Landmines in Bosnia-Herzegovina and Croatia

James Trevelyan August 1999

This report combines my experiences in both Bosnia-Herzegovina and Croatia. While there are many differences, the similarities and common factors are so strong that it would be repetitious to write two separate articles about them.

To understand the mine problem in these countries, it is first necessary to know something of the terrain and recent history. Both contribute to a unique situation and make this region quite unlike any other mine-affected region in the world. The cost of mine clearance in dollar terms could exceed the entire cost of clearing mines in all other parts of the world.

To make an effective contribution to solutions also requires some understanding of both the current political realities in these countries and their recent history.

Terrain

Diagram: Map (400 kb) of region showing physical features of interest in military operations. Most rivers have swampy areas on either side, and many minefields were laid in and around these swamps. (full-size map in new window)

The northern third of this region is either flat or gently rolling agricultural land of exceptional fertility, with good rainfall and interspersed by large rivers (Sava, Drava and Danube), and occasional hilly areas covered in forests. The soils are fine clay and silt mixtures.

The middle third is hilly, with forests on the limestone hills and fertile river valleys. Fine clay soils are thinner. In both these regions the winters are very cold and snowy, with little work possible from late November till March.

The southern third is more mountainous, with craggy, rocky limestone peaks, steep gorges, and somewhat drier valleys in the south eastern coastal region. The soil is stony and much shallower. The climate is warmer, being close to the sea, with demining work possible through the winter.

The coast is rocky, with few beaches, and numerous rocky islands. It is very attractive for summer tourists, and has been extensively developed with hotels, bathing and boating facilities. Road access is relatively difficult though spectacular gorges, with numerous tunnels and high viaducts.

In the summer, vegetation growth is extremely rapid, particularly in the northern regions. Warm weather, rain and fertile soils can support up to 30cm or more growth per week.

Historical notes relevant to mine problems

An economic collapse in the late 1980s nourished regional, ethnic and religious differences that led to the disintegration of Yugoslavia in 1990 and 1991. Slovenia was the first region to declare itself an independent state, leaving the Serbs with overwhelming political power in the remaining parts. Croatia followed Slovenia, but there were large numbers of ethnic Serbs in several districts. As violence spread, irregular militias were formed for self-defence, using arms stolen from Yugoslav army stores and improvised weapons. The Serbs withdrew what was left of the Yugoslav army from Croatia and Bosnia-Herzegovina, but many military assets fell into the hands of the militias, and many Serbians in the army joined Serb militias defending Serb majority areas in Croatia.

Fighting broke out between Croatian and Serb militias in Western Slavonia in 1991. The militias concentrated along confrontation lines surrounding Serb population pockets, and landmines were used as a cheap way to consolidate defences by most groups. The Serbs were forced to withdraw to the South of Western Slavonia, where they were backed up by Serb districts in the north west of Bosnia-Herzegovina.

Diagram: Map of region showing major towns and rivers. (full-size map in new window)

As units on both sides forced people from their homes, many houses were blown up or burnt (after looting) and some were mined to deter people from trying to return.

After Serb units advanced around Vukovar and Osijek in Eastern Slavonia, the fighting bogged down, and confrontation lines were extensively mined throughout Croatia and Bosnia-Herzegovina. In the latter region, Muslims had formed their own militias and fought both Serb and Croatian units, again along extended confrontation lines, with all groups using mines to deter surprise attacks.

In 1995, the Croatian army mounted a surprise attack on Serb-held areas with unofficial outside assistance. In a few days they had secured almost the entire country and huge numbers of refugees fled to Serbia. After NATO air intervention in Bosnia, all parties signed the Dayton peace accords. Among the conditions was that all the armies would provide details of minefields they laid, and would proceed to remove their mines. In 1996 and 1997, the entity armies, assisted by I-FOR, S-FOR and United Nations units from many countries, carried out "mine lifting" and mine clearance operations in known mine fields, mainly along the former confrontation lines.

The entity armies, in particular suffered high casualty rates in carrying out these directives. Since all sides inherited the same military doctrine from the former Yugoslav Army, tactical advantage relied on improvisation. One tactic was to lay additional mines in enemy minefields. As suppliers were always short, orders would often be issued to lift a minefield and re-lay the mines in another location, resulting in some unpleasant surprises for the mine lifters.

Part of mine contamination map showing central BH near Sarajevo. The minefield locations are predominantly concentrated along former confrontation lines. (full-size map in new window) However, many mine incidents occur away from the known minefields, so we know that the contaminated areas are more extensive than these maps indicate. Blue dots represent cleared minefields (current at August 1999).

A reference point is always needed to mark out and define a minefield. So another tactic was to lay mines around likely reference points, again causing unpleasant surprises for the side lifting the minefield later on.

Mine lifting relied on the memory of the minelayers or paper records if available. These records were often sketchy, prepared in haste, using reference marks that were sometimes destroyed in fighting. Therefore, mines remained behind in most declared minefields. The mine clearance records prepared by foreign military units are also unreliable in many cases. Records were made in the units’ own languages, eg. Sanskrit, Urdu, Arabic, Russian and others. They used a variety of maps with different grid systems, and the reliability of some maps has turned out to be questionable. It has been found that grid references were often read incorrectly.

Part of mine contamination map showing most of the city of Sarajevo. The minefields along the controntation lines are marked in red. Note that the yellow stars with black outlines are mine incident locations. There are several mine incident reports which are not near known minefields. (full-size map in new window)

Records of mine incidents involving civilians show many incidents where there are no recorded minefields. These are either due to UXO, or indicate unrecorded minefields. It is known that current records and databases are still incomplete. In Bosnia-Herzegovina, current estimates allow for 40% of mined areas being unknown. Mines were often laid in a hurry or records were lost. Mines were also used by non-military units and other groups not following military procedures.

Estimates of suspected mine contamination in Bosnia-Herzegovina run between 400 and 2000 square kilometres, and around 6000 — 12000 sq km in Croatia. Most of these areas contain relatively few mines, however.

Types of mine threats

The majority of mines being encountered by deminers seem to be fragmentation mines, especially stake mines. There are many bounding mines as well, especially the feared PROM-1 which has taken the lives of many deminers in accidents. AP blast mines are found near fragmentation mines: they were used to deter attempts to remove the visible (or easily detectable) fragmentation mines.

Anti tank mines were laid in areas where tank warfare was expected. I-FOR and S-FOR troops cleared the road network, but many AT mines remain wherever natural obstacles were considered insufficient.

The TMRP-6 mine in particular, with its shaped charge, is of particular concern because demining machines are particularly vulnerable to severe damage: the machines survive ordinary AT blast mine explosions. However, they were more expensive and relatively few seem to have been laid.

The shortage of munitions fostered improvisation: booby traps are commonly found using both mines and ordinary grenades or even commercial explosives.

Finally, none of these common patterns can be relied on by deminers. Mine incidents occur widely outside known minefield areas; mines were laid unconventionally, often by untrained people, sometimes under fire.

Demining Organisations

Demining in Croatia and B-H is organised differently to most countries. Most work is carried out by private companies, military units, civil defence workers and non-government organisations (NGO’s). The work is supervised and controlled by mine action centres: in BiH there is the BHMAC, under which the Federation MAC controls work in the Federation area and the RSMAC controls work in Republika Srbska. In Croatia, CROMAC controls all demining work. In Croatia, the United Nations has direct involvement in demining through UNMAPP which operates alongside CROMAC, helping to supervise the work and providing technical assistance. UNMAPP staff sit alongside CROMAC counterparts in regional offices, but for historic reasons, CROMAC head office sits apart in Sisak, about 45 minutes drive from Zagreb where UNMAPP head office is still located.

The situation in Croatia is complex, with three main sources of funding: direct government funded work, work funded by World Bank and other institutional loans, and more recently there have been projects funded by NGO’s, humanitarian agencies and direct bilateral assistance programs, often negotiated with the help of UNMAPP field staff. From 2000 onwards, UNMAPP has to raise its own operating costs from these programs because external UN funding has been diverted to Kosovo.

There is no direct participation of the military or police in demining work. However, many of the deminers were recruited from these agencies. MUNGOS is a semi-government demining company. Other companies are local private companies or local branches of international companies such as MECHEM (South Africa), ABC Construction (Italy), Ronco (USA), Tamar Consulting (France, Germany). Companies bid for work with fixed price contracts. The companies must have government approvals to operate as deminers (now issued by CROMAC).

In BiH the situation is even more complex, reflecting the complexities of the Dayton Accords which govern political and military arrangements in the former Yugoslavian republics. Elements of the entity armies are still involved in demining. With their deminers paid very low rates compared with private companies, their performance has been questioned by many people. At the time of my visit, deminers in the Republika Srbska army were on strike as they had not been paid for some time. The civil defence organisation (which performed EOD work before and during the war) is also performing demining work. Then there are private companies (as in Croatia): some local (eg Piper, Oktol) and others foreign (Ronoc, Mechem, Minetech etc.). There are NGO’s such as HELP (Germany) and Nowegian Peoples Aid that receive money from their parent organisations, and also bid for other demining contracts. Finally, UNHCR operate their own demining teams to work on refugee resettlement projects.

Although the BHMAC supervises demining work, it is not responsible for administering contracts: this is the responsibility of a special agency operating under the BiH Demining Commission.

The United States government provided substantial funding for demining in both countries, but has now formed a new agency: the Slovenian Trust Fund. CROMAC and BHMAC submit projects to the trust fund which will provide up to US$14 million each year, but only with matching funding from other sources. In Croatia this new arrangement is being treated with caution: several smaller contracts have been arranged and more will come if these operate well.

The complexities of these organisational structures are reflected in the roles of senior staff, especially in BiH. At the time of my visit there were obvious strains and uncertainties, with several ex-patriate staff feeling insecure. With Croatia now providing much of the leadership for its demining program, there is pressure for more local staff to take the lead in BiH. They have the necessary skills, but the current administrative structure retains ex-patriate staff in key roles (advisors in name, but executives in practice) which makes hand-over difficult.

Clearance Priorities

Although four years have passed since the end of fighting, most clearance work is dictated by short-term emergency priorities:

Survey

Survey work is restricted to what is called "Level 1" though there is debate about whether to increase level 2 capabilities.

A survey, therefore, is a process of gathering information that might eliminate the possibility of mines being present in a suspect area.

This term is widely used in demining, but the meaning varies according to time and place.

In Kosovo, HALO Trust is completing a country-wide "level 1" survey. They are collecting data on the likely locations of mines and suspect areas where there is a significant probability that mines may have been used.

Photos: hr1-07a: Level 1 Survey Team

In Croatia, I joined a group of people conducting a "level 1" survey. They group consisted of:

We looked at several destroyed houses in mainly deserted, former Serb villages, though two houses were located in a Croat village. In each case, we looked for signs, which would indicate explosive threats:

The survey party did not have to decide on the likelihood of mines or whether the houses would be checked with dogs. Thirty houses would cost about US$10,000 to check, including removing vegetation from the approaches and inside the ruined houses which was almost impenetrable in some instances.

Photo: hr1-15a: Demolished house designated for reconstruction (no. PR42). The front yard vegetation has been cleared, but no one knew who had risked mine injuries to do that. Notice the rusting domestic appliance near the front gate which was left behind by looters. All usable equipment was taken from the houses before destruction and left in front for others to collect.

After visually inspecting all the houses from the road (without stepping off the road), and interviewing any nearby residents, the team met at the council planning office to review the numerous 1:5000 maps of the district, mostly prepared in the Austrian-Hungarian period. Known minefields and confrontation lines had been marked on these. After some time, it was concluded that the mine threat was low, but some houses should be checked for UXO.

The socio-economic factor was important in this assessment. It would take about one year for dog team resources to become available for this work. However, the reconstruction funding would only be available if work commenced within six months. All the house owners (displaced persons or refugees) had satisfied the council planning officers and EU Reconstruction Commission of their need for reconstruction and bona-fide intentions to remain in their new homes after reconstruction. They could still have to wait for 18 months to two years before moving in. Population records are also checked during the process: if village populations were steadily declining in the two decades before the war, then resettlement and reconstruction might, in the end, be futile.

Before clearance starts, demining team leaders carry out further assessments that can reduce the size of the area requiring clearance.

Evidence of people using land for agriculture, or simply clearing vegetation, without mine incidents, reduces the apparent risk of mines. In some instances, this is taken as sufficient evidence to declare the area "safe". However, mines and UXO have been found in ploughed paddocks in other cases. There is anecdotal evidence that some demining teams mark any agricultural land being used as "cleared", but as procedures are being enforced more rigorously in both countries, this practice is less likely.

The use of level 2 technical survey is quite low in both countries. There are two reasons:

The vegetation and destruction of mines makes area reduction equivalent to clearance. All current detection techniques require complete vegetation removal (dogs, metal detectors, manual probing), and this can constitute 70% of the cost. Also, since there are few mines left in almost all cases, there is unlikely to be an identifiable "densely mined area" which can be fenced, marked or left for longer-term clearance.

For the time being, therefore, there seems to be little to be gained from level two survey and area reduction methods, according to mine action staff.

This issue is moderately controversial, however, and there are many different opinions. Perhaps the key to this question is the current 99.6% clearance criterion, which cannot be verified, and needs to be re-assessed according to Paddy Blagden, technical director of GICHD, who accepts responsibility for inventing it.

Demining Operations in the Region — Some Common Approaches

Although there are differences between demining operations in Bosnia Herzegovina and Croatia, there are many common approaches.

First, most demining is being performed by private or semi-private companies. In Bosnia Herzegovina army units also carry out demining. In BiH, several foreign funded NGO’s such as Norwegian Peoples Aid (NPA) and HELP (German) undertake demining work. Some NGO’s join companies in bidding for fixed price-demining contracts.

The main priorities for demining currently are houses and yards for returning refugees or displaced people and national infrastructure.

Vegetation poses a major problem in most demining operations. Planning is essential because of rapid growth in the summer months. The type of terrain is often classified as:

In Bosnia Herzegovina, demining contracts specify the respective areas of each terrain type in a given demining task. However, ground which is category A in March can turn into B or even C later in the year. Machines are being increasingly used to turn level and gently sloping category C ground back into category A for manual demining or dogs.

Differences

There are some significant differences between Bosnia and Herzegovina (BiH) and Croatia demining.

The first difference is that the Croatian government contributes a major part of the cost of demining in Croatia, currently about 40% of the total cost. The BiH government contributes very little in comparison, but has no revenue to support expenditure. Industries in BiH before the war were predominantly defence-related armament manufacture, and under the Dayton Accords, these industries are no longer permitted. Croatia, on the other hand, has a more commercially oriented economy with a large tourism industry along the southern coast.

The second difference is legal. Croatia has developed a legal framework for demining which allows the government, acting through the Croatian Mine Action Centre (CROMAC), to enforce minimum standards.

The legislation covers deminer registration and licensing, supervision, working hours, and provides legitimate authority for CROMAC in supervising the work of demining organisations.

While this is a commendable step, there are weaknesses which have to be recognized:

The third difference is in the cost of labour. In Croatia, there are social security taxes which amount to about 133% of a deminer's pay. These taxes are paid by the employer. One rationalisation for this is that it provides government revenue from the large contracts for reconstruction in the country. However, it means that a deminer costs about DeM 65,000 yearly in Croatia and about 27,000 in BiH. This has a major impact on demining costs.

The fourth difference is in the level of technical and scientific support. CROMAC works closely with ‘The Croatian Scientific Council for Demining’ which consists of about 30 academics who are supporting the demining effort with research and development or are providing other inputs. These people are, in turn, connected with foreign institutions who are providing some useful assistance (principally from Britain). There are several companies which are developing machinery and other equipment to help with demining. A good example of this is the MV2 remotely operated mini-flail, now being exported to other countries. Although Bosnia and Herzegovina has many academics and substantial engineering infrastructure, it seems that little of this potential has been harnessed to help with demining. One reason for this could be the very low level of government contributions to demining, and language barriers which have separated these potential assets from the demining community which is largely run by foreigners.

Methods

Each demining company has its own standard operating procedures (SOP's) and there are significant differences between them. SOP's have to be approved by the BHMAC or CROMAC in the respective countries. The following sections summarise the common techniques.

Manual probing

This is the most basic method. A deminer marks a 1 metre wide entry to a lane with a wooden index rod 1 metre long. He probes at 25mm intervals along the rod, using index marks painted on the rod as a guide. Then he rotates the rod, rolling it forwards 180° and repeats the probing across the 1m wide lane width. The index marks may be offset by 12mm (half spacing) on the underside of the rod.

Vegetation has to be cut carefully work advances. There may be trip wires or mines with projecting fuses. Deminers use hand shears (similar to those used for shearing wool off sheep) for grass as they can do the cutting carefully. Larger hedge clippers (used widely in Combodia, for instance) are quite dangerous.

Prodding alone cannot detect mines to the normally specified depth of 200mm. Most probes being used are 200-250 mm long and are inserted over than 10cm below the surface, even in very soft conditions. When soils in the effective probing depth.

The depth of probing can be increased by carefully stripping 5-10 cm of soil after probing and repeating the process. However, this significantly increases the cost (already high) and difficulty.

Thin whisker wires are used to check for trip wires. These are straight and inserted into the vegetation and lifted carefully. They are usually about 600mm long.

Break wires can be harder to detect. These are very fine wires made from brittle material, closing or opening an electric circuit. Some deminers have used simple aids like party streamers. The streamers catch on the near invisible wires.

Time works in favour of deminers, sometimes. Trip wires are often to have been pulled by animals or wind action.

Protective Equipment

There has been a big increase in the use of protective equipment between 1998 and 1999. Most deminers now wear ballistic helmets and Kevlar aprons which provide some protection for the head and torso from fragmentation mines. The visor is a weakness, but still provides protection from blast effects. All the demining organisations are wearing similar equipment.

Procedures

There is clearly a problem in conforming with SOP's in both countries. A spate of accidents in 1999 is causing deep concern in the MAC's because procedural faults are the cause almost every time. Since the only guarantee of demining quality is conformance with procedures, there are also concerns about demining quality. For example, the 25 metre safe working distance is often compromised. This has been evident in some recent accidents in B-H where two or more deminers have been killed or injured by powerful PROM mine explosions. A PROM can kill at a distance of 100 metres, so one might argue that 25 metres is not a safe distance. However, deminers seem to want company, believing (at the moment) that "there is safety in numbers" or "two sets of eyes are better than one". Deminers seem afraid to be alone.

Part of the reason for this may lie in the military procedures many deminers were first trained to follow. These required deminers to work shoulder to shoulder, with 10 or 20 deminers working alongside each other, prodding the ground ahead of them to breech a minefield.

However, there has been steady improvement, and with some retraining, it is thought that local deminers could equal the best in the world.

There are different opinions on this. Some expatriate observers are disappointed that these issues are not taken as seriously as they "ought" to be. However, I was impressed by the extent to which CROMAC is concerned and looking for answers which go beyond attempts to enforce military-style discipline. They are currently trying to improve recruitment and training for supervisors and team leaders. To do this, they are asking field deminers to tell them what makes a good leader, and by this they intend to improve the quality of team level supervision. It is said that one difficulty lies with the management of some companies who do not visit minefields, and are only concerned with making a profit on current contracts.

Some companies, in consultation with technical experts, adapt their SOP's quite well to local conditions without necessarily obtaining formal approval from the MAC's. They only revert to old procedures when MAC staff visit to check on procedural compliance.

Metal Detector + Probing

Practically all mines in the region contain enough metal to be detected to a depth of at least 15cm. Fragmentation mines and UXO can be detected much deeper.

Deminers who use metal detectors still need to be cutting vegetation unless they are working with mechanical assistance. Vegetation has to be removed to allow the deminer to scan the lane in front of the 1-m index rod.

There is some room for variation in SOP's. Depending on requirements, deminers may ignore deeper targets after investigating to the required depth of 20cm. Metal does not have to be removed from the area in most operations where dogs can be used for quality assurance. This differs from Combodia and Afghanistan where manual demining usually involves removing all metal so that QA can be done with metal detectors. Deminers may spot a fragment of metal on the surface, in which case they simply remove it and may discard it on the ground behind them. Deminers may collect fragments in piles or "frag pits" or leave them in the cleared lanes.

One company in Croatia uses "metal free" SOP's, meaning that all metal fragments detected are removed so quality control can be implemented by using a metal detector to re-check the cleared area. Other companies leave metal in the ground, but they are all required to "investigate" all metal detector readings. One or two companies in B-H also have metal free SOP’s but the rest leave metal in the ground most of the time and rely more on dogs for quality assurance checking.

Companies choose their own detectors. There have been trials to help guide equipment selection procedures, but the results seem to have been inconclusive. Many interesting weaknesses were exposed in detectors, but the weaknesses do not necessarily preclude companies from using them. For example, some detectors will not work well when they are wet, or are being used in wet grass.

There are anecdotes which tell of companies who claim to have used metal detectors in very unlikely situations: between the rails of a railway and around steel electricity pylons.

Rates of progress range from 30 to 60 metres per day (on a 1 metre wide lane) depending on vegetation. With 5 hours work, this corresponds to 6-12 metres per hour. In Afghanistan, comparable rates of progress range from 8 —13 meters per hour, depending on the fragment density which is usually 0.1 to 3 fragments per metre.

Metal detectors run into difficulties in three common situations:

• Soil has high minerals content causing a continuous indication from most detectors. The Australian minelab mine detector overcomes this, also recent offering from Schiebel and Vallon.
• Household refuses pollution-cans, bottle tops, cigarette papers/ wrappers, nails, screws, etc. This is often a problem when clearing around houses and adjacent yards.
• Shell, mine and grenade fragmentation, both primary and secondary. This problem can occur where there has been heavy fighting or bombardment. Because many minefields lie along former confrontation lines, this is a frequent problem, which can nullify the advantages of metal detectors over probing alone.

One demining site I visited was flat land with trees adjacent to a river. A Mig-21 jet had crashed at the site, exploding into thousands of Aluminium fragments with some larger pieces (engine undercarriage). Metal detectors were used on the site (though I did not see them in use) showing that fragmentation can be quite high before metal detects becomes useless.

Dogs

Dogs are being used across the region, though more seem to be needed.

Like other methods, the vegetation has to be removed if it is more than 200 — 300mm high or if it is prickly and too dense for the dogs to get their noses close enough to the ground.

In areas with dense tree canopy (forests), the ground vegetation may not be significant making it relatively easy for dogs.

Most groups I spoke with use the dogs to check square or rectangular areas of land 10m or 15m square (boxes) These areas are first defined by cleaning 1m safe lanes in a grid pattern using metal detectors and prodding. The vegetation may have been removed or mulched with a machine. The size of the boxes is determined by the length of the dog’s lead, and by the need to locate trip wires. Dogs can activate trip wires so they must be found by deminers before they let the dogs work in the boxes. When enough boxes have been prepared, a dog team may be brought in to check each of the boxes in turn.

UNHCR, for example, suggested that they may prepare 50-60 boxes for dogs to check in a single operation.

I did not see dogs in use, nor did I go into SOP’s for dogs in any detail.

I did see a test area prepared by the Italian owned Croatian Company ABC. This had been sat up 10 days before the dogs were first used at the site. Photographs

Boris Katic, Director of ABC, explained how he had inserted a series of targets into the ground with the deepest, hardest target to smell first. Targets become progressively easier as the dog progresses around the test area. The dog and handler are ignored by onlookers for some time, even if the dog correctly indicates a target to be sure that the dog is not simply "trying to please and play up to his audience". Only if the dog senses a difficult enough target is it allowed to work on clearing the adjacent mined areas.

Since I did not examine SOP's for dogs in detail, I cannot be sure if there are significant differences between different organizations. However, I had the impression that dogs are being used to locate more than one target in a "box" in Bosnia whereas in Croatia only one target is located by a dog in a box. If any target is found by a dog, (i.e. if the dog indicates by sitting) the whole box is cleared by normal mental detector and prodding methods.

Certainly, there were doubts at the time of my visit about the effectiveness of dogs in the minds of some. There are accounts of missed mines, a dog handler who was killed during a quality assurance check, and concerns that little action seems to have been taken to prevent this recurring. These doubts have increased since my visit. The BHMAC introduced a registration performance test for mine detecting dogs and one source indicated that less than 25% of the tested dogs passed before the winter weather closed down operations. No one seems to know what to do about land which has been "cleared" by dogs which subsequently fail performance tests.

It was summed up by one observer this way: "We don't know how much to trust the dogs, and the handlers agree. Managers often don't want to even think about it".

This is a major issue which concerns most people working with dogs in both countries.

Mechanical Assistance

While few people believe machines can clear mines and unexploded ordnance to make land completely safe, Machines have proved to be very effective for "ground preparation" before using dogs and manual clearance.

Several different machines are being used. Some have been purchased and others donated.

Croatia offers more scope with significantly higher labour costs (see separate notes).

I met several people with first hand experience of using machines and visited a site in Croatia where the Belarty flail machine (Bosnia, Slovakia) had been used.

The principal application of demining machinery is removing vegetation. However, the machines can also be used in different ways. Several people emphasized the importance of using an appropriate combination of methods for the condition:

- Machines (several different kinds)

- Dogs

- Manual demining (metal detection and /or probing)

Machinery is expensive to buy and needs maintenance by skilled people who are expensive to employ and retain. Skilled maintenance staff is being keenly sought by many companies, not just demining organizations.

It was also clear to me that companies operate on different financial and technical approaches. Some companies have high costs with purchased machinery, skilled people and comprehensive on-site support; others are using donated machinery for which they only pay maintenance costs. In the former case, much more management skill is needed to achieve profitable operation when both types of company are competing for similar contracts.

MV2 remote controlled robot flail, now becoming well used in Croatia. (full-size)	Mine rollers are being used, with caution. Note the chain between the rollers which catches the TMRP-6 mines. (full-size)	Rhino ground milling machine. There is enthusiasm for these machines in Croatia, but there is also some concern about an unexpected high number of missed mines. (full-size)
Reference for photographs: Technika Za Razminiranje (Technology for Demining), Dinko Mikulic, Hvratski Centar Za Razminiranje (HCR), Sisak, Croatia. 1999, 270 pp. ISBN 953-97854-0-5 (HCR).

There are many different opinions on the utility of different machines. While most people now accept the use of flails, some insist that the flail only be used to cut vegetation, leaving the ground untouched. Others are happy for the flail to beat the ground as well. Many people I met are strongly opposed to the use of ground milling equipment such as the Rhino because these machines have left mines behind during tests. CROMAC however are strongly in favour of using these machines in combination with manual demining teams and dogs to locate mines which have been missed by the machines.

Quality Assurance

Inspection of cleared areas after demining teams have finished their work is essential for the time being. There are anecdotes of mines being found after certain organisations have finished their work, of land being declared clear when no work has been done and other less likely results. The cost of QA inspection will be significant if 10 — 20% of the cleared area has to be checked. Currently CROMAC employ 7 deminers and 2 dogs for QA work which seems a very small group to check 30 sq km of outstanding unverified land.

Given this situation, there seems to be a pressing requirement to find ways of confirming the quality of demining work while it is being done, rather than relying on a QA check which may take months or years to be completed. Many industries have learned an important lesson in the last few decades: it is much more cost effective to ensure that work is done correctly the first time so subsequent QA checking is unnecessary.

There are many arguments and strong opinions on the quality of demining work being performed in both countries. Ex-patriate staff and foreign-funded NGO’s are often skeptical of work performed by local demining organisations, citing instances where seemingly enormous areas have been cleared by a handful of people in almost no time at all. Foreign companies allege that local companies and semi-government teams shelter behind corrupt government officials. NGO’s and some UN staff allege that the commercial operators are all cutting corners to make profits as competition tightens and profit margins fall to a few percent. Private companies allege that NGO’s are wasting money by being too cautious and are not managing their resources efficiently. Many of these opinions are strongly voiced. When one hears of accidents being caused by failures to follow procedures, and missed mines, one tends to give more credence to the critical comments. This seems to be reinforcing skepticism in funding agencies who are uncertain whether they are getting value for their money.

One response to this, in BiH, is to focus on productivity. A local workshop discussed this extensively in November 1998, but the results were inconclusive. This probably reflects the divergent aims of this analysis: on the one hand seeking more efficiency, but on the other hand seeking to highlight excessive clearance rates which might indicate that corners are being cut. A subsequent analysis of the data collected by the BHMAC (by Saad Soliman) has confirmed that the data supplied by different companies cannot be compared or related to a fixed standard. For example, many companies do not count fragments so one cannot relate the square metres cleared to the difficulty of the task.

This led us to the conclusion that the only way to resolve this issue is through direct measurement of demining quality by an independent method. This has now become a major research focus for our group.

Mine Awareness

Mine awareness programs are being run in both countries. A major television awareness campaign was due to go to air late in 1999 in B-H. The ICRC and other NGO’s have been very active in this aspect, especially in schools.

One issue which has caused some difficulties is the need for minefield marking along tourist routes in Croatia. Most minefields are not marked: the cost of maintaining fences and marking around known minefields would far exceed current funds available for mine clearance. It is normal for travellers to relieve themselves standing right on the edge of the sealed road surface because most people are aware of the potential danger lurking just off the sealed roads. (The road network was the first priority for clearance when I-FOR and S-FOR troops arrived, but this work was confined to the roadways, not the verges.)

Many people were concerned that marking minefields (which are numerous along popular tourist routes to the south coast) would deter tourism. Others argued that warnings were essential: one injured tourist could spell disaster for the industry. In the end, markings were laid, but with an unanticipated difficulty: tourists showed a frustrating tendency to take the signs and marking tape as souvenirs!

In Croatia there are a few pemanent road signs. However, one technical expert confessed that he had driven past a permanent road sign warning of mines every day for several months before he finally worked out what the sign meant!

Notes on Performance and Costs-Croatia

Reference: Katic, Boris

Director, ABC Construction - Italian owned demining company based in Croatia.

Belarty machine typically clears 2,000 sq. per day, if it is being carefully used. "Some people claim they can do 10,000 sq. per day, but I do not think they are being used carefully".

ABC was offered US$ 3,000 per day to lease the machine with full support. They decided to purchase the machine instead and have contracted the local Croatian Army tank workshop to undertake a complete overhaul and replacement (up grade) of the hydraulics system, which powers the flail.

He said ABC had cleared 500,000-sq.m., of which the Belarty machine had cleared 380,000-sq.m denoting about 40 AT mines. Just the overhaul would cost about DEM 110,000.The recommended policy was a complete overhaul after every 10 AT mine blasts. However they were still learning about maintenance requirements.

PROM mines had caused some damage. One fragment hit the sealing ring of a hydraulic cylinder, gouging the piston rod as well. The cylinder had to be replaced. Another fragment had penetrated 10mm steel plate.

Because the Belarty is mounted on a 755-tank base, it can easily be maintained in an East European Military workshop where spares and skilled technicians are readily available.

Taking these figures, we can calculate the following:

Belarty clears 380,000 sq.m

2,000 sq.m per day

190 days

21 days per month

assume 6 hrs/day 1,140 hours operating

Operators:

Driver

Mechanic (spare driver)

annual cost DeM 6500 each

4 labourers for pulling, burning vegetation and

4 Assisting manual deminers

annual cost DeM 40,000 each

Labour cost total: DeM 290,000

Overhaul cost

DeM 110,000

Total direct costs:

DeM 600,000

Assuming purchase cost is DeM 1,000,000, could add annual cost of DeM 200,000 to cover depreciation and interest cost on capital invested.

This brings total cost to about DeM 2 per square metre for mechanical assistance.

Deminer cost is about DeM 65,000 per year as follows:

Equipment, clothing: DeM 5,000

Salary: DeM 27,000 (about 3,000 per month for 8 months)

Tax, pension, social security DeM 33,000

Working on his own, a deminer can clear 40 sq.m per day average through 8 usable months (weather, holidays etc), 21 days per months (i, iii) reference to notes at end.

6,500 sq.m/ year

or DeM 10 per sq. metre.

Two dogs do about 2,000 - 3,000 sq.m /day at a cost of about 100,000 DeM per year each. Both dogs cover some area. Assume dogs cannot work everyday (weather, attitude etc) but they can keep up with machine clearance. (380,000 sq.m)

Cost Approx. DeM 0.50 per sq. metre.

Assume that manual demining is needed for boxes (approx. 20 sq. m per box)

And one in ten boxes - total 30% of area checked by dogs. At 2000-sq.m. per day, we need 600 sq.m. Of manual demining which requires 6 deminers - since all vegetation and trip wires have been removed. Assume too that mechanical labourers also help deminers with simple non-technical tasks such as vegetation removal and burning, marking etc.

Cost for 6 manual deminers @ 65000 DeM each

DeM 390,000

Add costs for paramedics, Other unskilled labour (non - deminers) transport, supervision, on- site accommodation:

DeM 600,000

Total estimated cost for 380,000 sq.m. Mechanically assisted demining with dogs:

Machinery - direct costs: 600,000

Capital cost 200,000

Dogs 200,000

Deminers 390,000

DeM 1,990,000 (approx. DeM 4.6/ sq.m)

Notes:

1. Deminers (manual) work for 5 hours per day total, typically working a 40 minute spell followed by 20 minutes rest. Total time per day is restricted by government regulations.
2. Katic placed great emphasis on the need to perform to high standards. He did not want to risk breaking law and regulations as he has too many stakes in the commercial venture.
3. Tile layers earn DeM 15/sq.m. And bricklayers 12/ sq.m.
4. If, for interest, recast these cost estimates with Bosnia- Herzegovina labour rates, we find the cost is reduced to about 2.9 DeM per square metre.

In Bosnia, deminers are paid about 1200 to 1800 DeM per month. Assuming a well run company will pay at the top rate, and government charges are 30% of salary instead of 130% Croatia, the annual cast of a deminer will be (8 months)

DeM 17300

Add allowance for equipment, clothing etc: DeM 5,000

Total: DeM 22300

Extrapolating this difference to the estimates above, the labour component reduces from about 1,320,000 to about 450,000. This reduces the total cost to about 1,120,000 reducing the cost per square metre to about 2.9 DeM.

This corresponds with rates being currently quoted in Bosnia Herzegovina.

5. Katic is experimenting with innovations to reduce is costs. One is to use conventional tank rollers attached to the Belarty machine to detonate AT mines. He plans to drive the machine forwards, then execute a 180-degree turn using the flail to clear enough turning space in which the mine roller will not be effective. The machine will then comeback alongside the original track marks.

Notes on Performance and Costs

Reference: Nurija Memic, Director, Oktol - Demining company based in Sarajevo, Bosnia Herzegovina.

Demining costs range from 3-25 DeM/ sq.m. deminers are paid 1800 DeM per month, and supervisors 2100.

His company operates as a policy of full and open disclosure with careful attention to all aspects of

With DeM 2,000,000 of his own capital invested in the company mostly in the form of retained profits, he cannot afford to take risks which would damage the company's reputation. So far, no mines or UXO have been found in land cleared by Oktol, and there have been no mine incidents reported after clearance. His company has cleared the second largest area of all companies operating in Bosnia Herzegovina.

Oktol estimate their price for demining contracts using the following approach:

From these estimates, staff prepares a cost estimate, which is reviewed at a meeting of all senior staff. Finally personally decided on the tender price after thinking through all the issues. Sometime he will allow other companies to "find out" what he may have submitted / as a tender price, but he will deliberately mislead the competition, to gain commercial advantage.

He never discloses the final contract price to his colleagues until after the contract is won (or lost!).

The company implements a training programme for all staff, which requires two days per month. Staff at all levels take part in training appropriate for their level and technical skills. All for top company staff have university degrees (mainly Mechanical Engineering). The company's priorities are:

• recruit top quality staff
• use best equipment available
• use best organization techniques available

He said that some companies lack highly trained staff and their performance reflects this.

He aims for a profit margin of 3-6% normally, but will quote at a zero profit margin for "strategic jobs" which will lead to further opportunities for contracts, or up to 16% when he is confident of winning the contract. He said that some companies have tried trimming costs to lift profit margin to 60%, but the have failed. This approach leads to accidents. Oktol has never lost money on a contract. He said that the staff at the BHMAC don't understand " productivity. He intends to show that productivity can be improved by 40% by improving training and organization.

On Improved Detectors

This is the biggest problem facing deminers: the current detectors find all metal, not just the mines. Asked whether a low false alarm rate detector would be attractive, Oktol staff were skeptical, and do not believe such a detector would be feasible. "Well if it really works that would solve our problem, but I can’t see it happening". They would like a detector which can show the shape of the metallic components. The common item in all the minimum metal mines is the alumium case around the booster charge: a cylinder 5 — 7 mm in diameter and length.

Notes on performance and costs

Reference: Per Kvarsvik, UN Mine Action Assistance Programme Osijek, Croatia.

The question was: what options are open to a demining company when realize that they have under - estimated the cost of a demining contract? How can they save costs?

Response:

The simplest option is to tell the deminers they have to work unpaid hours for part of the job. There are not many other job opportunities except possibly in the tourist season on the south coast (this has not been a good year). If a deminer refuses, there are others who will happily take this job.

The work can be speeded up to some extent through this may increase the risks for the deminers themselves. Sometime companies have declared land cleared when it was not been completed, but this can often be detected by visual inspection.

Another option is to "pull strings " to arrange for government resources (either cash or labour) to help out. But this is now more difficult than it was at the start.

In Croatia, the 99.6%clearance requirement is interpreted as follows (sometimes):

100% clearance of mines and UXO from 99.6% of the task area. Mines found afterwards obviously from part of the 0.4% allowable tolerance. Thus, in a task, which requires 100,000 sq. metres to be cleared, up to 400 sq. metres can be " Missed". That is a big margin.

Information Issues

Several issues raised by people I spoke with concern aspects of information technology.

First, there is a need for information systems which distribute data to field offices and which allow field information to be entered promptly into databases. Some regional managers told me about how they read about new demining operations in their own areas in the newspapers because their own data was so out of date. Data entry is slow and incomplete, and distribution of the data (particularly forward plans and schedules) is also slow.

Each organisation seems to be using different software, and while better software is available from various donors, the effort needed to translate existing data and fill in the missing data has to be met from internal resources.

There have been registration problems with maps. Different mapping authorities have used different projections with the result that the maps in computer databases do not match up with each other at the joins, nor do minefields appear in the correct locations on the maps. I saw maps of Croatia with Croatian minefields appearing beyond the borders of neighbouring countries! There is a need for software which can overcome these problems. Western military maps use UTM grid projections, and most GIS systems can accommodate this. However, good quality local maps are available on paper in Gauss Krieger projection which cannot easily be transformed to UTM. There are also humidity effects on the paper maps and scanning errors to take into account. Croatia (and presumably Bosnia as well) have excellent hand-drawn local maps (approx 1:5000 scale) produced at the time of the Austrio-Hungarian empire, but it is not currently possible to use these in databases.

| Demining Home Page | James Trevelyan's Home Page | UWA Home page |
Updated January 2000
For more information on any of this material look at our enquiries page
Webmaster: Edin Tabak All graphics by Demining Research Team. Copyright University of Western Australia 2000