- Email: [email protected]
Building materials as a source of PCB pollution in Bergen, Norway
Science of the Total Environment 325 (2004) 139–144
Building materials as a source of PCB pollution in Bergen, Norway M. Andersson*, R.T. Ottesen, T. Volden Department of Geochemistry and Hydrogeology, Geological Survey of Norway, 7491 Trondheim, Norway Received 26 September 2003; received in revised form 17 November 2003; accepted 21 November 2003
Abstract Although PCB is a globally recognised pollutant, an understanding of its transport from man-made building materials to the environment is poorly constrained. This paper presents data from a study that was conducted in order ¸ to determine the extent of PCB usage in plaster on building facades in the Bergen area, Norway. The study was to determine whether PCB concentrations vary according to building usage type and age. One aim was also to determine ¸ the nature and extent of displacement of PCB from the facade into the surrounding soil. Buildings built between 1952 and 1979 were chosen for the study. Three different media were sampled during the study; surface soil, plaster and paint. Samples were then analysed for PCB7 content. The results show that there is a difference in PCB usage in buildings of different age and usage type. Residential buildings and schools demonstrated higher PCB concentrations in both soil and plaster than buildings designated for office use, storage, or for industrial purposes. Buildings erected in the 1950s and 1960s also show a higher PCB concentration than buildings from a later date. It appears that the usage of PCB for these purposes decreased in the 1970s. Thirty percent of the soil samples showed a higher PCB concentration than the Norwegian action level. The soil samples tend to have a higher concentration than the corresponding plaster from the adjacent wall, which probably has its cause in the high soil organic matter contents that retains PCB. Plaster has not been considered a pollution source in previous studies; therefore this study demonstrates a new source that needs to be considered in emission calculations. 䊚 2003 Elsevier B.V. All rights reserved. Keywords: Polychlorinated biphenyls; Buildings; Construction materials; Plaster; Soil
1. Introduction There are no naturally occurring polychlorinated biphenyls (PCB); therefore all PCBs found in nature can be accredited to man-made materials. PCBs were commonly used in Norway from the early 1950s until the end of the 1970s, in building *Corresponding author. Tel.: q47-7390-4321; fax: q47-73921620. E-mail address: [email protected] (M. Andersson).
materials such as joint sealing materials, glue used in the production of double glazing, concrete, paint and plaster. PCB was, amongst other things, used in polyvinyl acetate (PVA) mixtures to improve the building properties of concrete and plaster (Waldum and Engelsen, 2003). The positive improvement of PVA on plaster and concrete was better flexibility of the material in the building stage as well as dried, increased resistance against mechanical erosion and improved adherence to a
0048-9697/04/$ - see front matter 䊚 2003 Elsevier B.V. All rights reserved. doi:10.1016/j.scitotenv.2003.11.014
140
M. Andersson et al. / Science of the Total Environment 325 (2004) 139–144
variety of other building materials. The use of PCBs was banned in Norway in 1979. Our study was conducted in order to determine the extent of PCB usage in plaster on building facades and whether PCB concentration varies ¸ according to building usage type and age. It was also critical to determine the nature and extent of displacement of PCB from the facade into the ¸ surrounding soil. The investigation concentrated on plaster and paint in buildings that lacked joint sealing material in order to avoid the influence from this source. It has been calculated that around 85 tonnes, or approximately 18% of existing PCB, is present in plaster in Norwegian buildings (SFT, 2002). Other possible sources of PCB are, as calculated, glue in double-glazing (200 t), condensators (105 t), joint sealants (50 t) and paint (10 t). The use of PCB in plaster and concrete cannot be accurately calculated as the level of PCB usage in the building processes differed from builder to builder. The south-western part of Norway, with Bergen as the centre, has demonstrated a higher usage of PVAy PCB in construction materials than any other area in Norway, mainly because of more severe climatic factors, such as rainfall, that necessitate easily applied building materials (Waldum and Engelsen, 2003). Most previous work regarding PCB in construction material in the Nordic countries has concentrated on PCB in joint sealing material ˚ (Jansson et al., 1997; Astebro, 1999; Sverud and Estensen, 2000; Hellman and Puhakka, 2001) and how the PCB has spread into the environment and surrounding materials. Studies have been conducted on the background level of PCB in Norway that show the soil concentration level of PCB to correlate with proximity to source region, land use and soil organic matter content (Meijer et al., 2002). Urban soil concentrations are very heterogeneous, but generally a factor of ;5–10 above soil in rural areas (Meijer et al., 2003). Average soil background PCB concentrations from remote areas on mainland Norway were determined to be 0.52 mgykg dry weight (total PCB) (Meijer et al., 2002). Although studies have attempted to calculate PCB emission data (Breivik et al., 2002), buildings as a source for
soil contamination were not taken into consideration. A number of studies of the sediments in the harbour area of Bergen show a high degree of PCB contamination (Lone and Systad, 1998a,b, 2001). However, these studies did not determine any specific source for the contamination. A large dredging project took place in 2002 in the fiord outside Bergen to remove PCB-contaminated sediments. A subsequent study of the harbour sediments conducted after the dredging still demonstrated an increasing level of PCB-contamination (Johnsen et al., 2003). This was likely due to an unknown source on land that had not been recognised previously. This unidentified source was, therefore, still active as a source for PCB being deposited into the fiord. To date, limited research has been conducted relating to the distribution of PCB in plaster in buildings of different age groups or in buildings of different usage type. Ottesen et al. (2000) showed that plaster is a source of severe PCB contamination to the urban environment. One other study focusing on PCB in paint in a restoration project analysed a single plaster sample that showed a very high PCB concentration (PCB7 336 mgykg) (Ruus and Mage, 2002). The Nor˚ wegian PCB concentration level for hazardous waste is 50 mgykg and the recommended action level in soil is 0.5 mgykg PCB. 2. Study area Bergen is situated in the south-west part of Norway (Fig. 1). A town with 220 000 inhabitants, it historically has had a lot of trading contact with different parts of Europe throughout the centuries. Therefore it is a town that is not typically Scandinavian in building style and is generally built in stone and concrete. This probably contributes to the enhanced problem of PCB pollution in Bergen when compared with other parts of Norway. More severe weather conditions may also have contributed to an increased usage of PCB. The average rainfall in the Bergen area is 2250 mm in contrast to the average of 770 mm for the rest of the country (information from Statistics Norway).
M. Andersson et al. / Science of the Total Environment 325 (2004) 139–144
141
time period 1950–1980. Buildings with an unknown or uncertain building year were excluded from the comparison of buildings built in different decades. A combined total of 46 official and residential buildings were chosen for this study. These were organised into three types: (i) residential buildings; (ii) schools; and finally (iii) office, storage and industrial buildings. A minimum of two samples were taken from each building, one plaster sample ¸ from the facade and one surface soil sample. Where possible, a paint sample was also collected. In total PCB was determined in 35 plaster samples, 43 surface soil samples and five paint samples. In cases where the extracted core only consisted of painted concrete no analysis was performed, since the analysed media was plaster. Soil samples that had been sampled where the soil surrounding a building recently had been changed were not analysed either, since the soil in those cases, in our opinion, had not had time to accumulate PCB in a comparative manner. 3.1. Sampling Fig. 1. The location of the investigation area in Bergen. The enclosed map shows the sampling points.
The harbour area of Bergen is an important economic centre, also in the production of seafood. However, at present, seafood consumption restrictions exist to limit human intake of potentially damaging chemicals, with PCB as a major contributor, taken up by the marine biota. Therefore, the lifting of these restrictions is critically dependent on the localisation of the PCB pollution sources and a better understanding of the nature of its migration pathways from the sources on land into the fiord. 3. Methods In order to determine a possible pattern in usage of PCB in the Bergen area, sampling was conducted in different parts of the town, in different types of buildings, and in buildings of different ages. Documentation was provided by Bergen Council detailing official buildings (council offices, schools, etc.) that were constructed in the
The surface soil samples (0–5 cm in depth) were collected adjacent to the buildings. The samples were then shipped in glass jars to the laboratory for PCB analysis. Samples of plaster and concrete were collected using a diamond core drill, where a core of 6 cm length and 4 cm in diameter was extracted from the facade. Where possible, the samples were taken using a hammer. The plaster was then separated from the concrete with a stone saw and manually crushed. The paint samples were collected from buildings where the paintwork was flaking. Two of the samples were manually crushed and sent to TAUW laboratory, while five samples were analysed as intact flakes at AnalyCen in Moss, Norway. 3.2. Analytical methods The PCB analyses were performed by TAUW laboratory in the Netherlands. The samples were first extracted using acetone and hexane. The extract was then dried using sodium sulphate,
142
M. Andersson et al. / Science of the Total Environment 325 (2004) 139–144
vaporised and subsequently cleaned over deactivated aluminium oxide and then analysed using a gas chromatograph with an electron capture detector (GC-ECD). The seven Dutch congeners of PCB (PCB7) were analysed with a detection limit of 1 mgykg. 4. Results The concentration of PCB7 detected in plaster ranged from non-detected to 290 mgykg, while the surface soil samples ranged from non-detected to 320 mgykg. Table 1 shows an overview of PCB concentrations in the different sampling media as well as differences in buildings of different age and usage. There is a clear tendency for the soil sample to show a higher concentration of PCB than the corresponding plaster on the wall. The same tendency exists for the paint samples that show a higher grade of contamination than the corresponding plaster. The high PCB concentration of the soil sample with the maximum concentration (320 mgykg) is possibly due to contamination from double-glazing windows, which had been stored in this location. This was found out after the analysis was made. If that sample were to be removed, because of contamination from another source than plaster, the median for the soil samples
Fig. 2. Graph showing the distribution of PCB concentrations in buildings of different age. Note the logarithmic scale on the y-axis.
would be 0.135 mgykg and the maximum concentration 41 mgykg. Buildings built during the 1970s show a much lesser degree of PCB contamination than buildings erected during the two previous decades (Table 1 and Fig. 2). This also applies to the surface soil samples. Residential buildings and schools show a higher level of PCB usage in the building process than do office-, storage- or industrial buildings (Table 1 and Fig. 3). 5. Discussion This is the first study of this kind and magnitude that has been conducted and therefore there are no
Table 1 An overview of the analysed samples, combined into several categories, by sampling media, building age and type of building Sampling medium
Min (PCB7, mgykg)
Max (PCB7, mgykg)
Median (PCB7, mgykg)
Number of samples
Surface soil Plaster Paint
-0.001 -0.001 0
320 290 1940
0.15 0.06 2.1
43 35 5
Building decade
Min (PCB7, mgykg) in plaster -0.001 -0.001 -0.001
Max (PCB7, mgykg) in plaster 1.1 290 0.012
Median (PCB7, mgykg) in plaster 0.028 0.015 0.001
Number of samples 15 14 4
Min (PCB7, mgykg) in plaster -0.001
Max (PCB7, mgykg) in plaster 0.044
Median (PCB7, mgykg) in plaster 0.018
Number of samples 9
-0.001 -0.001
290 75
0.42 0.012
10 19
1950–59 1960–69 1970–79 Type of building Officesystoragey industrial use Residential buildings Schools
M. Andersson et al. / Science of the Total Environment 325 (2004) 139–144
Fig. 3. Graph showing the distribution of PCB concentrations in buildings of different usage type. Note the logarithmic scale on the y-axis.
reference concentrations that could be used as comparison. Plaster as a medium, has not been considered a serious source of contamination in any of the Nordic countries. Most previous studies of PCB in construction materials have totally concentrated on joint sealants or glue used in double-glazing. However, this study demonstrates that building plaster is also a potential source for PCB pollution. Our data demonstrate that there is a significant difference in PCB usage between buildings of different use. Offices, storage and industrial buildings all show a lesser degree of contamination, relative to schools and residential housing, which show a marked increase of PCB in the building materials used. At some sampling sites the relative occurrence of the 7 PCB congeners varied between the plaster and the surrounding soil. However, it was also noted that in many sampling sites there was no variation in the congeners between the plaster and the soil. It is, as yet, uncertain which processes are responsible for this variation. It may be possible that this phenomenon is due to a renewal of the soil, since this difference in congener variation mostly took place at residential buildings where the exterior had been refurbished. Therefore, it did not seem appropriate to correlate the amount of PCB that has been transported into the soil with the age of the building. In official buildings, where no refurbishment of the soil around the building occurred, the PCB congeners show the same rela-
143
tive pattern, thus demonstrating that the PCB product used in plaster is reflected in the soil. A higher PCB concentration was found in soil than in the corresponding wall. It has been fairly well documented that PCB is retained well in soil with a high organic matter (OM) content (Meijer et al., 2002, 2003; Krauss and Wilcke, 2002), which probably stands in contrast to the other man-made media, such as plaster and paint, studied here. This is probably the reason why the PCB concentration is higher in the soil than in the corresponding plaster. There is also a differing degree of PCB contamination in buildings of different age. This difference in PCB concentration probably reflects the level of PCB use. The use must have been at its peak in the 1950s and 1960s gradually decreasing during the 1970s. Analyses of the paint flakes showed that these samples should be collected by another method. For example, on some paint samples there were three layers of material present, outermost was the paint itself (possibly consisting of several layers), then a PVA-liquid that was applied on top of the plaster for better adherence, and finally a thin layer of plaster is potentially present. It therefore proved difficult to distinguish the paint from any possible pre-painting treatment material. 6. Conclusions ● There is a difference in the pattern of PCB consumption in building material in Bergen. Buildings of different types and age show differing levels of PCB usage, so that the peak of PCB use occurred in the 1950s and 1960s and decreased in the 1970s. ● Residential buildings and schools show an increased level of PCB concentration in relation to buildings used for offices, storage or for industrial purposes. ● Thirty percent of the soil samples analysed exceed the Norwegian action level of 0.5 mgy kg in PCB concentrations. ● PCB concentrations in soil are higher than the plaster samples from the corresponding wall. This is likely due to higher OM contents in the soil.
144
M. Andersson et al. / Science of the Total Environment 325 (2004) 139–144
● Some locations demonstrate a difference in the relative occurrence of the seven congeners in soil and corresponding plaster. This is probably due to a renewal of soil around some buildings. Where the soil obviously was original, the pattern of the congeners was the same. ● This study shows that plaster as pollution source should be considered during refurbishing and demolition of buildings. More extensive studies are required to determine the complete usage pattern. References Breivik K, Sweetman A, Pacyna JF, Jones KC. Towards a global historical emission inventory for selected PCB congeners—a mass balance approach 2. Emissions. Sci Total Environ 2002;290:199 –224. Hellman S, Puhakka JA. Polychlorinated biphenyl (PCB) contamination of apartment building and its surroundings by construction block sealants. In: Salminen R, editor. ECOGEO 2000: International conference on practical applications in environmental geotechnology. Special Paper, Geological Survey of Finland, vol. 32 2001. p. 123 –127. ˚ Jansson B, Sandberg J, Johansson N, Asterbro A. PCB i fogmassor—stort eller litet problem? (translated title: PCB in joint sealants—large or small problem?) Naturvardsverket ˚ report. 1997. (p. 4697). Johnsen A, Rossland HK, Søbye E, Longva KS. Diffuse kilder til PCB og effektstudier i torsk og blaskjell ved Haakonsvern ˚ orlogsstasjon (translated title: Diffuse PCB sources and studies of cod and mussels by Haakonsvern naval base). FFI, Forsvarets forskningsinstitutt report 2003y01595. 2003. Krauss M, Wilcke W. Sorption strength of persistent organic pollutants in particle-size fractions of urban soils. Soil Sci Soc Am J 2002;66:430 –437. Lone S, Systad H. Bergen og Omland Havnevesen, utfyllingsomrade Nøstebukten, Miljøteknisk grunnundersøkelse ˚ (translated title: The harbour of Bergen and surrounding areas, filling area Nøstebukten). Noteby: Norsk Teknisk byggekontroll AyS report 51925-1, 1998. Lone S, Systad H. Mjellem & Karlsen Verft AS, Kai 4, forlengelse. Miljøteknisk grunnundersøkelser. (translated
title: Mjellem & Karlsen Verft AS, extended Dock 4). Noteby: Norsk Teknisk byggekontroll AyS report 52239-1, 1998b. Lone S, Systad H. Frydenlund, Damsgardsgaten 163-169. ˚ Miljøteknisk grunnundersøkelse. Noteby: Radgivende ˚ ingeniører MRIF report, 400511-1, 2001. Meijer SN, Steinnes E, Ockenden WA, Jones KC. Influence of Environmental variables on the spatial distribution of PCBs in Norwegian and U.K. soils: implications for global cycling. Environ Sci Technol 2002;36:2146 –2153. Meijer SN, Ockenden WA, Sweetman A, Breivik K, Grimalt JO, Jones KC. Global distribution and budget of PCBs and HCB in background surface soils: implications for sources and environmental processes. Environ Sci Technol 2003;37:667 –672. Ottesen RT, Volden T, Haugland T, Alexander J. Jordforurensning i Bergen—Oppfølgende undersøkelser av jordforurensning i barns lekemiljø i Sentrum-, Laksevag-, Løvstakken-, ˚ Sandviken og Landas ˚ bydeler. Helserisikovurderinger. (translated title: Soil pollution in Bergen—continuing studies of soil pollution in children’s play areas in the town areas: Centre of Bergen, Laksevag, ˚ Løvstakken, Sandviken and Landas), ˚ Geol Surv Norway Rep 2000.089. 2000. Ruus A, Mage ˚ A. Vedrørende oppfølging av funn av forhøyde PCB-konsentrasjoner i blaskjell fra Tyssedal. (translated ˚ title: Continuing studies of heightened PCB concentrations in mussels from Tyssedal). NIVA and Hardanger Environmental Centre Report. 2002. Sverud T, Estensen ASG. Identifisering, prøvetaking og analyse av fugemasse i bygninger for PCB. Rapportering av prosjekt utført i 1998 og oppfølgingsprosjekt utført i 1999. (translated title: Identification, sampling and PCB analyses of joint sealants in buildings. Project report from 1998 and continuing studies in 1999). Kobygg: Veritas report 20003073, 2000. SFT. Miljøgifter i produkter i 2000. (translated title: Environmental poisons in products in 2000). SFT: Statens forurensningstilsyn report, TA-1894y2002, 2002. Waldum AJ, Engelsen CJ. PCB-holdige materialer i puss og betong. En historisk oppsummering og kjemisk analyse av 10 utvalgte malingsprøver. 2003 (translated title: A historical summary and chemical analysis of 10 paint samples). Norges byggforskningsinstitutt report, O-10786. 2002. ˚ ¨ Astebro A. Inventering av fogmassor med PCB. Handbok for ¨ fastighetsagare. 1999. (translated title: Inventory of joint ¨ sealants with PCB. Handbook for property owners). Miljo¨ forvaltningen i Stockholm report, 1999.
Building materials as a source of PCB pollution in Bergen, Norway M. Andersson*, R.T. Ottesen, T. Volden Department of Geochemistry and Hydrogeology, Geological Survey of Norway, 7491 Trondheim, Norway Received 26 September 2003; received in revised form 17 November 2003; accepted 21 November 2003
Abstract Although PCB is a globally recognised pollutant, an understanding of its transport from man-made building materials to the environment is poorly constrained. This paper presents data from a study that was conducted in order ¸ to determine the extent of PCB usage in plaster on building facades in the Bergen area, Norway. The study was to determine whether PCB concentrations vary according to building usage type and age. One aim was also to determine ¸ the nature and extent of displacement of PCB from the facade into the surrounding soil. Buildings built between 1952 and 1979 were chosen for the study. Three different media were sampled during the study; surface soil, plaster and paint. Samples were then analysed for PCB7 content. The results show that there is a difference in PCB usage in buildings of different age and usage type. Residential buildings and schools demonstrated higher PCB concentrations in both soil and plaster than buildings designated for office use, storage, or for industrial purposes. Buildings erected in the 1950s and 1960s also show a higher PCB concentration than buildings from a later date. It appears that the usage of PCB for these purposes decreased in the 1970s. Thirty percent of the soil samples showed a higher PCB concentration than the Norwegian action level. The soil samples tend to have a higher concentration than the corresponding plaster from the adjacent wall, which probably has its cause in the high soil organic matter contents that retains PCB. Plaster has not been considered a pollution source in previous studies; therefore this study demonstrates a new source that needs to be considered in emission calculations. 䊚 2003 Elsevier B.V. All rights reserved. Keywords: Polychlorinated biphenyls; Buildings; Construction materials; Plaster; Soil
1. Introduction There are no naturally occurring polychlorinated biphenyls (PCB); therefore all PCBs found in nature can be accredited to man-made materials. PCBs were commonly used in Norway from the early 1950s until the end of the 1970s, in building *Corresponding author. Tel.: q47-7390-4321; fax: q47-73921620. E-mail address: [email protected] (M. Andersson).
materials such as joint sealing materials, glue used in the production of double glazing, concrete, paint and plaster. PCB was, amongst other things, used in polyvinyl acetate (PVA) mixtures to improve the building properties of concrete and plaster (Waldum and Engelsen, 2003). The positive improvement of PVA on plaster and concrete was better flexibility of the material in the building stage as well as dried, increased resistance against mechanical erosion and improved adherence to a
0048-9697/04/$ - see front matter 䊚 2003 Elsevier B.V. All rights reserved. doi:10.1016/j.scitotenv.2003.11.014
140
M. Andersson et al. / Science of the Total Environment 325 (2004) 139–144
variety of other building materials. The use of PCBs was banned in Norway in 1979. Our study was conducted in order to determine the extent of PCB usage in plaster on building facades and whether PCB concentration varies ¸ according to building usage type and age. It was also critical to determine the nature and extent of displacement of PCB from the facade into the ¸ surrounding soil. The investigation concentrated on plaster and paint in buildings that lacked joint sealing material in order to avoid the influence from this source. It has been calculated that around 85 tonnes, or approximately 18% of existing PCB, is present in plaster in Norwegian buildings (SFT, 2002). Other possible sources of PCB are, as calculated, glue in double-glazing (200 t), condensators (105 t), joint sealants (50 t) and paint (10 t). The use of PCB in plaster and concrete cannot be accurately calculated as the level of PCB usage in the building processes differed from builder to builder. The south-western part of Norway, with Bergen as the centre, has demonstrated a higher usage of PVAy PCB in construction materials than any other area in Norway, mainly because of more severe climatic factors, such as rainfall, that necessitate easily applied building materials (Waldum and Engelsen, 2003). Most previous work regarding PCB in construction material in the Nordic countries has concentrated on PCB in joint sealing material ˚ (Jansson et al., 1997; Astebro, 1999; Sverud and Estensen, 2000; Hellman and Puhakka, 2001) and how the PCB has spread into the environment and surrounding materials. Studies have been conducted on the background level of PCB in Norway that show the soil concentration level of PCB to correlate with proximity to source region, land use and soil organic matter content (Meijer et al., 2002). Urban soil concentrations are very heterogeneous, but generally a factor of ;5–10 above soil in rural areas (Meijer et al., 2003). Average soil background PCB concentrations from remote areas on mainland Norway were determined to be 0.52 mgykg dry weight (total PCB) (Meijer et al., 2002). Although studies have attempted to calculate PCB emission data (Breivik et al., 2002), buildings as a source for
soil contamination were not taken into consideration. A number of studies of the sediments in the harbour area of Bergen show a high degree of PCB contamination (Lone and Systad, 1998a,b, 2001). However, these studies did not determine any specific source for the contamination. A large dredging project took place in 2002 in the fiord outside Bergen to remove PCB-contaminated sediments. A subsequent study of the harbour sediments conducted after the dredging still demonstrated an increasing level of PCB-contamination (Johnsen et al., 2003). This was likely due to an unknown source on land that had not been recognised previously. This unidentified source was, therefore, still active as a source for PCB being deposited into the fiord. To date, limited research has been conducted relating to the distribution of PCB in plaster in buildings of different age groups or in buildings of different usage type. Ottesen et al. (2000) showed that plaster is a source of severe PCB contamination to the urban environment. One other study focusing on PCB in paint in a restoration project analysed a single plaster sample that showed a very high PCB concentration (PCB7 336 mgykg) (Ruus and Mage, 2002). The Nor˚ wegian PCB concentration level for hazardous waste is 50 mgykg and the recommended action level in soil is 0.5 mgykg PCB. 2. Study area Bergen is situated in the south-west part of Norway (Fig. 1). A town with 220 000 inhabitants, it historically has had a lot of trading contact with different parts of Europe throughout the centuries. Therefore it is a town that is not typically Scandinavian in building style and is generally built in stone and concrete. This probably contributes to the enhanced problem of PCB pollution in Bergen when compared with other parts of Norway. More severe weather conditions may also have contributed to an increased usage of PCB. The average rainfall in the Bergen area is 2250 mm in contrast to the average of 770 mm for the rest of the country (information from Statistics Norway).
M. Andersson et al. / Science of the Total Environment 325 (2004) 139–144
141
time period 1950–1980. Buildings with an unknown or uncertain building year were excluded from the comparison of buildings built in different decades. A combined total of 46 official and residential buildings were chosen for this study. These were organised into three types: (i) residential buildings; (ii) schools; and finally (iii) office, storage and industrial buildings. A minimum of two samples were taken from each building, one plaster sample ¸ from the facade and one surface soil sample. Where possible, a paint sample was also collected. In total PCB was determined in 35 plaster samples, 43 surface soil samples and five paint samples. In cases where the extracted core only consisted of painted concrete no analysis was performed, since the analysed media was plaster. Soil samples that had been sampled where the soil surrounding a building recently had been changed were not analysed either, since the soil in those cases, in our opinion, had not had time to accumulate PCB in a comparative manner. 3.1. Sampling Fig. 1. The location of the investigation area in Bergen. The enclosed map shows the sampling points.
The harbour area of Bergen is an important economic centre, also in the production of seafood. However, at present, seafood consumption restrictions exist to limit human intake of potentially damaging chemicals, with PCB as a major contributor, taken up by the marine biota. Therefore, the lifting of these restrictions is critically dependent on the localisation of the PCB pollution sources and a better understanding of the nature of its migration pathways from the sources on land into the fiord. 3. Methods In order to determine a possible pattern in usage of PCB in the Bergen area, sampling was conducted in different parts of the town, in different types of buildings, and in buildings of different ages. Documentation was provided by Bergen Council detailing official buildings (council offices, schools, etc.) that were constructed in the
The surface soil samples (0–5 cm in depth) were collected adjacent to the buildings. The samples were then shipped in glass jars to the laboratory for PCB analysis. Samples of plaster and concrete were collected using a diamond core drill, where a core of 6 cm length and 4 cm in diameter was extracted from the facade. Where possible, the samples were taken using a hammer. The plaster was then separated from the concrete with a stone saw and manually crushed. The paint samples were collected from buildings where the paintwork was flaking. Two of the samples were manually crushed and sent to TAUW laboratory, while five samples were analysed as intact flakes at AnalyCen in Moss, Norway. 3.2. Analytical methods The PCB analyses were performed by TAUW laboratory in the Netherlands. The samples were first extracted using acetone and hexane. The extract was then dried using sodium sulphate,
142
M. Andersson et al. / Science of the Total Environment 325 (2004) 139–144
vaporised and subsequently cleaned over deactivated aluminium oxide and then analysed using a gas chromatograph with an electron capture detector (GC-ECD). The seven Dutch congeners of PCB (PCB7) were analysed with a detection limit of 1 mgykg. 4. Results The concentration of PCB7 detected in plaster ranged from non-detected to 290 mgykg, while the surface soil samples ranged from non-detected to 320 mgykg. Table 1 shows an overview of PCB concentrations in the different sampling media as well as differences in buildings of different age and usage. There is a clear tendency for the soil sample to show a higher concentration of PCB than the corresponding plaster on the wall. The same tendency exists for the paint samples that show a higher grade of contamination than the corresponding plaster. The high PCB concentration of the soil sample with the maximum concentration (320 mgykg) is possibly due to contamination from double-glazing windows, which had been stored in this location. This was found out after the analysis was made. If that sample were to be removed, because of contamination from another source than plaster, the median for the soil samples
Fig. 2. Graph showing the distribution of PCB concentrations in buildings of different age. Note the logarithmic scale on the y-axis.
would be 0.135 mgykg and the maximum concentration 41 mgykg. Buildings built during the 1970s show a much lesser degree of PCB contamination than buildings erected during the two previous decades (Table 1 and Fig. 2). This also applies to the surface soil samples. Residential buildings and schools show a higher level of PCB usage in the building process than do office-, storage- or industrial buildings (Table 1 and Fig. 3). 5. Discussion This is the first study of this kind and magnitude that has been conducted and therefore there are no
Table 1 An overview of the analysed samples, combined into several categories, by sampling media, building age and type of building Sampling medium
Min (PCB7, mgykg)
Max (PCB7, mgykg)
Median (PCB7, mgykg)
Number of samples
Surface soil Plaster Paint
-0.001 -0.001 0
320 290 1940
0.15 0.06 2.1
43 35 5
Building decade
Min (PCB7, mgykg) in plaster -0.001 -0.001 -0.001
Max (PCB7, mgykg) in plaster 1.1 290 0.012
Median (PCB7, mgykg) in plaster 0.028 0.015 0.001
Number of samples 15 14 4
Min (PCB7, mgykg) in plaster -0.001
Max (PCB7, mgykg) in plaster 0.044
Median (PCB7, mgykg) in plaster 0.018
Number of samples 9
-0.001 -0.001
290 75
0.42 0.012
10 19
1950–59 1960–69 1970–79 Type of building Officesystoragey industrial use Residential buildings Schools
M. Andersson et al. / Science of the Total Environment 325 (2004) 139–144
Fig. 3. Graph showing the distribution of PCB concentrations in buildings of different usage type. Note the logarithmic scale on the y-axis.
reference concentrations that could be used as comparison. Plaster as a medium, has not been considered a serious source of contamination in any of the Nordic countries. Most previous studies of PCB in construction materials have totally concentrated on joint sealants or glue used in double-glazing. However, this study demonstrates that building plaster is also a potential source for PCB pollution. Our data demonstrate that there is a significant difference in PCB usage between buildings of different use. Offices, storage and industrial buildings all show a lesser degree of contamination, relative to schools and residential housing, which show a marked increase of PCB in the building materials used. At some sampling sites the relative occurrence of the 7 PCB congeners varied between the plaster and the surrounding soil. However, it was also noted that in many sampling sites there was no variation in the congeners between the plaster and the soil. It is, as yet, uncertain which processes are responsible for this variation. It may be possible that this phenomenon is due to a renewal of the soil, since this difference in congener variation mostly took place at residential buildings where the exterior had been refurbished. Therefore, it did not seem appropriate to correlate the amount of PCB that has been transported into the soil with the age of the building. In official buildings, where no refurbishment of the soil around the building occurred, the PCB congeners show the same rela-
143
tive pattern, thus demonstrating that the PCB product used in plaster is reflected in the soil. A higher PCB concentration was found in soil than in the corresponding wall. It has been fairly well documented that PCB is retained well in soil with a high organic matter (OM) content (Meijer et al., 2002, 2003; Krauss and Wilcke, 2002), which probably stands in contrast to the other man-made media, such as plaster and paint, studied here. This is probably the reason why the PCB concentration is higher in the soil than in the corresponding plaster. There is also a differing degree of PCB contamination in buildings of different age. This difference in PCB concentration probably reflects the level of PCB use. The use must have been at its peak in the 1950s and 1960s gradually decreasing during the 1970s. Analyses of the paint flakes showed that these samples should be collected by another method. For example, on some paint samples there were three layers of material present, outermost was the paint itself (possibly consisting of several layers), then a PVA-liquid that was applied on top of the plaster for better adherence, and finally a thin layer of plaster is potentially present. It therefore proved difficult to distinguish the paint from any possible pre-painting treatment material. 6. Conclusions ● There is a difference in the pattern of PCB consumption in building material in Bergen. Buildings of different types and age show differing levels of PCB usage, so that the peak of PCB use occurred in the 1950s and 1960s and decreased in the 1970s. ● Residential buildings and schools show an increased level of PCB concentration in relation to buildings used for offices, storage or for industrial purposes. ● Thirty percent of the soil samples analysed exceed the Norwegian action level of 0.5 mgy kg in PCB concentrations. ● PCB concentrations in soil are higher than the plaster samples from the corresponding wall. This is likely due to higher OM contents in the soil.
144
M. Andersson et al. / Science of the Total Environment 325 (2004) 139–144
● Some locations demonstrate a difference in the relative occurrence of the seven congeners in soil and corresponding plaster. This is probably due to a renewal of soil around some buildings. Where the soil obviously was original, the pattern of the congeners was the same. ● This study shows that plaster as pollution source should be considered during refurbishing and demolition of buildings. More extensive studies are required to determine the complete usage pattern. References Breivik K, Sweetman A, Pacyna JF, Jones KC. Towards a global historical emission inventory for selected PCB congeners—a mass balance approach 2. Emissions. Sci Total Environ 2002;290:199 –224. Hellman S, Puhakka JA. Polychlorinated biphenyl (PCB) contamination of apartment building and its surroundings by construction block sealants. In: Salminen R, editor. ECOGEO 2000: International conference on practical applications in environmental geotechnology. Special Paper, Geological Survey of Finland, vol. 32 2001. p. 123 –127. ˚ Jansson B, Sandberg J, Johansson N, Asterbro A. PCB i fogmassor—stort eller litet problem? (translated title: PCB in joint sealants—large or small problem?) Naturvardsverket ˚ report. 1997. (p. 4697). Johnsen A, Rossland HK, Søbye E, Longva KS. Diffuse kilder til PCB og effektstudier i torsk og blaskjell ved Haakonsvern ˚ orlogsstasjon (translated title: Diffuse PCB sources and studies of cod and mussels by Haakonsvern naval base). FFI, Forsvarets forskningsinstitutt report 2003y01595. 2003. Krauss M, Wilcke W. Sorption strength of persistent organic pollutants in particle-size fractions of urban soils. Soil Sci Soc Am J 2002;66:430 –437. Lone S, Systad H. Bergen og Omland Havnevesen, utfyllingsomrade Nøstebukten, Miljøteknisk grunnundersøkelse ˚ (translated title: The harbour of Bergen and surrounding areas, filling area Nøstebukten). Noteby: Norsk Teknisk byggekontroll AyS report 51925-1, 1998. Lone S, Systad H. Mjellem & Karlsen Verft AS, Kai 4, forlengelse. Miljøteknisk grunnundersøkelser. (translated
title: Mjellem & Karlsen Verft AS, extended Dock 4). Noteby: Norsk Teknisk byggekontroll AyS report 52239-1, 1998b. Lone S, Systad H. Frydenlund, Damsgardsgaten 163-169. ˚ Miljøteknisk grunnundersøkelse. Noteby: Radgivende ˚ ingeniører MRIF report, 400511-1, 2001. Meijer SN, Steinnes E, Ockenden WA, Jones KC. Influence of Environmental variables on the spatial distribution of PCBs in Norwegian and U.K. soils: implications for global cycling. Environ Sci Technol 2002;36:2146 –2153. Meijer SN, Ockenden WA, Sweetman A, Breivik K, Grimalt JO, Jones KC. Global distribution and budget of PCBs and HCB in background surface soils: implications for sources and environmental processes. Environ Sci Technol 2003;37:667 –672. Ottesen RT, Volden T, Haugland T, Alexander J. Jordforurensning i Bergen—Oppfølgende undersøkelser av jordforurensning i barns lekemiljø i Sentrum-, Laksevag-, Løvstakken-, ˚ Sandviken og Landas ˚ bydeler. Helserisikovurderinger. (translated title: Soil pollution in Bergen—continuing studies of soil pollution in children’s play areas in the town areas: Centre of Bergen, Laksevag, ˚ Løvstakken, Sandviken and Landas), ˚ Geol Surv Norway Rep 2000.089. 2000. Ruus A, Mage ˚ A. Vedrørende oppfølging av funn av forhøyde PCB-konsentrasjoner i blaskjell fra Tyssedal. (translated ˚ title: Continuing studies of heightened PCB concentrations in mussels from Tyssedal). NIVA and Hardanger Environmental Centre Report. 2002. Sverud T, Estensen ASG. Identifisering, prøvetaking og analyse av fugemasse i bygninger for PCB. Rapportering av prosjekt utført i 1998 og oppfølgingsprosjekt utført i 1999. (translated title: Identification, sampling and PCB analyses of joint sealants in buildings. Project report from 1998 and continuing studies in 1999). Kobygg: Veritas report 20003073, 2000. SFT. Miljøgifter i produkter i 2000. (translated title: Environmental poisons in products in 2000). SFT: Statens forurensningstilsyn report, TA-1894y2002, 2002. Waldum AJ, Engelsen CJ. PCB-holdige materialer i puss og betong. En historisk oppsummering og kjemisk analyse av 10 utvalgte malingsprøver. 2003 (translated title: A historical summary and chemical analysis of 10 paint samples). Norges byggforskningsinstitutt report, O-10786. 2002. ˚ ¨ Astebro A. Inventering av fogmassor med PCB. Handbok for ¨ fastighetsagare. 1999. (translated title: Inventory of joint ¨ sealants with PCB. Handbook for property owners). Miljo¨ forvaltningen i Stockholm report, 1999.