Andy Smith
MIne-action specialist

Example risk assessment

What follows is a risk assessment that I made on behalf of a demining NGO for a research group. The research group were not staffed by experienced deminers and so wanted all risks outlined as clearly as possible, with strategies for risk reduction as detailed as possible. It was not appropriate to make "requirements" over the manner in which the research was to be conducted, so changes to the group's actual operating procedure were limited to suggestions rather than requirements.

April 2003, AVS Consultants

Risk Assessment for APOPO/MgM live-area tests

MgM is co-operating with APOPO in its assessment of the ability of trained pouched-rats to detect buried explosive items. To facilitate this, MgM previously prepared extensive Safe-test areas for the rats in an area near to the mine-belts but where there were no live devices. The vegetation in the area was cut and the area searched, then test-boxes were marked out and safe devices with an explosive content were concealed in November 2002.

APOPO have used the safe-test areas extensively and recorded the performance data to their satisfaction (not to mine).

APOPO then asked MgM for the opportunity to assess the explosive detection abilities of its trained pouched-rats in as realistic a setting as possible. The purpose of the tests is for APOPO to assess the efficiency of the rats and the practicality of their deployment method in a live area.

To facilitate this, MgM have prepared test lanes that cut through the known mine-belts in the live mined-area. Safe lanes for the handlers to stand in as they drag the animals to and fro have been cleared. These test lanes are called Hazardous-test areas because they are known to contain live and functional mines. [This did not occur because the rats failed to pefrom in the Safe-test areas.]

The purpose of this document is to indicate to APOPO the level of hazard that is perceived in their testing so that APOPO staff are fully informed of any risks they may run. MgM will do everything they can to ensure that the risk is minimised, but cannot accept liability for any residual risk, whether predicted or otherwise.


THREAT Assessment

With large parts of the minefield already cleared, the known mines in the Hazard-test areas are the GYATA-64 and the POMZ-2 and POMZ-2M. The PP-Mi-Sr is another possible threat. There is no significant recorded ordnance threat, with small arms ammunition being the only UXO located to date. If other explosive devices are found the threat assessment may need to be reviewed.

The GYATA-64 is a large AP blast mine with an explosive content of 300g TNT. It is 106mm in diameter and 60mm high. If detonated by stepping on the pressure plate, the GYATA-64 will remove the foot and damage the lower leg extensively. Past demining accidents in Mozambique have involved three "lower-leg" and one "through-knee" amputations. No deminer deaths from stepping on a GYATA-64 are recorded. If the mine is detonated in or beside a hand, amputation should be expected. Eye damage may occur if within 15 metres of a detonation without using a protective visor. All eye damage can be permanent and severe. Ear damage could occur if within a similar distance, but is unlikely if more than 10 metres away. Ear-damage from blasts of this size at more than five meters is usually temporary and results in a very small permanent hearing loss.

The picture Below shows the metal content from a GYATA-64 alongside an example of that mine. The leaf-spring that holds up the firing-pin is on the left.

Assessment: slightly elevated risk
Hundreds of these mines have been found in MgM's clearance task, and almost all are believed to have been functional. There is occasional evidence that the exposed rubber tops of the mines have been degraded by sunlight, weather and/or bushfires. There is also evidence of internal corrosion to the thin steel leaf-spring and coil-spring that hold the firing-pin in place. Weight on the mine compresses the springs and allows the firing pin to line up with a hole and fire into the detonator. Corrosion of the leaf-spring and the coil-spring above it could reduce resistance and so make the mine more sensitive to pressure applied from above. MgM have successfully exposed and either disarmed or destroyed many hundreds of these mines and so my current judgement is that the condition of the mines does not significantly increase deminer risk during normal manual operations, but that disarming should not be attempted without good reason.


POMZ-2 and POMZ-2M

The POMZ-2 and POMZ-2M are stake mounted fragmentation mines with an explosive content of 75g TNT. The difference between them is in the way the fuze attaches, with the 2M model having a threaded hole for the fuze to screw into. In the POMZ-2 the fuze pushes into place and is more likely to be found separate from the mine body. Sometimes locally sand-cast, the cast-iron bodies vary in quality and they fragment into rice-sized grains or larger chunks.

The picture shows a POMZ-2M fresh from the store.

If detonated by removing the pin, a POMZ will spread fragments all around. The density of fragments being greater at the sides of the mine. Fragment speeds have been estimated at up to 600 metres a second, with very few fragments actually achieving that speed. Lacking direction and being an irregular shape, almost all fragments slow down quickly after leaving the mine. At 600 metres a second, a fragment could penetrate the standard body armours (offering protection from STANAG fragments moving at 450 metres a second) and kill. But there is no record of this having happened with POMZ anywhere in the world. There are records where light armour and a 5mm polycarbonate visor have stopped 95% of POMZ fragments at a one metre distance in tests.

Unprotected people have been seriously injured and killed by POMZ. There is evidence of this in the mine-belts where MgM works in Mozambique, with a partial human skeleton having already been located (and buried according to local custom) and another reported to be ahead of the clearance group.

If a POMZ detonates within five metres of a person, severe penetrating injuries to unprotected parts of their body may be expected. Armour penetration could occur and fatal injuries result. The visor is the weakest part of conventional body armour (providing protection from STANAG fragments moving at around 240 metres a second) and may fail, resulting in severe eye or head injury. Ear damage could occur at up to 15 metres distance but is unlikely if more than 5 metres away. Ear-damage from blasts of this size is usually temporary and results in a very small permanent hearing loss.

Assessment: Low risk
Hundreds of POMZ mines have been found in MgM's clearance task, and almost all are believed to have been non-functional due to fuze corrosion. None have had more than a few centimetres of heavily rusted intact tripwire attached. None have been found with an intact wooden stake. Some are found with the roll of TNT having fallen out of the cast casing. Some have been found partly buried by silt or leaf-litter, others are lying on the surface. All the POMZ fuzes have been members of the MUV family, with a very few having a stainless steel casing, the rest an aluminium alloy that is heavily corroded. The extent of the common corrosion has meant that the pin has fused into the shaft of the striker and the striker and spring are corroded to the fuze casing. Except when the fuzes have stainless steel casings, these mines are considered safe to move cautiously for later destruction. Only the Chief of Demining and the Chief of Operations are normally permitted to move them.

The picture shows an old POMZ-2, dug up some twenty years after it was placed.



Two PP-MI-SR mines have been found recently in the mine-belts. None had been found in the belts up to and including the area where APOPO Hazard-Test lanes have been prepared. The PP-MI-SR mines were found lying on their side on the surface. They have either MUV or bakelite RO-1 style fuzes. One of the mines found had been fitted with two fuzes, apparently in error.

The PP-MI-SR mines are of the bounding fragmentation type. Usually placed with the lower half in the ground and the top of the mine protruding, those found in Mozambique usually have RO-1 style fuzes. RO-8 pressure fuzes can be used with the mine, but none have been reported in Mozambique. With an RO-1 fuze, initiation is by pulling a tripwire, which pulls the pin from the fuze, allowing the spring loaded striker to fire into the detonator. When the mine is initiated, a small charge in its base throws the body of the mine into the air. The body of the mine is secured to the base with a wire cable. When the wire cable is pulled tight (usually at 60-100 cms) the main charge of approximately 310g TNT in the body of the mine detonates. The thick steel of the mine body shatters, spreading hundreds of metal fragments in a 360 radius around the mine. The speed of the PP-MI-SR fragments has been estimated at up to 1200 metres a second, with very few fragments actually achieving that speed. Lacking direction and being irregular shapes, most fragments slow down quickly after leaving the mine. At 1200 metres a second, a fragment will penetrate the standard body armours (offering protection from STANAG fragments moving at 450 metres a second) and kill. This has occurred with many similar mines elsewhere in the world.

During demining in Mozambique, there have been five recorded accidents with bounding fragmentation mines (but none with the PP-MI-SR model) and four have resulted in fatalities. In the fifth it seems that the mine did not bound properly. Four other people were injured in these accidents, despite being some distance away. One deminer was killed by a single fragment that struck him in the neck at 31 metres from the detonation.

If any bounding fragmentation mine detonates within five meters of a person, that person will be at very high risk of suffering multiple penetrating injuries to head and body regardless of the protective equipment worn. At up to 50 metres, severe damage could be inflicted on unprotected areas. Ear damage could occur at up to 20 metres distance but is unlikely if more than 10 metres away.

The picture shows a PP-MI-SR with an MUV type fuze.

Assessment: Low risk
The undergrowth in the APOPO Hazard-test lanes was mechanically cut when the lanes were made and any PP-MI-SR mines on the surface would have been easily identified. None were found. Because the rats require undergrowth to be cut to ground level, the fresh undergrowth will be cut again before the tests are conducted, and a second visual search conducted. If any PP-MI-SR is visibly present, the test lane will not be used by APOPO and a new Risk Assessment may be made. Any PP-MI-SR present that is entirely concealed underground is unlikely to be initiated by the shallow scratching of a rat, so is considered to be a very low risk.



The Hazard-test lanes comprise part of a known mined-area alongside a railway line in the Limpopo river valley. MgM has been active in the area for two years and has cleared large sections of the defensive mine-belts present. Clearance is being done by manual deminers using metal-detectors, assisted when appropriate by a range of Mine Protected area-preparation machines and explosive detecting dogs (EDDs).

To date the mines located have been in two or three belts (varied) and have been GYATA-64 AP blast mines and POMZ-2 and 2M fragmentation mines. At the start of April, two PP-MI-SR bounding fragmentation mines were found. They were several hundred metres from the area prepared for use in the APOPO trials. Defensive mine-belts are sometimes "maintained" by the defending forces, and may be "patched" with other available mines - which seems to have happened here.

By cutting lanes through the mine-belts and leaving uncleared strips, MgM has prepared live Hazard-test lanes where they have a high degree of confidence that there are mines. The test lanes are approximately five metre wide strips, with safe areas all around. The uncleared areas are clearly marked with pickets and marking tape.

The Hazard-test lanes are clearly marked. There is resurgent undergrowth and the stumps of previously cut undergrowth in the lanes. Access is via cleared and QA checked safe areas. The area undulates slightly but is largely level. No further significant rainfall is expected this season.

Before the testing is conducted, the undergrowth will be cut again (using mechanical assets reaching into the area without driving over it).


Conditions of assessment

1) Should an employee or associate of APOPO suffer an accident anywhere within MgM's working area, MgM cannot and does not take any responsibility for the injuries, medical costs and suffering that may result. At all times, APOPO staff and associates operate at their own risk and on the understanding that they carry their own appropriate insurance.

2) When the undergrowth in the Hazard-test lanes is cut, the perimeter marking will be made good and a further set of pickets and tape shall be established outside all four sides of the test lane and at least 40cm from it. No APOPO personnel shall be allowed inside the outer marking. This is because the rat-handlers stand on both sides of the lane to control the movements of a rat across the hazardous area. While doing this they watch the progress of the rat intently. With their attention on the rat, they may place their feet under the perimeter tape (as has been observed in the Safe-test area). By placing a second perimeter marking 40cm back, any accidental movement across that marking should not be catastrophic.

3) The third APOPO staff member (who records the rats signals) must stand as far as possible from the rat, and at least five metres from either handler, moving as they move.

4) All APOPO staff in the mined-area must wear protective equipment and a visor that does not allow the wearer to see the ground without looking through the visor. Short, "dog-handler" visors may be sourced as an alternative to those currently used by APOPO. If this is done, APOPO should recognise the slightly increased risk to the lower face that this may represent.

5) No APOPO personnel shall approach the Hazard-test area without the express permission of the Chief of Operations or the Chief of Demining, and shall always be accompanied, obeying all safety conditions in force.

6) No APOPO personnel should approach the Hazard-test area without the presence of an MgM paramedic and ambulance at the site.

7) Any observers invited by APOPO (and approved by MgM) shall obey all safety rules and maintain safety distances as directed. APOPO should ensure that they are adequately insured. The method under trial requires three people to be inside what is usually considered to be the minimum working distance for a blast-mine threat. Their risk is minimised by the wearing of body armour and visors. All others must be at a distance not less than 25 metres from the working area.

8) At all times during any activity in the MgM-Task area that includes the Hazard-test lanes, MgM must remain in control and retain the right to halt all activity and withdraw persons from the field. If a person behaves in an unsafe manner and ignores safety directions, MgM reserves the right to permanently exclude that person from the Hazard-test area, the MgM-Task area, and the safety cordon around it.

9) The indications of the pouched-rats will be recorded but will not be made available to the deminers who will clear the area manually in the normal manner after the APOPO tests are completed. The way that they record their finds of fragments and "innocent" items may be changed to reflect APOPO's needs.

10) POMZ mines that are on the surface or have visible fuzes that appear to have stainless-steel sleeves should be removed before the APOPO tests begin, or the lane where they are should be removed from the APOPO test area. When a mine is removed from a test-lane, its original position will be recorded in the manner required by APOPO.

11) APOPO shall draft operating procedures, including basic safety features and procedures in the event of an accident, and present them for MgM and IND's approval or adjustment prior to the testing. Accident procedures must be practised before work begins in the Hazard-test area.



The conditions outlined in the previous section represent an attempt (agreed with MgM) to reduce risk to the bare minimum during the APOPO tests. However, there remains a small identified risk that results from the method under test. The residual risk results from the weight of the pouched-rats and the way that the rats indicate a "signal". They commonly (but not always) scratch the ground close to the concealed explosive. The force that a rat applies to the ground when doing this has not been measured and recorded. It is possible that a heavily corroded GYATA-64 could become very sensitive so that the pressure exerted by the rat detonated it. It is even possible that a rat's weight alone could be enough to initiate the mine (the rats are large and fat).

MgM cannot prevent this occurring.

If it does occurs, I anticipate that the rat would be killed and the three APOPO staff in close proximity may suffer some slight hearing loss (temporary). They may also suffer light injuries to their legs where they are unprotected by the armour apron. They may also suffer a state of shock.

A further slight risk arises if a rat digs the pin from the fuze of a POMZ or were to roll an exposed POMZ around. While most of the fuzes are inoperative, if the POMZ were to detonate there is a significant risk of injury to the APOPO staff and a small risk of severe or life-threatening injury. By removing exposed POMZ, MgM will do all it can to reduce this risk.

With MgM's agreement, I recommend that the APOPO handlers seek to minimise any scratching that a rat may make in any one area, if at all possible.

With MgM's agreement, I recommend that APOPO consider supplying their handlers and the data recorder with frontal leg protection to be worn as additional protection during the tests.

[Despite extensive preparation and every effort being made to accommodate APOPO's needs, the organisation decided not to conduct the trial in the hazard-areas. Having watched the rats fail to find the disarmed target mines in the practice/test area, I believe that the APOPO staff decided not to complete the trial because they would be obliged (under the terms of their contract with MgM) to publish the results if the trial were completed.]