Sie sind auf Seite 1von 117


t. ~ .•.

Information Circular 9135

·"" :•
: · " · - ·J

Surface Mine Blasting

Proceedings: Bureau of Mines Technology Transfer
Seminar, Chicago, IL, April 15, 1987

Compiled by Staff, Bureau of Mines

US Department of Interior
Office of Surface Mining
Reclamation and Enforcement
Kenneth K. Eltschlager
Mining/Blasting Enginee1
3 Parkway Center
Pittsburgh, PA 15220

Phone 412.937.2169
Fax 412.937.3012


Donald Paul Hodel, Secretary

Robert C. Horton, Director


Preface......................................................................... i
Abstract....................................................................... 1
Bureau of Mines Surface Mine Blasting Research, by Dennis V. D'Andrea.......... 2
Reducing Accidents Through Improved Blasting Safety, by Larry R. Fletcher and
Dennis V. D'Andrea............................................................ 6
Blaster's Training Manual for Metal and Nonmetal Miners, By Michael A. Peltier,
Larry R. Fletcher, and Richard A. Dick........................................ 19
Delayed Blasting Tests To Reduce Rockfall Hazards, by Virgil J. Stachura and
Larry R. Fletcher••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••• 25
Effects of Blast Vibration on Construction Material Cracking in Residential
Structures, by MarkS. Stagg and David E. Siskind............................. 32
Blast Vibration Measurements Near Structures, by David E. Siskind and
Marks. Stagg................................................................. 46
Initiation Timing Influence on Ground Vibration and Airblast, by John W. Kopp.. 51
Vibrations From Blasting Over Abandoned Underground Mines, by David E. Siskind
and Virgil J. Stachura........................................................ 60
Computer Modeling of Rock Motion, by Stephen A. Rholl.......................... 73
Influence of Blast Delay Time on Rock Fragmentation: One-Tenth-Scale Tests,
by MarkS. Stagg and Michael J. Nutting....................................... 79
Blasting Effects on Appalachian Water Wells, by David E. Siskind and
John W. Kopp.................................................................. 96
Fiber Optic Probe to Measure Downhole Detonation Velocities of Explosive
Columns, by David .L. Schulz................................................... 103
Stemming Ejection and Burden Movements of Small Borehole Blasts, by
John w. Kopp.................................................................. 106

atm atmosphere (in/s) /in inch per second

per inch
degree Celsius
K kelvin
dB decibel
kHz kilohertz
degree Fahrenheit
lb pound
ft foot
lbf/in 2 pound (force)
square foot per square inch

ft/lb 11 2 foot per square root m meter

pound (scaled distance)
mg/L milligram per liter
ft/lb 11 3 foot per cube root
pound (scaled distance) mi/h mile per hour

ft/s foot per second min minute

G gravity (32.2 ft/s2) uin/in microinch per inch

g gram mm millimeter

gal/d gallon per day ms millisecond

gal/min gallon per minute ms/ft millisecond per

(gal/min)/ft gallon per minute per
foot (specific well us microsecond
mV millivolt
gram per cubic centimeter
ns nanosecond
h hour
pet percent
Hz hertz
s second
in inch
w watt
cubic inch
yr year
in/s inch per second

cubic inch per second


Proceedings: Bureau of Mines Technology Transfer Seminar,

Chicago, IL, April 15, 1987

Compiled by Staff, Bureau of Mines


The Bureau of Mines has sponsored a comprehensive research program to

enhance the safe, effective, and efficient use of blasting technology by
the mining industry. Recent research results of the surface mine blast
ing program were presented at a seminar on April 15, 1987, in Chicago,
IL. Many of the topics discussed at the seminar are presented in this
proceedings. The new research described in these papers includes com-
puter monitoring of rock motion, the influence of blast delay times on
rock fragmentation, blasting effects on Appalachian water wells, blast
vibration measurement near structures, and the reduction of accidents
through improved blasting safety.


By Dennis V. D'Andrea1

The Bureau of Mines Twin Cities Re- and equipment, for potential improvements C01
search Center has a comprehensive re- in blasting practices. This paper out- n(
search program on the efficient and safe lines surface mine blasting research com- St1
application of explosives in mining. Re- pleted since the Bureau's last Technology v:
searchers combine an understanding of the Transfer seminar on blasting in December I!
basic principles of dynamic rock fragmen- 1980. Three programmatic areas--produc-
tation with new blast design technology tivity technology, blasting vibrations, Fa1
and recent developments in both methods and blasting safety--are reviewed. b


Blasting research has been conducted safety research began at TCRC as one con- BL
at the Twin Cities Research Center (TCRC) tract project in fiscal year 1978 and v
since the center opened in 1959. Re- grew to involve four in-house projects
search during the 1960's and early 1970's during the years 1984 to 1986. Bl.
established TCRC as the leading Bureau of The heavily field-oriented blasting re- f
Mines center in the area of blasting for search program at TCRC has included 45
improved fragmentation and increased pro- in-house and contract project efforts
ductivity. During the period from fiscal since 1975, resulting in 122 publications
year 1974 through fiscal year 1979, pro- and numerous presentations at profes-
ductivity research was on blasting to sional meetings. Report of Investiga-
prepare ore bodies for in situ leaching. tions (RI) 8507, on structural response
The major effort at TCRC from fiscal year and damage from blasting vibrations, won it
1975 through fiocal year 1983 was in the the 1981 Applied Research Award from the
area of environmental effects of blasting U.S. National Committee for Rock Mechan-
(ground vibrations and airblast). Re- ics. TCRC personnel have responded to
search since fiscal year 1983 has been over 650 requests for technical assis- mi
mostly on blasting fundamentals for tance and advice on blasting since 1981. pr
improvements in productivity. Blasting Na
Mining Technology examining the fundamentals of blasting mu
and blast-produced rock fragmentation. fi
Major research efforts at TCRC on im- The fundamental research includes a study se
proved productivity and blasting vibra- of high-precision delay initiators to im- st
tions in surface mines are listed in ta- prove fragmentation. vi
ble 1. Included are projects intended to The Bureau is frequently asked to as- ef
improve mining productivity and to p:::-c- sist other Federal agencies, such as the or
vide information on good blasting prac- Office of Surface Mining, the Bureau of ur
tices. The projects that started in the Reclamation, and the Mine Safety and
late 1970's addressed environmental is- Health Administration, on environmental
sues, with indirect implications for and safety issues associated with blast-
mining costs and productivity. More re- ing practices. These assistance efforts
cent long-range, high-risk research is usually do not involve pure research, but cc
rather providing technical information at
1 Research
supervisor, Twin Cities Re- and, on occasion, measurement, analysis, mj
search Center, Bureau of Mines, Minne- and advice. mi
apolis, MN. si

TABLE 1. - Blasting research at TCRC on improved productivity

and blasting vibrations

Research area Fiscal Key researchers Significant

y_ears _p~ublications I
Blasting effects on 1978-80 P. R. Berger and D. Robert son (_!).
Appalachian water wells. Associates. 2
ents Contour mine blasting 1978-81 V. J. Stachura •••••••••• Stachura, RI 8892
out- noise and vibrations. (__?).
com- Standards efforts on 1980-81 D. E. Siskind, Acousti- ANSI Standard
logy vibrations, ANSI and cal Society of America. S3.29-1983 (_~).
mber ISO. Other ANSI and ISO
duc- Standards.
ons, Fatigue from repeated 1979-83 M. s. Stagg, National Stagg, RI 8896 ( 4).
blasting. Bureau of Standards. 2 Siskind ( 5).
Woodward (~).
Vibration measurement 1983 D. E. Siskind ••••••••••• Siskind, RI 8969 (]_).
near buildings.
con- Blast designs to control 1979-83 J. \¥. Kopp • ••••••••••••• Kopp, RI 9026 (~).
and vibrations.
Blasting fragmentation 1984- M. S. Stagg, Sandia Nut t i. ng (2) •
; re- fundamentals. (ongoing)Laboratories, 2 Univer-
:d 45 sity of Maryland.2
orts Low-frequency vibrations. 1985- D. E. Siskind, In preparation.
ions (ongoing) V. J. Stachura.
)fes- ANSI American National Standards Instttute.
:iga- ISO International Standards Organization.
10nly the senior author is listed here. Underlined numbers in parentheses refer to
won items in the list of references at the end of this paper.
1the 2 Work conducted for the Bureau under contract.
~d to Projects listed in table 1 range from Center, which focuses on the safe use and
;sis- minor efforts (such as providing an im- evaluation of permissible explosives and
~81. proved technology basis for the American permissible blasting methods for under-
National Standards Institute and the ground coal mines, and research on the
International Standards Organization to properties and explosives chemistry re-
use in establishing standards) to major lated to safe explosive performance,
;ting multiphase projects involving up to storage, and transportation.
:ion. five supporting industry and Government Table 2 summarizes the major blasting
:;tudy service contracts (such as the fatigue projects at TCRC in the Bureau's Health
) im- study). The research on low-frequency and Safety Program. Initially, all work
vibrations is a TCRC technical assistance was contracted out. More recently, the
::>as- effort for the Office of Surface Mining TCRC has conducted in-house research on a
3 the on blasting vibrations above abandoned wide range of significant problems re-
3U of underground mine workings. Bu. ."-'\':>r~):::. lated to safer blasting procedures.
and Examples of this in-house research are
ental Blasting Safety blast designs for safer and more stable
last- highwalls and the development of materi-
forts Research on blasting safety at TCRC is als for blasters training.
, but concerned with safer blasting practices Research on blasting safety is guided
at ion and blast designs primarily for surface by the analysis of blasting accident
ysis, mines and underground metal-nonmetal statistics. This analysis now covers
mines. The Bureau also conducts explo- 8 yr, and has determined the most fre-
sives research at the Pittsburgh Research quent causes of blasting accidents and

TABLE 2. - Research at TCRC since 1980 on blasting safety

Research area Fiscal Key researchers Significant

years publicationsl
Certification of blasters 1978 E. I. DuPont de Nemours Coulson and Southall
& Co. 2 (.!_Q_).
Blasting manual •••••••••• 1980-81 R. A. Dick ••.. •...•••••• Dick, IC 8925 (!.!).
Misfires ••••••••••••••••. 1982 L. R. Fletcher •••••••••• Fletcher (12).
Blast area security •••••• 1981-84 Mining & Marketing Bennett (13-14).
Associates. 2
Highwall stability ••••••• 1983-84 v. J. Stachura, Stachura, RI 8916
1. R. Fletcher. <.!1)' RI 9008 (16).
Blasters training •••••••• 1984-86 1. R. Fletcher and In preparation.
M. A. Peltier.
Flyrock control •••••••••• 1986 L. R. Fletcher •••••••••• Fletcher (17).
Blasting accident 1981- D. v. D'Andrea •••••••••• D'Andrea (18).
analysis. (ongoing) Peltier (19).
Only the senior author is l1sted here. Underlined numbers in parentheses refer to
items in the list of references at the end of this paper.
2work conducted for the Bureau under contract.

identified where the most hazardous situ- these needs through research efforts in
ations exist. Through this analysis, re- areas such as blast area security, mis-
searchers have been able to determine the fires, and flyrock control.
most critical industry needs and address


The blasting research effort at TCRC is house showing that low-level vibrations
concerned with improved productivity, from repeated blasting did not damage
blasting vibrations, and blasting safety. structures. Blast design studies found
The following papers summarize research that the- best fragmentatio.n was achieved
projects on surface mine blasting that when delays between blastholes were at
have been carried out and reported on least 1 ms/ft of burden. Highwall sta-
since the Bureau last held a Technology bility was improved using longer periods
Transfer seminar on blasting in December in the row of blastholes that formed the
1980. highwall. Accident analysis indicated
Highlights of the TCRC research include that failure of the blast area security
evaluations and recommendations for vi- system is the major cause of blasting
bration measurement methods near build- accidents.
ings and a comprehensive study of a test


1. Robertson, D. A., J. A. Gould, Contour Mine Blasting. BuMines RI 8892,

J. A. Straw, and M. A. Dayton. Survey 1984, 31 pp.
of Blasting Effects on Ground Water Sup- 3. Acoustical Society of America (New
plies in Appalachia. Volume I (contract York). Guide to the Evaluation of Human
J0285029, Philip R. Berger and Associ- Exposure to Vibrations in Buildings.
ates, Inc.). BuMines OFR 8(1)-82, 1980, American National Standards Institute
159 pp. (ANSI) S3.29. 1983, 9 PP•
2. Stachura, V. J., D. E. Siskind, and 4. Stagg, M. S,, D. E. Siskind, M. G.
J. W. Kopp. Airblast and Ground Vibra- Stevens, and C. H. Dowding. Effects of
tion Generation and Propagation From

Repeated Blasting on a Wood-Frame House. TX, Jan. 31-Feb. 4, 1983). Soc. Explos.
BuMines RI 8896, 1984, 82 PP• Eng., Montville, OH, 1983, pp. 123-132.
5. Siskind, D. E., and M. S. Stagg. 13. Bennett, J. Survey of Safety Pro-
Blast Vibration Damage to Structures. cedures for Guarding Blast Affected Areas
hall Paper in Proceedings 16th Annual Insti- (contract J0205019, Mining & Marketing
tute on Coal Mining Health, Safety, and Associates, Inc.). BuMines OFR 88-82,
1_). Research. VA Polytech. Inst., Dep. Min. 1981, 74 pp.; NTIS PB 82-246042.
and Miner. Eng., Blacksburg, VA, 1985, 14. D'Andrea, D. V., and J. Bennett.
PP• 199-213, Safeguarding of Blast-Affected Areas.
6. Woodward, K. A., and F. Rankin. Paper in Proceedings of the lOth Confer-
6 Behavior of Concrete Block Masonary Walls ence on Explosives and Blasting Tech-
16 Subjected to Repeated Cyclic Displace- nique, ed. by C. J. Konya (Lake Buena
ments. NBS (Tech. Rep.) NBSIR 83-2780, Vista, FL, Jan. 29-Feb. 2, 1984), Soc.
1983, 178 PP• Explos. Eng., Montville, OH, 1984,
7. Siskind, D. E., and M. S. Stagg. PP• 110-119.
Blast Vibration Measurements Near and On 15. Stachura, V. J., and L. R. Flet-
Structure Foundations. BuMines RI 8969, cher. Delayed Blasting Tests To Improve
er to 1985, 20 PP• Highwall Stability--A Progress Report.
8. Kopp, J. W., and D. E. Siskind. BuMines RI 8916, 1984, 24 pp.
Effects of Millisecond-Delay Intervals on 16. Stachura, V. J., L. R. Fletcher,
Vibration and Airblast From Surface Coal and M. A. Peltier. Delayed Blasting
s in Mine Blasting. BuMines RI 9026, 1986, Tests To Improve Highwall Stability--A
mis- 44 PP• Final Report. BuMines RI 9008, 1986,
9. Nutting, M. J., M. S. Stagg, and 12 PP•
D. R. Barlet. Effects of Delay Time on 17. Fletcher, L. R., and D. v.
Fragmentation. Paper in Rock Mechanics: D'Andrea. Control of Flyrock in Blast-
Key to Energy Production, (Proc. 27th ing. Paper in Proceedings of the 12th
U.S. Symp. on Rock Mechanics, ed. by Conference on Explosives and Blasting
tions H. C. Hartman Univ. AL, Tuscaloosa, AL, Technique, ed. by C. J. Konya (Atlanta,
a mage June 23-25, 1986). Soc. Min. Eng., 1986, GA, Feb. 9-14, 1986). Soc. Explos. Eng.,
found PP• 449-454. Montville, OH; 1986, pp. 167-177.
.ieved 10. Coulson, J. R., and L. T. South- 18. D'Andrea, D. V., and L. R. Flet-
re at all, II. Considerations for the Certifi- cher. Analysis of Recent Mine Blasting
sta- cation of Blasters (contract J0285012, Accidents. Paper in Proceedings of the
riods E.I. DuPont de Nemours & Co., Inc.). Bu- 9th Conference on Explosives and Blasting
d the Mines OFR 59-81, 1980, 77 PP•; NTIS PB Technique, ed. by C. J. Konya (Dallas,
cated 81-214116. TX, Jan. 31-Feb. 4, 1983). Soc. Expl.
urity 11. Dick, R. A., L. R. Fletcher, and Eng., Montville, OH, 1983, pp. 105-122.
.sting D. V. D'Andrea. Explosives and Blasting 19. Peltier, M. A., L. R. Fletcher,
Procedures Manual. BuMines IC 8925, and D. V. D'Andrea. Coal Mine Blasting
1983, 105 pp. Accidents. Paper in Symposium on Engi-
12. Fletcher, L. R., and D. V. neering Health and Safety in Coal Mining,
D'Andrea. A Study of Misfires in Min- ed. by A. Wahab Khair (New Orleans, LA,
ing. Paper in Proceedings of the 9th Mar. 2, 1986). Soc. Min. Eng. AIME,
8892, Conference on Explosives and Blasting 1986, pp. 174-181.
Technique, ed. by C. J. Konya. (Dallas,

M. G•
.ts of


By Larry R. Fletcher1 and Dennis v. D'Andrea 2


The Bureau of Mines investigated three flyrock. The operator must change blast- to
of the major causes of mine blasting ac- ing methods when shooting in geology that bL
cidents: inadequate blast area security, favors the production of flyrock. bli
excessive flyrock, and misfires. Most misfire accidents are caused by OCI
Accidents resulting from inadequate drilling into bootlegs in underground US!
blast area security occur during sched- metal and nonmetal mines. Improper dis- ha,
uled blasting because of failure to clear posing of misfires is the second most th<
the blast zone, inadequate guarding, and frequent cause of misfire accidents, and
failure ~f personnel to follow instruc- some accidents are due to impact initia-
tions, retreat to a safe location, and/or tion of explosives in the muckpile. Mis-
take adequate cover. fires are usually caused by misunder-
Excessive flyrock is produced when standing, improper use, or some failure me:
there is too much explosive energy for of the initiation system. Other causes in
the amount of burden, stemming is inade- are cutoffs, insufficient firing current, Wh1
quate, or the explosive energy is too inadequate priming, improper explosive Mo:
rapidly vented through a zone of weakness storage, and damage to the initiation Unt
in the rock. Geology, improper blast de- system. bl;
sign, or carelessness can cause unwanted si:
Analysis of mine blasting accidents with poorly controlled shots. A safety in:
shows that the rive most frequent causes factor is added to the normal flyrock un.
of accidents are premature blasts, in- range to determine the blast area to be th;
adequate blast area security, excessive cleared ana secured before .the blast is bL
flyrock, misfires, and fumes. Over 70 detonated. Rock that travels beyond the te:
pet of all blasting accident injuries secured blast area is excessive flyrock
from 1978 through 1985 have been caused (fig. 1). A distinction is made between Gl
by three of these: inadequate blast area injuries that occur within the estab-
security, flyrock, and misfires, all lished blast area and injuries that are
of which are discussed in this paper, the result of excessive flyrock projected Set
based on earlier Bureau of Mines research beyond the blast area. Un(
(l_?_).3 Anyone who remains in the blast area, ne:
Each mlning operation has a normal fly- such as the shot firer, must have ade- fr<
rock range, the distance from the blast quate protection from flyrock such as met
at which flyrock can be expected, based that provided inside a blast shelter. ZOI
on blasting experience at that operation. The shot firer is frequently guilty of pa~
The distance of the normal flyrock range
will vary from a few feet in an area fly rock cor
strip coal mine blast to more than a mile t i'

Mining engineering
2 . -
technician. J:
Research supervisor. sife cl<
'I'win C i tiP.s Resea:rch C: enter, Bureau of I
Mines, M1nneapolis, MN. th~
3 underlined numbers in parentheses re
fer to Ltems in the list of references at FIGURE 1.-Biast site, normal flyrock range, secured blast I
the end of this paper. area, and excessive flyrock region for surface mine blast. ent

shooting and observing blasts from within because many people feel that misfires
the blast area without adequate protec- reflect the quality of their work, they
tion. When a shot firer pays inadequate are reluctant to report them. In addi-
attention to the hazards of flyrock, a tion, the increased use of non-cap-sen-
permissive attitude toward guarding and sitive blasting agents with lower shock
protecting the blast area is created. sensitivity has generated an attitude of
A misfire results when explosives fail indifference about misfires at some oper-
>last- to detonate as planned during a mine ations. Because of these two factors, it
' that blast. It is difficult, if not impossi- is rare that a misfire that does not in-
ble, to determine how frequently misfires volve an injury is ever reported.
;ed by occur. When operators are asked, they This paper discusses the elements of
sround usually reply, "Rarely, if ever, do we effective blast guarding, the causes and.
~ dis- have a misfire." The truth is, however, control of flyrock, and the causes, de-
most that misfires are fairly common, but tection, and disposal of misfires.
;, and
mder- The blast area security system is the keep personnel from moving into the blast
lilure means by which a mine operator prevents zone.
:::a uses injury to people or damage to equipment
rrent, when a scheduled blast is detonated. Elements contributing to the accom-
Losi ve Most blasting accidents in surface and plishment of the blast-site security sys-
tat ion underground mines occur during scheduled tem requirements are as follows:
blasting and are due to inadequate blast
site security. The result is often un- Management commitment to safety.
necessary injury and/or death, caused, in Training.
most cases, by flying rock. Mine person- Definition of blasting authority.
nel, visitors, and even trespassers can Planning.
safety inadvertently wander int0 a blasting zone Blast zone boundaries.
lyrock unless proper procedures exist to ensure Selection and placement of guards.
to be that all personnel are cleared from the Clearing procedure.
ast is blast zone and kept safely away until af- LocaLion of blast initiation sice.
nd the ter blasting is completed. Blasting time.
lyrock Blast signals.
estab- Communication of blast location and
at are The basic requirements of a blast-site time.
jected security system at either a surface or Blasting crew communications.
underground mine are (1) to move person-
area, nel and equipment out of range of flyrock Each of these elements is discussed
ade- from the blast and (2) to prevent move- below.
uch as ment of personnel back into the blast
elter. zone. This includes visitors and tres- MANAGEMENT COMMITMENT TO SAFETY
ty of passers, along with mine personnel.
The most effective procedures for ac- The mines that have developed the best
ive flyrock complishing blast-site security objec- blast-guarding systems appear to be those
"' tives are to..;..- in which management has made a definite
commitment to safety. It has been recog-
Have blasting personnel physically nized at those mines that it is manage-
clear the blast-affected zone; ment's responsibility to provide mining
Account for personnel, to establish personnel with a systematic, safe blast-
that no one is present in the blast zone guarding method and to provide sufficient
at blast time; and training to allow m~n~ng personnel to
Jred blast Place guards beyond the blast zone, at operate intelligently within the estab-
blast. entries leading into the blast site, to lished system.

Blast-site security objectives are gen- SELECtiON AND PLACEMENT OF GUARDS

erally accomplished through the develop-
ment of a blast-guarding standard operat- It is the general consensus throughout
ing procedure specifically designed to the industry that human guards should be
fit the requirements of a given mine. used to guard blast-site entries whenever
possible. Entrances to blast areas are
TRAINING sometimes left unguarded because it is
either impractical or unsafe to place a
Having developed a standard blast- person on post. When an entry is un-
guarding procedure, mine management is guarded, a barricade or sign is used to
obligated to inform its employees of warn people, but these methods are gener-
their responsibilities. Operations with ally considered inadequate. i.' ,
good blasting security systems utilize Persons selected as guards should be ~
information regarding proper blast guard- sufficiently trained to understand the
ing as training material for new hires serious responsibility they have to pre-
and for retraining experienced miners. vent any movement of personnel back into
It is important that blast guarding be the blast zone. The absolute authority
treated specifically and in detail as a of the guard to prevent movement beyond
real safety problem. the post must be established. The guards
themselves must be placed in a location
DEFINITION OF BLASTING AUTHORITY safe from flyrock and fumes.

One experienced individual must be re- CLEARING PROCEDURE

sponsible for the blast. Whether this
person is a blasting superintendent, Mine management should establish a def-
foreman, or crew leader, it is extremely inite clearing procedure as part of the
important that one key individual direct blast-guarding standard operating proce-
the blast-guarding process for any given dure. The clearing path will depend on
blast. mine layout, and considerable variance
PLANNING can exist within a given mine. Clearing
personnel in surface mines are assigned
Preblast planning is an important as- to drive ~or walk through the blast zone
pect of blast guarding. Mines that have prior to the blast. In mines where low-
standardized their blast-guarding methods mobility equipment such as drills and
reduce the requirement for daily planning shovels are left in the blast zone, this
by making certain blast-associated deci- equipment should be inspected to be sure
sions routine. This is particularly true all personnel have left.
at mines that have adopted central blast- Clearing procedures in underground
ing systems and consistent methods for mines vary considerably. Since blasters
identifying crew member locations before frequently finish loading underground
a blast. rounds well before the scheduled shift-
change blast time, last-minute clearing
BLAST ZONE BOUNDARIES must be done. If effective clearing pri-
or to blasting is impractical, it is es-
The boundary around the blast beyond sential to use personnel accounting and
which personnel and equipment will be keep miners informed of the time and lo-
safe from flyrock must be determined. cation of the blast. Blast-guarding sys-
Guard posts must be located outside that tems generally benefit from the inherent
boundary at all entries into the blast redundancies associated with requiring
zone. Flyrock range, of course, is de- both clearing and personnel accounting in
pendent on a number of factors including the blast zone.
blast configuration, rock hardness, stem-
ming, geology, etc., and is determined
for each particular location and blast.

)S LOCATION OF BLAST INITIATION SITE done concurrently with the production

blast •
.ghout A definite initiation site (blasting
.ld be location) should be designated for all BLASTING SIGNALS
never blasting areas in the mine or pit. Loca-
.s are tion of the blast initiation site should Audible signs are used at most surface
i t is be based on a conscious management deci- mines to warn of the impending blast. In
ace a sion as part of the preblast planning some cases, an air horn or electric siren
un- process. is mounted at a fixed power source. In
ed to The purpose of designating location is others, particularly when the pit is
ener- twofold. First, it provides management large, sirens or horns are mounted on
with a means of fulfilling its respon- vehicles used for clearing the blast
ld be sibility to ensure that the blaster re- zone. In either case, the devices used
the treats to a safe location before blast- should be loud enough to be heard by per-
pre- ing. The recent Bureau analysis of sonnel approaching the blast zone from ·
into blasting accidents revealed that many any direction. In addition, considera-
ority surface and underground accidents were tion should be given to prevailing wind
eyond caused by unsafe location of the shot directions and velocities. Many surface
uards firer. Second, the location of a desig- mines use flashing lights on blast-asso-
at ion nated blast initiation station can easily ciated vehicles as a supplement to sirens
be communicated to all mine personnel as and horns.
the required checkpoint for personnel ac- The duration and pattern of audible
counting when persons are entering or signals used at the mines vary consid-
leaving that area of the mine or pit. erably. Some start 10 min prior to the
def- Minimum recommended requirements for a blast, while others start 2 min before
f the blasting site are as follows: the blast. Some signals continue
throughout the blast, while others are
ld on The site should be located outside the interrupted during the blast. A second
tance blast zone, protected from flyrock and signal after the blast signifies that all
uing upwind of blast fumes. Blast initiation is clear.
tgned from inside the blast zone should be per- The signal must fulfill three require-
zone mitted only if a secure blast shelter is ments, -as follows:
low- used.
and The site should be equipped with a It must be loud enough to warn person-
this highly visible sign giving the blast nel inside and near the zone of the im-
sure time. The sign should also notify visi- pending blast.
tors that they are required to check in In the event that the clearing process
·ound and out at that station. has failed and persons remain exposed to
;ters the impending blast, the signal should
::ound BLASTING TIME allow sufficient time for them to take
lift- cover or leave the area.
Iring Predictability is an essential element It must clearly indicate when it is
pri- in a process with the destructive poten- safe to reenter the blast zone.
es- tial of blasting. Therefore, it is rec- The air-powered whistle used at the
; and ommended that efforts be made to estab- Union Carbide Creek Pine Creek Mine (fig.
l lo- lish consistent blast times. Whether the 2) is a simple and effective means of
sys- blast is initiated at lunch time or the warning that a blast is about to be ini-
!rent end of a shift, blast scheduling should tiated, particularly in areas where en-
.ring be strictly controlled where possible. tries are left unguarded. A whistle im-
tg in Timing for secondary blasting varies ac- proves blasting-site security when used
cording to the needs of the particular as a backup to human guards and is a very
mine. Many mines, both surface and un- cost-effective device. Easy to use, the
derground, have developed systems in whistle is simply screwed onto a drop on
which secondary blasting is regularly the air line and turned on as the blaster

~----------------6in------------------~ during the blast. The board contains a

tag for every person assigned to a par-
ticular station plus tags for visitors.
When blast time arrives, supervisors at to
each safe location merely need to look at bL
~ 6 -in schedule 40 pipe the board to ensure that all personnel no
Weld plug have returned from the immediate blast co:
in end si
FIGURE 2.-Air whistle used at Pine Creek Mine (NPT-National Surface mines seem to place less empha- all
Pipe Taper). sis on personnel accounting than do the at
underground mines. This is somewhat jus- no
retreats from the face to begin the tified by the fact that clearing proce- sh
clearing process. It remains on through- dures are more effectively supervised and ca
out the blasting procedure and is turned performed within surface blast areas be- ea
off by the first person entering the area cause of better visibility. It is also
after the blast. impractical for some large surface mines ca
that blast several times each day to ac- th
PERSONNEL ACCOUNTING count for all mine personnel before each ra
blast. Fortunately, most of these mines di
Personnel accounting refers to deter- have a limited flyrock range and can re- tr
mination of the locations of personnel liably use the clearing process. How-
throughout the mine site. In essence, it ever, where practical, personnel account-
is a tracking system and should be de- ing should act as a backup system to the
signed, at the very least, to account for clearing process. Blasting safety at
the presence of personnel within the zone many surface mines would benefit from an je
of a blast. Whenever reasonably possi- improvement in personnel accounting. ar
ble, accounting stations should be estab- mu
lised near all blast zones, preferably at COMMUNICATION OF BLAST bu
the locations from which the blasts are LOCATION AND TIME or
regularly initiated. At a minimum, an VE
effective personnel accounting system It is important that all personnel en- ce
must establish that no personnel are oc- tering a mine be aware of the exact loca- of
cupying the blast zone at blast time. tion of blasting sites and the schedule oc
The personnel accounting system is an im- for blast initiation. Surface and under- CE

portant backup to the clearing process. ground mines that blast on a daily basis bl
However, some large open pit operations can list the blasting times on signs at ti
that blast several times each day rely on appropriate entrances into the mining lE
clearing as the only practical method of areas. Since blast location changes fre- cz
ensuring that there are no people in the quently at most mines, mine personnel si
blast area. When clearing is the only should be informed on a daily basis of he
method used, the clearing personnel must the exact locations of all scheduled PI
conscientiously cover the entire blast- blasts.
affected area prior to every blast. Notices regarding blast-site locations
Many mines account for personnel by in underground mines should be placed at
means of a head count at some safe loca- the personnel accounting station for the
tion prior to the blast. While a head blasting area. As personnel check in at tt
count is better than not accounting for the blasting station, they are informed tt
personnel at all, it allows considerable of the blast time and location. This bE
potential for error. particularly benefits personnel who are tt
The system of using a blasting board is not regularly assigned to a section for Mt
applicable at many underground and sur- which blasting is scheduled but who are WE
face mines. Blasting boards are placed required to move throughout the mine. cz
at safe locations where miners congregate fr

,ins a BLASTING CREW COMMUNICATIONS by reference to the Institute of Makers

par- of Explosive (IME) Publication 20, "Safe-
.tors. Some surface mines utilize hand signals ty Guide for the Prevention of Radio Fre-
;rs at to communicate between the shot firer and quency Radiation Hazards in the Use of
•OK at blast guards. However, hand signals do Electric Blasting Caps" (4). Another
.onnel not provide the positive, instantaneous source is the Du Pont "Blaster's Hand-
blast communication so necessary to blasting- book" (5). Radio communication for blast
site security. A hand signal will not guarding is safe if the IME guidelines
mpha- always immediately attract the blaster's are followed. Transmitters should be
.o the attention. Poor visibility, a blaster's limited to 5 W or less, and a boundary
jus- not wearing corrective eyeglasses, or sun should be defined around the blast pat-
•roce- shining in the blaster's eyes are all tern within which the use of radios is
,d and cases in which a hand signal might not be forbidden.
.s be- easily seen • An alternative is the use of nonelec-
also Guards should use warning devices tric blast initiation systems, which
mines capable of quickly and positively gaining are widely used in the mining industry.
o ac- the blaster's attention. The two-way These systems cannot be initiated by
each radio does this best. Safe separation radio frequency energy and are com-
mines distances between transmitters and elec- pletely compatible with · close radio
n re- tric blasting circuits can be determined communi cat ion.
o the
y at Excessive flyrock is rock that is pro- become filled with too much explosive for
om an jected beyond the normal blast-affected the amount of rock burden, resulting in
area. It is generated when there is too large flyrock distances. Blastholes can
much explosive energy for the amount of penetrate openings from abandoned under-
burden, when stemming in insufficient, ground mines to create a dangerous con-
or when the explosive energy is rapidly dition similar to that resulting from
vented through a plane of weakness. Ex- natural cavities. Fracturing due to
1 en- cessive flyrock is responsible for 24 pet backbreak or overbreak from previous
loca- of the blasting accident injuries that blascing can also cause dangerous planes
edule occur in surface mining (1978-85). Ex- of weakness. A ragged highwall face or
nder- cessive flyrock can be the result of an overhang can result in diminished bur-
basis blast site geology and/or rock condi- den along the front row of holes.
ns at tions, improper blast design, or care- Excessive flyrock can be generated if
ining lessness. Flyrock control is achieved by blasts are not properly designed. Any
fre- careful attention to blast design, blast- blast design feature that results in in-
onnel site inspection, blasthole layout, blast- sufficient explosive confinement or the
is of hole drilling, and blasthole loading rapid venting of the explosive gases can~,
duled practices. create a problem. Blast design errors
such as too high a powder factor, an in-
tions FLYROCK CAUSES adequate burden, too short a stemming re-
ed at gion, failure to use stemming, improper
r the Geology and rock conditions can cause delays between rows, or the wrong blast-
in at the generation of flyrock. Geologic fea- hole delay sequence can result in un-
ormed tures such as mud seams, natural joint or wanted flyrock. The wrong delay sequence
This bedding planes, fractures, or cavities in can cause diminished burden if the delay
are the rock can result in excessive flyrock. is too long. Cratering and blowouts can
n for Mud seams and fractures are planes of occur when back holes fire before front
o are weakness through which explosive gases holes. A very short delay can result in
can rapidly vent and accelerate rock too much confinement and, again, crater-
fragments. Cavities can accidentally ing and blowouts.

Unfortunately, carelessness is a ~-~~­ ruggedness, overhangs, fractures, zones

ing cause of excessive flyrock. Care- of varying competence, and amount of toe he
lessness during any part of the blast de- burden. The blast site should also be DE
sign, the blasthole pattern layout, the inspected for backbreak, jointing, mud rot
drilling of the blastholes, the loading seams, voids, and other zones of weak- dl
of the blastholes, or the hookup of the ness. Any of these blast-site features tl
initiation system can create a dangerous could cause excessive flyrock. dl
situation. The loading of explosives too The layout of the blasthole pattern tl
near the collar of the blasthole is a starts with the front row. If the ver- dl
common cause of flyrock. tical face has overhangs, is concave, or
is irregular, the burden at some point sj
CONTROL OF FLYROCK may be reduced and violent cratering lc
could occur. Faces with backbreak, open tr
The control of flyrock starts with joints, weak zones, and mud seams will If
proper blast design. The correct burden allow rapid venting of explosion gases pJ
must be used. A small burden will not with flyrock. In addition, the burden cc
contain the explosive energy, while using will not pull (be removed) as planned, re
a large burden may result in cratering causing an increase in the burden for be
and/or blowouts. The bench height, bur- later holes, which results in cratering a
den, and stemming region must be such at the top of the bench. If the face is m,;
that the blasted rock movement is primar- sloped, the toe burden will be larger Be
ily horizontal and outward, and not up- than the crest burden unless angled holes he
ward. In multiple-row shots, the delay are used. If the normal column load is ti
between rows must be long enough to allow used when there is a sloped face, there th
rock from an earlier row to move out so will be flyrock because of the short an
that the next row will have adequate re- crest burden. Also, the toe may not cc
lief. Insufficient relief can cause fly- pull, producing a buildup in front of
rock. However, the delay must not be so later holes, again resulting in flyrock. a
long that cutoffs occur and cause mis- Adjustments in hole locations and powder of
fires that increase the burden on later columns in the front row should be made ti
firing holes, again resulting in blowouts when conditions exist near the face that un
and flyrock. favor the- generation of flyrock. Once th
In designing a blast, relationships be- the front row is established, the bal- fr
tween charge diameter, burden, spacing, ance of the shot can be laid out. A tape re
subdrilling, stemming region, and bench should be used to ensure accurate spacing ho
height are available for initial esti- and burden distances. th
mates (~). The type of explosive, the Locating a blasthole close to an open th
priming, and the initiation system must fracture will provide a weak zone. The uc
be selected. Toe priming reduces fly- shot will break into the fracture, vent rna
rock, compared with collar priming. with flyrock, and produce poor fragmenta-
Decisions must be made on the type of tion. The same kind of venting can occur ri
blasthole pattern, square or staggered, when a hole is abandoned and a second sl
and on the delay sequence. The powder hole is drilled a few feet away. To pre- vo
factor is calculated to ensure that the vent this, the first hole should be back- bl
quantity of explosive being used is with- filled. Where open fractures are pres- ex
in the range of that normally used in ent, they can be backfilled, but this is en
surface mine blasting. These initial ap- difficult and time consuming. The best bl
proximations must be modified for the way to handle fractures from previous qu
particular blasting situation and may be blasting is to eliminate the cause of de
further modified after experience with a overbreak. ag
number of blasts. Accurate drilling is essential. The mu
Before the blasthole pattern is laid holes must be located in accordance with du
out in preparation for drilling, a care- the blast design and drilled at the cor- rna
ful inspection of the blast site should rect angle and to the proper depth. With An
be made. The face should be examined for a high face and smaller diameter holes,

zones extra care should be taken to ensure that continued. If the void is too large for
f toe holes are drilled at the proper angle. this to be practical, the hole may have
o be Deviated blastholes can result in burdens to be abandoned and another hole drilled
mud much smaller or larger than planned. The nearby. Redrilling should not be done if
weak- driller should provide a log of each hole there is a possibility of drilling into
tures that includes depth drilled, problem explosives. In some cases, the hole can
drilling zones, and any changes in pene- be plugged just above the void. Once a
ttern tration rate. This information could in- plug is formed, the explosive loading
ver- dicate voids or zones of weakness. can resume. In operations where voids
e, or All holes must be checked before explo- are common, a special system for borehole
point sives are loaded, to ensure that their plugging should be developed.
ering location and depth are in accordance with Checking the column rise will prevent
open the blast design and the driller's log. accidental overloading of the blasthole.
will If blocked holes are undetected, the ex- Maintaining sufficient stemming is an
~ases plosive column can be loaded too near the important factor in flyrock control.
1rden collar, leaving an insufficient stemming Stemming lengths of 0.7 to 1.0 times the
1ned, region. Partially blocked holes can also burden are commonly used. When collar
for be a problem and may not be detected with priming is used, the stemming length may
~ring a weighted tape. On sunny days, a mirror need to be increased because of the
:::e is may be used to check for obstructions. greater potential for violence with top
uge r Before an attempt is made to load a short priming. Crushed and sized rock is the
10les hole or a partially blocked hole, correc- best material to use for stemming, but
ld is tive action such as redrilling, changing drill cuttings are commonly used because
:here the planned explosive load, or abandoning of availability and economy. Large
>hort and backfilling of the hole should be pieces of rock or other material should
not considered. never be mixed with the stemming as they
of Open joints and cracks can extend into can become missiles if there is a blow-
·ock. a blasthole. The presence of cracks can out. Large rocks can also cut off or
>wder often be detected by a lack of drill cut- damage the initiation system and cause a
made tings at the top of the hole. However, misfire.
that unless corrective action is taken during Care must be taken to ensure that the
Once the loading of bulk products, these open initiation system is properly hooked up
hal- fractures can be loaded with explosives, and that the delays are correct. A final
tape resulting in excessive energy for the check of the hookup is imperative.
tcing hole. There will be less confinement of The secondary blasting of boulders too
the explosive charge. This may affect large for the loading equipment or crush-
open the performance of some explosive prod- er is required at some operations. Sec-
The ucts as well as provide a weak zone that ondary blasting can produce dangerous
vent may blow out. flyrock even though ~he charge~ are
mta- Frequent checks of the powder-column small. Determ1ning the blast area for
•ccur rise during loading are important. A this kind of shooting is very difficult,
~cond slower than normal rise may indicate a and careful clearing and guarding are re-
pre- void, while a sudden rise indicates a quired. Although secondary blasting is
•ack- blockage. When a blockage occurs during frequently done on-shift and as needed,
,res- explosive loading, care must be taken to it is best to shoot at a standard time
a is ensure that the explosive load above the such as during shift change. Second
best blockage will detonate. TI1is will re- blasting can be done at the same time as
'ious quire the placing of another primer and primary blasting if the same immediate
of detonator in the column above the block- area is involved. However, if the blasts
age. When a void is indicated, loading are widely separated, there will be two
The must be stopped. If the void is not un- sources of flyrock to guard against.
with duly large, backfilling with stemming The reshooting of misfired blastholes
cor- material will correct the problem. can generate dangerous flyrock when there
With Another primer must be added and loading is reduced burden and reduced confinement
of the explosive charge. When reduced required, which must be cleared and
burden, distance can exceed the normal guarded. It is best to shoot misfired
flyrock range, from production blasts, blastholes during shift change or at the
and a larger-than-normal blast area is same time as primary blasts.


A misfire results when explosives fail in additional drilling, explosives, prim-

to detonate as planned during a blast. ers, detonators, and labor. In some un-
A misfire has two basic effects on an derground mines, the failure of a "cut"
operation: the safety hazard it presents hole could result in the loss of the en-
and the increase in mining costs. tire round.
The handling of boulders that require
EFFECTS OF MISFIRES secondary breakage is another cost fac-
tor. Increased digging time and greater
To most people, a misfire primarily wear and damage to equipment (especially
represents a safety hazard. With the in- bucket teeth) result in lower productiv-
creasing use of non-cap-sensitive blast- ity and higher maintenance costs. Mis-
ing agents, the possibility of accidental fires frequently are the cause of high
initiation is reduced. However, based on bottom, which results in reduced produc-
U.S. Mine Safety and Health Administra- tion and higher maintenance costs on
tion (MSHA) data, there are still many mucking equipment. In addition, haulage
injuries sustained because of misfire ac- vehicles traveling over rough terrain
cidents. In an 8-yr period (1978-85), 56 will increase the haulage cost and vehi-
misfire accidents resulted in 63 injuries cle maintenance. Often, these humps must
and 6 fatalities. The majority of mis- be drilled and blasted, which constitutes
fire accidents (75 pet) occur in under- another cost.
ground mines, with 54 pet in underground The misfiring of one hole will increase
metal-nonmetal mines. These numbers are the burden on a later hole, causing cra-
not surprising because it is not the tering with excessive flyrock and over-
total amount of explosives used, but the break. Flyrock is a leading cause of
number of shots or holes fired that personal injury and equipment damage.
provides the opportunity for misfires. Overbreak may extend beyond the burden
Also, underground mines use smaller diam- for the first row of holes for the next
eter charges, fired on smaller spacings, shot, which can cause problems in drill
which are prone to misfire. In addition, setup for the next shot. Overbreak at
visibility in underground mines is gener- the final pit wall could produce ground
ally poor, so detection of misfires is control problems with very high cost and
hampered. even loss of ore.
The remaining 25 pet of misfire acci-
dents occurred in surface mines, with 11 CAUSES AND AVOIDANCE OF MISFIRES
pet in surface coal mines and 14 pet in
surface metal-nonmetal mines. During the preparation and initiation
Misfires that result in an accident of blasts, there are many aspects that
with injuries, fatalities, and/or equip- may result in misfires. The most fre-
ment damage involve obvious costs; how-/ quently stated cause of misfires is the
ever, there are other costs due to mis- incorrect use of the initiation system, a
fires that are not obvious. Because the problem common to all initiation systems.
effects of misfires will vary greatly, A major contributing factor is the lack
operators should conduct probable cost of understanding by blasting personnel of
analyses of misfires at their mines. how the system works. Unless the blast-
Those who do will place greater emphasis ing personnel have a full understanding
on avoiding misfires. of the initiation system, even minimal
In addition to the cost of disposal of changes in a shot can result in poor
the misfire, there are the direct costs blast performance and misfires.

Damage to the initiation system or ex- Hole blockage can also be the result of
plosives column is another common source careless work habits, such as knocking
of misfires. Causes include poor work material into the hole. Rapidly loaded
practices on the blast site and rock cartridged products will often bridge,
movement that produces cutoffs. particularly in holes that are partially
Damage can often occur while stemming filled with water. Careless work prac-
is shoveled in and when wires, cords, or tices are often due to the rush placed on
tubes are stepped on. Even driving over the blasting crew to get the shot off·by
the initiation systems is not uncommon at a given time.
some mines. The wiring-in or hookup of a Poor priming practices frequently cause
shot, regardless of the initiation system detonation failure. Each product has
m- used, is a very important part of blast- minimum priming requirements. Even when
ing. It is important that each initia- proper priming is used, if the primer
.re tion system be checked after the hookup. sinks into the mud at the bottom of the
The method of system checkout depends on hole or water enters a hole loaded with
the system used. All electrical hookups non-water-resistant products, a misfire
should be checked with a bLaster's meter. can occur. Many operators of surface
All shots should be checked visually. mines using larger diameter blastholes
Good housekeeping and neat and consistent place the primer at the floor level and
hookup practices are helpful in accom- not at the bottom of the hole in the sub-
plishing the system checkout. drilled region. In the event of a mis-
Rock movement may cut off the explo- fire, the more sensitive primer and deto-
sives columns and result in misfired ex- nator are more easily retrieved.
plosives. Uplift, as well as horizontal Storage is also an important factor in
displacement, is a factor in cutoffs. An avoiding misfires, since improper storage
area surface coal mine may have 100 or can alter the performance of many prod-
more holes drilled and loaded, which are ucts. The sensitivity of some products
then divided into 4 or 5 shots. Where is dramatically reduced by low-temper-
rock movement is generally up, with lit ature storage, which can result in mal-
tle or no horizontal movement, the uplift functioning. Malfunctioning of explo-
can damage loaded holes waiting to be sives products is usually due to improper
wired in for the next shot. storage rather than quality control prob-
te• The proper use of delays is very impor- lems in manufacturing, which are rare.
tant in preventing cutoffs. The longer Hydrostatic pressure as well as compres-
the delay between holes, the greater the sion from the firing of adjacent holes
probability of cutoffs. Shorter delay has caused some products to lose sensi-
at times are needed when surface delays are tivity and misfire. If misfires are to
used. If longer delay times are required be avoided, the blaster must have a com-
md for fragmentation and rock displacement, plete understanding of the products and
in-the-hole delays should be consid- conditions under which they can be used.
ered. Many systems or combinations of This information is available from the
systems are available to meet delaying supplier.
lon Geology is another factor that plays DETECTION OF MISFIRES
1at a major role in causing cutoffs. Frac-
:e- tures, faults, joints, and bedding planes Sometimes misfires are obvious, some-
:he are all zones of weakness that may cause times they are not. Each shot must be
. a burden movement to occur in much shorter checked for misfires before mucking is
ns. times than considered normal. The use begun. When the explosive is lying on
;ck of decking and multiple primers is some- the muck, detection of a misfire is no
of times advantageous to avoid cutoffs. probln.m. However, when the explosive is
3t- Rock falling from the walls of the bore- buried in broken rock, visual detection
lng holes can cause bridging and explosive is unreliable.
nal column separation.

There are a number of clues that may difficult to relocate and check for mis- will
indicate a misfire. Most operations have fires. Because of potential misfires, prim;
standardized blasting practices that have blastholes must never be collared in of W<
fairly uniform results from each shot. A bootlegs. ter 1
change from the norm could indicate mis- Also
fires. In surface mines where the shot DISPOSAL OF MISFIRES read:
can be observed from a safe location, dis at
watching and listening to the shot is There. are two basic methods used to and (
worthwhile. A change in the sound-- dispose of undetonated explosives: to poor
louder or quieter--may indicate misfires. recover and destroy the explosives or to 'When
Ejection of stemming, cratering, and fly- detonate the misfire in place. Any ex- air <
rock may result from too much burden, due plosive product removed from a misfire is fract
to misfiring of earlier holes. considered damaged and must be destroyed creal
In surface and underground operations, in a safe manner. The manufacturer is chaq
the muckpile profile can reveal areas of the best source of information on de- muckj
possible misfires. Muck lying mostly to stroying an explosive product. Many ThE
one side of the shot, less displacement water-based products do not burn readily firir
than expected, abnormal backbreak, and and are difficult to destroy. The most
humps and valleys in the muckpile can all common method is in-place detonation.
be due to misfires. Change in fragmenta- This is good practice in underground
tion is a very good indicator of possible mines where flyrock is less of a problem. Bl.:
misfires. Boulders across the top of However, in many operations, the disposal terms
the muckpile are easy to see and could of misfires is the only blasting that is tion,
indicate misfires. Operators of loading done on-shift, and this creates blast accic
equipment should be aware that boulders area security problems. invoJ
uncovered in lower sections of the pile When the original initiation system is derst
also may be the result of misfires. This still intact, it can be used to refire resul
is common where multiple decks are used. the charge, except with cap and fuse of tt
Multiple priming can minimize misfires, blasting. When a misfire results with apprc
although this method is sometimes not cap and fuse blasting, the blaster The
used because of additional costs. Con- should never relight the fuse because it are c
sidering the cost of misfires, perhaps may have been shortened, causing unex- groun
multiple priming should be used more fre- pected premature initiation, or it may be of bl
quently. A double-trunkline or loop sys- damaged and could produce a hangfire. persc
tem must be used with detonating cord With cap and fuse blasting, repriming is (2) t
systems. Even with the two paths of det- essential. back
onation, all of the cord should be con- Misfires are occasionally refired in Ele
sumed in the blast. Finding detonating surface mines, in which case flyrock is a opmen
cord in the muck is a strong indication major consideration. Because of reduced ment
of a misfire. burden on the missed hole, violent fly- initi
In underground mines, detection is ham- rock may result. Normally, removal of plann
pered by poor lighting. It is difficult the explosive load is recommended in sur- aries
for the loader operator to spot explo- face mines. Mucking out a misfire must clear
sives in the muck. Checks for explosives be done with caution, with the minimum initi
should be made before and during the number of personnel in the area, and warni
mucking operation. Every bootleg must be under the supervision of a competent commu:
examined carefully fo~ misfires. A mis- person. and b
fire may have occurred even though there Removal of stemming and explosives from Man
is no cap legwire or tubing protruding blastholes is more difficult in vertical ' flyro.
from the hole. Lifter holes are of par- holes than in horizontal holes. Two com- 'i
ticular concern because they slope down- mon techniques used are washing material Blastl
ward and are often water filled, and it out with water or blowing it out with must 1
is easy for loose material to fall into air. When the main charge is ammonium count
them. These conditions promote misfires. nitrate and fuel oil (AN-FO), washing
In addition, lifter holes are the most with water has an advantage because water

Ls- will desensitize the AN-FO. But when re- technique that has been used. This meth-
~s, priming of the hole is planned, the use od is not recommended and should be used
in of water is a disadvantage and it is bet- only as a last resort, because the danger
ter to blow out the stemming with air. of drilling into the misfire is always
Also, air has an advantage in that it is present. In blasts with angled holes,
readily available in many mines. The such as vee cuts, or where hole length-
disadvantage of air is that it blows dirt to-burden ratios are high, this technique
to and dust into the atmosphere, creating should not be used. When it is used, re-
to poor working conditions around the hole. covery and disposal of the explosive must
to When the explosive is to be removed, both still be perfornmed in a safe manner.
ex- air and water may deposit the charge in If explosives from a misfire must be
is fractures around the borehole, which may stored, all detonators should be removed
yed create more of a problem than if the and stored separately. Explosives and
is charge is left and dug out during the detonators removed from a misfire should
de- mucking operation. not be stored with other explosives or
any The removal of an explosive charge by detonators.
ily firing a nearby charge is another
em. Blasting accidents are very costly in must be checked before loading of explo-
sal terms of human suffering, lost produc- sives, and loading must be monitored
is tion, and damage to equipment. Blasting closely so that any problems encountered
ast accident can be avoided if all persons can be corrected.
involved in the. blast have a thorough un- Most misfires are due to some problem
l is derstanding of the conditions that can with the initiation system such as fail-
ire result in an accident and an appreciation ure to make a connection, a broken lead,
use of the hazards involved, and take the or simply not understanding the initi-
•ith appropriate precautions. ation system. Other cauaes of misfires
:ter The requirements of blast site security are cutoffs, inadequate priming, and mal-
: it are common to both surface and under- functioning of the explosives due to im-
tex- ground mines. The two basic requirements proper storage.
' be of blast guarding are (I) to move mining Detection of a misfire is no problem if
.re. personnel out of range of the blast and none of the holes detonate. However, if
is (2) to prevent the movement of personnel only a few holes or portions of a single
back into the blast zone. hole fail to detonate, detection of the
in Elements that contribute to the devel- misfire can be very difficult. In these
~s a opment of safe .blast guarding are manage- cases, visual inspection of the muckpile
teed ment commitment to safety, training, def- for undetonated explosives and boulders
:ly- inition of blasting authority, preblast or other muckpile irregularities that
l of planning, definition of blast zone bound- suggest possible misfires is the most re-
;ur- aries, selection and placement of guards, liable detection method.
!lUSt clearing plans, location of the blast Disposal of detected misfires is accom-
Lmum initiation site, blasting time, blast plished by removing the explosives with
and warning signals, personnel accounting, water washing or air flushing, repriming,
:ent communication of blast location and time, and reshooting, or by detonating a nearby
and blasting crew communications. charge. However, detonating a nearby
:rom Many factors can affect the amount of charge can be very dangerous and is not
Leal flyrock produced by a blast. The blast recommended.
:om- design must be appropriate for the site. The best way to avoid misfire accidents
dal Blasthole pattern layout and drilling and costs is to eliminate their causes.
dth must be accurate and must take into ac- This can be done by knowing the charac-
'lium count blast-site conditions. All holes teristics of the explosives, delays, and


initiation system, by proper blasting de- and by good housekeeping practices at the
sign, by taking care in loading the shot blasting site.
and hooking up the initiation system,


1. D'Andrea, D. V., and J. Bennett. Conference on Explosives and Blasting

Safeguarding of Blast-Affected Areas. Technique, ed. by C. J. Konya (Dallas,
Paper in Proceedings of the lOth TX, Jan. 31-Feb. 4, 1983). Soc. Explos. b:
Conference on Explosives and Blasting Eng., Montville, OH, 1983, PP• 123-132. at
Technique, ed. by c. J. Konya (Lake 4. Institute of Makers of Explosives r:
Buena Vista, FL, Jan. 29-Feb. 2, 1984). Safety Library (Washington, DC) •. Safety ffi!
Soc. Explos. Eng., Montville, OH, 1984, Guide for the Prevention of Radio Fre- s:
PP• 110-119. quency Radiation Hazards in the Use of cl
2. Fletcher, L. R., and D. v. D'Anrea. Electric Blasting Caps. Pub. 20, Oct. p:
Control of Flyrock in Blasting. Paper in 1978, 20 PP• Ot
Proceedings of the 12th Conference on Ex- 5. E.I. duPont de Nemours & Co., Inc. gt
plosives and Blasting Techniques, ed. by (Wilmington, DE). Blaster's Handbook. Ol
C. J. Konya (Atlanta, GA, Feb. 9-14, 16th ed., 1978, 494 pp. fit
1986). Soc. Explos. Eng., Montville, OH, 6. Dick, R. A., L. R. Fletcher, and t
1986, PP• 167-177. D. V. D'Andrea. Explosives and Blast-
3. ___ A Study of Misfires in Min- ing Procedures Manual. BuMines IC 8925, m
ing. Paper in Proceedings of the 9th 1983, 105 pp.





By Michael A. Peltier,1 Larry R. Fletcher,2 and Richard A. Dick3

,s, The Bureau of Mines has developed a a site-specific blasters' training pro-
•S. blaster's training manual for the metal gram. Each module contains text material
and nonmetal mining industry. The mate- that comprehensively covers the topic, as
·es rial is divided into 6 chapters and 47 well as a paraphrased section highlight-
ty modules, with each module covering a ing the major ideas of the text. Also
e- single topic. (For example, the second included with each module are line draw-
of chapter, which deals with initiation and ings and test questions with answers.
t. priming, is subdivided into nine modules. The objective of this material is to
One module covers initiations systems in increase hazard awareness and foster the
c. general, another covers delay series, and use of safe blasting practices, with the
:<.. one discusses priming. The remaining six anticipated end result being accident-
modules deal with each of the six initia- free and productive blasting.
... _ tion systems.)
The modules were structured to enable
5, mine training personnel to easily develop


Based on accident data obtained from property is based on experience at that

the U.S. Mine Safety and Health Adminis- mine and is done without the aid of ade-
tration (MSHA), most blasting accidents quate training materials. An improved
are caused by human error, lack of hazard and more meaningful blasters' training
awareness, or lack of general blasting program is essential in assisting opera-
knowledge. A lack of understanding as to tors to properly train blasters and meet
how explosives function can contribute to MSHA training regulations.
higher mining costs because of inadequate The blasters' training material was de-
fragmentation or lost production. veloped to · aid industry in the prepara-
Federal regulations require that every tion of a site-specific training course
person who uses or handles explosive ma- and is based on a previous Bureau of
terials be experienced and understand the Mines Information Circular titled "Explo-
hazards involved. Trainees should do sives and Blasting Procedures Manual".4
such work only under the supervision of The intent is to help individuals using
and in the immediate presence of experi- explosives and blasting agents to develop
enced miners. Federal regulations also a better understanding of the various as-
require hazard and task training for min- pects of blasting that contribute to a
ers. Most training given on mining safe and efficient blast.


The blaster's training manual has been blasters course. The material consists
constructed to be easily used in develop- of discrete modules that contain text ma-
ing a site-specific and comprehensive terial, a paraphrased section, line draw-
ings, and test questions with answers.
1 Mining engineer.
2 Mining engineer technician. 4o·l.C k , R. A., L. R. Fletcher, an d
3staff engineer. D. v. D'Andrea. Explosives and Blasting
Twin Cities Research Center, Bureau of Procedures Manual. BuMines IC 8925,
Mines, Twin Cities, MN. 1983, 105 pp.

Individual pages have been dividen !--Explosives Products" section of the

lengthwise with the comprehensive text checklist, ammonium nitrate and fuel oil is
material on the left-hand side of the (AN-FO) and emulsion maybe noted next to bl
page. Each paragraph of the text mate- the subsection "Blasting Agents." By
rial is numbered for quick reference. reading the list of modules in the table ·~ pr
The right-hand side of the page consists
of paraphrased text material with a main
heading and a paragraph number. The
of contents under the "Chapter !--Explo-
sives Products" section, the trainer
will notice that module 4 discusses AN-FO
il O'f

person preparing the training course can and module 5 discusses emulsions. ~
read the paraphrased material quickly in The second step is to gather the train- ·~ pJ
order to grasp the main ideas of the text ing material needed. The information ·~ pi
material. If an explanation is needed, gathered from blasting personnel through '~
the individual, by noting the paragraph
number, can go directly to the paragraph
the use of the checklist will indicate
which modules should be included in the .t
discussing a particular point. course. In addition to the modular mate-
Line drawings are included with the ma- rial, slides and other visual aids from
terial to illustrate specific concepts. the actual operation should be used. Ad-
The line drawings can be easily converted ditional technical information concerning
to overhead transparencies for use in the specific blasting products can be ob-
training course. tained from either the explosives sup-
The first step in preparing a blasters plier or manufacturer.
training course is to determine what ma- The third step is to write lesson plans
terial must be covered. This can be ac- for the course and arrange the training
complished by talking with the blasting material into a cohesive unit. The writ-
supervisor and blasters, and by observing ing of the lesson plans can be simplified
the blasting operation. To help deter- by making extensive use of the para-
mine what topics need to be covered in phrased sections in the modules.
the course, a checklist is included with Since the experience, knowledge, and
the material. It is arranged to parallel ability of individual blasters vary
the chapters. By completing the check- widely, both the length and amount of ma-
list, the trainer will be able to lo- terial to be included in the course will 1
cate the modules to be included in the have to be determined by mine
course. For example, under the "Chapter management.



Purpose and Description Upon completion of. this chapter, the

blaster should be able to
The purpose of this chapter is to help
the blaster develop an understanding of 1. Give a concise explanation of the
various types of explosives. Chemical nature of various explosive products;
and physical properties of seven types of 2. List the basic reactive ingredi-
explosive products are discussed. Addi- ents of an explosive product;
tional information explains nine proper- 3. Explain how the detonation pres-
ties of explosives that are used to de- sure and explosive pressure cause the
termine how an explosive product will rock to be broken;
function under field conditions. Mate- 4. Explain the importance of oxygen
rial explaining how to select an explo- balance as it relates to both the energy
sive product is included in this released and the formation of toxic
chapter. gases;

f the 5. Describe the individual character- 1. Name the three basic parts of an
l oil istics of the explosive products the initiation system;
'<t to blaster may be using; 2. Explain the difference in sensi-
By 6. Briefly explain why a particular tivity to initiation between high
.able product is being used at the blaster's explosives and blasting agents.
rplo- operation; 3. State the difference between an
.ner 1. State and explain nine basic prop- instantaneous and a delay detonator;
\N-FO erties of explosive prodncts; and 4. List the various components of the
8. Relate the basic properties of ex- initiation system the blaster will be
:ain- plosives to the types of explosive using;
~tion products being used on the job. 5. Explain how the initiation system
·ough functions;
cate Chapter Modules 6. Explain how to check the final
the hookup of the system;
,ate- Module Title 7. Discuss the potential hazards to
from the initiation system;
Ad- 1 •••••••••••• Chemistry and Physics of 8. Give the definition of a primer;
ning Explosives. 9. Name some types of explosives used
ob- 2 •••••••••••• Types of Explosives and as primers;
sup- Blasting Agents. 10. Explain the proper procedure for
3 •••••••••••• Nitroglycerin-Based High making primers; and •
lans Explosives. 11. Explain why the proper location of
:ling 4..••..••.••• Ammonium Nitrate-Fuel Oil the primer in the borehole is important.
:-it- (AN-FO).
=ied 5 •••••••••••. Slurries, Water Gels, Chapter Modules
1ra- Emulsions.
6............ Heavy AN-FO. Module Title
and 1 •••••••••••• Primers and Boosters.
1ary 8 •••••••••••• Liquid Oxygen Explosives. 12 •••••••••• Initiation Systems.
ma- 9 •••••••••••• Black Powder. 13 •••••••••• Delay Series.
rill 10 •••••••••••• Properties of Explosives. 14 •••••••••• Electric Initiation.
1ine 11 •••••••••••• Explosives Selection 15 •••••••••• Detonating Cord Initiation.
Criteria. 16 •••••••••• Detaline Initiation System.
17 •••••.•••• Cap-and-Fuse Initiation.
CHAPTER TWO--INITIATION AND PRIMING 18 •••••••••• Hercudet Initiation.
19 •••••••••• Nonel Initiation.
Purpose and Description 20 •••••••••• Priming.

the The purpose of this chapter is to help CHAPTER THREE--BLASTHOLE LOADING

the blaster develop an understanding of
six initiation systems. The blaster Purpose and Description
the will learn the various components of each
initiation system, how each individual The purpose of this chapter is to exam-
di- system functions, and the advantages and ine proper blasthole loading techniques.
disadvantages of the six systems. Infor- The chapter discusses loading procedures
es- mation about the two basic delay series for both small- and large-diameter
the and material concerning priming are also blastholes. Also included in the chapter
included in this chapter. is material that discusses how to check
gen blastholes for proper depth, water,
r:gy Objectives voids, and obstructions, and how to miti-
de gate these problems.
Upon completion of this chapter, the
blaster should be able to

Objectives Objectives

Upon completion of this chapter, the Upon completion of this chapter, the
blaster should be able to blaster should be able to

1. Explain why blastholes should not 1. Discuss how geology affects

be loaded and workers should retreat from fragmentation;
the blast area during the approach or 2. Name the most significant geologic
progress of an electrical storm; features to consider when designing a
2. Describe how to check the borehole blast;
for proper depth, obsructions, water, and 3. Discuss the importance of a well-
voids; detailed drilling log;
3. Explain how to remedy problems, 4. Explain how to determine the
such as improper borehole depth, obstruc- burden;
tions, water, and voids; 5. Explain why geologic structure is
4. State why stemming is _important and the major factor in determining blasthole
how to estimate the amount of stemming diameter;
needed; 6. Explain how collar distance affects
5. Explain when plastic borehole fragmentation size;
liners or water-resistant cartridges 7. Explain the relationship of collar
should be used; distance to airblast and flyrock;
6. Explain the proper technique for 8. Explain the relationship between
loading the explosive or blasting agent burden flexing and rock fragmentation;
the blaster will be using. 9 Discuss the problems of either ex-
7. Describe the characteristics of the cessive or insufficient subdrilling;
type of pneumatic loading the blaster 10. Explain how spacing is determined;
will use; 11. Explain the advantages of milli-
8. Explain the potential problem of second delays;
static electricity if the blaster is go- 12. Discuss the two classifications of
ing to use a pneumatic loader; and opening cuts;
9. List the advantages and disadvan- 13. Explain how to design an angled
tages of using bulk-loaded products in cut;
large-diameter blastholes. 14. Explain how to design a parallel-
hole cut;
Chapter Modules 15. Discuss the two types of delays for
underground blasting;
Module Title 16. Name the two main advantages of us-
ing controlled blasting;
21 •••••••••• Introduction. 17. List the four primary methods of
22 •••••••••• Checking the Blasthole. controlled blasting; and
23 •••••••••• General Loading Procedures. 18. Discuss the advantages and disad-
24 •••••••••• Small-Diameter Blastholes. vantages of the various methods of con-
25 •••••••••• Large-Diameter Blastholes. trolled blasting.


Purpose and Description Module Title

The purpose of this chapter is to exam- 26 •••••• Introduction to Blast Design.

ine the factors that influence safe and 27 •••••• Properties and Geology of the
effective blast design. In addition to Rock Mass.
the discussion of design factors for sur- 28 •••••• Surface Blasting.
face and underground blasting, four con- 29 •••••• Underground Blasting.
trolled blasting techniques are also 30 •••••• Controlled Blasting Techniques.


Purpose and Description
Purpose and Description
The purpose of this chapter is to help
The purpose of this chapter is to the blaster develop a better understand-
examine the environmental effects of ing of blasting safety, by examining a
blasting. The material will discuss fly- number of auxiliary blasting functions.
rock, ground vibrations, airblasr-., and A number of precautions related to
dust and gases. Methods to reduce the previous modules are mentioned. Four ac-
potential health and safety hazards they cident types that occur frequently are
may present will be discussed. also discussed.

Objectives Objectives

Upon completion of this chapter, the Upon completion of this chapter, the
blaster should be able to blaster should be able to
1. Explain why a knowledge of all cur-
1. Explain the importance of conduct- rent blasting safety regulations is im-
ing a preblast survey, maintaining portant;
comprehensive records, and good public 2. Name the agencies that regulate and
relations; enforce the use and storage of explosives
2. Discuss the causes of flyrock; and blasting agents;
3. Discuss methods to alleviate 3. Describe the requirements for ve-
fly rock; hicles used to transport explosives and
4. Discuss the causes of ground blasting agents from the magazine to the
vibration; job site;
5. Discuss design techniques to mini- 4. Explain the importance of marking
mize vibrations; the blast area and keeping nonessential
6. State some methods to monitor personnel away;
ground vibrations; 5. Explain when to check for extrane-
7. Discuss the causes of airblast; ous electricity;
8. Discuss methods to monitor 6. Discuss why electrical storms are a
airblast; hazard regardless of the type of initia-
9. List techniques to reduce airblast; tion system;
10. Explain why an adequate amount of 7. Explain the importance of proper
time must be given for dust and gases to primer makeup;
be diluted before returning to the blast 8. List a number of checks to be made
site; and before borehole loading begins;
11. List the two common toxic gases 9. Describe various methods to check
produced by blasting and list techniques column rise during borehole loading;
to reduce them. 10. Describe some precautions to con-
sider before and during the hookup of the
Chapter Modules shot;
11. Explain some good methods for blast
Module Title area security;
12. Describe the potential hazards to
31 ••••••••• Introduction to Environmen- check for when reentering the blast site
tal Effects of Blasting. after the shot has been fired;
32......... Flyrock. 13. Discuss methods for disposing of
33 ••••••••• Ground Vibrations. misfires; and
34......... Airblast. 14. Discuss the principal causes of
35......... Dust and Gases. blasting accidents.

Chapter Modules Module Title

Module Tit 41.......... Borehole Loading.

42 •••••.••.• Hooking Up the Shot.
36 •••••••••• Introduction to Blasting 43 •••••••••• Shot Firing.
Safety. 44 ••.•••••.• Postshot Safety.
37 •••••••••• Explosives Storage. 45 •••.•••••• Disposing of Misfires.
38 •••••••••• Transportation From Maga- 46 •••••••••• Disposal of Explosive
zine to Job Site. Materials.
39 •••••••••• Precautions Before Loading. 47 ••••••••.• Principal Causes of Blast-
40 •••••••••• Primer Safety. ing Accldents.


A training manual for metal and Supplementing the modules are a 73-item
nonmetal mining has been developed by bibliography, a list of regulatory au-
the Bureau. This program consists of 47 thorities and their responsibilities,
modules or topics under 6 major headings additional information on HSHA and OS~1
(chapters). The modules consist of a (U.S.Office of Surface Mining), a glos-
text and outline on a single blasting sary, and 65 illustrations suitable for
topic, plus questions and answers. duplication.


By Vi 1 J. Stachura1 and Larry R. Fletcher2


The Bureau of Mines conducted delayed 100 ms longer than nominal, and 50 and
blasting experiments at a contour coal 100 ms longer in the two rows of holes
mine, which were designed to reduce over- nearest to the highwall. The mine's nom-
break without special drilling or signif- inal blast design was a flat V-pattern
icant additional costs. In the standard with 17-ms surface delays between holes,
layout of the blast pattern at this mine, 42-ms surface delays between rows, and
the ends of the rows formed the highwall. ZOO-ms in-the-hole delays in each hole.
Overbreak was reduced by increasing the All three test designs produced highwall
delays on the last row of holes at the improvements, compared with results using
highwall, which changed the effective de- the nominal design, with occasional ex-
lay pattern geometry and the direction of ceptions because of geologic variations.
burden movement. These experiments re- Observations and terrestrial photogram-
sulted in smoother highwalls, which were metry showed that the delay changes pro-
also inherently safer because of the re- duced generally smoother vertical pro-
duced likelihood of rockfall. files with less loose material.
Three delay combinations were tested:
50 ms longer than the nominal design,


One of the major hazards found in sur- Inc. evaluated blasting practices at nine
face mining is rockfall from highwalls. contour mines (2). Almost all the mines
This hazard occurs in all forms of exca- visited had highwall instability prob-
vation in rock, especially where explo- lems that were aggravated by poor blast
sives are used. The explosive energy not ing -practices. Engineers International
only fractures the rock to be excavated conducted eight test blasts that demon-
but also damages the rock that borders strated that good blasting practices did
the excavation. This reduces the stabil- improve highwall stability. However, af-
ity of the highwall and increases the po- ter the tests were complete and the con-
tential of rockfall. The rockfall hazard tractor was off the site, the mine per-
is normally attributed to blasting prac- sonnel reverted to their old blasting
tices, geologic conditions, and adverse practices.
weather in 65 pet of accidents resulting To avoid a similar result, the tests in
from fall of rock (1).3 Of these three this report emphasize simple, easily un-
factors, only blasting is controllable, derstood changes that minimize economic
and therefore, blasting was the subject and procedural impact and maximize opera-
of this investigation. tor acceptance. The experiments are di-
In earlier research sponsored by the rected at reducing overbreak without spe-
Bureau of Mines, Engineers International cial drilling or significant additional
cost. They use simple changes in blast-
1Geophysicist. hole initiation timing, which improve
2Mining engineering technician. relief by changing the direction and
Twin Cities Research Center, Bureau of time of burden movement. In this report,
Mines, Minneapolis, MN. overbreak is defined as excessive break-
3underlined numbers in parentheses re- age of rock beyond the desired excavation
fer to items in the list of references at limit (3).
the end of this paper.


The approach selected for devising the were used in the highwall holes (figs.
experimental blast designs was to take 2-3). Design 4, another Bureau design,
the design in use at the mine site and was the same as design 1 except that 300-
make minor delay-period changes. No oth- ms in-the-hole delays were used in the
er parameters were changed intentionally, highwall holes and 250-ms in-the-hole de-
but in the mine's normal blasting proce- lays were used in the second row of holes
dures the accuracy of spacings and bur- from the highwall (fig. 4). Figures 1
dens varied, more than one hole diameter through 4 show the cumulative delay times
was occasionally used, and powder column for each blast hole and include arrows to
heights also varied. The mine where the indicate the observed direction of burden
tests were conducted used a Nonel4 (also movement. In figures 2 through 4, this
called shock tube) initiating system, so direction is perpendicular to the plane
the original delay times reflected those of the highwall, a sign of improved re-
available using that system. lief over that illustrated in figure 1.
The blast design used by the mine, de- Design 4 was tried because of reports of
sign 1, was a flat V with surface delays overbreak extending far beyond a distance
of 17 ms between holes, 42 ms between equal to one burden Ci-2)· It was antic-
rows, and 200-ms in-the-hole delays (fig. ipated that the individual delay in the
1). Designs 2 and 3, experimental de- second row of holes from the highwall
signs used by the Bureau, were the same would provide additional relief, reducing
as design 1 except that in-the-hole de- the damage to the highwall.
lays of 250 and 300 ms, respectively,


To determine a criterion for evaluating Typical examples of individual profiles

the test blasts, discussions were held from two different highwalls are shown in
with u.s. Mine Safety and Health Admini- figures 5 and 6. Figure 5 shows a large
stration (MSHA) inspectors, mine super- ledge at the top_ of the test area, a re-
intendents, safety officers, and blast- sult of shot design 1. Figure 6 shows
ers. The general theme found in these an adjacent highwall, which resulted from
discussions was that a smooth highwall of shot design 2 and has a much smoother
competent appearance is a safer one. A profile. Figures 7 and 8 show a test
competent appearance is achieved by re- highwall and an array of profiles that
ducing overbreak. In addition to visual are the result of shot design 1. Fig-
inspection criteria, stereo photography ures 9 and 10 show a test highwall and an
techniques were used to analyze the array of profiles that are the result of
blasting effects in this study. The use design 2. The photographs (figs. 7, 9)
of highwall profiles generated from ste- were used to evaluate amounts of loose
reo photographs provided for more analyt- material and the appearance of compe-
ical and consistent comparison of high- tence. The arrays of profiles (figs. 8,
walls than was possible with visual 10) illustrate the degree of overhanging
inspection alone. The details of the material and the size of ledges.
procedure used to take and analyze stereo A total of 59 test blasts and their re-
photographs may be found in RI 8916 and sulting highwalls were evaluated by vari-
RI 9008 (!!._-]_). ous combinations of stereo photographs,
on-site observation, photographs taken
ference to specific products floes by the blaster, and notes made by the
not imply endorsement hy the Bureau of blaster.


0 0 0 0 0 0 0 0
0 0 0 0 0 0 0
l I
I 0 0 0 0
0 0 0 0 0 0 0 0
~Ne,l section
0 0 0 0
~Nex~ section
of hiqhwoll of hHJhwol!

0 0 0 0 0 0 0 0 I 0 0
343 326
360 377
4~1 I
343 326 343 360 377 394 411
0 0
30! 84
352 3~91
I 0 /
284 352

I 0
0 0 0
3~71 Note: Number.s by holes are
0 0
259 242 259
259 242 j Note: Number~ by holes ore
cumulative de loy limes. cumvlollve de loy times.
0 0
1 Auow 1t1dicotes
burden movement
0 0
I Affow lndicoles
burden movement.


FIGURE 1.-Design 1: break line prior to 285 ms with 200-ms FIGURE 2.-Design 2: break line prior to 335 ms with 250-ms
in-the-hole delays throughout. in-the-hole delays in highwall holes and 200-ms in-the-hole delays
in remaining holes.

0 0 0
0 0 0
" 0 0 0
444 461
0 0 A 0 0 0
5~ ~Ne<l seclton 0

0 o I
r-Nexl seclton
3/3:8~-~:2- 419
3J of hiqhwoll 486 553 of hiqhwoH
385 3:8 385 402 ·- 419

343 326 343 360 31 0
343 326 343 360 377
0 0

335 352
o I 0 0
318 335

0 0 0 0 o I 0 0 0 0 0 0 0 I
259 242 276 310 4271 Note: Numbers by holes.ore
cumulohve delay t1mes.
259 242 259 276 293 360 427
I Noie: Numbers by holes ore
cumulolive deloy limes.
0 0
J Arrow indicates 0 0 0 0 0 0 --o I Arrow mdit;;Oies
268 --· 3 ~ 5
0 0 0 0 burden movement
217 bu(den movement 217 200 217 234 251 318
200 217 234 251 3851

FIGURE 3.-Design 3: break line prior to 385 ms with 300-ms

FIGURE 4.-0eslgn 4: break line prior to 385 ms with 250· and
in-the-hole delays in highwall holes and 200-ms in-the-hole delays 300-ms in-the-hole delays in two rows of holes nearest to highwall
in remaining holes. and 200-ms in·the-hole delays In remaining holes.


The results demonstrated that greater were either 50 or 100 ms longer than
highwall stability can be achieved by those previously used in the delay pat-
changing the burden movement to a direc- tern. Another variation was to increase
tion closer to perpendicular to the plane the delay time by 50 ms in the second
of the highwall and by allowing more time row of holes from the highwall and by
for movement of the burden in front of 100 ms in the highwall row of holes. The
the highwall blastholes. The changes in lengthened delays allowed more time for
blast design described in this report can the burden to move, thereby reducing
be implemented without increased costs or overbreak, which causes irregular and
technical complication. unstable highwalls. At the test sites
The burden movement was redirected by used, the 50-ms-longer delays or a combi-
using delays in all highwall holes that nation of 50- and 100-ms longer delays

90~~--~----------------~ 90~----~----------------~

-Coal -Coal

Test Test
area area

0 50 0 50
FIGURE 5.-Profile from stereo analysis, no additional delays. FIGURE 6.-Profile from stereo analysis, 50-ms-longer delays
(Numbers indicate distance in feet.) in highwall holes. (Numbers indicate distance in feet.)

worked better than when the delay time heights varied, the geology changed con-
was increased by 100 ms only. This was tinually (because the test site was a
because of the better shearing action contour mine), and scaLing practices var-
obtained with the 50-ms incremental in- ied. During the course of study, the
crease. Since the test mine had one par- safety officer and the mine operator both
ticular geology and used the Nonel ini- observed that the highwalls were notice-
tiating system for blasting, other mine ably improved and required less cleanup
sites and initiating systems may require time by the dozers used to scale the
a different adjustment of the delay time highwalls at the test site. Precise
to obtain optimum results. The reader evaluation of the blast effects on the
should also note that, for these tests, highwalls proved to be difficult because
the highwall was located at the ends of of the variables mentioned above. How-
the rows rather than the back row of ever, after observing the results of 59
holes. test blasts, the authors recommend an in-
The test results showed a general im- creased delay of 50 to 100 ms in the
provement in the highwalls even though highwall holes, as described in the
drill-hole alignment and powder column "Blast Designs" section.

FIGURE 7.-Test highwall, no increase in delay time (blast design 1).

Test highwall

FIGUREs.-Test highwall profiles, no increase in delay time.


FIGURE 9.-Test highwall, 50-ms-longer delays in highwall holes (blast design 2).

Test highwall

20 yl

10~\'> z ')_0
5 - .,

····--'-----'-----. X
0 5 10 15 20

FIGURE 10.-Test highwall profiles, 50-ms-longer delays in highwall holes.



1. Theodore Barry and Associates. Design and Ground Control, Reno, NV, Nov.
Industrial Engineering Study of Hazards 12-13, 1984, 10 pp.; available upon re-
Associated With Surface Coal Mines (con- quest from V. J. Stachura, BuMines,
tract H0230004). BuMines OFR 48-74, Minneapolis, MN.
1974, 265 pp.; NTIS PB 235 927. 5. Holmberg, R., and K. Maki. Case
2. Kendorski, F. S., and M. F. Dunn. Examples of Blasting Damage and Its In-
Safety and Cost Benefits From Improved fluence on Slope Stability. SveDeFo,
Highwall Blasting Practice (contract Stockholm, Sweden, Rep. DS 1981:9, 1981,
H0282011, Eng. Inst. Inc.). BuMines OFR 20 PP•
54-82, 1981, 169 PP•; PB 82-205691. 6. Stachura, V. J., and L. R. Flet-
3. Dick, R. A., L. R. Fletcher, and cher. Delayed Tests To Improve Highwall
D. V. D'Andrea. Explosives and Blast- Stability--A Progress Report. BuMines
ing Procedures Manual. BuMines IC 8925, RI 8916, 1984, 24 pp.
1983, 105 PP• 7. Stachura, V. J., L. R. Fletcher,
4. Hoek, E. Impact of Blasting Damage and M. A. Peltier. Delayed Blasting
on the Stability of Rock Structures. Stages To Improve Highwall Stability--
Pres. at 2d Annu. Workshop, Generic A Final Report. BuMines RI 9008, 1986,
Miner. Technol. Centr., Mine Systems 12 PP•



By MarkS. Stagg1 and David E•. Siskind2


The Bureau of Mines studied the prob- superstructure, an increase in the rate
lems of blasting-vibration-induced struc- of·cracking is not likely to result·from
tural response and cracking of low-rise blasts generating vibrations of less than
residential structures in a series of re- 0. 5 in/s·~ Data on cracks in masonry
search projects between 1976 and 1983. walls suggest that blast-induced vibra-
This paper summarizes the published tion levels of up to 3.0 in/s may be a
Bureau findings and presents them from threshold for local block-length cracks.
the point of view of the cracking and However, additional data are needed to
failure of the construction materials quantify vibration level effects neces-
used for homes. sary to generate stair-stepped cracks in
The damage data suggest that, for masonry walls, which indicate loss of
plaster and wallboard attached to the shear load capacity.


Ground vibrations from blasting have This paper summarizes the material on
been a continual problem for the mining cracking of construction materials used
industry, the public living near the min- in low-rise residential structures; the
ing operations, and the regulatory agen- data are excerpted from two comprehensive
cies responsible for setting environmen- Bureau vibration studies, RI 8507 on dy-
tal standards. Since 1974, when the namic response and damage and RI 8896 on
Bureau of Mines began to reanalyze the fatigue and long-term influences (1-~).
blast damage problem, several field and Specifically, the paper discusses the
laboratory studies have been conducted; cracking of plaster, wallboard, and mas-
the results of the most recent were pub- onry from blasting and other influences,
lished in RI 8969, in 1985 (l).3 The giving an overall perspective to the
studies examined blast vibrations with blast vibration impacts as part of the
respect to generation, propagation, total lifetime dynamic load for such
structural response, cracking potantial, materials.
instrumentation, and fatigue (2-4). A
similar series of studies was conducted
for airblast (1-i).


Current residential construction prac- nonstructural minor or cosmetic cracks.

tices address basic human safety and Many of these practices were derived from
not specifically the occurrence of allowable deflection criteria, in which
material cracking potential is considered
1civil engineer.
(7-9). However, cosmetic cracks do de-
2supervisory geophysicist.
velop, and in 1948, Whittemore (10) dis-
Twin Cities Research Center, Bureau cussed the lack of guidelines for-vibra-
of Mines, .Minneapolis, MN. tions of floors and pointed out that
Underl~ned numbers in parentheses re- "deflection and vibration can be de-
fer to items in the list of references creased, but only at an increase in
at the end of this paper. price."


PREVIOUS STUDIES (!l) in a study of 43 single-story con-

crete block houses over a 26-week period;
Structures crack naturally over time. he reported a crack rate of 2. S cracks
Holmberg (11) analyzed blasting inspec- per day for the 43 houses (<1 crack per
tion reports-to estimate a crack rate for week per house).
apartment buildings in Sweden. Two The large variation in the crack rates
apartment buildings were inspected for reported in the separate studies by Holm-
cracks three times between 1968 and 1980. berg, Andrews, and Wall is indicative of
The number of observed cracks is plotted the wide variation of susceptibility of
as a function of time in figure 1. An houses to cracking. The rates ranged
average of 12 to 13 new cracks per year from near zero to 23 cracks per week.
occurred for these particular structures. (The yearly rate reported by Holmberg in-
Holmberg did not report any specifics on dicates a cracks-per-week rate of less
the building construction, although con- than one.) None of the investigators re-
crete is a reasonable assumption. ported rates of zero. The la~ge differ-
The crack rate depends upon the type of ences in the rates reported are partially
structure. Rates for 11 wood frame a result of the difficulty of defining
houses that were subjected to 26 weeks of "cracks". For example, in Wall's report,
sonic booms and 13 weeks when there were shrinkage cracks were ignored, and only
no booms, as reported by Andrews (~), new cracks in the moderate (easily dis-
are listed in table l. Crack rates at tinguishable) range were reported.
homes 1 through 4, which were studied These data point out that when months
during both periods, were generally lower pass between preblast and postblast in-
during the 13-week nonboom period, which spections, any postblast inspection is
is similar to Bureau findings discussed likely to find some new cracks that are
later. The investigators also found evi- the result of natural aging.
dence of the possibility that relative
humidity and th~ number of sonic booms BUREAU LONG-TERM FATIGUE STUDY
may together have harl an effect on the
occurrence of cracks. Blast effects on long-term crack rates
The rates of 1.4 to 23 cracks per week were monitored over a 2-yr period at a
during the nonboom period are quite high Bureau of Mines test house <i>· Bureau
compared with the rate observed by Wall researchers developed two types of data
in terms of the expected damage mecha-
200 I I
nisms: (l) fatigue damage from accumu-
KEY 8 lated exposure, assessed by periodic in-
o Apartment house I spections, and (I) triggering effects of
o Apartment house 2 discrete blast events assessed by inspec-
(/) tions immediately before,, and after
(..) blasts, where the strains from blasting
0:: are added to already existing environ-
(..) 0
mental strains. Researchers found that
0 100 f- - long-term repetition of the low-level
w blasts (peak particle velocity <0.5 in/s)
~ produced no significant effect; however,
z blasts with velocities greater than about
1.0 in/s were associated with higher
0 cracking rates, as shown in table 2.
The crack rate, or number of new cracks
0 I I per inspection, along with the number of
1965 1970 1975 1980 blasts that produced ground vibrations
FIGURE 1.-Building age versus crack occurrences, after greater than 0.50 in/s and greater than
Holmberg ( 11). 1. 0 in/s, is shown in figure 2. Sixty

~ABLE 1. - Crack rates for houses subjected to sonic booms, after Andrews (~).

Number Number of cracks

louse of Area, Foundation Age, Finish Occu- per week
stories ft2 yr Interior Exterior pied Boom Non boom
period period
1 • •• 1 1,560 Concrete 5 Wallboard •• Brick ••• Yes •• 3.7 1. 9
2 • •• 2 1,750 • •• do • ••••• New • • • do • ...•. .•• do ••• No ••• 8.2 3.3

3 • .. 1 1' 4 70 • •• do • ••••• 8 • • • do •••••• . • • do ••. No ••• 8.8 1.5

4 • •• 1 1,160 Concrete 18 ••• do • ••••• • •• do • •• No ••• 6.1 1. 8

stem wall.
5 • •• 2 2,870 Masonry >50 Plaster Asbestos No ••• NM 23
stem wall. and lath. siding.

6 ••• 1 1, 100 Concrete 25 • •• do •••••• Stone ••• Yes •• NM 2.6

stem wall.

7 ••• 1 1, 090 ••• do •••••• 30 Lath and Wood lap Yes •• NM 1.4
8 . •. 1 1,280 • .. do • .•.•• 30 Plaster Brick ••• Yes •• NM 3.3
and lath.
9 • •• 2 2,000 Masonry 40 Paper on Wood lap Yes •• NM 3.0
stem wall. plaster
and lath.
o. .. 2 2, 3 70 Concrete 35 Plaster . . • do • •• Yes •• NM 14
stem wall. and lath.
. 1 • •• 1 1, 330 Concrete 8 Wallboard •• Brick ••• Yes •• NM 2.2
rM Not measured. -
human activity. The data that exclude
TABLE 2. - Crack versus vibration (~) corner cracks are more realistic indica-
tions of blasting influences for homes
,last level, in/s Cracks per week other than new construction, i.e., within
Without corner Total 6 months.
0.5............. 0.28 0.84 Differences were found in the number of
-0.5, <1.0....... .33 .89 cracks observed by the two teams of in-
1.0............. 1.0 1.8 spectors, Vibration Measurement Engineers
(VME) and Bureau personnel, during peri-
hots had levels between 0.5 and ods 1, 15, and 36. The most pronounced
.0 in/s, while 48 shots had levels above difference was for period 15. The deci-
in/s. Some of the crack rates shown in sion to include small corner cracks was
igure 2 include small hairline corner made after VME had completed its inspec-
racks, and some do not. The majority of tion for that period but before the Bu-
orner cracks occurred in the first 8 reau had completed it inspection for pe-
onths. Cracks were found in nearly riod 15. Other than for that period,
very corner in the house, but were ig- differences in the number of cracks ob-
ored until inspection period 15. Then served were an inevitable consequence of
t was decided to observe them rigorously the difficulty of observing hairline
espite their minuscule size. Corner (0.01- to 0.1-mm) cracks. Periods 1, 15,
racks are an inevitable consequence of and 36 were omitted in calculations of
he curing of the tape compound and are crack rates. However, periods in which
nhanced by dynamic strains induced by there were unusual external influences,

lL ·- !5


lL <=
0 ·-
w 1\
zo 5
25 57

- Including corner crocks
- - - E<cluding corner crocks
a:: 8 Bureau of Mines present
0 15
lL during inspection
0 0 Earthquake, N E. Kentucky
a:: E Earthwork


FIGURE 2.-Number of cracks and blasts >0.50 in/sand >1.0 in/s versus inspection period.

including an earthquake and soil removal about 1.0 in/s. The low crack formation
by a scraper 40 ft from the test house, rates reported are reasonable since the
were included. The self-triggering seis- test house was new, showed no differen-
mograph recorded a 0.06-in/s vibration tial settlement, and was not regularly
for the scraper activity but did not occupied. These conditions resulted in
trigger during the earthquake. low rates of natural crack formation,
The increase in crack rate with ground which allowed a few blast-related cracks
vibration level indicates that the blast- to significantly affect crack formation
ing produced a triggering strain, at rates.
Cosmetic cracks result when a dynamic- up to failure (cracking). Most mate-
induced strain (blast vibration or other rials, including masonry, plaster, and
transient vibration) added to a preexist- wallboard, respond linearly up to the

.. ing strain (static load) exceeds the

strain level necessary to initiate a
initial yield point. A linear response
means that deformation (strain) is di-
crack. Differential foundation settle- rectly proportional to load (stress).
ment, excessive structural loads, and ma- Beyond initial yeild, plastic deformation
terial shrinkage all induce strains that or creep occurs until ultimate failure
can produce random and/or patterned (fig. 3). The yield point damage is of-
cracking. For analyzing blasting ef- ten not visually noticeable because of
fects, these strain-inducing forces are limited naked-eye resolution of 0.01 to
considered static and the resulting 0.1 mm, particularly in textured surfaces
strains are called "prestrains." such as masonry.
Stress-strain curves are used to de-
scribe response of materials under load
36 '1

Initial paper
failure '~
(\) Visual
c:: paper
....... 1
~ 75
0 failure


0 200 400 600 800 1,000 1,2 00 1,400 1,600 1,800 2,000 2,200 2,400 2,600 2,800
STRAIN, J.Lin/in . I
FIGURE 3.-Tensile stress-strain curve for 112-in-thick wallboard.

PLASTER visual observation of buckling or crack-

ing was not possible until a slightly
Plaster was not studied extensively be- higher strain level was reached.
cause of its widespread replacement by 3. Strain rate seemed to affect ulti-
wallboard for modern construction. How- mate or total failure, but the paper
ever, many of the older homes analyzed yield point was relatively constant.
for RI 8507 (l) were plastered and pro- This allows comparison of various loading
vided some insight into cracking poten- factors (e.g., blasting versus other ac-
tial. Also, wallboard is a gypsum plas- tivities and environmental factors).
ter faced with paper on both sides. 4. Once the wallboard cracked, cyclic
Tests run on stripped wallboard are sug- opening and closing of the crack of up to
gestive of plaster failure Cl)• Of all O.l mm was observed. These movements
construction materials, plaster is con- were unaffected by blasting activities.
sidered most susceptible to damage and 5. Data on cyclic loading behavior of
exhibits fatigue at stress levels less wallboard are limited, but results of
than 50 pet of the static failure level tests on wood products indic~ted that fa-
(14). tigue effects can occur at stress (or
strain) levels equivalent to 50 pet of
WALLBOARD static failure conditions, but over
100,000 cycles are required.
The Bureau studied wallboard cracking
both in the laboratory and as part of the MASONRY
fatigue study of the test house (4). For
wallboard in the test house, researchers Bureau researchers also studied the
found threshold cracks occurring primar- cracking of concrete block walls, both at
ily in the wall corners and around nail the test structure with its full-size
neads. They found for wallboard-- basement and through a series of tests in
cooperation with the National Bureau of
The gypsum core failed at strains Standards (NBS) in Gaithersburg, MD
~f about 350 ~in/in in tension and at (~, ~). Generally, two types of cracks,
about 1,000 ~in/in in bending, based on local and steplike, were identified. Lo-
the nonlinear response points. cal block-length cracks less than 0.2 mm
2. For visible cracking, paper failure wide were difficult to discern from ex-
~as the controlling factor. Its nonlin- isting mortar joint separations and are '
~ar response point occurred at strains of usually not observed by homeowners.
l,OOO to 1,200 ~in/in (fig. 3). However, Steplike masonry cracks transverse the

wall along the mortar joint interface 0.01 mm

but 67.0 ~in/in,
and, over time, open beyond 0. 2 mm in 150 mm
Previous work by Cranston (16), Green where 13 and 150 mm are gauge lengths,
(17), and Wroth (18) noted that-all brick and the visible crack width is 0.01 mm.
walls have small,-o.l-mm cracks upon com- Because strain gauge readings can be mis-
pletion. Green stated that 0.1-mm cracks leading, crack growth is best described
are difficult to see and "therefore, do in terms of displacement.
not cause concern." As reported by Wood- 3. Local-site strains across the wall
ward (12), local cracks opened and closed vary considerably from global strains.
throughout the cyclic and monotonic in- For in-plane shear failure, global
plane shear tests of a 5- by 5-ft con- strain is measured or calculated across
crete blvck wall. It was not until a the wall diagonally.
steplike crack propagated the length of 4. Local cracks can occur at low glo-
the wall specimen that shear load failure bal strains, and global assessment of
occurred. these cracks is not recommended. But,
Although findings by Bureau researchers for the assessment of steplike cracks
on masonry failure provide some insight, that propagate across the entire wall,
further work at NBS on torsion and out- the global strain approach appears
of-plane loading is recommended. Key reasonable.
findings for tests with masonry are given 5. Global failure strain levels for
below. steplike cracks are not available. Lim-
ited testing to date has shown that in-
1. Observations of tensile cracks at plane shear failure may not occur in
strain-monitored sites showed that such homes because of the relatively light
cracks were first detected visually at vertical load available to prevent rota-
strain levels well above the first non- tion from the shear couple and at least
linear response point because of naked- a partial conversion of the shear to
eye limitations (~0.01 to 0.1 mm). tension.
2. Strains read at the threshold of 6. For cosmetic cracks that do not af-
visual cracking using different gauge fect load-carrying capacity, a crack-
lengths gave different overall strain widt~ criterion has been proposed (17).
readings as illustrated below. However, the 3Cceptability of crack
widths varies with material. For con-
= _, crete, 0.25 mm is the limit of accept-
Based on the equation £
1 ability (19), while 1 mm is the limit of
0 ' 01 mm = 770 ~in/in, acceptability for brickwork (~).
13 mm


Bureau researchers studied structure mines with thick soil overburdens and
responses and cracking associated with large-diameter blastholes, cases which
blasting vibrations in an investigation had not been studied previously. In all,
involving a relatively few measurements about 900 shots produced useful data on
at each of a wide variety of residential- structural responses and damage potential
type structures (3). Following this, fa- from blast vibrations.
tigue from repeated loading of one house
over a long period of time was studied ENVIRONMENTAL STRAINS
(4). For both efforts, measurements were
made of wall, floor, and racking re- Houses are subject to a variety of dy-
sponses, and observations of damage were namic loads, in addition to static or
made that could be correlated to specific slightly variable loads from settlement,
'·'' vibration events. A significant part of soil changes, and aging. Among the dy-
the work was done near large surface coal namic forces considered significant are


daily and annual temperature and humidity -817 uin/in, or up to 82 pet of failure.
cycles, wind, and human household activ- "Failure" is defined as the strain level
ity. Bureau researchers monitored the of 1,000 uin/in, found to produce wall-
weather and inside environment during the board cracking as previously discussed
2-yr test period and, in more detail, for (fig. 3).
short periods. For one test, they took
readings at 3-h increments for a 2-day HUMAN-ACTIVITY-INDUCED STRAINS
period, simultaneously measuring strain
at site Kz, over a major doorway (fig. Activities within the home can produce
4). Because there were at least four significant vibration and strain in local
factors influencing the strain, re- structural members (l-~). In severe
searchers used multiple linear regression cases, such as a hard door slam, the en-
analyses. Maximum strains from daily en- tire superstructure resonates producing
vironmental changes were found to be a strains in every wall, corner, and floor.
significant fraction of those needed for By contrast, nail pounding produces a
wallboard core failure or paper cracking. strong response only on the wall af-
The maximum strain observed at Kz was fected. Strains range up to about 100
+385 uin/in or 39 pet of failure. The uin/in, with typical values being 50
total maximum strains calculated from the uin/in in critical areas over windows and
correlation equation, assuming the worst doorways.
case for each of the factors, are +675 to


' 150
-= ol._ 100
z 50
<! 0
t; -50
_____ l_ _ _ __ l_ _ _ __ L_ _ _ __ L_ _ _ _

,._____ .,._____ ______,,',

-150L---~ ~----~----~----~----~----L_ _ _ _ _L __ _ _ _L __ _ _ _L __ _~

_,. -------
w <>-
> ..,,. . . ____ ...,..,""'.,.,. ', t",, ,
- >--
1- 1- ;
_Je- 60
a: ::>
:I: 50 ~----~~ln~s~id~e----~----_.-----.-----.----~----~----~----~----~----~----~----~----1
40L---~----~-----l-----L-----L----~----~----~----J_ ____~____L __ _ _ _~_ _ _ _L __ _ _ _L __ _~

~ 90.---------------------------------------------------------------------------------------,
~ 70
:::: t .,.,.,."" ............ ----.~--
Outside - - ... ,._ ... ___ .. _____ • _____ .....,........ ,. ..-----------·----... ............
<( ----- ~-----
15 60
:;: 50
1- 40~--~----~----~----~-----L-----L-----L----~----J_----~----~----~----~----~--~

..,. ____ ...____ _. ____ ........ __ ,._... _.......

,*"'"' --........ __ -,.,.____________ .......... .,.,
......... ;; East- west
-15 ' ....... -----·
3(8/IOl 6 9 12 15 II' 21 24 3(8/11) 6 9 18 21 24

FIGURE 4.-Wallboard joint strain and environmental factors versus time, site K2 over a doorway.


Blasting responses and strains in resi-
line dential structures were reported in de-
tail in Bureau Rl's 8507 and 8896 (3-4).
An example of blast-vibration-indu~ed
210 strains from the fatigue study reported
in Rl 8896 is shown in figure 5. Struc-
ture vibration responses can be transi-
tional, torsional, vertical uplift, or at
times a combination of all three. In
150 0 blasting, both the superstructure and
0:: foundation are typically affected. Non-
(/) blasting causes of vibration and strain
120 -
act only on the superstructure, except
for slowly acting soil changes and set-
90 0 tlement. Because initial damage involves
0 cosmetic cracks on supersturcture in-
terior walls, it is appropriate to com-
pare supersturcture strains from blasting
and other sources (table 3). These com-
parisons are only approximate. A given
vibration level does not always produce
the same strain even at a single monitor-
0 2 3 4 5 6 7 ing point, much less throughout the
MAXIMUM GROUND VIBRATION, in/s structure, probably because of different
response modes for different blast
FIGURE 5.-Piaster and wallboard strain versus maximum
ground vibration at site S1 over a doorway. angles, and wave characteristics.

TABLE 3. - Comparison of strain levels induced by daily environmental

changes, household activities, and blasting (4)

Induced strain, Corresponding

Loading phenomena Site Jlin/in blast vibration
level, 1 in/s
Daily environmental changes Bedroom midwall 149 1.2
Do • •••••• «; • • " ••••••••••• Over doorway ••• 385 ' 3.0
Household activities:
Walking • .••••••.•••..•••• 9
Over a window ••<.03
Heel drop • .••.•..••..••.. 20 .03
•.. do . .........
Jumping . ••..•.••.••••.•.• 37 .28
••. do . .•.•••.••
Door Slam •••••••••••••••• Over doorway ••• 49 .so
Pounding a nail . ......... Over a window •• 89 .88
1Vibration velocities are based on highest observation strains for a given
velocity. Use of mean or "typical" values from regression analysis gives ve-
locities considerably higher • For example, the door slam produces a level of
strain typically observed at 1.44 in/s ground-measured particle velocity. See
figure 5.


As discussed earlier, environmental table 10 (3), Langefors did not separate

factors induce most of the strain neces- the cracking and fall-of-plaster cases.
sary for the generation of cracks trig- Dvorak's study produced observations of
gered by household activities or blast- cracking at some of the lowest peak par-
ing. Crack rates did not increase until ticle velocities, and questions have been
blast vibration levels rose above normal raised about data reliability. However,
threshold levels o f 1.0 in/s. It is not Dvorak used the same seismic monitoring
surprising then that both wallboard and system as Langefors. Dvorak's orick
plaster cracked at low vibration levels, structures were likely different from
even though failure strain levels for Langefors' unspecified structures (proba- I
wallboard are about 3 times those of bly concrete), and the vastly different I
plaster. measured frequencies are indicative of a
In reviewing both past and newly avail- soil versus rock foundation. The lowest
able data on dynamic vibration response, vibration level at which cracking was ob-
researchers noticed irregular and some- served was 0.5 in/s with Dvorak's data
times high-amplitude responses when the and about 0.7 in/s without.
vibration frequencies matched structure
resonances (fig. 6). A similar effect,
noticed for the cracking data, was one of
the most significant findings in RI 8507
6 o I -story structure
(2)• Consequently, coal mine and quarry 6 2- story struciure
production blasts that are typically 10
to 25 Hz produce a greater damage risk
than smaller scale blasts often used for
construction, excavation, and secondary
In a departure from earlier analyses
and reports, the following review quanti-
fies damage separately for each of the
three major construction materials:
plaster, wallboard, and concrete block.
The reader is directed to the original
reports for procedure and analysis de-
tails 3-4 ..!2).
Mid wall


Threshold and minor cracking data are

summarized in figure 7 for pre-1975
studies and in figure 8 for recent Bureau
research. All these data have been pre; 5
0 0

viously published in RI 8507 (3) and

RI 8896 (4). However, in a departure 4
from the -earlier reports, these figures
identify each data point as to source, 0

degree of cracking damage, and type of "' 0

D80 6 0 6 0
material involved.
0 "'
Langefors (12) presents the only sig-
nificant amount of high-frequency data.
These data, suggest that vibration levels
as high as 4 in/s may be safe for fre- Of------L--~-L-L~LL~IO----~--_L~~-L~~IOO

quencies above about 70 Hz. In the de- GROUND VIBRATION FREQUENCY, Hz

scriptions of damage, in RI 8507's FIGURE 6.-Corner and midwall amplification factors (3).

strains about three times those required

for core failure. Wallboard cracks are
also influenced by how well panels are
attached to the superstructure frame.
Not being structural elements, they are
not always put under in-plane stress when
: 10.0
the frame flexes. The core around the
v nailhead is, at best, partially crushed
-' upon attachment to the studs, and when
w the studs are uneven major core cracking
•8"0 oooo
can occur. The response from superstruc-
a: o8 go 00
~ 1.0
4 Plaster crocKs and foil of plas1er. ture vibration is additional wallboard
o Dvorak {20) core crushing around the nailheads, re-
0 • Morris (iL}
" .c. Thoenen (22} sulting in a "loose" attachment.
o LanqeforsT~)
v Jensen (unpublished
Bureau contract repori)
At the test house, it was observed that
Wallboard ond joint crocks: cracks developed primarily at the plas-
Y Jensen (unpublished
Bureau contract report) tered joints at wall corners and in plas-
Q,J l _. _l____L.._j________l__.ILLJIr_jj!'--L--'-'---L~.-LIlllj_WI_ss_(_,¥_) .~LL..L.LLIJ ter covering coating over nailheads
I 10 100 1,000
FREQUENCY, Hz (table 4). The high rate of naturally
FIGURE 7.-Velocity and frequency levels for threshold plaster occurring cracks was caused primarily
and wallboard cracking, pre-1975 studies. from curing of the tape compound. As the
tests on the structure continued, a de-
crease of natural frequency of about
20 pet, e.g., 7.5 to 6Hz at one loca-
tion indicated a loss of rigidity and
general flexure-induced loosening (i).
o Plaster crocks The lowest levels of observed blast-
• Wollboord crock-s 1 oil types
o Wallboard fatigue cracks over nailheads
vibration-induced cracking occurred at a
v Wotlboord fatigue <::racks at tope Joints wall corner as crack extensions and when
J. Wallboard panel fatigue
a new crack was observed beneath a win-
10.0 dow, at amplitudes of 0.79-1.1 in/s
.... 0
(fig. -s) •
0 0 0
w 0
Fatigue-induced cracks were observed at
0.3 to 1.0 in/s. However, this cracking
required a large number of vibration
<( cycles, such as over 50,000 at a 0.5-in/s
equivalent ground vibration. This
a. equates to decades of typical blasting
with one blast per day producing 10 cy-
cles per blast.

.I '---L-L-.l-LL..L..l..U...----1....."-1-.-'-..~1JIJI..!.IJ..Ld---'--·-'--'----'-'-U--"
I 10 100 1,000
FREQUENCY, Hz Cracks produced in block masonry walls
FIGURE a.-Velocity and frequency levels for threshold plaster by blasting are given in figure 9 for
and wallboard cracking, recent studies (3-4). past work and figure 10 for recent Bu-
reau studies Cl-~). Most cracks observed
were local, typically shorter than one
WALLBOARD CRACKS block length, and about 0.2 mm in width.
Cracks of this magnitude were observed
The state of cracking of wallboard is from blast vibrations up to 6.2 in/s and
hard to identify because the interior were not of concern, being indistinguish-
plaster core will crack long before any able from normal construction and shrink-
surface effect is visible. Visible age effects. Their observation is dif-
cracking of paper covering occurs at ficult, which accounts for the high

. ..
"'• t • 'c 10.0
,: e ,:

. .
A Ali
.... ....
--' A
....t ~·.
A e e
A <:;
"':> A > • 0 •

"'u--' "'

......"' KEY
"'<t a.
Local masonry wall cracks:
• Full aca le fatigue, Stagg lil
0.. Creek classification:
A Minor domoqe-blostlng
• Major damage • Model scale fatigue, Koerner (til
A Minor damage o Threshoid doma'ile- blostin9
• T hroshold domoge


I 10 100 1,000 10 100 1,000


FIGURE 9.-Velocity and frequency levels for local masonry FIGURE 10.-Velocity and frequency levels for local masonry
wall cracks, pre-1975 studies. wall cracks, recent studies (3-4).

TABLE 4. - Wallboard cracks observed in fatigue test house (~).

Initial Cracks developed Blasting Mechanical shaker tests 1

Material cracks, during testing ievel, Cracks Number of cycles
before Naturally From in/s developed at cracking
testing occurring blasting
Taped corners ••• 39 35 5 0.88-3.5 N0 2 NAp
Nailheads ••••••• 5 4 3 1.8 -2.2 >3 56,000, 339,500
Taped joints •••• 2 6 NO NAp 1 56,000
Wallboard ••••••• 3 6 NO NAp 1 361,500
NAp Not appl1cable. NO None detected.
1Shakers run at resonant frequency at equivalent vibration levels of 0.3 to 1.0 in/s.
2Corners almost completely cracked before shaker study.

TABLE 5. - Masonry wall mortar joint cracks observed in fatigue test house (~)

Initial Cracks developed Blasting Mechanical shaker tests 1

Material cracks, during testing level, Number of Number of cycles
before Naturally From in/s cracks at cracking
testing occurring blasting
Brick •••••••••• 20 28 7 3.4 -6.2 2 229,000, >293,500
Fireplace •••••• 21 11 1+ -6.9 2 56,000, >108,500
Block . ......... NA NA S+ 6.2 -6.9 3+ >339,500
Steplike crack. NO NO 1 .96-1.5 N02 NAp
Separation ••••• NO NO 2 6.9 1 >339,500
NA Not ava1lable--see text. NAp Not applicable. NO None detected.
Shakers run at resonant frequency at equivalent vibration levels of 0.3 to 1.0 in/s.
2Existing steplike crack functioned as an area of stress relief.

number of naturally occurring cracks propagations .were observed across brick

(table 5). Also, these local cracks be- or block walls. The existing steplike
came more apparent during the cyclic crack functioned as an area of strain re-
test. Differential motion along the lief during shaker runs. Energy trans-
block interfaces was easily observed dur- mitted by the shakers into the super-
ing continued cyclic motion, which ac- structure and foundation was primarily
counts for the low vibration levels, 0.3 dissipated in areas of previous cracking.
to 1.0 in/s. However, in the test house, Observations were also made of chimney
a blast vibration of 6.9 in/s produced a and brick veneer responses during cyclic
crack of significant magnitude, widening shaker tests. The masonry walls were
a crack beyond the width that was ob- relatively stationary, with the super-
served in the absence of a blast. structure cyclically bumping the chimney
and a brick veneer wall near the roof
SHEAR LOAD FAILURE line. Mortar joint cracks developed at
the chimney-roof interface and horizon-
Shear load failure of the basement wall tally across the brick veneer just above
of the test house was observed after four door height.
shots in one day. A diagonal steplike Crack data from Edwards and Northwood
crack propagated in the southwest base- (~) do not specify crack widths. If
ment wall, starting at ground level and these crack data correspond to observa-
proceeding upward. When these four shots tions exceeding 0.2 mm (excessive crack
were detonated, their vibration levels widths), it would suggest that cracks can
(ranging from 1.0 to 1.5 in/s) were the occur at particle velocity levels of 3
highest recorded in the study up to that to 7 in/s with no effect of frequency.
time. But because observation of cracks Additional data are needed to qualify
in masonry is difficult, it remains un- frequency effects and the generation of
known whether blasting or other events stairstep crack patterns across the wall
caused this steplike crack. It is note- signifying shear load failure.
worthy that no additional steplike crack


Bureau studies of the response and of external transient vibrations on wood-

cracking of low-rise residential struc- frame, low-rise residential structures
tures from blasting indicated that crack- typical of those studied by the Bureau.
ing of plaster and wallboard is not Data on the response and cracking of
likely below about 0.5 in/s peak particle masonry walls from blasting indicated
velocity for the worst case· of structure that local cracking (block-length) may
condition and typical vibration fre- not be noticeable until particle velocity
quency. levels are up to 3.0 in/s. However, ad-
This safe-level criterion also appears ditional research is needed to quantify
independent of the number of blasting vibration levels that promote the genera-
events and their durations. Researchers tion of stair-stepped cracks that propa-
also noticed that high strains are pro- gate across the wall and reduce its shear
duced in structure walls by normal load capacity.
weather conditions, such as wind, temper- The authors encourage, where possible,
ature, and humidity cycling. Dynamic direct measurements or assessment of
events such as door slams or blasting strains or loads on members likely to
produce additional strain, which can fail. Alternatively, estimates of re-
trigger a crack in a structure already sponses should be based on realistic
under strain. Human activities, such as transfer functions relating measured vi-
door slams, can be equivalent to blast brations and reasonably expected re-
vibrations of up to 0.5 in/s. The vibra- sponses. In particular, applications be-
tion level of 0.5 in/s thus provides a yond the scope of the original Bureau
minimum value of concern for the impact studies are to be done only with caution.



1. Siskind, D. E., and M. S. Stagg. 12. Andrews, D. K., G. W. Zumwalt,

Blast Vibration Measurements Near and On R. L. Lowery, J. H. Gillespie, and D. R.
Structure Foundations. BuMines RI 8969, Low. Structure Response to Sonic Booms.
1985, 20 pp. (U.S. FAA contract FA-64-AC-6-526, An-
2. Stagg, M. S., and A. J. Engler. drews Associates Inc. and Hudgins, Thomp- \
Measurement of Blast-Induced Ground Vi- son, Ball and Associates Inc., Oklahoma :
brations and Seismograph Calibration. City, OK). Rep. AD618022, Feb. 5, 1965,
BuMines RI 8506, 1980, 62 PP• 228 pp.; available from Defense Documen-
3. Siskind, D. E., M. s. • tation Cent., Alexandria, VA.
J. W.Kopp, and C. H. Dowding. Structure 13. \\Tall, J. R., Jr. Seismic-Induced
Response and Damage Produced by Ground Architectural Damage to Masonry Struc-
Vibration From Surface Mining Blasting. tures at Mercury, Nevada. Bull. Seismol.
BuMines RI 8507, 1980, 74 PP• Soc. Am., v. 57, No. 5, Oct. 1967,
4. Stagg, M. S., D. E. Siskind, PP• 991-1007.
M. G. Stevens, and C. H. Dowding. Ef- 14. Leigh, B. R. Lifetime Concept of
fects of Repeated Blasting on a Wood- Plaster Panels Subjected to Sonic Boom.
Frame House. BuMines RI 8896, 1984, Univ. Tononto, Ontario, Canada, UTIAS-IN-
82 PP· 191, July 1947, 78 PP•
5. Siskind, D. E., V, J. Stachura, 15. Woodward, K. A., and F. Rankin.
M. S. Stagg, and J. w. Kopp. Structure Behavior of Concrete Block Masonry Walls
Response amd Damage Produced by Airblast Subjected to Repeated Cyclic Displace-
From Surface Mining. BuMines Rl 8485, ment. NBSIR 83-2780, 1983, 178 pp.; NTIS
1980, 111 pp. PB 84-122092.
6. Stachura, V. J., D. E. Siskind, 16. Cranston, W. B. Masonry Research
and A. J. Engler. Airblast Instrumenta- and Codes in the United Kingdom. Paper
tion and Measurement Techniques for in Earthquake Resistant Masonry Con-
Surface Mine Blasting. BuMines RI 8508, struction: National Workshop (NBS, Boul-
1981, 53 pp. der, CO, Sept. 13-16, 1976), ed. by R. A.
7. Sabnis, G. M. Vibrations of Con- Crist and L. E. Cattaneo. NBS Build.
crete Structures. Sec.: Introduction Sci. Series 106; Sept. 1977, pp 166-176.
and Background. Am. Caner. Inst., 17. Green, D. G., I. A. Macleod, and
s p 6 0- 1 , 19 79 ) p p. 1-12 • W. G. Stark. Observation and Analysis of
8. Tuomi, R. I., and W. J. McCutchen. Brick Structures on Soft Clay. Paper
Testing of a Full-Scale House Under Simu- in Performance of Building Structures
lated Snowloads and Windloads. U.S. (Proc. Int. Conf., Glasgow Univ., Mar.
For. Serv., Res. Paper FPL 234, 1974, 32 31-Apr. 1, 1976). Pentech Press, 1976,
PP• PP• 321-336.
9. Warwaruk, J. Vibrations of Con- 18. Wroth, C. P. General Report Ses-
crete Structures. Sec.: Deflection sion IliA: Response of the Structure
Requireements--History and Background to Foundation Movements. Paper in Per-
Related to Vibrations. Am. Concr. Inst., formance of Building Structures (Proc.
Sp 60-2, 1979, pp. 13-41. Int. Conf., Glasgow Univ., Mar. 31-
10. Whittemore, H. L., J. B. Cotter, Apr. 1, 1976). Pentech Press, 1976,
A. H. Stang, and V. B. Phelan. Strength pp. 489-508.
of Houses--Application of Engineering 19. Haldane, D. The Importance of
Principles to Structural Design. NBS Cracking in Reinforced Concrete Members.
Build. Mater. and Struct. Rep. BMS 109, Paper in Performance of Building Struc-
Apr. 1948, 131 PP• tures (Proc. Int. Conf., Glasgow Univ., \
11. Holmberg, R., N. Lundberg, and Mar. 31-Apr. 1, 1976). Pentech Press,
G. Rundqvist. Ground Vibration and Darn-
age Criteria. Constr. Res. Counc.,
1976, PP• 99-109.
20. Dvorak, A. Seismic Effects of
Stockholm, Sweden, Rep. R 85:81, 1981, Blasting on Brick Houses. Pr. Geofys.
30 PP•


Ustance Cesk. Akad. Ved. No. 169. Geofys. 25. Stagg, M. S., and D. E. Siskind.
Sb., 1962, pp. 189-202. Repeated Blasting: Fatigue Damaging or
21. Morris G., and R. Westwater. Not? Paper in Proceedings of the 11th
Damage to Structures by Ground Vibrations Annual Conference on Explosives and
Due to Blasting. Mine and Quarry Eng., Blasting Technique, ed. by C. J. Konya
v. 24, Apr. 1958, pp. 116-118. (San Diego, CA, Jan. 27-Feb. 1, 1985).
22. Thoenen, J. R., and S. L. Windes. Soc. Explos. Eng. Montville, OH, 1985,
Seismic Effects of Quarry Blasting. PP• 96-110.
BuMines B 442, 1942, 83 PP• 26. Edwards, A. T., and T. D. North-
23. Langefors, U., B. Kihlstrom, and wood. Experimental Studies of the
H. Westerberg. Ground Vibrations in Effects of Blasting on Structures.
Blasting, Water Power, v. 10, 1958, Engineer, v. 210, Sept 30, 1960,
Sept., pp. 335-338; Oct., 390-395; PP• 538-546.
Nov., 421-424. 27. Koerner, R. M., and J. L. Rosen-
24. Wiss, J. F., and H. R. Nicholls. farb. Feasibility of Fatigue Assessment
A Study of Damage to a Residential of Block Walls From Laboratory Scale
Structure From Blast Vibrations. Res. Methods (contract J0285013, Drexel
Counc. for Performance of Struct., ASCE, Univ.). BuMines OFR 144-80, 1980,
New York, 1974, 73 PP• 96 pp.: NTIS PB 81-140139.




By David E. Siskind1 and Mark S. Stagg2


Blasting near structures often involves With the exception of the basement t
vibration measurement to assess damage floor measurements and some of the dis- (

potential. Several methods of measure- tant measurements, waveforms were similar

ment are used worldwide; however, there and amplitudes were generally within 10
is no. consensus as to which methods are to 30 pet of each other. The low-
technically sufficient and yet practical frequency part of the wave (5 to 10 Hz)
for all situations. was particularly uniform in measurements
The Bureau of Mines studied five place- obtained at all five locations. Differ-
ment locations for vibration transducers ences in peak values were mostly from
to determine the best method for monitor- minor shifts in phase of the high-
ing blasting vibrations. The locations frequency components, which are less sig-
were--burial in soil next to the struc- nificant to structural response and
ture, attachment to the foundation at potential damage than the low-frequency
ground level, to the basement floor, or waves. Shallow surface burial resulted
to a surface slab, and burial at a dis- in good signal detection and the least
tance from the structure in undisturbed chance of mechanically induced error.
soil. Typical surface mine production
blasts were used as vibration sources.


The Bureau of Mines studied vibration (2) shallow burial of the transducer in
measurement methods applicable to produc- the soil next to the foundation, and
tion blasting in surface mines (1).3 (3) measurement on a nearby concrete slab
Blast vibrations are routinely monit~red such as a driveway or walkway. The
for one of two purposes: (l) to assess specific practices followed are often
damage risk for nearby structures and based on convenience. However, at suf-
(2) to derive predictive equations for ficiently low vibration amplitudes, all
vibration generation and propagation. three methods will give similar results.
Despite years of practice and several The many factors involved, such as trans-
published studies on measurement, the in- ducer shape and size, soil or ground
dustry has not adopted a uniform and con- strength, and density, make for varied •
sistent methodology. Occasionally, those measurement requirements. Past Bureau
monitoring blasts fail to obtain repro- studies did identify the need to anchor,
ducible and accurate vibration records. attach, or bury vibration transducers
Three measurement methods are in common for blasts exceeding 0.2 G acceleration
use: (1) direct attachment of the trans- amplitude (_~).
ducer to the foundation of the structure Because of problems of practicality,
to be monitored, at or near ground level, cost, and site access, it is not likely
that a single measurement method will be
1 supervisory geophysicist. universally adopted, or needs to be. The
2civil engineer. Bureau, therefore, evaluated the most
Twin Cities Research Center, Bureau of common methods to determine which give
Mines, Minneapolis, MN. accurate indications of vibration energy
3underlined numbers in parentheses re- transferred into structures while also
fer to items in the list of references at being representative of the blast as a
the end of this paper. vibration source (l).


Previous research on blast vibration particularly at high accelerations (com-

monitoring has been concerned with sta- binations of high velocity and high
bility and slippage (2-4 , spiking in frequency).
soil (5-6), partial resonances (7-8), im- The research described in this report
pedancematching (9), Rayleigh wave depth involves side-by-side comparisons of vi-
effects (6-10 , burial in soil, (9, 11- bration time histories from a series of
12), and measurement for response spectra surface coal mine production blasts. A
... (13-16). The overall conclusions from more complete description of the study
:ill-these studies are that burial of and related background was published in
Lransducers is desirable and that cau- 1985 in RI 8969 (_.!_).
tion is required for other methods,


Four structures were studied at three this could be significant. The decrease
operating mines, two near St. Clairs- depends not simply on depth, but on depth
ville, OH, and one near Evansville, IN. compared to wavelength (fig. 1). There-
All had concrete block or stone basements fore, measurements were made of propaga-
with 5- to 7-ft depths. Researchers mon- tion velocity, allowing the calculation
itored a total of 23 blasts, containing of wavelengths from
22 to 2,200 lb per delay, at distances
from 425 to 4,900 ft. A = c/f,
Monitoring methods varied among the
sites because of differences of accessi- where A wavelength, ft,
bility and sui~ability. Concrete slabs
were near the structures at two sites; c propagation velocity, ft/s,
however, at the other two, the slabs were
22 and 60 ft away from the main struc- and f frequency, Hz.
ture. Direct comparisons were made as
closely as possible, by simultaneous mon- According to Clark, a deeper foundation
itoring, with transducers at the follow- "sees" a decreased amplitude vibration.
ing locations: This is also consistent with observed
low-amplitude vibrations on basement
1. Buried in the soil with 2 ft of the floors (!Q, ~-17 •
2. Mounted on the foundation wall at
or near ground level, usually inside. RELATIVE VELOCITY AMPLITUDE
3. On the basement floor, 0 --~0=.5~-----e~------1.5
4. On a nearby surface slab, a walk-
way, driveway, or garage floor, and
5. Buried near the structure but re-
moved from the influence of disturbed
ground around the foundation, 72 to Horizontal/ \Vertical
100 ft distant. component component

Special measurements were made of vi-

bration propagation velocity to examine
the implications of depth-dependent vi-
bration amplitudes. Clark had described
.·• the decrease of Rayleigh wave amplitudes
... with increasing depth (1Q). For situa- DEPTH/WAVELENGTH
tions when the blast vibration is domi- FIGURE 1.-Rayleigh wave amplitude profile versus depth,
nated by the Rayleigh surface waves, after Clark, Larson and Lande (10).

Waveforms were found to be very similar Amplitudes varied more than waveform
for measurements made near and on the frequency characteristics. These ampli-
structure, and also on a nearby surface tude differences of about 25 pet resulted
slab. A set of vibration wave compar- from phase changes and minor amplitude

isons is shown in figure 2. Seven addi- differences from the more strongly af-
tional record sets are shown in RI 8968 fected higher frequencies. In terms of .It
(1). For structure response~ the vibra- structure response, they are not signifi-
tion's low frequencies (less than 20 Hz) cant. However, they do complicate mon-
are most critical. The low-frequency itoring by increasing the scatter in peak ,.
characteristics of the measured records velocity values. Table 1 summarizes the I

were nearly identical, peak for peak and peak vibration amplitudes. As expected, '
wiggle for wiggle. However, high fre- basement floors are consistently low, by
quencies present in the outside buried about 25 pet. Most of the peak vibration
gauge records were absent inside, as the amplitude differences in table 1 are from
structure filtered the vibrations by its minor phase changes and variations in the
frequency-dependent response. These high relatively insignificant high frequen-
frequencies provided minor changes in the cies. However, they do suggest the ad-
waveform details and peak values. vantages of using some method of vibra-
Records collected at distances from the tion analysis to compensate for these
house differ, mainly from phase changes waveform differences, such as signal
between arriving vibration modes and smoothing, weighting, velocity exposure,
travel-time shifts of energy from various response spectra, or another frequency-
parts of the blast. dependent method.

Location of

---·------/""~,}'1._/\VfJ'\.V{ \\.M~rv-LA'v.,/fiiV\r·~ (\~~{'\ /"...._./·---~---------
Buried n_. ear

~Basement floor

\ ..

Midspan of wall

0 0.2

Time, s
FIGURE 2.-Vibration records obtained at Fador house, shot. 6, longitudinal component of motion, at different locations (1).

TABLE 1. - Vibration levels measured near and on test structures, peak particle
... velocities, inch per second •

Measurement location
Structure Shot Component Ground Ground Ground Basement Away from
level, surface, level, on floor structure,
buried on slab foundation buried
Schnegg house ••• 1 v 0.32 NM NM 0.21 NM
L .26 0.35 NM .13 NM
T .28 .41 NM .15 NM
Mine office ••••• 2 v .12 NM 0.060 NM NM
L .19 .14 .13 NM 1 o. 11
T .18 .12 .15 NM I .13
Fador house ••••• 4 v NM NM .060 .054 NM
L NM NM .051 .040 NM
T NM NM .079 .072 NM
Do •••••••••••• 5 v • 11 NM .090 .080 2 .Q90
L .13 NM .14 .12 2 .160
T .11 NM .11 .080 2 .160
Do •••••••••••• 6 v .13 NM .12 • 11 NM
L .18 NM • 17 .15 NM
T .13 NM .12 NM NM
Schnegg house ••• 6 v .034 NM .035 NM NM
L .074 NM .075 NM NM
T .070 NM .060 NM NM
Training house •• 13 v .017 .015 .016 NM 3 .020
L .058 .060 .035 NM 3 .067
\, T .057 .059 .047 NM 3 .063
NM Not measured. .74 ft from structure.
V Vertical. 2 72 ft from structure.
L Longitudinal. 3 98 ft from structure.
T Transverse.


At the four sites examined in this Considerable differences were noted for
study, the specific measurement methods the high-frequency part of the waves,
used around structures appear not to be mostly in the beginnings, which corre-
critical at low vibration levels. Five spond to the multiple arrivals from the
gauge locations were examined: on a various delayed holes. It is likely that
surface slab, buried in the ground at the these differences were primarily a result
structure, mounted on the foundation of the varied wave paths to the different
wall, on the basement floor, and buried monitoring positions, leading to uneven
at a distance of 72 to 98 ft from the and irregular wave interference. The
structures. The waveforms for all three high frequencies are of less concern for
,I components, vertical, longitudinal, and structural response, as discussed in a
transverse, were found to be similar for
the five measurement locations. This was
previous Bureau report on structure re-
sponse from blasting (16).
particularly true for the low-frequency Because peak values-are influenced by
.. part of the waves, which is of most con- the way specific vibration or wave modes
cern for vibrational response of struc- interact, they were found to vary ir-
tures. Low frequency for this study was regularly between the different methods.
5 to 10 Hz. However, they were generally within


20 to 30 pet of each other and not con- displacement for the wall top and bottom.
sistent with the systematic 0.40 factor Apparently, vibrational energy does de-
for foundation depths of 6 to 9 ft pre- crease with depth.
dicted by Clark (10). This demonstrates Rather than recommending a specific
one of the weakne;ses of a simple peak- measuring method, the researchers recom- I
particle-velocity criterion and over- mend that consistency be used at any one !
reliance on interpretation of precise site. Where the option is available,
shallow soil burial is still the desired
Vibration levels below grade, such as method and was found least likely to in- I
on the basement floor, were
tently lower,
suggesting differential
troduce additional mechanical error. II

1. Siskind, D. E., and M. S. Stagg. Soc. Explos. Eng., Montville, OH, 1983,
Blast Vibration Measurements Near and On PP• 27-62.
Structure Foundations. BuMines Rl 8969, 11. Skipp, B. 0. Ground Vibration In-
1985, 20 pp. strumentation--A General ,View. Paper 2
2. Stagg, M. S., and A. J. Engler. in Instrumentation for Ground Vibration
Measurement of Blast-Induced Ground Vi- and Earthquakes (Proc. Conf. for Earth- /
brations and Seismograph Calibration. quake and Ci v. Eng. Dyn. , Keele Uni v., ~'
8uMines RI 8506, 1980, 62 PP• Staffordshire, England, July 4, 1977).
3. Duvall, W. I. Design Criteria for lost. Civ. Eng., London, 1978, pp. 11-34.
Portable Seismographs. BuMines RI 5708, 12. Nicholls, H. R., c. F. Johnson,
1961, 6 pp. and w. I. Duvall. Blasting Vibrations
4. Fogelson, D. E., and C. F. and Their Effects on Structures. BuMines
Johnson. Calibration Studies of Three B 656, 1971, 105 pp. ,
Portable Seismographs. BuMines RI 6009, 13. Dowding, C. H., and P. G. Corser.
1962, 21 pp. Cracking and Construction Blasting. Im-
5. Johnson, C. F. Coupling Small Vi- portance of Frequency and Free Response.
l:lration Gages to Soil. Earthquake Notes, J. Constr. Div., ASCE, v. 107, No. Co. 1,
v. 33, Sept. 1962, PP• 40-47. 1981, pp. 89-106:
6. Gutowski, T. G., L. E. Witting, 14. Dowding, C. H., P. D. Murray, and
~nd C. L. Dym. Some Aspects of the D. K. Atmatzidis. Dynamic Properties
;round Vibration Problem. Noise Control of Residential Structures Subjected
:<.:ng., v. 10, No. 3, 1978, PP• 94-100. to Blasting Vibrations. J. Struct.
7. Bycroft, G. N. The Magnification Div., ASCE, v. 107, No. St. 7, 1981,
~aused by Partial Resonance of the Foun- PP• 1233-1249.
jation of a Ground Vibration Detector. 15. Langan, R. T. The Use of Single-
rrans. Am. Geophys. Union, v. 38, No. 6, Degree-of-Freedom System as a Dynamic
l957, pp. 928-931. Model for Low-Rise Residential Structures
8. Washburn, H., and H. Wiley. The Subjected to Blast Vibrations. M.S.
~ffect of the Placement of a Seismometer Thesis, Northwestern Univ., Evanston, IL,
)O Its Response Characteristics. Geo- 1980, 79 pp.
lhysics, v. 6, 1941, pp. 116-131. 16. Siskind, D. E., M. S. Stagg, J. W.
9. Duvall, W. I. Design Requirements Kopp, and c. H. Dowding. Structure Re-
:or Instrumentation To Record Vibrations sponse and Damage Produced by Ground
>roduced by Blasting. BuMines Rl 6487, Vibration From Surface Mine Blasting.
L964, 7 pp. BuMines RI 8507, 1980, 74 pp.
10. Clark, D., B. Larson, and G. 17. Siskind, D. E., v. J. Stachura,
.ande. Vibration: Its Effect and Mea- and K. S. Radcliffe. Noise and Vibra- '·
;urement Techniques at or Near Buildings. tions in Residential Structures From
>aper in Proceedings of the 9th Confer- Quarry Production Blasting. BuMines
!OCe on Explosives and Blasting Technique RI 8168, 1976, 17 PP•
:Dallas, TX, Jan. 31-Feb. 4, 1983).


By John W. Kopp 1


A major concern with blasting at sur- frequency content. Various delay inter-
face mines is generation of ground vibra- vals were used within and between rows of
tions and airblast and their effects ou blastholes. Delay intervals within rows
nearby residences. This Bureau of Mines were 17 and 42 ms, and those between rows
report looks at the use of millisecond ranged from 30 to 100 ms; these intervals
delays in blast design and their effect are equivalent to burden reliefs of 0.5
on the resulting ground vibrations and 1.3 ms/ft within rows and 1.2 to 4.3
and airblast. A total of 52 production ms/ft between rows. Subsonic delay in-
blasts were instrumented and monitored at tervals within rows reduced airblast by
a surface coal mine in southern Indiana. 6 dB. Large delay intervals between rows
Arrays of seismographs were used to gath- reduced the amplitude of ground vibra-
er time histories of vibrations and air- tions; vibration frequency depended pri-
blast. The data were analyzed for peak marily upon the geology of the mine site.
values of vibration and airblast and for
Millisecond-delay blasting caps were natural frequencies of oscillation of
introduced to the mining industry in the structures. A study was then undertake~
1940's and gairred wide acceptance as a at a surface coal mine to determine if
tool for improving rock fragmentation. the predominant frequency of ground vi-
The use of these delays also reduced brations could be controlled by an appro-
ground vibration levels. The Bureau of priate choice of blast delay intervals.
Mines reported on this technique, which This paper summarizes that study, pub·
allowed the explosive in each delay peri- lished as RI 9026 in 1986 (l)·
od to be treated separately in its con- During the analysis of the Bureau 1 !::
tribution to the ground vibrations, in coal mine data, another study of dela)
1963 <.JJ.2 control of blasting was undertaken b;
The Bureau undertook a major research Reil (4) through a Bureau contract.
effort during the 1970's to quantify Reil's study in two stone quarries in
these ground vibrations and their effects valved precise timing control. Result:
on structures (2). Out of this study were mixed with respect to both ampli·
came the fact that not only amplitude is tudes and frequencies; results also ap·
important in preventing damage, but also peared to be both distance and measuring·
the frequency content of the vibrations, site specific.
because resonanr.~s were occurring at the


INSTRUMENTATION AND MEASUREMENT ground motion and the airblast overpres

TECHNIQUES sure on standard cassette audiotapes
The tape recorder for each machine wa
Airblast and ground vibrations from automatically activated when the groun
52 production blasts were measured with vibration reached a predetermined level
•. self-triggered seismographs. These seis-
mographs recorded three components of 2underlined numbers in parentheses re
fer to items in the list of references a
1Mining engineer, Twin Cities Research the end of this paper.
Center, Bureau of Mines, Minneapolis, MN •

The frequency range of the transducers in four directions from the shot. Each
used to monitor ground vibration was 1 to array used three instruments, located at
200 Hz. The maximum amplitude that could distances of 300 to 5,000 ft from the
be recorded was 4 in/s. For low~level blast. The complete ground vibration and
signals, an alternative range could airblast waveforms were recorded at each
be selected with a maximum amplitude of station. From these, the frequency spec-
1 in/s. tra and peak particle velocities and air-
The airblast channel used a 1-1/8-in blast could be determined. Peak particle
ceramic microphone. The frequency re- velocities were plotted versus scaled
sponse of the system was 0.2 to 200 Hz, distance for each array direction. A
with a maximum peak overpressure of 137 least squares fit of the regression line
dB. was determined for each set of data. A
The blasts were also monitored using a one-way analysis of variance test was
16-mm high-speed cinecamera. The rotat- then performed on the data sets to deter-
ing-prism camera was capable of speeds in mine if the blast parameter under study
excess of 8,000 frames per second, but a was significant.
rate of 1,000 frames per second was suf- The second objective was to determine
ficient for this study. The initiation the effects of varying the delay inter-
system used was Nonel3 with surface de- vals between holes and rows. Airblast
lays. Nonel tubing was also tied into and ground vibration measurements were~
the delay detonators in order to provide made as before with seismographs deployed
a flash signal for the camera to record. in arrays in the four directions. Delay
This allowed computation of the firing intervals used were 17 and 42 ms between
time for each delay to the nearest holes in an echelon and 30, 42, 60, 75,
millisecond. and 100 ms between echelons. Normal mine
production shots used 17 ms between holes
TEST SITE in an echelon and 42 ms between echelons.
The Nonel Primadet system was used for
The project test site was a surface these delays for shots 1 through 36. De-
coal mine in southern Indiana. The mine lay intervals between rows for shots 37
utilized two large draglines to remove 50 to 52 were obtained by using electric
to 100 ft of overburden from a 4- to 5-ft caps, all of one period, with a sequen-
coal seam. The overburden was primarily tial blasting machine. The high-speed
shale with some sandstone intermixed, and camera described previously was used to
required blasting to facilitate digging determine actual firing times for each
by the draglines. Blasting was accom- hole. Propagation plots were made of the
plished using 12-1/4-in holes normally airblast and vibration data.
drilled on a 30-ft-square pattern and
shot en echelon into a buffer. The ter- DIRECTIONAL EFFECTS
rain is flat to gently rolling hills.
The layout of the pit is not influenced The blast design examined for direc-
by topography; it is about 3 miles long tional effects was the one normally used
in a north-south direction. The movement by the mine. A delay of 17 ms was used
of mining is toward the west. between holes in the echelon, and a delay
time of 42 ms was used. between echelons.
TEST PROCEDURE Ten shots were examined for this experi-
ment.4 The ground vibration data are
This series of tests had two objec- shown as a propagation plot in figure 1.
tives. Firs~, to determine if orienta- The airblast data are presented in figure
tion of the shot affected vibration lev- 2. A least squares regression analysis
els, seismograph arrays were established
4shots 14 through 23. (Shots 1 through
4 are discussed in the following section;
not imply endorsement by the Bureau of shots 5 through 13 are not discussed in
Mines. this paper. )

0 0

o North
\1 South
t> East
o West

~ 1.00
I ::!:
~ 1.00
.s c
>- ~
1- 1-
0 0
...J ...J
w w
> >
w w
...J ...J
cr: 1-
a:: KEY
-;;_ .10 .10
rt 0 North
A East
+ South
South 0
deviation .& East
.& East

I I Ill _l_. ._._J.n.-_ ___.__ .01

10 100 500 5 10 100 500
11 I

FIGURE 1.-Propagation plot of peak particle velocity show- FIGURE a.-Directional effects on propagation of ground vibra-
ing directional differences. tions. All regression lines have a common slope.

was used to determine the regression line

of each set of data. Analysis of vari-
ance tests performed on the vibration
KEY data determined that the data can be rep-
0 North resented by four regression lines with a
\1 South common slope (fig. 3). This indicates
w 130
t> East
that the vibration level is dependent on
::::J West
r.n +standard direction from the blast, but attenuation
,,f. w deviation
of the vibrations is independent of di-
a. rection, with the possible exception of
w 120
> the western direction. This may be due
to a geologic anomaly west of the mine.
The westen part of the mine is overlain
[J[J by lacustrine and sand and gravel depos-
a:: [J [J[J its associated with a large creekbed
drainage area. These deposits tended to
produce lower predominant frequencies of
ground vibrations (fig. 4) in che crans-
100--~~ ~~~~~-----~-~J-~~~ verse axis than those produced for the
10 100 1,000
I other arrays on undisturbed ground (north
CUBE ROOT SCALED DISTANCE, ft/lb /3 and south directions). Frequencies of
FIGURE 2.-Propagation plot of peak airblast showing direc- vertical and radial vibrations did not
tional differences. appear to be affected. The frequency of

vibrations in the reclaimed spoil or The highest vibration levels were found
eastern direction was also predominantly in the western direction, with levels in
lower. the north array direction the next high-
est. The results in figure 3 suggest
that vibration levels in the direction or
initiation can be twice those in the op-
60 posite direction.
40 North station
Two different delay intervals were used
30 between adjacent holes in each echelon:
17-ms from shots 1 and 2 and 42-ms delays
from shots 3 and 4. A 100-ms delay in-
terval was used between rows. The blasts
were all shot at the same location in the
mine using the same blast pattern.
For these shots, the mine used a square
pattern drilled on 25-ft centers. The
pattern was fired en echelon, giving an
effective burden of 18 ft and spacing ov
35 ft. The actual firing times averaged
Eo~' station
23 and 44 ms for the nominal 17- and 42-
ms delays, respectively. This gave a
relief of 0.5 ms/ft of spacing for the
17-ms-delay shots and 1.3 ms/ft for the
42-ms-delay shots. The burden I delays
. averaged 96 ms, giving a burden relief of
5. 3 ms /ft.
w The direction of the measurement arrays
from the shot did not appear to signifi-
u cantly affect _the airblast data. Analy-
sis showed the design using 42-ms delays
South stat ion
produced 6 dB less airblast than the 17-
ms design (fig. 5).
An analysis of variance was also per-
formed for each array direction comparing
che two delay intervals. The data were
sufficiently different to cequire sepa-
rate regression lines with a common slope
for the west and north arrays, but showed
50 no difference in the east array (figs. 6-
8). The south array had insufficient
40 data for analysis. The direction of ini-
West station tiation of the holes in each row was to-
ward the northwest. The airblast trace
20 velocity for the 17-ms delay design was
supersonic in the north and west direc-
10 tions but subsonic in the east direc-
tion. The airblast trace velocity for
0 5 25 45 the 42-ms design was subsonic in all di-
FREQUENCY, Hz rections. The airblast from the 17-ms
FIGURE 4.-Predominant frequencies of transverse ground design was 7 dB higher in the north array
vibrations. and 6 dB higher in the west array, but no

)- 0

co 0
w 140 0 w 140
d 0::
::> (/)
if) (/)
w w
0:: 0::
(L 0...
0:: 0::
w w
0 >
1- 1-
lf) (/)
<( <(
..J ...J
~ 120 A
Ol 120
KEY o I 7 ms

I! o 17ms
A 42ms
I± standard deviation A
A 42 ms

:! standard deviation 17

A 42

r IOOL-~~---L-L-L~~----~--L-J-~~
10 100 1,000
I 1

FIGURE 5.-Propagation plot showing differences in airblast FIGURE G.-Propagation plot of peak airblast for the west
levels for the two different blast designs. arrays.

different in the other directions. This The delay interval between holes should
would indicate that the reduction in be selected such that the trace velocity
airblast is attributable to the fact along the free face is subsonic. This
that the trace velocity along the free resulted in a reduction of airblast of up

I face was subsonic for the longer delay

to 6 dB in these tests. The delay inter-
val between holes did not affect ground

The two blast designs also show some vibration amplitudes in these tests.
difference in the predominant frequencies
of the airblast. The design using 17-ms DELAY INTERVALS BETWEEN ROWS
delays had more airblast energy in the
10-Hz range than the 42-ms delay design, Shots 24 through 52 used the same 17-ms
as shown in figure 9. delay between holes in a row (the average
Statistical analysis showed no signifi- value of the actual delay interval be-
cant differences in the ground vibration tween holes was 23 ms), but the delays
levels from the two blast designs. How- between the rows were varied, in five
ever, spectral analysis did show a dif- steps from 30 to 100 ms. The shot pat-
ference in the predominant frequencies tern was 33 ft square, shot en echelon,
of the two designs (fig. 10). The 17-ms giving an effective burden of 23 ft and
design has its predominant frequencfes effective spacing of 47 ft. Five dif-
around 10 Hz, while the 42-ms design ferent delay intervals between rows
has more scatter in its predominant were used to study the effect of burden
frequencies. delay timing on vibration levels. The

ro 0
~ 0 """ 140 0
w 140 w
0::: 0:::
:J :J 0
{f) (/)
(/) (/)
w w
0::: 0:::
a. !::.
0::: 0:::
w w
0 >
f- f-
(/) (/)

<( <(
120 KEY
ro 120 - rn
0::: 0::: o 17 ms
<! KEY
<( t::. 42 ms
0 I 7 ms I'
:!: standard deviation
t::. 42 ms I
I :!: standard deviation

100 ····~ ... _______l___L. .. 100

10 100 1,000 10 100 1,000
I 11

FIGURE 7.-Propagation plot of peak airblast for the north FIGURE B.-Propagation plot of peak airblast for the east
arrays. airways.

intervals used were 42 ms, which was the spoils) had the lowest vibration levels.
delay used by the mine, and 30, 60, 75, The highest levels were toward the west,
and 100 ms. The 42 ms was a pyrotechnic where the ground was undisturbed. The
delay, while the others were selected us- vibration levels of the other arrays were
ing a sequential blasting machine. Table intermediate between these levels. The
1 gives values of burden relief for the western and northern vibration arrays
different burden delays used. were chosen for further analysis.
Vibration data for each design were Vibration levels of the different de-
compared to determine if direction of signs were compared for the north and
orientation of the seismograph array was west arrays using regression analysis
important. The eastern array (in the (figs. 11-12). Table 2 gives values of
intercepts for regression lines with com-
TABLE 1. - Effective values of burden mon slopes and shows that the two longest
delay intervals 1 delay intervals resulted in the lowest
vibration levels.
Shot Delay interval, ms Burden relief, Analysis of the regression lines shows
Nominal Actual ms/ft (actual) that only the two longest delay inter-
42-44 30 27.5 1.2 vals resulted in significant reductions
24-36 42 48.5 2.1 of ground vibration. This is probably
37-41 60 58.5 2.5 the result of reduced confinement be-
45-49 75 76.0 3.3 cause sufficient time was allowed by the
50-52 100 99.5 4.3 longer delays to move the burden before
Effect1ve burden was 23 ft for all the next echelon of holes was initiated.
shots. This study found an average vibration
70 70 17·ms delays

60 17-ms delay
30 -u
20 ~

a. 10
z 20
w 0:::
u 0 0::: 10
z ::J
w u
0::: u 0
0::: 0
::::> 60
u 50
0 42-ms delay 42-ms delays




0 5 10 15 20 25 30 35 0 5 10 15 20 25 30
FIGURE 9.-Histograms showing the spectra differences of FIGURE 10.-Histograms showing the spectra differences of
airblast for the two designs. ground vibrations for the two designs.

TABLE 2. - Comparison of regression lines for various burden delay intervals

Burden delay Regression line Regression line

Shot interval, ms Array direction I Slope Intercept with common slop_e
Slope Intercept
42-44 •••••• 30 North • .•••••••.••••.•• -1.69 214 -1.50 116
24-36 •••••• 42 ••• do • •••••••••••••••• -1.29 53 -1.50 104
3 7-41 •••••• 60 •.• do • .•.•••.••••••••• -1.44 102 -1.50 122
45-49 •••••• 75 ••• do ••••••••••••••••• -1.74 212 -1.50 87
'50-52 •••••• 100 ••• do ........... "'••••••• -1.41 51 -1.50 71
42-44 •••••• 30 West • ••••••••••••••••• -1.11 42 -1.25 71
24-36 •••••• 42 .•• do • ••.••.•.••.••.•• -1.27 77 -1.25 73
3 7-41 •••••• 60 .•. do . ••.•.•..•....••. -1.47 171 -1.25 80
45-49 •••••• 75 •.• do • .••••••• ~ •..•.•• -1.06 30 -1.25 63
50-52 •••••• 100 ••. do • •..•.•....••.••• -1.03 19 -1.25 50

+ 30 ms
o 42 ms A 30 ms
o 42 ms
"' 60ms 0 o 60 ms
x 75 ms
...... o 100 ms "'c
...... " 75 ms
c <> 100 ms
.; I :!:: standard
deviation >-
:!: standard
...... deviation
(.) 1.00 g 1.00
0 w
w >
> w
w (.)
(.) 1-
i= <(
0::: ¢X a.


.10 -42
~100 .0 5 L_L_l_LIL___ __,l__ _L_-.L__,l___.L.L.UJ.--:----'-----'----'----'
5 10 500
JQ 100 I 500 1

FIGURE 11.-Propagation plot of peak particle velocity show· FIGURE 12.-Propagation plot of peak particle velocity show·
ing differences caused by burden timing, for the north array. ing differences caused by burden timing, for the west array.

reduction of 30 pet using the 100-ms de- predominant influence on the frequency of
sign compared with the 42-ms design nor- vibrations.
mally used by the mine. Airblast was also analyzed, but no sig-
The frequency content of the ground vi- nificant differences were found in fre
brations was also studied. No reduction quency or amplitude a~ ~delays between
in low-frequency vibrations was observed, rows were varied.
which suggests that geology was the


Careful attention to blast design prac- of the airblast wave fronts from the in-
tices can help reduce airblast and ground dividual holes. Care must also be taken
vibrations generated by mine blasting. to avoid selection of delay intervals
This study examined blasthole delay in- that can cause airblast frequencies equal
tervals and their effects on airblast and to the natural frequencies of midwalls of
ground vibrations. nearby structures (about 11 to 25 Hz).
Airblast was influenced by the trace Delay intervals of less than 40 ms will
velocity along the free face. To reduce usually not present a problem.
airblast, the trace velocity, which is a Orientation of the blast and direction
function of delay interval and spacing of initiation had a noticeable effect on
between holes in an echelon, should be the magnitude of vibrations. Vibration
less than the speed of sound in air. levels in the direction of initiation
Airblast was reduced by about 6 dB by were about twice the level of those away
choosing delays giving a trace velocity from the direction of initiation. Vibra-
of 80 pet of the speed of sound rather tion levels across the pit from the blast
than a supersonic velocity. were also lower.
Delays between holes in each row or Vibration levels were also dependent on
echelon should be greater than 1 ms/ft of the delay interval between rows. Ade-
spacing, in order to prevent reinforcing quate time must be provided for burden

relief for each row. This investigation The timing of delay intervals between
found that the delay interval between rows had no influence on the frequency
rows should be as long as practical for content of the ground vibrations. Geol-
the burden involved. The longest burden ogy was the controlling factor for pre-
relief value, 4.3 ms/ft, gave the lowest dominant frequencies of ground vibrations
vibration levels. in this investigation.


Duvall, W. I., C. F. Johnson, A. V. Vibration and Airblast From Surface Coal
c. Meyer, and J • F. Devine. Vibra- Mine Blasting. BuMines RI 9026, 1986,
tions From Instantaneous and Millisecond- 44 PP•
Delayed Quarry Blasts. BuMines RI 6151, 4. Reil, J. W., D. A. Anderson, A. P.
1963, 34 PP• Ritter, D. A. Clark, S. R. Winzer, and
2. Siskind, D. E., M. S. Stagg, J. W. A. J. Petro. Geologic Factors Affect-
Kopp, and C. H. Dowding. Structure Re- ing Vibration From Surface Mine Blast-
sponse and Damage Produced by Ground Vi- ing (contract H0222009, Vibra-Tech Eng.,
bration From Surface Mine Blasting. Bu- Inc.). BuMines OFR 33-86, 1985, 204 pp.;
Mines RI 8507) 1980, 74 pp. NTIS PB 86-175858.
3. Kopp, J. W., and D. E. Siskind.
Effects of Millisecond-Delay Intervals on


By David E. Siskind1 and Vigil J. Stachura1


Vibration wave frequency from surface durations, and lower-than-normal attenua-

mine blasting is an important influence tions of amplitude with distance. The
on the response and potential cracking of observed low-frequency waves were consis-
nearby structures. The Bureau of Mines tent with predictive theoretical models
studied blasting vibrations in a midwest- of surface wave generation using the
ern coal mine that occasionally produces depths to the old mines.
4-Hz surface waves in its production Blast designs also contributed to the
blasting and has received numerous com- vibrations problem. The complex multide-
plaints from neighbors. The mine and layed blasts generated vibration ampli-
nearby town are underlain by abandoned tudes up to three times those of same-
underground coal mines 85 to 325 ft below weight-per-delay single charges, particu-
the surface. larly for the echelon designs. By con-
Blast vibration measurements at the trast, the heavy casting blasts generated
site and analysis of mine and regulatory more of the unusual low frequencies. Be-
agency records indicated that the propa- cause of these low-frequency, long-period
gating medium was primarily responsible waves, the widely adopted 8-ms minimum
for the vibration wave characteris- charge separation criterion may not apply
tics, including low frequencies, long at this site.


The Bureau of Mines studied a site at some vibration records showed data unlike
the western Indiana town of Blanford, any from previous Bureau studies, with
where surface coal mine blasting was pro- very low frequencies down to 4 Hz and
ducing unusual low-frequency, long- long durations of over 4 s.
duration vibrations. At the request of Causes for the unusual data were then
two regulatory agencies, the Indiana sought, including the ground and struc-
Department of Natural Resources (DNR) and ture conditions and factors in the sur-
the U.S. Office of Surface Mining (OSM), face blast designs. By identifying rela-
the Bureau investigated the influence of tive influences, design guidance could be
the ground structure, which includes ex- provided to minimize the low frequencies
tensive abandoned underground workings generated at this and similar sites.
under both the active mining and the town Seismologists have recognized that low-
of Blanford. These abandoned workings frequency vibrations are produced in
are at several depths, the most signifi- soft, weak materials with l~w propagation
cant being the extensively mined coalbed velocity (1).2 Thick soil layers, fill
No. 5 at 225 ft and the partia1ly mined areas, and-weak sedimentaryJrocks are ex-
No. 4 at 325 ft (fig. 1). The local sed- amples of such materials. Furthermore,
imentary rocks and the coal seams are es- such layering typically has contrasting
sentially horizontal. acoustic properties and produces surface
The initial objective of the study was waves of several types. These waves are
to determine if blasting vibrations at inherently low in frequency and develop
this site were unusual, as local home- from multiple reflections of conventional
owners claimed. Researchers noticed that body waves as they propagate in the rock

st. ~Underlined numbers in parentheses re-

Twin Cities Research Center, Bureau of fer to items in the list of references at
<lines, Minneapolis, MN. the end of this paper.

/~..... -1
I ./
/'I c:~,-":J
I 1 r
-.J ~- - ,
I '"1
I r 1


Boundary of coolbed 4 mining

---- Boundary of coalbed 5 mining
=Roads and streets
I 0 1000
'- _r-\.-""\
Scale, ft
I '-,
I '-
II \
..._ __
I '

FIGURE 1.-Town of Blanford, current surface mining highwall, abandoned underground workings, and some of the homes studied
for vibrations and settlement.

layers. Extensive horizontal mine work- 9 ms (as opposed to 17 and 34 ms) pro-
ings also provide a strong reflecting duced lower frequencies.
layer. Research by Anderson (l-4) and Reil (i)
Recent research suggests the selection investigated the use of single-charge
of delay intervals as a useful method for blasts to "calibrate" a site and providE
controlling blast vibration frequencies. a model record for seismogram syntheses.
Early research by the Bureau investigated Anderson developed a "Fourmap" scheme,
delay effects on vibration amplitudes which uses superposition of appropriate!;
(2). However, these researchers also delayed single-charge time histories that
noticed that the shorter delays of 0 and are then converted into Fourier spectr•

(~). Accurate delay times are assumed. frequency; however, he did notice the
This method graphically shows the occurrence of constructive wave interfer-
relationship between delays and pre- ence when using the 9-ms delay separa-
dieted spectral distributions and is tion, but. not when using 17-ms delays
used commercially by several blasting (7). Kopp specifically sought the influ-
consultants. ence of delay times on frequency for coal
Contrasting with Anderson's and Reil's mine blasts (~). He concluded that vi-
work in rock quarries, a small amount of bration frequencies did not generally re-
work was done by the Bureau in surface flect the between-row or within-row delay
coal mines using standard production intervals. However, his closest measure-
blasts. Here, the thick soil cover and ments at this thick overburden site were
relatively weak and complex sedimentary at 300 ft, and it is likely that, at that
rock layers represent a far more influen- distance and in that material, the pri-
tial propagation medium. Absorption, mary influence on the waves is the propa-
dispersion, and the generation of second- gation medium itself. In the light of
ary waves through multiple reflection current knowledge, Kopp's field program
quickly complicate the wave character and in 1981-82 should have included closer-in
make it more difficult to control through monitoring and tests with more precise
blast design. These conditions, plus the initiation timing.
larger holes and charges, tend to gener- This paper briefly summarizes an in-
ate lower frequencies, which are less vestigation in the Blanford area reported
amenable to change. more fully in a recent Bureau Report of
Wiss' extensive 1979 coal study did not Investigations (1) and earlier as an un-
specifically examine delays and vibration published Bureau report to OSM.
Vibration records from 235 production attenuation of particle velocity ampli-
blasts were available for analyses of tudes. Researchers did this for each
amplitudes, frequencies, and total dura- home, for various groups (by neighbor-
tions. These were originally collected hoods), and as an overall summary for all
by the Indiana DNR and the surface mine homes. Propagation plots were also pre-
(Peabody Coal Co., operating the Univer- pared for various blast designs and, in
sal Mine) at seven Blanford homes over a the report to OSM, according to vibration
period of 11 months. Distances of the frequency characteristics.
homes from the blast source ranged from In addition to vibration amplitudes,
about 980 to 9,800 ft, and as many as six frequency characteristics recorded at
homes were monitored at one time various homes and with various shot de-
(fig. 1). signs were compared. Single charges and
For purposes of identifying the genera- close-in monitoring revealed the ground's
tion and propagation influences, Bureau natural response frequency of 8 Hz. Shot
researchers instrumented seven blasts, design data were compared for the four
including two specially fired single- blasting methods employed by Peabody dur-
charge shots. For these tests, measure- ing the period covered by the records.
ments were made as close as 56 ft to de- This allowed identification of delay-
termine the wave characteristics before sequencing influences on vibration fre-
they were modified by the propagation quency, as opposed to ground influence.
medium. The single charges were used to Other site data were collected to iden-
identify the specific vibration- tify areas around Blanford and the nearby
generation influences of the complex surface mine that had been previously un-
multideck, multidelayed blasts, as con- dermined. Depths to these deep coalbeds
trasted to the simple vibration sources were available from Peabody, which is
represented by the single-charge blasts. conducting an extensive study of one
All vibration records were plotted for Blanford home. Bureau researchers con-
propagation showing the generation and ducted level-loop surveys of eight homes

in September 1985 and the following April determine if observed low frequencies
to identify possible blast-vibration-in- were consistent with depths to the old
duced subsidence and longtime stability. workings (IQ).
Finally, a predictive model was used to


VIBRATION AMPLITUDES intervals. Echelon blast amplitudes ex-

ceeded the historical coal mine summary
Propagation plots based on the Peabody
and DNR reords revealed higher vibration
amplitudes than were found in previous
Maximum velocify1
Bureau studies at surface coal mines (1). Rl8507, cool
Figure 2 is a summary of Peabody- ;nd
DNR-collected maximum velocities. The
data are clustered, having been collected ~

"'c 1.00
at seven Blanford homes rather than with ,..
widely spaced seismograph arrays. Most 1-
notable is that virtually all the mea- _J
surements exceeded the mean of the pre- w
vious coal mine vibration summary (shown _J

as maximum velocity, RI 8507, coal). a: KEY
Some even exceeded the envelope line of "
0. .10 o 17 by lOOms

highest measurements, which is approxi- w

"'" a 17 by 200 ms
Q. a 17+42 by 200 IllS

mately two standard deviations (2cr) above v 11 by 42 ms

the mean.
Separate propagation plots were also
made for the various blast designs (IQ). '0 I __L_-"-.-'--'~- ,LUL__ _.__,__._L...LL UL__L__L_ll-L"-l.JW
I 10 100 1,000
Figures 3 and 4 summarize these plots SQUARE ROOT SCALED DISTANCE, ft/1~ 12
based on square root scaled propagation
FIGURE 3.-Propagation plot of Peabody and DNA data for
and all charge weights within 8-ms echelon pmduction blasts.

Maximum velocity1
Rf 8507, cool

1.00 --
,..- .... 1.00
u ,:
0 1-
..J u
w 0
> .J
w w
..J >
~ w
f- .J
0:: u
"- .10 I-
;t .I 0
"- ~ ~

10 100 1,000 ...LL-1-...LL.UL. -'-'-'-'lu.l'-1- - ' -

SQUARE ROOT SCALED DISTANCE, fl/lb 112 10 100 1,000
FIGURE 2.-Propagation plot of Peabody and DNA data for all
homes compared with surface coal mine blasting data summary FIGURE 4.-Propagation plot of Peabody and DNA data for
from Rl 8507 (1). casting production blasts.

TABLE 1. - Production blast designs used at the Universal Mine,

July 1984 through April 1985

Design type Number Charge weight per delay, lb

and delays, ms l ,2 of decks Typical Exception or maximum
17 by 42 • •.••.•..•• 1 1,500 2, 258 maximum.
17 by 100 • ••••••••• 2-4 325 Exception: 625 on 4-22-85
17 by 200 • ••.••.••. 2-4 200- 400 Exception: 1, 4 7 5 on l-05-85
l' 911 on 1-12-85
Casting: 8, 10, or
12.5 by 140 to 210 •• 1 2,000 3, 842 maximum.
17 by 42 ind1cates 17-ms delays between holes 1n a row and 42-ros de-
lay between rows.
2 The 17+42 by 200-ros echelon had too few values to analyze separately.

from RI 8507 more so than did the casting 2. The casting blast vibration ampli-
blasts. This appears to be related to tudes, although high, were closest to the
the blast complexity. Both the casting historical data of the various blast de-
shots and the 17- by 42-ros echelon (17 ms signs plotted.
between holes in a row and 42 ms between 3. If the high amplitudes were from
rows) were full-column charges. The low surface wave attenuation, the cast-
other echelon blasts had 2 to 4 decks ing blasts would be the worst case, not
(table 1). The shots using full-column the least abnormal. Apparently, the
charges produced vibrations close to the problem with the casting blasts was not
historical mean, while the other three their amplitudes but their unusual low
echelon designs produced most of the frequencies and long durations.
higher amplitude vibrations.
Although decking is used to reduce the Production and single-charge data were
charge weight per delay, the increased obtained on-site from the widely spaced
complexity and resulting charge interac-
tions prevent a corresponding decrease in
vibration amplitude. Hence, the vibra-
array of seismographs (fig. 5). Vibra-
tion propagations from the single charges
were close to those of previous Bureau I
tion observed from these blasts was
strong relative to the charge weights per
delay. These charge weights per delay
were computed in the traditional manner,
studies. However, the produ_ction vibra-
tions, for the same charge weight per de-
lay, averaged three times higher. Evi-
dently, vibrations from the delayed
based on a minimum time separation of 8 charges in the multihole, multideck pro-
ms. Apparently, this amount of sepa- duction blasts were interacting. The
ration is insufficient to separate ef- somewhat shallower line slopes for the
fects for such low-frequency, long-period propagation means compared with those
waves. from previous studies summarized in
It is logical to question if the ab- RI 8507, are indicative of a low-vibra-
normal amplitudes resulted from the domi- tion attenuation with distance. This is
nance of low-frequency and inherently likely from surface waves being gener-
low-attenuation surface waves instead of ated at the longer propagation distances.
constructive wave interference between These results are consistent with the
delayed charges. Surface waves were previous measurements. Based on con-
judged a minor influence on vibration am- structive wave interference for various
plitudes (if at all) through the follow- designs used by Peabody, the number of
ing analysis: interacting charges were estimated from
the delay sequencing shown in figures 6
1. Most of the very low frequency through 10 (table 2). Sequence times are
cases were from casting blasts. based on nominal delays as shown on the

TABLE 2. - Analysis of charge weights

for production blasts
(Maximum number of charges
per time interval)

Design type and 8ms 17 ms 60 ms

delays, ms
1- Echelon:
u 1.00
17 by 42 . •.••...••. 2 3 7
> 17 by 100 • ••••••••• 2 3 9
u 17 by 200 •.•• ••.••. 2 3 7
a: Casting:
KEY Short array •..•.... l 2 5
~ .fO
o Single-charge shot dafo Long array ••••••••• 3 6 13
o Production shot dala

Researchers compared vibration rec-

ords for various blast designs and moni-
toring locations (figs. 7-10). Although
the larger casting blasts frequently pro-
.0 I L..~.......l........l.-'-..J....J...'-'-l.'------'---'-'-..J....J...J .. LLL_--'--'-..(_J_-"-'-ill duced low frequencies, the three tested
10 100 1,000
SQUARE ROOT SCALED DISTANCE, ft/lbl;2 echelon designs occasionally also did so.
Generally, periodicities in the blast
FIGURE 5.-Propagation plots of recent Bureau tests (9) com-
oared with surface coal mine summary from Rl 8507 (1). design, such as between-row spacing, were
identifiable in some of the records (fig.
6) but not in others. At the larger dis-
tances of measurement, the vibrations
tops of figures 6 through 10 and the oc- were strongly influenced by the propaga-
currence of constructive interference tion medium. Researchers believe that
when delays are within one-half the sinu- low-frequency cases can be moderated by
sodial vibration period. minimizing the amount of energy at such
frquency that is produced by the blast
FREQUENCY OF VIBRATIONS design as an energy source (_~)·

Production Blasts as Vibration Vibration Wave Types

Source Functions
The researchers attempted to identify
All vibration measurements available as the types of waves observed, based on ex-
time history records were analyzed for pected components and their phases.
frequency content. Some blasts had very Rayleigh waves are vertically polarized
low frequencies of about 4 Hz on all with retrograde elliptical particle mo-
three components (vertical, longitudinal, tions. They have significant motion in
and transverse) (fig. 6). Other blasts the longitudinal and vertical directions,
had from zero to two components with such and little in the transverse. The gener-
low frequencies. The source function for ation of these waves requires only a
the blast shown in figure 6 lasted about single free surface such as the ground-
0.7 s, which is also the approximate dur- air interface. Rayleigh waves can also
ation of the high-frequency part of the propagate where there is a sharp acoustic
vibration record. The low-frequency wave contracting layer at depth.
began at about 1.7 sand lasted at least Love waves are horizontally polarized
2 s, for a total vibration duration of shear waves. They are strong only in the
abollt 4 s at the measurement distance of transverse direction. Generation of Love
7,5~0 ft. waves requires a layer with top and


. . ..., . ... ,.
~~ '52'4 494 464 449 434 419 404 389 374 359 329 299 269 239 209
306 292 278 264 250 <22 208
o------;,..... I
510 496 466 ..0 426 412 ~09 479

724 710 696 682

••• 60. 640 626 694 681 667 654 640 627 613

••• .,,
-;79 ••
906 894 880 838 824 810


464 ms
Ll_IL...JI'--1-LI-LI-LI-L.....L......L...L..L...LI-LI.....L.I_IL.J'--1--L---L--L-----~.~._Il!.st row
208ms 412 ms S26ms
1 1 1
Fi rs~ r~~------L-'--:!s!--e""co_n.Ld,-rLo-wL-"------,T>"hi'""rd~roL-w
1 I I I I I I I I I Second row

0 _____.______.. 0.1 Third row

Time, s
! I l l ! ! I! II !II lllltlll f Ill Ill Ill Ill II Ill I f I ! All
Time, s
~ Tron'Sverse =0.051 in/s

Vertical' 0.072 in/s

Longitudinal' 0.082 in/s

Longitudinal• 0.15 in/s

0 I 0 I
-------Time,$ ,......., ~
Time 1 s


FIGURE 6.-Casting blasts, 200 ms between rows, 10 ms be- FIGURE 7.-Casting blasts, 200 ms between rows, 10 ms be-
tween holes in a row, January 21, 1985. tween holes in a row, February 16, 1985.



All 0 0.1
0 0.1 Time, s

Transverse • 0.39 in/t

Transverse • 0.05 7 in/s

Vertico I ' 0.22 in /s

r-., Vert I col • 0.018 in/s
J '-;------~--------------------
Longiludinol • 0.35 in/s

Longitudinal •0.062 in/s -~----

0 I 0 I
liiiiiiiiiil ......---, ........-~
Time, s Time. s

FIGURE a.-Echelon blast, 42 ms between rows, 17 ms be- FIGURE 9.-Echelon blast, 100 ms between rows, 17 ms be-
tween holes in a row, January 4, 1985. tween holes in a row, March 4, 1985.

- --~- DECK
1:1\ 3.0 (in/s)/in

65fl-''i'i~ 0
!'I ~~~
1,371 4,431 ,,..
No6 i! Scale, in
583 787 99l 1,195 1,.399 "'t4~'S .• '!
!',1 ~·.
1.0 (in/s)/in
1 1 1 Second deck 226 ft··-!fil':('t'~~
L.LJ_ ~-~I! lll II t I 11!!1 II I All II
first row Second rO'iii' Third row Fourth row l
(one hole) (one hole) 0.2 \, ..J: f. .. 0.5 (in/s)/in
SEQUENCE OF CHARGES Time, s 407ft-:.•:;,l\i\/v"JVV'~
,,.y if
------ .


Transverse~ 0.072 in/s

VerUcal :0.033 in/s


LonQiludinal =0.053 in/s

0 I
Time, s
0.3 (in/s)/in

FIGURE 10.-Echelon blast, 200 ms between rows, 17 ms be·

tween holes in a row, January 12, 1985.

bottom boundaries raving good reflect- ··. 0. I (in/s)/in

~'·\·- • .: /~-..--,"-",AI~/'.r>p.,\.......,_......,/"1.-..,...........-L.</"---..•~~----_."-• .. -.""""~
ing properties. Extensive underground

voids could provide such a reflecting
surface, as could any low-velocity layer. FIGURE 11.-Vibration records from single charge, vertical.
Horizontal scale is 500 ms/in.
The amounts of low frequency appeared to
be related to the distance from the
blast; however, trends were not consis- the site and the blasting. Table 3 sum-
tent from site to site. marizes all available time history
Site Differences for Vibration Waves
Single-Charge Blasts
Six of the seven measurement sites
had patterns of low frequencies resem- The two special single-charge shots
bling Rayleigh waves or Rayleigh and were fired to identify the blast initi-
Love waves together. Two of these sites ation-sequencing influences on the wave
also occasionally had only high frequen- forms as well as the previously discussed
cies (e.g., 20Hz) at the same time and vibration amplitudes. They also showed
for the same shots when low frequencies the complexity of the propagating medium.
were being measured elsewhere. One site Figure 11 is the vertical vibration rec-
(Jackson) had every kind of combination ord from a 1-m column of explosive that
at different times. The seventh site took about 0.3 ms to fully detonate. Af-
(Verhonik) had only transverse low fre- ter propagating 65 ft, the vibration
quencies, suggesting Love waves. This duration was over 150 ms. At a distance
seventh site was in a different location of 1,165 ft, the vibration duration was
from the others, east of the blasting dominated by a 6.5-Hz wave lasting about
rather than north. There was a 100- 2 s. This observation suggests that any
ft-thick layer of mine spoils between blast at this site is a potential

TABLE 3. - Vibration components showing very low frequencies (~4 Hz), 1

by measurement site and blasting method

Date Massa Hollings- Jackson Verhonik

12-18-84 •••••••••••••••• HF
12-19-84 •••••••••••••••• L,V
12-22-84 •••••••••••••••• HF
12-28-84 •••••••••••••••• v HF

1- 4-85 •••••••••••••••• v HF
1- 9-85 •••••••••••••••• L,V HF
3- 2-85 •••.•••••••••••• HF
3- 2-85 •.•••••••••••••• HF
3- 2-85 •••••••••••••••• HF HF HF
3-14-85 ••••.••••••••••• 8 Hz

3-16-85 •••••••••••••••• 8 Hz
3-18-85 •••••••••••••••• HF
3-21-85 ••••.••••••••••• HF
3-25-85 •••••••••••••••• / HF

3-26-85 ••••••.••••••••• HF
9-10-86 ••••..•••••••••• HF HF ~10 Hz
9-10-86 •••••••••••••••• 8 Hz HF ~10 Hz
9-10-86 •••••••••••••••• HF ~5 Hz ~14 Hz

9-12-86 ••••••••••.••••• HF ~9 Hz
9-13-86 . ............ ..
7- 2-84 •••••••••••••••• T
7- 3-84 •••••••••••••••• T
7- 6-84 •••••••••••••••• L,V,T
7- 7-84 •••••••••••••••• L,V,T

7- 7-84 . ..•....••...•.. L,V,T

7-11-84 •••••••••••••••• T
7-11-84oooor••••••••••• T
7-11-84 •••••••••••••••• T

7-13-84 •••••••••••••••• T
7-14-84 •••••••••••••••• T
7-28-84 •••••••••••••••• L,V,T
1- 5-85 •••••••••••••••• HF

1-12-85 •••••••••••••••• L,V

3- 5-85 •••••••••••••••• HF
3- 6-85~ ••••••••••••••• HF L,V,T HF
3- 7-85 •••••••••••••••• HF

3- 9-85 •••••••••••••••• HF
See explanatory notes at end of table.

TABLE 3. - Vibration components showing very low frequencies (~4Hz), 1

by measurement site and blasting method--Continued

Date Massa Jackson Verhonik

8- 9-84 •••••••••••••••• L,V,T

8-13-84 •••••••••••••••• L,V,T
20-30-84 •••••••••••••••• v
11- 5-84 . ....•.......... v
11- 7-84 . .......•..•.... L,V T
11-10-84 ••.••••••••••••• L,V,T L,V,T T
11-15-84 ••••••••••••.••• L,V,T
11-17-84 •..•..•••••••••• L,V,T L,V
11-19-84 .........••..•.. L,V,T L,V,T T
11-23-84 •••••••••..••••• L,V,T L,V T
11-24-84 ••••••••••• ~ •••• L,V
11-26-84 ••.••••••••••••• L,V
12- 1-84 . ..........•.... L,V,T L,V,T L,V,T
12- 4--84 ................. L,V,T
12- 6-34 . ............... L,V,T L,V L,V,T T
12- 8-84 . ............... T
12-10-84 ••••.••••••••••• L,V,T T L,V,T T2
12-12-84 ••••.••••••..••• L,V,T L,V,T T2
12-15-84 •••••••••••••••• v T2
12-17-84 ••••.••••••••••• L,V,T L,V,T

1-12-85 ••••••••••.•.••• L,V T

1-14-85 •••••••••••••••• L,V,T - L,V T
1-17-85 •••••••••••••••• L,V,T L,V L,V,T T
1-21-85 •••...•••.•••... L,V,T L,V,T T

1-25-85 •••••••••••••••• L,V,T L,V,T L,V,T v T

1-28-85 •••••••••••••••• L,V,T L,V,T L,V,T v T
1-31-85 •••••••••••••••• L,V,T L,V L,V L,V T
2- 2-85 •••••••••••••••• L,V,T L,V,T L,V,T L,V,T T

2- 6-85 •••••••••••••••• L,V,T L,V,T L,V,T L,V,T v T

2- 9-85 . ..•.....•..•... L,V,T L,V,T v L,V,T v T
2-14-85 •••••••••••••••• L,V,T L,V,T L,V,T T2
2-16-85 •••••••••••••••• L, V, T L,V,T T2

2-19-85 •••••••••••••••• L,V L,V,T L,V,T T2

2-21-85 •••••••••••••••• L,V L,V L,V
3- 1-85 •••••••••••••••• HF
1 Longitud1nal (or radial). V Vertical. T Transverse.
Rayleigh surface waves would dominate 1 and V components; Love surface waves would
dominate T component.
HF Higher frequency vibration with no clear components below 10 Hz.
2Vertical record may be defective; may have been present, but was not visible•

NOTE.--No entry records available.


TABLE 4. - Summary of two level loop surveys of eight

Blanford residence.

Maximum elevation Maximum angular

House differences, 1 ft distortion 2
Sept. 1985 April 1986 Sept. 1985 April 1986
Ahlmeyers •••• 0.06 0.07 1/340 1/340
Albrecht ••••• .12 .11 1/250 1/260
Finger, E ••• • .08 .07 1/300 1/340
Finger, 0 • ..• .09 .11 1/200 1/164
Jovanovich ••• .30 .26 1/108 1/125
Marietta ••••• .09 .09 1/560 1/533
Skorich •••••• .06 .05 1/250 1/300
Zell • ....•... .37 .39 1/65 1/65
1Accuracy is ±0.01 ft.
2 1/340 = Distortion of 1 part in 340.

low-frequency problem as the ground there wave, h is 167 ft. Depths to the coal-
favors such frequencies. beds as measured by Peabody at one site
ranged from 85 to 394 ft. The exten-
Theoretical Model sively mined No. 5 was at 226 ft.

The O'Brien model computes surface LEVEL-LOOP SURVEYS OF HOMES

waves generated by multiple reflections
of compressive body waves in a low-veloc- Researchers surveyed eight Blanford
ity layer (lQ). Another version by Gupta home foundations to identify surface and
is for shear waves. For a strong veloc- structural changes from possible subsid-
ity contrast (strong reflector), the sim- ence over abandoned workings (2). Only
plified relationship is one of the eight homes was also monitored
for vibration (Zell house). The primary
T 4h,
= _ selection criterion was the existence of
vl a clear foundation horizon for surveying.
Most of the homes were out of level by
where T is the surface wave period, h is significant amounts of up to one part in
the layer thickness, and vl is the propa- 65 (table 4). Boscardin cites that angu-
gation velocity for the low-velocity lay- lar distortions or deflection ratios of 1
er. Assumed for this simple model is part in 300 can cause cracking in panel
that vl << v2, with v2 being the high- and load-bearing walls (ll)· A resurvey
velocity layer. 7 months later showed no appreciable ele-
At Blanford, researchers measured a V2 vation changes. With these data alone,
of 10,000 ft/s and, using a1.6-s arri- it is not possible to tell if the homes
val time difference, calculated a V1 of are distorted or strained, or if these
2,700 ft/s. For a 6.5-Hz surface wave differences were originally built in.
(T = 0.15 s), an h of about 102ft is in- Apparently, no significant long-term
dicated. Similarly, for a 4-Hz surface changes are occurring.


The propagating medium appears pri- vibration attenuation (higher amplitudes)

marily responsible for the adverse vibra- with distance compared with other coal
tion impacts in Blanford, through three mine sites, and (3) it produces interac-
mechanisms: (1) It favors generation of tions between delayed charges beyond
low-frequency surface waves of several those expected from the blasts as de-
types, with frequencies between 4 and 10 signed, because of constructive wave in-
Hz, (2) it has the appearance of reduced terference for these long-period waves.

Although further study of the subsur- rapid vibratiion amplitude attenuation

face conditions is needed in order to observed between 407 and 807 ft (fig. 5).
completely understand all the factors, Blast designs are also significant.
the observed surface waves are consistent The widely adopted 8-ms charge-separation
with a strongly reflecting subsurface in- criterion appears not to apply for this
terface at a depth of about 175 ft, or low-frequency site, as it was also pre-
about the same as the depth of the exten- viously found suspect in a 1979 study by
sively mined No. 5 coalbed. This agrees Wiss of large-scale surface coal mine
with theoretical models that predict low- blasting (~). Vibration frequency char-
frequency waves from strongly reflecting acteristics appear to reflect periodici-
near-surface horizontal layers. ties in the blast design timing, partic-
The existing geologic structure between ularly close to the shot. The larger
the main part of Blanford and the active casting blasts produced the clearest 4-Hz
pit is another possible or contributing surface waves. By contrast, the more
cause of the low-frequency blast vibra- complex echelon blasts produced the larg-
tions. This region has a coal cutout or est vibration amplitudes for a given
zone where the coal and other sedimentary charge weight per delay. Results suggest
rock beds are missing, replaced by fill that longer delays be used between
characterized as sandy, gravelly drift. charges to'prevent constructive addition
Vibrations propagating through such mate- of vibration amplitudes for such low-
rial often have abnormally low frequency. frequency cases.
Such a medium could also explain the


1. Siskind, D. E., M. S. Stagg, J. W. 5. Reil, J. H., D. A. Anderson, A. P.

Kopp, and c. H. Dowding. Structure Re- Ritter, D. A. Clark, S. R. Winzer, and
sponse and Damage Produced by Ground Vi- A. J. Petro. Geologic Factors Affecting
bration From Surface Mine Blasting. Bu- Vibrations From Surface Mine Blasting
Mines RI 8507, 1980, 74 PP• (contract 80222009, Vibra-Tech Eng.,
2. Duvall, W. I., C. F. Johnson, A. V. Inc.). BuMines OFR 33-86, 1985, 204 pp.;
C. Meyer, and J. F. Devine. Vibrations NTIS-PB 86-175858.
From Instantaneous and Millisecond- 6. Anderson, D. A., S. R. Winzer, and
Delayed Quarry Blasts. BuMines RI 6151, A. P. Ritter. Synthetic Delay Versus
1963, 34 PP• Frequency Plots for Predicting Ground Vi-
3. Anderson, D, A., S. R. Winzer, and bration From Blasting. Paper in Proceed-
A. P. Ritter. Blast Design for Optimiz- ings of the Third International Confer-
ing Fragmentation While Controlling Fre- ence on Computer-Aided Seismic Analysis
quency of Ground Vibration. Paper in and Discrimination (Catholic University,
Proceedings of Eighth Annual Conference Washington, DC, June 16-17, 1983), lEE
on Explosives and Blasting Technique, ed. Computer Society Press, Silver Spring,
by C. J. Konya (New Orleans, LA, Jan. 31- MD, 1983, pp. 70-74.
Feb. 4, 1982), Soc. Explos. Eng., Mont- 7. Wiss, J. F., and P. Linehan. Con-
ville, OH, 1982, PP• 69-89. trol of Vibration and Blast Noise Frorr.
4. Anderson, D. A., A. P. Ritter, S. Surface Coal Mining (contract J0255022,
R. Winzer, and J. W. Reil. A Method for Wiss, Janney, Elstner and Associates,
Site-Specific Prediction and Control of Inc.). Volume II. BuMines OFR 103(2)-
Ground Vibration From Blasting. Paper in 79, 1978, 280 pp.; NTIS PB 299 888.
Proceedings of First Mini-Symposium on 8. Kopp, J. w., and D. E. Siskind.
Explosives and Blasting Research, ed. by Effects of Millisecond-Delay Intervals ot
c. J. Konya (San Diego, CA, Jan. 31-Feb. Vibration and Airblast From Surface Coa:
1, 1985). Soc. Explos. Eng., Montville, Mine Blasting. BuMines RI 9026, 1986
OH, 1985, pp. 28-43. 44 PP·

9. Siskind, D. E., V. J. Stachura,

Ge0phy. Prospect., v. 5, 1957, pp. 371-
and M. J. Nutting. Low-Frequency Vibra- 380.
tions Produced by Surface Mine Blasting
Over Abandoned Underground Mines. Bu-
11. Boscardin, M. D. Building Response
Mines RI 9078, 1987. to Excitation-Induced Ground Movements.
Ph.D. Thesis, Univ. IL, Urbana-Cham-
10. O'Brien, P. N. S. Multiply-Re- paign, IL, 1980, 279 pp.
flected Refractions in a Shallow Layer.


By Stephen A. Rholl1


A computer model of rock motion due to to simulate bench blasting during full-
blasting is presented. The code, CAROM, scale fLagmentation tests at a neaLby
was developed at Sandia National Labora- rock quaLry. Results of the code are
tories (SNL) to predict rock motion and shown for the fiLst 2 s following explo-
final muckpile distribution. Researchers sive detonation.
at the Bureau of Mines applied the code
A contributing factor to the overall which can quickly analyze technical
efficiency of surface mining is the final problems, involves computer-simulated
muckpile distribution of the blasted rock blasting.
fragments. This is especially true in Much of the early work to develop sim-
mining operations utilizing overburden ple and effective computer models to sim-
casting to reduce stripping costs. As ulate blasting was done by SNL. In 1983,
the overburden-to-coal ratio increases, researchers developed BUMP (!),2 the
it is essential, in many operations, that first code to use very simple interaction
as much fragmented rock as possible be laws to reduce computational times. A
thrown into the pit or onto the spoil stabilized and improved code, CAROM (1),
bank. was later introduced. Both codes weLe
To optimize overburden casting, mine written to study rock motion in oil
operators most often vary powder factor, shale crater tests.
explosive type, drill patterns, and/or The Bureau of Mines initiated contract
delay timing. These variations are nor- J02450ll with SNL to modify these codes
mally tested by trial and error based on for u~e in modeling bench blasting. Sub-
the experience of blasting personnel. sequently CAROM was modified and used in
Unfortunately, this can often be time blast design research. SNL staff also
consuming and expensive. An alternative provided appropriate values for input
approach to the optimization problem, parameters to execute the code.


The CAROM code is a two-dimensional influence on the calculated material mo-

distinct-element code. That is, a group tion for modeling oil shale cratering
of distinct elements is used to describe experiments.
the system of rock fragments undergoing The initial configuration of the ele-
motion. The shape of these elements is ments used in the CAROM code is specified
in theory arbitrary, but circles are most by the user. In addition, the initial
often used to simplify time-of-collision conditions (both the velocity and accel-
calculations. The size of the elements eration of each element) must likewise be
is unrestricted, and often a single specified. CAROM simply lets the ele-
element is used to represent a group of ments move about until the code detects a
fragments. Studies by Gorham-Bergeron collision between two elements. The col-
(1) have shown that the size of the lisions are handled by the theories of
material elements does not have a strong classical mechanics, namely, conservation

Research physicist, Twin Cities Re- numbers in parentheses re-

search Center, Bureau of Mines, Minneap- fer to items in the list of references at
olis, MN. the end of this paper.

of linear momentum and Newton's collision Although it is possible to include
rule. The collisions themselves can be friction in the calculations, CAROM usu-
treated elastically or inelastically, al- ally assumes that the coefficient of
though the latter is more often assumed. friction between the elements is zero.
Because CAROM calculates the time the However, it should be noted th~t at the
next collision will occur, the code is boundaries representing the pit floor,
inherently stable. When a collision is the elements are not allowed to slide.
detected, CAROM redistributes the linear Too much sliding was deemed unrealistic,
momentum of the elements and then repeats and therefore the code assumes infinite
the process. Collisions with a fixed friction between the elements and the pit
boundary (or boundaries) are also permit- floor boundaries.
ted and treated in a similar way.


The CAROM code was applied by Bureau Research at SNL has shown that a dynam-
researchers to simulate rock motion dur- ic finite-element code, DYNA2D (3), can
ing bench blasting at a local quarry. A be used to describe the explosive detona-
two-dimensional discrete rock model de- tion, rock fragmentation, borehole pres-
scribing the 22-ft highwall is shown in sures, and initial rock motion. The in-
figure l. A series of 33 elements, each formation calculated by DYNA2D can be
with a radius of 1 ft, represent the used as the input necessary for the CAROM
highwall. The explosive column is also code. Both the momentum imparted to the
indicated in the figure. rock fragments and the blast pressures
After detonation of the explosives, due to the explosive gases are chosen
shock waves are generated in the sur- based on DYNA2D calculations. DYNA2D
rounding rock, which not only fracture cannot be used to model the entire blast
the rock, but also impart momentum caus- because the code is a continuous repre-
ing unconfined rock to begin motion. In sentation, meaning it cannot describe
addition, large gaseous pressures are material that breaks apart.
created within the boreholes that exert Beside the initial conditions, further
forces on the elements causing them to assumptions were necessary prior to run-
accelerate. ning the CAROM code. For the quarry
tests, the size of each element was fixed
and not allowed to change. This meant
that elements were not permitted to break
l0 ~~~,-·,---,-·,--~--.---r-.---r-.....----,r--- ,--,--,----, up into smaller fragments. Also, to sim-
NOTE.--C1rcles represent ~ock g]emgnts~ plify calculations, the elements were not
allowed to roll and therefore, could not
e possess any angular momentum. The co-
u efficient of restitution for collisions
< between elements was chosen to be a con-

0 5 stant for all of the elements. A value

< of 0.01 was assumed based on research at
SNL. This implies that all of the col-
lisions are treated as nearly perfectly
inelastic. One final assumption deals
with the collision criterion. Based on
0 "'-"~~--'----1 ___l__.,____..___..___,_ . _L_...L--.-J'---....L._-L__j
work at SNL, a collision is defined to
0 5 10 15
occur when the distance between the cen-
HORJZONT~L OJST~NCE. m ters of two elements equals 1.9 ft.
FIGURE 1.-Highwall representation by discrete rock elements. The quarry tests were conducted as sin-
Heavy line on left indicates explosive column. gle-row bench blasts to study the effect

of delay timing on fragmentation. Four of delay timing between holes in a single

blastholes were fired in each of six row. However, once a three-dimensional
full-scale shots. Because CAROM is at code becomes available, these studies can
the moment only a two-dimensional code, be done.
it cannot be used to predict the effects


All of the test shots were monitored The output from the CAROM code is shown
with two high-speed cameras operating at in figures 2-11 for the first 2 s of rock
500 frames per second to record the actu- motion at 200-ms intervals. The average
al rock motion. Analysis of the result- velocity for any element can easily be
ing films yielded data providing a basis calculated from the figures. The element
for confirmation of the CAROM code. Of representing the leading edge of the
primary interest was the initial outward blasted rock face, as shown in figure 2,
velocity of the bench face material. has an average velocity determined to be
This was observed in all tests to be 16.5 m/s, which is in excellent agreement
about 15 m/s. with the cinematographic data.

10 10

NOTE. --Circles represent ,~ack alemo:nts. NOTE.--Circlo:s represent rock elements.

z z
< <
;- ;-
(J) U)

0 5 0

o0 o
< <
u u

;: ;:
> "'>w 8
0 0
0 0
0 5 10 15 0 5 10 15


FIGURE 2.-Predicted rock motion at time equals 0.2 s. FIGURE 3.-Predicted rock motion at time equals 0.4 s.

10 10
NDTE.--Circles ~epres~nt rock G]ements. NOTE.--Circles represent rock elemGnts.

z z
< <
1- 1-
U) Ul

0 5 0 5
..J _j
< <
u u
1- 1-
"'>w "'>
w oo
0 0 CXX::O, d1cS!fl8°oCb.r 90 ·---'------'--
0 5 10 15 0 5 10 15


FIGURE 4.-Predicted rock motion at time equals 0.6 s. FIGURE 5.-Predicted rock motion at time equals 0.8 s.

NOTE.--Cjrcles represent rock glements.

0 5


0 5 lO IS 20

FIGURE 6.-Predicted rock motion at time equals 1.0 s.


NOTE.--Circles represent rock el


0 5 ~·

> 0

0 5 10 15 20


FIGURE ?.-Predicted rock motion at time equals 1.2 s.


NOTE.--C!rcl~s represent rock elements.


0 5


0 5 10 15 20 25


FIGURE B.-Predicted rock motion at time equals 1.4 s.


NOTE.-~Circles r@present rock

;:'.; 5

0 5 10 15 20 25


FIGURE 9.-Predicted rock motion at time equals 1.6 s.

10 ,...---,---~ -,-··
NDTE.--Clrcles represent rock elements~

~ 5
0 0
__O_~~oco .. ~l ]___L_L..___._Q
0 5 10 _15 20 25
II FIGURE 1D.-Predicted rock motion at time equals 1.8 s.

j 10
NOTE. --Circles repr·gsont rock

~ 5

o cx-ftffb81oco 0 ._,_Q1D_
0 5 10 15 20 25

FIGURE 11.-Predicted rock motion at time equals 2.0 s.

In the fragmentation test shot utiliz- little or no vertical rock motion in the
ing 4.0 ms/ft of burden as the delay tim- field tests.
ing, the maximum rock throw was measured One of the deficiencies of the code ap-
to be 30 m. The CAROM code predicted, as pears to be its inability to predict the
shown in figure 9, at least one element profile of the rock motion of the bench
thrown to about 24 m after 1.6 s of time face. As shown in figure 2, CAROM pre-
passage in the calculations. CAROM is dicts almost the entire face to break off
also capable of predicting vertical mq- as a slab. It is likely that this is due
tion of the bench top. However, in the to improper balancing of the input param-
series of results presented here no up- eters. That is, there is too much empha-
ward motion is observed. It is possible sis on the imparted momentum due to the
that CAROM still did indicate such mo- shock wave impulse, and not enough empha-
tion, but that it occurred prior to sis on the accelerations due to the
200 ms (f 2). At this time CAROM forces created by the pressures of the
shows the surface material falling down- explosives gases. Research continues to
ward under the influence of gravity. The develop a step function approximation for
high-speed films confirmed that there was the high-pressure gases.


The computer code CAROM, which models maximum throw of the rock fragments.
rock throw in blasting, has been briefly CAROM is still being developed, and many
discussed. The code was used to simulate improvements are being implemented. The
rock motion at B limestone quarry, and objective continues to be to provide
the results of the analysis have been mine operators with accurate predic-
presented. The code accurately predicted tions of rock motion and final muckpile
the velocity of burden movement and the distribution.


1. Schamaun, J. T. An Engineering Applications-in Rock Masses • . Proc. 26th

Model for Predicting Rubble Motion During U.S. Symp. on Rock Mechanics, Rapid City,
Blasting. r in Proceedings of the SD, June 1985, Soc. Min. Eng., 1985,
9th Conference on Explosives and Blasting pp. 151-159.
Technique, ed. by C. J. Konya. (Dallas, 3. Ha 11 qui s t , J. 0 • User's Manual for
TX, Jan. 31-Feb. 4, 1983). Soc. Explos. DYNA2D - An Explicit Two-Dimensional Hy-
Eng., Montville, OH, 1983, pp. 199-222. drodynamic Finite Element Code With
2. Gorham-Bergeron, E. Rock Motion Interactive Rezoning. Lawrence Livermore
Modeling oE Oil Shale Cratering Experi- Nat. Lab., Livermore, CA, UCRL-52997,
ments. r in Research and Engineering 1982, 109 pp.


By Mark S. Stagg1 and Michael J. Nutting2


The Bureau of Mines is studying blast ms/ft of burden. In this range, stress-
delay timing influences on rock frag- wave-induced strains interacted with
mentation in a series of tests that longer lasting gas-pressure strains from
started in 3-ft concrete blocks and in- earlier holes. Coarse fragmentation re-
cludes reduced-scale and full-scale bench sulted from short delays (<1 ms/ft),
blasts. This paper reports on the where breakage approached presplit con-
reduced-scale tests. In a 45-in-high ditions with a major fracture between
dolomite bench, 18 single-row blasts were blastholes and large blocks in the burden
fired with 15-in burdens. Spacings were region. Coarse fragmentation also re-
21 and 30 in. Delay intervals ranged sulted from long delays (>17 ms/ft), with
from 0 to 45 ms, equivalent to 0 to 36 explosive charges acting independently.
ms/ft of burden. Each shot was instru- The broad acceptable range provides blast
mented for strain and pressure for both design tools for a variety of purposes,
in situ dynamics and interactions between including optimum muckpile displacement
blastholes. All fragmented rock was and vibration control.
The finest fragmentation occurred at
blasthole delay intervals of 1 to 17
The explosives industry is developing interaction between shotholes. Initial
and testing delay blasting caps of im- testing in the laboratory provided an
proved accuracy. Precise delays have effective means for establishing a meth-
been cited as factors in controlling odology of controlled experimentation.
blast vibration amplitude and frequency The tests at reduced scale in the field
and improving fragmentation (l-I).3 How- provided experience in fragmentation
I ever, data on complete fragmentation as- assessment techniques and results that
sessment of shots initiated with precise
days is limited. As part of .its blasting
could be used to optimize the expensive
full-scale field tests. The full-scale
research program, the Bureau of Mines is field tests are currently in progress.
I examining the influence of timing inter-
vals by completely screening the blast-
This paper discusses the reduced-scale
field tests and results.
induced rock. Three- and four-hole shots The reduced-scale field tests were con-
I have been conducted in concrete blocks in ducted at the University of Missouri's
the laboratory and at reduced (45-in Experimental Mine in Rolla. This site
bench) and full (22-ft bench) scale in was chosen because of its accessibility
the field. Tests thus far have mostly and geology, and the cooperation avail-
been concerned with the effect of de- able from the University. Furthermore,
lay time on fragmentation and on the the results of previous research con-
ducted at the mine on blast design and
Civil engineer. fragmentation were reported in several
2Geophysicist (now with Philip R. Ber- theses Cl-2). These studies investigated
ger & Associates, Inc., Warrendale, PA). various design factors affecting fragmen-
Twin Cities Research Center, Bureau of tation, such as coupling, initiation se-
Mines, Minneapolis, MN. quence, primer location, and airgap, anc
3underlined numbers in parentheses re- provided a comparison with the BureaL
fer to items in the list of references at test results.
the end of this paper.


The experiment was conducted in a 45-in typical full-scale bench blasts with di-
bench of dolomite, part of the Jefferson mensions about 10 pet of full scale.
City Formation. The rock is of irregular However, at reduced scale, rock struc-
grain size, 10 pet calcite, and thick tures such as bedding and jointing are
hedded, with a specific gravity of 2.65 exaggerated and can have an unrealistic
and compressional and shear velocities of effect. The massive dolomite at the Rol-
14,800 and 8,100 ft/s, respectively (2). la site provided a good medium for the
Bureau researchers verified these values reduced-scale testing since only three
by in situ seismic measurements behind bedding planes were in the 45-in bench
the blastholes, finding 14,700 ft/s (com- and no jointing was observed within any
pressional) and 8,100 ft/s (shear). of the test shots.
The reduced-scale tests were designed
to be geometrically proportional to


In order to prepare the pit and bench with air to remove loose material, and
for the tests, development work was nec- the entire area was swept clean prior to
essary to provide consistent geometry of each shot.
the working faces. Horizontal holes were Each hole contained 144 g of 60 pet
drilled 45 in above the pit floor to give extra-high-density dynamite tamped into a
the proper bench height, and vertical 1/2-in-ID plastic tube 40 in long. Each
holes were drilled to obtain a vertical charge was bottom-primed and initiated by
face. For each test shot, an open end an exploding bridgewire detonator (EBW).
at an angle of approximately 135° was Complete coupling was assured by placing
formed, as shown in figure 1. To form the charge into grout-filled holes. The
the single-row pattern, three shotholes initiation system for the EBW 1 s consistei
were drilled to 50-in depth, including of a power supply, firing module, anc
5 in of subdrill. The burden was held digital delay generator with a firing
constant at 15 in, and spacings were 21 accuracy of 0.0025 pet times the delay
and 30 in, spacing-to-burden ratios (S/B) time or ±50 ns, whichever is greater.
of 1.4 and 2.0, respectively. All faces Most shots were instrumented with
were cleaned with a scaling bar and blown dynamic-strain and pressure gauges (figs.
2-3). The strain gauges were a six-com-
ponent type, built after a design by Reed
(~) as modified by Anderson (l). These
gauges were grouted into the burden re-
gion at various locations between the
boreholes. Pressure gauges were placed
either above the strain gauges or in in-
clined holes in the face, which were
filled with a water-revert mixture. The
pressure gauges were of two types, carbon
resistors dipped in liquid tape (insulat-
ing coating) and Navy-built tourmaline
gauges in an oil-filled boot. Also, a
fiber optic system was used to measure
60 pet exlro- detonation velocities. Data were re-
dynamite corded on a 28-channel Wide Band I (80

kHz) recorder and a digital oscilloscope,
with a O.S~s response rate.
FIGURE 1.-Shot pattern and blast design for reduced-scale To contain flyrock and minimize second-
tests. ary breakage, the entire shot was covered

FIGURE 3.-Carbon resistor (left and right) and tourmaline

(center) pressure gauges.

FIGURE 2.-Stages of assembly of six-component strain


with a blasting mat held in place with

timbers and anchoring cables (fig. 4).
The area in front of the shots was cov-
ered with a plastic sheet to aid in .iden-
tifying blasted material. Flyrock went
beyond the sheeting for only a few shots.
This rock was identified when possible
and included with the muckpile.
Screening of the muckpile began immedi-
ately after each shot. All fragments
3 in and larger were sized and weighed in
the pit. The pit size fractions were
minus 3, plus 3, minus 6, plus 6, minus
12, and plus 12 in. Material passing the
3-in screen was loaded into containers,
removed from the pit, and mechanically
shaken through screens. These size frac-
tions were plus 1-1/2 minus 3, plus 3/4
minus 1-1/2, plus 3/8 minus 3/4, plus
3/16 minus 3/8, and minus 3/16 in. From
the weight of each fraction, its percent- FIGURE 4.-Pit area with blasting mat and timbers covering
age of the total muckpile was calculated. test shot.


A total of 24 blasts were detonated. single-hole test shot. Delay intervals

Two were development shots. Of the re- ranged from simultaneous to 36.0 ms/ft of
mainder, nine at 21-in spacing and burden. Table 1 lists the reduced-scale
nine at 30-in spacing were completely (RS) shots.
screened, three misfired, and one was a

Since these tests were designed to de- came from outside the shot pattern, it
termine the effect of delay time on frag- was removed from the corresponding size
mentation, i.e., the interaction between fraction and not included in the analy-
shotholes, it was decided to identify any sis of fragmentation. Table 2 lists the
overbreak from each shot and to size and sieve weights for the shots.
weigh it separately from the muckpile. A The cumulative percent passing versus
third of the shots produced no back- or sieve size is plotted in figure 5 for
end-overbreak. There was no obvious cor- four shots at each spacing, covering the
relation between spacing or timing and range of delay times tested. The mate-
those shots producing overbreak, which rial that passed through the 24-in sieve
was generally oversized and averaged 10 would in most cases have passed through a
pet of the total blasted rock weight. much smaller sieve, even down to 13 in.
Because the overbreak skewed the particle Since the largest size piece was not mea-
size distributions at the high end and sured and 24 in is too large, the 24-in

TABLE 1. - Reduced-scale (RS) test shots

Spacing,Delay time Spacing, Delay time

Shot 1 inBetween ms/ft of Shot 1 in Between ms/ft of
holes, ms burden holes, ms burden
RS- 3 • ••..• 30 20.0 16.0 RS-14 ••••• 30 1. 25 1.0
RS-4 •••••• 30 7.5 6.0 RS-15 ••••• 30 0 0
RS- 5 •••••• 30 5.0 4.0 RS-16 ••••• 30 45.0 36.0
RS-6 •••••• 30 20.0 16.0 RS-1 7 ••••• 30 2.5 2.0
RS-9 •••••• 21 0 0 RS- f 8 ••••• 21 21.0 16.8
RS-10 ••••• 21 30.0 24.0 RS-19 ••••• 21 1. 7 5 1.4
RS-11 ••••• 21 14.0 11.2 RS- 21 ••••• 21 .4375 .35
RD-12 ••••• 21 3.5 - 2.8 RS-22 ••••• 21 8.75 7.0
RS-13 ••••• 30 30.0 24.0 RS-24 ••••• 21 45.0 36.0
1 Not listed: Shots RS-1 and RS-2 (development shots), RS-7, RS-8, and RS-20 (mis-
fires), and RS-23 (single-hole shot).

TABLE 2. - Weight of rock fragments at various screen sizes, pounds

+3/16 +3/8 +3/4 +1-1/2 t3 +6 +12

Screen size, in ••• -3/16 -3/8 -3/4 -1-1/2 -3 -6 -12 -24
RS- 3 . ..•..•.•••••. 208 243 268 535 639 1,389 1,607 241
RS-4 •••••••••••••• 172 243 263 570 675 1,987 2,022 241
RS-5 •••••••••••••• 215 237 258 498 523 1,628 1,941 159
RS-6 . ••.•••.•••.•• 188 216 257 466 596 1,656 1,745 117
RS-9 . ..••..••••• • • 196 182 196 356 421 1,197 I, 17 3 1, 415
RS-10 ....•........ 232 259 320 552 631 1, 3 72 2,100 172
RS-11 • .•.•..••...• 305 318 391 587 572 1,149 800 0
RS-12 . ••..••..••.• 233 255 298 557 639 1,521 1,147 0
RS-13 •••••.••••••• 238 238 276 452 516 1,550 1, 857 520
RS-14 • . . . . . . . . . . . . 217 240 253 541 612 1, 911 1, 718 266
RS-15 •••••••••.••• 167 182 208 466 521 l, 657 2, 092 462
RS-16 •• ~ •••••••••• 244 250 246 471 512 1,607 2,019 536
RS-1 7 • .•••...•.... 242 247 309 522 630 1,801 1,546 446
RS-18 • ••..•.•...•. 246 275 342 522 590 1, 197 876 0
RS-19 •.•••••••..•• 202 234 302 548 635 1, 219 911 0
RS- 21 • ..........•. 203 194 198 377 449 1,132 1,507 688
RS- 2 2 • •••.•.•..•.. 268 235 258 454 501 1,131 1, 022 143
RS-23 •.••••••..••• 71 60 61 123 138 364 550 178
RS-24 •••••••.••... 295 249 270 486 541 1, 060 1, 110 422

t 100 -~··~---·~··-···r--~-~-T~--,--,--r--r-r-.rr?-----r-cf>-1
e 21-in spacin'J I

21-in spacing
I '' - - a - i O p c t passing

e ao
KEY I '' o 50 pet passing
:. 80 pet possinq
o RS-9, 0 ms/ft I '' 30-in spacing
o RS-19, 1.4 ms/ft
'' ~----~~-26 pc t poni ng
A RS-IS. 16.8 ms/fr ' • 50 pel passing
v RS-24, 36.0 ms/ft 10 • SO pet passing
GO ---Extrapolation of dato. set text -~-f>lropolot;on ol dolo,:.~·~_.


0 10 40
100 r-'f-~ ~,-,---------~-~...___,.-~~~
DELAY~ mslft of burden
30-in spocino
FIGURE 6.-Size at 20, 50, and 80 pet passing versus delay
KEY period for shots at 21- and 30-in spacings.
BO <> RS-15, 0 ms/lt
a RS-14, LO ms/ft
A RS-6 1 16.0 ms/ft

I 60
v RS-24, 36.0 ms/fl
that delays of 1 to 17 ms/ft produced
the finest fragmentation. Poorest or
I 40
coarsest fragmentation was observed for
simultaneous shots and for shots with

I 20
delay intervals >24 ms/ft, although the
differences at the longer delays were
smaller with the 30-in spacings. The im-
proved fragmentation obtained for- most of
0 ~·---'--L~-'-~·~·~·--""--·--'...__.L-.L..L.J.....LL.L_.-·~-'- the shots at 21-in spacings, as compared
0.1 1.0 10.0 30.0
SIEVE Sl ZE, in with 30-in spacings, was expected and was
FIGURE 5.-Particle size distribution for shots at 21- and 30-in
due in part to a higher powder factor-.
spacings. However, the coarsest fragmentation re-
sulted at the very short delay times for
the 21-in spacing tests, even though
point has been omitted from the curves, there was a higher powder factor.
except that it was used to obtain the 80- To quantify the test results and ulti-
pct-passing value for shot RS-9, as shown mately develop an equation to predict
in figure 5. The curves in figure 5 are fragmentation, it is necessary to develop
representative of the results, which a mathematical description of the cumu-
showed that a dramatic improvement, a 20- lative percent-passing data. Previous
to 50-pet reduction in average size, oc- researchers have used regression analy-
curred at delays of ~1.0 to 17.0 ms/ft sis to fit observed blast fragmentation
compared with simultaneous initiation. data to logarithmic, power (8), Gaussian
At delays longer than 17.0 ms/ft, the (2), and Weibull distributi~ns (9-10).
average size increased ~20 to 50 pet. Analysis-of-variance tests can then-be
The 20-, 50-, and 80-pct passing sizes used to determine the statistical signif-
versus delay period, shown in figure 6, icance of the effect of delay time on the
were determined for all the tests from cumulative size distribution functions.
percent-passing curves similar to those These tests determine whether a single
shown in figure 5. The delay period had regression line can be used to represent
little effect on the size of fragments the combined results of several shots.
in the 20-pct-passing fraction, except If pooling or combining the results tests
that they were slightly coarser at the positive, then there exists a statistical
simultaneous shots. At both spacings, inference that the delay time has not
the 50- and 80-pct-passing fractions show significantly affected fragmentation.

Statistical tests were run on several log-normal distribution. Analysis-of-

regression line fits to the data, and the variance tests were run for all finer
correlation coefficients (R) for the var- size data from delays of 1 to 36 ms/ft,
ious distributions are given in table 3. and one curve could be used to represent
The data shown in figure 6 suggest that the data at both spacings. For shots at
the finer size material was less affected delays of less than 1 ms/ft, the weight
by delay time. An examination of the of fines was reduced.
weights of the material passing the 3-in The correlation coefficients in table 3
sieve found minimal variability, as shown show that a Gaussian distribution for
in figure 7, except for shots at delays the coarser size data usually produced
less than 1.0 ms/ft. Excluding these the best fit. The Gaussian, power, and
shots and an extremely high (2,170 lb) Weibull fits to the coarse size data are
outlier value, the weight of material shown in figure 9 for shots RS-3 and
passing the 3-in sieve ranged from 1,716 RS-13. It was observed that the 1-1/2-in
to 1,994 lb and 1,720 to 1,950 for the sieve material also fit the Gaussian dis-
21- and 30-in spacing shots, respective- tribution, and this material was included
ly. The nearly constant weight of finer in the regression analysis of table 3.
material suggests that a fractured zone The coarse material is assumed to be gen-
extends around the borehole, which, as- erated primarily between the boreholes
suming a cylindrical nature, would have outside the extended fractured zone that
a radius of 10 in. This equates to 40 exists around the borehole.
explosive radii, within the range of An attempt to pool all of the data us-
the damage zone predicted by Siskind (ll- ing a Gaussian distribution indicated (at
12), Olson (13), and others in terms of a 95-pct-confidence level) that the delay
charge radii (i.e. , 20 to 40 radii). As time was indeed significant. Further
shown in figure 8, the fines best fit a analysis-of-variance tests were conducted

TABLE 3. - Correlation coefficients for Weibull,

power, and Gaussian distributions

Shot Wei bull Power Gaussian

AI B2 AI B2 B2
RS- 3 • ••..•.•.•••. o. 9923 0.9861 0.9963 0.9980 0.9999
RS-4 ••••••••••••• .9906 .9858 .9961 .9944 .9995
RS- 5 . .••...•..... .9833 .9758 • 9972 • 9941 .9993
RS-6. •••••••••••• .9856 .9815 .9874 .9953 • 9998
RS-9 .•••.•.••• •· •• .9961 .9936 • 9978 • 9942 .9837
RS-10 •••••••••••• .9859 .9720 .9966 .9980 .9979
RS-11 • •••..••.... .9966 .9852 .9876 .9956 .9996
RS-12 • .•.•.••.••. .9958 .9845 .9948 .9941 • 9992
RS-13 •••••.•••••. .9893 .9847 .9973 .9946 .9988
RS-14 •••.••••.••• .9893 .9868 .9972 • 9923 .9981
RS-15 • ••.•••••••• .9895 .9848 .9982 .9948 .9993
RS-16 ....••.••..• .9872 .9830 .9971 .9940 • 9992
RS-17 •••.•.•••..• .9832 .9906 .9970 • 9928 .9945
RS-18 ••.....•.•.. .9966 .9875 .9902 .9933 .9998
RS-19 •.••.••••••. .9978 .9909 • 9913 .9932 1. 0000
RS- 21 • ••••••••••• • 9921 .9887 .9981 .9970 .9980
RS-22 ..••..••..•• .9904 .9862 .9966 .9957 .9994
RS-24 ••...•....•. .9956 • 9921 .9968 • 9980 .9964
I - po~nts were used
All data ~n regression
' analysis except
the +12-, 24-in data point.
2 0nly data at 1-1/2-, 3-, 6-, and 12-in sizes were used;
100-pct polnt was also excluded.


95 o RS-3, 16.0 ms/ft

80 o RS-11, 11.2 ms/fl
90 o RS-13, 24.0 ms/ft
80 Gaussian fit
50 70
~ 40 60
lJJ 40
3: 20 30
~ 10 ~-----L-------L-------L------~----~
~ to~-----,~v-~3------6~----------~,2~----~
~ 90 ~-----.-------.-------.-------.------~ w
z 3100,-----.------.------.------r.-----~
t; 80 I -36 ms/fl >- 90
a: 21- and 30-in spacing I]) 80 Power fit
~ 70 (!) 70
z 60
60 ~ 50
50 <(
0.. 40
40 1-
z 30
30 w

20 a: 20
0.. ilt2 3 6 12

SIZE, in 90 Weibull fi1

FIGURE B.-Log-normal distribution for material passing 1112-in 80

sieve size. 70
to see if certain shots could be pooled
to form one regression line. For exam-
ple, the 21-in shots at 11.2 and 16.8 20
ms/ft could be combined (i.e., there was 15
11/2 3 6 12
no effect due to delay time). The 30-in
SIZE, in
shots at 1,2, 4, and 6 ms/ft could also
be combined. Figure 10 is a plot of the FIGURE 9.-Comparison of Gaussian, power, and Wiebull
50- and 80-pct-passing value determined
from the Gaussian distributions. A hori-
zontal line is drawn for delays that which improved fragmentation over that
could be pooled into one regression line. resulting from single-hole shot.
Although not shown in the figure, the The 50- and 80-pct curves of figure 6
single-hole fragmentation distribution and 10 are quite similar except for the
pooled with the 30-in spacing curves simultaneous shot at 21-in spacing, RS-9.
at delays of 24 and 36 ms/ft. Apparent- The Gaussian distribution was not the
ly, firing holes at delays of >24 ms/ft best fit for this shot, and the regres-
can be considered as firing single-hole sion line predicts a higher value than
shots. It is noteworthy that the dis- the data suggest.
tance from the corner to the first hole As mentioned earlier, the University of
was 21 in (fig. 1), but the fragmentation Missouri's Experimental Mine has been
data pooled with the 30-in spacing tests. used by several researchers (1-~) to con-
The single-hole test reflected the 30-in duct investigations of blast design and
results because at 21 in, the first fragmentation. Where possible, these
hole's breakout angle, >135°, reduced the data have been compared with Bureau re-
burden distance for subsequent holes, sults, as shown in figure 11. Since the

15 r-~~.-­ improvement in fragmentation as delays

21·in spacing increase to 1 ms/ft, slightly coarser
' '
o 50 pet passing
c SO pel passing
fragmentation between 6 and 7 ms/ft, and
' continued improvement to 10 ms/ft.
' 30-in spacing
--...-:so pc1 passing Strain and pressure records obtained
• 80 pel passing
for the reduced-scale tests tend to con-
c firm the fracture development mechanisms
observed and reported by Holloway in
work done under contract to the Bureau
(80245046, 1986). Initially, a fracture
zone develops around the borehole because
of the development of radial fractures
and fracturing caused by reflected stress
waves. The radial fractures propagate at
0 I 10 40
speeds down to 12 pet of the P-wave ve-
DELAY, ms/lt of burden locity (~). The fractured region for
FIGURE 10.-Size at 50 and 80 pet passing versus delay period these tests appeared to coincide with the
from Gaussian distribution fits to the 21- and 30-in spacing data. finer material zone, which was generated
within about 40 charge radii. From pres-
sure gauges installed in this zone, the
velocity of explosive gases penetrating
fractures was found to be approximately
1,800 to 2,700 ft/s, as shown in figure
n 8uMines, S;8 :: 2.0 12. The large impulsive signals on the
o BuMines,SJ6 z 1.4 records were due to the pickup of elec-
A 8ergmorm,S;8 ::c I .5
<3 v Winle(, S;8 :: I .5 trical noise from the EBW initiation sys-
• Unlv.MOtes!s,SJa"' 1.4 tem and were used to confirm the delay
V> 10 Univ, MO tests, s; 6 "'2.0
S/8 spocingwto·burden rotio
The P-wave velocities were determined
:i from the arrival times and the distances
from shotholes to gauges. The distance
Vi and arrival time measurements used to
calculate the velocity of gas movement
through the rock were adjusted to correct
for the travel time of the bottom initi-
10 50
ated explosive detonation (8,000 ft/s) to
DELAY, ms/ft of burden the height of the gauge.
FIGURE 11.-Comparison of reduced-scale data with results
The borehole pressure and radial crack
obtained by other researchers. pressurization produce stresses in the
material beyond the near fracture zone,
and this leads to additional failure and
shot designs for these tests were similar extension of the radial cracks to the
to those used by the Bureau, the fragmen- free face. The velocity of gas penetrat-
tation results compare very favorably. ing fractures was estimated to be 146
Other research, such as Bergmann's multi- ft/s for shot RS-14, as shown in figure
hole tests in granite blocks (~), showed 13. This velocity was determined by sub-
a similar significant improvement in tracting from the arrival time the travel
fragmentation as delay times increased time of the explosive to the gauge height
from simultaneous to 1 ms/ft of burden. and the time of gas penetration (1,800
Winzer's tests in limestone blocks and in ft/s) out to 10 in. The remaining dis-
a small bench (15) resulted in a rela- tance to the gauge and the remaining time
tionship between~elay time and fragmen- were used to determine the velocity.
tation that correlates well with Bureau Strain data from the shots were pro-
results. The character of the data in cessed into resultant principal strains,
figure 11 is similar, showing substantial octahedral shear strain and dilatation

300 hole 3
.... Firing
.c pulse, Firing
hole I pulse, Gas pressure, 1,760ft/s
a:: 200 hole 2

0 2 3 4 5 6 7 8 9 10 II 12 13 14
TIME, ms


Gouge hole drilled 2oo-ro-,- 1

r I
down from horizontal 1 115"1

FIGURE 12.-Pressure measured for shot RS·19.

(7). Two of the principal strains are a major crack developed in line with
plotted in figures 14 through 17 for test holes 2 and 3, causing the observed gas
shots with simultaneous and 1.4-, 2.0-, velocity of ~2,000 ft/s. Excessive end-
and 16.0-ms/ft delays. Shown on the rec- break was noted for this development
ords are the calculated P-wave and gas shot.
velocities determined from arrival times Optimum fragmentation occurred when a
and shothole-to-gauge distances as dis- hole fired such that its stress wave in-
cussed earlier. The strain pulses in- teracted with the stress induced by the
creased in amplitude with decreasing dis- expanding gas pressurization from the
tances from shothole to gauge. These previous hole. Shots RS-2 and RS-6 show
pulses correspond to the arrival of a long-term strain believed to be induced
stress waves, which often arrived in- by the late-arriving gas pressure inter-
creasingly later in time as holes 2 and 3 acting with the stress wave from hole 3.
were fired. The observed decrease in Similar measurements and observations
compressional (P) and shear (S) veloc- have been made at full scale (~). The
ities is probably due to flaws in the interactions of strains induced by the
rock, such as fractures and cracks pro- stress waves and strains induced by gas
duced by stress waves from an earlier pressure were not always observed for all
shothole. shots of optimum delay (1 to 17 ms/ft),
The gas velocities observed on the because the gauge is stressed only if
strain records agreed with pressure gauge near a pressurized crack and well-coupled
gas velocity observations, except for to the rock. These interactions were
those in shot RS-2, where it is believed also not observed for shots with delay


... 60

401 Gas pressure,
P-wave, 6,880 ft/s

TIME, ms

pulse, I
hole I I

Gauge hole drilled i ~T,.
20° down from ~ 20
horizon ta I _..;..1 ---~--...__ __,_.~

Firing Firing
pulse, pulse,
hole 2 hole 3
FIGURE 13.-Pressure measured for shot RS-14.

times outside the optimum range. Even velocity for crack development in the lO-
though shot RS-19, shown in figure 15, in zone around the borehole, plus th£
does not show an interaction at the gauge explosive detonation time (0.42 ms), th£
location, pressure effects are still ob- next hole should not be fired until after
served later in the record. Gas effects 0.9 ms or 0. 7 ms/ft. Enhanced crackint
are not as apparent for the.simultaneous appears to last as long as the gas is re-
shot, shown in figure 14. tained. Gas velocities through the rocl
Fragmentation results at short delays suggest the process will last up to 20 m:
(less than 1 ms/ft of burden) suggest or 16 ms/ft, based on a 45° breakout an·
that fracturing in the zone around the gle and gas penetration velocities o
borehole must be completed before the 1,800 ft/s for the first 10 in and 5
next hole fires. Using 1,800 ft/s as the ft/s for the next 11 in.

2,000 !Fidng P"'"· 0 11 holes

---- P-wove, 15 •600 ft/s


z- 0 ____..
6.0 8.0

FIGURE 14.-Principal strains meas ured for shot RS-9. Simultaneous shot.

JJJ--- ---®-~-----([)------ : : o l : , _ ; - - -

Firing pulse, hole I "

P-wave, 13,100 ft/s
1,200 Firing pulse, hole 2
P-wave, ll,800 ft/s
P-wave, 7,060 ft/s
Gas pressure, 60 ft/s

0 .0 s.o/f
rf .
-400 TIME, ms

Firing pulse, hole 3


FIGURE 15.-Principal strains measured for shot RS-19. Delay time, 1.4 ms/ft.

..... -
<D (()
hole I .....
Firing ~ ...

Firing ~ 0I() pulse, ~ ;)


pulse, :: 10 .. hole 3 d.'
0 hole 2 o (\J
~ 0
ell ...
1,000 !() ;)
,; ,; Cll

3 3
a. a.
. 0

-1,000 i

Shot material
~-'"'"""'""'-~"'u-:;,' ,,.
Strain gouge,
25 in deep

FIGURE 16.-Princlpal strains measured for shot RS-2. Delay time, 2.0 ms/ft.

P-wave,IO,OOO ft Is

P-wove, 7,170 ft/s


Fire pulse, hole 3

-1.400 Fire pulse, hole 2

Fire pulse, hole I

-2,800 l

in deep

FIGURE 17.-Principal strains measured for shot RS-6. Delay time, 16.0 mslft.


An investigation of the effect of delay determined to come from the immediate

time on fragmentation was conducted at blasthole vicinity and showed little var-
a reduced scale using three blastholes iability from shot to shot.
per shot in a 45-in bench. With the bur- Analysis-of-variance tests showed that
den constant at 15 in, delay intervals delay time did influence the distribution
were varied from 0.0 (simultaneous) to of fragment sizes for both spacings, but
36.0 ms/ft of burden, and the entire more so for the 21-in spacing. Fragmen-
muckpile was screened to assess frag- tation was coarsest for shots fired
mentation for tests with 21- and 30-in simultaneously and at delay times of
blasthole spacings. 24 ms/ft of burden and greater. Better
Attempts were made to mathematically fragmentation was observed for delay
describe the distribution of fragment times from 1 to 17 ms/ft, with the tests
sizes using power and Weibull functions, at 11.2 and 16.8 ms/ft resulting in the
which have been used by other research- best fragmentation. Dynamic strain and
ers. However, regression analysis indi- pressure measurements indicated that this
cated the Bureau data were best described improved fragmentation may be the result
by a Gaussian or simple normal distribu- of strains induced by stress waves con-
tion. Materials excluded from the analy- structively interacting with strains in-
sis were overbreak, which came from out- duced by gas pressure from an earlier
side the shot area, and fines, which were detonated hole.


1. Chiappetta, R. F., S. L. Burchell, 5. Smith, N. S. Burden Rock Stiffness

D. A. Anderson, and J. W. Reil. Effect and Its Effect on Fragmentation in Bench
of Precise Delay Times on Blasting Pro- Blasting. Ph.D. Thesis, Univ. MO-Rolla,
ductivity, Ground Vibrations, Airblast, 1976, 148 PP• _
Energy Consumption and Oversize. Paper 6. Reed, R. P. Triaxial Measurement
in Proceedings of the 12th Annual Con- of Stress Haves in the Free-Field. San-
ference on Explosives and Blasting Tech- dia Lab., Albuquerque, NM, Rep. SAND-78-
nique, ed. by C. J. Konya (Atlanta, GA, 212236, 1979, 21 pp.
Feb. 1986). Soc. Explos. Eng., Mont- 7. Anderson, D. A., S. R. Winzer, and
ville, OH, 1986, pp. 213-240. A. P. Ritter. Time-Histories of Princi-,
2. Reil, J. lv., D. A. Anderson, A. P. pal Strains Generated in Rock by Cylin-
Ritter, D. A. Clark, S. R. Winzer, and drical Explosive Charges. Paper in Rock
A. J. Petro. Geologic Factors Affect- Mechanics in Productivity and Pro~ection,
ing Vibration From Surface Mine Blast- ed. by C. H. Dowding and M. M. Singh
ing (contract H0222009, Vi bra-Tech Eng., (Proc. 25th U.S. Symp. on Rock Mechanics,
Inc). · BuMines OFR 33-86, 1985, 204 PP•; Evanston, IL, June 25-27, 1984). Soc.
NTIS PB 86-175858. Min. Eng. AIME, 1984, PP• 959-968.
3. Bleakney, E. L. A Study of Frag- 8. Da Gama, D. Use of Comminution
mentation and Ground Vibration With Air Theory To Predict Fragmentation of
Space in the Blasthole. M.S. Thesis, Jointed Rock Masses Subjected to Blast-
Univ. MO-Rolla, 1984, 92 PP• Paper in First International Sympo-
4. Brinkman, J. R. The Influence of sium on Rock Fragmentation by Blasting.
Explosive Primer Location on Fragmenta- Lulea Univ. Techno!., Lulea, Sweden,
tion and Ground Vibrations for Bench 1983, pp. 565-579.
Blasts in Dolomitic Rock. M.S. Thesis, 9. Cunningham, C. The Kuz-Ram Model
Univ. MO-Rolla, 1982, !33 pp. for Prediction of Fragmentation From

Blasting. Paper in First International Rock Damage From Small Charge Blasting in
Symposium on Rock Fragmentation by Blast- Granite. BuMines RI 7751, 1973, 44 pp.
ing. Lulea Univ. Technol., Lulea, Swe- 14. Bergmann, 0. R., F. C. Wu, and
den, 1983, PP• 439-453. J. W. Edl. Model Rock Blasting Measures
10. Hjelmberg, H. Some Ideas on How Effect of Delays and Hole Patterns on
To Improve Calculations of the Fragment Rock Fragmentation. Eng. and Min. J., v.
Size Distribution in Bench Blasting. 175, June 1974, PP• 124-127.
Paper in First International Symposium on 15. Winzer, S. R., D. A. Anderson, and
Rock Fragmentation by Blasting. Lulea A. P. Ritter. Rock Fragmentation by Ex-
Uni v. Technol. , Lulea, Sweden, 1983, plosives. Paper in First International
PP• 469-494. Symposium on Rock Fragmentation by Blast-
11. Siskind, D. E., R. C. Steckley, ing. Lulea Univ. Technol., Lulea, Swe-
and J. J. Olson. Fracturing in the Zone den, 1983, pp. 225-249.
Around a Blasthole, White Pine, Mich. 16. Holloway, D. C., D. B. Barker, and
BuMines RI 7753, 1973, 20 PP• W. L. Fourney. Dynamic Crack Propagation
12. Siskind, D. E., and R. R. Fumanti. in Rock Plates. Paper in the State of
Blast-Produced Fractures in Lithonia the Art in Rock Mechanics (Proc. 21st
Granite. BuMines RI 7901, 1974, 38 pp. U.S. Symp. on Rock Mechanics, Rolla, MO,
13. Olson, J, J., R. J. Williard, May 28-30, 1980). Univ. MO-Rolla, 1980,
D. E. Fogelson, and K. E. Hjelmstad. PP• 311-379.


By David E. Siskind1 and John W. Kopp2


The Bureau of Mines, in a contract causes. In the field tests, research-

study, examined blasting vibration im- ers found no .significant direct effects
pacts on low-yield domestic water wells from the blasting. However, in three of
in the Appalachian coal mining region. the four cases, they did observe changes
Researchers surveyed 36 case histories to in the static water levels and specific
determine if blasting was likely to have well capacities as the excavations ap-
caused the claimed or observed changes, proached to within 300 ft. Researchers
ranging from turbidity to loss of water. attributed these changes to mass rock
Following these investigations, they con- movement resulting from downslope lateral
ducted field studies at four sites where stress relief in the low-yield fracture
the impacts of surface mine blasting system aquifers. With sufficient re-
could be directly measured on operating charge, static levels recovered and ca-
wells of known capacities. pacities increased, provided that the
Researchers found no evidence of blast- nearby mine excavations did not drain the
ing effects at the 36 well sites; aquifers.
instead, they observed other more likely
At about the time the' Bureau of Mines headed by Donelson Robertson, reported
was studying the problems of dynamic vi- their research in a series of three con-
bration response and safe levels for tract final reports available for inspec-
houses near surface mine blasting, al- tion at Bureau centers and for pur-
legations were being made that resi- chase through the National Technical
dential water wells were also being Information Service (!-~).3 This paper
damaged by blasting. Technical experts summarizes the key findings, which were
believed that such effects were unlikely. published in November 1980 as volume 1 of
However, there had never been a carefully the contract final report (l).
designed and controlled study of this The study consisted of three parts:
problem. Such a study appeared justified (1) a background review of vibration and
by the number of alleged cases, particu- other impacts on water wells, such as
larly in the Appalachia coal mining earthquakes, earth tides, and nuclear
region. blasts, (2) examination of 36 cases of
The Bureau contracted with Philip R. alleged damage from blasting in Ap-
Berger and Associates, Inc., to examine palachia, and (3) a careful study of
the problem of possible vibration damage blasting effects on wells at four surface
to residential water wells from nearby mine sites in Appalachia.
surface mine blasting. The Berger team,


The background review found little that ground displacement, such as land slid-
was directly applicable. Observed cases ing, rather than vibration. The types of
of well damage were caused by permanent effects observed required vibration

3underlined numbers in parentheses re-

') . . .
~i'·11n.tng fer to items in the list of references at
Twin Cities Research Center, Bureau of the end of this paper.
"1ines, Minnen.polis, MN.

levels many orders of magnitude higher specifically involving mining were con-
than typical blasting vibrations and were cerned with pit excavations and included
listed as "casing collapse, earth dis- interception with the aquifer, pumping
placement, pump base displacement, mis- from bit bottoms, and ground water
alignment of pump column," etc. Cases pollution.



Inquiries to Appalachian surface mines, tested in any quantitative way.

regulatory agencies, explosives sup- That test was inadequate and made
pliers, coal companies, insurance compa- the owner think he had a much bet-
nies, and trade associations identified ter well than was actually the
36 reports of blast damage to wells. Of case.
these, 24 sites were visited for either Consequently, it was very diffi-
direct well measurements or discussions cult to confirm or deny that blast
with owners and/or mine operators. damage had occurred, but among the
In the Berger report, Robertson states: 36 examples, some of the well his-
tories suggested two scenarios in
In many cases, it was apparent which blasting might cause damage.
that the damage claimed was caused The first is that the ground vibra-
by something other than blasting. tions might be sufficient at times
In other cases, it was clear that to cause loose material such as
there had been a general lowering drill cuttings to slough off the
of the water table, possibly as the uncased borehole and cause the
result of unplugged flowing test water to become temporarily turbid,
hofes, drainage at the high wall, or if enough material was involved,
or a two-to-threefold increase in to bury pump components at the bot-
the number of residences utilizing tom of the well. The second con-
a limited supply, combined with cerns those wells that obtain their
seasonal changes. water from flooded and abandoned
In nearly every case, there was a deep mine workings. Ground vibra-
lack of good benchmark data. Many tions might be sufficient at times
residents have only a vague idea of to cause roof falls that could stir
the depth of their wells. Fewer up sediment in the water or disturb
know the depth of the casing. None an existing potable water-mine acid
of the residents interviewed knew stratification. Of course, slough-
the source of the water in their ing of the well bore and mine roof
wells. About 50 pet had a vague falls can occur in the absence of
idea of the static water level in blasting, so these scenarios are
the well when it was initially com- not exclusive.
pleted. Only one well had been


GROUND WATER OCCURRENCE IN APPALACHIA relief associated with natural topo-

graphic development. The coal seams of-
Most ground water used for domestic ten serve as the primary water conduit,
supplies in Appalachian coal-bearing being low in tensile strength and having
strata are in vertical fractures, joints, extensive vertical fractures. Often, the
and along bedding planes. Some of these coal is underlain by relatively fracture
joints are tectonic in origin and have a free claylike rock preventing further
regional pattern. However, local frac- vertical migration. In their study, Ber-
ture systems exist from lateral stress ger engineers found that local systems

TABLE 1. - Appalachian well water characteristics

Iron •.••••••••••.••••••• Commonly exceeded recommended 0.3 mg/1.

Manganese •••••••••..•••• Often exceeded standard of O.OS mg/L.
Sulfates •••••••••••••••• 14 to 240 mg/L, below recommended level of 2SO.
Total solids:
Suspended ••••••••••.•• Within acceptable ranges.
Dissolved •••••••..•••• Do.
pH levels ••••••••••••••• Most were within 6 to 8, with the total range
of S to 8.7.
Color •••••.••••••••••••• Within acceptable ranges.
Odor ••••••.••••••••••••• Do.
Turbidity ••••••••••••••• Commonly exceeded standard limit of S units.

did not always interact, with static wa- become popular to line all wells deeper
ter levels somet imes varying between than 100 ft to control sloughing prob-
wells only 10 to 3S ft apart. lems.) Important considerations that to-
gether determine the continuous capacity
WATER CHARACTER AND QUALITY of the well are the pump depth below the
static water table, storage capacity of
Information on water quality was ob- the well, natural recharge rate, and pump
tained from the literature and tests made size.
on wells for this project. The general The response to pumping these low-yield
results are summarized in table 1. The fracture water-table systems is a rapid
only observed problem was an occasional drawdown until a near-equilibrium situa-
instance of rusty or reddish-colored tion is reached. The pump must ~e suf-
water seemingly unrelated to the specific ficiently below the static water table
iron content. Jn other studies, this (submergence) to allow this drawdown and
coloration (red slime) was found to cor- still retain water above it. Larger
respond to the presence of iron bacteria pumps produce greater drawdown and re-
and was a problem when wells went un- quire correspondingly deeper submergence,
pumped for ~ long time. The sulfate unless the flow is restricted by a valve
levels suggest no serious acid mine arrangement. Additionally, rapid draw-
drainage nor influence from acid rain, downs from high pumping rates can cause
which is typically of pH 4.S to Sin the abrupt pressure changes, turbid water,
region. and possible "sanding up" of the pump.
As an example, Robertson calculates
TYPICAL DOMESTIC WELLS IN APPALACHIA that an increase of pump submergence from
SO to 100 ft allows a constant pumping
Residential wells in Appalachia are rate from a well with only one-fourth the
typically 6~in-diameter rotary or cable- specific capacity. He states that a
tool drilled and 100 to ISO ft deep pumping rate of 5 gal/min could be
(maximum about 400 ft). Normally, only maintained by the typical Appalachia
the top 20 ft or so is cased. The re- fracture-system well (specific capacity
mainder is unprotected from sloughing off of 0.093 (gal/min)/ft of drawdown) pro-
the well sidewalls, which is a common and vided it has 100 ft of pump submergence.
normal occurrence. (Recently, it has
EXPERIMENTAL DESIGN shallow wells were drilled to obtain wa-
ter from the coal measures. Additional-
Researchers conducted an evaluation of ly, deep wells drew water from the layers
four producing wells to specifically below the coal and were isolated from the
determine the influence of blasting on upper measures. For each of these test
well productive capacity. At four sites, wells, two or three observation wells

were used to monitor drawdown at varying

Dynamic effects were measured by 10-h
drawdown tests before, during, and fol-
lowing blasting. The static water level
was monitored with float gauges. Water
quality was also sampled periodically.
To monitor blasting vibrations, research-
ers placed seismic instrumentation on the
surface and at the bottom of an observa-
tion well for the 1-yr test duration.


Four sites were found to be suitable

for the tests, allowing sufficient moni-
toring time and standoff distance, and FIGURE 1.-Location of test sites.
within convenient travel distance from
Pittsburgh, PA, where Philip R. Berger experimental parameters is given in table
and Associates is headquartered (fig. 1). 2.
A summary of the four test sites and


Essentially, the same study was done at iron and drill cuttings, and disturbance
the f~ur sites in Appalachia. A capsule from the manual sample-collecting proce-
summary is given in table 3. dure. A continuously used well would not
have some of these problems.
Researchers observed a correlation be-
tween the mining and well changes. The Results were similar to those at Bro-
closest blast removed supporting material therton, except for smaller changes in
(e.g., toe), allowing lateral stress re- static water levels. Specific capacities
lief and the widening of vertical frac- increased during the 12-month test for
tures. In other words, the pit excava- both shallow and deep wells. They went
tion and not the vibrations influenced from initial values of about 0.065 and
the fracture system. Changes were ob- 0.020 (gal/min)/ft to 0.66 and 0.05, re-
served in the shallow well but not in the spectively. As at Brotherton, this was
deep well, likely because of the more ex- attributed to removal of downslope rock,
tensive vertical fractures near the sur- resulting lateral stress relief, and the
face. Significantly, the drops in static consequent widening of vertical frac-
water level corresponded to increased tures. Along with the increased capacity
specific capacity. The wider cracks al- was the availability of increased re-
lowe~ a temporary drop in static level, charge from the coal seam into the tight
which would recover with sufficient rain- sandstone formation, possibly accounting
fall. However, the cracks also provided for the little observed change in static
increased flow .(specific capacity). water levels. Figures 2 through 5 show
Turbidity results were harder to inter- the site plan, profile, well arrangement,
pret. Irregular fluctuations occurred and a drawdown test record for the date
during the 11-month study, and some tem- June 9, 1979, when an 0.80-in/s particle
porary increases could have been from velocity was recorded.
blasting. The nature of the tests con- Researchers concluded that ground vi-
tributed to the turbidity problem: in- brations produced no deleterious effects
frequent and periodic pumping, suspended on either the deep or shallow well, with
TABLE 2. - Blasting and pumping tests at four Appalachian mines 0
Shallow well Deep well Blast Scaled Resultant
Site Well Pump Well Pump distances, distance, vibration Notes
depth, rate, depth, rate, ft ft/lb 1 !2 velocity,
ft gal/min ft .e:al/min in/s
Brotherton, 109 2.5 -7.9 169 4.0 -4.6 165- 500 12.1- 98 0.04..:.2.20 18 drawdown tests. Well
PA. site was 40 ft above
initial mining. Pit was
pumped at 500,000 gal/d •
Tenmile, WV. 146 .21-4.1 187 .27-1.0 64- 580 2.4- 34 • 11-5. 44 23 drawdown tests. Well
site was on steep slope
above blasting •
Rose Point, None None 158 1.8 -6.8 175- 775 12. 8- 60 • 64-2.14 13 drawdown tests. Well
PA. site was below mining.
St. Clairs- 69 .24- .64 163 .24- .64 425-1,000 32 -153 .25- .84 14 drawdown tests. Well
ville, OH. site was on slope above
mining. Study ended be-
fore blasts reached
--- wells.

TABLE 3. - Summary of results at four test sites

Site Static water level Specific capacity, Water chemistry Turbidity
Brotherton, Dropped 21 ft in 11 months from Initially 0.34, steady for No changes ••••••• Possible tempo-
PA. pit pumping. Slight drops for 11 months and then started rary increases
closest blasts (shallow well to increase (shallow well from blasting.
only). Recovery related to only).
Tenmile, WV. Minor variation. Some drop for Improved as test progressed ••• do •••••••••••• Highly varia-
shallow well with recovery by by factor of 10 for shallow ble. Possi-
recharge. Little change for well and 3 for deep well. ble increases
deep well. Highest value was 0.66. from blasting
within 300 ft.
Rose Point, Very minor variation except for Initially 0.33. Improved Unexplained vari- Little
PA. one temporary increase. late in test except for one ations and dis- variation.
temporary decrease. Then crepancy between
recovered to initial value. field and labo-
ratory tests.
St. Clairs- Incomplete data from equipment Very low at 0.007 to 0.017 Little variation. Insufficient
ville, OH. failure. except for unexplained ini- data.
tial high values.

- -~- ------- - -----



12 0
0 0
14 0
0 IS lost (0.80 in/s)
z 16 0 t
~ 0
18 0
20 0

I 22


I FIGURE 2.-Tenmile test site plan.

FIGURE 5.-Tenmile drawdown test for deep well (D-1).


This site differed from the previous
two in that the well was below the blast-
ing. Consequently, researchers expected
little stress relief effect. Indeed,
little change occurred in either static
water level or specific capacity until 5
1,$00' months into the test, when specific capa-
Well D-1
0 200 city jumped from about 0.33 to nearly
Scole, ft 0.60 (gal/min)/ft and later decreased.
FIGURE 3.-Tenmile test site profile. Researchers attributed the changes to
nearby removal and replacement of over-
burden, which occurred at that time. As
Well Well Well Well Well with previous sites, no direct effect of
D-2 D-1 S-1 S-2 · S-3 blasting was evident.

cp__ ,o ft_cp_,o tt-cp~7 u-Ls n---cp ST. CLAIRSVILLE SITE

200 ft 200 ft 160 ft 160 ft 160 ft
This site was characterized by very low
0 10 capacity wells, which were expected to be
E~ J
Scale, ft susceptible to outs.ide disturbances. The
FIGURE 4.-Tenmile well layout. Deep and shallow pumped initial drawdown test in the shallow well
wells are D-1 and S-1. Others are observation wells. appeared satisfactory, with the well
able to sustain pumping rates of
one possible exception. A blast that 0.86 gal/min, equivalent to a specific
went offscale at 2 in/s (distance of capacity of 0.041 (gal/min)/ft.
85 ft) could have dislodged and caused a The deep well was worse, and was only
loose rock in the sidewall to shift into able to maintain 0.21 gal/min for about
the hole, producing a partial bridge and half the normal test period of 600 min.
preventing the sounding of the bottom. (For this site, as well as the others, an
Turbidity results were also similar to unperforated liner and packer were used
those from the Brotherton tests, with to isolate the deeper well from recharge
wide variations, difficult measurement from the shallower coal measure being
conditions, and the suggestion of tem- mined.)
porary increases from close-in blasts, One month after the initial drawdown
less than 300 ft away. test, the shallow well was again pumped,

but this time it had a specific capacity summarizes the average relative resultant
of only about one-third as much and at velocity amplitudes.
about one-half the pumping rate. Re-
searchers suspected that air had been TABLE 4. - Reduced vibrations
entrapped from initial tests, preventing measured at depth
full recharge.
In summary, researchers observed no Site Depth, ft Relative
clear blasting effects at this site. Un- amplitude 1
fortunately, work stopped short of dis- Brotherton, PA •••• 149 zo.34
tances found to produce effects observed 3 .68
at the three other sites. Some equipment Tenmile, WV ••••••• 160 .44
failures were also experienced. Rose Point, PA •••• 168 .14
St. Clairsville,
DOWN-THE-HOLE VIBRATIONS OH ••••••••••••••• 180 ~.25
1 Depth vibration divided by surface
Vibrations were monitored at the bottom vibration.
of shallow observation wells as well as 2Blasting in poorly confined upper
on the surface. Downhole vibrations layers.
3 Blasting in well-confined lower
were, as expected, of lower amplitudes,
suggesting less risk to subsurface as layers.
opposed to surface structures. Table 4


Research at four sites in Appalachia noticed. Of benefit to the well user is

found no catastrophic effects on water the increased storage and flow as shown
wells from blasting at vibration levels by higher observed specific capacities.
up to about 2.0 in/s . . At three of the Shallow wells exhibited this effect more
sites (and possibly the fourth, had test- than deep wells, consistent with expecta-
ing continued), long-term changes were tions that more extensive fracture sys-
observed and were attributed to the tems exist at shallow depths. At one
removal of confining rock. site, backfilfing reversed the improve-
As blasting and the pit excavation ap- ment, either from clogging by fines or
proached within 300 ft of the wells, the reintroduction of crack-closing lateral
mechanism of lateral stress relief al- confinement.
lowed vertical fractures to open. Be- Blasting may cause some temporary in-
cause these fracture systems are typi- creases in turbidity, of the same order
cally the abode and conduit of shallow as those occurring in the absence of
Appalachian ground water, the static blasting. Results were uncertain on this
water levels then dropped over a period because of the difficulty of sampling
of weeks. With sufficient rainfall, the without causing disturbance and natural
water levels would return. Where suf- sloughing. Plastic well liners were
ficient submergence exists, such minor recommended to control turbidity.
changes in static level would not be
l. Robertson, D. A., J. A. Gould, J. 2. Berger, P. R., D. T. Froedge, J. A.
A. Straw, and M. A. Dayton. Survey of Gould, and L. F. Kreps. Survey of Blast-
Blasting Effects on Ground Water Supplies ing Effects on Ground Water Supplies in
in Appalachia (contract J02BS029, Philip Appalachia. Part II (contract J02B5029,
R. Berger and Associates, Inc.). Volume Philip R. Berger and Associates, Inc.).
I. BuMines OFRS(l)-82, 1980, 159pp.; BuMines OFR 188-83, 1982, 114 pp.; NTIS
NTIS PB 82-152125. Volume II. BuMines PB 84-113182.
OFR 8(2)-82, 1980, 266 pp.; NTIS PB 82-



By David L. Schulz1


Following ideas developed by research- fiber optic probe for downhole measure-
ers at the University of Maryland, the ment of explosive shock front position
Bureau of Mines assembled a versatile, and velocity. The accuracy of the probe

II readily available, and very inexpensive was determined in field testing.

Detonation velocities of explosive col- University of Maryland researchers used a
l umns are often measured to determine in fiber optic system to measure detona-
' situ explosive performance. Several tion velocity. Discussions with these

I methods are in use, ranging from simple

resistance probes to very sophisticated
and expensive electromagnetic resonance
researchers and further Bureau work led
to the development of the fiber optic
probe described in this paper.
measuring systems.
In a recent study on blasting, under
Bureau of Mines contract S0245046,
The fiber optic probe consists of a data transmission applications, the mod-
fiber optic cable to detect and carry ules cost about $16 each and are re-
the detonation-zone light emission and usable. Bright light through the system
a sensor to detect and convert the
light signal into an electrical signal.
Figure 1 shows the Bureau's system.


The fiber cable is a DuPont Crofon2

lightguide consisting of sixteen 0.010-
in-diameter optic strands in a plastic
tube with an overall outside diameter of
0.087 in. The lightguide used in the
Bureau's tests was obtained from the
Edmund Scientific Co. at a cost of about
$0.70/ft in September 1986.


The light detector is a fiber optic

data link receiver module-100 series with
the trade name Optolink. Designed for

Electronics technician, Twin Cities

Research Center, Bureau of Mines, Minne-
apolis, MN.
2 Reference to specific products and
sources does not imply endorsement by the FIGURE 1.-Fiber optic detector and cable. Cable outside
Bureau of Mines. diameter is 0.087 in.

gives an output of about 100 to 300 mV. receiver-integrated circuit chip to am-
These particular modules by Interoptics plify the signals in one package. The
were obtained through the Newark Elec- overall dimensions are approximately 1/2
tronics catalog. by 5/8 by 7/8 in.
The receiver module incorporates a
photo diode to detect the signals, and a


The fiber optic cables are cut to sensor, the light signals are converted
length with a sharp tool to get a good to millivolt signals. Typical rise time
light-conducting surface, and one end is is 0.01 ms. The millivolt signals are
inserted into the light detector. The measured by an oscilloscope using a mem-
other end of the cable is embedded in the ory or hold feature to retain both the
explosive column from the top down to the voltage spikes for each bundle of fibers
predetermined distance from the initia- and an initiation-time spike. From known
tion point, usually the bottom of the ex- insertion distances and measured times,
plosive column. Intense light is given velocities are simply calculated. A
off as the detonation front moves up the four-channel Nicolet digital oscillo-
explosive column. This light is picked scope, model 4094A, set to a time scale
up by the fiber optic as the detonation of 0.5 us per point, was used for the
passes the ends of the fiber cables. Bureau's tests.
Carried by the fiber optic cable to the


Bureau researchers used the fiber optic A second set of tests was run on larger
system for two series of field studies of charge diameters. Explosive columns were
explosively produced rock fragmentation. 16 ft long and consisted of 2-in sticks
The first tests used explosive columns of of 60-pct-extra dynamite. Holes for the
7/16-in-inside-dlameter, 40-in-long plas- fiber optic b~ndles were punched in the
tic tubes filled with 60-pct-extra dyna- sticks at desired distances from the ini-
mite. The measured detonation velocity tiation point. The optic cables were
was 8,680 ft/s at the bottom of the col- then inserted and taped in place and the
umn where the dynamite was densely powder cartridges carefully loaded. Ac-
packed. At the top where the powder was curacy was determined by the dimensional
less well packed, the measured veloci. ty stability of the column of powder car-
was 7,350 ft/s (fig. 2). In 5/8- tridges (some compaction can occur). Re-
in-inside-diameter columns, ~easured ve- searchers measured a steady-state ve-
locities were 9,920 to 11,400 ft/s. locity of 15,140 ft/s.

Probe 8 -----.FP<ebe A



0 100
TIME, p.s
FIGURE 2.-0utputs from fiber optic probe for bottom-initiated blasthole showing measured velocity between probes of 7,353 ft/s.


Advances in fiber optic technology have of detonation velocity for as little as

provided an inexpensive and versatile $10 per blast. Only a memory oscillo-
method to measure explosive detonation scope_is needed, in addition to the fiber
velocity. An off-the-shelf system can be optic cable and detector module.
assembled and accurate mesurements made



by John W. Kopp 1


Stemming is used in blasting operations properly stemmed blasts, stemming is con-

to help contain explosive gases as long tained until some burden movement has oc-
as possible. Stemming can reduce air- curred. Test blasts at two surface lime-
blast, improve fragmentation, and reduce stone quarries were evaluated using
the chances of hot explosive gases ignit- high-speed photography. For the condi-
ing methane and dust explosions in under- tions of these tests, a stemming length
ground mines. of at least 26 charge diameters was found
The types and amounts of stemming ma- to prevent premature stemming ejection.
terial that are desirable in underground In tests with stemming lengths of 16
metal and nonmetal mine blasting to im- charge diameters, the stemming was ef-
prove fragmentation while containing the fective but there was early venting of
hot gases are largely unknown. This Bu- hot gases through fractures in the rock.
reau of Mines research examined the ef Further testing with other rock types,
fectiveness of differing lengths of stem- hole diameters, explosive types, and
ming by measuring stemming ejection times stemming materials to determine their
as related to burden movement. With effect on incendivity is recommended.


Methane emissions in underground mines noncoal mines. Most of these operations

can present hazards, especially when ig- use conventional explosives in standard
nition sources such as hot gases from underground blasting practices. Safety
explosives are present. The problems is sometimes ensured by evacuating all
associated with blasting in underground personnel to the surface during the
coal mines have been addressed by use of blast. However, this is often not prac-
permissible explosives and permissible tical for large mines utilizing mining
procedures for their use. However, meth- methods such as room and pillar. Some
ane also occurs in some noncoal under- mines require 20 or more blasts per day,
ground mines, particularly oil shale, involving large amounts of explosives.
trona, salt, potash, copper, limestone, In order to maintain production, blasts
and uranium mines. At present, such must be scheduled while personnel are
mines receive a variance from the u.s. working in the mine.
Mine Safety and Health Administration The contractor made a number of recom-
(MSHA) blasting regulations, depending on mendations for blasting underground with
the source of methane, associated ore personnel present in the mine. Important
body, and the method of mining. Conven- among these was use of stemming to con-
tional explosives and blasting agents, tain the hot gases and flame from the ex-
rather than permissible explosives, are plosion in the borehole until expansion
normally used for both practical and of the burden sufficiently cooled the
economic reasons. gases to prevent ignition of methane.
A recent Bureau of Mines contract 2 The contractor made some predictions of
examined blasting practices in gassy stemming behavior, based on a simple
mathematical model, and recommended that
neer, Twin Cities Research his calculations be confirmed by field
Center, Bureau of Mines, Minneapolis, MN. studies.
2 contract J0215031; Bauer, Calder & This Bureau study was conducted as
Workman, Inc. a followup to the previous study, to

measure the retention time of various careful control of the test blast design
lengths of stemming in the borehole dur- variables and adequate lighting for high-
ing normal blasting and to relate stem- speed photography. While this study is
ming retention time to the burden move- directed at blasting underground, the
ment caused by the expanding gases of the efficient use of stemming will improve
explosive. Tests were conducted at two blasting in surface mines also.
surface limestone quarries, which allowed


Early experimentation had shown that cratering shots were detonated in a fac-
high-speed cinematography was the best torial experiment to test two types of
method for measuring stemming movement. stemming material at three lengths of
This method also allowed observation of stemming and with two explosive types.
the burden displacement and dust and A high-energy explosive and a rela-
smoke generated by the blast. tively low energy explosive were used for
Field experiments were filmed with two this series of tests. The low-energy ex-
cameras, a 16-mm rotating-prism camera plosive was chosen to produce results
capable of speeds up to 11,000 frames per similar to those from ammonium nitrate
second and a 16-mm registering-pin camera and fuel oil (ANFO). Both explosives
with filming speeds up to 500 frames per were in 1-1/4-in-diameter cartridges.
second. The registering-pin camera has Enough explosive was used in each test to
better resolution than the rotating-prism make a charge 16 in long. The volume of
camera and thus provides a much clearer explosive was not varied for this test
picture. series. Properties of the explosives
The time of detonation of the explosive used are shown in table 1.
was recorded on film with Nonel3 shock The blastholes had a 1-l/2-in diameter
tubing. A known length of shock tubing and ranged from 36 to 72 in deep. The
was attached to the explosive charge and holes were drilled vertically in a lime-
passed through the stemming to the sur- stone quarry floor. This represents the
face and coiled to allow the flash to be worst_ case for blasting efficiency, al-
recorded on film. Detonation of the ex- lowing relief in only one direction,
plosive initiated the tubing, which deto- upward.
nated at 6,000 ft/s. Thus, the time of The stemming material consisted of
detonation was determined by noting the crushed limestone in one of two sizes:
flash of the coiled tubing on the film The material was either drill cuttings
and calculating the time required for the screened to less than minus 10 Tyler
detonation to reach the surface. series mesh size (0.0661 in) or limestone
Full-scale field tests were performed gravel between 3/8- and 3/16-in size.
at a surface limestone quarry in order to The stemming material was added above the
eliminate lighting problems when filming explosive and filled the hole to the col-
with the high-speed cameras. Twelve lar. Table 2 shows stemming length and
type, and assignment of shot numbers.
to specific products does The stemming was lightly tamped and had a
not imply endorsement by the Bureau of density of about 1.5 g/cm 3 •

TABLE 1. - Properties of explosives used in the test series

Explosive Density, Detonation Relative Explosive Borehole

type g/cm 3 velocity, bulk temperature, pressure,
ft/s st~ngth 1 K atm
High energy I 1.16 15,000 148 3,000 30,000
Low energy , 1. 07 10,500 115 ··---~--
2,870 19,000
ANFO 100.
108 i

TABLE 2. Experimental design and assignment
of test numbers

Length of stemming . •••.........• in • • 20 32 50-60

Fine drill cuttings:
High-energy explosive ••••••••••••• S-3 S-2 S-11
Low-energy explosive •••••••••••••• S-6 S-12 S-1
Coarse crushed stone:
High-energy explosive •••••••••••• • I S-4 S-8 S-10
Low-energy explosive ••••••••••• ·~. S-5 S-7 S-9


Twelve cratering tests were conducted The burden velocity was calculated to be
according to the factorial design in 27 ft/s. The increase in burden volume
table 2. All shots were monitored with was calculated by assuming the burden
the two high-speed cameras described pre- movement to be in the shape of a cone and
viously, one running at 500 frames per measuring this increase on figure 1. It
second and the other at 1,000 to 3,000 is apparent from the figure that the bur-
frames per second. Analysis of the films den movement is closely approximated by a
showed that stemming was usually ejected cone. The rate of volume increase was
from the shallow holes. With stemming found to be 600,000 in 3 /s. A plot of
lengths of 20 in, the stemming material stemming movement and burden expansion
was ejected as follows: versus time elapse from initiation of
detonation is shown in figure 2 for shot
Ejection, S-3. It is apparent from figure 2 that
Shot ms considerable burden movement occurred be-
fore the stemming was completely ejected
S-3........................... 13 after 13 ms. In this case, some cooling
S-6........................... 32 of the hot explosive gases had occurred
S-4••••••o•••••••••••••••••••• 8.8 because of volume expansion as the gases
S-5........................... Retained worked their way into the fractured bur-
den region.
Longer stemming lengths resulted in re~ An estimate of the amount of cooling of
tention of the stemming. When stemming the explosive gases can be obtained by
is retained, it has done its job in terms assuming the thermodynamics of the expan-
of confining the hot eKplosive gases. sion to be adiabatic. The expansion is
However, when stemming is ejected, fur- rapid and allows little time for heat to
ther analysis is required to determine if be exchanged between the gases and the
an adequate length of stemming was used. surrounding rock. From the first law of
Further analysis of the films indi thermodynamics it can be shown that tem-
cated the motion of burden. Not only can perature and volume of the gas are
velocities of various parts of the burden related as follows:
be calculated, but an estimate of the in-
creased burden volume caused by the ex- 1
panding gases can be made. Figure 1 T 2 T- 1 V 2'
shows stemming and burden movement for
shot S-3, which was a test with 20 in of which on rearranging becomes
fine stemming, using the high-energy ex-
plosive, which resulted in a stemming
ejection time of 13 ms. Initial movement
of the stemming was obscured by dust
caused by venting through cracks in the
burden. The initial velocity of the where T 1 , T2 are the initial and final
stemming was calculated to be 190 ft/s. temperatures, V1 , V2 are the initial and

I ---Burden

I ---Dust
4, 6, etc. Film frame number

o----------- 0

FIGURE 1.-Stemming and burden movements of shot 5·3.

final volumes, and T is the ratio of the new spaces created by fracturing of the
heat capacities of the expanding gases. burden, this is approximately balanced,
The actual value of T depends on the as no allowance is made for expansion of
molecular structure of the gases in- the gases into existing voids or for
volved. Most of the gas products of the porosity of the rock. The estimated tem-
explosives used are diatomic and poly- perature is thus an approximation but
atomic gases: nitrogen, carbon dioxide, provides some insight into the phenomena
and water vapor. The average value of T involved.
for these gases is approximately 1.3. Figure 2 also shows the predicted gas
The equation describing the gas temper- temperature decrease based on the use of
ature line in figure 2 becomes equation 3 and the burden volume increase
as determined from analysis of high-speed
films. The stemming remained in the
borehole for 13 ms, at which time the gas
temperature is estimated to have cooled
where V1, T1 are the initial volume and from the detonation temperature of 3,000
temperature of the explosive at detona- K (2,727° C) to 550 K (277° C), which
tion, V is the increase in volume, and T would be sufficient to prevent ignition
is the temperature at that volume. of a methane-air mixture, since the ig-
It is now assumed that the gas expands nition temperature of methane is 905 K.
into all of the new volume created by the However, from figure 1, it is evident
expanding burden. Though the expanding that venting, probably through a major
gases may not actually fill all of the fracture, occurred before expulsion of

30 12
I ,;

,."" "'
'''o ,;

\ - - - - Burden
"' c:
\ - - Temperature ,;
<t 2 t-" 20
z \ ,; "'
,; w
8 (j)
::E \ ,;
,; <t
> \ ,; "' 0::
,; "' z
::E \ "' "'
"' w
<t <.::) \ .-"'"' :2

\ ,;" "" ...J

'' '
<t ::E ,;
u w 10
4 z
,. .. "' w
w ......... ;"' 0
0:: ;"' 0::
0 -..........::"' :::;)

..... ;
-------- ---------- CD

,.," ""
, ... Stemming ejected
,. .. ""
0 0 :a
2 4 6 8 10 12 14 16 18 20
TIME, ms
FIGURE 2.-Relative movements of stemming and burden for shot S-3 and the associated gas temperature.

the stemming, at 3.8 ms after initiation. Shot S-6 had 20 in of fine stemming and
Figure 2 shows that the explosive gases the lower volume-energy explosive. Stem-
would have cooled only to 1,300 K after ming was ejected from shot S-6, though
3.8 ms, not sufficient to prevent methane at a slower rate than with the higher
ignition. energy explosive (fig. 4). The time of
A similar analysis was performed for ejection was 32 ms, and the estimated gas
shots S-4 and S-6. Results are presented temperature at the time was 625 K. Vent-
in figures 3 and 4. Figure 3 presents . ing of smoke or dust was also observed,
the analysis of shot S-4 where 20 in of starting at 4.8 ms after initiation.
coarse stemming was ejected using the From figure 4, the estimated gas temper-
high-energy explosive. At the time of ature at the beginning of venting was
stemming expulsion at 8.8 ms, the explo- about 1,300 K, high enough for a methane
sive gases were estimated to have cooled ignition.
to 650 K, within a safe temperature Four shots were fired with 32 in of
range. However, venting of dust and stemming in each hole. The stemming
smoke was observed at 3.5 ms after initi- remained intact for all these shots. The
ation. From figure 3, the estimated tem- film analysis did not show any stemming
perature of the gases would be 980 K, movement. Burden movement was slower
above the safe limit. Again, the venting than in the previous shots using less
probably occurred through existing frac- stemming. With the exception of shot S-
tures in the rock. This shot and the 12, no venting of smoke or dust occurred.
previously discussed shot used a high- Also, a rubble zone of broken material
energy explosive, but the previous shot was left at the surface of each hole.
used a different stemming size. The The zone was about 2 ft in diameter for
finer materi~l held longer. three of these shots, smaller thAn the

:::.::: KEY
w c
a:: "'g
~ 2 ~ 40 4 w
<r z (/)
a:: w <l:
w :! w
0... w a::
:! > u
w 0 z
1- :!
<r l? :!
l? z ::J
...J :! 0
u ~ 20 2 ~
1- 1- w
w (J)
a:: a::
------ ::J

e jecled

FIGURE 3.-Relative movements of stemming and burden for shot S-4 and the associated gas temperature.

30 I'


:::.::: I -----Burden
--Stemming I'

''b I --Temperature

w \ /
/ "'o
·=t-=' 20 \ I'
1- I' <r
a:: w
\ /
/ w
\ /
r- :! 15 \ /
I' w
z \ I'
/ :::>

l? / 0
~ >
.J /
:! / z
<r w
u 1- '~/
/....._ w
f= (/)
w /
a:: I' ro
0 /
w /
:X: /
1- Venting /

0 ..______.~.~ 0
0 5 35
TIME, ms

FIGURE 4.-Relative movements of stemming and burden for shot S-6 and the associated gas temperature.

2 to 5 ft for 20-in stemmed blasts. The retained in all cases with stemming re-
presence of rubble zones at the surface gions of 32 in or greater. When 20 in of
for both the 20- and 32-in stemming cases stemming was used, there was stemming
indicates that with 32 in of stemming, ejection in three of the tests after 8.8
the explosive is still sufficiently close ms or more, but venting of gases through
to the surface to allow fragmentation of fractures occurred before 8.8 ms. Esti-
the burden. mates were made of explosive gas cooling
The final four shots of this series associated with volume increases as the
used stemming lengths of 50 to 60 in. No explosive gases expanded into the frac-
stemming movement was detected for any of tured burden. In the three tests where
these shots. Burden movements were also stemming ejection occurred, explosive
smaller than for the previous shots. gases were estimated to have cooled suf-
Figure 5 shows a comparison between bur- ficiently to prevent ignition of methane
den velocities and the length of stemming in the time required for stemming ejec-
used in each hole. The type of explosive tion (8.8 ms) but not at the time when
used made a difference in burden velocity venting of gases through fractures occur-
only at the two shortest stemming red. There was no premature venting of
lengths. Differences caused by stemming gases with a 32-in stemming region. It
type were inconclusive. is concluded then that for the conditions
In this crater test series with 1-1/4- of these tests, a stemming length-
in charges in 1-1/2-in- diameter blast- to-charge-diameter ratio of 26 (32 in of
holes, it was found that stemming was stemming) was adequate to prevent igni-
tion of methane. However, a stemming
length of 16 charge diameters (20 in of
stemming) could have resulted in methane
KEY ignitions be~ause of early venting of hot
o High-energy explosive gases through fractures in the limestone
o Low-energy explosive rock.
Bauer, Calder & Workman, Inc., sug-
gested a simple physical model to predict
the time required to eject stemming.
20 This model d~pends only on inertia of the
...... stemming material and not on frictional
::: forces to resist movement, and thus the
>- acceleration, a, of the stemming is given
u by

_J F
w a (4)
> M
a:: where F is the force exerted on the stem-
(I) 10 ming by explosive gases and M is the mass
of the stemming.
The equation of motion is thus

s V0 t + 1/2 a t2, (5)

where S is distance traveled by the stem-

0 ming, V0 is initial velocity, and t
= time in seconds.
Combining equations with V0 0 gives

FIGURE 5.-0bserved burden velocities versus length of stem-

ming for the 1 '12-in-diameter shot series.
t = flfi· (6)

The force F can be estimated from the All observed times for ejection were
borehole pressure, P, times the cross- much longer than the calculated times.
sectional area of the hole, A, or In many cases, no stemming movement was
observed before the calculated time was
F = PA, (7) past. This calculation method then
should be used only to obtain an estimate
and the mass of stemming equals of the minimum stemming ejection time.
Improved estimates will require the
M = Ps Al, (8) addition of frictional forces in the
where Ps is the density of stemming ma- A clue to this late stemming movement
terial, A is the cross-sectional area, is provided by the instrumentation used
and £ is the length of stemming. in a 6-in vertical cratering blast by
Substituting gives Sandia National Laboratories4 (shot V-1).
The blasthole was instrumented with the
SLIFER system to monitor the detonation
rate of the explosive. This instrument

This prediction method yields ejection D. I.. , J. E. Uh l, and R. I. •

times for the shots in this investiga- Parrish. Atlas Cratering Tests. Ch. in
tion as shown in table 3. Also shown in Oil Shale Program Quarterly Reports, Oct-
table 3 is the time of first observed ober 1984 through March 1985. Sandia
stemming movement and observed time to Nat. Lab., Rep. SAND 85-2768, 1986,
ejection. PP• 5-26.




I 0
j :lE

(!) 2

52 53 54 55 56 57 58 59 60 62
TIME, ms
FIGURE 6.-SLIFER data from hole 3 of shot V-1 showing detonation of explosive column and crushing rate of stemming. From
Sandia National Laboratories.

TABLE 3. - Calculated and actual stemming ejection times

Hole Stemming Calculated First stemming Observed

Shot diameter, in length, in ejection movement, ms ejection
time, ms time, ms
s- 3 • •••• 1-1/2 20 o.s 3.4 13
S-4 . .... 1-1/2 20 .s 4.6 9
S-6 ••••• 1-1/2 20 .6 6.1 32
v-1 1 •••• 6 108 2. 1 10 Retained
1 6-in cratering shot from Sandia National Laboratories.

also observed the rate of crushing or takes 7.5 ms to complete. The time taken
compaction of the stemming in the bore- for the crushing to propagate through the
hole. The record for shot V-1 is shown stemming is close to but somewhat less
in figure 6. The first 6 ft of the rec- than the obseryed time for the start of
ord shows the detonation of the explo- stemming movement. The stemming appears
sive. The detonation velocity is approx- to be bridging in the hole and preventing
imately 19,000 ft/s. Above 6 ft, the movement until it is crushed by the pres-
record shows the crushing of the stem- sure pulse in the stemming column, and
ming. This crushing extends to 12-1/2 only then does it start to move.
ft, or 2-1/2 ft from the hole collar, and


High-speed films of single-hole crater fractures occurred before stemming ejec-

test blasts in two surface limestone tion and the temperature of the vented
quarries were analyzed to evaluate the gas was estimated to be above the methane
ability of stemming to contain explosive ignition temperature. For the conditions
gases. Stemming ejection and burden mo- of these tests, it is thus concluded that
tions were examined. When sufficient a stemming length of 16 charge diameters
stemming was used, ejection of stemming could have resulted in methane ignition.
was prevented. A ratio of length of Stemming ejection takes much longer and
stemming to charge diameter of 26 or more is more complex than would be predicted
was found to prevent premature ejection by a simple calculation based on inertia
of stemming and venting of gases. of the stemming material. Ejection times
If release of stemming does occur, the were three times or more longer than
time required for stemming ejection may those predicted by a simple inertia
be sufficient to permit burden movement model. Reasons for this are the addi-
to start, with expansion of explosive tional time required for the stress wave
gases into the fractured burden and cool- to crush the stemming material and cause
ing of the explosive gases. An estimate it to start to move and the subsequent
of this cooling was made using thermo- decrease of borehole pressure through
dynamic principles. In three tests with crushing and expansion of the bore-
a ratio of stemming length to charge di- hole. A better understanding of these
ameter of 16, the explosive gases cooled mechanisms requires further research with
below the ignition temperture of methane more sophisticated instrumentation, such
in the time required for stemming as the SLIFER system.
ejection, but venting of gases through