Dust Handling

Addressing the Hazards

of Data Omission
from MSDSs
James C. Mulligan If certain chemicals lack physical hazard data,
Lockheed Martin Corp.
suppliers must be proactive to keep workers safe.
Here’s how to determine dust flammability and
reactivity data for inclusion on MSDSs.

T HE U.S. OCCUPATIONAL SAFETY AND HEALTH to make MSDSs more useful references for process hazards analy-
Administration’s (OSHA’s) Hazard Communication sis (PHA).
Standard (HCS) (29 CFR 1910.1200) is intended to promote Companies where hazardous chemicals are used are mandated
safety in the workplace by providing workers with information by HCS to communicate hazard information to workers in four
about the physical and health hazards posed by the chemicals they ways. First, organizations must maintain a list of hazardous chem-
handle or to which they may otherwise be exposed (1). However, it icals that they use. Second, companies must maintain copies of the
does not encourage companies to develop physical hazard data — MSDS for each of these chemicals and make them readily acces-
including fire, explosion, and reactivity hazard data — beyond sible to workers in the work area where the chemicals are used.
those explicitly listed in ANSI Z400.1 (2), even when those data Third, companies must provide training to workers about physical
are likely required for safe manufacturing of the substance. If these and health hazards in the workplace, including: 1) explaining the
data are not developed by the manufacturer and are not otherwise hazards posed by hazardous chemicals; 2) instructing workers in
readily available from the scientific literature, they are unlikely to how to understand MSDSs; 3) describing methods for safe han-
be included on material safety data sheets (MSDSs). dling of hazardous chemicals; and 4) demonstrating suitable pro-
Thus, MSDSs do not often serve as a useful reference doc- tective equipment.
ument for conducting process hazards analysis as required Finally, when hazardous chemicals are transferred from ship-
under OSHA’s Process Safety Management (PSM) standard ping containers to smaller containers, these smaller containers
(29 CFR 1910.119) (3), including fire and explosion hazard must be properly labeled in accordance with the specific require-
assessments and chemical reaction hazard assessments, and ments of the standard.
under the U.S. Environmental Protection Agency’s (EPA’s)
Risk Management Plan (RMP) program (40 CFR 68) (4). This Addressing health and physical hazards
is clearly an inherently hazardous situation, the implications Despite the fact that the standard addresses both physical and
of which are discussed herein. health hazards in the workplace, OSHA and the regulated com-
This article discusses data that would be useful for companies munity have seemingly focused for the last 10 years on health
to include on MSDSs, including flammability and reactivity data. hazards under the HCS, including on MSDSs. For example,
In addition, OSHA and ANSI need to provide more explicit guid- many MSDSs include numerous pages of toxicological data and
ance as to which specific fire, explosion, and reactivity data should ecological data, i.e., environmental fate and transport. In con-
be included on MSDSs and protect companies providing such trast, MSDSs often contain relatively little data or information
data from liability due to the improper application or interpretation about fire and reactivity hazards posed by hazardous chemicals.
of these data. The objective of the additional guidance should be Unfortunately, the omission of such data is permissible under

36 www.cepmagazine.org April 2005 CEP

the HCS and simply suggests that the data were not readily avail- that may result in a hazardous situation and should be avoided,
able from the scientific literature. When fire and reactivity informa- such as heat, pressure, light, freezing, moisture, shock, or other
tion is included, it is often in the form of general warnings, such as, physical stresses.” ANSI also recommends that “undesirable con-
“Powder may form explosive dust-air mixture” or, “Reactivity: sta- ditions should also be specified with appropriate data if applicable,
ble.” These qualitative statements beg the question, “Under what e.g., auto-decomposition temperature...etc.”
conditions is the substance flammable, unstable or explosive?”
According to data compiled by the U.S. Chemical Safety and Interpretation of ANSI standards — too strict?
Hazard Investigation Board (CSB), the lack of data about the fire, The main problem with the ANSI guidance is that the regulat-
explosion and reactivity hazards posed by hazardous chemicals ed community appears to be strictly interpreting it, and not consid-
has lead to several recent fire and explosion incidents that have ering the need for data or warnings beyond those specifically stat-
caused 12 deaths and 79 injuries (5). These deficiencies in hazard ed in the guidance document, despite suggestions in the document
communication included: not providing customers with an to more broadly consider additional data and warnings as appro-
MSDS at all; not training personnel so that they are able to priate. Similarly, manufacturers may be inclined to omit other use-
understand MSDS data; not providing data in a format that work- ful physical hazards data that do explicitly correspond to one of
ers can understand; inclusion of incorrect data on the MSDS; and the ANSI subheadings.
lack of adequate data to describe the flammability and reactivity For example, ANSI Z400.1 lists “flash point” as one of the
of hazardous chemicals. physical hazard data that should be included on an MSDS. The
flash point is the temperature at which a sufficient concentration
Omission of data from MSDSs poses problems of vapor is evolved from a substance to form a flammable atmos-
Allowing companies to provide hazard data only when such phere in air at standard pressure. Flash points can be determined
data are “readily available” can be interpreted by unscrupulous using any number of standard methods, including American
companies as a disincentive to develop hazards data at all. That is, Society for Testing and Materials (ASTM) Methods D-56 (Tag
if hazard data are “not readily available from the scientific litera- Closed Cup), D-92 (Cleveland Open Cup), D-93 (Pensky-Martens
ture” and the company does not develop such data, no hazard data Closed Cup), and D-3278 (Setaflash).
will be available to include on an MSDS. The flash point is commonly thought of as applicable only
The lack of adequate data for assessing potential fire, explo- to liquids. As a result, companies often list “not applicable”
sion, and reactivity hazards is also a result of the regulated com- for flash point on the MSDS for bulk solids and powders. In
munity’s strict interpretation of the American National Standards this vein, it is crucial for companies to be aware of the fact
Institute’s (ANSI’s) guidance document on the preparation of that sublimable solids may also have a flash point. Those
MSDSs. ANSI Z400.1 addresses fire and reactivity hazards in four solids that do not have a flash point, (i.e., where the vapor
sections of the guidance document, which correspond to four sec- evolved from a solid is not flammable) may instead have a
tions of the model MSDS: flash ignition temperature (FIT) when tested in accordance
• Section 5: Fire Fighting Measures with ASTM Method 1929.
• Section 7: Handling and Storage It would be helpful for the purposes of PHA for such addi-
• Section 9: Physical and Chemical Properties tional data to be included on the MSDS. Further, the inclusion of
• Section 10: Stability and Reactivity these data would be consistent with the intent of the HCS to pro-
ANSI Z400.1 states that Section 5 of the MSDS should vide data that enable companies receiving hazardous chemicals
identify “qualitative flammable properties and reactivity haz- to design suitable programs for protecting workers. Simply put,
ards that enhance fire and explosion potential,” and cites the it accomplishes little to provide qualitative warnings that a
aforementioned sample warning: “Powder may form explosive chemical poses a physical hazard if data are not provided
dust-air mixture.” For Section 7, ANSI states that the MSDS describing the conditions under which it poses the hazard.
should include “handling practices, such as how to prevent
vapor release, the need for a totally-enclosed system, and Liability concerns inhibit MSDS data inclusion
other useful practices” and “statements about use of non- Companies likely refrain from including data beyond those
sparking tools, explosion-proof equipment and grounding specifically outlined in ANSI Z400.1 because of legal liability
when handling flammables.” concerns. That is, companies comply with the so-called “letter
For Section 9 of the MSDS, ANSI states that “the following of the law” by including only those categories of data in their
properties subheadings shall appear in this section, where appro- MSDSs that are explicitly listed in ANSI Z400.1. Companies
priate and relevant according to the countries or regions for are concerned that by providing data that appear to go beyond
which the MSDS is intended and into which the product is being what is explicitly outlined by the HCS, they may expose
supplied...flash point;...flammability (solid, gas); upper/lower themselves to liability if customers misinterpret or misuse the
flammability or explosive limits;...auto-ignition temperature; data. Companies are also concerned about the applicability of
decomposition temperature.” The standard suggests that its rec- data they include on MSDSs.
ommended list of physical and chemical properties is not all- For example, the explosibility of a combustible dust
inclusive, however. refers to its flammability in the air when it is dispersed in
Section 10 should indicate “if the material is stable or danger- the appropriate concentration. Explosibility of the dust par-
ously unstable under normal ambient temperature and pressure or ticles is a factor of their particle size and moisture content,
anticipated storage and handling conditions,” and “list conditions among other factors. A coarse bulk solid may not be

CEP April 2005 www.cepmagazine.org 37

Dust Handling

explosible at its nominal particle size, e.g., greater than and powders with relatively higher melting points can pro-
400–500 mm. However, dust finer than this size that is vide greater thermal insulation as layer thickness increases.
evolved may be explosible. Accordingly, the manufacturer As a result, less heat can be dissipated from the particles
of this bulk solid or powder may include a suitable warning closest to the heat source.
on the MSDS: “Dust evolved from bulk solid may form Should a supplier be liable for injuries and damages
explosive dust-air mixture.” incurred by a customer when providing in the MSDS the MIT
But what if the manufacturer reasonably interprets ANSI of a dust layer as determined in accordance with ASTM E-
Z400.1 and wishes to include data describing the conditions 2021 if the customer or its employees are harmed by a fire or
under which the dust is explosible for purposes of aiding cus- explosion during use of a product, even though the supplier
tomer PHA? For example, the ignition sensitivity of a dust could not reasonably foresee that the customer would allow
cloud is described by its minimum ignition energy (MIE) and accumulations of combustible dust exceeding a thickness of 5
minimum (auto-) ignition temperature (MIT or MAIT). mm on hot equipment? What obligation does the customer
The MIE describes the sensitivity of an explosible dust cloud have to reasonably interpret the MSDS data, i.e., to limit accu-
to ignition by electrical arcs and electrostatic discharges, and can mulations of combustible dust to less than 5 mm or preferably
be determined in accordance with ASTM Method E-2019. The eliminate them altogether.
MIT describes the sensitivity of an explosible dust cloud to igni-
tion from hot surfaces and can be determined in accordance with Additional data to consider for MSDSs
ASTM Method E-1491. Considered together, the MIE and MIT Questions may also arise regarding which data are the most
describe the sensitivity of a dust cloud to ignition from frictional useful and relevant to include on the MSDS to describe physical
and impact sparks, such as that which may occur in rotating or hazards. For example, ANSI Z400.1 indicates that AIT and
moving metal equipment in the event of maloperation or failure. decomposition temperature should be included in Section 9 of
These methods generally recommend that testing be per- an MSDS. For combustible dusts, does this mean the MIT of a
formed for a sample fraction having a gross particle size less than dust cloud and/or MIT of a dust layer, or another property —
75 µm (200 ASTM mesh), since this is the fraction most likely to namely the spontaneous ignition temperature (SIT), which can
remain suspended in the form of an explosible dust cloud. Thus, a be determined, along with FIT, in accordance with ASTM
manufacturer may be inclined to include the MIE and MIT for Method D-1929?
<75-mm dust on the MSDS for a combustible bulk solid. SIT is the temperature at which approximately 3 g of a solid
However, what if the customer grinds or mills the solid? The substance will glow or ignite solely in response to the prevailing
MIE and MIT of a dust cloud generally decrease as particle size environmental temperature. The method does not distinguish,
decreases, so the MIE and MIT may be lower during and after however, between ignition of the solid sample and ignition of
grinding or milling. Consequently, additional precautions may be flammable vapor that may be evolved. Further, even the SIT, MIT
required to minimize the risk of fire and explosion during these of a dust cloud, and MIT of a dust layer may not represent the
operations beyond those that may be suggested by the MSDS data. lowest hot-surface temperature capable of posing a hazard to a
The MIT of a dust cloud is analogous to the autoignition substance.
temperature (AIT) for a gas or vapor, which is presently one Often, the maximum safe exposure temperature for a substance
of the physical hazard data recommended by ANSI Z400.1. should be based on its self-accelerating decomposition tempera-
While the AIT of a gas or vapor evolved from a combustible ture (SADT) or self-heating onset temperature (To) — i.e., the
liquid is sometimes included on the MSDS, the MIT of a temperature at which self-heating commences. These are likely
dust cloud rarely is. Perhaps this stems from the fact that it what ANSI means when it refers to “decomposition temperature”
could be confused with the MIT of a dust layer, which can in Section 9 of ANSI Z400.1. The SADT and To are important to
be determined in accordance with ASTM E-2021. The MIT know for PHA purposes because they represent the temperatures
of a dust layer describes the sensitivity of a 5-mm dust layer at which substances will continue to decompose or self-heat
to ignition from hot surfaces. respectively, even if the source of heat that initiated these reactions
The inclusion of both the MIT of a dust cloud and MIT is subsequently removed.
of a dust layer would be useful for a PHA. For example, the Self-heating occurs when the rate of heat generated by oxida-
maximum safe exposure temperature for a substance should tion exceeds the rate at which a bulk solid or powder can dissipate
in some applications be based on the lower of the MIT of a heat to the environment, even in the absence of an increase in the
dust cloud and MIT of a dust layer, less a suitable safety ambient temperature. This increases the temperature of the sub-
factor. Specifically, it is generally recommended that sub- stance and, as a result, increases the rate of oxidation. The reaction
stances not be exposed to temperatures exceeding two-thirds rate can increase by two- or three-fold for every 10°C increase in
of their MIT (or AIT). the reaction temperature (6). Ultimately, the temperature of the
However, in the same way that the MIT of a dust cloud substance can exceed its AIT or MIT, causing it to ignite and burn.
is a factor of the particle size and moisture content of the The period for which a bulk solid or powder must be
dust, the MIT of a dust layer is a factor of the thickness of exposed to its To for self-heating to commence is known as
the dust layer on a surface. While ASTM E-2021 recom- the induction time. Self-heating of bulk solids and powders is
mends testing a 5-mm-thick (100-mm dia.) dust layer, it is a form of relatively slow combustion (oxidation) that can
known that ignition temperature generally decreases as occur over hours, days or even weeks. In contrast, the MIT
layer thickness increases. This is because many bulk solids describes the temperatures at which a dust evolved from such

38 www.cepmagazine.org April 2005 CEP

bulk solids and powders will ignite instantaneously for MIT of
Table. Useful fire, explosion, and reactivity data for
a dust cloud, and within 30 min for MIT of a dust layer.
material safety data sheets.
Like the MIT of a dust layer, the To for a bulk solid or powder
generally decreases as the amount of material increases, owing
again to an increase in thermal insulation. Thus, unlike melting Combustible Gases and Liquids (Vapor)
point and other physical hazard data recommended for inclusion • Flash point
on the MSDS, the To and MIT of a dust layer are not inherent • Flammable limits (LFL and UFL)
• Minimum ignition energy (MIE)
properties of a substance. Rather, they are factors of the quantity
• Autoignition temperature (AIT)
and configuration of the substance, among other factors.
• Self-accelerating decomposition temperature (SADT)
The To for bulk solids and powders can be estimated using • Maximum explosion (deflagration) pressure (Pmax)
differential scanning calorimetry (DSC) and ASTM Method E- • Explosion severity index (Kg)
2009. It can also be estimated using larger-scale diffusion cell • Limiting oxidant concentration (LOC)
tests in accordance with methods developed by Beever and
Thorne and published by the Institution of Chemical Engineers Combustible Bulk Solids and Powders (Dusts)
(IChemE) (7) or by isothermal basket testing based on a tech- • Flash ignition temperature (FIT)
nique developed by Bowes (8). • Minimum explosible concentration (MEC)
Whereas To describes the temperature at which the combus- • MIE of a dust cloud
tion reaction will irrevocably commence within a bulk solid or • MIE of a dust layer
powder, the SADT describes the temperature at which the • Minimum ignition temperature (MIT) of a dust cloud
decomposition reaction will irrevocably commence. Like To, • MIT of a dust layer
SADT is not an inherent property of the substance, but rather a • Spontaneous ignition temperature (SIT)
• Self-heating onset temperature (To)
factor of the solid’s quantity and configuration, as well as the
• Self-accelerating decomposition temperature (SADT)
heat transfer properties of the package or container in which the • Pmax
solid is contained, among other factors. The SADT for sub- • Explosion severity index (Kst)
stances can be determined using the adiabatic Dewar calorimeter • LOC
in accordance with United Nations (UN) Test Method H.2 (9).
Since “auto-ignition temperature” may mean the AIT, MIT
of a dust cloud, MIT of a dust layer, or SIT, and “decomposi- While these lists are not inclusive, including these and
tion temperature” may mean self-heating To, SADT or some other fire, explosion, and reactivity data on MSDSs would
other decomposition temperature, a supplier could consider promote a better understanding of the hazards posed by a
including any or all of these data on the MSDS for the sub- substance, facilitate PHA, and thereby enable companies to
stance. Other useful fire, explosion, and reactivity data are dis- design more effective and comprehensive programs for pro-
played in the Table. tecting workers from those hazards. CEP

Literature Cited JAMES C. MULLIGAN is a senior systems safety engineer with

Lockheed Martin Corp.’s Systems and Sensors Div. (199 Borton
1. U.S. Code of Federal Regulations, Hazard Communication, Vol. 29, Landing Road; Mail Stop 127-307; Moorestown, NJ 08057; Phone:
Part. 1910, Subpart. 1200. (856) 722-2539; Fax: (856) 273-5569; E-mail: james.c.mulligan@
2. American National Standards Institute (ANSI), “American lmco.com). He has been working in the fields of process safety and
National Standard for Hazardous Industrial Chemicals (ANSI) environmental management for over 18 years, specializing in fire
Z400.1; Material Safety Data Sheets – Preparation,” Section 5.1, and explosion hazard assessment, incident investigation, and
Washington, DC (2003). prevention and protection. Mulligan is a former technical manager
3. U.S. Code of Federal Regulations, Process Safety Management of with the Chemical Manufacturers Assn. (now American Chemistry
Highly Hazardous Chemicals, Vol. 29, Part. 1910, Subpart. 119. Council) in Washington, DC. He has authored testimony given
4. U.S. Code of Federal Regulations, Chemical Accident Prevention, before Congress and represented the chemical industry before U.S.
Vol. 40, Part. 68. government agencies during the development of numerous safety
5. Merritt, C. W., Testimony Before the Subcommittee on Employment, and environmental laws and regulations. Mulligan has written
Safety, and Training, U.S. Chemical Safety and Hazards numerous articles on chemical process safety, is a regular speaker
Investigation Board, Committee on Health, Education, Labor, and at industry training courses and events, and is a member of AIChE,
Pensions – U.S. Senate; Washington, DC (March 25, 2004). NFPA, ASSE, and ASTM. He received a bachelor’s degree in
6. Expert Commission on Safety in the Swiss Chemical Industry chemical engineering degree from the Catholic Univ. of America
(ESCIS), “Thermal Process Safety — Data, Assessment Criteria, and and a master’s degree in business administration from Villanova
Measures,” Booklet 8, Lucerne, Switzerland, p. 8 (1993). Univ.
7. Abbott, J. A., Prevention of Fires and Explosions in Dryers — A User
Guide, 2nd Ed., IChemE, Rugby, U.K. (1991).
8. Bowes, P. C., Self-Heating: Understanding and Controlling the
Hazards, Elsevier, Amsterdam, The Netherlands (1984). Acknowledgments
9. United Nations, ST/SG/AC.10/11/Rev. 4, “Recommendations on the
Transport of Dangerous Goods — Tests and Criteria,” 4th Ed., United The author gratefully acknowledge the valuable contributions of
Nations, Geneva (2003). Ms. Rita D’Aquino to this article. He also wishes to thank his wife
and family, who have helped make all of his achievements possible.

CEP April 2005 www.cepmagazine.org 39