- Email: [email protected]
Aqueous leaching of polycyclic aromatic hydrocarbons from bitumen and asphalt
PII: S0043-1354(01)00216-0
Wat. Res. Vol. 35, No. 17, pp. 4200–4207, 2001 r 2001 Elsevier Science Ltd. All rights reserved Printed in Great Britain 0043-1354/01/$ - see front matter
AQUEOUS LEACHING OF POLYCYCLIC AROMATIC HYDROCARBONS FROM BITUMEN AND ASPHALT H. C. A. BRANDT* and P. C. DE GROOT Shell International Oil Products BV, Shell Research and Technology Centre Amsterdam, P.O. Box 38000, 1030 BN Amsterdam, Netherlands (First received 10 May 2000; accepted in revised form 23 April 2001) AbstractFThe application of bitumen in, e.g. asphalt roads, roofs and hydraulic applications will lead to the leaching of compounds from the bitumen/asphalt into the environment. Because polycyclic aromatic hydrocarbons (PAHs) are present in bitumen, static and dynamic leach tests have been performed to study the leaching behaviour of this class of compounds. Nine petroleum bitumens covering a representative range of commercially available products and one asphalt made from one of the bitumens have been tested in a static leach test. The asphalt has been also subjected to a dynamic leach test. The main conclusions are that a 30 h dynamic leach test is sufficient to determine the equilibrium concentration that will be reached after bitumen or asphalt has been in contact with the water for more than 3–6 days. As an alternative to performing a leach test, this concentration can be calculated from the PAH concentrations in the bitumen, and their distribution coefficients, as calculated here, or from their aqueous solubilities. The equilibrium PAH concentrations in the leach water from bitumens stay well below the surface water limits that exist in several EEC-countries and are also more than an order of magnitude lower than the current EEC limits for potable water. r 2001 Elsevier Science Ltd. All rights reserved Key wordsFbitumen, asphalt, PAH, static leaching, dynamic leaching, partitioning
INTRODUCTION
The main application of bitumen/asphalt is in the coating of surfaces, e.g. of roads, roofs, linings of water basins and pipes. All these applications have a potential for contact with water. Such contact will lead to the leaching of components from the bitumen/asphalt into the environment. Many substances are potentially harmful for man or environment and in several countries limits have been set for their release, e.g. in surface water and/or drinking water. Because PAHs are present in bitumen, static and dynamic leaching tests have been performed to study the leaching behaviour of this class of compounds. Mu¨nch (1992, 1993) analysed the soil next and below to bitumen based asphalt roads and found that the PAH concentrations below the road were less than 30% of the concentrations measured beside the road at the same depth. The PAH concentrations found near bitumen bound forest roads were of similar magnitude as those found in the middle of the forest and below Dutch and German limits. A major
*Author to whom all correspondence should be addressed. Z. Jansestraat 46-2, 1097CN Amsterdam, Netherlands. Tel.: +31-20-665-6909; e-mail: [email protected]
part of the PAH contamination was attributed to the spreading of asphalt particles abraded from the roads. More recently, Sadler et al. (1999) found PAH concentrations in the soil immediately below the asphalt cover of an electric trolly bus depot of the order of 1–10 mg/kg. These concentrations were a factor of ten higher than those found by Mu¨nch. A direct logarithmic relationship could be demonstrated between soil content of a particular PAH and molecular properties such as water solubility, suggesting that leaching from asphalt surfaces is a contributor to roadside PAH contamination. Few papers about aqueous leaching of PAHs from bitumen into water exist. Miller et al. (1982) circulated water for 232 h through a 10 m length, 15 cm i.d. iron pipe coated internally with two different bitumen pipe coatings: 225 L water circulated every 10–12 s. Analysis of the leach water after enrichment over polyurethane plugs was done by HPLC-UV fluorescence. The sum of the six World Health Organisation (WHO) PAHs (fluoranthene, benzo(b)- and benzo(k)-fluoranthene, benzo(a)pyrene, benzo(g,h,i)perylene and indeno(1,2,3-c,d)pyrene) was less than 10 ng/L, the highest concentration for any one individual PAH was less than 5 ng/L. The detection limit was 0.2–0.75 ng/L. Kriech (1997)
4200
Aqueous leaching of polycyclic aromatic hydrocarbons from bitumen and asphalt
reports leach tests of hot mix asphalt. Except for naphthalene (250 ng/L) all PAHs were below the detection limit that ranged from 194 ng/L for the 2ring PAHs to 20 ng/L for the 3–6-ring PAHs. Brantley and Townsend (1999) report leaching from reclaimed asphalt pavement, the concentration of none of the 16 EPA PAHs was found above the detection limit that ranged from 0.25 to 5 mg/L. In the past coal-tar has been used in many applications, where now bitumen is used. Aqueous concentrations in leaching experiments with coal-tars or tar-liquids range from mg/L to mg/L. For example, Picel et al. (1988) report values ranging from 5 mg/L for pyrene up to 1.6 mg/L for naphthalene and Lee et al. (1992a,b) values ranging from 1 mg/L for benzo(a)pyrene up to 14 mg/L for naphthalene. Mu¨nch (1993) reports PAH concentrations in soil at the road side of roads covered with coal-tar based asphalt that are ten times higher than those found at the sides of similar roads covered with bitumen based asphalt. The main objective of this study was to establish the bitumen contribution to PAH emissions into the aqueous environment using laboratory leach tests and if possible, develop an understanding of the underlying principles. In principle two types of leaching methods can be distinguished: 1. Dynamic extraction methods. These are generally methods of wide applicability for solid wastes. Most methods prescribe a sampling technique, a size reduction step, the extraction medium, the solid/liquid ratio, the extraction time and sometimes the apparatus. 2. Static migration tests. These are generally methods which measure the diffusion out of an intact surface into a known volume of water of specified quality. In order to avoid high concentrations in the leachate, in some methods the water is refreshed a number of times. The primary use of bitumen is in intact structures, therefore has been chosen for a static leach test. Because the data had to be accepted by the authorities a normalised method has been used as far as possible. For comparison we have subjected an asphalt made from one of the bitumens to both the static and a dynamic test. Nine bitumen samples have been selected, comprising a representative range of commercially available products. In addition to the bitumens, one asphalt, made from one of the bitumens was tested. In this paper we present data on the leaching of PAHs from those bitumens, showing that after a number of days, equilibrium is reached and we give an empirical relation describing the leach concentrations as function of time. Bitumen water distribution coefficients will be calculated for PAHs.
4201
MATERIALS AND METHODS
Methods The static leach test chosen follows a Dutch standardised test for building materials, Dutch Normalisation Institute (1995) as far as possible. This test consists of serial batch leaching with a fixed liquid/solid ratio of 4.5 : 1, using water acidified to pH 4, the leachate is refreshed after 0.25, 1, 2.25, 4, 9, 16 and 36 days. Because of the liquid-like nature of bitumens the test had to be slightly modified: instead of using a block of material, a layer of bitumen (about 140 g of bitumen) contained in a covered glass dish (diameter 134 mm, height 55 mm) was used. As water for the leaching double-distilled deinonised water, purified over a 150 4.6 mm column filled with XAD-2 resin, and subsequently acidified to pH 4, was used. Great care was given to thorough cleaning of the glassware, preventing contamination with PAHs from the environment and containers, etc. In all tests a blank leaching experiment was run in parallel and the data were corrected for the blanks. For the static leaching test of the asphalt a cylindrical block of asphalt (diameter 100 mm, height 64 mm) was placed on glass rods in a covered glass dish (diameter 170 mm, height 94 mm). The amount of water was chosen to keep the amount of water comparable to that of the tests with the bitumens (4 mL/cm2). The dynamic leach test was performed following a draft European test subsequently issued as 185CEN/TC292 (European Centre for Normalisation, 1999) using the ‘‘Zero Head Space Extractor’’ from Millipore, Bedford USA. The CEN test conditions are: size reduction of the sample to r4 mm, leaching with water acidified to pH 4, with a liquid solids ratio of 10 : 1, agitating for 30 h at 30 rpm. For the PAH analysis the dilute leachate was preconcentrated before the actual analysis: 200 mL leach water was pumped over an inline HPLC-precolumn (C18), followed by backflushing to the analytical column. PAH analysis was performed on a Hewlett-Packard 1090 series M Liquid Chromatograph with a Vydac C18 column, using 35/65 acetonitrile/water for 0.1 min, followed by a linear gradient to 100% acetonitrile over 12 min, and 13 min isocratic with 100% acetonitrile, at a flow rate of 2 mL/min. Quantitative detection was performed with an HP 1046A programmable fluorescence detector. For each PAH an excitation and emission wavelength was used to optimise peak to background ratio in the sample matrix. All results were from duplicate PAH analyses. Calibration was performed twice a day by injecting a standard PAH mixture in tetrahydrofurane on the enrichment column followed by washing the PAHs onto the column with 200 mL bi-distilled water purified inline over a 80 7 mm column filled with XAD-2 resin. If the concentration of a PAH was too high for the determination, a second run with a lower amount of leachate was done. The test procedure is described in the Amsterdam Method Series (AMS, 1065-1) (leaching) and (AMS, 1051-1) (PAH analysis). Because naphthalene already starts to elute with water during the enrichment step, during calibration the same amount of water was used to wash the calibration mixture onto the enrichment column as was used for the analysis of the leachate. For the determination of PAHs in the bitumens a preseparation is necessary before HPLC analysis. After precipitation of the C7 asphaltenes (IP 143/90, 2000), an aromatic fraction, including the PAHs, was extracted from the maltenes with dimethyl sulphoxide (IP 346/80, 1980). Before injection into the liquid chromatograph the PAH fraction was further isolated by chromatography over SepPak silica gel cartridges (Waters Associates, Inc.). The PAH analysis was performed following Amsterdam Method Series (AMS, 1051-1).
4202
H. C. A. Brandt and P. C. de Groot Table 1. Description of samples
Code
Description
A B C D E
Conventional Middle East penetration bitumen Conventional Heavy Venezuelan penetration bitumen Conventional Heavy Venezuelan penetration bitumen Straight run vacuum residue of Middle East origin Deeply vacuum flashed conversion residue from a thermal cracker ex D Conventionally blown roofing bitumen ex D Conventionally blown roofing bitumen of Middle East origin Blown roofing bitumen of middle east origin Multiphalte bitumen
F G H I
Bitumens tested Nine bitumens have been tested, covering a representative range of commercially available products. In addition to the bitumens one asphalt, made from bitumen A, plus the aggregate the asphalt was made of, was tested. A description of the samples is given in Table 1. For the static leach test a ‘‘Marshall’’ block made from a dense asphalt mix with aggregate o4 mm was used. The block weighed 1192 g and contained 94 g bitumen (7.9%). For the dynamic leach test the same mix was used uncompacted. RESULTS
PAHs Generally only the PAHs with 4 rings or less, which have the highest water solubility could be found in concentrations above 0.1 ng/L (Table 2). Naphthalene is found in concentrations ranging from 1 to 400 ng/L and dominates the other PAHs. Therefore, in plots, the concentrations of the P PAHs with more than 2 rings are summed ( 2+ring PAHs) and naphthalene is reported separately. Three ring PAHs range from 0.1 to 180 ng/L, the four ring PAHs from 0.1 to 5 ng/L. The five and heavier ring PAHs have not been found at concentrations above
0.4 ng/L. Plots of the PAH concentrations in the aqueous phase against the corresponding leach period are P given in Fig. 1 for naphthalene and Fig. 2 for 2+ring PAHs. Of the 2+ ring PAHs generally phenanthrene has the highest concentration. Precision data that have been obtained from duplicate leachtests are given in Table 3. For naphthalene, acenaphthene, fluorine, pyrene and sum 2+ ring PAHs, the relative precisions are similar to those determined for PAH analysis in bitumen fumes. For phenanthrenes, anthracenes, fluoranthene, benz(a)anthracene and chrysene the relative precisions for the leach tests are 2.5–5 times higher. Analysis of some leachate samples by GC-MS in a certified laboratory gave similar or slightly higher concentrations for naphthalene to pyrene. For chrysene 3–10 times higher concentrations were found with GC-MS, this may be due to codetermination of tryphenylene. The general pattern for the concentration with leach time is that the concentration in the leachate increases and reaches a steady state concentration between 3 and 6 days. The average plateau concentrations, calculated from the concentrations in the samples taken after 9, 16, 36 and 64 days are given in Table 2. The amounts leached in the whole static test as % of present in the bitumen or asphalt ranges from maximum 1.3% for naphthalene, o0.5% for acenaphthene and fluorene to less than 0.1% for phenanthrene and anthracene. For the higher molecular weight PAHs this is less than 0.01%. Assuming equilibrium for the longer leach periods the equilibrium concentration will depend on the concentration of the PAH in the bitumen (Cbit ) and on the distribution coefficient of the PAH between bitumen and water (KBitWater ): KBitWater ¼ Cbit =Cw
ð1Þ
Table 2. Average plateau concentrations reached in static leach test, calculated from the concentrations in the samples taken after 9, 16, 36 and 64 daysa Bitumen code type of sample/test 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
a b
Naphthalene Acenaphthene Fluorene Phenanthrene Anthracene Fluoranthene Pyrene Benz(a)anthracene Chrysene Benzo(b)fluoranthene Benzo(k)fluoranthene Benzo(a)pyrene Dibenz(a,h)anthracene Benzo(g,h,i)perylene Indeno(1,2,3-c,d)pyrene Coronene P P2+ring PAHs 6 WHO PAHs
A
B
C
35 1.3 2.1 4.1 0.1 0.4 0.4 0.1 0.3 b.d.l. b.d.l. b.d.l. b.d.l. b.d.l. b.d.l. 0.03 8.8 0.43
371 17 42 180 12 1.7 3.9 1.4 5.3 0.4 0.2 0.1 b.d.l. b.d.l. b.d.l. 0.05 263 2.33
51 5 7 47 5 0.8 1.4 0.45 0.83 0.14 0.14 0.3 b.d.l. b.d.l. b.d.l. 0.09 68 1.38
D E F Bitumen static (ng/L) 175 2.4 3.6 2.9 0.5 0.16 0.47 0.04 0.13 0.01 b.d.l. b.d.l. b.d.l. b.d.l. b.d.l. 0.04 10 0.17
30 0.6 0.8 3.3 0.5 0.1 0.5 0.1 b.d.l. b.d.l. b.d.l. b.d.l. b.d.l. b.d.l. b.d.l. 0.01 5.9 0.08
n.v. 2.7 19.4 15.9 6.1 1.7 4.3 0.5 0.5 b.d.l. b.d.l. b.d.l. b.d.l. b.d.l. b.d.l. b.d.l. 51 1.7
n.v.: not valid; b.d.l.: below detection limit. if less than 200 ml of leach water is used the detection limit increases proportionally.
G
H
I
120 0 11 5 0.3 0.1 0.3 0.1 0.1 b.d.l. b.d.l. b.d.l. b.d.l. b.d.l. b.d.l. b.d.l. 17 0.07
0.9 0.3 3.8 1.0 0.1 0.1 0.1 b.d.l. b.d.l. b.d.l. b.d.l. b.d.l. b.d.l. b.d.l. b.d.l. b.d.l. 5.4 0.07
168 11 44 82 28 1 4 b.d.l. b.d.l. b.d.l. b.d.l. b.d.l. b.d.l. b.d.l. b.d.l. b.d.l. 172 1.47
Detection limitb
0.3 0.04 0.02 0.06 0.07 0.05 0.04 0.05 0.06 0.01 0.02 0.02 0.04 0.02 0.1 0.03
Aqueous leaching of polycyclic aromatic hydrocarbons from bitumen and asphalt
4203
Fig. 1. Aqueous leaching of naphthalene from bitumen in static leach test.
Fig. 2. Aqueous leaching of
P
2+PAHs, in static leach test.
Table 3. Precision of the static leach test, calculated from duplicate leach tests Analyte
Naphthalene Acenaphthene Fluorene Phenanthrene Anthracene Fluoranthene Pyrene Benz(a)anthracene Chrysene S2+ring PAHs
Relative standard deviation
Degrees of freedom
Precision for single leaching
Based on concentration range (ng/L)
(%) RSD
P=0.95 % relative
Min
Max
df
14 19 20 58 58 61 37 76 94 24
10 10 10 10 10 10 10 10 10 10
32 43 45 130 129 136 83 169 209 54
14 0.9 0.9 1.4 0.04 0.03 0.3 0.01 0.02 4
466 3.0 21 14 5.2 1.5 4.6 0.4 0.8 50
4204
H. C. A. Brandt and P. C. de Groot Table 4. Leaching of bitumen and asphalt (static and dynamic) compared Bitumen A Concentrations in leachate (ng/L) Static Average plateau concentrations for Bitumen Asphalt
Naphthalene Acenaphthene Fluorene Phenanthrene Anthracene Fluoranthene Pyrene Benz(a)anthracene Chrysene Benzo(b)fluoranthene Benzo(k)fluoranthene Benzo(a)pyrene Dibenz(a,h)anthracene Benzo(g,h,i)perylene Indeno(1,2,3-c,d)pyrene Coronene
35 1.3 2.1 4.1 0.09 0.4 0.4 0.06 0.3 bdl bdl bdl bdl bdl bdl 0.03
Cw ¼ Cbit =KBitWater ð1@exp
Þ
Asphalt
33 0.8 bdl 1.2 0.1 0.08 0.1 bdl 0.09 bdl bdl bdl bdl bdl bdl bdl
The whole curve can be empirically described by @kt
Dynamic
ð2Þ
where t is the leach time and k the dissolution rate constant that determines how fast equilibrium is reached. It was not possible to calculate a reliable kvalue for predictive purposes, because the outcome in the first days seems to depend on the bitumen. Bitumen vs. asphalt, static and dynamic In Table 4 the average plateau concentrations reached in the static tests on bitumen A and the asphalt made from it, and the concentrations measured in the dynamic leach test have been given. No significant difference exists between the outcome of the static test with bitumen or asphalt, also the build up of the concentration with time is very similar as is shown in Fig. 1. A comparison of the static leach tests with the dynamic test shows that for most of the PAHs the concentrations are similar, except for the 5-ring PAHs, where in the dynamic test concentrations in the order of 0.1 ng/L are found, whereas in the static test these are at least a factor of 10 lower. Because these results are based on one leach test, more tests would be necessary to establish whether this difference is significant. Distribution coefficients For three of the bitumens tested PAH analyses in the bitumens were available. From the bitumen PAH concentrations (Table 5), and the measured aqueous concentrations of the PAHs leached from those bitumens, we have calculated the bitumen-water distribution coefficient Kbitwater (equation (1)) for those PAHs which had leach concentrations above the detection limit (Table 5). In Fig. 3 the average
52 1 1.9 2.5 0.3 bdl 0.03 0.2 0.2 0.2 bdl 0.2 0.1 0.2 bdl bdl
distribution coefficients have been plotted (on a log– log scale) against the subcooled liquid solubility (Sscl ), calculated from Pearlman’s crystalline solubility data (Sc ), (Pearlman et al., 1984) and the melting point (MP, 1C) of the PAH (Lane and Loehr, 1995) using log Sscl ¼ log Sc þ 0:01ðMP@25Þ
ð3Þ
The subcooled liquid solubility is used, because in bitumens the PAHs are already present as dissolved molecules, whereas for aqueous dissolution from the crystalline state, one has to take into account the melting energy. In Fig. 3 we have added the data from distribution coefficients calculated from the partitioning of PAHs between water and coal-tars (Lane and Loehr, 1992) and a coal-oil (Picel et al., 1988). If we leave out the single value for benz(a)anthracene (found systematically below the predicted line (see also (Lee et al., 1992a,b)), the line through the combined coal-tar data runs practically parallel to the regression line through our bitumen data: log Kbitwater ¼ 1:65@0:96log Sscl ;
R2 ¼ 0:9469 ð4Þ
In the same graph the points for the distribution of PAHs between diesel fuel and water (24) are given. The regression line through the diesel points has a slope that slightly deviates from @1. In the plot (not shown) of the cumulative emission (ng/m2) in the static leach test against the square root of leach time (days) for bitumen A, already between 3 and 4 days the line starts to deviate from linear, indicating that only in the first days the leaching is diffusion-controlled.
Aqueous leaching of polycyclic aromatic hydrocarbons from bitumen and asphalt
4205
Table 5. log Kbitwater values for 8 PAHs and the average and standard deviation for three bitumens Bitumen code PAH Naphthalene Acenaphthene Fluorene Phenanthrene Anthracene Fluoranthene Pyrene Benz(a)anthracene Chrysene Benzo(b)fluoranthene Benzo(k)fluoranthene Benzo(a)pyrene Dibenz(a,h)anthracene Benzo(g,h,i)perylene Indeno(1,2,3-c,d)pyrene
PAH concentrations in bitumen
Distribution coefficients
A (mg/kg)
E (mg/kg)
G (mg/kg)
A (log K)
E (log K)
G (log K)
Average (log K)
(SD) %rel
2.7 0.2 0.3 1.8 0.2 0.9 0.9 0.7 2.4 1.0 0.4 0.7 0.5 2.0 0.5
3.0 0.7 0.4 2.0 0.2 0.8 1.0 0.2 1.0 0.7 0.3 0.5 0.3 2.0 0.2
2.5 bdl 0.4 1.1 0.1 0.3 0.3 bdl 0.5 0.4 bdl bdl bdl 0.8 bdl
4.8 5.3 5.3 5.9 6.1 6.7 6.9
5.0 6.1 5.7 5.8 5.6 7.1 6.3
4.3 4.6 5.3 5.5 6.7 6.1
4.7 5.7 5.2 5.7 5.7 6.8 6.4
8 10 11 5 5 3 7
7.1
7.3
6.8
7.1
4
Fig. 3. Log distribution coefficient (log Kd) between bitumen (coal-tars and gasoils) and water for 8 (2–4) ring PAHs vs. log subcooled aqueous solubilities of these PAHs.
DISCUSSION
Leach tests have been performed on bitumen and on one asphalt made with one of the bitumens. The determination of PAHs at low levels (down to 0.01– 0.06 ng/L) has allowed us to measure concentrations of a number of 2–6-ring PAHs in the leach water from bitumen. The general pattern observed in the static leach test shows a diffusion-controlled process in the first few days, after which equilibrium is reached between 3 and 6 days. It may well be that it was not possible to calculate a dissolution rate constant for the diffusion-controlled period, because of the presence of natural surfactants in bitumen (Acevedo et al., 1992). The nature and concentration
of those surfactants will vary depending on the crude origin of the bitumen. Because surfactants can influence the phase transfer process this will have its influence on the dissolution rate constant. The equilibrium concentration depends on the PAH concentration in the bitumen and the aqueous subcooled liquid solubility of the PAH. The empirical equation for the time course of the aqueous concentration shows to be very similar to that used in a mass transfer model describing the dissolution of organics from oil films into water (Southworth et al., 1983). The observation that the concentrations in the dynamic test are similar to the plateau values in the static test shows that in the dynamic test also, equilibrium is reached under the test conditions
4206
H. C. A. Brandt and P. C. de Groot
Table 6. Averages for chemical type analysis, number average molecular weights (Mn) and densities for coal tar, bitumens and automotive diesel fuels (AGO)a
Coal-tar Bitumen AGO a b
Saturates
Aromatics
Resins
2 13 70
30 42 30
11 32 0
C7-asphb 56 13 0
Fraction aromatic carbon 0.97 0.30 0.25
Mn 624 839 227
Density
Molar volume
1.2 1.0 0.86
0.53 0.83 0.26
Data from Zander (1995), Puzinauskas and Corbett (1978), CONCAWE (1995) and Lee et al. (1992a,b) and own unpublished work. The fraction that precipitates when the sample is diluted with n-heptane.
(i.e. 30 h agitation, L/S ratio 10). This is in line with the observation of Lane and Loehr (1992) that a 24 h agitation period was sufficient to reach equilibrium in PAH desorption experiments from soil. Distribution coefficients for the distribution between bitumen and water (Kbitwater ) have been calculated for a number of PAHs. Log Kbitwater relates linearly with the log subcooled liquid solubility of the PAH. The distribution coefficients are well in line with those reported for the same PAHs for coal-tars and coal-oils. The slope of the plot of the distribution coefficients for bitumen and water is virtually equal to that for coal-tar products and water and approached the ideal slope of @1 (Chiou et al., 1982). The differences in the intercepts may originate from differences in molar volumes, densities and activity coefficients. The activity coefficients differ because the chemical composition is different (Table 6). Coal-tars consist mainly of aromatic molecules with short- or no- side chains (aromatic carbon 97% of total carbon) (Zander, 1995), whereas bitumens do not only contain saturated compounds, but the aromatic molecules are also heavily alkylated (aromatic carbon 30% of total carbon leading to quite different affinity for PAHs. Although the PAH distribution coefficients for coal-tars only differ by a factor of about 4 (Dlg Kd ¼ 0:65) from those for bitumens, the aqueous leach concentrations from coal-tars are much higher (mg/L–mg/L) than those from bitumens (ng/L), because the PAH concentrations in coal-tars are much higher (order g/kg) than those in bitumens (order mg/kg) (Brandt et al., 1985). The different slope of the regression line through the diesel fuel data remains unexplained, it may be due to the imprecision of the data, for example the fluoranthene Kd has a great influence on the slope of the diesel fuel line. The PAH concentrations in the leach water from bitumens (plateau values, Table 2) stay well below the surface water limits that exist in several EECcountries and also are more than an order of magnitude lower than the current EEC limits for potable water (EEC, 1998). For example the current Dutch limits for groundwater, above which action is required, range from 70 mg/L for naphthalene to 5 for phenanthrene (Dutch State Journal, 1994), which is more than a factor of 20 higher than the equivalent PAH concentrations in the leach water from bitumens. A comparison with the current limits for
drinking water shows that the range found for the sum of the 4 EEC PAHs (benzo(b)fluoranthene, benzo(k)fluoranthene, benzo(ghi)perylene and indeno(1,2,3-cd)pyrene) amounts from E0.01 to 0.64 ng/ L compared to a maximum tolerable concentration for potable water of 0.1 mg/L. The benzo(a)pyrene concentrations in the leach water range from o0.015 to 0.3 ng/L compared to the limit of 10 ng/ L that is regulated separately from the other PAHs (EEC, 1998). The range of bitumens we have tested is representative of current commercial practice. All bitumens showed similar leaching behaviour against time. This indicates that our findings will be valid for other petroleum bitumens as well, eliminating the need for using the very time- and manpower- consuming static leach test. If one wants to determine the equilibrium concentration one can use a less time consuming dynamic test. As an alternative for performing a leach test one can calculate the aqueous concentrations from the PAH content of the bitumen and its distribution coefficient or its aqueous solubility. For example, the aqueous PAH concentrations predicted from the PAH concentrations in bitumens (pooled data from Table 4 and Brandt et al. (1985)) range from 0.08 to 0.6 ng/L for the sum of the six WHO PAHs. This compares well with the experimental range of E0.1–2.4 ng/L found in this study. Because no difference was found between leaching from bitumen and asphalt one can use these calculated values also for predicting the bitumen contribution to the leaching from asphalt.
CONCLUSIONS
In the static leach test a range of petroleum bitumens, representative for current commercial practice all show the same leaching behaviour against time. In the first days the concentration in the aqueous leachate increases and reaches a steady state concentration between 3 and 6 days. Distribution coefficients for the distribution between bitumen and water (Kbitwater) have been calculated for a number of PAHs. Log Kbitwater relates linearly with the log subcooled liquid solubility of the PAH. Bitumen and asphalt, the latter both in the static and a dynamic leach test give similar results. For
Aqueous leaching of polycyclic aromatic hydrocarbons from bitumen and asphalt
obtaining the equilibrium aqueous concentration, instead of the static leach test the less time consuming dynamic leach test can be used. The equilibrium PAH concentrations in the leach water from bitumens stay well below the surface water limits that exist in several EEC-countries and also are more than an order of magnitude lower than the current EEC limits for potable water. Instead of performing a leach test the aqueous concentration of a PAH can also be calculated from its concentration in the bitumen and the Kbitwater or the aqueous solubility of the PAH.
REFERENCES
Acevedo S., Escobar G., Gutie´rrez L. and Rivas H. (1992) Isolation and characterization of natural surfactants from extra heavy crude oils, asphaltenes and maltenes. Interpretation of their interfacial tension-pH behaviour in terms of ion pair formation. Fuel 71, 619–623. Amsterdam Method Series. The leaching behaviour of PAC from bitumen, static leaching test, AMS 1065-1. (Available from Shell Research and Technology Centre Amsterdam, SIOP OGMB, P.O. Box 38000, 1030 BN, Amsterdam.) Amsterdam Method Series. Determination of polycyclic aromatic compounds in condensed bitumen fumes, AMS 1057-1. (Available from Shell Research and Technology Centre Amsterdam, SIOP OGMB, PO Box 38000, 1030 BN Amsterdam.) Brandt H. C. A., de Groot P. C., Molyneux M. K. B. and Tindle P. E. (1985) Sampling and analysis of bitumen fumes. Ann. Occup. Hyg. 29, 27–80. Brantley A. S. and Townsend T. (1999) Leaching of pollutants from reclaimed asphalt pavement. Environ. Eng. Sci. 16(2), 105–116. Chiou C. T., Schmedding D. W. and Manes M. (1982) Partitioning of organic compounds in octanol-water systems. Environ. Sci. Technol. 16(1), 4–10. CONCAWE (1995) Gas oils (diesel fuels/heating oils). CONCAWE Brussels, product dossier no. 95/107. Dutch State Journal (1994) Staatscourant 95 d.d. 24 mei1994: circulaire interventiewaarden Bodemsanereing (9 mei 1994/nr. DBO/07494013). Dutch Normalisation Institute (1995) Leaching characteristics of shaped building materials. NEN 7345, Dutch Norm (in Dutch). EEC (1998) COUNCIL DIRECTIVE 98/83/EC on the quality of water intended for human consumption. Official Journal of the European Communities L 330/32, Brussels, 3 November.
4207
European Centre for Normalisation (1999) Compliance test for leaching of granular waste materials, including pastelike materials and sludges. 185CEN/TC292. IP 346/80 (1980) Polycyclic aromatics in petroleum fractions by dimethyl sulphoxide-refractive index method. Institute of Petroleum, London. IP 143/90 (2000) British Standards 2000. Determination of asphaltenes (heptane insolubles), Part 143. Kriech A. J. (1997) Evaluation of hot mix asphalt for leachability. Asphalt Institute.http://www.asphaltinstitute. org/environ/leachma.html. Lane W. F. and Loehr R. C. (1992) Estimating the equilibrium aqueous concentrations of polynuclear aromatic hydrocarbons in complex mixtures. Environ. Sci. Technol. 26(5), 983–990. Lane W. F. and Loehr R. C. (1995) Predicting aqueous concentrations of polynuclear aromatic hydrocarbons in complex mixtures. Water Environ. Res. 67(2), 169–173. Lee L. S., Hagwall M., Delfino J. J. and Rao S. C. (1992a) Equilibrium partitioning of polycyclic aromatic hydrocarbons from coal-tar into water. Environ. Sci. Technol. 26(11), 2104–2110. Lee L. S., Hagwall M., Delfino J. J. and Rao S. C. (1992b) Partitioning of polycyclic aromatic hydrocarbons from diesel fuel into water. Environ. Sci. Technol. 26(11), 2110– 2115. Miller H. C., Barret W. J. and James R. H. (1982) Investigating potential water contamination by petroleum-asphalt coatings in ductile-iron pipe. J. AWWA 74, 151–156. Mu¨nch D. (1992) Soil contamination beneath asphalt roads by polynuclear aromatic hydrocarbons, zinc, lead and cadmium. Sci. Total Environ. 126, 49–60. Mu¨nch D. (1993) Vergleich von asphaltierten und teerhaltig befestigten waldwegen bezuglich Ausbreitung von polycyclischen aromatischen Kohlenwasser-stoffen (PAH) und Schwermetallen. Bitumen 4, 152–158. Pearlman R. S., Yalkowsky S. H. and Bannerjee S. (1984) Water solubilities of polynuclear aromatic and heteroaromatic compounds. J. Phys. Chem. Ref. Data 13(2), 555–562. Picel K. C., Stamounidis V. C. and Simmons M. S. (1988) Distribution coefficients for chemical components of a coal–oil/water system. Water Res. 22(9), 1189–1199. Puzinauskas, V. P. and Corbett, L. W. (1978) Differences between petroleum asphalt, coal-tar pitch and road tar. The Asphalt Institute, Research Report, p. 78–1. Sadler R., Delamont C., White P. and Connel D. (1999) Contaminants in soil as a result of leaching from Asphalt. Toxicol Environ. Chem. 68, 71–81. Southworth G. R., Herbes S. E. and Allen C. P. (1983) Evaluating a mass transfer model for the dissolution of organics from oil films into water. Water Res. 17(11), 1647–1651. Zander M. (1995) Aspects of coal-tar chemistry/a review. Polycyclic Aromatic Compounds 7, 209–221.
Wat. Res. Vol. 35, No. 17, pp. 4200–4207, 2001 r 2001 Elsevier Science Ltd. All rights reserved Printed in Great Britain 0043-1354/01/$ - see front matter
AQUEOUS LEACHING OF POLYCYCLIC AROMATIC HYDROCARBONS FROM BITUMEN AND ASPHALT H. C. A. BRANDT* and P. C. DE GROOT Shell International Oil Products BV, Shell Research and Technology Centre Amsterdam, P.O. Box 38000, 1030 BN Amsterdam, Netherlands (First received 10 May 2000; accepted in revised form 23 April 2001) AbstractFThe application of bitumen in, e.g. asphalt roads, roofs and hydraulic applications will lead to the leaching of compounds from the bitumen/asphalt into the environment. Because polycyclic aromatic hydrocarbons (PAHs) are present in bitumen, static and dynamic leach tests have been performed to study the leaching behaviour of this class of compounds. Nine petroleum bitumens covering a representative range of commercially available products and one asphalt made from one of the bitumens have been tested in a static leach test. The asphalt has been also subjected to a dynamic leach test. The main conclusions are that a 30 h dynamic leach test is sufficient to determine the equilibrium concentration that will be reached after bitumen or asphalt has been in contact with the water for more than 3–6 days. As an alternative to performing a leach test, this concentration can be calculated from the PAH concentrations in the bitumen, and their distribution coefficients, as calculated here, or from their aqueous solubilities. The equilibrium PAH concentrations in the leach water from bitumens stay well below the surface water limits that exist in several EEC-countries and are also more than an order of magnitude lower than the current EEC limits for potable water. r 2001 Elsevier Science Ltd. All rights reserved Key wordsFbitumen, asphalt, PAH, static leaching, dynamic leaching, partitioning
INTRODUCTION
The main application of bitumen/asphalt is in the coating of surfaces, e.g. of roads, roofs, linings of water basins and pipes. All these applications have a potential for contact with water. Such contact will lead to the leaching of components from the bitumen/asphalt into the environment. Many substances are potentially harmful for man or environment and in several countries limits have been set for their release, e.g. in surface water and/or drinking water. Because PAHs are present in bitumen, static and dynamic leaching tests have been performed to study the leaching behaviour of this class of compounds. Mu¨nch (1992, 1993) analysed the soil next and below to bitumen based asphalt roads and found that the PAH concentrations below the road were less than 30% of the concentrations measured beside the road at the same depth. The PAH concentrations found near bitumen bound forest roads were of similar magnitude as those found in the middle of the forest and below Dutch and German limits. A major
*Author to whom all correspondence should be addressed. Z. Jansestraat 46-2, 1097CN Amsterdam, Netherlands. Tel.: +31-20-665-6909; e-mail: [email protected]
part of the PAH contamination was attributed to the spreading of asphalt particles abraded from the roads. More recently, Sadler et al. (1999) found PAH concentrations in the soil immediately below the asphalt cover of an electric trolly bus depot of the order of 1–10 mg/kg. These concentrations were a factor of ten higher than those found by Mu¨nch. A direct logarithmic relationship could be demonstrated between soil content of a particular PAH and molecular properties such as water solubility, suggesting that leaching from asphalt surfaces is a contributor to roadside PAH contamination. Few papers about aqueous leaching of PAHs from bitumen into water exist. Miller et al. (1982) circulated water for 232 h through a 10 m length, 15 cm i.d. iron pipe coated internally with two different bitumen pipe coatings: 225 L water circulated every 10–12 s. Analysis of the leach water after enrichment over polyurethane plugs was done by HPLC-UV fluorescence. The sum of the six World Health Organisation (WHO) PAHs (fluoranthene, benzo(b)- and benzo(k)-fluoranthene, benzo(a)pyrene, benzo(g,h,i)perylene and indeno(1,2,3-c,d)pyrene) was less than 10 ng/L, the highest concentration for any one individual PAH was less than 5 ng/L. The detection limit was 0.2–0.75 ng/L. Kriech (1997)
4200
Aqueous leaching of polycyclic aromatic hydrocarbons from bitumen and asphalt
reports leach tests of hot mix asphalt. Except for naphthalene (250 ng/L) all PAHs were below the detection limit that ranged from 194 ng/L for the 2ring PAHs to 20 ng/L for the 3–6-ring PAHs. Brantley and Townsend (1999) report leaching from reclaimed asphalt pavement, the concentration of none of the 16 EPA PAHs was found above the detection limit that ranged from 0.25 to 5 mg/L. In the past coal-tar has been used in many applications, where now bitumen is used. Aqueous concentrations in leaching experiments with coal-tars or tar-liquids range from mg/L to mg/L. For example, Picel et al. (1988) report values ranging from 5 mg/L for pyrene up to 1.6 mg/L for naphthalene and Lee et al. (1992a,b) values ranging from 1 mg/L for benzo(a)pyrene up to 14 mg/L for naphthalene. Mu¨nch (1993) reports PAH concentrations in soil at the road side of roads covered with coal-tar based asphalt that are ten times higher than those found at the sides of similar roads covered with bitumen based asphalt. The main objective of this study was to establish the bitumen contribution to PAH emissions into the aqueous environment using laboratory leach tests and if possible, develop an understanding of the underlying principles. In principle two types of leaching methods can be distinguished: 1. Dynamic extraction methods. These are generally methods of wide applicability for solid wastes. Most methods prescribe a sampling technique, a size reduction step, the extraction medium, the solid/liquid ratio, the extraction time and sometimes the apparatus. 2. Static migration tests. These are generally methods which measure the diffusion out of an intact surface into a known volume of water of specified quality. In order to avoid high concentrations in the leachate, in some methods the water is refreshed a number of times. The primary use of bitumen is in intact structures, therefore has been chosen for a static leach test. Because the data had to be accepted by the authorities a normalised method has been used as far as possible. For comparison we have subjected an asphalt made from one of the bitumens to both the static and a dynamic test. Nine bitumen samples have been selected, comprising a representative range of commercially available products. In addition to the bitumens, one asphalt, made from one of the bitumens was tested. In this paper we present data on the leaching of PAHs from those bitumens, showing that after a number of days, equilibrium is reached and we give an empirical relation describing the leach concentrations as function of time. Bitumen water distribution coefficients will be calculated for PAHs.
4201
MATERIALS AND METHODS
Methods The static leach test chosen follows a Dutch standardised test for building materials, Dutch Normalisation Institute (1995) as far as possible. This test consists of serial batch leaching with a fixed liquid/solid ratio of 4.5 : 1, using water acidified to pH 4, the leachate is refreshed after 0.25, 1, 2.25, 4, 9, 16 and 36 days. Because of the liquid-like nature of bitumens the test had to be slightly modified: instead of using a block of material, a layer of bitumen (about 140 g of bitumen) contained in a covered glass dish (diameter 134 mm, height 55 mm) was used. As water for the leaching double-distilled deinonised water, purified over a 150 4.6 mm column filled with XAD-2 resin, and subsequently acidified to pH 4, was used. Great care was given to thorough cleaning of the glassware, preventing contamination with PAHs from the environment and containers, etc. In all tests a blank leaching experiment was run in parallel and the data were corrected for the blanks. For the static leaching test of the asphalt a cylindrical block of asphalt (diameter 100 mm, height 64 mm) was placed on glass rods in a covered glass dish (diameter 170 mm, height 94 mm). The amount of water was chosen to keep the amount of water comparable to that of the tests with the bitumens (4 mL/cm2). The dynamic leach test was performed following a draft European test subsequently issued as 185CEN/TC292 (European Centre for Normalisation, 1999) using the ‘‘Zero Head Space Extractor’’ from Millipore, Bedford USA. The CEN test conditions are: size reduction of the sample to r4 mm, leaching with water acidified to pH 4, with a liquid solids ratio of 10 : 1, agitating for 30 h at 30 rpm. For the PAH analysis the dilute leachate was preconcentrated before the actual analysis: 200 mL leach water was pumped over an inline HPLC-precolumn (C18), followed by backflushing to the analytical column. PAH analysis was performed on a Hewlett-Packard 1090 series M Liquid Chromatograph with a Vydac C18 column, using 35/65 acetonitrile/water for 0.1 min, followed by a linear gradient to 100% acetonitrile over 12 min, and 13 min isocratic with 100% acetonitrile, at a flow rate of 2 mL/min. Quantitative detection was performed with an HP 1046A programmable fluorescence detector. For each PAH an excitation and emission wavelength was used to optimise peak to background ratio in the sample matrix. All results were from duplicate PAH analyses. Calibration was performed twice a day by injecting a standard PAH mixture in tetrahydrofurane on the enrichment column followed by washing the PAHs onto the column with 200 mL bi-distilled water purified inline over a 80 7 mm column filled with XAD-2 resin. If the concentration of a PAH was too high for the determination, a second run with a lower amount of leachate was done. The test procedure is described in the Amsterdam Method Series (AMS, 1065-1) (leaching) and (AMS, 1051-1) (PAH analysis). Because naphthalene already starts to elute with water during the enrichment step, during calibration the same amount of water was used to wash the calibration mixture onto the enrichment column as was used for the analysis of the leachate. For the determination of PAHs in the bitumens a preseparation is necessary before HPLC analysis. After precipitation of the C7 asphaltenes (IP 143/90, 2000), an aromatic fraction, including the PAHs, was extracted from the maltenes with dimethyl sulphoxide (IP 346/80, 1980). Before injection into the liquid chromatograph the PAH fraction was further isolated by chromatography over SepPak silica gel cartridges (Waters Associates, Inc.). The PAH analysis was performed following Amsterdam Method Series (AMS, 1051-1).
4202
H. C. A. Brandt and P. C. de Groot Table 1. Description of samples
Code
Description
A B C D E
Conventional Middle East penetration bitumen Conventional Heavy Venezuelan penetration bitumen Conventional Heavy Venezuelan penetration bitumen Straight run vacuum residue of Middle East origin Deeply vacuum flashed conversion residue from a thermal cracker ex D Conventionally blown roofing bitumen ex D Conventionally blown roofing bitumen of Middle East origin Blown roofing bitumen of middle east origin Multiphalte bitumen
F G H I
Bitumens tested Nine bitumens have been tested, covering a representative range of commercially available products. In addition to the bitumens one asphalt, made from bitumen A, plus the aggregate the asphalt was made of, was tested. A description of the samples is given in Table 1. For the static leach test a ‘‘Marshall’’ block made from a dense asphalt mix with aggregate o4 mm was used. The block weighed 1192 g and contained 94 g bitumen (7.9%). For the dynamic leach test the same mix was used uncompacted. RESULTS
PAHs Generally only the PAHs with 4 rings or less, which have the highest water solubility could be found in concentrations above 0.1 ng/L (Table 2). Naphthalene is found in concentrations ranging from 1 to 400 ng/L and dominates the other PAHs. Therefore, in plots, the concentrations of the P PAHs with more than 2 rings are summed ( 2+ring PAHs) and naphthalene is reported separately. Three ring PAHs range from 0.1 to 180 ng/L, the four ring PAHs from 0.1 to 5 ng/L. The five and heavier ring PAHs have not been found at concentrations above
0.4 ng/L. Plots of the PAH concentrations in the aqueous phase against the corresponding leach period are P given in Fig. 1 for naphthalene and Fig. 2 for 2+ring PAHs. Of the 2+ ring PAHs generally phenanthrene has the highest concentration. Precision data that have been obtained from duplicate leachtests are given in Table 3. For naphthalene, acenaphthene, fluorine, pyrene and sum 2+ ring PAHs, the relative precisions are similar to those determined for PAH analysis in bitumen fumes. For phenanthrenes, anthracenes, fluoranthene, benz(a)anthracene and chrysene the relative precisions for the leach tests are 2.5–5 times higher. Analysis of some leachate samples by GC-MS in a certified laboratory gave similar or slightly higher concentrations for naphthalene to pyrene. For chrysene 3–10 times higher concentrations were found with GC-MS, this may be due to codetermination of tryphenylene. The general pattern for the concentration with leach time is that the concentration in the leachate increases and reaches a steady state concentration between 3 and 6 days. The average plateau concentrations, calculated from the concentrations in the samples taken after 9, 16, 36 and 64 days are given in Table 2. The amounts leached in the whole static test as % of present in the bitumen or asphalt ranges from maximum 1.3% for naphthalene, o0.5% for acenaphthene and fluorene to less than 0.1% for phenanthrene and anthracene. For the higher molecular weight PAHs this is less than 0.01%. Assuming equilibrium for the longer leach periods the equilibrium concentration will depend on the concentration of the PAH in the bitumen (Cbit ) and on the distribution coefficient of the PAH between bitumen and water (KBitWater ): KBitWater ¼ Cbit =Cw
ð1Þ
Table 2. Average plateau concentrations reached in static leach test, calculated from the concentrations in the samples taken after 9, 16, 36 and 64 daysa Bitumen code type of sample/test 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
a b
Naphthalene Acenaphthene Fluorene Phenanthrene Anthracene Fluoranthene Pyrene Benz(a)anthracene Chrysene Benzo(b)fluoranthene Benzo(k)fluoranthene Benzo(a)pyrene Dibenz(a,h)anthracene Benzo(g,h,i)perylene Indeno(1,2,3-c,d)pyrene Coronene P P2+ring PAHs 6 WHO PAHs
A
B
C
35 1.3 2.1 4.1 0.1 0.4 0.4 0.1 0.3 b.d.l. b.d.l. b.d.l. b.d.l. b.d.l. b.d.l. 0.03 8.8 0.43
371 17 42 180 12 1.7 3.9 1.4 5.3 0.4 0.2 0.1 b.d.l. b.d.l. b.d.l. 0.05 263 2.33
51 5 7 47 5 0.8 1.4 0.45 0.83 0.14 0.14 0.3 b.d.l. b.d.l. b.d.l. 0.09 68 1.38
D E F Bitumen static (ng/L) 175 2.4 3.6 2.9 0.5 0.16 0.47 0.04 0.13 0.01 b.d.l. b.d.l. b.d.l. b.d.l. b.d.l. 0.04 10 0.17
30 0.6 0.8 3.3 0.5 0.1 0.5 0.1 b.d.l. b.d.l. b.d.l. b.d.l. b.d.l. b.d.l. b.d.l. 0.01 5.9 0.08
n.v. 2.7 19.4 15.9 6.1 1.7 4.3 0.5 0.5 b.d.l. b.d.l. b.d.l. b.d.l. b.d.l. b.d.l. b.d.l. 51 1.7
n.v.: not valid; b.d.l.: below detection limit. if less than 200 ml of leach water is used the detection limit increases proportionally.
G
H
I
120 0 11 5 0.3 0.1 0.3 0.1 0.1 b.d.l. b.d.l. b.d.l. b.d.l. b.d.l. b.d.l. b.d.l. 17 0.07
0.9 0.3 3.8 1.0 0.1 0.1 0.1 b.d.l. b.d.l. b.d.l. b.d.l. b.d.l. b.d.l. b.d.l. b.d.l. b.d.l. 5.4 0.07
168 11 44 82 28 1 4 b.d.l. b.d.l. b.d.l. b.d.l. b.d.l. b.d.l. b.d.l. b.d.l. b.d.l. 172 1.47
Detection limitb
0.3 0.04 0.02 0.06 0.07 0.05 0.04 0.05 0.06 0.01 0.02 0.02 0.04 0.02 0.1 0.03
Aqueous leaching of polycyclic aromatic hydrocarbons from bitumen and asphalt
4203
Fig. 1. Aqueous leaching of naphthalene from bitumen in static leach test.
Fig. 2. Aqueous leaching of
P
2+PAHs, in static leach test.
Table 3. Precision of the static leach test, calculated from duplicate leach tests Analyte
Naphthalene Acenaphthene Fluorene Phenanthrene Anthracene Fluoranthene Pyrene Benz(a)anthracene Chrysene S2+ring PAHs
Relative standard deviation
Degrees of freedom
Precision for single leaching
Based on concentration range (ng/L)
(%) RSD
P=0.95 % relative
Min
Max
df
14 19 20 58 58 61 37 76 94 24
10 10 10 10 10 10 10 10 10 10
32 43 45 130 129 136 83 169 209 54
14 0.9 0.9 1.4 0.04 0.03 0.3 0.01 0.02 4
466 3.0 21 14 5.2 1.5 4.6 0.4 0.8 50
4204
H. C. A. Brandt and P. C. de Groot Table 4. Leaching of bitumen and asphalt (static and dynamic) compared Bitumen A Concentrations in leachate (ng/L) Static Average plateau concentrations for Bitumen Asphalt
Naphthalene Acenaphthene Fluorene Phenanthrene Anthracene Fluoranthene Pyrene Benz(a)anthracene Chrysene Benzo(b)fluoranthene Benzo(k)fluoranthene Benzo(a)pyrene Dibenz(a,h)anthracene Benzo(g,h,i)perylene Indeno(1,2,3-c,d)pyrene Coronene
35 1.3 2.1 4.1 0.09 0.4 0.4 0.06 0.3 bdl bdl bdl bdl bdl bdl 0.03
Cw ¼ Cbit =KBitWater ð1@exp
Þ
Asphalt
33 0.8 bdl 1.2 0.1 0.08 0.1 bdl 0.09 bdl bdl bdl bdl bdl bdl bdl
The whole curve can be empirically described by @kt
Dynamic
ð2Þ
where t is the leach time and k the dissolution rate constant that determines how fast equilibrium is reached. It was not possible to calculate a reliable kvalue for predictive purposes, because the outcome in the first days seems to depend on the bitumen. Bitumen vs. asphalt, static and dynamic In Table 4 the average plateau concentrations reached in the static tests on bitumen A and the asphalt made from it, and the concentrations measured in the dynamic leach test have been given. No significant difference exists between the outcome of the static test with bitumen or asphalt, also the build up of the concentration with time is very similar as is shown in Fig. 1. A comparison of the static leach tests with the dynamic test shows that for most of the PAHs the concentrations are similar, except for the 5-ring PAHs, where in the dynamic test concentrations in the order of 0.1 ng/L are found, whereas in the static test these are at least a factor of 10 lower. Because these results are based on one leach test, more tests would be necessary to establish whether this difference is significant. Distribution coefficients For three of the bitumens tested PAH analyses in the bitumens were available. From the bitumen PAH concentrations (Table 5), and the measured aqueous concentrations of the PAHs leached from those bitumens, we have calculated the bitumen-water distribution coefficient Kbitwater (equation (1)) for those PAHs which had leach concentrations above the detection limit (Table 5). In Fig. 3 the average
52 1 1.9 2.5 0.3 bdl 0.03 0.2 0.2 0.2 bdl 0.2 0.1 0.2 bdl bdl
distribution coefficients have been plotted (on a log– log scale) against the subcooled liquid solubility (Sscl ), calculated from Pearlman’s crystalline solubility data (Sc ), (Pearlman et al., 1984) and the melting point (MP, 1C) of the PAH (Lane and Loehr, 1995) using log Sscl ¼ log Sc þ 0:01ðMP@25Þ
ð3Þ
The subcooled liquid solubility is used, because in bitumens the PAHs are already present as dissolved molecules, whereas for aqueous dissolution from the crystalline state, one has to take into account the melting energy. In Fig. 3 we have added the data from distribution coefficients calculated from the partitioning of PAHs between water and coal-tars (Lane and Loehr, 1992) and a coal-oil (Picel et al., 1988). If we leave out the single value for benz(a)anthracene (found systematically below the predicted line (see also (Lee et al., 1992a,b)), the line through the combined coal-tar data runs practically parallel to the regression line through our bitumen data: log Kbitwater ¼ 1:65@0:96log Sscl ;
R2 ¼ 0:9469 ð4Þ
In the same graph the points for the distribution of PAHs between diesel fuel and water (24) are given. The regression line through the diesel points has a slope that slightly deviates from @1. In the plot (not shown) of the cumulative emission (ng/m2) in the static leach test against the square root of leach time (days) for bitumen A, already between 3 and 4 days the line starts to deviate from linear, indicating that only in the first days the leaching is diffusion-controlled.
Aqueous leaching of polycyclic aromatic hydrocarbons from bitumen and asphalt
4205
Table 5. log Kbitwater values for 8 PAHs and the average and standard deviation for three bitumens Bitumen code PAH Naphthalene Acenaphthene Fluorene Phenanthrene Anthracene Fluoranthene Pyrene Benz(a)anthracene Chrysene Benzo(b)fluoranthene Benzo(k)fluoranthene Benzo(a)pyrene Dibenz(a,h)anthracene Benzo(g,h,i)perylene Indeno(1,2,3-c,d)pyrene
PAH concentrations in bitumen
Distribution coefficients
A (mg/kg)
E (mg/kg)
G (mg/kg)
A (log K)
E (log K)
G (log K)
Average (log K)
(SD) %rel
2.7 0.2 0.3 1.8 0.2 0.9 0.9 0.7 2.4 1.0 0.4 0.7 0.5 2.0 0.5
3.0 0.7 0.4 2.0 0.2 0.8 1.0 0.2 1.0 0.7 0.3 0.5 0.3 2.0 0.2
2.5 bdl 0.4 1.1 0.1 0.3 0.3 bdl 0.5 0.4 bdl bdl bdl 0.8 bdl
4.8 5.3 5.3 5.9 6.1 6.7 6.9
5.0 6.1 5.7 5.8 5.6 7.1 6.3
4.3 4.6 5.3 5.5 6.7 6.1
4.7 5.7 5.2 5.7 5.7 6.8 6.4
8 10 11 5 5 3 7
7.1
7.3
6.8
7.1
4
Fig. 3. Log distribution coefficient (log Kd) between bitumen (coal-tars and gasoils) and water for 8 (2–4) ring PAHs vs. log subcooled aqueous solubilities of these PAHs.
DISCUSSION
Leach tests have been performed on bitumen and on one asphalt made with one of the bitumens. The determination of PAHs at low levels (down to 0.01– 0.06 ng/L) has allowed us to measure concentrations of a number of 2–6-ring PAHs in the leach water from bitumen. The general pattern observed in the static leach test shows a diffusion-controlled process in the first few days, after which equilibrium is reached between 3 and 6 days. It may well be that it was not possible to calculate a dissolution rate constant for the diffusion-controlled period, because of the presence of natural surfactants in bitumen (Acevedo et al., 1992). The nature and concentration
of those surfactants will vary depending on the crude origin of the bitumen. Because surfactants can influence the phase transfer process this will have its influence on the dissolution rate constant. The equilibrium concentration depends on the PAH concentration in the bitumen and the aqueous subcooled liquid solubility of the PAH. The empirical equation for the time course of the aqueous concentration shows to be very similar to that used in a mass transfer model describing the dissolution of organics from oil films into water (Southworth et al., 1983). The observation that the concentrations in the dynamic test are similar to the plateau values in the static test shows that in the dynamic test also, equilibrium is reached under the test conditions
4206
H. C. A. Brandt and P. C. de Groot
Table 6. Averages for chemical type analysis, number average molecular weights (Mn) and densities for coal tar, bitumens and automotive diesel fuels (AGO)a
Coal-tar Bitumen AGO a b
Saturates
Aromatics
Resins
2 13 70
30 42 30
11 32 0
C7-asphb 56 13 0
Fraction aromatic carbon 0.97 0.30 0.25
Mn 624 839 227
Density
Molar volume
1.2 1.0 0.86
0.53 0.83 0.26
Data from Zander (1995), Puzinauskas and Corbett (1978), CONCAWE (1995) and Lee et al. (1992a,b) and own unpublished work. The fraction that precipitates when the sample is diluted with n-heptane.
(i.e. 30 h agitation, L/S ratio 10). This is in line with the observation of Lane and Loehr (1992) that a 24 h agitation period was sufficient to reach equilibrium in PAH desorption experiments from soil. Distribution coefficients for the distribution between bitumen and water (Kbitwater ) have been calculated for a number of PAHs. Log Kbitwater relates linearly with the log subcooled liquid solubility of the PAH. The distribution coefficients are well in line with those reported for the same PAHs for coal-tars and coal-oils. The slope of the plot of the distribution coefficients for bitumen and water is virtually equal to that for coal-tar products and water and approached the ideal slope of @1 (Chiou et al., 1982). The differences in the intercepts may originate from differences in molar volumes, densities and activity coefficients. The activity coefficients differ because the chemical composition is different (Table 6). Coal-tars consist mainly of aromatic molecules with short- or no- side chains (aromatic carbon 97% of total carbon) (Zander, 1995), whereas bitumens do not only contain saturated compounds, but the aromatic molecules are also heavily alkylated (aromatic carbon 30% of total carbon leading to quite different affinity for PAHs. Although the PAH distribution coefficients for coal-tars only differ by a factor of about 4 (Dlg Kd ¼ 0:65) from those for bitumens, the aqueous leach concentrations from coal-tars are much higher (mg/L–mg/L) than those from bitumens (ng/L), because the PAH concentrations in coal-tars are much higher (order g/kg) than those in bitumens (order mg/kg) (Brandt et al., 1985). The different slope of the regression line through the diesel fuel data remains unexplained, it may be due to the imprecision of the data, for example the fluoranthene Kd has a great influence on the slope of the diesel fuel line. The PAH concentrations in the leach water from bitumens (plateau values, Table 2) stay well below the surface water limits that exist in several EECcountries and also are more than an order of magnitude lower than the current EEC limits for potable water (EEC, 1998). For example the current Dutch limits for groundwater, above which action is required, range from 70 mg/L for naphthalene to 5 for phenanthrene (Dutch State Journal, 1994), which is more than a factor of 20 higher than the equivalent PAH concentrations in the leach water from bitumens. A comparison with the current limits for
drinking water shows that the range found for the sum of the 4 EEC PAHs (benzo(b)fluoranthene, benzo(k)fluoranthene, benzo(ghi)perylene and indeno(1,2,3-cd)pyrene) amounts from E0.01 to 0.64 ng/ L compared to a maximum tolerable concentration for potable water of 0.1 mg/L. The benzo(a)pyrene concentrations in the leach water range from o0.015 to 0.3 ng/L compared to the limit of 10 ng/ L that is regulated separately from the other PAHs (EEC, 1998). The range of bitumens we have tested is representative of current commercial practice. All bitumens showed similar leaching behaviour against time. This indicates that our findings will be valid for other petroleum bitumens as well, eliminating the need for using the very time- and manpower- consuming static leach test. If one wants to determine the equilibrium concentration one can use a less time consuming dynamic test. As an alternative for performing a leach test one can calculate the aqueous concentrations from the PAH content of the bitumen and its distribution coefficient or its aqueous solubility. For example, the aqueous PAH concentrations predicted from the PAH concentrations in bitumens (pooled data from Table 4 and Brandt et al. (1985)) range from 0.08 to 0.6 ng/L for the sum of the six WHO PAHs. This compares well with the experimental range of E0.1–2.4 ng/L found in this study. Because no difference was found between leaching from bitumen and asphalt one can use these calculated values also for predicting the bitumen contribution to the leaching from asphalt.
CONCLUSIONS
In the static leach test a range of petroleum bitumens, representative for current commercial practice all show the same leaching behaviour against time. In the first days the concentration in the aqueous leachate increases and reaches a steady state concentration between 3 and 6 days. Distribution coefficients for the distribution between bitumen and water (Kbitwater) have been calculated for a number of PAHs. Log Kbitwater relates linearly with the log subcooled liquid solubility of the PAH. Bitumen and asphalt, the latter both in the static and a dynamic leach test give similar results. For
Aqueous leaching of polycyclic aromatic hydrocarbons from bitumen and asphalt
obtaining the equilibrium aqueous concentration, instead of the static leach test the less time consuming dynamic leach test can be used. The equilibrium PAH concentrations in the leach water from bitumens stay well below the surface water limits that exist in several EEC-countries and also are more than an order of magnitude lower than the current EEC limits for potable water. Instead of performing a leach test the aqueous concentration of a PAH can also be calculated from its concentration in the bitumen and the Kbitwater or the aqueous solubility of the PAH.
REFERENCES
Acevedo S., Escobar G., Gutie´rrez L. and Rivas H. (1992) Isolation and characterization of natural surfactants from extra heavy crude oils, asphaltenes and maltenes. Interpretation of their interfacial tension-pH behaviour in terms of ion pair formation. Fuel 71, 619–623. Amsterdam Method Series. The leaching behaviour of PAC from bitumen, static leaching test, AMS 1065-1. (Available from Shell Research and Technology Centre Amsterdam, SIOP OGMB, P.O. Box 38000, 1030 BN, Amsterdam.) Amsterdam Method Series. Determination of polycyclic aromatic compounds in condensed bitumen fumes, AMS 1057-1. (Available from Shell Research and Technology Centre Amsterdam, SIOP OGMB, PO Box 38000, 1030 BN Amsterdam.) Brandt H. C. A., de Groot P. C., Molyneux M. K. B. and Tindle P. E. (1985) Sampling and analysis of bitumen fumes. Ann. Occup. Hyg. 29, 27–80. Brantley A. S. and Townsend T. (1999) Leaching of pollutants from reclaimed asphalt pavement. Environ. Eng. Sci. 16(2), 105–116. Chiou C. T., Schmedding D. W. and Manes M. (1982) Partitioning of organic compounds in octanol-water systems. Environ. Sci. Technol. 16(1), 4–10. CONCAWE (1995) Gas oils (diesel fuels/heating oils). CONCAWE Brussels, product dossier no. 95/107. Dutch State Journal (1994) Staatscourant 95 d.d. 24 mei1994: circulaire interventiewaarden Bodemsanereing (9 mei 1994/nr. DBO/07494013). Dutch Normalisation Institute (1995) Leaching characteristics of shaped building materials. NEN 7345, Dutch Norm (in Dutch). EEC (1998) COUNCIL DIRECTIVE 98/83/EC on the quality of water intended for human consumption. Official Journal of the European Communities L 330/32, Brussels, 3 November.
4207
European Centre for Normalisation (1999) Compliance test for leaching of granular waste materials, including pastelike materials and sludges. 185CEN/TC292. IP 346/80 (1980) Polycyclic aromatics in petroleum fractions by dimethyl sulphoxide-refractive index method. Institute of Petroleum, London. IP 143/90 (2000) British Standards 2000. Determination of asphaltenes (heptane insolubles), Part 143. Kriech A. J. (1997) Evaluation of hot mix asphalt for leachability. Asphalt Institute.http://www.asphaltinstitute. org/environ/leachma.html. Lane W. F. and Loehr R. C. (1992) Estimating the equilibrium aqueous concentrations of polynuclear aromatic hydrocarbons in complex mixtures. Environ. Sci. Technol. 26(5), 983–990. Lane W. F. and Loehr R. C. (1995) Predicting aqueous concentrations of polynuclear aromatic hydrocarbons in complex mixtures. Water Environ. Res. 67(2), 169–173. Lee L. S., Hagwall M., Delfino J. J. and Rao S. C. (1992a) Equilibrium partitioning of polycyclic aromatic hydrocarbons from coal-tar into water. Environ. Sci. Technol. 26(11), 2104–2110. Lee L. S., Hagwall M., Delfino J. J. and Rao S. C. (1992b) Partitioning of polycyclic aromatic hydrocarbons from diesel fuel into water. Environ. Sci. Technol. 26(11), 2110– 2115. Miller H. C., Barret W. J. and James R. H. (1982) Investigating potential water contamination by petroleum-asphalt coatings in ductile-iron pipe. J. AWWA 74, 151–156. Mu¨nch D. (1992) Soil contamination beneath asphalt roads by polynuclear aromatic hydrocarbons, zinc, lead and cadmium. Sci. Total Environ. 126, 49–60. Mu¨nch D. (1993) Vergleich von asphaltierten und teerhaltig befestigten waldwegen bezuglich Ausbreitung von polycyclischen aromatischen Kohlenwasser-stoffen (PAH) und Schwermetallen. Bitumen 4, 152–158. Pearlman R. S., Yalkowsky S. H. and Bannerjee S. (1984) Water solubilities of polynuclear aromatic and heteroaromatic compounds. J. Phys. Chem. Ref. Data 13(2), 555–562. Picel K. C., Stamounidis V. C. and Simmons M. S. (1988) Distribution coefficients for chemical components of a coal–oil/water system. Water Res. 22(9), 1189–1199. Puzinauskas, V. P. and Corbett, L. W. (1978) Differences between petroleum asphalt, coal-tar pitch and road tar. The Asphalt Institute, Research Report, p. 78–1. Sadler R., Delamont C., White P. and Connel D. (1999) Contaminants in soil as a result of leaching from Asphalt. Toxicol Environ. Chem. 68, 71–81. Southworth G. R., Herbes S. E. and Allen C. P. (1983) Evaluating a mass transfer model for the dissolution of organics from oil films into water. Water Res. 17(11), 1647–1651. Zander M. (1995) Aspects of coal-tar chemistry/a review. Polycyclic Aromatic Compounds 7, 209–221.