Energy for Sustainable Development 23 (2014) 32–45

Contents lists available at ScienceDirect

Energy for Sustainable Development

Energy and GHG emission reduction potential of power generation fromsugarcane residues in Thailand

Salakjai Jenjariyakosoln, Shabbir H. Gheewala ⁎, Boonrod Sajjakulnukit, Savitri GarivaitThe Joint Graduate School of Energy and Environment, King Mongkut's University of Technology Thonburi, 126 Prachauthit Road, Bangmod, Tungkru, Bangkok 10140, ThailandCenter of Excellence on Energy Technology and Environment, Ministry of Education, Thailand

⁎ Corresponding author at: The Joint Graduate School oMongkut's University of Technology Thonburi, 126 PrachaBangkok 10140, Thailand.

E-mail address: shabbir_g@jgsee.kmutt.ac.th (S.H. Ghe

http://dx.doi.org/10.1016/j.esd.2014.07.0020973-0826/© 2014 International Energy Initiative. Publish

a b s t r a c t

a r t i c l e i n f o

Article history:Received 27 September 2013Revised 15 July 2014Accepted 15 July 2014Available online 6 August 2014

Keywords:Greenhouse gasSugarcane residuesBiomass supply chainAdvanced sugar mill

Electric power production using biomass residues from agricultural production using high efficiency electricitygeneration technologies would reduce greenhouse gas (GHG) emissions and contribute to climate change miti-gation. This study investigated the case of the sugar industry in Thailand and identified scenarios offering GHGemissions reduction benefits. Electricity generation potential from using sugarcane residues and/or upgradingpower generation systems represent beneficial options. The largest potential of electricity export to the na-tional grid can be achieved by upgrading boiler systems of all sugar mills to 103 bar and 515 °C. Using 19% ofthe generated sugarcane tops and leaves along with bagasse can generate 9 TWh electricity and would re-duce GHG emissions by 4.8 Mt CO2 equivalent a year. The economic analysis shows that using high steampressure boiler configurations for power generation results in substantial reduction in production costand increase in benefit.

© 2014 International Energy Initiative. Published by Elsevier Inc. All rights reserved.

Introduction

Renewable energy technologies have an important role in miti-gating climate change through reduced anthropogenic greenhousegas (GHG) emissions (IEA, 2011). Among the various renewable en-ergy sources, biomass is currently gaining considerable interestamong energy policy makers (Bakos et al., 2008; WEC, 2013). Agri-cultural residues, especially agro-industrial wastes, are widely usedas fuel for power generation. Due to the increase in power demandand saturation of agro-waste utilization for power generation, theincrease in electricity efficiency along with the utilization of fieldresidues are receiving increased interest from researchers and policymakers (Bocci et al., 2009; Deepchand 2001; Guzman and Valdes,2000; Hassuani et al., 2005; Khatiwada et al., 2012; Larson et al.,2001; Macedo et al., 2001; UNFCCC, 2014). In Thailand, electricitygeneration, being highly dependent on fossil fuels (67% natural gas,and 20% coal and lignite), is one of the major sectors contributingto GHG emissions (EPPO, 2012). Thailand formulated the Power De-velopment Plan (PDP) for the period of 2010–2030 for energy secu-rity and adequacy by considering environment concerns, energyefficiency, and renewable energy. Promotion of both renewable

f Energy and Environment, Kinguthit Road, Bangmod, Tungkru,

ewala).

ed by Elsevier Inc. All rights reserve

energy and nuclear power was initially considered in this plan. How-ever, following Japan's Fukushima incident in 2011, the revised planin 2012 promoted natural gas cogeneration systems in the initialphase considering the existing infrastructure and domestic re-sources. This was planned to replace the older gas-fired stations bycombined cycle power plants (US.EIA, 2013). Thailand is an agricul-tural country, so biomass sources especially agro-industrial wasteshave been used as fuel for generating electricity for exporting tothe national grid through the Small Power Producers (SPP)(10–90 MW) and Very Small Power Producers (VSPP) (b10 MW)schemes. Some of the supporting schemes and incentives for SPPand VSPP are the feed-in premium tariff, exemption of investmenttax scheme, soft loans for renewable energy, and fund provisions forrenewable energy investments (Asawachintachit, 2012; Sutabutr,2010). An Alternative Energy Development Plan for 2012–2021(AEDP 2012–2021) was also established by the Thai government totarget increasing the share of renewable and alternative energy to25% of total energy demand within 10 years, particularly increasingbiomass power generation to 14,008 GWh, and reducing GHG emis-sions by 76 Mt CO2 equivalent (CO2e) annually (DEDE, 2012).

Increased power generation from agro-waste e.g. woodchips, ricehusk, and bagasse has recently been achieved by upgrading technolo-gies to increase electric efficiency. Many researchers have consideredthe feasibility of improving electric efficiency by upgrading technolo-gies and using additional fuel to meet the increased feedstock demand.The surplus electricity production from sugarcane residues (bagassecombinedwith tops and leaves)with increasing efficiency byupgrading

d.

33S. Jenjariyakosoln et al. / Energy for Sustainable Development 23 (2014) 32–45

power generation technologies has successfully been developed inseveral countries e.g. Brazil, Cuba, India, Mauritius, and Thailand(Bocci et al., 2009; Deepchand 2001; Guzman and Valdes, 2000;Hassuani et al., 2005; Khatiwada et al., 2012; Larson et al., 2001;Macedo et al., 2001; UNFCCC, 2014). Especially, the cogenerationsystems of sugar mills in Brazil had been developed from 22 barboiler with back pressure turbines to 105 bar with extraction con-densing steam turbine and using tops and leaves as additional fuelthat increased surplus electricity export 16 folds (Khatiwada et al.,2012).

Thailand is the world's fourth largest sugar producer, producing98 Mt and 10.43 Mt of sugarcane and sugar in 2012 respectively. Also,sugarcane production has been increasing 10% annually for the last5 years (OAE, 2012;OCSM, 2013). Thewaste from sugarcane processingi.e. bagasse, is already being used as the main fuel for heat and powergeneration for sugar production with excess electricity being sold tothe national grid (Mendoza et al., 2002; OCSB, 2007b). The amount ofbagasse varies from 23% to 37% of the sugarcane (Deepchand, 2002),with anaverage of 28% (Larson et al., 2001; PDTI, 2011). Another residuethat is interesting is the field residue from sugarcane cultivation i.e. topsand leaves which varies between 17 and 30% of the sugarcane (DEDE,2005; DEDP, 1992; Junginger et al., 2001; KMUTT, 2006). The availableamount of tops and leaves is approximately 74–98% of the total amountgenerated, the rest being left in thefield for incorporation into the soil asorganic fertilizer and weed control (DEDE, 2005; KMUTT, 2006;Sajjakulnukit et al., 2005). Most of the tops and leaves are currentlyopen burnt in the field in order to facilitate sugarcane pre-harvesting,and post-harvesting for land preparation (KMUTT, 2006; Yuttithamet al., 2011). Utilizing tops and leaves as fuel for heat and power gener-ation can help alleviate the open field burning problem, avoid GHGemissions and contribute to reducing energy imports (Bocci et al.,2009; DEDE, 2005, Gheewala et al., 2011; Guzman and Valdes, 2000;Mendoza et al., 2002). The tops and leaves can be used as a secondaryfuel combined with bagasse in existing boilers, thereby avoidingthe need for storing excessive amounts of tops and leaves whichwould be required if they are used as the primary fuel for thewhole year operation due to their seasonal availability limitedto 4–6 months annually in the harvesting season (Guzman andValdes, 2000; OCSM, 2013). However, the cost of the supply chain

process, including baling, field hauling and loading, truck transport,shredding and storage stacking, of the low bulk density tops andleaves is a major obstacle. Therefore this cost should be factored inconsiderations for using this residue.

Currently, most sugar mills in Thailand operate low efficiency grateboilers and back pressure steam turbines with steam pressure about20 bar and temperature 350–360 °C; the plants produce energy fortheir ownneeds (sugarmilling) for themost part with only some excesselectricity being exported to the national grid (PDTI, 2011). The averageelectricity export is only 14.5 kWh/ton sugarcane (tc) as compared to70 kWh/tc and 158 kWh/tc that have been observed for the most ad-vanced sugar mills in Thailand and Brazil respectively (Khatiwadaet al., 2012; Siemers, 2010). For new units recently equipped in themore advanced sugar mills in Thailand with boilers that producesteam at 103 bar and 515 °C, high amount of surplus electricity can beproduced for export to the grid; but additional fuel is required whichcan possibly be provided by tops and leaves (ONEP, 2013). Siemers(2010) evaluated the increasing surplus electricity generation andGHG reduction using the best available boiler technology in Thailand.However, this study was based on the existing technology in 2006 atthe highest boiler steam pressure of 70 bar and considering eitherbagasse or tops and leaves as feedstock, but not a combination ofboth. Hence, a need was perceived for updating the study withmore current technology (103 bar boiler pressure), using both resi-dues (bagasse as well as tops and leaves) in combination accountingfor actual demand and availability, and considering actual data viaextensive site surveys.

The study aims to evaluate the electricity generated from sugar-cane residues (tops and leaves, and bagasse) by upgrading the boilerconfigurations in the existing power plants with high pressure steamturbines. The associated GHG emissions from increasing surplus elec-tricity are evaluated and compared to natural gas combined cyclepower plants that are expected to be constructed following theThailand PDP 2010 plan, and be the marginal power plants displacedby the surplus electricity. The electricity export potential of differentproduction systems is also evaluated in terms of economic analysis toencourage sugar mill owners or investors to consider exporting moreelectricity. The cost models used are general and can be also appliedto other biomass feedstocks and locations.

Methodology

The methodology is organized into four parts. First, the tops and leaves availability was assessed followed by the estimation of the arearequired for collecting these residues. This was followed by the assessment of surplus electricity potential from different scenarios dependingon different technologies of cogeneration systems at sugar mills. Sugar mills in Thailand were grouped roughly based on different levels ofboiler pressure and one sugar mill of each group was selected as the representative of that group for assessing energy balance (PDTI, 2011;Siemers, 2010). After that, the overall GHG emissions assessment was carried out for the entire life cycle of power generation from bagassecombined with tops and leaves. For tops and leaves, the supply chain (life cycle) includes collection, field hauling, road transport, shreddingand storage whereas for bagasse, only storage is required as it is generated in the sugar mill itself. The avoided GHG emissions from replacingfossil-fuel power generation were also estimated. Finally, the cost assessment was conducted comprising the costs of tops and leaves supply chainas well as different configurations of power plants.

Tops and leaves quantity and area estimation

The quantity of tops and leaves generated was evaluated based on the quantity of sugarcane delivered to the mill using residue to product ratio(RPR) and surplus availability factor (SAF) as shown in Eq. (1) (Bhattacharya et al., 2005; Sajjakulnukit et al., 2005).

Tops and leaves available tð Þ ¼ Sugarcane amount tð Þ � RPR � SAF� collection efficiency: ð1Þ

RPR varies with plant structure, seasonality, harvesting methods, irrigation practices, soil quality, moisture content, and various otherminor factors (Koopmans and Koppejan, 1998). The RPR of tops and leaves range between 17 and 30% as mentioned earlier. The averagevalue of 22%was selected for estimating tops and leaves generation. The amount of tops and leaves that has to be left in the field for agriculturalpurposes (soil fertility and weeds control) depends on sugarcane variety, climate, soil, etc. (Hassuani et al., 2005). The SAF factor representsthe proportion of unused amount of tops and leaves divided by the annual total amount of tops and leaves generated. A significantly large

34 S. Jenjariyakosoln et al. / Energy for Sustainable Development 23 (2014) 32–45

amount of tops and leaves is being open burned in the field, which could be considered as unused amount (DEDE, 2005). Collection efficiencyrefers to the fraction of residues that can be collected from the field as compared to the total amount generated. The tops and leaves recoverywas calculated based on the collection efficiency of the baler of 70% (DEDE, 2005; EPPO, 2010a; Mendoza et al., 2002) and the surplus avail-ability factor of 74% (DEDE, 2005).

The tops and leaveswere assumed to have a constant distribution over the circular catchment area and distributed around sugarmills. The radiusof the catchment area was estimated from Eq. (2) which accounted for the winding nature of the road and the straight line distance for the radius bymultiplying with the square root of two (Delivand et al., 2011a; Rentizelas et al., 2009).

Radius kmð Þ ¼ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

Top sand leaves tð ÞSAF� Collection efficiency� Farmland factor� π� Average yield of top sand leaves

tkm2

� �vuuut �

ffiffiffi2

p: ð2Þ

The assumption was that 25% of the land is occupied by infrastructure (road, building, etc.), leaving 75% of the farm area to be actually cultivatedby sugarcane which indicates the farmland factor. These assumptions are based on Delivand et al. (2011a). The average yield (t/km2) of tops andleaves is based on sugarcane production in 2011/12 using the average RPR value above (OAE, 2012).

Electricity production in sugar mill

Description of power production from bagasse in ThailandSugar mills have generally used bagasse as fuel to cogenerate heat and power using boilers and turbines (back pressure or condensing turbines

with or without controlled extraction). In Thailand, the bagasse generated is about 25–31% of sugarcane with an average of 28% (PDTI, 2011).Currently, some of the bagasse generated in the crushing season is stored to serve as fuel for the sugar re-melting process and power generationin the off-season. A few advanced sugarmills equippedwith high boiler steampressures produce electricity in the off-season by using stored bagassecombinedwith other biomass (e.g.wood bark).Moreover, the other byproduct of sugar i.e.molasses, is actually sold by the sugarmills to producers ofliquor and animal feed, but site survey and literature data revealed that at the advanced sugarmillswith 70bar and 103 bar cogeneration systems, themolasses produced is used for ethanol production as an additional income source (ONEP, 2013). Therefore, these advanced sugarmills can sell part ofthe surplus steam and electricity to the ethanol plant, the remaining being sold to the national grid.

Options for increasing energy output from sugar millsThe potential of electricity generation depends on type of boiler and turbine, and operating configuration (pressure and temperature) of the co-

generation systems (Bhatt and Rajkumar, 2001; Bocci et al., 2009; Khatiwada et al., 2012; Siemers, 2010). Review of local research revealed that thesugar mills in Thailand use 20 bar, 30 bar, 40 bar, 70 bar, and 103 bar steam pressure boilers as shown in Table 1 (DEDE, 2008; OCSB, 2007a; ONEP,2013; PDTI, 2011; Siemers, 2010; Tossanaitada, 2008). It must be noted that the 20 bar configuration, which is used in about half of the sugar mills,actually represents a small range of boilers with pressures varying between 20 and 28 bars. The turbine technologies used for boilers with pressure20, 30 and 40 bars are back pressure steam type whereas those with 70 and 103 bars are extraction condensing type.

In this study, based on the above review, three scenarios have been developed, S1 is the baseline scenario representing the current situation, S2 isthe scenario where the feedstock is increased to extend the number of operating days and S3 is the scenario where the power production technologyis upgraded to 103 bar pressure boiler and feedstock increased. The details of each scenario are described as follows.

Scenario S1 represents the prevailing conditions for sugar milling and power production in Thailand with steam generation pressures at 20 bar,30 bar, 40 bar, 70 bar, and 103 bar and different turbine technologies mentioned above. The number of operating days varies between 120 and240 days as shown in Table 2, with an average of 140 days.

Scenario S2 is similar to S1 but the operating time is increased to 300 days per year (140 days crushing season and 160 days off-season). Theadditional fuel requirement due to the increased number of operating days is met by tops and leaves.

Scenario S3 assumes the upgrading of all power production systems to 103 bar boiler pressure combined with extraction condensing steamturbine. The operating time is 300 days per year. Here also, the additional fuel requirement is met by tops and leaves.

To estimate the additional fuel for Scenarios S2 and S3, a sugar mill representative from each group of boilers (20, 30, 40, 70 and 103 bars) wasselected for conducting energy balance calculations. These representative mills were surveyed for the collection of primary data such as operatingtime, crushed sugarcane amount, the average data per hour of bagasse consumption, heat and electricity requirements, and data related to steaminlet and outlet. The surplus electricity (to be exported to the national grid) for each representative group was then calculated as kWh/tc. The

Table 1Grouping of current cogeneration technologies used in sugar mills in Thailand.

Technologya Sugar mill Sugarcane delivered in 2011/12b Electricityc export

Amount (t) % (kWh/tc)

Group 1: no export 4 2,732,300 2.79Group 2: back pressure steam turbine, 20 bar, 360 °C 28 49,822,070 50.85 5.13Group 3: back pressure steam turbine, 30 bar, 390 °C 4 11,321,521 11.55 16.25Group 4: back pressure steam turbine, 40 bar, 485 °C 4 10,379,022 10.59 22.57Group 5: extraction condensing steam turbine, 70 bar, 507 °C 3 8,566,992 8.74 69.34Group 6: extraction condensing steam turbine, 103 bar, 515 °C 3 15,157,784 15.47 88.78Total/average 46 97,979,690 100.00 26.67

a Derived current technologies in used in Thailand from PDTI (2011), OCSB (2007a), Siemers (2010), Tossanaitada (2008), ONEP (2013) and a 5-site survey (2010–2012).b Delivered data of sugarcane delivered to sugar mills in Thailand in 2011/12 (OCSM, 2013).c Estimated based on the current excess electricity generated according to a 5-site survey.

Table 2Gross generated electricity and electricity production export for S1 and S2.

Items Unit S1 S2

Group 2 Group 3 Group 4 Group 5 Group 6 Group 2 Group 3 Group 4 Group 5 Group 6

20 bar 30 bar 40 bar 70 bar 103 bar 20 bar 30 bar 40 bar 70 bar 103 bar

Gross generated electricity MWh/y 33,745 89,288 246,671 394,092 385,081 61,585 119,875 309,725 503,549 534,900Own consumption MWh/y 27,719 54,050 146,886 145,831 111,140 53,908 81,468 186,617 198,073 155,392Electricity export into the grid MWh/y 6026 35,238 99,785 238,625 265,565 7677 38,407 123,108 295,840 371,204Electricity export to the ethanol plant MWh/y 9636 8376 9636 8376Steam export to the sugar mill t/y 60,752 347,446 691,508 937,793 1,044,155 79,344 378,697 853,140 1,038,061 1,223,262Steam export to the ethanol plant t/y 125,611 109,182 125,611 109,182Total bagasse demand t/y 329,216 607,134 1,237,849 963,591 837,561 329,216 607,134 1,237,849 963,591 837,561Bagasse demand for electricity exportinto the grid

t/y 30,381 150,405 341,028 542,139 536,806 26,590 133,090 354,683 525,957 540,177

Bagasse demand for electricity export forethanol plant

t/y 21,892 16,931 17,131 12,188

Total tops and leaves demand t/y 74,795 69,566 114,830 125,990 128,376Tops and leaves demand for electricityexport into the national grid

t/y 6454 16,416 35,199 74,021 89,071

Tops and leaves demand for electricityexport for the ethanol plant

t/y 2411 2010

Operating time, (24 h/day) Days/y 123 216 241 194 183 300 300 300 300 300

35S. Jenjariyakosoln et al. / Energy for Sustainable Development 23 (2014) 32–45

total electricity export potential for each group was then calculated by summing up the amount of electricity exported by each representative mill.Finally, the results of surplus electricity were multiplied by the amount of sugarcane to obtain the total electricity exported for each group(see amount of crushed sugarcane in each group in Table 1).

Life cycle GHG emissions estimation method

Goal and scope of the life cycle GHG emissions studyOne goal of the study is the estimation of the life cycle GHG emissions from power production using sugar residues as fuel. The life cycle GHG

emissions for the different scenarios outlined in Options for increasing energy output from sugar mills section were compared with those from anequivalent amount of electricity produced from natural gas combined cycle power plant which the former are supposed to substitute followingthe Thailand PDP 2010 plan. The data of life cycle GHG emissions from conventional natural gas combined cycle power plant in Thailand weresourced from Phumpradab et al. (2009). The GHGs included in the study are CO2, CH4 and N2O with 100 year-global warming potentials of 1, 25and 298, respectively (IPCC, 2007). The additional data and assumptions used in the Life Cycle Assessment (LCA) are shown in Table 3.

The study covered 42 sugar mills which produce and sell electricity; these account for 97% of the national sugar production. Most of the existingsugarmills in Thailand use bagasse as primary fuel for producing steam and electricitywhich are used in the sugarmilling process aswell as exportedoutside themill in case of excess (EPPO, 2010b; ONEP, 2013;Mendoza et al., 2002; OCSB, 2007b). GHG emissions for power production from bagasseresult only from the combustion in boilers for power production. Tops and leaves are used as additional fuel for Scenarios S2 and S3 tomeet the targetof the scenarios. The GHG emissions for tops and leaves include those from the supply chain and combustion in boilers for power production. TheGHG emissions from the construction of sugar mills and associated power plants have not been included in the system boundary because thelarge throughput and long lifetime make the impacts of their establishment and decommissioning per unit of product insignificant. The systemboundaries are shown in Fig. 1.

Table 3Factors for estimating GHG emissions from surplus electricity generation.

Item Value/assumptions Remark/references

Average sugarcane productsSugarcane harvest per hectare 76.80 t sugarcane OAE (2012)Bagasse produced 28% of sugarcane PDTI (2011)Tops and leaves available 11.4% of sugarcane Estimated value in Tops and leaves quantity and area estimation

sectionSugar produced (raw sugar) 10.45% of sugarcane OCSM (2013)Molasses produced 4.48% of sugarcane OCSM (2013)

Tops and leaves supply chain process (unit/t tops and leaves)Baling 1.75 L diesel/t Calculated based on ASAEa methodLoading 0.60 L diesel/t Pilot project data of a Thai advanced sugar millTruck transport 0.46–1.90 L diesel/t Adapted from Delivand et al. (2011a)Stacking (storage) 0.60 L diesel/t Calculated based on the ASAE methodShredding 21 kWh/t Pilot project data of a Thai advanced sugar mill

Average lower heating value (LHV) of bagasse 7.6 MJ/kg, around 50% moisture content (MC) 5 Sugar mills from site surveyAverage LHV of tops and leaves 14.12 MJ/kg, around 10% MC A Thai advanced sugar mill; EFE (2006); Junginger et al. (2001)Diesel combustion in vehicles 3.2 kg CO2e/L GEMIS 4.8 (2013)Bagasse combustion in boilers 11.23 kg CO2e/MWh National Greenhouse Gas Inventory Committee (2007),

Australian Government DepartmentTops and leaves combustion in boilers 20.78 kg CO2e/MWh National Greenhouse Gas Inventory Committee (2007),

Australian Government DepartmentThai natural gas combined cycle power plant 539.46 kg CO2e/MWh Phumpradab et al. (2009)

a Stands for American Society of Agricultural and Biological Engineers.

Sugar Milling

Sugar Molasses

Steam & Electricity generation

Ethanol plant

Sugarcane cultivation

Mill mud

Boiler ash

Tops and leaves (collecting,

transportation, shredding)

Bagasse

Electricityto grid

Excess steam &

electricity

Stea

m &

El

ectri

city

Fig. 1. System boundary indicated by the dashed outline.

36 S. Jenjariyakosoln et al. / Energy for Sustainable Development 23 (2014) 32–45

The GHG emissions of surplus electricity for the representative of each scenario were calculated using economic allocation with 1 MWh as thefunctional unit following Eqs. (3) and (4) (Nguyen and Gheewala, 2008; Ramjeawon, 2008; Renouf et al., 2011).

AFEE ¼MEE � Pelectricity

� �

ðMrawsugar � Psugar þXni¼1

ðMi � PiÞð3Þ

where AFEE is the allocation factor of the surplus electricity,MEE is the surplus electricity (MWh) per 100 tc, Pelectricity is the purchasing electricityprice (USD/MWh), Mraw sugar is the average raw sugar generated per 100 tc, Psugar is the price of raw sugar (USD/t raw sugar), Mi is the mass ofco-product i to n per 100 tc, Pi is the price of co-product i to n (USD per unit of co-product), and n is the number of co-products.

GHG emissions kg CO2e=MWhð Þ ¼ ð FB � EFGHG Bð Þ þ FTL � EFGHG TL þ EFGHG BTLð Þð Þ � AFEEMEE

ð4Þ

where FB (t) is the amount of bagasse consumption per 100 tc, EFGHG_B (kg CO2e/t) is the GHG emission factor of bagasse used as fuel, FTL (t) isthe amount of top and leaves consumption per 100 tc, EFGHG_TL (kg CO2e/t) is the GHG emission factor of top and leaves used as fuel, and EFGHG_BTL(kg CO2e/t) is the GHG emission factor of biomass supply chain process.

Power plant (MW)

Cost (USD/MW)

Selling steam to sugar mill

Sale

Selling bagasse

Grid

Own use

Tops and leaves baling

Road transportation

Cost (USD/MW)

Pretreatment size reduction

Machinery cost Operating cost Labor cost Fuel cost

Fuel cost Labor cost O&M cost

Contingency cost Interest & capital cost

Ethanol plant

Fig. 2. Scope of the project cost evaluation.

Table 4Machinery and related costs.

Machinery and parameter Unit Value

Big rectangular balerPurchasing cost of a new machinea USD 57,258Investment costb USD 25,726Insurance cost, 0.6%b USD 2205Repair and maintenance costb USD 41,367

Tractor, PTOc-84HPPurchasing cost of a new machined USD 35,472Investment costb USD 15,261Insurance cost, 0.9%b USD 2943Repair and maintenance costb USD 35,756

Crab loader, PTO-70 HP (made in Thailand)Purchasing cost of a new machinee USD 30,390Investment costb USD 11,207Insurance cost, 0.9%b USD 2522Repair and maintenance costb USD 30,634

ShredderPurchasing cost of a new machinef USD 31,990Investment costb USD 3915Insurance cost, 0.9%b USD 503Repair and maintenance costb USD 14,715

a The purchasing price of the big rectangular baler was derived from (Hassuani et al., 2005), the USD price is escalated to the base year 2013.b Calculations and parameters based on the ASAE approach and (Delivand et al., 2011a).c PTO (power take off) is the maximum power (PTO-hp) of the machine.d Kubota tractors models M8540.e The crab loader was made in Thailand (a sugarcane farmer, Kanchanaburi province).f The purchasing price of shredder was derived from a site survey.

37S. Jenjariyakosoln et al. / Energy for Sustainable Development 23 (2014) 32–45

GHG emission reduction analysisThe estimated GHG emissions of electricity export in S1 and S2 for national grid were summed from the results of multiplying GHG emis-

sions (kg CO2e/MWh) of each group with the generated electricity export for the grid for that group. The capacities of sugar mills in Thailand arebetween 2300 and 40,000 tcd; however, the trend is towards larger capacity mills. Hence, the largest scale mill was used for the GHG emissioncalculations in S3. As the excess electricity generated will displace natural gas power generation as explained above, the GHG emission reductionsfor S2 and S3 were calculated using Eq. (5).

GHG emission reductioni ¼ GHGSi– GHGS1ð Þ−DESi−S1 � GHGNG ð5Þ

where GHG emission reductioni (kg CO2e) is the GHG emission reduction of scenario i (i= 2, 3), GHGSi (kg CO2e) is the GHG emissions of the elec-tricity exported in scenario i (i = 2, 3), GHGS1 (kg CO2e) is the GHG emissions of the electricity exported in S1, ΔESi − S1 (MWh) is the difference ofelectricity exported between scenario i (i=2, 3) and S1,GHGNG (kg CO2e/MWh) is theGHGemissions perMWhof electricity generated fromnaturalgas combined cycle power plant (Phumpradab et al., 2009).

Life cycle economic analysis method

Another goal of the study is to evaluate the economic aspect of surplus electricity production of the different sugar mill configurations in the threescenarios. The economic analysis considered costs and incomes of the entire processes related to excess electricity generation for export. The costs werecategorized into fuel costs and the costs of cogeneration system for electricity production. Fig. 2 shows the boundary of the project economic evaluation.

Table 5Major parameters of the agricultural machinery; ASAE standards (Adapted from Painter, 2011; Delivand et al., 2011a).

Machinery Estimated life (h)–years RFV a (% of new costs) Repair factor

RF1b RF2b

Four-wheel tractor (12,000)–15 22.92 0.007 2.0Large square baler (3000)–10 28.37 0.10 1.8Shredder (2000)–10 34.97 0.23 1.0

a Remaining on farm value after their economic life time.b RF1 and RF2 refer to repair and maintenance factors — ASAE standards.

Table 6Major assumptions for the activity duration time (h) in a round trip (Adapted from Delivand et al., 2011a).

Activity Unit Time

Stop time, loading and uploading of bales h 0.40Baling time of the bale h/t 0.125Hauling time and staking time h/t 0.166

Table 7Transportation parameters (adapted from Delivand et al., 2011a).

Parameter and activities Unit Value

Purchasing of the trailera USD 100,128Salvage costb 10% of purchasing costMaximum volume capacity m3 100Loading weightc ton 21.2Average travel speed km/h 40Insurance and maintenance costb 8% of purchasing costMiscellaneous costb 4% of purchasing costLife time Years 7Average annual traveled distanceb km 80,000Fuel consumptiond L diesel/100 km 41.7Diesel coste USD/l 0.97Driver hourly pay rate — 8 h shift USD/h 2.0

a Obtained from a dealer of Hino Co. in Thailand (20.08.2013). Approximate dimensions: 2.5 × 2.5 × 14 (m); maximum loading capacity 37 t.b Adopted from Huisman et al. (1997), Delivand, et al. (2011a).c Based on the permitted volume capacity of the truck and number of bales (a bale dimension and weight is 0.8 × 0.87 × 1.9–295 kg) that can be loaded.d The value was obtained from IPCC (1996).e Average retail price of the first six months in 2013 for H-diesel (EPPO, 2013).

38 S. Jenjariyakosoln et al. / Energy for Sustainable Development 23 (2014) 32–45

Fuel costsBagasse alone is used as fuel for heat and power generation in the sugar mills in Scenario S1 whereas bagasse along with tops and leaves are

used in Scenarios S2 and S3. Fuel costs of bagasse and top and leaves are different because bagasse is a process based residue generated at thesugar mill, while the tops and leaves are field based residues. As bagasse is generated at sugar mills, it was considered as a cost-free fuel forsugar mills in the past. However, after the extensive use of bagasse for power production, some of which is exported to the grid, it has acquireda substantial market price. In this study, the fuel cost of bagasse was assumed 8.0 USD/t (250 THB/t). Tops and leaves are used as the secondaryfuel for power generation. As they are field based residues, the cost of collecting, transportation, and pretreatment processes must be included.Although tops and leaves are not traded at the commercial market in Thailand, the price of 1.6 USD/t (50 THB/t) of tops and leaves is assumedfrom the experience of a pilot project that paid farmers at the field. The overall fuel costs of the tops and leaves are calculated as a sum of theprice paid to farmers at the field along with the subsequent cost of handling (baling and field hauling), transport, processing (shredding) andstorage (stacking) at the power plant.

The purchasing prices ofmostmachinery such as tractors, flatbed trailers and crab loaderswere obtained from companies in Thailand (Kubota, Hino,local company) as shown in Table 4. Studies have shown the economic advantage of using large straw baling systems (Delivand et al., 2011a; Hassuaniet al., 2005); hence, large balers were assumed for collecting tops and leaves at the field. As these are currently not available in Thailand, the priceinformation was used from literature after adjusting for the base year 2013 (for July 2013; 1 USD= 31.26 THB) (Hassuani et al., 2005). The machinerycosts are divided into ownership costs (depreciation costs, insurance and interest rates) and operating costs (repair andmaintenance, fuel and lubrica-tion, and labor costs).

The calculation of ownership costs and operating costs of agricultural machinery was based on themachinery management data of the AmericanSociety of Agricultural and Biological Engineers (ASAE) Standards (Delivand et al., 2011a; Painter, 2011). Based on the aforementioned assumptionsand calculations, the input data related tomachinery are listed in Table 5. The transport distanceswere different dependingon the radius of the catch-ment area (see Tops and leaves quantity and area estimation section). The major assumptions in assessing the time duration of the trip for the pro-jectionmodel are summarized in Table 6. The used data andmajor assumptions for the road transportation assessment are presented in Table 7. It isalso assumed that the portion of the tops and leaves that need to be stored for use in the off season share the same facility as the bagasse. Thus there isno additional cost for storage space, only some cost associated with stacking. On average, a wage of 2 USD/h was assumed for the drivers of tractors,crab loaders, and shredders. The specific cost components of biomass supply chain of tops and leaves were calculated for delivering one ton of topsand leaves from the field to preparing the residues as a ready fuel for sugar mills.

Cost of electricity exported to the gridThe cost consideration of electricity exported to the grid for the three scenarios is mainly composed of operating and fuel costs. The method of

calculating fuel costs is already shown in Fuel costs section. The Scenarios S1 and S2 do not include the capital costs because the mills generateheat and electricity from the existing steam boilers and system configurations. In the third scenario (S3), all boilers were upgraded to the 103-barsteam pressure for which the sugar mills required an investment for setting up a new power plant. Thus, the costs included the capital cost ofinstalling new boilers and turbines along with operating and fuel costs.

Capital costs. The costs of the boilers, turbines, and other related machinery were obtained from DEDE (2008), ONEP (2013), and PDTI (2011). Thecost of boilers and turbines also increase with the steam temperature (Bhatt and Rajkumar, 2001). The boiler and turbine costs of 392 USD/kW,782 USD/kW, 1011 USD/kW, 1278 USD/kW, and 1393 USD/kW were assumed for 20 bar, 30 bar, 40 bar, 70 bar, and 103 bar models, respectively.The total investment cost including equipment costs, and related costs of capital parameters is shown in Table 8. The major assumptions listed inTable 8 were adapted from PDTI (2011). The costs of electricity transmission installation were also included in the investment cost. It was assumedthat around 70% of total investment costwill come from a bank loanover 10 years at an annual interest rate of 7%. The remaining 30%will be investedby the owner in the first year of project.

Operating costs. Operating costs included the maintenance and repair costs, insurance costs, labor costs, and fuel costs.

Maintenance and repair, and insurance cost. The annual maintenance and repair costs of the power plants were 4% of equipment costs. An annualinsurance cost of 1.5% of equipment costs was also assumed in the financial analysis models.

Table 8Basic of calculations for financial analysis (adapted from PDTI, 2011).

Cost component Unit Value

Equipment costCost of electric motorsa 106USD S&M: 3.07, L: 3.57Boilers & turbinesb USD/kW

Economic parametersLoan % 70Equity % 30Interest rate on loanc % 7Payback period of loan years 10

Capital parametersGrid connection costd

Electricity export capacity:≤20 MW 106USD 1.22≥20 MW 106USD 6.77

Life time Years 20Cost of engineering and civil work % 1.0% of equipment costCost of construction time insurance % 0.5% of equipment costContingency cost % 2.0% of equipment costDiscount rate % 10

Operating parametersAverage annual incomes of a labor USD/person 12,796Annual operating and maintenance cost % 4.0% of equipment costAnnual insurance cost % 1.5% of equipment costFuel costBagasse USD/t 8Tops and leaves e

Revenue parametersInternal steam soldf USD/t 2.06Unit price of electricity sold of SPP-firmg USD/kWh 0.09Feed-in premium (adder) for first 7 yearsg USD/kWh 0.01

a S, M, and L are the sizes of sugar mill capacities: small size (b10,000 tcd), medium size (N10,000 tcd and b20,000 tcd), and large size (N20,000 tcd) (PDTI, 2011).b The costs were shown in Fuel costs section.c Based on the MLR interest rate of term loans for corporate customers of typical commercial banks.d Derived from literatures (DEDE, 2008; ONEP, 2013).e Tops and leaves fuel costs were shown in Power export potential from sugarcane residues to the national grid and GHG reduction potential of utilizing the surplus bagasse, and tops

and leaves for power production sections.f Derived from the ONEP, 2013.g Derived from Sutabutr, 2013.

39S. Jenjariyakosoln et al. / Energy for Sustainable Development 23 (2014) 32–45

Labor cost. The number of laborers in the sugar mill for Scenario S2 did not increase despite the increase in number of operating days as they arealready employed all year round for maintenance, etc. For Scenario 3, the operator requirement for new power plant was estimated by Eq. (6)(Delivand et al., 2011b). In addition, the number of laborers was allocated to excess electricity.

Number of laborers ¼ 13:761 MWeð Þ0:4328: ð6Þ

For this estimation, three shifts were considered along with the total annual labor and operator requirements, consisting of the numbers ofordinary workers, skilled workers, engineers, supervisors, managers, and office staff. The average annual income and annual bonus per personwere assumed at 12,796 USD based on PDTI (2011).

Income conditionsIn the three scenarios, income was generated in two ways, surplus electricity and surplus steam, with several different prices. The purchasing

price of 0.09 USD/kWh was assumed based on the firm electricity tariff regulations of the Thailand Energy Policy and Planning Office (EPPO) for

S3 -

20

40

60

80

100

120

140

S1 S2

20 bar

30 bar

40 bar

70 bar

103 bar

average

kWh/tc

Elec

tric

ityex

port

pot

en�a

l (kW

h/tc

)

Fig. 3. Comparison of electricity export per ton of sugarcane for S1, S2 and S3.

Table 9Gross generated electricity and electricity production export for S3.

Items Unit S3

103 bar (2300 tcd) 103 bar (40,000 tcd)

Gross generated electricity MWh/y 57,162 941,538Own consumption MWh/y 16,596 273,416Electricity export into the grid MWh/y 39,671 653,451Electricity export to the ethanol plant MWh/y 895 14,745Steam export to the sugar mill t/y 132,212 2,163,426Steam export to the ethanol plant t/y 11,669 192,213Total bagasse demand t/y 83,194 1,370,346Bagasse demand for electricity export into the grid t/y 57,737 951,032Bagasse demand for electricity export for ethanol plant t/y 1303 21,460Total tops and leaves demand t/y 13,721 226,001Tops and leaves demand for electricity export into the national grid t/y 9521 156,828Tops and leaves demand for electricity export for the ethanol plant t/y 215 3539Operating time, (24 h/day) Days/y 300 300

40 S. Jenjariyakosoln et al. / Energy for Sustainable Development 23 (2014) 32–45

all surplus electricity export. The adder premium of 0.01 USD/kWh for first 7 years was assumed for surplus electricity supplying into the nationalgrid (Sutabutr, 2013). The steam selling rate was assumed as 2.08 USD/t for all sugar mills and some ethanol plant (ONEP, 2013).

Evaluation of the economics of excess electricity productionEconomic criteria for this analysis comprised production cost, NPV (net present value), and IRR (internal rate of return). Themethods considered

the most general scheme, and have been frequently used in many technical-economic analyses (Delivand et al., 2011b; Moomaw et al., 2011;Siemers, 2010). A discounted future stream of net incomes was used by applying 10% project-specific discount rate to evaluate the desirability ofthe project fiscal measures. This is the highest rate including the risk premium reflecting a longer time perspective for investments from threesuggested values (i = 3, 7 and 10%) of the IPCC special report on renewable energy sources and climate change mitigation for the cost evaluationof biomass power plants (Moomaw et al., 2011).

The production cost (USD/MWh) is the ratio of the total life time costs including the value of the capital, fuel costs, and operating costs to thelife-time net delivered electricity (to the grid) (Moomaw et al., 2011). The production cost model can be determined through following Eq. (7).

Production cost USD=MWhð Þ ¼

Xnt¼0

Costt= 1þ rð Þt

Xnt¼0

Electricity export= 1þ rð Þtð7Þ

where Costt (USD) is the cost in the year t, n (year) is the project life-time (20 years), and r is 10% discount rate.

Table 10Quantities of products for S1: electricity export using pure surplus bagasse as fuel (from 100 tc), including the economic allocation factor.

Unit Bagasse power generating systems

Group 2 Group 3 Group 4 Group 5 Group 6

20 bar 30 bar 40 bar 70 bar 103 bar

Flow of bagasse within furnaceBagassea t 28.00 28.00 28.00 28.00 28.00

Quantities of productsRaw sugarb t 10.45 10.45 10.45 10.45 10.45Molasses (to animal feed)b t 4.48 4.48 4.48Molasses (to ethanol)b t 4.48 4.48Electricity exporta MWh 0.51 1.63 2.26 6.93 8.88Electricity export (to ethanol)a MWh 0.28 0.28LP steam to ethanola t 3.65 3.65

Economic allocation factorsRaw sugarc % 92.32 90.83 90.00 83.91 81.78Molasses (to animal feed)c % 6.92 6.81 6.75Molasses (to ethanol)c % 6.29 6.13Electricity exportc % 0.76 2.36 3.25 9.68 11.98LP steam to ethanolc % 0.11 0.11

GHG emissionElectricity export kg CO2e/MWh 4.64 4.57 4.53 4.22 4.11Electricity exportd tCO2e 1186 841 1061 2507 5531

a Based on the energy balance of each representative sugar mill.b Derived from the average of sugar and molasses products of sugarcane production statistics in 2011/12 (OCSM, 2013).c Based on the average economic values in 2012–2013: raw sugar 533.86 USD/t (OAE, 2012), molasses 93.35 USD/t (BOT, 2012), electricity price 0.09 USD/kWh (Sutabutr, 2013), and

internal steam price 2.06 USD/t (ONEP, 2013).d For each group, the GHG emissions from electricity export are calculated for all the sugar mills within the group.

Table 11Quantities of products for S2: electricity export after expanding operating times and using tops and leaves as a secondary fuel (from 100 tc), including the economic allocation factor.

Unit Power generating systems of sugarcane residues

Group 2 Group 3 Group 4 Group 5 Group 6

20 bar 30 bar 40 bar 70 bar 103 bar

Flow of sugarcane residues within furnaceBagassea t 28.00 28.00 28.00 28.00 28.00Tops and leavesa t 6.03 3.29 2.60 3.66 4.29

Quantities of productsRaw sugarb t 10.45 10.45 10.45 10.45 10.45Molasses (to animal feed)b t 4.48 4.48 4.48Molasses (to ethanol)b t 4.48 4.48Electricity exporta MWh 0.65 1.77 2.78 8.60 12.31Electricity export (to ethanol) MWh 0.28 0.28LP steam to ethanol t 3.65 3.65

Economic allocation factorsRaw sugarc % 92.12 90.62 89.30 82.01 78.18Molasses (to animal feed)c % 6.91 6.79 6.69Molasses (to ethanol)c % 6.15 5.86Electricity exportc % 0.97 2.59 4.01 11.73 15.96LP steam to ethanolc % 0.11 0.11

GHG emissionTops and leaves supply chain processd kg CO2e/t 12.53 12.53 12.53 12.53 12.53Electricity export kg CO2e/MWh 7.81 6.15 5.77 5.77 5.76Electricity exporte tCO2e 2543 1233 1668 4249 10,739

a Own calculation based on the energy balance of each representative sugar mill.b Derived from the average of sugar and molasses products of sugarcane production statistics in 2011/12 (OCSM, 2013).c Based on the average economic values of the references in Table 10.d Based on the estimated GHG emission amount of tops and leaves supply chains process of the average daily sugar mill capacity in Thailand (16,200 tcd).e For each group, the GHG emissions from electricity export are calculated for all the sugar mills within the group.

41S. Jenjariyakosoln et al. / Energy for Sustainable Development 23 (2014) 32–45

The Thailand Board of Investment (BOI) provides a policy to support the renewable energy investment by offering the application of a privilegecorporate income tax exemption. Based on that, the tax payment was assumed as follows: 0% tax rate for first 8 years, 15% for the next 5 years, and30% for the remaining years (Asawachintachit, 2012).

NPV is the sum of the initial investment and the present value of all future cash flows at a particular discount rate, and it is calculated throughfollowing equation (Chau et al., 2009).

NPV ¼ −I0 þXnt¼1

Ct

1þ rð Þn ð8Þ

where I0 (USD) is the initial investment (new constructing investment of power plant), Ct (USD) is the cash flow in period t (year), n (year) is theproject life-time (20 years), and r is discount rate of 10%.

Table 12Quantities of products fromminimum andmaximum sugar mill capacity (tcd) of S3: upgrading power generation configurations as 103-bar and 515 °C and using tops and leaves as thesecondary fuel (from 100 tc), including the economic allocation factor.

Unit Power generating systems of sugarcane residues

(2300 tcd) (40,000 tcd)

Flow of sugarcane residues within furnaceBagassea t 28.00 28.00Tops and leavesa t 4.29 4.29

Quantities of productsRaw sugarb t 10.45 10.45Molasses (to ethanol)b t 4.48 4.48Electricity exporta MWh 12.31 12.31Electricity export (to ethanol) MWh 0.28 0.28LP steam to ethanol t 3.65 3.65

Economic allocation factorsRaw sugarc % 78.26 78.26Molasses (to ethanol)c % 5.87 5.87Electricity exportc % 15.76 15.76LP steam to ethanolc % 0.11 0.11

GHG emissionTops and leaves supply chain processd kg CO2e/t 10.06 14.75Electricity export kg CO2e/MWh 5.59 5.85Electricity export tCO2e – 68,535

a Own calculation based on the energy balance of each representative sugar mill.b Derived from the average of sugar and molasses products of sugarcane production statistics in 2011/12 (OCSM, 2013).c Based on the average economic values of the references in Table 10.d Based on the estimated GHG emission amount of tops and leaves supply chain process of smallest and largest scales of daily sugar mill capacities.

42 S. Jenjariyakosoln et al. / Energy for Sustainable Development 23 (2014) 32–45

IRR was used to decide whether or not a project would be feasible for investment towards a new power plant. Project IRR is frequently used formaking investment decisions; it is a specific rate calculated by the sum of cash flows after tax in which the NPV (net present value) of the project iszero. A project can be economically feasiblewhen the IRR is higher than the accepted required rate of return for investors. The accepted IRR should beabove 11% as theminimum required rate of return for private investors in the case of Thailand as suggested by the Department of Alternative EnergyDevelopment and Efficiency (Delivand et al., 2011b).

Results and discussion

Power export potential from sugarcane residues to the national grid

Scenario 1Table 1 provides the summary of cogeneration technologies in

Thailand. The data show a significant increase in export potential ofsurplus electricity generated under the new technology of extractioncondensing turbines. Even with the older technology, increasing pres-sure and temperature contributes to an increased export potential.The electricity export and operating time of each group in S1 areshown in Table 2.

The operating times of the first three configurations i.e. sugar millmodels of 20 bar, 30 bar, and 40 bar were based on their current opera-tion using bagasse fuel. The electricity from sugar mills in Thailand iscurrently being generated mainly from bagasse fuel (EPPO, 2010b).The numbers of operating days of sugar mill models of 70 bar, and103 bar were estimated based on their bagasse availability with the ac-tual conditions of heat and power generation and consumption. Theelectricity export of each sugar mill model in each group is shown inTable 1. The 28 sugar mills with 20 bar boilers, covering 51% of the sug-arcane processing, could deliver approximately 5 kWh/tc of electricityto the grid. Themills with 30 and 40 bar boilers, comprising 22% of sug-arcane processing, could deliver 16–23 kWh/tc of electricity to the grid.The high electricity export potential group of 70 bar and 103 bar steampressure representing 24% of sugarcane processing was 69–89 kWh/tc.The average electricity exported by sugar mills in Thailand increasedfrom 14.5 kWh/tc in 2006 to approximately 26.67 kWh/tc in 2012.

Scenario 2The different increases of electricity export potential for each boiler

pressure in S2 are shown in Table 2 and Fig. 3. S2, which increases elec-tricity export by expanding operating time to 300 operating days andusing tops and leaves as the secondary fuel, could export about34.86 kWh/tc, or 3416 GWh/y. About 23% of the generated tops andleaves were sufficient as the secondary fuel to generate heat andpower generation for sugarmill use, and for electricity export. The elec-tricity export of Thailand was increased by approximately 802 GWh/yor 31% compared to the former systems characterized by less efficientlow pressure boilers with back pressure turbines.

Scenario 3The electricity export potentials of the two sugar mill capacities

(2300 and 40,000 tcd) in S3 are shown in Table 9. Fig. 3 confirms thatthe electricity exported in S3 at as high as 123 kWh/tc could be achievedvia an improvement in power generation potential from upgrading allboiler and system configurations to 103 bar and 515 °C with extractioncondensing steam turbine. The tops and leaves fuel required for heat

Table 13Estimated machinery ownership and operating costs for handling tops and leaves.

Machine Ownership & operatin

(USD/t)

Big rectangular baler 4.60Tractor, PTO 84 hp 0.85Crab loader, PTO 70 hp (hauling and loading) 0.47Crab loader, PTO 70 hp (stacking 50% of tops and leaves fuel) 0.24Shredder 3.99

and power generation of sugar mills in this case is only 19% ofthe generated amount. The electricity export was approximately11,715 GWh/y, increasing by 9102 GWh/y from the currentsituation.

GHG reduction potential of utilizing the surplus bagasse, and tops andleaves for power production

The co-products of surplus electricity, and molasses generated insugar mills are significant in economic value (Ramjeawon, 2008). Theissue of surplus steam and electricity from the power plant being usedat the ethanol production plant in sugar mills of 70 bar and 103 barmodels in Thailand as aforementioned in Description of powerproduction from bagasse in Thailand section was also investigated.The GHG emissions of the biomass supply chain for tops and leaves forS2 were based on the average sugarcane crushing capacity of16,200 tcd whereas for S3 were based on 2300 tcd and 40,000 tcd. Theamount of feedstock requirement for increasing electricity export byexpanding operating time was evaluated by electric efficiency, and theaverage lower heating value (LHV) of bagasse, and tops and leaves(shown in Table 3). The quantities of generated products and the resultsof the economic allocation factors of each of the groups in each scenarioare shown in Tables 10, 11, and 12. The life cycle GHG emissions forpurely bagasse-derived electricity are estimated at 4.11–4.64 kg CO2e/MWh whereas those from mixed residues (bagasse combined withtops and leaves) are 5.75–7.81 kg CO2e/MWh (Table 11). On the otherhand, GHG emissions from natural gas combined cycle power plantsare about 540 kg CO2e/MWh (Phumpradab et al., 2009), two orders ofmagnitude higher than those from electricity generation from sugar-cane residues.

The results of GHG emission reduction from the electricity exportgenerated by sugarcane residues replacing the electricity generated bynatural gas combined cycle power plants in S2 and S3 are 423 ktCO2e,and 4853 ktCO2e respectively. Obviously, the more biomass-basedpower generated, the greater the reduction in GHG emissions ascompared to conventional fossil-based power.

Economic analysis results

The cost considerationwas separated into two categories i.e. costs offuel and power production using improved technology. The fuel cost isthe cost of feedstock and of biomass supply chain process varyingwith the amount of biomass in that area and distance of transportation,specific for tops and leaves. For bagasse, which is generated in the millitself, only the cost of feedstock is considered. The difference in thecost of power production is not only from the cost of main equipment(boiler and turbine) but also the cost of operation and fuel.

g cost Fuel and lubrication cost Labor cost Twine cost Total

(USD/t) (USD/t) (USD/t) (USD/t)

0.55 5.151.94 0.25 3.040.65 0.17 1.290.33 0.08 0.65

0.40 4.39

17.00

17.50

18.00

18.50

19.00

19.50

20.00

20.50

21.00

21.50

22.00

2,300 5,000 10,000 15,000 20,00 0 25,000 30,000 35,000 40,000

Cost

s of t

ops a

nd le

aves

(USD

/t)

Sugarmill capacities (ton sugarcane per day)

21 km

37km48 km

61 km

15 km

30 km

43km

53km 57 km

Fig. 4. Trend in tops and leaves fuel costs with various sugar mill capacities and travel-distances.

43S. Jenjariyakosoln et al. / Energy for Sustainable Development 23 (2014) 32–45

Fuel costsThe fuel cost of tops and leaves included a feedstock price of

1.6 USD/t at the field which is paid to the farmers and the costs of bio-mass supply chain process. The biomass supply chain model was devel-oped for delivering sugarcane to sugarmills across the entire range from2300 to 40,000 tcd. This range is the current smallest and largest actualdaily crushed sugarcane amount at sugarmills in Thailand during 2011/12. The results of estimation based on Tables 4–7 and methodology ofASAE standards are shown in Table 13. The costs included specific own-ership and operating costs, and twine cost of baler for handling tops andleaves in thefield, at the storage site, and for shredding. The specific costof biomass supply chain excluding the transportation cost would be14.5 USD/t. The cost was mainly from big rectangular baler (35%), andshredder (30%). Moreover, the increase of tops and leaves fuel costshown in Fig. 4 was from a minimum of 18.8 USD/t for 2300 tcd to21.5 USD/t for 40,000 tcd of sugar mill capacity. The reason for thecost variations in the two capacities of sugar mills was the variation ofthe estimated round radius distance from 15 km for the smallest scaleof sugar mill capacity to a maximum of 61 km for the largest scale. Ifwe assume daily sugarcane capacity of 5000, 10,000, 20,000, and40,000 tcd, the specific biomass supply chain costs would be 19.2,

Table 14Specific production costs and NPV values for S1 and S2.

Items Unit S1

Group 2 Group 3 Group

20 bar 30 bar 40 bar

1. Operating costOperating and maintenance cost USD/MWh 5.31 6.04 6.99Insurance cost USD/MWh 1.99 2.26 2.62Total operating cost 7.30 8.30 9.61

2. Fuel cost USD/MWhBagasse fuel cost USD/MWh 40.32 34.14 27.33Hypothetical price of tops and leavesFuel cost (@ 1.6 USD/t of feedstock) USD/MWhFuel cost (@ 3.2 USD/t of feedstock) USD/MWhFuel cost (@ 8.0 USD/t of feedstock) USD/MWh

3. Production cost USD/MWh 47.62 43.34 36.94@ 1.6 USD/t of feedstock USD/MWh@ 3.2 USD/t of feedstock USD/MWh@ 8.0 USD/t of feedstock USD/MWh

4. NPV value 106USD 3.00 18.62 52.23@ 1.6 USD/t of feedstock 106USD@ 3.2 USD/t of feedstock 106USD@ 8.0 USD/t of feedstock 106USD

19.7, 20.5, and 21.5 USD/t, respectively. The economic analysis showsthat increasing the system capacity from 2300 tcd to 40,000 tcd, a 17.4fold increase, resulted in an increase of the specific tops and leavesfuel costs by 14.6%.

Technology improvementBased on the financial evaluation of the configurations in S1 and S2,

the economic criteria including the production cost, and NPV over thelife-time of the systems have been computed, and the results areshown in Table 14. Site survey revealed that sugar mill capacities of9600, 13,000, 24,000, 22,000 and 22,000 tcdwere equippedwith boilershaving pressures 20 bar, 30 bar, 40 bar, 70 bar and 103 bar, respectively.From the estimated tops and leaves fuel cost varying in terms of sugarmill capacities in Fuel costs section, the tops and leaves costs of 19.70,20.00, 20.60, 20.50 and 20.50 USD/t were assumed for models of20 bar and 30 bar, 40 bar, 70 bar and 103 bar respectively. Table 14shows the production costs including operating costs and fuel costs ofeach group in S1 and S2. The fuel costs of the groups in S2 were higherthan those in S1. However, the operation cost per unit of electricity(USD/MWh) of each group in S2 was significantly lower than that inS1. The range varied from 19 to 59%, (note 59% reduction of 20-bar

S2

4 Group 5 Group 6 Group 2 Group 3 Group 4 Group 5 Group 6

70 bar 103 bar 20 bar 30 bar 40 bar 70 bar 103 bar

11.33 13.04 2.18 4.35 5.60 7.38 8.004.25 4.89 0.82 1.63 2.10 2.77 3.00

15.58 17.93 3.00 5.98 7.70 10.14 11.00

18.17 16.17 27.70 27.71 23.04 14.22 11.64

16.49 8.51 5.9 5.13 4.9117.83 9.21 6.35 5.54 5.321.87 9.87 7.69 6.73 6.45

33.75 34.1047.19 42.20 36.64 29.49 27.5548.53 42.90 37.09 29.90 27.9452.57 43.56 38.43 31.09 29.09

125.40 137.563.84 20.03 64.78 161.55 204.543.77 19.83 64.40 160.68 203.493.55 19.26 63.18 158.06 200.38

Table 15Production costs and NPV values for S3.

Items Unit S3

103 bar (2300 tcd) 103 bar (40,000 tcd)

1. Capital cost USD/MWh 42.26 33.032. Operating cost

Operating and maintenance cost USD/MWh 9.44 7.89Labor cost USD/MWh 8.52 1.88Insurance cost USD/MWh 3.54 2.96Total operating cost USD/MWh 21.49 12.72

3. Fuel costBagasse USD/MWh 11.64 11.64Hypothetical price of tops and leavesFuel cost (@ 1.6 USD/t of feedstock) USD/MWh 4.51 5.16Fuel cost (@ 3.2 USD/t of feedstock) USD/MWh 4.89 5.54Fuel cost (@ 8.0 USD/t of feedstock) USD/MWh 6.05 6.70

4. Production cost@ 1.6 USD/t of feedstock USD/MWh 79.90 62.55@ 3.2 USD/t of feedstock USD/MWh 80.28 62.93@ 8.0 USD/t of feedstock USD/MWh 81.44 64.09

5. NPV value@ 1.6 USD/t of feedstock 106USD 6.94 196.23@ 3.2 USD/t of feedstock 106USD 6.81 194.37@ 8.0 USD/t of feedstock 106USD 5.98 188.90

6. IRR value@ 1.6 USD/t of feedstock % 20.00 31.26@ 3.2 USD/t of feedstock % 19.38 31.08@ 8.0 USD/t of feedstock % 18.82 30.50

44 S. Jenjariyakosoln et al. / Energy for Sustainable Development 23 (2014) 32–45

sugar mill configuration, 39% of 103 bar, and 35% of 70 bar). The resultsshould convince investors of the benefits of expanding operating timeto 300 days. The advantages achievable through reduction in produc-tion costs and increasing positive values of NPV for the groups in S2without capital investment for newer technologymust be taken note of.

Assuming that a market of tops and leaves is established, it can beexpected that the price paid to the farmers may also increase. Apartfrom the modeled base price of 1.6 USD/t, calculations were also donefor a 3.2 USD/t as well as 8 USD/t (equalling the current assumedprice of bagasse following the experience of rice straw in Thailand).The increase from 1.6 to 8 USD/t resulted in an increase of only 4–11%of the production costs and decrease in NPV by 2–8% (Table 14).

Table 16Percentage variation of the individual variables to yield at IRR = 11% for S3.

Variable (Δ% of base values) S3

103 bar (2300 tcd) 103 bar (40,000 tcd)

Fuel price, Δ% +135 +235Electricity selling price, Δ% −24.2 −43.8Capital cost, Δ% +49.5 +115Plant factor, Δ% −27.2 −49.7

Investment analysis results for S3The systems in S3 were examined for the effects of the different

scales on new investment of equipment and installation. For S3, thespecific production costs varied from a minimum of approximately63 USD/MWh for the large capacity sugar mill to a maximum of80 USD/MWh for the small scale. The production cost of biggest sugarmill (40,000 tcd) was around 22% lower than the smallest one(2300 tcd) (Table 15). The change inmarket prices of tops and leaves as-sumed earlier generates a 2–3% increase in production costs, and 4–14%decrease in NPV values (Table 15). These figures confirm the financialviability of the interventions.

The IRR values for the power plants associated with the 2300 and40,000 tcd sugar mills were 17% and 27% respectively which, beinghigher than 11%, are highly favorable. If the feed in tariff of0.01 USD/kWh (0.3 THB/kWh) offered by the government is includedin the calculations, the IRR values further increase to 20% and 31% forthe small and large systems respectively. These values clearly showthe rationalization for investments.

The sensitivity analysis of investment appraisal condition on IRR≥ 11% was considered i.e. in case of 40,000 tcd capacity of 103-barsugar mill, even if fuel price, selling price of electricity, capital cost,and the plant factor were changed by +2.35 folds (26.79 USD/t of ba-gasse fuel cost, and 72.07 USD/t of tops and leaves fuel cost), −43.80%(0.051 USD/kWh), +1.15 folds (308 Million USD) and −49.72%(3620 h/y), respectively, investment will still remain appealing

(see Table 16). The larger scale sugar mill capacity is financially morerobust and is less sensitive to all factors compared to the smaller one.

Conclusions

Increasing the number of operating days for electricity generation inexisting sugar mills using top and leaves as the supplementary fuelcould provide a potential gain of 31% in surplus electricity (8.2 kWh/tc).Additionally, using upgraded technology of boiler steam pressure of 103bar and 515 °C and extraction condensing steam turbine technologyshowed a potential of a 3.5 fold increase in surplus electricity generationfor export as compared to the current situation. For this case, the electric-ity export to the grid is 123 kWh/tc and the secondary fuel used is only19% of generated tops and leaves. The electricity export from sugarmills with state-of-the-art technology and use of bagasse and tops andleaves as fuel would be 9 TWh or 65% of biomass power target of theAEDP 2012–2021 with 4.8 Mt CO2e of GHG emission reduction or 6% ofthe AEDP 2012–2021 target. The electricity generation costs and NPVvalues show that the high boiler steam pressure configuration and thereduction in production cost can lead to higher benefits than the currentsituation. The results confirmed that despite the large investments for thestate-of-the-art technology, the IRR was still higher than 11% (the mini-mum pre-requisite rate of return). Therefore, these provide an attractiveinvestment option, especially for large sugar mill capacities. The result incost analysis of tops and leaves supply chains in different scales of sugarmill capacities (2300–40,000 tcd) also showed less sensitivity to the dif-ferent scales of sugar capacities and different transportation distances.

45S. Jenjariyakosoln et al. / Energy for Sustainable Development 23 (2014) 32–45

Acknowledgments

The authors would like to thank Dr. Bundit Fungtammasan for histechnical comments and suggestions. This Ph.D. work was financiallysupported by the Joint Graduate School of Energy and Environment(JGSEE), King Mongkut’s University of Technology Thonburi (KMUTT),Bangkok, Thailand.

References

Asawachintachit D. Thailand, a perfect place for your business to grow. Thailand: ThailandBoard of Investment; 2012 [Retrieved January 21, 2014, from: http://www.boi.go.th/upload/content/Tax%20seminar_Feb%202012_53851.pdf].

Bakos GC, Tsioliaridou E, Potolias C. Technoeconomic assessment and strategic analysis ofheat and power co-generation (CHP) from biomass in Greece. Biomass Bioenergy2008;32:558–67.

Bhatt MS, Rajkumar N. Mapping of combined heat and power systems in cane sugarindustry. Appl Therm Eng 2001;21:1707–19.

Bhattacharya SC, Abdul Salam P, Runqing H, Somashekar HI, Racelis DA, Rathnasiri PG,et al. An assessment of the potential for non-plantation biomass resources in selectedAsian countries for 2010. Biomass Bioenergy 2005;29:153–66.

Bocci E, Di Carlo A,Marcelo D. Power plant perspectives for sugarcanemills. Energy 2009;34:689–98.

BOT. Molasses price in Thailand in 2012. Bank of Thailand; 2012 [Retrieved September 01,2013, from: http://www.bot.or.th/Thai/EconomicConditions/Thai/Northeast/com-modities/Doclib_CommodityMonthly/Ethanol%20Monthly–01-56.pdf].

Chau J, Sowlati T, Sokhansanj S, Preto F, Melin S, Bi X. Techno-economic analysis of woodbiomass boilers for the greenhouse industry. Appl Energ 2009;86:364–71.

DEDE. The assessment of agricultural residues availability for energy resources in Thailand.Department of Alternative Energy Development and Efficiency; 2005 [RetrievedMay 20, 2013, from: http://e-lib.dede.go.th/mm-data/Bib10630-1.pdf].

DEDE. Co-generation biomass. Department of Alternative Energy Developmentand Efficiency; 2008 [Retrieved May 20, 2013, from: http://e-lib.dede.go.th/mm-data/Bib11329.pdf].

DEDE. The Renewable and Alternative Energy Development Plan for 25 percent in10 years (AEDP 2012–2021). Thailand: Department of Alternative Energy Develop-ment and Efficiency; 2012 [Retrieved December 05, 2012, from: http://www.dede.go.th/dede/images/stories/dede_aedp_2012_2021.pdf].

DEDP. The study of behavior and patterns of energy use in plantation: summary report.Thailand: Department of Energy Development and Promotion; 1992.

Deepchand K. Commercial scale cogeneration of bagasse energy inMauritius. Energy Sus-tain Dev 2001;1:15–21.

Deepchand K. Promoting equity in large-scale renewable energy development: the caseof Mauritius. Energy Policy 2002;30:1129–42.

Delivand MK, Barz M, Gheewala SH. Logistics cost analysis of rice straw for biomasspower generation in Thailand. Energy 2011a;36:1435–41.

Delivand MK, Barz M, Gheewala SH, Sajjakulnukit B. Economic feasibility assessment ofrice straw utilization for electricity generating through combustion in Thailand.Appl Energ 2011b;88:3651–8.

EFE. Biomass analysis values. Bangkok, Thailand: Energy for Environment Foundation;2006 [Retrieved November 11, 2007, from: http://www.efe.or.th].

EPPO. Study of a small-scale biomass power plant for rural communities (Phase II).Thailand: Energy Planning and Policy Office, Ministry of Energy; 2010a.

EPPO. Power Plant Community Development Fund: statistics of monthly fee of biomasspower producers paid into the fund based on the amount of monthly electricity ex-port into the grid. Thailand: Energy Planning and Policy Office, Ministry of Energy;2010b [Retrieved January 11, 2013, from: http://www.eppo.go.th/cdf/Document/Fund%20Income.xls].

EPPO. Power generation by type of fuel. Thailand: Energy Planning and Policy Office, Min-istry of Energy; 2012 [Retrieved August 18, 2013, from http://www.eppo.go.th/info/5electricity_stat.htm].

EPPO. H-diesel price statistics of January–July 2013. Thailand: Energy Planning and PolicyOffice, Ministry of Energy; 2013 [Retrieved August 18, 2013, from: http://www.eppo.go.th/index-E.html].

GEMIS 4.8. Global Emission Model Integrated System Version 4.8. 2013. Oko-Institute forapplied Ecology; 2013 [Retrieved August 18, 2013, from: http://www.iinas.org/gemis.html].

Gheewala SH, Bonnet S, Prueksakorn K, Nilsalab P. Sustainability assessment of a biorefinery complex in Thailand. Sustainability 2011;3:518–30.

Guzman PL, Valdes A. Heat and power cogeneration at a Cuban sugar mill based onbagasse and trash as fuel: the “Hector Molina” project. Energy Sustain Dev2000;4:90–2.

Hassuani SJ, Leal MRLV, Macedo IC. Biomass power generation — sugarcane bagasse andtrash; 2005. p. 57–63 [Brazil].

HuismanW, Venturi P, Molenaar J. Cost of supply chains ofMiscanthus giganteus. Ind CropProd 1997;6:353–66.

IEA. Renewable energy policy considerations for deploying renewable. Paris: Organizationfor Economic Co-operation and Development/International Energy Agency; 2011.

IPCC. Revised 1996 IPCC Guidelines for National Greenhouse Gas Inventories. The Inter-governmental Panel on Climate Change; 1996.

IPCC. IPCC fourth assessment report (AR4) — climate change 2007. IntergovernmentalPanel on Climate Change; 2007.

Junginger M, Faaij A, Broek VDR, Koopmans A, Hulscher W. Fuel supply strategies forlarge-scale bio-energy projects in developing countries. Electricity generation fromagricultural and forest residues in Northeastern Thailand. Biomass Bioenergy 2001;21:259–75.

Khatiwada D, Seabra J, Silveira S, Walter A. Power generation from sugarcane biomass— a complementary option to hydroelectricity in Nepal and Brazil. Energy 2012:1–14.

KMUTT. Assessment of potential amount solid-biomass from agricultural residues andagro-residues from sawed timber, furniture and short rotation tree for heat andpower generation in Thailand. Thailand: King Mongkut's University of TechnologyThonburi; 2006. p. 2–42.

Koopmans A, Koppejan J. Agricultural and forest residues-generation, utilization andavailability. FAO; 1998 [Retrieved November 22, 2012, http://wgbis.ces.iisc.ernet.in/energy/HC270799/RWEDP/acrobat/p_residues.pdf accessed November 10,2010].

Larson ED, Williams RH, Leal MRLV. A review of biomass integrated-gasifier/gas turbinecombined cycle technology and its application in sugarcane industries, with an analysisfor Cuba. Energy Sustain Dev 2001;1:54–76.

Macedo IC, Leal MRLV, Hassuani SJ. Sugar cane residues for power generation in thesugar/ethanol mills in Brazil. Energy Sustainable Dev 2001;1:77–82.

Mendoza TC, Samson R, Elepano AR. Renewable biomass fuel as “Green Power” alterna-tive for sugarcane milling the Philippines. Philipp J Crop Sci 2002;27:23–39.

Moomaw W, Burgherr P, Heath G, Lenzen M, Nyboer J, Verbruggen A. Annex II:methodology. In: Edenhofer O, Pichs-Madruga R, Sokona Y, Seyboth K,Matschoss P, Kadner S, Zwickel T, Eickemeier P, Hansen G, Schlömer S, vonStechow C, editors. IPCC Special Report on Renewable Energy Sources andClimate Change Mitigation. United Kingdom and New York, NY, USA:Cambridge University Press, Cambridge; 2011.

National Greenhouse Gas Inventory Committee. Australian methodology for the estima-tion of greenhouse gas emissions and sinks 2006. Canberra: Energy (stationarysources) Australian Government Department of Climate Change; 2007.

Nguyen TLT, Gheewala SH. Life cycle assessment of fuel ethanol from cane molasses inThailand. Int J Life Cycle Assess 2008;13:301–11.

OAE. Basic information of agricultural crops in Thailand in 2009/10–2011/12.Thailand: Office of Agricultural Economics, Ministry for Agricultural and Cooper-atives; 2012.

OCSB. In-depth study energy efficiency improvement for sugar industries. Faculty ofEngineering Chulalongkorn University, Bangkok: Office of the Cane and SugarBoard, Ministry of Industry, Thailand; 2007a.

OCSB. The value added for sugarcane and sugar industry. Thailand: Office of the Cane andSugar Board. Ministry of Industry; 2007b.

OCSM. Statistics of sugarcane production in 2002/03–2012/13. Thailand: Office of theCane and Sugar Management. Ministry of Industry; 2013.

ONEP. Environmental impact assessment reports. Thailand: Office of Natural Resourcesand Environmental Policy and Planning; 2013.

Painter K. Costs of owning and operating farmmachinery in the Pacific Northwest. U.S.: APacific Northwest Extension Publication University of Idaho, Washington StateUniversity, Oregon State University; 2011

PDTI. Assessment of energy efficiency development from heat and power gener-ation in sugar factories in Thailand. Thailand: The Pilot Plant Developmentand Training Institute (PDTI), King Mongkut's University of TechnologyThonburi; 2011.

Phumpradab K, Gheewala SH, Sagisaka M. Life cycle assessment of natural gas powerplants in Thailand. Int J Life Cycle Assess 2009;14:354–63.

Ramjeawon T. Life cycle assessment of electricity generation from bagasse in Mauritius. JClean Prod 2008;16:1727–34.

Renouf MA, Pagan RJ, Wegener MK. Life cycle assessment of Australian sugarcaneproducts with a focus on cane processing. Int J Life Cycle Assess 2011;16:125–37.

Rentizelas AA, Tolis AJ, Tatsiopoulos IP. Logistics issues of biomass: the storage problemand the multi-biomass supply chain. Renew Sust Energ Rev 2009;13:887–94.

Sajjakulnukit B, Yingyuad R, Maneekhao V, Pongnarintasut V, Bhattacharya SC, AbdulSalam P. Assessment of sustainable energy potential of non-plantation biomassresources in Thailand. Biomass Bioenergy 2005;29:214–24.

Siemers W. Greenhouse gas balance for electricity production from biomass resources inThailand. Sust Energ Environ 2010;1:65–70.

Sutabutr T. Business opportunities in Thailand's renewable energy, ASEAN clean energytrade, technology and investment forum. Manila, The Philippines: USTDA-ASEAN-BCIU; 2010. p. 19–21 [April].

Sutabutr T. Updated renewable energy policies: case of Thailand. 2013 International Con-ference on Alternative Energy in Development Countries and Emerging Economies.Bangkok, Thailand: AEDCEE; 2013. [May].

Tossanaitada W. Cogeneration efficiency enhancement in sugar mills [MD Dissertation]Bangkok, Thailand: King Mongkut's University of Technology Thonburi; 2008.

UNFCCC. CDM Project documents, 2014. United Nations Framework Convention onClimate Change. Retrieved June 13, 2014, from: http://cdm.unfccc.int/Projects/projsearch.html.

US.EIA. Overview Thailand is a net importer of oil and natural gas, although the country isa grower producer of natural gas. Retrieved February 25, 2014 http://www.eia.gov/countries/analysisbriefs/Thailand/thailand.pdf, 2013.

WEC. World energy resources. Retrieved June 12, 2014 http://www.worldenergy.org/wp-content/uploads/2013/10/WEC_Resources_summary-final_180314_TT.pdf, 2013.

YuttithamM, Gheewala SH, Chidthaisong A. Carbon footprint of sugarcane produced fromsugarcane in eastern Thailand. J Clean Prod 2011;19:2119–27.