Ex d – a non solution to a hot problem
Background
Very occasionally someone notices that the IEC apparatus standards usually only consider ambient temperature conditions of -20 °C to + 40 °C. This range can be extended when this is considered desirable and noted on the certificate and by marking. The usual limits are –40 °C and + 75 °C. Some concern has recently been raised as to whether measuring temperature outside this range using thermocouples [TCs] and resistance thermometers [RTDs] is adequately covered by the certification and/or is adequately safe.
Note: An interesting aside is that both the ATEX Directives apply only to gas/air mixtures ‘under atmospheric conditions’ so the Directives are not applicable and do not consider this problem
Some thermocouple installations are located at the interface between Zone 0 and Zone 1. In these circumstances the inside of the thermowell is usually a Zone 1 and the outside a Zone 0. Hence from a thermal ignition risk viewpoint it must be suitable for Zone 0
The intrinsically safe solution
It is known that ignition energy of a gas/air mixture decreases with temperature, although the available experimental data is limited. At the ignition temperature it is zero, and the way it changes between atmospheric temperature and that point is not well defined [a translation of that statement is that the author of this note does not know the answer since the limited information is partially contradictory]. Within the normal testing range for apparatus [up to +75 °C] it is usual to assume that the change is small and adequately accommodated by the large safety factors inherent in the intrinsic safety technique.
There are obvious thermal ignition problems where the process temperature approaches the ignition temperature of the gas and consequently the temperature rise of the measuring element under fault conditions must be restricted to as low a level as practicable.
Usually the TC or RTD is regarded as simple apparatus and it is the output parameters of the transmitter or IS interface, which determines the acceptability of the measurement. Fortunately most of the recently designed temperature converters, temperature alarms &c have very low safety output parameters, which make them useable with TCs and RTDs at higher temperatures. For example the MTL5074 temperature converter in TC mode has output parameters of Uo: 6.6V, Io: 76mA and Po:17mW. The voltage is well below the ignition threshold of any gas [10 -12V], and the current is too low to create an inductance problem. The limited power means that the possible rise in temperature of the TC under fault conditions is small.
It seems reasonable therefore that provided that the output parameters of the connected instrumentation are small then the measurement of temperatures approaching the ignition temperature of the gas/air mixture is acceptably safe. The output parameters of the connected device are in accordance with the certificate; hence it can be argued that the installation is partially ‘certified’. It is adequately safe to the normally accepted level and no problem exists.
The flameproof [Ex d] non-solution
The performance of flameproof enclosures at temperatures other than those for which they are tested is not well documented. The emphasis on the need for testing in the Ex d standards suggests that the number of interacting factors make the prediction of performance difficult. Most of the recently expressed concern relates to low temperatures and hence possibly high temperature effects are not so critical if metal enclosures are used. In the particular case of TCs and RTDs then the relatively small heat capacity [compared with a large metal box] and the possible intimate thermal contact with the process fluid calls into question the adequacy of the flameproof practice of allocating a temperature classification without considering faults.
If this argument is accepted then the use of flameproof thermocouple wells and heads is only marginally acceptable at atmospheric temperatures and the possibility of an unspecified temperature rise under fault conditions makes them unsuitable for use at elevated temperatures. The questionable performance of flameproof enclosures at temperatures lower than that for which they have been tested makes their use in these circumstances not acceptable.
Some thermocouple installations are located at the interface between Zone 0 and Zone 1. In these circumstances the inside of the thermowell is usually a Zone 1 and the outside a Zone 0. Hence from a thermal ignition risk viewpoint the installation must be suitable for Zone 0, which means that a flameproof installation is not acceptable.
The inevitable conclusion is that the majority of flameproof TC or RTD installations do not achieve the level of safety implicit in the IEC standards. Fortunately they do not have to comply with the ATEX Directives but they are probably not adequately safe. However there are no explosion incidents attributable to the use of this type of installation known to the author so perhaps there is no call for undue panic.
Conclusion
The only acceptable technique for the measurement of temperature in using TCs or RTDs in Zone 0 and Zone 1 hazardous areas is to use intrinsically safe equipment where the interconnected interface has output parameters which have a considerable safety factor over those permitted for gas/air mixtures at atmospheric temperatures. If this is done then the safety level achieved is not well defined, but it is covered by the well-known practice of ‘making it as safe as is practically possible’. However it is unlikely to be covered by the certification.
January 22, 2007 at 8:54 pm
I agree that most suppliers do not take into account the Thermocouple or RTD when they have their terminal heads certified we Okazaki have taken this into account an dhave considered all fault conditions an dhave had our range of temperature senors certified EExde which takes into account the two fault conditions EExd for the enclosure and E increased safty fo the senor so we can issue a cert without any thermowell
January 23, 2007 at 9:59 am
It is interesting and commendable that at least one other supplier has considered the problem. There remains the fundamental problem that area classification of a location which contains a flammable gas at temperatures above atmospheric is not covered by the IEC or CENELEC standards and hence a risk analysis is the only appropriate technique. It is reasonable to argue that gases at a high temperature are easier to ignite than those at ambient temperature and hence a higher than normal level of protection is required. I would argue that the only method of protection that meets this requirement is the enhanced ‘ia’ which was previously discussed and that this needs to be used even where the gas/ air mixture being monitored is not always within the flammable range. Personally I would not consider a thermocouple assessed only against the Exe standard to be adequately safe. It may be that whoever certified the device took into consideration other factors not included in the Exe standard and thus ensured that it is adequately safe. If this is so then the test report and installation manual will list the additional factors considered and would make interesting reading if it can be published.
On a more provocative note: It is possible that the IEC standard that suggests that a combination of Exd and Exe has got it wrong when considering thermal ignition. Exd apparatus is temperature tested in normal operation and this is accepted as being adequate in Zone 1 because the electrical protection will remove the fault in a relatively short time. If the Exe equipment inside an Exd enclosure within a Zone 0 develops a fault it seems questionable that the electrical protection will operate fast enough to prevent a possible explosion. The minimum requirement is that the electrical protection should be clearly specified.