Nearly 50% of India's population depends on variants of pit-toilet systems for human waste disposal. Nitrate contamination of groundwater by pit-toilet leachate is a major environmental concern in the country as it sources a major proportion (50–80%) of potable water from aquifers. Therefore, minimizing nitrate contamination of groundwater due to leachate infiltration from pit-toilet systems is essential. Batch and column experiments demonstrated the capability of bentonite-enhanced sand (BES) specimens to reduce nitrate concentrations in synthetic solutions (initial NO3-N concentration = 22.7 mg/L, C/N = 3) by about 85–90% in 10 to 24 hour by a heterotrophic denitrification process. Based on the laboratory results, it is recommended that use of a BES-permeable reactive barrier layer at the base of pit-toilets will facilitate heterotrophic denitrification and mitigate nitrate contamination of the underlying aquifer.

INTRODUCTION

Nearly 50% of India's population is categorized as un-sewered communities that rely on variants of pit-toilets for disposal of human waste (Chourasia 2008). A central component of these on-site systems is a single or double pit that facilitates anaerobic decomposition of solid waste and leaching of nitrate-rich effluents into surrounding soil (Garg 1988; Metcalf & Eddy 2003). Assessment of groundwater quality in Mulbagal town, Kolar District, Karnataka that relies on pit-toilet systems for sewage disposal showed excessive Escherichia coli and nitrates in drinking water wells located inside the town due to leachate infiltration from pit-toilets (Rao et al. 2013); the results of the study underlined the need to develop barriers to mitigate groundwater nitrate contamination by pit-toilets.

Slow sand filters have traditionally been used as a water treatment technique, where colloidal particles and micro-organisms are trapped by the sand matrix (Huisman & Wood 1974). The large surface area and cation exchange capacity of clay particles facilitates bacterial adhesion (Stevik et al. 2004). Inclusion of clay particles in a sand matrix may not only act as supporting media but also act as inocula for the growth of denitrifying bacteria. Studies by Malini (2015) have shown that use of bentonite contents larger than 5% restricts the flow rate of clay–sand mixes. Hence, a 95% sand +5% bentonite mix (on a dry mass basis) was examined to support heterotrophic denitrification of nitrate-contaminated septage. The 95% sand +5% bentonite mix is termed bentonite-enhanced sand (BES). Batch experiments were performed to understand the kinetics and mechanism of heterotrophic denitrification, while a permeable reactor barrier (PRB) constructed using 95% sand and 5% bentonite evaluated the barrier's potential for denitrification of contaminated water. Based on the laboratory results, a design was proposed for the construction of a PRB layer at the base of pit-toilets to mitigate nitrate contamination of underlying aquifers.

EXPERIMENTAL PROGRAM

Materials

BES specimens containing 95% sand +5% bentonite were prepared using commercially obtained natural sand and bentonite. Analytical reagent grade potassium nitrate was used as the nitrate source in the laboratory experiments. Laboratory grade ethanol was used as it is the most suited C source for nitrate removal from contaminated groundwater (Gomez et al. 2000). The total dissolved solids (TDS), pH, cation-exchange capacity (CEC), and organic matter of the BES specimens were 624 mg/L, 8.77, 5 meq/100 g, and 0.88%, respectively. The X-ray diffraction (XRD) pattern of the BES specimens showed the presence of quartz, kaolinite, and montmorillonite minerals.

Batch experiments

120 g batches of BES specimens (114 g sand and 6 g bentonite) were mixed with 60 mL of ethanol-spiked KNO3 solution (NO3-N concentration = 23.7 mg/L, C/N = 3). The moist specimens were hand compacted to a bulk density of 1.57 Mg/m3 in centrifuge bottles and incubated for periods ranging up to 48 hour at 40 °C. The incubation temperature (40 °C) and C/N ratio (3) was arrived at from parametric studies performed to identify optimal laboratory conditions for denitrification (Malini 2015). At the end of the given contact period, the moist sand specimen was centrifuged at 6,800 rpm for 12 minutes to separate the solid and solution phases. The residual nitrate and nitrite ion concentrations in the extracted pore water were determined using a Dionex ion chromatograph (ICS 2000). The ammonium ions in soil–water extracts were measured by the phenate method.

PRB experiments

Oven-dry BES samples were mixed with distilled water to yield a gravimetric water content of 17% to ensure that most voids (between particles) were saturated. After equilibrating for 24 hour, the moist specimens were statically compacted to a dry density of 1.36 Mg/m3 in permeameter cells (height and diameter: 60 and 82 mm) to form the PRB.

23.76 mg/L nitrate-nitrogen solution, spiked with ethanol (C/N mass ratio = 3) was permeated through the PRB. Outflow pore volumes were periodically analyzed for pH, EC, nitrate, and nitrite concentrations. Single pore volume of the PRB corresponded to 127 cm3. After 9.6 hour of continuous permeation, the nitrate-nitrogen concentration in the outflow solution decreased to 0 mg/L indicating that the colony of denitrifying bacteria was well established and was able to completely reduce the available nitrate. It was desired to have the nitrate-nitrogen concentration in the outflow solution below 10.2 mg/L as that corresponds to the permissible limit of nitrate-nitrogen in drinking water. The maturation period of the PRB (9.6 hour) needed passage of 2.5 pore volumes of nitrate solution (318 cm3). Subsequently, nitrate solution was permeated for about 7 hour every day and resumed the next day. The soil voids were saturated with pore solution even during the no-flow period; hence, anaerobic conditions prevailed during flow and no-flow conditions.

The outflow rate (cm3/h) was initially slow (40–45 cm3/h) but then increased to a steady-state value of 170–180 cm3/h after the passage of 40 pore volumes. Whenever the flow rate fell <70 cm3/h (from microbial growth on soil particles and accumulation of gas molecules), the PRB was backwashed by reversing the direction of permeant flow. However, there was no attempt to conduct nitrogen balance in the experiments by capturing off-gases being generated. About 380 pore volumes of ethanol-spiked nitrate solution were permeated through the PRB before stoppage.

RESULTS AND DISCUSSION

Batch experiment results

Figure 1 plots the variations in NO3-N and NO2-N concentrations in pore solutions of the BES specimens as a function of contact period. The NO3-N concentration decreased from 23.5 to 1.56 mg/L after 24 hours, while the NO2-N concentration peaked at 12 hour (10.7 mg/L) and thereafter decreased to 0.55 mg/L at 24 hours. The duration of the batch experiments was restricted to 24 hour as this period was sufficient to reduce the NO3-N concentrations in pore solutions below the drinking water limit (10.2 mg/L). Nitrate concentration decreased in the narrow pH range of 7.85 to 7.9 and the reduction commenced when the redox potential of the pore solution decreased to 90 mV.

Figure 1

Variation of nitrate-nitrogen, nitrite-nitrogen, pH, and Eh with time.

Figure 1

Variation of nitrate-nitrogen, nitrite-nitrogen, pH, and Eh with time.

Minimal NH3-N concentrations (0.094–0.13 mg/L) were measured during the batch experiments implying that ammonia (as ammonium ions) is not evolved and nitrate reduction occurs via heterotrophic denitrification rather than by a dissimilatory nitrate reduction to ammonia (DNRA) process (Calderer et al. 2010). The pore solution was rendered increasingly alkaline (pH increase from 7.82 to 8.35; Figure 1) with contact period from oxidation of organic C that released bicarbonate ions. The initial rise and decrease in nitrite concentrations imply their formation as initial by-products that subsequently convert to gaseous nitrogen. The overall bacterial denitrification reaction can be represented as:
formula
1

PRB experiment results

Figure 2 shows that the outflow nitrate-nitrogen concentration decreased to 0 mg/L (from 23.76 mg/L) after the passage of 2.5 pore volumes. Figure 1 shows that NO3-N concentration decreased from 23.5 to 7.9 mg/L after 12 hour. Comparatively, the slow permeation (33 mL/h) of nitrate solution in the PRB column achieved complete nitrate removal after 9.6 hours. The greater nitrate removal efficiency of the PRB column was possibly the result of the match between hydraulic conductivity and microbial kinetics. Vallino & Foreman (2008) and Conca et al. (2002) observed near complete nitrate removal (initial NO3-N concentration 3.16–11.3 mg/L) upon permeation of nitrate-contaminated groundwater through a permeable reactive barrier at flow rates of 12–104 mL/h.

Figure 2

Variation of nitrate-nitrogen with pore volume.

Figure 2

Variation of nitrate-nitrogen with pore volume.

After passage of 29 pore volumes, the outflow NO3-N concentration increased to 7.68 mg/L and was attributed to loss of bacterial cells attached to soil particles during intermittent backwashing at 72 and 207 pore volumes (Figure 2). Subsequent passage of permeant allowed regeneration of denitrifying bacteria colonies that reduced the NO3-N concentration to lower values (0–2.26 mg/L). In comparison, the nitrite-nitrogen concentrations showed lesser variations and ranged between 0 and 4.86 mg/L with an average value of 1.52 mg/L.

Suggested design for incorporation of PRB layer in pit-toilets

Although Rao et al. (2013) reported that the average groundwater NO3-N concentration in Mulbagal town was 33.35 mg/L, dilution effects could have rendered the anion concentration to be lower than that encountered immediately below pit-toilets. The total organic carbon (TOC) of the untreated septage (806 mg/L) and the average NO3 (148 mg/L) concentration indicates that an adequate C/N ratio (approximately 5) would be available for heterotrophic denitrification by the PRB layer installed below the base of the pit-toilets. At lower levels, water transports organic matter and microbes to high recharge sites that harbor different culturable bacteria (Brockman et al. 1992). It is probable that anoxic conditions conducive to denitrification are encountered even at levels below the PRB, which in combination with carbon source and capable bacteria population will reduce nitrates that may escape the PRB zone. In the case of an inadequate carbon source, residual ammonia could act as a reducing source for nitrate (Miller et al. 2006).

To minimize nitrate migration, it is proposed that a 0.2 m thick PRB layer be introduced at a depth of 1.3 m below the pit-base (Figure 3). To circumvent the need for back-wash, the PRB layer should be preceded by a 0.975 m thick gravel/brick bats layer and a 0.325 m thick sand layer. Trapping of suspended material in septage by brick bats and sand layers would increase the life of the PRB layer. To vent accumulated gas bubbles, a perforated pipe (perforations only in the sand and PRB layers) should be introduced. During de-sludging (every 2–3 years), the base of the pit can be excavated and the gravel/brick bat, sand and PRB layers replaced.

Figure 3

PRB layer construction for pit-toilet.

Figure 3

PRB layer construction for pit-toilet.

CONCLUSIONS

BES specimens reduced nitrate loads in batch and column experiments through denitrification reactions. The initial rise and decrease in nitrite concentrations implied their formation as an initial by-product that rapidly converted to gaseous nitrogen. The PRB experiment showed near complete nitrate removal after passage of 2.5 pore volumes in 9.6 hour. Microbial growth on soil particles and accumulation of gas molecules reduced the flow rate that necessitated back-washing of the PRB column resulting in the outflow NO3-N concentration varying from 0 to 9.49 mg/L. Based on the heterotrophic denitrification ability of BES particles, it is proposed that incorporation of a PRB barrier at the base of pit-toilets can reduce nitrate loads on underlying aquifers.

REFERENCES

Brockman
F. J.
Kieft
T. L.
Fredrickson
J. K.
Bjornstad
B. N.
Li
S. M. W.
Spangenburg
W.
Long
P. E.
1992
Microbiology of vadose zone in south-central Washington State
.
Microbiol. Ecol.
23
,
279
301
.
Calderer
M.
Gibert
O.
Martí
V.
Rovira
M.
de Pablo
J.
Jordana
S.
Duro
L.
Guimerà
J.
Bruno
J.
2010
Denitrification in presence of acetate and glucose for bioremediation of nitrate-contaminated groundwater
.
Environ. Technol.
31
,
799
814
.
Chourasia
H. S.
2008
Low cost options for disposal of human excreta
. In:
Advances in Water Quality and Management
, (
Rai
S. M.
Mani
M.
Ravindranath
N. H.
, eds).
Research Publishing
,
Singapore
, pp.
87
102
.
Conca
J.
Strietelmeier
E.
Lu
N.
Ware
S.
Taylor
T.
Kaszubal
J.
Wright
J.
2002
Treatability study of reactive materials to remediate ground water contaminated with radionuclides, metals and nitrates in a four-component permeable reactive barrier
. In:
Groundwater Remediation of Trace Metals, Radionuclides, and Nutrients, with Permeable Reactive Barriers
, (
Naftz
D. L
Morrison
S. J.
Davis
J. A.
Fuller
C. C.
eds).
Academic Press, Amsterdam
,
the Netherlands
, pp.
221
252
.
Garg
S. K.
1988
Sewage and Waste Disposal Engineering
.
Khanna Publishers
,
New Delhi
.
Gómez
M. A.
González-López
J.
Hontoria-García
E.
2000
Influence of carbon source on nitrate removal of contaminated groundwater in a denitrifying submerged filter
.
J. Hazard. Mater.
80
(
1–3
),
69
80
.
Huisman
L.
Wood
W. E.
1974
Slow Sand Filtration
.
World Health Organization
,
Geneva
.
Malini
R.
2015
Granular Media Supported Microbial Remediation of Nitrate Contaminated Drinking Water
.
PhD dissertation
,
Indian Institute of Science
,
Bangalore
.
Metcalf, Eddy
.
2003
In:
Wastewater Engineering. Treatment and Reuse
, 4th edn,
Revised by (
George
T.
Burton
F. L.
David
S. H.
, ed.).
Tata McGraw-Hill Publishing Company Limited
,
New Delhi
, p.
1819
.
Miller
J. H.
Ela
W. P.
Lansey
K. E.
Chipello
P. L.
Arnold
R. G.
2006
Nitogen transformations during soil–aquifer treatment of wastewater effluent-oxygen effects in field studies
.
J. Environ. Eng. ASCE
32
,
1298
1306
.
Vallino
J.
Foreman
K.
2008
Effectiveness of reactive barriers for reducing N-loading to the coastal zone
.
Technical Report, Ecosystems Center, Marine Biology Laboratory
,
Woods Hole, MA
,
USA
.