Fuel 103 (2013) 421–429

Contents lists available at SciVerse ScienceDirect

Fuel

journal homepage: www.elsevier .com/locate / fuel

Oxy-fuel technology: An experimental investigations into oil shalecombustion under oxy-fuel conditions

L. Al-Makhadmeh a,⇑, J. Maier b, M. Al-Harahsheh c,d, G. Scheffknecht b

a Department of Environmental Engineering, Al-Hussein Bin Talal University, Ma’an 71111, Jordanb Institute of Combustion and Power Plant Technology (IFK), University of Stuttgart, Pfaffenwaldring 23, 70569 Stuttgart, Germanyc Department of Mining Engineering, Al-Hussein Bin Talal University, Ma’an 71111, Jordand Department of Chemical Engineering, Jordan University of Science and Technology, Irbid 22110, Jordan

h i g h l i g h t s

" First investigation on oil shale combustion under oxy-fuel conditions." A 20 kW Once-Through reactor was used." SO2 emissions under oxy-fuel conditions is lower than air-firing by around 30%." NOx emission was found to be lower under oxy-fuel conditions than air-firing." NOx emission can be reduced efficiently by adopting staged combustion.

a r t i c l e i n f o

Article history:Received 12 March 2012Received in revised form 26 May 2012Accepted 29 May 2012Available online 15 June 2012

Keywords:Oxy-fuelCombustionOil shaleSO2

NO

0016-2361/$ - see front matter Crown Copyright � 2http://dx.doi.org/10.1016/j.fuel.2012.05.054

⇑ Corresponding author. Tel.: +962 797076387.E-mail addresses: l.al-makhadmeh@ahu.edu

(L. Al-Makhadmeh).

a b s t r a c t

Oil shale utilization has received much attention all over the world due to the rise of oil-prices; Jordan isone of the countries that have just started an intensive research for oil shale utilization. In this study thefeasibility of oil shale combustion under oxy-fuel conditions was investigated using a 20 kW Once-Through reactor at a combustion temperature of 1200 �C. To the investigators best knowledge, this isthe first time that oxy-fuel technology is applied for oil shale combustion. So this study is considered aunique with respect to the conditions and the scale of the combustion experiments. Jordanian oil shalesamples from El-Lajjun was used. Unstaged air-firing and oxy-fuel combustion were investigated to studyoil shale combustion behaviour. It is found that direct combustion of oil shale under oxy-fuel conditionsis feasible, 100% oil shale burnout was achieved for OF27 combustion as well as air-firing. In addition, thehigh S content in oil shale is a well known problem; our study aims to find if oxy-fuel conditions willaffect SO2 emissions as well as NO emissions. It is found that SO2 emissions during oil shale combustionunder oxy-fuel conditions is lower than air-firing by around 30%. In addition, NOx emission was alsofound to be lower and can be reduced efficiently by adopting staged combustion technology underoxy-fuel conditions as well as air-firing, however, the oxy-fuel investigations were carried out withoutflue gas recirculation.

Crown Copyright � 2012 Published by Elsevier Ltd. All rights reserved.

1. Introduction

Jordan is largely affected by the world energy situation since itis a non-oil producing country. It has a future plan or prospect thatoil shale utilization will be one of Jordans’ future energy sources.Oil shale resources in Jordan are estimated to be more than 40 bil-lion tons, in which the oil content is about 4 billion tons [1]. Itslarge deposits are widely distributed all over the country, particu-larly in the central region [2]. As confirmed by geological surveys,

012 Published by Elsevier Ltd. All r

.jo, leema_30@yahoo.com

the existing oil shale reserves cover more than 60% of Jordan’s ter-ritory. Preliminary studies have shown that underground miningand open pits are possible for either direct combustion to generateelectricity or retorting process to produce shale oil [2]. Recently,there has been an increasing interest to study the characteristicsand behaviour of oil shale combustion, but in a lab scale.

Oil shale, a fine-grained sedimentary rock, often contains a pro-portionally large amount of kerogen, which can be converted intooil by thermal degradation. It has been estimated that over2.7 � 1014 tons of oil potentially exists in the known worldwide re-sources of oil shale, and this represents an enormous energy poten-tial [3]. This is 10 times more than the proven liquid petroleumreserves [4]. Small quantities of clay, quartz, pyrite and sulphur

ights reserved.

422 L. Al-Makhadmeh et al. / Fuel 103 (2013) 421–429

are usually found within the shale matrix. Each type of shale con-tains a certain quantity of volatile material. The inconsistency andthe variety in composition of raw oil shale will complicate themechanism of combustion. Oil shale is employed in China, Estoniaand Russia to yield shale oil through retorting processes and usedfor electric power generation by direct combustion.

Oil shale in Jordan and elsewhere faces many obstacles that hin-der its utilization. Among these are the high content of inorganicmaterials, mainly carbonates, which could reach values greaterthan 55% of its content. The high sulphur content in oil shale is con-sidered one of the biggest challenges for utilization of Jordanian oilshale [5]. The use of existing technologies for combustion createsserious environmental and health problems. For commercial useof oil shale deposits in the world, a significant improvement ofwaste removal and recycling system is needed [6]. Mineral carbon-ation has been proposed as a means of reducing CO2 emissions inthe Republic of Estonia, whose energy sector is predominantly(up to 67%) based on oil shale, a locally-available low-grade fossilfuel [7]. Even though, former workers have done little related tothe combustion characteristics of oil shale and reducing its envi-ronmental pollution [8].

Energy production from fossil-fuels combustion, including coal,oil, oil shale and natural gas, results in the emission of greenhousegasses, the dominant contributor being CO2. Model predictionsshow that the emissions of CO2 will continue to increase over thecoming decades as fossil fuels continue to be the major source ofenergy [9]. Because of the large contribution of CO2 to climatechange, large reductions in CO2 emissions will be necessary to sta-bilize the atmospheric CO2 concentration. Greenhouse-gas emis-sions from energy production can be reduced by the use ofalternative energy sources such as nuclear power and renewableenergy. Renewable energy sources are increasingly used, however,until these sources can reliably produce significant amounts of en-ergy, the immediate energy demand is likely to be met by ad-vanced fossil-fuel-utilization technologies. One of the severaltechnologies developed for CO2 capture and storage is the oxy-fueltechnology. During oxy-fuel combustion, a combination of oxygen,with a purity of more than 95%, and recycled flue gas is used forcombustion of the fuel. By recycling the flue gas consisting mainlyof CO2 and water vapour is generated, which is ready for storageafter purification. The recycled flue gas is used to control the flametemperature and ensure a proper heat transfer without majorchanges in the layout of the boiler. Pilot-scale and full-scale studieshave demonstrated the feasibility of using oxy-fuel combustion asa technology applicable to pulverized-fuel (PF) power plants for

Table 1Instrumental techniques used for characterization and analysis.

Analysis Measurement principle

Higher heatingvalue

Adiabatic calorimeter

Proximateanalysis

Thermogravimetric analysis

Ultimate analysis(elementalanalysis)

Incineration with oxygen and thermal conductivity detectors

Ash mainelements

ICP used to produce excited atoms and ions that emitelectromagnetic radiation at wavelengths characteristic of aparticular element.

Ash minorelements

CO2 recovery or capture. A part from enriching the flue gas withCO2, this technology also offers other benefits, including a potentialdecrease in NOx emissions and lower net combustion gas volumeat higher O2 feed concentration [10].

This work offers oxy-fuel technology as one solution for oilshale combustion environmental problems, a pilot scale test facil-ity was used. This study aims to investigate the feasibility of oilshale combustion using oxy-fuel technology, to figure out oxy-fuelconditions compared to air-firing. In addition, it is aimed to test ifthis technology will be of zero emission technology for the dirtyfuel Jordanian oil shale; to find out if oxy-fuel conditions will affectSO2 emissions due to S-capture in the ash and affect NO emissionas well. The question if CO2 capture technologies can be adoptedto oil shale combustion is of great interest all over the world.

2. Experimental work

2.1. Oil shale characterization

Oil shale sample used in this study was obtained from El-Lajjunarea in Jordan. The collected sample (150 kg) was crushed using ajaw crusher and then milled using an open hammer mill to a sizefraction of 673 lm. Detailed analysis for a representative samplewas carried out at the Institute of Combustion and Power PlantTechnology (IFK)/Stuttgart University. Several instrumental tech-niques were used for characterization as summarized in Table 1.The characterization included proximate, ultimate analyses andthe heating value as shown in Table 2. The ash is mainly composedof calcite (46.1%) and silicate (28.3%), whereas, the main trace ele-ments present in the ash are Sr, Zn, Cr and Mo. From the elementalanalysis, S content in El-Lajjun oil shale is 9.03 waf%. Therefore, it isof particular importance to take into account such high content of Swhen using this oil shale as energy source, either for oil productionor direct combustion. It was reported that the sulphur in El-Lajjunoil shale is mainly organic with some pyritic sulphur [11].

2.2. Description of the 20 kW Once-Through furnace

Oil shale combustion was performed in a 20 kW Once-Throughfurnace (Fig. 1). It mainly consists of a ceramic tube reaction zoneof 2500 mm length and 200 mm diameter. The facility is electri-cally heated around the reaction zone enabling constant walltemperature and uniform temperature profile. Constant walltemperature, up to 1400 �C, can be maintained thus enabling reli-able investigation at different temperatures. The combustion air is

Equipment Species Standard

IKA C 4000 – DIN 51900

Leco TGA-500 Water DIN 51718Ash DIN 51719Volatiles DIN 51720Fixed-C Calculated

Vario el (elementar) CarbonHydrogen DIN 51732NitrogenSulphur DIN

instrumentOxygen By

differenceInductively Coupled PlasmaOptical Emission Spectrometry(ICP-OES)

SiO2, Al2O3, CaO, MgO,K2O, Na2O, Fe2O3, P2O5,TiO2

DIN51729-11

ICP-OES/AAS Mn, Sr, Ba, Cu, Zn, Cr.etc. DIN22022-(1-6)

Table 2Proximate and ultimate analyses of El-Lajjun oil shale.

Proximate analysisWater (ar,%)a 0.78Ash (wf,%)b 54.99Volatile (waf,%)c 99.95Fixed carbon (waf,%) 0.0

Ultimate analysisC (waf,%) 54.33H (waf,%) 5.32N (waf,%) 0.86S (waf,%) 9.03O (diff,%)d 30.44LHV (waf, kJ/kg) 19514.98

a As received.b Water free.c Water ash free.d By difference.

Fig. 1. Schematic diagram of the 20 kW Once-Through furnace.

Table 3Input parameters during El-Lajjun oil shale unstaged combustion.

Air OF27a OF27 OF27

Gas feeding rateCarrier gas (Nm3/h) 1.5 1.8 1.8 1.8Primary gas (Nm3/h) 3.40 3.30 3.30 3.30Secondary gas (Nm3/h) 5.10 4.90 4.90 4.90

Oil shale feeding rate (kg/h) 3.36 4.48 4.70 4.90Overall oxygen ratio (k) 1.19 1.14 1.09 1.04O2 excess (vol.%) 3 3 2 1Wall temperature (�C) 1200 1200 1200 1200

a OF27: 27% O2/73% CO2 environment.

L. Al-Makhadmeh et al. / Fuel 103 (2013) 421–429 423

injected through annular clearances, divided into primary and sec-ondary air. To simulate an oxy-fuel combustion environment, al-most pure O2 and CO2 was mixed in a highly flexible mixingstation and supplied to the furnace, however there was no fluegas recirculation in this study. Gas samples are collected by meansof an oil-cooled sampling probe which can be moved verticallyalong the central axis of the reactor from the burnout section tothe burner mouth with high flexibility. The ash samples are alsocollected via the oil-cooled sampling probe. They are separatedfrom the flue gas at the end of the oil-cooled probe by means ofa filter and are immediately collected in glass containers for labanalysis. Detailed information about the feeding system and theburner can be found in [12].

2.3. Experimental parameters and approach

Oil shale combustion experiments were performed under airand oxy-fuel conditions. Unstaged and staged oil shale combustionwere investigated. All the combustion experiments were per-formed at a wall temperature of 1200 �C. Tables 3–5 illustratethe input parameters for El-Lajjun oil shale combustion under airand oxy-fuel firings, unstaged and staged conditions.

For the analysis of oil shale burnout, ash samples were collectedduring unstaged combustion at different reactor positions. Thesesamples were submitted to ultimate and proximate analyses. Toestimate oil shale burnout, the total ash tracer technique is used.The oil-shale burnout is calculated using the following equation:

xi ¼1� Ash0

Ashi

1� Ash0ð1Þ

where Ash0 and Ashi: represent the ash content of oil shale andpartly burnt char on a dry basis, respectively.

3. Results and discussion

In results explanation and interpretation it was referred to coalresults, since to the authors best knowledge there is no publishedwork on oil shale combustion under oxy-fuel conditions.

3.1. Combustion behaviour, kinetics, and emission during unstagedcombustion

Fig. 2 shows the axial gaseous concentration profiles during un-staged combustion for air-firing and OF27 combustion of El-Lajjunoil shale. OF21 (O2 = 21%) combustion was not investigated, since itis found from previous studies on oxy-coal combustion that adestabilized flame and a delay is encountered in combustion dur-ing bituminous coals combustion using a self-sustained reactor[13]. From Fig. 2A of oxygen consumption, it is clear that the oxy-gen concentration reaches the experimental setup concentration ataround 0.9 m from the burner under air-firing and OF27 combus-tion conditions. The oxygen consumption ratio (relative to the ini-tial input oxygen concentration) is compared when the exit oxygenconcentration reaches the experimental setup value (3%), it isfound to be 86% and 84% for OF27 combustion and air-firing,respectively. In addition, it is noticed that O2 consumption is fasterfor OF27 combustion compared with air-firing.

On the other hand, as expected, the CO concentration in theburner zone is much higher during OF27 compared with air-firing,which is similar to the findings during coal combustion [12]. This isexplained by the gasification reactions in a CO2-rich medium. COprofile during OF27 combustion is wider than that of air-firing; amaximum CO value formed at around 0.5 m from the burner, whilefor air-firing a peak is formed at 0.7 m from the burner. Neverthe-less, the exit CO concentration for both cases is lower than

Table 4Input parameters during El-Lajjun oil shale staged air-firing.

Staged combustion- air

O2 excess 3% (vol)Overall oxygen ratio (k) 1.19Burner oxygen ratio 0.75 0.85 0.95

Volume (Nm3/h) Mass (kg/h) Volume (Nm3/h) Mass (kg/h) Volume (Nm3/h) Mass (kg/h)

Oil shale – 4.54 – 4.00 – 3.58Carrier air 1.8 – 1.8 – 1.8 –Primary air 2.7 – 2.7 – 2.7 –Secondary air 4.1 – 4.1 – 4.1 –Staged aira 4.98 – 3.38 – 2.13 –

aBurnout probe position at 2.20 m (3 s).

Table 5Input parameters during El-Lajjun oil shale staged combustion with 27% O2/73% CO2.

Staged combustion- 27%O2/73%CO2

O2 excess 3.00% [vol]Overall oxygen ratio k 1.14Burner oxygen ratio 0.75 0.85 0.95

Volume (Nm3/h) Mass (kg/h) Volume (Nm3/h) Mass (kg/h) Volume (Nm3/h) Mass (kg/h)

Oil shale – 5.83 – 5.14 – 4.60Carrier air 1.8 – 1.8 – 1.8 –Primary air 2.7 – 2.7 – 2.7 –Secondary air 4.1 – 4.1 – 4.1 –Staged aira 4.48 – 2.95 – 1.74 –

aBurnout probe position at 2.20 m (3 s).

0

5

10

15

20

25

30

0 0.5 1 1.5 2 2.5

O2,

vol%

distance from burner, m

air-3%

OF27-3%

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

2

CO

, vol

%

distance from burner, m

air-3%

OF27-3%

0

100

200

300

400

500

600

700

NO

, ppm

distance from burner, m

air-3%

OF27-3%

0

1000

2000

3000

4000

5000

6000

7000

8000

9000

0 0.5 1 1.5 2 2.5

SO2,

ppm

distance from burner, m

air-3%

OF27-3%

DC

A B

0 0.5 1 1.5 2 2.5

0 0.5 1 1.5 2 2.5

Fig. 2. Axial concentration profiles of oxygen, carbon monoxide, nitrogen oxide and sulphur dioxide for air-firing and OF27 combustion (excess O2 = 3%, T = 1200 �C).

424 L. Al-Makhadmeh et al. / Fuel 103 (2013) 421–429

100 ppm, which means almost no problem for CO controlling forOF27 combustion as well as air-firing. From the error bars, it isclear that there are large fluctuations in CO concentrations in theburner zone especially during OF27 combustion, but the trend isthe same.

From NO concentration profiles, it is found that the NO forma-tion rate near burner region is faster and higher for air-firing,

and reaches a peak at approximately 0.5 m from the burner. How-ever, during OF27 combustion the NO formation rate is lower andthe NO concentration reaches a peak at 0.9 m from the burner.Moreover the NO concentration at the exit of the reactor forOF27 is lower than air-firing.

Fig. 2D shows the axial SO2 concentration profiles during OF27combustion and air-firing. The emission rate is almost the same in

0

5

10

15

20

25

30

0 0.5 1 1.5 2 2.5

O2,

vol%

distance from burner, m

OF27-3% OF27-2% OF27-1%

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

0 0.5 1 1.5 2 2.5

CO

, vol

%

distance from burner, m

050100150200250300350400450500

0 0.5 1 1.5 2 2.5

NO

, ppm

distance from burner, m

0

1000

2000

3000

4000

5000

6000

7000

8000

0 0.5 1 1.5 2 2.5

SO2,

ppm

distance from burner, m

Fig. 3. Axial concentration profiles of oxygen, carbon monoxide, nitrogen oxide and sulphur dioxide for oil shale combustion OF27 at different excess oxygen concentrationsand at 1200 �C.

L. Al-Makhadmeh et al. / Fuel 103 (2013) 421–429 425

the burner zone (till 0.3 m), a peak is formed at 0.9 m from the bur-ner during air-firing, while during OF27 the peak is formed around0.7 m. The exit SO2 concentration is lower during OF27 combustioncompared to air-firing. One should keep in mind that OF27 com-bustion experiment was carried out without flue gas recirculation;NOx and SO2 concentrations during combustion are higher due tothe accumulation with the recycled flue gas.

The effect of excess O2 concentration has been studied for OF27combustion in the range 1–3% to find the optimum conditions foroil shale combustion under oxy-fuel conditions, Fig. 3 shows theaxial gaseous concentration profiles. It is found that the oxygenconcentration reaches the experimental setup concentration ataround 0.9 m from the burner for all cases. The O2 consumption ra-tios at that position are 95%, 92% and 86% for excess O2 1%, 2% and3%, respectively. It is clear that as the excess O2 concentration de-creases the CO concentration in the burner zone is higher. Whenthe excess oxygen is 1% a peak of CO is formed at 0.3 m from theburner, while with excess O2 concentration 2% the peak is formedat 0.5 m which is the same for the case of combustion with 3% ex-cess O2. On the other hand, the CO concentration at the exit of thereactor increases with decreasing the excess O2; the CO concentra-tion at the exit of reactor is 784 ppm when the excess O2 concen-tration is 1%. This forms a problem in CO controlling as the excessO2 concentration decreases to 1% and lower. Therefore, oil shalecombustion can be operated with excess oxygen concentration2% and higher but not lower than 2%.

For NO it is found that as the excess O2 concentration decreasesto 1% the NO concentration at the exit of the reactor is lower com-pared to the other two cases while it is almost the same when theoxygen excesses are 2% and 3%. In the burner zone it is not easy todistinguish the effect, but it is clear that for the case when theexcess O2 concentration is 1%, a peak is formed at 0.5 m from theburner, and then a decrease in the NO concentration is noticed.The peak is formed at 0.7 m and 0.9 m from the burner for thecases 2% and 3% excess oxygen concentration, respectively.

As for SO2 profiles, it is again not easy to distinguish the differ-ence in the burner zone, but the exit concentrations shows oppo-site trend as the excess O2 concentration decreases. In otherwords, as the excess O2 decreases, slightly higher SO2 concentra-tion at the exit of the reactor is found. So emission rate will be used(as illustrated below) to find the effect of different excess oxygenon NO and SO2 emissions.

Since the inlet O2 concentrations during air-firing and OF27combustion are not the same and for different excess O2 concentra-tion, there are differences in the oil shale feeding rate during air-firing and OF27 combustion experiments. To have a representativecomparison of the emission, this difference should be considered,therefore, an expression that takes into account the difference inenergy input is required. Emission rate, which is defined as themass of pollutant emitted per energy input, is used to express pol-lutant emission and is given by the following expression:

uk ¼ yk;d � qnk �VG;d

Huð2Þ

where uk is the emission in the flue gas of the species k related toenergy input, g/MJ, yk,d is the volume fraction of species k in the fluegas (dry), ppmv, qnk is the standard density of species k (qnk = Mk/Vmnk), kg/m3 stp, VG,d is the dry flue gas volume, excess air included,related to fuel mass, m3 stp/kg, Mk is the molecular weight of spe-cies k, kg/mol, Vmnk is the standard molar volume of species k, m3

stp/mol, and Hu is the net calorific value of El-Lajjun oil shale, MJ/kg.Fig. 4 shows NO emission at the exit of the reactor under air-fir-

ing and OF27 combustion conditions with different excess oxygenconcentration. It is found that the NO emission rate is significantlyhigher for air-firing case compared to OF27 combustion cases.Lower NO emission under OF27 combustion condition can be ex-plained by the absence of thermal NO, a decrease in fuel NOx for-mation due to lower flame temperature in a CO2 environmentand a possible reduction of NO on char. On the other hand, it is no-ticed that as the excess oxygen decreases from 3% to 2%, a slight

425.4

236.7 228.5

163.3

26.8

14.614.0

10.0

0

5

10

15

20

25

30

0

50

100

150

200

250

300

350

400

450

500

air-3% OF27-3% OF27-2% OF27-1%

Fue

l-N

con

vers

ion

to N

O

NO

, mg/

MJ

Fig. 4. NO emission for El-Lajjun oil shale combustion at 1200 �C.

426 L. Al-Makhadmeh et al. / Fuel 103 (2013) 421–429

decrease in NO emission is obtained. A larger reduction is noticedwhen the excess O2 concentration is reduced to 1%.

This trend can be explained by the overall oxygen ratios sincethey are 1.14, 1.09 and 1.04 for excess oxygen 3%, 2% and 1%,respectively. This means when the excess O2 is lower, there is adeficiency of oxygen to react with N precursors leading to lowerNO emissions. Oil shale nitrogen conversion to NO shows the sametrend, the NO conversion is 26.8% for air-firing, while it is 14.6%during OF27-3% which is significantly lower during oxy-fuel casewith the same excess oxygen ratio. A slight difference in the con-version is noticed as the O2 excess decreases to 2%, the conversionratio is 14%. The NO conversion is 10% for the case of OF27-1%. Itshould be pointed out here that these results are dependent onthe test facility and burner configuration, the burner is basicallya circular jet burner and is mounted at the top of the test facility.Previous air-firing experiments have been carried out using 21 dif-ferent coals in the same test facility, it has reported a conversionrate in the range of 20–50% during un-staged combustion [14].

Moreover, El-Lajjun oil shale combustion under oxy-fuel condi-tion (OF27) results in a significantly lower SO2 emission comparedto air-firing as shown in Fig. 5, which suggests good sulphur reten-tion in ash. It is found that El-Lajjun oil shale S conversion to SO2 is71.8% during air-firing while it is 42.1% under OF27-3% combustionconditions; SO2 emission during oxy-fuel conditions is lower byabout 30%. On the other hand, as the excess O2 concentration de-creases the SO2 emission slightly increases, SO2 conversion is42.7% and 43.5% for excess oxygen 2% and 1%, respectively. Oneshould keep in mind that this investigation was carried out in aOnce Through furnace, the SO2 concentration in the furnace duringOF27 combustion will not represent the SO2 concentrations of anoxy-fuel plant with flue gas recirculation, as there was no accumu-lation of SO2 (there is no dilution with N2) in the furnace.

An interesting fact that oil shale contains high concentrationof calcite CaCO3. Calcite decomposes producing free lime which

7386

4427 4505 4616

71.8

42.1 42.7 43.5

0

10

20

30

40

50

60

70

80

0

1000

2000

3000

4000

5000

6000

7000

8000

air-3% OF27-3% OF27-2% OF27-1%

Fue

l -S

conv

ersi

on t

o SO

2

SO2,

mg/

MJ

Fig. 5. SO2 emission for El-Lajjun oil shale combustion at 1200 �C.

then reacts with SO2 forming CaSO4 according to the followingreactions:

CaCO3 ! CaOþ CO2 ð3Þ

2CaOþ 2SO2 þ O2 ! 2CaSO4 ð4Þ

In oxy-fuel combustion, calcite decomposition is inhibited dueto the high CO2 concentration [15,16], and direct sulphation ofCaCO3 is favored according to:

2CaCO3 þ 2SO2 þ O2 ! 2CaSO4 þ 2CO2 ð5Þ

On the other hand, different characteristics such as larger porediameter of lime produced under oxy-fuel combustion than that un-der air-firing could result in higher S retention. Larger pore diameteryields better sulphation of CaO due to reduced pore filling and plug-ging [17]. Moreover, CaSO4 decomposition may influence S reten-tion; calcium sulphate dissociates at temperatures higher than850 �C depending on the carbonate fraction and the composition ofthe surrounding atmosphere, however the higher concentration ofSO2 in oxy-fuel combustion stabilizes the formed CaSO4 [16,18,19].

Fig. 6 shows S retained in ash at 2.5 m from the burner for thefour investigated unstaged combustion cases, S retention is higherunder OF27-3% conditions compared to air-firing. On other hand,the slight increase in SO2 emission and conversion as the excessO2 decreases can be explained by the low oxygen availability forsulphation, which is important for sulphation reaction as notedfrom Eq. 5 above. Ca has a dominant role in sulphur retention;the Ca/S molar ratio is one of the main characteristics that governsthe retention of sulphur in the ash increasing with the increase ofthis ratio. The molar Ca/S ratio of El-Lajjun oil shale is 5.21.

It can be concluded that oxy-fuel condition is one solution forminimizing SO2 emissions during oil shale (especially, with highS content) combustion. However, the results of this study areobtained using a Once-Through reactor and not a recycled fluegas case which is the real one. This means an expectable lowerSO2 emission values could be obtained than those obtained here.Similar trends were found by Dhungel [12] using different coals.Studies on SO2 emissions during O2/RFG (Recycled Flue Gas) com-bustion reported lower conversion of sulphur to SO2 [20–22].

Fig. 7 shows El-Lajjun oil shale burnout at 1200 �C under un-staged air-firing and OF27 combustion conditions with differentexcess O2 concentration. High volatile and ash contents make com-bustion characteristic of oil shale different from that of coal. Com-bustion intensity of oil shale is strong at the early stage ofcombustion, but significantly reduced at the following burningstage of coke because mass transfer resistance of ash layer willgradually increase and become major controlling factor in burningrate of fixed carbon [23]. Nevertheless, 100% oil shale burnout isachieved for all investigated cases.

0

0.5

1

1.5

2

2.5

3

3.5

air -3% OF -3% OF -2% OF -1%

S in

ash

at

2.5m

, %

Fig. 6. S content of oil shale unburned char at 1200 �C during air-firing and OF27combustion with different excess O2.

0

50

100

150

200

250

300

350

400

450

0 0.5 1 1.5 2 2.5

NO

, ppm

Distance from burner, m

air-0.75 air-0.85 air-0.95OF-0.75 OF-0.85 OF-0.95

Fig. 9. Axial nitrogen oxide concentration profiles for oil shale air-firing and OF27combustion at 1200 �C (burner oxygen ratio: 0.75, residence time: 3 s).

425.4

236.7

200

250

300

350

400

450

500

NO

, mg/

MJ

air

OF27

0

0.2

0.4

0.6

0.8

1

0 0.5 1 1.5 2 2.5

Bur

nout

, %

Distance from burner, m

air -3%

OF -3%

OF -1%

Fig. 7. Oil shale burnout at 1200 �C during unstaged air-firing and OF27 combus-tion at different excess O2 concentrations.

L. Al-Makhadmeh et al. / Fuel 103 (2013) 421–429 427

3.2. Staged combustion

3.2.1. NOx reduction during oil shale combustion under oxy-fuelconditions

As noticed from unstaged results NO emission and conversiondecreases with the decrease of overall oxygen ratio. To investigatethis parameter further and to study NO formation and reductionmechanism, staged combustion has been investigated for oil shaleOF27 combustion as well as air-firing. Burner oxygen ratios were0.75, 0.85 and 0.95.

Fig. 8 shows the axial concentration profiles of NO, O2 and COduring OF27 combustion with a burner oxygen ratio of 0.75 andresidence time of 3 s. It is noticed that a peak is formed at around0.3 m from the burner, then a sharp decrease in NO concentrationis found reaching zero at around 1.1 m from the burner, followedby an increase at the position of staging. The reduction of NO is ex-plained by conversion of NO to N precursors due to reactions withhydrocarbons and tar (N pyrolysis in oxygen deficient zone). It isfound that NO concentration is smaller in the burner zone forOF27 combustion case compared to air-firing case.

As for oxygen profile, it decreases rapidly reaching zero concen-tration around 0.9 m, the area before that is the region where mostof the oil shale mass and nitrogen are devolatilized [24]. An in-crease of oxygen concentration is found at the position of injectionthe burnout oxygen which explains the increase in NO concentra-tion; so N precursors react with available O2 resulting in higher NOconcentration at the exit of the reactor. On the other hand, CO con-centration profile shows high values in the reduction zone. CO con-centration decreases to almost zero after injecting burnout oxygenwhich supply O2 for oxidizing CO to CO2. However, CO concentra-tions in the sub-stoichiometric zone are higher for OF27 combus-tion as compared to air-firing. This is due to an enhancedreaction of CO2 with carbon and consumption of H2 to produceCO by the water–gas reaction [25,26]. Similar trends of NO, O2

0

5

10

15

20

25

30

0

50

100

150

200

250

300

350

0 0.5 1 1.5 2 2.5

O2

and

CO

, vol

%

NO

, ppm

distance from burner, m

NO

O2

CO Staging position

Fig. 8. Axial concentration profiles of oxygen, carbon monoxide and nitrogen oxidefor oil shale OF27 combustion at 1200 �C (burner oxygen ratio: 0.75, residence time:3 s).

and CO profiles for the case of air-firing were found (profiles arenot shown here).

Fig. 9 shows NO concentrations profiles under air-firing andOF27 combustion conditions at different burner oxygen ratios(0.75, 0.85 and 0.95). As seen, the lower the burner oxygen ratio,the lower the final NO concentration, for OF27 combustion. Whileit is noticed that there is no difference in NO concentration at theexit of reactor between the cases 0.75 and 0.85 burner oxygen ratiofor air-firing. A peak of NO is formed in the fuel-oxidant mixingzone for all studied experimental conditions, as sufficient oxygenis still present in this region. The location of NO peak, for mostcases, is slightly delayed for OF27 combustion. It is noticed thatat burner oxygen ratio of 0.95 NO reduction continues even afterinjecting the burnout oxygen, this can be explained by the hetero-geneous reduction on char active surface.

Emission based on the mass of pollutant per energy input, isused when comparing NO emissions at the furnace exit as shownin Fig. 10. It is clear that staging causes a high reduction in NOemission and as the burner oxygen ratio decreases higher NOreduction is noticed for both air-firing and OF27 combustion. Thisis due to the higher hydrocarbons content as burner oxygen ratiodecreases. Another reason could be the heterogeneous NO reduc-tion on char active sites. The high volatile content of El-Lajjun oil

31.333.1

74.7

24.2 35.7

77.5

0

50

100

150

0.75 0.85 0.95 unstaged

Fig. 10. NO emissions for oil shale during staged and unstaged combustion at1200 �C.

Table 6El-Lajjun oil shale N conversion to NO under air-firing and OF27 conditions.

Condition Fuel-N conversion to NO,%

Air-firing OF27

0.75 2.82 1.910.85 2.74 2.610.95 5.71 5.26Unstaged 26.8 14.6

428 L. Al-Makhadmeh et al. / Fuel 103 (2013) 421–429

shale may has a rule as well since oil shale nitrogen releases rap-idly in the gas phase which is then reduced during the reductionzone. Kluger et al. [14] have also reported a sharper decrease inthe NO concentration during staged air-firing with coals containinghigher volatile content, or coals that release more fuel-N via thegas phase. Table 6 shows El-Lajjun oil shale N conversion to NOas well.

It can be concluded that more reduction of NO is achieved underoil shale staged air-firing as well as staged OF27 combustion con-ditions. Oxidant staging during OF27 combustion without flue gasrecirculation is efficient for NO reduction as well as air-firing overthe investigated burner stoichiometry. In addition, as the burneroxygen ratio decreases from 0.85 to 0.75 a slight difference is no-ticed, so such technique can be applied at non-extreme reducingconditions to avoid corrosion problem due to the high H2S concen-tration in the reducing zone as shown later.

3.2.2. SO2 and H2S formation mechanism during oil shale air-firing andoxy-fuel combustion

Fig. 11 shows SO2 and H2S axial concentration profiles under oilshale OF27 combustion condition with burner oxygen ratio of 0.75and residence time of 3 s. OF27 combustion experiment was car-ried out without flue gas recirculation. It is noted that SO2 concen-tration reaches a peak around 0.3 m from the burner as well, thenit decreases till 0.9 m from the burner. While H2S concentration in-creases reaching a peak around 0.5 m from the burner then de-creases reaching zero at the exit of the reactor. From the sumvalues of H2S and SO2, it is found that the maximum release of Sis around 0.15–0.5 m from the burner. Therefore, it can be con-cluded that sulphation takes place after 0.5 m from the reactor.

It is found that H2S concentration is extremely high under allinvestigated conditions especially for OF27 combustion cases. Inaddition, the maximum peaks are delayed for air-firing at 0.75and 0.85 burner oxygen ratios. This result suggests that attention

0.250.13

0.56

0.840.94 0.94

0.00

0.20

0.40

0.60

0.80

1.00

air -0.75 OF -0.75 air -0.85 OF -0.85 air -0.95 OF -0.95

SO2

or H

2S f

ract

ion

at 0

.9m

fr

om th

e bu

rner

, %

H2S

SO2

Fig. 12. SO2 and H2S volumetric fraction at 0.9 m from the burner during oil shaleair-firing and oxy-fuel staging combustion.

0

5

10

15

20

25

30

0

2000

4000

6000

8000

10000

12000

14000

16000

18000

0 0.5 1 1.5 2 2.5

O2,

vol

%

SO2,

H2S

(ppm

vd)

distance from burner, m

SO2 H2S SO2+H2S (SO2+H2S)max O2

Fig. 11. Axial concentration profiles of SO2, H2S and O2 for oil shale OF27combustion at 1200 �C (burner oxygen ratio: 0.75, residence time: 3 s).

should be paid to corrosion issue especially if staging is involvedfor NO reduction under oxy-fuel conditions, in particular for suchfuel with high S content. SO2 and H2S volumetric fractions at0.9 m from the burner for air-firing and OF27 combustion at differ-ent burner oxygen ratios 0.75, 0.85 and 0.95 are shown in Fig. 12.The SO2 and H2S fraction at 0.9 m from the burner shows that asthe burner oxygen ratio increases from 0.75 to 0.95, the H2S frac-tion decreases from (0.75 to 0.06 vol.%) and (0.87 to 0.06 vol.%),while SO2 fraction increases from (0.25 to 0.94 vol.%) and (0.13 to0.94 vol.%) for air-firing and OF27 combustion, respectively.

4. Conclusion

Direct combustion of Jordanian oil shale under oxy-fuel condi-tions is the first of its kind. Air-firing as well as staged and un-staged combustion at 27% O2/73% CO2 (OF27) was conductedsuccessfully. A 20 kW electrically heated Once Through furnace,enabling highly flexible parametric studies as well as reliable andrepeatable measurements, was used for this investigation. To sim-ulate an oxy-coal combustion environment, almost pure O2 andCO2 was mixed in a highly flexible mixing station and suppliedto the furnace. OF27 combustion tests were conducted without fluegas recirculation. It was found that Jordanian El-Lajjun oil shalecombustion under oxy-fuel conditions is feasible; 100% oil shaleburnout is achieved for OF27 combustion with 3% and 1% as wellas air-firing.

The CO concentration in the burner zone is much higher underOF27 compared to air-firing, which is explained by the gasificationreactions in a CO2-rich medium. Meanwhile, the exit CO concentra-tion for both cases is lower than 100 ppm, which means almost noproblem for CO controlling for OF27 combustion as well as air-fir-ing. On the other hand, during OF27 combustion with different ex-cess O2 it is found that the CO concentration at the exit of thereactor increases with the decrease of the excess O2. The CO con-centration at the exit of reactor is 784 ppm when the excess O2

concentration is 1%. This means that controlling CO will be a prob-lem as the excess O2 concentration decreases to 1% and lower.

The conversion of N to NO in the furnace exit during unstagedcombustion is seen to be lower for OF27 combustion at differentexcess oxygen compared to air-firing. In addition, staging can beused efficiently also for oil shale combustion during oxy-fuel con-ditions to lower NO emission; a considerable reduction in NO re-lease is achieved during oil shale staged OF27 combustion aswell as under staged air-firing condition. Attention should be paidto corrosion issue so that such technique can be applied at not ex-treme reducing conditions to avoid corrosion problem due to thehigh H2S concentration in the reducing zone especially underoxy-fuel conditions.

SO2 emission during OF27 combustion at different excess oxy-gen concentration is significantly lower than air-firing; SO2 emis-sion during oxy-fuel conditions is lower by about 30%.Sulphation have been occurred during OF27 which was by theanalyses of S in the collected ash at 2.5 m from the burner. There-fore, unstaged oxy-fuel combustion presents a solution to lowerSO2 emission for such high S content in oil shale. In other words,oil shale combustion with lower SO2 and NO emissions can beachieved using oxy-fuel conditions.

The current results add basic knowledge to the understandingof the behaviour of oil shale combustion under oxy-fuel conditionsas well as air-firing, although further conditions and parametersespecially injections of NO and SO2 through the burner to simulatethe real situations during oxy-fuel combustion in terms of gasquality and its effects should be studied. So the next focus in thisfield will be, therefore, the determination of the fate of NO andSO2 recycled back into the furnace by injecting a known concentra-tion of pure NO and SO2 via the burner with O2/CO2 mixture.

L. Al-Makhadmeh et al. / Fuel 103 (2013) 421–429 429

Acknowledgments

The authors would like to thank Mattias Pagano, Roland Kuhn,Mario Krautz, Wolfgang Ross and Mario Wolf from Institute ofCombustion and Power Plant Technology for their help in perform-ing the experimental program and lab analysis. Thanks also goes toJordanian Natural Resources Authority (Dr. Ziad Hamarneh,Mr. Marwan Mdanat, Mr. Ahmad Abu Risha and Eng. Samer Mne-sel) for facilitating collection of oil shale samples from El-Lajjunarea in Jordan.

References

[1] Rssei PL. Oil shales of the world oxford. Pergamon Press; 1990.[2] NERC. A guide to NERC. Amman, Jordan: National Energy Research Center;

2002.[3] Williams PT, Nazzal JM. Polycyclic aromatic compounds in shale oils: influence

of process conditions. Environ Technol 1998;19:775–87.[4] Külaots I, Goldfarb JL, Suuberg EM. Characterization of Chinese, American and

Estonian oil shale semicokes and their sorptive potential. Fuel2010;89:3300–6.

[5] Howard JR. Fluidized beds: combustion and applications. London: AppliedScience Publishers; 1983.

[6] Teinemaa E, Kirso U, Strommen MR, Kamens RM. Atmospheric behaviour ofoil-shale combustion fly ash in a chamber study. Atmos Environ2002;36:813–24.

[7] Kallaste T, Liik O, Ots A. Possible energy sector trends in Estonia. Context ofclimate change, Tallinn. Estonia; 1999.

[8] Miao Z-Y, Wu G-G, Li P, Zhao N, Wang P-C, Meng X-L. Combustioncharacteristics of Daqing oil shale and oil shale semi-cokes. Mining SciTechnol (China) 2009;19:380–4.

[9] Kattenberg A, Giorgi F, Grassl H, Meehl GA, Mitchell JF, Stouffer RJ, et al.Climate models – projections of future climate, in climate change. In:Houghton JT, Filho LGM, Callander BA, Harris N, Kattenberg A, Maskell K,editors. The science of climate change. Cambridge, UK: Cambridge UniversityPress; 1996. p. 285–357.

[10] Scheffknecht G, Al-Makhadmeh L, Schnell U, Maier J. Oxy-fuel coal combustion– a review of the current state-of-the-art. Int J Greenhouse Gas Control2011;5:S16–35.

[11] Al-Harahsheh A, Al-Otoom AY, Shawabkeh RA. Sulfur distribution in the oilfractions obtained by thermal cracking of Jordanian El-Lajjun oil Shale. Energy2005;30:2784–95.

[12] Dhungel B. Experimental investigations on combustion and emissionbehaviour during oxy-coal combustion, Institute of Combustion and PowerPlant Technology/University of Stuttgart. PhD Thesis; 2010.

[13] Liu H, Zailani R, Gibbs BM. Comparisons of pulverized coal combustion in airand in mixtures of O2/CO2. Fuel 2005;84:833–40.

[14] Kluger F, Foertsch D, Spliethoff H, Schnell U, Hein KRG. Comparison of coals forunstaged and air staged combustion with respect to NOx emission. In: 23rdInternational technical conference on coal utilization and fuel systems,Florida; 1998.

[15] Fleig D, Normann F, Andersson K, Johnsson F, Leckner B. The fate of sulphurduring oxy-fuel combustion of lignite. Energy Procedia 2009;1:383–90.

[16] Yrjas P, Iisa K, Hupa M. Comparison of SO2 capture capacities of limestones anddolomites under pressure. Fuel 1995;74:395–400.

[17] Chen C, Zhao C, Liang C, Pang K. Calcination and sintering characteristics oflimestone under O2/CO2 combustion atmosphere. Fuel Process Technol2007;88:171–8.

[18] Cheng J, Zhou J, Liu J, Zhou Z, Huang Z, Cao X, et al. Sulfur removal at hightemperature during coal combustion in furnaces: a review. Prog EnergyCombust Sci 2003;29:381–405.

[19] Liu H, Okazaki K. Simultaneous easy CO2 recovery and drastic reduction of SOx

and NOx in O2/CO2 coal combustion with heat recirculation. Fuel2003;82:1427–36.

[20] Kiga T, Takano S, Kimura N, Omata K, Okawa M, Mori T, et al. Characteristics ofpulverized-coal combustion in the system of oxygen/recycled flue gascombustion. Energy Convers Manage 1997;38:S129–34.

[21] Croiset E, Thambimuthu KV. NOx and SO2 emissions from O2/CO2 recycle coalcombustion. Fuel 2001;80:2117–21.

[22] Woycenko DM, van de Kamp WL, Robert PA. Combustion of pulverized coal ina mixture of oxygen and recycled flue gas [Summary of the APG ResearchProgram]. Joule II program, IFRF Doc F98/Y/4; 1995.

[23] Dockter L, Turner TF, Combustion rates for oil shale carbonaceous residue. Insitu: oil-coal-shale minerals, vol. 2; 1978.

[24] Bose AC, Dannecker KM, Wendt JOL. Coal composition effects on mechanismsgoverning the destruction of NO and other nitrogenous species during fuel-rich combustion. Energy Fuel 1988;2:301–8.

[25] Al-Makhadmeh L, Maier J, Scheffknecht G. Coal pyrolysis and char combustionunder oxy-fuel conditions. In: 34th International technical conference on coalutilization & fuel system, Clearwater, Florida; 2009.

[26] Al-Makhadmeh L. Coal pyrolysis and char combustion under oxy-fuelconditions. Institute of Combustion and Power Plant Technology/Universityof Stuttgart. PhD Thesis; 2009.