paper was first published in the JMU Journal of Mine Action
in April 2000. The number of recorded accidents
in the DDAS has increased by a factor of 10 since this was written 15 years ago, but the general conclusions remain
very similar. The only significant change has been in the number
of deaths caused by ordnance and IEDs following recent conflict, usually involving
devices that could not be realistically protected against with
have approached this subject by studying the risks that deminers
really face and the injuries that result, then working out how
to minimise risk and protect against any residual danger. I
bore in mind that there was no point is prescribing an action
or a garment that would not be used. This is not an approach
widely endorsed in the protective equipment industry, which
apparently prefers to base its assessment of risk on experimental
data and a scale of injury borrowed from the automobile industry.
If the injuries commonly predicted by them were accurate, most
of the accident victims I have worked with would have been dead.
Anyone considering this matter objectively should bear in mind
that demining accidents are rare. While I do not have all the
relevant data, I believe that severe accidents occur at the
rate of one per 25-30 man-years of actual demining. This statement
ignores the facts that some groups have more accidents than
others (perhaps working in more dangerous areas) and only serves
to explain why it is that most actual deminers have not seen
a serious accident and will not wear equipment that they believe
The following paper draws on my five years of field research
and my knowledge of the Database of Demining Incident Victims
(DDIV). The DDIV covers all recorded explosive incidents that
have occurred while demining in Angola, Mozambique, Cambodia,
Bosnia Herzegovina, Laos and Zimbabwe. It also covers all the
accidents that occurred in Afghanistan between 1997 and 1999
and those made available from Kosovo. It does not include details
of civilian accidents and injuries. Often with considerable
detail about the circumstances surrounding an accident, the
records provide a reference for informed analysis.
The DDIV has been accepted as an authoritative resource by GICHD
in its work advising on the revision of UN standards for HD.
The DDIV is available on CD. Contact me for details. [The DDIV became the Database of Demining Accidents when it ecame a relational database in 2003.]
The first part of this paper gives facts in support of the discussion
1 The facts
1.1 Threat activities
There are many opinions of what constitutes the greatest threat
in demining. Using the DDIV, genuine threat activities can be
listed in terms of incident type and frequency.
The type of accident is followed by the number of recorded victims
[The activity titles are defined in full in the DDIV.]
In the DDIV, injuries likely to be life-threatening, to require
surgery or to result in permanent disability are rated as “severe”.
All others are rated as “minor”.
the whole database the following injuries are recorded:
1.3 Devices involved
The following records the devices that are most commonly involved
in recorded incidents.
The Blast mine threat in the various theatres can be summarised
Afghanistan, Iraq, Mozambique – PMN (240g TNT).
Angola – PPM-2 (110g TNT), PMN.
Bosnia-Herzegovina – PMA-3 (35g Tetryl), PMA-2 (100g
Cambodia – PMN-2, Type 72 (51g TNT), M14, MD82B (27/28g).
Kosovo – PMA-2 (100g TNT).
Zimbabwe – R2M2 (58g RDX/WAX).
The PMN represents the largest AP blast threat and is present
in most theatres.
The fragmentation mine threat is far less common and can be
Afghanistan, Angola, Cambodia – POMZ (75g TNT)
Bosnia-Herzegovina – PROM-1 (425g TNT).
Iraq – the Valmara-69 (450g Comp B) and PROM-1 (425g
Laos – a mortar features in the only recorded injury.
Mozambique – OZM-4 (170g TNT).
The PROM-1 represents the greatest threat (in terms of numbers
of accidents). The smallest fragmentation mine, the POMZ, is
the most universal threat.
The ordnance threat (crudely defined as explosive devices that
are not classed as mines) in the various theatres can be summarised
as “fuzes”. There have been very few UXO accidents,
almost all of which have involved fuzes and been minor.
2 Reducing risk
There are two obvious ways to reduce the risk of injury in an
accident. The first is to avoid the accident. The second is
to provide effective protective equipment.
Avoiding risk can be achieved by revising operating procedures,
or by enforcing the application of safe operating procedures.
The DDIV records 272 injuries where the primary cause was a
management inadequacy at some level. This was generally either
the failure to provide appropriate equipment or training or
the failure to ensure that deminers worked as trained. Clearly,
improving controls at all levels could be a very effective way
of reducing accident numbers – and it would take another
paper to explain how I would approach that.
When everything has been done to avoid an accident, provision
must be made to protect against residual risk.
For brevity I only discuss protection needs in the four most
common accident types in the following. The other accident types
were rare and my conclusions are similar to those expressed
Excavation of a suspect mine is the most common accident activity.
This is an activity that must be carried out and accidents have
occurred when no “mistake” was attributed to the
victim. The “duty of care” of an employer requires
that a deminer be protected appropriately when working as directed
on a required task.
Missed-mine accidents are the second most common and indicate
that clearance has not been effective. Some time-served groups
have not had any missed-mine accidents: others have had many.
This implies that it is possible to work in a way that avoids
them. (Incidentally, there is no evidence that there is a greater
risk of missing a mine when demining in areas with minimum-metal
mines). The evidence in the DDIV suggests that the best defence
against the “missed-mine” risk is to avoid it by
using proven working methods that are adequately supervised.
Handling is the next most common accident and occurs when a
device or part of a device explodes in the hand. In several
accidents the victim did not recognise the risk, which could
have been avoided with appropriate training. Some groups seek
to avoid the risk altogether by not allowing devices to be handled
(these groups have still suffered these accidents). Practical
protection is impossible, so avoidance is the only way to reduce
Victim inattention is the next most common accident and covers
deminers behaving in a thoughtless manner. While close supervision
and rigorous training might have prevented some accidents, it
has to be accepted that moments of inattention occur. Given
that it is impossible to predict the nature of these accidents,
the only practical protection seems to be that used for other
The only kind of accident that occurs despite a deminer acting
properly in accordance with his training and instructions is
“excavation” – which is also the most common
accident. This is why I believe it should provide the benchmark
for protection needs (an opinion shared by many field workers).
Protection while excavating
To protect a deminer against accidents that occur when excavating
we must know the position he is in and the injuries he risks.
The data in the DDIV clearly illustrates that almost all deminers
excavate in a kneeling or squatting position (whatever their
SOP). This is good news for the deminer because he avoids the
whiplash acceleration injuries that might have been associated
with having his head within a few centimetres of a blast while
his body remained stationary. The exploding device is almost
invariably directly in front of and below his body and head.
Often his hand is above or alongside the device.
The following lists the number of severe (disabling) injuries
recorded while excavating.
|Face & neck
The low number of lower limb injuries illustrates the way that
a fragment cone rises from a seat of initiation and its core
often misses the legs (minor leg injuries were more common -
36). The low number of trunk/body injuries illustrates how the
main torso is not threatened as much as the upper limbs and
the head. (Half of the severe body injuries were caused by parts
of the victim’s hand-tool.)
Face and neck protection
Despite the fact that some form of eye protection was issued,
it was not worn in almost half the recorded blast accidents
and more than one in three victims suffered eye injury.
Eye protection issued varies from industrial safety spectacles
to 5mm polycarbonate visors. Visors made from 5mm polycarbonate
have been used by the most responsible groups for some years
and their use is spreading (MAG, HALO Trust, NPA, MgM, Koch
MineSafe, MineTech, INAROEE, most Bosnian and Croatian groups,
etc). Some of these are short and attach to helmets –
usually leaving the wearer’s throat exposed. Others are
long and worn without helmets. These offer some protection to
the throat when kneeling and looking down.
I have tested 5mm polycarbonate visors in over 40 tests using
mines. In one test, the visor was penetrated by a steel fragment
in the earth covering the mine. In several further tests against
POMZ fragmentation mines, the visor was not penetrated at all.
One 5mm visor broke in two in a recorded incident. These facts
illustrate the unpredictability of mines, but also shows that
even 5mm polycarbonate does not guarantee protection to a deminer
excavating an AP blast mine. It is, however, light enough for
sustained wear (thousands of deminers do so) and is the best
available option until a lighter, stronger material is developed.
The evidence suggests that 5mm polycarbonate visors that are
fixed in the “down” position should be provided
for deminers excavating AP blast mines.
Upper limb protection
The DDIV records 51 severe upper-limb injuries from blast mine
detonations (including 14 amputations of fingers and hands,
and 10 of arms). The injuries were worst when the tool was short
and used vertically. When the tool broke up, deminers were struck
in the chest, upper arm and face. At least five deminers died
after their hand-tool fragmented in a blast.
The DDIV also provides evidence that tools which stay in one
piece do not injure the user.
Demining hand-tools should be designed so that they:
• are easiest to use a low angle to the ground;
• stay in one piece;
• are long enough to keep the user’s hand at
least 30cm from the blast;
• incorporate a flexible blast shield whenever possible
without reducing utility.
Examples of such tools exist and are available on the commercial
Body protection against fragmentation
Protection designed to reach a STANAG V50 of 450m/s (current
UN standard) has proved less than adequate against bounding
fragmentation mines. Deminers who let one off at close quarters
invariably die even when wearing protection. However, bounding
fragmentation mine incidents occur rarely outside Europe and
there are no records of a bounding fragmentation mine incident
having occurred while excavating (although I have anecdotal
evidence of one such incident in Kuwait). Protection against
the close quarter detonation of a bounding fragmentation mine
would involve such a weight of body armour that it is not practical,
but the use of an angled steel shield when setting charges on
such devices might be practical.
3.4 Body protection against blast
In the recorded excavation accidents where body armour was worn,
it did not fail, illustrating the fact that the current STANAG
450m/s standard of body protection is sufficient (or more than
sufficient) against the largest AP blast-mine threat.
But a fragmentation V50 of 450m/s is no measure of blast protection.
Blast is a significantly different threat and the materials
used to protect against it may lack fragmentation resistance
while being highly effective against blast.
In an attempt to use more practical armour, there has been a
general move away from flak-jackets to the use of frontal “aprons”.
Some have a V50 as low as 380m/s, others over 500m/s. The only
sort to have failed in my tests had the higher V50 but was made
up of discrete panels that the blast separated. A one-piece
apron with a 380m/s V50 has performed well in 7 tests and in
at least 15 real incidents.
The evidence shows that the need for body protection may not
be a high priority, but it is desirable, especially when it
is comfortable enough for a deminer to wear. Simple, frontal
blast aprons have proven capable of protecting an excavating
deminer – and are comfortable enough to be worn without
The evidence suggests that deminers should be issued with frontal
blast protection (240g TNT at 30cm) for use when excavating.
No protection because no proven risk
There are a number of products available that offer protection
against risks that the facts suggest are not real. There is,
for example, no evidence of over-pressure internal injuries
from any AP blast mine. There is also no evidence to suggest
that blast-proof boots would have significantly reduced injury
(most occurred with mines far larger than those used in “successful”
boot trials). There is no evidence that wearing a helmet or
an armour back-panel has ever significantly reduced injury.
Protection against hearing damage is sometimes suggested. There
were many claims of hearing damage in Afghanistan during a period
when compensation was paid for small, unverifiable hearing loss.
Excluding Afghanistan, there is only one claim of severe hearing
damage resulting from a single blast in the DDIV (the claimant
was in close proximity to a very large ambush device). The risk
is, at worst, very low.
Wrapping it up
As I see it, there are two practical approaches to meeting deminer
protection needs. These are:
Reducing the number of accidents that occur.
b) Reducing the severity of injury when an accident occurs.
The first can be pursued via changes to working methods and
improved supervision – and is likely to have most effect.
The second can be pursued via the provision of PPE appropriate
for use at times when risk cannot be avoided.
The practical personal protective equipment I recommend is:
• Fixed eye protection with a blast performance and
fragmentation protection equal to that offered by untreated
• Hand-tools that are fit for purpose and that are
designed to minimise the risk of adding to injury.
• Comfortable frontal blast protection (against 240g
TNT at 30cm/12”) for use when excavating. The inclusion
of a collar that overlaps the visor and closes any access
to the throat from below is desirable.
Some groups already do most of the above. A few have done so
for many years. This provides evidence that my suggestions are
practical, and the DDIV provides evidence that they are needed.