Propellants, Explosives, Pyrotechnics 26, 69 Ä… 74 (2001) 69
Some Factors InŻuencing Toxic Fume Generation by NG-based
Semigel Explosives in Laboratory Studies
M. M. Bhattacharyya*, P. K. Singh, P. Ram, and R. K. Paul
Central Mining Research Institute, Barwa Road, Dhanbad Ä… 826 001 (India)
Summary generally used with NG-based explosives also inŻuence
the toxic fume production of such explosives(1,13,14).
In a continued study on the generation of toxic fumes by commercial
Because ammonium nitrate (AN), an oxygen positive
explosives the inŻuence of some variations in a nitroglycerine (NG)
major ingredient is used as a solid in manufacturing NG-
based semigel explosive formulation was investigated in a cylindrical
based semigel explosives, its grain size and hence the
laboratory fume study chamber with a capacity of 6 m3. The inŻuences of
varying quantities of wax coated cartridge wrapper, the grain sizes of reactivity appear to inŻuence the generation of toxic fumes
ammonium nitrate and the sampling time interval on the generation
of such explosives(8).
of oxides of nitrogen (NOx) and carbon monoxide (CO) in the deto-
A detailed study at the Central Mining Research Institute,
nation products were studied by analyzing a large number of sample
India, on the generation of toxic fumes of a single component
data. A graphical analysis of the resulting fume data led to interesting
®ndings, indicating the possibility of designing NG-based semigel secondary explosive and a composite NG-based=slurry type
explosives with a better fume quality through the selection of coarser
explosive under laboratory and ®eld conditions lead to the
grains of AN rather than ®ner grains as ingredient and smaller quan-
development of an ideal laboratory method, where the
tities of waxed wrapping paper as cartridge. A further reduction of the
explosive was ®red in the maximum con®nement of a steel
total toxic fume level in the post blast atmosphere may be achieved
with lapse of time.
cannon bore with clay plug stemming, producing least toxic
fumes(15).
This paper presents a part of a continued study high-
lighting the inŻuence of varying wrapper contents, the grain
1. Introduction
size of AN and the length of the time interval of sampling of
toxic fumes of an NG-based semigel explosive and seven
Commercial explosives produce physiologically harmful
more formulations modi®ed with respect to the grain size
gases, generally called as fumes, in measurable quantities
of AN. The results indicate the scope for the design of
under the conditions of blasting in spite of taking necessary
explosives with a better fume quality.
precautions(1). The minimum of constituents in such compo-
site explosives are an oxidizer and a fuel, made up of carbon,
hydrogen, oxygen and nitrogen. Care should be taken to
2. Experimental Details
make the explosives oxygen balanced in order to achieve a
better fume quality, particularly when they are meant for use
2.1 Laboratory Setup
in underground blasting(2,3). The major toxic components of
the post blast gases are carbon monoxide (CO) and oxides of
A cylindrical steel fume chamber with a capacity of
nitrogen (NOx) under laboratory and ®eld conditions(4 Ä… 7).
6m3 constitutes the laboratory setup. It has a circular
Extensive laboratory studies have been undertaken by
hinged door, a close circuit circulatory blower, a fan with
researchers in many countries to understand the phenomenon
a motor at one end and a cannon (1 m long with an axial
of toxic fume generation by various explosives and classi®-
bore of 38 mm in dia. and 75 cm long) on the bottom for
cation or acceptance criteria are developped with respect to
®ring the explosive under the desired con®nement. It is
the toxic fume content in the detonation products(8 Ä… 12).
designed to contain the detonation fumes and to mix them
The amount of toxic fume generation depends on the
uniformly before sampling. The chamber has ten surface
loading density, the mode of initiation, the detonation
holes to take samples for an instrumental analysis. A
velocity, the con®nement, whether detonation gases
schematic diagram of the experimental arrangement is
expand adiabatically and reversibly, freely or against any
shown in Figure 1. An infrared gas analyzer of The
burden and on physico-chemical parameters of the sur-
Foxboro Company, USA, a single beam spectrophotometer
rounding medium(8). The composition of the explosive
of the Electronics Corporation of India Ltd, Hyderabad,
and the quantity of the wax coated wrapping material
India and a Graham Lawrence Apparatus of Meghna
Industries Ltd., Calcutta, India were employed for the
* Corresponding author; e-mail: [e-mail] instrumental analysis of the post detonation gas samples.
# WILEY-VCH Verlag GmbH, D-69469 Weinheim, 2001 0721-3115/01/0204 Ä… 0069 $17.50‡:50=0
70 M. M. Bhattacharyya, P. K. Singh, P. Ram, and R. K. Paul Propellants, Explosives, Pyrotechnics 26, 69 Ä… 74 (2001)
Figure 1. Experimental arrangements in the laboratory fume chamber.
2.2 Explosives Studied plate. Dry and wet bulb thermometer readings inside the
fume chamber were noted before and after the experiment.
An NG-based semigel explosive (sample A) of ICI India Barometric pressure reading was also noted. From these
was studied in the cannon con®ned by a clay plug and under thermometer readings relative humidity (RH) prevailing
prevailing relative humidity in the fume chamber after inside the chamber was computed. About eight to ten gas
varying the quantity of the wax coated manila paper of the samples taken in stoppered glass tubes after 5 minutes of
cartridge. Then two formulations of the same explosive mixing were analysed for each of NOx and CO instrumen-
composition (®rst batch) wrapped with usual manila paper tally. In the case of the modi®ed explosive samples B to H a
but modi®ed with respect to the grain size of AN, sample B 50 g portion of a cartridge with its usual wrapper was ®red
containing coarser grains and sample C with ®ner grains, under similar conditions and the fume samples were taken at
were studied under similar conditions. Later ®ve more two time intervals 5 and 35 min after ®ring and analysed.
modi®ed formulations (second batch) of the semigel NG- Total toxic fumes produced in each experiment were com-
based explosive were studied under similar con®nement puted from the average values of NOx and CO at 25 C and
conditions in the laboratory chamber. The ®ve samples D 0.101 MPa in litres per kg of the explosive by the equation:
to H, contained varying percentages of coarse grains (mix-
Y ˆ xCO ‡ kxNOxÄ…
tures of 85% of 1676 mm and 15% of 500 mm) and ®ne grains
(mixtures of 85% of 500 mm and 15% of 251=152 mm) of AN. where x CO and x NOx are the average volumes of carbon
Freshly made cartridges of 32 mm diameter from carefully monoxide and oxide of nitrogen in litres per kg respectively,
controlled batches were used in these studies. Details of the k is the toxicity factor and Y is total fumes in litres per kg of
explosives are given in Table 1 below. the explosive under study. In the present study the value of k
has been taken as 6.5 to calculate the total toxic fumes as
suggested earlier by Rossi and Usachev(9). The arithmetic
2.3 Experimental Procedure and Data Analysis mean value of a set of total fume data generated from a
number of experiments under similar conditions at a parti-
A number 6 strength copper coated steel tube instanta- cular time was calculated.
neous electric detonator was used to ®re the charge. The With the explosive sample A ®ve sets of experiments were
resulting post detonation gases were uniformly mixed with conducted under cannon con®nement with a 2.5 cm clay plug
the help of the circulatory blowers and the fan on the end by ®ring 50 g of the explosive without any wrapper and then
Table 1. Explosives with Varying Grain Sizes of Ammonium Nitrate
Code number of explosive Major ingredients Type of explosive
Sample A low freeze NG, NC, NaCl, AN P5 semigel
Sample B low freeze NG, NC, NaCl, coarser grains of AN P5 semigel
Sample C low freeze NG, NC, NaCl, ®ner grains of AN P5 semigel
Sample D low freeze NG, NC, NaCl, 100% coarse AN P5 semigel
Sample E low freeze NG, NC, NaCl, 75% coarse with 25% ®ne AN P5 semigel
Sample F low freeze NG, NC, NaCl, 50% coarse with 50% ®ne AN P5 semigel
Sample G low freeze NG, NC, NaCl, 25% coarse with 75% ®ne AN P5 semigel
Sample H low freeze NG, NC, NaCl, 100% ®ne AN P5 semigel
Propellants, Explosives, Pyrotechnics 26, 69ą74 (2001) Factors InŻuencing Toxic Fume Generation 71
Table 2. Summary of the Total Fume Data from the First Batch of Modi®ed Explosives
Code number of explosive Total fumes in l=kg Mean total fumes in l=kg Standard deviation
Sample B 79.84, 51.71, 64.04, 57.94, 55.55, 85.73, 72.36, 71.01, 71.12 9.64
69.37, 83.11, 73.76, 72.31, 72.58, 72.98, 75.64, 80.03
Sample C 74.82, 168.13, 173.34, 69.56, 153.63, 121.89, 116.79, 108.47 37.88
89.49, 88.44, 91.66, 81.70, 72.12
50 g of the explosive with varying quantities (1, 2, 3 and 4 g) for ®ve modi®ed samples on total toxic fume generation after
of wrapper adhered to the charge. Average CO values were a time interval of 5 min. The illustrations in Figures 7 Ä… 9 refer
calculated for each experiment with a particular quantity of to the results with the same samples after two different time
wrapper used and the arithmetic mean values for each set intervals.
of experiments were then computed.
4. Discussion
3. Experimetal Results
Cannon con®nement with 2.5 cm clay plug stemming was
The results of the fume studies with the explosive samples established as the most favourable experimental condition in
A, B, C, D, E, F, G and H to assess the inŻuence of various a previous study for any composite type of commercial
factors on the toxic fume generation have been illustrated explosive to undergo detonation ideally generating least
graphically in the Figures 2 Ä… 9. Average values of NOx and toxic fumes(15). The present studies were carried out under
CO and arithmetic means of the total fume of about ten this particular set of experimental conditions. The experi-
experiments conducted under similar con®nement condition ments were carried out under the prevailing relative humidity
have been depicted by histograms. condition.
The inŻuence of the various amounts of waxed wrapper Graphical plottings of the arithmetic mean values of the
content on the CO generation of the explosive sample A total fume data against the varying quantities of waxed
under the cannon con®nement has been depicted in Figure 2. manila paper wrapping material of the explosive sample A,
A summary of the total fume data from ®rst batch of the as illustrated in Figure 2, show that there was an increase in
modi®ed explosives B and C containing more of coarser CO generation with the increase of the waxed paper wrap-
grains and more of ®ner grains of AN respectively is given in ping the NG-based semigel explosive when detonated under
Table 2. The results of the studies at two time intervals with the con®nement of the cannon bore with clay plug stemming.
these two modi®ed samples are illustrated by histograms in Initially two samples of the NG-based semigel explosive
Figures 4 Ä… 6 for average NOx, CO and total fumes, respec- modi®ed with respect to the grain size of ammonium nitrate
tively. Figure 3 shows the inÅ»uence of varying ®neness of AN the major ingredient of the formulation were studied. It is
Figure 2. InŻuence of waxed wrapper on CO generation.
72 M. M. Bhattacharyya, P. K. Singh, P. Ram, and R. K. Paul Propellants, Explosives, Pyrotechnics 26, 69 Ä… 74 (2001)
Figure 3. Variation of mean total toxic fumes with ®neness of ammonium nitrate (5 min).
the formation of more oxides of nitrogen contributing to the
increasing total toxic fume generation(3).
Thus the study of the two modi®ed samples of a NG-
based semigel explosive composition prompted the chief
investigator to continue the studies with ®ve more samples
of the same explosive composition modi®ed in terms of
its content of varying grain sizes of AN. The samples D,
E, F, G and H were made under carefully controlled
manufacturing conditions having mixtures with varying
percentages from 0 to 100% of coarse or ®ne grains of
AN, respectively (Table 1).
As the arithmetic means of the total toxic fume data
Figure 4. InÅ»uence of time interval on NOx generation by modi®ed
obtained from ten or more experiments conducted with
explosives (®rst batch).
each of the ®ve explosives in cannon con®nement with
stemming are plotted against the increasing ®neness of the
AN grains, the resulting graph is a linear curve having a high
index of determination (Fig. 3). The graph indicates an
increasing generation of toxic fumes by the modi®ed explo-
sives with their AN content of increasing ®neness (i.e., from
100% coarse ‡ 0% ®ne to 0% coarse ‡ 100% ®ne in steps
of 25% variation) under the same experimental conditions.
Thus the results corroborates the earlier ®ndings observed
with the two modi®ed samples (1st batch), one containing
more of coarse (sample B) and the other containing more of
®ne grains of AN (sample C).
In an attempt to study the effect of the length of the time
interval on the sampling two sets of samples were taken for
the determination of NOx and CO after 5 and 35 min of ®ring.
Figure 5. InÅ»uence of time interval on CO generation by modi®ed
For all the seven modi®ed explosives similar sets of data have
explosives (®rst batch).
been generated under the prevailing relative humidity con-
dition of the chamber. Comparative pictures of the average
values of NOx and CO determinations for the samples B and
observed that the sample B containing more of the coarser
C illustrated in Figures 4 and 5, respectively reveal that there
grains of AN generated less total fumes than the sample C
is a decrease of NOx and CO values with time. The arithmetic
made of more of the ®ner grains of AN (Table 2).
mean values of their corresponding total fume data also
This interesting behaviour of the coarser and ®ner grains
showed a decreasing trend accordingly (Fig. 6).
of AN could be explained by the fact that the ®ner grains of
The reduction in the NOx values with time inside the
AN would react faster making more oxygen available in the
chamber can be well explained by the presence of water
con®nement of the apparatus. The availability of this excess
vapour in the vicinity of the water soluble NO2 molecules and
oxygen in the extremely short reaction time(8) would lead to
also by the gradual dissolution of NO2. NO converted to NO2
Propellants, Explosives, Pyrotechnics 26, 69ą74 (2001) Factors InŻuencing Toxic Fume Generation 73
Figure 6. Comparison of mean total toxic fumes generated by mod-
Figure 9. Comparison of mean total toxic fumes generated by mod-
i®ed explosives (®rst batch) at different time interval.
i®ed explosives (second batch) at different time interval.
down in the larger volume of the fume chamber. Cooling salt
ions of the permitted explosive play also a role in the process.
On the other hand in the case of the second batch of ®ve
modi®ed samples the average values of CO showed a mixed
trend of increasing with three samples (E, F and G) and
decreasing in the other two cases with time (Fig. 8). The
increase of the CO values with time appears more probable
because of the possibility that solid products of the detona-
tion undergo secondary reactions resulting in the formation
of more CO at the prevailing temperature inside the chamber.
The average NOx values of all ®ve samples exhibited similar
trends of reduction with time (Fig. 7), as in the case of the ®rst
Figure 7. InÅ»uence of time interval on NOx generation by modi®ed batch of the two samples B and C. The arithmetic mean
explosives (second batch).
values of the total fume data computed from the average NOx
and CO values of all ®ve samples as illustrated in Figure 9,
however, showed a reducing trend with time.
5. Conclusion
NG-based semigel explosives generate less quantities of
toxic fumes when coarser grains of ammonium nitrate are
used as ingredient and detonate under ideal condition of
con®nement. A further reduction in the total fume level may
be achieved with increasing time after detonation upto a
certain period limited to the experimental conditions adopted
in this study. Moreover, with the increase of the wax coated
wrapping paper content of the cartridge the production of
Figure 8. InÅ»uence of time interval on CO generation by modi®ed
carbon monoxide increases. Therefore, a NG-based semigel
explosives (second batch).
explosive with better fume quality can be designed with less
content of waxed paper.
due to the reaction of NO with the excess oxygen present in Thus blasting in underground mines with NG-based
the con®ned space: semigel explosives can be made safer through designing
such explosives with better fume quality. Waiting time for
2 NO‡ O2 ˆ 2 NO2
mining personnel may be judiciously adopted taking advan-
tage of the reducing total toxic fume level at the blasted face
The decrease in the CO values with time (Fig. 5), however,
usually with high moisture content.
cannot be explained with certainty. Although CO2 is water
soluble and its gradual dissolution in water vapour and
simultaneous conversion of CO to CO2 could lead to such a
6. References
decrease theoretically, but the latter reaction, which is
favoured around 500 C, might not take place at the prevailing
(1) B. D. Rossi, ``The Present State of the Study of Noxious Gases
much lower temperature 5 min after the ®ring of the explosive.
in Blasting Operations and the Control of Such Gases'', in:
The products of detonation undergo an expansion and cool B. D. Rossi (ed), ``Control of Noxious Gases in Blasting Work
74 M. M. Bhattacharyya, P. K. Singh, P. Ram, and R. K. Paul Propellants, Explosives, Pyrotechnics 26, 69 Ä… 74 (2001)
and New Methods for Testing Industrial Explosives''. Translated Testing Industrial Explosives''. Translated from Russian by the
from Russian by the Israel Program for Scienti®c Translation Israel Program for Scienti®c Translation Ltd., Jerusalem, for
Ltd., Jerusalem, for National Science Foundation, Washington National Science Foundation, Washington D.C., 1971, pp. 57 Ä… 59.
D.C., 1971, pp. 5 Ä… 7. (10) E. Eitz and R. Zimmermann, ``Die in der Bundesrepublik
(2) S. H. Davidson and I. O. Lewis, ``British Commercial Explo- Deutschland benutzte Methode zur Bestimmung der toxischen
sives'', in: K. H. Fraenkle (ed), ``Manual on Rock Blasting, Bestandteile der Sprengschwaden'', Propellants and Explosives,
Vol. II'', Atlas Copco AB, Stockholm and Sandvikens Jernverks 3, 17 Ä… 19 (1978).
AB, Sandviken, Sweden, 1957, pp. 1 Ä… 45. (11) ``Precautions against Blasting Fumes'', Circular No. 4 (Tech. 84)
(3) M. A. Cook, ``The Science of Industrial Explosives'', Graphic issued by Directorate General of Mines Safety (DGMS) in India,
Service & Supply Inc. Publication, USA, 1974, pp. 411 Ä… 429. 1984.
(4) F. G. Gagauz and A. V. Drebnitsa, ``Composition and Quantity of (12) P. M. Vuillaume and J. U. Bigourd, ``French Permitted Explo-
Noxious Gases During Blasting Operations in Underground sives'', Propellants, Explosives, Pyrotechnics, 11, 123 Ä… 128
Mining'', in: B. D. Rossi (ed),``Control of Noxious Gases in (1986).
Blasting Work and New Methods for Testing Industrial Explo- (13) R. McDam and R. Westwater, ``Mining Explosives'', Oliver and
sives''. Translated from Russian by the Israel Program for Sci- Boyd Publication, Edinburgh, London, 1958, p. 21.
enti®c Translation Ltd., Jerusalem, for National Science (14) C. E. Gregory, ``Explosives for North American Engineers'',
Foundation, Washington D.C., 1971, pp. 14 Ä… 15. Trans Tech Publications, Federal Republic of Germany, 1984,
(5) A. G. Streng, ``Evaluation of Toxic after Detonation Gases pp. 33, 37 and 170.
Formed by Industrial Explosives'', Explosivstoffe, 34, 58 Ä… 64 (15) M. M. Bhattacharyya, ``Studies on Fume Characteristics of
(1971). Explosives under Laboratory and Field Conditions''. Unpub-
(6) R. Meyer, ``Explosives'', Verlag Chemie, Weinheim, New York, lished Ph.D Thesis, Indian School of Mines, India, 1995,
1977, p. 118. pp. 1 Ä… 133.
(7) F. Volk, ``Investigation of the Detonation Reaction Products of
Different Explosives'', Propellants and Explosives, 3, 9Ä…13
Acknowledgements
(1978).
The authors express their sincere thanks to the Director, Central
(8) M. A. Cook, ``The Science of High Explosives'', ACS Mono-
Mining Research Institute, Dhanbad, India and scientists of its
graph No. 139, Reinhold Publishing Company, USA, 1974, pp.
Explosives Laboratory for the support and co-operation in carrying out
275 Ä… 300.
the investigations.
(9) B. D. Rossi, and V. A. Usachev, ``Quantitative Determination of
Noxious Gases Formed During the Explosive Conversion of
Explosives under Laboratory Conditions'', in: B. D. Rossi (ed),
``Control of Noxious Gases in Blasting Work and New Methods for (Received September 30, 1999; Ms 1999/62)