VdS-Schadenverhuetung Technische Pruefstelle

Test methods and practical fire tests

Design of an oxygen reduction system

Unlike automatic fire extinguishing systems, which react to fires that have already broken out, an active fire prevention system actively extinguishes fires from the outset by reducing the oxygen content in the area to be protected.

Text: Heike Siefkes, VdS Schadenverhütung GmbH, Product Group Manager for Gas and Special Extinguishing Systems in Product Management at the Technical Inspection Services

The fire prevention system essentially consists of the nitrogen generator, the oxygen sensors and a control and regulation device. Sensors continuously measure the oxygen content in the protected area. The nitrogen supply is interrupted via the evaluation in the control and regulation device when a defined oxygen concentration is reached. This is only restarted when the oxygen concentration rises again, for example due to removal or storage processes.

The respective residual oxygen concentration required for the flammable material, the so-called ignition limit, must be determined in tests. The flammability of a flammable substance is directly related to the oxygen concentration in the ambient air and this decreases as the oxygen concentration decreases. If the oxygen content in the ambient air is reduced, considerably more energy is required to ignite a combustible. The energy required for ignition is higher than that required to maintain a fire. Lowering the oxygen concentration therefore significantly slows down the speed of the chemical and physical processes involved in a fire.

Test methods

There is a constant debate as to which is the most suitable test method for determining the ignition limits. Approaches such as the test method according to ISO 4589-2 “Determination of burning behavior by the oxygen index” were quickly discarded.

Test scenarios with solid and fluid combustibles based on the internationally accepted test methods of ISO 14520 have proven to be suitable for determining the corresponding ignition limits. In addition to VdS 3527, this approach can also be found in the European standard for design, installation, planning and maintenance EN 16750 and other national guidelines in various countries. These tests have proven to be practicable and reproducible.

The test scenarios are not 1:1 tests, but reflect the actual storage configurations, i.e. including packaging and, if necessary, storage aids such as pallets. The environmental conditions of the real situation in the protected area are also taken into account in these test scenarios. For this purpose, the atmosphere in the test room is set to the appropriate temperature.

With the publication of test results by FM Global “Evaluation of Oxygen Reduction Systems (ORS) in Large Scale Fire Tests”, there are repeated discussions about the extent to which the test methods used to date actually cover a corresponding risk. The reason for this is the residual oxygen concentrations listed in the FM Global technical report, some of which deviate from the values already published in various guidelines and standards.

This raises the question of how the different concentrations can occur with supposedly identical materials. When comparing the test scenarios of VdS and FM Global, the first thing that stands out is the test size itself. Even though FM Global carried out the tests for storage risks with cardboard packaging on a large scale, this is not the cause of the differing results for the necessary oxygen concentrations. A closer look at the tests carried out shows that the approach and the associated framework conditions are different, which led to the different results.

Terms and definitions

Starting with the definitions, the first differences are recognizable. The definition of an oxygen reduction system (ORS) in the FM Global plant is a fire protection system that reduces or prevents the risk of fire development and spread due to the oxygen-reduced atmosphere. The concept described in the VdS guidelines, on the other hand, describes the framework conditions that prevent the spread of fires.

Another point when analyzing the test report is the result itself and how it is used for further interpretation in an oxygen reduction system. The terms used here are different and are nevertheless compared with each other. According to FM, the decisive parameter for oxygen reduction systems is the Limiting Oxygen Concentration (LOC), defined as the lowest O2 concentration that can support combustion for a particular fuel.

VdS, on the other hand, uses the terms design concentration (CA) and ignition limits (EZG), whereby the design concentration is the ignition limit minus a safety factor. The ignition limit itself is defined as follows: Maximum oxygen concentration in a mixture of a flammable substance with air and nitrogen in which ignition does not occur, determined under specified test conditions.

The LOC and the EZG are therefore similar in terms of definitions. However, the decisive difference in the result is based on the VdS definition of “determined under defined test conditions”. The test conditions and measurement criteria differ greatly from the conditions in the test scenarios and therefore the results are also different.

General conditions

In addition to the differences in the definitions, the framework conditions of the tests carried out are also different and therefore need to be scrutinized.

The selected experiments with

  • Cardboard packaging (class 3, CUP and CEP) and
  • unpackaged materials (UUP and UEP)

only show applications for warehouses with pallet storage. Other applications are not described.

The basis for the tests themselves are test set-ups in accordance with the protocol of FM Approvals, “Approval Standard for Automatic Control Mode Sprinklers for Fire Protection”.

The identical test setups and ignition methods from sprinkler system test protocols are intended to provide a direct comparison of oxygen reduction systems and sprinkler systems, but do not do justice to the specific and different nature of fire development in an oxygen-reduced atmosphere.

Unterschiede zu Sprinkleranlagen

With sprinkler systems, it is necessary for a certain amount of fire development to take place after the ignitions, otherwise the sprinklers cannot be activated.

An oxygen reduction system is always active. The constantly oxygen-reduced atmosphere is therefore effective from the first spark. If the oxygen content in the atmosphere is reduced, considerably more energy is also required to ignite the fuel.

For this reason, it is unlikely that a full fire can occur, as is assumed in the tests with sprinkler systems.

Choice of ignition source

The selected ignition source and the ignition duration must therefore be viewed critically. A propane burner with an output of 33 kW is used as the ignition source. It is intended to simulate an external heat source such as an electric arc or hot work.

This condition with the high heat release rate and the corresponding temperatures is typical of a full fire and therefore does not represent a fire in a room that is protected by an oxygen reduction system.

Period of occupancy

Another point is the fire duration. If the burner is switched off after 55 seconds in the tests, which is comparable to the burning time of 60 seconds according to VdS 3527, the fire is extinguished in a short time and no further fire spread occurs. The heat release rate drops from 250 to 24 kW. The FM tests thus prove that in an oxygen-reduced atmosphere, fire spread is extinguished without an external fire source. With a pre-burning time of 210 seconds, on the other hand, the maximum heat release rate is 870 kW, which is equivalent to a full fire, and remains above 600 kW for more than 2 minutes, which would not correspond to extinguishing.

Supply of oxygen and nitrogen

Another critical point within the test scenario is the oxygen-nitrogen supply rate. The selected 5-fold rate of the stoichiometric value also does not represent a room protected by an oxygen reduction system in which the oxygen concentration in the protected room is constantly low. By supplying this air mixture directly under the fire source, the oxygen concentration near the ignition point or flames is higher compared to the rest of the room. (Effect of “blowing into the fire”). This also contradicts the real situation of the self-sustaining fire: even in an “infinite” store with a reduced oxygen atmosphere, the air supply is not constant, but develops depending on the fire behavior.

Series of tests with cardboard packaging

Despite the different conditions, the LOCs determined in the test for the various types of packaging are compared with the design concentrations determined by VdS for individual materials that were not listed in the VdS guidelines. For this reason, VdS carried out a series of tests with cardboard packaging.

Various cardboard boxes with different thicknesses and flute pitches were used in the test series. The cardboard boxes were filled with plastic cups to rule out any influence of the storage material itself. The plastic cups were “European Standard Cups” made of unfoamed polystyrene (PS), which are also used in other internationally recognized test scenarios.

The ignition limits for cardboard packaging were confirmed in these tests. The series of tests - similar in design to the FM tests - which were intended to simulate the “chimney effect” also confirmed the concentration of the ignition limit. However, the damage pattern was significantly greater than in the one-sided ignition in the first series of tests with the identical cardboard packaging.


The LOC for cardboard packaging in the FM tests (Class 3, CUP and CEP) with a continuous fire load was determined to be 11.1% by volume and is therefore identical to the extinguishing concentration according to the VdS 2380 planning and installation guidelines for gas extinguishing systems. This confirms that the test scenario with the fire load and the extinguishing concentration does not correspond to realistic fire behavior in an oxygen-reduced atmosphere, as the fire must be extinguished and is not suppressed in the development phase.

If one compares the test results for cardboard (Class 3, CUP and CEP and UUP and UEP) when the flame is switched off, these are similar to values determined according to the test protocol for ignition limits for oxygen reduction systems from VdS.

In addition to the ignition limits for cardboard, LOCs for PPMA, polyethylene, methanol and ethanol are also listed in the FM Global report. Looking at the concentration for plastics and foams, the design concentration at VdS is even lower than based on the FM tests.


In summary, it can be said that the tests described in the report by FM Global and their results are not comparable with published values from standards and guidelines. In particular, since the test setup and the implementation with parameters such as the oxygen supply and the combustion duration do not correspond to those in an area protected by an oxygen reduction system.

Both the theoretical consideration of combustion processes in an oxygen-reduced atmosphere and a large number of tests carried out under real, comparable conditions show that the development or spread of fire can be prevented with a residual oxygen concentration of around 15% by volume. Similar to the determination of the extinguishing gas concentration, no 1:1 tests are necessary for this.

Rather, it is necessary to consider scaling and corresponding safety margins for smaller tests, as is the case with gas extinguishing systems and their test methods for determining the extinguishing gas concentration.

The residual oxygen concentration to be determined in tests and the corresponding safety factors are an essential component of a functioning system, but also require other parameters such as the reliability of the system components and the room conditions. Therefore, in addition to the tests to determine the ignition limits, the testing of the individual components - in particular the control unit and the oxygen sensors - as well as their interaction in the system is important for reliable fire protection in oxygen reduction systems.