7
Occupational exposure in the fluorescent lamp recycling sector in France François Zimmermann , Marie-Thérèse Lecler, Frédéric Clerc, Alain Chollot, Eric Silvente, Jérome Grosjean Department of Process Engineering, Institut National de Recherche et de Sécurité, 54500 Vandoeuvre-les-Nancy, France article info Article history: Received 20 September 2013 Accepted 28 March 2014 Available online xxxx Keywords: WEEE Lamp recycling Occupational exposure Ambient concentration abstract The fluorescent lamp recycling sector is growing considerably in Europe due to increasingly strict regulations aimed at inciting the consumption of low energy light bulbs and their end-of-life manage- ment. Chemical risks were assessed in fluorescent lamp recycling facilities by field measurement surveys in France, highlighting that occupational exposure and pollutant levels in the working environment were correlated with the main recycling steps and processes. The mean levels of worker exposure are 4.4 mg/m 3 , 15.4 lg/m 3 , 14.0 lg/m 3 , 247.6 lg/m 3 , respectively, for total inhalable dust, mercury, lead and yttrium. The mean levels of airborne pollutants are 3.1 mg/m 3 , 9.0 lg/m 3 , 9.0 lg/m 3 , 219.2 lg/m 3 , respectively, for total inhalable dust, mercury, lead and yttrium. The ranges are very wide. Surface samples from employees’ skin and granulometric analysis were also carried out. The overview shows that all the stages and processes involved in lamp recycling are concerned by the risk of hazardous substances penetrating into the bodies of employees, although exposure of the latter varies depending on the processes and tasks they perform. The conclusion of this study strongly recommends the development of a new generation of processes in parallel with more information sharing and regulatory measures. Ó 2014 Elsevier Ltd. All rights reserved. 1. Introduction Whilst the availability of clear incandescent and conventional halogen lamps has fallen since the EU Ecodesign Directive (2009) prohibited the sale of certain traditional style energy intensive light bulbs, the growth in sales of discharge lamps is likely to continue in the years to come. Fluorescent lamps come under the category of discharge lamps in which the light is generated by an electric discharge in a gas or vapor. In this illumination process, a droplet of liquid mercury is introduced in the fluorescent lamp and a phosphor-coated glass tube is used to convert the ultra- violet light into visible light output (Aman et al., 2013). The phosphor coating is formed by special luminescent powder containing rare earth elements and metals. In 2011, 30 million (4040 tons) spent lamps were collected in France, representing 35% of overall end-of-life recyclable lamps (Recylum, 2012). Compliance with the 2003 European WEEE direc- tive (Directive 2002/96/EC) and the accreditation of Recylum in 2006, the French eco-organization devoted to the waste manage- ment of spent lamps, has led to the increased collection of spent lamps and better organization of the lamp recycling sector, reinforcing the activity of existing facilities and leading to the cre- ation of facilities capable of handling this waste efficiently. The minimum targets of recycling efficiency of spent lamps fall within the scope of the new WEEE directive (Directive 2012/19/EU), i.e. 75% shall be recovered and 55% shall be prepared for re-use and recycled. Although the RoHs directive (Directive 2002/95/EC) prohibited the use of certain hazardous substances in electrical and electronic equipment, the use of mercury is still permitted in fluorescent lamps because low-pressure mercury vapor is essential for them to work properly (European Commission, 2009). The amount of mercury in compact fluorescent lamp should be lower than 5 mg per lamp, while Santos et al. (2010) showed that 40% of the lamps analyzed contained more than this limit allowed by European Community. By crossing information on mercury distribution (Rey-Raap and Gallardo, 2012) and mercury speciation (Raposo et al., 2003) inside spent fluorescent lamps it is possible to high- light the fact that treating this e-waste cleanly is a difficult prob- lem. Worldwide, various fluorescent lamp recycling scenarios (Rasapo and Roeser, 2001; Chang et al., 2007; Apisitpuvakul et al., 2008; Asari et al., 2008) have shown that a significant part of mercury is released into the environment by various media including emission into the air. Initiatives have emerged to remove mercury (Chang et al., 2009; Rey-Raap and Gallardo, 2013) and to http://dx.doi.org/10.1016/j.wasman.2014.03.023 0956-053X/Ó 2014 Elsevier Ltd. All rights reserved. Corresponding author. Tel.: +33 383509837. E-mail address: [email protected] (F. Zimmermann). Waste Management xxx (2014) xxx–xxx Contents lists available at ScienceDirect Waste Management journal homepage: www.elsevier.com/locate/wasman Please cite this article in press as: Zimmermann, F., et al. Occupational exposure in the fluorescent lamp recycling sector in France. Waste Management (2014), http://dx.doi.org/10.1016/j.wasman.2014.03.023

Occupational exposure in the fluorescent lamp recycling sector in France

  • Upload
    jerome

  • View
    215

  • Download
    0

Embed Size (px)

Citation preview

Waste Management xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

Waste Management

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

Occupational exposure in the fluorescent lamp recycling sector in France

http://dx.doi.org/10.1016/j.wasman.2014.03.0230956-053X/� 2014 Elsevier Ltd. All rights reserved.

⇑ Corresponding author. Tel.: +33 383509837.E-mail address: [email protected] (F. Zimmermann).

Please cite this article in press as: Zimmermann, F., et al. Occupational exposure in the fluorescent lamp recycling sector in France. Waste Mana(2014), http://dx.doi.org/10.1016/j.wasman.2014.03.023

François Zimmermann ⇑, Marie-Thérèse Lecler, Frédéric Clerc, Alain Chollot, Eric Silvente,Jérome GrosjeanDepartment of Process Engineering, Institut National de Recherche et de Sécurité, 54500 Vandoeuvre-les-Nancy, France

a r t i c l e i n f o a b s t r a c t

Article history:Received 20 September 2013Accepted 28 March 2014Available online xxxx

Keywords:WEEELamp recyclingOccupational exposureAmbient concentration

The fluorescent lamp recycling sector is growing considerably in Europe due to increasingly strictregulations aimed at inciting the consumption of low energy light bulbs and their end-of-life manage-ment. Chemical risks were assessed in fluorescent lamp recycling facilities by field measurement surveysin France, highlighting that occupational exposure and pollutant levels in the working environment werecorrelated with the main recycling steps and processes.

The mean levels of worker exposure are 4.4 mg/m3, 15.4 lg/m3, 14.0 lg/m3, 247.6 lg/m3, respectively,for total inhalable dust, mercury, lead and yttrium. The mean levels of airborne pollutants are 3.1 mg/m3,9.0 lg/m3, 9.0 lg/m3, 219.2 lg/m3, respectively, for total inhalable dust, mercury, lead and yttrium. Theranges are very wide. Surface samples from employees’ skin and granulometric analysis were also carriedout. The overview shows that all the stages and processes involved in lamp recycling are concerned bythe risk of hazardous substances penetrating into the bodies of employees, although exposure of thelatter varies depending on the processes and tasks they perform. The conclusion of this study stronglyrecommends the development of a new generation of processes in parallel with more informationsharing and regulatory measures.

� 2014 Elsevier Ltd. All rights reserved.

1. Introduction

Whilst the availability of clear incandescent and conventionalhalogen lamps has fallen since the EU Ecodesign Directive (2009)prohibited the sale of certain traditional style energy intensivelight bulbs, the growth in sales of discharge lamps is likely tocontinue in the years to come. Fluorescent lamps come under thecategory of discharge lamps in which the light is generated by anelectric discharge in a gas or vapor. In this illumination process, adroplet of liquid mercury is introduced in the fluorescent lampand a phosphor-coated glass tube is used to convert the ultra-violet light into visible light output (Aman et al., 2013). Thephosphor coating is formed by special luminescent powdercontaining rare earth elements and metals.

In 2011, 30 million (4040 tons) spent lamps were collected inFrance, representing 35% of overall end-of-life recyclable lamps(Recylum, 2012). Compliance with the 2003 European WEEE direc-tive (Directive 2002/96/EC) and the accreditation of Recylum in2006, the French eco-organization devoted to the waste manage-ment of spent lamps, has led to the increased collection of spentlamps and better organization of the lamp recycling sector,

reinforcing the activity of existing facilities and leading to the cre-ation of facilities capable of handling this waste efficiently. Theminimum targets of recycling efficiency of spent lamps fall withinthe scope of the new WEEE directive (Directive 2012/19/EU), i.e.75% shall be recovered and 55% shall be prepared for re-use andrecycled.

Although the RoHs directive (Directive 2002/95/EC) prohibitedthe use of certain hazardous substances in electrical and electronicequipment, the use of mercury is still permitted in fluorescentlamps because low-pressure mercury vapor is essential for themto work properly (European Commission, 2009). The amount ofmercury in compact fluorescent lamp should be lower than 5 mgper lamp, while Santos et al. (2010) showed that 40% of the lampsanalyzed contained more than this limit allowed by EuropeanCommunity. By crossing information on mercury distribution(Rey-Raap and Gallardo, 2012) and mercury speciation (Raposoet al., 2003) inside spent fluorescent lamps it is possible to high-light the fact that treating this e-waste cleanly is a difficult prob-lem. Worldwide, various fluorescent lamp recycling scenarios(Rasapo and Roeser, 2001; Chang et al., 2007; Apisitpuvakulet al., 2008; Asari et al., 2008) have shown that a significant partof mercury is released into the environment by various mediaincluding emission into the air. Initiatives have emerged to removemercury (Chang et al., 2009; Rey-Raap and Gallardo, 2013) and to

gement

2 F. Zimmermann et al. / Waste Management xxx (2014) xxx–xxx

recover yttrium from spent fluorescent lamps (De Michelis et al.,2011). In fact, dust and powdery materials (e.g. yttrium-containingluminescent powder) can also be spread into the occupationalatmosphere throughout the fluorescent lamp recycling process.

It is therefore assumed that handling and processing lamps cangenerate occupational health issues, depending on the processes,procedures and operations used, and on the degree they causeemissions. Despite growing awareness of health risks for workersin the e-waste recycling industry as a whole (Searl and Crawford,2012; Lundgren, 2012) and the knowledge gained on the risks ofmercury escaping from broken fluorescent lamps (Nance et al.,2012; Hu and Cheng, 2012; Salthammer et al., 2012), there is a lackof extensive exposure measurement data in the spent lamprecycling sector, especially in developed countries where manage-ment policies are being implemented.

The aim of this article is to review an exposure assessment inthe five French spent fluorescent lamp recycling facilities thathighlights occupational exposure and pollutant levels in the work-ing environment and correlates them with the main recycling stepsand processes. The locations of pollutant sources assumed in thediscussion of this article were complemented using real-timemeasurements.

2. Material and methods

2.1. Description of treatment facilities

Two main types of treatment processes are involved in devel-oped facilities for recycling fluorescent lamps:

� The ‘‘end cut’’ processes. A guide-chain leads the lamps into anenclosed, depressurised chamber. The metallic end-caps areremoved by cutting and the luminescent powders depositedon the internal face of the glass are flushed out with jets ofcompressed air. The cleaned glass is milled. All the outputsare stored in separate containers. This type of process requirespreliminary sorting of straight fluorescent lamps by lengthand diameter.� The crushing processes. These processes are used either for

fluorescent or compact fluorescent lamps. Batches of lamps arefed into a shredder or crusher where the degree of separationof the outputs varies as a function of the process. These typesof processes generally function without any preliminary sorting.

Table 1 presents an overview of the facilities and processesreviewed in the lamp recycling sector.

2.2. Sampling, analysis and measurement methods

Assessing levels of exposure and pollution with dusts and met-als was performed by taking personal and airborne samples,respectively. Particulate aerosols corresponding to the inhalablefraction (aerodynamic diameter <100 lm) were sampled at aflow-rate of 2 L/min in 37 mm closed cassettes on PVC capsulesfused to a cellulose acetate filter (Accu-CapTM).

2.2.1. Metal-containing inhalable dust and mercury vaporsInhalable dust concentrations were determined by gravimetry

according to the INRS Metropol 002 method (INRS Metropol,2009). Chemical analyses of metals were performed by plasmaemission spectrometry using ICP Varian 720-ES. The latter requireddissolving and mineralizing the capsules according to the INRSMetropol 003 method (INRS Metropol, 2008).

Mercury was sampled using Hydrar SKC Anasorb C300/500 mgat a flow-rate of 2 L min�1. Plasma emission spectrometry analysis

Please cite this article in press as: Zimmermann, F., et al. Occupational exposu(2014), http://dx.doi.org/10.1016/j.wasman.2014.03.023

was performed by displacing the cold vapor after extractionaccording to the INRS Metropol 024 method (INRS Metropol,2000).

The range of elements analyzed in the samples includes the fol-lowing: Hgvapour, Al, Ba, Be, Cd, Ce, Cr, Er, Eu, Fe, Gd, La, Mn, Ni, Pb,Pr, Sb, Sn, Si, Tb, Ti, Y, Yb, Zn.

In this study, the level of personal dust exposure and airbornedust corresponding to total inhalable dust and the level of personalexposure and airborne dust of key agents (e.g. Hgvapor; and Pb, Y, Bain the inhalable dust) were used to assess the atmosphericexposure of workers and evaluate process efficiency. Yttrium canbe considered as a tracer of contamination caused by the releaseof fluorescent powders. Lead mainly stems from glass dustalthough it has been identified in fluorescent powders, too. Thecurrent French occupational exposure limit values (OEL) foryttrium, lead, barium and mercury are 1000 lg/m3, 100 lg/m3,500 lg/m3 and 20 lg/m3, respectively.

2.2.2. Surface contaminationSurface samples from operators’ skins and work surfaces were

also taken using GhostWipes. Surface samples from workstationswere taken over a 10 cm � 10 cm area. The skin of operators’ handsand necks were mainly swabbed.

The amount of metal-containing dust deposited on operators’skin (hands and neck – in lg) was used as an indicator of surfacecontamination. There are currently no occupational regulatory limitvalues and no consensus values for surface pollution at national,european and international levels. In France, a reference value of10 lg/dm2 of lead is used as the limit set by the public healthauthorities in houses after construction (decree of May 12, 2009).

2.2.3. Granulometric analysisA granulometric analysis was carried out to complement the

quantitative atmospheric exposure data with particle size informa-tion. The particle size distribution was measured in liquid suspen-sion using the Malvern Mastersizer X technique. This optical lasertechnique measures particle distribution by scattered lightanalysis. The distribution of particle volume is measured as a func-tion of optical diameter. Imaging using a Jeol 7400-F field-emissiongun scanning electron microscope was also performed.

2.2.4. Real-time measurementIn addition, real-time measurements were also taken by direct

reading of atmospheric pollutant concentrations (dusts, mercuryvapors) and dust granulometry using portable optical analyzers(MIE Personal DataRam photometer, LightHouse Handheld 3016-IAQ) and mercury detectors (Mercury Tracker 3000). Real-timemeasurements were performed during specific operations andperiodically in the treatment facilities to support the hypothesisof emission sources.

3. Results

The granulometry of dust and powder emitted during the differ-ent steps of the treatment process showed that a significant pro-portion of the dust is in the alveolar fraction (Table 2), includinga sizeable proportion of nanoparticles (Fig. 1).

Workers in lamp recycling facilities are commonly involved inversatile tasks covering all the processing stages: storage manage-ment, lamp sorting, process feeding, output management andother various activities (e.g. cleaning and maintenance operations).They are, however, devoted to only one type of treatment process.Their main work area is around this process, feeding it with spentlamps and managing the outputs when required. Table 3 gives theoccupational atmospheric exposures of workers versus the two

re in the fluorescent lamp recycling sector in France. Waste Management

Table 1Overview of the facilities and processes involved in the lamp recycling sector.

Facilities Pre-treatments Treatment processes Output storagea (glass, luminescentpowder, metal end parts)

Annualtreatmentactivity

Lampstoragea

Lampsorting

End cut Crushing

A Yes Yes - MRT end cut machine 5000 (ECM5000)

Compact fluorescent lampshredders

Yes 1600 tons

- Mobile on-site recycling systemfrom Herborn GmbH)

B No No – Lamp shredder from MRTsystem (hammer mill)

No 300 tons

C Yes Yes MRT end cut machine 5000 (ECM5000)

– No 560 tons

D No No – Lamp shredder No 1000 tons

E Yes No – Lamp crusher (blade mixer) Yes 300 tons

a Defined areas in facilities that can be assessed.

Table 2Granulometric analysis of dusts and luminescent powders.

Luminescent powder(sample 1)

Luminescent powder(sample 2)

Luminescent powder(sample 3)

Dust (sample 1: endcut)

Dust (sample 2:crushing)

Median diameter 8.2 lm 6.4 lm 7.9 lm 6.8 lm 6.5 lmParticles <5 lm (alveolar

fraction)30% 40% 40% 30% 40%

Fig. 1. SEM Images of luminescent powders from fluorescent lamps.

F. Zimmermann et al. / Waste Management xxx (2014) xxx–xxx 3

types of treatment process. Fig. 2 shows the repartition of all theexposure measurements of workers versus the France OELs (Expo-sure Index = pollutant individual concentration/pollutant OEL).

During an 8-h workday, the exposure levels of workers reflectthe combination of their various activities and moving in and outof process areas. The airborne pollutant levels in specific areas ofthe various processing stages, including an evaluation of processemissions, are given in Table 4.

The results show that operator exposure to the pollutants mea-sured can reach very high levels compared with OEL values, withpersonal dust exposures varying between 0.7 and 13.2 mg/m3.The maximal exposures to key agents were 86.7 lg/m3 for mer-cury, 85.6 lg/m3 for lead and 1010.2 lg/m3 for yttrium.

The analysis of surface samples confirmed that the skin of oper-ators becomes polluted as it revealed the presence of metal pollu-tants (Table 5).

4. Discussion

The purpose of this study was to determinate the occupationalexposure levels and the emission levels of the different operationsinvolved in lamp recycling processing.

Please cite this article in press as: Zimmermann, F., et al. Occupational exposu(2014), http://dx.doi.org/10.1016/j.wasman.2014.03.023

Table 3 shows that all the processing steps and areas in whichoperators circulate are affected by the emission of pollutants,although degrees of pollutant levels vary widely inside andbetween the different stages investigated.

4.1. Lamp storage

Lamp storage is an area where the activity is usually limited toloading and unloading operations. Background pollution is usuallylow, but pollution peaks can occur (up to 1740 lg/m3 for yttrium)if (1) some lamps already broken in the containers emit pollutants,especially when they are shaken during transfers, (2) further acci-dental breakages can occur during handling, and (3) the storagearea is not usually confined from the other activities.

4.2. Lamp sorting

Lamps must be sorted before being introduced into the end cutair push machine. Lamp sorting can be a hazardous operation.Short time exposure may exceed OEL (up to 2430 lg/m3 foryttrium). This is due both to the breakage of tubes during sorting,and to broken tubes already present in the cases before sorting. The

re in the fluorescent lamp recycling sector in France. Waste Management

Fig. 2. Individual atmospheric measurements to key metallic elements divided byFrance OELs (Exposure Index).

Tabl

e3

Indi

vidu

alat

mos

pher

icex

posu

res

todu

stan

dm

etal

licel

emen

tsve

rsus

type

ofpr

oces

s.

Tota

lin

hal

able

dust

(mg/

m3)

Hg

(lg/

m3)

Ba

(lg/

m3)

Pb(l

g/m

3)

Y(l

g/m

3)

OEL

sFr

ance

:20

USA

(Nio

sh):

50l

g/m

3O

ELs

Fran

ce&

USA

(Nio

sh):

500

lg/

m3

OEL

sFr

ance

:10

0U

SA(N

iosh

):50

lg/

m3

OEL

sFr

ance

&U

SA(N

iosh

):10

00l

g/m

3

NM

ed.

AM

GM

Ran

geN

Med

.A

MG

MR

ange

NM

ed.

AM

GM

Ran

geN

Med

.A

MG

MR

ange

NM

ed.

AM

GM

Ran

ge

Indi

vidu

alm

easu

rem

ents

End

cut

82.

83.

83.

21.

1–7.

530

7.7

10.3

7.1

0.1–

50.6

89.

713

.810

3–45

.58

3.1

3.3

1.8

0.1–

7.9

817

3.8

208.

416

7.0

35.3

–471

.2

Cru

shin

g32

4.3

4.6

3.7

0.7–

13.2

4715

.618

.612

.41.

9–86

.732

15.0

19.3

12.5

0.2–

109.

932

8.4

16.7

6.9

0.1–

85.6

3212

2.8

257.

477

.60.

1–10

10.2

Lam

pre

cycl

ing

chan

nel

(all

)40

4.2

4.4

3.6

0.7–

13.2

7711

.815

.410

.00.

1–86

.740

13.8

18.2

42.4

0.2–

109.

940

5.0

14.0

5.3

0.1–

85.6

4015

0.4

247.

690

.50.

1–10

10.2

N:

nu

mbe

rof

sam

ples

;M

ed.:

med

ian

;A

M:

arit

hm

etic

mea

n;

GM

:ge

omet

ric

mea

n.

4 F. Zimmermann et al. / Waste Management xxx (2014) xxx–xxx

Please cite this article in press as: Zimmermann, F., et al. Occupational exposu(2014), http://dx.doi.org/10.1016/j.wasman.2014.03.023

pollutant level at this step depends to a great extent on (1) thethoroughness of sorting and avoiding breakage, (2) how breakageis managed, and (3) the storage conditions of the broken lamps.

4.3. End cut processes

The end cut processes are enclosed in a low pressure confine-ment to avoid any contamination to its surrounding area. Highconcentrations of toxics in these atmospheres have been mea-sured, especially for inhalable dust (149 mg/m3) mercury vapor(82.4 lg/m3) and lead (486.5 lg/m3). Using an end-cutting tech-nique, Rhee et al. (2013) showed that the mass mercury releasedto vapor phase is decreased as air flow rate is increased. Thus, itis important to work at high airflow rate.

In addition, during cleaning or maintenance episodes, operatorscan be exposed to high concentrations of pollutants, that requirethe use of specific respirators and protective equipment being suit-able for both mercury vapors and harmful dusts.

4.4. Crushing processes

A different technology is used to crush fluorescent lamps. Pollu-tion and operator exposure are significant around the treatmentprocess, whatever the latter. Operators devoted to crushing pro-cesses are on average more exposed to inhalable dust and metallicpollutants than those working on the end cut machines.

Crushing lamps can cause greater emissions depending on howthe crusher is confined, and because no decontamination opera-tions are performed during crushing. This is in contrast with theend cut procedures. In all cases, the handling of outgoing fractionsand the cleaning and maintenance of the processing areas andequipment are linked to particularly significant exposure. A similarfinding has been established by Lucas and Emery (2006) whorecommended to consider the entire process, not just the crushing

re in the fluorescent lamp recycling sector in France. Waste Management

Tabl

e4

Air

born

edu

stan

dke

yel

emen

tle

vels

vers

uspr

oces

sing

stag

esan

dty

peof

trea

tmen

tpr

oces

s.

Tota

lin

hal

able

dust

(mg/

m3)

Hg

(lg/

m3)

Ba

(lg/

m3)

Pb(l

g/m

3)

Y(l

g/m

3)

NM

ed.

AM

GM

Ran

geN

Med

.A

MG

MR

ange

NM

ed.

AM

GM

Ran

geN

Med

.A

MG

MR

ange

NM

ed.

AM

GM

Ran

ge

Are

am

easu

rem

ents

Lam

pst

orag

e19

0.7

0.8

0.6

0.2–

1.6

172.

22.

52.

31.

3–4.

314

1.3

6.3

2.0

0.4–

61.4

196.

712

.36.

10.

2–10

1.3

1920

.125

3.0

6.5

0.1–

1740

Lam

pso

rtin

g26

0.6

0.7

0.6

0–3.

419

5.5

6.3

5.6

2.5–

17.7

212.

311

.42.

90.

6–18

0.1

261.

45.

11.

50–

87.1

2649

.220

9.3

69.1

14.9

–243

0

End

cut

proc

esse

sIn

put

160.

40.

50.

50.

1–1.

216

11.1

10.1

7.9

0.8–

19.8

142.

64.

23.

21.

2–12

160.

91.

40.

50.

1–5.

316

47.3

134.

469

.820

.8–9

43In

side

552

.063

.335

.23.

3–14

94

33.0

37.9

20.5

3.2–

82.4

537

5.0

338.

723

3.9

29.1

–581

.45

6.2

100.

128

.80–

486.

55

237.

920

9.9

117.

56.

8–38

4.5

Ou

tpu

t15

0.4

0.5

0.5

0.1–

1.1

148.

99.

67.

30.

8–19

143.

74.

83.

71.

5–14

.715

0.7

2.1

0.7

0.1–

7.9

1558

.919

4.6

77.2

20.8

–177

0

Cru

shin

gpr

oces

ses

Inpu

t16

1.1

1.5

1.1

0.3–

416

5.8

8.9

6.6

2.3–

29.4

165.

79.

45.

20.

6–42

.316

5.9

6.5

2.7

0.1–

22.5

1611

1.3

243.

461

0.1–

739

Insi

de4

2.5

3.4

3.1

2.2–

6.5

420

.226

.921

.711

.4–5

64

27.2

25.5

23.1

11.2

–36.

44

3.7

5.6

4.7

3–12

.14

885.

510

56.7

719.

527

6–21

80O

utp

ut

260.

50.

70.

50.

1–3.

426

3.1

8.7

4.0

0.3–

6226

2.8

3.5

2.6

0.3–

13.1

313.

65.

03.

60.

6–18

.931

93.3

192.

625

.60.

1–96

0

Ou

tpu

tst

orag

e12

0.6

0.7

0.6

0.2–

1.2

123.

45.

64.

01.

4–19

.67

1.5

1.9

1.5

0.6–

3.9

75.

87.

16.

43.

5–13

.17

0.1

1.7

0.2

0.1–

6.7

Lam

pre

cycl

ing

chan

nel

(all

)13

90.

63.

10.

70–

149

128

5.0

9.0

5.4

0.3–

82.4

121

2.9

20.7

3.8

0.3–

581.

413

93.

29.

02.

30–

486.

513

952

.821

9.2

32.9

0.1–

2430

N:

nu

mbe

rof

sam

ples

;m

ed.:

med

ian

;A

M:

arit

hm

etic

mea

n;

GM

:ge

omet

ric

mea

n.

F. Zimmermann et al. / Waste Management xxx (2014) xxx–xxx 5

Please cite this article in press as: Zimmermann, F., et al. Occupational e(2014), http://dx.doi.org/10.1016/j.wasman.2014.03.023

xposu

phase, and to perform lamp processing and crushing in well-venti-lated or outdoor work locations. We shall not support this very lastoption as it is not environmentally friendly, and moreover ventila-tion systems are more and more efficient.

4.5. Lamp input and output zone

Operators at the lamp feed (input) and output areas of treat-ment processes are particularly exposed and these areas are pol-luted. This pollution is mainly due to the breakage of largequantities of lamps during handling and feeding, while emissionsfrom outputs continue. Indeed, mercury vapors are emitted as soonas lamps are opened, broken or crushed. The portion of mercuryreleased almost instantly to vapor phase can vary from 2% to 14%with the lamp manufacturers (Rhee et al., 2013) and their life time.Vapors then diffuse continuously reaching a peak concentrationduring the first minutes following equalization of the pressureinside and outside the lamp. Vapors may continue to be releasedlong after the lamp is opened, broken or treated. The kinetics ofmercury emission is also affected by the distribution of its chemi-cal forms and its physical states. Mercury present in a lamp at theend of its life is distributed as follows: either as a vapor releasedduring the first instants after opening, or as adsorbed forms onmetallic end-pieces, glass, powders (Jang et al., 2005; Aucottet al., 2003). These forms are released as vapors over the hoursand days following lamp rupture. The outputs containing lumines-cent powders are the most polluting since the portion of mercuryadsorbed in powders is higher than 80% of the total mercury inlamp (Rhee et al., 2013; Rey-Raap and Gallardo, 2012).

Finally, the output storage area is not spared by pollution forthe reasons already highlighted: mercury vapors diffuse continu-ously, pollutants are again emitted when the outputs are shakenor moved, and the open (no-confined) storage zones imply back-ground pollution.

4.6. Additional routes of contamination

Regarding the analysis of surface samples (Table 5), skin can bea major route of penetration into the human body. The contamina-tion can be either directly or indirectly by oral absorption, as it hasbeen observed that workers do not systematically wash theirhands and faces before breaks.

Therefore the entry of identified pollutants into the body isdiverse and includes inhalation, and cutaneous and oral exposures.Respirable and ultra-fine particles increase the risk of penetrationof hazardous substances into the body, therefore raising the levelof concern.

Mercury pollution is all the more worrying as the workplace canappear to be clean with no visible signs of contamination. A casestudy in fluorescent lamp manufacture (Guthrie et al., 2006)showed that mercury was being spread around the workroom ontosurfaces away from the working area. It means that extended areafar from the core treatment process can be concerned aboutcontamination.

5. Conclusion

The findings described in this work were provided by an over-view of the exposure of workers and their working environmentto pollutants in the fluorescent lamp recycling sector in France.Measurement campaigns were carried out in five treatment facili-ties, representing an exhaustive panorama of the French situationin this challenging industry.

We assessed the exposure of workers to the emissions of vari-ous processing steps and processes. The results clearly show that

re in the fluorescent lamp recycling sector in France. Waste Management

Table 5Metal-containing dust contamination of surface samples taken from operators’ necks and hands.

Ba (lg) Pb (lg) Y (lg)

Med. AM GM Range Med. AM GM Range Med. AM GM Range

Individual measurementsEnd cut (N = 28) 48.8 75.5 27.3 0.2–460.3 18.8 22.9 16.6 3–83.2 484.1 677.5 426.0 88.4–3113.4

Crushing (N = 19) 50.0 38.6 26.0 2.5–69.2 33.3 45.6 29.3 4.5–128.5 285.7 370.4 283.3 121.1–893.7

Lamp recycling channel (all; N = 47) 50.0 60.6 26.7 0.2–460.3 22.7 32.1 20.9 3–128.5 393.7 553.4 361.2 88.4–3113.4

N: number of samples; med.: median; AM: arithmetic mean; GM: geometric mean.

6 F. Zimmermann et al. / Waste Management xxx (2014) xxx–xxx

all the stages and processes involved are affected by worrying lev-els of pollutants, especially mercury vapors and dust containinglead and yttrium. The exposure of workers differs as a functionof the processes and tasks in which they are involved.

The organization of the sector and widely used technologiesimplemented for lamp recycling in France are definitely amongthe most operational and efficient as they rely on the best availabletechnologies. However, it can be assumed that the internationalsituation regarding lamp recycling is doubtful (due to the consider-able difficulties involved in its management). This leads us to con-clude it is very necessary to encourage the development of a newgeneration of processes, with more information sharing and regu-latory measures. In France, a regulatory text (Journal Officiel de laRépublique Française, 2009) described by Ogden and Lavoue(2011) obliges all companies to respect a strict sampling protocoland to apply a statistical evaluation of measurement values forestablishing a diagnostic per similar exposure group. The sub-stances concerned by this regulation include Hg and Pb, which cer-tainly are the most harmful for the employees’ health. The DEEErecycling facilities such as the ones presented in this study are con-cerned and must be compliant with this regulation protecting theworkers.

Meanwhile, exposures of workers and pollutants levels in thework environment should be mitigated by following these basicrecommendations:

1. Inputs and outputs should be stored and handling in ventilatedareas. Broken lamps and outputs releasing mercury vaporsshould be confined in airtight containers.

2. Accidental breakages should be avoided by handling softly.3. Existing processes should be improved by implementing

source-capture methods and/or keep the core process in a vac-uum confined enclosure.

4. Vacuum lock or a semi-automatic feed at the lamp input areashould be implemented to keep the workers away from pol-luted area.

5. The mercury level should be continuously controlled at specificpoints in the workplace by real time measurement device con-nected with alarm system.

6. A general exhaust ventilation system should be implementedfor all the workplace.

7. In addition to prevention by source reduction and collectiveprotection, suitable Personal Protective Equipment may berequired especially for limited and hazardous activities (e.g.cleaning and maintenance operations). Finally, good conditionsof hygiene should be ensured (e.g. hand washing before break,shower and clothe changing at the end of workday, workingclothes supplied and washed by company).

In order to support the French lamp recycling sector to developsafely, INRS provides technical advice and has drafted brochureswhich help to identify risks. These brochures are published in col-laboration with the French Fund for Retirement and Occupational

Please cite this article in press as: Zimmermann, F., et al. Occupational exposu(2014), http://dx.doi.org/10.1016/j.wasman.2014.03.023

Health (CARSAT), the Regional Health Insurance Fund (CRAM)and Recylum, an eco-organization (INRS, 2008; Zimmermannet al., 2011).

Acknowledgements

We are grateful to the eco-organisation, Récylum, and to thetreatment facilities which participated in this study. We wouldalso like to thank the technical teams at the Ingénierie des Procé-dés department at INRS, Juliette Jannot, Nathalie Monta, YvesMorèle, Thérèse Nicot and Isabelle Subra, who contributed to theanalyses and to developing the technical preventive solutions.

References

Aman, M.M., Jasmon, G.B., Mokhlis, H., Bakar, A.H.A., 2013. Analysis of theperformance of domestic lighting lamps. Energy Policy 52, 482–500.

Apisitpuvakul, W., Piumsomboon, P., Watts, D.J., Koetsinchai, W., 2008. LCA of spentfluorescent lamps in Thailand at various rates of recycling. J. Cleaner Prod. 16,1046–1061.

Asari, M., Fukui, K., Sakai, S., 2008. Life-cycle flow of mercury and recycling scenarioof fluorescent lamps in Japan. Sci. Total Environ. 393, 1–10.

Aucott, M., McLinden, M., et al., 2003. Release of mercury from broken fluorescentbulbs. J. Air Waste Manage. Assoc. 53 (2), 143–151.

Chang, T.C., You, S.J., Yu, B.S., Kong, H.W., 2007. The fate and management of highmercury-containing lamps from high technology industry. J. Hazard. Mater.141, 784–792.

Chang, T.C., You, S.J., Yu, B.S., Chen, C.M., Chiu, Y.C., 2009. Treating high-mercury-containing lamps using full-scale thermal desorption technology. J. Hazard.Mater. 162, 967–972.

De Michelis, I., Ferella, F., Varelli, E.F., Veglio, F., 2011. Treatment of exhaustfluorescent lamps to recover yttrium: experimental and process analyses.Waste Manage. (Oxford) 31, 2559–2568.

Directive 2009/125/EC of the European parliament and the Council of 21 October2009 establishing a framework for the setting of ecodesign requirements forenergy-related products.

European Commission, 2009. Full impact assessment – commission staff workingdocument accompanying document to the commission regulationimplementing Directive 2005/32/EC.

Guthrie, G., Dilworth, M., Sen, D., 2006. Reducing mercury exposure in fluorescentlamp manufacture – a workplace case study. J. Occup. Environ. Hyg. 3, D15–D18.

Hu, Y., Cheng, H., 2012. Mercury risk from fluorescent lamps in China: current statusand future perspective. Environ. Int. 44, 141–150.

INRS, 2008. La filière des lampes usages – Aide au repérage des risques dans lespoints de collecte et les entreprises de collecte et de recyclage. ED 6043.

INRS Metropol, 2000. Fiche Métropol 024. Mercure sur tubes de charbon actif.<http://www.inrs.fr> (accessed September 2013).

INRS Metropol, 2008. Fiche Métropol 003. Métaux-Métalloïdes. <http://www.inrs.fr> (accessed September 2013).

INRS Metropol, 2009. Fiche Métropol 002. Concentration pondérale d’un aérosol.<http://www.inrs.fr> (accessed September 2013).

Jang, M., Hong, S.M., Park, J.K., 2005. Characterization and recovery of mercury fromspent fluorescent lamps. Waste Manage. (Oxford) 25 (1), 5–14.

Journal Officiel de la Republique Francaise, 2009. Arrete du 15 décembre 2009relatif aux contrôles techniques des valeurs limites d’exposition professionnellesur les lieux de travail.

Lucas, A., Emery, R., 2006. Assessing occupational mercury exposures during the on-site processing of spent fluorescent lamps. J. Environ. Health 68 (7), 30–34.

Lundgren, K., 2012. The global impact of e-waste – addressing the challenge.International Labour Office. Program on Safety and Health at Work and theEnvironment. <http://www.ilo.org/wcmsp5/groups/public/—ed_dialogue/—sector/documents/publication/wcms_196105.pdf> (accessed March 2013).

re in the fluorescent lamp recycling sector in France. Waste Management

F. Zimmermann et al. / Waste Management xxx (2014) xxx–xxx 7

Nance, P., Patterson, J., Willis, A., Foronda, N., Dourson, M., 2012. Human health risksfrom mercury exposure from broken compact fluorescent lamps (CFLs). Regul.Toxicol. Pharmacol. 62, 542–552.

Ogden, T., Lavoue, J., 2011. Testing compliance with occupational exposure limits:development of British–Dutch guidance. J. Occup. Environ. Hygiene 9, D63–70.

Raposo, C., Carvalhinho, Windmöller C., Durao Junior, W.A., 2003. Mercuryspeciation in fluorescent lamps by thermal release analysis. Waste Manage.(Oxford) 23, 879–886.

Rasapo, C., Roeser, H.M., 2001. Contamination of the environement by the currentdisposal methods of mercury-containing lamps in the state of Minas Gerais.Brazil Waste Manage. 21, 661–670.

Recylum, 2012. Communiqué de presse. Bilan Récylum – 30 millions de lampescollectées en 2011. <http://www.recylum.com/wp-content/uploads/PRESSE-CP-2012-02-27-Bilan-2011.pdf> (accessed March 2013).

Rey-Raap, N., Gallardo, A., 2012. Determination of mercury distribution inside spentcompact fluorescent lamps by atomic absorption spectrometry. Waste Manage.(Oxford) 32, 944–948.

Please cite this article in press as: Zimmermann, F., et al. Occupational exposu(2014), http://dx.doi.org/10.1016/j.wasman.2014.03.023

Rey-Raap, N., Gallardo, A., 2013. Removal of mercury bonded in residual glass fromspent fluorescent lamps. J. Environ. Manage. 115, 175–178.

Rhee, S.-W., Choi, H.-H., Park, H.-S., 2013. Characteristics of mercury emission fromlinear type of spent fluorescent lamp. Waste Manage., in press. Available online17 september 2013.

Salthammer, T., Uhde, E., Omelan, A., Lüdecke, A., Moriske, H.-J., 2012. Estimatinghuman indoor exposure to elemental mercury from broken compact fluorescentlamps (CFLs). Indoor Air 22, 289–298.

Santos, E.J., Herrmann, A.B., Vieira, F., Sato, C.S., Corrêa, Q.B., Maranhao, T.A.,Tormen, L., Curtius, A.J., 2010. Determination of Hg and Pb in compactfluorescent lamp by slurry sampling inductively coupled plasma optimalemission spectrometry. Microchem. J. 96, 27–31.

Searl, A., Crawford, J., 2012. Review of health risks for workers in the waste andrecycling industry. Institut. Occup. Medicine.

Zimmermann, F., Lecler, M-T., Ihoua, S., Silvente, E., Chollot, A., 2011. Risqueschimiques dans la filière de valorisation des lampes usagées. Documents pour lemédecin du travail. Collection DMT assistance, TP 12.

re in the fluorescent lamp recycling sector in France. Waste Management