Mechanism of low nox emission in circulating fluidized bed decoupling combustion

The lowered NOx emission was believed to result from the combined contributions of DSL pyrolysis products including char, tar, and pyrolysis gas py-gas to the reduction of NOx via their

Mechanism of Low-NOx Emission in Circulating Fluidized-Bed

Decoupling Combustion

作者姓名 : DO HAI SAM

指导教师 : 许光文 (研究员，中国科学院过程工程研究所)

高士秋 (研究员，中国科学院过程工程研究所)

学位类别: 工学博士

学科专业: 化学工程

培养单位: 中国科学院过程工程研究所

2018 年 6 月

Decoupling Combustion

A dissertation submitted to University of Chinese Academy of Sciences

in partial fulfillment of the requirement

for the degree of Doctor of Philosophy

in Chemical Engineering

By Hai-Sam Do Supervisors: Professor Guangwen Xu

Professor Shiqiu Gao

Institute of Process Engineering Chinese Academy of Sciences

June 2018

摘 要

摘 要

循环流化床解耦燃烧（CFBDC）实现低氮氧化物（NOx）排放的技术可行性已在处理白酒糟（DSL）的工业示范装置上得到了很好验证。低 NOx 排放被认为是由包括焦炭，焦油和热解气（py-gas）的 DSL 热解产物在燃烧器中再燃时共

Abstract

Abstract

The technical feasibility of low-NOx circulating fluidized-bed decoupling combustion (CFBDC) has been well proved in an industrial demonstration plant treating distilled spirit lees (DSL) The lowered NOx emission was believed to result from the combined contributions of DSL pyrolysis products including char, tar, and pyrolysis gas (py-gas) to the reduction of NOx via their reburning in the riser combustor

of the CFBDC system In order to further understand the mechanism of NOx reduction

in CFBDC, this study is devoted to investigating the capabilities of biomass char, tar and py-gas for NO reduction through experiments in a lab-scale drop-tube reactor (DTR) that simulates the reburning conditions occurring in CFBDC

The work performed pyrolysis of DSL at 500 °C to produce the tested char and tar reactants while py-gas was prepared by mixing pure gases from cylinders according to the analyzed py-gas composition In order to ensure the sufficient reduction of NO, a total mass feeding rate of reagents, which is 0.15 g/min, was adopted for most experiments We first investigated in Chapter 4 the variations of acquired NO reduction

efficiency (η e) with major parameters including reagent feeding rate, reburning stoichiometric ratio (SR), reaction temperature, residence time, and initial flue gas composition It was found that tar enabled the best NO reduction in comparison to char and py-gas did at the same mass feeding rate of reagent (0.15 g/min) The presence of

CO in py-gas inhibited the homogeneous NO reduction reactions to cause lower η e For

DSL-derived char and tar, their realized η e were facilitated through by higher temperatures and higher initial NO and CO concentrations The main conclusion of this

chapter is that the suitable SR values for obtaining the highest η e by reburning of pyrolysis products were found to be 0.6 − 0.8

Based on preceding results, the synergetic effects among char, tar and py-gas reagents on NO reduction were evaluated and discussed in Chapter 5 The comparison

at given total mass flow rate of NO-reduction reagent indicated that the char/py-gas (binary reagent) enabled the best synergetic NO reduction than the others did Its realized efficiency elevated with increasing of the py-gas proportion The tar/py-gas or tar/char mixture caused a positive effect only when the tar proportion was necessarily lowered to about 26% In addition, there existed obvious interactions between char and some species in py-gas (i.e., H2, CxHy) for NO reduction by pyrolysis products The

synergetic effects were closely related to the molar ratio of C and H elements in reagents over the fed NO (CH/NO ratio)

The NO reduction capabilities of char and tar reagents derived from other fuels such as sawdust (SD) and Xianfeng (XF) lignite were further investigated in Chapter 6

At the specified mass flow rate of reductant, say, 0.15 g/min, the SD char or XF lignite char were less efficient than the DSL char did for reducing NO because of the lower contents of ash (containing catalytic matters) in the SD and XF lignite chars However,

the SD tar enabled the highest η e among the three tested tars Above all, tar as an attractive reagent always exhibits the better NO reduction than char does Testing model tar compounds including phenol, benzene, acetic acid, methyl acetate and heptane for

NO removal revealed that phenol plays an important role in enabling the good NO reduction by the SD tar Our major understanding from testing the NO reduction by tar

is that the compounds containing at least one aromatic ring (e.g phenol, benzene) are the major contributor for reducing NO in either biomass tar or coal tar

In conclusion, the results of this study would be significant in the operation of CFBDC technology treating N-rich fuel By considering the pyrolysis-generated tar as

a dominant factor in lowering NOx emission in a CFBDC system, further studies are suggested to focus on kinetic analysis of the NO reduction by tars and also on the combined action of tar with other reagents Additionally, the effect of pyrolysis temperature on NO reduction activity by various reagents should be investigated

Key words: pyrolysis, NO reduction, reburning, decoupling combustion, circulating

fluidized bed, biomass, coal, low-NOx combustion, drop-tube reactor

Table of Contents

Table of Contents

Chapter 1 Introduction 1 

1.1  Background 1 

1.2  Objectives and Significance 3 

1.3  Thesis Outline 4 

Chapter 2 Literature Review 7 

2.1  Nitric Oxides 7 

2.1.1  Sources of NOx 7 

2.1.2  NOx Emission in China 7 

2.2  Low-NOx Emission Strategy 9 

2.2.1  NOx Formation During Fuel Combustion 9 

2.2.2  NOx Reduction Technologies 11 

2.3  Decoupling Combustion (DC) for Lowering NOx Emission 14 

2.3.1  Principle of Decoupling and DC Technology 14 

2.3.2  Low-NOx Emission in Grate-Based DC 16 

2.3.3  Low-NOx Emission in CFBDC 18 

Chapter 3 Material and Methodology 25 

3.1  Preparation of NO-Reduction Reagents 25 

3.1.1  Feedstock Material 25 

3.1.2  Pyrolysis Setup and Procedure 25 

3.1.3  Characteristics of NO-Reduction Reagents 28 

3.2  Experimental Drop-Tube Reactor for NO-Reduction Evaluation 30 

3.2.1  Main Chamber 30 

3.2.2  Heating Control System 31 

3.2.3  Reagent-Feeding System 32 

3.2.4  Flue-Gas Supplying System 35 

3.2.5  Sampling and Analyzing System 36 

3.3  Experimental Procedure 37 

3.3.1  Procedure and Analysis 37 

3.3.2  Validation of Experimental Setup Conditions 39 

Chapter 4 NO Reduction by Biomass Pyrolysis Products 45 

4.1  Introduction 45 

4.2  Experimental Conditions 45 

4.3  Results and Discussion 46 

4.3.1  NO Reduction Varying with Reagent Feeding Rate 46 

4.3.2  NO Reduction Varying with SR 49 

4.3.3  NO Reduction Varying with Reaction Temperature 53 

4.3.4  NO Reduction Varying with Residence Time 54 

4.3.5  NO Reduction Varying with Flue Gas Composition 55 

4.4  Conclusions 61 

Chapter 5 Synergetic Effect Among Pyrolysis Products in Reducing NO 63 

5.1  Introduction 63 

5.2  Experimental Conditions 63 

5.3  Results and Discussion 65 

5.3.1  Synergetic Effect of Binary Reagent 65 

5.3.2  Synergetic NO Reduction Varying with Reaction Temperature 70 

5.3.3  Synergetic NO Reduction Varying with Residence Time 71 

5.3.4  Synergetic NO Reduction Varying with Gas Species 72 

5.4  Conclusion 76 

Chapter 6 NO Reduction by Reagents Derived from Different Fuels 77 

6.1  Introduction 77 

6.2  Materials and Experimental Conditions 77 

6.2.1  Materials 77 

6.2.2  Experimental Conditions 79 

6.3  Results and Discussion 81 

6.3.1  NO Reduction by Char Reagents and Effect of Ash Content 81 

6.3.2  NO Reduction by Tar Reagents and Model Tar Compounds 88 

6.4  Conclusions 98 

Chapter 7 Conclusions and Recommendations 99 

7.1  Conclusions 99 

7.2  Innovation 101 

7.3  Recommendations for Future Work 101 

Nomenclatures 103 

References 105 

Appendix A Chemical Compositions of Tested Tar Reagents 115 

A.1 GC–MS Spectra 115 

Table of Contents

A.2 Identified Compounds of Tar 116 

Appendix B Calibration Curves of Feeders for Different Reagents

119 

Acknowledgement 121 

Résumé 123 

Fig 1.1 Distilled spirit lees (DSL) disposal

In term of beverage industries, China is a large spirits-producing country Therefore, the output of solid residue such as distilled spirit lees (DSL) generated in the unique solid-state fermentation process (Fig 1.1) is enormous all over the country In fact, DSL amounts to 20 million tons per year (Deng and Luo, 2004) and can be considered as a good biomass resource due to its being rich in cellulose and hemicellulose Traditionally, the DSL have sufficiently high nutrition content for being feedstuff or protein substrate of animals such as pig With progress in the biotechnology, the digestible nutrition in the lees becomes so low that it cannot meet the requirement

of animal feedstuff Nowadays, except for apart used as the filler material for animal feedstuff or as fertilizer, the DSL is mainly landfilled or discharged to the open air

(Deng and Luo, 2004; Zhang et al., 1997) However, the high moisture in DSL, even

up to 60 wt %, and the properties of easy putrescibility and strong acidity can cause serious environmental problems, such as smelly gas release and underground water pollution (Xu et al., 2009a) Therefore, highly reliable technologies for clean, rapid and large-scale utilization of DSL are in a great demand in the spirits industry of China, especially for large distiller factories

The nature of rich in cellulose in DSL makes it hardly treated via biological conversion technologies The thermal conversion of the lees into the energy usable in the distiller factories, such as steam or fuel gas, is thus considered to be viable Burning DSL to produce steam in fact not only limits the pollution as a result of DSL disposal but also offers a part of energy required by distilled spirits production (Xu et al., 2007) However, as a result of its relatively high N content (about 3 – 5 wt % on a dry basis), direct combustion of DSL via the traditional way has to release high NOx emission, which is one of the main gas-phase pollutants released in fuel combustion, not only injures human health but also forms acid rain and photochemical smog (Winter et al., 1999; Calvert, 1997) The NOx content in the flue gas can even reach up to 550 ppm during burning this material in a laboratory fluidized bed (Zhu et al., 2015)

In order to reduce NOx emission in combusting high-N biomass waste and also improve the combustion efficiency of a conventional circulating fluidized bed, the so-called circulating fluidized-bed decoupling combustion (CFBDC) technology has been investigated and developed by our Advanced Energy Technology Laboratory (AET Lab), Institute of Process Engineering (IPE), Chinese Academy of Sciences (CAS) The technology has been well proven by an industrial system treating DSL The actual running data show that the NOx emission was lowered by about 70% comparing

to the traditional CFB combustion (Han et al., 2015; Yao et al., 2011; Xu et al., 2010), making the NO content in its flue gas be 120 – 170 mg/Nm3 for the DSL containing N

of about 4.0 wt.% This kind of technologies can also be applied to many other lees and residues such as vinegar lees and Chinese herb residues generated in various light industrial processes (Yao et al., 2011)

The CFBDC technology is based on the reaction decoupling concept which separates the combustion process into drying/pyrolysis of fuel and combustion of pyrolysis-generated char and volatile Thus, the system is composed of a fluidized-bed pyrolysis reactor and a riser combustor, the pyrolysis-generated volatile consisting of non-condensable pyrolysis gas (py-gas) and condensable tar is sent to an intermediate

Fig 1.2 Principle conception of circulating fluidized bed decoupling combustion (CFBDC)

1.2 Objectives and Significance

The approaches of the decoupling principle for innovating combustion technologies were unquestionably effective, namely in term of lowering NOx emission The NOx reduction mechanisms in the developed coal/biomass combustion devices have been systematically studied by AET Lab to determine better control strategies associated with the technologies involving decoupling conception Thereby, the reduction effects of char and reducing gas on NO (as the main component of NOx in most practical flue gas) were analyzed in detail (Cai et al., 2013; Dong et al., 2010,

2009, 2007; He et al., 2006) Recently, tar species derived from the pyrolysis of DSL was found to reduce NOx significantly by the analysis in micro-fluidized bed reactor (Song et al., 2014) However, the experimental conditions in earlier investigations were actually not close enough to the reburning condition in CFBDC due to the limitations

Flue gas

Distilled spirit lees

Riser Combustor

Fluidized Bed Pyrolyzer

Heated HCPs + unburnt char

Py-gas & Tar

Temperature-lowered HCPs + char

of batch or semi-batch reactors adopted In addition, those studies are still far from fully understanding the interactions occurred within the practice system since the combined actions of NOx reduction in the reburning zone by different agents including char, tar and py-gas have not been previously considered Above all, the complexity of the NO reduction reactions required that further experimental work must be carried out in order

to understand some key aspects of the process Therefore, it is highly worthwhile to deepen such studies for further understanding the mechanism of low-NOx emission in CFBDC treating not only DSL but also other fuels such as sawdust or coal To implement this plan, a continuous drop-tube reactor (DTR) was indeed adopted to facilitate the investigations on the characteristics of NO reduction using char, tar and py-gas from pyrolysis of different fuels as reagents, these aim to

(i) figure out the dominant NO reduction reactions for CFBDC,

(ii) provide the optimal conditions for operating CFBDC system in term of achieving high NOx reduction,

(iii) reveal the synergetic effect among char, tar and py-gas on NO reduction occurring during their reburning in CFBDC,

(iv) understand the combined homogeneous–heterogeneous reaction mechanism for NO reduction,

(v) understand the different NO reductions for tars, chars derived from different fuels and the contributions of their constituents to the achieved NO reduction

1.3 Thesis Outline

This thesis discusses basically on two following topics

(i) reduction of NO by DSL pyrolysis products including char, tar and py-gas in both individual and combined actions,

(ii) reduction of NO by reagents derived from different fuels in comparison with that by DSL-derived products

The contents are divided into seven chapters The first chapter served as an

introduction where the brief background of CFBDC are given, followed by the research objectives and approach

Chapter 2 provides a literature review about the theoretical background and

current state of research essential for the discussion of experiment results

Chapter 3 introduces the materials and the main experimental techniques used in

Chapter 1 Introduction

this study Namely, the preparation of NO-reduction reagents from DSL, the design of experimental apparatus, the procedure and analysis of experiments were described

in detail

Chapter 4 begins presenting the results of this study The NO reduction

characteristics of DSL-derived char, tar and py-gas are shown and discussed, from which the influences of reagent feeding rate, reburning stoichiometric ratio (SR), reaction temperature, residence time, initial composition of flue gas on NO reduction

by each reagent are put into evidence

Chapter 5 covers the synergetic effect of reagents on NO reduction The results of

NO reduction tests using the mixtures of binary reagents among char, tar and py-gas derived from DSL are given and discussed in function of SR, temperature and residence time Moreover, the effects of gas species in py-gas such as CO, H2, CH4 on NO reduction by char/py-gas mixture are presented

Chapter 6 examines the NO reduction by tars, chars derived from different fuels

such as such as SD and XF coal The results in this chapter are brought together with those in Chapter 4 in order to compare the capabilities of different reagents for reducing

NO The difference in NO reduction by chars is discussed based on the analysis of catalytic matter in ashes, while the contribution of some components in tars to the overall NO reduction is provided in addition to explain the reactivity of tars for NO removal

Chapter 7 briefly summarizes the key conclusions of this study and provides a few

recommendations for future research directions

Chapter 2 Literature Review

Chapter 2 Literature Review

2.1 Nitric Oxides

Nitrogen oxides (NOx), some of the main gas-phase pollutants released in fuel combustion, are a very interesting and important family of air pollutant NOx not only injures human health but also forms acid rain and photochemical smog (Winter et al.,

1999; Calvert, 1997) Therefore, a strict legislation is thus being considered or has been

implemented by many countries, including China, to lower the emission of NOx 2.1.1 Sources of NOx

There are two sources of NOx emission generated by human activities such as mobile sources and stationary sources Automobiles and other mobile sources contribute about half of the NOx that is emitted, while electric power plant boilers produce about 40% of the NOx emissions from stationary sources (United States Environmental Protection Agency, 1999) The substantial emissions are also added by such sources as industrial boilers, incinerators, gas turbines, reciprocating spark ignition and diesel engines in stationary sources, iron and steel mills, cement manufacture, glass manufacture, petroleum refineries, and nitric acid manufacture In addition, biogenic or natural sources of nitrogen oxides including lightning, forest fires, grass fires, trees, bushes, grasses, and yeasts are also involved in the global NOx emission These various sources produce differing amounts of NOx but almost three-quarters of the total amount

of NOx emission is contributed by human activities through combustion of fossil and alternative fuels (including field burning and forest fires) (Topsoe, 1997; Bosch and Janssen, 1988)

2.1.2 NOx Emission in China

China is the largest NOx emission country in Asia contributing 41% – 57% of Asian NOx emissions (Fei et al., 2016; Wang et al., 2014; Wang and Hao, 2012; Zhao

et al., 2008; Ohara et al., 2007; Mauzerall et al., 2005) With the rapid growth of energy consumption, NOx emissions were estimated to more than double from 11.0 Mt in 1995

to 26.1 Mt in 2010, with an annual growth rate of 5.9% Power plants, industry and transportation were major sources of NOx emissions, accounting for 28.4%, 34.0%, and 25.4% of the total NOx emissions in 2010, respectively (Zhao et al., 2013b) In 2014,

2.5 million tons of NOx were emitted from industrial boilers, which accounted for 12.3% of total Chinese NOx emissions Industrial boilers are the third largest emission source, after power plants and vehicles (Ministry of Environmental Protection of the People’s Republic of China, 2014) Based on current legislation and implementation status, defined as a business as usual (BAU) scenario according to Zhao et al (2013b), NOx emissions in China are estimated to increase by 36% in 2030 from 2010 level In detail, the trend in NOx emissions and that prediction for 2020 and 2030 of each province in China are listed in Table 2.1

Table 2.1 Provincial NOx emissions during 2005 – 2030 (Mt) in China (Zhao et al., 2013b)

Chapter 2 Literature Review

To mitigate the adverse effect of air pollution, the Chinese government released many policies assuring enforcement of the control programs to reduce NOx emissions during the 12th Five-Year Plan (2011 – 2015) (The State Council of the People’s Republic of China, 2011b) This implied rapid installation of control measures, namely most coal-fired power plants are requested to be equipped with flue gas denitrification (The State Council of the People’s Republic of China, 2011a, 2010) In addition, emission standard for thermal power plants has been continually updated, and the Euro

IV emission standard for heavy-duty diesel vehicles was implemented (Ministry of Environmental Protection of China, 2011) As a result, the governmental target of reducing 10% NOx emissions by 2015 (compared to 2010) has been achieved (Jin et al., 2016; Shi et al., 2014; Zhao et al., 2013a) However, the proportion of industry increases despite the control measures, reflecting the lower ambition of explored policies with respect to control of industrial NOx emissions For example, as shown in Table 2.1, Shandong, Henan, Jiangsu, Guangdong, and Hebei provinces were top five emitters of NOx emissions in 2010, each of which had over 1.5 Mt NOx emissions and contributed together over 35% of total emissions, and those continue to have the largest emissions in the next 5 – 10 years (Zhao et al., 2013b)

2.2 Low-NOx Emission Strategy

2.2.1 NOx Formation During Fuel Combustion

NOx emissions from combustion are commonly considered to be comprised of nitric oxide (NO) and nitrogen dioxide (NO2) For most combustion systems, significant evidence exists to show that NO is the predominant NOx species (over 95% of the total) (Meadows et al., 1996) However, for purposes of emissions control, NOx is defined as the sum of NO and NO2 fully converted to NO2 or vice versa This corresponds to the output of most NO measurement technique

The formation of NOx during fuel combustion is a complex interaction among chemical, physical, and thermal processes occurring simultaneously within the device Generally, NOx is formed either from fixation of N2 in the combustion air at high temperatures or from oxidation of nitrogen chemically bound in fuel (Glarborg et al., 2003) To help simplify the understanding of NOx formation and assist in identifying control strategies, NOx is typically considered to form through three following mechanisms

Thermal NOx is formed by the oxidation of atmospheric nitrogen by free oxygen

atoms in the higher temperature regions of the combustion flame The formation

of thermal NOx proceeds through the following reactions sequence known as Zeldovich mechanism (Miller and Bowman, 1989; Zeldovich, 1946) The first step (2.1) has high activation energy so that the formation of thermal NO is most important at temperatures above 1800 K

(2.1) (2.2) (2.3)

Prompt NOx is formed by chemical reactions between atmospheric nitrogen and

fuel-derived hydrocarbon radicals and subsequent oxidation In this way, the CH radicals attack on the N2 triple bond, and then the reactive nitrogen compounds may subsequently be oxidized to NOx or recycled to N2 dependent on reaction conditions (Williams et al., 1994; Miller and Bowman, 1989; Glarborg et al., 1986),

Those reactants continue to react with O to form NOx, as

(2.8) (2.9)

The total mechanism of the prompt NOx formation is complex It was reported that the prompt NOx is usually formed in the fuel-rich zone close to the burner that

is of high concentration of CH and low concentration of O2 However, the formation is often assumed negligible during biomass fuel combustion (Nussbaumer, 2003; Skreiberg et al., 1997)

Fuel NOx is formed from chemical reactions involving nitrogen atoms chemically

bound within the fuel component species (fuel-N) The detailed fuel-N behavior during the thermal conversion of carbonaceous fuel is illustrated in Fig 2.1 Initiating combustion process, a part of fuel-N is released as N-containing volatile including NOx precursor gases such as ammonia (NH3), hydrogen cyanide (HCN), isocyanic acid (HNCO) and nitrogen oxides (NO, NO2, N2O), toxic contaminant

Chapter 2 Literature Review

like organic nitrogenated compounds (tar-N) and N2 The rest part of fuel-N is left

in the char (char-N) In turn those NOx precursors, tar-N, char-N are emitted as NOx/N2O after being combusted and gasified (Mukadi et al., 2000; Tan and Li, 2000) Considering fuel-NOx is the main part which cover about 80% of NOx formed during fuel combustion, the control of NOx emission in thermal conversion of carbonaceous fuel is usually based on the fuel-N content

Fig 2.1 Fuel NOx released during the thermal conversion of carbonaceous fuel (Chen, 2012)

2.2.2 NOx Reduction Technologies

Various techniques for reducing NOx emission, which are summarized in Fig 2.2, have been applied on fuel-fired systems (Baukal, 2004; Park et al., 2001; Tomita, 2001; Hill and Smoot, 2000; Smoot et al., 1998; Teng and Huang, 1996; Mereb and Wendt, 1994) Those can be classified into two fundamentally different categories: combustion controls reducing NOx formation during the combustion process and post-combustion controls reducing NOx after it has been formed Combustion control techniques mainly include low-NOx burners (LNBs), reburning, overfire air (OFA), flue gas recirculation (FGR), operational modifications and so on, while selective catalytic reduction (SCR) and selective noncatalytic reduction (SNCR) are two well-known technologies in post-combustion controls (United States Department of Energy, 1999)

Char-N

Fig 2.2 Most popular technologies for NOx emission control

Low-NOx Burners – LNBs are designed to control the mixing of fuel and air so as

to achieve staged combustion This results in a lower maximum flame temperature and a reduced oxygen concentration during some phases of combustion, thus resulting in both lower thermal NOx and lower fuel NOx production

Overfire Air – OFA is the air that is injected into the burnout zone located above

the normal combustion zone of furnace OFA is generally used in conjunction with operating the burners under fuel-rich conditions, which reduces NOx formation The OFA is then added to achieve complete combustion In addition, OFA is frequently used together with LNBs in combustion system

Fuel Reburning (or fuel-staging) combustion system consists of three zones in

series: (1) a combustion zone operating under slightly fuel-lean conditions; (2) a reburning zone, where added fuel results in a fuel-rich, reducing condition; and (3)

a burnout zone, where the OFA is injected leading to complete of the fuel combustion With a typical reburning process, about 10% – 30% total heat input

is added into reburning zone, where fuel-rich conditions allow reducing up to 70% NOx formed in the normal combustion zone (Smoot et al., 1998) Coal, oil, gas and also biomass can all be used as the reburning fuel

Low NOx burners

Fuel staging (Reburning)

Operational modification

Overfire Air

Selective Catalytic Reduction (SCR) Selective Noncatalytic Absorptions

Post-combustion controls Combustion controls

NOx Control Technologies

Flue gas recirculation

Others

Chapter 2 Literature Review

Flue Gas Recirculation – FGR, in which part of flue gas is recirculated to the

combustion zone, can be used to modify the fuel-burning conditions, namely lowering temperature and reducing oxygen concentration to reduce NOx formation In practical system, FGR is also used as a carrier to inject fuel into the reburning zone to increase penetration and mixing

Operational Modifications involve several methods changing the operational

parameters to create conditions in the furnace that can lower NOx formation, e.g., burners-out-of-service (BOOS), low excess air (LEA), and biased firing (BF) In BOOS, selected burners are suspended by temporarily stopping fuel injection, but air flow is maintained to create staged combustion in the furnace LEA involves operating at the lowest possible excess air level but maintaining good combustion, and BF creates staged combustion conditions in the furnace by injecting more fuel

to some burners (typically the lower burners) while reducing fuel to other burners (typically the upper burners)

Selective Noncatalytic Reduction – SNCR is a simple process, referred to as

“thermal deNOx”, and involves the reduction of NOx to N2 in the presence of oxygen by reaction with a reducing agent (typically amine-based reagent such as ammonia or urea at 1073 – 1273 K The critical factors in applying SNCR are sufficient residence time in the appropriate temperature range and even distribution and mixing of the reducing agent across the full furnace cross section

Selective Catalytic Reduction – SCR is a typical technology which has been

applied on stationary combustion units for effectively reducing NOx emission In SCR, a catalyst bed is installed downstream of the furnace A reducing agent such

as ammonia is injected into the flue gas before it passes over the fixed-bed catalyst The catalyst promotes a reaction between NOx and ammonia to form nitrogen and water By using SCR, NOx reductions as high as 95% are achievable (Pan et al., 2013; Hans and Peter, 2010), but careful design and operation are necessary to keep ammonia emissions (referred to as “ammonia slip”) to a concentration of a few ppm

Hybrid systems An effective way is combining the above technologies in a

combustion system that can highly promote NOx reduction efficiency, as illustrated in Fig 2.3 For example, LNBs can be used in conjunction with SCR or

SNCR to achieve a greater overall efficiency than those achieve alone, fuel reburning and SNCR/SCR can be used together to maximize NOx reduction (Yang

et al., 2009) The choice depends upon the type of combustion system, type of boiler or other energy conversion device, and especially type of fuel used Available technologies will be limited by considering turndown ratio, stability of combustion, availability or access to burners, air supply controls, fuel impurities, and cost among other factors (United States Environmental Protection Agency, 1999)

Fig 2.3 Location of NOx control techniques in a fuel-fired system (Biarnes, 2018)

2.3 Decoupling Combustion (DC) for Lowering NOx Emission

2.3.1 Principle of Decoupling and DC Technology

Thermochemical conversion of solid carbonaceous fuels is generally shown with three types of process technologies: pyrolysis (including coking and carbonization),

gasification, and combustion These processes actually involve a similar complex

reaction network as shown in Fig 2.4, those are resulted from differentiating the oxygen amount fed to the conversion system and the typical operation temperatures, as follows

- The combustion process requires an excessive oxygen supply to convert all

C and H elements in fuel into CO2 and H2O

- The gasification refers to the conversion of a solid fuel into CO, H2 and light

hydrocarbon gas under an insufficient oxygen supply but enough to maintain the required temperatures for accelerating adiabatic reactions

Chapter 2 Literature Review

- The pyrolysis process indicates the conversion of fuel without oxygen requirement or with a certain amount but much less than its demand

CH4 Then, other reactions start to occur, accelerating a series of interactions among various reactions such as cracking, hydrogenation, reforming and so on Among these interactions, some can facilitate the conversion to lower pollutant emission, increase process efficiency and product quality, and widen fuel adaptability, but some others are not In order to avoid or weaken the effects of undesired interactions for optimizing the conversion process, the related reactions should separately occur in the rearranged reaction zones This idea of reaction control has been termed “decoupling” and further developed by AET Lab (Zhang J et al., 2013; Zhang Y et al., 2013; Xu et al., 2010, 2009b, 2008; Zhang et al., 2010)

In conventional combustion (CC) process, solid carbonaceous fuels are generally

burned in a single reaction vessel (combustor) and completely coupling all the involved

reactions, resulting in difficult control of pollutant emission and combustion efficiency

Many technical ways have been developed to optimize the combustion process, such as

air-staging (OFA), reburning, FGR, LNBs; as described in Subsection 2.2.2, aiming to lower NOx emission Those advanced combustion technologies are actually correlative with decoupling, but it was not commonly reported in literatures This conception was only introduced into the combustion area for the first time as a so-called

decoupling combustion (DC) by researchers in IPE (Li et al., 1997) The technology

has been developed to reduce the NOx emission during fuel combustion through separating the involved reactions of combustion process according to the principle illustrated in Fig 2.5

Fig 2.5 Principle of decoupling combustion

Up to now, many studies on the NOx emission and reduction during the solid fuel combustion have been carried out to demonstrate the advantage of DC in efficiently lowering NOx emission (Han et al., 2015; Xu et al., 2010; Dong et al., 2009, 2007; He

et al., 2006) The decoupling conception is in fact implemented in combustion process via two typical ideas (decoupling modes) which are isolating and staging the fuel pyrolysis and char combustion reactions Table 2.2 thus summarizes two DC technologies representative for such two decoupling modes and their realized decoupling effects

Table 2.2 Typical combustion technologies and their features by involving decoupling (Zhang

J et al., 2013)

Decoupling

Staging Grate-based DC (fixed bed) (Li et al., 1997) Burning the released volatiles in combusting char bed to lower NO and CO emissions

bed) (Xu et al., 2010)

Reburning the released volatiles in transporting bed

of char in burning to lower NO to CO emissions

2.3.2 Low-NOx Emission in Grate-Based DC

Fig 2.6 shown the DC implemented in coal-fired grate furnace (i.e., DC domestic stove) (Dong et al., 2009, 2007; He et al., 2006; Li et al., 1997) Instead of being in a single vessel and single layer, the furnace is divided into three zones such as pyrolysis

Chapter 2 Literature Review

zone, a combustion zone and a post-combustion zone Firstly, fuels are pyrolyzed in the pyrolysis zone to produce char and volatiles including tar and py-gas Char is in turn moved into the combustion zone due to gravity so that it could be burned together with the released volatiles Then the unburned volatiles are forwarded to burnout in post-combustion zone In this way, the NOx reduction effects are believed to occur when the volatiles are passing through the char bed in combustion zone

Fig 2.6 Configuration of a domestic stove using DC

By comparing different combustion modes that may occur in real stove, He et al (2006) reported that the obtained low NOx emission in DC stove burning coal is mainly attributed to the catalytic reduction of NO over hot char particles The reactions and mechanisms of NO reduction over char have been widely reported in literatures (Chambrion et al., 1998; Suzuki et al., 1994; Yamashita et al., 1993), also the NO generated in char combustion can be reduced by CH4, H2, and CO in py-gas and gaseous tar component as well (Giral and Alzueta, 2002; Rüdiger et al., 1997) Above all, the surface and structure of char particles were found to be the dominant factor in such DC stove because most of NO reduction reactions take place on char surfaces in combustion zone (Dong et al., 2010; Dong et al., 2007)

In view of application, the NOx emission from DC stove burning coal was reduced

to less than 200 mg/m3 (excess O2 at 7%), which is up to 40% lower than that of traditional stoves (Li et al., 1997) Moreover, complete combustion of the pyrolysis gas

in DC also led to lower emissions of CO and soot, thus resulting in the higher combustion efficiency Dong et al (2009) also applied DC on grate furnace for burning

biomass (rice husk, corn straw and sawdust) and biomass–coal blends Both their tests

in a quartz dual-bed reactor or in a real stove demonstrated that the NO emission and the corresponding conversion of fuel-N to NO were obviously lower for DC than for

CC Currently, different types of household stoves and industrial boilers using DC in capacities of 0.01 − 0.7 MW are widely equipped in rural areas of China for cooking, heating, and hot water supply (Zhang J et al., 2013)

2.3.3 Low-NOx Emission in CFBDC

Technical Description of CFBDC

The DC can be also implemented in circulated fluidized-bed (CFB) combustor for lowering NOx emission and promoting combustion efficiency, which is in fact representative for the “isolating” decoupling approach The property of low NOx emission in the so-called circulating fluidized-bed decoupling combustion (CFBDC) was fully verified by the actual running data of a CFBDC system treating DSL in demonstration plant (35000 ton/a), namely the NOx emission was lowered by about 70% comparing to the traditional CFB combustion (Han et al., 2015), making the NOx content in its flue gas be 120 – 170 mg/Nm3 for the DSL containing N of about 4.0 wt.%

Fig 2.7 A schematic diagram of CFBDC system and possible mechanisms for NO reduction inside the reburning zone

Chapter 2 Literature Review

Fig 2.7 illustrated the schematic diagram of a CFBDC system, it involves two fluidized-bed chambers, such as a fuel pyrolyzer (or gasifier) and a riser char combustor, and the heat carrier particles (HCPs) between the two chambers are circulated to carry the heat of combustion into the pyrolyzer for the reactions of fuel pyrolysis Different from CFB combustion, in CFBDC the fuel is supplied into pyrolyzer and the generated volatiles are fed into the middle of the riser char combustor while the produced char mixed with HCPs is forwarded to the bottom of the riser combustor to generate heat Therefore, the combustion in the middle of riser combustor could be considered as a reburning zone following commercialized reburning technology, in which not only the homogeneous interaction between py-gas and flue gas (i.e., gas reburning) (Cai et al., 2013) but also the reduction effect of the entrained char on NOx can contribute to lowering the NOx emission (Dong et al., 2010) In addition, the tar species derived from fuel pyrolysis can reduce NOx significantly (Song et al., 2014)

NOx-Reduction Chemistry of Reburning

Fuel reburning is a well-known low-NOx combustion technique Many studies have been performed on reburning of different fuels including gas (e.g., natural gas and other hydrocarbon fuels), liquid (e.g., residual fuel oil) and solid (e.g., coal and biomass) (Meadows et al., 1996) Considering NO as the main component of NOx in most practical flue gas, the mechanisms of lowering NOx in reburning were almost studied through considering NO reduction reactions in terms of gas-phase homogeneous reactions between volatiles (gaseous fuel, tar and py-gas) and NO and gas-solid heterogeneous reactions between char and NO

As mentioned above, the pyrolysis products including char, tar and py-gas are considered as reburning fuels conducting to an intermediate position of the riser char combustor of CFBDC system Therefore, the occurred mechanism of NO reduction in the system is quite similar to that in the commercialized reburning technology The chemistry of reburning is very complex although many fundamental studies on it have been reported (Luan T et al., 2009; Casaca and Costa, 2005; Glarborg et al., 2000; Smoot et al., 1998) The technologies with different types of reburning fuel or other features have their respective characteristics, for example in CFBDC system both heterogeneous and homogeneous mechanisms were involved to form a combined action

of char, tar, and py-gas in reducing NO, as also illustrated in Fig 2.7 Following are the summaries of NO reduction reactions which may occur in CFBDC

Homogeneous Mechanism of NO Reduction:

In the homogeneous mechanism, light hydrocarbons from existing components and gaseous tar cracking decompose into C-containing radicals (e.g.,

CHi and HCCO) via partial oxidation (Shu et al., 2015b; Zhang R et al., 2014), as

Analyzing kinetics of reactions between NO and single or blended hydrocarbons showed that the conversion of NO was highly sensitive to reactions (2.11) and (2.12) under moderate fuel-rich conditions, while high temperatures also facilitate such reactions (Dagaut and Lecomte, 2003; Dagaut et al., 2000b, 1999) Thus, the essential conditions for NO reduction are the formation of sufficiently many CHi

and HCCO radicals

On the other hand, the nitrogen bounds in tar (tar-N) can form nitrogen species in the reburning zone to release more nitrogenous radicals (e.g., HCN and

NHi) during tar decomposition, as (Lu et al., 2009; Liu et al., 1997; Wendt, 1995)

It is well known that the availability of nitrogen species is responsible for the NO reduction (into N2) by hydrocarbons through the following sequence of reactions (Dagaut and Lecomte, 2003; Dagaut et al., 2000a, 1998; Glarborg et al., 2000):

HCNO + H → HCN + OH , HCN + O/OH → CN/NCO/NH + … ,

The elementary reactions (2.14) – (2.19) indicate the importance of oxygen presence in homogeneous NO reduction by tar Oxygen not only takes part in oxidative cracking of hydrocarbons via reaction (2.10) but also facilitates

Chapter 2 Literature Review

conversion of HCN into NH in reactions (2.15) – (2.17) This in turn facilitates

NO reduction via reactions (2.18) – (2.19)

However, when oxygen in the reaction zone is excessive, the formation of

CO and CO2 from oxidation of hydrocarbons becomes dominant so that CHi and HCCO are consumed in oxidation rather than in reducing NO Meanwhile, the nitrogenous radicals would be oxidized to form NO Kinetic analysis by Dagaut and Lecomte (2003) revealed that the following reactions are highly sensitive in biomass pyrolysis gas reburning and varying the outlet NO concentration:

While the NO-reduction radicals (i.e., CHi, HCCO) are consumed, the outlet NO concentration has to increase Consequently, controlling oxygen content for an appropriate reburning SR is a key factor for promoting the contribution of homogeneous NO reduction in a practical reburning process

Heterogeneous Mechanism of NO Reduction:

Regarding the heterogeneous reaction mechanism, the char derived from pyrolysis (coal and biomass) has been widely tested as an effective reagent for NO reduction The free-active sites C* and surface complexes C(O) generated by partially oxidizing carbon on char surface would reduce NO through following gas-solid heterogeneous reactions (Dong et al., 2007; Suzuki et al., 1994; Yamashita et al., 1993):

(2.24) (2.25) (2.26) (2.27) (2.28) (2.29) (2.30)

In addition, CO existing in flue gas can remove the chemisorbed oxygen from the surface complexes C(O) and also reacts with NO on the char surface, as follows (Dong et al., 2007; Liu et al., 1997)

CO + C(O) → CO 2 + nC* , (2.31)

Studies on NO Reduction by Char, Tar and Gas

NO Reduction by Gaseous Reagents:

Many studies have been conducted regarding the homogeneous NO reduction reaction using gaseous reactants Fan et al (2006) reported that hydrocarbon gases such as CH4 and C3H6 have great impact on NO reduction, whereas CO or H2 are hard to reduce NO efficiently The simulation conducted by Glarborg et al (2000) indicated that the gas mixture of CO and H2 removed 20% – 30% of the NO entering the reburning zone and the NO conversion increased with rising the CH4

content in such a mixture Wang et al (2008) showed that the NO reduction efficiency increased with raising the reburning fraction, hydrocarbon content in the reburning fuel injected, initial NO concentration and residence time of flue gas

in the reburning zone Yu et al (2008) reported that hydrocarbon radicals such as

CHi and HCCO released from reaction (2.10) are the important intermediates in NOx reduction by CH4 and C2Hi Concerning the interactions among gas components during NO reduction by biomass volatile, Wu et al (2015) clarified that CO had no contribution to NO reduction and there was a synergetic effect for

H2 and CH4 Most of literatures have revealed in general that the reduction ability

of NO is evident for hydrocarbon gases, even though arguments exist on whether non-hydrocarbon gases have effective reduction capability for NO or not

NO Reduction by Tar Reagents:

The pyrolysis tar, which is gaseous product at high temperature, also take part

in the homogeneous NO reduction reactions The tar derived from fuel pyrolysis

is composed of many aromatic species including phenols, BTX and other compounds such as aliphatic, carboxylic acid, ester groups and N-bound components as well (Zhang H et al., 2011) As presented above, the decomposition

of these species at high temperatures releases a huge amount of hydrocarbon and non-hydrocarbon species like CHi and HCCO, NHi radicals, HCN, CO and H2

gaseous (Brezinsky et al., 1998; Lovell et al., 1989; Brezinsky, 1986) to cause high effect in reducing NO (reactions (2.10) – (2.19)) However, there were very few studies on the NO reduction by tar Rüdiger et al (1997, 1996) reported that the

Chapter 2 Literature Review

nitrogen content in tar components (tar-N) has a positive effect on NOx reduction

in the reburning zone The research group of Luo et al (Zhang R et al., 2014, 2011; Liu et al., 2009; Duan et al., 2007) revealed that tar is highly helpful in increasing the NO reduction efficiency by biogas, while the effectiveness of NO reduction by model tar compounds was also observed

NO Reduction by Char Reagents:

The residual char derived from fuel pyrolysis has been widely tested as an effective heterogeneous reducing reagent for NO reduction The NO-char reaction has been reported to play an important role in suppressing the conversion of fuel

N into NO in fuel combustion (Thomas, 1997; Tullin et al., 1993) The influential factors of heterogeneous NO reduction by char in the presented mechanism include the char types, structure of char particles and presence of gas species on char surface Dong et al (2010, 2007) reported that biomass char was more active

in reducing NO than coal char did, and the NO reduction over biomass char can

be enhanced with the gas components CO, O2 and SO2 present in the reaction atmosphere The NO reduction activity of char was correlated with the specific surface area and porosity of char particles (Lu et al., 2011; Zhong et al., 2002) In addition, the catalytic effect of inherent metal in char on the NO–char reaction cannot be ignored Illangomez et al (1996, 1995a, b, c, d) found that the group metals of K, Ca, Fe, Cr, Co, Ni, and Cu catalyzed NO reduction reaction by char and resulted in a decrease in the activate energy Zhong et al reported that KOH and NaOH could reduce the activate energy of NO–char reaction and the catalytic activity was slightly better for KOH than for NaOH (Zhong and Tang, 2007; Zhong et al., 2002)

Regarding the combined action of char, volatiles in reducing NO, Shu et al (2015a) and Lu et al (2011, 2009) have studied the contribution of char to the total

NO reduction during reburning biomass and coal Results indicated that the contribution of char to the total NO reduction was higher for higher-volatile fuel

On the contrary, Liu et al (1997), by using a self-proposed mechanistic model, revealed that NO reduction by volatiles predominates, but as the reburning SR increases, the contribution of char to overall NO reduction increases Above all, the combined action of different constituents should be a complex process, it is difficult to distinguish the effects of the heterogeneous and homogenous reactions

when they simultaneously occurred in the reburning zone Also, the synergetic reduction of NO by char and volatile matters has not been well understood This study indeed investigates the NO reduction characteristics of char, tar, py-gas and also their synergetic effects to further understanding low-NOx emission mechanism in CFBDC

Chapter 3 Material and Methodology

Chapter 3 Material and Methodology

3.1 Preparation of NO-Reduction Reagents

3.1.1 Feedstock Material

As mentioned above, the experiments performed in this study simulate the reburning process that occurs within the CFBDC system treating distilled spirit lees (DSL) (Han et al., 2015) This kind of biomass waste, provided by Luzhou Laojiao Group Company of China, was thus chosen as the feedstock material to prepare NO-reduction reagents The characteristic of the adopted material was firstly examined

by the proximate analysis following the national standard method of China for solid biofuels – GB/T 28731-2012 In addition, the ultimate analysis was analyzed by the CHNS element analyzer (VARIO MACRO CHNS analyzer with an oxygen kit; Elementar Co., Langenselbold, Germany) Table 3.1 shows the proximate and ultimate analyses of the received DSL sample, the sample contained about 51.22 wt % water and had a relatively high content of N, say, 3.5 wt % on the dry and ash free basis (daf) The size of DSL as received was in range of 2.0 – 5.0 mm

Table 3.1 Proximate and ultimate analyses of raw distilled spirit lees (DSL)

Proximate analysis (wt %, ar)

3.1.2 Pyrolysis Setup and Procedure

NO-reduction reagents used in this study were prepared from DSL in a pyrolysis setup which was mainly composed of a reaction system and a volatile-collection system,

as illustrated in Fig 3.1 In the reaction system, a horizontal fixed bed quartz reactor (reactor tube) of 42 mm in inner diameter and 1250 mm in length was mounted inside

an annular electric furnace together with a K-type thermocouple to measure the temperature at an intermediate position of the furnace, and this temperature was treated

as the pyrolysis temperature The volatile-collection system consisted of two condensing tubes, three acetone scrubbing bottles and a tar collection flask These bottles and flask were immersed in an ice-water bath to keep the expected cooling conditions for possibly highest condensable volatile (tar) yield while the condensing tubes were cooled down by cold-water stream from a circulating water bath.

(1) Reactor (quartz tube); (2) Horizontal furnace; (3) Condensing tubes; (4) Ice-water bath;

(5) Tar collection flask; (6) Acetone scrubbing bottles; (7) Wet gas meter

Fig 3.1 A schematic diagram of pyrolysis experiment

The pyrolysis temperature was maintained at 500 oC by a temperature control unit, this temperature is quite close to those adopted for fuel pyrolysis in CFBDC that ensure the stable combustion in the riser (Yao et al., 2011) In addition, at high temperatures, cracking of tar should occur to decrease tar yield and make the collected tar be mainly heavy tars As in coking, the produced tar is mainly pitch or asphaltene The pyrolysis temperature is thus set at 500 °C in this study with the concerns of having higher tar yield (with least secondary reactions)

The temperature profiles obtained from maintaining the furnace temperature of

500 oC and 600 oC are shown in Fig 3.2 As shown, there was a central zone (approximately 400 mm in length) where the temperature could be considered fairly uniform Therefore, the material should be placed in this zone of reactor (reaction zone)

to gain a high efficiency of pyrolysis process as well as a homogeneous feature for char particles obtained With the given sizes in range of 2.0 – 5.0 mm, about 100 g of the dried DSL, obtained by exposing raw material in air for 24 h to reach a moisture of approximately 10%, was fully loaded into the defined reaction zone for a batch process

of DSL pyrolysis

Chapter 3 Material and Methodology

Fig 3.2 Temperature profiles along the horizontal axis of furnace

After loading a certain amount of material, the air inside reactor tube was removed

by introducing high purity N2 gas at a flowrate of 1 L/min for at least 10 min to form

an inert atmosphere The pyrolysis process was started with heating the furnace to a preset temperature of 500 oC at which the reactor tube loaded with the given amount of DSL was quickly placed into the furnace and in turn quickly connected to the volatile-collection system The gaseous pyrolysis product coming out from the reactor was cooled immediately in the tar collection flask immersed in ice-water bath and the condensing tubes The formed liquid tar including water was collected in the flask while the other condensable light tar was further absorbed by acetone in the scrubbing bottles The remaining non-condensable gas was burned before being released into the atmosphere The process was ended after about 30 min as no gas bubble appeared in the scrubbing bottle, this ensured the complete release of volatile matter from the residual char inside reactor tube The reactor tube was taken out of the furnace and cooled to room temperature in inert atmosphere by a N2 flow similar to that in the preparing step

The residual char from the cooled reactor tube was ground and sieved to gain a particle size range of 0.05 – 0.1 mm suitable for NO-reduction experiments The liquids from tar collection flask, from scrubbing bottle and from washing the entire pipeline and condensing tubes (by acetone) were blended together and the blend was then treated

in a rotary vacuum evaporator to primarily remove acetone The evaporated liquid was further dehydrated using MgSO4 and filtrated to remove any dust intake Finally, the

filtrate was evaporated in the rotary vacuum evaporator to fully recover pure tar ready for NO-reduction evaluations Because the storage of non-condensable pyrolysis gas (py-gas) required a complicate system, the used py-gas for NO-reduction tests was a model gas made according to the analyzed composition of the sampling gas which was obtained in a batch pyrolysis

3.1.3 Characteristics of NO-Reduction Reagents

Similar to the raw material, the characteristics of its pyrolysis products were firstly examined by proximate and ultimate analyses Table 3.2 shows the results of proximate and ultimate analyses for DSL pyrolysis-derived char, tar, and py-gas

Table 3.2 Proximate and ultimate analyses of DSL pyrolysis products

Table 3.3 Composition analyses of DSL-derived py-gas (vol %)