Environment International, Vol. 14, pp, 407-416, 1988 0160-4120/88 $3.00 + .00 Printed in the USA. All rights reserved. Copyright © 1989 Pergamon Press plc

AN APPRAISAL OF ATMOSPHERIC POLLUTION BY ALUMINIUM FUMES EMANATING FROM SMELTER WORKS IN WESTERN NORWAY

F. B. Pyatt a and D. Lacy b aDepartment of Life Sciences and, bDepartment of Physical Sciences, Trent Polytechnic, Clifton, Nottingham, UK

(Revised 8 September 1988; Accepted 30 September 1988)

Material including birch leaves, twigs, soil, and grass was collected from various sites around the Ovre /~rdal aluminium smelter in Western Norway. Samples were analysed by meazas ,of electron probe X- ray microanalysis to ascertain the relative amounts of deposited aluminium. The prevailing wind together with the steep sided valleys controlled the fate of the pollumm, ~uad material was found to sediment out from the atmosphere in lowland sites particularly to the ,no~llla-east, north-west, and east of the industrial complex; most deposition occurred within several kilometers of the plant. Some effects of aluminium on plant surfaces are discussed.

~,tnt[oduction

,OMuminium, an important metal, constitutes approxi- mately 8% of the earth's crust. It was first extracted by a chemical method in 1825 by Orsted, as reported in the /~rdal og Sunndal Verk report (1971). Later, work by Wohler and Sainte-Claire DeVille resulted in the production of aluminium on a commercial scale in 1855.

In 1886, H6ro~flt and Hall independently invented a process for producing aluminium by the electrolysis of alumina (alumimum oxide) dissolved in molten cryolite (Na3A1F6); subseqaea~y, smelters were developed in France, Germany, and ~a~re USA and, by 1900 world production of aluminiurn ~ , Jn excess of 7,000 tonnes per annum.

It was in 1908 that lhe first Norwegian smelter was initiated at Stongfjord, Sunnfjord and, in 1909 one was also developed near Kristiansand. In 1947 A/S ,~rdal og Sunndal Verk were established and developed from plants initiated by the Germans during the last War. Energy is efficiently provided by cables from the major power stations (hydro-electric) situated at Fortun and Tyin. The ~pl, a.nts (Figs. 1 and 2) are located in Western Norway ~o the south of the mountainous Jotunheimen area, with its important National Park (Fig. 2) and are located in the towns of Ardalstangen and Qvre ,~rdal (Figs. 2 and 3). Ardalstangen possesses excellent dock- ing facilities at the northern end of the Ardalsfjord; this fjord joins the Sognefjord, q.v., and thence leads to the

sea. The works at /~rdalstangen receives alumina from ships; in addition the plant at ,~rdalstangen also pro- duces the electrode paste and prebaked carbon for an- o d es and c a t h o d e s for the pots e m p l o y e d in the production process. Material is then transferred to the smelting and other facilities of Ovre .Ardal (Fig. 4) some 15 km to the north. The plants possess techniques for the removal of pollutants (such as aluminium fume and fluorine) from emissions emanating from the smelting plant.

According to the A I S ]krdal og Sunndal Verk report (1971), in 1970 the production capacity exceeded 500,000 tonnes which made Norway the fifth largest producer of aluminium in the world.

The group has a number of major smelters including • ~rdal, Sunndal Verk (Sunndals0ra), H0yanger Verk (H0yanger), and mines including Surnadal and Lasse- dal.

Boehlen (1968) noted that the most important ex- traneous substances in the atmosphere resulting from alumina reduction are inorganic fluorine compounds, alumina, tar, carbon, sulphur dioxide, and carbon monoxide. McCabe (1955) described the electrolytic reduction of alumina - - the oxygen liberated in the cry- olite bath combines with the carbon of the anode yield- ing carbon dioxide. As the carbon dioxide rises through the blanket of alumina covering the electrolytic cells, small amounts of alumina dust and fluorides become entrained. He noted the evolution of heat and stressed

407

408 F.B. Pyatt and D. Lacy

T ~(.4-- • Sunndalsqbra

• Fortun H1~y,,nger ~ & ~rda, " ~

~: A, Tyin A • Oslo / )

I8 Fig. 1. Western Norway; the Sognefjord.

• JOTUNHEIMEN

N SUNNFJORD*

;TONGFJORD

)ROE

VADHEIM

SOGNEFJO~RDEN

ca

URNES

LUSTERFJORD

o • (~VRE ARDAL

, ARDALSTANGEN

ARDALSFJORD

LAERDALS~RA

DAD

Fig. 2. The ,~rdalstangen and Ovre Ardal region; the smelter occurs at Ovre/~rdal to the east.

Atmospheric pollution 409

N

T

%%

~ . . ~ . JOTUNHEIMEN ~ . . % NATIONAL

PARK

s s ~ ~

1096m 1124m •

HELLERLIHALSEN •

VETTI•

1199m

HOYSTOLEN

ME

• 133gm

TRONTEIGEN

ALUMINIUM PLANT

<)

&

• 4 9 6 m

HJELLE

&1222m

U 0

(~VRE ARDAL

1Km | I

1089m

MOADALEN

RESIDENTIAL AREAS

MAJOR ROADS Fig. 3. ~vre ~rdal and associated valleys at the northern end of A, rdalsvatnet.

410 E B. Pyatt and D. Lacy

Fig. 4. The plant at Ovre ,~rdal (viewed from east).

the need for adequate ventilation in pot rooms; how- ever, he indicated that the high velocities of an induced draught will lift dust from the top of the pots.

The Ministerium for Arbeits report (1972) examined emissions in nonferrous industries in North Rhine- Westphalia and noted that, in the aluminium industry, the following quantities of pollutants were produced:

gaseous fluorine componds: 0.8--1.5 kg/tonne aluminium- and fluorine-bearing dust: 9-20

kg/tonne sulphur dioxide: 3-15 kg/tonne, and carbon mon-

oxide.

It should, however, be stressed that the plant at Ovre ,~rdal has the advantage of hydro-electric power.

Less and Waddington (1971) investigated the nature of aluminium reduction cell fume and found that the particulate matter exhibited a bimodal size distribution with one fraction consisting mainly of dust particles with- a diameter greater than 51xm, and the other of material with a diameter smaller than 1 Ixm. The coarse fraction included alumina, carbon, and cryolite droplets, whilst they describe the fine fraction as consisting mainly of condensed fluoride vapour approximating in composi- tion to chiolite (5NaF. 3A1F3) and that this fine partic- ulate accounted for approximately 35% of the total fluorine emission from the cells.

Location of Sampling Sites

~rdalstangen lies at the north-eastern end of ,~r- dalsfjorden; this fjord, to the west joins Lustrafjorden

and Lcerdalsfjorden and then leads ino the Sognefjor- den. To the north of the town lies the lake ,~rdalsvatnet which terminates at Ovre Ardal. The latter is located in a steep sided valley which leads into valleys including Fardalen (to the north-west), Utledalen (to the north- east), and Moadalen (approximately to the east); these are illustrated in Fig. 3. The restricting topography and wind direction are factors which will have major effects on the fate of atmospheric pollutants in this region.

The population of .~rdalstangen and Ovre/~rdal is approximately 7,000. Industrial activity in the latter town dates back to the early eighteenth century when this represented an important site for copper mining. With these towns are situated a number of smaller com- munities including Fardal, Utladal, Seimsdal, Nadvik, Ofredal, and Vetti; some of these are shown in Fig. 3.

North and north-east of these sites is the mountain- ous Jotunheimen region characterized by glaciers such as Storbreen, Sm6rstabbreen, and Tverr~breen, and mountain peaks including Galdh0piggen (2469 m), and Glittertind (2452 m). West lies the Hurrungane with glaciers such as Styggesdalsbreen; associated glaciers include Jervvassbreen (to the east), Maradalsbreen (to the south-east), Slingsbybreen (to the south), and Skardst61sbreen (to the south-west). Peaks in the region include Fannarhki (1068 m), Storen (2403 m), and Styg- gedalstindane (2387 m).

In Utladalen, near Vetti (Fig. 3), is situated the fa- mous Vettisfossen; a major waterfall possessing a 275 m free fall. This area, like many other sites in this region has vegetation exhibiting severe effects of fluorine pol- lution; the aluminium (q.v.) appears to have a more limited sphere of influence.

Atmospheric pollution 411

Field Work and Experimental Procedures

Samples were collected in clean plastic screw top containers during the course of ecological field work in the summer of 1987. They were obtained during dry weather and carefully harvested, subsequently air dried and thence transferred to the laboratory. In the various sites selected, the following material was col- lected (when available):

1. Birch (Betula): leaves and twigs - - all at a height of 2 m .

2. Soil: samples were collected with a nonmetal tool from the top 5 cm of the soil.

3. Grass: this was in each case an approximately equal mixture of Agrostis, Festuca, and Deschampsia.

Leaf material was also examined microscopically in an attempt to determine whether stomatai penetration and accumula t ion of par t icu la tes had occur red or whether deposition of particulates was confined to cu- ticular surfaces.

Results and Discussion

The results are presented in Tables 1 and 2 and three sample results are given in Figs. 5, 6, and 7. It must be stressed that the technique only records elements higher than sodium; replicates were used throughout.

The print out will list a variety of elements; thus, for example, one of the birch twig samples near Svalheim gave (in simplified format):

In addition, from a few sites were collected cores of birch or cores of Norway Spruce (Picea abies); these were obtained by means of a Pressler borer.

Preparation and analysis Each sample was mounted on a 13 mm diameter

carbon stub and held in place using conductive carbon cement. The samples were given a light coating of Du- ron Spray (an antistatic material produced by Hansa- werke of Bremen-Hemelingen) to eliminate problems caused by electrostatic charges and when dry placed in a Cambridge Stereoscan 600.

Samples were analysed by means of electron probe X-ray microanalysis with a Link System 860 series 2 computer using a ZAF-4 program, (where Z is the atomic number, A the absorption, and F the fluores- cence). In this process of X-ray microanalysis, an elec- tron beam strikes the solid specimen and a number of interactions occur including the production of X-rays. These X-rays are detected by a lithium drifted silicon detector and are passed on to a multi-channel analyser. Suitable areas on each sample (10,000 ixm -~) were se- lected for analysis using the microscope visual display monitor and analysed at a magnification of 500X for 100 seconds of live time at 25kV (electron beam en- ergy).

The quantitative accuracy of this technique on bio- logical material may be restricted to ± 10% relative of the true value (Goldstein et al., 1984). The results may be influenced by a combination of factors:

1. The variable geometry of the sample surface, 2. The sample itself may be a cause of contamination, 3. The majority of elements in biological material (i.e.,

carbon, oxygen, nitrogen, hydrogen), are not de- tected by the type of detector used, as the energy they generate is too low.

The technique is otherwise highly accurate; to main- tain maximum acccuracy of data 10 random areas were examined in each case.

Table l. An example of a printout of elements.

Element % element

Mg 0 AI 33.232 Si 6.967 P 0 S 10.110 K 4.685

Ca 36.586 Mn 0 Fe 7.031 Zn 1.398

Aluminium will be represented as AI as indicated above, but also as Si in, for example, the form of aluminosilicates (e.g., in the soil). Thus, soil samples will include in their percentage occurrence values not just aluminium from any pollution sources but also al- uminium from the normal soil sources.

In this article, little emphasis is placed on the soil samples and all data will only be examined in a com- parative manner.

Particulate pollution by aluminium fume was con- fined to sites close to the smelter (Table 2). To the north-east they occurred within 8 km of the works, to the north-west within 5 km, and, to the east, within 3 km. The fate is affected, as noted by Bovay and Bolay (1965), working on fluoride pollution, by the direction and nature of air movement , but also the topography of the area is important. The valley above the smelter is steep sided and consequently low altitude winds mov- ing in from Sognefjord (and the west) will largely move any atmospheric pollutants into one of the three major valley systems emanating from Ovre Ardal (i.e., Utle- dalen, Fardalen, and Moadalen). Of these, the former appears to be the major sink for such pollutants as the access into the valley is not steep as is the case with Fardalen and Moadalen. In fact, this is reflected by the results expressed in Table 2.

412 F.B. Pyatt and D. Lacy

C:. ~ 1

L i r,~::

: : : : : : : I : : : . . . . . . . . . . . . f p ~i~]~ ".-~i~!~.

: I : I ; I l l : l l i l ; ; I :1:1:1 11 : :1 : : : 1 : . : i

i~ ]]iiii iii]iiiiiii ii- ~: : : : : . : : : : : : : : : : : : : : : : : : : : : : : : : : : : :

" . ' . ~ - " " - 7 - ' . - " - - : " E ! : . E " ' " ~ ' " ! ' ! ! ! ' ! . . ' ! : " ' ! : . . ~ ! : . - ! ' E : ; : : : : : : . : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : :

iiifi!i!!!'i~ii~!iii~iiiii]iiiiii!iiiiii!iiiiiiiiiiiiiiiiii!~:i

I iiiii i~iii iliil

: : : : : : :

: : : : : : : : : : : : : :

!!!!ii!i

iiiH!i!~ !!!!!!Hi - - C a

?iiiiiiii?i

:i, iiiii!iiiii . . . . . . . . . : : :222:2222

!i!!iii!!!i !!!i!!i!!!i!!!

, iiiiiiiii! ,i]i]i ii iiiiii , ili ---ca i~!iiiiiii~i!iiiii~iiiii~iiiiiiiiiiiiiiii~i~ii!~!!~i!i~ii~ii]i~i~i~].~ii]i%~.~. ~.-' ~..iiiiiiiii~iii~..iiiiiiig~iiiiii!.." iiii~i~ i!!iiii~ii!ii!!..~iii~i!i!!!!i..'i!!!!~!!!!~ii!~ii]~i~.i~iiii~!~i~ii~!~i!ii~i~i!!..~ ~i'! :!iiii!!i!!i!'z.'i!ii~ii~i!ii~!i~i~!ii .:iiiiii~:

ii.~iFi'..i~i~i.!]~i~.'...iiii~]i~ii!i!i~iiii.i!~ii]ii~..i~ii~i~iiiii!]! iiii..~iii]~iii.-=..".4i~iii!ii]ii!i~i~ii~L "i~!~iii. - ~ . . : . . . . . . : . . . . . : : r . . . . . . : r .~ . . . r . . . . . . . . . . . . . . :~ . . . . . . . . . . . . . . : . . . . ~ . . . . . : r :~ r : . . : . . : . . : : .~ : . r . . . . . . . :~ . : .~ : .~ .~ .~ . . . -~ . . a . r . : .~ r r . .~ . . . . : . : r : . . - . . v . . . . . . .

i i i i i i . i i ~ i i i i i i i i i i i i i i . ' . i i i i i i ~ i i i i i ~ i i i i i i i i i ! i ~ i i ~ i i i i i i ~ i i i i i i i i i i i ~ i i i i ~ i ~ i i ~ i i i i i ~ i i i ~ i ! i ! . i i i i i i i i i ~ i ~ ! ~ i i i ~ i i i i ! i i i ! i i ! i i i ! i i i i i i i i ~ i i i i i i i . i i ~ i i i i i : i ! . . ~ i r ' . . . . . ~ . ~ . . ' ' .... = ; . .:

. . . . . . ~ . o l 1 . . . . . . . . . . . . . . . . ~ . , , . . . ~ . * i i ! ; i l ; ; l . % - ; ; : , ~ ; ; ; I I ; l ; ~ I I l l l l ; ~ ' ~ I ~ ; ~ ; : ; ~ ' ~ ' ~ . ~ t ; ~ ; ~ ; ; ~ : ~ ; ; ~ ; ~ I ~ I ~ t ~ ; ~ I I ~ I ~ I ~ t ; ~ I ~ I ~ I ~ I I ~ : ~ I ~ I ~ I ~ I . . . . . . . . . . . . . . . i . . . . .

.~i~i~iiii!i~iii!~ii~iiiiii~i~iii!~!iiiii}~iiiiiiiii~iii!i!~i~ii!!ii!i!i!iiiii~i!ii!~i!!i!i~iii!~!iiii~ii!!~i~4~ii~ii~ !iii!!i!iiiiiiiiiiiiii!iiiii~!iiii~!i !!i!!iii!!!!!!!i!!!i !! ! !i ! !! !!igg~!!-ill ..r..!!~i]~;!i]ii..~]]i~]iiii]i~i]]iii]]!ii!i]!ii!]iiii~ii~iigiii!~]!!]i]~i!iii]ii]iiiiiiii~ii]~ii!~iii~ii~iiiiiiiii~i.~.~]~]iiii~iiiii]i!!~!]i!i~i~igii!i~ii]ii~ii~i~iii]i~]ii]~ii!ii]~iiii~.i~iiii~!!]ii]i]!!ii]i!]ii~].~

~iiei!i~Niiigi~i~i~iiiiiiif~ifggiig~$~i~i~i~iiiii!g~g~$f~i~i$i$$i$i~ii$iiiiN$H$i!ii~ii$Hi!iiiH$i~iH~i~HHH~iNN~$iH~i~iiiii~iiHiHiHH~iWHiiiiiiiiiiHHHiHiiHHiH~i~H~ill~$H~H~$iH$i~ ! ] i i l l ] i i i , I I , l i , I ~ , i , l , i l

~ _ ~ ~ _ , ~ I " I I E I ' I i~ i " : 3 - 1 c

Fig. 5. Major elements in a sample of birch leaves•

The Ovre Ardal sites located approximate ly 100 m from the plant had a lumin ium c o n t a m i n a t i o n o f Betula twigs, leaves , and soil; note the earlier c o m m e n t s about soil a luminium values . The birch leaves carried more a luminium than the twigs and the average birch value (twigs and leaves) a m o u n t e d to 7 .0%.

The site near Sva lhe im (alt i tude approx imate ly 50 m) is located in the low altitude val ley Ut leda len at approximate ly 2.5 km north-east o f the plant. The re-

suits indicate major depos i t ion o f a lumin ium in this area which will reflect not only a tmospher ic po l lut ion but also such processes as soil eros ion , etc. The birch twigs here had a m u c h higher load than the birch leaves and the average birch value a m o u n t e d to 19.68%.

Samples of birch cores were also r e m o v e d to ascer- tain whe ther m e a s u r e a b l e quantit ies of a luminium had been accumula ted in the w o o d y t issues o f this tree; in addit ion, twigs f rom which the bark had been r e m o v e d

Atmospheric pollution 413

C N I . ::3 K F .?, -" B 3 ~ - . ~ E V 2 0 E V / C : H A N

r. ~ ¢. L i n k c, Hs era= 8~(~ A n a L , H s e r 5 - . M a r - B B

!

f S i

AI i f S f C a i i i i iiiiiiiiiiii!ii!iii~ii,ilili!iii . . . . i~i ~ ] :iii~i

i ii!i!iiiiiiiii i iiiiii iiiii i iiii iii ii i i iiiiiii ii ii ii iiii iiiii iiii iiiiiiiiiiiiiiiiiiiii ii iiiiiiiiii iiiiiiii!ill ii iiiiii iiiii! i iiiiiii i i i!!i !!! i i iii!! ii R E I W I ~-'% = 3 2 D

Fig. 6. Major elements in a sample of birch twigs.

were also examined. Aluminium was not found to be accumulated (in any age wood) from the atmosphere into the tissue or indeed from the soil system into the tissues of the stem or twigs.

The grass samples contained no aluminium and this conceivably can result from various factors such as; (a) the grass is sheltered topographically from wind car- rying particulates (would not apply during still condi- tions when particulate sedimentation would occur); (b)

leaf presentation, as noted by Pyatt (1973) is not con- ducive to accumulation of particulate pollutants; (c) some of the grass can be sheltered by adjacent taller species in a stratified system - - the sampling procedure avoided this; and (d) grass grows constantly from a basal meristem and replaces portions (which might be con- taminated) removed by processes such as grazing pres- sure or indeed by cutting.

The final sites in Utledalen (at approximately 8.0

414 F. B. Pyatt and D. Lacy

3 , ' 4 . ~ - ~ E V 2 0 E V J C l q l A N L i~k S~j~tems 86e AnaLtjser 5 - M a r - 8 8

1 ::::: !!~!

iiiil ilf~

~fiii i{i~i

i,~!i iiiii ii

Oa--i }}ii! iiH ge ii}}}, il}i}}iii}} ~ i}ii

iiHii i!iiiiiiiii !iH, fill iii!i} iili !!1!!1! ii!il !iii

. . . . . . . : : : : : : : : : : : : : 1 : :

!!i!!!i !!!! ~!i!ill !!!!! !~ : : : : : : : : : : : : : : : : : : ~ i ~ ~!E

!!!!1!! !1!1.1!!!!!! ~!!!!~ !i!! !!!i!!! ig!i!ii!!!!! i!!~!i iiiii . . . . . . . : : : : : : : : : : : : : : : : : : : : : : : :

: : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : :

iiiiii! ~i. i!i!iiiiiii~i i iiiii! i~

t!!!!!i !!!!i!ilt!!!!!tH! ii!i ~iiilt!! ~!. i!!iill l l l l ] i ~ l l l l l l l l i l l l l l t l l ~ l . ; , . l i l t : i .~I i~I II~ l l l l l | ~

iiiiiig iiiiiiiiiiiiiiiiiiiiii iiiii, iiiiiiiii ~iii iiiigiii ;?~!!! !!!!!!!!i!!!!!!~!{!!i:. : i!!i!!::!!~!!~ii :!: . !!}~!!! :=

: ~ l t t t

i!~!iiii ~ii~i~iii~iii~i!}!!~}~{i~!!!!i!!!iiii!ii~!!!!!!!!i~1i!i{!i!ii}iiiii~ii~iiii~iiii~8Rii{iii!{!ii~iiii~iiii!!!ii!1!~}i1~i~!!~ii{i!]iiiii!iiiii1~iiii~i!~iii!ii{~ii~iiiii~ii~ii]ii

l ' " ' l ' ' " l " " l ' l " l ' ' " l " " l ' ' " l ' ' " l " _ ~ 8 _ 5

M L ~ M A : 2 - ~ 5 A

Fig. 7. Major elements in a sample of soil.

and 9.0 km f rom the plant) had vege ta t ion which did not carry any a lumin ium deposits ; the par t iculates , thus sed iment out or impact well be fore these areas are reached.

To the nor th-west of the plant two sites were ex- amined as indicated. These sites, on the s teep valley side, were at al t i tudes of app rox ima te ly 274 m (Mel- helm) and 535 m (H~ystOlen). F r o m Table 2 it is ap- paren t that only birch leaves and twigs in Melhe im

showed con tamina t ion ; it appea r s that the dense par- t i cu l a t e s f r o m the r e l a t i v e l y low a l t i t ude e m i s s i o n sources (works are on the val ley f loor at approx ima te ly f jord level) do not genera l ly reach sites with an al t i tude much in excess of that found at Melhe im. The same appl ied to sites to the east of the works with alt i tudes of app rox ima te ly 325 m and 525 m.

Birch twigs then are a good indicator of pol lut ion of this type. They have the advan tage over the leaves, that

Atmospheric pollution 415

Table 2. Aluminium content of samples (dry weight) from various sites in the vicinity of the Owe Ardal smelter.

Average AP Approximate Content (of 10

Approximate Distance From Compass random areas) Site Location Centre of Smelter Bearing Material mg/g

~vre ~rdal 100 m N birch twigs 61.5 birch leaves 78.5 soil 169.5

Svalheim 2.5 km NE (Utledalen)

birch twigs birch leaves soil grass birch cores birch twigs without bark

291.9 101.7 111.9

0 0 (throughout)

0

Hellerlihalsen 8.0 km NE birch twigs (Utledalen) birch leaves

soil

Vetti 9.0 km NE (Utledalen)

birch birch soil grass

twigs leaves

Melheim 2.0 km NW (Fardalen)

birch birch grass birch

twigs leaves

cores

152.5 104.2

0 0 (throughout)

H6ystflen 5.0 km (Fardalen)

NW birch birch grass soil

twigs leaves

East of Works 0.5 km E birch twigs birch leaves soil grass birch cores

84.6 38.0 0 0 0 (throughout)

Tyin Road 3.0 km E birch twigs 0 (Moadalen) birch leaves 0

soil 0 grass 0

~The average aluminium contents are expressed in terms of a weight per unit area of sample; 10 replicates being employed in each case.

despite a relatively low surface area to volume ratio, of being exposed to the a tmosphere th roughou t the year. Also, as can be seen f rom Fig. 8 factors, including flu- orine pol lut ion, in the area have resulted in massive Birch d e f o l i a t i o n - - t h i s can lead to the twigs being more exposed and consequent ly more accessible to at- mospher ic pollutants.

The leaf samples f rom Svalheim in Ut ledalen were examined to assess part iculate distribution. It was noted that 60 - - 84% of the part iculates were associated with the upper leaf surface; this is in ag reement with Pyatt (1973) and Gilber tson, Kent , and Pyat t (1985). Sections

th rough the leaves, and direct leaf examinat ion, re- vealed that part iculate mat te r of the smaller d iameters had pene t ra ted the s tomatal openings; also cuticular accumulat ion was well represented. This conceivably can limit product ivi ty and also may affect herbivores on leaf surfaces and ult imately decomposers after leaf fall.

Soil samples f rom the sites furthest f rom the works conta ined no a luminium (below detectable level of the ins t rument) ; it appears then that long- term processes of climatic weather ing have mobil ised this aluminium. However , the occur rence of a luminium in sites near the

416 E B. Pyatt and D. Lacy

Fig. 8. Birch woodland, approximately 500 m, to east of plant; note severe defoliation which is to some extent the result of fluorine pollution.

works would indicate atmospheric deposition of alu- minium has exceeded the rate of climatic weathering.

Acknowledgements - - We thank Mr. Martin Hutchings for assistance in preparing the diagrams and Mrs. Joan Witter for typing the manu- script.

References

,~,rdal og Sunndal Verk Report (1971) Grcndahl & SOn, Oslo. Boehlen, B, (1968) Fluorine emission at aluminium works, Chem.

Eng. 221, CE 266-268. Bovay, E. and Bolay, A. (1965) Dispersion of fluorinated gases in

the Central Valley. Agr. Romande. Ser. A, 4(S) 33-36.

Gilbertson, D. D. Kent, M. and Pyatt, F. B. (1985) Practical Ecology, Hutchinson, London, UK.

Goldstein, I., Newbury, E., Echlin, P., Joy, D.C., Fiori, C., Lifshin, E. (1984) Scanning Electron Microscopy and X-Ray Microana- lysis, Plenum Press, London.

Less, L. N. and Waddington, J. (1971) The characterisation of alu- minium reduction cell fume. American Inst. of Mining, Metal- lurgical, and Petroleum Engineers publication, New York, N.Y.

McCabe, L. C. (1955) Atmospheric pollution. Ind. Eng. Chem. 47 95A-96A. Ministerium for Arbeits, Gesundheit und Soziales des Landes Nordrheim-Westfalen, Dusseldorf (1972). Nonferrous Metallurgy In: Reine Luft for morgen. Utopie oder Wirklichkeit. Moehnesee-Wamel, West Germany, K yon Saint George, 60-65.

Pyatt, E B. (1973) Some aspects of plant contamination by airborne particulate matter. Intern. J. Environmental Studies 5,215-220.