The Explosibility

of Flour, Gluten
and Wheat Dust

First edition
January 2004
The Explosibility
of Flour, Gluten
and Wheat Dust
Summary
Dust explosions require the presence of an explosible dust, air/oxygen
and a source of ignition. Flour, wheat and related dusts have explosive
qualities - but fall within the category of the least explosible of dusts.
Recent research commissioned by nabim has presented consistent and
logical results for the Kst values (the rate of maximum pressure rise) for
a range of dusts to be found in a flour mill; the research confirms that
flour is low on the level of explosible dusts. Appendix 2 sets out the
three explosion classes and gives examples of each.

The research found Kst values for various flour dusts that ranged from
53-100bar.m/sec; for wheatfeed in the middle of that range; and for
various wheat dusts from 105-122bar.m/sec, However, a sample of gluten
at 6.7% moisture recorded an unexpectedly high value of 149bar.m/sec.
As a result, nabim has encouraged flour millers to obtain accurate data
from their gluten suppliers.

The minimum ignition energy levels recorded in the research gave no

cause for concern, other than that for gluten for which the value was
identified as 30-100mJ. However, even this level does not require millers
to take any additional precautions.

The information contained in this booklet should provide flour millers

with all the data they require to identify the explosibility characteristics
of their raw materials and products in the Explosion Protection
documents that they will be drawing up in line with the requirements of
the Dangerous Substances and Explosive Atmospheres Regulations
(DSEAR). The exception is where millers use uncommon processes; for
example, producing very dry flours. In such instances, they should carry
out their own tests to establish whether there is any additional hazard.
Also, millers may use some ingredient dusts which were not tested in
this research; companies should seek information on the explosibility
of these ingredients from their suppliers.

Responsible milling companies should not find the implications of either

DSEAR or the nabim research too onerous. However, both should
encourage millers to assess risk systematically and to ensure that the
proper management procedures are in place.
Introduction
The fact that flour and other wheat products such as gluten, when
dispersed in air, are capable in some circumstances of giving rise to a
dust explosion has been known for a great number of years. In the
UK, nabim has always been alive to this hazard and has, since the 1960s,
produced several pieces of guidance on the prevention of fire and dust
explosions.

In spite of the long-term awareness of the hazard, until the late 1980s
relatively little had been published on the explosibility properties of flour
and wheat gluten; the data which was in the public arena was generic
and unspecific both in terms of substance and test methodology. It
was generally accepted that flour fell into the least hazardous class (St1)
of explosive dusts but little more was known.

Then within a period of a few years, two programmes of tests aimed at

increasing understanding of the subject were undertaken:

1. Measurements of the maximum pressure and maximum rate of

pressure rise for a number of types of flour (sponsored by nabim);

2. Measurements of the minimum ignition energy of flours, glutens

and collector dusts (sponsored by Rank Hovis Ltd).

Both these series of tests were carried out by Chilworth Technology

Ltd, Southampton, and some results are shown in tables 2 and 3.

In 2001, prompted by the European ATEX Directive, now implemented

in UK law by the Dangerous Substances and Explosive Atmospheres
Regulations (DSEAR), nabim commissioned further research into the
explosive properties of flour, gluten and wheat dusts. This project was
undertaken by Burgoyne Consultants Ltd, Runcorn, and was far more
extensivethanp re viousr esea rch,loo ki
nga t–i nit
ially–twe lved iffer
ent
sample dusts.

The results of the tests are contained in table 1 and summarised along
with their significance in the following paragraphs. (The full Burgoyne
Consultants report is available to nabim member companies free of
charge on request. Other organisations wishing to purchase a copy
(price £500) should contact nabim.)
The dusts and the tests
nabim wished its new research project to cover the whole range of
wheat-related dusts that might be found in the mill. The costs of testing
for explosive characteristics is high; nabim therefore sought to ensure
that the range of samples chosen provided a complete picture of the
milling process. Ten samples were chosen:

1. a high protein white breadmaking flour (no added gluten);

2. a Chorleywood breadmaking flour;
3. a biscuit flour;
4. a heat-treated cake flour;
5. wholemeal;
6. gluten;
7. (mill head end) dust collector stock;
8. wheatfeed;
9. wheat dust;
10. screenroom filter dust.

All the samples were tested for maximum pressure (Pmax), maximum rate
of pressure rise (Kst) and minimum ignition energy (MIE).

In order to make the research as relevant as possible to the situation

foun dinami ll
,s amp l
e5wa st ested‘ asr eceived’fromt hemi llwh ils
t
sampl es1a n
d8we retestedb oth‘ asr eceived’a n
dh a
vingbe ens ie
v ed
in accordance with standard scientific protocol. This meant that twelve
sample dusts were tested. Following completion of the project, nabim
commissioned further tests on a thirteenth sample, a 90:10 mixture of
samples 1 and 6.

Other tests
Some of the samples were also tested for other explosive characteristics.
Samples 1 and 8 (both as received and sieved) were tested for both
minimum ignition temperature (MIT) and layer ignition temperature (LIT).
Samples 5 (as received), 6, 9 and 10 were tested for MIT. Sample 1 (as
received) was tested for minimum explosive concentration.
Preparation of the samples
All the samples were collected through the generous assistance of Rank
Hovis Ltd. Each sample was tested for moisture and protein content
before being coded and despatched, in sealed containers, to Burgoyne
Consultants, who again tested each sample for moisture and protein.
Once the tests had been completed, the remainder of the samples were
returned to Rank Hovis for repeat testing to ensure that there could
have been no possibility of samples having become misidentified during
the research process. (In the event, nabim had reservations about the
particle sizes, as calculated by Burgoyne Consultants, using mechanical
sieving. The check testing carried out by Rank Hovis, using the
Malvern apparatus, revealed nothing untoward about the samples.)

The Test Results

Maximum rate of pressure rise
The Kst values for the various flour samples were generally lower than
previously published research and confirmed that values substantially
above 100 are unlikely. Kst values for wheat dust and all other dusts
were a little higher, though still well within the St1 explosion class.

For the various flours, the range was between 53bar.m/sec for wholemeal
(sample as received) and 100bar.m/sec for heat-treated cake flour (6.6%
moisture). Wheatfeed recorded Kst values of 79 (as received) and 69
(sieved according to standard testing protocol). Samples with higher
values were dust collector stock (105); wheat dust (112); screenroom
filter dust (122); and gluten at 6.7% moisture (149).

Minimum Ignition Energy

All samples with Kst values of under 100 recorded MIEs in excess of
1000mJ. Samples with Kst vales between 100 and 122 recorded MIEs
of between 300 and 1000mJ.

The MIE for gluten was recorded at 30-100mJ because one sample, of
1200mg, was ignited at the tenth attempt at 100mJ. (Samples of 600mg
and 900mg did not ignite in ten attempts.) Previous research findings
have been highly variable but Cerestar quote 120mJ.
Gluten
nabim raised its concerns over the Kst and MIE results for gluten with
Bu rgoy neCon sultant
s,wh os tat
edt hat“ t
heo nlyc o nclus
ion( we )wo uld
draw is that the actual MIE ... may lie closer to 100mJ than 30mJ at
which no ignitions occurred. However, in practice this is not significant
as European guidance on electrostatic hazards does not require any
additional control measures for materials with an MIE of 30mJ over
ma terialswi thanMI Eof100mJ .
”Bu rgoynesa lsos t
atedth at“theg luten
sample tested is more sensitive (than the other samples tested) to
electrostatic sources of ignition but not sufficiently so that additional
preca uti
on sa rer equiredt opr e
cludea ne lectros t
atici gnit
ionr is
k ”.

Burgoynes saw no practical benefit in conducting further research into

either the Kst or MIE of gluten. nabim agreed but considered that further
MIE and Kst tests should be commissioned on a mix of 90% high
protein white breadmaking flour and 10% gluten, both substances being
drawn from the original test samples. This new mixed sample had a
Kst value of 84 bar.msec; a Pmax of 7.3bar; and a MIE of >1000mJ, so
confirming that any likely high protein flour with added gluten would
have explosibility characteristics far closer to flour than to gluten.

Maximum Pressure
Readings range from 6.6bar (wheatfeed) to 8.8bar (wheat and screenroom
filter dust); similar to previous findings although a little higher for the
wheat/filter dusts.

Minimum Ignition Temperature

Lowest amongst any samples was 430C (flour). In line with previous
research.

Layer Ignition Temperature

None of the four samples tested gave a LIT of below 450C. This data
is new and, though needed, gives little cause for concern.

Minimum Explosible Concentration

Recorded for white breadmaking flour as 50g/m3.“ Th i
sr esultc onfir
ms
that a flammable dust cloud will be very dense and is unlikely to occur
int heg ene
ralwo rkplace .
”
 Implications of the Test Results
The risk of explosion from wheat dust and related components of wheat
has been understood by the flour milling industry for many years. The
evidence presented by the latest research project has strengthened this
k nowl edge.Ane ncou ragi
n ga s
pe c
toft hepr oject’sc onclusionsist hat
although there are measurable risks, they can be controlled and the
likelihood of an explosion is low.

With MIEs in excess of 1000mJ for most of the tested samples, no special
arrangements are required other than the earthing of metal containers.
There are no further requirements given the MIEs of 300-1000mJ for
wheat dusts and filter stock. The finding of a possible MIE for gluten
below 100mJ is of concern but still does not require further precautions
to be taken. However, increased vigilance of gluten plant is advised.

The Kst values confirm that flour and wheat dusts generally lie around
the halfway mark in explosion class St1. There appears to be a
correlation between Kst value and moisture content / particle size.

The MIT and LIT values are used in the selection of equipment. The
maximum thickness of powder deposits on the surface of equipment
should not exceed 5mm. The temperature of the surface: should not
exceed two-thirds of the MIT value; and should be at least 75C below
the LIT value. The values are used in the selection of the T1 - T6
rating of equipment.

A minimum explosion concentration of 50g/m3 confirms that a high

density dust cloud is required for there to be a significant risk of a dust
explosion.

The figures for maximum pressure rise are of most concern to silo and
milling equipment manufacturers; millers should ensure that their
suppliers are using the correct HSE-approved formulae to calculate the
strength of equipment needed.

sr
Thus, nabim’ e
sea
rchde
mon
str
ate
sth
att
her
eisn
otah
ighr
iskof
explosion with wheat dusts, provided there is the proper management
of what risk there is. The potential for a dust explosion is present in
all mills and, given the appropriate circumstances, can and will occur.
Millers should therefore be vigilant to the problems and consider what
is required to prevent dust explosions.

The priorities in turn should be:

1. avoidance of dust clouds;

2. elimination of ignition sources;
3. containment;
4. suppression;
5. venting.

Greatest care should be given to those areas or processes with a higher

measurable risk, i.e. drying, filter stocks, bin filling. Good housekeeping
within a milling plant are essential, not only for the control of infestation
but for the reduction of explosion risk.

Elimination of ignition sources should include consideration of:

- mechanical friction;
- hot work (e.g. welding);
- grinding;
- electrostatic discharge;
- smoking;
- lighting and electrical equipment;
- spontaneous combustion;
- risk of fire.

Training and the provision of information and instruction to all staff

are valuable components of an explosion prevention strategy, preventing
ignorance from contributing to an incident.

Finally, adherence to the requirement of DSEAR will concentrate activity

on the systematic assessment of risk and the provision of safe
procedures for milling plant.
 APPENDICES
1. Glossary
2. Explosibility Classification
3. 2002 Test Results
4. Previous nabim / Rank Hovis Research

Appe
ndi
x1-‘
Glo
ssa
ry’

Kst and Pmax:

measuring the maximum rate of pressure rise should a dust explosion
occur, the Kst value is determined in a 20litre sphere, in accordance with
ISO 6184/1-1985 (BS 6713 Part 1). The maximum explosion pressure
(Pmax) is also determined as an integral part of this test.

Minimum Ignition Energy (MIE):

is the lowest energy value of a high-voltage capacitor discharge required
to ignite the dust cloud at atmospheric pressure and room temperature.
The MIE is determined by testing in the MIKE 3 apparatus, expected
to be the basis of the European standard.

Minimum Ignition Temperature (MIT):

the minimum temperature at which a dust cloud will explode. The MIT
is determined using a Godbert-Greenwald furnace; testing is carried out
ina cco r
dancewi thBSEN 5 02 81-
2 -
1:1 999.No te–Eu ropea ns t
a ndards
require that surface temperatures of electrical equipment should not
exceed two-thirds of the MIT of the dust cloud which might occur.

Minimum Layer Ignition Temperature (LIT):

the minimum temperature at which a dust layer will ignite when exposed
to a heated surface, determined using a 200mm diameter hotplate and
5mm thick dust layer. Testing is carried out in accordance with BS EN
50281- 2-1:1999. Not e–Eu rope ans t
a ndardsr equirethatthes ur
fa c
e
temperature of electrical equipment should not exceed 75K (=75C) below
the LIT.
 Appendix 2 - Explosibility Classification

Explosion severity can be characterised by the maximum

explosion pressure generated in a closed vessel and the maximum
rate at which the pressure increases in the explosion.

Dusts can thus be classified by their explosibility, as defined by

their Kst value, i.e. the maximum rate of pressure rise in the
event of an explosion. The Kst value is measured in bar.m/s.

The most explosive dusts are classified as St3, applied to those

with Kst values in excess of 300bar.m/s; examples are typically
metals, aluminium for example.

The next classification is St2, applied to dusts with Kst values

ofbe t
we en201a nd300ba r.m/ s
; ma ny‘ma n-ma de ’pr
oducts
fall within this classification, as do some coal dusts.

Classification St1 is applied to dusts with Kst values of between

1 and 200 bar.m/s. Examples range from coal dust (c150-220)
through wheat, flour and other cereals and cereal products (50-
100) to dusts which are barely explosible. Most materials
handle d‘ inbul k ’fa l
lint otheSt 1c lass.

For a dust to be classified as St0, it must have a Kst value of

0 bar.m/s; in other words, it is not explosible.
 Appendix 3 - Summary of Testing carried out for European Fl

Test Material Moisture Protein Protein Ash Particle Size * MIE Pmax
Median (tested)
%w/w %as is% d.b.% d . b . μm Mj bar

HPwhi
tebr
eadm’
g 13.4 12.7 14.8 0.58 (<63) >1000 6.9

HPwhi tebr eadm’ g 13.4 12.7 14.8 0.58 >63 >1000 6.8
(tested as received) as received

HPwhi tebr
eadm’ g 12.1 - - - 63-90 >1000 7.3
(with 10% gluten)

Chorleywood flour 13.6 9.8 11.4 0.70 >63 (<90) >1000 7.2

Biscuit flour 13.1 9.5 11.0 0.73 >90 (<90) >1000 7.5

Heat-treated cake flour 6.6 8.6 9.3 0.60 >63 (<90) 300-1000 7.8

Wholemeal 12.9 14.2 16.4 - >63 >1000 6.8

(tested as received) as received

Gluten 6.7 71.9 77.5 >38 (<63) 30-100 7.3

Collector stock 12.9 >90 (<125) 300-1000 7.7

Wheatfeed 13.5 (<250) >1000 7.0

Wheatfeed 13.5 >250 >1000 6.6

(tested as received) as received

Wheat dust 11.5 >63 (<38) 300-1000 8.8

Screenroom filter dust 11.5 >38 (<38) 300-1000 8.8

* The particle size data from Burgoyne Consultants, quoted above, were obtained using mechanical sieving.
Data obtained in the industry for these typical products, using Malvern apparatus, may differ from the above.
ting carried out for European Flour Explosibility Project

Pmax Kst Explosion MIT LIT MEC Combustion

Class Rating
b a r bar.m/s C C g/m3

6.9 65 St1 430 >450

6.8 68 St1 430 >450 50 1

7.3 84 St1

7.2 66 St1

7.5 91 St1

7.8 100 St1 1

6.8 53 St1 430

7.3 149 St1 470

7.7 105 St1

7.0 69 St1 470 >450

6.6 79 St1 510 >450 1

8.8 112 St1 470

8.8 122 St1 450

btained using mechanical sieving.

paratus, may differ from the above.
Appe
ndi
x4–Pr
evi
ousnabi
mandRankHovi
sRe
sear
chDat
a
(undertaken by Chilworth Technology Ltd c1988-1991)

Type of Moisture Protein Median Maximum Kst Minimum

flour Content Content Particle Pressure Ignition
(14%m.c.) Size Energy

nabim
Bakers 13.5 12.3 64 6.6 95 -
Bakers 8.4 12.2 61 6.9 101 -
Bakers 4.3 12.3 62 6.7 112 -
Chorleywood 14.7 10.6 71 7.1 80 -
Chorleywood 7.9 10.5 66 7.0 74 -
Chorleywood 3.6 10.4 72 6.7 107 -
Biscuit 12.6 8.8 46 7.1 108 -
Cracker 13.7 10.3 81 6.3 96 -
Filter 13.7 15.9 21 6.4 108 -
Cake
(heat-treated) 4.1 8.9 39 6.6 118 -
Wheat dust 11.8 9.6 45 6.4 89 -

Rank Hovis
Chorleywood 7.8 - - - - >500
Dried 1.5 - - - - 250-280
Gluten 7.3 - - - - 400-480
Gluten 8.3 - - - - 450-500
Pneumatic
collector dust 5.3 - - - - 200-210
Collector dust 5.5 - - - - 400-450
Produced by nabim