Methods for Sampling and Measurement of Compressed Air … · 2015-03-30 · AE-511 METHODS FOR SAMPLING AND MEASUREMENT OF COMPRESSED AIR CONTAMINANTS Lars Ström SYNOPSIS In order

AE-511 UDC 533.6.011:

661.91 532.51

Methods for Sampling and Measurement of Compressed Air Contaminants

L. Ström

AKTIEBOLAGET ATOMENERGI

STUDSVIK, NYKÖPING, SWEDEN 1976

AE-511

METHODS FOR SAMPLING AND MEASUREMENT

OF COMPRESSED AIR CONTAMINANTS

L a r s Ström

SYNOPSIS

In o rde r to improve the technique for measur ing oil and water

entrained in a compressed a i r s t r eam, a labora tory study has been

made of some methods for sampling and m e a s u r e m e n t . F o r this pu r ­

pose water or oil as art if icial contaminants were injected in thin

s t r e a m s into a tes t loop, car ry ing dry compressed a i r . Sampling was

performed in a ver t ica l run, down-s t ream of the injection point. Wall -

attached liquid, coa r se droplet flow, and fine droplet flow were s a m p ­

led separa te ly . The resu l t s were compared with two-phase flow theory

and d i rec t observat ion of liquid behaviour.

In a study of sample t r anspor t through na r row tubes, it was o b s e r ­

ved that, below a cer ta in liquid loading, the sample did not move, the

liquid remaining s tat ionary on the tubing wall .

The basic analysis of the collected samples was made by g rav i ­

m e t r i c methods . Adsorption tubes were used with success to m e a s u r e

water vapour. A humidity m e t e r with a sensor of the aluminium oxide

type was found to be unre l iab le . Oil could be m e a s u r e d selectively by

a flame ionization detector , the sample being p re t r ea ted in an evapo­

rat ion-condensat ion unit.

P r in ted and dis t r ibuted in October 1976

ISBN-91-7010-007-1

- 1

CONTENTS

page

PREFACE 2

1. INTRODUCTION 3

2. SAMPLING 7

2. 1 P r inc ip l e s 7

2 .2 Charac te r of flow 7

2 .3 Wall flow sample r 11

2 .4 Coarse a i r -bo rne liquid sampler 16

2. 5 Gas and fine a i r - b o r n e liquid sample r 21

2 .6 Conclusions concerning sample r s 23

2. 7 Calculation of vapour and liquid flows from 24 sample flows

3. SAMPLE TRANSPORT 24

3. 1 Est imation of t r anspor t velocity 2 6

3.2 Exper iments on t r anspor t velocity 27

4. SAMPLE PREPARATION 28

4. 1 Requirements regarding sample flow to 28

ins t ruments

4. 2 Mechanical homogenization of sample flow 30

4 . 3 Sample vaporization 31 4 . 4 Evapora tor 34 4. 5 Condensation unit 35

5. INSTRUMENTS 37

5. 1 Investigated methods 37

5.2 Humidity m e a s u r e m e n t 37

5. 3 Hydrocarbon m e a s u r e m e n t 39

6. NOTES ON CONTINUED INVESTIGATIONS 43

7. REFERENCES 45

- 2 -

PREFACE

The investigation repor ted here was init iated by the Swedish A s s o ­

ciation of Metal Working Indust r ies , Working Group for Fluid Systems

Technology.

The project has been directed by a committee with the following

m e m b e r s :

I Ahlberg AB Westin & Backlund

K C Blomqvist AB Mecman (par t - t ime)

B Grans t röm Atlas Copco AB

L Innings AB Mecman (par t - t ime)

O More L. M Er ics son AB (par t - t ime)

L Ström A B Atomenergi (project leader)

Labora tory work has been performed by G Gebert and A Dahlgren

of AB Atomenergi .

The project has received financial support from the Swedish Board

for Technical Development.

This repor t is also published in the publication s e r i e s of the

Swedish Associat ion of Metal Working Indus t r ies .

- 3 -

1. INTRODUCTION

Most compressed a i r applications requ i re a l imit to be set on the

amount of contaminants in the a i r s t r e a m . Also the a i r may c a r r y a

lubricant , which is n e c e s s a r y for the operat ion of an a i r -d r iven tool.

In such a case the application also demands that a lower l imi t is set

on the amount of oil . But too much oil may lead to contamination of

the air at the workplace . It accordingly becomes evident that technical ,

economical and occupational health aspects combine to requi re wel l -

defined contamination^ levels in different applicat ions. This situation

has been recognized by the International Organization for Standardiza­

tion (ISO) which has formed a commit tee for the s tandardizat ion of

compressed air quality (Technical Commit tee 118, Working Group 2:

Quality of compressed a i r and influence on environment from the use of

pneumatic equipment).

Some standardizat ion of compressed air has a l ready been c a r r i e d

out (ref 22, 24, 2 5), but it is re levant only to cylinder a i r .

The n e c e s s a r y m e a s u r e m e n t technique has received l i t t le at tention.

Only a few relevant publications have been found (ref 14, 20, 21 , 22,

26). Of these , the d i sse r ta t ions of Sittel and Muno a r e the mos t i m ­

portant . They have both studied oil m i s t lubricat ion. The difficulties

of sampling were avoided by measur ing droplets optically in the p r e s ­

sur ized conduit. A l a rge body of data on oil droplet generat ion and

t r ansmis s ion is repor ted .

The publications by Evans and from CGA a r e concerned with cylin­

der a i r , and the observat ions of these papers a r e not of immedia te in­

t e r e s t in this context.

The project repor ted he re was undertaken with the main objective

to improve the m e a s u r e m e n t technique. It is des i rab le to develope

methods which can be applied in the field without r e c o u r s e to skilled

personnel and labora tory faci l i t ies . Measurements should preferably

be p resen ted as continuous record ings .

F o r the sake of simplici ty the t e rm "contamination" is used he re for any liquid or vapour in the a i r s t r eam, whether i ts p r e sence is intentional o r not.

- 4 -

Measurement range

Compressed a i r is used for many purposes , with widely different

to lerance to contamination. To l imit the scope of this investigation,

medica l and food process ing applications w e r e excluded, and also

breathing a i r . P a r t i c l e s were not regarded as impor tant . Still, the

range of the p a r a m e t e r s to be covered is considerable . The following

contaminant l imi ts were set, to comprise mos t indust r ia l applications:

TABLE 1:1

Contaminant l imi ts

Water as liquid 300 - 1, 000 m g / m 3 (Norm)»

Water as vapour 10 - 3,000 " "

Oil as liquid 10 - 1, 000 " "

The maximum liquid water concentration cor responds to a comp­

r e s s o r with an af tercooler and a cyclone water s epa ra to r . The m i n i ­

mum water vapour concentration corresponds to a dew point at a t m o s ­

pher ic p r e s s u r e of about -55 C.

The upper l imit of the oil range can occur at oil m i s t lubricat ion,

while the lower concentration can descr ibe "oi l - f ree a i r " in many app­

l ica t ions .

The vapour concentration of oil is not specified. Lubricat ing oil

has a very low vapour p r e s s u r e at room t empera tu re , see F igure 4 . 2 .

Som information on oil concentration can be found in ref 19.

The a i r flow va r i e s considerably from smal l a i r tools to l a rge

c o m p r e s s o r s . The flow of contaminant va r i e s under these conditions

over m o r e than four decades , F igure 1. 1. It has not been possible to

cover this la rge dynamic range in one set of equipment.

Na ture of contaminants

P a r t i c l e s of sand, rus t , scale , etc may be des t ruct ive to pneuma­

t ic equipment. But as a l ready mentioned, they a r e left out of conside-

"Norm" impl ies that the volume is r e f e r r ed to the s tandard con ditions of 10132 5 Pa (760 mm Hg) and 20°C.

- 5 -

ration in this investigation, mainly because the m e a s u r e m e n t technique

is very different from the technique for liquids and vapours .

<P|oiio o f liqur'd ^*v\q / s )

\ooo

\oo

•10

OA

ö.oi

mm mm

i£ry v*

7

r « 5 - • . • ' . -" • . - , ! • '? '

••••>^W,:

my**?/ We/

^g|| r ' -- - •,*•-! " J " '

/

/¥V«?V

4o -loo 1000 -»oooo Afr f W (T £ <$W*>) /5 )

Fig i . 1 Assumed contamination concentration and flow for liquid water and oi l .

The oil in compressed a i r is mainly of two kinds: compres so r oil

and a i r tool oi l . In our exper iments we have used both, Table 1:2.

TABLE 1:2 SPECIFICATIONS OF OILS EMPLOYED

Designation in this r epor t

Manufacturer

Manufac turer ' s designation

Density, k g / m

Viscosi ty , c St

O

Ol j ekon sum ent e rn a

K-14

870 (20°C)

35 (20°C)

11.2 (50°C)

3. 6(99°C)

N

Nynäs -Pe t ro l eum

Compres so r oil 35

875 (15°C)

58 (38°C)

7. 6(99°C)

- 6 -

Tes t loop

To facil i tate exper iments on sampling technique, a tes t loop was

set up in the labora tory as shown in F i g u r e 1. 2. The loop consis ts of

one ve r t i ca l and two horizontal sections of s ta in less s tee l tubing, inter­

nal d iameter 16 m m . In the f i rs t horizontal section, oil and water can

be injected into the dr ied, f i l tered, and p r e s s u r e - r e g u l a t e d a i r s t r e a m . -5

F i l t e r penetrat ion is stated by the manufacturer to be l e s s than 10

for submicron pa r t i c l e s .

1 D=Q QsxutrteroJklva differ ftettune ._., A

a = ^ v

tUJWtcr

= $ > < $

"Plow FIQK>

droplet* Sample

Fig 1.2 Compressed air loop for investigation of sampling and sample t r anspor t .

Different sampling devices a r e being tes ted in the ver t ica l section.

The ver t ica l pa r t before the s ample r s i s c 400 m m long.

- 7 -

The a i r flow is m e a s u r e d and regulated in the second horizontal

section. The p r e s s u r e is usually kept constant at 600 kPa above a t m o s ­

pher ic p r e s s u r e .

2. SAMPLING

2. 1 P r inc ip l e s

The flow of contaminants may appear as liquid flowing on the con­

duit wal ls , as f ree droplets in the a i r s t r eam, or as vapour. Sampling

of wal l -a t tached liquid and of coa r se droplets mus t be done by complete

extraction, because it is impossible to make sure that any par t ia l flow

is r ep resen ta t ive . On the other hand, vapour and fine par t ic les cannot

be extracted completely and made available at a measur ing ins t rument .

But for these fractions a represen ta t ive par t ia l flow can be withdrawn,

under suitable precaut ions . Thus in o rde r to m e a s u r e the contaminant

flow in all of its phases , the two ca tegor ies wal l -a t tached flow and

coa r se a i r borne drople ts , and fine a i r -bo rne droplets and vapour,

mus t be separa te ly sampled.

2. 2 Charac te r of flow

To gain a bet ter foundation for the design of s a m p l e r s , a study

has been made of l iqu id- in-a i r flow.

A mixed s t r eam of gas and liquid is a case of two-phase flow.

Because of i ts impor tance in some engineering applications (heat

t rans fe r , gas- l iquid react ions) it has received considerable attention,

from a prac t ica l as well as from a theore t ica l point of view (refs 3, 6,

8, 9).

The mode of flow i s influenced by the p roper t i e s of the liquid, the

gas , and the conduit. This i s i l lus t ra ted by the well-known Baker

d iagram (ref 2, ref 3 p 727), showing observed flow pa t te rns in a ho r i ­

zontal c i rcu lar tube. To incorpora te the very low liquid flows of i n t e r ­

est he re , the d iagram has been extended to low flows, F i g u r e 2. 1. The

contaminant flow l imi ts expressed by Table 1:1, and common a i r flows

pe r unit a r ea in compressed a i r conduits form the boundaries of the

region of in te res t . The region is drawn as a rectangle in F igu re 2. 1

(marked — • — • —).

- 8 -

A X (*$/*>")

ICf

io

Riv/OU6.T P L O W

Muno e*ipar-i*vN2*\ta

^ = 3 r'

iz

IO<t US 12*,

\ \ \ lot fl6| \I2I

f >v- oil . TaWe. (=1

IO

Wall f lovo

-A lO . - • 2 .

\o - 2

to JO >o 10"

A

9 cr

Liquid wvA&«> ? 5 o w ( W $ / s )

L-iawidl SvAr'Kxce.+O.KvVlOA ^ N / w \ )

IO"

1, /« water

7 wovter"

n U^yud ulsoe&VhA C ^ ^ / " ^ v

F i g 2. 1 B a k e r d i a g r a m showing flow m o d e s in a h o r i z o n t a l t u b e .

By m e a n s of the B a k e r d i a g r a m i t i s p o s s i b l e to c o m p a r e r e s u l t s

f r o m e x p e r i m e n t s m a d e u n d e r d i f f e ren t cond i t ions of l iqu id s u r f a c e

t e n s i o n , v i s c o s i t y and d e n s i t y , and gas d e n s i t y . O t h e r s i m i l a r c o r r e ­

l a t i o n s a r e d i s c u s s e d in ref 6.

R e s u l t s apply ing to low l iqu id f lows a r e v e r y s p a r s e . L e v y ( re f 12)

h a s d e r v i e d e x p r e s s i o n s for f i lm t h i c k n e s s and l iqu id e n t r a i n m e n t in to

the c o r e . T h e s e da t a ex tend to r a t h e r th in f i l m s , l e s s than 1 % of p ipe

r a d i u s . At a low l iqu id flow the f i lm b r e a k s u p into th in s t r e a m s ,

t e r m e d r i v u l e t s . A c c o r d i n g to Quand t ( r e f 16) t h i s o c c u r s at a l iqu id

load ing of l e s s than 450 m g of w a t e r p e r s e c o n d and p e r cm of condui t

p e r i p h e r y . In ref 15 a l i m i t of about 100 m g / c m , s h a s been o b s e r v e d ,

a l so for w a t e r . A l i m i t of 200 m g / c m , s i s p lo t t ed into F i g u r e 2 . 1 ,

c a l c u l a t e d for the 16 m m i n t e r n a l d i a m e t e r of tube of the t e s t l oop .

- 9 -

, T|his re luctance of the liquid to spread on the conduit wall is of i m ­

por tance when designing a wall flow sampler , and also for the t r anspor t

velocity for wal l -a t tached oil .

In an investigation on oil m i s t lubricat ion, Muno (ref 14) has m e a ­

sured deposition and resuspens ion in a conduit. The oil, injected in the

air line by means of a lubr ica tor , was f i rs t deposited, but further down

the conduit became almost completely a i r -bo rne again (F igure 2 .2 ) . The

conditions for this observat ion a r e also shown in F igu re 2. 1 (marked

— ) .

T w b a l<Z.*\.Qt-K (*y-\

Fig 2.2 Resuspension of oil in a compressed a i r l ine. The oil is f i r s t deposited, but further down the line is par t ly r e -suspended. a = core oil f low/total oil flow. Steel tube, in terna l d iameter 8 m m . Air flow 20 m 3 (Norm)/h . (ref 14).

To fur ther clarify the mode of flow at low l iquid- to-gas ra t ios ,

some exper iments were made in the t e s t r ig . The s ample r s in

F i g u r e 1. 1 were replaced with a piece of glass tubing so that the

flowing liquid could be observed. F igu re 2 .3 shows* some photographs

of the flow modes and also gives details on a i r and oil flow.

F r o m the photographs it is c lear that the oil does not form a con­

tinuous film inside the glass tube. Instead, the liquid forms a na r row

rivulet , occupying only a smal l fraction of the conduit per iphery . The

same phenomenon of contraction has been observed at the sample

t r anspor t tubes , F igu re 3. 2. It is queer to note that an inc reased a i r

- 10 -

8.3 l(Norm)/s

111 U i L 126 %

5.6 — >j —

104^^™ 113 JrL 124

2.8 - » -

102

n

115 122

1.4 )J

116 L 121

2mg/s 3.4 mg/s 8.7mg/s

Fig 2. 3 Observed flow pa t te rns of oil r ivulet in t es t loop glass sec­tion. The a i r and oil s t r eam ver t ica l ly downwards. Atmos­pher ic p r e s s u r e .

- 11 -

s t r eam makes the r ivulet thinner , except at the highest liquid flow,

where it instead becomes b roader and tends to spread . The oil flow of

8. 7 m g / s corresponds to about 2 mg per second and cm of per iphery .

As mentioned above, a continuous water film will develop only above

c 200 mg per sec and cm of per iphery .

The a i r and oil flows corresponding to the photographs in F igu re

2. 3 have been plotted in the Baker d iagram. F igu re 2. 1 (numbers 101 -

126). The p ic tures support the predict ion of the Baker d iagram, that

at high flows of a i r and liquid, r ipples and waves a r e c rea ted and some

liquid is torn loose forming an a i r - b o r n e m i s t (d i spersed flow). Also

Muno noted high a i r - b o r n e fractions at these air and oil flows. As will

be seen below (p 15), the t e s t s of the wall flow sample r also indicated

a higher a i r - bo rne fraction above an air loading of 10 k g / s , m .

The information on two-phase flow of low liquid contents cited

above can be summar ized thus:

At the a i r and liquid loadings of in t e re s t to this project the liquid

is completely wal l -a t tached at low a i r flows and may be a lmost comp­

letely a i r - b o r n e at high a i r flows. The fraction of a i r - bo rne depends

strongly on the length of conduit through which the liquid has t rave l led .

When wal l -a t tached, the liquid moves in one o r seve ra l r ivu le t s . At

the lowest a i r flow the liquid may be strat if ied in horizontal tubes, but

at higher has flows conduit orientat ion is of l i t t le impor tance .

The fraction of a i r - b o r n e depends on the liquid loading, but ev i ­

dence is contradic tory . F i g u r e 2 .2 shows a dec reased a i r - b o r n e

fraction for m o r e liquid, but the photographs 106, 111 and 126 of

F i g u r e 2. 3 to higher resuspens ion at a higher liquid flow. In F i g u r e 2. 1

these photographs and the Muno exper iments cover about the same a r ea .

2. 3 Wall flow sampler

Design

Some different a r r angemen t s to ex t rac t wal l -a t tached flow a r e

shown in F igu re 2 . 4 . The porous wall models a r e easy to build. A gas

s t r eam is maintained into the annular chamber from the flow co re .

This s t r eam sweeps the wall flow through the porous wal l . A main diffi­

culty with this type of sampler is to de te rmine the n e c e s s a r y sweeping

flow. With an o i l /wa te r mix ture , problems may also a r i s e if the oil

makes the porous m a t e r i a l hydrophobic.

- 12 -

C . Rsron-% ww.ll. 6«.*v-vple*~

I^^WN

iSjlpS b . Porow.'fc <*>oM &a.mp\o.r voifK iwtemoA

L W W ^ ^

C.. C-OW'CQ.1 6I0V 6c\w^p\<i«- LQ.a?. H « y g K e )

Fig 2 .4 Wall flow sample r s from the l i t e r a tu r e .

A sampler of the conical slot type is shown in F igure 2 . 4 c . The

liquid is conveyed into the slot by means of a concurrent a i r flow, and

also by surface tension fo rces .

The kind of sampler employed in this project i s shown in F i g u r e

2 . 5 . It is s imi l a r to the Huyghe model , except that the tube d iameter

i s slightly sma l l e r downstream of the slot than u p s t r e a m . This s impl i ­

fies manufacture , in that the centering of the u p s t r e a m and downstream

tubes i s then not so very c r i t i ca l . (It has been observed that ups t r eam

tube "hangover" causes liquid to skip the s lo t . )

There is no theore t ica l foundation for the design of the wall flow

sample r . Several opposing requ i rements mus t be met : a wide slot to

accommodate l a rge r ivulets of liquid, but a na r row slot to get a high

gas velocity into the slot; a high sampling gas flow, also to i nc rease

the slot gas velocity, but a low gas flow to min imize in te r ference

from liquid and vapours in the core flow.

The sampler i s thus designed on ra the r weak foundations. Some

exper iments to t es t i t s per formance were deemed to be n e c e s s a r y .

- 13 -

Fig 2. 5 Wall flow sample r employed in the exper iments .

Exper iments on wall flow sampler

The sampler was mounted in the ver t i ca l section of the tes t loop,

the flow being downwards for a i r and liquid. The p r e s s u r e in the loop,

re la t ive to the a tmosphere , was 600 kPa .

A flow of liquid was injected in the horizontal pa r t of the tes t loop.

Care was taken to place the liquid on the tube wall, avoiding the fo rma­

tion of droplets in the a i r s t r e a m .

Liquid extracted by the wall flow sampler was collected in a cylinder

at A, F igu re 2. 6. Liquid pass ing the sampler was collected at B.

The liquid flows were calculated as collected volumes divided by

sampling t ime . F o r injected flow a special f low-meter was used .

- 14 -

Q

LiqiAic\

Pilter- fov valvs profecticm

wam/vple -floiw <o*\+rol välva

dvAlJrvdo:»- A

öylmd^t- g>

WAU flow

liqiAid Sa\n/vp\<z.r~

Fig 2 .6 Exper imenta l a r r angement s for t e s t s of wall flow sample r .

Sampling efficiency T] is calculated as the rat io between the ob­

tained sample flow q . and the total flow q. + q_

Tl ^A

*A + q B

The liquid flows employed a r e equal .to or exceed the flows c o r r e -•2

sponding to the upper l imit of concentration, 1 000 m g / m (Norm). It

is felt that this should suffice for the whole concentration range, since

a lower flow should be eas ie r to deflect into the sampling slot .

- 15 -

Exper iments with water

Observed efficiency at the slot width 0. 3 m m is shown in F igu re

2. 7 as a function of extracted a i r flow. Efficiency i n c r e a s e s with in­

creas ing a i r flow, but complete collection cannot be achieved at this

slot width.

E$*f tcierxcy, E

5 1o

Fig 2 .7 Wall flow sample r efficiency at diffe­rent slot widths and a i r flows.

With the slot width inc reased to 0. 7 m m the efficiency rose to

a lmost 100 % at all sample gas flows. Only smal l amounts of liquid

w e r e found at B .

This resu l t also shows that, under the c i r cums tances of the ex­

per iment , the re was no liquid in the core flow. In the tes t s with the

slot width 0. 3 m m the flow conditions were the s ame , and the liquid

found at B was thus wall flow, which had skipped the sample r .

The invest igated range of a i r and liquid flows is plotted in F i g u r e 2. 1

(marked o — o — o).

Exper iments with oil

Oil of type 0 was used . The exper imenta l a r r angemen t s were the

same as with wa te r , but only one slot width, 0. 7 mm, was tes ted .

The t e s t s showed that with a high a i r flow in the conduit the in­

jec ted oil (about 8 m g / s ) could not be extracted completely, Table 2 : 1 .

- 16 -

TABLE 2:1 INFLUENCE OF SAMPLE AIR FLOW ON WALL

FLOW SAMPLER EFFICIENCY

Main, a i r flow Sample a i r flow Extracted through wall flow sampler

1 (Norm)/s ml (Norm)/s %

2.8 100 100

2.8 200 99

2.8 385 99

5.6 100 69

5.6 200 75

5.6 385 67

8.3 100 66

8.3 200 67

8.3 385 66

The extracted gas flow, intended to c a r r y the wal l -a t tached liquid

into the slot and through the sampling l ine, was var ied . The table above

shows that this did not influence the amount of oil collected. It is con­

cluded that the oil, as opposed to water , i s par t ly in an a i r -bo rne s ta te ,

not access ib le for sampling as a wall flow.

This s e r i e s of t es t s corresponds to the photographs 121 - 124 of

F i g u r e 2. 3. The investigated range of a i r and liquid flows is also shown

in F i g u r e 2. 1 (marked o — o — o).

2 .4 Coarse a i r -bo rne liquid sampler

This sampler is intended for collection of coarse a i r -bo rne m a t e ­

r ia l that cannot be sampled by the Levin nozzle (p 21). This fraction

has a s trong tendency to become deposited. If the sample r were not

t he re , this liquid would soon be deposited anyway and take pa r t in the

exchange of m a t e r i a l between wall flow and core flow. Because of th is ,

the sample could be said to be a pa r t of the wall flow. On the other

hand, it has a higher velocity of t r anspor t and is in this way m o r e s i ­

m i l a r to the fine a i r -bo rne m i s t .

This sampler could also be used to collect the wall flow if the re

i s no in t e re s t in the subdivision of these f rac t ions .

- 17 -

Design

The sample r works through the combined action of iner t ia forces

and gravitat ion. The outlines a r e shown in F igu re 2 . 8 .

The unit is intended to opera te with the gas and liquid flowing v e r ­

tically downwards. At high gas flows, coa r se pa r t i c l e s a r e impacted

on the surfaces below the nozz les ; at low gas flows sedimentation will

deposit the l a r g e r pa r t i c l e s .

Coarse droplet

O 5 10cm i i i i I i i i i I

do<u<$e droplet

Fine droplef-ScMvtpUVtcv oriftc«

TMMC droplet Scwvxple JJ Sa**vpl«

Fig 2 .8 Sedimentat ion-impaction sample r for coarse a i r -bo rne liquid.

- 18 -

Impaction

Impaction efficiency can be calculated from published data, (ref 13,

p 229). This kind of deposition is governed by the Stokes number :

Stk

A2

v d v o PYP 9Tp

w h e r e

Y P

n D

velocity in the nozzle ( m / s )

par t ic le d iameter (m)

par t i c le density (kg /m )

gas viscosi ty ( N s / m )

nozzle d iameter (m)

About 50 % of the par t i c les of the s ize in question will be deposited

at a Stokes number of 0 .2 . If the Stokes number becomes lower, the

deposition efficiency will be sma l l e r , and vice v e r s a . The impaction

efficiency for the nozzles in F i g u r e 2 .8 has been calculated, F i g u r e 2. 9.

Trwr\btv\!&6for\

1

\o 30 \oo

Fig 2. 9 Calculated t r ansmiss ion through the sedimentat ion-impact ion unit, based on the impaction effect. Ai r flow 0. 3 - 30 1 ( N o r m ) / s . Absolute a i r p r e s s u r e 700 kPa .

- 19 -

S edim entation

At low a i r flows sedimentation rep laces impaction. To calculate

the deposition effect, the sedimentat ion-impact ion unit i s imagined to

consist of six consecutive boxes, F igu re 2. 10. We a s sume for a moment

that all pa r t i c l e s a r e of the same s ize . If the ae roso l in each box is well

mixed, the ra te of deposition in a box i s

c. A. v (kg/s)

The notation is explained in the f igure. The m a s s balance for a

box becomes

Q ' C i - l = C i A i V s + Q ' Ci

and for all boxes

A ,

c. c,_,

Q' A, A;

-on*-—oiM-

Air £low «t- pressure. |_ t^ (6 J

P«rHd<£ oemc. \v\ tt\e \¥r\ b o * [ k ^ , / m J

•l-lorizonfoil a r e a vt\

V% 4Qec\\r*OJHb*Hov\ velocihj L ^ / s J

F i g 2. 10 Sedimentat ion-impaction unit. Sedimenta­tion in a number of se r ies -connec ted boxes.

- 20 -

"out

in n

Q' Q* + A. v

i s

F o r the g e o m e t r y in q u e s t i o n , w e can a s s u m e a l l A. to b e about

t h e s a m e . With s i x b o x e s , the c o n c e n t r a t i o n r e d u c t i o n b e c o m e s

ou t c i n

A v

"Q7 Mi-t-Tv^r6

The e x p r e s s i o n h a s been e v a l u a t e d f o r a n u m b e r of a i r f lows ,

F i g u r e 2. 1 1 . By v a r y i n g the p a r t i c l e d i a m e t e r t h e t r a n s m i s s i o n

t h r o u g h the u n i t i s o b t a i n e d for a l l s i z e s .

i

0.5

- \ \

- -»oA

i 1 1 1 1 1

A 3 \

1

1 0 \

I 1

?A

1 1 1 >

\0 hO 100

F i g 2 . 11 C a l c u l a t e d t r a n s m i s s i o n due to the s e d i m e n t a t i o n effect . A i r flow 0. 3 - 30 1 ( N o r m ) / s . Abso lu t e a i r p r e s s u r e 700 k P a .

To ta l depos i t i on

It now r e m a i n s to c o m b i n e the depos i t i on effects of g r a v i t y and

i n e r t i a . Th i s canno t b e done on t h e o r e t i c a l g r o u n d s , bu t a q u a l i t a t i v e

p i c t u r e i s e a s i l y d r a w n , F i g u r e 2 . 12.

- 21 -

O 5 ho \% lo 1% 30

Fig 2. 12 Es t imated t r ansmiss ion for pa r t i c l e s through the sedimentat ion-impact ion unit, gravity and iner t ia working s imul ­taneously. Absolute a i r p r e s s u r e 700 kPa .

The penetrat ion of pa r t i c l e s va r i e s with the flow. 10 |j,m pa r t i c l e s

a r e t r ansmi t ted at all flows but the lowest, while only a smal l fraction

of the par t i c les g r e a t e r than 40 ^m will pass at any flow.

Tes ts

The design descr ibed above has not been prac t ica l ly tes ted . A s i m i ­

la r sampler has been used in the t e s t s with the wall flow sample r to

collect passing liquid (Figure 2. 6).

2 . 5 Gas and fine a i r - b o r n e liquid sampler

Design

The usua l way to withdraw a par t icula te sample from a s t reaming

ae roso l is to a r r ange an isokinet ic p robe . This is a nozzle with i t s

thin-walled opening d i rec ted against the flow, and with a velocity in

the inlet equal to the velocity in the main s t r e a m . This r equ i r e s that

the sampling flow is proport ional to the main flow, and under varying

flow conditions this is not easi ly achieved.

Instead it is advantageous to use a special kind of sampler , h e r e

called Levin sample r (ref 11). Here the aeroso l is extracted through

a smal l orif ice at a high veloci ty. The high p r e s s u r e available makes

this easy. The je t of a i r from the nozzle leads to high deposition in

- 22 -

the sampling line, but this is not so ser ious with a liquid as with solid

p a r t i c l e s . Deposition takes place anyway in the fi l ter protect ing the

flow-controlling valve in the sample t r anspor t l ine.

The sampler ext rac ts a predictable fraction of pa r t i c les of a given

s ize , the fraction being expressed by

p, = 1 - 0.8 K + 0.008 K2 - . . .

K = T ( 4 T r / q ' ) l / 2 U 3 / Z

where

sampling efficiency = (cone in sample) /conc in main s t r e a m (-)

= d p /i8T] par t ic le relaxation t ime (s)

par t ic le d iameter (m)

par t i c le density (kg/m )

gas viscosi ty ( N s / m )

sample flow (m / s )

vector ia l sum of gas velocity and sedimentation velocity ( m / s )

The formula is applicable only at efficiencies g r ea t e r than 0 . 8 .

The velocity of the oncoming pa r t i c l e s , U, is made low by using a wide

exit opening of the coarse droplet separa to r (Fig 2 .8 ) .

Sampling efficiency has been calculated at a sample flow of 100 m l

( N o r m ) / s , F i g u r e 2. 13. When the a i r flow is high, only smal l pa r t i c les

a r e sampled cor rec t ly , but this is of no impor tance , since the sedimen-

tat ion- impact ion unit effectively removes any la rge par t i c les at high

flows.

T_e_sts

A Levin sample r has been built and tes ted in the tes t loop. Unfor­

tunately the amount of liquid s t reaming into the sampler was too smal l

for sample t r anspor t . As is explained in some detail in Chapter 3

(Sample t r anspor t ) , the liquid does not move in a sample t r anspor t

line if the amount of liquid is too smal l . The maximum liquid flow from

the Levin sampler will be

1 000 m g / m 3 (Norm) • 100 m l (Norm) /s = 0. 1 m g / s

T

d P

PP T]

q*

u

- 23 -

100

30

w to /

Fig 2 .13 L a r g e s t pa r t i c l e s to be sampled with an e r r o r l e s s than 0.8 in the fine core flow sample r . Absolute a i r p r e s s u r e 700 kPa .

The flow is then normal ly l e s s than the min imum acceptable flow

of 0. 1 m g / s (Figure 3. 1).

Under the given conditions, collected liquid cannot move in the

sample t r anspo r t l ine . But on the other hand, with a low concentration

of liquid, the liquid can be evaporated and moved in the gas phase . It

has been calculated that, under reasonable assumptions regarding the

oil, a t empe ra tu r e of about 2 50 C would suffice to get a vapour p r e s s u r e

corresponding to 1 000 m g / m (Norm). Thus a heated sampling l ine

can solve this p rob lem. No exper iments have been done on th i s .

2 . 6 Conclusions concerning sample r s

Wall flow sample£

About 50 t e s t s on the wall flow sampler have been recorded . It is

believed that the unit i s effective. But as the design is empir ica l , the

bas i s for scaling to other pipe d i ame te r s is uncer ta in .

- 24 -

Corse a i r - b o r n e liquid sampler

The sedimentat ion-impaction unit has only a theoret ica l foundation,

and no prac t ica l t e s t s have been made! F r o m experience, sedimenta­

tion and impaction effects can be predicted with some confidence.

Vapour and fine a i r -bo rne liquid sampler

Also this sampler is a theore t ica l design, and only qualitative ex­

pe r imen t s have been ca r r i ed out. F u r t h e r t e s t s a r e advisable .

2 .7 Calculation of vapour and liquid flows from sample flows

To calculate the t r anspo r t of vapour and liquid in the compressed

a i r line from the sample flows, the following formulas apply approxi­

mately:

cp (wall) = q [c (wall) - c (core , fine)]

cp (core , coarse) = q [c (core , coarse) - c (core , fine)]

cp (core , fine) = Q • c (core , fine)

The notation is explained in F igure 2. 14. All concentrat ions and

flows a r e r e f e r r ed to s tandard conditions (Norm = 101325 Pa , 20 C).

The formulas apply to oil as well as wa te r . The concentrat ions c a r e

supposed to be indicated on the in s t rumen t s , chapter 5.

3. SAMPLE TRANSPORT

F r o m the sampling points on the compressed a i r l ine, a mix tu re

of a i r and liquid is conducted to the measur ing in s t rumen t s . In the p r e ­

sent investigation, teflon tubing of 1. 7 mm internal d iameter was used.

P r e s s u r e reduction to a n e a r a tmospher ic level takes place in a needle

valve, protected by a s in tered meta l f i l ter . The flow from the needle

valves is d i rec ted along m o r e teflon tubing to different ins t ruments by

means of ball va lves .

These components for sample t r anspor t de termine the t ime of

response for the measur ing sys tem. The t ime of r e sponse t is given

by:

t = m / m

- 25 -

Sa»v\pl<2

coarse cor«u

J' qp (wall)

4- <}? ( W e , $\ni)

^ Q ~ H

v ^ . -f me c o r e ?lou> QKCI

where

m contaminant m a s s flow (kg/s) 3

Q main s t r e a m a i r flow (m (Norm) /s )

q sample a i r flow (m (Norm) /s )

c (core , fine) smal l par t ic le and vapour m a s s concentration in main s t r e a m and sample s t r eam (kg/m^ (Norm))

F ig 2 . 14 Notation for calculation of conta­minant flows.

m = the amount of liquid accumulated on the in te r io r surfaces of the sampling sys tem

m = m a s s flow of the liquid

It can be assumed that only a minor pa r t of the liquid is t r anspor t ed

as a i r - b o r n e d rop le t s . Due to r ivulet formation, and the amount of liquid

in the s in tered fi l ter and in joints and other c rev ices of the manifold,

i t is difficult to pred ic t how much liquid will accumulate in the t r anspor t

sys tem.

- 26 -

3. 1 Est imat ion of t r anspor t velocity

A crude es t imate of the t r anspor t velocity for low flows of liquid

can be made if one a s sumes a laminar liquid film on the wall of the

sample conduit. The shear s t r e s s from the s t reaming a i r is t r ansmi t ted

through the film to the solid tube wall . The s t r e s s is (ref 17, p 553),

with thin f i lms:

To 4 dx

where

T shear s t r e s s in the film on the wall (N/m )

D conduit d iamete r (m)

p a i r p r e s s u r e (N/m )

x coordinate along conduit axis (m)

The p r e s s u r e drop along the conduit, dp/dx, can be expressed by

the friction factor X

, -2

To ~ 8

where

Y a i r density (kg/m )

u mean a i r velocity ( m / s )

There a r e severa l ways to obtain the friction factor, e .g . the

Blas ius equation

_ 0.3164 A —

(Re(air)) 4

where

u Y D Re = °

T) air viscosi ty (Ns /m )

- 27 -

With the liquid film moving in a l aminar flow over the tube wall ,

the velocity i s zero at the wal l . At the interface a i r - l iqu id the liquid

veloci ty i s l a rge s t .

T

° ' a max T] l i q

and, for a thin film

v = v / 2 mean max '

a is the film th ickness , which depends on the m a s s flow:

m , . = p.. v n Da liq r l i q mean

Eliminating the film th ickness , one gets

T m , • v =V 2 liS

m e a n Pliq \ i q 2 T T D

F r o m this express ion the t r anspor t velocity in sampling l ines can

be calculated. F igu re 3. 1 shows some resu l t s obtained for oi l . Below

1 mg of oil pe r second the oil film is expected to move at a velocity of

l ess than 1 m m / s .

3. 2 Exper iment on t r anspor t velocity

In prac t ica l exper iments on t r anspor t velocity the liquid was o b s e r ­

ved to move about five t imes fas ter than predicted by the l aminar film

theory . As could be seen through the teflon tube wall, this situation

a r i s e s because the liquid is not evenly dis tr ibuted around the tube

per iphery , but r e s t r i c t ed to a nar row rivulet , F igu re 3 .2 . This i s a

phenomenon of the same kind as descr ibed in connection with the s a m p ­

l e r s , p 8. It i s probably not a non-wetting effect of oi l-on-teflon, b e ­

cause the velocity of oil in s teel tubing is also about five t imes higher

than theoret ical ly predic ted . The steel i s easi ly wetted by oi l .

- 28 -

Transport V«io«4Hj (?***/$)

O OX | Liquid -flou; (iwsft/s)

Fig 3. 1 Observed and calculated t r anspor t . velocity for oil in sampling l ines . Oil viscosi ty 0. 1 N s / m , gauge p r e s s u r e 600 kPa, gas flow 100 m l ( N o r m ) / s .

The velocity in tubes is only par t of the answer to the question of

response t ime . The total response t ime during an exper iment with the

tes t loop turned out to be 12 minu tes . In this case the oil was injected

into the sample line at the wall flow collector (m = 0. 5 m g / s ) and ob­

served at the flame ionization detector . The sample t r anspor t c ircui t

included about 2 m of tubing, one s intered me ta l f i l ter , two valves and

the evaporat ion-condensation unit, containing approximately 2 m of

heated s teel tubing.

4. SAMPLE PREPARATION

4. 1 Requirements regarding sample flow to ins t ruments

The sample s t r eam consisting of a mix ture of a i r , water , and oil

i s not immediately suitable for admission to measur ing ins t rument s .

Thus for the determinat ion of water and water vapour it i s f i r s t n e c e s ­

sa ry for the oil to be removed.

It is also n e c e s s a r y to shunt the sample flow of 100 m l / s . The e lec ­

t ronic humidity m e t e r r equ i re s a flow of about 10 m l / s , and the hydro­

carbon analyzer about 1 m l / s . Achievement of this sub-sampling has

- 29 -

m

Fig 3 . 2 Rivulet of oil in a teflon sampling l ine.

entailed some effort. An evaporat ion-condensation technique was even­

tually adopted.

It was found n e c e s s a r y to dilute the samples in o r d e r to come within

the measur ing range of the i n s t rumen t s .

- 30 -

4. 2 Mechanical homogenization of sample flow

Shunting of the sample s t r eam implies the same difficulties as

sampling of the main s t r eam discussed e a r l i e r . In an at tempt to form

a m o r e homogeneous mix ture of a i r and liquid, the sample s t r e a m was

forced through an u l t rasonic nozzle . F igu re 4. 1 shows the exper imen­

tal a r r a n g e m e n t s .

Wall wafer -f lovo wiVk 6«h\pl» eur flow 100 iv\l £ H O M V \ ) A

•¥lovO y a\r- -f lovo . / N / J L Ultra UltraöOAic rvoWlizer

»\icong. OS<5H

l^i^uicl deposited

3 5 liVrefe

• Capillaries -for _. , ,

tvMi*ttr"^

Fig 4. i Exper iment on sample homogenization by means of u l t rason ic nebul izer .

The p r i m a r y size of the u l t rason ic nozzle was below 10 p,m. Although

the total a i r flow was inc reased to 1 l ( N o r m ) / s , the concentration of

a i r - b o r n e liquid in the tank could reach 30 000 m g / m at maximum flow

of liquid in the tes t loop. At such concentrat ions droplet growth through

coalescence is fast, and the enlarged droplets a r e deposited to an appre ­

ciable extent.

- 31 -

A subsample withdrawn from the tank would thus not be r e p r e s e n ­

tat ive of the total liquid flow, escpecial ly if deposition was select ive

with r e spec t to oil o r wa te r . The spray tank was not a good solution to

the sub sampling problem.

4. 3 Sample vaporizat ion

Some exper iments on the vaporizat ion of the liquid have been made

to invest igate the possibi l i ty of subsampling in the gas phase . The con­

ditions for evaporation of wa te r a r e well known, but the oil has m e r i t e d

a special examination.

The medium viscosi ty lubr icant fraction of m i n e r a l oi ls has 25 - 35

carbon atoms per molecu le . The oil, being a mix ture of ve ry many

hydrocarbon compounds, has no defined boiling point. The range of

vapour p r e s s u r e s covered by the oil components i s shown in pr inc ip le

in F igu re 4. 2. When the oil i s heated to 350 - 400 C, mos t of the oil

compounds reach a p r e s s u r e sufficient for fairly rapid evaporation,

even if the liquid does not boi l .

vAxpour pt-ese*A»-Ä (_ P a )

\o

\<T

10"

10'

\o

|0

io .-* i r i i i i i i i I i i i i i i i i i I — n

210 3 l o /r(K-'j

Fig 4 .2 Approximate vapour p r e s s u r e for m e ­dium viscosi ty lubricat ing oi l .

- 32 -

When oil i s heated above 350 - 400 C, some l e s s des i rab le effects

might also occur , namely cracking, polymerizat ion, oxidation and coke

formation.

Hydrocarbon cracking gives smal le r molecules and the p rocess is

used for t ransforming oil to gasoline and gas . The p r o c e s s is thus ad­

vantageous for the purpose of oil evaporation and sample subdivision.

Together with the b reak -up of molecules , the re is also a po lymer i ­

zation p r o c e s s , leading to hydrocarbons of higher viscosi ty . These

semi - so l id substances might be deposited Li the evaporator and mus t

be avoided. The same applies to the solid coke.

Using the cracking as a m e a s u r e of the chemical react ion taking

place on heating, we can express the react ion velocity as

A = A e " k t

o

where

A amount of m a t e r i a l not yet reac ted (kg)

A s tar t ing amount of m a t e r i a l (kg)

The react ion velocity t ime constant k is obtained through the

Arrhen ius equation

k = b e " R T

where

E . heat of activation ( j /mo l )

R gas constant, 8.31 ( j / m o l K)

T absolute t empera tu re (K)

F o r a light pe t ro leum fraction, gas oil , it has been observed (ref 7,

p 3 5 )

E A = 200 k j / m o l

b = 3.22 . 10 1 2 1/s

- 33 -

In o rde r to achieve evaporation of the oil with smal les t possible

cracking, polymerizat ion and coking, the t ime at high t empe ra tu r e

should be as shor t as poss ib le . F igu re 4. 3 shows the calculated con­

vers ion at short retention t i m e s . Under the assumed conditions, a

surface t empe ra tu r e up to 450 C can be allowed in the evapora tor .

Aoo boo eoo TémperexViM-e ( "Cj

Fig 4. 3 Tempera tu re and t ime influence on hydrocarbon cracking

A -kt

o E A

k = b • e

E A = 2 • 105 ( j /mo l )

R = 8 . 3 1 ( J /mol )

T = 273 + temp (°C)

t = t ime (s)

Technical oils vary considerably in composition, and F i g u r e s

4. 2 and 4. 3 a r e only intended to give an indication of the t e m p e r a t u r e s

and t imes involved. But it seems c lear that many oils can be vapor ized

without des t ruct ion.

Inc reased t empe ra tu r e also leads to oxidation of the oil . It has not

been possible to evaluate the impor tance of this factor. Some l a t e r

r e fe rences a r e 1, 4, 18.

- 34 -

4 . 4 Evaporator

As concluded in the previous paragraph , the o i l - w a t e r - a i r mix ture

from the sampling point should be heated as fast as poss ib le . As the

oil par t ly flows on the evaporator wal ls , the wall t empera tu re mus t not

exceed the des i red final a i r t empe ra tu r e very much. Instead the heat

t ransmit t ing a r ea mus t be made l a r g e .

The design chosen is shown in F igu re 4 . 4 . A hel ical coil of s ta in­

l e s s s teel tubing is wound on a cast i ron core and heated in a cyl indr i ­

cal oven. The tubing has an in ternal d iameter of 2. 3 m m and a length

of 2 m. A thermocouple m e a s u r e s the steel tubing t empera tu re close

to the outlet . A second thermocouple in the oven is used for automatic

t empera tu re regulation.

In the working range of in te res t and at 100 ml (Norm) / s , the gas

t empera tu re at the outlet i s c 30 C below wall t e m p e r a t u r e .

To investigate the r i sk of deposit formation in the evaporator , an

exper iment was ca r r i ed out with a s impler evaporator , containing only

0. 5 m of tubing. At an a i r flow of 100 ml (Norm) / s and oil 0 the amount

of injected oil which did not evaporate but dripped out as liquid was ob­

served to be:

Wall t empera tu re -Approx a i r Not evaporated t empera tu re

°C ^C %

305 215 4.5

350 260 3.2

395 305 1.9

Each t empera tu re was maintained for 6 hours . The p r e s s u r e drop

over the unit was observed to be constant at the two lower t e m p e r a t u r e s ,

but at 395 C the p r e s s u r e drop inc reased during the course of the expe­

r iment . This implies that in this case a deposit was formed in the

heating coil.

F o r work with the l a r g e r evaporator , it was decided to use the

wall t empera tu re 325 C, giving an a i r t empera tu re of 295 C. The amount

not evaporated should then be insignificant, and the r i sk of deposits

smal l .

- 35 -

I l i • i I « l t I \

Ccisf tror» core

Cerwmcc -hAb«

Sub-S<*w\ple +©

tWK^evAporated oil

F i g 4 . 4 E v a p o r a t i o n - c o n d e n s a t i o n un i t for s u b -s a m p l i n g of s a m p l e s o b t a i n e d a t c o m ­p r e s s e d a i r l i n e .

4 . 5 C o n d e n s a t i o n un i t

S a m p l e cond i t ion ing for o i l m e a s u r e m e n t

Oil cannot r e m a i n in t h e v a p o u r p h a s e a t the r e l a t i v e l y low t e m p e ­

r a t u r e r e q u i r e d by the f l a m e ion iza t ion d e t e c t o r . When m e a s u r i n g o i l ,

- 36 -

therefore , the t empera tu re was reduced by mixing cool a i r with the hot

sample s t r eam from the evaporator . This caused the oil vapour to con­

dense, forming an ae roso l of very smal l pa r t i c l e s . Such an ae roso l i s

easily sampled. F igu re 4 . 4 shows the form taken by the evaporat ion-

condensation unit in p rac t i ce .

Apart from cooling the effluent gas from the evaporation unit, the

admixed a i r also dilutes the o i l - a i r mix tu re . This i s n e c e s s a r y to get

within the measur ing range of the flame ionization detector .

Sample conditioning for humidity m e a s u r e m e n t

A s t r eam of compressed a i r , free of oil and liquid water , is easily

sampled. If it i s not very dry (dew point above -20 C at a tmospher ic

p r e s s u r e ) , the water vapour content is also readily m e a s u r e d . De te r ­

mination of the humidity in a i r of lower dew point demands considerable

ca re if e r r o r s , due to water diffusing from or through sampling line

wal l s , a r e to be avoided.

If the re is liquid in the compressed a i r , it must be incorpora ted

in the sample . Any oil p resen t may contain adsorbed or emulsified

wa te r . Lubricants for compressed a i r tools often have emulsifying

addi t ives . While it is impor tant for this water to be re leased , oil mus t

at the same t ime be prevented from reaching the humidity m e t e r .

With a t empera tu re of about 100 C in the evaporat ion-condensation

unit, mos t water p resen t will be vaporized, while mos t of the oil will

remain in the liquid phase . The addition of a i r for purposes of cooling

and dilution can be done in the condensation section, thus complying

with the requi rement for keeping the dew point below the highest con­

centrat ion accepted by the humidity m e t e r .

The a i r is then sampled and fi l tered to remove any oil droplets

p resen t . It is essent ia l that the t empe ra tu r e at the fil ter is lower than

at the humidity m e t e r , so as to prevent oil condensation in the m e t e r .

This system does not pe rmi t the m e a s u r e m e n t of humidity in very

dry a i r . The evaporator , the condensation unit and the fi l ter offer l a rge

a r e a s for deposition and desorption of water vapour, and the t ime of

response becomes too l a rge for p rac t ica l pu rposes .

In this case the evaporation-condensation unit has to be by-passed

and the sample fed direct ly to the humidity m e t e r .

- 37 -

5. INSTRUMENTS

5. 1 Investigated methods

It has been one of the a ims of this project to find ins t ruments suitable

for making continuous m e a s u r e m e n t s without r e c o u r s e to manual methods

and labora tory fac i l i t ies . These ins t ruments were required to be por table

and s imple to u s e .

F o r humidity m e a s u r e m e n t s , the re i s a wide choice of i n s t rumen t s ,

which a r e m o r e o v e r in continuous use in indus t ry . An ins t rument of this

c lass was obtained and tes ted .

Adsorption tubes for water vapour, with an indicator for the amount

of water absorbed, were also tes ted . Similar tubes for other vapours

a r e common in occupational safety work.

The s tandard labora tory method for the determinat ion of smal l

quantit ies of oil is the m e a s u r e m e n t of the infrared adsorption in a

liquid sample . It was not thought that this method could be easi ly adap­

ted to pe rmi t continuous m e a s u r e m e n t s on an a i r -bo rne s t r e a m of o i l .

Instead a hydrocarbon analyzer of the flame ionization type was chosen.

This ins t rument r eac t s to the g rav ime t r i c concentration of hydrocarbons

in a i r .

5.2 Humidity m e a s u r e m e n t

The range of in t e re s t is 1 0 - 3 000 m g / m (Norm), but the i n s t r u ­

ment* s m e a s u r e m e n t range should extend to higher humidity, which

may occur when the re is liquid water in the sample . The dew point

m e t e r with a cooled m i r r o r is applicable, but it i s fairly sensi t ive to

contamination.

Instead the preference lay with another type equipped with a sensor

in the form of an aluminium oxide f i lm. The oxide film adsorbs water in

proport ion to the water content of the surrounding a i r . E lec t rodes , on

ei ther side of the film, m e a s u r e film capacitance o r r e s i s t ance , then

being a function of the amount of water adsorbed [ref 10] .

The sensor should be protected from high re la t ive humidity and is

therefore mounted in a chamber operat ing at 50 C. The sample s t r e a m

m u s t be free of oil which, should it be deposited on the sensor , would

block i t s p o r e s . The sample s t r eam should be about 10 m l / s .

- 38 -

It was found n e c e s s a r y to r eca l ib ra te the humidity m e t e r at regular

i n t e rva l s . It is not known whether the unsat isfactory stabili ty was a

feature only of the ins t rument used o r whether it i s cha rac te r i s t i c of

these ins t ruments general ly . F o r calibration purposes a humid-a i r

genera tor was constructed, where sa tura ted a i r was mixed with very

dry a i r in known propor t ions .

The humid-a i r genera tor was also used to invest igate the behaviour

of indicating adsorption tubes for water vapour, manufactured by

Drägerwerk , Ltfbeck, BRD. These proved very handy to use and a r e

specific to water vapour. But the calibration factor depends on the

water vapour concentration, F ig 5. 1. It thus becomes n e c e s s a r y to

u s e different calibration curves for different sample volumes, Fig 5 .2 .

X •

i

^ ^ • •

i> i 1 1 1 1 1 1 1 1J-L

OA

i Vapour GoiXÄKvtmVioK^ Ov

Fig 5. 1 Observed rat io between indicated (<*) and suppled (m v ) amount of water vapour in Drage r indicating absorption tubes . Air flow 35 ml (Norm) / s , gauge p r e s s u r e at tube inlet c 20 kPa .

- 39 -

IndtcoiKonof

oj i 10

F ig 5.2 Calibration curves for water vapour in Drage r indicating absorption tubes , at different volumes of sample (V). Air flow 35 ml ( N o r m ) / s , gauge p r e s s u r e at tube inlet c 20 kPa .

5. 3 Hydrocarbon m e a s u r e m e n t

The flame ionization detector responds to carbon in organic chemi­

cal compounds. It is not sensi t ive to carbon oxides or water vapour as

long as the concentrat ions a r e not so high as to dilute the a i r sample .

Essent ia l ly the ins t rument compr i ses a smal l hydrogen-a i r bu rne r ,

an ion collecting e lect rode around the burner f lame, and a sensi t ive

amplif ier for the ion cur ren t . The cur ren t i s d i rec t ly propor t ional to

the number of organic carbon atoms brought into the flame pe r unit 3

t ime . The measur ing range extends from 1 mg of hydrocarbon per m

of sample to about 1 g / m .

After having reviewed p re sen t theor ies on the operat ion of the f lame

ionization detector it is concluded that cracked and oxidized hydroca r ­

bons should give a response only slightly smal le r than a low-molecular

hydrocarbon.

- 40 -

A f l a m e i o n i z a t i o n d e t e c t o r w a s o b t a i n e d f r o m F O A , S tockho lm

( T h e R e s e a r c h I n s t i t u t e of N a t i o n a l Defence) [ r e f 5] . T h e F O A d e s i g n

h a s the s p e c i a l a d v a n t a g e of p e r m i t t i n g a e r o s o l p a r t i c l e s to e n t e r the

f l a m e . A s w a s m e n t i o n e d e a r l i e r , t h i s i s e s s e n t i a l if h i g h - b o i l i n g h y d r o ­

c a r b o n s a r e to be m e a s u r e d . The i n s t r u m e n t r e q u i r e s a s a m p l e flow of

p r e c i s e l y 1. 17 m l / s , and t h e i n l e t p r e s s u r e shou ld n o t v a r y m o r e than

± 200 P a , gauge p r e s s u r e .

A f l a m e ion i za t i on d e t e c t o r w a s c a l i b r a t e d wi th p r o p a n e to d e t e r ­

m i n e the r e l a t i o n s h i p b e t w e e n ion i za t ion c u r r e n t and c a r b o n c o n c e n ­

t r a t i o n in the s a m p l e . It w a s found tha t

I. = 2 . 5 0 • 1 0 ' 5 c ^ ion C

w h e r e

I. = i on i za t i on c u r r e n t (A) ion v ' 3

c - = c a r b o n c o n c e n t r a t i o n in s a m p l e ( k g / m ( N o r m ) )

When dea l ing wi th s a t u r a t e d h y d r o c a r b o n s of h igh m o l e c u l a r

w e i g h t (C H_ , _, C H., , n = 25 - 35), such a s t h o s e in l u b r i c a t i n g ° x n 2n+2 n 2n ' e

o i l , t h e above c o r r e l a t i o n caji a l s o b e e x p r e s s e d a s

I. = 2 . 14 * 1 0 " 5 c „ „ ion HC

w h e r e in t ha t c a s e

c „ p = h y d r o c a r b o n c o n c e n t r a t i o n in s a m p l e ( k g / m ( N o r m ) )

With the e v a p o r a t i o n - c o n d e n s a t i o n un i t c o n n e c t e d to the f l a m e i o n i ­

za t ion d e t e c t o r a s in F i g u r e 5. 3 , t h e r e l a t i o n b e t w e e n o i l m i s t c o n c e n t r a ­

t ion (CTT~) and i on i za t i on c u r r e n t w a s i n v e s t i g a t e d .

A m e a s u r e d flow of o i l w a s i n j ec t ed in a s a m p l e a i r s t r e a m of 100

m l ( N o r m ) / s . H y d r o c a r b o n c o n c e n t r a t i o n in the c o n d e n s a t i n g c h a m b e r

w a s c a l c u l a t e d a s

_ (inj&cted o i l flow) x (oi l d e n s i t y ) HC ~ ( s a m p l e a i r flow) + (d i lu t ion a i r flow)

- 4 1 -

J Vnoiw £«a.po»-ei.KoiA JAK!V

SI

fc*k«w.sl-

,e*

CWNCI recorder

Fig 5. 3 Connection of the flame ionization de t ec ­tor to the evaporat ion-condensation unit .

The observed ionization cu r ren t i s plotted against hydrocarbon

concentration, F igu re 5 .4 .

The constant k^-, in the express ion

I = K • c ion ^HC HC

has been calculated. The re su l t s a r e as follows:

K HG,

A rn /kg

Manufacturer (Foa)

Calibrat ion with propane 1974-11-25

Calibration with propane 1976-01-27

Calibrat ion with propane 1976-01-29

Measurements on oil N 1976-01-28--29

Measurements on oil O 1976-02-02

2. 14

2. 14

2 .24

2 .44

2. 1

1.9

10"

10*

10'

10*

10*

10 -5

- 42 -

ion

10

ft 3

o.< 30

i v~tf—•

100 300 1000

/c HC, (yw^/^Cwo^y

Fig 5.4 Observed ionization cur ren t as a function of oil concentration in sample flow. Oven wall t empera tu re 325 C. Oven a i r t empera tu re 295°C. Oven a i r flow i00 ml ( N o r m ) / s . Diluting and cooling a i r flow 6 l ( N o r m ) s .

Compared with the mos t recent propane cal ibra t ions , i t s eems

that the flame ionization detector indicates on oil concentration 13 - 20 %

lower than expected.

Unevaporated oil and wall l o s se s in the condensation unit a r e co l lec­

ted at the bottom of the condensation unit . Very l i t t le oil was found h e r e .

L o s s e s in the t r anspor t from the condensation unit to the flame

ionization detector cannot be evaluated, but might have had some in ­

fluence.

There is also the possibil i ty that the flame ionization detector has

a lower sensibili ty to (possibly cracked and oxidized) oil than to propane .

The propane used for cal ibrat ion may have par t ly oxidized in i ts

s teel container, and the calibrat ion of 1974 might be m o r e co r r ec t . In

- 43 -

such case the re i s much be t te r agreement between observed and expec­

ted response to oil m i s t .

F u r t h e r examination of this method of sample prepara t ion will no

doubt d isc lose the reason for observed d i sc repances , but it s eems a l ­

ready c lear that the method can be used .

6. NOTES ON CONTINUED INVESTIGATIONS

Pr inc ip le of m e a s u r e m e n t

Originally this projec t was a imed at the design of a cheap, easi ly

applied ins t rument for oil and water contamination in compres sed a i r

l ines . After having considered a number of poss ible p r inc ip les , it was

decided to employ sampling, with sample t r anspor t to the ins t ruments

p rope r . As i s evidenced by this repor t , this approach involved som

difficulties. These hindrances were surmounted, but i t is sti l l des i rab le

to develop some other pr inciple which does not include sampling.

The s amp le r s tes ted have been designed to provide information not

only on the total flow of contaminants , but also on the mode of t r an spo r t .

If this information is not needed, the wall flow sample r can be excluded,

getting the wall flow in the " coa r se , a i r - b o r n e " sample r .

The s ample r s for a i r - b o r n e m a t e r i a l have not been as thoroughly

tes ted as the wall flow sample r . Here som work remains to be done.

Sample t r anspor t

As pointed out ea r l i e r , the "fine, a i r - b o r n e " sampling chain does

not work p roper ly at low concentrat ions because the collected smal l

amounts of liquid do not move in the sampling l ine. Especial ly at f i l ter

tes t ing, the concentrat ion of liquid downstream of the fil ter is ve ry low.

It i s believed that a heated sampling line would also make f i l ter test ing

possible with the p resen t equipment.

Ins t ruments

The humidity m e t e r employed did not function to sat isfact ion. It i s

advisable to incorpora te in an installat ion some means for cal ibrat ion.

- 44 -

This could be, e. g . , a manual dew-point m e t e r , o r a humid air gene­

r a to r .

The indicating reagent tubes tes ted proved to be ve ry handy for

modera te ly dry a i r . To simplify thei r use , and to extend the range of

m e a s u r e m e n t to very dry a i r , a unit compris ing a p r e s s u r e contain­

ment for the tube should be developed, and also a sample volume m e t e r .

The flame ionization detector functioned sats i factor i ly . But it is

a complicated ins t rument , requir ing a skilled ope ra to r . Its range of

m e a s u r e m e n t does not extend to the ra ther high concentrat ions s o m e ­

t imes encountered in the sample a i r s t r e a m . In cases where highest

sensit ivity i s not n e c e s s a r y and liquid water i s absent , a catalytic

combustion ins t rument (a kind of the rmal analyzer) might be m o r e

sui table .

In conclusion, much work sti l l r ema ins to be done on the subject

of compressed a i r contaminant m e a s u r e m e n t . Specifically, it should

be noted that considerable savings could be made if one o r m o r e con­

taminant l imi ts of Table 1:1 could be re laxed. If, for example, the

liquid phase could be excluded, sampling would be much simplified. Or,

if oil (water) were absent , the m e a s u r e m e n t of wa te r (oil) could probably

be m a d e with s impler equipment. The work should thus be d i rec ted to

m o r e specific conditions and applicat ions, e . g .

t es t of f i l ter efficiency

tes t of d rye r efficiency

- . m e a s u r e m e n t s on workplace a i r

oil entrained from c o m p r e s s o r s

a i r quality a s su rance in the food industry

oil m i s t lubricat ion

- 45 -

7. REFERENCES

Additives Von D Chris takudis u a Hrsg von H P r i n z l e r E W Berghoff. Deutscher Ver l f Grundstoffindustrie VEB, Leipzig 1966.

BAKER O Simultaneous flow of oil and gas . Oil gas J 53 (1954) p 185.

BRAUER H Grundlagen der Emphasen- und Merhphasens t römmungen. Ver lag Sauer länder , Frankfur t am Main 1971.

CIBULA G Die Oxydation bei Schmierölen. Mineralöl tech 11 (1966) p 8.

FROSTLING H and BRANTTE A A Mobile analysis ins t rument for the m e a s u r e m e n t of organic vapours and ae roso l s in a i r . J Phys E Sei i n s t rum 1972 (5) p 251.

GOVIER G W and AZIZ K The flow of complex mix tu re s in p ipes . New York 1972.

GRUSE W A and STEVENS D R Chemical Technology of pe t ro leum. 3rd ed New York I960.

HUHN J and WOLF J Zweiphasenst römmung. Gasförmig/ f lüss ig . Leipzig 1975.

ISHII M Thermo-fluid dynamic theory of two-phase flow. P a r i s 1975.

JASON A C Some p rope r t i e s and l imita t ions of the aluminium oxide hygro ­m e t e r . Humidity and m o i s t u r e . Ed by Wexler, New York 1965 Vol 1 p 372.

LEVIN L M The intake of ae roso l samples . Bull Acad Sei. USSR Geophys Ser 1957:7 p 87.

LEVY S Pred ic t ion of two-phase annular flow with liquid ent ra inment . Int J Heat Mass Transfer 9(1966) p 171.

MERCER T T Aerosol technology in hazard evaluation. New York 1973.

MUNO H Das Transpor tverha l ten von ölnebel in pneumatichen V e r t e i l e r ­sys temen. Mainz 1973.

- 46 -

15. NORMAN W S and McINTYRE V heat t r ans fe r to a liquid film on a ve r t i ca l sur face . Trans Inst Chem Eng 38(1960) p 301.

16. QUANDT E Analysis of gas- l iquid flow pa t t e rns . Chem Eng P r o g r Symp Ser 61(1965)57 p 128.

17. SCHLICHTING H Grenzschicht - Theor ie . Kar l s ruhe 1965.

18. SCOTT G Atmospher ic oxidation and antioxidants. Ams te rdam 1966.

19. SWELL A P How much oil is in your a i r sys tem ? Power 119(1975)10 p 39.

20. MORGAN P G The m e a s u r e m e n t of oil in compressed a i r l ines . Compress a i r Hydraul 27(1962)312 p 98.

2 1 . EVANS A P Sampling and analyzing of high p r e s s u r e gas s t r e a m s for lub r i ­cant content. J Amer Ass Contam Contr 4(1965) July p 10.

22. CGA specification G-7 . 1. Commodity specifications for a i r . Compressed Gas Associat ion, Inc. New York 1973.

23 . HUYGHE J and MINK T Q Ecoulement en tube d* un melange de liquide et de gaz en reg ime annulaire d i spe r se : Mesure du debit de liquide en film a la paro i . C R Acad Sci 260(1965) p 2405.

24. ISO International organization for s tandardizat ion. Compressed non-breathing a i r for use in a i rc raf t . ISO 2434-1973(E).

25. DIN Deutsche Normen . Druckluft (Press luf t ) ftir A temgerä te . DIN 3188. Januar i 1974.

26. SITTEL P Methoden der Ölnebel-Erzeugung und - Vermessung . Diss Mainz 1970.

AE:s kontorstryckeri Nyköping 1976

LIST OF PUBLISHED AE-REPORTS

1—440 (See back cover earlier reports.)

441. Neutron capture gamma ray cross sections for Ta, Ag, In and Au between 30 and 175 keV. By J. Hellström and S. Beshai. 1971. 30 p. Sw. cr. 15:-.

442. Thermodynamical properties of the solidified rare gases. By I. Ebbsjö. 1971. 46 p. Sw. cr. 15:—.

443. Fast neutron radiative capture cross sections for some important standards from 30 keV to 1.5 MeV. By J. Hellström. 1971. 22 p. Sw. cr. 15:-.

444. A Ge (Li) bore hole probe for in situ gamma ray spectrometry. By A. Lau-ber and O. Landström. 1971. 26 p. Sw. cr. 15:-.

445. Neutron inelastic scattering study of liquid argon. By K. Sköld, J. M. Rowe, G. Ostrowski and P. D. Randolph. 1972. 62 p. Sw. cr. 15:-.

446. Personnel dosimetry at Studsvik during 1970. By L. Hedlin and C.-O. Widell. 1972. 8 p. Sw. cr. 15:-.

447. On the action of a rotating magnetic field on a conducting liquid. By E. Dahlberg. 1972. 60 p. Sw. cr. 15:-.

448. Low grade heat from thermal electricity production. Quantity, worth and possible utilisation in Sweden. By J. Christensen. 1972. 102 p. Sw. cr. 15:-.

449. Personnel dosimetry at Studsvik during 1971. By L. Hedlin and C.-O. Widell. 1972. 8 p. Sw. cr. 15:-.

450. Deposition of aerosol particles in electrically charged membrane filters. By L. Ström. 1972. 60 p. Sw. cr. 15:- .

451. Depth distribution studies of carbon in steel surfaces by means of charged particle activation analysis with an account of heat and diffusion effects in the sample. By D. Brune, J. Lorenzen and E. Witalis. 1972. 46 p. Sw.cr. 15:-.

452. Fast neutron elastic scattering experiments. By M. Salarna. 1972. 98 p. Sw. cr. 15:-.

453. Progress report 1971. Nuclear chemistry. 1972. 21 p. Sw. cr. 15:-. 454. Measurement of bone mineral content using radiation sources. An annotated

bibliography. By P. Schmeling. 1972. 64 p. Sw. cr. 15:-. 454. Measurement of bone mineral content using radiation sources. An annotated

bibliography. Suppl. 1. By P. Schmeling. 1974. 26 p. Sw. cr. 20:-. 455. Long-term test of self-powered detectors in HBWR. By M. Brakas, O. Strirt-

dehag and B. Söderlund. 24 p. 1972. Sw. cr. 15:-. 456. Measurement of the effective delayed neutron fraction in three different

FRO-cores. By L. Moberg and J. Kockum. 1972. Sw. cr. 15:-. 457. Applications of magnetohydrodynamics in the metal industry. By T. Robin­

son, J. Braun and S. Linder. 1972. 42 p. Sw cr. 15:-. 458. Accuracy and precision studies of a radiochemical multielement method

for activation analysis in the field of life sciences. By K. Samsahl. 1972. 20 p. Sw. cr. 15:—.

459. Temperature increments from deposits on heat transfer surfaces: the thermal resistivity and thermal conductivity of deposits of magnetite, calcium hydro­xy apatite, humus and copper oxides. By T. Kelén and J. Arvesen. 1972. 68 p. Sw. cr. 15:—.

460. Ionization of a high-pressure gas flow in a longitudinal discharge. By S. Palmgren. 1972. 20 p. Sw. cr. 15:-.

461. The caustic stress corrosion cracking of alloyed steels — an electrochemi­cal study. By L. Dahl, T. Dahlgren and N. Lagmyr. 1972. 43 p. Sw. cr. 15:-.

462. Electrodeposition of "point" Cu"sl roentgen sources. By P. Beronius, B. Johansson and R. Söremark. 1972. 12 p. Sw. cr. 15:-.

463. A twin large-area proportional flow counter for the assay of plutonium in human lungs. By R. C. Sharma, I. Nilsson and L. Lindgren. 1972. 50 p. Sw. cr. 15:-.

464. Measurements and analysis of gamma heating in the R2 core. By R. Carls-son and L. G. Larsson. 1972. 34 p. Sw. cr. 15:-.

465. Determination of oxygen in zircaloy surfaces by means of charged particle activation analysis. By J. Lorenzen and D. Brune. 1972. 18 p. Sw. cr. 15:—.

466. Neutron activation of liquid samples at low temperature in reactors with reference to nuclear chemistry. By D. Brune. 1972. 8 p. Sw. cr. 15:—.

467. Irradiation facilities for coated particle fuel testing in the Studsvik R2 re­actor. By S. Sandklef. 1973. 28 p. Sw. cr. 20:- .

468. Neutron absorber techniques developed in the Studsvik R2 reactor. By R. Bodh and S. Sandklef. 1973. 26 p. Sw. cr. 20:- .

469. A radiochemical machine for the analysis of Cd, Cr, Cu, Mo and Zn. By K. Samsahl, P. O. Wester, G. Blomqvist. 1973. 13 p. Sw. cr. 20:- .

470. Proton pulse radiolysis. By H. C. Christensen, G. Nilsson. T. Reitberger and K.-A. Thuomas. 1973. 26 p. Sw. cr. 20:- .

471. Progress report 1972. Nuclear chemistry. 1973. 28 p. Sw. cr. 20:- .

472. An automatic sampling station for fission gas analysis. By S. Sandklef and P. Svensson. 1973. 52 p. Sw. cr. 20:- .

473. Selective step scanning: a simple means of automating the Philips diffrac-tometer for studies of line profiles and residual stress. By A. Brown and S. A. Lindh. 1973. 38 p. Sw. cr. 20: - .

474. Radiation damage in CaF: and BaFz investigated by the channeling tech­nique. By R. Hellborg and G. Skog. 1973. 38 p. Sw. cr. 20:- .

475. A survey of applied instrument systems for use with light water reactor-containments. By H. Tuxen-Meyer. 1973. 20 p. Sw. cr. 20:- .

476. Excitation functions for charged particle induced reactions in light elements at low projectile energies. By J. Lorenzen and D. Brune. 1973. 154 p. Sw. cr. 20:- .

477. Studies of redox equilibria at elevated temperatures 3. Oxide/oxide and oxide/metal couples of iron, nickel, copper, silver, mercury and antimony in aqueous systems up to 100°C. By Karin Johansson, Kerstin Johnsson and Derek Lewis. 1973. 42 p. Sw. cr. 20:- .

478. Irradiation facilities for LWR fuel testing in the Studsvik R2 reactor. By S. Sandklef and H. Tomani. 1973. 30 p. Sw. cr. 20:- .

479. Systematics in the (p,xn) and (p,pxn) reaction cross sections. By L. Jéki. 1973. 14 p. Sw. cr. 20: - .

480. Axial and transverse momentum balance in subchannel analysis. By S. Z. Rouhani. 1973. 58 p. Sw. cr. 20: - .

481. Neutron inelastic scattering cross sections in the energy range 2 to 4.5 MeV. Measurements and calculations. By M. A. Etemad. 1973. 62 p. Sw. cr. 20: - .

482. Neutron elastic scattering measurements at 7.0 MeV. By M. A. Etemad. 1973. 28 p. Sw. cr. 20: - .

483. Zooplankton in Tvären 1961-1963. By E. Almquist. 1973. 50 p. Sw. cr. 20:- . 484. Neutron radiography at the Studsvik R2-0 reactor. By I. Gustafsson and E.

Sokolowski. 1974. 54 p. Sw. cr. 20:- . ISBN 91-7010-006-3

485. Optical model calculations of fast neutron elastic scatterinq cross sections for some reactor materials. By M. A. Etemad. 1974. 165 p. Sw. cr. 20: - .

486. High cycle fatigue crack growth of two zirconium alloys. By V. S. Rao. 1974. 30 p. Sw. cr. 20:- .

487. Studies of turbulent flow parallel to a rod bundle of triangular array. By B. Kjellström. 1974. 190 p. Sw. cr. 20: - .

488. A critical analysis of the ring expansion test on zircaloy cladding tubes. By K. Pettersson. 1974. 8 p. Sw. cr. 20:- .

489. Bone mineral determinations. Proceedings of the symposium on bone mine­ral determinations held in Stockholm-Studsvik, Sweden, 27-29 may 1974.

Vol. 1. Presented papers. 1974. 170 p. Sw. cr. 20:—. Vol. 2. Presented papers (cont.) and group discussions. 1974. 200 p. Sw. cr. 20:—. Vol. 3. Bibliography on bone morphometry and densitometry in man. By A. Horsman and M. Simpson. 1974. 112 p. Sw. cr. 20:—.

490. The over-power ramp fuel failure phenomenon and its burn-up dependence — need of systematic, relevant and accurate irradiation investigations. -Program proposal. By H. Mogard. 1974. Sw. cr. 20:—.

491. Phonon anharmonicity of germanium in the temperature range 80—880 K. By G. Nelin and G. Nilsson. 1974. 28 p. Sw. cr. 20:- .

492. Harmonic lattice dynamics of germanium. By G. Nelin. 1974. 32 p. Sw. cr. 20:- .

493. Diffusion of hydrogen in the /?-phase of Pd-H studied by small energy transfer neutron scattering. By G. Nelin and K. Sköld. 1974. 28 p. Sw. cr. 20:- .

494. High temperature thermocouple applications in the R2-reactor, Studsvik. By B. Rohne. 1974. 20 p. Sw. cr. 20:- .

495. Estimation of the rate of sensitization in nickel base alloys. By J. Wiberg. 1974. 14 p. Sw. cr. 20:- .

496. A hort-el-complex in Sweden. By J. Christensen. 1974. 82 p. Sw. cr. 20: - . 497. Effect of wall friction and vortex generation on radial void distribution - the

wall-vortex effect. By Z. Rouhani. 1974. 36 p. Sw. cr. 20:—. 498. The deposition kinetics of calcium hydroxy apatite on heat transfer surfaces

at boiling. By T. Kelén and R. Gustafsson. 1974. 30 p. Sw. cr. 20:- . 499. Observations of phases and volume changes during precipitation of hydride

in zirconium alloys. By G. Ostberg, H. Bergqvist, K. Pettersson, R. Attermo, K. Norrgård, L-G. Jansson and K. Malén. 1974. 16 p. Sw. cr. 20:- .

500. X-ray elastic constants for cubic materials. By K. Malén. 1974. 25 p. Sw. cr. 20:- .

501. Electromagnetic screening and skin-current distribution with magnetic and non-magnetic conductors. By E. Dahlberg. 1974. 44 p. Sw. cr. 20:—.

502. Depth distribution studies of carbon, oxygen and nitrogen in metal sur­faces by means of neutron spectrometry. By. J. Lorenzen. 1975. 54 p. Sw. cr. 20:—.

503. A systematic study of neutron inelastic scattering in the energy range 2.0 to 4.5 MeV. By E. Almén-Ramström. 1975. 108 p. Sw. cr. 20:—.

504. Analysis of EPR with large quadrupole interaction. By K.-A. Thuomas. 1975. 28 p. Sw. cr. 20—.

505. Observations on deformation systems in zircaloy-2 deformed at room tem­perature. By K. Pettersson and H. Bergqvist. 1975. 20 p. Sw. cr. 20.—.

506. Study of a tritium-fueled battery utilizing the difference of workfunction between electrodes. By J. Braun. 1975. 20 p. Sw. cr. 20.—.

507. X-ray Characterization of non-equilibrium solid solutions. By A. Brown and O. Rosdahl. 1975. 30 p. Sw. cr. 20:—.

508. Statistic performance of dichromatic scanners for absorptiometric determin­ation of bone mineral content using low energy gamma rays. By E. Dissing. 1975. 12 p. Sw. cr. 20:—.

509. Fast reactor blanket experiments in FRO. By T. L. Andersson, R. Håkans­son, R. Richmond and P. Stevens. 1976. 51 p. Sw. cr. 20.—.

510. Two phase sintering of UOi—20 Ce02. A homogenization model for mixed-oxide fuel pellets. By Allan Brown. 1976. 72 p. Sw. cr. 20:—.

511. Methods for sampling and measurement of compressed air contaminants. By L. Ström. 1976. 46 p. Sw. cr. 20:—.

List of published AES-reports (In Swedish)

1. Analysis by means of gamma spectrometry. By D. Brune. 1961. 10 p. Sw. cr. 6:- .

2. Irradiation changes and neutron atmosphere in reactor pressure vessels-some points of view. By M. Grounes. 1962. 33 p. Sw. cr. 6:—.

3. Study of the elongation limit in mild steel. By G. Östberg and R. Atter­mo. 1963. 17 p. Sw. cr. 6 : - .

4. Technical purchasing in the reactor field. By Erik Jonson. 1963. 64 p. Sw. cr. 8 : - .

5. Agesta nuclear power station. Summary of technical data, descriptions, etc. for the reactor. By B. Lilliehöök. 1964. 336 p. Sw. cr. 15:-.

6. Atom Day 1965. Summary of lectures and discussions. By S. Sandström. 1966. 321 p. Sw. cr. 15:-.

7. Building materials containing radium considered from the radiation pro­tection point of view. By Stig O. W. Bergström and Tor Wahlberg. 1987. 26 p. Sw. cr. 10:- .

8. Uranium market. 1971. 30 p. Sw. cr. 15:-. 9. Radiography day at Studsvik. Tuesday 27 april 1971. Arranged by AB Atom-

energy, IVA's Committee for nondestructive testing and TRC AB. 1971. 102 p. Sw. cr. 15:-.

10. The supply of enriched uranium. By M. Mårtensson. 1972. 53 p. Sw. cr. 15:—. 11. Fire studies of plastic-insulated electric cables, sealing lead-in wires and

switch gear cubicles and floors. 1973. 117 p. Sw. cr. 35:-. 12. Soviet-Swedish symposium on reactor safety problems. Sfudsvik, march 5—7,

1973. P a r t i . Swedish papers. 1973. 109 p. Sw. cr. 20:-Part 2. Soviet papers. 1973. 120 p. Sw. cr. 20:—.

13. International and national organizations within the nuclear energy field. By S. Sandström. 1975. 24 p. Sw. cr. 2 0 — .

14. Energy analysis and power growth patterns. By K. Jirlow. 1975. 44 p. Sw. cr. 20.—.

Additional copies available from the Library of AB Atomenergi, Fack, S-611 01 Nyköping 1, Sweden.

Pogo Print, Stockholm 1975