semi合集-English.pdf - 第6810页

SEMI S2-0703a E © SEMI 1991, 2004 39 method is to make two traverses across t he diameter of the duct at right angles to each other. Reading a re taken at the center of annular rings of equal area. Whenever possible, the…

SEMI S2-0703a

manifold boxes, and secondary gas panel enclosures,

are typically six-sided fully enclosed enclosures with

access panels/doors on at least one side. These

ventilated enclosures are designed to contain and

remove hazardous gases from the work area in the

event of a gas piping failure or leak. Gas panel

enclosures are typically of two types, those requiring no

access while gas systems are charged, and those that

must be opened during processing while gas systems

are charged. There is also a distinct difference in

control specifications for those with pyrophorics or

other flammables vs. other HPMs, specifically in the

control of pocketing.

A2-6.1 Design Recommendations

A2-6.1.1 Compartmentalize potential leak points.

A2-6.1.2 Minimize the total size of the panel and its

enclosure.

A2-6.1.3 Minimize size and number of openings.

A2-6.1.4 Minimize static pressure requirements of the

enclosure; control has been shown to be achievable

with –1.3 to –2.5 mm ( –0.05 to –0.1 in.) w.g.

A2-6.1.5 Design for sweep. Minimize the number and

size of openings. Seal unnecessary openings (e.g.,

seams, utility holes).

A2-6.1.6 Where routinely used access doors are

required:

 Make the access door as small as practical.

 Place the openings to the enclosure in the access

door to minimize air flow requirements.

 Provide baffles behind the door to direct leaks

away from the door and openings.

 Compartmentalize the enclosure so that access to

one area does not affect air flow control in other

areas.

A2-6.2 Control Specifications

A2-6.2.1 Exhaust volumes as low as 4–5 air changes

per minute or less can be specified and meet the S2

criteria in Section 23.5 if the design principles listed

above are considered when designing equipment and

enclosures.

A2-6.2.2 Where there is potential for chemical

exposure during access which can be controlled by face

velocity, the enclosure should also provide a minimum

face velocity of 0.36 to 0.76 m/s (70 to 150 fpm) when

open. Face velocity should not be relied upon to control

emissions from a pressurized fitting.

A2-6.2.3 Enclosures for pyrophoric or flammable

gases should be designed to ensure adequately uniform

dilution (i.e., prevent “pocketing”) and to prevent

accumulation of pyrophoric and flammable gases above

their lower explosive limit. Uniform dilution can

generally be verified through exhaust vapor

visualization techniques. Ventilation flow rate should

be adequate to maintain concentrations below 25% of

the lower explosive limit for the gas with the lowest

LEL that is used in the enclosure. This can generally be

verified using engineering calculations to verify

dilution, and vapor visualization to verify mixing.

A2-6.3 Recommended Measurement/Validation

Method

A2-6.3.1 Preferred validation by tracer gas testing per

SEMI F15.

A2-6.3.2 Additional confirmation by visualization

check of air flow, mixing and sweep using smoke or

vapor testing.

A2-6.3.3 Measurement of average face velocity at

inlet(s), opening(s), or routinely used access doors.

Measurements should be taken with a velometer or

anemometer. For larger openings, multiple

measurements are taken in a grid, at least 10 to 40 per

square meter (1–4 per square foot) of open area. Useful

equation: V = 4.043 (VP/d)

0.5

, where V = velocity in

m/s, VP = velocity pressure in mm H

O, and d =

density correction factor (unitless).

A2-7 Equipment Exhaust Ventilation

Specifications and Measurements

A2-7.1 Specifications for equipment exhaust should be

provided by the supplier and define:

A2-7.1.1 The control specification or standard for the

hood or enclosure, i.e., face velocity or capture velocity

if applicable.

A2-7.1.2 The airflow in the duct required to maintain

the control volume or flow required. Measurements

should be made using the ACGIH pitot traverse method

described below.

A2-7.1.3 The location where the Pitot traverse

measurement in the duct was made.

A2-7.1.4 Static pressure requirements.

A2-7.1.5 Coefficient of Entry (C

) (see definitions and

Section 22.3).

A2-7.1.6 Hood Loss Factor (K or F

) (see definitions

and Section 22.3).

A2-8 Duct Traverse Method

A2-8.1 Because the air flow in the cross-section of a

duct is not uniform, it is necessary to obtain an average

by measuring velocity pressure (VP) at points in a

number of equal areas in the cross-section. The usual

SEMI S2-0703a

method is to make two traverses across the diameter of

the duct at right angles to each other. Reading are taken

at the center of annular rings of equal area. Whenever

possible, the traverse should be made 7.5 duct

diameters downstream and 3 diameters upstream from

obstructions or directional changes such as an elbow,

hood, branch entry, etc. Where measurements are made

closer to disturbances, the results should be considered

subject to some doubt and checked against a second

location. If agreement within 10% of the two traverses

is obtained, reasonable accuracy can be assumed and

the average of the two readings used. Where the

variation exceeds 10%, a third location should be

selected and the two air flows in the best agreement

averaged and used. The use of a single centerline

reading for obtaining average velocity is a very coarse

approximation and is not recommended. If a traverse

cannot be done, then the centerline duct velocity should

be multiplied by 0.9 for a coarse estimate of actual

average duct velocity. Center line duct velocity should

not be used less than 5 duct diameters from an elbow,

junction, hood opening, or other source of turbulence.

A2-8.2 For ducts 150 mm (6 in.) and smaller, at least 6

traverse points should be used. For round ducts larger

than 150 mm (6 in.) diameter, at least 10 traverse points

should be employed. For very large ducts with wide

variation in velocity, 20 traverse points will increase the

precision of the air flow measurement.

A2-8.3 For square or rectangular ducts, the procedure

is to divide the cross-section into a number of equal

rectangular areas and measure the velocity pressure at

the center of each. The number of readings should not

be less than 16. Enough readings should be made so the

greatest distance between centers is less than 150 mm

(6 in.).

A2-8.4 The following data are required:

A2-8.4.1 The area of the duct at the traverse location.

A2-8.4.2 Velocity pressure at each point in the traverse

and/or average velocity and number of points measured.

A2-8.4.3 Temperature of the air stream at the time and

location of the traverse.

A2-8.4.4 The velocity pressure readings obtained are

converted to velocities, and the velocities (not the

velocity pressures) are averaged. Useful equation: V =

4.043 (VP/d)

0.5

, where V = velocity in m/s, VP =

velocity pressure in mm H

O, and d = density

correction factor (unitless). Some monitoring

instruments conduct this averaging internal to the

instrument.

A2-8.5 Flow measurement taken at other than standard

air temperatures should be corrected to standard

conditions (i.e., 21C [70F], 760 mm [29.92 in.] Hg).

SEMI S2-0703a

APPENDIX 3

DESIGN GUIDELINES FOR EQUIPMENT USING LIQUID CHEMICALS

— Design and Test Method Supplement Intended for Internal and

Third Party Evaluation Use

NOTICE: The material in this appendix is an official part of SEMI S2 and was approved by full letter ballot

procedures on December 15, 1999 by the North American Regional Standards Committee.

A3-1 Introduction

A3-1.1 This appendix provides specific technical

information relating to Section 23. In general, it

provides information on potential hazards,

recommended control methods, and design

considerations.

A3-1.2 This appendix is not intended to limit hazard

evaluation methods or control strategies (e.g., design

principles) employed by manufacturers. Alternative

methods are acceptable if they provide an equivalent

level of hazard control.

A3-1.3 This appendix is intended to be used as a

starting point for reference during equipment design.

An example would be during a formal hazard analysis

in a brainstorming session.

Table A3-1 Liquid Chemicals

Potential Hazard Recommended Control Method Design Considerations

Containment, control, and alarm

notification for spills, leaks or vapors.

Appropriately sized secondary containment (minimum 110%

volume of entire contents)

Equipment exhaust

Leak sensors to initiate auto shutdown.

Controlled access to chemical

containment areas.

Door/access cover interlocks that automatically depressurize the

area of the system being accessed.

Exposure to

operators

Control of access to point-of-

operation hazards.

Physical guarding/presence-sensing devices

Control of chemical delivery

pressure; control of residual

chemicals.

Depressurization upon system failure, interlock activation, or

normal shutdown

Transparent doors/covers allow visual inspection.

Exposure to

maintenance

personnel

Serviceability Built-in system purge and flush capabilities

System components accessible and easy to service.

Chemical resistance/compatibility

Appropriate materials used for equipment construction and

components.

General

equipment and

component

failure

Pressure rating

Pressurized systems designed to withstand 150% of maximum

foreseeable pressure, or provide a suitable relief valve.

Durable bulk chemical containers

Use of approved (e.g., DOT, UN Dangerous Goods) containers

in bulk distribution systems.

Control of pressurized vessels and

piping.

Provide visual pressure indicators with or without alarms.

Pressurized vessels and piping are designed and built to

recognized standards.

Spill control Automatic system pressure check prior to allowing dispense.

Use of normally closed valves on distribution lines.

Chemical

delivery system

leak

Drum change-out controls Over-fill sensors on chemical baths

Monitoring for excess flow.

Keyed and color-coded quick-connects