An advanced church heating system favourable to artworks: A contribution to European standardisation

Journal of Cultural Heritage 11 (2010) 205–219

Original article

An advanced church heating system favourable to artworks: A contribution to European standardisation Dario Camuffo a,∗ , Emanuela Pagan a , Sirkka Rissanen b , Łukasz Bratasz c , Roman Kozłowski c , Marco Camuffo d , Antonio della Valle a 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, 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 Vicenza, 21, 35121 Padova, Italy Received 16 December 2008; accepted 24 February 2009 Available online 13 November 2009

Abstract The European project Friendly-Heating (FH): 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. © 2009 Elsevier Masson SAS. All rights reserved. Keywords: Church heating; Preservation of cultural property; Warm-air heating; Pew heating; Heating carpets; Localized heating; Thermal comfort; European standardization

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 favorable 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 tem-

∗

Corresponding author. Tel.: +39 049 8295902. E-mail address: [email protected] (D. Camuffo).

1296-2074/$ – see front matter © 2009 Elsevier Masson SAS. All rights reserved. doi:10.1016/j.culher.2009.02.008

perature (T) 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

206

D. Camuffo et al. / Journal of Cultural Heritage 11 (2010) 205–219

heating is especially efficient in preventing dampness in buildings during the rainy seasons. Continuous heating is sometimes considered a favorable strategy, which avoids cyclical changes in T 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 T is +10 ◦ C, the indoor RH drops to extremely low levels e.g. 10 to 20% when outdoor RH is 50 to 100%, respectively. Low RH levels are damaging artworks. For example, they can cause the excessive shrinkage of wood leading to irreversible stretching and eventual cracking. Rapid heating for 1 or 2 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. The European Funded Commission project FriendlyHeating (FH), implemented between 2002 and 2005, started with analysing the main risks for preservation and 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 proximity 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 3-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 T between −10 and −20 ◦ C. The church was built in the 15th century from local stone, and features onemetre-thick walls, and a nave with a chapel on either side, which were added in the 19th 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 altarpiece, paintings on canvas and panels, choir stalls, a decorated organ-loft with modern organ, and frescoes from the 15th 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 selfprotection. 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 Saint-Mary Magdalene and Saint-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 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 14th 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.

D. Camuffo et al. / Journal of Cultural Heritage 11 (2010) 205–219

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 prototypes 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 volume to provide a thermally comfortable environment before people entered for a service, and then keeping this T 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 lowtemperature 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 temperature. 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 reduces the current intensity. Consequently, the maximum T is self-regulated (Fig. 1) at levels specifically

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

207

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 T (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 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 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 walking: 2.0 MET [16]. One MET is equivalent to metabolic rate consuming one 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, 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, T and location were provided, accordingly. A number of low-temperature emitters were considered for each pew and strategically placed to satisfy the warmth request 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

208

D. Camuffo et al. / Journal of Cultural Heritage 11 (2010) 205–219

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.

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

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.

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 T 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 submicrometric layer of sputtered metal oxides inside, were tried for the purpose. They are heated up by the Joule effect when electric power is supplied. A thermostat maintains T at the desired level of 40 ◦ C and a second thermostat guarantees extrasafety 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 underkneeler pad for feet and the under-seat emitter for legs are still useful. In the FH prototypes installed in 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 (TS ) is in the range 20 to 25 ◦ C [3]. The T 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

D. Camuffo et al. / Journal of Cultural Heritage 11 (2010) 205–219

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 Ts 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, 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 T 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 The 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 the 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 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

209

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. heating foils under the seats and kneeling pads, hand-warmer, heating carpet and quartz tubes. The T 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 quasicontact 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 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 T 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 T 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 T derived from a balance between radiant and TA 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

210

D. Camuffo et al. / Journal of Cultural Heritage 11 (2010) 205–219

have a very slow response (20–30 minutes) 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 ␮m (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 T using the Rayteck pyrometric transducer and was in the order of 1 to 2 minutes, 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 T measured by the black strips represents the effective T reached by clothing, if we ignore the 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 highresolution 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 prEN 15758 [18]. RH 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 psychrometer sampling in the mountains. 5. The prevailing indoor climate The climatic patterns are evaluated by plotting the daily mean values of indoor and outdoor Ts, 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.

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.

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.

D. Camuffo et al. / Journal of Cultural Heritage 11 (2010) 205–219

211

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. 6. The old warm-air heating system and its negative effects on artworks 6.1. Floor The floor T 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

Fig. 5. Ceiling temperature before (left) and after (right) the warm-air heating has been operated for 30 minutes.

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.

and very low heating efficiency. The T 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 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 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 to 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 T 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

212

D. Camuffo et al. / Journal of Cultural Heritage 11 (2010) 205–219

evaporation, which is forced by the WA heating, and has a visual appearance that might be mistaken for a slight moisture capillary rise. The damaged 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 T there often drops below the DP. 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 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 ongoing 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 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 (c).

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 air-plaster interface 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

7.1. Floor The floor T 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 T of the frescoes and the ceiling showed no detectable changes, because during the heating, the ceiling T 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

D. Camuffo et al. / Journal of Cultural Heritage 11 (2010) 205–219

213

Fig. 9. Micromapping of the air temperature in a vertical cross section of the church during the operation of the Friendly-Heating system.

Fig. 8. Ceiling temperature before (left) and after (right) the Friendly-Heating system has been operated for 70 minutes.

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 dissolutioncrystallization cycles of the soluble salts present in the frescoes and the masonry. 7.3. Vertical T and RH profiles 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 T, 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 T was at its maximum within the pews, where it reached the level required for human comfort. Above the pews, T decreased

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.

214

D. Camuffo et al. / Journal of Cultural Heritage 11 (2010) 205–219

8. Draughts, comfort and the European standardization 8.1. Negative effects of draughts: wall blackening and thermal discomfort

Fig. 11. Air temperature (a) and relative humidity (b) measured at a 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.

and the 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 T 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 T 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 pews.

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 T 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 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 T 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 Ts and has two negative effects. Draughts increase the deposition rate of suspended particles, i.e. wall blackening [20]. T 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

D. Camuffo et al. / Journal of Cultural Heritage 11 (2010) 205–219

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 T: Brownian deposition, because the random motion of fine particles becomes more intense with higher energy of impinging air 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, TA , 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 T 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 Ts , air velocity and turbulence in the environment of the church; • measurements of skin T, recorded at several parts of the body with 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. 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 T 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

215

Fig. 12. Effective temperature measured at different parts of the body for the Friendly-Heating (FH) and warm-air (WA) heating systems.

heating system are schematically shown in Fig. 12. The WA heating was characterized by thermal stratification: with warm head (TA increasing by 5 to 10 ◦ C) and cold feet (almost no increase in TA ). This inverse T 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 T in the unheated interior is close to 0 ◦ C, the occupied area is mild on average, with warm feet (TA 18–25 ◦ C), moderately cool leg (TA 10–15 ◦ C), and cool head (TA 6–8 ◦ C). A large vertical T difference between head and ankles may cause discomfort. The allowable T difference is 3 ◦ C, which applies to situations where the T 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 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 TA , < 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.

216

D. Camuffo et al. / Journal of Cultural Heritage 11 (2010) 205–219

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 to 70%. As the requirement was originally defined for working sites which are characterized by mild T (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 T. The skin T 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 T was calculated as an area weighted average [25]. 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 T of people adequately clothed for cold environments to define the desired level of the thermal comfort rather than the TA . The skin T 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 T of 30 to 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 T drops near to, or below 0 ◦ C. An increase in the power

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.

of the heat sources in the pews 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 integrative 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 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 to specify these values was proposed in the 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.

D. Camuffo et al. / Journal of Cultural Heritage 11 (2010) 205–219

217

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.

Fig. 15. Rocca Pietore church. Percentile distribution of the daily variability in air temperature (a) and RH (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.

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 fluctuations is calculated as the difference between each RH reading and the corresponding smoothed value of the running 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 (Figs. 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

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 average (the RH target level) is marked by a horizontal line.

218

D. Camuffo et al. / Journal of Cultural Heritage 11 (2010) 205–219

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 eighth and the 92nd percentiles of the fluctuation amplitudes, respectively.

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.

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. 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 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 TA , incoming radiation and convection to measure the effective T 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 temperature 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 having 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 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 T to define the desired level of the thermal comfort. The 30 to 33 ◦ C mean skin T, 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 places of worship. Acknowledgements This research was carried out within the FH 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, Directors of the National Office for the Preservation of the Ecclesiastic Cultural

D. Camuffo et al. / Journal of Cultural Heritage 11 (2010) 205–219

Heritage of 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. References [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, Fraunhofer IRB Verlag, Stuttgart, 2007. [6] D. Limpens-Neilen, Bennch 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

[13]

[14] [15]

[16]

[17]

[18]

[19] [20] [21]

[22] [23]

[24]

[25] [26]

[27]

[28]

219

deposition: the case of churches situated in a cold climate, J. Cult. Herit. 8 (2007) 7–12. 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. A.M. Spiazzi, G. Toffoli, Rocca, Pietore, La chiesa di Santa Maria Maddalena, Canova, Treviso, 1993. 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. AHSRAE Standard 55: Thermal Environment Conditions for Human Occupancy, Am. Soc. Heat., Refrigerating and Air-Conditioning Engineers, 2004. 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. 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. Italian Standard UNI 11131, Cultural heritage – Field measurement of the air humidity, Ente Nazionale Italiano di Unificazione, Milan, 2005. D. Camuffo, Microclimate for Cultural Heritage. Developments in Atmospheric Science 23, Elsevier, Amsterdam, 1998. 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. J. Toftum, R. Nielse, Draught sensitivity is influenced by general thermal sensation, Int. J. Ind. Ergon. 18 (1996) 295–305. 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. 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. J.D. Hardy, E.F. DuBois, The technique of measuring radiation and convection, J. Nutr. 15 (1938) 461–475. 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. Draft European Standard CEN/TC346 prEN 15759, Conservation of Cultural Property – Specification and control of indoor environment – Heating of places of worship, European Committee for Standardisation, Brussels, 2008. 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.