Determinations of PCB within a project to develop cleanup methods for PCB-containing elastic sealant used in outdoor joints between concrete blocks in buildings† M. Sundahl*, E. Sikander, B. Ek-Olausson, A. Hjorthage, L. Rosell and M. Tornevall SP Swedish National Testing and Research Institute, P.O. Box 857, SE-501 15 Bora° s, Sweden. E-mail: mikael.sundahl@sp.se; Fax: 46-33-120254; Tel: 46-33-165666 Received 29th March 1999, Accepted 1st June 1999 Determinations of PCB were carried out as part of a project aimed at developing cleanup methods for PCBcontaining elastic sealant used in outdoor joints between concrete blocks.The goals of the project were to develop methods, which minimise the spread of PCB to the outdoor environment and to indoor air, and which keep the PCB levels as low as reasonably possible in the workplace environment whilst removing the elastic sealant.The following PCB determinations were carried out: (1) concentration in the elastic sealant; (2) concentration in the concrete close to the sealant; (3) concentration in soil; (4) concentration in the indoor air; and (5) concentration in the air in the workplace environment.The cleanup process consisted of a number of diVerent steps: (1) cutting the elastic sealant with an oscillating knife; (2) grinding the concrete with a mechanical machine; (3) sawing the concrete with a mechanical saw and (4) cutting the concrete with a mechanical chisel. In all these diVerent steps a high capacity vacuum cleaner connected to the machines was used. The elastic sealant contained 4.7 to 8.1% total PCB of a technical product with a composition most similar to Clophene A40.The concrete close to the sealant (first 2 mm) contained 0.12 and 1.7% total PCB at two diVerent places. The pattern of the PCB in the concrete resembled that of the sealant. PCB concentrations in the soil from the ground close to the building were 0.1 and 0.3 ppm at two diVerent places before the remedial action.The source of the PCB in the soil is most likely the sealant as the PCB pattern is similar for the two materials. The PCB levels in the workplace air at the beginning of the project, when the techniques were not fully developed, were generally above the occupational exposure limit of 10 mg m-3 (up to 120 mg m-3). Later when the techniques were optimised to better take care of dust and gases produced during the cutting and grinding etc., the levels were below or close to 10 mg m-3.The pattern of the PCB in the workplace air was diVerent from that of the sealant and contained higher levels of lighter components. The PCB concentrations in the indoor air were measured before and during the remedial process. The levels were around 600 ng m-3 and there was no significant increase during the removal of the sealant.The PCB level after the remedial action will be measured later. The pattern of the PCB in the indoor air was diVerent from that of the sealant as well as from that of the workplace air. Higher levels of the lighter PCB were present indoors compared to the composition in both workplace air and in sealant.Extracts of PCB were analysed by GC-MS with a SIM method (selected ion monitoring). Standard procedures were used for extraction of solid materials. For the air samples an OVS tube was used with XAD-2 as adsorbent. The filter and adsorbent were extracted with toluene. This work shows that it is important to perform remedial action of PCB-containing elastic sealant as: (1) there is a spread of PCB to the indoor air giving high enough concentrations to make this the main PCB load on humans living in the apartments studied in this project; (2) large amounts are spread to the soil from these sealants; and (3) many of the PCBcontaining elastic sealants used need to be changed from a functional perspective.amount of PCB added in such sealants ranges from 100 to Introduction 500 tonne.3 It has been shown that PCB from sealants leak to PCB are persistent organic pollutants; their eVects on the the environment, into the surrounding ground4 as well as to environment have been described in numerous papers.In this indoor air.5 It is most likely that PCB from sealants have paper we refer to a single source with the aim of neither already spread to refuse dumps for used building materials, choosing the most extensive papers nor of selecting the first due to a previous lack of knowledge about the problem.Many or most recent ones.1 PCB is a collective name for 209 diVerent of the sealants are getting old and need to be changed from a chemical compounds, congeners, with biphenyl and chlorine functional perspective and it is far more expensive to take care as the common parts in the molecules.Several diVerent PCB of PCB when it is diluted with other garbage on a refuse dump technical products have been produced; these include Clophene than to take care of it at source. Thus, there is a need for A40, A50, A60 and Arochlor 1242, 1248, 1254 and 1260. eYcient methods for exchanging these sealants with new PCB free materials, which minimise the spread of PCB to the From 1956 to 1972 PCB were extensively used in Sweden as environment, to indoor air and to the workplace environment.plasticizers in elastic sealants used in joints between concrete In this paper results are presented mainly from chemical blocks in buildings. In particular, products based on thiokol,2 measurements from a project to develop cleanup methods for a polysulfide, often contain PCB.In Sweden, the estimated PCB-containing elastic sealants. Technical descriptions of the cleanup process, where relevant, are given as background information.6 †Presented at AIRMON ’99, Geilo, Norway, February 10–14, 1999. J. Environ. Monit., 1999, 1, 383–387 3831 l min-1.The flow rate was checked using a calibrated flow Methods and background information meter. The filter and first section of the adsorbent were Taking away the sealant extracted together and the backup section of the adsorbent was extracted separately. The materials were mixed with The cleanup process consisted of diVerent steps: (1) cutting toluene (5 ml ) with 1,2,3,4-tetrachloronaphthalene (50 ml with the elastic sealant with an oscillating knife; (2) grinding the concentration 10 ng ml-1) added and extracted in an ultrasonic concrete with a mechanical machine; (3) sawing the concrete bath for 0.5 h.The method is partly validated by OSHA. with a mechanical saw; and (4) cutting the concrete with a Extraction eYciencies were greater than 93%. Breakthrough mechanical chisel.In all these diVerent steps a high capacity was not detected, for 125 l air, and the average recoveries were vacuum cleaner connected to the machines was used. The 95%. The breakthrough in all our measurements was less than vacuum cleaner had a capacity of 425 m3 h-1 (maximum 2%. Blank tubes were also extracted and no PCB could flow). The dust was separated in three steps with cyclone, fine be detected. filter and micro filter.Analysis of PCB extracts Description of the building All extracts were analysed by a gas chromatography mass The building was eight storeys high and was located in spectrometry selected ion monitoring (GC-MS-SIM) method. Bergsjo� n, Go�teborg, Sweden. Wall structures were constructed The instrument used was an HP 5890 GC with an MSD 5970.of concrete blocks and the joints were filled with a PCBThe column used for the GC separation was a BPX containing elastic sealant which was in bad condition and had 25 m×0.25 mm. The parameters used were as follows: to be replaced. The length of the sealants was about 4000 m 70 ml min-1 split flow, 300 °C injector temperature splitless which corresponds to about 1500 kg sealant; with an average injection, 100 °C for 1 min then 5 °Cmin-1 to 300 °C.MSD content of 6% PCB this corresponds to about 90 kg of PCB. parameters: SIM detection 3- and 4-Cl congeners starting at Extraction and sampling of PCB 6 min, Cl m/z 256.0, 258.0, 254.0, 220.0 and 292.0; 4-Cl congener starting at 23.4 min, m/z 292.0, 290.0 and 288.0; To determine the PCB concentration in the sealant an extraction internal standard starting at 24.3 min, m/z 263.9, 265.d with heptane was performed. Sealant (100 mg) was immersed in 194.0; 5-Cl starting at 25.5 min, m/z 326.0, 324.0 and 322.0; heptane (3 ml ) and extracted in an ultrasonic bath (1 h).The 6-Cl congener starting at 30.4 min, m/z 360.0, 362.0 and 356.0; extraction procedure was repeated twice.The extracts were col- 7-Cl congener starting at 33.0 min, m/z 393.9, 395.9 and 390.0. lected and washed with concentrated sulfuric acid (1 ml ) and To secure the identification, retention times and relative peak then diluted with heptane (100 ml final volume). heights of the SIM masses were used. The PCB congeners 28, 1,2,3,4-Tetrachloronaphthalene (100 ml with concentration 10 ng 52, 101, 118, 138, 153 and 180 were quantified by this method ml-1 to 10 ml extract) was added as an internal standard.The and the sum of these is denoted by S7PCB. To determine the internal standard was added after the extraction procedure. The total PCB content the results were compared with the PCB extraction eYciency was not determined, but the repeated procontents in the technical products and then S7PCB was ceduremost likely gives eYciencies greater than 90% (by comparimultiplied by the appropriate correction factor.7 For air son with similar compounds in similar matrices).samples which do not have the same composition as any To determine the PCB concentration in soil an extraction with technical product a conversion factor of 6 was used to obtain heptane was performed.Larger stones were taken away and soil the total PCB concentration from S4PCB.5 S4PCB is the sum (1–3 g) was mixed with heptane (3 ml ) and extracted in an of the congeners 28, 52, 101 and 138. ultrasonic bath (1 h). The extraction procedure was repeated three times. The extracts were collected and washed (50 ml 0.2M Quantification and detection limits and accuracy and precision NaCl in 0.1M H3PO4).The water phase was extracted with in the measurements heptane. The collected heptane phase was evaporated (ca. 3ml) and washed with concentrated sulfuric acid (1 ml ) and then The quantification limit of a single congener was 4 pg ml-1 diluted (5 ml ). 1,2,3,4-Tetrachloronaphthalene (50 ml with con- and the detection limit was 1 pg ml-1.The total accuracy for centration 10 ng ml-1 to 5 ml extract) was added as an internal the diVerent measurements was estimated to be better than standard. The internal standard was added after the extraction ±21%. The accuracy was estimated by adding an estimated procedure. The extraction eYciency was not determined, but the accuracy for the sampling–extraction step of ±15% and for repeated procedure most likely gives eYciencies greater than the analytical step of±15%. To obtain a value of the combined 90%.To determine the dry weight (dw) of the soil sample, soil repeatability in the analysis and extraction step, 18 sealants was dried (24 h, 105 °C) and weighed before and after drying. were extracted and analysed twice. The results gave at 95% To determine the PCB concentration in concrete, ground confidence level a spread of ±12%.concrete was collected and the powder was extracted with heptane. Ground concrete (0.3 g) was mixed with heptane (3 ml ), Results and discussion extracted in an ultrasonic bath (1 h) and the sediment was then Elastic sealant allowed to settle. The extraction procedure was repeated twice.The extractswere collectedandwashedwith concentrated sulfuric The elastic sealant contained from 4.7 to 8.1% PCB of a acid (1 ml) and then diluted with heptane (100 ml final volume). technical product similar to Clophene A40 (Table 1). Most 1,2,3,4-Tetrachloronaphthalene (100 ml with concentration likely, the original PCB content in the sealant was higher as 10 ng ml-1 to 10 ml extract) was added as an internal standard.PCB certainly was added from a functional perspective to The internal standard was added after the extraction procedure. plasticize the sealant. The sealant is 30 years old and PCB The extraction eYciency was not determined, but the repeated have leaked to the surrounding ground and to indoor air as procedure most likely gives eYciencies greater than 90%.is shown below. To determine the PCB concentration in air an OVS tube with XAD-2 as an adsorbent was used. An air sampling pump, Soil SKC aircheck sampler model 224-PCXR7, was used and depending on whether the air was from indoor or from the The soil from the ground was contaminated by PCB. The results are summarised in Table 2. A 2 cm portion was taken workplace environment the air volumes sampled were around 60 l and around 1500 l, respectively at a flow rate of around from the top layer of the soil from a 0.5×0.5 m surface.The 384 J. Environ. Monit., 1999, 1, 383–387Table 1 Partial content (normalised to highest) of the 7 selected PCB congeners in sealant and the technical product Clophene A40 and sum concentrations of PCB. For Clophene A40 the conversion factor from S7PCB to total PCB concentration is 5.3.The results reported in the same row are results obtained for the same sample but with separate extractions and analyses, thus giving an indication of the repeatability Partial content of the respective PCB in sample and product content PCB Congener 28 52 101 118 138 153 180 Content S7PCB Total content Clophene A40 0.19 0.39 0.14 0.13 0.05 0.06 0.01 Sealant 0.17 0.30 0.29 0.17 0.05 0.03 0.01 9.4 and 8.8 g kg 5.1 and 4.7% Sealant 0.19 0.38 0.24 0.15 0.04 0.02 0.01 14 and 15 g kg 7.4 and 8.1% Table 2 Results obtained for soil samples taken from the ground before and after the refinishing process.The results reported in the same row are results obtained for the same sample but with separate extractions and analyses, thus giving an indication of the repeatability S7PCB (mg kg-1 dw), before S7PCB (mg kg-1 dw), after Soil 1 1–1.5 m from building 0.013 and 0.028a 0.080 and 0.088 Soil 2 1–1.5 m from building 0.048 and 0.062 Soil 3 10–10.5 m from building Below detection limit 0.026 and 0.025a aThese values are below the quantification limit, at this dw, S7PCB 0.04 mg kg-1 dw.The detection limit is, at this dw, S7PCB 0.01 mg kg-1 dw. results obtained indicate that the soil was contaminated during We conclude that no significant increase of the PCB levels indoors occurs during the remedial process. The PCB sample the work involved in removing the sealant. See below for an estimate of the amount of PCB that was spread during the of indoor air taken after the remedial action was taken shortly after the process was finished and we believe that the concen- remedial process.tration will drop considerably within the next half year. Concrete Workplace air The concrete close to the sealant was contaminated by PCB The results are summarised in Table 4. The air was sampled in but the concentrations drop quickly with the distance from the breathing zone of the worker.The quantification limit of the joint. The results are summarised in Fig. 1. The first layer PCB in used air was #2 mg m-3 with an operating range of up from which the sealant was cut has a high PCB concentration to #80 mg m-3. An important conclusion drawn from the which shows the need for grinding the concrete surface to get measurements is that workers should use some kind of breathing rid of the PCB in this layer.It is reasonable from a practical protection due to the levels of PCB present in workplace air. point of view to grind a few millimetres of the concrete but Furthermore, smallmistakes in the remedial procedure can result not much more. in high PCB concentrations; therefore it is important to be careful during all the steps involved in the procedure.Indoor air Measurements of the PCB concentration in the air located after The results are summarised in Table 3. The quantification limit the vacuum cleaner at the vent were also carried out and the with the used air sampling volumes was #50 ng m-3 with an total PCB concentrations were found to be 15 and 17 mg m-3 in operating range up to #2000 ng m-3.Rather high levels of the new and used filter, respectively. PCB were found in the indoor air in the building. This can most likely be coupled to the PCB present in the sealant. Comparison of the PCB pattern in diVerent samples Previous investigations of buildings in Germany showed that The relative contents of the 7 indicator congeners are diVerent PCB from sealants can leak to indoor air and particularly for diVerent types of samples taken in the project reported in when lower chlorinated technical products are used as sealants this paper (Table 5).Some important conclusions can be higher concentrations are found indoors.8 There is also a drawn. The concentration of the PCB with low chlorine correlation between PCB concentration in indoor air and content is important for the level of the concentrations in air outdoor temperature.5 At higher temperatures the concenas the PCB found in air is dominated by lighter more volatile tration indoors gets higher.We have not attempted to study PCB. The PCB found in soil and concrete most likely originate this correlation further in this project due to limited resources.from the sealant as there is a similarity between the PCB pattern found for these three materials. The diVerent air measurements that is, workplace environment, air located after the vacuum cleaner and indoor air, show diVerent patterns. Looking at the 7 indicator congeners, indoor air is dominated by PCB 28 and 52, air after the vacuum cleaner is dominated by 28, 52 and 101 and workplace environment is dominated by PCB 28, 52, 101 and 118.It is also important that there is no significant diVerence between the PCB pattern of the indoor air before and during the remedial action which indicates that the remedial action does not aVect the PCB level indoors. Fig. 1 Total PCB concentration in diVerent layers of concrete close to Estimation of the PCB spread during the remedial action the surface where the sealant was cut away.The results reported in the same column at the same depth are results obtained for the same The estimations are rough but give a feeling of the magnitude sample, but with separate extractions and analyses, thereby giving an of the PCB spread and the amount of PCB left in the building indication of repeatability. The diVerent columns represent results obtained for samples taken from diVerent locations.after remedial action had been taken. J. Environ. Monit., 1999, 1, 383–387 385Table 3 PCB concentrations in indoor air before, during and after remedial action Total PCB concentration Total PCB concentration Total PCB concentration Outdoor temperature in indoor air in indoor air in indoor air before, during before remedial during remedial after remedial and after action action action remedial action Apartment a, 530 ng m-3 910 ng m-3 11 °C, 12.5 °C, — uninhabited Apartment b, 610 ng m-3 570 and 660 ng m-3 810 ng m-3 15.5 °C, 7–8 °C, 0–5 °C inhabited Apartment in <50 ng m-3 7–10 °C, —, — building without PCB-containing sealant Long term target 30 ng m-3 level in Germany9 Guideline level 300 ng m-3 in Germany9 Level of 3000 ng m-3 concern in Germany9 Table 4 PCB concentrations in workplace air.The results in the same row were not obtained during the same period, thus giving an indication of the variation in the concentrations during the work Total PCB concentration in Working operation Vacuum cleaner breathing zone Comments Cutting with an oscillating knife Not connected 14 and 49 mg m-3 Not optimised Grinding with a mechanical machine Connected method, beginning Cutting with a mechanical chisel Not connected of the project Cutting with an oscillating knife Connected 5.6 and 9.4 mg m-3 Optimised method Grinding with a mechanical machine Connected 11.0 and 5.5 mg m-3 Optimised method Threshold limit value for 10 mg m-3 workplace environment in Sweden Cutting with an oscillating knife Connected 42 and 30 mg m-3 Vacuum cleaner tubing got loose Sawing with a mechanical saw Connected 34 mg m-3 Vacuum cleaner Cutting with a mechanical chisel Connected tubing got loose Grinding with a mechanical machine Connected 120 mg m-3 a Vacuum cleaner tubing got loose Grinding with a mechanical machine Connected 80 mg m-3 Sampling during Sawing with a mechanical saw Connected all the time of the Cutting with a mechanical chisel connected two measurements above aConcentrations of individual PCB congeners in samples higher than those in the highest standard might give a higher uncertainty in the value compared to other values.The amount of PCB spread to the air during the remedial We estimate the increase in the PCB concentration in the soil as being an average of 0.05 ppm in the first 4 cm, no action should mainly be that located in the air after the vacuum cleaner.The sealants in the entire building took 120 d increase deeper into the soil, and in the first 20 m from the house. The soil samples were taken at 0–2 cm depth, and 1 to cleanup and change.In a normal day, two vacuum cleaners were run 4 h each. A flow of 300 m3 h-1 per machine at a and 10 m from the house. The house was 77 m long×13 m wide. The amount of soil contaminated by 0.05 ppm is equival- concentration of 16 mg m-3 gave a total amount spread to the air of 120×2×4×300×16×10-6=5 g PCB during the entire ent to [(77+20+20)×(13+20+20)-(77×13)]×0.04= 208 m3×1000 kg m-3×0.05×10-6×5.3 (from S7PCB to cleanup process.It is reasonable to assume that the PCB spread to the air during the remedial action mainly goes this total concentration)=55 g PCB. We estimate the amount left in the concrete to be as follows: way as the flow is so high through the vacuum cleaner. This can be compared to the amount that is oV-gassed to the indoor first 2 mm ground away; then, 2 mm×2 (number of edges)×3 cm (width of ground surface)×4000 m× environment during one year: Volume of the house×air exchange rate (not measured, 0.5 per hour is a usual value 2300 kg m-3 (density of concrete)=1104 kg concrete, 2 to 4 mm contains 400 ppm, 4 to 6 mm contains 90 ppm and for mechanical ventilated apartments)×hours in 1 year ×PCB concentration indoors (representative value for deeper no PCB, giving 1104×(400+90)×10-6=0.54 kg PCB.The 2 mm grinding of the concrete is a minimum value, in the measurements 600 ng m-3)=77×13×22×0.5×24×365 ×600×10-9=60 g PCB spread to the indoor air during one most cases more than 2 mm are ground. Therefore 0.54 kg is probably too large an estimate. year. 386 J. Environ. Monit., 1999, 1, 383–387Table 5 The relative contents of the 7 indicator PCB for diVerent samples PCB28/31a PCB52 PCB101 PCB118 PCB138 PCB153 PCB180 Concrete 0.67 1 0.99 0.60 0.12 0.063 0.010 Concrete 0.49 0.85 1 0.60 0.19 0.13 0 Concrete 0.80 1 0.79 0.49 0.11 0.055 0.007 Concrete 1 0.83 0.98 0.63 0.14 0.069 0.008 Concrete 0.44 0.84 1 0.56 0.20 0.11 0.036 Soil 0.47 0.76 1 0.56 0.88 0.81 0.53 Soil 0.52 1 0.90 0.56 0 0 0 Soil 0.64 1 0.68 0.34 0.17 0 0 Soil 0.66 1 0.95 0.51 0.18 0.14 0 Vacuum cleaner 1 0.63 0.19 0 0 0 0 Vacuum cleaner 1 0.49 0.15 0.040 0 0 0 Sealant 0.53 1 0.96 0.56 0.14 0.091 0.027 Indoor air before work 1 0.44 0 0 0 0 0 Indoor air before work 1 0.50 0.042 0 0 0 0 Indoor air during work 1 0.52 0.048 0.012 0 0 0 Indoor air during work 1 0.42 0.033 0 0 0 0 Workplace environment 1 0.94 0.49 0.22 0 0 0 Workplace environment 1 0.88 0.37 0.19 0.037 0.016 0 Workplace environment 1 0.90 0.36 0.14 0.051 0 0 Workplace environment 1 0.90 0.35 0.13 0 0 0 Workplace environment 1 0.82 0.35 0.16 0 0 0 Workplace environment 1 0.88 0.40 0.17 0.054 0.022 0 Workplace environment 1 0.89 0.40 0.17 0.045 0.022 0 aIn this table the sum of PCB28 and 31 are used for the comparison, since there is a poor chromatographic separation between these two congeners.To summarise 540 g+55 g+5 g=0.6 kg PCB is either spread developing the cleanup methods and cooperation during the measuring periods. Bjo�rn Lundgren is gratefully acknowledged or left in the building compared to the 90 kg PCB taken away. Thus more than 99% of the PCB are removed, which must be for his valuable comments. considered a good result.References Conclusions 1 C. Bernes, in Monitor 16, Persient Organic Pollutants, A Swedish View of an International Problem, Swedish Environment Protection It is important to remove PCB-containing elastic sealant in Agency, Stockholm. order to reduce the spread of PCB to indoor air and to the 2 Thiokol is a polysulfide which often has PCB added as a plasticizer. outdoor environment.Some kind of breathing protection 3 T. Hammar, PCB i fogmassor, La�nstyrelsen i Kalmar informerar should be used by workers during the work involved in (The County Government Board in Kalmar, Sweden informs), removing the sealant. There is no significant increase in PCB 1992, 10. 4 B. Jansson, J. Sandberg, N. Johansson and A. A° stebro, PCB i levels indoors during the remedial process. A few millimetres fogmassor—Stort eller litet problem? Swedish Environmental of the concrete should be taken away as the concentration in Protection Agency, Stockholm. the concrete closest to the sealant is high. 5 C. Benthe, B. Heinzow, H. Hessen, S. Mohr and W. Rotard, It is possible to remove PCB-containing elastic sealant with Chemosphere, 1992, 25, 1481. minor spread of PCB to the environment. A rough estimate 6 E. Sikander, M. Sundahl, M. Tornevall, B. Ek-Olausson, suggests that more than 99% of the PCB is removed. A. Hjorthage, L. Rosell and P.-O. Johnnesson, Utveckling och utva�rdering av metoder fo�r PCB-haltiga fogmassor, SP Swedish National Testing and Research Institute, Bora°s, Sweden, 1999, 07. Acknowledgements 7 D. E. Schultz, G. Petrick and J. C. Dulnkner, Environ. Sci. Technol., 1989, 23, 852. ‘Kretsloppsmiljarden’ a Swedish governmental fund for a 8 E. Balfanz, J. Fuchs and H. Kieper, presented at Dioxin ’93, 13th sustainable development, ‘Miljo� teknikdelegationen’ a Swedish International Symposium on Chlorinated Dioxins and Related governmental fund for environmental technology, The Swedish Compounds, Organohalogen Compounds 1993. 9 Internet address: http://www.mtm.de/pcbnrw.htm Council for Building Research, and others are gratefully acknowledged for financial support. Hans Zackrisson, Fogspecialisten AB, is gratefully acknowledged for help in Paper 9/02528F J. Environ. Monit., 1999, 1, 383&ndash