- Email: [email protected]
Thermal and energy performance evaluation of underground bunkers: An adaptive reuse approach
Sustainable Cities and Society 46 (2019) 101444
Contents lists available at ScienceDirect
Sustainable Cities and Society journal homepage: www.elsevier.com/locate/scs
Thermal and energy performance evaluation of underground bunkers: An adaptive reuse approach
T
⁎
Rudina Breçani , Sokol Dervishi Department of Architecture, Epoka University, Rruga Tiranë-Rinas, Km 12, 1039, Tirana, Albania
A R T I C LE I N FO
A B S T R A C T
Keywords: Thermal performance simulation Scenarios Energy Underground nuclear bunker Adaptive reuse
The quantitative research on bunker tunnels is extremely limited nowadays, though this typology of structure is vastly found globally. The present study follows up on the existing literature by providing an innovative insight on the subject: evaluation of the thermal indoor comfort and energy consumption through the reality of living in a bunker. The potential adaptive reuse of the unexploited extensive underground bunker network of Albania is investigated. The approach is illustrated using a 150-m long cross-section of the underground network selected for the parametric computational simulation on thermal and energy performance. The model simulation involves data collection on local climate, design typology, and building materials to generate a finite-element simulation model. Different scenarios including insulation of the outer walls, occupancy patterns, and ventilation regimes are examined in order to assess the respective thermal and energy performance. Two sets of simulations are held simultaneously: at 325 m above sea level (current altitude of the underground tunnels), and 350 m above sea level (the current altitude of the city above ground). Insulation does not affect the heat flux through the outer walls. Furthermore, the results establish the same building consumes 44-145% more energy when located above ground as opposed to underground.
1. Introduction A special kind of underground shelters can be observed in Albania, i.e. nuclear bunker tunnels. These capillaries, which were built in several cities during the communist period and never used to fulfill the purpose for which they were built (refuge in case of war), constitute a large underground network of concrete leftover space. Therefore, the question on the efficiency with which they can fulfill their original purpose i.e. that of sheltering, comes out naturally. The present study assesses the thermal and energy performance and evaluates the adaptive reuse efficiency of the underground shelters through computational simulation modeling, by using the tunnel network found in the city of Kukës, Albania, as a base for conducting simulations. Past researches on underground spaces have tapped into various topics. Underground sheltering has been subject to various analyses throughout the years, ranging from researches on the psychology and behavior of isolated humans (Poulios, 2009; Romanova, 2016), urban planning (Admiraal & Cornaro, 2016; Delmastro, Lavagno, & Schranz, 2016), to the adaptive reuse of mines, railways, and other similar underground facilities after their activities disconnect (Sasmito, Kurnia, Birgersson, & Mujumdar, 2014). Li, Yuan, Li, Han, and Zhang (2018) have monitored the physical reactions of people while staying in refuge ⁎
chambers, concluding that the number of people in the enclosed space directly affects the humidity level and temperature of the said area; the heart rate and temperature of the individuals subjected to the experiment remained constant (Li et al., 2018, Delzendeh, Wu, Lee, & Zhou, 2017). However, he recognizes the enclosed space, constructed specifically for the study, is small and the prejudice that accompanies the concept of underground is major Li et al. (2018). The studies on the computational evaluation of either thermal or energy performance for such spaces are fewer in number. Zhu has carried out thermal simulations on a sample of atrium plan earth sheltered buildings with additional heating systems named Yaokang concluding that the buildings constitute a step forward towards low energy building design (Zhu & Tong, 2017). Another thermal performance simulation and evaluation study has been carried out by Ip, but the building in question has only one wall in contact with the earth (Ip & Miller, 2009). Tan has investigated the thermal comfort conditions in underground spaces in four major Chinese cities characterized by different climate conditions in order to assess the influence of the specific climatic variables (Tan et al., 2018). He recognizes passive design strategies as effective measures in reducing energy consumption especially during transitional periods (Tan et al., 2018). Tan however also recognizes his study as only qualitative and evaluates a building energy
Corresponding author. E-mail addresses: [email protected] (R. Breçani), [email protected] (S. Dervishi).
https://doi.org/10.1016/j.scs.2019.101444 Received 26 August 2018; Received in revised form 19 January 2019; Accepted 19 January 2019 Available online 23 January 2019 2210-6707/ © 2019 Elsevier Ltd. All rights reserved.
Sustainable Cities and Society 46 (2019) 101444
R. Breçani, S. Dervishi
temperature is always higher than the air temperature during winter and lower during summer. Due to the solar radiation and the soil’s relative capacity, there is a 5-hour time lag a day regarding temperature distribution (Florides & Kalogirou, 2005). Whereas previous researches have provided significant insight on different topics related to underground structures, there is a lack of quantitative analyses providing close examination of thermal performance and energy consumption evaluation. Furthermore, the present study constitutes the first research on thermal and energy performance evaluation on the typology of underground bunkers, a phenomenon present not only in Albania, but in a global scale, and the number of similar studies worldwide is limited. Additionally, this research indicates and examines alternatives for low energy consumption adaptive reuse functions, appropriate for the design typology, and space quality of the tunnels. The biggest flaw of these constructions is simultaneously a great possibility for future development: the underground bunkers are massive leftover space which does not serve any function currently. For the context in which these tunnels are situated nowadays, the importance of the findings from this research relate to the significance they hold for the hosting country. Since the study aims to prove these tunnels are energy efficient structures which do not disrupt the city environment, a positive outcome and future step to this research would be their consideration and implementation into the country’s trading and city governance policies. A good example of this is the transformation of the underground network into international rentable space for goods, laboratories etc. Similarly to Albania, several other countries in Europe and wider share an analogous historical background, due to which the presence of underground bunker like structures is spread, and thus, the findings from this study could efficiently serve as a comparative case study for future research in an international context. It is also important to emphasize that thermal simulation and energy efficiency assessments are definite for every study. In the case of Albania, the specificities that influence the results of the study are: local climate, design typology, percentage of contact with the earth, and altitude above sea level. Appropriate functions noted by the proposals of previous studies, the characteristics of the adaptive functions, the design typology of the tunnels in Albania, as well as the urban composition and quality of the city of Kukës have been considered in order to decide on the modeling simulation scenarios. Two sets of software based simulations are held simultaneously for the same building and scenarios, in order to compare energy efficiency and temperature performance results: at 325 m above sea level (current altitude of the underground tunnels of Kukës), and 350 m above sea level (the current altitude of the city of Kukës above ground).
modeling would provide more solid data since qualitative analyses often depend on real life restrictions (Tan et al., 2018). Zhao, Peng, Wang, Zhang, and Jiang (2016); Tezuka and Seoka (2003), as well as Tan et al. (2018) investigate the potential integration of underground space examination in urban city planning. Tan et al. insist that underground planning should be an integral part of any National Regulation Plan since the active use of underground space can positively affect the natural city landscape (Tan et al., 2018). Shan et al. have identified through questionnaires the top four advantages of underground spaces that make these spaces attractive: space saving, improved indoor thermal comfort, resistance to external noises, and increased level of privacy (Shan, gang, & Wong, 2017). The major disadvantages they have found are limited amount of daylight and negative psychological effects (Shan et al., 2017), which are also supported by other researches tackling the psychological consequences on humans living in isolated underground spaces by Romanova (2016), and Poulios (2009). However, Shan et al. recognize that results might be affected by bias or prejudice. Furthermore, Tan et al. point out that sufficient natural light is not provided in many above ground buildings either (Tan et al., 2018). Alkaff et al. clarify the factors that most notably affect the energy conservation quality of the underground structures, such as: design typology, depth below ground surface, ventilation system, climate and soil thermal properties, insulation and infiltration, and altitude above sea level (Alkaff, Sim, & Ervina Efzan, 2016). The design typology affects the sun penetration into the building thus influencing its energy performance according to Alkaff et al. (2016) and Anselm (2012). A structure that is located deeper underground might need waterproofing and it benefits from superior temperature performance. As for the altitude above the sea level, the depth of 300–750 m above sea level results in a more comfortable temperature inside the underground structures (Paramananthan, 2000). Mechanical ventilation is a factor that could negatively affect energy performance in underground spaces due to the limited amount of natural ventilation, however, Hait (1983) and Anselm (2012) have offered cost effective solutions to such a problem, such as precooling the fresh air that enters the building. The ground temperature distribution is also affected by the physical properties of the ground, the local meteorological variables, and the surface cover of the said portion of the ground (Popiel, Wojtkowiak, & Biernacka, 2001). Some scholars have carried out scientific research with a purpose to create predictive formulas for the thermal capacity and heat flux of the soil. As Alkaff et al. point out, a common misconception depicts the earth as a good insulator when in reality the earth has high heat conductivity. Owing to that earth is usually extended in vast areas, the thermal energy loses transfer, and the temperature fluctuations lower with the increment of depth (Alkaff et al., 2016). Moreover, the heat capacity of the soil is considerably higher than that of the air. Florides and Kalogirou claim that because of the high thermal inertia of the soil, the temperature fluctuations at the surface of the ground are diminished as the depth of the ground increases (Florides & Kalogirou, 2005). According to Kajtar et al. the temperature of the soil changes with amplitude of 0.6 °C in the depth of 8 m and 0.2 °C in the depth of 10 m (Kajtar, Nyers, & Szabo, 2015). Popiel states that after the depth reaches more than 20 m the temperature becomes practically constant (Popiel et al., 2001). The formula depicted in Eq. (1) illustrates the ground heat flux formula used by him. Here, x is the depth below the ground surface, T is the temperature, k is the heat conductivity of the ground, and q the heat flux.
q = −k
∂T ∂x
2. Methodology 2.1. Tunnel description The underground tunnels used for building simulations in this study are part of the ex-secret military map of Albania, which has been disclosed to the public only recently, in 2014. The military reinforcements adopted by the communist Albanian government during 1945–1990 are present throughout the territory of the country in the shape of singular secluded mushroom-like concrete bunkers or in the shape of extended underground galleries designed for sheltering a large part of a city’s population. Thus, the latter design typology can be found in various Albanian cities. The case study buildings for conducting simulations can be found in Kukes, in North Eastern Albania. This network of tunnels includes 30 galleries which make up to 2400 m in length as established in Fig. 3. Each gallery is composed of a corridor with a row of rooms on one side. The rooms and corridors were built inside an outer concrete shell (Municipality of Kukes, 2000;
(1)
Florides and Kalogirou depict a time lag between the temperature fluctuations above and below the ground (Florides & Kalogirou, 2005). This explains why, after reaching a specific depth, the soil’s 2
Sustainable Cities and Society 46 (2019) 101444
R. Breçani, S. Dervishi
Fig. 4. Close-up partial axonometric drawing of the inner construction of the tunnels.
Fig. 1. Underground network of tunnels for Kukës (in red) together with the building fabric of the city (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article).
Fig. 5. Partial axonometric drawing of the section used for the computational simulations in the hypothetical situation it was located at 350 m above sea level.
level. The section, illustrated in Fig. 5, is enclosed on one side by a 25 m tall staircase which connects the existing section to the existing hospital of the city located above ground, whilst the other side of the corridor is a dead end. The section is surrounded by earth on all sides; the only connection of the gallery with the environment above ground is by the staircase that connects it to the hospital above. The corresponding hypothetical section of the same characteristics, when located at 350 m above sea level, is illustrated in Fig. 5. The two building sections differ in construction only in the presence of staircase well, and presence of windows. The energy loads delivered by the software provide a valuable insight into the hypothetical heating and cooling demand of such a structure. Nowadays, the underground network is unoccupied and it does not use any mechanical ventilation of HVAC systems. In the different scenarios tested during simulations, the energy loads inflicted on the building include mechanical ventilation, retrofitting HVAC systems, electricity supply systems, as well as lighting have been taken into consideration.
Fig. 2. Partial section of the underground network of tunnels for Kukës.
Polytechnic University of Bari, 2017). The internal part of the circumferential walls was also built with concrete, whilst the indoor walls were built with bricks. The tunnels have corridors which are 270 cm high and 120 cm wide, and rooms of an average of 3.4 × 3.4 m. The network is located from 319 m to 336 m above the sea level (the different levels are connected by staircases), whilst the ground level of the city of Kukës is located 350 m above the sea level (Municipality of Kukes, 2000; Polytechnic University of Bari, 2017). The outer walls of the tunnels are consisting of 40 cm thick reinforced concrete. The inner walls are consisting of bricks and amount to 27 cm thickness. The doors are made of stainless steel, and the entrances to the tunnels are consisting of three doors with 170 cm distance from one another. Fig. 1 illustrates the general layout of the underground network of tunnels overlapping with the urban fabric of the city above. The buildings circled in red are connected to the network by the basement. Figs. 2–4 illustrate partial axonometric drawings and sections of the network of tunnels. In this study, a representative section of one gallery is used to conduct simulations. The gallery is 150 m long, and with rooms on one side of the corridor. The section’s altitude above sea level is in accordance with the scenarios chosen for comparative simulation analyses: once at the actual high of the tunnels, 325 m above sea level, and secondly on the level of the city above ground, at 350 m above sea
2.2. Climate characteristics The city of Kukës is located in North-Eastern Albania. The city is located in the Mediterranean mountainous climate zone, characterized by a severe winter and hot summer. The duration of sunny hours is 61 h in January, and 312 h in July. The total amount of solar radiation per year is 1422.28 kWh*m−2. Kukës is distinguished for a low number of
Fig. 3. Partial axonometric drawing of the section used for the computational simulations. 3
Sustainable Cities and Society 46 (2019) 101444
R. Breçani, S. Dervishi
level of the city above) without insulation, and S3.3-S5.3 when the building is located at 350 m above the sea level with insulation. Simulations for scenarios S1-S5 are conducted on the model located at 325 m above sea level, whilst simulations for the scenarios S3.2-S5.2, as well as S3.3-S5.3 are conducted on the model located at 350 m above sea level. Differences in the two models are:
• Presence of windows; • Sun radiation; • Contact with the earth; and • Altitude above sea level.
Fig. 6. Annual temperatures for the city of Kukës.
The respective construction data assumptions (materials and Uvalue) in these scenarios are resumed in Table 1. Assumptions regarding occupancy patterns and ventilation regimes for the scenarios (S3- S5, S3.2-S5.2, and S3.3-S5.3) are based on energy-use data supported by Design Builder. Table 2 illustrates the assumptions including the schedule (h), density (people. m−2), internal gains (W m−2), and heating and cooling set point (◦C). To run the simulations, weather files were generated based on the Meteonorm weather file (Anonymous, 2019a). Table 3 illustrates the input data for the lighting used in different activities.
clear days, with 3.2 days in January, and 14.4 in August, whereas the number of cloudy days is 18.5 in January, and 3.2 in July. The average annual temperature is 11.8 °C and the annual amplitude is 21.5 °C. The absolute maximum temperature in Kukës was marked in August 1957, reaching 39.5 °C, whereas the absolute minimum was recorded on January 27, 1963, reaching −21 °C. The actual mean temperature of the earth in Kukës is 13.8 °C, fluctuating from 5 °C in January to 22.7 °C in July. Fig. 6 illustrates the annual temperature of Kukës. The atmospheric precipitations are mainly concentrated on the western and southwestern slopes of the mountain ranges. The average amount of yearly rainfall reaches 946 mm, with over 70% falling during fall and winter (570 mm) and only 15% during summer (143 mm). From October to April, a part of the precipitations fall in the form of snow. Snow thickness varies from 80 cm to 1 m. The snow layer averages 35–40 days. The average wind speed is 3.5 m s−1 during winter, 3.1 m s−1 during spring and 2.3 m s−1 during fall.
3. Results and discussions Figs. 7–10 introduce an overview of the thermal conditions in the selected scenarios based on two reference days in January and July. Figs. 7, 9 show the temperature performance for the first two scenarios: S1 (base case without thermal insulation) and S2 (base case with 10 cm thermal insulation) respectively. Subsequently, the results from the thermal performance of the activity scenarios when situated underground and above ground are established in Figs. 8, 10. 11, 12 resume the information regarding undergrounds’ monthly and yearly energy consumption compared to the same activity scenarios above ground. A lot of effort is put to accurately describe the network of tunnels which are used in the present case study. The tunnels described in this research paper are inaccessible to the public, thus an on-site calibration of the simulation data provided by the model is impossible. Nevertheless, the temperature figures as a result of this study are in accordance with the respective scientific theoretical research on underground heat flux and heat transmission in underground environments (Florides & Kalogirou, 2005; Kajtar et al., 2015; Popiel et al., 2001). The results are in accordance to figures described in preceding related literature which has considered similar cases. Various researchers have assessed the temperature of the soil as quasi-stationary and thus the spaces enclosed in it benefit from comfortable indoor thermal performance (Alkaff et al., 2016; Anselm, 2012; Kajtar et al., 2015; Popiel et al., 2001). Paramananthan had predicted the 300–750 m altitude above sea level to result in the most comfortable indoor temperature for underground structures (Paramananthan, 2000). This assumption is supported by the simulation results since the bunker network of Kukës is located 325–350 m above sea level. The heat flux of the soil in the depth where the tunnels are located, 25 m underground, is almost constant and allows the spaces enveloped in it to remain cool during summer and warm during winter. This assumption is further sustained by the graphs established in Figs. 7, 9 where the lines for 10 cm and no thermal insulation are either overlapping or so close in value that the difference is not detectable, thus implying the effect of thermal insulation on the thermal performance of the building is negligible. Simultaneously, when thermal insulation of 5 cm, two times narrower, is applied to the scenarios above ground (S3.2-S5.2; S3.3-S5.3), it results in a change of 1–3 °C as illustrated in Figs. 8, 10. As depicted in Fig. 8, during January all temperatures establish an increment during the day, and cooling down after 18:00. The hospital’s temperatures are more constant and 2–3 °C lower. The laboratory’s and museum’s temperatures vary considerably during the 12/24 h of
2.3. Thermal performance simulation 2.3.1. Simulation software Initial simulation models were generated based on collected geometry and construction data. Assumptions were made based on in situ observations and historical documents. The simulations are carried out with Design Builder (Anonymous, 2019b), a software tool used to perform building energy, lighting, and comfort performance. It has been developed to simplify the process of building, modeling, and simulation for maximum productivity, by allowing users to rapidly compare the function and performance of building designs, different scenarios, and deliver results quickly and easily. Meteonorm software (Anonymous, 2019a) was used to generate the local weather file in accordance to the information displayed in Section 2.2. A digital performance simulation model of the underground tunnel is generated including the modeling, occupancy patterns and ventilation regimes. Only a tunnel section of 150 m (see Figs. 4,5) is used in the view of simulations. 2.3.2. Scenarios To illustrate the effect of the interventions on the thermal performance of the tunnels, a set of five scenarios were established for a parametric study (see Table 1). The first scenario (S1) represents the existing conditions of the tunnels, at 325 m above sea level. A second scenario (S2) was defined involving thermal insulation on the existing base case of the tunnels. These first two simulation scenarios are conducted in order to define the actual thermal performance of the base case tunnel section and to make an evaluation of the thermal insulation effect on its performance. The three following scenarios (S3-S5) involve different potential activities of the underground tunnels including laboratory (S3), hospital (S4) and museum (S5) respectively. Three simultaneous sets of simulations are conducted for scenarios S3-S5 in order to verify energy reduction potential of underground buildings with comparison to above ground buildings: S3-S5 when the building is located at 325 m above the sea level (the actual level of the tunnels), S3.2-S5.2 when the building is located at 350 m above the sea level (the 4
Sustainable Cities and Society 46 (2019) 101444
R. Breçani, S. Dervishi
Table 1 Description of the simulation scenarios (S1–S5, and S3.2-S5.2) with associated U-value assumptions regarding the pertinent building components. Code
Scenario
U-value (W/m²-K)
Description
S1
Base Case without thermal insulation
S2
Base Case with 10 cm thermal insulation
Uinner = 1.42 Uintermediate = 1.89 Uouter = 2.31 Uinner = 0.27 Uintermediate = 0.30 Uouter = 0.29 Uinner = 1.42
Inner corridor walls (0.28 m): Brickwork, plastered, no insulation; Room walls (0.16 m): Brickwork, plastered, no insulation; Outer walls (0.45 m): Reinforced concrete, plastered, no insulation (including roof and floor). Inner corridor walls (0.38 m): Brickwork, plastered, thermal insulation; Room walls (0.26 m): Brickwork, plastered, thermal insulation; Outer walls (0.55 m): Reinforced concrete, plastered, thermal insulation (including roof and floor). Inner corridor walls (0.28 m): Brickwork, plastered, no insulation;
Uintermediate = 1.88 Uouter = 2.31 Uinner = 0.63 Uintermediate = 0.707 Uouter = 0.528 Uwindows = 3.157 Uinner = 1.42
Room walls (0.16 m): Brickwork, plastered, no insulation; Outer walls (0.45 m): Reinforced concrete, plastered, no insulation (including roof and floor). Inner corridor walls (0.31 m): Brickwork, plastered, thermal insulation; Room walls (0.185 m): Brickwork, plastered, thermal insulation; Outer walls (0.465 m): Reinforced concrete, plastered, thermal insulation (including roof and floor); Windows: double glazing. Inner corridor walls (0.28 m): Brickwork, plastered, no insulation;
Uintermediate = 1.88 Uouter = 2.31 Uinner = 0.63 Uintermediate = 0.707 Uouter = 0.528 Uwindows = 3.157 Uinner = 1.42
Room walls (0.16 m): Brickwork, plastered, no insulation; Outer walls (0.45 m): Reinforced concrete, plastered, no insulation (including roof and floor). Inner corridor walls (0.31 m): Brickwork, plastered, thermal insulation; Room walls (0.185 m): Brickwork, plastered, thermal insulation; Outer walls (0.465 m): Reinforced concrete, plastered, thermal insulation (including roof and floor); Windows: double glazing. Inner corridor walls (0.28 m): Brickwork, plastered, no insulation;
Uintermediate = 1.88 Uouter = 2.31 Uinner = 0.63 Uintermediate = 0.707 Uouter = 0.528 Uwindows = 3.157
Room walls (0.16 m): Brickwork, plastered, no insulation; Outer walls (0.45 m): Reinforced concrete, plastered, no insulation (including roof and floor). Inner corridor walls (0.31 m): Brickwork, plastered, thermal insulation; Room walls (0.185 m): Brickwork, plastered, thermal insulation; Outer walls (0.465 m): Reinforced concrete, plastered, thermal insulation (including roof and floor); Windows: double glazing.
S3
Laboratory S 3.2
S 3.3
Laboratory
S4
Hospital S 4.2
S 4.3
Hospital
S5
Museum S 5.2
S 5.3
Museum
Table 2 Assumptions pertaining to schedule (h), density (people. m − 2), internal gains (W. m − 2), and heating and cooling set point (◦C) for the study of activity impact on the undergrounds tunnel thermal load. Activity
Code
Altitude
Laboratory
S S S S S S S S S
325 350 350 325 350 350 325 350 350
Hospital
Museum
3 3.2 3.3 4 4.2 4.3 5 5.2 5.3
above above above above above above above above above
sea sea sea sea sea sea sea sea sea
level level level level level level level level level
Schedule
Density (people/m²)
Gain (W/m²)
Radiant fraction
Heating (20 °C)
Cooling (20 °C)
12/24h 12/24h 12/24h 24/24h 24/24h 24/24h varying varying varying
0.11 0.11 0.11 0.10 0.10 0.10 0.15 0.15 0.15
8.73 8.73 8.73 3.58 3.58 3.58 3.50 3.50 3.50
0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2
20 20 20 18 18 18 20 20 20
23 23 23 25 25 25 24 24 24
Table 3 Records on the lighting used for different activities. Assumptions pertaining to lighting normalized power density, radiant and visible fraction for the lighting impact on the underground tunnel thermal load. Code
Luminaire type
Normalized power density (W/m²-100 lux)
Radiant fraction
Visible fraction
S 3, S 3.2, S 3.3 S 4, S 4.2, S 4.3 S 5, S 5.2, S 5.3
suspended suspended suspended
2 2.67 5
0.42 0.42 0.42
0.18 0.18 0.18
Kalogirou (2005) can be observed: the temperature alternates between the earth and the ground, when above ground the temperatures are high, they are lower underground and vice versa. The time lag is of approximately 10 h. However, this phenomenon cannot be observed during July because the temperatures are constantly high and do not vary. Furthermore, the temperatures of the underground scenarios (S1S5) are overall more constant compared to the scenarios above ground, as predicted by the theoretical background (Kajtar et al., 2015; Popiel et al., 2001). The performance of the S3-S5 regarding energy consumption varies. The average energy consumption per month is 14, 19.2, and 8.24
activity, most possibly due to the difference in nature of the activities held which require more physical interaction and movement from the people involved. Fig. 10 illustrates how during July the temperature variation lines are considerably more constant and closer in value to one another. During January, the temperatures of the scenarios above ground are lower by 3–5 °C in comparison to the same scenarios located underground, whilst during July they are 5–8 °C higher. The average temperatures during July are 23.1 °C, 24.2 °C, and 23.3 °C for S3, S4, and S5 respectively. The average temperatures during January are 21 °C, 18.3 °C, and 19.1 °C for S3, S4, and S5 respectively. During January, the time lag phenomenon described by Florides and
5
Sustainable Cities and Society 46 (2019) 101444
R. Breçani, S. Dervishi
Fig. 7. Simulated indoor temperatures (whole building) for S1 (without insulation) and S2 (with insulation) together with the external temperature data from the weather file (15th and 16th of January, 2016).
Fig. 10. Comparison of the simulated indoor temperatures (whole building) of the activity scenarios (S3-S5, S3.2-S5.2, S3.3-S5.3) together with the external temperature data from the weather file (15th and 16th of July, 2016).
Fig. 8. Comparison of the simulated indoor temperatures (whole building) of the activity scenarios (S3-S5, S3.2-S5.2, S3.3-S5.3) together with the external temperature data from the weather file (15th and 16th of January, 2016).
Fig. 11. Comparison of simulated monthly energy demand (kWh*m−2) for all activity scenarios (S3-S5, S3.2-S5.2, S3.3-S5.3).
Fig. 12. Comparison of simulated yearly energy demand (kWh*m−2) for all activity scenarios (S3-S5, S3.2-S5.2, S3.3-S5.3).
the museum is the scenario which consumes the least amount of energy. 4. Conclusions
Fig. 9. Simulated indoor temperatures (whole building) for S1 (without insulation) and S2 (with insulation) together with the external temperature data from the weather file (15th and 16th of July, 2016).
The majority of information on underground spaces is provided by theoretical or qualitative analyses. This paper investigates for the first time quantitative data on nuclear shelters, with focus on a case study situated in Albania. The emphasis is on indoor thermal performance and energy consumption evaluation carried out through simulation software in order to assess the feasibility of adaptive reuse. Using the example of thermal simulation of a case study underground network in Kukës, the paper explores the process and the recent results from the data collection. As such, a building simulation model is applied toward the assessment of the buildings’ performance and prediction of the consequences of alternative options for its renovation, reuse, and adaptation. Two simultaneous sets of simulations for the same building and
kWh*m−2 for S3, S4, and S5 respectively. Regarding yearly energy consumption, the hospital’s energy consumption results are 38% higher than the laboratory’s, and 135% higher than the museum’s. Figs. 11, 12 compare the monthly and yearly energy consumption rates for scenarios S3-S5, S3.2-S5.2, and S3.3-S5.3. The findings from the simulations establish the same building consumes 44–145% more energy when located above ground as opposed to underground, thus confirming the underground network of tunnels could efficiently be adapted into a low energy consumption facility. As a conclusion, the laboratory scenario performs better in terms of thermal comfort, whilst 6
Sustainable Cities and Society 46 (2019) 101444
R. Breçani, S. Dervishi
activity scenario are carried out to verify energy reduction potential of underground buildings with comparison to above ground buildings. The scenarios used for the simulations are appropriate to the design typology, the urban context, and the quality of the space of the underground network of tunnels found in Kukës. The 150 m long section used for the simulations on thermal performance is situated 25 m below the level of the city, and it is connected to the city’s hospital by the basement. Different occupancy scenarios are tested in order to assess the thermal performance of the underground tunnels for activities of 12/24 h, 24/24 h, and indefinite occupancy hours such as laboratories, hospital extensions, and museums respectively. The results establish that the average energy consumption fluctuates between 55 kWh*m−2 for base case and 98–230 kWh*m−2 for design scenarios. Indoor air temperature ranges from 15 to 18 °C during winter and 23–28 °C during summer. Insulation does not affect the heat flux through the outer walls. Furthermore, the results establish the same building consumes 44–145% more energy when located above ground as opposed to underground. The thermal performances of all cases are close in values; nevertheless, the 24/24 h activity scenario leads to higher energy consumption due to lower temperature rates during winter. Considering the results from this study, the tunnels can be efficiently subjected to adaptive reuse for functions such as extensions to hospitals, laboratories, museums etc. A preferable scenario for adaptive reuse would be an activity held only 12/24 h, resulting in less energy consumption and better thermal performance. In the present specific case, the 150 m section rooms can be proficiently repurposed as laboratory sectors that could serve to the hospital situated above ground. The present paper uses a contextual case study, with definite altitude above sea level, structure typology, and climate, but nevertheless, the approach tracked here may be followed in a wider context. Although simulation and evaluation through software may face limitations due to the inability to make comparison with real life situations, the study represents an effective and well documented first step towards energy efficiency buildings and possible adaptive reuse of abandoned underground or earth sheltered bunkers, globally.
planning policy–and how this will enhance an urban underground future. Tunnelling and Underground Space Technology, 55, 214–220. Alkaff, S. A., Sim, S. C., & Ervina Efzan, M. N. (2016). A review of underground building towards thermal energy efficiency and sustainable development. Renewable and Sustainable Energy Reviews, 60, 692–713. https://doi.org/10.1016/j.rser.2015.12. 085. Meteonorm n.d. http://www.meteonorm.com/ (accessed February 2, 2018). Design builder n.d. https://www.designbuilder.com/https://www.designbuilder.com/ software/product-overview (Accessed February 2, 2018). Anselm, A. J. (2012). Earth shelters ; a review of energy conservation properties in earth sheltered housing. InTech, 125–148. https://doi.org/10.5772/51873. Delmastro, C., Lavagno, E., & Schranz, L. (2016). Energy and underground. Tunnelling and Underground Space Technology, 55, 96–102. https://doi.org/10.1016/j.tust.2015.10. 021. Delzendeh, E., Wu, S., Lee, A., & Zhou, Y. (2017). The impact of occupants’ behaviours on building energy analysis: A research review. Renewable and Sustainable Energy Reviews, 80, 1061–1071. Florides, G., & Kalogirou, S. (2005). Measurements of Ground temperature at various depths. 8th REHVA World Congr. Hait, J. (1983). Passive annual heat storage: Improving the design of earth shelters. USA: Rocky mountain research centre. Ip, K., & Miller, A. (2009). Thermal behaviour of an earth-sheltered autonomous building—The Brighton Earthship. Renewable Energy, 34, 2037–2043. https://doi. org/10.1016/j.renene.2009.02.006. Kajtar, L., Nyers, J., & Szabo, J. (2015). Dynamic thermal dimensioning of underground spaces. Energy, 87, 361–368. https://doi.org/10.1016/j.energy.2015.04.112. Li, Y., Yuan, Y., Li, C., Han, X., & Zhang, X. (2018). Human responses to high air temperature, relative humidity and carbon dioxide concentration in underground refuge chamber. Building and Environment, 131, 53–62. https://doi.org/10.1016/j.buildenv. 2017.12.038. Municipality of Kukes (2000). General Urban plan for the New City of kukes, 1, 1973. Paramananthan, S. (2000). Malaysia soil taxonomy—A unified malaysian soil classification system. Soil Malaysia, 1. Polytechnic University of Bari (2017). Notes on the underground network of tunnels of Kukes. Popiel, C., Wojtkowiak, J., & Biernacka, B. (2001). Measurements of temperature distribution in ground. Experimental Thermal and Fluid Science, 25, 301–309. https://doi. org/10.1016/S0894-1777(01)00078-4. Poulios, N. (2009). The psychology of living underground. Romanova, E. (2016). Negative influence of underground constructions on mental health of the person. Procedia Engineering, 165, 1176–1183. https://doi.org/10.1016/j. proeng.2016.11.836. Sasmito, A. P., Kurnia, J. C., Birgersson, E., & Mujumdar, A. S. (2014). Computational evaluation of thermal management strategies in an underground mine. Applied Thermal Engineering, 90, 1–7. https://doi.org/10.1016/j.applthermaleng.2015.01. 062. Shan, M., gang, Hwang B., & Wong, K. S. N. (2017). A preliminary investigation of underground residential buildings: Advantages, disadvantages, and critical risks. Tunnelling and Underground Space Technology, 70, 19–29. https://doi.org/10.1016/j. tust.2017.07.004. Tan, Z., Roberts, A. C., Christopoulos, G. I., Kwok, K. W., Car, J., Li, X. Z., et al. (2018). Working in underground spaces: Architectural parameters, perceptions and thermal comfort measurements. Tunnelling and Underground Space Technology, 71, 428–439. https://doi.org/10.1016/j.tust.2017.09.002. Tezuka, M., & Seoka, T. (2003). Latest technology of underground rock cavern excavation in Japan. Tunnelling and Underground Space Technology, 18, 127–144. Zhao, J., Peng, F., Wang, T., Zhang, X., & Jiang, B. (2016). Advances in master planning of urban underground space (UUS) in China. Tunnelling and Underground Space Technology, 55, 290–307. Zhu, J., & Tong, L. (2017). Experimental study on the thermal performance of underground cave dwellings with coupled Yaokang. Renewable Energy, 108, 156–168. https://doi.org/10.1016/j.renene.2017.02.051.
Acknowledgements Special thanks are extended to Mr. Artan Hysa, professor at Epoka University, for his assistance during the preliminary phase of this research. The authors would like to thank the Polytechnic University of Bari for facilitating the research with data on the tunnels of Kukës. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. References Admiraal, H., & Cornaro, A. (2016). Why underground space should be included in urban
7
Contents lists available at ScienceDirect
Sustainable Cities and Society journal homepage: www.elsevier.com/locate/scs
Thermal and energy performance evaluation of underground bunkers: An adaptive reuse approach
T
⁎
Rudina Breçani , Sokol Dervishi Department of Architecture, Epoka University, Rruga Tiranë-Rinas, Km 12, 1039, Tirana, Albania
A R T I C LE I N FO
A B S T R A C T
Keywords: Thermal performance simulation Scenarios Energy Underground nuclear bunker Adaptive reuse
The quantitative research on bunker tunnels is extremely limited nowadays, though this typology of structure is vastly found globally. The present study follows up on the existing literature by providing an innovative insight on the subject: evaluation of the thermal indoor comfort and energy consumption through the reality of living in a bunker. The potential adaptive reuse of the unexploited extensive underground bunker network of Albania is investigated. The approach is illustrated using a 150-m long cross-section of the underground network selected for the parametric computational simulation on thermal and energy performance. The model simulation involves data collection on local climate, design typology, and building materials to generate a finite-element simulation model. Different scenarios including insulation of the outer walls, occupancy patterns, and ventilation regimes are examined in order to assess the respective thermal and energy performance. Two sets of simulations are held simultaneously: at 325 m above sea level (current altitude of the underground tunnels), and 350 m above sea level (the current altitude of the city above ground). Insulation does not affect the heat flux through the outer walls. Furthermore, the results establish the same building consumes 44-145% more energy when located above ground as opposed to underground.
1. Introduction A special kind of underground shelters can be observed in Albania, i.e. nuclear bunker tunnels. These capillaries, which were built in several cities during the communist period and never used to fulfill the purpose for which they were built (refuge in case of war), constitute a large underground network of concrete leftover space. Therefore, the question on the efficiency with which they can fulfill their original purpose i.e. that of sheltering, comes out naturally. The present study assesses the thermal and energy performance and evaluates the adaptive reuse efficiency of the underground shelters through computational simulation modeling, by using the tunnel network found in the city of Kukës, Albania, as a base for conducting simulations. Past researches on underground spaces have tapped into various topics. Underground sheltering has been subject to various analyses throughout the years, ranging from researches on the psychology and behavior of isolated humans (Poulios, 2009; Romanova, 2016), urban planning (Admiraal & Cornaro, 2016; Delmastro, Lavagno, & Schranz, 2016), to the adaptive reuse of mines, railways, and other similar underground facilities after their activities disconnect (Sasmito, Kurnia, Birgersson, & Mujumdar, 2014). Li, Yuan, Li, Han, and Zhang (2018) have monitored the physical reactions of people while staying in refuge ⁎
chambers, concluding that the number of people in the enclosed space directly affects the humidity level and temperature of the said area; the heart rate and temperature of the individuals subjected to the experiment remained constant (Li et al., 2018, Delzendeh, Wu, Lee, & Zhou, 2017). However, he recognizes the enclosed space, constructed specifically for the study, is small and the prejudice that accompanies the concept of underground is major Li et al. (2018). The studies on the computational evaluation of either thermal or energy performance for such spaces are fewer in number. Zhu has carried out thermal simulations on a sample of atrium plan earth sheltered buildings with additional heating systems named Yaokang concluding that the buildings constitute a step forward towards low energy building design (Zhu & Tong, 2017). Another thermal performance simulation and evaluation study has been carried out by Ip, but the building in question has only one wall in contact with the earth (Ip & Miller, 2009). Tan has investigated the thermal comfort conditions in underground spaces in four major Chinese cities characterized by different climate conditions in order to assess the influence of the specific climatic variables (Tan et al., 2018). He recognizes passive design strategies as effective measures in reducing energy consumption especially during transitional periods (Tan et al., 2018). Tan however also recognizes his study as only qualitative and evaluates a building energy
Corresponding author. E-mail addresses: [email protected] (R. Breçani), [email protected] (S. Dervishi).
https://doi.org/10.1016/j.scs.2019.101444 Received 26 August 2018; Received in revised form 19 January 2019; Accepted 19 January 2019 Available online 23 January 2019 2210-6707/ © 2019 Elsevier Ltd. All rights reserved.
Sustainable Cities and Society 46 (2019) 101444
R. Breçani, S. Dervishi
temperature is always higher than the air temperature during winter and lower during summer. Due to the solar radiation and the soil’s relative capacity, there is a 5-hour time lag a day regarding temperature distribution (Florides & Kalogirou, 2005). Whereas previous researches have provided significant insight on different topics related to underground structures, there is a lack of quantitative analyses providing close examination of thermal performance and energy consumption evaluation. Furthermore, the present study constitutes the first research on thermal and energy performance evaluation on the typology of underground bunkers, a phenomenon present not only in Albania, but in a global scale, and the number of similar studies worldwide is limited. Additionally, this research indicates and examines alternatives for low energy consumption adaptive reuse functions, appropriate for the design typology, and space quality of the tunnels. The biggest flaw of these constructions is simultaneously a great possibility for future development: the underground bunkers are massive leftover space which does not serve any function currently. For the context in which these tunnels are situated nowadays, the importance of the findings from this research relate to the significance they hold for the hosting country. Since the study aims to prove these tunnels are energy efficient structures which do not disrupt the city environment, a positive outcome and future step to this research would be their consideration and implementation into the country’s trading and city governance policies. A good example of this is the transformation of the underground network into international rentable space for goods, laboratories etc. Similarly to Albania, several other countries in Europe and wider share an analogous historical background, due to which the presence of underground bunker like structures is spread, and thus, the findings from this study could efficiently serve as a comparative case study for future research in an international context. It is also important to emphasize that thermal simulation and energy efficiency assessments are definite for every study. In the case of Albania, the specificities that influence the results of the study are: local climate, design typology, percentage of contact with the earth, and altitude above sea level. Appropriate functions noted by the proposals of previous studies, the characteristics of the adaptive functions, the design typology of the tunnels in Albania, as well as the urban composition and quality of the city of Kukës have been considered in order to decide on the modeling simulation scenarios. Two sets of software based simulations are held simultaneously for the same building and scenarios, in order to compare energy efficiency and temperature performance results: at 325 m above sea level (current altitude of the underground tunnels of Kukës), and 350 m above sea level (the current altitude of the city of Kukës above ground).
modeling would provide more solid data since qualitative analyses often depend on real life restrictions (Tan et al., 2018). Zhao, Peng, Wang, Zhang, and Jiang (2016); Tezuka and Seoka (2003), as well as Tan et al. (2018) investigate the potential integration of underground space examination in urban city planning. Tan et al. insist that underground planning should be an integral part of any National Regulation Plan since the active use of underground space can positively affect the natural city landscape (Tan et al., 2018). Shan et al. have identified through questionnaires the top four advantages of underground spaces that make these spaces attractive: space saving, improved indoor thermal comfort, resistance to external noises, and increased level of privacy (Shan, gang, & Wong, 2017). The major disadvantages they have found are limited amount of daylight and negative psychological effects (Shan et al., 2017), which are also supported by other researches tackling the psychological consequences on humans living in isolated underground spaces by Romanova (2016), and Poulios (2009). However, Shan et al. recognize that results might be affected by bias or prejudice. Furthermore, Tan et al. point out that sufficient natural light is not provided in many above ground buildings either (Tan et al., 2018). Alkaff et al. clarify the factors that most notably affect the energy conservation quality of the underground structures, such as: design typology, depth below ground surface, ventilation system, climate and soil thermal properties, insulation and infiltration, and altitude above sea level (Alkaff, Sim, & Ervina Efzan, 2016). The design typology affects the sun penetration into the building thus influencing its energy performance according to Alkaff et al. (2016) and Anselm (2012). A structure that is located deeper underground might need waterproofing and it benefits from superior temperature performance. As for the altitude above the sea level, the depth of 300–750 m above sea level results in a more comfortable temperature inside the underground structures (Paramananthan, 2000). Mechanical ventilation is a factor that could negatively affect energy performance in underground spaces due to the limited amount of natural ventilation, however, Hait (1983) and Anselm (2012) have offered cost effective solutions to such a problem, such as precooling the fresh air that enters the building. The ground temperature distribution is also affected by the physical properties of the ground, the local meteorological variables, and the surface cover of the said portion of the ground (Popiel, Wojtkowiak, & Biernacka, 2001). Some scholars have carried out scientific research with a purpose to create predictive formulas for the thermal capacity and heat flux of the soil. As Alkaff et al. point out, a common misconception depicts the earth as a good insulator when in reality the earth has high heat conductivity. Owing to that earth is usually extended in vast areas, the thermal energy loses transfer, and the temperature fluctuations lower with the increment of depth (Alkaff et al., 2016). Moreover, the heat capacity of the soil is considerably higher than that of the air. Florides and Kalogirou claim that because of the high thermal inertia of the soil, the temperature fluctuations at the surface of the ground are diminished as the depth of the ground increases (Florides & Kalogirou, 2005). According to Kajtar et al. the temperature of the soil changes with amplitude of 0.6 °C in the depth of 8 m and 0.2 °C in the depth of 10 m (Kajtar, Nyers, & Szabo, 2015). Popiel states that after the depth reaches more than 20 m the temperature becomes practically constant (Popiel et al., 2001). The formula depicted in Eq. (1) illustrates the ground heat flux formula used by him. Here, x is the depth below the ground surface, T is the temperature, k is the heat conductivity of the ground, and q the heat flux.
q = −k
∂T ∂x
2. Methodology 2.1. Tunnel description The underground tunnels used for building simulations in this study are part of the ex-secret military map of Albania, which has been disclosed to the public only recently, in 2014. The military reinforcements adopted by the communist Albanian government during 1945–1990 are present throughout the territory of the country in the shape of singular secluded mushroom-like concrete bunkers or in the shape of extended underground galleries designed for sheltering a large part of a city’s population. Thus, the latter design typology can be found in various Albanian cities. The case study buildings for conducting simulations can be found in Kukes, in North Eastern Albania. This network of tunnels includes 30 galleries which make up to 2400 m in length as established in Fig. 3. Each gallery is composed of a corridor with a row of rooms on one side. The rooms and corridors were built inside an outer concrete shell (Municipality of Kukes, 2000;
(1)
Florides and Kalogirou depict a time lag between the temperature fluctuations above and below the ground (Florides & Kalogirou, 2005). This explains why, after reaching a specific depth, the soil’s 2
Sustainable Cities and Society 46 (2019) 101444
R. Breçani, S. Dervishi
Fig. 4. Close-up partial axonometric drawing of the inner construction of the tunnels.
Fig. 1. Underground network of tunnels for Kukës (in red) together with the building fabric of the city (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article).
Fig. 5. Partial axonometric drawing of the section used for the computational simulations in the hypothetical situation it was located at 350 m above sea level.
level. The section, illustrated in Fig. 5, is enclosed on one side by a 25 m tall staircase which connects the existing section to the existing hospital of the city located above ground, whilst the other side of the corridor is a dead end. The section is surrounded by earth on all sides; the only connection of the gallery with the environment above ground is by the staircase that connects it to the hospital above. The corresponding hypothetical section of the same characteristics, when located at 350 m above sea level, is illustrated in Fig. 5. The two building sections differ in construction only in the presence of staircase well, and presence of windows. The energy loads delivered by the software provide a valuable insight into the hypothetical heating and cooling demand of such a structure. Nowadays, the underground network is unoccupied and it does not use any mechanical ventilation of HVAC systems. In the different scenarios tested during simulations, the energy loads inflicted on the building include mechanical ventilation, retrofitting HVAC systems, electricity supply systems, as well as lighting have been taken into consideration.
Fig. 2. Partial section of the underground network of tunnels for Kukës.
Polytechnic University of Bari, 2017). The internal part of the circumferential walls was also built with concrete, whilst the indoor walls were built with bricks. The tunnels have corridors which are 270 cm high and 120 cm wide, and rooms of an average of 3.4 × 3.4 m. The network is located from 319 m to 336 m above the sea level (the different levels are connected by staircases), whilst the ground level of the city of Kukës is located 350 m above the sea level (Municipality of Kukes, 2000; Polytechnic University of Bari, 2017). The outer walls of the tunnels are consisting of 40 cm thick reinforced concrete. The inner walls are consisting of bricks and amount to 27 cm thickness. The doors are made of stainless steel, and the entrances to the tunnels are consisting of three doors with 170 cm distance from one another. Fig. 1 illustrates the general layout of the underground network of tunnels overlapping with the urban fabric of the city above. The buildings circled in red are connected to the network by the basement. Figs. 2–4 illustrate partial axonometric drawings and sections of the network of tunnels. In this study, a representative section of one gallery is used to conduct simulations. The gallery is 150 m long, and with rooms on one side of the corridor. The section’s altitude above sea level is in accordance with the scenarios chosen for comparative simulation analyses: once at the actual high of the tunnels, 325 m above sea level, and secondly on the level of the city above ground, at 350 m above sea
2.2. Climate characteristics The city of Kukës is located in North-Eastern Albania. The city is located in the Mediterranean mountainous climate zone, characterized by a severe winter and hot summer. The duration of sunny hours is 61 h in January, and 312 h in July. The total amount of solar radiation per year is 1422.28 kWh*m−2. Kukës is distinguished for a low number of
Fig. 3. Partial axonometric drawing of the section used for the computational simulations. 3
Sustainable Cities and Society 46 (2019) 101444
R. Breçani, S. Dervishi
level of the city above) without insulation, and S3.3-S5.3 when the building is located at 350 m above the sea level with insulation. Simulations for scenarios S1-S5 are conducted on the model located at 325 m above sea level, whilst simulations for the scenarios S3.2-S5.2, as well as S3.3-S5.3 are conducted on the model located at 350 m above sea level. Differences in the two models are:
• Presence of windows; • Sun radiation; • Contact with the earth; and • Altitude above sea level.
Fig. 6. Annual temperatures for the city of Kukës.
The respective construction data assumptions (materials and Uvalue) in these scenarios are resumed in Table 1. Assumptions regarding occupancy patterns and ventilation regimes for the scenarios (S3- S5, S3.2-S5.2, and S3.3-S5.3) are based on energy-use data supported by Design Builder. Table 2 illustrates the assumptions including the schedule (h), density (people. m−2), internal gains (W m−2), and heating and cooling set point (◦C). To run the simulations, weather files were generated based on the Meteonorm weather file (Anonymous, 2019a). Table 3 illustrates the input data for the lighting used in different activities.
clear days, with 3.2 days in January, and 14.4 in August, whereas the number of cloudy days is 18.5 in January, and 3.2 in July. The average annual temperature is 11.8 °C and the annual amplitude is 21.5 °C. The absolute maximum temperature in Kukës was marked in August 1957, reaching 39.5 °C, whereas the absolute minimum was recorded on January 27, 1963, reaching −21 °C. The actual mean temperature of the earth in Kukës is 13.8 °C, fluctuating from 5 °C in January to 22.7 °C in July. Fig. 6 illustrates the annual temperature of Kukës. The atmospheric precipitations are mainly concentrated on the western and southwestern slopes of the mountain ranges. The average amount of yearly rainfall reaches 946 mm, with over 70% falling during fall and winter (570 mm) and only 15% during summer (143 mm). From October to April, a part of the precipitations fall in the form of snow. Snow thickness varies from 80 cm to 1 m. The snow layer averages 35–40 days. The average wind speed is 3.5 m s−1 during winter, 3.1 m s−1 during spring and 2.3 m s−1 during fall.
3. Results and discussions Figs. 7–10 introduce an overview of the thermal conditions in the selected scenarios based on two reference days in January and July. Figs. 7, 9 show the temperature performance for the first two scenarios: S1 (base case without thermal insulation) and S2 (base case with 10 cm thermal insulation) respectively. Subsequently, the results from the thermal performance of the activity scenarios when situated underground and above ground are established in Figs. 8, 10. 11, 12 resume the information regarding undergrounds’ monthly and yearly energy consumption compared to the same activity scenarios above ground. A lot of effort is put to accurately describe the network of tunnels which are used in the present case study. The tunnels described in this research paper are inaccessible to the public, thus an on-site calibration of the simulation data provided by the model is impossible. Nevertheless, the temperature figures as a result of this study are in accordance with the respective scientific theoretical research on underground heat flux and heat transmission in underground environments (Florides & Kalogirou, 2005; Kajtar et al., 2015; Popiel et al., 2001). The results are in accordance to figures described in preceding related literature which has considered similar cases. Various researchers have assessed the temperature of the soil as quasi-stationary and thus the spaces enclosed in it benefit from comfortable indoor thermal performance (Alkaff et al., 2016; Anselm, 2012; Kajtar et al., 2015; Popiel et al., 2001). Paramananthan had predicted the 300–750 m altitude above sea level to result in the most comfortable indoor temperature for underground structures (Paramananthan, 2000). This assumption is supported by the simulation results since the bunker network of Kukës is located 325–350 m above sea level. The heat flux of the soil in the depth where the tunnels are located, 25 m underground, is almost constant and allows the spaces enveloped in it to remain cool during summer and warm during winter. This assumption is further sustained by the graphs established in Figs. 7, 9 where the lines for 10 cm and no thermal insulation are either overlapping or so close in value that the difference is not detectable, thus implying the effect of thermal insulation on the thermal performance of the building is negligible. Simultaneously, when thermal insulation of 5 cm, two times narrower, is applied to the scenarios above ground (S3.2-S5.2; S3.3-S5.3), it results in a change of 1–3 °C as illustrated in Figs. 8, 10. As depicted in Fig. 8, during January all temperatures establish an increment during the day, and cooling down after 18:00. The hospital’s temperatures are more constant and 2–3 °C lower. The laboratory’s and museum’s temperatures vary considerably during the 12/24 h of
2.3. Thermal performance simulation 2.3.1. Simulation software Initial simulation models were generated based on collected geometry and construction data. Assumptions were made based on in situ observations and historical documents. The simulations are carried out with Design Builder (Anonymous, 2019b), a software tool used to perform building energy, lighting, and comfort performance. It has been developed to simplify the process of building, modeling, and simulation for maximum productivity, by allowing users to rapidly compare the function and performance of building designs, different scenarios, and deliver results quickly and easily. Meteonorm software (Anonymous, 2019a) was used to generate the local weather file in accordance to the information displayed in Section 2.2. A digital performance simulation model of the underground tunnel is generated including the modeling, occupancy patterns and ventilation regimes. Only a tunnel section of 150 m (see Figs. 4,5) is used in the view of simulations. 2.3.2. Scenarios To illustrate the effect of the interventions on the thermal performance of the tunnels, a set of five scenarios were established for a parametric study (see Table 1). The first scenario (S1) represents the existing conditions of the tunnels, at 325 m above sea level. A second scenario (S2) was defined involving thermal insulation on the existing base case of the tunnels. These first two simulation scenarios are conducted in order to define the actual thermal performance of the base case tunnel section and to make an evaluation of the thermal insulation effect on its performance. The three following scenarios (S3-S5) involve different potential activities of the underground tunnels including laboratory (S3), hospital (S4) and museum (S5) respectively. Three simultaneous sets of simulations are conducted for scenarios S3-S5 in order to verify energy reduction potential of underground buildings with comparison to above ground buildings: S3-S5 when the building is located at 325 m above the sea level (the actual level of the tunnels), S3.2-S5.2 when the building is located at 350 m above the sea level (the 4
Sustainable Cities and Society 46 (2019) 101444
R. Breçani, S. Dervishi
Table 1 Description of the simulation scenarios (S1–S5, and S3.2-S5.2) with associated U-value assumptions regarding the pertinent building components. Code
Scenario
U-value (W/m²-K)
Description
S1
Base Case without thermal insulation
S2
Base Case with 10 cm thermal insulation
Uinner = 1.42 Uintermediate = 1.89 Uouter = 2.31 Uinner = 0.27 Uintermediate = 0.30 Uouter = 0.29 Uinner = 1.42
Inner corridor walls (0.28 m): Brickwork, plastered, no insulation; Room walls (0.16 m): Brickwork, plastered, no insulation; Outer walls (0.45 m): Reinforced concrete, plastered, no insulation (including roof and floor). Inner corridor walls (0.38 m): Brickwork, plastered, thermal insulation; Room walls (0.26 m): Brickwork, plastered, thermal insulation; Outer walls (0.55 m): Reinforced concrete, plastered, thermal insulation (including roof and floor). Inner corridor walls (0.28 m): Brickwork, plastered, no insulation;
Uintermediate = 1.88 Uouter = 2.31 Uinner = 0.63 Uintermediate = 0.707 Uouter = 0.528 Uwindows = 3.157 Uinner = 1.42
Room walls (0.16 m): Brickwork, plastered, no insulation; Outer walls (0.45 m): Reinforced concrete, plastered, no insulation (including roof and floor). Inner corridor walls (0.31 m): Brickwork, plastered, thermal insulation; Room walls (0.185 m): Brickwork, plastered, thermal insulation; Outer walls (0.465 m): Reinforced concrete, plastered, thermal insulation (including roof and floor); Windows: double glazing. Inner corridor walls (0.28 m): Brickwork, plastered, no insulation;
Uintermediate = 1.88 Uouter = 2.31 Uinner = 0.63 Uintermediate = 0.707 Uouter = 0.528 Uwindows = 3.157 Uinner = 1.42
Room walls (0.16 m): Brickwork, plastered, no insulation; Outer walls (0.45 m): Reinforced concrete, plastered, no insulation (including roof and floor). Inner corridor walls (0.31 m): Brickwork, plastered, thermal insulation; Room walls (0.185 m): Brickwork, plastered, thermal insulation; Outer walls (0.465 m): Reinforced concrete, plastered, thermal insulation (including roof and floor); Windows: double glazing. Inner corridor walls (0.28 m): Brickwork, plastered, no insulation;
Uintermediate = 1.88 Uouter = 2.31 Uinner = 0.63 Uintermediate = 0.707 Uouter = 0.528 Uwindows = 3.157
Room walls (0.16 m): Brickwork, plastered, no insulation; Outer walls (0.45 m): Reinforced concrete, plastered, no insulation (including roof and floor). Inner corridor walls (0.31 m): Brickwork, plastered, thermal insulation; Room walls (0.185 m): Brickwork, plastered, thermal insulation; Outer walls (0.465 m): Reinforced concrete, plastered, thermal insulation (including roof and floor); Windows: double glazing.
S3
Laboratory S 3.2
S 3.3
Laboratory
S4
Hospital S 4.2
S 4.3
Hospital
S5
Museum S 5.2
S 5.3
Museum
Table 2 Assumptions pertaining to schedule (h), density (people. m − 2), internal gains (W. m − 2), and heating and cooling set point (◦C) for the study of activity impact on the undergrounds tunnel thermal load. Activity
Code
Altitude
Laboratory
S S S S S S S S S
325 350 350 325 350 350 325 350 350
Hospital
Museum
3 3.2 3.3 4 4.2 4.3 5 5.2 5.3
above above above above above above above above above
sea sea sea sea sea sea sea sea sea
level level level level level level level level level
Schedule
Density (people/m²)
Gain (W/m²)
Radiant fraction
Heating (20 °C)
Cooling (20 °C)
12/24h 12/24h 12/24h 24/24h 24/24h 24/24h varying varying varying
0.11 0.11 0.11 0.10 0.10 0.10 0.15 0.15 0.15
8.73 8.73 8.73 3.58 3.58 3.58 3.50 3.50 3.50
0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2
20 20 20 18 18 18 20 20 20
23 23 23 25 25 25 24 24 24
Table 3 Records on the lighting used for different activities. Assumptions pertaining to lighting normalized power density, radiant and visible fraction for the lighting impact on the underground tunnel thermal load. Code
Luminaire type
Normalized power density (W/m²-100 lux)
Radiant fraction
Visible fraction
S 3, S 3.2, S 3.3 S 4, S 4.2, S 4.3 S 5, S 5.2, S 5.3
suspended suspended suspended
2 2.67 5
0.42 0.42 0.42
0.18 0.18 0.18
Kalogirou (2005) can be observed: the temperature alternates between the earth and the ground, when above ground the temperatures are high, they are lower underground and vice versa. The time lag is of approximately 10 h. However, this phenomenon cannot be observed during July because the temperatures are constantly high and do not vary. Furthermore, the temperatures of the underground scenarios (S1S5) are overall more constant compared to the scenarios above ground, as predicted by the theoretical background (Kajtar et al., 2015; Popiel et al., 2001). The performance of the S3-S5 regarding energy consumption varies. The average energy consumption per month is 14, 19.2, and 8.24
activity, most possibly due to the difference in nature of the activities held which require more physical interaction and movement from the people involved. Fig. 10 illustrates how during July the temperature variation lines are considerably more constant and closer in value to one another. During January, the temperatures of the scenarios above ground are lower by 3–5 °C in comparison to the same scenarios located underground, whilst during July they are 5–8 °C higher. The average temperatures during July are 23.1 °C, 24.2 °C, and 23.3 °C for S3, S4, and S5 respectively. The average temperatures during January are 21 °C, 18.3 °C, and 19.1 °C for S3, S4, and S5 respectively. During January, the time lag phenomenon described by Florides and
5
Sustainable Cities and Society 46 (2019) 101444
R. Breçani, S. Dervishi
Fig. 7. Simulated indoor temperatures (whole building) for S1 (without insulation) and S2 (with insulation) together with the external temperature data from the weather file (15th and 16th of January, 2016).
Fig. 10. Comparison of the simulated indoor temperatures (whole building) of the activity scenarios (S3-S5, S3.2-S5.2, S3.3-S5.3) together with the external temperature data from the weather file (15th and 16th of July, 2016).
Fig. 8. Comparison of the simulated indoor temperatures (whole building) of the activity scenarios (S3-S5, S3.2-S5.2, S3.3-S5.3) together with the external temperature data from the weather file (15th and 16th of January, 2016).
Fig. 11. Comparison of simulated monthly energy demand (kWh*m−2) for all activity scenarios (S3-S5, S3.2-S5.2, S3.3-S5.3).
Fig. 12. Comparison of simulated yearly energy demand (kWh*m−2) for all activity scenarios (S3-S5, S3.2-S5.2, S3.3-S5.3).
the museum is the scenario which consumes the least amount of energy. 4. Conclusions
Fig. 9. Simulated indoor temperatures (whole building) for S1 (without insulation) and S2 (with insulation) together with the external temperature data from the weather file (15th and 16th of July, 2016).
The majority of information on underground spaces is provided by theoretical or qualitative analyses. This paper investigates for the first time quantitative data on nuclear shelters, with focus on a case study situated in Albania. The emphasis is on indoor thermal performance and energy consumption evaluation carried out through simulation software in order to assess the feasibility of adaptive reuse. Using the example of thermal simulation of a case study underground network in Kukës, the paper explores the process and the recent results from the data collection. As such, a building simulation model is applied toward the assessment of the buildings’ performance and prediction of the consequences of alternative options for its renovation, reuse, and adaptation. Two simultaneous sets of simulations for the same building and
kWh*m−2 for S3, S4, and S5 respectively. Regarding yearly energy consumption, the hospital’s energy consumption results are 38% higher than the laboratory’s, and 135% higher than the museum’s. Figs. 11, 12 compare the monthly and yearly energy consumption rates for scenarios S3-S5, S3.2-S5.2, and S3.3-S5.3. The findings from the simulations establish the same building consumes 44–145% more energy when located above ground as opposed to underground, thus confirming the underground network of tunnels could efficiently be adapted into a low energy consumption facility. As a conclusion, the laboratory scenario performs better in terms of thermal comfort, whilst 6
Sustainable Cities and Society 46 (2019) 101444
R. Breçani, S. Dervishi
activity scenario are carried out to verify energy reduction potential of underground buildings with comparison to above ground buildings. The scenarios used for the simulations are appropriate to the design typology, the urban context, and the quality of the space of the underground network of tunnels found in Kukës. The 150 m long section used for the simulations on thermal performance is situated 25 m below the level of the city, and it is connected to the city’s hospital by the basement. Different occupancy scenarios are tested in order to assess the thermal performance of the underground tunnels for activities of 12/24 h, 24/24 h, and indefinite occupancy hours such as laboratories, hospital extensions, and museums respectively. The results establish that the average energy consumption fluctuates between 55 kWh*m−2 for base case and 98–230 kWh*m−2 for design scenarios. Indoor air temperature ranges from 15 to 18 °C during winter and 23–28 °C during summer. Insulation does not affect the heat flux through the outer walls. Furthermore, the results establish the same building consumes 44–145% more energy when located above ground as opposed to underground. The thermal performances of all cases are close in values; nevertheless, the 24/24 h activity scenario leads to higher energy consumption due to lower temperature rates during winter. Considering the results from this study, the tunnels can be efficiently subjected to adaptive reuse for functions such as extensions to hospitals, laboratories, museums etc. A preferable scenario for adaptive reuse would be an activity held only 12/24 h, resulting in less energy consumption and better thermal performance. In the present specific case, the 150 m section rooms can be proficiently repurposed as laboratory sectors that could serve to the hospital situated above ground. The present paper uses a contextual case study, with definite altitude above sea level, structure typology, and climate, but nevertheless, the approach tracked here may be followed in a wider context. Although simulation and evaluation through software may face limitations due to the inability to make comparison with real life situations, the study represents an effective and well documented first step towards energy efficiency buildings and possible adaptive reuse of abandoned underground or earth sheltered bunkers, globally.
planning policy–and how this will enhance an urban underground future. Tunnelling and Underground Space Technology, 55, 214–220. Alkaff, S. A., Sim, S. C., & Ervina Efzan, M. N. (2016). A review of underground building towards thermal energy efficiency and sustainable development. Renewable and Sustainable Energy Reviews, 60, 692–713. https://doi.org/10.1016/j.rser.2015.12. 085. Meteonorm n.d. http://www.meteonorm.com/ (accessed February 2, 2018). Design builder n.d. https://www.designbuilder.com/https://www.designbuilder.com/ software/product-overview (Accessed February 2, 2018). Anselm, A. J. (2012). Earth shelters ; a review of energy conservation properties in earth sheltered housing. InTech, 125–148. https://doi.org/10.5772/51873. Delmastro, C., Lavagno, E., & Schranz, L. (2016). Energy and underground. Tunnelling and Underground Space Technology, 55, 96–102. https://doi.org/10.1016/j.tust.2015.10. 021. Delzendeh, E., Wu, S., Lee, A., & Zhou, Y. (2017). The impact of occupants’ behaviours on building energy analysis: A research review. Renewable and Sustainable Energy Reviews, 80, 1061–1071. Florides, G., & Kalogirou, S. (2005). Measurements of Ground temperature at various depths. 8th REHVA World Congr. Hait, J. (1983). Passive annual heat storage: Improving the design of earth shelters. USA: Rocky mountain research centre. Ip, K., & Miller, A. (2009). Thermal behaviour of an earth-sheltered autonomous building—The Brighton Earthship. Renewable Energy, 34, 2037–2043. https://doi. org/10.1016/j.renene.2009.02.006. Kajtar, L., Nyers, J., & Szabo, J. (2015). Dynamic thermal dimensioning of underground spaces. Energy, 87, 361–368. https://doi.org/10.1016/j.energy.2015.04.112. Li, Y., Yuan, Y., Li, C., Han, X., & Zhang, X. (2018). Human responses to high air temperature, relative humidity and carbon dioxide concentration in underground refuge chamber. Building and Environment, 131, 53–62. https://doi.org/10.1016/j.buildenv. 2017.12.038. Municipality of Kukes (2000). General Urban plan for the New City of kukes, 1, 1973. Paramananthan, S. (2000). Malaysia soil taxonomy—A unified malaysian soil classification system. Soil Malaysia, 1. Polytechnic University of Bari (2017). Notes on the underground network of tunnels of Kukes. Popiel, C., Wojtkowiak, J., & Biernacka, B. (2001). Measurements of temperature distribution in ground. Experimental Thermal and Fluid Science, 25, 301–309. https://doi. org/10.1016/S0894-1777(01)00078-4. Poulios, N. (2009). The psychology of living underground. Romanova, E. (2016). Negative influence of underground constructions on mental health of the person. Procedia Engineering, 165, 1176–1183. https://doi.org/10.1016/j. proeng.2016.11.836. Sasmito, A. P., Kurnia, J. C., Birgersson, E., & Mujumdar, A. S. (2014). Computational evaluation of thermal management strategies in an underground mine. Applied Thermal Engineering, 90, 1–7. https://doi.org/10.1016/j.applthermaleng.2015.01. 062. Shan, M., gang, Hwang B., & Wong, K. S. N. (2017). A preliminary investigation of underground residential buildings: Advantages, disadvantages, and critical risks. Tunnelling and Underground Space Technology, 70, 19–29. https://doi.org/10.1016/j. tust.2017.07.004. Tan, Z., Roberts, A. C., Christopoulos, G. I., Kwok, K. W., Car, J., Li, X. Z., et al. (2018). Working in underground spaces: Architectural parameters, perceptions and thermal comfort measurements. Tunnelling and Underground Space Technology, 71, 428–439. https://doi.org/10.1016/j.tust.2017.09.002. Tezuka, M., & Seoka, T. (2003). Latest technology of underground rock cavern excavation in Japan. Tunnelling and Underground Space Technology, 18, 127–144. Zhao, J., Peng, F., Wang, T., Zhang, X., & Jiang, B. (2016). Advances in master planning of urban underground space (UUS) in China. Tunnelling and Underground Space Technology, 55, 290–307. Zhu, J., & Tong, L. (2017). Experimental study on the thermal performance of underground cave dwellings with coupled Yaokang. Renewable Energy, 108, 156–168. https://doi.org/10.1016/j.renene.2017.02.051.
Acknowledgements Special thanks are extended to Mr. Artan Hysa, professor at Epoka University, for his assistance during the preliminary phase of this research. The authors would like to thank the Polytechnic University of Bari for facilitating the research with data on the tunnels of Kukës. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. References Admiraal, H., & Cornaro, A. (2016). Why underground space should be included in urban
7