CECOD Guide for performing a Hazardous Area Assessment as required by Directive 1999/92/EC:

Transcription

1 G ATEX zoning around Gasoline/Diesel/LPG/DME and CNG/LNG/H2 dispensers Rev 3 CECOD Guide for performing a Hazardous Area Assessment as required by Directive 1999/92/EC: ATEX Zoning around Gasoline/Diesel/LPG/DME and CNG/LNG/H2 Dispensers G Version: 3 Date: 29/07/2016

2 Document Identification Title Reference G Prepared by SG6 Creation Date 16/12/2013 Version 3 G ATEX zoning around Gasoline/Diesel/LPG/DME and CNG/LNG/H2 dispensers Rev 3 CECOD Guide for performing a Hazardous Area Assessment as required by Directive 1999/92/EC: ATEX Zoning around Gasoline/Diesel/LPG/DME and CNG/LNG/H2 Dispensers Document Status comment Draft draft revisions are from 0a to 0z... Final 3 released revisions are from 1 and up History of Changes Date Author Version Description of Changes 16/12/2013 Ph.CLOUTIER 0.a Created for CECOD SG6 brainstorm 21/01/2014 Ph.CLOUTIER 0.b Added T3 class or better for all zones 10/04/2014 Ph.CLOUTIER 0.c Completed with calculations 14/04/2014 Ph.CLOUTIER 0.d Added forecourt arrangement suggestions 05/06/2014 Ph.CLOUTIER 0.e Small adjustments in editorials (yellow) 06/08/2014 Ph.CLOUTIER 1.0 First valid released revision 08/09/2014 Ph.CLOUTIER 1.1 Disclaimer adjustment 29/09/2014 Ph.CLOUTIER 1.1 beta2 Adjustement for CNG and H2 29/01/2015 Ph.CLOUTIER 1.1 beta3 Including LNG in the CNG chapter 23/02/2015 Ph.CLOUTIER 1.1 beta4 Explanation for installed Z2 height 02/03/2015 Ph.CLOUTIER 1.2 Improvements after APEA London 02/15 16/10/2015 M. Melnyk 1.3 Reformatted, calcs put into appendix 10/02/2016 Ph.CLOUTIER 2 For official release, with informative annex 29/07/2016 M. Melnyk 3 Alternative approaches added Revised By Name Title/Role Ola THORKELSSON (Wayne) SG6 chair Mike MELNYK (GVR) Publishing Destination Level Date Authorized by (name) Signature Study Group only SG CHAIR 10/10/2016 Ola Thorkelsson TC decision of 4/10/2016 Internal TC CHAIR 10/10/2016 Philippe Cloutier and TC meeting TC decision of 4/10/2016 Public CECOD CHAIR 10/10/2016 Informative to Public on CECOD Website As per GA decision TC decision of 4/10/2016

3 G ATEX zoning around Gasoline/Diesel/LPG/DME and CNG/LNG/H2 dispensers Rev 3 DISCLAIMER The information contained in this document corresponds to best available practice. The task of CECOD study groups is to compile common guidance without disclosure of proprietary information or knowledge. CECOD assumes no responsibility due to any misuse of this information, and takes no responsibility on any consequences. It is the responsibility of the person or company using this document to fully endorse the design of their products, processes and services. It is the responsibility of the person, or companies, legally in charge of the initial and continued compliance with applicable directives and laws and using this document to check and address any legal issues related to (list is not exhaustive) - Liability - Compliance to applicable directives and laws (from transposition or national specific) - Patent infringement - H&S issues for public and workers - Consequences of design selection, design changes, impact on verification procedures and instructions The use of this document is only authorised if the person or company using this document fully takes ownership of legal and commercial responsibility of the resulting study, product, service, and will not seek any liability of CECOD. The person or company shall also take all and any necessary steps to prevent any liability to CECOD.

4 G ATEX zoning around Gasoline/Diesel/LPG/DME and CNG/LNG/H2 dispensers Rev 3 INDEX Scope 1) Domain 2) References to standards and Directives 3) Terms, acronyms and symbols 4) Hazardous Areas in the Refuelling Area Annex : Supporting Calculations 1/7

5 G ATEX zoning around Gasoline/Diesel/LPG/DME and CNG/LNG/H2 dispensers Rev 3 SCOPE: Dispensers manufactured by CECOD members are subject to ATEX 94/9/EC certification conducted with reference to harmonized EU standards, thereby demonstrating ATEX compliance - Gasoline and diesel dispensers: using harmonized standard EN LPG dispensers: using harmonized standard EN Both standards comprise an annex ZA for correspondence between the clauses in the standard and the essential requirements of the ATEX directive 94/9/EC for products. However, the specific ATEX directive for working places (1992/92/EC) is transposed into national requirements. Such transposition either leaves definition of hazardous areas to the owner of the fuel filling station, or some countries may impose local zoning constraints. When zoning around dispensers is defined by national legislation (as part of transposition of 1999/92/EC), it is assumed that legislators have performed their own risk assessment and calculations to establish requirements in the most adequate way. When zoning around dispensers is left by national legislation for the site owner to determine, it is not always easy for owners to establish this, as this requires very good understanding of the refuelling process and other existing hazards on the site. The purpose of this guide is to help owners establish their own assessment for 1999/92/EC ATEX zoning for their sites. The purpose of this guide is also to bring CECOD knowledge and hazardous area assessment work to EU member state legislators, to help them establish their own risk assessment when they impose the zoning for 1999/92/EC in their national legislation and/or rules. The use of this guide is only permitted if the person or company using this document fully takes ownership of legal and commercial responsibility of resulting study, risk assessment, decisions and any consequences, and will not seek any liability of CECOD or CECOD members, either companies or individuals. The person or company using this guide as a whole, or only part, also commits to take all necessary steps to prevent any liability to CECOD. 2/7

6 G ATEX zoning around Gasoline/Diesel/LPG/DME and CNG/LNG/H2 dispensers Rev 3 1) Domain Filling stations in Europe sites for refueling with gasoline, diesel, LPG, DME, CNG, LNG and Hydrogen where the dispensing equipment and vehicle are in free open air. 2) References to Standards and Directives EN ATEX series EN EN Directives 94/9/EC (ATEX) and NLF 2014/34/EU - concerning equipment and protective systems intended for use in potentially explosive atmospheres Directive 1992/92/EC (ATEX working places) - concerning Protection of Workers at Risk from Potentially Explosive Atmospheres 3) Terms, acronyms and symbols LEL: Lower Explosive Limit - the lowest concentration (percentage) of a gas or a vapour in air capable of producing a flash of fire in presence of an ignition source 3/7

7 G ATEX zoning around Gasoline/Diesel/LPG/DME and CNG/LNG/H2 dispensers Rev 3 4) Hazardous Areas in the Refuelling Area 4.1. Fuels with Vapours heavier than Air Possible Zoning Approaches based on this Guide As part of a proposed method for hazardous area assessment, this guide considers the likely area extent over which a vapour-air mix will remain above a particular concentration (percentage of LEL) during the refueling of vehicles. The second step is to convert these areas to traditional hazardous area zones. Two possible approaches are documented within this guide; although other hazardous area assessments are feasible based upon using the same formulae but by selecting different concentration limits. Approach A : a cautious approach based on o Zone 0 for any potential concentration above LEL o Zone 1 for where concentration by calculation under ideal conditions (no wind, no moving vehicles etc) would be above 20% of LEL o Zone 2 for where concentration by calculation under ideal conditions would be above 5% of LEL. However, note that the spread across the forecourt is based upon the average vapour volume associated with 20% and 5% LEL concentrations, which is defined in column xc of Table 1. Approach B : a traditional approach in some countries based on o Zone 0 for any potential concentration above LEL o Zone 1 for where concentration by calculation under ideal conditions would be above 50% of LEL under normal operating conditions, or under foreseeable fault conditions (primary releases) o Zone 2 for where concentration would be above 50% of LEL in the event of a fault, in particular, complete failure of a vapour recovery system (secondary release). Approach A is likely to be useful for dispenser designers looking at the detailed positioning of electronic components or user interfaces external to the dispenser housing. For example, equipment mounted on nozzles, or non-ex equipment at relatively low heights for disabled persons access. Approach B is likely to be useful for generic forecourt zoning. 4/7

8 G ATEX zoning around Gasoline/Diesel/LPG/DME and CNG/LNG/H2 dispensers Rev Hazardous Areas A permanent hazardous area exists, which is a combination of 1. the hazardous area associated with idle dispensers installed on the fuel filling station. 2. a general hazardous area across the forecourt associated with small fuel spills and residual vapour from the refueling of vehicles. Temporary hazardous areas also exist due to the refueling process. The point of vapour release is dependent upon the position at which a vehicle parks for refueling. Fixed electrical equipment needs to be suitably selected taking into account the permanent hazardous area, and the temporary hazardous areas associated with all potential vehicle positions for refueling Permanent Hazardous Area The hazardous area associated with an idle dispenser on the fuel filling station is as defined in the ATEX type certification for the dispenser. A Zone 2 hazardous area exists covering an area defined by the maximum extent of the dispenser nozzle reach, and further extended by the Zone 2 associated with the refueling process. The height of this Zone 2 is assumed to be 100mm from the forecourt level. 5/7

9 G ATEX zoning around Gasoline/Diesel/LPG/DME and CNG/LNG/H2 dispensers Rev Temporary Hazardous Area associated with the Refueling of Petrol and Diesel As fuel is delivered into a vehicle tank, vapour is expelled from the tank filler neck. The immediate hazardous area at above LEL is centred at the filler neck and approximates to a sphere, and as vapour is diluted by air, spheres of lower concentration are centred at the same point. This vapour then falls under gravity to the forecourt, forming a cylindrical projection. This is always assumed to be a Zone 2. The vapour then spreads across the forecourt. The hazardous area associated with refueling is deemed to terminate where it intersects with the permanent Zone 2 hazardous area across the forecourt. This temporary hazardous area disperses within 10 to 15 seconds of releasing the nozzle trigger. Example of zoning using Approach A 6/7

10 G ATEX zoning around Gasoline/Diesel/LPG/DME and CNG/LNG/H2 dispensers Rev 3 Example of zoning using Approach B Non-vapour recovery and diesel dispensing Dispensing with vapour recovery Temporary Hazardous Area associated with the Refueling of LPG. There is potential for small leaks at the interface of the nozzle to vehicle filler, and in particular, a discharge of gas when the nozzle is released at the end of a delivery. The resultant hazardous area is a similar shape to that defined for petrol and diesel. 7/7

11 Summary of Hazardous Area Extent (Table 1) Description Rzone0 (mm) R50% (mm) R20% (mm) R10% (mm) R5% (mm) x50% (mm) (Approach B) x20% (mm) x10% (mm) x5% (mm) xc% (mm) (Approach A) Petrol 40 litres per minute, with active closed loop vapour recovery Petrol 40 litres per minute, open loop vr with monitoring Petrol 40 litres per minute, open loop vr without monitoring Petrol 40 litres per minute, without vapour recovery Petrol 40 to 80 litres per minute, without vr Petrol 80 to 130 litres per minute, without vr Petrol 130 to 200 litres per minute, without vr Diesel 40 litres per minute, Diesel 40 to 80 litres per minute, Diesel 80 to 200 litres per minute, LPG 40 litres per minute, LPG 40 to 130 litres per minute, h (mm) Note that for Approach A the blue and yellow columns are applied. (for Approach B the blue and green columns are applied) 8/30

12 Illustrative Examples of applying Table 1 G ATEX zoning around Gasoline/Diesel/LPG/DME and CNG/LNG/H2 dispensers Rev 3 9/30

13 G ATEX zoning around Gasoline/Diesel/LPG/DME and CNG/LNG/H2 dispensers Rev Illustrative Example of Refueling Process 10/30

14 G ATEX zoning around Gasoline/Diesel/LPG/DME and CNG/LNG/H2 dispensers Rev Fuels with Vapours lighter than Air A single CECOD approach is offered within this guide for consideration where local or national rules/guidance is not available Hazardous Areas A permanent hazardous area exists, which is a combination of 1. the hazardous area associated with idle dispensers installed on the fuel filling station. 2. a general hazardous area across the forecourt associated with potential small gas releases from nozzle disconnections and from dispenser pressure relief valve operation. Temporary hazardous areas also exist due to the connection and disconnection of nozzles, and the possible operation of safety breaks/ disconnect devices in dispenser hose assemblies.. The point of vapour release is dependent upon the position at which a vehicle parks for refueling. Fixed electrical equipment needs to be suitably selected taking into account the permanent hazardous area and the temporary hazardous areas associated with all potential vehicle positions for refueling Permanent Hazardous Area The hazardous area associated with an idle dispenser on the fuel filling station is as defined in the ATEX type certification for the dispenser. A Zone 2 hazardous area exists covering an area defined by the maximum extent of the dispenser nozzle reach. The height of this zone is deemed to be within 2 metres and 4 metres above forecourt level, although care should be taken to review any potential vapour traps under the forecourt canopy. It should be noted that the hazardous area associated with hydrogen dispensers is equipment group and temperature class IIC T1. 11/30

15 G ATEX zoning around Gasoline/Diesel/LPG/DME and CNG/LNG/H2 dispensers Rev Temporary Hazardous Areas Vapour expelled from the connection/disconnection of the nozzle, or operation of the safebreak, approximates to a Zone 0 sphere, and as vapour is diluted by air, Zone 1 and Zone 2 spheres centred at the same point. This vapour then rises, forming a cylindrical upward projection of the hazardous area until it intersects with the permanent forecourt hazardous area. The temporary hazardous zone will disperse within 15 seconds for CNG and 120 seconds for LNG Example Illustration of Permanent and Temporary Hazardous Areas 12/30

16 Summary of Hazardous Area Extent G ATEX zoning around Gasoline/Diesel/LPG/DME and CNG/LNG/H2 dispensers Rev 3 Description Rzone0 Rzone1 Rzone2 (mm) (mm) (mm) CNG or LNG 20kg per minute, CNG or LNG 20 to 60 kg per minute Hydrogen 10kg per minute, Hydrogen 10 to 40 kg per minute, /30

17 ANNEX A Supporting Calculations Note that the values in Table 1 were created using formulae in an excel spreadsheet. A limited selection of examples are provided here in order to explain and support the general methodology. Fundamentals of the CECOD approach Approach A, is to consider: - Zone 0 for any potential concentration above LEL - Zone 1 for where concentration by calculation under ideal conditions (no wind, no moving vehicles etc) would be above 20% of LEL - Zone 2 for where concentration by calculation under ideal conditions would be above 5% of LEL Applying calculations to a traditional approach Approach B, is to review - Zone 0 for any potential concentration above LEL - Zone 1 for where concentration by calculation under ideal conditions would be above 50% of LEL under normal operating conditions, or under foreseeable fault conditions (primary releases) - Zone 2 for where concentration would be above 50% of LEL in the event of a fault, in particular, complete failure of a vapour recovery system (secondary release). 14/30

18 Overview of Calculations Calculations for gasoline and diesel (vapours heavier than air) include: Calculation of the radius of the spheres at various concentrations around the filler point during fuelling of the vehicle. Calculation of the extent (x) of how far the vapour will spread across the forecourt as it falls. Figure A1 Calculations for gases and LPG relate to the radius of the spheres around any potential disconnection points. i.e. nozzle to vehicle, and safebreaks, and the associated spread of vapour across the forecourt. 15/30

19 Gasoline dispensing : calculation of radius of spheres around filler point Assumption 1: Vapours of gasoline expelled from car filler neck fall to the floor at typically 0.4 metres/second (Worst case scenario is butane which is the lighter fraction of gasoline). Assumption 2: LEL is 0.7% for gasoline. Assumption 3: The speed at which gasoline vapours are expelled creates a spheres of varying minimum concentrations. Such spheres are limited on one side by the vehicle cladding, plus it is unlikely that the speed of projection will really result in forming the upper section of the sphere. To create the worst case radius from the calculations, (even if a full sphere is illustrated in the figures), consider that the vapour expelled creates a quarter sphere. - ½ because vapours are falling under gravity, - ½ because dissipation is limited on one side by the vehicle cladding As a consequence, surface where the vapours mix with air at bottom border of sphere is only 25% of total surface of sphere. Sphere surface is S = 4 π R 2 Therefore surface of interface vapour to fresh air (Si) is π R 2. Assumption 4: If filling flowrate is Q (in liters per second), amount of vapours expelled during the fill process is also Q. However, for a closed loop vapour recovery, at least 95% is captured by the VR2 recovery system (based upon the vapour/liquid ratio for systems), so at a fuel flowrate of 40lpm, only 40 x 0.05 = 2 liters per minute of vapour is assumed to be released to atmosphere. For an open loop system with monitoring, at least 85% is captured. (6lpm vapour release) For an open loop system without monitoring, it is assumed that it will spend part of its life operating at below 85% efficiency, and that typically efficiency may fall to 70% (12lpm vapour release) Assumption 5: The vapours are considered to be saturated at 100% to account for worst case (a more typical scenario is between 30% and 65% when expelled from vehicle tank filler neck). Assumption 6: Safe rounding will be applied to calculations to round upwards to the nearest 0.05m (50mm). 16/30

20 Formula: Relative speed of expelled volume through the interface surface (not considering gravity) is Ps = (Q/1000)/Si = (Q/1000)/ (π * R 2 ) Considering gravity (and migration speed of 0.4m/s of vapours, ie: butane worst case), concentration of vapours at the interface surface is C = Ps/0.4 C shall be lower than targeted LEL concentration for such zone C = Ps /0.4 = ((Q/1000)/ (π * R 2 ))/ 0.4 R = Q / (400 * π * C) Applying the Formula for gasoline - Zone 0 assumes any potential concentration above LEL, i.e. >0.7% gasoline vapour concentration - 50% of LEL therefore relates to a gasoline vapour concentration of 0.35% - 20% of LEL relates to a gasoline vapour concentration of 0.14% - 10% of LEL relates to a gasoline vapour concentration of 0.07% - 5% of LEL, relates to a gasoline vapour concentration of 0.035% Example 1: At a fuel flowrate of 40lpm with closed loop vapour recovery, applying Approach A and considering extent of 5% of LEL, the concentration target is 0.035%, R5% = ((0.05 * 40 (l/mn) / 60 (s in a minute) )/ (400 * π * ) R5% = 0.275m Rounded to 0.3m (300mm) Example 2: At a fuel flowrate of 40lpm without vapour recovery, considering extent of sphere close to filler at 50% of LEL (Approach B Zone 1 extent), concentration target is 0.35%, R5% = (40 (l/mn) / 60 (s in a minute) )/ (400 * π * ) = 0.67 / (400 * 3.14 * ) R5% = 0.389m rounded to 0.4m (400mm) Results for all concentrations and gasoline flowrates are recorded in Table 1. 17/30

21 Diesel dispensing : calculation of radius of spheres around filler point Assumption 1: LEL is 0.6%, and saturating pressure is lower than 40 C. As for gasoline, the worst case scenario is butane content. This will assume to be expelled from car filler neck and fall to the floor at typically 0.4 metres/second Assumption 2: Maximum likely volume of vapours expelled from filler neck is 1% (10hPa / 1013 C) of the nozzle flow. For the calculation, Q (flowrate of vapours expelled) shall be considered as 2% of flowrate of nozzle (applying a safety factor of 2). Formula: C shall be lower than targeted LEL concentration for evaluated zone C = Ps /0.4 = ((Q/1000)/ (π * R 2 ))/ 0.4 R = Q / (400 * π * C) Applying the Formula for diesel - Zone 0 assumes any potential concentration above LEL, i.e. 0.6% - 50% of LEL therefore relates to a vapour concentration of 0.3% - 20% of LEL, relates to a vapour concentration of 0.12% - 10% of LEL, relates to a vapour concentration of 0.06% - 5% of LEL relates to a vapour concentration of 0.03% Example 1: At a fuel flowrate of 40lpm, considering extent of 5% of LEL (Approach A Zone 2 extent), concentration target is 0.03%, R5% = (0.02 * 40 (l/mn) / 60 (s in a minute) )/ (400 * π * ) R5% = 0.188m rounded to 0.2m (200mm) Example 2: At a fuel flowrate of 80lpm, considering extent of 50% of LEL (Approach B Zone 1 extent), concentration target is 0.3%, R50% = (0.02 * 80 (l/mn) / 60 (s in a minute) )/ (400 * π * 0.003) R50% = 0.084m rounded to 0.1m (100mm) Results for all concentrations and diesel flowrates are recorded in Table 1. 18/30

22 Extent of gasoline or diesel vapour spread across the forecourt Once vapour (heavier than air) has been expelled from the vehicle tank, it will fall under gravity. We assume this action will create a cylinder shape. Since the vapour is mixing with air as it falls, we assume this cylinder is a Zone 2. However, as the vapour hits a layer of previously expelled vapour at ground level, the result will be an accumulation of vapour which will spread outwards across the forecourt. This forms a blanket or cushion. (See Figure A1 : spread of Zone 2 ). It is assumed that vapours are expelled from tank at saturated vapour pressure (which is a maximum of 65% of atmospheric pressure). A worst case scenario is assumed of this hazardous area being present immediately from the commencement of the fuelling operation and for 10 seconds after fuelling process is finished (end of flow). A worst case scenario is assumed of the hazard remaining at Zone2 as it spreads away from the source point, and that it does not dilute further. Calculation is to evaluate the thickness of the Z2 vapour blanket/cushion till it reaches 100mm height (the permanent Z2 height is 100 mm, and vehicle clearance will be higher than this. It should be noted that the highest point of the blanket/cushion only depends of the volume of vapours and not the height of the filling point. Formula: Q, quantity vapour = Flowrate * (saturation/lel of vapour) * (1-vr efficiency/100) * (100/concentration target as % of LEL) Height (H in m) of cushion at a distance R (in m) from the center of the Zone 2 cylinder (if descending speed of vapours is V in m/s, and Q (l/s) is the quantity of vapours at average between 1/5 th and 1/20 th of LEL for Approach A, or quantity of vapour at 50%LEL for Approach B) Formula: H = Q / ( * R * V * 1000) The radius at the intersection with the generic forecourt zone at a height of 0.1m, defined as x in figure A1, is such that: 0.1 = Q / ( * x * V * 1000) x (as shown in figure A1) = Q/ ( * V * 100). 19/30

23 Applying the formula for gasoline Gasoline at 40 l/mn without vapour recovery, using Approach A - 40 l/mn of saturated vapours are expelled from tank, at 65% saturation, creating between approx. (40 * (65/0.7) *1 *100/20) = and (40 * (65/0.7) * 1 *100/5) = liters per minute of vapours that can be considered as Zone 2 (limits of 1/5 th of LEL and 1/20 th of LEL). Average is liters - this is approx 785 liters per second - Using approach A, pattern starts at radius of R=1.25m (projection of Z2 of active dispenser, see sphere calcs ) - Formula: H = 785 / ( pi * R * 0.4 * 1000) (descending speed of butane is 0.4 m/s) - x = value of R when H=0.1 R=1.25 R=1.50 R=2.00 R=2.50 R=3.00 R=4.00 R=6.2 H=0.50 H=0.42 H=0.32 H=0.25 H=0.21 H=0.15 H=0.10 X = 6.2m Gasoline at 40 l/mn (with closed loop VR), using Approach A - volume ratio efficiency of VR2 systems is between 95 and 105% so worst case scenario is when VR2 system is at 95% - remaining 5% of 40 l/mn of saturated vapours are expelled from tank, at 65% saturation, creating between 950 and 3750 liters per minute of vapours that can be considered as Z2 Average is 2350 liters - this is approx 40 liters per second - Pattern starts at radius of R=0.3m (projection of Z2 of active dispenser, see sphere calcs ) - Formula: H = 2 * 40 / ( pi * R * 0.4 * 1000) X = 0.35m (i.e. minimal spread beyond Zone 2 cylinder) Gasoline at 40 l/mn without vapour recovery, using Approach B - 40 l/mn of saturated vapours are expelled from tank, at 65% saturation, creating approx. (40 * (65/0.7) *1 *100/50) = 7500 lpm of vapours that can be considered as above 50% of LEL. - this is approx 125 liters per second - Using approach A, pattern starts at radius of R=0.4m (projection of Z1 sphere) - Formula: H = 125 / ( pi * R * 0.4 * 1000) (descending speed of butane is 0.4 m/s) - x = value of R when H=0.1 X = 1.0m 20/30

24 Applying the formula for diesel Diesel at 40 l/mn using Approach A - 40 l/mn of saturated vapours are expelled from tank, at 1% saturation, creating between. (40 * (1/0.6) *100/20) = 333 and (40 * (1/0.6) *100/5) = 1333 liters per minute of vapours that can be considered as Zone 2. - Average is 833 liters - this is approx 14 liters per second - pattern starts at radius of R=0.2m (projection of Z2 of active dispenser, see sphere calcs ) - Formula: H = 14 / ( pi * R * 0.4 * 1000) (descending speed of butane is 0.4 m/s) X < 0.2m (so we assume no spread beyond the Zone 2 cylinder) Similarly calculations at other flowrates up to 200lpm suggest there is no spread of vapour beyond the radius of the projection of the Zone 2 sphere around the filling point 21/30

25 LPG/ DME Dispensing. Calculation of radius of spheres around nozzle connection Assumptions: Any leak of LPG from the nozzle interface to the vehicle will turn to vapours immediately, and these vapours fall to the floor at around 0.3m/s (worst case is Propane). Assume LEL is 1.8% for Butane (worst case). A similar formula can be used for estimating the hazardous area extent as for gasoline. Anticipated leakage calcs: Light vehicle LPG refuelling (flowrate 40 liters per minute) Public Self Service or attended: Any leak at the nozzle interface is considered as being clearly detected if the dispenser operator is attending and controlling the fill (dead man push button) and if the leak rate is greater than 10 cm 3 per minute of liquid LPG. If leak is above 10 cm 3 per minute, it is assumed operator will stop flow by releasing dead-man push button, or activating the emergency stop. If leak rate is 10 cm 3 per minute of liquid LPG, it will generate (worst case scenario) a volume of saturated LPG gas (gaseous form) of : - Propane C3H8 density of liquid at 20 C = cm 3 of liquid creates 22.4 * 10 * / 44 = 2,59 litres of 100% saturated gas - Butane C4H10 density of liquid at 20 C = cm 3 of liquid creates 22.4 * 10 * / 58 = 2.24 litres of saturated 100% gas Therefore worst case scenario is a Propane leak of 2.59 liters/minute Heavy vehicles (flowrate > 40 liters per minute and 130 liters per minutes) Attended or professional drivers: Any leak at nozzle interface is considered as being clearly detected if the dispenser operator is in close proximity of the vehicle connection and if the leak rate is greater than 30 cm 3 per minute of liquid LPG. If leak is 30 cm 3 per minute of liquid LPG, it will generate (worst case scenario) a volume of saturated LPG gas (gaseous form) of: - Propane C3H8 : 30 cm 3 of liquid creates 22.4 * 30 * / 44 = 7.77 liters of saturated 100% gas - Butane C4H10 30 cm 3 of liquid creates 22.4 * 3 * / 58 = 6,71 liters of saturated 100% gas Therefore worst case scenario is a Propane leak of 7.77 liters/minute. 22/30

26 Formula: C shall be lower than targeted LEL concentration for evaluated zone. Considering gravity (and migration speed of 0.3m/s of propane vapours, worst case scenario), concentration of vapours at the interface between vapour and air is C = Ps/ C = Ps /0.3 = ((Q/1000)/ (π * R 2 ))/ 0.3 R = Q / (300 * π * C) Applying the Formula for LPG - Zone 0 assumes any potential concentration above LEL, i.e. 1.8% - 50% of LEL therefore relates to a vapour concentration of 0.9% - 20% of LEL, relates to a vapour concentration of 0.36% - 10% of LEL, relates to a vapour concentration of 0.18% - 5% of LEL relates to a vapour concentration of 0.09% LPG liquid turns into vapours at atmospheric pressure. Q is the maximum invisible leak rate of vapour. Self service mode with dead man push button Q = 2.59 liters per minute. Attended mode with activation switch Q = 7.77 liters per minute R = Q / (300 * π * C) Example 1: At a LPG fuel flowrate of 40lpm, where there will be a deadman button, considering extent of a concentration target of 0.09%, (Approach A : Zone 2 sphere calculation) R5% = (2.59 (l/mn) / 60 (s in a minute) )/ (300 * π * ) R5% = 0.22m rounded to 0.25m (250mm) Example 2. At a LPG fuel flowrate of 40lpm, where there will be a deadman button, considering extent of a concentration target of 0.9%, (Approach B : Zone 1 sphere calculation) R50% = (2.59 (l/mn) / 60 (s in a minute) )/ (300 * π * 0.009) R50% = 0.07m rounded to 0.10m (100mm) Results for all zones and for standard and high flow scenarios are recorded in Table 1. For simplicity, assume that the hazardous areas associated with the safe break are of a similar size. 23/30

27 Extent of LPG vapour spread across the forecourt Applying similar principles to gasoline refueling, using Approach A At 40 l/mn assume an invisible leak of 2.59 liters per minutes at nozzle or break-away: 2.59 l/mn of 100% saturated vapour with an LEL of 1.8% will create between approx. (2.59 * (100/01.8) *100/20) = 720 and (2.59 * (100/01.8) *100/5) = 2880 liters. Average = 1800litres - This is approx 30 liters per second - Pattern starts at radius of R=200mm (as calculated above) - Formula: H = 30 / ( pi * x * 0.3 * 1000) (descending speed of propane is 0.3 m/s) X = 0.35m At 130 l/mn assume an invisible leak of 7.77 liters per minutes at nozzle or break-away: - Leak of 7.77 liters per minute of propane (worst case), is creating between 2160 and 8640 liters per minute of gas-air mixture, which can be considered as Zone2. - Average is 5400 liters/min - This is approx 90 liters per second - Pattern starts at radius of R=400mm (as calculated above) - Formula: H = 90 / ( pi * x * 0.3 * 1000) (descending speed of propane is 0.3 m/s) X = 1.00 m Note : Similar calculations using Approach B suggest there is no spread of vapour at a concentration of 50% of LEL at above 0.1m above the forecourt beyond the radius of the projection of the Zone 2 sphere around the nozzle / safebreak. 24/30

28 CNG/LNG (CH4) dispensing A similar approach to that applied to gasoline, diesel and LPG could be applied to CNG and LNG, although care should be taken not to conflict with any harmonized standards or national regulations which may be introduced. Assumption 1: The calculations here assume open air sites using odorized CH4 dispenser Assumption 2: The flow is controlled by a dead-man push button, or dispenser attended by trained staff Assumption 3: The dispenser is deemed active from the moment the nozzle is connected to the vehicle filler neck (start of transaction) until: - for CNG : 15 seconds after the nozzle is disconnected from the filler neck of the vehicle (end of flow and dissipation delay for CH4). - for LNG : 120 seconds after the nozzle is disconnected from the filler neck of the vehicle (end of flow and dissipation delay for CH4). (see footnote on this page) Assumption 4: Any leak of CNG/LNG from the nozzle interface with the vehicle or at the hose safety breakaway system will dissolve to the atmosphere and start ascending at around 0.8 metres/second. LEL is 5% for Methane. Assumption 5: A similar approach to that used for the assessment of gasoline zones can be used, but in the case of CNG/LNG, the quarter sphere around the filler interface as used in the calculations is based upon the underside of the sphere not being formed due to the gas rising. A worst case approach assumes the hazardous areas associated with safe break assemblies are the same size as those calculated using the quarter sphere approach used for the nozzle interface. Note on LNG: LNG gas is very cold (approx -160 C) making density of CH4 (gaseous form) more than 2 times the usual density of CH4 at normal ambient temperature. This prevents normal upward dissipation of CH4 till the gas is capable of warming up to ambient. This process takes time. Cold CH4 needs to mix with enough warmer ambient air before it will start to ascend and dissipate. At nozzle disconnection, any liquid drops of LNG also need time to evapourate and dissipate. 25/30

29 Assumption 6 (detection): Light vehicle fueling with CNG (flowrate 20 kg per minute) Public Self Service or attended: A leak at nozzle interface or hose breakway will be audible, visible and smell if the dispenser operator is in a position controlling the fill (dead man push button) and if there is a leak of more than 100cm 3 /s (at 1 atm) of gas. This equates to approx. 0.4cm 3 /s of natural gas at 250 bar (expanding instantaneously to 100cm 3 at 1 atm with significant noise). In the calculations, a safety factor of 2 is added : Q = 200cm 3 /s = 0.2litres/s. Heavy vehicles fueling with CNG (flowrate > 20 kg per minute and 60 kg per minutes) Attended: A leak at nozzle interface or hose breakway will be audible, visible and smell if the dispenser operator is in a position controlling the fill (dead man push button) and if there is a leak of more than 250cm 3 /s (at 1 atm) of gas. This equates to approx. 1cm 3 /s of natural gas at 250 bar. In the calculations, a safety factor of 2 is added : Q = 500cm 3 /s = 0.5litres/s. If odorizing is not technically feasible (incompatibility of odorizing compound with cryogenic temperatures or the source of CH4), it is strongly suggested that special warning signs are affixed to inform operators of this fact, and raise awareness of possible signs of malfunctions/leaks. Note: the odorizing compound used in CNG/LNG (Méthanethiol / Mercaptan ) is detectable by human nose at levels as low as ppm (ref: INRS-France FT190 edition 2007) and is heavier than air, so the smell of a leak remains longer than leak itself, and is nose detectable. Ascending speed of CH4 gas information Thanks to: Association Suisse de l Industrie Gazière - Verband der Schweizerischen Gasindustrie essiggas_d.pdf 26/30

30 Formula: Relative speed of expelled volume through the interface surface (not considering gravity) is Ps = (Q/1000)/Si = (Q/1000)/ (π * R 2 ) Considering gravity (and migration speed of 0.8m/s of methane), concentration of CH4 at top side of the interface between the vapour and air is C = PS/0.8 C shall be lower than targeted LEL concentration for such zone C = Ps /0.8 = ((Q/1000)/ (π * R 2 ))/ 0.8 R = Q / (800 * π * C) Applying the Formula for CNG/LNG - Zone 0 assumes any potential concentration above LEL, i.e. 5% - Zone 1 assumes any potential concentration above 20% of LEL, i.e. 1% - Zone 2 assumes any potential concentration above 5% of LEL, i.e. 0.25% Example 1: At a fuel flowrate of 20 kg per minute, considering extent of Zone 2, concentration target is 0.25%, Rzone2 = 0.2/ (800 * π * ) Rzone2 = 0.178m rounded to 0.2m Example 2: At a fuel flowrate of 60 kg per minute, considering extent of Zone 2, concentration target is 0.25%, Rzone2 = 0.5/ (800 * π * ) Rzone2 = 0.282m rounded to 0.3m Results for all zones and for standard and high flow scenarios are recorded in Table 1. Extent of CH4 spread across the forecourt To limit risk, and because any pockets of CH4 are not visible, and predicting where the gas will travel is highly dependent upon site conditions, CECOD propose to assume a permanent zone 2 between 2m and 4m above the forecourt. 27/30

31 Hydrogen dispensing A similar approach to that applied to gasoline, diesel and LPG could be applied to Hydrogen; although care should be taken not to conflict with any harmonized standards or national regulations which may be introduced. Assumption 1: The dispenser is deemed active from the moment the nozzle is connected to the vehicle filler neck (start of transaction) until 15 seconds after the nozzle is disconnected from the filler neck of the vehicle (end of flow and dissipation delay for H2). Assumption 2: The flow is controlled by a dead-man push button, or dispenser attended by trained staff Assumption 3: Any leak of H2 from nozzle interface with the vehicle or at the hose breakaway system will enter the atmosphere and start ascending at approx 1 metre/second. LEL is 4% for hydrogen. Ascending speed information for H2 Thanks to: INERIS and TAS (Daisy Bell Avalanche control system) (ref at page 3, fig 4) Assumption 4: A similar approach to that used for the assessment of gasoline zones can be used, but in the case of Hydrogen, the quarter sphere around the filler interface as used in the calculations is based upon the underside of the sphere not being formed due to the gas rising. A worst case approach assumes the hazardous areas associated with safe break assemblies are the same size as those calculated using the quarter sphere approach used for the nozzle interface. 28/30

32 Assumption 5 (detection): Light vehicle fueling with Hydrogen (flowrate 10 kg per minute) Public Self Service or attended: A leak at nozzle interface or hose breakway will be audible and possibly visible if the dispenser operator is in a position controlling the fill (dead man push button) and if there is a leak of more than 100cm 3 /s (at 1 atm) of gas. This equates to approx. 0.25cm 3 /s of hydrogen at 400 bar (expanding instantaneously to 100cm 3 at 1 atm with significant noise). In the calculations, a safety factor of 5 is applied because of very low energy needed to ignite H2 and the likely absence of odorizing compound. Q = 500cm 3 /s = 0.5litres/s. Heavy vehicles fueling with Hydrogen (flowrate 40 kg per minute) Attended: A leak at nozzle interface or hose breakway will be audible and possibly visible if the dispenser operator is in a position controlling the fill (dead man push button) and if there is a leak of more than 250cm 3 /s (at 1 atm) of gas. This equates to approx. 0.6cm 3 /s of hydrogen at 400 bar (expanding instantaneously to 250cm 3 at 1 atm with significant noise). In the calculations, a safety factor of 5 is added : Q = 1250cm 3 /s = 1.25litres/s. Formula: Relative speed of expelled volume through the interface surface (not considering gravity) is Ps = (Q/1000)/Si = (Q/1000)/ π R 2 Considering gravity (and migration speed of 1m/s of H2), concentration of H2 at top side of the interface between the hydrogen and air is C = PS/1.0 C shall be lower than targeted LEL concentration for such zone C = Ps /1.0 = ((Q/1000)/ (π * R 2 )) R = Q / (1000 * π * C) Applying the Formula for Hydrogen - Zone 0 assumes any potential concentration above LEL, i.e. 4% - Zone 1 assumes any potential concentration above 20% of LEL, i.e. 0.8% - Zone 2 assumes any potential concentration above 5% of LEL, i.e. 0.2% 29/30

33 Example 1: At a fuel flowrate of 10 kg per minute, considering extent of Zone 2, concentration target is 0.2%, Rzone2 = 0.5/ (1000 * π * ) Rzone2 = 0.282m rounded to 0.3m Example 2: At a fuel flowrate of 40 kg per minute, considering extent of Zone 2, concentration target is 0.25%, Rzone2 = 1.25 / (800 * π * ) Rzone2 = 0.446m rounded to 0.45m Results for all zones and for standard and high flow scenarios are recorded in Table 1. Extent of H2 spread across the forecourt To limit risk, and because any pockets of H2 are not visible, and predicting where the gas will travel is highly dependent upon site conditions, CECOD propose to assume a permanent zone 2 between 2m and 4m above the forecourt. 30/30