Professional Documents
Culture Documents
a r t i c l e i n f o a b s t r a c t
Article history: U.S. municipal solid waste incinerator (MSWI) ash is generally disposed of in landfills, due to the absence
Received 20 November 2018 of approved avenues to repurpose the material through recycling. This work investigates the utilization
Received in revised form of MSWI bottom ash sands in concrete that have been reclaimed from a recycling plant in Central
24 April 2019
Pennsylvania. Physical and chemical properties of the reclaimed sands are examined for suitability as a
Accepted 29 April 2019
Available online 3 May 2019
partial replacement for fine aggregates in concrete. The performance of raw incinerator ash vs. Reclaimed
sands in concrete are compared through microstructure SEM analysis, shrinkage, compression strength,
and hydrogen gas production due to metallic aluminum present in incinerator ash. It is found that
Keywords:
Solid waste
reclaimed sands satisfy most ASTM requirements to be used as a lightweight concrete aggregate. Despite
MSWI this, it is currently not recommend to use reclaimed sands in concrete due to the formation of empty
Waste-to-energy holes in the cement matrix microstructure due to hydrogen gas production. These holes cause cracks in
Recycled the concrete which leads to reduced structural capacity of the material. It is also not recommended that
Ash raw ash be used in concrete due to significant amounts of drying shrinkage, leaching concerns, and the
unclean nature of the ash.
© 2019 Elsevier Ltd. All rights reserved.
https://doi.org/10.1016/j.jclepro.2019.04.387
0959-6526/© 2019 Elsevier Ltd. All rights reserved.
G. Mathews IV et al. / Journal of Cleaner Production 229 (2019) 838e849 839
Table 1 ash in concrete and associated issues, and the studies state that
Reclaimed sand concrete leachates. MSWI ash is unusable in concrete without being refined to remove/
Heavy Reclaimed Sand EPA Non-Hazard DEP Permit (WMGM061) pretreated for significant amounts of ultra-fine particles (fly ash),
Metal Concrete (mg/L)a Limits (mg/L) Option 2 Limits (mg/L) salts, heavy metals, and organic components. To this point, studies
Aluminum 0.5667 NA 5 on using MSWI ash in concrete have not explored utilizing an ash
(Al) product that is produced through industrial post processing after
Arsenic 0.0225 5 0.25 incineration. In addition, current studies do not explore the
(As)
behavior of raw ash from a MSWI incinerator vs. the same raw ash
Barium 1.3933 100 50
(Ba) treated at a post-processing facility.
Cadmium 0.0016 1 0.125 This works explore the performance of concrete that in-
(Cd) corporates reclaimed sands that are refined from MSWI ash. The
Chromium 0.0262 5 2.5
MSWI ash is refined at the Pure Recovery Group recycling plant
(Cr)
Lead (Pb) 0.1012 5 0.375
(Pure Recovery, 2018) in York, Pennsylvania. These reclaimed sands
Mercury 0.0002 0.2 0.05 are classified as a 100% post-consumer material and have been
(Hg) approved for construction use via DEP permit WMGM061. There is
Selenium 0.0190 1 1 a significant potential to utilize reclaimed sands in concrete due to
(Se)
sustainability initiatives. One possible initiative is through Leader-
Silver (Ag) 0.0049 5 2.5
ship in Engineering and Environmental Design (LEED) (LEED
a
Testing results obtained from (Sweeney, 2016a).
USGBC, 2018) credits, if the material meets LEED regulations. In
addition, in 2018 it cost $34.80-$68.48/ton (EREF, 2018) to dispose
of MSWI ash in a landfill which provides an opportunity to subsi-
measured leachates (i.e. more metals, organics). It can be seen that dize the material. To the best of the authors’ knowledge, the Pure
the DEP limits are much more severe than the TCLP requirements Recovery group is one of only a few recycling plants in the U.S.
for landfill disposal. In this work, MSWI ash must conform to the refining MSWI ash for use in construction materials. The goal of this
leaching requirements of DEP permit WMGM061, which authorizes work is to assess the feasibility of using reclaimed sands in con-
the processing and beneficial use of MSWI aggregates for con- crete. This assessment is done by evaluating reclaimed sands per
struction use in Pennsylvania. This permit provides 3 sets of fine aggregate material standards, and evaluating concrete mixed
leaching limits based on the specific application of MSWI ash in a with reclaimed sands though microstructure analysis, compression
construction material, and the material in this work must meet strength, and shrinkage. The process to produce reclaimed sands
option 2 for encapsulated products (i.e. recycled aggregates and material properties are discussed in the next section.
encapsulated in concrete). From Table 1, it can be seen that leach-
ates from concrete incorporating reclaimed sands meet the DEP 2. Materials
permit limits. It should be noted that a Synthetic Precipitation
Leaching Procedure (SPLP) (USEPA, 1994) test was used to obtain 2.1. Reclaimed sand production
the reclaimed sand leachates, and the DEP permit accepts results
from both the TCLP and SPLP tests. The SPLP test was used in this The reclaimed sands in this work are refined from raw inciner-
work since it is more representative of actual exposure condition of ator ash obtained from the York County Solid Waste Authority
it-situ concrete. The TCLP test exposes materials to a solution of (YCSWA) (YCSWA, 2018) incinerator plant. The YCSWA plant is a
acetic acid before measuring leachate amounts, which is supposed WTE mass burn facility that processed 431,412 tons of waste and
to represent the harsh exposure conditions of a landfill. The SPLP produced approximately 141,176 tons of incinerator ash by weight
test exposes materials to a solution that simulates acid rain expo- in 2016 (Sweeney, 2016a unpublished report; YCSWA, 2016).
sure, which is representative of normal weather conditions. Separate streams are produced for bottom ash and fly ash at the
The DEP limits are not adequate to mitigate strength reduction YCSWA plant, and for landfill disposal the streams are combined
in construction materials due to the presence of aluminum. In proportionally as needed to satisfy EPA non-hazard leaching limits.
concrete, the aluminum in MSWI ash produces hydrogen gas In this work, the combined bottom ash þ fly ash stream is sent to
(Nithiya et al., 2018) that pushes the material outwards leading to the Pure Recovery Group recycling plant where the ash is refined
volume changes, shrinkage, and cracking. This has a deleterious using a patented process from the Pure Recovery Group
effect on the shrinkage and compression strength of concrete (Kim (Vandemierden, 2014). At the Pure Recovery Group, MSWI ash
et al., 2015). In the particular case of aerated concrete, hydrogen gas particles are subjected to a density separation process and treat-
production from MSWI bottom ash can have a beneficial effect by ment procedures which are designed to remove the majority of
reducing the time needed for gas foaming (Li et al., 2018). In a large ferrous and non-ferrous impurities. The density separation
comparison of several MSWI ash concrete studies (Lynn et al., process utilizes a drum with internal flights and an internal water
2016), shows that concrete strength consistently increases when wash system (Sweeney, 2016b unpublished report) to separate
treatment methods are applied that address hydrogen gas pro- material into separate particle fractions by density. Ferrous com-
duction due to aluminum. The vast majority of studies on MSWI ash ponents removed using a drum magnetic system, while non-
utilization have been focused on low cementitious construction ferrous metals are removed using an Eddy Current Separator. This
materials where hydrogen gas production from aluminum is not a refinement process is outlined in Fig. 1. At the end of the process
concern; such as road subbase (Lynn et al., 2017; Van Praagh et al., gravel and sand size fraction materials are produced, and this work
2018), embankment fill (Zhang et al., 2016), and masonry block is focused on the sand size fractions. This refinement process is
(Holmes et al., 2016; Lynn et al., 2017). The utilization of MSWI ash outlined in Figs. 1, and Fig. 2 shows the density separator and Eddy
in these applications has been heavily investigated in the U.S., Current system equipment.
Europe, and Asia as discussed in (An et al., 2014; Dou et al., 2017; The reclaimed sands used in this study are sand size aggregates
Song et al., 2015) and referenced within. Studies are provided by taken from the recycled aggregate product bin (Fig. 1(b)) at Pure
(An et al., 2014, 2017; Kim et al., 2015; Minane et al., 2017; Rubner Recovery. Material produced from this combination of systems has
et al., 2008; Verbinnen et al., 2017) that investigate the use of MSWI been evaluated over several years in (Sweeney, 2016a unpublished
840 G. Mathews IV et al. / Journal of Cleaner Production 229 (2019) 838e849
Fig. 1. Pure Recovery recycling process: (a) procedure A; (b) procedure B. Modified from (Vandemierden, 2014) with permission from Jay Berman, Pure Recovery Group.
Fig. 2. Recycling equipment: (a) Liquid Density Separator (LDS); (b) Drum Magnetic Separator (DMS). Photos used with permission from Jay Berman, Pure Recovery Group.
report), which ultimately lead to DEP approval of the reclaimed Recovery process removing large metal pieces from the ash, that
sands. Raw MSWI ash undergoes a significant filtering process significant improvement is needed in the process to remove small
when refined through Pure Recovery's recycling procedure, and particle fractions of metal. These particles tend to be inherently
this can be observed by noting the difference in appearance of raw weaker than the rest of the material that make up reclaimed sands.
incinerator ash vs. Reclaimed sands shown in Fig. 3. Raw incinerator Currently it is not practical to remove these particles via recycling
ash contains large metal pieces and organics that are filtered out practices. The full XRF analysis results will be discussed in section
through the reclaimed sand refinement process, and it can be seen 2.2. The friability of both raw ash and reclaimed sands depends on
from Fig. 3 that reclaimed sands are a much cleaner material than whether the small metal particles in the material are removed. The
raw ash. In addition, the raw ash has significantly more fly ash in it metal particles (up to 39.5% of the material) are very friable while
than reclaimed sands. the rest of the material is not friable. This implies that the material
An X-ray fluorescence (XRF) analysis on reclaimed sands vs. Raw will have friability issues if the metal particles aren't removed.
ash shows that both materials are composed of approximately Pretreatment methods are available to reduce hydrogen gas
39.50% metal compounds. This suggests that despite the Pure production from the aluminum in reclaimed sands as discussed in
G. Mathews IV et al. / Journal of Cleaner Production 229 (2019) 838e849 841
Table 2
Aggregate physical properties.
Bulk Density kg/m3 (lb/ft3) Specific Gravity Absorption Capacity (%) Void Content (%)
Table 3
XRF analysis (by concentration).
Gardner Color Value In this work reclaimed sands are used as a partial replacement
Sample 1 5
for sandstone in concrete at replacement ratios of 90/10, 80/20, and
Sample 2 5 70/30, with the second number denoting the percentage of
Sample 3 6 reclaimed sands in the mix. Corresponding mix designs are pro-
vided in Table 5, and all reclaimed sand proportions are calculated
based on a control mix with a design 28 day compression
strength ¼ 24 MPa (3500 psi). These replacement ratios were
report), provides a detailed SPLP analysis on the organics of selected to represent a practical utilization of reclaimed sands in
reclaimed sands, and it is found that organic impurities are not a concrete mixes, since recycled materials tend to be structurally
significant concern. The loss on ignition test evaluates the amount weaker than natural aggregates.
of volatile compounds in fine aggregate by heating material sam-
ples up to 950 ± 50 # C in a muffle furnace and measuring how much
2.5. Testing of reclaimed sand concrete
material was burned off. Per ASTM C330 the loss on ignition of fine
aggregate samples must be less than 5%. Preliminary loss of ignition
A scanning electron microcopy (SEM) analysis was performed to
testing on the reclaimed sands indicates that more than 5% of the
assess the effect that raw ash and reclaimed sands have on the
material is burned off during the test. However, this test requires
microstructure of concrete. It is expected that hydrogen gas effects
further investigation with respect to the evolving chemical nature
and shrinkage cracks will be prevalent in the reclaimed sand and
of reclaimed sands during cement hydration. Soundness and
raw ash samples. SEM images were obtained for the control mix
Freeze-Thaw testing are used to assess the durability of concrete
and 80/20 reclaimed sand and raw ash samples. Samples were
when subjected to simulated environmental conditions through
made by coating concrete in epoxy and then polishing the surface
accelerated aging tests, i.e. rapid freezing and thawing cycles.
using grinding paper and diamond polishing paste. The surface of
PennDot is currently updating publication 408 to specific Freeze-
all samples were coated in iridium to negate any charging effects
Thaw testing over sulfate soundness, and freeze-thaw testing will
from materials in the concrete. Images are provided at 500x, 2500x,
be performed on concrete incorporating reclaimed sands in a future
and 5000x magnification in Fig. 7 e 9. Boxes drawn on the SEM
study.
images indicate locations that are zoomed in on at larger
Table 5
Concrete mix proportions a.
Concrete Mix Design Coarse Aggregate Fine Aggregate Reclaimed Sands Raw Ash Cement Water
kg/m3 (lb/yd3) kg/m3 (lb/yd3) kg/m3 (lb/yd3) kg/m3 (lb/yd3) kg/m3 (lb/yd3) kg/m3 (lb/yd3)
Fig. 8. 80/20 raw ash SEM: (a) 500x; (b) 2500x; (c) 5000x
Fig. 9. 80/20 reclaimed sands SEM: (a) 500x; (b) 2500x; (c) 5000x
magnifications. An Energy-Dispersive X-ray Spectroscopy (EDXS) magnification images, and a more representative composition of
was performed on all SEM samples to obtain the chemical the can be obtained by analyzing serval different spots on an image
composition of the concrete, and the results are shown in Table 6. It through EDXS.
should be noted that the EDXS analysis was only run on the 250X Shrinkage is one of the main concerns in using MSWI bottom
G. Mathews IV et al. / Journal of Cleaner Production 229 (2019) 838e849 845
Table 6
Concrete composition (EDXS) (by weight).
Compound Control Mix (%) 80/20 Reclaimed Sands (%) 80/20 Raw Ash (%)
Table 7
Mortar mix proportions.
Concrete Mix Design Fine Aggregate g Reclaimed Sands g Raw Ash g Cement g Water g
ash in concrete due to the material's high absorption capacity. The voids as shown in Fig. 12(a)e(b). These voids are created due to
drying shrinkage of reclaimed sand mixes are investigated in this trapped hydrogen gas in the concrete, and in general form towards
study through 2.54 cm (1 in) x 2.54 (1 in) x 29.21 cm (11.5 in) the top of the cylinders in this work. In prior studies exploring the
mortar bars, and the corresponding mix designs are provided in use of MSWI ash in concrete (An et al., 2014, 2017; Kim et al., 2015;
Table 7. A 50/50 mix of sandstone to reclaimed sands and a control Minane et al., 2017; Rubner et al., 2008; Verbinnen et al., 2017), the
mix were used for the shrinkage tests. In addition, 50/50 raw behavior of trapped hydrogen gas with respect to concrete height is
incinerator ash mortar samples are tested to evaluate the effects of not investigated. In (An et al., 2014), surface voids in cylinders made
the reclaimed sand refinement process on shrinkage. In general a with MSWI bottom ash are evenly distributed along the height of
50/50 mix would not be practical for recycled materials, but the mix the samples. Other studies referenced in this work do not provide
is used in this study to assess shrinkage under a worst case con- enough information to make this assessment. It should be noted
dition. Four samples were prepared for each mortar mix per ASTM that MSWI bottom ash in (An et al., 2014) was obtained from a
596 and cured in lime water. Samples were tested at 1, 7, and 28 refuse-derived fuel (RDF) WTE facility, which processes trash
days, and the shrinkage results are shown in Fig. 10. differently than the YCSWA mass-burn incinerator. In addition, the
The 28-day compression strength of reclaimed sand, raw ash, material in (An et al., 2014) does not undergo a secondary refine-
and control mix concrete samples has been determined for each ment process. For the materials in this study, it is suspected that the
mix design in Table 4. Compression testing was performed per behavior of trapped hydrogen gas varies with respect to height, and
ASTM C39. Unbounded metal caps with an elastomeric insert this variation will cause a corresponding change in concrete den-
meeting the requirements of ASTM C1231 were used to cap the sity. This variation in density has significant implications on pour-
cylinders. Three 10.16 cm (4 in) x 20.32 cm (8 in) cylinders were ing concrete containing reclaimed sands, as it is generally assumed
tested for each mix design and the results are provided in Fig. 11. that poured concrete maintains a constant density. To better un-
When making concrete cylinders with reclaimed sands, it was derstand the effect of hydrogen gas production in reclaimed sand
observed that the surface of the cylinders contained pockets of concrete, 10.16 cm (4 in) x 20.32 cm (8 in) cylinders were made and
tested for density with respect to sample height. The cylinders were
Fig. 10. Shrinkage of Reclaimed Sands vs. Raw Ash. Fig. 11. Compression strength of Reclaimed Sand and Raw Ash concrete.
846 G. Mathews IV et al. / Journal of Cleaner Production 229 (2019) 838e849
Fig. 12. Reclaimed Sand concrete cylinders: (a) Mushroom top effects; (b) Trapped hydrogen gas voids; (c) Cylinder top after mushroom top removal.
cut into five equal thickness disks to measure density with respect more pronounced on the reclaimed sand samples, despite the raw
to height, and a typical disk sample is shown in Fig. 13(a). The ash actually producing more hydrogen gas (see Fig. 6). In general,
density of the disks were found utilizing standard specific gravity the hydrogen gas effects are much more damaging to the cement
testing procedures similar to those outlined in ASTM C128, and the matrix than shrinkage cracks. The EDXS analysis in Table 6 shows
testing setup is shown in Fig. 13(b). In this study the testing setup is that the composition of the concrete does not changes significantly
used to measure the density of disk samples in air and in a vacuum, between the control and 80/20 mixes. In particular, the change in
which is needed to eliminate the contribution of internal air voids silica content is negligible between the mixes, which would imply
to disk density. Three control mix and 80/20 reclaimed sand con- that ASR will not be an issue.
crete samples have been tested and the results are provided in
Fig. 14.
3.2. Drying shrinkage
Fig. 14. Disk density vs. height: (a) Vacuum samples; (b) Non-vacuum samples.
848 G. Mathews IV et al. / Journal of Cleaner Production 229 (2019) 838e849
shrinkage in raw ash concrete is problematic and potentially not the Pure Recovery Group, L.P. on this project. The authors would
practical. also like to thank Sean Sweeny and Alison L'Hommedieu D'Airo
from Barton & Loguidice, D.P.C. for their continued support and
3.4. Disk test collaboration.