DRAFT MINUTES OF THE BVAA/CEIR/VDMA MEETING - MACHINERY DIRECTIVE

Transcription

1 DRAFT MINUTES OF THE BVAA/CEIR/VDMA MEETING - MACHINERY DIRECTIVE 16 JULY 2015 Attendees (see signatures & business cards in Annex) Christophe Bochaton Martin Greenhalgh Thomas Kraus Alessandro Maggioni Duncan Pudney Hartmut Tembrink Pascal Vinzio Benoit Wagner PROFLUID BVAA VDMA ANIMA/AVR KOHLER MIRA VDMA CEIR (KSB) EMERSON Process Management Opening of the meeting A roll call of participants is organized. Since all participants know the situation and earlier developments, it is not necessary to remind in details the context elements. Then, it is proposed to VDMA representatives to present their position and the work that was produced. Presentation of VDMA Position by Mr. Tembrink (see Annex) Mr. Kraus first explains that even if the reversal of position from VDMA is quite a new story, the issue of industrial valves being within or outside the scope of MD is old in Germany and that debates have been open for a dozen years. Mr. Tembrink adds that the revision of VDMA papers was initiated due to increased pressure from third parties, end-users and valve manufacturers. Consequently, a Working Group was created, gathering representatives from valve industry, power industry (RWE), chemistry industry (BASF) and VDMA. Once the work was completed, it was presented during a seminar. 40 attendees from the Valve industry + 60 more were informed. Although some questions were raised, the German Valve industry agrees with the new VDMA guidelines. VDMA considers that actuated valves fit with the 3 criteria of the definition of machinery in the Directive. Therefore, manufacturers have to comply with the requirements. In particular, the paper states that valves do always have a specific application because the range of use is indicated to the user. The status of valves intended to be actuated is dependent on the knowledge of the actuator that will be used. Such valves can be machines or PCMs depending on the case. Page 1

2 Discussions The question of the specific application is the cornerstone of the different interpretations. VDMA considers that industrial valves do always have a specific application since they are designed to control or stop a flow. The other participants agree this doesn t make a specific application, but defines the basic function of the product. Multi-purpose products (such as catalog products) do not have a specific application. Examples taken from the Machinery Directive guidelines seem to be in line with this second interpretation. Mr. Kraus insisted on the fact these are only examples and that only the text from the Directive itself is binding. However, several participants agree that the Directive content is not obvious and that examples from the Machinery Directive guidelines are here to help. This remark is also valid for the example of solenoid valves, hydraulic cylinders and gear boxes which are namely cited as being out of the scope of the Directive. These products are relatively similar to valves. Mr Greenhalgh explained that the very fact this meeting is necessary clearly shows the difficulty to interpret the European legislation. However, back to the products themselves, the safety aspects can be divided in two: the pressure aspects in one side and the movement risks on the other side. Mr Greenhalgh considers that the Pressure Equipment Directive already covers these movements risks in its Essential Requirements. Then, in virtue of article 3 of Machinery Directive, only the PED would apply. Indeed, in the context of conformity with PED, a risk assessment is made by the manufacturers and will highlight the main risks associated with industrial valves. Page 2

3 However, participants are not convinced that PED coves explicitely the risk linked to movement of parts. Mr Greenhalgh proposes the question is raised to the PED WG from the European Commission, to get a firm answer. At this occasion, it is also reminded that Machinery WG was supposed to question PED WG on the interaction between PED and MD and no reply has been dispatched. The interaction between PED and MD has been extensively discussed by participants. If some valves were to be considered as machines within the scope of MD, they would be excluded from PED as far as they are not higher than category I. Mr Wagner and Mr Greenhalgh insist that such valves can be quite large and highly pressurized. Giving the priority to MD (and minor mechanical risks) instead of PED (and major pressure risks) would be counter-productive and would create an important loophole for lower quality products. Mr Kraus disagrees, based on the fact that even in MD, the state-of-the-art would apply for the risk of break-up during operation, covering what is related to pressure. Other participants argue that this notion of state-of-the-art is very weak and diverse and will not prevent from the legal putting on the market of products with poor materials and poor design in comparison with PED harmonized standards. Mr Bochaton insists on the fact that if loopholes are created, they will be filled! Mr Vinzio adds that market surveillance for non-compliant products is already quite weak in Europe, so we shouldn t lower the regulatory requirements, especially on the pressure aspects. The status of partly completed machinery is also discussed, because several participants consider it is very difficult to understand how to apply it. According to VDMA, a valve intended to be actuated is at least a PCM. It can be a machine if the actuator is completely defined and the risk assessment considered it. However, one must remember that most manual valves have a mechanical interface for actuators they are intended to be actuated, for instance in retrofitting operations. Uncertainty on responsibility is high if an actuator is actually assembled on such valves. It probably can be rules contractually; however, the fact that regulatory statements would depend on business aspects is not a good solution. Mr Wagner exposes the case of control valves: to be working, they depend on the complete infrastructure that will actually pilot them. By themselves, they do not have any possible use and are not autonomous. Mr Kraus replies that this argument is not enough: the lack of control system does not mean the product is not a machine according to the MD. Mr Tembrink says he is surprised of the reaction of CEIR and BVAA. According to him, being covered by the MD is a good thing that helps designing safer products in addition to the application of PED. Mr Bochaton replies that some requirements from the MD could be confusing when applied to industrial valves. This is typically the case of the safety of the control system that might lead to systematic request of SIL levels on products. At the end, spending possible redesign time of products on risks that exist probably only on paper is maybe not a good idea, especially considering the likely increased risk for category I pressure equipment. Mr Kraus replies that there have been accidents with actuated valves (trapped fingers in arcades). Page 3

4 Research for consensus It seems obvious there is a lot of disagreement between participants. However, for the sake of the Industry, it is necessary to build a common approach, based on what is understood without discrepancy. Mr Vinzio proposes an approach based on the majority, but considering the potential risks related to pure machinery aspects : o Machinery Directive doesn t apply (opinion of the large majority). However, manufacturers agree to apply best practices and to prevent all possible risks o PED applies in all cases Pressure risk is obviously covered Movement risk could be added in the hazards analysis (if not already considered) under responsibility of each manufacturer o Thanks to this : Focus on PED without creating a dangerous loophole All requirements are met under VDMA disagrees with this approach, based on the fact that the starting point shall be the legal text. But how to define the specific application, which is the diverging point between the 2 approaches? Mr Greenhalgh proposes that the question of what is a specific application is formally asked to the Machinery WG. However, on such an abstract question, participants think there will be no reply or a reply which doesn t really help. Mr Maggioni reminds the statement from the former Machinery Directive guidelines which seemed to state that industrial valves don t generally have a specific application. Mr Bochaton reminds it is always up to the manufacturer to choose if he has to apply or not a Directive. Then, the best the group can probably do is to present him a clear choice, focusing on the specific application vs. not specific application. He proposes to draft a new position calling manufacturers to determine if their product is a machine based on the Directive and the guidelines: o Catalog products do not have a specific application : the manufacturer doesn t know in what conditions it will be use Customised products are derived from catalog products and do not have a specific application o Tailor made products (based on contractual technical specifications from the customer) can have a specific application and in such case could be covered by MD Even if this position does not solve everything (PED vs MD), it is considered worth discussing internally Conclusions. After strong discussions, the group finally identified a common point and the dialog continues. It is important that the Industry finds an agreement to avoid that other interests decide instead of them. It is proposed that : A draft paper is prepared by CEIR by mid September and communicated to the whole group for internal discussions, The group meets again on 19 October (in ORGALIME, Brussels), Hopefully, an agreed paper is finalised and presented at the next Machinery Working Group meeting (most likely in November 2015). Page 4

5 Page 5

6 Page 6