Controlled Ventilation of Historic Buildings: Assessment of Impact on the Indoor Environment via Hygrothermal Building Simulation

Abstract

Historic buildings, when they are unheated, often face problems of summer condensation. After the cold season warm, humid air enters the building and condenses on the walls, a problem that can occur during the entire warm season. This leads to moisture related problems such as mould or algae growth on building surfaces. High humidity can also damage works of art inside the building. One possibility to lower the level of relative humidity is to ventilate the building every time when the dewpoint outdoors is lower than indoors. On the other hand, the resulting humidity fluctuations may also cause damages on works of art and bringing in cold air will further lower the temperatures of the whole building and its walls thus being counter productive to the drying process. An automatic system for ventilating historic, unheated buildings is assessed using building simulation software (WUFI®Plus) on two case studies, the St. Margaretha church in Roggersdorf, near Holzkirchen, Germany and the Gatehall of Lorsch, Germany.

1 Introduction

A high number of unheated historic buildings all over Europe show problems with mould or algae growth. This is very often due to summer condensation. Several options to reduce the risk for mould or algae growth exist. One is, to dehumidify the air to keep the relative humidity inside the buildings on a low level. But dehumidification leads on one hand to a high energy demand; on the other hand it is very often not possible in the affected buildings due to missing space for the equipment and/or missing power supply for operation. Another option is using a controlled ventilation system. Controlled ventilation systems measure the absolute humidity inside and outside the building. If the absolute humidity outside is lower than inside, there is potential to dry the building by an increased ventilation.

In this article two buildings in Germany are assessed. Both show moisture related problems. A hygrothermal whole building simulation software, WUFI®Plus, is used to assess the building temperature and humidity conditions with different ventilation strategies.

2 Humidity Control in Historic Buildings and Hygrothermal Whole Building Simulation

Adaptive ventilation is used to lower the relative humidity in buildings. The aim is usually to reduce the risk for biodeterioration. Adaptive ventilation means to ventilate only when the exterior absolute humidity level is lower than the interior absolute humidity level. This way humidity can be removed from the inside. On the other hand it is important to stop the ventilation if the exterior absolute humidity is higher than the interior to avoid an additional moisture supply into the building. Furthermore it is important to also measure the temperatures and stop ventilation if the exterior temperature is too low to reduce an additional cooling effect or freezing due to ventilation with exterior air. Thus a mechanical fan and damper with temperature and humidity sensors attached to a control unit are required to conduct controlled ventilation.

For the assessment of mould growth the times where a combination of temperature and relative humidity prevail are critical (Sedlbauer 2001). As a general rule of thumb the relative humidity should not exceed 80 % RH at room temperature. Algae growth usually requires liquid water on the surface. This means, that the dew point temperature on a surface is higher than the actual temperature. But not only mould and algae growth are critical for the assessment of historic buildings, also the fluctuations of the conditions, especially of relative humidity, is important. According to [Holmberg 2001] the daily relative humidity fluctuations should be kept below 15 % RH.

Broström et al. (2011) assessed the application of adaptive ventilation on a case study building in Gotland, Sweden. They concluded that the adaptive ventilation has had a significant drying effect and that the mould risk was kept at an acceptable level for most of the times. Even the short term RH variations were acceptable. But they also mention that further research is required to understand the effects of hygrothermal inertia and the integration of heating and dehumidification.

These parameter variations can be studied by using hygrothermal whole building simulation software. The advanced hygrothermal whole building simulation tool WUFI®Plus is used for simulation (Holm et al. 2003). This software couples whole building energy modeling with hygrothermal component modeling. It calculates the coupled heat and mass transfer, i.e. mass transfer by diffusion and liquid water transport, for every enclosure assembly and couples heat and moisture fluxes from the inner surfaces with the zone temperature and relative humidity. This model allows the combined assessment of hygrothermal conditions of the building envelope, indoor climate and energy demand of the building. It requires in general hourly exterior climate data, which is provided by the climate database that comes with the software or by measured data. Materials are selected from a material database or specified by providing thermal and hygric properties, so that assemblies can be built from one or more layers of different materials. As the indoor climate is a result of the simulation, appropriate assumptions for inner sources and set-points need to be made. The set-points can be met by specifying an ideal HVAC system. A predefined ventilation control algorithm is available that allows the application of an adaptive ventilation strategy.

3 Case Study Buildings

For the simulation of the hygrothermal building behaviour the software tool WUFI®Plus was used (Holm et al. 2003). In the following the general conditions, the assumed building geometry and materials as well as the heating, ventilation and air conditioning operations are described.

3.1 Church of Roggersdorf

After a complete renovation of the chapel of ease St. Margaretha (see Fig. 1) in September 2004, the churchwarden again noticed moisture damage on the walls. A subsequent climate measurement showed that the damage was a matter of condensation that occurs mainly in the transitional period during spring-time. At that time of the year the building is still cold because of the winter. If warm, humid air enters the building due to natural air flow or uncontrolled ventilation, it condenses on the cold wall surfaces. But even in the summer and fall warm humid weather conditions can cause condensation problems.

Fig. 1

Picture and screenshot of the computer model of the exemplary church

From December 2004 to August 2006 the temperature and relative humidity were measured inside and outside the church. To asses condensation events the wall surface temperature was also measured on the Western wall at the joint to the floor. Weather data was available from the Fraunhofer IBP outdoor testing facility at Holzkirchen, only 5 km away from Roggersdorf. During the period when measurements were taken the climate inside the church showed a high average humidity with values over 75 % RH for more than half of the time of the year (see Fig. 2). During winter time freezing conditions with temperatures below 0 °C are recorded for more than six weeks in a row (Kilian 2007). Subsequently, the church starts with very low wall temperatures into spring time and the warm season.

Fig. 2

Temperature and humidity in the church during 2005. Half of the time RH lies above 75 %, during winter the indoor temperature falls below freezing conditions for 6 weeks

Another problem that became evident was the uncontrolled opening of windows and doors over the warming period of spring that was supposed to reduce the moisture levels in the space by ventilating. In the summer 2005 “ventilation traffic lights” showing the times when the water content of the outdoor air was lower than indoors were used to give advice to the warden of the church. This significantly increased the daily fluctuations of RH; in the spring and summer 2005 the fluctuations were above 15 % RH for more than 30 days. Daily changes above 15 % RH are thought to be critical to works of art (Holmberg 2001), as the risk of structural damages due to swelling and shrinking of materials increases with the range of RH change per day. Also the newly restored altarpiece from the nineteenth century started to show additional damage to the gilding.

As a consequence, the installation of a controlled ventilation system was discussed with the church authorities. Such systems permit to compare the absolute humidity inside and outside the building. Whenever the outer humidity is lower, a ventilator starts bringing dryer air into the building. Also, boundary conditions for the lower temperature can be given, as well as a maximum allowed range of RH change in the last 24 h. Practical experience has shown that this type of ventilation system can help reduce the RH range and reduce the overall moisture in a room over the year, as was done for the church located in Urschalling, Bavaria (Künzel 2009). However, the open questions are: how good is the system for the respective location; what can it achieve; how much energy does it use; and, what is the overall impact when bringing colder air into the building during winter and in summer nights? To assess the possibilities and limitations of these systems, whole building simulation was used to predict the indoor environment in the church using weather data from Holzkirchen.

3.1.1 Boundary Conditions

The time of calculation lasted from January 1st to December 31st of the year 2005. The simulation was carried out in time steps of 5 min to reproduce a realistic ventilation control. Given that the location of the church was in Roggersdorf, which is located near Holzkirchen, the meteorological data of Holzkirchen for the year 2005 was used for the simulation. A weather file was created which consisted of 5-min values of exterior temperature, relative humidity (both in Fig. 3), global and diffuse radiation, rain, wind and barometric pressure. Additionally a file with ground temperatures at a depth of 0.5 m in the soil was generated to take into account conditions below the foundation of the church. Inner loads were not taken into account for the base case, assuming that there is no church service.

3.1.2 Building Geometry and Material Data

A detailed model of the chapel’s geometry was created in WUFI®Plus (Fig. 1). The main building is about 12 m long and 9 m wide. The height from the ground to the top of the roof is about 6 m and to the top of the tower about 9 m. For the simulation the main body was divided into three zones where the climate was modeled: the nave, sacristy and the attic. A mechanical ventilation system in the nave will provide outside air to this zone with an air exchange rate of 5 h⁻¹, as long as the outside absolute humidity is lower than the absolute humidity inside and the exterior temperature remains above 0 °C. The infiltration ventilation air was taken from outside, with an infiltration air exchange rate of 0.4 h⁻¹ for the nave. The tower as well as the entrance area was treated as an attached unheated zone with assumed exterior temperature and relative humidity in the zones.

The outer walls of the chapel are built from sandstone with a thickness of 0.52–0.77 m. At the inner surfaces there is a layer of lime plaster which is 2 cm thick. The base plate consists of a layer with loose material (0.15 m) which is mainly covered by natural stone plates with a thickness of 5 cm. Under the benches at the north and south side of the church there are panels of hard wood instead. The ceiling to the attic has an overall thickness of 0.3 m consisting of 2.5 cm softwood, a 15 cm thick air layer, 10 cm mineral wool and again 2.5 cm softwood. The chapel’s roof is mainly made of softwood with shingles on the outside. The windows have a single glazing with an overall thermal transmission value of 3.7 W/m²K.

3.1.3 Parameter Variations

The simulations were run with different parameter variations. The base case is a simulation as described in the previous section. The simulation results of this variant are always compared to a variant without any ventilation control, i.e. only natural ventilation mainly by infiltration. Further simulations were run with air change rates of 1 h⁻¹ and 10 h⁻¹ to show the influence of the capability of the ventilation system. A further simulation was run where only the energy transport was modeled whereas the coupled equations for heat and moisture transport in the building envelope were not modeled. This simulation allowed assessing the necessity of a simulation model capable of modeling the coupled transport equations.

3.1.4 Results

Figure 3 shows the modeled temperature and relative humidity inside the nave without additional ventilation and with an assumed air change rate of 5 ACH for ventilation during all periods with lower exterior absolute humidity than inside and while exterior temperatures are above 0 °C. The inner climate conditions are compared to the outer climate conditions.

Fig. 3

Temperature and relative humidity outside and modelled conditions inside the church during the whole simulation period with (Air Exchange 51/h) and without mechanical ventilation

Expectedly lower fluctuations for temperature as well as for relative humidity are found for the interior conditions compared to the outside climate. Not only the peaks are damped, but also in longer periods with very low exterior temperatures the inner temperatures do not reach the mean low temperature. The fluctuations are lowest in the model without additional mechanical ventilation. Especially in spring, the relative humidity is reduced in the controlled ventilation case, but also the temperature inside the church. No effect is found in the cold month due to the temperature limitation at 0 °C of the ventilation control.

The modeled ventilation control acts on absolute humidity. A comparison in Fig. 4 for absolute humidity inside and outside shows also a dampening of the fluctuations from outside to inside for the case without ventilation. The general trend is followed pretty close. The above mentioned effectiveness of the ventilation control in the cold month is reflected by the related ventilation status. Almost no ventilation actions are performed from mid November until end of March. In the warm months the peaks in exterior absolute humidity appear reduced at the inside. The benefits of ventilation control become evident by comparing the absolute humidity with and without controlled ventilation from April to November.

Fig. 4

Absolute humidity outside and inside the church as well as ventilation status during the whole simulation period

A closer look to the absolute humidity conditions in October is given in Fig. 5; details are provided of the resulting higher fluctuations of absolute humidity inside the building because of active ventilation control. But it also clearly shows that lower interior absolute humidity’s are achievable with effective ventilation during periods with lower exterior humidity.

Fig. 5

Absolute humidity outside and inside the church and ventilation status for simulation in October

A variation of different air change rates as well as a simulation without moisture buffering—standard building energy simulation—was performed and the results for October are shown in Fig. 6. An improvement in efficiency is achievable by increasing the air change rate during control action from 1 to 5 ACH. In the modeled case, a further increase up to 10 ACH does not further reduce the absolute humidity indoors, as the exterior minimum absolute humidity is also reached with 5 ACH. In the case of a simulation without modeling the coupled heat and mass transfer in the building envelope but using only a mass balance for interior humidity, the interior absolute humidity follows very closely the exterior absolute humidity.

Fig. 6

Absolute humidity outside and inside the church with different parameters for simulation during representative month October

3.1.5 Discussion

It was found, that daily fluctuations of temperature and relative humidity are higher with mechanical ventilation control compared to the climatically free floating church. Especially in cases with valuable interior artifacts that are sensitive to high RH fluctuations, this may cause problems. The daily fluctuations are compared in Fig. 7 for the case without mechanical ventilation and the ventilation control case with 5 ACH during operation of the mechanical ventilation. The average daily variation is 3 % RH without and 7 % RH with ventilation with maximum daily fluctuations of 15 and 39 % RH respectively.

Fig. 7

Boxplots of simulation results for all daily fluctuations of relative humidity in the space without ventilation and the space with ACH 1/h

3.2 Gatehall Lorsch

The Torhalle (“Gate Hall”) of Lorsch, also called King’s Hall, is one of the best preserved examples of Karoling architecture in Germany, dating back to 760 A.C. The hall is located on the first floor above an entrance gate to a monastery and is believed to have been used as a church originally. It features valuable Romanic wall-paintings. In the 1980s thorough investigations were conducted on the building and its indoor environment which shows high relative humidity all year round Kießl (1987). In 1991 a controlled ventilation system was introduced that promised to lower RH significantly. Already in 1992 it was turned off by a sceptical conservator. Now the building faces the same problems as before requiring new investigations including collecting climate data and evaluating past experience. The aim is to develop a new climatisation strategy for the Gatehall. For this purpose the whole building simulation tool WUFI®Plus is applied, to assess the use of controlled ventilation for the specific location at Lorsch and also to research the possible effects on indoor environmental fluctuations that might be harmful to the ancient wall-paintings. The King’s Hall is very well documented regarding its historic building materials and past climate. Also, its special construction, sitting above a gate and thus being exposed to exterior climate from all sides without any side effects—like rising damp—makes it especially interesting for whole building simulation. This research is part of the European project Climate for Culture on the effects of climate change on cultural heritage.

The building is a two-story building (see Fig. 8), which has in the lower part a pathway to the monastery church with three big round arches. Above it is the Gatehall, a room with eight small round-arch-windows, which is accessible from both sides trough stairway-towers from the pathway. The Towers are closed down- and upstairs by wooden doors which are probably not airtight. Under the steep roof the room closes with a barrel vaulting, the roof beams are visible. Especially the ancient plaster and fragments of wall-paintings represent the value of the inner surfaces. The facade is designed following the example set by the Romans in a varied “opus reticulatum” technique with coloured natural stone.

Fig. 8

Picture of the Gatehall of Lorsch

3.2.1 Boundary Conditions

3.2.1.1 Visitors

From the visitor statistics of the last 10 years provided by the Hessische Schlösserverwaltung a generic visitor profile has been specified for the year 2008. Factors for every day of the year and every hour between visiting hours from 10 am to 4 pm were calculated to predict the part of the yearly number of visitors. Depending on the seasonal clothing the generic visitors represent different heat and moisture loads.

3.2.1.2 Weather Data

On-location the weather was only measured by a simple temperature and humidity data logger, which delivered no satisfying information for use in building simulation. To guarantee the usability of the weather data from station in Bensheim, 5 km away from Lorsch, both datasets were compared and were almost identical.

3.2.1.3 Ventilation

The air change rate was measured in a former research project. After an interview with employees of the monastery it became clear, that currently at this time neither controlled ventilation nor any kind of ventilation strategy is established. Windows and doors are opened and closed completely arbitrarily by the employees.

Table 1 shows the results of the measured air change rate in the 1980s. The researchers proceeded in different steps, gradually sealing all openings. The air change rate is low, only 0.4 h⁻¹ for normal conditions. It is reduced to 0.26 h⁻¹ if only the doors are sealed and to 0.08 h⁻¹ if also all windows are sealed.

Table 1 Measured air change rates in the Gatehall of Lorsch (Kießl 1987)

3.2.1.4 Outdoor Climate

The grey lines in Fig. 9 show the outdoor temperature and relative humidity in Lorsch. Year 2008 shows a moderate climate, where the outdoor temperature falls rarely below 0 °C and never below −10 °C. The outdoor relative humidity fluctuates considerably, in the summer from <40 % up to 100 % RH. In winter the lower limit is a little bit higher, around 60–80 % RH.

Fig. 9

Measured inner and outer temperature and relative humidity for the years 2008–2010 for the Gatehall Lorsch

3.2.1.5 Indoor Climate

The indoor climate was measured over the last 3 years in several shorter time periods. Through the free natural ventilation of the King’s Hall the indoor climate follows damped the outdoor climate. In Fig. 9 one can see also, that the indoor temperature follows the upper region of the outdoor temperature, but the indoor relative humidity follows the lower region of the outdoor humidity. Altogether the indoor relative humidity is around 40–80 % RH in summer, but sometimes clearly below. In winter the relative humidity is around 80–100 %.

3.2.2 Parameter Variations

To compare the influence of natural ventilation and controlled ventilation the air change rate was modified in several simulations. A simulation was run with a constant air change rate of 0.0 h⁻¹. Further simulations were run with controlled ventilation and an air change rate 1.0, 5.0 and 10.0 h⁻¹.

3.2.3 Results

In Fig. 10 you can see the difference between an airtight building with an air change rate 0.0 h⁻¹ and a building with a controlled ventilation air change rate 1.0 h⁻¹. Even with this low air change rate the level of relative humidity is most of the time lower than without an exchange. The relative humidity nearly fluctuates with the outdoor relative humidity. You can also see, that both the range of the daily fluctuations and the seasonal fluctuations of the relative humidity increases with a controlled ventilation.

Fig. 10

Simulated inner temperature and relative humidity with an air change rate 0.0 h⁻¹ and with a controlled ventilation air change rate 1.0 h⁻¹ for the year 2008

Figure 11 shows the absolute humidity and the ventilation status for the airtight and the ventilated building. The controlled ventilation can decrease the absolute humidity in the Gatehall by 1–2 g/m³ with an air change rate of 1.0 h⁻¹ over the whole year.

Fig. 11

Simulated absolute humidity and ventilation status with an air change rate 0.0 h⁻¹ and with a controlled ventilation air change rate 1.0 h⁻¹ for the year 2008

In the detailed Fig. 12 it is obvious that the absolute humidity with controlled ventilation is always below the air tight simulation. But the difference between indoor and outdoor absolute humidity is small. In times when the absolute humidity inside is lower than the outside conditions the ventilation is off.

Fig. 12

Simulated absolute humidity and ventilation status with an air change rate 0.0 h⁻¹ and with a controlled ventilation air change rate 1.0 h⁻¹ for two weeks in August 2008

Higher controlled ventilation air change rate causes lower absolute humidity in the Gatehall. The biggest difference shows the comparison between air change rate 0.0 h⁻¹ and air change rate 1.0 h⁻¹, but also between air change rate 5.0 h⁻¹ and air change rate 10.0 h⁻¹ a difference is still discernible (see Fig. 13).

Fig. 13

Simulated absolute humidity for several air change rates for two weeks in August 2008

The level of absolute humidity decreases with higher air change rates for the controlled ventilation. But also the inner temperature decreases. This causes higher fluctuations of the relative humidity (see Fig. 14). The difference can be up to 10 % RH between air change rate 1.0 h⁻¹ and air change rate 10.0 h⁻¹. A specific limit of the ventilation has to be found for every situation and building.

Fig. 14

Simulated inner temperature and relative humidity for several air change rates for 2 weeks in August 2008

3.2.4 Discussion

A high salt load in the plaster and stone of the walls of the Gatehall was found in the research project in the 1980 and 1990s. The salt was examined and it came out, that it was nitrate of ammonium, brought in by pigeon excrements. The deliquescence relative humidity (RHD) can be calculated with a simple model for this salt (Nieveler 1998). Approximately one-third of the measurement points are above the limit. So there is the assumption, that temperature and relative humidity will change inopportunely, that the salt in the walls will get in solution and crystallize several times in a year. This leads into progressive damage like chipping off pieces of plaster or paintings, described in the earlier research projects.

Under the assumption that the surface conditions are close to the temperature and relative humidity conditions inside the room Fig. 15 shows that most of the time steps for the no ventilation case conditions above the RHD are found. In this case the salt in the plaster is in solution. With a ventilation control the range of relative humidity is bigger than without. This could end up in more phase changes of the salt and more damage. In this case the first thoughts of reducing the humidity to save the wall paintings may cause more damage by salt phase changes than it could avoid.

Fig. 15

Simulated inner temperature and relative humidity for an air change rate 0.0 h⁻¹ and a controlled ventilation air change rate 1.0 h⁻¹ for the year 2008

4 Summary and Conclusions

A simulation model of a church and another unheated historic building was implemented. Parameter variations were undertaken, to assess the effectiveness of controlled ventilation to lower humidity conditions inside the buildings. A model of the free floating buildings without mechanical ventilation was compared to the same building with a mechanical ventilation system. This ventilation system was controlled to be active while inner absolute humidity was higher than exterior absolute humidity with the limitation that exterior temperature had to be above 0 °C.

The results of the simulation permitted determining that it is possible to lower the inner absolute humidity during some times of the year with the ventilation control system assessed in this study. The main limitation of the system is that moisture removal can never reduce the inner absolute humidity to a level below exterior conditions. This limits its effectiveness in cases of a free floating building with low inner moisture loads.

Critical conditions due to moisture vented or infiltrated into the building leading to high relative humidity or even liquid water on the envelope surface cannot always be avoided. The daily cycle and other fluctuations in exterior absolute humidity are low. Potential for active ventilation is not always available. Therefore also ventilation systems that create low air change rates allow the full removal of excess moisture. A further increase in air change rate is not necessary. In comparable free floating cases with no inner loads humidity conditions can still be critical. Improvement in the indoor humidity level is bought at the expense of stabile humidity conditions. Fluctuations of temperature and humidity are much higher with controlled ventilation than without. This effect has to be balanced with the small improvements in moisture level. Even if an improvement in terms of moisture level is achievable it is to be assessed if these improvements really reduce possible damages or if it is counterproductive.

A more general understanding of adaptive ventilation as a dehumidification measure in unheated historic buildings is required. As for most of these buildings only low moisture loads prevail, the moisture excess inside the buildings and with it the capacity to remove moisture by controlled ventilation is usually low. Different results can be expected if higher moisture loads, e.g. due to building usage or moisture from the ground, are found. Further research is also required to assess the effects of the interaction of controlled ventilation and heating.

It was shown that it is critical to model coupled heat and moisture transport processes in the envelope and its interaction with the room to achieve realistic simulation results. Relative humidity level and fluctuations can only be represented if moisture buffering effects are taken into account in the simulation.