Explosive atmospheres

On the 1st July 2003 the ATEX Directive of the European Union became a part of UK law. Whilst its provisions can, and do, af­fect all equipment used in hazardous areas, this Section con­centrates on its requirements for fans. They are, after all, the prime mover in ventilation and air conditioning systems.

Unfortunately there is a lot of ignorance as to its intentions and requirements. Even some fan manufacturers appear to be un­aware of its effect on them. Some have been heard to say that it only applies to the electric driving motor — NOT TRUE. Others appear to believe that it has no effect on materials of construc­tion, running clearances between stationary and rotating parts, bearing selection and so on. They couldn’t be more wrong!

Just as unfortunate has been the reaction of many designers and ventilation system users. You know the sort of response — “Oh heck, another piece of legislation to worry about. Put some­thing in the specification like must comply with the ATEX Direc­tive”.

YOU CAN’T DO IT!

The Directive requires that the customer and the manufacturer each carefully consider their response to the particular prob­lem. The main duties of the end user (who will no doubt ap­praise his plant designer) are:

• to prevent the formation of explosive atmospheres

• to make an assessment of explosion risks

• to categorise the work place area and divide it into appropri­ate zones

To select appropriate products according to the zone

подпись: to select appropriate products according to the zoneFigure 21.80 Positioning of slide dampers 352 FANS & VENTILATION

• to prepare an explosion protection document

• to identify hazardous areas and “sign-post" them with warn­ings

Having completed all this, he is only then in a position to ap­proach a fan supplier with an appropriate enquiry giving all the relevant information.

And here another set of problems raises its ugly head. The Eu­ropean Commission “mandated” CEN (Comitй Europйan de Normalisation) to produce Standards covering the require­ments for all types of machinery. Fans were assigned to a com­mittee (what else?) designated CEN/TC305/WG2/SC1, which comprised representatives from a number of European Stan­dards bodies including BSI, DIN, AFNOR, UNI and SIS. The author was privileged (?) to lead the UK delegation.

The need for a Standard

If a piece of equipment, such as a fan, has been manufactured to a relevant mandated Standard, then it is deemed to comply with the Directive. Engineers are much happier working with Standards than the Europeanised legalese in which the Direc­tive is couched! The committee set to work with a will and pro­duced a draft within the allotted timescale. There were argu­ments, some bitter, between DIN (Germany) and the rest of the nations represented, principally over clearances and material pairings between rotating and stationary parts. However, even­tually an amicable compromise was reached, and the draft sent upstream to its parent CEN/TC305/WG2.

Whilst apparently endorsing its technical content, the draft was rejected as it was not in the format approved for all ATEX man­dated standards. Evidently it was felt that it more closely resem­bled a standard produced for the Machinery Directive. It was then rewritten in the correct format, which gave a further oppor­tunity for delegates to re-raise objections that they might still have! The fan manufacturers would have preferred the “ma­chinery” format as it enabled easy comparisons to be made be­tween “standard” and “explosion proof fans.

Zone classification and fan categories

Users of flameproof electrical equipment will have been familiar for a number of years with the zone classification and gas grouping used to delineate the required features. These are mirrored in the draft fan Standard pr EN 14986 by three catego­ries suffixed with a G or D to identify a gas or dust mixed with the air.

It must be emphasised that the choice of category is ultimately the user’s, but in the absence of specific information, Table

21.10 would be used by fan manufacturers. The user may need to make an assessment based on his knowledge of the fan site. A fan sited in the middle of the Sahara desert, well away from other habitation, might not be the same risk as if it were sited close to a school, for example.

In Zone

Applicable category

If designed for

0

1G

Gas/air mixture or vapour/air mixture or mist/air mixture

1G or 2G

Gas/air mixture or vapour/air mixture or mist/air mixture

2

1G or 2G or 3G

Gas/air mixture or vapour/air mixture or mist/air mixture

20

1D

Dust/air mixture

21

1D or 2D

Dust/air mixture

22

1D or 2D or 3D

Dust/air mixture

It has been established that there shall never be more than one category difference (step lower) for the outside of a fan than for inside the fan casing and that for a ducted fan located in an un­ventilated room, the same category shall be applied for the out­side and the inside fo the fan casing

These provisos have been established to take account of situa­tions for example where the fan is handling a more dangerous gas but is located in a safer area.

PrEN 14986 — contents of this draft Standard

Fans for operation in all such atmospheres have to be of a rigid design. This requirement is considered as fulfilled for casings, support structures, guards, protective devices and other exter­nal parts, if the deformation resulting from a single impact test at the most vulnerable point is so small that the moving parts do not come into contact with the casing (see EN13463-1). All im­pellers, shafts, bearing, pulleys, cooling discs, etc, have to be positioned by positive locking devices.

Fan casings must be of substantially gas-tight construction (de­fined as category E from ISO 13349:1999 Table 4). A gas tight seal at the shaft entry will be necessary where there is a differ­ence between the gas inside and around the fan. The material pairings between stationary and rotating parts must be taken from a comprehensive list, (see Table 21.11) it being noted that not all plastics are necessarily flameproof.

It must be understood that the pairings in the Table are “hedged” with a number of requirements and footnotes. It is es­sential to refer to prEN 14986 for full details. Impellers have to be rigid with all calculated stresses less than 2/3 of the yield stress.

Item

Material (1)

Material (2)

Category

3

2 and 1

1

Leaded brass CuZn39Pb

Carbon or stainless steel or cast iron

Yes

Yes

2

Copper

Carbon or stainless steel or cast iron

Yes

Yes

3

Tin

Carbon or stainless steel or cast iron

Yes

Yes

4

Aluminium alloy

Aluminium alloy

Yes

Yes

5

Aluminium alloy

Naval brass CuZn39Sn

Yes

Yes

6

Aluminium alloy

Leaded brass CuZnPb3/CuZn39Pb

Yes

Yes

7

Nickel based alloy

Nickel based alloy

Yes

Yes

8

Stainless steel

Stainless steel

Yes

Yes

9

Any other steel alloy or cast iron

Any other steel alloy or cast iron

Yes

Yes

10

Any steel alloy

CuZn37

Yes

No

11

Plastic

Plastic

Yes

Yes

12

Plastic

Naval brass CuZn39Sn

Yes

Yes

13

Plastic

Aluminium alloy

Yes

Yes

14

Plastic

Nickel based alloy or nickel based steel alloy

Yes

Yes

15

Plastic

Leaded brass CuZnPb3

Yes

Yes

16

Plastic

Any steel alloy or cast iron

Yes

Yes

17

Plastic

Stainless steel

Yes

Yes

18

Rubber

Any steel alloy or cast iron

Yes

Yes

19

Rubber coated metal

Rubber coated metal

Yes

‘ ‘

Yes

Table 21.11 Permissible material pairings for gas explosion groups 11Aand 11B

Category 3 fans must be designed for easy inspection and cleaning. There must be a clearance of 1 % of the possible con­tact diameters between rotating and stationary parts with a min­

Explosive atmospheres

100

подпись: 100

0 25 50 75 100

Volumetric Flowrate % FfO max

Figure 21.82 Influence of clearance at inlet on the pressure against flowrate curve of a typical backward bladed centrifugal fan

подпись: 0 25 50 75 100
volumetric flowrate % ffo max
figure 21.82 influence of clearance at inlet on the pressure against flowrate curve of a typical backward bladed centrifugal fan

* 76 E

#

©

3

M

Ј 50 ы.

O

Ra

To

® 25

подпись: * 76 e
#
©
3
m
ј 50 ы.
o
ra
to
® 25
Explosive atmospheres

0 25 50 75 100

подпись: 0 25 50 75 100

Volumetric Flowrate % FIO max

подпись: volumetric flowrate % fio max

Figure 21.81 Influence of tip clearance on the pressure against flowrate curve of a typical axial flow fan

подпись: figure 21.81 influence of tip clearance on the pressure against flowrate curve of a typical axial flow fan

*

©

3

Vi

M

Ј SO

0l

O

«

55

® 25

подпись: *
©
3
vi
m
ј so
0l
o
«
55
® 25

Imum of 3 mm and a maximum of 20 mm. This requirement does not of course apply to shaft seals where the rubbing speed is very low. Maximum temperature of surfaces shall be less than 75% of the gas ignition temperature. The Li0 bearing life shall be greater than 20,000 hours. Direct drive is preferred, al­though belt drives may be used with appropriate precautions. These fans may be self-certified by the manufacturer.

Category 2 fans are generally similar to Category 3 but L-|0 bearing lives have to be greater than 40,000 hours. Belt drives are not allowed. Casings must be continuously welded with gaskets at all openings and splits. Design documentation, risk assessment data etc. must be deposited (in a sealed envelope) with a Notified Body. This will be opened and used in evidence in the even of any accident etc.

Category 1G fans, typically used for atmospheres containing hydrogen or acetylene, are generally similar to Category 2 but also require flame arresters on the inlet and outlet or internal ar­resters. These have to be tested and witnessed by a third party (Notified Body). Gas tightness also has to be witness tested.

Where the fan is also handling dust, i. e. Category 1D there will also have to be apparatus for frequent automatic cleaning of ar­resters. An alternative is to fit quick acting leakproof dampers on the inlet and outlet (closing in microseconds).

Electric motors in the airstream are not acceptable in any Cate­gory 1 fan, but pneumatic or hydraulic motors may be possible with appropriate safeguards.

It need hardly be added that whilst a Category 3 fan may have only a relatively small price premium compared with a "stan­dard" fan, that for Category 2 and Category 1 fans will be con­siderable. But then the penalties for non-compliance can also be considerable!

Clearances between rotating and stationary parts

Of the above requirements, perhaps that specifying the clear­ance between the running and stationary parts has created the most anguish. Some manufacturers have chosen to ignore it, whilst others perhaps do not appreciate its significance.

The effects of such increased clearances are especially severe on the performance of axial flow fans. These depend, for the development of high pressures, on a minimum tip gap between the blade periphery and the circular casing. The magnitude of this degradation in performance is very much dependent on the blade design. So called forced-vortex blades (which have a large blade chord at the tip when compared with the chord at the hub) are severely affected. Free vortex blades (which have

100 minimum chord at the tip and a very large chord at the hub with considerable twist) are less affected. For a given duty, they are however usually larger in any case.

However, the majority of axial flow fans have arbitrary vortex blades. Figure 21.81 shows the typical performance of such fans with different tip gaps. It will be instantly appreciated how important this factor is. A normal tip gap for a “standard” fan would be 0.2 to 0.3%. For a fan complying with ATEX it is 1 %.

It might be thought that centrifugal fans would be unaffected. This is certainly not the case, especially with high efficiency fans where there is an overlap between the inlet cone and the impeller shroud. The dimension of the gap between the lip of the venturi inlet cone and the impeller shroud is again critical to minimize recirculation and thus maintain efficiency. In this case Figure 21.82 shows the typical degradation according to the ori­entation of the clearance.

Actions required by manufacturers and users

So can we ignore the ATEX Directive until the finally approved standard is published? The answer is very definitely NO. Ever since 1st July 2003 we have had to comply. Most of the Notified Bodies have taken the stance that the last published version of the draft represents the agreed state of the art as known to the reputable manufacturers. It should therefore be followed. In any event we have the non-specific EN 13463 Standards to which all products must comply in the absence of a product specific standard.

Purchasers of explosion proof fans are recommended to obtain the position statement on ATEX of the UK Fan Manufacturers Association (FMA). They should also consider carefully the purchase of such fans and make sure that they are not laying themselves open to future trouble by ignoring both their duties and the manufacturer’s obligations. Such fans should not be chosen simply on first cost considerations — its too dangerous on so many counts. Perhaps they should ask what features the fan manufacturer has included, what tip gaps his performance is based on and what material combinations he has used. If he is evasive, hesitant or says these questions are not important, be very wary.

The draft standard prEN14986 has at the time of writing been in the public domain for over 6 months, during which time com­ments were invited. It perhaps demonstrates the widespread interest generated to note that over 50 such letters were re­ceived by the committee. They reflected the fact that there are quite a number of manufacturers with an interest in this market. There was little commonality in design for explosion protection
before work on this Standard started, a general reluctance to change existing designs, a scepticism about some of the scien­tific justification for particular requirements, and some who do not appreciate the need for the Standard to follow the logic of the Directive, and the basic requirements standard for non­electrical equipment, EN 13463 part 1.

Considerable comment the scope was received. Some felt there was confusion that the Standard tried to address only the mechanical parts of fans, and yet almost all fans were sold or used with a motor. There was also felt to be confusion that elec­tric motors often had integral fans that might not meet the re­quirements of this Standard.

There was pressure not to exclude Group IIC gases from its scope, and this was done, but it was realised later that there was not the technical basis for designing a safe fan for this ex­panded scope, and the requirements for the small number of gases in Group IIC might not always be the same. So there are now references later to requirements specific to hydrogen ser­vice.

There were also discussions about the working temperature range. The wide range in the enquiry draft arises from align­ment with mining service, where a 60°C ambient is possible. However, many types of electric motor are only available off the shelf with an ambient range up to 40°C. This conflict will proba­bly need to be resolved by marking many fans with the actual in­tended range of service temperatures and may require X mark­ing of a large number of fans.

Probable changes to prEN 14986

It is apparent that in the past many manufacturers had no data about the leakage rates either from joints in the casing, or from around the fan shaft. The draft Standard now will provide es­sentially 3 options:

• If the casing is not designed or tested to be leak tight, the fan should have the same category inside as out.

• If casing construction requirements are met, with effectively sealed joints, there can be a one category difference be­tween inside and out.

• If the fan is additionally tested, and the user is provided with measured information about the leakage rate around the shaft seal, then there can be a 2 category difference be­tween inside and out. This only applies of course to fans with a ducted inlet and outlet.

The section on gas temperatures has been reworded, mainly with the intention that is clearer.

Material pairings, predictably again, generated much discus­sion, with some strong views about what was or was not accept­able. To some degree this reflects a belief that in the past appro­priate material pairings were the principle means by which a standard fan was modified for an ATEX type application. As now drafted, that is nor longer the case; many other features also being relevant.

To some degree the difference of opinion arises from fragmen­tary information; there are no complete data sets about the risk, for all material pairings, for all gases, and probably never will be. Even if all the experimental work was done, it would have to be recognised that there is no sharp line between safe and un­safe.

The consequence was that the committee chairman rightly de­clared that the table in the public enquiry draft was already a compromise; and no progress would be made if a debate were reopened on all issues. So changes are limited. The use of alu­minium inlet cones was however definitely “banned”.

Footnotes to the table regarding paint were split, to prevent alu­minium being used where rust could be present, and also to prevent iron oxide in the paint, where aluminium was used for construction. For stainless steel, Germany presented data showing the chrome content could be safely reduced from 18.5% to 18.1 % and this was accepted.

A note about different plastics not being automatically permissi­ble was deleted, but it was not normative, and should not create new problems.

The reference to a test in IS01210 was changed to making it an option rather than a requirement.

A proposal to provide different minimum thicknesses for linings where these are used for category 2 and category 3 fans was accepted.

For category 1 fans, the German Notified Bodies have certified some small designs for vapour recovery plants, particularly in petrol handling facilities. Safety is based on a very strong, rigid casing; flame arrestors on the inlet and outlet, and full testing.

They proposed to remove a requirement for vibration monitor­ing. The UK argued strongly against this, as vibration monitor­ing seems a good way of identifying incipient faults, before a fan fails in a way that creates an ignition risk. The counter argument is that with flame arrestors and an explosion resistant casing, full testing shows the fan remains safe, even if the impellerdoes disintegrate. The compromise text agreed now says the manu­facturer must make a special ignition hazard assessment con­cerning this aspect, for category 1 fans.

Vibration monitoring is retained as a requirement for category 2D fans, because of the risk of dust deposits on moving parts.

Markings

This generated strong views, about how the category number inside and out should be identified. Some wanted a scheme that says (for instance) cat 2 (I) and cat 3(o).

However, the EU Commission guide, currently under revision, has an elaborate extension to the current marking scheme, and would in this case mean the category would be 2/3. Should CEN automatically fall in line with the Commission? Probably CENELEC would not be willing to do this. This is a power strug­gle, and the outcome cannot be predicted in the market and the standards committees. The author’s own view is that you can­not squeeze all the necessary information for safe installation into any sensible marking scheme, and when it comes to explo­sion safety, people really must read the instructions.

Annexes

Annex A will make clear the distinction between routine tests and type tests.

Conclusions

It would seem that all discussion is now at an end and finality has been reached. We can look forward to the publication of EN14986 in the near future. It must be emphasised that there will be some important changes from prEN 14986 the most re­cent document in the public domain. Whilst the Sections above give the flavour of the discussions, the copyright naturally re­sides with CEN. Interested parties are therefore strongly rec­ommended to purchase EN14986 as soon as it is published.

Posted in Fans Ventilation A Practical Guide