Water">
82 801 1 PB
82 801 1 PB
82 801 1 PB
Abstract
In this study the removal of control parameters and basic pollutants of domestic wastewater
in artificial wetlands of subsurface flow was evaluated using two different support media,
round-ridge gravel and sedimentary-gap gravel, with vegetation Sagittaria latifolia
(swallow's tail), with retention times of 4.8 days and 4.9 days respectively. Three wetlands
with round-ridge vegetation and gravel (HACC), three wetlands with vegetation and
sedimentary-gap gravel (HABC); three control wetlands without vegetation and gravel of
round ridge (HACS) and three without vegetation and sedimentary gap gravel (HABS). For
the support media the variables were determined as apparent density, real, porosity and
electrical conductivity (N=28). At the beginning of the operation of the artificial wetlands,
the round ridge support medium presented an apparent density of 1390.79 ± 54.36 Kg m-3
(mean ± SD.), The actual density of 2626.01 ± 75.43 Kg m-3, porosity 51.87 ± 3.90% and an
Electrical Conductivity (CE) 125.46 ± 11.81 dS m-1, the sedimentary gap support medium
showed an apparent density of 1415.63 ± 43.94 Kg m-3 (mean ± SD), actual density of
2678.16 ± 36.67 Kg m-3 , porosity 52.86 ± 1.56% and an EC of 94.13 ± 3.58 dS m-1. At the
end of the operation the support medium round edge proved to be efficient with only a loss
of 5% porosity, while the gap presented a loss of 23% of this. In terms of water quality, 160
composite samples were analyzed for the four wetland treatments and 40 simple samples in
the feed distribution tank (N=40). After the first year of operation it is seen that he HACC is
the most efficient, the removal reached was 96.85% for SST, 95.85% for BOD5, 96.78% for
NT and 96.79% for PT. There is a high potential for these support means to be implemented
in HAs in southeastern Mexico, reducing construction and maintenance costs, since the
means of support currently used are not native to the region, they present erosion and attrition,
they do not allow a adequate growth of the biofilm and vegetation and, as a consequence,
have low efficiencies.
Resumo
O objectivo deste estudo foi avaliar a remoção dos parâmetros de controlo de base e
contaminantes das águas residuais domésticas com fluxo alagados construídos
subsuperficial, para o qual foram utilizados dois meios de suporte diferentes: borda redonda
cascalho e brita lacuna sedimentar vegetação Sagittaria latifolia (cauda de andorinha) e com
tempos de retenção de 4,8 dias e 4,9 dias, respectivamente. Especificamente, três vegetação
húmida e borda redonda cascalho (HACC), três vegetação e lacuna cascalho sedimentar
(HABC) e três controlos zonas húmidas sem vegetação e borda redonda cascalho (HACS) e
três sem vegetação e alocados Cascalho de Brechas Sedimentares (HABS). Para o suporte,
as variáveis foram determinadas como densidade aparente e real, bem como porosidade e
condutividade elétrica (N = 28). No início da operação de zonas húmidas artificiais, os meios
de suporte cantando rodada forneceu os seguintes valores: densidade a granel de 1390,79 ±
54,36 kg m-3 (média ± SD), uma densidade efectiva de 2626,01 ± 75,43 kg m-3, porosidade
51,87 ± 3,90% e condutividade elétrica (CE) 125,46 ± 11,81 dS m-1. O apoio médio hiato
sedimentar mostrou uma densidade a granel de 1415,63 ± 43,94 kg m-3 (média ± SD), uma
densidade efectiva de 2678,16 ± 36,67 kg m-3, porosidade 52,86 ± 1,56% e CE 94,13 ± 3.58
dS m-1 . Ao final da operação, o suporte de borda arredondada mostrou-se eficiente com
apenas 5% de perda de porosidade, enquanto o gap obteve uma perda de 23%. Em termos de
qualidade da água, 160 amostras compostas foram analisadas para os quatro tratamentos de
terras húmidas e 40 amostras simples no tanque de distribuição de alimentos (N = 40). Após
o primeiro ano de operação, observou-se que o HACC é o mais eficiente porque a remoção
alcançado foi 96,85% de TSS, DBO5 95,85%, 96,78% e 96,79% para o NT para PT. O
tratamento com HABC, por outro lado, alcançou eficiências de remoção de 95,52% para
SST, 95,02% para DBO5, 95,45% para NT e 95,36% para PT. Portanto, considera-se que
existe um grande potencial para estes meios de apoio pode ser implementado em HA no
sudeste do México, que os custos de construção e manutenção seria reduzido, como
Introduction
Wastewater is a problem that affects many terrestrial and aquatic ecosystems when
they are deposited without any prior treatment, which directly affects the self-purification of
bodies of water. Therefore, the need to implement technologies that are not aggressive to the
environment and that are economically viable (National Water Commission [Conagua],
2016a, Vymazal, 2014) has been raised.
In this regard, artificial wetlands (HA) are an excellent alternative in the treatment of
domestic wastewater, as natural technology results in low operating costs and easy
maintenance, with minimum personnel requirements (Crites, Middlebrooks and Bastian,
2014) . In these systems, the treatment occurs by a natural process called phytodepuration,
which consists of developing a macrophytic plant culture on a support medium within a
control volume, where various physical, chemical and biological reactions occur through
which the wastewater is progressively purified (Delgadillo, Camacho, Pérez and Andrade,
2010), or also within the wetlands where the removal of contaminants is achieved by
sedimentation, absorption and bacterial metabolism (Llagas and Gómez, 2006).
In HA, different plant species have been used, which are the basis of the process,
since they are capable of degrading, absorbing and assimilating organic matter and nutrients
in their tissues, in addition to the support medium providing an extensive surface where
bacterial growth is facilitated, which facilitates the retention of suspended solids, since they
also function as a filtering medium (Upadhyay, Bankoti and Rai, 2016).
𝑑𝑒𝑛𝑠𝑖𝑑𝑎𝑑 𝑎𝑝𝑎𝑟𝑒𝑛𝑡𝑒
𝑃𝑜𝑟𝑜𝑠𝑖𝑑𝑎𝑑 (%) = (1 − 𝑋 100)
𝐷𝑒𝑛𝑠𝑖𝑑𝑎𝑑 𝑟𝑒𝑎𝑙
Nomenclatura: HABS es HA con medio de soporte de brecha sedimentaria sin vegetación (control); HABC es HA con
medio de soporte de brecha sedimentaria con vegetación (tratamiento); HACS es HA con medio de soporte de canto redondo
sin vegetación (control), y HACC es HA con medio de soporte de canto redondo con vegetación (tratamiento).
Experimental design
For this research, a random design of a factor was required to analyze the treatment
systems (support medium with vegetation) and their controls (support medium without
vegetation). For each of these was run in triplicate during the four-month period (August-
December 2017).
Statistic analysis
In the support media, the variables of apparent density and real density followed a
normal and homocedastic behavior, for which an ANOVA was determined. Regarding the
porosity and electrical conductivity, by not presenting the attributes of normality, we
proceeded to determine its nonparametric analysis by performing the Kruskal-Wallis test and
the contrast of Mann-Whitney medians.
Likewise, and because for water variables the only parameter that fulfills the
postulates of normality and homoscedasticity is temperature, a one-way ANOVA was
performed, while the variables of quantitative responses of BOD5, PT, NT, SST, color,
turbidity, EC and pH in the different types of treatments present in HA did not follow the
normal distribution, so they were applied the non-parametric statistical analysis, the Kruskal-
Wallis test and the contrast of Mann-Whitney medians , for which the statistical package was
used Statghapics 16MR.
Tabla 1. Valores promedio y desviación estándar de los parámetros físicos evaluados en los
medios de soporte (N = 28)
Canto redondo (can) Brecha sedimentaria (bre)
Parámetro
M DE M DE
Densidad real (kg m-3) 2626.01 75.43 2678.16 36.67
Densidad aparente (kg m-3) 1390.79 54.36 1415.63 43.94
Porosidad (ɳ) (%) 51.87 3.90 52.86 1.56
Conductividad E. (dS m-1) 125.46 11.81 94.13 3.58
Fuente: Elaboración propia
1250 A A
1200 A A
1150
1100
1050
1000
brec-C brec-S can-C can-S
Tratamiento
Nota: El tamaño de muestras compuesta es N = 28 para cada tipo de medio. Letras desiguales indican diferencias
estadísticamente significativas.
Fuente: Elaboración propia
Figura 5. Valores promedio (± DE) de la densidad real de los medios de soporte en cada
Medias y 95.0% de Fisher LSD
tratamiento
2900
A
Densidad Real (kg m-3)
2800 A
2700 B B
2600
2500
brec-C brec-S can-C can-S
Tratamiento
.
Nota: El tamaño de muestras compuesta es N = 28 para cada tipo de medio. Letras desiguales indican diferencias
estadísticamente significativas.
Fuente: Elaboración propia
52
AB
49 B
Porosidad (%)
46 A
A
43
40
37
brec-C brec-S can-C can-S
Tratamiento
Nota: El tamaño de muestras compuesta es N = 28 para cada tipo de medio. Letras desiguales indican diferencias
estadísticamente significativas.
Fuente: Elaboración propia
Electric conductivity
The Kruskal-Wallis test evaluates the hypothesis that the electrical conductivity
medians (dS m-1) within each of the four treatment levels are equal, since the p-value is
greater than or equal to 0.05, so that there is a statistically significant difference between
medians with a 95% confidence level. The gravel with median of greater electrical
conductivity was the round ridge without vegetation (can-S) 166.00 ± 52.25 dS m-1, while
the sedimentary gap without vegetation (bre-s) was the lowest 66.0 ± 13.47 dS m-1 (figure
7).
300
Conductividad Eléctrica (dS m-1)
250
A A
200 A
150
B
100
50
0
brec-C brec-S can-C can-S
Tratamiento
Nota: El tamaño de muestras compuesta es N = 28 para cada tipo de medio. Letras desiguales indican diferencias
estadísticamente significativas.
Fuente: Elaboración propia
Parámetro M DE
Turbiedad (NTU) 93.98 22.89
Color (UC) 1282.39 254.09
pH 8.35 0.18
CE (µ S cm-1) 1381.11 140.19
Temperatura (°C) 26.84 1.08
SST (mg L-1) 250.10 59.91
DBO5 (mg L-1) 289.91 65.51
NT (mg L-1) 146.6 23.50
PT (mg L-1) 5.79 1.31
Fuente: Elaboración propia
Turbiedad
The Kruskal-Wallis test evaluates the hypothesis that turbidity medians (NTU) are
equal within each of the five treatment levels. Since the p-value is less than 0.05, there is a
statistically significant difference between the medians with a 95% confidence level (Figure
8). Among the wetlands, the treatment with the lowest median ± SD was HACC with 2.89 ±
2.46 NTU, followed by HABS with 3.42 ± 4.07 UTN, and HABC with 4.09 ± 2.23 NTU.
The wetland with the highest average value was the HACS with 6.67 ± 4.29 NTU and with
regard to the water entering the distribution tank, it had a mean value of 93.0 ± 22.89 NTU
(Figure 8).
120 B
Turbiedad (NTU)
80
40
A A A A
0
HABC HABS HACC HACS T. distr.
Tratamiento
Nota: El tamaño de muestras compuesta es N = 40 para cada tipo de medio. Letras distintas representan
diferencias estadísticamente significativas.
Fuente: Elaboración propia
Color
The Kruskal-Wallis test for the color variable (UC) shows that the p-value is less than
0.05, so there is a statistically significant difference between the medians with a confidence
level of 95%. The treatment with the lowest value of median ± SD was HACC with 132.0 ±
123.49 UC, followed by HABC with 149.5 ± 43.04 UC and HABS with 149.5 ± 92.78 UC.
The wetland with the highest median value was the HACS with 187.5 ± 104.03 UC.
Regarding the water entering the distribution tank, a median value of 1333.25 ± 466.59 UC
was presented (Figure 9).
(X 1000)
2
1.6 B
Color (UC)
1.2
0.8
A A A A
0.4
0
HABC HABS HACC HACS T. distr.
Tratamiento
Nota: El tamaño de muestras compuesta es N = 40 para cada tipo de medio. Letras distintas representan
diferencias estadísticamente significativas.
Fuente: Elaboración propia
pH
The Kruskal-Wallis test shows that the p-value is less than 0.05, so there is a
statistically significant difference between the medians with a confidence level of 95%
(Figure 10). The behavior of the tributary during the monitoring period was slightly alkaline
(8.34 ± 0.02). The HA that exhibited the most stable values of neutral pH were the HABC
and the HACC (7.83 ± 0.02 and 7.93 ± 0.02) during the whole operation phase; nevertheless,
the HACC presented an atypical behavior, since it reached a maximum value in September
of 8.68, although it subsequently stabilized until October, where it had a minimum value of
7.01. For HA control (HABS and HACS) the average behavior during the operation phase
was slightly alkaline trend (8.05 ± 0.02 y 8.00 ± 0.02).
8.8
C
A B B B
8.5
8.2
pH
7.9
7.6
7.3
7
HABC HABS HACC HACS T. distr.
Tratamiento
Nota: El tamaño de muestras compuesta es N = 40 para cada tipo de medio. Letras distintas representan
diferencias estadísticamente significativas.
Fuente: Elaboración propia
Electric conductivity
The Kruskal-Wallis test shows that the p-value is less than 0.05, so there is a
statistically significant difference between the medians with a confidence level of 95%. The
median values of ± SD were lowest in the HABS treatment (915.0 ± 204.47 μ S cm-1),
followed by HACS (925.5 ± 169.65 μ S cm-1), HACC (1014.0 ± 201.66 μ S cm-1) , HABC
(1055.0 ± 153.92 μ S cm-1) and the highest average value was the distribution tank with
1371.0 ± 140.19 μS cm-1 (figure 11).
(X 1000)
2
1.6 E
CD D
CE (µ S cm-1)
A AB
1.2
0.8
0.4
0
HABC HABS HACC HACS T. distr.
Tratamiento
Nota: El tamaño de muestras compuesta es N = 40 para cada tipo de medio. Letras distintas representan
diferencias estadísticamente significativas.
Fuente: Elaboración propia
Temperature
The ANOVA analysis shows that there is a statistically significant difference (p>
0.05) in the average values of the temperature variables (° C) between the different treatments
evaluated in the artificial wetlands. The wetlands that presented the lowest average value at
the temperature were the HABC (25.66 ± 1.66 ° C), followed by the HACC (25.77 ± 1.62 °
C). The highest average was presented in the distribution tank with 26.84 ± 1,008 ° C (Figure
12).
27.4
B
27
Temperatura (°C)
26.6
A A A A
26.2
25.8
25.4
25
HABC HABS HACC HACS T. distr.
Tratamiento
Nota: En todos los casos N = 40. Letras distintas indican diferencias estadísticamente significativas entre
tratamiento (p < 0.05) con 95 % de nivel de confianza.
Fuente: Elaboración propia
500
400
SST (mg L-1)
B
300
200
100 A A A A
0
HABC HABS HACC HACS T. distr.
Tratamiento
Nota: El tamaño de muestras compuesta es N = 40 para cada tipo de medio. Letras distintas representan
diferencias estadísticamente significativas.
Fuente: Elaboración propia
500
400
DBO (mg L-1)
300
200
B
100
A A A A
0
HABC HABS HACC HACS T. distr.
Tratamiento
Nota: El tamaño de muestras compuesta es N = 40 para cada tipo de medio. Letras distintas representan
diferencias estadísticamente significativas.
Fuente: Elaboración propia
Total nitrogen
The one-way Kruskal-Wallis analysis shows the existence of a statistically significant
difference (p <0.05) between the medians of the NT variable of the treatments evaluated in
artificial wetlands with 95% reliability. The wetlands that presented the NT parameter with
a value of median ± SD lower were the HACC (4.74 ± 11.24 mg L-1), followed by the HABS
(5.6 ± 10.24 mg L-1) and the HABC (6.7 ± 3.65 mg L -one). The record of the highest median
in NT was presented in the feed tank or raw water (147.47 ± 23.50 mg L-1) and later in the
HACS (10.92 ± 7.18 mg L-1) (figure 15). The wetland that had the support media
sedimentary gap with vegetation was the most efficient with 96.14% removal efficiency,
followed by the wetland with support medium with sedimentary gap without vegetation and
round ridge with vegetation, with 95.59% and 95.42%, respectively.
160
NT (mg L-1)
120
80
40 B
A A A A
0
HABC HABS HACC HACS T. distr.
Tratamiento
Nota: El tamaño de muestras compuesta es N = 40 para cada tipo de medio. Letras distintas representan
diferencias estadísticamente significativas.
Fuente: Elaboración propia
Total phosphorus
The one-way Kruskal-Wallis analysis shows the existence of statistically significant
differences (p <0.05) in the medians of the PT variables between the treatments evaluated in
the HA, with 95% reliability. The wetlands that presented in the PT parameter the lowest
median value were the HACC (0.18 ± 0.42 mg L-1), followed by the HABS (0.21 ± 0.24 mg
L-1) and the HABC (0.26 ± 0.19 mg L-). one). The record of the highest median phosphorus
in HA was presented in the HACS (0.41 ± 0.27 mg L-1) (Figure 16); however, the highest
PT value was present in the incoming water with a median of 5.61 ± 1.31 mg L-1. The
wetland that had the sedimentary gap support medium without vegetation was the most
efficient with 95.94% removal efficiency, followed by the wetland with support medium,
sediment gap with vegetation and round with vegetation, with 95.59% and 95.42%,
respectively.
8
PT (mg L-1)
B
2
A A A A
0
HABC HABS HACC HACS T. distr.
Tratamiento
Nota: El tamaño de muestras compuesta es N = 40 para cada tipo de medio. Letras distintas representan
diferencias estadísticamente significativas.
Fuente: Elaboración propia
Discussion
Color and turbidity
During the operation phase, HA with vegetation presented lower values of color and
turbidity of up to 54 UC and 1.32 NTU, while the maximum value was 505 UC and 18.40
NTU. With this, a removal efficiency for the HABC of 88.91% of color and 96.38% of
turbidity was reached, while the HACC reached 87.32% of color and 95.71% of turbidity.
The control HAs achieved an efficiency of removal with the media support gap sedimentary
(HABS) of 88.93% UC and 95.87% NTU, and with the round-ridge HA (HACS) the
efficiency was 85.45% CU and 94.40% NTU.
According to García and Corzo (2008), color and turbidity are related to the presence
of solids in suspension in the wastewater in the HA. In this sense, the support means fulfills
a filtering function of these solids, retaining them by adhesion, while the same flow due to
its low velocity allows sedimentation, which favors the HAFS to have a yield of more than
90% in effectiveness. the removal of suspended matter (Conagua, 2016b).
Total nitrogen
The results of the effluents of the HA for the variable NT were presented with a
variation of 7.5 mg L-1 to 11.6 mg L-1, reaching a removal of 94% to 96%. These findings
are comparable with those presented by Bai et al. (2017), who in their study of artificial
multilayer wetlands treating wastewater from a university campus in Guilin, China, reported
NT removal efficiencies of 74% with respect to the quality of wastewater that fed their HA.
These researchers consider that the effect of nitrification is a limiting factor for NT
elimination. These efficiencies are probably linked to the effect of temperature because it
influences microbial activity, which is inhibited at low temperatures, while at warm
temperatures it increases due to the aerobic, anoxic and aerobic conditions of these systems
(Fan et al. al., 2016). This is represented by the study by Wu, Ma, Kong and Liu (2018), who
using HA in a climate of 0 ° C to 10 ° C obtained removal efficiencies of 59.92%. In this
study, the researchers used a combination of coarse gravel and sand as a support medium
with a vegetation of Phragmites australis.
Total phosphorus
The removal efficiencies of PT were presented from 94% to 95.9%, and can be
compared with those obtained by Wang, Dong, Liu, Liu and Zhu (2013), who achieved a
maximum removal efficiency of 95.88% with a medium of oyster shell support, and whose
feeding affluent had a PT load of 83.64 mg L-1. In this sense, Yin, Yan and Gu (2017) have
also reported that calcium-rich and thermally modified attapulgite can achieve a removal
efficiency of 94.3% with an HRT of 8 hours. The differences of these two investigations in
relation to the present one are in the dimensioning of the units (given that the scale presented
in this study is larger) and in the absorption by the vegetation Sagittaria latifolia (swallow's
tail). This parameter is discharged into the effluents of the HA within the limits established
by NOM-001-SEMARNAT-1996 for the protection of aquatic life.
Conclusion
According to the results of the present investigation, it can be concluded that HAFS
on a pilot scale with Sagittaria latifolia (swallowtail) vegetation with round ridge support
means and sedimentary gap are useful for removing basic pollutants and control parameters
from domestic wastewater. In fact, it can be affirmed that during the first year of evaluation,
the round-topped HA with vegetation is more efficient for that purpose because it achieved
the following removal percentages: 96.85% for SST, 95.85% for BOD5, 96.78% for NT and
96.79% for PT, while in the experiments with sediment gap the removal efficiencies reached
were 95.52% for SST, 95.02% for BOD5, 95.45% for NT and 95.36% for PT.
References
Abou-Elela, S. I., Golinielli, G., Abou-Taleb, E. M. and Hellal, M. S. (2013). Municipal
wastewater treatment in horizontal and vertical flows constructed wetlands.
Ecological Engineering, 61, 460-468. Doi: 10.1016 / j.ecoleng.2013.10.010.
Bai, S., Lyu, T., Ding, Y., Li, Z., Wang, D., You, S. and Xie, Q. (2017). Campus sewage
treatment in multilayer horizontal subsurface flow constructed wetlands: Nitrogen
removal and microbial community distribution. CLEAN – Soil, Air, Water, 45(11).
Doi:10.1002/clen.201700254.
Bedoya, J. C., Ardila, A. N. y Reyes, J. (2014). Evaluación de un humedal artificial de flujo
subsuperficial en el tratamiento de las aguas residuales generadas en la Institución
Universitaria Colegio Mayor de Antioquia, Colombia. Revista Internacional de
Contaminación Ambiental, 30(3), 275-283. Recuperado de
http://www.redalyc.org/articulo.oa?id=37031522004.
Chung, A., Wu, Y., Tam, N. and Wong, M. (2008). Nitrogen and phosphate mass balance in
a sub-surface flow constructed wetland for treating municipal wastewater. Ecological
Engineering, 32(1), 81-89. Doi:10.1016/j.ecoleng.2007.09.007
Comisión Nacional del Agua (Conagua) (2016a). Estadística del agua en México. Secretaría
de Medio Ambiente y Recursos Naturales. Recuperado de
http://201.116.60.25/publicaciones/eam_2016.pdf.