Life cycle assessment of coal based direct-reduced iron production in India

(1)

Master of Science Thesis

KTH School of Industrial Engineering and Management Energy Technology EGI-2018-2020

Division of Energy Systems SE-100 44 STOCKHOLM

Life cycle assessment of coal based direct-reduced iron production in

India

Sabarish Elango

(2)

-2-

Master of Science Thesis EGI 2018:2020 TRITA-ITM-EX 2020:509

Life cycle assessment of coal based direct- reduced iron production in India

Sabarish Elango

Approved

October 2020

Examiner

Dilip Khatiwada

Supervisor

Dilip Khatiwada

Commissioner Contact person

Deepak Yadav Tirtha Biswas

Abstract

The production of iron through direct reduction (Direct-Reduced Iron; DRI) involves the use of natural gas or coal to reduce iron ore to iron through carbothermic reactions at a temperature below its melting point, negating the need for a blast furnace as otherwise required. In India, around 25% of iron is produced through direct reduction. However, there is a high reliance on coal (79% of DRI production capacity) causing significant energy use and emissions from production. Also, a large portion of raw materials (especially coal) is imported due to low quality of domestic resources. To understand the overall supply chain impact of DRI or sponge iron production, a life-cycle assessment is carried out on four stages:

mining, transport, beneficiation (i.e. ore pellet-making and coal washing) and DRI production. Around 315 coal-based DRI production plants are mapped using GIS to identify clustering of plants in major iron- producing regions. Weighted average specific energy use and emissions is calculated for seven such clusters (using total cluster capacity), based on regional raw material qualities and transport distances from various mines, ports and beneficiation plants. The results suggest an overall specific (per tonne DRI) energy consumption of 27.24 GJ with an emission of 2.8 tCO2eq, 2.6 kgNOx, 1.8 kgSOx and 1.4 kgPM2.5. The specific energy and emission values are used to calculate the total annual emissions by multiplying with the 2019 DRI production amount of 27.8 million tonnes. The annual midpoint and endpoint impacts as per ReCiPe 2016 (country-wise factors where applicable) are then calculated. The DRI industry causes 77.31 million tCO2eq/year in global warming potential, 59.02 thousand tSO2eq/year in acidification potential and 287.2 thousand tPM2.5eq/year in fine dust formation potential. It is estimated to cause approximately 270,000 years of reduction in overall human life and 230 species years of species loss (mainly in terrestrial ecosystems). Different sensitivities are carried out to understand the impact of some key influencing parameters (effect of ore quality and coal quality, effect of imports of ore and coal). Some development scenarios, such as increasing coal washery capacity, shifting land transport from road to rail, increasing waste-heat recovery penetration, effect of stricter regulations, etc. are discussed, along with pathways for fuel-switching from coal to natural gas, and then from natural gas to hydrogen.

(3)

-3-

Abstrakt

Järnsvamp producerat genom direktreduktion (Direct-Reduced Iron; DRI) innefattar användning av naturgas eller kol för reduktion av järnmalm till järn genom karboterma reaktioner vid en temperatur under dess smältpunkt, vilket eliminerar behovet av en masugn. Ca 25% av järnproduktionen i Indien sker genom direkt reduktion. Energin kommer i 73% av fallen från kol, vilket gör processen är väldigt kolberoende. I sin tur bidrar detta till en betydande energianvändning och stora koldioxidutsläpp. Låg kvalité på inhemska resurser (framförallt kol) kräver import av en stor del av processråvarorna. För att förstå den övergripande effekten av försörjningskedjan av DRI- eller svampjärnproduktionen genomfördes en livscykelbedömning i fyra steg - gruvdrift, transport, anrikning (dvs tillverkning av malmpellets och koltvätt) och DRI- produktion. Cirka 315 kolbaserade DRI-produktionsanläggningar kartlades med GIS för att identifiera kluster av produktionsanläggningar i större järnproducerande regioner. Vägt genomsnittligt specifikt energianvändning och utsläpp beräknades för sju sådana kluster (med total klusterkapacitet) baserat på regionala råvarukvaliteter och transportavstånd från olika gruvor, hamnar och förädlingsanläggningar.

Resultaten indikerar en total specifik energiförbrukning på 27,24 GJ/ton DRI med ett utsläpp på 2,8 tCO2eq, 2,6 kg NOx, 1,8 kg SOx och 1,4 kg PM2,5 per ton DRI. De specifika energi- och utsläppsvärdena användes för att beräkna de totala årliga utsläppen genom att multiplicera med DRI-produktionsmängden för 2019, vilken motsvarade 27,8 miljoner ton. Den årliga mitt- och slutpunktseffekten enligt ReCiPe 2016 (landsmässiga faktorer där så är tillämpliga) beräknades sedan. DRI-industrin har en global uppvärmningspotential motsvarande 77,31 miljoner tCO2eq/år, 59,02 tusen tSO2eq/år i försurningspotential och 287,2 tusen tPM2,5eq/år i potential för bildning av fint damm. Detta beräknas förkorta livslängden hos människor världen över med total cirka 270 000 år och 230 år av artförlust (främst i markbundna ekosystem). En känslighetsanalys utfördes för att förstå effekterna av viktiga påverkande parametrar (effekt av malmkvalitet och kolkvalitet, effekt av import av malm och kol). Vissa utvecklingsscenarier, såsom ökad koltvättkapacitet, skifte i transport från väg till järnväg, ökad penetration av spillvärmeåtervinning, effekten av strängare föreskrifter etcetera diskuterades tillsammans med tillvägagångssätt för bränsleomkoppling från kol till naturgas och sedan från naturgas till väte.

(4)

-4-

Acknowledgment

I extend my sincere gratitude to my KTH supervisor, Dr. Dilip Khatiwada for the patience and support extended to me over the period of my work. His guidance regarding the methodology for the life-cycle assessment and the structuring of the study has been indispensable. He played a major role in ensuring the timely progress of work and correctness of the data and calculations used.

I convey my heartfelt thanks to my internship supervisors, Dr. Deepak Yadav and Mr. Tirtha Biswas at the Council on Energy, Environment and Water (CEEW) in New Delhi, India, for making this work possible.

Their joint efforts and support extended to me during my internship at CEEW have provided immeasurable value to my thesis work. Such extensive data collection would have been difficult without their support.

Their suggestions and comments have helped provide narrative and structure to the work overall. Their efforts in reading and checking the model have been useful to ensure the validity of the data, assumptions and calculations used in the study.

I express my gratitude to CEEW, its administrators and staff, for allowing me to develop this work in parallel with my internship responsibilities, and for providing a warm and welcoming atmosphere over the course of the six months of my work.

(5)

-5-

List of Figures

Figure 1a: Locations of various DRI plants and facilities in India... 8

Figure 1b: Locations of various coal-DRI plants in India ... 8

Figure 2: Rotary kiln (SL/RN process) ... 9

Figure 3: LCA phases according to ISO14040:2006 ... 11

Figure 4: Process flowchart for LCA ... 11

Figure 5: Locations of DRI plant clusters ... 20

Figure 6: Transport schemes ... 22

Figure 7: Specific life cycle midpoint impacts ... 24

Figure 8: Effect of iron ore grade... 26

Figure 9: Effect of coal grade... 26

Figure 10: Effect of coal beneficiation share ... 27

Figure 11: Effect of ore imports ... 28

Figure 12: Effect of coal imports ... 28

Figure 13: Effect of road vs. rail for land transport ... 29

Figure 14: Effect of WHR share ... 29

List of tables

Table 1: India-average grid emission factors ... 13

(6)

-6-

Table 2: Energy and emission factors for coal mining – India ... 14

Table 3: Energy and emission factors for coal mining – Indonesia ... 14

Table 4: Energy and emission factors for coal mining – South Africa ... 14

Table 5: Energy and emission factors for iron ore mining – India ... 15

Table 6: Energy and emission factors for iron ore mining – Australia ... 15

Table 7: Energy and emission factors for iron ore mining – South Africa ... 15

Table 8: Energy and emission factors for dolomite mining ... 16

Table 9: Number of trucks in operation ... 16

Table 10: Truck emission norms ... 16

Table 11: Energy and emission factors for trucks ... 17

Table 12: Energy and emission factors for railways ... 17

Table 13: Fuel-specific emission factors and fuel consumption of dry-bulk carriers ... 17

Table 14: Energy and emission factors for ships ... 18

Table 15: Energy and emission factors for coal beneficiation ... 18

Table 16: Energy and emission factors for iron ore beneficiation ... 18

Table 17: Typical raw material input quantities for DRI production ... 19

Table 18: Energy and emission factors for DRI production ... 19

Table 19: Potential for electricity from WHRB ... 19

Table 20: Potential energy and emissions offset by WHR power plant ... 20

Table 21: Capacities of different representative plants ... 21

Table 22: Assumptions of iron ore and coal quality ... 21

Table 23: Example of overall transport emissions (for Jamshedpur cluster) ... 22

Table 24: Overall specific energy use and emissions ... 23

Table 25: Midpoint and endpoint life cycle impacts ... 23

Table 26: National annual life-cycle energy use and emissions (with and without WHR) ... 25

Table 27: National annual life-cycle impacts (with and without WHR) ... 25

List of abbreviations

DALY disaility-adjusted life years DRI direct-reduced iron LCA life cycle assessment LCI life cycle inventory

LCIA life cycle impact assessment GWP global warming potential AP acidification potential FP formation potential GCV gross calorific value VM volatile matter TPD tonnes per day WHR waste-heat recovery WHRB waste-heat recovery boiler MTPA million tonnes per annum FC fixed carbon

LNG liquefied natural gas HFO heavy fuel oil FO furnace oil

(7)

-7-

1. Introduction

On a global level, the iron and steel industry is known for its significant energy use and emissions, amounting to around 8.3% (IEA, 2019) to the total worldwide energy consumption (IEA Statistics, 2019) and 7-9% of total CO2 emissions (World Steel Association, 2019a) as of 2017. India is the 2^nd largest steel producer in the world as of 2018, after China, producing 106.5 MT (The Economic Times, 2019). India’s iron and steel industry is highly dependent on coal for its energy needs, thus causing it to be highly GHG intensive. It emits approximately 2.3 tCO2/tonne of iron (Pal, Gupta and Kapur, 2016), well above the global average of 1.83 tCO2/tonne of steel (Word Steel Association, 2019a).

India is the world's largest producer of Direct-Reduced Iron (DRI) or sponge iron, with the coal-based production method accounting for 79% of the total DRI production in the country (the remainder is gas- based) in 2018 (Indian Bureau of Mines, 2019). The DRI industry in India has seen enormous growth, with installed capacity increasing from 1.52 million tonnes per annum (MTPA) in 1990-91 to around 49.62 MTPA in 2017-18. Production of sponge iron grew from 0.9 million tonnes in 1990-91 to 30.51 million tonnes in 2017-18 (Indian Bureau of Mines, 2019). The National Steel Policy of 2017 projects an increase in domestic demand for sponge iron to 80 million tonnes by 2030-31; sponge iron capacity may increase to 114 million tonnes by 2030-31 with nearly 30% share of gas-based plants (Ministry of Steel, 2017). Coal use in India is projected to grow by nearly 3% per annum from 2015-2040 due to iron and steel industry growth and relatively intensive coal use in DRI (U.S. EIA, 2017). Due to the large percentage of DRI production for ironmaking in India, there is a great potential for emissions and energy use reductions/offsets in the future. Considering that the large DRI share is contributed to mainly by small- and medium-scale producers, upscaling of the DRI process and integration with modern heat recovery, renewable energy and energy conservation techniques could likely be the way forward. Hence, the direct reduction process for sponge iron (DRI) production is the focus of the LCA study.

The study analyses the impact of national DRI production – from the mining of coal, iron ore and dolomite;

beneficiation of coal and iron ore; transportation of raw materials through road, rail and sea (a large quantity of coal and significant amounts of iron ore are imported), and the DRI production process itself. The LCA estimates the energy use and emissions. GHGs and local pollutants (SOx, NOx and PM2.5) provide important and interesting insights into the impacts caused to human life and ecosystems through global warming, acidification and particulate matter pollution.

1.1. Literature review

Published research on the iron and steel industry from the LCA perspective exist, such as the work of Olmez et al, 2016. Also, works such as by Cavaliere, 2019, detail the different GHG reductions possible in the iron and steelmaking processes. Studies specific to the DRI industry in India are important, since there are many smaller plants which do not have integrated steelmaking capabilities and are typically more inefficient than larger integrated steel plants. Recent works on the DRI industry such as those by Chatterjee, 2010, Sarangi and Sarangi, 2011, Dey et al, 2015, Kumar and Khanam, 2012 and Prasad et al, 2011 look at the DRI production process, its specific parameters and the potential for energy conservation primarily through waste heat recovery. However, a study quantifying the impact of the industry in terms of overall energy inputs and emissions (from resource extraction, transformation and consumption) has not been carried out.

A robust LCA that captures the process variability, its effect on energy use and emissions is required to understand the true opportunities and challenges over the future years of growth for the sector. A cradle- to-gate LCA is developed for this purpose, to understand the impacts of DRI production from the energy use and emissions produced from the supply chain.

1.2. Data collection

The direct reduction process of iron production is modelled based mainly the work of Sarangi and Sarangi, 2011. Indian Bureau of Mines, 2019 suggests that the Indian industry predominantly relies on coal based DRI production using the rotary kiln technology (MIDREX, 2019). Only 3 facilities use the natural gas- based process. Based on data from the Indian Bureau of Mines, 2019 and online mapping services, a

(8)

-8-

database is built for major sponge iron plants, coal mines, iron ore mines, dolomite mines, coal washeries, pellet plants and import sources (Figures 1a and 1b). These are mapped using the open source geographic information system software QGIS (QGIS, 2020) to identify clusters and spatial spread across the country.

1.3. Research objectives

The objective of the study is to understand the environmental impact of coal-based sponge iron plants across different locations in India from a life cycle perspective (cradle-to-gate). The inventory of emissions from mining of raw materials, transportation, beneficiation of raw materials and production of DRI are developed. Life cycle impact assessment methodology (RIVM, 2018) is used to quantify the environmental impact of this industry on various parameters, at midpoint and endpoint levels. The midpoint level impacts show the environmental problems (such as global warming, acidification and PM formation), while the endpoint level impacts show the impacts on higher levels such as human health and biodiversity. Overall, the study aims to improve understanding of the impact of the Indian DRI industry on national and global emissions, provide impetus for stricter policymaking and to point out gaps in public knowledge of the functional aspects of the industry.

This thesis will answer the following key research questions:

RQ1. What are the energy use and emissions (of GHG, SOx, NOx and PM2.5) from the Indian DRI industry?

RQ2. What are the environmental impacts of the Indian DRI industry on human life and ecosystems?

RQ3. What are the potential changes and improvements to reduce the impact of DRI production?

The sensitivity analysis considered the variations of different factors such as efficiency, transport modes, technologies, etc. of the product supply chains of coal, iron ore and limestone, and of the core iron production process.

1.4. Thesis structure

The thesis provides an overview of the direct reduction process (Section 2) using the rotary kiln technology and the reactions involved in the production of iron. The methodology section (Section 3) defines the goal and scope of the thesis and details the emission inventory data of various life cycle steps used for the life cycle impact assessment. Section 3.3 discusses the method used to apply the inventory data to obtain the

Figure 1b: Locations of various coal-DRI plants in India (Source: Mapped by author based on Indian

Bureau of Mines, 2018 and QGIS,2020) Figure 1a: Locations of various DRI plants and

facilities in India (Source: Mapped by author based on Indian Bureau of Mines, 2018 and QGIS, 2020)

(9)

-9-

national contribution of the DRI industry. The results are provided in Section 4 in terms of average emissions and life cycle impacts per tonne of DRI, and total emissions and life cycle impacts resulting from one year of national DRI output. A sensitivity analysis is performed in Section 4.1 to account for uncertainties in some variables considered. Section 5 discusses the results and outcomes of the thesis, and provides some suggestions to abate the impact of the DRI industry in future.

2. Overview of direct reduction process

The basic mechanism behind iron production involves two main pathways,

i. Using a blast furnace (heated using coal or natural gas) for reduction of iron ore (iron oxides) into pig iron by reaction with coke and fluxes (usually limestone) (SAIL, 2012). The molten pig iron is then converted to steel (through the steelmaking process, usually with a basic oxygen furnace) or processed and sold as such. In 2019, 46.7% of India’s steel industry utilised the blast furnace-basic oxygen furnace (BF-BOF) method (World Steel Association, 2019b).

ii. Using coal (solid or gas) or reformed natural gas to perform a direct reduction of the iron ore into Direct-Reduced Iron (DRI) or Sponge iron at high heat (but below melting point) (Sarangi and Sarangi, 2011). The sponge iron is then converted to steel (with an electric arc or electric induction furnace) or processed and sold. The share of electric induction/arc furnace processes in India constituted 53.3% in 2019 (World Steel Association, 2019b).

The SL/RN process (developed by Steel Company of Canada, Lurgi Chemie, Republic Steel Company and National Lead Corporation in 1964) forms the basis of rotary kiln technologies used in India (Sarangi and Sarangi, 2011); the process uses a rotary kiln into which iron ore pellets, non-coking coal (for reduction) and limestone/dolomite (flux) is supplied. From the other end, air and coal (for combustion) are supplied.

The resulting high temperatures (900 to 1020 °C) form a reducing atmosphere of CO which reduces the iron ores to sponge iron. The sponge iron is subsequently separated out of the remaining reaction products through magnetic separation. The kiln is inclined at an angle of ~2.5° to facilitate movement of the charge

from the feed end to the exit. The rotary motion encourages even reaction of the charge through mixing with the reducing gases (Dey et al, 2015). The basic process is shown in Figure 2.

Around a third of the kiln length is typically required for preheating the charge consisting of iron ore, coal and dolomite. The dolomite flux is added to control sulphurisation. The coal supplied along with the ore is mainly meant to produce reducing gas by reacting with atmospheric oxygen at high temperature. In this stage, the iron ore (predominantly haematite - Fe2O3) is partially reduced to ferrous oxide. After reaching the ideal reaction temperature of 900-1100 °C, the ore is reduced to metal in the latter portion of the kiln through further reduction. The following are the main reactions taking place within the kiln, at a temperature of 1067 °C (Sarangi and Sarangi, 2011).

3𝐹𝑒₂𝑂₃+ 𝐶𝑂 → 2𝐹𝑒₃𝑂₄+ 𝐶𝑂₂− 44.46 𝑘𝐽/𝑚𝑜𝑙 (1)

Figure 2: Rotary kiln (SL/RN process) (Source: Dey et al, 2015)

(10)

-10-

𝐹𝑒₃𝑂₄+ 𝐶𝑂 → 3𝐹𝑒𝑂 + 𝐶𝑂₂+ 3.07 𝑘𝐽/𝑚𝑜𝑙 (2)

𝐹𝑒𝑂 + 𝐶𝑂 → 𝐹𝑒 + 𝐶𝑂₂− 11.12 𝑘𝐽/𝑚𝑜𝑙 (3)

The CO required for the above reduction reactions is produced when fixed carbon of the feed-end coal reacts with CO2 produced by the reductions, in a perpetual, reversible reaction called Boudouard reaction.

𝐶 + 𝐶𝑂₂⇋ 2𝐶𝑂 + 167.52 𝑘𝐽/𝑚𝑜𝑙 (4)

This reaction is crucial to maintaining the reducing atmosphere and kiln temperature. The ratio of CO/(CO+CO2) depends on the temperature inside the kiln; ideally a CO concentration of ~50-60% is maintained (Dey, Prasad and Singh, 2015) to ensure optimum reduction of ore. Since the forward reaction (4) is highly endothermic, it serves to maintain kiln temperature for a regulated combustion of injection coal. By combining the above reactions, we get

2𝐹𝑒₂𝑂₃+ 3𝐶 → 4𝐹𝑒 + 3𝐶𝑂₂+ 432.52 𝑘𝐽/𝑚𝑜𝑙 (5)

Note that only one part of CO produced in (4) is used for the reduction, whereas the other part is combusted into CO2 resulting in a net output of CO2 from the kiln. Various other reactions take place due to the combustion of injection coal fixed carbon and volatiles, causing the formation of additional CO and CO2 along with H2O and CH4. The sulphur present in coal is removed by dolomite, as the CaCO3 and MgCO3 decompose into CaO and MgO to act as desulphurising agents. The addition of dolomite is crucial to control the sulphur content in the DRI (to prevent embrittlement in steel production), and also to control SOx emissions (Sarangi and Sarangi, 2011).

After the reduction process, the metal (now known as sponge iron or DRI) is separated from the remaining slag (consisting of coal char, unreacted coal, sulphurated dolomite) through magnetic separation. The product CO2 reacts further with incoming/excess coal to produce more CO. Thus, for a low ash coal with high reactivity, the reduction efficiency will be higher as the quantity of coal input would be reduced. Also, the retaining time of the ore within the kiln can be lower, thus improving output (Dey et al, 2015).

3. Methodology and data sources

A framework is developed for the life-cycle assessment based on the literature review and data collection. The LCA will be cradle-to-gate in nature (Haque and Norgate, 2015) since sponge iron is an intermediate product in steel production. The LCA will be carried out using Microsoft Excel, as per the guidelines set forth in the ISO 14044 standard. Due to the nature of the study (in associating a part of environmental impact with the investigated product), availability of data in averages, consideration of factors such as feasibility, accuracy, comprehensibility, inspiration and robustness have caused the chosen LCA type as attributional (Ekvall, 2019). The four stages of the LCA applied in this work as per ISO 14044 include,

i. determination of the goal and scope ii. inventory analysis of inputs and outputs iii. assessment of environmental impact iv. interpretation of results

3.1. Goal and scope

The research objective is to calculate the environmental impact of energy use and emissions from coal- based sponge iron plants across different locations in India from a life cycle perspective (cradle-to-gate).

Various sensitivity parameters can help address the uncertainties, such as process efficiency, share of import, etc. Additionally, potential avenues for reductions in energy and emissions intensity through various policy instruments and technological developments are explored. The study will fill a lacuna in understanding of the life cycle impacts of the Indian DRI industry. The results of the study will aid Figure 3: LCA phases according to

ISO14040:2006

(11)

-11-

policymaking for environmental and emissions controls in the DRI industry and will provide impetus for transformation of the DRI industry towards decarbonisation. The scope of the LCA is cradle-to-gate in nature, as DRI is not the final product in the value chain. The functional unit (FU) for the study is taken as the production of one tonne of sponge iron (DRI).

The study will assess the production of DRI in India at a specific level and at a national level. The product system (shown in Figure 4) includes the mining of resources (coal, iron ore and dolomite), transportation (by road, rail and sea), beneficiation of resources (iron ore pellet-making and coal washing) and the production of DRI through the rotary kiln process.

Possible energy use and emissions reduction techniques are noted, such as waste heat recovery for onsite power generation (Cavaliere, 2019). Onsite power generation is quite common among larger DRI production facilities (>200 TPD) (National Productivity Council, 2017), so the offset of emissions from grid electricity is considered. The impact of introducing/scaling up technologies such as natural gas-based DRI (which has much lower emissions) (Cavaliere, 2019) and hydrogen-based DRI (which can have zero emissions) (Hybrit, no date) are also explored.

3.1.1. System boundaries and allocation

The product system is restricted to energy use and emissions from direct consumption of resources, except for the generation of grid electricity. Energy use and emissions from secondary processes such as oil refining, import of oil for fuel, production of coke for pellet-making, etc. are not considered due to lack of pertinent data and information, and lack of bandwidth to carry out additional data collection. The effect of

Figure 4: Process flowchart for LCA

Note: Red – processes, grey – sub-processes, blue – fuels, orange – electricity

(12)

-12-

by-products and wastes of DRI production, such as dust, slag and char are also not considered, due to the high degree of variability in the processing of such materials and lack of regulations (Bhushan et al., 2011).

In most cases, waste materials are landfilled onsite (Bhushan et al., 2011).

For industrial processes with more than one material output, the impacts are proportionally allocated to each output as per ISO 14040:2006. However, the only significant by-product of DRI production is char, which has value as a combustion fuel. However, as mentioned above, the variations in handling of such materials prevented allocation factors from being applied. Similar is the issue with coal bed methane, as the status of methane capture in coal mines being considered is unknown.

3.1.2. Data and limitations

For carrying out an LCA at the national level, data regarding the DRI production process through rotary kiln technology is required. The major resources being consumed by the production process include iron ore, coal, dolomite and electricity. Hence, data on extraction of raw materials (domestic and foreign) is required, as is the fuel consumption and distance covered in transporting the raw materials (road, rail and sea). Since a portion of coal is washed and iron ore pelletised, data on beneficiation processes are collected.

Some limitations of the study include the constant grid emissions and vehicular emissions inventory for mining of raw materials in both India and other countries. For vehicular emissions, the emission norms have been used as a baseline for the typical emission levels. The prevalence and techniques used for iron- ore pellet making in India are not well understood, hence, only the effects of furnace oil and electricity use are considered (KIOCL, 2019). The grade of raw materials being consumed in each region of the country is assumed – more granular data would improve the accuracy of results. For DRI production, the prevalence of emissions control technologies is not known; older data of SOx, NOx and PM2.5 have been adapted in this study (CPCB, 2007). The complexities of the DRI process require real time data from various production facilities to be represented more accurately.

3.2. Life-cycle inventory and impact assessment

To conduct a detailed Life Cycle Assessment, the effects of various processes further up the supply chain need to be considered and quantified. In this case, the processes being considered are shown in Figure 4.

i. Grid emission factors for electricity consumption are calculated from published LCA studies of different power generating technologies. For consumption of other fuels (diesel, FO, etc.) the emissions from point of combustion are considered, using emissions norms (DieselNet, 2018;

CPCB, 2017; Transportpolicy.net, 2019).

ii. In the primary level, impacts from mining/extraction (and import, if applicable) of coal, iron ore and dolomite are considered for the sponge iron production. In the case of coal, imports from South Africa and Indonesia are considered (Indian Bureau of Mines, 2019); for iron ore, imports from South Africa and Australia are considered (Indian Bureau of Mines, 2019). The transport of extracted resources through road, rail and sea are considered. The mining and transport technologies are assumed to be constant across different countries for simplicity.

iii. In the secondary level, iron ore processing (pellet-making) and coal beneficiation (washing) are considered.

iv. At the final use level, the iron production process and its energy use and emissions are included.

For this, details regarding the production facility such as capacity, location and key process metrics are found from published data. The processing of wastes from production are not addressed, since there is a large degree of variability involved (such as the combustion of char, local landfilling, selling of char, etc.) (Bhushan et al., 2011).

Life cycle inventory (LCI) data is derived from literature sources. For the DRI production process, literature references are used to develop a heat and mass balance model, along with certain reaction parameters to obtain the emission factors for given inputs. Since the process deals with the production of a single product, allocation issues are considered predominantly insignificant (Haque and Norgate, 2015).

(13)

-13-

The LCI will provide Scope 1 (direct) and Scope 2 (indirect, from grid electricity) energy use in MJ and emissions in GHG (CO2, CH4) as kgCO2eq; SOx, NOx and PM2.5 in g (respectively). Impacts of the life cycle will be assessed through the ReCiPe 2016 LCIA methodology. ReCiPe LCIA translates emissions and resource extractions into environmental impact scores using characterisation factors at two levels – midpoint level and endpoint level (RIVM, 2018). Of the three time-horizon options within ReCiPe, a hierarchical (100-year) time horizon is considered for this study. The impacts are assessed in terms of country-specific global warming potential (GWP) in kgCO2eq, acidification potential (AP) in gSO2eq and fine dust formation potential (fine dust FP) in gPM2.5eq, in the midpoints; effect on human health (in disability-adjusted life years or DALY), terrestrial ecosystems and aquatic ecosystems (in time integrated species loss, i.e., species years) in the endpoints.

3.2.1. Grid electricity

Emissions factors are estimated for the electric power grid using the energy mix in generation of India (Central Electricity Authority, 2019) and specific emission factors from literature for certain exemplary power generation technologies such as coal, nuclear, hydro, renewables, etc. Table 1 shows the emissions factors of various pollutants calculated using weighted average of the generation mix.

Table 1: India-average grid emission factors Generation

share %

GHG g/kWh

SOx g/kWh

NOx g/kWh

PM2.5

g/kWh

Reference(s)

Thermal 74.52% 953.77 4.72 3.22 0.94 (Singh, Sharma and Mahapatra, 2016)

(Guttikunda and Jawahar, 2018) (Seal and Bose, no date)

Hydro 10.46% 28.40 0.11 - - (Pang et al., 2015)

Nuclear 2.76% 18.50 0.02 0.04 0.01 (Hatch, 2014)

Gas 3.63% 626.00 0.02 0.32 0.04 (GREET, 2019)

Solar 2.86% 69.39 0.57 0.20 - (Yu and Halog, 2015)

Wind 4.52% 10.50 0.03 0.03 0.02 (Hatch, 2014)

Weighted average 748.63 3.55 2.42 0.70

3.2.2. Mining

For the mining processes, diesel and electricity are the fuels considered, as typically cited in the given references. The impact of explosives is not included. For diesel consumption, the emission factors prescribed for non-road diesel engines (construction) (DieselNet, 2018); CO2 emissions are calculated as stoichiometric CO2 output from combustion minus carbon going as CO, to reflect the incomplete combustion. For electricity, the grid emission factors found in Table 1 are applied. In case of fugitive emissions, data from GREET, 2019 is used.

a. Coal mining India

Energy use and emissions data for coal mining are obtained from Coal India Limited, 2017, which is the world’s largest single coal producer. In India, Coal India Limited has a market share of 83% in coal production (Coal India Limited, 2020). The sustainability report published by CIL in 2016-17 provides the amount of diesel and electricity consumed in mine activities (the two main fuels), for which specific consumption data is calculated based on the coal production in 2016-17. Fugitive methane emissions data from coal mines is obtained from the GREET database (Argonne National Laboratory, 2019).

Diesel consumption: 0.79 l/tonne coal (or) 8.12 kWh/tonne coal Electricity consumption: 8.82 kWh/tonne coal

(14)

-14-

Table 2: Energy and emissions factors for coal mining – India (Source: Coal India Limited, 2020; Argonne National Laboratory, 2019; Author’s calculations)

Fugitive Diesel Electricity Total

Energy (MJ/tonne coal) 29.22 31.75 60.96

GHG (g/tonne coal) 32796.75 2045.69 6602.03 41444.47

NOx (g/tonne coal) 32.46 21.36 53.82

SOx (g/tonne coal) 0.55 31.28 31.84

PM2.5 (g/tonne coal) 21.20 1.62 6.19 29.01

Imports of coal from Indonesia and South Africa are considered. The emission factors for diesel and electricity are kept constant for simplicity, but the specific consumption data is obtained for each country.

Indonesia

One of the largest coal mines at Kideco is operated by Indika Energy. Data regarding fuel usage is obtained from their sustainability report (Indika Energy, 2020). Of note is that 97% of diesel is biodiesel, hence the biogenic CO2 is not considered. Also, most of the energy requirement is met by onsite diesel generators.

Table 3: Energy and emissions factors for coal mining – Indonesia (Source: Indika Energy, 2020; Author’s calculations)

GHG (g/tonne coal) 32796.75 868.25 217.74 33882.4

NOx (g/tonne coal) 459.28 0.70 459.98

SOx (g/tonne coal) 7.84 1.03 8.87

PM2.5 (g/tonne coal) 21.20 22.96 0.20 44.36

South Africa

For South Africa, data is obtained for the largest coal mining company, Exxaro (Exxaro, 2019).

Table 4: Energy and emissions factors for coal mining – South Africa (Source: Exxaro, 2019; Author’s calculations)

GHG (g/tonne coal) 32796.75 5640.05 10331.15 48767.95

NOx (g/tonne coal) 89.50 33.42 122.93

SOx (g/tonne coal) 1.53 48.95 50.48

PM2.5 (g/tonne coal) 21.20 4.48 9.69 35.36

b. Iron ore mining India

Energy use and emissions data for iron ore mining are obtained from National Mineral Development Corporation, 2018, which is a publicly owned mining company. NMDC is the largest iron ore mining company in India, with a production of 35.57 million tonnes in 2017-18 (National Mineral Development Corporation, 2018a). The sustainability report published by NMDC in 2018 specifies the diesel and electricity consumption in its mines. Note that the same fugitive PM2.5 values are taken as in coal mining.

(15)

-15-

Diesel consumption: 0.44 l/tonne ore (or) 4.46 kWh/tonne ore Electricity consumption: 2.47 kWh/tonne ore

Table 5: Energy and emissions factors for iron ore mining – India (Source: National Mineral Development Corporation, 2018a; Author’s calculations)

Diesel Electricity Total

Energy (MJ/tonne ore) 16.06 8.89 24.95

GHG (g/tonne ore) 1124.13 1849.13 2973.25

NOx (g/tonne ore) 17.84 5.98 23.82

SOx (g/tonne ore) 0.30 8.76 9.07

PM2.5 (g/tonne ore) 0.89 1.73 10.50 (incl. fugitive)

Imports of iron ore from Australia and South Africa are considered. As before, the emission factors are kept constant, but the energy consumption data is country specific.

Australia

For Australia, data is obtained from (Norgate and Haque, 2010).

Diesel consumption: 2.68 l/tonne ore (or) 27.47 kWh/tonne ore Electricity consumption: 3 kWh/tonne ore

Table 6: Energy and emissions factors for iron ore mining – Australia (Source: Norgate and Haque, 2010; Author’s calculations)

GHG (g/tonne ore) 6924.86 2245.90 9170.76

NOx (g/tonne ore) 109.89 7.27 117.16

SOx (g/tonne ore) 1.88 10.64 12.52

PM2.5 (g/tonne ore) 5.49 2.11 28.80 (incl. fugitive) South Africa

For South Africa, data is obtained for Kumba Iron Ore company, operating some of the largest mines in the world, such as the Sishen mine (Kumba Iron Ore Limited, 2020).

Diesel consumption: 4.58 l/tonne ore (or) 46.91 kWh/tonne ore Electricity consumption: 10.61 kWh/tonne ore

Table 7: Energy and emissions factors for iron ore mining – South Africa (Source: Kumba Iron Ore Limited, 2020;

Author’s calculations)

GHG (g/tonne ore) 11823.61 7945.41 19769.02

NOx (g/tonne ore) 187.63 25.71 213.34

SOx (g/tonne ore) 3.20 37.65 40.85

PM2.5 (g/tonne ore) 9.38 7.45 38.03 (incl. fugitive) c. Dolomite mining

In this study, it is assumed that the DRI production process uses only dolomite and no limestone. Due to a lack of energy use data specific to dolomite mining, a calculation is made based on data published by Directorate General of Mine Safety [DGMS], 2015. This database lists the total production of dolomite in India and the number of pieces of equipment and their average horsepower. Based on this information and

(16)

-16-

assumptions regarding the working hours of a dolomite mine (Shruti Seva, 2019) and the load factor of equipment (Klanfar, Korman and Kujundžić, 2016), specific consumption data is calculated as follows:

𝑆𝑝. 𝑑𝑖𝑒𝑠𝑒𝑙 𝑢𝑠𝑒 = 𝑁𝑜. 𝑜𝑓 𝑑𝑖𝑒𝑠𝑒𝑙 𝑚𝑎𝑐ℎ𝑖𝑛𝑒𝑠 × 𝐴𝑣𝑔. ℎ𝑝 × 𝐴𝑣𝑔. 𝑙𝑜𝑎𝑑 𝑓𝑎𝑐𝑡𝑜𝑟 × 𝑊𝑜𝑟𝑘𝑖𝑛𝑔 ℎ𝑜𝑢𝑟𝑠 𝑝𝑒𝑟 𝑦𝑒𝑎𝑟 × (1 + 𝑡ℎ𝑒𝑟𝑚𝑎𝑙 𝑒𝑓𝑓𝑖𝑐𝑖𝑒𝑛𝑐𝑦 𝑜𝑓 𝑑𝑖𝑒𝑠𝑒𝑙 𝑒𝑛𝑔𝑖𝑛𝑒)

𝑆𝑝. 𝑒𝑙𝑒𝑐𝑡𝑟𝑖𝑐𝑖𝑡𝑦 𝑢𝑠𝑒 = 𝑁𝑜. 𝑜𝑓 𝑒𝑙𝑒𝑐𝑡𝑟𝑖𝑐 𝑚𝑎𝑐ℎ𝑖𝑛𝑒𝑠 × 𝐴𝑣𝑔. ℎ𝑝 × 𝐴𝑣𝑔. 𝑙𝑜𝑎𝑑 𝑓𝑎𝑐𝑡𝑜𝑟 × 𝑊𝑜𝑟𝑘𝑖𝑛𝑔 ℎ𝑜𝑢𝑟𝑠 𝑝𝑒𝑟 𝑦𝑒𝑎𝑟

Diesel consumption: 0.32 l/tonne dolomite (or) 3.26 kWh/tonne dolomite Electricity consumption: 2.42 kWh/tonne dolomite

Table 8: Energy and emissions factors for dolomite mining (Source: DGMS, 2015; Author’s calculations) Diesel Electricity Total

Energy (MJ/tonne dolo) 11.75 8.71 20.46

GHG (g/tonne dolo) 815.28 1810.57 2625.85

NOx (g/tonne dolo) 13.06 5.86 18.92

SOx (g/tonne dolo) 0.22 8.58 8.80

PM2.5 (g/tonne dolo) 0.98 1.70 23.87 (incl. fugitive) 3.2.3. Transport

Transport is an important part of the supply chain, especially considering that the iron industry, to an extent relies on import of coal and iron ore. Moreover, due to the generally remote locations of mines and the plants themselves, a fleet of trucks and/or a nearby railway is relied upon to transport most of the raw material. Thus, the transport emissions can be broken down into road transport, rail transport and maritime transport. For all these cases, emission factors have been estimated as elaborated in the sections that follow.

Based on location of a certain cluster, these factors are to obtain the total energy use and emissions.

a. Road transport

To estimate the emissions from road transport, the currently operating fleet of trucks in India are considered based on the lifetime of such vehicles (taken as 20 years) and annual sales data since 2005 till 2018 (MoRTH, 2017). Based on the year of sale, the number of trucks following the then-active emission standards is found. A weighted average emission factor is then found to estimate for the current fleet based on the historical emission norms in India (CPCB, 2017; Transportpolicy.net, 2018).

Table 9: Number of trucks in operation (Source: MoRTH, 2017) Years active No. of vehicles % share

India 2000 2000-2005 847,142 19.50%

BS-II 2005-2010 1,178,737 27.13%

BS-III 2010-2017 1,978,604 45.53%

BS-IV 2017-2018 340,781 7.84%

Total 4,345,264 100.00%

Table 10: Truck emission norms (Source: CPCB, 2017; Transportpolicy,net, 2018)

(Units in g/kWh) CO HC NOx PM SO2

India 2000 (2000) 4.5 1.1 8 0.36 0.395

BS-II (2005) 4 1.1 7 0.15 0.079

BS-III (2010) 2.1 1.6 5 0.1 0.055

BS-IV (2017) 1.5 0.96 3.5 0.02 0.008

Weighted avg. for

current vehicle mix 3.036 1.317 6.010 0.158 0.008 (fuel dependent)

(17)

-17-

CO2 emissions are estimated by considering the stoichiometric requirement of diesel fuel to produce CO2, i.e. 2640 gCO2/l diesel. From the CO production given by Table 10, the carbon in CO is subtracted from that of CO2 to account for the incomplete combustion. SO2 emissions are estimated as a stoichiometric formation from the sulphur in fuel (TransportPolicy.net, 2018). Fuel consumption data is taken for N3 rigid vehicles tested at 40 kph (International Council on Clean Transportation, 2017), calculated as 0.1516 kWh diesel/tonne-km. Thus, the emission factors for Indian trucks is found, given in Table 11.

Table 11: Energy and emission factors for trucks (Source: Author’s calculations) Energy (MJ/tonne-km) 0.546

GHG (g/tonne-km) 37.570

NOx (g/tonne-km) 0.911

SOx (g/tonne-km) 0.001

PM2.5 (g/tonne-km) 0.024 b. Rail transport

A small percentage of materials is transported through railways (for most of the distance) in instances where the plant/mine is located near a railway system; railways are typically significantly cheaper and more reliable (Freightera, 2019). Indian Railways operates a fleet of electric and diesel-electric locomotives (depending on route), emissions factors for which are identified (Indian Railways, 2019) as being 0.003 litres/net tonne-km (diesel) and 0.01 kWh/net tonne-km (electric), by applying a conversion from the given gross tonne-km to net tonne-km values. Average daily running km of electric (387 km) and diesel goods locomotives (351 km) provided the usage share of electric vs diesel at 52.44% vs 47.56% respectively.

While India does not yet have emission norms for diesel locomotives, an interim report from the Research Designs and Standards Organisation prescribes some suggested standards for diesel locomotives based on tests carried out to a certain section of the fleet (Research Designs and Standards Organisation, 2017).

These standards are applied to the specific consumption figures mentioned.

Table 12: Energy and emission factors for railways (Source: Research Designs and Standards Organisation, 2017) Diesel Electric Weighted avg.

Energy (MJ/tonne-km) 0.122 0.035 0.077

GHG (g/tonne-km) 8.640 10.384 9.554

NOx (g/tonne-km) 0.656 0.019 0.322

SOx (g/tonne-km) 0.000 0.019 0.010

PM2.5 (g/tonne-km) 0.019 0.017 0.018

c. Maritime transport

A portion of the Indian sponge iron industry relies on imported coal and iron ore (small %) to obtain better quality. Shipping predominantly occurs from South Africa, Australia and Indonesia. Such materials are transported using dry-bulk carriers, running predominantly on heavy fuel oil (HFO). The emissions factors for shipping are calculated based on engine characteristics detailed by Kristensen, 2015. An average value of the emissions per kg fuel for different engine configurations is considered as an estimate. Fuel demand and annual tonne-km of dry-bulk carriers is given by Psaraftis and Kontovas, 2009, under the assumption that fuel consumed per tonne-km has not changed significantly since 2009. Using the fuel-specific emission factors of the former, and fuel consumption and tonne-km of the latter, emissions are found in g/tonne- km; except for SOx emissions, for which recent IMO standards limit it at 0.5% fuel weight (IMO, 2020).

Table 13: Fuel specific emission factors and fuel consumption of dry-bulk carriers (Source: Author’s calculations)

(Kristensen, 2015) CO2 SOx NOx PM2.5

g/kg fuel g/kg fuel g/kg fuel g/kg fuel

Average emissions 3116.73 5.00 43.74 2.00

(18)

-18-

Dry bulk fuel consumption 52,376 tonnes

(Psaraftis and Kontovas, 2009) Total tonne-km 37,131,605,092 tonne-km

Specific consumption 1.411 g/tonne-km

Table 14: Energy and emission factors for ships (Source: Author’s calculations) Energy (MJ/tonne-km) 0.059

GHG (g/tonne-km) 37.570

NOx (g/tonne-km) 0.911

SOx (g/tonne-km) 0.001

PM2.5 (g/tonne-km) 0.024 3.2.4. Beneficiation

a. Coal washery

Coal beneficiation involves several processes of washing and drying the coal to remove excess ash content, thus improving the coal quality. Data for the coal beneficiation process is obtained from Chaturvedi, 1997.

The coal washery process involves primarily electrical equipment such as pumps, conveyors, cyclones, etc.

It is assumed that 20% of the coal required for DRI production is washed. The coal washery allows a typical reduction in ash content of ~28-30% (Zamuda and Sharpe, 2007).

Electricity consumption: 17.67 kWh/tonne raw coal

Table 15: Energy and emission factors for coal beneficiation (Zamuda and Sharpe, 2007; Author’s calculations) Energy (MJ/tonne coal) 63.60

GHG (g/tonne coal) 13225.56

NOx (g/tonne coal) 42.79

SOx (g/tonne coal) 62.67

PM2.5 (g/tonne coal) 12.41 b. Iron ore pellet making

Iron ore is in the form of lumps and fines of various sizes, and as-mined may have some impurities which reduce the Fe %. Thus, ore is sent to pellet making plants where it is concentrated and impurities are removed; the output (ore pellets) are of higher Fe % and the uniform size improves reducibility (Cavaliere, 2019). Data for the pellet making process is obtained from KIOCL, 2019, a government owned iron ore mining company. This company makes calcined pellets using furnace oil, electricity, coke and bentonite (the effects of production/mining and transport of the latter two are not considered due to lack of data).

Note: Furnace oil emissions calculated using non-road engine emission norms (DieselNet, 2018).

Furnace oil consumption: 16.65 l/tonne ore Electricity consumption: 66.83 kWh/tonne ore

Table 16: Energy and emission factors for iron ore beneficiation (KIOCL, 2019; Author’s calculations) Furnace oil Electricity Total

GHG (g/tonne ore) 40,784.51 70724.852 135462.164

NOx (g/tonne ore) 645.94 161.86 807.80

SOx (g/tonne ore) 11.03 237.07 248.10

PM2.5 (g/tonne ore) 48.45 46.94 95.38

(19)

-19- 3.2.5. DRI production

For the DRI production step of the life cycle, literature data (Sarangi and Sarangi, 2011) is utilised to develop a heat and mass balance. The reactions, kiln conditions and DRI quality used by the model provided in this reference are replicated on Excel. The following are major assumptions as found in the reference:

i. ~5.3% of iron ore is lost in flue dust

ii. ~17% of fixed carbon is lost in flue dust, char and DRI

iii. Fixed carbon of injection coal combusts to increase temperature and produce CO2

iv. Fixed carbon of feed coal produces CO by reacting with CO2 formed by injection coal combustion v. 67% of injection coal volatile matter (VM) reacts in kiln, remaining VM combusted in the ABC vi. 1/3^rd of feed coal’s volatile hydrogen is reduced to methane - assumed to remain un-combusted in

the ABC. If combusted, GHG emissions would be reduced by nearly 25%.

vii. Kiln temperature is 1340 K, DRI/char is 1325 K, Flue gas/dust is 1100 K

The input parameters (ore quality and coal quality) are varied for different datapoints for the DRI clusters (see Section 3.3). Using the built-in ‘Goal seek’ tool, the model allowed the calculation of specific energy consumption and emissions per tonne of DRI production, by estimating the coal and ore requirement based on quality of raw materials (i.e. ultimate analysis-based composition). Other process parameters (such as losses) are kept constant. Below are the input parameters for an average case, the cluster of Jamshedpur.

Table 17: Typical input quantities for DRI production (Source: Chatterjee, 2010; Author’s calculations) Input material Quantity Quality

Iron ore 1483 kg 65% Fe

Coal 1155 kg 35% Fixed carbon, 34% volatiles, 9% moisture, 21% ash GCV = 5214 kcal/kg

Dolomite 98 kg 28% CaO, 20% MgO

Electricity 70 kWh (Outotec, no date)

For the above inputs, the net heat is balanced using the ‘Goal seek’ function by adjusting coal input, such that the heat supplied from exothermic reactions is equated to heat consumed for reactions/losses. By computation for the above inputs, the typical emissions from DRI production are shown in Table 18. Note that due to the presence of various types of emissions control technologies, it is difficult to estimate NOx, SOx and PM2.5 from the heat/mass balance. These are adapted from an earlier survey (CPCB, 2007) as average values per Nm³ of waste gas reported for various DRI plants.

Table 18: Energy and emission factors for DRI production (Author’s calculations, CPCB, 2007)) Energy (MJ/tonne DRI) 25,447.12

GHG (kg/tonne DRI) 2623.77

NOx (g/tonne DRI) 596.15

SOx (g/tonne DRI) 1637.41

PM2.5 (g/tonne DRI) 1197.18

Whilst the prevalence of waste-heat recovery for power generation is unknown, it is common for plants having more than 200 TPD of DRI capacity to have a waste-heat recovery boiler (WHRB) to generate steam, for electricity generation using a steam turbine (National Productivity Council, 2017). This would account for emissions offsets, as it negates a portion of grid-generated electricity. Considering a typical configuration and steam conditions for a WHRB (BEE, 2011), the following is the potential for power generation and associated emissions offsets.

Table 19: Potential for electricity from WHRB (Source: Author’s calculations; National Productivity Council, 2017)

Component Value Comments

(20)

-20-

Heat in flue gas 6700 MJ/tonne DRI Flue gas from 1 tonne DRI produced

Boiler efficiency 80% Typical value

Heat supplied to steam 5360 MJ/tonne DRI Heat contained in steam Turbine heat rate 2356 kcal/kWh (IGEN, no date)

Net electricity output 426 kWh/tonne DRI Incl. 70 kWh for DRI plant and 10% parasitic loss Table 20: Potential energy and emissions offset by WHR power plant (Source: Author’s calculations)

Energy (MJ/tonne DRI) -1534.53

GHG (kg/tonne DRI) -319.11

NOx (g/tonne DRI) -1032.40

SOx (g/tonne DRI) -1512.12

PM2.5 (g/tonne DRI) -299.36

3.3. Application of LCI data

A calculation model for the life cycle assessment is developed using Microsoft Excel. As given in Figure 4, based on the total material consumption of the cluster plants, the supply chain data is applied. For coal mining, it is assumed that 50% of coal is procured through imports and 50% is domestically sourced (Argus, 2019). For iron ore mining, it is assumed that 85% is domestic and 15% is international, based on the overall average share of domestic vs imported ore used by the iron and steel industry (Indian Bureau of Mines, 2019). Of the imported materials, the share of different exporting countries is applied based on data from Chatham House, 2018. It is arbitrarily assumed that 20% of domestic iron ore and 20% of domestic coal underwent beneficiation into pellets and beneficiated coal respectively.

To obtain a broader picture of the Indian DRI industry, the specific energy and emissions are expanded to a national level. From the developed database containing 317 sponge iron plants, their locations and capacities, it is identified that the states of Odisha, Jharkhand, Chhattisgarh, West Bengal and Karnataka accounted for ~80% of the total installed capacity. Formation of clusters is observed in the GIS mapping (Figure 1b), such as around the cities of Jamshedpur and Rourkela. Based on these formations, predominantly observable clusters are selected (Figure 5):

Figure 5: Locations of DRI plant clusters (Source: Author’s creation)

Bellary

Durgapur Jamshedpur

Rourkela

Angul Raigarh/

Jharsuguda Raipur

(21)

-21- i. Angul, Odisha

ii. Rourkela, Odisha iii. Jamshedpur, Jharkhand

iv. Raigarh/Jharsuguda, Chhattisgarh/Odisha v. Raipur, Chhattisgarh

vi. Durgapur, West Bengal vii. Bellary, Karnataka

Once clusters are identified, representative plants are considered in each of the clusters, which are arbitrarily considered to be an agglomerate of all plants in that cluster. Essentially, each cluster is assumed as one large plant for the purpose of simplifying the calculation of distances and total energy consumption. Then, based on the total DRI production capacity, the energy consumption and emissions produced by each cluster are calculated; the above clusters represent ~77% of the total coal based DRI capacity in India. By summing up these values, and extrapolating to the total DRI production in India (Statista, 2020), the total energy use and emissions for DRI production is calculated.

Also, the calculation of mining emissions varies between clusters – it is assumed that the Southern cluster of Bellary utilises imported coal from South Africa, whereas the other clusters in the North are assumed to use Indonesian coal due to their proximities to the West and East coasts respectively.

The calculation for energy use and emissions from the DRI production are varied by considering the typical qualities of iron ore and coal in the respective source regions of these materials (Srinivasan et al., 2018; Indian Bureau of Mines, 2019). For e.g., in the case of Raipur, ore with 65% Fe is chosen since that is the most common type mined in the region. The change in material quality also affects transport and beneficiation indirectly. Dolomite quality is kept constant due to the limited quantity considered.

Table 22: Assumptions of iron ore and coal quality

Cluster Iron ore quality

(Indian Bureau of Mines, 2019)

Domestic coal source (20% beneficiated) (Srinivasan et al., 2018)

Imported coal source (Srinivasan et al., 2018)

Angul 62% Fe MCL Indonesian

Rourkela 62% Fe MCL Indonesian

Jamshedpur 65% Fe CCL Indonesian

Raigarh/Jharsuguda 65% Fe SECL Indonesian

Raipur 65% Fe WCL Indonesian

Durgapur 65% Fe ECL Indonesian

Bellary 62% Fe SCCL South African

Cluster Capacity (MTPA)

Angul 5.39

Rourkela 3.35

Jamshedpur 2.34

Raigarh/Jharsuguda 6.16

Raipur 5.02

Durgapur 2.14

Bellary 3.01

Total 27.41

Total coal DRI capacity

in India (2019) 37.60

% Represented 72.88%

Coal DRI produced in

India (2019) 27.81

Table 21: Capacities of different representative plants (Source: Indian Bureau of Mines, 2018)

(22)

-22-

Figure 6 shows the transportation scheme considered. From each representative plant, the distances to the nearest coal port, iron ore port, coal mine region, coal washery, iron ore mine region, iron ore pellet plant and dolomite mine region are found using QGIS and online mapping services. Such distances are calculated for both road and rail (from the nearest railway station at each end, last mile by road). For shipping, the sea distances are calculated from major exporting ports in the exporting countries to major importing ports in India, then the remaining distance is covered by road. Thus, in this manner, there is the possibility of two scenarios – transport by road or by rail; the initial results are calculated with an arbitrary 50% road and 50%

rail scenario. Table 23 shows the typical values obtained in the case of Jamshedpur cluster.

Table 23: Example of overall transport emissions (for Jamshedpur cluster) (Source: Author’s calculations) Coal transport emissions Truck Train Ship Total

Energy (MJ/tonne DRI) 70.22 17.82 577.84 621.86

GHG (g/tonne DRI) 4811.66 1572.14 40788.80 43980.70

NOx (g/tonne DRI) 117.23 55.46 832.75 919.10

SOx (g/tonne DRI) 0.15 2.41 25.81 27.10

PM2.5 (g/tonne DRI) 3.08 2.02 26.15 28.70

Iron ore transport emissions Truck Train Ship Total

Energy (MJ/tonne DRI) 101.54 28.56 225.12 290.17

GHG (g/tonne DRI) 6957.66 2378.49 16133.01 20801.08

NOx (g/tonne DRI) 169.51 78.55 297.68 421.71

SOx (g/tonne DRI) 0.22 2.91 15.03 16.60

PM2.5 (g/tonne DRI) 4.46 2.74 10.35 13.94

Dolomite transport emissions Truck Train Total

Energy (MJ/tonne DRI) 20.13 3.53 11.83

GHG (g/tonne DRI) 1379.27 350.72 864.99

NOx (g/tonne DRI) 33.60 13.87 23.74

SOx (g/tonne DRI) 0.04 0.74 0.39

PM2.5 (g/tonne DRI) 0.88 0.54 0.71

Figure 6: Transport schemes

(23)

-23-

4. Results and sensitivity analysis

By applying the emissions factors to each of the different stages from cradle to gate (mining > transport >

beneficiation > DRI production), the overall specific emissions per tonne of DRI is obtained.

Table 24: Overall specific energy use and emissions

Category Mining Transport Beneficiation DRI

production Total Energy (MJ/tonne DRI) 337.80 995.77 79.31 25,828.57 27,241.45

GHG (kg/tonne DRI) 49.80 70.86 23.51 2,635.95 2,780.12

NOx (g/tonne DRI) 360.57 1,500.68 200.58 600.87 2,662.70

SOx (g/tonne DRI) 38.95 46.55 67.28 1,652.79 1,805.57

PM2.5 (g/tonne DRI) 81.91 47.44 24.53 1,209.89 1,363.77 From Table 24, the contributions of different steps along the supply chain to the total energy use and emissions per tonne of DRI can be compared. In all metrics except NOx emissions, the DRI production process itself is the largest contributor. Thus, it highlights the need for stricter regulations and better monitoring to control the emissions from this process. This is especially true since the most efficient coal- DRI processes can reach a SEC as low as 17.5 GJ/tonne DRI (Chatterjee, 2012) from the above calculated 25.8 GJ/tonne DRI. Note that, as mentioned in Section 3.2.5, the combustion of any methane generated during DRI production would reduce GHG emissions by 25%.

Efficiency improvements can be possible by enhanced waste-heat recovery systems that allow preheating the kiln feed materials, leading to 8-9% reduction in coal consumption (Kumar and Khanam, 2012).

Efficiency improvements can also be had by increasing ore beneficiation, by which the Fe % in ore is increased, the reducibility is improved and ore losses through dust are reduced (Chatterjee, 2010). Coal beneficiation can be employed to reduce the ash content of local coals (Zamuda and Sharpe, 2007) to improve their quality, thus reducing reliance on imports. Making WHR mandatory (currently it is a recommended guideline) (Bhushan et al., 2011) for power generation for plants with >200 TPD capacity (National Productivity Council, 2017) could allow emissions offsets (Table 26) by negating the corresponding amount of electricity produced using the grid (Plant SEC reduces to 24.2 GJ/tonne DRI); a WHR power plant could produce a net output of 426 kWh/tonne DRI (see Annex).

Of note in Table 24 is the significant contribution of transport systems towards the total energy use and emissions. Especially in terms of NOx emissions, the transport contribution is even higher than the DRI production process (the latter is usually reduced through emissions control technologies such as wet scrubbers, electrostatic precipitators, etc.). The overall energy use and emissions of transport is significantly influenced by shipping, which contributes 80-90% of the total value, despite being less emissions intensive than other modes. Thus, while imported coals are of higher quality, a better option would be to deploy larger capacity of clean raw material beneficiation technologies to improve the quality of domestic ores and coals (Argus, 2019). This will be a key step to reduce reliance on foreign trade and to still maintain emissions reductions targets. It remains to be seen how the mining emissions will progress; in India, mining is still highly labour-intensive (hence the lower fuel consumption than foreign mines, see Section 3.1.2). If mechanisation increases in the coming years, there is a potential for steady increase in mining emissions.

Thus, mining processes must be improved to ensure accelerated adoption of sustainable fuels and practices.

Table 25: Midpoint and endpoint life cycle impacts

Category Mining Transport Beneficiation DRI

production Total Midpoint impacts

100 yr GWP (kgCO2eq/tonne

DRI) 49.80 70.51 23.51 2,635.95 2,779.78

(24)

-24-

AP (gSO2eq/tonne DRI) 151.91 554.44 114.61 1,301.15 2,122.10 100 yr Fine dust FP

(gPM2.5eq/tonne DRI) 627.28 996.28 301.43 8,402.28 10,327.26 Endpoint impacts

GWP Human health

(DALY/tonne DRI) 4.62E-05 6.54E-05 2.18E-05 2.45E-03 2.58E-03

GWP Terrestrial ecosystems

(species years/tonne DRI) 1.39E-07 1.97E-07 6.58E-08 7.38E-06 7.78E-06 GWP Aquatic ecosystems

(species years/tonne DRI) 3.81E-12 5.39E-12 1.80E-12 2.02E-10 2.13E-10 AP Terrestrial ecosystems

(species years/tonne DRI) 2.41E-08 6.90E-08 2.57E-08 4.66E-07 5.85E-07 Fine dust FP Human health

(DALY/tonne DRI) 4.26E-04 6.77E-04 2.03E-04 5.68E-03 6.98E-03

Table 25 provides the midpoint and endpoint life cycle impacts from each stage of the process as per the ReCiPe 2016 methodology. The midpoint impacts are quantified in terms of global warming potential, acidification potential and PM formation potential. Across all three indicators, the DRI production process is the major contributor. However, the transportation process (shipping, mainly) contributes significantly to acidification. These trends can be observed in Figure 7. In terms of endpoint impacts, human health and ecosystem loss are measured in disability-adjusted life years (DALY) and species years respectively. The impact on both human life is highest due to the formation of fine dust, whereas global warming causes most of the species loss. The national annual impacts can be found in Table 27.

49.80 1.8%

70.86

2.5% 23.51 0.9%

2 635.95 94.8%

1 10 100 1 000 10 000

kgCO2eq/tonne DRI

GWP 100 yr

81.91

6% 47.09

9.7% 24.53 3%

1 209.89 81.4%

1 10 100 1 000 10 000

gPM2.5eq/tonne DRI

Fine dust FP 100 yr

151.91 7.2%

554.44 26.1%

114.61 5.4%

1 301.15 61.3%

1 10 100 1 000 10 000

gSO2eq/tonne DRI

AP

Figure 7: Specific life cycle midpoint impacts

Life cycle assessment of coal based direct-reduced iron production in India

Life cycle assessment of coal based direct-reduced iron production in

India

Sabarish Elango

Abstract

Abstrakt

Acknowledgment

Table of Contents

List of Figures

List of tables

List of abbreviations

1. Introduction

1.1. Literature review

1.2. Data collection

1.3. Research objectives

1.4. Thesis structure

2. Overview of direct reduction process

3. Methodology and data sources

3.1. Goal and scope

3.2. Life-cycle inventory and impact assessment

3.3. Application of LCI data

4. Results and sensitivity analysis

GWP 100 yr

Fine dust FP 100 yr

AP