Impact of Envelope Structure on the Solutions of Thermal Insulation from the Inside

The subject of the authors’ research involves timberframed wall structures with a ceramic brick filling, referred to as Prussian wall. An example of such a half-timbered structure is the building of the existing State Music School, dated the beginning of the 20^th century, located in Gliwice (Fig. 1). The frame structure was applied in the walls of the first floor, and the ground floor was made entirely of a brick wall. The rooms on the first floor contain a library and music rooms for students to practice.

Figure 1.

The structure of the envelopes made in Prussian wall technology in Silesia depended on where they were placed in the building. Residential and utility floors of multi-family buildings and public utility buildings (further floors) were made with the thickness of one brick. The timber frame (the most commonly used cross-sections of studs: 12/12 cm – 16/16 cm) visible from the façade’s side was filled with bricks and covered from the inside with a layer of bricks laid flat. The thickness of the envelope was approx. 25 cm. In unused attics or staircases, the walls were made with the thickness of 1/2 brick (12 cm). The fields within the wooden frame were left as brickwork structure or they were covered with layered plaster, often textured.

Due to the historic character of many buildings, thermal insulation must be placed from the inside. Such a solution is not efficient in view of building physics. High thermal resistance of thermal insulation brings about a significant drop in the temperature of the wall from the inside and increases the risk of interstitial condensation of the diffusing water vapor. The research shows that with the mass moisture of wood at the level of over 20%, being persistent over longer time periods, there is a risk of biological corrosion [13, 14]. The extent of such moisture significantly increases insulation thickness [12, 13]. It is assumed that in the climatic conditions of Poland, the insulation thickness should not exceed 8 cm. The optimal thickness is 3 cm [13, 14].

When using interior wall insulation, we can choose three main solution concepts (Fig. 2):

Insulation from the inside, preventing the development of condensation.

The literature [15] recommends that the value of the diffusively equivalent air layer thickness sd of the thermal insulation or the applied vapor barrier should exceed 1500 m.

Figure 2.

The standard DIN 4108-3 [15] permits the use of diffusion resistance materials for which the diffusively equivalent air layer thickness s_d is contained between 0.5 and 1500 m. Such large variation in the value of s_d affects ambiguously the assessments involving the suitability of the applied thermal insulation.

DIN 4108-3 [15] permits the use of materials for which diffusively equivalent air layer thickness s_d is lower than 0.5 m. The thermal insulation materials used in such solutions are capillary-active and enable the developed condensate to accumulate in the material structure, without deterioration of their physical properties.

The systems with vapor barrier from the inside prove most effective in buildings with high humidity. Since the diffusion of water vapor through the surface is completely blocked, possibly the highest efficiency of the ventilation system should be ensured.

The envelopes in the Prussian wall system are characteristic of the presence of gaps between the frame and the adjacent elements, allowing the penetration of rainwater into the wall (Fig. 3). Studies have shown [17] that it can penetrate the wall to the depth of 20 cm. Such a phenomenon was described by Kozakiewicz [18], who also pointed to the aging processes and slow losses in the cross-section of wooden elements.

Figure 3.

The properties of external wall coatings in terms of rain protection are defined by their water absorption coefficients, equivalently thick layer of air s_d and the product of both these quantities C_RP [17].

Initially, the suggested C_RP value was to be 0.1 [kg/m·h^0.5] [19]. However, most products and materials were not able to meet such criteria, and therefore the value was increased to 0.2 [kg/m·h^0.5]. For envelopes insulated from the inside, WTA guidelines recommend that C_RP should not be higher than 0.1 [kg/m·h^0.5] with the W_w value not higher than 0.2 [kg/m²·h^0.5] and s_d lower than 1 m [17].

The explained earlier limit values and requirements for W_w and s_d refer to water-resistant plaster and paint systems as defined in part 3 of the standard DIN 4108 [15].

In 2009, the document WTA – Merkblatt 6-4 2009-05 [20] has been published in which selected issues concerning the design of thermal insulation from the inside were presented. Such solutions lead to coolingout of the structural part of the wall, and hence they reduce the drying-out process of the existing structure. The presented procedure allows, in a simplified way, to estimate the correctness of the choice of a material solution, in terms of water absorption of the existing envelope and its outer layer. The first element of the procedure is to determine the impact of rain on the outer coating of the wall, i.e. to determine its water absorption expressed in W_w [kg / m² h^0.5]. If there is sufficient protection against rain in accordance with DIN 4108 [15], (part 3), it is usually sufficient. If the condition is not met, the following diagram is applied (Fig. 4) [20].

Figure 4.

However, the diagram – methodology [20] can only be applied if:

The envelopes insulated from the inside require a detailed assessment of their interaction with additionally designed thermal insulation layers in terms of durability of the entire structure.

The preparation of project documentation should be preceded by a thorough examination of the building and its elements. The authors made a number of research studies for the investigated building. Some of them are presented in this work. Basing on the said studies, and in accordance with the recommendations presented in section 2.2, the material for thermal insulation of the walls was selected.

Figure 5.

The measurements were made in three opencasts, at three heights, after the removal of plaster off the brickwork. The measured surface humidity was the highest close to the flooring, and it decreased with the height. On the façade, damage to the bricks was identified, which is characteristic of progressing capillary action. The strongest damage was observed by the ground and up to the height of about 150 cm. The measurements confirmed the diminishing level of moisture (17% by the ground, 2.5% at the height of 175 cm). Also brick brittleness and discolorations were found. The lowest bricks were chipped. At moistened places, biological growth was visible. Similar damage occurred in the zone near the ground on the remaining façades.

The research was conducted for the front façade. The measurement was carried out for the wall part of the wall. The study used 4 Karsten samples in two measuring series.

The measurement defined the extent of water penetration in the plane of brickwork. The measurement results for the brickwork are 5–7.3 (kg/m²h^0.5).

It is often impossible to determine all material parameters required in the assessment. This is due to restricted sampling possibility of materials (in the case of historic buildings, etc.) and research costs. For this reason literature data is applied in the modeling process of transfer phenomena in the envelope.

Figure 6.

Climatic conditions during measurements

During the tests, both the outside air temperature and the temperature inside the building were measured before and in the course of the tests.

Measurement performed: December 18, 2014.

Outside ambient temperature, t_e: 5.5°C–5.0°C

The temperature inside the building, t_i: 20.0°C–22.0°C

Thermovision measurements were performed at a temperature difference, Δt = 14.5–17.0 K.

During 72 hours preceding thermal measurements of the building, favorable weather conditions prevailed:

The obtained temperature difference between external and internal air (about 15 K) meets the requirements of the PN-EN 13187 standard – the minimum temperature difference at the level of Δt_min> 10 K. Thermovision tests were conducted from 7 pm to 9 pm.

Analysis of the obtained thermograms showed no gaps or cracks at places where wooden elements got into contact with the wall for the investigated building.

On the basis of the carried out research and visual inspection of building elements, following the design recommendations for thermal insulation of external walls, light cellular concrete was selected. The thickness of the thermal insulation material was adopted in the following variants: 3, 6, 8 cm. For comparison, the modeling results for a wall insulated with polyurethane foam of the thickness of 3, 6, 8 cm have been presented. The parameters of the material are presented in Table 1.

The modeling tests were carried out on the basis of hygrothermal simulations for transient boundary conditions, using the software WUFI 2D. [23] From the described physical principles of heat and moisture transport a closed differential equation system can be developed with which the moisture behaviour of multi-layered building components can be calculated under natural climatic boundary conditions. Since it is a non-linear equation system whose coefficients are greatly dependent on the potentials, an analytical solution is not possible. Described in detail below is the derivation of the coupled equation system and the numerical solution technique which forms the basis for the computer program called WUFI [23]. For non-stationary flows and conditions of two-dimensional heat transport and mass equations, they have the form:

where

The program has its own material database comprising basic types of building products. Some specific information involving physical features of the envelope components was supplemented basing on measurements.

Simulation tests were carried out for Prussian wall envelope – a building envelope with a brick filling 25 cm thick, covered with a lime-cement plaster 2 cm thick, 14/14 cm wood studs, the interior covered with lime plaster, 2 cm thick, insulated from the inside with light cellular concrete blocks of the thickness 3, 6, 8 cm. Material parameters were adopted from the WUFI database: for historical bricks and hardwood. The distance between the studs was determined on the basis of inventory measurements of the building.

The initial moisture content of the materials was accepted as for the equilibrium state. The parameters of the indoor climate were determined as for average-humidity living quarters based on the standard EN 15026:2007. External climate conditions for the area of Upper Silesia (from the meteorological station in Katowice), were determined using outside temperatures, air humidity, atmospheric precipitation and insulation. For the northern wall, the influence of the slanting rain was taken into account (according to the most frequent wind direction for the region of Silesia). A three-year period of hygrothermal simulations was adopted.

The results of the calculations made in the WUFI program are presented in a graphic form. The changes in the water content [kg/m³] (Figs. 7–13) in the selected planes of the insulated envelope were determined.

Figure 7.

Figure 8.

Figure 9.

Figure 10.

Figure 11.

Figure 12.

Figure 13.

In the presented case (Figs. 7–12), the analysis of the proposed solution did not confirm the accumulation of moisture inside the envelope for a given calculation period. Both the application of porous concrete materials (Figs. 7, 8, 9) and barriers in the form of polyurethane panels (Figs. 10, 11, 12) did not raise the water content of the envelope. It should be mentioned that the envelope modeling was carried out assuming full tightness of the external wall (its facade layer). The graphs demonstrate (Figs. 8, 11) that the water content in the cross-section of wooden element exceeds 100 kg/m³ and shows a downward tendency – drying out to the level of ∼75 kg/m³. With respect to the biodegradation of wood, the moisture content in wood should not exceed 20% of the volumetric weight. Slightly higher values are permitted for a period of 6 months.

Fig. 13 presents changes of water content in the particular layers of the 12 cm-thick walls insulated with 8 cm-thick thermal insulation material (lightweight cellular concrete). Water content in wooden elements exceeds 20% (∼130 kg/m³ in the winter season). In the summer season, the level of water content drops to 90 kg/m³. We must add that such a situation becomes disturbing with heavy rainfall. Rainwater can penetrate to the depth of about 20 cm [13, 14, 17] and can extensively dampen the envelope, resulting in the degradation of its elements. In such a situation the insulating material itself (light cellular concrete) loses its insulation qualities since the pores are filled up with water.

One of the criteria needed to ensure an effective design of thermal insulation from the inside is the efficient elimination of thermal bridges. As part of the carried out analysis, an exemplary thermal bridge was presented in the form of Prussian wall corner (Figs. 14, 15). The assumptions given in Table 2 were adopted for the calculations. The modeling of joints was made in accordance with the principles of ISO 10211: 2017 [24].

Linear coefficient of thermal coupling L^2D [W/(m·K)]

Figure 14.

Figure 15.

The following qualities were accepted as boundary conditions: temperature of outdoor air t_e: -20°C, temperature of indoor air t_i: +20°C, thermal surface coefficients h_e = 25 [W/(m²K)], h_i =7.69 [W/(m² K)], for calculations f_Rsi: h_i =4 [W/m²K] was assumed.

The research results are presented for a 25 cm-thick wall with the 8 cm-thick thermal insulation.

The obtained results (Tab. 2) demonstrate that there is a risk of surface condensation and mold growth for the non-insulated wall. The value of the temperature factor f_Rsi is lower than the permissible value of 0.72 [1]. For the thermally insulated wall the risk is eliminated, but we can observe a high impact of external temperatures on the structural part of the wall (masonry).