E-Beam SO2 and NOx removal from flue gases in the presence of fine water droplets

Page 1

Radiation Physics and Chemistry 85 (2013) 130–138

Contents lists available at SciVerse ScienceDirect

Radiation Physics and Chemistry

0969-80

http://d

n Corr

E-m

chmiele

daniel.i

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

E-Beam SO2 and NOx removal from flue gases in the presenceof fine water droplets

Ioan Calinescu a, Diana Martin b, Andrezj Chmielewski c, Daniel Ighigeanu b,n

a University POLITEHNICA of Bucharest, ]149 Calea Victoriei St., 010072 Bucharest, Romaniab National Institute for Lasers, Plasma and Radiation Physics, Electron Accelerators Laboratory, P.O.Box: MG-36, ]409 Atomistilor St., 077125 Magurele, Ilfov, Romaniac Institute of Nuclear Chemistry and Technology, ]16 Dorodna St., 03-195 Warsaw, Poland

H I G H L I G H T S

c The medium-energy EB accelerators are proposed for flue gases treatment.c The energy losses in the windows and in the air gap between them are reduced.c To increase the density of the reaction medium and to reduce the penetration depth of EB fine water droplets (FWD) are used.c Determining the energy efficiency the favorable effect of the method was demonstrated.c The maximum amount of FWD was determined from the total energy balance of the process.

a r t i c l e i n f o

Article history:

Received 30 May 2012

Accepted 11 October 2012Available online 20 November 2012

Keywords:

Flue gases

NOx and SO2 removal

Electron beam

Fine water droplets

6X/$ - see front matter & 2012 Elsevier Ltd.

x.doi.org/10.1016/j.radphyschem.2012.10.008

esponding author. Tel.: þ40 214574346; fax

ail addresses: [email protected] (I. Calinescu

[email protected] (A. Chmielewski),

[email protected] (D. Ighigeanu).

a b s t r a c t

The Electron Beam Flue Gas Treatment (EBFGT) has been proposed as an efficient method for removal of

SO2 and NOx many years ago. However, the industrial application of this procedure is limited to just a

few installations. This article analyses the possibility of using medium-power EB accelerators for off-

gases purification. By increasing electron energy from 0.7 MeV to 1–2 MeV it is possible to reduce the

energy losses in the windows and in the air gap between them (transformer accelerators can be applied

as well in the process). In order to use these mid-energy accelerators it is necessary to reduce their

penetration depth through gas and this can be achieved by increasing the density of the reaction

medium by means of dispersing a sufficient amount of fine water droplets (FWD). The presence of FWD

has a favorable effect on the overall process by increasing the level of liquid phase reactions. A special

reactor was designed and built to test the effect of FWD on the treatment of flue gases with a high

concentration of SO2 and NOx using high-energy EBs (9 MeV). By determining the energy efficiency of

the process the favorable effect of using FWD and high-energy EB was demonstrated.

& 2012 Elsevier Ltd. All rights reserved.

1. Introduction

New technologies and processes have promoted our economicgrowth and have been remarkably changing our lifestyles. How-ever, technological innovation is not always welcome. It hasimpact on the environment and may have effects related to thepeople’s health. The emission of inorganic pollutants such asnitrogen oxides (NOx) and sulfur dioxide (SO2) has been remark-ably reduced by application of state-of-the-art technologiesduring the past 60 years. However, industrialized countries havebeen using the big amounts of oil and coal for energy production

All rights reserved.

: þ40 214574243.

),

via combustion of fossil fuels to convert chemical energy of thesubstrate into electricity, vehicles fuels and process heat to beused in industries.

Although the developing countries also try to increase theconsumption of environment-friendly energies instead of fossilfuels, the steep demand for energy due to the recent economicgrowth makes it difficult or even impossible to reduce the usageof fossil fuels. This causes a large amount of emission of NOx, SO2,as well as carbon dioxide to the atmosphere (Gaffney et al., 1987;Kato and Akimoto, 2007; Ramanathan and Feng, 2009; Streets andWaldhoff, 2000).

While the industrial countries have achieved a sufficientreduction of NOx and SO2 emissions, it is still an urgent issuefor the developing countries to follow this trend. To meet thestrict regulations established by local governments, the wet limescrubber method and the selective catalytic reduction method

Page 2

I. Calinescu et al. / Radiation Physics and Chemistry 85 (2013) 130–138 131

have been applied to treat SO2 and NOx. However, the wet limescrubber method requires wastewater treatment, and the cata-lysts have to be replaced periodically. New technology is expectedfor simple and simultaneous treatment processing of the bothpollutants. The electron-beam irradiation process for flue gaspurification (EBFGT) has been proposed as an efficient methodbecause it has the following advantages (Hatano et al., 2011):

�

has simultaneous denitrification and desulfurization possibilities, � no wastewater treatment is required, � no expensive catalyst is required, � provides simple process and its operation, �

Fig. 1. The main reactions used in EBFGT to convert SO2 and NOX (Namba et al., 1998).

produces profitable products.

The application of electron beams to treat flue gases, such asremoving sulfur dioxide, was started by Ebara Corporation inJapan and in the US (Chmielewski, 2011).

There are only few pilot and industrial installations for fluegases treatment by electron beam (EB) irradiation (Chmielewskiet al., 2004) operated at this moment. Its drawback, like in otherconventional technologies, is high energy consumption (thenecessary power for the electron beam is around 2–4% from thetotal electrical energy produced by the plant) (Hackman andAkiyama, 2000; Licki et al., 2003) and the difficulties in operationof very high power accelerators. That is why it is mandatory tofind new solutions and to develop strategies for diminishingradiation dose absorbed in the flue gas and to optimize theprocess for better uses of EB energy.

Subjected to the electron beam irradiation process, the maincomponents of flue gases (N2, O2 and H2O), are transformed intodivergent ions and radicals. The primary species include: e� , N2

þ ,Nþ , O2

þ , Oþ , H2Oþ , OHþ , CO2þ , COþ , N2

n, O2n, N, O, H, OH and CO

(Matzing and Paur, 1992; Tokunaga and Suzuki, 1988). In the caseof high water concentration, oxidizing radicals HO and HO2 andexcited species such as O(3P) are the most important productformed (Person and Ham, 1988). In the presence of waterdroplets, the radiolytically produced hydrated electron reactsvery fast with the dissolved oxygen to produce the superoxide(O2�) radical. Since the O2

� has a pKa value of 4.7, it will beconverted to HO2 in acidic medium. HO2 is also produced by thereaction of H-atom with oxygen. However the yield from thisreaction is only about 0.06 mmol per joule.

N2,O2,H2O vapors_at_high_concnð Þ þe�-HO�,HO2�,On,ions,excited-species

ð1Þ

In the presence of these reactive species, NOx and SO2 fromflue gases are oxidized and produce nitric acid and sulfuric acid,respectively, as intermediate products. These acids are neutra-lized with ammonia, giving powders of ammonium nitrate andammonium sulfate, respectively (Namba et al., 1998) (Fig. 1).

The total yield of SO2 removal consists of the yield of thermaland radiation induced reactions that can be written (Chmielewskiet al., 1995a; Matzing et al., 1993):

ZSO2¼ Z1 F,Tð ÞþZ2 D,aNH3

,T� �

ð2Þ

The yield of the thermal reaction depends on the temperatureand humidity. The yield of the radiation induced reaction dependson the dose, temperature and ammonia stoichiometry. The mainparameter in NOx removal is the dose.

The presence of water vapors is mandatory for the efficientremoval of SO2 and NOx from gases. By radiolysis of water the HOradical is obtained and this radical will oxidize both SO2 and NOx

(Genuario, 2009):

NOþHO�þM-HNO2þM (3)

NO2þHO�þM-HNO3þM (4)

SO2þHO�þM-HS�O3þM (5)

HS�O3þO2-SO3þHO2� (6)

SO3þH2O-H2SO4 (7)

NOþHO2�-NO2þHO� (8)

A gas humidity 8–10 vol% is necessary to obtain the optimalremoval-efficiencies of both pollutants (Chmielewski, 2007).

The selection of an adequate accelerator is an important issuein the process engineering. The primary effect of any ionizingradiation is based on its ability to excite and ionize molecules, andthis leads to the formation of free radicals, which then initiatechemical reactions.

Accelerated electron beams suitable for flue gas treatmenthave sufficient energy (up to 5 MeV) to affect the electrons in theatom shell, but not its nucleus, and can therefore only initiatechemical reactions (Zimek, 1995). Typically, the reactionsinitiated by electron beam are extremely fast and are completedin fractions of a second.

Electrons that are capable of electronically exciting and ioniz-ing molecules, such as N2, O2, CO2, H2O etc., must have energies inthe range from 12 to 16 eV. Such electrons can be produced fromfast electrons by the energy degradation process in solids, liquids,and gases. These secondary electrons show energy distributionwith the maximum in the range from 50 to 100 eV. In contrast tofast electrons, exhibiting energies in the keV and MeV range,secondary electrons are capable of penetrating solids and liquidsonly a few nanometers. Consequently, they generate ions, radi-cals, and excited molecules in ‘‘droplets’’ along the paths of thefast electrons (Drobny, 2010). Fig. 2 illustrates schematically theprocess of generation of reactive species.

The basic fundamental properties for the EB machine; beamcurrent and beam energy should be derived from the processrequirements to ensure satisfactory results with minimum capitaland operating costs. The power of EB is determined function ofthe flow rate of the treated material and the absorbed doseneeded for the process:

DðaveÞ ¼ FðpÞP= M=T� �

ð9Þ

Page 3

Fig. 2. The process of generation of reactive species by high energy electrons.

Fig. 3. Electron energy losses function of incident electron energy for flue gases

treatment by EB (Cleland, 2007).

Fig. 4. Electron energy deposition vs. thickness�density in water at 1.0, 1.5, 2.0,

2.5, 3.0, 3.5 4.0, 4.5 and 5.0 MeV incident electron energy (IAEA, 2010).

I. Calinescu et al. / Radiation Physics and Chemistry 85 (2013) 130–138132

where D(ave) is the average absorbed dose in kGy, P is the emittedradiation power in kW, T is the treatment time in s and M is themass of irradiated material in kg.

The factor F(p) is the fraction of emitted power absorbed by thematerial, which depends on the size, shape, thickness and densityof the object and the penetrating quality of the radiation. Theenergy absorption in the primary and secondary window shouldbe considered in the case of EBFGT. This quantity is difficult tomeasure (by using dosimeters placed inside the reaction vessel),but it can be calculated by Monte Carlo simulation. It can rangefrom 0.25 to 0.75, depending on the particular application.

EB energy losses in the beam windows and in the air spacebetween them are high for the low energy–high power accel-erators that are presently used for air pollutants removal(Chmielewski et al., 1995b). The total EB energy losses (back-scattering, beam windows and air gap) are substantially lowerwith higher EB incident energy (see Fig. 3). The useful EB energyfrom input energy is around 50% for 0.5 MeV and about 95% above3 MeV. The lower electrical efficiencies of accelerators withhigher energies are partially compensated by the lower electronenergy losses in the beam windows. In addition, accelerators with

higher electron energies can provide higher beam powers withlower beam currents (Cleland, 2007).

Much attention should be paid to the range of the electron beampenetration. The penetration of high-energy (relativistic) electronbeams in irradiated materials increases linearly with the incidentenergy. The electron pathway range also depends on the atomiccomposition of the irradiated material. The energy deposition iscaused mainly by collisions of the incident electrons with atomicelectrons. Therefore, materials with higher electron contents (elec-trons per unit mass) will absorb higher doses near the entrancesurface, but shorter electron penetration ranges (Cleland, 2005).

Fig. 4 shows the depth–dose distribution curves for beamenergies between 1.0 and 5.0 MeV in centimeters of water asderived from Monte Carlo calculations using the ITS3 (IntegratedTiger Series code) (Cleland, 2004).

For mid-energy (500 keV to 5 MeV) and high-energy (5 MeV to10 MeV) electron accelerators, it is common to express beampenetration on the basis of equal-entrance and exit doses in unitdensity material. In Fig. 5 are presented the calculated values ofpenetration depth (on the basis of practical range R(p) for fluegases and for flue gases with fine water droplets (FWD), thevolume fraction of water is designated as L¼(Vaq/Vgas), where Vaq

and Vgas refer to the irradiated volumes for the two phases,L¼4�10�3.

The range parameters can be correlated with the incidentelectron energy E with sufficient accuracy for industrial applica-tions by using the following linear equations (for polyethylene)(Cleland, 2005):

RðpÞ ¼ 0:510E�0:145 ð10Þ

Electron ranges in other materials can be estimated by multi-plying the polyethylene range with the ratio of their CSDA ranges.

RðmaterialÞ ¼ RðpolyethyleneÞ � CSDAðmÞ=CSDAðpeÞ ð11Þ

CSDA ranges for many materials with a wide range of electronenergies can be obtained from ICRU Report 37 or from (http://www.nist.gov/pml/data/star/index.cfm).

Supposing exhaust gases at a temperature of 60 1C (densityr¼1.06�10�3 g/cm3) are treated by an electron acceleratorsequipped with a Ti foil (50 mm, r¼4.5 g/cm3) window, air gap15 cm and another Ti foil for the reactor (50 mm), the range ofelectrons generated from 750 and 2500 kV are 235 and 1240 cm,respectively (see Fig. 5). The depth of the reactor in the irradiationdirection must be less than these values to avoid non-irradiationspace in the reactor.

Page 4

Fig. 5. The penetration depth of electron beam function of electron energy in flue

gases (r¼1 g/L) and in flue gases with FWD (L¼4n10�3, r¼5 g/L).

Fig. 6. The influence of liquid water fraction (L) in flue gases on the density of flue

gases with FWD and penetration depth for EB of 2.5 MeV.

I. Calinescu et al. / Radiation Physics and Chemistry 85 (2013) 130–138 133

Although the medium-energy EB (2.5 MeV) are used with highefficiency (with lower total energy losses, see Fig. 3), the penetra-tion depth of these EB is too high (approx. 12.4 m, see Fig. 5). Inorder to reduce the penetration depth, it is necessary to increasethe density of the medium. This can be achieved by dispersingFWD in flue gases. The influence of volume fraction of water (L) influe gases on the density of the gas medium and on the penetra-tion depth of EB with 2.5 MeV is presented in Fig. 6.

Several reasons showing the potential importance of droplet-phase oxidation in the EBFGT process can be outlines (Cooperet al., 1998):

�

High driving force for the most oxidation process due to lowGibbs free energy of the products obtained in liquid phase; � High solubility of the mixture components (SO2 and the

intermediates NO2, N2O4) resulting in an increase in itsliquid-phase concentrations;

� Rapid rates of the aqueous-phase oxidation process initiated

by ionizing radiation.

Regarding the primary partitioning of the dose rates betweenthe irradiated phases, denoted DRgas and DRaq, one notes (Cooperet al., 1998) that its magnitude is governed mainly by the ratio of

the phases’ densities, rgas and raq. Since the ratio raq/rgas is typicallyZ103, then DRaq/DRgas�raq/rgasZ103, assuming that the stoppingpowers of the gas and liquid phase are quite the same. Thus the doserate in the droplet is higher than that of the gas by a factor of at leastthree orders of magnitude. This means that the ratio of the chemicalreactions rates of the pollutants’ oxidation in the two phase systemis controlled by the phases’ densities.

With regard to the portioning of absorbed doses between the twophases, which determines the extent of pollutant removal duringirradiation, one gets (Cooper et al., 1998): Daq/DgasZraq*L/rgas.

When L is quite low (in a typical EBFGT process Lr10�6), it isunlikely that the radiation-induced reactions in the liquid phasecan a play a significant role in the removal of pollutants from theirradiated gas, but if L�10�3 the reactions in liquid phase becomeimportant. For such conditions, the chain oxidation takes place inliquid phase, but the chain is initiated by active particles formedin the gas and liquid phase (Potapkin et al., 1995).

It was found (Yermakov et al., 1995) that the extents of NOx andSO2 depletion on simultaneous irradiation and spraying of water(or water solution) are sensitive not only to absorbed dose but alsoto the amount of water existing in the gas as liquid droplets.

The conversion of SO2 can be obtained by reactions in gasphase (thermal and radio-induced) and by reactions in liquidphase. Because of high value of solubility of SO2 in water (Henry’sconstant¼1.23 Mnatm�1) SO2 is easily absorbed in water andhere the SO2 is oxidized by a chain reaction in the presence ofdissolved oxygen (Potapkin et al., 1995). The conversion of NOtakes place only by radio-induced reactions. In gas phase NO isoxidized to NO2 by reactions of NO with O and O3. In the absenceof water droplets the removal efficiency of NO dropped sharplyafter NO conversion to NO2 (Potapkin et al., 1995). This is becauseof the influence of the reactions with N and NH2 (Matzing, 1989):

NO2þN-2NO ð12Þ

NO2þN-N2OþO ð13Þ

NO2þNH2-N2OþH2O ð14Þ

In the presence of fine water droplets the quenching of NO2 takeplace by dissolution and reaction in droplets of water (the Henry’sconstant of NO2 is 6 times higher than the value for NO:1.23n10�2 Mnatm�1 respectively 0.195n10�2 Mnatm�1 (Squadritoand Postlethwait, 2009).

The dissolved NO2 can be oxidized by OH radicals obtained bywater radiolysis or reduced by reaction with S(IV) compounds(Potapkin et al., 1995):

NO2þOH-HNO3 ð15Þ

No2þHSO�3 -NO�2 þHSO�3 ð16Þ

No�2 þHSO�3þHþ-H2O

SO2�4 þNHþ4 ð17Þ

NO�2 þHSO�3--HSO�4 þN2 ð18Þ

However, the experiments of Yermakov were done at quitelow value of L (about 10�4), at such values the influence of FWDon density of the gas-liquid system is quite low and is notpossible to use the accelerators with mid-energy (about 2 MeV)due to high penetration depth of electrons (see Fig. 5).

Our purpose is to treat flue gases at high values of L (about10�3). At this level it is possible to achieve a significant increasein the density of the medium (4–5 times higher than the densityof flue gases without FWD) and also it is possible to use electronbeam with mid-energy in irradiation reactors with a smalldiameter. Moreover, the radiation-induced reactions developedin liquid phase will play a significant role and will increase theoverall efficiency of the process.

Page 5

I. Calinescu et al. / Radiation Physics and Chemistry 85 (2013) 130–138134

2. Experimental

The experimental part took place in the irradiation labora-tories of the National Institute for Lasers, Plasma and RadiationPhysics Bucharest using an ALIN-10 EB accelerator.

The ALIN-10 accelerator is a laboratory installation designedfor fundamental research. It is located in a horizontal positioninside the irradiation room and is equipped with a post-acceleration beam focusing and bending to project accelerated

Fig. 7. The schematic drawing of ring-shaped EB collection monitor.

0 5 10 15 20 25 30 350

2

4

6

8

10

12

14

EB loading characteristicEB power characteristic (100 Hz)

26.25 µA

12.46 MeV

6.23MeV

EB a

vera

ge p

ower

, PE

B (W

)

EB a

vera

ge e

nerg

y, E

EB (M

eV)

EB average current, IEB (µA)

0

26

52

78

104

130

156

182

9.17 MeV

13 µA

PEB max

Fig. 8. EB average energy (EEB) and electron average beam power (PEB) versus EB

average current (IEB).

Fig. 9. The schematic drawing of EB reactor for treatm

EB either on the accelerating structure axis (through a horizontalvacuum window exit port) or at right angles to the acceleratorstructure (through a vertical vacuum window exit port) ALIN-10is current used to develop new technologies in the field ofenvironment pollution control, material processing, biomedicineand health-care products synthesis studies. The ALIN-10 accel-erator is of traveling-wave type, driven by 2 MW peak powertunable EEV M5125 type magnetrons (English Electric ValveCompany Ltd.) operating in S-band (2992–3001 MHz). In theALIN-10 output arrangement with horizontal and vertical exitports, the electron beam passes a ring-shaped EB collectionmonitor (RS-EBCM) shown in Fig. 7.

The RS-EBCM collection monitor that intercept only a fractionof exit accelerator EB give, together with its associated electronicsinstrumentation, a relative value of the EB average current(EBAC). The EBAC relative level has been calibrated with a Faradaycup placed at the EB exit end of the ALIN-10. Fig. 8 shows theALIN-10 beam loading characteristic and beam power character-istic. As seen in Fig. 8, the optimum values of the EB averagecurrent IEB and EB average energy EEB to produce maximumaverage output power PEB for a fixed pulse duration tEB andrepetition frequency fEB are as follows: EEB¼6.23 MeV;IEB¼26.25 mA; PEB¼164 W (fEB¼100 Hz, tEB¼3.5 ms).

The flue gases are obtained in a burner and had a compositionsimilar to gases from power plants that burn coal or oil with highsulfur content. The sulfur and nitrogen content from the diesel-fuel used was assured by adding carbon disulfide and aniline,respectively (Calinescu et al., 2012).

The laboratory installation for flue gases treatment with EB,generated by ALIN-10 accelerator, consists of the followingtechnological units: (i) the fuel burning unit; (ii) the flue gasconditioning system; (iii) the reactor for flue gas irradiation withEB; (iv) the liquid product separation unit; (v) the continuousgases monitoring system; and (vi) the treated gas elimination unitand is described in our previous paper (Calinescu et al., 2012).

In Fig. 9 is presented the EB reactor with devices for FWDformation. The EB reactor is designed like a Faraday cage in orderto permit the EB average current monitoring during gaseousmixture irradiation. It consists of two concentric stainless steelcylindrical vessels: a gas and electrically insulated inner vessel of0.2 m inner diameter and 2.967 m long and a grounded outervessel of 0.28 m inner diameter and 3.145 m long.

On the inner cylinder were fixed 15 nozzles of 0.2 mm bywhich is introduced water as fog curtains with an Aero Mist pumpof 0.5 gallons per minute (GPM) and 700 W, that have workingpressure of 70 bar. The pump contains a 5 mm particulate filter.Water is pushed through a pressure-resistant nylon hose in nozzles

ent of flue gases by EB in the presence of FWD.

Page 6

Table 1Treatment conditions and results obtained for flue gas treatment by EB, in the presence of ammonia and of FWD.

No. Treatment conditions Results Obs.

Flow rate (l/h) NH3 stoecha Initial conc. (ppmv) Treatment type RE SO2 RE NOx

Gases Liq. water SO2 NO

1 1000 0 0.8 2000 400 EB 0.98 0.62 Ighigeanu et al. (2012)

2 5200 0 0.9 812 44 EB 0.85 0.78

3 5200 50 0.9 812 44 EBþFWD 0.94 0.984 14,000 0 0.6 1116 102 EB 0.70 0.36

5 14,000 50 0.6 1116 102 EBþFWD 0.88 0.886 14,000 0 0.95 423 135 EB 0.97 0.32

7 14,000 50 0.95 423 135 EBþFWD 0.98 0.848 14,000 0 0.95 520 292 EB 0.97 0.20

9 14,000 50 0.95 520 292 EBþFWD 0.98 0.8

a NH3 stoech.¼[NH3]/(2� [SO2]þ[NOx]), with all concentrations in ppmv.

I. Calinescu et al. / Radiation Physics and Chemistry 85 (2013) 130–138 135

that are mechanically fitted with screw clamps and helically andequidistant distributed to 1201. The high pressure causes water to besprayed through nozzles as small droplets of the diameter close to10 mm.

The electrons are introduced in the EB reactor through two exitwindows, one of each in the form of cylinder, made of 100 mm-thick aluminum foils. The EB current is collected on the electricalinsulated inner vessel of the EB reactor, integrated and displayedon the control desk.

The characteristics of ALIN-10 EB accelerator used in ourexperiments (I—beam current, E—electrons energy and P—EBpower) highlight on Fig. 8, are

Iexit accelerator¼1371 mAEexit accelerator¼9.17�0.23/þ0.1 MeVireactor as Faraday cup (without FWD): 10–11 mAIreactor as Faraday cup (with FWD): 0.5–1 mAPexit accelerator¼13 mA�9.17 MeV¼119.21 WPreactor inside power (without FWD)¼10 (11) mA�9.17 MeV¼91.7(100.87) W (about 27.3–17.34 W are losses in the twowindows of the reactor, the two air gaps, and other uncon-trolled losses).

3. Results and discussion

In our experiments, fine water droplets, that are producedwith a high pressure pump, are injected simultaneously with ahigh energy (6–9 MeV) EB into a proper modified EB reactorreported in our previous paper (Ighigeanu et al., 2012) for thefollowing purposes:

�

To promote the use of medium to high-energy EB acceleratorsfor air pollutants removal in order to be in accordance withCleland’ recommendations: the ‘‘electron energy losses in thedual beam windows and the air space will be substantiallylower with higher electron energies’’ (Cleland, 2007). � To reduce the length of useful penetration of high energy

electrons that is in air much bigger than the length of the usualEB reactors (EBRs) that are presently used for air pollutantsremoval. Therefore, the presence of FWD simultaneously witha high energy electron beam makes possible the use of EBreactors with suitable size.

In view of above conditions, the injection of FWD in conjunctionwith high energy EB in the gaseous mixture is able to shorten EBranges and to improve high energy EB absorption and consumptionin useful chemical reactions. A laboratory accelerator of high

energy (6–9 MeV) was used to reveal the effect of fine waterdroplets. Research must be pursued using mid-energy accelerators(1–2 MeV).

As evaluation criterion of the process were calculated pollu-tant (P) removal efficiency (RE), reactor energy density (RIED), andenergy efficiency (EE) by using the following expressions(Zhu et al., 2009):

RE %ð Þ ¼½P�inlet�½P�outlet

½P�inlet

� 100% ð19Þ

RIED kJ=L� �

¼EB_input_power ðWÞ

gas_f low_rate L=min� �� 60� 10�3

ð20Þ

EE g=kWh� �

¼½P�inlet � RE

RIED� 10�3

ð21Þ

where [P]inlet is in mg/N m3.A number of experiments were conducted using flue gases

flow rates ranging between 5.2 and 14.0 m3/h. Two operationmodes were tested EB only or EB combined with FWD. Thetreatment conditions and the results are shown in Table 1. Forcomparison, the results obtained previously on the same installa-tion but without FWD, are presented.

It can be noticed that the presence of FWD causes an increaseof the pollutants removal efficiency. This observed effect is biggerin the conditions when the values of the removal efficiency in theabsence of FWD are lower, i.e. lower doses regions.

For different reaction conditions the density of reaction mix-ture will be changed and from this reason also penetration depthand the EB power effectively used. Table 2 presents the treatmentconditions, the effectively used EB power (determined accordingto the accelerator power, the losses in windows and the air gap,and function of the ratio between the reactor length and thepenetration depth) and also the results obtained: reactor energydensity and energy efficiency determined for effectively used EBpower and for total EB power.

From the values presented in Table 2 it can be noticed that thepresence of FWD leads to an increase in the density of themedium and thus, the penetration depth of EB in the reactionmedium is significantly changed. Because the reactor length islimited to 300 cm, we can consider that the ratio between thereactor length and the penetration depth provides informationabout the effectively used power. From this value of power usedwe can determine the real values of RIED and EE.

From the analysis of values listed in Table 2 we can notice thatin the presence of FWD the real EE values are lower than thoseobtained in the absence of water droplets. However, this is

Page 7

Table 2RIED and EE values for different treatment conditions.

No Treatment conditions Results for effectively used EB power Results for total EB power (96.3 W)

Liquid water

fraction, L

Density (g/L) Penetration

depth, (cm)

Absorbed EB

power (W)

RIED (kJ/L) EE (g/kWh) RIED (kJ/L) EE (g/kWh)

SO2 NOx SO2 NOx

1 0 1.06 4530 9 0.032 172.8 10.3 0.347 16.2 1.0

2 0 1.06 4530 9 0.006 316.5 7.4 0.067 29.6 0.7

3 0.0096 10.64 426 82.6 0.057 38.1 1.0 0.067 32.7 0.9

4 0 1.06 4530 9 0.002 964.4 21.3 0.025 90.1 2.0

5 0.0036 4.6 1260 25 0.006 436.5 18.7 0.025 113.3 4.9

6 0 1.06 4530 9 0.002 506.6 25.0 0.025 47.3 2.3

7 0.0036 4.6 1260 25 0.006 184.2 23.6 0.025 47.8 6.1

8 0 1.06 4530 9 0.002 622.7 33.8 0.025 58.2 3.2

9 0.0036 4.6 1260 25 0.006 226.5 48.7 0.025 58.8 12.6

Table 3Values for industrial EBFGT plant of Pomorzany (Chmielewski et al., 2004).

Results for total EB power Results for effectively used EB power (�21.8% losses)

SO2 NOx SO2 NOx

Flow rate for one EB accelerator, (N m3/min) 1,125,000 1,125,000 1,125,000 1,125,000

EB power (kW) 259,700 259,700 203,085 203,085

RE (%) 0.9 0.7 0.9 0.7

Pollutant inlet (mg/N m3) 1500 391 1500 391

RIED (kJ/L) 0.014 0.014 0.011 0.011

EE (g/kWh) 97.47 19.76 123.31 25.00

Table 4Values of beam power losses function of EB energy obtained at pilot plant operated at EPS Kaweczyn (for 0.5, 0.6 and 0.7 MeV) and calculated for different energies of

electron beam in the presence of FWD (for 1–2 MeV).

EB energy

(MeV)

Beam power losses

(windowsþair)

(kW or %)

Losses due to vessel diameter

(1.6 m)

Volume ratio

liquid:gas, L

Power used to obtain

FWD (P) (kW)

Proper diameter

to avoid losses, m

Energy economy

(kW or %)

Without FWD

(L1) (kW or %)

With FWD

(L2) (kW or %)

EnEc1 EnEc2

With out

FWD

With FWD (L1�(L2þP))

0.5 47.4 0.4

0.6 29.2 2.9

0.7 21.8 14.8

0.7 21.8 14.8 5.2 0.0001 3.9 2.1 1.8 5.70

1 14.3 46

39 0.0001 3.89 3.81 3.47 3.00 10.5

19 0.0005 19.45 3.81 2.54 8.04 15.54

4 0.001 38.90 3.81 1.91 3.08 10.58

1.5 8.4 74

56 0.0005 19.45 6.71 4.47 �1.83 11.57

39 0.001 38.90 6.71 3.35 �4.21 9.19

25 0.0015 58.35 6.71 2.68 �9.09 4.31

10 0.002 77.80 6.71 2.24 �13.71 �0.31

1 0.003 116.70 6.71 1.68 �43.56 �30.16

2 6.2 82

74 0.0005 19.45 9.56 6.37 �11.52 4.08

65 0.001 38.90 9.56 4.97 �21.83 �6.23

53 0.0015 58.35 9.56 3.98 �29.90 �14.3

39 0.002 77.80 9.56 3.32 �35.21 �19.61

19 0.003 116.70 9.56 2.49 �53.77 �38.17

I. Calinescu et al. / Radiation Physics and Chemistry 85 (2013) 130–138136

compensated by the low level of EB power used in the absenceof FWD.

From the analysis of the energy efficiency values function ofthe EB power provided by the accelerator the positive effect ofFWD presence in the reaction mixture can be noticed. This effectis more important for NOx. In order to compare our experimentaldata with those obtained in an industrial installation, we madesome calculations for the operational conditions of EBFGT plant inEPS Pomorzany, Poland (Chmielewski et al., 2004). The values arepresented in Table 3.

Our experiments 8 and 9 are carried out with pollutantsconcentrations similar with those existing in the Pomorzany powerplant. It can be noticed that the values obtained for the energyefficiency (EE) are higher when FWD are used as compared to thoseobtained in the industrial plant (for effectively used EB power).

It is also important to determine the power of the pump usedto obtain FWD. The output power of this pump can be determinedby the equation (Perry, 1999):

P¼H*Q=3:599*106ð22Þ

Page 8

I. Calinescu et al. / Radiation Physics and Chemistry 85 (2013) 130–138 137

where P is the output power of the pump (kW), H is the totaldynamic head (Pa) and Q is the capacity (m3/h).

Table 4 contains some data reported from the EBFGT pilotplant experiments at EPS Kaweczyn, Poland (20,000 N m3/h,100 kW beam power and initial electrons energy of 0.7 MeV)(Chmielewski et al., 1995a) and calculated values of EB losses fordifferent energies of electron beam in the presence of FWD (for1 to 2 MeV).

It can be noticed how the losses in windows and air gap aredecreasing while the electrons energy raises but the reactordiameter has to be very large in order to trap the full energy ofelectrons. The losses due to the reactor diameter are lower in thepresence of FWD. In determining the power necessary to obtainFWD it can be seen that there is a limitation in the amount ofliquid water used.

Thus using EB of 1 MeV in a reactor with 1.6 m diameter in theabsence of FWD the power losses are 46% which means 46 kW forthe pilot plant that has 2 EB accelerators of 50 kW. In the presenceof FWD the losses are reduced at 39%, 19% or 4% (function ofLiquid:gas ratio used) for obtaining FWD a power of 3.9–38.9 kWis used. The energy economy due to the use of FWD (EnEc1) canbe determined as the difference between the losses due to vesseldiameter in the absence of FWD and the sum of losses due tovessel diameter in the presence of FWD and the power used toobtain FWD.

The energy economy due to the use of mid-energy electronbeams and the use of FWD (EnEc2) also considers the reduction ofbeam power losses (windowsþair) as compared to the situationin which EBs of 0.7 MeV are used.

We should remember, however, that this application leads toapplication of so called semi-wet process and the byproduct is inthe form of solution. The mist can be eliminated from the outletflue gases through condensation ESP application (Porowski et al.,1995). To obtain solid state fertilizer (crystals) the water has to beevaporated. This can be done in the humidification tower usingthe heat of the inlet flue gases.

4. Conclusions

Electron beam accelerators most often applied for flue gastreatment are those with medium electrons energy (o1 MeV) inorder to obtain low penetration depths (�3 m). At this energylevel of electrons a significant level of electron beam power is lostin the metallic foil windows and in the air gap between them. Ifhigher electron beam accelerators can be used (minimum1.5 MeV), these losses would be significantly reduced. At thislevel of energy, however, the penetration depth increases sig-nificantly (approx. up to 7.5 m), which means that the treatmentof gases becomes due to the process vessel design difficulties.However if FWD are added to the reaction mixture, the density ofthis mixture increases largely, which determines the reduction ofthe penetration depth to reasonable values (2–3 m).

The overall efficiency of the process can be measured using theenergy efficiency (g pollutant/kWh of effectively used EB). The useof this parameter allows us to compare the classical experiments(in which only EB treatment was used) with the experiment inwhich EB treatment in the presence of FWD was used at volu-metric ratios liquid/gas between 0.0035 and 0.0096.

The presence of FWD determines the increase of the EB powerused, and the energy efficiency values obtained are higher thanthe ones obtained in an industrial plant in which only EBtreatment is used.

When the power necessary to pump water in order to obtainFWD is included in the total energy balance it can be noticed thatthere is a limit value of the volumetric ratio liquid:gas (L). For a

higher amount of water it can be noticed that the energy neededfor obtaining FWD is higher than the economy of energy obtainedby using higher energy electrons. The optimal value of L isfunction of the EB energy and of the maximum diameter of theirradiation reactor.

The proposed solution requires some changes in the processengineering since the mist has to be removed from the outletgases and then water evaporated to obtain solid product which isfertilizer component.

References

Calinescu, I., Martin, D., Ighigeanu, D., Bulearca, A., 2012. Flue gases treatment byirradiation with electron beam in the presence of fine water droplets. Rev.Chim. Bucharest. 63 (6), 576–579.

Chmielewski, A.G., 2011. Electron accelerators for environmental protection. In:Chao, A.W., Chou, W. (Eds.), Reviews of Accelerator Science and Technology,vol. 4. World Scientific Publishing Company, pp. 147–160.

Chmielewski, A.G., 2007. Industrial applications of electron beam flue gastreatment—from laboratory to the practice. Radiat. Phys. Chem. 76,1480–1484.

Chmielewski, A.G., Licki, J., Pawelec, A., Tyminski, B., Zimek, Z., 2004. Operationalexperience of the industrial plant for electron beam flue gas treatment. Radiat.Phys. Chem. 71, 439–442.

Chmielewski, A.G., Tyminski, B., Licki, J., Iller, E., Zimek, Z., Radzio, B., 1995a. Pilotplant for flue gas treatment—continuous operation tests. Radiat. Phys. Chem.46, 1067–1070.

Chmielewski, A.G., Zimek, Z., Panta, P., Drabnik, W., 1995b. The double window forelectron beam injection into the flue gas process vessel. Radiat. Phys. Chem.45 (6), 1029–1033.

Cleland, M.R., 2007. Technical aspects of flue gas irradiation. Presented at the IAEATechnical Meeting on Prospects and Challenges in Application of Radiation forTreating Exhaust Gases, Warsaw, Poland.

Cleland M.R., 2005. Industrial applications of electron accelerators. Presented atthe CERN Accelerator School Small Accelerator Course Zeegse, Netherlands.

Cleland, M.R., 2004. Comparisons of Monte Carlo and ICRU electron energy vs.range equations. Radiat. Phys. Chem. 71 (1–2), 585–589.

Cooper, W.J., Curry, R.D., O’Shea, K.E., 1998. Environmental Applications of IonizingRadiation, 1st ed. John Wiley &Sons, New York.

Drobny, J.G., 2010. Radiation Technology for Polymers, second ed. CRC Press.Gaffney, J.S., Streit, G.E., Spall, W.D., Hall, J.H., 1987. Beyond acid rain. Do soluble

oxidations and organic toxins interact with SO2 and NOX to increase ecosystemeffects? Environ. Sci. Technol. 21, 519–524.

Genuario, R.D., 2009. Wet discharge electron beam flue gas scrubbing treatment.US Patent 2009/0188782.

Hackman, R., Akiyama, H., 2000. Air pollution control by electrical discharges. IEEETrans. Dielectr. Electr. Insul. 7 (5), 654–682.

Hatano, Y., Katsumura, Y., Mozumder, A., 2011. Charged Particle and PhotonInteractions with Matter. CRC Press.

IAEA (International Atomic Energy Agency), 2010. Industrial Electron BeamProcessing. Working material—Revision 4.

Ighigeanu, D., Martin, D., Calinescu, I., Bulearca, A., Manaila, E., Craciun, C., 2012.Gaseous pollutants removal by electron beam based hybrid systems. Rev.Chim. Bucharest 63 (2), 183–188.

Kato, N., Akimoto, H., 2007. Anthropogenic emissions of SO2 and NOX in Asia:emission inventory. Atmos. Environ. 41, S171–S191.

Licki, J., Chmielewski, A.G., Iller, E., Zimek, Z., Mazurek, J., Sobolewski, L., 2003.Electron-beam flue-gas treatment for multicomponent air-pollution control.Appl. Energy 75, 145–154.

Matzing, H., Namba, H., Tokunaga, O., 1993. Kinetics of SO2 removal from flue gasby electron beam technique. Radiat. Phys. Chem. 42 (4–6), 673–677.

Matzing, H., Paur, H.R., 1992. Chemical mechanisms and process parameters offlue gas cleaning by electron beam. In: Nriagu, J.O. (Ed.), Gaseous Pollutants:Characterization and Cycling. Wiley, New York, pp. 307–331.

Matzing, H., 1989. Chemical Kinetics of Flue Gas Cleaning by Electron Beam.Keruforschunfzektrum Karslruhe. KFK494.

Namba, H., Hashimoto, S., Tokunaga, O., Suzuki, R., 1998. Electron beam treatmentof lignite-burning flue gas with high concentrations of sulfur dioxide andwater. Radiat. Phys. 53, 673–681Chem 53, 673–681.

Perry, H.R., 1999. Perry’s Chemical Engineeres’ Handbook, 7th edition McGraw-Hill.Person, J.C., Ham, D.O., 1988. Removal of SO2 and NOx from stack gases by electron

beam irradiation. Radiat. Phys. Chem. 31 (1–3), 1–8.Porowski L., Chmielewski A.G., Iller E., Luty, L., 1995. Method for removing acid gas

impurities from industrial gases stream and equipment for removal of acid gaspollutants from industrial gases waste stream. Polish Patent 900427.

Potapkin, B.V., Deminski, A., Fridman, A.A., Rusanov, D., 1995. The effect of clustersand heterogeneous reactions on non-equilibrium plasma flue gas cleaning.Radiat. Phys. Chem. 45 (6), 1081–1088.

Ramanathan, V., Feng, Y., 2009. Air pollution, greenhouse gases and climatechange: global and regional perspectives. Atmos. Environ. 43, 37–50.

Page 9

I. Calinescu et al. / Radiation Physics and Chemistry 85 (2013) 130–138138

Squadrito, G., Postlethwait, E., 2009. On the hydrophobicity of nitrogen dioxide:could there be a ‘‘lens’’ effect for NO2 reaction kinetics? Nitric Oxide 21 (2),104–109.

Streets, D.G., Waldhoff, S.T., 2000. Present and future emissions of air pollutions inChina: SO2, NOX, and CO. Atmos. Environ. 34, 363–374.

Tokunaga, O., Suzuki, N., 1988. Radiation chemical reactions in NOX and SO2

removals from flue gas. Radiat. Phys. Chem. 24 (1), 145–165.

Yermakov, A.N., Zhitomirsky, B.M., Sozurakov, D.M., Poskrebyshev, G.A., 1995.Water aerosols spraying for SO2 and NOX removal from gases under e-beamirradiation. Radiat. Phys. Chem. 45 (6), 1071–1076.

Zhu, T., Li, J., Liang, W., Jin, Y., 2009. Synergistic effect of catalyst for oxidationremoval of toluene. J. Hazard. Mater. 165 (1–3), 160–1258.

Zimek, Z., 1995. High power electron accelerators for flue gas treatment. Radiat.Phys. Chem. 45 (6), 1013–1015.