Dispersion Modeling of Commodity and Structural Fumigation - - PowerPoint PPT Presentation

1

Dispersion Modeling of Commodity and Structural Fumigation Applications

Rick Reiss, Exponent Presented at Kansas State University May 12th, 2010

2

Overview of Presentation

• Risk assessment process for bystander exposure to

fumigants

• Use of dispersion modeling to estimate downwind

concentrations

• Example studies with methyl bromide used to

characterize emissions

3

Background

• Fumigants are generally highly volatile
• Emissions after treatment can lead to downwind exposures to

bystanders

• Regulators are interested in minimizing exposures
• The solution was to establish buffer zones around applications,

which restrict entry for a period of time after the application

• EPA recently established national buffer zones for most

fumigants

• The principal tool used by EPA was PERFUM

4

4

Hazard Identification Hazard Identification

Can the substance cause Can the substance cause illness or disease? illness or disease?

Dose Dose-Response Response Assessment Assessment

What dose is necessary? What dose is necessary?

Exposure Assessment Exposure Assessment

What levels are people What levels are people exposed? exposed?

Risk Characterization Risk Characterization

What is the risk of disease? What is the risk of disease?

Source: Modified from NAS 1983 (pg. 21)

The Risk Assessment Process

5

Basis for Buffer Zone Estimation

• Use of air dispersion modeling to estimate downwind

concentrations over range of meteorological conditions

• Comparison of concentration estimates with toxicity

reference concentrations to estimate risk

6

Possible Buffer Zone Definitions

MOE < 100 MOE > 100 5% Maximum concentration buffer zone Whole field buffer zone

7

Factors Influencing Downwind Concentrations

• Source-specific
• Application rate
• Treatment and aeration length
• Air exchange rates
• Volatility of fumigant
• Sealing
• Location-specific
• Meteorological conditions
• Wind speed, wind direction, atmospheric stability
• Terrain
• Nearby buildings (downwash)

8

Dispersion Modeling Theory

9

Building Downwash

10

Building Downwash – Another Look

11

Steps in Modeling Analysis

• Estimating emission rates using real-world field data
• Characterization of source of interest
• Application rate
• Length of treatment and aeration
• Air exchange rates
• Estimating range of downwind concentrations using

historical meteorological datasets

12

General Methodological Options for Emissions

• Option #1: Assume a percent release during treatment

and aeration

• Option #2: Assume an air exchange rate during

treatment and aeration and use it to calculate the hourly emissions.

13

V=volume C(t)=concentration

Schematic of Building Air Flow – Ventilation Model

Q=air flow Q Air Exchange Rate (R) = Q/V

14

1

Dose Response

• Traditional Approach with Animal Studies
• Exposure animals for a range of doses and measure

chemical-related effects

• Determine the No Observed Effect Level (NOEL)
• Apply a 100X uncertainty factor to NOEL to determine the

“safe” dose for risk assessment

• Includes 10X uncertainty factor to account for uncertainty in

extrapolating between animals and humans

• Includes a 10X uncertainty factor to account for variation in

the human population

15

Dose-response for methyl bromide

• NOAEL of 40 ppm based on a rabbit study exposures

at Days 6-16 of gestation

• Agenesis (failure to develop) of the gall bladder
• Fused sternebrae
• Uncertainty factor of 30
• Reference concentration = 1.3 ppm (1300 ppb) over 4

hours

16

Using Field Studies to Estimate Emissions – Dispersion Modeling in Reverse

17

Review of Methyl Bromide Historical Studies (early 1990s)

Location Location Volume Volume (ft (ft3) Treatment Treatment Period (hr) Period (hr) Aeration Aeration Period (hr) Period (hr) Watsonville Watsonville 1065 1065 0.2 0.2-0.5 0.5 0.2 0.2-0.5 0.5 Bakersfield Bakersfield 18,290 18,290 2 2 Madera Madera 320,000 320,000 90 90 2 Stanislaus Stanislaus 1,450,000 1,450,000 24 24 24 24 Sutter Sutter 3,100,000 & 3,100,000 & 6,800,000 6,800,000 24 24 24 24

18

Loss Rates During Treatment in Mid-Sized Warehouse in Madera

C(t) = 7710 ppm exp(-0.05 h-1 t) R2 = 0.99 1 10 100 1000 10000 20 40 60 80 100 Time (hr) Building Concentration (ppm)

19

Loss Rates During Aeration in Large Processing Plant in Stanislaus

y = 3979 ppm exp(-1.7 h-1 t) R

2 = 0.90

1000 10000 10 20 30 40

Time (mins) Exhaust Stack Concentration (ppm)

20

Air Exchange Rates from Historical Studies

Study Study Scenario Scenario Duration Duration (hrs) (hrs) ACH (hr) ACH (hr) Watsonville Watsonville A 0.17 0.17 0.01 0.01 A 0.28 0.28 0.04 0.04 Bakersfield Bakersfield A 2 0.01 0.01 Madera Madera T 90 90 14 14 A 2 0.9 0.9 Stanislaus Stanislaus T 24 24 23 23 A 24 24 0.4 0.4 Sutter County Sutter County T 24 24 41 41 A 24 24 0.3 0.3-0.7 0.7

21

Characteristics of New Methyl Bromide Studies

Site A Site B Site C Application Rate (lbs/1000 ft3) (nominal) 1.0 1.0 1.0 Initial Concentration 24 g/m3 6000 ppm 30 g/m3 8000 ppm 18 g/m3 5000 ppm Change in Concentration during Treatment

• 60%
• 50%
• 65%

Duration to aerate 50% of fumigants 0.5 h (passive) 3.0 h (active and passive) 1.0 h (active)

22

23

24

Emission Rates for New Studies

y = 5550 e-0.05x R2 = 0.98 1000 10000 5 10 15 20 25 30 Time from Treatment (h) Methyl Bromide Concentration (ppm)

Building air- exchange rate is defined by the rate of change in fumigant concentration.

25

Emission Rate Estimates for New Studies

Site A Site B Site C

Treatment Air-Exchange Rate 14-35 hr 17 hr 17 hr Fumigant mass loss 44 to 70% over 24 hours 55% over 20 hours 62% over 24 hours Aeration Air-Exchange Rate 0.5-3.5 hr (passive) 0.3-2.8 hr(both active and passive) 0.7 hr (active

• nly)

Fumigant mass loss 18 to 78% in 1 hours 22 to 86% in 1 hour 63% in 1 hour

26

Summary of Measured Air Exchange Rates

• Treatment
• Range of 14-41 hr, consistently across studies
• Length between 10 minutes and 90 hours
• Often lose >50% of mass, by 3-5% per hour
• Aeration
• Range of 0.01-3 hr
• Larger rates for smaller buildings
• Typically, similar, but somewhat less, than the rated fan

capacity

• Length between 10 minutes and 3 hours

27

Application to Mid-Sized Warehouse in Madera

16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1

28

Modeled versus Measured Fits for Madera Treatment Period Are Quite Good

200 400 600 800 1000 1200 1400 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Sampling Location Concentration (ppb) Measurements AERMOD ISCST3

29

Point Comparisons for New Studies

Treatment - 3 Sites Summary

y = 0.26x + 56.76 R2 = 0.07 50 100 150 200 250 300 50 100 150 200 250 300 Predictions (ppb) Measurements (ppb)

Aeration - 3 Sites Summary

y = 0.86x - 0.45 R2 = 0.43 100 200 300 400 500 600 100 200 300 400 500 600 Predictions (ppb) Measurements (ppb)

Model captures the range of concentrations well, but local wind pattern around building structures makes it difficult to predict the spatial distribution.

30

Maximum Concentrations for New Studies

Site A Site B Site C Treatment Measured (ppb) 170 280 120 Predicted (ppb) 317 326 290 Aeration Measured (ppb) 113 40 565 Predicted (ppb) 303 519 970

Comparison of predicted maximum concentration in the model domain and measured maximum concentration.

EPA Level of Concern (4 hrs) = 1300 ppb

31

Conclusions

• Risk assessment with dispersion modeling is a

practical method to address bystander exposure and establish buffer zones

• There is comparatively less data on emissions than for

field applications

• More information would be helpful to refine the risk

assessment