LEACHATE GENERATION AND TREATMENT AT A LANDFILL

LEACHATE GENERATION AND TREATMENT AT THE BUKIT TAGAR LANDFILL, MALAYSIA A P KORTEGAST*, S F ELDRIDGE*, B A RICHARDS*, S YONG**, E T CHOCK***, A BRYCE*, H ROBINSON****, M CARVILLE**** * Tonkin & Taylor Ltd, New Zealand ** TT Konsult, Malaysia *** Berjaya KUB, Sdn Bhd, Malaysia **** Enviros Consulting, United Kingdom SUMMARY: Development of the 130Mm3 mega-landfill at Bukit Tagar near Kuala Lumpur Malaysia, represents a major step forward in waste management technology for Malaysia’s largest city and the surrounding Klang Valley region. A review of leachate generation after the first year of operation highlighted the limitations of commonly adopted predictive methods (that are mainly derived from landfill performance in temperate to dry climates), when assessing leachate generation at landfills in tropical climates. The importance of landfill operational controls, the influence of tropical rainfall patterns, the significance of the water content of the waste as received are highlighted, and the influence of these variables on total and peak leachate generation are discussed. The volume and chemistry of the leachate imposed considerable demands on the design of a suitable leachate treatment plant. The on-site leachate storage that was required in order to cater for leachate generation prior to treatment plant commissioning, of itself resulted in significant pre-treatment of the leachate and promoted methanogenesis. The key elements of the leachate treatment plant design are outlined, and initial performance data are presented. These show the leachate treatment process to be very effective with a high quality final effluent being produced; suitable for spray irrigation to plantation land on site. 1. INTRODUCTION Following a fast-track development planning process, construction of an advance cell was completed in early 2005 at the Bukit Tagar Landfill, located 40km north of the Malaysian capital, Kuala Lumpur. Design of the Advance Phase was completed by TTKonsult in association with Tonkin & Taylor International. Construction commenced in 2004 and the site opened in April 2005. The site received about 0.5Mt of municipal solid waste in its first year of operation. The amount of waste received has now increased to an average of some 2000t/d. The first major cell of the main landfill has recently been constructed and is scheduled to commence accepting waste in late 2007. This paper summarises the original basis for leachate generation calculations, the principal findings of a review conducted after one year’s flow data became available, and the design operation and performance of the leachate treatment plant.

Sardinia 2007, Eleventh International Waste Management and Landfill Symposium

2. LEACHATE GENERATION The site was proposed as a modern, lined landfill generally following the USEPA Subtitle D design approach – although not rigorously. It was designed to receive a mix of non-hazardous commercial and domestic putrescible and inert waste. It is situated in a hot tropical climate and receives some 3m of rainfall annually, reasonably evenly distributed throughout the year. Monthly rainfall typically ranges from 150-360mm, with no distinct monsoon season. Estimates of leachate generation rates were undertaken at the preliminary design stage using the USEPA HELP model in conjunction with a simple water balance model. The preliminary design assessment of leachate generation for the 4 ha Advance Phase of the landfill was based on the following assumptions: Mean annual rainfall of 2767 mm 30% rainfall percolation through intermediate cover 100% rainfall percolation at the working face Rainfall always exceeds evaporation Incoming waste at field capacity Good cover and runoff management applied to minimise infiltration. From these assumptions, the average leachate flow from the Advance Phase was estimated as 90 to 160m3/d, with an assessed peak of 180m3/d. The operational methods that were initially adopted were aimed at trialling semi-aerobic landfill technology and hence cover practices were deliberately not optimal in terms of excluding rainfall. This factor, together with the even wetter than estimated nature of the waste, meant that the initial estimates of leachate generation proved to be low. Furthermore, commencement of operation of the Advance Phase was brought forward and consequently significant quantities of leachate were generated prior to commissioning of the Leachate Treatment Plant, which did not occur until April 2006. In the interim, a proportion of the leachate was recirculated into drier areas of waste, but by the time the treatment plant was commissioned, a total of some 150,000m3 of leachate had accumulated, stored in five lagoons. In April 2006 an updated water balance model was developed to back-analyse what had occurred in relation to leachate generation during initial Advance Phase development. This was aimed at providing a comparison with the leachate generation estimates made during the preliminary design phase, and at developing more accurate estimates of future leachate generation as guidance for site management practices. Back-analysis of actual leachate flow data to create a leachate generation model is becoming recognised as an important tool in predicting leachate generation in tropical climates. This is because conditions differ markedly from those typically experienced at landfills in the USA and Western Europe, and for which operational data are the most widely published. For any leachate generation model to continue to be useful throughout the operational life of a landfill, it must be continually updated and re-calibrated against actual site data. Thus the regular collection of accurate data is of considerable importance in the ongoing development of any landfill, and now forms part of the ongoing monitoring record at Bukit Tagar. The key influences on leachate generation at Bukit Tagar are: Rainfall – The site’s climate is tropical and characterised by high annual rainfall (average 2.8 m). There are four months of higher than normal rainfall, but overall rainfall occurs consistently throughout the year. This high rainfall quantity and regular pattern impose considerable constraints on the landfill development and operations. Waste moisture content - The average waste moisture content is very high (65% to 70%). Most loads exhibit free liquid and most waste deposited tends to be at greater than field capacity. This resulted in a higher than anticipated leachate base flow, which was a significant factor in initial under-estimation of leachate generation.


Operational practices – Landfill and operational practices are of critical importance in determining the amount of leachate produced and ultimately collected. Good surface water control during operation of the landfill, which includes minimising the area of the working face, diversion of clean stormwater away from the working face, and timely and appropriate use of daily and intermediate cover soil, are critical management methods needed to minimise the quantity of leachate produced at a landfill site. In this case there was significant pressure from government and environmental agencies to trial semi-aerobic landfilling techniques in the Advance Phase. The landfill was therefore designed to permit a high degree of rainfall ingress. However, this approach distorted the initial generation estimates, led to leachate management and other difficulties and was soon discontinued in favour of stringent water exclusion from the waste mass. Model development and findings An updated leachate generation model was developed and calibrated against leachate flow data recorded during the first year of site operation, to ascertain appropriate assumptions for infiltration and base flow. Rainfall data recorded on the site was supplemented with off site data where required. The infiltration rates and base flow assumptions were varied until the analysis provided a good correlation with the recorded leachate volume data. The resulting calibrated parameters are as follows: Infiltration: intermediate cover – slopes 30% of rainfall Infiltration: intermediate cover – platforms 40% of rainfall Infiltration:final cover (none placed to that point) 20% of rainfall Infiltration: tipping area and daily cover 100% of rainfall Direct flow from up-gradient slope 100% of rainfall Base flow during drier months (Jan - Mar, May - Sept) 15% of waste weight Base flow during wetter months (Apr, Oct - Dec) 20% of waste weight Average base flow 17% of waste weight Figure 1 summarises the results of the analysis and back-calibration. 160,000 140,000

Modelled leachate volume

Recorded leachate volume

120,000 100,000 80,000 60,000 40,000 20,000 0 Apr.05 May.05 Jun.05

Jul.05 Aug.05 Sep.05 Oct.05 Nov.05 Dec.05 Jan.06 Feb.06 Mar.06 Apr.06

Figure 1. Revised leachate generation model results for the first year’s operation.


The April 2006 model provided a good correlation with the cumulative leachate volume data and the analysis highlighted the following points: The lower leachate collection rate in the early months of operation of the landfill was probably a result of buffering within the waste mass. The base flow derived from the free leachate within the deposited waste was estimated to have contributed approximately 60% of the leachate generated to that point and is thus a crucial component. This component is significantly greater than the base flow typically attributed to similar waste streams even in Hong Kong (dry season 5%, wet season 10%). This is consistent with reports of the moisture content of the incoming waste mass of 65% to 70% at Bukit Tagar, compared with typical Hong Kong values of 60%. Furthermore, the regular rainfall pattern in Malaysia and the lack of a significant dry season is thought to generally increase the average moisture content of the waste mass as little is lost by evaporation. Limitations of the leachate flow data available to that point prevented more detailed back analysis such as estimation of peak daily flows as a result of individual storm events. Thus a clear delineation of the avoidable (i.e., that which can be avoided by altering landfill operation practices), and unavoidable (i.e., open face area and baseflow) leachate generation could not be achieved. However, the April 2006 analysis suggests that approximately 15% to 20% of the total leachate is likely to be the direct result of surface water from up-gradient formation slopes running into the waste mass. This is clearly avoidable to a significant extent with careful site management practices. Approximately 10% to 12% of the leachate volume is shown to be generated from rainfall on the tipping face, and approximately 13% to 18% from infiltration through intermediate cover areas. The modelled average daily leachate prediction based on the (then) site operational practices and waste volumes was in the order of 325 - 550m3/d. This is much closer to actual experience than the original estimates. Figure 2 summarises the results of the modelling of daily leachate flows. Typically within the landfill there is a lag and attenuation of leachate flow between the time when rain falls, and the time at which leachate emerges from the collection system. This effect is included in the model by calculating moving averages over 4 days. 2,500 Daily leachate volume

Attenuated

2,000 1,500 1,000 500 0 Apr.05

May.05

Jul.05

Aug.05

Figure 2. Daily leachate flow model results

Oct.05

Dec.05

Jan.06

Mar.06


The model indicated that based on the (then) current operational practices, daily flows for the Advance Phase might reach peaks as high as 1150m3/d following a significant rainfall event. This predicted peak flow was alarming and had major implications for the design and operation of future phases of the site as these were to be considerably larger. It also had significant implications for the physical requirements for leachate treatment, and both the capital and operating costs for the plant. This analysis confirmed that given the high cost of treating leachate, semi-aerobic landfilling methods were unlikely to be sustainable at the site and reversion to traditional anaerobic landfilling with meticulous operational control focused on exclusion of excess water would be essential to the ongoing viability of the site. 3. LEACHATE TREATMENT Leachate Treatment Plant Design In February 2004 Enviros Consulting was commissioned as specialist subconsultants to TTKonsult, to prepare detailed designs for a leachate treatment plant at Bukit Tagar. Based on extensive experience at similar tropical landfills (eg Robinson, 2007; Robinson and Luo, 1991), predictions were made for the composition of leachate anticipated at the site (Table 1). Table 1. Parameters used for the design of the Bukit Tagar leachate treatment plant Determinand pH-value COD BOD5 ammoniacal-N alkalinity (CaCO3) nitrate-N nitrite-N

Units pH mg/l mg/l mg/l mg/l mg/l mg/l

ACETOGENIC leachate effluent 6.0 8.0 20000 800 12000 10 1250 <2 7500 500 <1.0 <50 <0.1 0.2

METHANOGENIC leachate effluent 7.5 8.0 3000 1200 600 <30 1500 ~2 8000 500 <1 <1200 <0.1 <1

These estimated parameters were based on likely long-term mean values for the quality of each type of leachate, and recognised that peak values in leachates from specific site areas would exceed these figures on some occasions. The fact that leachate being treated would, in the longterm, be a blend of leachates from different cells and site areas, and that some control could be imposed on this, increased confidence in identifying an optimum design. Predictions were also made for the quality of effluent that could reliably be achieved by an on-site biological treatment plant, with some effluent polishing, and the predicted effluent characteristics are also included in Table 1. The initial plant was sized to provide treatment of up to 1000m3/d of methanogenic leachate, and about half that volume of acetogenic leachate. Subsequent pilot-scale treatment trials allowed this to be confirmed in practice, after the fullscale plant was actively treating methanogenic leachate. In Malaysia, discharges of treated wastewaters into surface watercourses are controlled by a set of national quality criteria known as Standards A and B. At Bukit Tagar, Standard B would be applied, and the treated leachate quality could readily comply with every limit, except that of 100mg/L for COD. Specific polishing treatment for removal of non-biodegradable COD would have involved the use of activated carbon and could not be justified based on technical and cost considerations. Arrangements were therefore made that final effluent would instead be irrigated


into the extensive areas of palm oil plantation surrounding the landfill, where concentrations of nitrate would be beneficial, and COD values would be attenuated in the soil. A treatment plant design was prepared that was based on experience from similar tropical regions (e.g., Robinson and Luo, 1991; Robinson and Carville, 1991). It comprised the following treatment processes: 1) Biological Treatment: Biological removal of biodegradable COD and nitrification of ammoniacal-N would take place in four 5000m3 HDPE-lined lagoons (see Plate 1), each of which is equipped with six floating surface aerators to provide oxygenation and mixing of solids. The lagoons operate in parallel as sequencing batch reactors (SBRs), such that temperatures within the lagoons do not exceed about 38oC, above which inhibition of treatment would occur. The lagoons are designed to operate in sequence on a 24-hour cycle, receiving inputs of leachate from a single large raw leachate balance tank (RLBT). Operation was arranged such that a discharge of SBR effluent was made by one lagoon every 6 hours, into a Treated Leachate Balance Tank (TLBT). The whole process was automated, with pH-values being maintained at optimum levels by means of dosing with sodium hydroxide solution. 2) DAF Treatment: SBR effluent is passed through a proprietary, automated dissolved air flotation (DAF) plant, sized and specified to remove almost all residual suspended solids, and some colloidal COD material. Dosing with polyelectrolyte and flocculant solution was optimised by on-site testing. 3) Reed Bed Polishing: Effluent from the DAF process is polished by passage through one of two banks of four reed beds, having a total area of 1Ha. Beds were lined with HDPE, and filled to a depth of 600mm with gravel. After planting the first four beds with reed plants, rapid growth was quickly established and it was possible to populate the second bank of beds entirely using reed cuttings from the first beds. The rapid growth allowed reeds in each bed to achieve a height in excess of 5m within one year of planting (see Plate 1). 4) Effluent Irrigation: Reed bed effluent flows by gravity to a storage lagoon. From there, it is pumped into a high level header lagoon, which feeds the extensive (100Ha) palm oil irrigation scheme.

Plate 1. Bukit Tagar Leachate Treatment Plant, Malaysia: one of the four SBR lagoons, and part of the extensive reed bed polishing system, April 2007. Leachate Characteristics at Bukit Tagar The site received about 1500t/d of waste over the first year of operation, prior to leachate treatment plant commissioning. During this period, substantial volumes of leachate were generated, and were stored in secure lined holding ponds. Table 2 presents detailed analytical data characterising the changes in the leachate chemistry in one such holding pond, Pond 3,


during this period. It can be seen from Table 2 that initial (acetogenic) leachate changed in composition during storage in the lagoon, in the tropical climate, and as development of methanogenic conditions occurred within the body of leachate in the holding pond, a substantial reduction in COD values occurred by anaerobic degradation. This is considered in more detail in a separate paper. Table 2: Results from the Bukit Tagar leachate treatment system, Kuala Lumpur, Malaysia Source: date:

Pond 3 15.3.06

Pond 3 29.3.06

Pond 3 12.4.06

Pond 3 12.12.06

RLBT 12.3.07

TLBT 12.3.07

Final Effluent 12.3.07

determinand suspended solids COD BOD20 BOD5 TOC fatty acids (as C)

930 33268 22340 17630 6804 3085

870 15900 12406 11360 3842 2127

445 9880 7931 6657 2159 1142

350 4636 2609 1834 540 461

700 4710 1922 1784 958 349

93 450 32 23 194 28

16 403 24 16 60 17

Kjeldahl-N ammoniacal-N nitrate-N* nitrite-N alkalinity (CaCO3) pH-value

2021 1923 0.2 8074 5.4

1853 1703 <0.1 6700 5.52

1729 1456 <0.1 6530 8.04

1330 1239 <0.1 5515 8.5

1904 1634 27.4 <0.1 9755 8.3

16.1 3.2 1286 0.1 136 7

13 0.9 1336 0.2 178 6.8

chloride sulphate (SO4) phosphate (as P) conductivity (µS/cm)

1568 50 40 22000

1790 32 11 16200

1687 70 8.2 15600

1615 1 8.8 15660

27036 1 12.8 23600

1889 6 4.4 15340

1873 21 6.2 15640

sodium magnesium potassium calcium

723 150 827 1038

726 132 752 302

853 113 1020 537

1175 126 1055 73

1294 155 1810 51.5

860 132 1450 61.4

936 142 1345 65.8

0.05 <0.01 12.4 0.21 0.03 0.23 0.05 0.16 <0.01 <0.001

<0.01 <0.01 0.94 0.16 0.02 0.22 0.07 0.14 <0.01 <0.001

0.02 <0.01 0.65 0.14 0.02 0.18 0.06 0.13 <0.01 <0.001

chromium manganese iron nickel copper zinc cadmium lead arsenic mercury Notes: • • • •

0.56 0.09 0.08 0.07 4.11 0.26 0.04 0.02 38.9 27.6 0.08 10.4 0.68 0.16 0.14 0.14 0.14 0.02 0.01 <0.01 6.24 0.74 0.06 0.2 0.04 <0.01 <0.01 0.03 0.38 0.12 0.03 0.20 0.09 <0.01 <0.01 <0.01 0.043 <0.001 0.005 <0.001 All results in mg/l except pH-value and conductivity (µS/cm) * nitrate results unreliable in raw leachates, presume <1 SBR eff. = effluent from SBR lagoons Final = final effluent from reed beds, awaiting irrigation

Therefore, by the time the treatment plant was commissioned during April 2006, the leachate had become methanogenic in nature. Figure 3 presents data for the changes in COD during the first year of operation. The data demonstrate the rapid change from acetogenic to methanogenic characteristics; they show impacts of flushing through more recently-formed leachate; but most


importantly, they demonstrate consistent and reliable treatment, with final effluent COD values of typically 300 - 500mg/L, in spite of variability in raw leachate COD values, as occasionally more acetogenic leachates were treated. Figure 4 shows equivalent data for removal of ammoniacal-N, concentrations of 1000 - 2000mg/l being routinely reduced to below 10mg/L in the final effluent. Removal was primarily by complete conversion to nitrate-N (see Figure 5). 5000 4500

14000 Raw Leachate COD value (mg/l)

4000 12000

Raw Leachate

SBR Effluent

Final Effluent 3500

10000

3000

8000

2500 2000

6000

1500 4000 1000 2000

500

Ju n07

M ay -0 7

A pr -0 7

M ar -0 7

Fe b07

Ja n07

D ec -0 6

N ov -0 6

O ct -0 6

S ep -0 6

A ug -0 6

Ju l-0 6

Ju n06

M ay -0 6

0

A pr -0 6

M ar -0 6

0

SBR Effluent and Final Effluent COD value (mg/l)

16000

2000

50 Raw Leachate

Raw Leachate NH4-N concentration (mg/l)

1800

SBR Effluent

Final Effluent

1600

40

1400 1200

30

1000 800

20

600 400

10

200

Ju n07

M ay -0 7

A pr -0 7

M ar -0 7

Fe b07

Ja n07

D ec -0 6

N ov -0 6

O ct -0 6

S ep -0 6

A ug -0 6

Ju l-0 6

Ju n06

M ay -0 6

0 A pr -0 6

M ar -0 6

0

SBR Effluent and Final Effluent NH4-N concentration (mg/l)

Figure 3. Removal of COD during passage through the leachate treatment plant

Figure 4. Removal of ammoniacal-N during passage through the leachate treatment plant


2500

Concentration - Nitrogen Species (mg/l)

Raw Leachate NH4-N

Final Effluent NO3-N

Final Effluent NO2-N

2000

1500

1000

500

Ju n07

M ay -0 7

A pr -0 7

M ar -0 7

Fe b07

Ja n07

D ec -0 6

N ov -0 6

O ct -0 6

Se p06

A ug -0 6

Ju l-0 6

Ju n06

M ay -0 6

A pr -0 6

M ar -0 6

0

Figure 5. Conversion of ammoniacal-N in leachate to nitrate-N in final effluent Future Development Planning During the first 12 months of treatment plant operation, over 300,000m3 of leachate was reliably and consistently treated, leaving the leachate holding ponds available again to balance peak flows of leachate, and also to provide some additional treatment of degradable COD by encouraging development of methanogenic conditions. This is likely to be important as filling progresses into Phase 1 of the main Landfill area. Initial high flow rates of acetogenic leachate from the new tipping area will benefit considerably from flow balancing and storage within these holding ponds, as was the case for leachates from the advance cell. Now that the leachate treatment plant has demonstrated it is able to treat at design flow rates and to specified standards, there are no immediate plans to extend it, but land is available if this is required at a future date. 4. CONCLUSIONS The Bukit Tagar Landfill is sited in Malaysia’s tropical climate, characterised by high annual rainfall reasonably evenly throughout the year. At the preliminary design stage a paucity of local or published leachatre generation data from sites with similar climates meant that initial generation estimates were approximate. Initial estimates were based on conservative assumptions, but still proved to be low due to the high proportion of leachate base flow generated by the saturated nature of the incoming waste. In addition the site management practices adopted for semi-aerobic landfill trials in the Advance Phase contributed significantly to total leachate flows. Consequently individual rainfall events resulted in high leachate flows over the days following major rain events. An updated water balance model prepared in April 2006 and calibrated against site data from the first year of operation, showed the significant influence of cover practices and baseflow on leachate generation. This reassessment of leachate generation rates in the Advance Phase was used as the basis for the next phase of the leachate management review aimed at addressing longer term leachate management for the main Landfill development. The findings of this review are presented in Eldridge et al (2007). The high moisture content of the incoming waste stream at Bukit Tagar has a significant


impact on the leachate generation. The resulting base flow ( 15 to 20 % of the mass of incoming waste), is significantly higher than experienced even in Hong Kong (5 % to 10 % of the waste mass) and is a major contributor to total flow (approximately 60% at Bukit Tagar on an annual basis). The paper also provides calibrated estimates of infiltration for intermediate cover slopes and confirms the critical nature of rigorous water exclusion measures in such a wet climate. The chemical characteristics of leachates at Bukit Tagar have proven to be very close to those predicted, and were used to design what is one of the largest leachate treatment plants in the world. Extended storage of leachate in holding ponds in advance of leachate plant construction has allowed strongly methanogenic conditions to become established, allowing increased volumes of leachate to be treated by the plant. The treatment plant provides biological treatment in four large aerated lagoons, operated as SBRs, to achieve reduction in COD values, and essentially complete nitrification of very high concentrations of ammoniacal-N. Biological effluent is treated by passage through a DAF plant, with final polishing through extensive reed beds, followed by safe irrigation onto a large area of palm oil plantation. The plant has performed in close accordance with the design, now routinely treating up to 1000 cubic metres of leachate per day, to the standards it was designed to achieve. The plant is operated automatically, by state-of-the-art software, which includes many fail-safe features, and allows for remote interrogation of plant operation. 4. ACKNOWLEDGEMENTS The Authors wish to thank KUB-Berjaya Enviro Sdn Bhd for permission to publish this paper and for the supply of the data presented. 5. REFERENCES Carville M S and Robinson H D (1991). Landfill leachate in Hong Kong: Characterisation and treatment with special reference to ammonia removal. Paper presented to Polmet ‘91, International Conference on Pollution in the Metropolitan and Urban Environment, Hong Kong, 9 - 13 December 1991, 30 pp. Eldridge J., Knox K., Chock E., Eldridge S., Richards B., Kortegast A. (2007) Leachate management strategies at large tropical landfills: A case study Proceedings Sardinia 2007, Eleventh International Waste Management and Landfill Symposium, S. Margherita di Pula, Cagliari, Italy; 1 - 5 October 2007 Knox, K. (2002), Development of a Novel Process for Treatment of Leachate with Very High Ammoniacal Nitrogen Concentrations. Hydrological Data, Rainfall and Evaporation Records for Malaysia, (1986-1990), Drainage & Irrigation Division, Ministry of Agriculture, Malaysia. Robinson, H and Luo, M (1991). Characterisation and treatment of leahates from Hong Kong landfill sites. Journal of the Institution of Water and Environmental Management, June 1991, 5, (6), pages 326-332. Robinson, H (2007). The composition of leachates from very large landfills: An international review. Communications in Waste and Resource Management, (UK CIWM Journal), June 2007, 8 (1), pages 19-32. Qasim, Syed R & Chiang, Walter (1994), Sanitary Landfill Leachate, Techtomic Publishing.

LEACHATE GENERATION AND TREATMENT AT A LANDFILL

Recommend Documents