is less.

The alternative method to assess thermal comfort was based on measurements of skin

temperature. The skin temperature was measured on 11 parts of the body (forehead, chest,

lower back, forearm, hand, finger, thigh front, thigh back, calf, foot and toe) for a number of

volunteers. The data were sampled at one-minute interval and stored in pocket data-loggers.

Mean skin temperature was calculated as an area weighted average [26]. The method is

objective and provides information about the thermal comfort of an individual and the

demand for heat supply, and how this varies with time and physical activity, e.g. when people

are sitting, standing or kneeling. Thermal sensation and comfort were also asked by using

subjective judgement scales [26].

At present, thermal comfort is regulated by ASHRAE 55 [16] and ISO 7730 [24] which cover

mild (TA > 18 °C) working places in which people stay for a long time. The standards specify

conditions acceptable to 80% of people exposed to the same conditions. The 80% overall

acceptability assumes 10% dissatisfaction for general thermal comfort plus 10%

dissatisfaction that may simultaneously occur from local thermal discomfort (i.e. unwanted

cooling of one particular part of the body). The FH research constitutes a valuable

contribution to the European standardisation concerning church heating. Basing on the

experience of the FH project the draft CEN/TC346 standard prEN 15759 [27] states that in the

specific case of churches, thermal comfort cannot be regulated by the above ASHRAE and

ISO standards. It was proposed to use the mean skin temperature of people adequately clothed

23

for cold environments to define the desired level of the thermal comfort rather than TA. The

skin temperature is a bulk parameter that takes into account all the environmental factors (TA,

radiation, draughts, RH), clothing, and physiological features. It was further stated that

thermal comfort should be limited to a slightly cool thermal sensation corresponding to an

average skin temperature of 30-33 °C. This target was reached by the FH prototype as

illustrated by the data measured in the church of Rocca Pietore (Fig.13). However, the FH

system might be incapable of providing enough comfort to the upper part of the body,

particularly the head in very cold indoor environments when the temperature drops near to, or

below 0 °C. An increase in the power of the heat sources will warm too much the lower part

of the body without a clear advantage for the upper one, and the dissipation of the excess heat

will produce unpleasant draughts. In such cases, additional heat should be supplied from an

additional intergrative system, e.g. overhead radiant heaters.

9. Tolerable and adverse microclimate variability for artwork preservation:

determination of the optimum climate target based on the historic microclimate

The good preservation state of artworks, kept over centuries in the same place, demonstrates

that these objects have adapted to their particular historic indoor microclimate. Such

adaptation might have produced fractures, acting as expansion joints which release tensions in

the materials generated by the variations of T and RH. If this particular historic climate is

changed, for example the variations of T or RH are increased, further physical damage can

occur until the artwork has acclimatised to the new conditions. Physical damage can be both

catastrophic and cumulative, especially when the stress develops in short-term cycles before

the complete relaxation of the material.

When a heating system is planned, a smaller or larger departure from the natural

microclimatic conditions should be expected. To evaluate whether the departure is safe or

may involve risk of damage, the climate target based on the historic microclimate, to which

artworks are acclimatised, should be defined and one should verify if the expected levels and

fluctuations of T and RH fall within that target or go outside it. A very good example of such

analysis can be provided using microclimatic variations induced by the WA and FH heating

systems, respectively.

The historic microclimate can be represented in terms of typical, most frequently occurring T

and RH levels and fluctuations. A method for specifying these values was proposed in the

24

draft CEN/TC346 standard prEN 15757 [28]. When stabilising RH has priority, the yearly

average RH, determined on the basis of the microclimate data continually sampled, is the

target level. The seasonal RH cycle is obtained by calculating a 30-day moving average, i.e.

the average for the previous 30-day period. The magnitude of a fluctuation is calculated as the

difference between each RH reading and the corresponding smoothed value of the moving

average. The lower and upper limits of the target range of RH fluctuations are determined as

the 8th and 92nd percentiles of the fluctuation amplitudes recorded in the monitoring period,

respectively. The 8th and 92nd percentiles are obtained by ordering the fluctuations by

amplitude, from the least to greatest value and selecting the values below which 8% or 92%

of observations may be found, respectively (Fig. 14 and 15). In this way, 16% of the greatest,

most risky fluctuations are excluded, which corresponds to one standard deviation in the

distribution of the fluctuation amplitudes. The cuts are equally applied to peaks and drops in

RH, yielding excessively moist or dry environments. However, if RH fluctuations depart by

less than 10% from the seasonal RH level, the 10% threshold should be accepted. The results

of the above procedure applied to the microclimate in the church in Rocca Pietore are reported

in Figs.16-18. The yearly average RH is 55%. A clear seasonal cycle in RH is observed. The

average RH decreases from the 60 to 70% summer range to 40 to 50% in winter (Fig.16). The

short-term fluctuations, superimposed on the seasonal variations, are shown in Fig. 17. The

lower and upper cut-off levels, i.e. the 8th and 92nd percentile respectively, are shown as

straight lines. They have been calculated from climatic data of the warm period only, during

which the indoor climate can be considered natural, not disturbed by heating episodes. The

target band is compared with the observed RH variations in Fig. 18. It is obvious that the

sporadically operated WA heating system is the major cause of departures from the climate

stability, because it generates deep short-term RH drops below the lower limit of the target

band of tolerable RH variations, as derived from the natural climate of the church.

25

10. Conclusions

The FH project offered a unique opportunity for a holistic approach to the problem of heating

historic churches which would safeguard the cultural heritage preserved in them. The outdoor

and indoor environmental parameters and the response of the building envelope and artworks

to the environmental variability were simultaneously monitored to establish clear cause-effect

relationships in order to improve the understanding of the main deterioration mechanisms

originating when a room is heated.

The experimental methodology of monitoring the effects of the heating episodes on the

artworks had some innovative aspects. The first was the use of a quasi-contact IR sensor that

provides very accurate measurements without any contact with the fragile artwork surface.

The second innovative methodology was the use of black strips reaching a rapid equilibrium

between air temperature, incoming radiation and convection to measure the effective

temperature reached by various surfaces. These two methodologies were included in the draft

European standard CEN/TC346 prEN 15758 Conservation of cultural property – Procedures

and instruments for measuring the temperatures of the air and of the surfaces of objects.

Another innovative aspect was establishing a procedure for determining the optimal climate

target based on the historic microclimate and of verifying whether a heating system is safe to

artworks, or induces variations falling within the risk-prone areas. This methodology was

included in the draft CEN/TC346 standard prEN 15757 Conservation of Cultural Property -

Specifications for temperature and relative humidity to limit climate-induced mechanical

damage in organic hygroscopic materials.

Church heating can cause problems in several areas: air movements create unpleasant

draughts and increase soiling of painted surfaces, periodic drops in the RH cause desiccation

of artworks of low thermal inertia like paintings on canvas and their shrinkage, whereas

moisture released can condense on cold walls having a slow thermal response. The FH project

has adequately addressed these problems by providing localised heat in the pew area where

people are without changing the general indoor microclimate, especially in the proximity of

the walls and the ceiling, where most valuable objects are located. A number of low-

temperature radiant emitters were mounted in pews to provide a desirable distribution of heat

to feet, legs and hands. Due to little heat dispersion and modular operation (single benches or

groups of them can be heated), the novel heating system is energy-efficient. The FH prototype

has brought an enormous improvement to the preservation of artworks. Furthermore, the

26

installation of the system produces limited mutilation to the church structure as it is powered

by electricity. The research has also proposed to use the skin temperature to define the desired

level of the thermal comfort. The 30 to 33°C mean skin temperature, corresponding to slightly

cool to neutral thermal sensations, should be the goal of any heating system and the

congregation should be encouraged to wear highly insulating clothing and footwear

The findings of the project have allowed formulating the key recommendations for the draft

CEN/TC346 standard prEN 15759 Conservation of Cultural Property - Specification and

control of indoor environment – Heating of churches.

Acknowledgements

This research was carried out within the Friendly Heating project (contract EVK4-CT-2001-

00067), supported financially by the European Commission 5th Framework Programme,

Thematic Priority: Environment and Sustainable Development, Key Action 4: City of

Tomorrow and Cultural Heritage. The authors also thank Mgr Giancarlo Santi and Don

Stefano Russo from the Italian Episcopal Conference (CEI) in Rome, Don Giacomo

Mezzorana of the Diocese in Belluno and Don Attilio De Zaiacomo, the parish priest in Rocca

Pietore, for their assistance in this study. A further European Commission SENSORGAN

project: Sensor system for detection of harmful environments for pipe organs, contract

022695, and COST Action D42 ‘Enviart’ gave the opportunity of further improvements,

international contacts, meetings and useful discussions.

27

Bibliography

[1] D. Camuffo (Ed.), Church Heating and the Preservation of Cultural Heritage, Guide to the

Analysis of the Pros and Cons of Various Heating Systems, Electa, Milan, 2007.

[2] H. Schellen, Heating Monumental Churches, Indoor Climate and Preservation of Cultural

Heritage, Eindhoven Technical University, Eindhoven, 2002.

[3] W. Bordass, C. Bemrose, Heating Your Church, Church House Publishing, London, 1996.

[4] C. Arendt, Raumklima in großen historischen Räumen: Heizungsart, Heizungsweise,

Schadensentwicklung, Schadensverhinderung, Rudolf Müller, Köln, 1993.

[5] H. Künzel, Bauphysik und Denkmalpflege, Frauenhoffer IRB Verlag, Stuttgart, 2007.

[6] D. Limpens-Neilen, Bench heating in monumental churches – Thermal performance of a

prototype, PhD Thesis, Eindhoven Technical University, Eindhoven, 2006.

[7] L. Bratasz, R. Kozlowski, Laser sensors for continuous in-situ monitoring of the

dimensional response of wooden objects, Studies in Conservation 50 (2005) 307-315.

[8] L. Bratasz, R. Kozlowski, D. Camuffo, E. Pagan, Impact of indoor heating on painted

wood: monitoring the mediaeval altar in the church of Santa Maria Maddalena in Rocca

Pietore, Italy, Studies in Conservation 50 (2007) 199-210.

[9] S. Jakiela, L. Bratasz, R. Kozlowski, Acoustic emission for tracing the evolution of

damage in wooden objects, Studies in Conservation 52 (2007) 101-109.

[10] Z. Spolnik, L. Bencs, A. Worobiec, V. Kontozova, R. Van Grieken, Application of

EDXRF and thin window EPMA for the investigation of the influence of hot air heating on

the generation and deposition of particulate matter, Microchim. Acta 149 (2005) 79-85.

[11] L. Bencs, Z. Spolnik, D. Limpens-Neilen, H. Schellen, B. Jütte, R. Van Grieken,

Comparison of hot-air and low-radiant pew heating systems on the distribution and transport

of gaseous pollutants in the mountain church of Rocca Pietore from artwork conservation

point of view, J. Cult. Herit. 8 (2007) 264-271.

[12] Z. Spolnik, A. Worobiec, L. Samek, L. Bencs, K. Belikov, R. Van Grieken, Influence of

different types of heating systems on particulate air pollution deposition: the case of churches

situated in a cold climate, J. Cult. Herit. 8 (2007) 7-12.

[13] L. Samek, A. De Maeyer-Worobiec, Z. Spolnik, L. Bencs, V. Kontozova, L. Bratasz, R.

Kozlowski, R. Van Grieken, The impact of electric overhead radiant heating on the indoor

environment of historic churches, J. Cult. Herit. 8 (2007) 361-369.

28

[14] A.M. Spiazzi, G. Toffoli, Rocca Pietore. La chiesa di Santa Maria Maddalena, Canova,

Treviso, 1993.

[15] D. Camuffo, G. Sturaro, A. Valentino, M. Camuffo, The Conservation of Artworks and

the Hot Air Heating System in Churches: Are they Compatible? The Case of Rocca Pietore,

Italian Alps, Studies in Conservation 44 (1999) 209-216.

[16] AHSRAE Standard 55: Thermal Environment Conditions for Human Occupancy,

American Society of Heating, Refrigerating and Air-Conditioning Engineers, 2004.

[17] Italian Standard UNI 11120, Cultural heritage - Field measurement of the air temperature

and the surface temperature of objects, Ente Nazionale Italiano di Unificazione, Milan, 2005.

[18] Draft European Standard CEN/TC346 prEN 15758, Conservation of cultural property –

Procedures and instruments for measuring temperatures of the air and of the surfaces of

objects, European Committee for Standardisation, Brussels, 2008.

[19] Italian Standard UNI 11131, Cultural heritage - Field measurement of the air humidity,

Ente Nazionale Italiano di Unificazione, Milan, 2005.

[20] D. Camuffo, Microclimate for Cultural Heritage. Developments in Atmospheric Science

23, Elsevier, Amsterdam, 1998.

[21] M. Santamouris, Passive Cooling of Buildings – The State of the Art, in: Advances in

Solar Energy, Vol.16, ISES, James and James Science Publishers, London, 2005.

[22] J. Toftum, R. Nielse, Draught sensitivity is influenced by general thermal sensation, Int.

J. of Ind. Ergon. 18 (1996) 295-305.

[23] ISO Standard 11079, Ergonomics of the thermal environment. Determination and

interpretation of cold stress when using required clothing insulation (IREQ) and local cooling

effects, International Organization for Standardization, 2007.

[24] ISO Standard 7730, Moderate thermal environments - Determination of the PMV and

PPD indices and specification of the conditions for thermal comfort. International

Organization for Standardization, 1994.

[25] J.D. Hardy, E.F. DuBois, The technique of measuring radiation and convection, J. Nutr.

15 (1938) 461-475.

[26] ISO Standard 10551, Ergonomics of the thermal environment - Assessment of the

influence of the thermal environment using subjective judgment scales. International

Organization for Standardization, 1995.

29

[27] Draft European Standard CEN/TC346 prEN 15759, Conservation of Cultural Property –

Specification and control of indoor environment - Heating of churches, European Committee

for Standardisation, Brussels, 2008.

[28] Draft European Standard CEN/TC346 prEN 15757, Conservation of Cultural Property -

Specifications for temperature and relative humidity to limit climate-induced damage in

organic hygroscopic materials, European Committee for Standardisation, Brussels, 2008.

30

Figures

Fig. 1. Temperature of two heating foils: a wide one under pew seats (black line) and a

narrow one under kneeler (grey line). They operated without a thermostat i.e. with a natural

cutout. The maximum foil temperature is self-regulated by the graphite granule resistance,

which increases with the foil temperature.

31

Fig. 2. Black-strip temperature between the pews. Each line refers to a slightly different

solution or position tested in the project. The vertical line B, refers to the room temperature

outside the pew area.

32

Fig. 3. The three IR emitters located on a pew: A. Under a kneeler, B. Under a seat, C. On the

back of the front pew to warm hands

33

a) b)

c)

Fig. 4. Outdoor - indoor daily averages (grey) and monthly averages (black) of air

temperature (a), relative humidity (b) and mixing ratio (c). “Out” stands for “outdoor” and

“in” for “indoor” in the axis description.

34

Fig. 5. Ceiling temperature before (left) and after (right) the warm-air heating has been

operated for 30 minutes.

35

Fig. 6. Distribution of the air temperature in a vertical plane along the nave after the warm air

system has been operated for 30 minutes. The entrance is on the left and the altar is on the

right.

36

a)

b)

c)

Fig. 7. Variation of air temperature, relative humidity and mixing ratio caused by the warm-air

heating episode in the absence of people. The warm air stratifies aloft (a) with the higher

temperature favouring evaporation from masonry in the upper part of the nave. The water vapour

is homogeneously distributed within the church (c). The combination of greater moisture content

and thermal layering produces a drop in RH aloft and an increase near the floor (b).

37

Fig. 8. Ceiling temperature before (left) and after (right) the Friendly-Heating system has

been operated for 70 minutes.

38

Fig. 9. Micromapping of the air temperature in a vertical cross section of the church during

the operation of the Friendly-Heating system.

39

Fig. 10. Daily ranges of temperature and relative humidity. Two groups of variations are

visible: small variations when the Friendly-Heating system was operated and large variations

for the warm-air system.

40

a)

b)

Fig. 11. Air temperature (a) and relative humidity (b) measured at several levels near the altar.

Perturbation is visible when the heating is operated, but this is minor with the Friendly-

Heating system (FH) and substantial with the warm-air (WA) heating.

41

Fig. 12. Effective temperature measured at different parts of the body for the Friendly-

Heating (FH) and warm-air (WA) heating systems.

42

Fig. 13. Mean skin temperature of several volunteers (averaged data) staying in a church

heated with the Friendly-Heating (FH) or warm-air (WA) heating systems. 31 °C corresponds

to slightly cool thermal sensation and 33 °C to neutral thermal sensation.

43

a)

b)

Fig. 14. Rocca Pietore church. Percentile distribution of the air temperature (a) and relative

humidity (b) representing the natural historic climate inside the church during a calendar year.

Heating episodes operated with the warm-air heating have been incorporated and are

responsible for the strong departure of the top temperature percentile and the lowest RH

percentile in the cold season.

44

a)

b)

Fig. 15. Rocca Pietore church. Percentile distribution of the daily variability in air

temperature (a) and relative humidity (b) representing the natural variability of the historic

climate inside the church during a calendar year. Heating episodes operated with the warm-air

heating are separately reported (black dots). They fall outside the natural variability.

45

Fig. 16. Rocca Pietore church. Indoor RH during 1 year (the jagged black line) and a seasonal

RH cycle (smooth grey line) obtained by calculating the 30-day moving average of the

readings. The yearly RH mean (the RH target level) is marked by a horizontal line.

46

Fig. 17. Rocca Pietore church. Short-term variations of RH around the general seasonal

tendency; the lower and upper limits of the range are calculated as the 8th and the 92nd

percentiles of the fluctuation amplitudes, respectively.

47

Fig. 18. Rocca Pietore church. Target RH level and range; the warm-air heating system

generates short-term RH drops hugely exceeding the lower limit of the target band of

tolerable RH fluctuations derived from the natural climate of the church.