Introducing a new GHG emission calculation approach for alternative methane reduction measures in the wastewater treatment of a palm oil mill
Palm oil mill wastewater treatment is a significant source of greenhouse gas (GHG) emissions. The wastewater, palm oil mill effluent (POME), carries substantial amounts of organic matter: If left in ponds, most of this organic matter would decompose relatively quickly into methane (CH4) and carbon dioxide. A belt filter press has been introduced as a means to separate organic matter from the wastewater. Solidified organic matter, the belt press cake recovered with a belt filter press, can be further used as organic fertilizer in the plantation. The aim of this study was to confirm that there is a reduction of CH4 emissions associated with the belt filter press and to find a simple method for determining the reduction at the palm oil mills. CH4 measurements and total organic carbon (TOC) analysis were conducted in a palm oil mill in Malaysia. The emissions were measured both in ponds with solid separation and in baseline open ponds without any CH4 reduction measures. TOC was measured in POME and in belt press cake. By removing organic matter, the measured palm oil mill CH4 emissions of the POME ponds were reduced by 11%. The reduction was as high as 54% in the single pond where the belt filter press was installed. A calculative 13% reduction in methane formation potential was confirmed with TOC measurement. TOC measurement was found to be a simple method worth considering for determining GHG reduction.
KeywordsPalm oil mill effluent GHG reduction Methane measurement Wastewater treatment Solids separation Belt filter press
At palm oil mills, the treatment of wastewater, often called palm oil mill effluent (POME), usually consists of a series of open ponds that are a significant source of greenhouse gas (GHG) emissions (Chin et al. 2013; Taylor et al. 2014), especially of methane, CH4, which has high global warming potential. The wastewater carries substantial amounts of degradable organic matter in the form of mill process residues. According to Loh et al. (2017), the dry matter content of fresh effluent entering the pond system is approximately 5%. This provides perfect anaerobic conditions and enables highly effective biodynamic processes: The carbon in POME is quickly turned into climate-relevant biogas. Thus, the management of a wastewater treatment plant (WWTP) for palm oil mills plays a substantial role when calculating the life cycle GHG emissions of crude palm oil (CPO).
The belt filter press is a device used for solid–liquid separation processes, particularly the dewatering of sludges in different industries and water treatment. The process of filtration is primarily obtained by passing a pair of filtering cloths and belts through a system of rollers, as illustrated in Fig. 1. The system takes sludge, effluent, or slurry as feed, often pretreated with flocculant, and separates it into a filtrate and a solid press cake. The filter cake can be used for fertilizer in the plantation, which reduces the need for GHG-intensive synthetic fertilizers.
The belt filter system also has significant potential for reducing GHGs from the wastewater. An installed belt filter system removes the organic matter out of the decaying and methane-producing conditions of open POME ponds. The aim of this study was to demonstrate that installing a belt filter system leads to reduced GHG emissions and to calculate the emission reduction based on the organic carbon in the belt press cake.
The arrangements for the actual CH4 emission measurement on the ponds required special equipment and both time and expertise; it was seen as too complicated a method for a palm oil mill to regularly identify the emission level of the wastewater as part of their business as usual: Regular CH4 measurements or monitoring cannot be integrated into the wastewater management. Therefore, there was a need to find a simple indicator to reliably determine the CH4 reduction effect of the belt filter system, in addition to confirming the CH4 reduction confirmation.
Two simple methods related to the assessment of wastewater methane formation potential were considered: chemical oxygen demand (COD) and total organic carbon (TOC). The TOC method was chosen from the perspective of trying to find the simplest method for the mill to analyze the methane reduction potential of their WWTP, without an extensive amount of sampling. The avoided methane formation potential of the belt press cake can be calculated based on its average carbon content.
COD analysis has been used previously for methane formation potential measurement for wastewaters (Yacob et al. 2006; Ahmad et al. 2003); however, the actual dependency may have significant site-specific differences as the method also oxidizes other substances besides carbon, and the methane conversion factor selection may need further wastewater analysis at each palm oil mill where the method is applied. Poh et al. (2010) state that TOC is essential for determining the biodegradability of POME. Also, the belt press cake has a high concentration of solids, which raises the question whether TOC is in fact the more suitable of the two methods for analyzing the emission potential of the belt press cake. COD has been used to assess the potential of a less solid raw effluent flow to form methane (Yacob et al. 2006), and a belt press cake is a much more homogenous sampling point than POME. The literature data available in general for POME and belt press cake TOC are basically nonexistent, and the usability of the indicator was in need of further assessment. In light of this study, TOC analysis appears to be a noteworthy method for analyzing GHG emission reduction potential in POME and belt press cake.
2 Materials and methods
The CH4 emissions and organic carbon from POME and belt press cake were measured at one palm oil mill in the state of Perak, Malaysia, which is owned by KLK. The criterion for selecting the palm oil mill was to have an installed belt filter system in operation in at least one of the anaerobic wastewater ponds. The study was conducted in January–March and June–July 2017.
The palm oil mill receives its raw material from several decentralized plantation facilities totaling approx. 10,000 ha. The average daily production of the palm oil mill was approx. 156 metric tons (t) of CPO during the measurement period. In this study, the mill’s POME discharge to the ponds was approx. 430 m3 per day, of which the belt filter press treated 100–150 m3. The production of belt press cake was approx. 27 metric tons per day.
The palm oil mill has five parallel anaerobic POME ponds. The belt filter press pump was installed in pond 1, to which it had been moved from pond 2 3 weeks prior to the measurement period. The belt filter press had been in use in pond 2 from October to December 2016. The remaining ponds 3, 4, and 5 have never had belt filter system in use. The palm oil mill had a system where fresh effluent was pumped permanently on a daily basis into anaerobic ponds and there was no rotation between the ponds. The POME load to each pond was assumed to be similar, and the only difference between the ponds was the belt filter press pump that was installed in pond 1 and that the pump had been in use in pond 2.
2.1 Methane measurement
Measurements were taken for 8 weeks between January and March 2017. Equipment tests and on-site calibration were performed during the week prior to commencement of the actual measurement. Measurements were done 5–6 days per week. CH4 measurements were carried out in a three-day rotational mode, starting on day 1 in anaerobic pond 1 and covering anaerobic POME pond 2 on the second day. On the third day, the measurements were taken in one of the remaining POME ponds 3, 4, or 5, which served as the baseline measurements with no CH4 reduction measures in use.
Due to the change in measurement points in the second half of the measurement period, and in order to harmonize the measured data, CH4 emissions were interpolated at five positions. The interpolation of values was calculated on each transect, based on the delta between two points and the according delta of pond depths. Pond depths were measured at all transects of each pond and associated accordingly.
Water temperatures were recorded with logging devices (LogTag) every 30 min at different depths (50, 200, 400, and 600) over several days in each pond. Air temperatures were recorded in the close vicinity of the POME ponds throughout the whole measurement period. The amounts of daily production of CPO, POME, and belt press cake were received from the mill’s own statistics.
2.2 Methane measurement equipment
2.3 Methane measurement emission calculations
Due to the amount of data produced and in order to avoid human bias in the data evaluation, the final calculation of emissions or fluxes is an automated process capable of presenting a full statistical data evaluation. The calculation was conducted using an automated approach by Hoffmann et al. (2017), which is available within the open-source statistics program R. The final output of the algorithm is the emission (or flux) given as a mass increase of the specific gas over time and surface. Having measured each pond repetitively in a close spatial and temporal grid, and taking into account the specific size of each pond, an average emission value for each pond was generated.
2.4 TOC measurement
TOC measurements were conducted after the actual CH4 measurement period. TOC was measured in the raw POME in the pond inlet, in the final effluent in the pond outlet, and in the belt press cake. The samples were taken in July and August 2017, and the amount of samples required was assessed based on the aim of the simplest plausible method for the palm oil mill. A total of 12 samples were taken, of which five were of the belt press cake. All 12 samples were analyzed in two commercial laboratories in Penang, Malaysia, using Nelson and Sommers and Walkley–Black titrimetry methods. The dry matter content of the TOC samples was analyzed using ASTM D 2974 and APHA 2540B.
2.5 Correlation between methane and TOC measurements
The theoretical CH4 emissions were calculated for a hypothetical POME treatment with five ponds without solid separation, using the average baseline emission for each pond as detected in the case palm oil mill. This was compared to emissions of another hypothetical mill with four similar average baseline ponds, plus one pond with a belt filter press and emissions equal to emissions from case palm oil mill pond 1. The reduction achieved with this exercise was compared to reduction of CH4 formation potential achieved by the removed organic carbon with a belt filter press. The CH4 formation potential calculation was based on TOC measurement from incoming POME and belt filter press cake.
Air temperature measurements showed the typical daily fluctuation of a tropical climate, from 22 °C at night to 34 °C during the day. All records were identical with regard to the hot temperatures, and no significant changes were observed throughout the observation period. The same observation applies for the records of POME temperatures in the ponds, where the temperature remained at the same level for the whole measurement period, the variation being within 1 °C at deeper levels. At the surface level, the total temperature delta is less than 3 °C at a depth of 50 cm below the surface, the variation being caused by sudden precipitation events.
All five ponds showed similar trends with regard to the decrease of CH4 emissions from the POME inlet to the POME outlet position. Emissions appeared to be highest in the center position of each transect and lower to the edges of the ponds, which was as expected due to the shape of the ponds (Fig. 2).
POME and belt press cake sample measurements showed averages of 1.6% TOC (variation of 1.26–1.75%) and 5.3% dry matter in raw effluent from the POME inlet; 0.5% TOC (variation of 0.39–0.74%) and 2.5% dry matter in the final POME discharge; and 3.3% TOC (variation of 2.45–3.71%) and 15.5% dry matter in the belt press cake. The averages are weighted per the amount of samples in each measuring month. The concentration of dry matter in the raw effluent is in the range of variation recorded in the literature (Poh et al. 2010; Chin et al. 2013; Loh et al. 2017). The effect of 27 metric tons of removed belt press cake equals 0.89 t of removed carbon per day with a 3.3% carbon content. This equals 13% of the incoming POME carbon content.
There was a significant difference between single-pond CH4 emissions from pond 1 and the average of ponds 3, 4, and 5 (Fig. 6). Pond 2 also resulted in lower CH4 emissions than the baseline ponds: It can be assumed from the average CH4 emissions from pond 2 that it was in some intermediate state between “baseline” conditions and “belt filter press” conditions due to the earlier utilization of the belt filter press prior to the measurement period. However, the emission dynamics were rather high and no temporal or other trends could be read from the data analysis during the observation period. The air and water temperatures were regarded as stable from which no influence on emission changes could have derived, and it is assumed that conditions within the ponds do not follow any daily dynamics.
In the mill in question, the effect of 27 metric tons of removed belt press cake is equal to 0.89 t of removed carbon per day. Further calculated with formula (2) with the IPCC methane global warming potential value of 28, the daily GHG formation potential of the belt press cake carbon would have been 24.9 t CO2e. Assuming 260 operational days per year, this would result in approx. 6500 t of prevented CO2e annually and a reduction of 0.16 kg CO2e per kg of CPO in this particular palm oil mill with a daily production of 156 metric tons of CPO. When using a global warming potential factor of 23 for methane as defined in the EU renewable energy directive (EC 2009), the reduction amounts to 0.13 kg CO2e per kg of CPO.
The carbon mass balance of all five ponds at the case mill based on TOC measurements of daily POME inlet and outlet and belt press cake flows in the mill’s wastewater treatment system indicates that one belt filter of this capacity removes 13% of the carbon from POME and that the POME outlet contains 22% of the organic carbon of POME inlet. This still leaves a theoretical maximum potential of half of the incoming POME to be converted to methane, and approximately 10% to be released as carbon dioxide, when applying the IPCC wastewater methodology for methane and carbon dioxide conversion (UNFCCC 2003, 2015).
There were 445 palm oil mills in Malaysia in 2015 (Loh et al. 2017), of which 246 had a biogas plant either installed, under construction, or being planned. If the remaining 199 mills were of average size and were to install a belt filter press into one of their POME ponds, it would prevent 1.3 million metric tons of annual GHG emissions in Malaysia when using the IPCC emission factor of 28 for methane (IPCC 2013). This represents 7% of Malaysian POME GHG emissions in 2015 (18.15 million t CO2e) (Loh et al. 2017).
According to the mill, the belt filter press has played a significant role in improving the wastewater quality and WWTP management practices. The theoretical calculation based on the carbon content of the wastewater shows that a belt filter press also has a significant role in climate change mitigation when removing the carbon from the wastewater. The actual CH4 measurement results give a similar indication with significantly decreased CH4 emissions in the ponds where the belt filter system was, or had been in use. Taking into account the complexity of CH4 measurements and uncertainties related to the measurements, analyzing TOC from belt press cake appears to provide a simple method worth consideration when estimating the GHG formation potential of palm oil mill wastewaters, based on the carbon load.
The methane capture system is more efficient in terms of methane reduction than a belt filter press. However, in cases where the mill is located too far from the electrical grid to feed the excess electricity, or where the mill is not large enough to efficiently utilize methane capture, the belt filter press method is a sensible solution and an excellent plan B for reducing methane emissions. Unlike methane digestion tanks, the belt filter press facility has a net energy requirement to operate, and some chemicals are needed to form good flocs and hence optimize organic matter separation from the POME. Based on KLK’s experience, the investment costs of a methane capture system, in contrast, are about 30 times higher than for the belt filter press system. This easily makes up for the moderate operational costs. Also, usage of the carbon-rich filter cake as a soil enhancer reduces fertilizer costs at the plantation.
Both the amount of belt press cake produced per day and its organic carbon content have a significant impact on the palm oil mill GHG emissions. Even further benefits could be achieved by installing more belt filter press capacity and possibly treating more than one of the palm oil mill’s anaerobic ponds at a time, so that more of the carbon is removed from the ponds. However, the calculation of the emission factor introduced in this study may only apply with proof of an actively running belt filter facility, including records of belt press cake volumes. In addition, the actual carbon content of the produced belt press cake must be analyzed by an authorized laboratory; TOC should be systematically measured several times per year at the palm oil mill in question, in both high and low crop seasons, for a longer period of time. If TOC level averages remain stable, the sampling frequency may be reduced.
The observations in this study represent small temporal frames and a small fraction of many variations of open POME pond management from KLK, and there are several uncertainties included in the measurements and in the results. Representative sampling is essential in emission calculation. Further comprehensive studies are suggested to investigate whether the findings made here can be further generalized.
The authors would like to thank the Sustainable Trade Initiative IDH for taking part in funding the research project.
- Drösler, M. (2005). Trace gas exchange and climatic relevance of bog ecosystem, Southern Germany. Ph.D. Dissertation, Lehrstuhl für Vegetationsokologie, Department für Ökologie, Technischen Universität München. https://mediatum.ub.tum.de/doc/603619/603619.pdf. Accessed 24 October 2017.
- EC. (2009). DIRECTIVE 2009/28/EC of the European Parliament and of the Council of 23 April 2009 on the promotion of the use of energy from renewable sources and amending and subsequently repealing Directives 2001/77/EC and 2003/30/EC.Google Scholar
- Hoffmann, M., Schulz-Hanke, M. V., Garcia Alba, D. J., Jurisch, N., Hagemann, U., Sachs, T., et al. (2017). A simple calculation algorithm to separate high-resolution CH4 flux measurements into ebullition- and diffusion-derived components. Atmospheric Measurement Techniques,10(1), 109–118.CrossRefGoogle Scholar
- IPCC. (2013). Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. In: T. F. Stocker, D. Qin, G.-K. Plattner, M. Tignor, S. K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex, P. M. Midgley (eds.) (p. 714). Cambridge University Press: Cambridge, United Kingdom and New York, NY, USA.Google Scholar
- Loh, S. K., Nasrin, A. B., Mohamad Azri, S., Nurul Adela, B., Muzzammil, N., Daryl Jay, T., et al. (2017). First report on Malaysia’s experiences and development in biogas capture and utilization from palm oil mill effluent under the Economic Transformation Programme: current and future perspectives. Renewable and Sustainable Energy Reviews,74, 1257–1274.CrossRefGoogle Scholar
- UNFCCC IPCC. (2003). Greenhouse gas inventories and additional information submitted by parties included in Annex I reporting, accounting and review requirements relating to the second commitment period of the Kyoto Protocol, version 01.10 (pp. 150).Google Scholar
- UNFCCC IPCC CDM AMS-III.H. (2015). Small-scale methodology methane recovery in wastewater treatment, version 18.0 (p. 13).Google Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.