INTRODUCTION TO MODELS AND

IH MOD, A NEW TOOL TO AID

INDUSTRIAL HYGIENISTS IN

EXPOSURE ASSESSMENTS Tom Armstrong, CIH, PhD

TWA8HR Occupational Hygiene Consulting, LLC

WWW.TWA8HR.COM

With contributions by many members of the

Exposure Assessment Strategies Committee

and MAJOR help on IH Mod from Daniel Drolet

UNDERSTANDING EXPOSURES AND HEALTH RISKS,

PAST, PRESENT AND FUTURE

1

WE WILL COVER THE

FOLLOWING ASPECTS

Brief theory of modeling to estimate exposures

What IH Mod is

Contents and layout of IH Mod

Starting and Navigating in IH Mod

Entering parameter values in IH Mod

Interpreting the results from IH Mod

Examples of what IH Mod can do

Workshop for you to try IH Mod

Tom

Arm

stron

g, W

WW

.TW

A8

HR

.CO

M

2

WHAT IS A MODEL AND WHY ARE THEY

IMPORTANT IN EXPOSURE ASSESSMENTS?

In an IH context, models are mathematical equations

that can estimate the concentration of a contaminant

in a work space based on physical and chemical input

parameters

Box models are one category of models and are

developed in several levels of complexity

The one compartment, “well mixed room” box model

is where we will start in this introduction to modeling

for exposure assessments

We will then cover more complex box models

There are also diffusion models, plume models and

generation rate models that we will briefly discuss

To

m A

rm

str

on

g, W

WW

.TW

A8H

R.C

OM

3

WHAT ARE THE MATHEMATICAL ASPECTS

OF THE ONE COMPARTMENT BOX MODEL?

Mass balance is a major part of it! Mass accumulated + Mass out = Mass generated + Mass in

See the book Mathematical Models for Estimating Occupational

Exposure to Chemicals for the differential equations and solutions

Contaminant

source (G)

Air In* Air & Contaminant Out Well-Mixed Air Inside

Contaminants released inside the box are

completely and instantaneous well-mixed. * The air in could bring some contaminant with it

Tom Armstrong, WWW.TWA8HR.COM

4

DID YOU REALLY MEAN…COMPLETELY AND

INSTANTANEOUS WELL-MIXED?!

In this model, YES!

Any model we use makes simplifying

assumptions (necessary for mathematical

tractability and we hope appropriate!)

More complex models may make fewer

assumptions but then they require more input

parameters

The art of modeling hinges on choosing a model

that is simple enough to be tractable yet

adequate to represent the conditions of the

scenario

To

m A

rm

str

on

g, W

WW

.TW

A8H

R.C

OM

5

A CONTAMINANT GENERATION RATE TERM G IS

COMMON TO MANY MODELS. WHAT IS G AND HOW

DO WE GET IT?

SEE Chapter 3 of the Book

“Mathematical Models for Estimating Occupational Exposure to Chemicals“

More details requires more time than we have today.

We (AIHA) have a 2 day course on Mathematical

Modeling that covers this “G” in more detail.

HOWEVER…we will give you a few hints about G

Remember, the book will tell you much more!

To

m A

rm

str

on

g, W

WW

.TW

A8H

R.C

OM

6

A FEW GENERATION (G) MECHANISMS…

Vapor displaced from a drum or container being filled

A spill of a chemical and its vaporization to air

Spray of chemical from a container (e.g. aerosol can)

Standard estimates of leaks from valves and flanges,

etc.

There are many other situations and mechanisms for

G

Units of measure are important. G is in mass per time

and in IH Mod, milligrams per minute

Particulates are particularly tough

To

m A

rm

str

on

g, W

WW

.TW

A8H

R.C

OM

7

You Should Review Both of These EASC

Publications (AIHA Press) to START (or

Continue) a Successful Career in Modeling

To

m A

rm

str

on

g, W

WW

.TW

A8H

R.C

OM

8

HERE’S THE WORK PLAN FOR THE

REMAINDER OF THIS SESSION

We will have a quick overview of loading and “activating” IH Mod, a spreadsheet of physical chemical exposure models

We will take a quick look at the IH Mod “README” file

We will take a quick look at the structure of the IH Mod spreadsheet, including the navigation controls.

We will look at a task exposure scenario with IH Mod

I will run IHMOD for you in a demonstration, primarily for the “Well-Mixed Box” and “Two-Zone” Models and look at a few other models

We will have open discussion of what we learned, and what comes next!

To

m A

rm

str

on

g, W

WW

.TW

A8H

R.C

OM

9

IHMOD INTRODUCTION

Disclaimer

Structure

Help System & User Documentation Screen

Data Entry / Data Sliders

Graphics

Printing Options

Parameter definition

Unit conversions

NOTE: the book is ESSENTIAL to REALLY use IH Mod with understanding!

To

m A

rm

str

on

g, W

WW

.TW

A8H

R.C

OM

10

IH MOD CONTAINS ELEVEN OF THE MODELS COVERED IN

THE TEXT “MATHEMATICAL MODELS FOR ESTIMATING

OCCUPATIONAL EXPOSURE TO CHEMICALS“

We will not cover all of them today

The text does a great job describing the

models and their uses

IH Mod includes short, specific HELP on

each model included

To

m A

rm

str

on

g, W

WW

.TW

A8H

R.C

OM

11

THERE ARE MULTIPLE MODEL PARAMETERS – THESE ARE

THE MOST PREVALENT (BESIDES G)

– BUT THE LIST VARIES BY MODEL!

V = “room” volume

Q = room air supply /exhaust ventilation rate

Alpha = an evaporation rate constant

Csat = saturation vapor concentration for the

contaminant

Mo = initial contaminant mass

Dt = turbulent diffusion coefficient

U = advective air speed

Beta = air exchange rate between zone one and zone

two (two zone model)

To

m A

rm

str

on

g, W

WW

.TW

A8H

R.C

OM

12

WHAT IS IN IH MOD?

WE WILL NOW RUN THROUGH

THE SETUP AND STRUCTURE WE WILL RUN A DEMO DURING THIS WEBINAR

AND WE WILL GET YOU TO THE POINT OF DOING

THAT ON YOUR OWN!

Some Example Screen Shots Follow in the PowerPoint Slides

(slight difference may appear with new versions of IH Mod)

To

m A

rm

str

on

g, W

WW

.TW

A8H

R.C

OM

13

BUT FIRST… IF YOU HAVE A PC AND A

COPY OF IH MOD ….

Can you enable macros on your PC? Windows XP, Excel 2003?

Windows 7, Excel 2007, Excel 2010

Windows 8?

Procedures differ! “Google” Enable Macros for your particular combination, or follow “Help” Links in IH Mod (I’ll show that in a following slide)

Those of you with PCs from some “locked down” corporation or other computer security sensitive organization MAY be OUT OF LUCK for now. Someone with “Administrative Rights” to your machine might be needed to “enable macros”.

To

m A

rm

str

on

g, W

WW

.TW

A8H

R.C

OM

14

START IH MOD IF

YOU WISH TO TRY IT

… WHILE

FOLLOWING

ALONG

To

m A

rm

str

on

g, W

WW

.TW

A8H

R.C

OM

15

THIS IS “THE FIRST SCREEN” IN IH MOD

… AFTER MACROS WERE ENABLED!

Tom

Arm

stron

g, W

WW

.TW

A8

HR

.CO

M

16

THE “CALLOUTS” ADDED TO THIS SCREEN CAPTURE

POINT TO A FEW IMPORTANT CONTROLS AND

FEATURES

To “General

Help and User

Input”

To “Help” re

“Enable Macros”

On right side of

General Help

Screen

We Force a

Visit to the

Disclaimer PER

SESSION

Scrolls to

Choose the

Model You

Want to

Run

Tom

Arm

stron

g, W

WW

.TW

A8

HR

.CO

M

17

PLEASE NOTE THE DISCLAIMER FOR THIS

FREEWARE PRODUCT…

DISCLAIMER: The material embodied in this software is provided

“as-is” and without warranty of any kind, expressed, implied, or

otherwise, including without limitation any warranty of

merchantability of fitness for a particular purpose.

In no event shall Thomas W. Armstrong, Daniel Drolet or the

American Industrial Hygiene Association (AIHA) be liable for any

direct, indirect, special, incidental, or consequential damages of any

kind, or any damages whatsoever, including without limitation loss

of profit, loss of use, savings or revenue, or the claims of third

parties, whether or not Thomas W. Armstrong, Daniel Drolet or the

AIHA has been advised of the possibility of such loss, however

caused, and on any theory of liability, arising out of or in connection

with the possession, use, or performance of this software.

18

Tom

Arm

stron

g, W

WW

.TW

A8

HR

.CO

M

THIS IS THE USER INPUT AND

GENERAL HELP SCREEN – TOP PART

Tom

Arm

stron

g, W

WW

.TW

A8

HR

.CO

M

19

IH MOD’S NAVIGATION CONTROLS T

om

Arm

stron

g, W

WW

.TW

A8

HR

.CO

M

20

USER SCREEN AND GENERAL

HELP – BOTTOM PART

Tom

Arm

stron

g, W

WW

.TW

A8

HR

.CO

M

21

THIS IS THE START OF THE

“README” FILE

To

m A

rm

str

on

g, W

WW

.TW

A8

HR

.CO

M

22

HERE IS AN EXAMPLE OF THE WELL MIXED ROOM MODEL!

To “Help” FOR

THIS MODEL!

Red

Mandatory!

Black Optional

Data Table and

User Calculations

Below Graph

Correct UNITS

are CRUCIAL!

Here is a

CONVERSION

Calculator

Tom Armstrong, WWW.TWA8HR.COM

23

THE WELL MIXED ROOM (ONE COMPARTMENT BOX

MODEL) – HELP SCREEN FROM IH MOD

To “Get Back”

to the Model in

Use

To

m A

rm

str

on

g, W

WW

.TW

A8H

R.C

OM

24

Here We Are Back in the Well-Mixed Room

Model

Sliders to

change

parameter

values

(Check

Positions!)

Concentration

at time T

User

“Inputs” go

in the

Green Cells

Results in

mg/M3!

CONVERT

if you need

PPM

TWA from T0 to Tx

Tom Armstrong, WWW.TWA8HR.COM

25

LET’S TAKE A LOOK AT THE INPUT

PARAMETERS

Question. If you have air changes per hour, how do you use that?

HINT: it involves the room volume and hours to minutes conversion.

Tom

Arm

stron

g, W

WW

.TW

A8

HR

.CO

M

26

THIS IS THE HELP SCREEN FOR THE TWO ZONE MODEL

(ANOTHER VERY USEFUL BOX MODEL) T

om

Arm

str

on

g, W

WW

.TW

A8H

R.C

OM

27

NF-FF (AKA 2 ZONE)

Note Short

Simulation Time and

Short Generation

Period To

m A

rm

str

on

g, W

WW

.TW

A8H

R.C

OM

28

IN THE NF-FF (AKA TWO ZONE)

MODEL, WHAT IS BETA (B)?

It is the Air Exchange Between the Near

Field and the Far Field

Beta = ½ FSA*s

FSA= Free Surface Area

S = random air velocity in m/min

For a sphere Beta = 2π(r2)s

More on this is available in the 2nd edition

math modeling book

To

m A

rm

str

on

g, W

WW

.TW

A8H

R.C

OM

29

LET’S QUICKLY LOOK AT A FEW

MORE OF THE MODELS IN

IH MOD

To

m A

rm

str

on

g, W

WW

.TW

A8H

R.C

OM

30

IH MOD SMALL SPILL MODEL

The HELP screen lists

IMPORTANT

constraints!

Tom

Arm

stron

g, W

WW

.TW

A8

HR

.CO

M

31

IH MOD WMR EXPONENTIAL

DECREASING EMISSION RATE

Tom

Arm

stron

g, W

WW

.TW

A8

HR

.CO

M

32

SCENARIO #1 - DRUMMING

Your boss asks for your judgment about exposures at a facility your company recently acquired*, and which you have not visited.

They drum mixed isomer xylene 55 gal (about 200 L), 10 drums per hour, all day, top fill via bung opening = 2000 L/hr = 2 M3/hr = 0.033 M3/min

General ventilation with some LEV, but there are complaints of strong xylene odor in the area

Density of mixed xylenes 0.87

TLV 100 PPM MW 106 so about 434 mg/m3

VP 7 Torr at 20 C (typical temp)

Csat = (VP*10^6)/Atmospheric pressure = 7000000/760 = 9210 ppm = 40000 mg/M3

* This is a “hypothetical” example and is not drawn from any known actual situation.s

To

m A

rm

str

on

g, W

WW

.TW

A8H

R.C

OM

33

Process Operator #1

Process Operator #2

Process Engineer

THIS IS AN ILLUSTRATION OF ONE TYPE

OF DRUM FILLING INSTALLATION

34

Tom

Arm

stron

g, W

WW

.TW

A8

HR

.CO

M

WHAT GOES ON IN DRUM FILLING?

What is the vapor displacement

rate and concentration?

Top “splash” loading

generates higher

concentration vapor

more quickly than

submerged filling.

Could there be a

difference in vapor

concentration out for

a new drum

versus one with a “heal”

of product?

Tom

Arm

stron

g, W

WW

.TW

A8

HR

.CO

M

35

GEE, WHAT IS G FOR THIS

SCENARIO?!

The Volume of vapor per time displaced from the drums is 0.033 M3/min

Csat in mg/M3 = 40000 mg/M3

See Chapter 3 for Container filling

(vol/time)*(mass/vol) = mass/time

Submerged filling of NEW drums, can reduce by up to 0.5. (What about drums shipped back for refill that have residue in them?)

Top splash can generate aerosol and vapor, so up to 1.45 of saturation vapor concentration*

Using a “factor” of 1 … 40000 mg/M3 x 0.033 M3/min = 1320 mg/min = G

* See Chapter 3 of “Mathematical Models for Estimating Occupational Exposure to Chemicals”

Tom Armstrong, WWW.TWA8HR.COM

36

YOU MAKE A PHONE CALL AND LEARN SOME

FACTS ABOUT THE DRUM FILLING STATION

Room volume about 600 m3

General room exhaust ventilation not

great, about 20 m3/min (2 ACH)

We already knew enough to estimate

the contaminant generation rate of

1320 mg/minute

To

m A

rm

str

on

g, W

WW

.TW

A8H

R.C

OM

37

LOOKING TO IH MOD … T

om

Arm

str

on

g, W

WW

.TW

A8H

R.C

OM

38

WHAT DO WE CONCLUDE?

Hint, TLV mixed Xylenes TWA = 100 ppm = 434 mg/M3

What if the fill rate was wrong and is actually 60 drums an hour? Easy, just change G to 6*1320 mg/min = 7920 mg/min

To

m A

rm

str

on

g, W

WW

.TW

A8H

R.C

OM

39

WHAT IF WE US THE TWO-ZONE MODEL? T

om

Arm

stron

g, W

WW

.TW

A8

HR

.CO

M

40

FOR THE TWO ZONE MODEL WE

NEED A FEW MORE PARAMETERS

Near Field Volume = a “virtual” volume of air

around the worker, smaller than the total room

volume

G remains as for the well mixed room

Beta is an important new parameter

Beta = ½ FSA*S

FSA = free surface area of the near field volume

S= the random air exchange between the near field and

the far field

Tom

Arm

stron

g, W

WW

.TW

A8

HR

.CO

M

41

HERE’S AN ESTIMATE OF BETA

FOR OUR BARREL FILL SCENARIO

We will assume “arm length” of the worker from the source,

and that is about 0.8 Meter, as radius of the sphere.

Vnear field = 2.1 cubic meter (WHY? V sphere = 4/3 pi r3)

Free Surface Area = the surface area of a sphere of volume

chosen, Surface Area = 4 pi r 2 = 8 m2

For S, default values based on studies of average air speeds in

workplaces are:

Absent strong sources of air motion near the source:

s = 3 to 4.5 m/min. We will assume 3 m/min

Given strong sources of air motion near the source:

s = 7.6 m/min

Then Beta = 1/2FSA*S = 12

Tom

Arm

stron

g, W

WW

.TW

A8

HR

.CO

M

42

WHAT DO WE CONCLUDE NOW??

The two zone model (undoubtedly correctly)

predicts a higher exposure to the worker

1. What if a second fill station is installed and

the production rate doubled with no other

changes?

2. What if the actual fill rate was 60 drums per

hour?

3. What if the material drummed had the same

Vapor Pressure but a the TLV was 5 mg/M3 ?

4. What if ? Modeling is a great tool for “what

if” analyses!

To

m A

rm

str

on

g, W

WW

.TW

A8H

R.C

OM

43

HOW DO WE ESTIMATE THE 95TH

PERCENTILE FROM MODEL RESULTS?

This is a GOOD question, one just recently

asked.

Research on this needs to be done.

For now:

If you have estimates of the parameter

distributions, it is easy to get the 95th percentile

from Monte Carlo simulation results

Lacking that, use the “rule of thumb” 2 to 4

times the mean and assume the model results are a

mean

Tom

Arm

stron

g, W

WW

.TW

A8

HR

.CO

M

44

LET US THINK ABOUT BETA FOR

ANOTHER MINUTE OR TWO

o A worker applies glue to objects on an

assembly line

o The work is a bit closer than arm’s length

o The workroom air shows no noticeable

directional draft

o What do we do to estimate a value for

Beta?

o Hint: Beta = ½ FSA*S

o How much of a spherical area does the

assembly line block? A full half? None?

Tom

Arm

stron

g, W

WW

.TW

A8

HR

.CO

M

45

HERE IS ANOTHER MODEL

YOU MAY FIND USEFUL

Tom

Arm

stron

g, W

WW

.TW

A8

HR

.CO

M

46

This model uses some different units of measure, so be

careful about conversions and input units.

SMALL SPILL MODEL HELP

Tom

Arm

stron

g, W

WW

.TW

A8

HR

.CO

M

47

AND HERE ARE TWO OTHER MODELS

YOU MAY FIND A USE FOR!

What does the HELP file say? Check “THE BOOK” for more details.

Tom

Arm

stron

g, W

WW

.TW

A8

HR

.CO

M

48

HERE ARE A FEW IMPORTANT

REMINDERS!

Watch the units!

IH Mod gives Mg/M3 not PPM … CAN YOU CONVERT?

Ventilation input is Q not ACH … CAN YOU CONVERT?

CAUTIONS?

Check the Input Value “slider” positions

Remember to clear prior results

Remember the difference between Ct and TWAt

DOCUMENT your thinking! You can use the “User Input”

screen and Save the file with a new name

What are your assumptions?

Where or how did you get your input parameter values

Remember the model’s stated limitations. Can you “live”

with them?

Tom

Arm

stron

g, W

WW

.TW

A8

HR

.CO

M

49

MORE RESOURCES! PDCS OFTEN OFFERED AT THE AMERICAN INDUSTRIAL HYGIENE CONFERENCE AND EXHIBITION

Exposure Assessment Strategies and Statistics (2-Day)

Bayesian Statistics: Overview and Applications in IH Data

Interpretation and Exposure Risk Assessment (1 Day)

Professional Judgment in Exposure Assessment (2 Day)

Dermal Exposure Assessment & Modeling (1 Day)

Using Mathematical Models to Estimate Exposure (IHMOD) (2 Day)

Advanced Excel for Industrial Hygiene Calculations

(1 Day)

NEW AIHCE 2013 Monte Carlo Techniques in Exposure and Risk Assessment (1

day)

NEW! IH Mod Discussion BLOG! www.ihmod.org

To

m A

rm

str

on

g, W

WW

.TW

A8H

R.C

OM

50

MORE RESOURCES!

Check the AIHA Exposure Assessment

Strategies Committee Website for

updated tools!

Link to that website

At that page, you will find links for

IH STAT

IH MOD

IH SkinPerm

To

m A

rm

str

on

g, W

WW

.TW

A8H

R.C

OM

51

RESOURCES! T

om

Arm

str

on

g,

WW

W.T

WA

8H

R.C

OM

52

THANK YOU!!!

QUESTIONS?

DISCUSSION? Tom Armstrong, CIH, PhD

53

TWA8HR Occupational Hygiene Consulting, LLC

“Understanding Exposures and Health Risks, Past, Present and Future”

WWW.TWA8HR.COM