Resources, Conservation & Recycling 174 (2021) 105736

Contents lists available at ScienceDirect

Resources, Conservation & Recycling

journal homepage: www.elsevier.com/locate/resconrec

Full length article

Utilization of municipal solid waste incineration bottom ash (IBA)

aggregates in high-strength pervious concrete
Peiliang Shen , Haibing Zheng , Jianxin Lu , Chi Sun Poon *
The Hong Kong Polytechnic University, Department of Civil and Environmental Engineering, Hung Hom, Kowloon, Hong Kong

A R T I C L E I N F O A B S T R A C T

Keywords: This study presents the use of IBA for the production of eco-pervious concrete, aiming at enhancing the recycling
Pervious concrete rate of IBA in construction materials. The IBA aggregates are used to replace the natural aggregates (NA) in
Incineration bottom ash aggregates various percentages (0%, 25%, 50%,75% and 100% by volume). The effect of IBA replacement ratios and casting
Compressive strength
methods on the mechanical properties, water permeability, porosity, density and thermal conductivity of the
Permeability
pervious concrete were determined. The volume stability of the pervious concrete incorporating IBA was eval­
Volume stability
uated. The results indicated that the appropriate content of IBA could benefit the mechanical properties due to
the internal curing effect of IBA. However, the excessive content of IBA in the pervious concrete led to reduced
water permeability, connected porosity and mechanical properties. Even so, the pervious concrete containing
100% IBA could achieve the required permeability and compressive strength for permeable pedestrian pavers
(JIS A 5371). In addition, the incorporated IBA in pervious concrete causes little volume expansion from the
reaction between alkalis and components consisting of metallic aluminum and glass in IBA owing to its high
connected porosity. Furthermore, the pervious concrete incorporated IBA had a low density and thermal con­
ductivity, showing the potential to be applied in thermal insulation areas.

1. Introduction and a lower strength level. Usually, the pervious concrete has a high
porosity of between 11% and 35% (Ćosić et al., 2015; Putman and
Waste incineration is widely used to manage municipal solid waste, Neptune, 2011; Tennis et al., 2004), and hence it possesses excellent
as the incineration process can reduce the volume of municipal solid permeability, good drainage property and high noise absorption ca­
waste by nearly 90% and the mass by 80% (Keulen et al., 2016). This pacity (Kim and Lee, 2010). Since the use of pervious concrete was re­
method is increasingly adopted all over the world including in devel­ ported to have multiple environmental benefits: facilitating stormwater
oping countries such as China (Wu et al., 2016). However, after the run-off, mitigating urban heat island effect, reducing road noise and soil
incineration process, there is still a large amount of incineration bottom pollution, etc. (Tennis et al., 2004; J.-X. Lu et al., 2019; Qin, 2015; Li
ash needed to be managed. According to the world waste production, the et al., 2013), it is widely used in various areas such as pavements,
annual municipal solid waste production was 1840 MT worldwide (data sidewalks, parks and building exteriors (Bhutta et al., 2012). Hong Kong
from 2012) (Lynn et al., 2016). As a result, approximately 360 MT of IBA is a densely built-up urban city and located in the region where heavy
was produced. The majority of IBA is currently disposed of at landfills rainfalls occur frequently. It is essential to promote the application of
(Chuang et al., 2018). Due to the shortage of landfill sites, it is an pervious concrete to avoid potential waterlogging.
increasingly important issue to develop novel solutions to reuse IBA as Since the over-exploitation of natural aggregates (NA) has last for a
resources. long time due to rapid urbanization, NA have become a scarce resource
Pervious concrete, which contains a large number of pores, was in Hong Kong and most cities in China. Reuse of inert wastes to replace
developed in the 1940s (Brown, 2006). A pervious concrete normally NA is one of the methods to relieve this concern. This study explored the
consists of coarse aggregate, cement paste and voids (Huang et al., feasibility of using IBA aggregates to replace NA to broaden the source of
2010). Compared to conventional concrete, it contains a lesser amount aggregates. Recently, some researches were carried out to use the IBA as
of cement paste and no fine aggregate, which leads to a higher porosity aggregates in pervious concrete. Due to the lower hardness of IBA than

* Corresponding author.
E-mail addresses: [email protected] (J. Lu), [email protected] (C.S. Poon).

https://doi.org/10.1016/j.resconrec.2021.105736
Received 12 March 2021; Received in revised form 29 April 2021; Accepted 6 June 2021
Available online 5 July 2021
0921-3449/© 2021 Elsevier B.V. All rights reserved.
P. Shen et al. Resources, Conservation & Recycling 174 (2021) 105736

that of NA, Kuo et al. used the washed IBA to replace the NA to prepare Table 2
low-strength pervious concrete (4.8 MPa-12.7 MPa) (Kuo et al., 2013). Physical properties of aggregates.
Wu (Wu et al., 2016) also used IBA in place of sandstone graded material Aggregates NA IBA
to prepare pervious concrete. The results indicated that the strength of 3
Loose Bulk Density (g/cm ) 1.33 1.049
permeable concrete blocks was less than 15 MPa, and it had lower Specific Gravity 2.54 1.88
permeability coefficients than that of conventional pervious concrete, Water Absorption (%) 1.63 11.64
which led to being impracticable in busy roads (Wu et al., 2016). Aggregate crushing value (%) 14.8 47.9
Generally, the mechanical properties of pervious concrete incorporating Los Angeles abrasion value (%) 17.6 49.5

IBA are usually much lower than that of pervious concrete prepared by
pure NA. Therefore, it is of great significance to explore methods to
improve the compressive strength of pervious concrete incorporating Table 3
IBA. Mix proportion of UHSC paste.
The objective of this research is to develop a high-strength pervious No. Cement Fly ash Silica fume w/b Superplasticizer
concrete incorporating IBA aggregates to replace NA. The ultra-high – 1 0.1 0.22 0.15 0.04
strength cement paste (UHSC) was prepared to improve its compres­
sive strength. The influences of IBA-NA replacement ratio and different
casting methods on the compressive strength, water permeability and cohesiveness of the fresh pervious concrete mixture was sufficient to
thermal conductivity of the pervious concrete were determined. Micro- hold its shape when it was hand-pressed, but the paste did not stick to
hardness, hydration heat and microscopy studies were also conducted to the hand itself. An aggregate to binder ratio (a/b) of 4 was set to achieve
explore the mechanisms controlling the performance of the pervious high strength and permeability according to our previous research
concrete incorporating IBA. (Shen et al., 2021). The IBA aggregate with a size ranging of 2.36–5 mm
was applied to replace the NA by volume. As the water absorbed in IBA
2. Materials and experiments was considered, the total w/b ratio was increased. The mixing process
was as following: 1) The IBA was immersed in water for 24 h and wiped
2.1. Materials and mixture preparation dry, then the saturated IBA in surface dry condition was collected; 2) All
the binders were dry mixed for 3 min, then the clean tap water with an
A 52.5 CEM I Portland cement (OPC) and fly ash (FA) sourced from appropriate amount of superplasticizer were poured into the mixer and
Hong Kong were used, and the silica fume (SF) produced by Elkem slowly mixed for 3 min followed by high speed mixing for another 3 min;
(China) was also used as well. The oxide compositions of OPC, FA and SF 3) The aggregates (including NA and IBA) were added into the mixer and
are shown in Table 1. Crushed granites with the size between 2.36 mm the mixture was further mixed for 3 min at a low mixing speed. After­
and 5 mm were served as the natural aggregates (NA). The IBA aggre­ ward, the freshly prepared mixtures were cast into molds.
gate (2.36–5 mm) was produced from a municipal solid waste inciner­ Two methods of casting and vibration were used to prepare the
ation facility in China. The compacted voids of granite and IBA were pervious concrete as follows: Method 1) The fresh mixture was poured
40.8% and 41.5%, respectively. The physical properties of these ag­ into the molds and compacted using a laboratory vibration table for 1
gregates are presented in Table 2. A superplasticizer high range water- min; Method 2) the fabrication procedure of the pervious concrete
reducing rate produced by BASF was also used to control the simulated the production process of concrete blocks in a real block
workability. precast plant using a relatively dry concrete mix with very low work­
In order to improve the mechanical properties of the UHSC paste, its ability. Thus, after casting half of the amount of the molds followed by
porosity is usually minimized by improving the packing density of the the normal compaction using the vibration table, the remaining half of
particles. In this study, the modified Andreasen and Andersen model was the mixtures were added, but the mixtures inside the molds were
employed to design a UHSC paste with excellent compressive strength transferred to a compression machine for stronger compaction with a
by using OPC, SF and FA (Yu et al., 2014). A target function obtained maximum molding pressure of 3 MPa (loading rate of 0.05 MPa/s, and
from this model was used for optimizing the composition of UHSC paste. kept at the maximum pressure for 60 s). This compaction parameter for
The mix proportions of each raw material can be chosen when an op­ fabricating the mixed specimens followed those used in previous studies
timum fit between the composed mix curve and the target curve is (J.-X. Lu et al., 2019; J.-X. Lu et al., 2019; Yang et al., 2019), which was
reached by Least Squares Method. Therefore, the mixture proportion of necessary to compact the constitutes. Afterward, the samples were
the UHSC paste will be obtained and show in Table 3. overlayed by preservative films. Then, the samples were demolded at 24
h and cured in a sealing condition at 25 ◦ C.

2.2. Preparation of pervious concrete

2.3. Experimental methods
The mix proportions of the pervious concrete are shown in Table 4. A
water to binder ratio of 0.15 (not include water in IBA) was chosen based 2.3.1. Compressive strength
on the trial tests that the mixtures condition could be agglomerated into One intended use of the pervious concrete could be precast pervious
a ball shape by hand without falling off of the aggregates (Tennis et al., blocks for pedestrian pavement. Thus, in this study, the geometry of the
2004). Meanwhile, the UHSC paste possessed a slump flow of 165 mm. pervious concrete specimens (200 mm × 100 mm × 60 mm) was pre­
After adding aggregate, a cohesive mix with very low workability (no pared according to BS 6717 and our previous study (Lu and Poon, 2019;
slump value) was produced. This condition occurred when the BS 6717, 2001). The compressive strength of pervious concrete with a

Table 1
Chemical compositions of cementitious materials (wt.%).
Oxide SiO2 Al2O3 CaO Fe2O3 SO3 MgO Na2O K2O LOI

OPC 19.61 7.32 63.15 3.32 2.03 2.14 0.13 0.32 2.12
SF 88.29 0.14 0.92 0.19 1.51 3.21 0.14 0.17 5.26
FA 54.69 12.71 17.48 4.61 0.97 1.68 1.75 3.18 1.79
IBA 40.00 9.42 29.40 6.69 1.17 1.37 2.78 1.97 5.72

2
P. Shen et al. Resources, Conservation & Recycling 174 (2021) 105736

Table 4
Mix proportion of pervious concrete (kg/m3).
Notation OPC FA SF w/b Water PCE Water in IBA IBA NA a/b Pressure Total w/b

0FP 262.2 65.6 72.2 0.15 56.8 4 0 0 1600 4 3 MPa 0.15

25FP 262.2 65.6 72.2 0.15 56.8 4 34.2 295 1200 / 3 MPa 0.24
50FP 262.2 65.6 72.2 0.15 56.8 4 68.4 590 800 / 3 MPa 0.32
75FP 262.2 65.6 72.2 0.15 56.8 4 102.7 885 400 / 3 MPa 0.41
100FP 262.2 65.6 72.2 0.15 56.8 4 136.9 1180 0 / 3 MPa 0.49
0FV 262.2 65.6 72.2 0.15 56.8 4 0 0 1600 4 0 0.15
50FV 262.2 65.6 72.2 0.15 56.8 4 68.4 590 800 / 0 0.32
0FV1 349.6 87.5 96.3 0.15 76.8 5.3 0 0 1600 3 0 0.15

size of 200 mm × 100 mm × 60 mm was measured according to the

procedure in ASTM C39 (ASTM C39, 2012). The pervious concrete was
loaded at a rate of 0.6 MPa/s with 200 × 100 mm2 surface upward. The
pervious concrete prepared with the two different casting methods and
amounts of IBA was chosen for strength measurement after curing for 28
d For samples casting with Method 2, the compressive strength of
samples incorporating with 0% and 50% IBA was determined at 3 d, 7
d and 28 d

2.3.2. Porosity
The connected porosity of pervious concrete with a size of 200 mm ×
100 mm × 60 mm was measured according to the method provided in a
previous study (Neithalath et al., 2006). Firstly, the volume (V) of
pervious concrete was obtained by measuring the dimension. Afterward,
the pervious concrete was immersed in water for one day to saturate the
pores with water. Then, the submerged weight (W1) of pervious concrete
was recorded. Finally, the weight of the specimens (W2) in the air was Fig. 1. Set-up of water permeability measurement.
measured after achieving a saturated surface dry condition. The con­
nected porosity was calculated as follows: VL
[ ] C = 10 × (3)
W2 − W1 AHT
P = 1− × 100% (1)
ρV
where C is the permeability coefficient, mm/s; A is the surface area of
where P and ρ are connected porosity of pervious concrete and bulk pervious concrete, cm2; V is the total volume of collected water, mL; L is
density of water, respectively. the thickness of specimens; H is the differentials between the height of
The total porosity of the specimens was determined in accordance water head in tube and water head in the bottle, cm; T is the recorded
with ASTM C 1754 (ASTM C 1754, 2012). Firstly, the specimens were time, s.
immersed in water for 24 h to make sure that the pores in IBA and the
cement matrix were saturated. Then the submerged weight (W3) of 2.3.4. Hydration heat
specimens was recorded. Finally, the weight of the specimens was The effect of IBA on hydration heat of pervious concrete was moni­
measured (W4) after drying in an oven at 105 ◦ C for 48 h, and. The total tored by using an I-CaI 4000 isothermal calorimeter. For each mixture,
porosity (P1) of specimens can be calculated as follows: about 20 g of binders and 80 g NA or IBA were placed into a plastic
[ ] ampule and put into the calorimeter. The IBA replacement ranged from
W4 − W3 0% to 100%. The isothermal calorimetry was started immediately after
P1 = 1 − × 100% (2)
ρV mixing and record for 168 h.

2.3.3. Water permeability 2.3.5. Density

The cylindrical specimens (Φ100 mm × 50 mm) were prepared and The density of pervious concrete was measured based on the method
used for water permeability measurement according to standard JIS A illustrating in ASTM C 1754 (ASTM C 1754, 2012). The weight of
5371 and GBT 25,993–2010 (J.-X. Lu et al., 2019; JIS A 5371, 2016; pervious concrete with a size of 200 mm × 100 mm × 60 mm was
GB/T 50082–2009, 2009). Fig. 1 shows the setup of permeability mea­ measured after drying in the oven at 105 ◦ C. In addition, the dimensions
surement. The water released from the bottle was collected, and the time of specimens were determined to calculate the volume. The density of
was recorded. The permeability coefficient could be calculated based on specimens can be calculated as follows:
Eq. (3). It should be noted that the influence of the thickness of speci­
M
mens was taken into account when the permeability coefficient was ρ= (4)
V
calculated. As the intended application of the pervious concrete was to
produce concrete blocks for pedestrian pavement, thus a similar thick­ where the ρ is the density of specimens, g/cm3; The M is the weight of
ness to the real blocks (60 mm) was produced to evaluate the perme­ pervious concrete after drying in an oven at 105 ◦ C, g; V is the volume of
ability in this study compared to the previous studies (Martin III et al., pervious calculated based on the dimensions, cm3.
2014). Because the pervious concrete was prepared by aggregates with
small size ranging 2.36–5 mm, leading to well-distributed pore structure 2.3.6. Microhardness
with small pore size. As a result, the specimens with a thickness of 50 The surface of pervious concrete containing different amounts of IBA
mm were representative and accurate for permeability measurement. was polished for the microhardness test. The pervious concrete was well
grinded to achieve a smooth surface. A digital Vickers microhardness
tester (HVX1000A, China) was used to determine the microhardness of

3
P. Shen et al. Resources, Conservation & Recycling 174 (2021) 105736

the matrix around aggregates. For pervious concrete incorporating IBA, effect is similar to the effect of lightweight aggregates on the strength of
50 points near IBA were tested. In case of pervious concrete without IBA, ultra-high performance concrete with low water to binder ratio (P. Shen
the 50 points were conducted in matrix near NA. et al., 2020).
These results indicated that the controlled addition of IBA could
2.3.7. Autogenous shrinkage and expansion behavior improve the compressive strength of the pervious concrete. When the
The effect of IBA on the expansion behavior of the pervious concrete content of IBA was lower than 50%, an improved compressive strength
was tested according to ASTM C 1260 (ASTM C 1260, 2014). The could be observed although the NA was replaced by the porous IBA with
specimens with a size of 25 mm × 25 mm × 285 mm were prepared and poor mechanical properties. This might be related to the pre-soaked in
demolded at 24 h after casting, and then placed into 80 ◦ C water for 24 IBA, as the release of pre-soaked water contributed to the improved
h. The zero value was recorded after the hot water curing. Afterward, the hydration degree. As a result, the bonding strength between the aggre­
specimens were soaked in 1 mol/L NaOH solution at 80 ◦ C for 28 d The gates and the cement paste was improved, contributing to the increased
dimensional development of the samples was determined at 1 d, 3 d, 7 d, compressive strength. Some of the pre-soaked water in IBA would be
14 d, 21 d and 28 d, respectively. released before setting due to the low relative humidity in UHSC paste
The autogenous shrinkage of the samples was measured according to according to the internal curing theory (Bentz and Weiss, 2011). This
the Chinese standard (GB/T50082-2009, 2009). The specimens with a would increase the volume of UHSC paste, resulting in the reduction of
size of 25 mm × 25 mm × 285 mm were also used. The autogenous porosity, which also contributed to the enhancement of compressive
shrinkage test was started at the age of 24 h, then until the age of 60 d strength. Although the compressive strength was reduced by 31.1%
while the IBA content reached 100%, it could still meet the requirement
3. Results and discussion of permeable concrete for pedestrian pavement according to JIS A 5371
(to JIS A 5371, 2016).
3.1. The mechanical properties
3.1.2. Effect of casting method
3.1.1. Effect of iba replacement As indicated in Fig. 3, the compressive strength of the pervious
Generally, as the IBA is a porous aggregate with low strength, it is concrete could be increased by applying molding pressure during pro­
expected to be the weakest zone in the prepared pervious concrete. duction. For the concrete prepared without IBA, the compressive
Hence, an understanding of the effect of IBA on the mechanical prop­ strength was only slightly improved. However, significant enhancement
erties of pervious concrete is necessary. The compressive strength of was observed with the use of IBA. It is expected that the use of molding
specimens containing different contents of IBA is shown in Fig. 2. It is
observed that the incorporation of IBA had a rather inconsistent effect
on the compressive strength. Compared to the reference sample, the
compressive strength of the pervious concrete incorporated with 25%
IBA was improved by 60%. This may be explained by the water absorbed
in IBA and the reduction in porosity (shown in Fig. 4). However, the
compressive strength was obviously reduced when the replacement of
IBA was higher than 75%. Fig. 2(b) shows the development of
compressive strength of the concrete prepared with and without IBA. It
can be observed that the 3 d and 7 d compressive strengths of the
reference sample were higher than that of the samples containing 50%
IBA, while the 28d-compressive strength of the reference sample was
lower than that of the concrete prepared with IBA.
For the reference sample, the UHSC paste was prepared with a low
w/b ratio of 0.15, and the mixing water would be exhausted at the early
age (Shen et al., 2019). Hence, the increment of compressive strength
was limited between 7 d and 28 d In contrast, the porous IBA can act as
an internal curing material in the pervious concrete, which increased the
internal relative humidity of the matrix at both the early and later ages.
Therefore, the hydration degree of the cement paste in pervious concrete
was improved due to the water released from IBA (Kevern and Now­ Fig. 3. Effect of casting methods on the compressive strength of
asell, 2018), thus leading to the higher 28 D-compressive strength. This pervious concrete.

Fig. 2. Effect of IBA on the compressive strength of pervious concrete.

4
P. Shen et al. Resources, Conservation & Recycling 174 (2021) 105736

pressure would produce a denser pervious concrete, thus resulting in a Fig. 5). The decreased connected porosity would contribute to the
higher strength compared to the reference sample prepared without reduced total porosity and increased density. The increased density from
molding pressure. Generally, the individual pellet strength of IBA was decrease of connected porosity surpassed the decreased density due to
quite low (Tang et al., 2019); thus, some of the particles would be replacement of the high-density NA by low-density IBA when 25% of NA
crushed when subjected to the molding pressure, leading to a low are replaced with IBA. As the connected porosity of pervious concrete
porosity (shown in Fig. 4) and a higher compressive strength. only shows a slight decrease with the increasing IBA content (shown in
Fig. 5), the increased in density resulted from the decrease of connected
3.2. Water permeability and connected porosity porosity stabilized when a higher IBA replacement was applied (>25%).
However, with the increasing IBA amount, the density of pervious
Fig. 4(a) shows the effect of IBA replacement on the water perme­ concrete was governed by the large amount of low density IBA. This
ability of the pervious concrete. It is indicated that the pervious concrete trend was consistent with the opposite trend of the total porosity shown
containing IBA had lower permeability than that of the reference sam­ in Fig. 8. The IBA had a porosity of 21.9%, which was much higher than
ple, which was attributed to the reduced connected porosity (shown in that of NA. Thus, the increased content of IBA would increase the total
Fig. 5(a)). Besides, for the pervious concrete prepared without IBA, the porosity of the pervious concrete. However, when 25% of NA was
casting method had little effect on the water permeability, and the replaced by IBA, the decreased in total porosity was caused by the
values were about 3 mm/s. But for the samples incorporating 50% IBA, decreased connected porosity (shown in Fig. 5). As a result, a sudden
the permeability coefficient was reduced to 2.4 mm/s (Fig. 4(b)), reduced total porosity could be observed in sample 25FP. As the con­
although the connected porosity was slightly decreased (Fig. 5(b)). It nected porosity of the pervious concrete only slightly decreased with the
might be due to the different grading of IBA and the increased volume of increasing IBA content, the decreased in total porosity resulted from the
cement paste due to desorption of water from IBA, which can reduce the decrease of connected porosity stabilized when a higher replacement
connected porosity and the connectivity of voids. Meanwhile, with the was applied (>25%). Meanwhile, many pores from IBA were induced
application of 3 MPa molding pressure, the water permeability of and increased as increasing IBA content, which overwhelmed the
pervious concrete incorporating 50% IBA was significantly reduced to decreased total porosity from decrease of connected porosity when the
0.6 mm/s. The damage of IBA under molding pressure and increase of IBA replacement reached 50%. Besides, both the density and total
cement paste are believed to induce the reduction in the porosity and porosity showed a linear relationship with the compressive strength
connectivity of the pores of the pervious concrete. It should be noted (shown in Fig. 9), indicating the critical influences of these two pa­
that the use of IBA is detrimental to the water permeability, but it was rameters on the compressive strength.
still higher than the permeability requirement of JIS A 5371 (0.1 mm/s). For the specimens prepared without IBA, the casting method had
Fig. 6 shows the relationship between the connected porosity and only a slight influence on the density and porosity, as well as the
permeability. Overall, the water permeability increased with the compressive strength and water permeability (shown in Fig. 2 and
increasing connected porosity, but no definitive relationship could be Fig. 4) of the pervious concrete. However, the density of pervious con­
found. The porosity of the pervious concrete incorporating 25% IBA was crete containing IBA was increased with the application of the molding
a little higher than that of the concrete containing 75% and 100% IBA. pressure. The density of FP50 prepared with 50% IBA was 1920.5 kg/
This might be attributed to the changes in pore structure between the m3, which was much higher than that of those prepared without
aggregate particles. Usually, the water permeability of pervious concrete applying the molding pressure (1781.3 kg/m3). This means that the
is not only related to the connected porosity but also its pore structure packing density was increased when the molding pressure was applied
(Neithalath et al., 2006). So, it showed the presence of IBA would likely to the mix prepared with the inhomogeneous, weaker and more porous
influence the characteristics of the pore structure. IBA. The total porosity of the concrete prepared with 50% IBA was
decreased by 30.4% when the molding pressure was used during sample
3.3. Density and total porosity preparation.

The density and total porosity of the pervious concrete are shown in 3.4. Thermal conductivity
Figs. 8 and 9, respectively. The density of the pervious concrete was
obviously affected by IBA content and the casting method. The density The thermal conductivity of pervious concrete incorporating with
values ranged from 1709.2 to 1994.2 kg/m3, which were much lower IBA is shown in Fig. 10. Apparently, the pervious concrete prepared with
than that of the normal concrete. In addition, as indicated in Fig. 7(a), 25% IBA showed the highest thermal conductivity. But the thermal
the density of the concrete incorporating IBA increased first and then conductivity decreased with the further increase in IBA replacement.
decreased with the increasing amount of IBA. the connected porosity of The thermal conductivity of the pervious concrete containing 100% IBA
the pervious concrete incorporating IBA was decreased (shown in decreased to 0.61 W/mK. The is mainly attributed to the low thermal

Fig. 4. Effect of IBA replacement ratio and casting methods on water permeability.

5
P. Shen et al. Resources, Conservation & Recycling 174 (2021) 105736

Fig. 5. Connected porosity of pervious concrete.

3.5. vol stability

Fig. 11 shows the effect of IBA replacement on the autogenous

shrinkage of the pervious concrete. Generally, ultra-high performance
concrete has a very high autogenous shrinkage owing to its character­
istics of low w/b ratio and high content of superfine particles (Shen
et al., 2018). It can reach 2000 µε at 28 d (Shen et al., 2019). Thus, the
UHSC paste would also possess a high shrinkage value. However, as
there is a large amount of coarse aggregate in the pervious concrete, the
autogenous shrinkage was mitigated by the skeleton of the aggregate.
The 28 D-autogenous shrinkage of the pervious concrete was very lower
(<140 µε (shown in Fig. 11)), which was also contributed by the internal
curing effect of the pre-saturated porous IBA. The autogenous shrinkage
value decreased with increasing IBA content.
The expansion behavior caused by the chemical reactions of the
contaminants in IBA (such as metallic aluminum, glass and gypsum (D.
Xuan et al., 2018)) was measured and shown in Fig. 12. An expansion
behavior could be observed in all the samples prepared with different
Fig. 6. Relationship between connected porosity and permeability. amounts of IBA. This expansion of the reference sample may be attrib­
uted to the alkali-silica reaction of the amorphous silica in the granite
conductivity of IBA aggregates. IBA is a type of porous material and its (Sreenivasulu et al., 2015). But the expansion values of samples con­
thermal conductivity is much lower than that of natural granite aggre­ taining different contents of IBA were relatively low compared to that of
gates (Cho et al., 2009). Therefore, the pervious concrete containing the expansion in conventional concrete incorporated IBA (D. Xuan et al.,
100% IBA presented the minimum thermal conductivity. It should be 2018). Meanwhile, the expansion of the samples consisting of high IBA
noted that the thermal conductivity of pervious concrete incorporating content was lower compared to concrete incorporating the low amount
100% IBA was comparable to the thermal conductivity of lightweight of IBA. This might be attributed to its porous structure, which could
concrete (Demirboga, 2007). accommodate the volume expansion induced by the formation micro
expansive materials (i.e., H2 gas and ASR gel etc.). This phenomenon is
in agreement with the expansion behaviors in dry-mixed concrete glass
blocks (J.-X. Lu et al., 2019).

Fig. 7. Effect of IBA replacement and casting method on the density of pervious concrete.

6
P. Shen et al. Resources, Conservation & Recycling 174 (2021) 105736

Fig. 8. Effect of IBA replacement and casting method on the total porosity of pervious concrete.

Fig. 9. Relationships between density or porosity and compressive strength.

Fig. 10. Thermal conductivity of pervious concrete containing different

amounts of IBA. Fig. 11. Autogenous shrinkage of pervious concrete containing different
amounts of IBA.

4. Further discussion
25% IBA and aggregate to binder ratio of 4 showed higher compressive
strength than the specimens with aggregate to binder ratio of 3. In other
4.1. The enhancement of strength by internal curing
words, the addition of IBA can promote the enhancement of compressive
strength. Although the IBA possesses a lower hardness than that of NA,
A pervious concrete with the a/b of 3 was also prepared by NA
the internal curing of IBA may overwhelm the reduced compressive
without molding pressure. The compressive strength, total porosity,
strength owing to the inhomogeneous, weak and porous characteristics
density and permeability of this pervious concrete were 46.15 MPa,
of IBA. Therefore, the micro-hardness of matrix and hydration heat of
17.85%, 2007.3 kg/m3 and 0.95 mm/s, respectively. The connected
pervious concrete containing different contents of IBA were measured to
porosity (13.67%) and water permeability (0.95 mm/s) were a little
explore mechanisms governing the compressive strength enhancement
smaller than those of pervious concrete containing 25% IBA and
of pervious concrete incorporating appropriate IBA.
aggregate to binder ratio of 4, but the pervious concrete prepared with

7
P. Shen et al. Resources, Conservation & Recycling 174 (2021) 105736

Fig. 12. Expansion of pervious concrete in 80 ∘ C NaOH solution.

Fig. 13 shows the hydration heat flow and cumulative heat of low strength of porous IBA. Usually, the compressive strength of
pervious concrete incorporating 0%, 25%, 50%, 75% and 100% IBA. It pervious concrete was controlled by the strengths of cement paste and
can be observed from Fig. 13(a) that the dormant period was reduced aggregates, and their bonding. Although the cement paste was
with the increased amount of IBA. Usually, the dormant period depends strengthened by the internal curing of porous IBA at an early age (3d and
on the water to cement ratio and the content of the superplasticizer in 7d), the low crushing strength of IBA would govern the overall strength
concrete (Hasanzadeh et al., 2016; Justs et al., 2015; Wang et al., 2015). of the pervious concrete. Compared to the reference sample, an
As the dosage of water and superplasticizer is the same in all the spec­ improved compressive strength could be observed in samples incorpo­
imens, the acceleration of dormant period is likely due to the water rated 50% IBA at 28d This was related to the continually internal curing
release and ion dissolution of IBA (D. Xuan et al., 2018). Meanwhile, effect of IBA at the age between 7 d and 28 d (a higher heat flow was
some water in internal curing materials would release before setting in observed at 7d) during which the hydration of the reference sample was
UHSC paste due to its low relative humidity (Ma et al., 2019). Similarly, limited due to lack of water (P. Shen et al., 2020). Because the improved
some pre-soaked water in IBA can also be released during the dormant compressive strength from the strengthened cement paste due to inter­
period. As a result, the increased w/b further influences the dormant nal curing surpassed the reduced compressive strength due to
period. low-strength IBA.
Furthermore, an accelerating period was followed by a main peak. Generally, the pervious concrete incorporating IBA can not only
The appearance of this main peak was ahead by adding more IBA accelerate the hydration of UHSC paste at the early age but also promote
obviously, and the location of this main peak shifts from 35 h to 15 h its hydration at the later age. The high internal relative humidity due to
while the IBA content increased from 0% to 100%. Accordingly, the the water desorption plays the main role in controlling the hydration,
cumulative heat of hydration increased with the increase of IBA addi­ leading to the improvement of hydration heat and hydration degree
tions, especially for the early age. Nevertheless, the heat flow of pervious (Espinoza-Hijazin and Lopez, 2011; Halleck and Bentz, 2006). It is ex­
concrete containing IBA was higher than that of the blank sample at 3 d, pected that the UHSC paste in pervious concrete incorporating IBA has a
which was due to the water desorption of IBA. However, the 3 d and 7 lower porosity compared to that of the reference sample. Although the
d compressive strengths of the reference sample were still higher than IBA is a type of porous material with low hardness and acts as the
that of the samples containing 50% IBA. This might be attributed to the weakest zone in pervious concrete, the enhancement of the matrix

Fig. 13. The heat hydration of pervious concrete.

8
P. Shen et al. Resources, Conservation & Recycling 174 (2021) 105736

owing to its internal curing effect partly compensated for the adverse
influence of IBA. Therefore, an enhancement or limited reduction of
compressive strength can be observed when the appropriate amount of
IBA is used.
The average microhardness of matrix around IBA in pervious con­
crete containing different IBA contents is shown in Fig. 14. The test
results indicated that the average microhardness increased with the
increasing amount of IBA. The increase of microhardness is mainly
attributed to the increasing degree of the matrix (shown in Fig. 13).
Generally, compared to the reference sample, a denser matrix around
the aggregate can be obtained due to the internal curing of IBA, which
exhibits excellent micro-mechanical performance (Shen et al., 2017).
Herein, although the addition of porous IBA gives rise to weakness zones
into pervious concrete, the improvement of matrix acts as a dense wall,
which can prevent the crack of IBA.
Fig. 15. The skeleton effect of UHSC paste.
The pre-saturated water in IBA not only promotes the hydration of
cement paste but also enhances its mechanical property. The micro­
hardness of cement paste in pervious concrete is significantly improved, Fig. 5), which benefited the compressive strength. Therefore, the above
which means that the cement paste around aggregate acts as the dense benefits surpassed the negative effects with 25% IBA incorporation. But
wall with low porosity and high mechanical strength. This dense wall is the adverse effect increased obviously with the further increase in IBA
formed, as shown in Fig. 15. As the w/b ratio of pervious concrete is very replacement. For pervious concrete incorporating a high amount of IBA
low, the pre-saturated water in IBA will be motivated by relative hu­ (larger than 75%), the reduction of compressive strength was higher
midity gradient and releases into the cement paste during hydration. than the enhancement. As a result, the highest compressive strength
The hydration of cement paste near IBA will be promoted and the could be obtained in pervious concrete incorporating 25% IBA.
microstructure becomes denser than that of the reference sample. This
dense wall has a higher mechanical strength than that of the reference 4.2. The mechanism of volume stability
sample, which may act as a skeleton in pervious concrete containing IBA
and will be of benefit to the bonding strength between aggregates and The incorporated IBA in pervious concrete has limited effect on the
mechanical strength of pervious concrete. expansion level. This is mainly attributed to the high porosity (shown in
In this study, the compressive strength of pervious concrete con­ Fig. 16), which is able to accommodate the alkali-silica-reaction gel
taining 25% IBA was higher than that of the other specimens. This was produced (Lee et al., 2011). In addition, the silica fume and fly ash in the
ironic because the presence of porous IBA aggregate with poor me­ matrix will also mitigate the alkali-silica-reaction expansion of the glass
chanical properties should have reduced the strength. This can be in IBA (Du and Tan, 2013). Furthermore, the aggregates (including NA
explained by a number of factors. Usually, the compressive strength of and IBA) are coated by cement paste, and the thickness of coatings
pervious concrete depended on the compressive strength of cement around aggregate is low. Therefore, the H2 gas from the
paste and aggregate, porosity and the bonding characteristic between alkaline-metallic Al reaction in IBA can release easily; thus, the expan­
cement paste and aggregate (Shen et al., 2021). From the hydration sion from H2 release can be reduced. As a result, a relatively low
results, the absorbed water in IBA promoted cement hydration which expansion can be obtained for pervious concrete incorporating IBA.
densified the UHSC paste. Also, the increased micro-hardness facilitated According to ASTM C1260, an expansion of less than 1000 µε is indic­
the formation of dense shell around the IBA and the interfacial transition ative of innocuous behavior of aggregates in most cases. Therefore, in
zone around the aggregates would be improved due to the internal case of pervious concrete, the IBA can be used as aggregate safely and
curing provided by the IBA and the pozzolanic reaction between IBA and effectively without considering the crack risk from the expansion of IBA.
cement paste (Qiao et al., 2008; Polettini et al., 2005), which was Indeed, the presence of heavy metals in IBA is an issue. The proposed
confirmed by the result of increased micro-hardness. The improvements procedure was able to address the potential expansion of concrete
of the UHSC paste and the interfacial transition zone contributed to the incorporated IBA, which produced pervious concrete with good quality
increased compressive strength. Furthermore, the connected porosity of even when the replacement of IBA reached 100%. However, the porous
pervious concrete incorporated with 25% IBA was decreased (shown in structure might provide the route for potential chemical leaching (i.e.
heavy metals and salts) of IBA (Verbinnen et al., 2017). But it is
generally recognized that the IBA from MSW incinerators is not a haz­
ardous waste and it can be reused/disposed of by normal means.
Furthermore, a dense wall with ultra-high micro-mechanical properties
was formed on the surface of IBA, which will be of benefit to the bonding
strength between aggregates and mechanical strength of pervious con­
crete. This dense structure of cement paste could certainly reduce the
potential leaching of heavy metals. A balance between adverse effect of
porous structure of pervious concrete and benefits from the dense wall
around IBA might be obtained when the appropriate amount of IBA with
a dense wall was were applied. Therefore, further studies will be carried
out to evaluate the leaching behaviors of heavy metals from pervious
concrete incorporated IBA, and the process and factors that influence the
potential leaching of heavy metals will be explored.

5. Conclusion

In this study, the physical properties, hydration and volume stability

Fig. 14. The micro-hardness of the matrix in pervious concrete. of pervious concrete incorporating with IBA were systemically

9
P. Shen et al. Resources, Conservation & Recycling 174 (2021) 105736

Fig. 16. The porous structure of pervious concrete with and without IBA.

characterized. From the performance point of view, it is potential to use Supervision, Funding acquisition.
IBA in the production of eco-pervious concrete. Based on the experi­
mental results and discussions, the conclusion can be drawn as follows.
Declaration of Competing Interest
(1) The use of IBA to replace the NA in pervious concrete led to an
The authors declared that they have no conflicts of interest to this
enhancement of compressive strength when the appropriate
work.
amount of IBA was utilized (lower than 50%). The IBA can act as
an internal curing material and motivated the hydration of UHSC
paste, leading to a higher hydration degree and microhardness. Acknowledgments
The use of molding pressure during production could produce a
denser pervious concrete, thus resulting in a higher strength. The authors wish to acknowledge the financial supports from the
(2) The use of IBA led to the reductions of water permeability and Innovation Technology Fund and the Environment and Conservation
connected porosity due to the damage of IBA during casting Fund.
under pressure and increased volume of cement paste. Although
100% IBA was used as aggregates, the compressive strength and Reference
the water permeability could meet the requirement of JIS A 5371
Keulen, A., Van, Z.A., Harpe, P., Aarnink, W., Simons, H.A., Brouwers, H.J., 2016. High
specification. performance of treated and washed mswi bottom ash granulates as natural aggregate
(3) The density of developed eco-pervious concrete ranged from replacement within earth-moist concrete. Wast. Manage. 49, 83–95.
1709.2 to 1994.2 kg/m3, which was lower than that of normal Wu, M.H., Lin, C.L., Huang, W.C., Chen, J.W., 2016. Characteristics of pervious concrete
using incineration bottom ash in place of sandstone graded material. Construct.
concrete. Both the density and the total porosity showed a linear Build. Mater. 111, 618–624.
relationship with compressive strength, which indicated that the Lynn, C.J., Obe, R.K.D., Ghataora, G.S., 2016. Municipal incinerated bottom ash
density and porosity played key roles in affecting the compressive characteristics and potential for use as aggregate in concrete. Construct. Build.
Mater. 127, 504–517.
strength. Chuang, K.H., Lu, C.H., Chen, J.C., Wey, M.Y., 2018. Reuse of bottom ash and fly ash
(4) The thermal conductivity decreased with the increase of IBA from mechanical-bed and fluidized-bed municipal incinerators in manufacturing
replacement due to the porous structure of IBA. The proposed lightweight aggregates. Ceram. Int. 44 (11), 12691–12696.
Brown, H.J., 2006. Pervious concrete research compilation: past, present and future.
pervious concrete with low thermal conductivity offered an
RMC Res. Found.. 1, 1–44.
alternative option to be used as a thermal insulating partition Huang, B., Wu, H., Shu, X., Burdette, E.G., 2010. Laboratory evaluation of permeability
wall. and strength of polymer-modified pervious concrete. Construct. Build. Mater. 24,
(5) The presence of the high amount of aggregates and internal 818–823.
Ćosić, K., Korat, L., Ducman, V., Netinger, I., 2015. Influence of aggregate type and size
curing of IBA compensated the autogenous shrinkage of pervious on properties of pervious concrete. Construction and Building Materials 78, 69–76.
concrete. The incorporated IBA in pervious concrete had limited Putman, B.J., Neptune, A.I., 2011. Comparison of test specimen preparation techniques
effect on the expansion level due to the high porosity and thin for pervious concrete pavements. Construct. Build. Mater. 25, 3480–3485.
P.D. Tennis, M.L. Leming, D.J. Akers, Pervious Concrete pavements, Portland Cement
coating around IBA. Hence, it was possible to use as aggregates in Association Skokie, IL 2004.
pervious concrete without considering the crack risk from the Kim, H.-.K., Lee, H.-.K., 2010. Influence of cement flow and aggregate type on the
expansion of IBA. mechanical and acoustic characteristics of porous concrete. Appli. Acoust. 71,
607–615.
Lu, J.-.X., Yan, X., He, P., Poon, C.S., 2019a. Sustainable design of pervious concrete
CRediT authorship contribution statement using waste glass and recycled concrete aggregate. J. Clean. Prod. 234, 1102–1112.
Qin, Y., 2015. A review on the development of cool pavements to mitigate urban heat
island effect. Renew. and Sustainab. Energ. Rev. 52, 445–459.
Peiliang Shen: Conceptualization, Methodology, Investigation, Li, H., Harvey, J.T., Holland, T., Kayhanian, M., 2013. The use of reflective and
Formal analysis, Writing - original draft, Validation. Haibing Zheng: permeable pavements as a potential practice for heat island mitigation and
Methodology, Formal analysis, Writing - review & editing. Jianxin Lu: stormwater management. Environ. Res. Lett. 8, 015023.
Bhutta, M.A.R., Tsuruta, K., Mirza, J., 2012. Evaluation of high-performance porous
Methodology, Resources, Formal analysis, Writing - review & editing.
concrete properties. Construct. Build. Mater. 31, 67–73.
Chi Sun Poon: Conceptualization, Writing - review & editing, Kuo, W.-.T., Liu, C.-.C., Su, D.-.S., 2013. Use of washed municipal solid waste incinerator
bottom ash in pervious concrete. Cem. and Concr. Composit. 37, 328–335.

10
P. Shen et al. Resources, Conservation & Recycling 174 (2021) 105736

Yu, R., Spiesz, P., Brouwers, H.J.H., 2014. Mix design and properties assessment of ultra- Demirboga, R., 2007. Thermal conductivity and compressive strength of concrete
high performance fibre reinforced concrete (uhpfrc). Cem. Concr. Res. 56, 29–39. incorporation with mineral admixtures. Build. Environ. 42, 2467–2471.
Shen, Peiliang, Zheng, Haibing, Lu, J.-.X., 2021. Chi sun poon, conceptual design and Shen, P., Lu, L., He, Y., Rao, M., Fu, Z., Wang, F., Hu, S., 2018. Experimental
performance evaluation of high strength pervious concrete. Construct. Build. Mater. investigation on the autogenous shrinkage of steam cured ultra-high performance
269, 121342. concrete. Construct. Build. Mater. 162, 512–522.
Lu, J.-.X., Zheng, H., Yang, S., He, P., Poon, C.S., 2019b. Co-utilization of waste glass Xuan, D., Tang, P., Poon, C.S., 2018a. Effect of casting methods and scms on properties of
cullet and glass powder in precast concrete products. Construct. Build. Mater. 223, mortars prepared with fine msw incineration bottom ash. Construct. Build. Mater.
210–220. 167, 890–898.
Yang, S., Ling, T.-.C., Cui, H., Poon, C.S., 2019. Influence of particle size of glass Sreenivasulu, C., Ramakrishnaiah, A., Jawahar, J.G., 2015. Mechanical properties of
aggregates on the high temperature properties of dry-mix concrete blocks. Construct. geopolymer concrete using granite slurry as sand replacement. Int. J. Adva. Eng.
Build. Mater. 209, 522–531. Technol. 8, 83–91.
Lu, J.-.X., Poon, C.S., 2019. Recycling of waste glass in construction materials. New Hasanzadeh, B., Liu, F., Sun, Z., 2016. Monitoring hydration of uhpc and conventional
Trends in Eco-Efficient and Recycled Concrete. Elsevier, pp. 153–167. paste by quantitative analysis on raman patterns. Construct. Build. Mater. 114,
BS 6717. Precast, Unreinforced Concrete Paving Blocks-Requirements and Test methods, 208–214.
British Standards Institution; 2001. Justs, J., Wyrzykowski, M., Bajare, D., Lura, P., 2015. Internal curing by superabsorbent
ASTM C39, Standard Test Method For Compressive Strength of Cylindrical Concrete polymers in ultra-high performance concrete. Cem. Concr. Res. 76, 82–90.
Specimens, (2012). Wang, D., Shi, C., Wu, Z., Xiao, J., Huang, Z., Fang, Z., 2015. A review on ultra high
Neithalath, N., Weiss, J., Olek, J., 2006. Characterizing enhanced porosity concrete using performance concrete: part II. Hydrat., Microstruct Propert., Construct. Build. Mater.
electrical impedance to predict acoustic and hydraulic performance. Cem. Concr. 96, 368–377.
Res. 36, 2074–2085. Xuan, D., Tang, P., Poon, C.S., 2018b. Limitations and quality upgrading techniques for
ASTM C 1754, Standard Test Method For Density and Void Content of Hardened Pervious utilization of msw incineration bottom ash in engineering applications – a review.
Concrete, (2012). Construct. Build. Mater. 190, 1091–1102.
JIS A 5371, 2016. Precast Unreinforced Concrete Products. Japanese Standards Ma, X., Liu, J., Shi, C., 2019. A review on the use of lwa as an internal curing agent of
Association. high performance cement-based materials. Construct. Build. Mater. 218, 385–393.
GB/T 50082-2009, 2009. Standard For Test Methods of Long-term Performance and Shen, P., Zheng, H., Xuan, D., Lu, J.-.X., Poon, C.S., 2020b. Feasible use of municipal
Durability of Ordinary Concrete. China Architecture and Building Press. solid waste incineration bottom ash in ultra-high performance concrete. Cem. Concr.
Martin III, W.D., Kaye, N.B., Putman, B.J., 2014. Impact of vertical porosity distribution Composit. 114, 103814.
on the permeability of pervious concrete. Construct. Build. Mater. 59, 78–84. Espinoza-Hijazin, G., Lopez, M., 2011. Extending internal curing to concrete mixtures
ASTM C 1260, 2014. Standard Test Method For Potential Alkali Reactivity of Aggregates with w/c higher than 0.42. Construct. Build. Mater. 25, 1236–1242.
(Mortar-Bar Method). ASTM International. Halleck, P.M., Bentz, D.P., 2006. Water movement during internal curing: direct
GB/T50082-2009, 2009. Standard For Test Methods of Long-term Performance and observation using x-ray microtomography. Concr. Int. Desi. Construction 28, 39–50.
Durability of Ordinary Concrete. China Architecture and Building Press. Shen, P., Chen, W., Lu, L., Geng, H., Li, Q., 2017. Effect of aggregate exposing and curing
P. Shen, L. Lu, Y. He, F. Wang, J. Lu, H. Zheng, S. Hu, Investigation On Expansion Effect agent on the performance of exposed aggregate concrete. Construct. Build. Mater.
of the Expansive Agents in Ultra-High Performance concrete, Cement and Concrete 156, 675–683.
Composites, (2019) 103425. Qiao, X.C., Tyrer, M., Poon, C.S., Cheeseman, C.R., 2008. Novel cementitious materials
Kevern, J.T., Nowasell, Q.C., 2018. Internal curing of pervious concrete using produced from incinerator bottom ash. Res., Conserv. Recycl. 52, 496–510.
lightweight aggregates. Construct. Build. Mater. 161, 229–235. Polettini, A., Pomi, R., Carcani, G., 2005. The effect of na and ca salts on mswi bottom
Shen, P., Lu, L., Wang, F., He, Y., Hu, S., Lu, J., Zheng, H., 2020a. Water desorption ash activation for reuse as a pozzolanic admixture. Res., Conserv. Recycl. 43,
characteristics of saturated lightweight fine aggregate in ultra-high performance 403–418.
concrete. Cem. Concr. Composit. 106, 103456. Lee, G., Ling, T.-.C., Wong, Y.-.L., Poon, C.-.S., 2011. Effects of crushed glass cullet sizes,
Bentz, D.P., Weiss, W.J., 2011. Internal curing: a 2010 state-of-the-art review. NIST. casting methods and pozzolanic materials on asr of concrete blocks. Construct. Build.
Interagency/Internal Rep. (NISTIR.) 7765. -. Mater. 25, 2611–2618.
to JIS A 5371, 2016. Precast Unreinforced Concrete Products. Japanese Standards Du, H., Tan, K.H., 2013. Use of waste glass as sand in mortar: part ii–alkali–silica reaction
Association. and mitigation methods. Cem. Concr. Composit. 35, 118–126.
Tang, P., Xuan, D., Poon, C.S., Tsang, D.C.W., 2019. Valorization of concrete slurry waste Verbinnen, B., Billen, P., Van Caneghem, J., Vandecasteele, C., 2017. Recycling of mswi
(csw) and fine incineration bottom ash (iba) into cold bonded lightweight aggregates bottom ash: a review of chemical barriers, engineering applications and treatment
(cblas): feasibility and influence of binder types. J. Haz. Mat. 368, 689–697. technologies. Waste Bioma. Valorizat. 8, 1453–1466.
Cho, W., Kwon, S., Choi, J., 2009. The thermal conductivity for granite with various
water contents. Eng. Geol. 107, 167–171.

11