TNO-PML: Operation Obong


Operation Obong: Mustard gas destruction in Indonesia


The cause

In 1975, the Dutch government learned from a former employee of the former Ministry of Colonies that there could be a stockpile of mustard gas at an artillery firing range near the village of Batujajar near the Javanese city of Bandung. In the years 1940-1941, this stockpile was synthesised and stored in underground tanks by the Dutch Artillerie-Inrichtingen (A.I.) for the benefit of the KNIL. With the outbreak of the Second World War and the Indonesian struggle for independence that followed, these tanks were completely forgotten.

Chemical laboratory Artillerie Inrichtingen, Bandung (photo courtesy of web site Artillerie Inrichtingen Hembrug)
Chemical laboratory Artillerie Inrichtingen, Bandung (photo courtesy of web site Artillerie Inrichtingen Hembrug)

Subsequently, the Indonesian government was alerted. The Indonesian Army’s NBC service conducted an on-site investigation in 1977. Indeed, five underground storage tanks with a capacity of approximately ten cubic meters each were discovered. Four tanks were still in good condition and appeared to contain mustard gas. The fifth tank was in poor condition and contained no mustard gas. The remains of the A.I. chemical laboratory were also found in the soil.

When the Japanese invaded Indonesia in 1941, most of the inventory of the chemical laboratory was buried in the ground. This included a number of 50-litre high-pressure chlorine gas cylinders. These were needed for the production of mustard gas. That discovery immediately explained the mysterious events that had taken place on the site since the early 1970s. Occasionally an explosion was heard. Around the place of the explosion, vegetation and animal life appeared to have been destroyed over a distance of 50 to 100 meters. The chlorine gas cylinders had rusted through after decades of storage.

Indonesia asked the Dutch authorities for assistance in destroying the mustard gas found. The Netherlands responded positively to this. At the beginning of 1978, TNO-PML expertise was called in. An orientation trip was made in April 1978. Based on the findings and reporting, the Ministry of Defence commissioned on June 9, 1978:

  • Elaborating the research results regarding the orientation trip into recommendations for the most responsible destruction measures.
  • Preparing, and, if necessary, supervising work intended to purify the investigated area of chemical laboratory residues and cylinders.
  • Providing active cooperation in the preparation and implementation of measures regarding the clearance of the found stockpiles.

Preparations and planning

A project team was formed within TNO-PML with employees with specific expertise. They had to be prepared to possibly go to Indonesia for a longer period of time for this top-secret operation. Because of the classification, it was forbidden to talk about the operation with colleagues under penalty of dismissal; even their own partners and children were not allowed to know about it. The project team consisted of eight people and a liaison between the deployed team and TNO. From the assignment in June 1978 until the end of the project a year later, the PML team worked almost full-time on this project.

First of all, the team investigated what the best way was to destroy the mustard gas and what equipment would be needed.
It was decided to burn the mustard gas in two ovens at such a high temperature that a complete combustion would occur. The combustion would be done by vaporising oil through a nozzle using a pump and spraying it into the oven. Once the oven was up to temperature, the oil would be replaced with mustard gas.

The Indonesian word for combustion led to the project name: Obong.

On the artillery firing range, a suitable furnace with a chimney would be located at a location approximately 4.5 km from the military buildings and the site of the storage tanks. The flue gases from the chimney, including the corrosive gases sulphur dioxide and hydrochloric acid gas, would not be purified. The remaining gases would be discharged directly into the atmosphere. According to the plan, the flue gas concentrations would be continuously sampled in the chimney. For safety reasons, detection measurements would also be carried out regularly in a circle of four kilometres around the combustion site.

Transport of the mustard gas from the storage tanks to the incinerator would be done with two tankers. The contents of the first tank would be burned, while the second tank would be filled from a storage tank using a pumping unit. Domestic heating oil would be used to fuel the oven and maintain the correct temperature.
In addition to many calculations, tables and graphs, the action plan also contained a large number of work and safety procedures, extensive overviews of equipment that would be used and a time schedule.

The planning assumed working in three shifts of eight hours, seven days a week. Taking into account setbacks, the total period of combustion was estimated at 20 days. In retrospect, this turned out to be far too optimistic: there were 36. The activities plan was approved by the Indonesian and Dutch governments in October 1978. It was also decided at that time that the NBC department of the Indonesian Army would provide extensive support such as a mobile crane for the construction of the installation, provision of oil storage tanks and the necessary oil, water supply, showers, disinfectants, levelling and improvements of the roads on the site, and so on.

In addition, security, medical care in general and in the event of mustard gas disorders in particular, means of transport, housing and care of the members of the Dutch team. Ultimately, it turned out to involve the deployment of almost 200 soldiers. Moreover, it was further agreed that three officers from Pusziad NUBIKA (NBC) would be added to the Dutch team in the context of their NBC training and education and as a liaison.

In October 1978, the project team started purchasing a lot of equipment and becoming familiar with it. Safety equipment had to be ordered in large numbers. Training had to be done, procedures had to be tightened, work instructions for the Indonesian team that would help had to be drawn up, etc. A portacabin was set up from which the entire installation and all detectors, safety valves and other peripheral equipment could be operated. All this had to be done under strict secrecy. All this happened under the eyes of TNO-PML colleagues who were not allowed to know what the project was about, why they had to make a certain device or why they had to answer all kinds of strange questions.
In February 1979, the team at the Central Technical Institute TNO in Apeldoorn, the builders of the ovens and accessories, followed an intensive training program, after which the equipment could be packed and shipped.

Transport and activities in Indonesia

Coincidentally, the Royal Dutch Navy supply ship “Poolster” (A835) was precisely would leave for Indonesia during this period as part of a small Dutch squadron. The TNO equipment (over 100 m3 with a weight of almost 23 tons) could be transported by this ship. The equipment arrived in Yogyakarta on April 28, 1979. Construction of the equipment started on May 2. It was planned in advance that this would take approximately three weeks. The estimate was right. On May 18, the rest of the team arrived. Construction and training continued until May 26.
The first experiences in terms of weather, physical strain and safety led to the decision to burn the mustard gas only in daylight. In concrete terms, this meant that work would be performed from 6 a.m. to 6 p.m. During the night, the ovens would be kept at the right temperature by switching from mustard gas to oil combustion. Every night, a team member was present at the incineration site. The next day he was free to rest.

On May 28, 1979, the first tanker was filled and driven to the oven. However, just before the mustard gas valve was to be opened, one of the two electricity generators broke down. It turned out that some mounting bolts had trembled loose. In principle, the installation could run on only one generator, but this was not considered responsible. After purchasing a new generator on the same day, it turned out that the second generator also had loose bolts. So a second generator was purchased and the start of burning was postponed until June 1.
That day, more than 800 litres of mustard gas were burned. Measurements of the air quality in the area showed no detectable concentrations of harmful substances. During all days of burning, a low, non-harmful concentration of sulphur dioxide was detected only once.
In the period up to June 23, an amount of 800 to 1,600 litres of mustard gas was burned per day.

However, many different problems occurred. After just a few days, it became apparent that the mustard gas pumps sometimes jammed and started running more slowly or not at all. Such a contaminated pump had to be replaced with a spare one. A job of more than an hour in a hazmat suit, mask and gloves. Then the pump ran again for a few days. Disinfection and overhaul of the broken=down pumps turned out to be quite easy. Also, occasionally a nozzle became clogged, resulting in mustard gas dripping from the oven door. With some skills and effort, a small oil burner was placed in front of the hole in the door (to keep the oven at the right temperature) and the nozzle was cleaned.
There also appeared to be water in the mustard gas, which settled to the bottom of the transport tank. When this water was sucked in and ended up in the nozzle, all alarms went off because the flame spontaneously extinguished. After that, the tanks were no longer emptied to the bottom but were refilled every day. Through trial and error, three of the four old storage tanks were emptied in this way.

On June 24, mustard gas was pumped from storage tank 4, the last storage tank, for the first time. A gauge glass was used to check the liquid level in the transport tank. It turned out that this mustard gas was a deep dark brown in colour whereas the mustard gas from the first three tanks had been light yellow. During that day, the smoke from the chimney, which until then had been white, turned red. The pressure of the mustard gas also decreased.
The next day both ovens broke down and there turned out to be a brown cake near the nozzle. Analysis showed that the mustard gas from the fourth tank contained very large amounts of iron. This iron was the source of all misery. With a lot of puzzling, trying, replacing and improvising, mustard gas was burned until July 3. However, the process could take weeks at the rate the burning was going.

On 4 July it was therefore decided that this approach could no longer be continued. The remaining mustard gas, approximately 4,000 litres, would have to be destroyed in another way. All equipment was turned off. The remaining mustard gas was ultimately destroyed by the Indonesian NBC unit by mixing it with caustic soda and hydrolysing it using air pumps for a long time.

The project ended with a ceremony on July 7, 1979. The team left for Jakarta, where the members were received at the Dutch embassy by HRH Prince Bernard, who was in Indonesia for work. After a 7-day holiday in Java and Bali, offered by the Indonesian government, the team left for the Netherlands on July 17.

The aftermath

A few years later, Operation Obong made it into the press. Team leader Van Zelm was able to tell the world about this successful operation. Together with Indonesia, the Netherlands drew up a document and sent it to the chemical disarmament conference in Geneva. That document played a role in drawing up the section of the Chemical Weapons Convention on ‘Old and abandoned chemical weapons‘. The Netherlands had cleaned up its old chemical weapons. It was up to other countries to do the same.

  • Indonesia and the Netherlands Working Paper (1982) Destruction of about 45 tonnes of `mustard’ agent at Batujajar, West Java, Indonesia. CD 270, Conference on Disarmament, Geneva.
  • Old Chemical Weapons and Abandoned Chemical Weapons, United Nations
  • PML Vaarwel.