The Joint TDG-GHS informal working group (IWG) dealing with the categorization of flammable gases was organized in Brussels from 9 to 11 March 2015 by the Belgian and the Japanese delegations to the TDG and GHS sub-committees. The participants (physically/by phone) list can be found in Annex 1of this report. The purpose of this IWG was to review the current criteria and discuss possible modifications to the GHS Flammable Gas Category 1 (extremely flammable gases). The participants were reminded of the mandate given to this IWG during the plenary sessions of the GHS and TDG sub-committees(1-12 December 2014, Geneva). The mandate can be found in Annex 2 of this report.
Mandate item 1(a): Analysis of the necessity to create GHS subcategories, within Category1, for flammable gases
Mandate Item 2 (a): Evaluation of the most appropriate additional parameters for modified classification criteria
Different presentations were given by governmental organizations and experts from industry and research institutes in order to give some background on the present situation and some recent evolutions:
The current criteria for flammable gases are based on the lower flammability limit and the flammability range. Transport only considers Flammable Gases Category 1, the others are considered by TDG as non-flammable. There are varying data regarding the flammable range of ammonia. In any case, ammonia and methyl bromide are considered as exceptions for some regulatory purposes because of historical reasons but arguably also fall under Category 1.
Currently, most flammable gases except some very specific mixtures, are classified as Category 1. Belgium and Japan identified a need to subdivide Category 1, rather than lump all flammable gases into one category for the reasons explained in the informal document -INF.10/Rev.1 (TDG, 46thsession) –INF.05/Rev.1(GHS28thsession).
From an industrial point of view, there is currently only one practical category of flammable gases. There is currently a practical inability to distinguish different levels of hazard within flammable gases. It would be useful for safety and knowledge reasons to have a distinction between higher and lower flammability gases. Even if such a distinction is made, all flammable gases should be labeled with a flame symbol and warning.
The Montreal Protocol of 1989 for the Protection of the Ozone Layer has led to the phasing out of ozone-depleting substances (e.g. CFC, HCFC, etc.) in a short period of time. Linked with the Climate Convention this caused the change from non-flammable high “global-warming potential”(GWP)gases to low GWP gases which are considered as mildly flammable. This has also led to the introduction of hydrocarbons, hydro-fluor-olefines and others, as solvents, refrigerants, blowing agents, etc. There is a tradeoff between the positive environmental impact and the flammability of these compounds. We need a new flammability index, similar to the GWP-index, ranking the flammability of the gases.
The current Flammable Gases Category 2 is virtually empty. With more subcategories, there can be a better consideration of the hazard. Maybe we should also think about revisiting Category 2, although it is not in the mandate.
Existing standards, such as ISO 5149, use a criterion based on the low flammability limit (LFL) 3.5 vol%, and ISO 5149-2014 and ISO 817 (2014) also use burning velocity of 10cm/s as a criterion: as well as the heat of combustion of 19 MJ/kg. ASHRAE 34 makes also a subcategory based on the LFL 100g/m³ and heat of combustion(HoC)19 MJ/kg, with a Subcategory 2L based on burning velocity, according to the measuring method described in ISO817. Those categorizations show that there was a need for subdivisions.
The IWG came to a principle agreement that there is a necessity to create an additional GHS subdivision within Category 1. This decision was based on:
- Safety considerations including the necessity to mark off reliable hazard areas for flammable gases and the necessity to provide hazard guidance for users of, for instance, blowing agents, solvents, cleaners and other process gases in hot and humid climates and high-temperature factory working environments,
- The reality of widespread adoption and further desirability of adoption of low GWP (but mildly flammable) gases to deal with climate change issues (Montreal Protocol/Kyoto Protocol) which arise with the currently used non-flammable gases.
The IWG further noted that:
- Additional data or testing should not be mandatory, any sub-categorization of gases should be optional to the producer/user, and the sub-categorization scheme should not be unnecessarily complicated.
- Specific cases, such as Ammonia and Methyl Bromide, which now attract a special treatment within GHS and TDG and are now held outside the standard category 1 framework should continue to be held outside.
- The necessity to create an additional sub-category is independent of any decision as to the specifics of that sub-category.•De-regulation in transport and unwanted downstream consequences must be avoided.
Mandate Item 2 (a): Evaluation of the most appropriate additional parameters for modified classification criteria
An overview was given by an expert about different flammability characteristics, prevention techniques, explosion/fire protection, sensitivity properties, and severity characteristics. The following parameters were retained as possible useful or relevant parameters as a basis for subcategorization:
- the flammability limits LFL/UFLand flammability range (UFL-LFL),
- the burning velocity,
- the minimum ignition energy (MIE)/the minimum ignition current (MIC)/the maximum experimental safe gap (MESG),
- the auto-ignition temperature AIT,
- the heat of combustion (HoC),
- the maximum explosion pressure (Pmax),
- the maximum rate of pressure rise Kg
BAM presented a proposal, based on lower flammability limits.
This was based on the idea that primary hazard identification should be a task of TDG-GHS sub-committees. Secondary and constructional flammability characteristics seem to be too detailed for use in a classification and labeling system. The determination methods of the flammability limits are well established:
- tube method,
- bomb method or
- glass flask
The international determination methods are sufficiently accurate and there is also the ISO 10156:2010 method available. There is a calculation method available for determining the flammability of gas mixtures and it can be extended with the Le Chatelier’s equation to calculate the LFL for gas mixtures (even though there are some issues in application for halogenated compounds). The LFL values of pure gases (see Table 2 in ISO 10156) and also LFL values of refrigerants are known (see ASHRAE 34-2013).
Proposal 1:
- subcategory 1A LFL <5%,
- subcategory 1B 5% < LFL <13% or flammable range >12%
- cat 2 : LFL > 13% and flammable range < 12%
Proposal 2:
- cat 1: LFL < 5%,
- cat 2: LFL > 5%,
In proposal 1, only LFL would be used for subdividing Category 1, and all typical fuel gases will bein subcategory 1A, and most halogenated gases will become 1B.
Proposal 2 would make more use of Category 2, which is now nearly empty.
The Chilworth study on “GHS Category of flammable gases: review and proposed modification” was presented. It was concluded that the fundamental burning velocity (BV) is an intrinsic property that takes into account both the likelihood and consequence of the flammability hazard. This study proposes a subcategory for flammability based on BV as follows:
hazard Subcategory 1B would include gases in Subcategory 1A with a BV < 10 cm/s with a modification of the hazard statement to H221-(“flammable gas”).
Mr. Scott Davis gave a presentation on the laminar or fundamental burning velocity (FBV). FBV helps to evaluate the likelihood and the consequences of a burning reaction and is essentially the reaction speed of this burning reaction. The combustion of hydrocarbons consists of chain reactions and halogens will stop these reactions and decrease reactivity. The transient state (linked to the speed of the reaction) is important, therefore FBV is an important parameter. Burning velocity is an intrinsic fundamental parameter. There is a good ranking in FBV of different gases, such as ethylene, methane, refrigerants, ammonia. FBV is also an important parameter to assess the turbulent flame velocity. NFPA has adopted FBV as the metric to determine safety venting in flammable gas environments. Testing FBV is feasible across the world with demonstrated testing methods. The total risk is the product of the likelihood and the consequence, and FBV is directly linked to both.
CONCLUSION on Item 2 of mandate subject (a)
Approximately 10 different parameters were brought forward in the IWG. Among them there was widespread support for the extension of the use of LFL and the Flammable Range (FR), and the use of FBV. There was also mention of using AIT as a parameter to assure that any pyrophoric gases are reverted to Subcategory 1a extremely flammable gases.