In celebration of its 20th anniversary, the International Society of Exposure Science has been including Exposure Science Digests - brief articles that "showcase successes" in the field. The July/August 2010 edition includes digests on:

'Ensuring the safety of chemicals'
http://www.nature.com/jes/journal/v20/n5/full/jes201028a.html

and 'The smoking gun: working to eliminate tobacco smoke exposure'.
http://www.nature.com/jes/journal/v20/n5/full/jes201034a.html

Kai-Chung gave a presentation on his recent work regarding the reconstruction of true CO levels from readings of a diffusion CO sampler with a slow response time. A report on this work is also forthcoming. The powerpoint presentation is attached.

I reworked one of my data processing scripts written in R to automatically process real-time SidePak aerosol data. You may recall I wrote the script originally for CO data processing.

With this script it is possible to very quickly create an R dataframe containing all of the real-time data present in a given directory. Attached is a ZIP file containing the new R processing scripts.

Neil, Wayne, and Kai-Chung visited the San Pablo Lytton casino in San Pablo, CA on January 12, 2008. They were able to measure aerosol data, casino dimensions, and nonsmoker/smoker counts. The visit was a success and informed protocols that can be used in our subsequent casino surveys.

Attached to this post is a detailed write-up of the visit, including the protocols used and the results.

All data from the visit have been uploaded to the archive server at the following FTP site:

domain: exposurescience.org
username: datastore

I have written a program for Linux called "SidePak LaBraT" that is able to simultaneously plot and save real-time SidePak aerosol data from multiple SidePak units. It seems currently impossible to connect multiple SidePaks to a single Windows computer, so this software is Linux-only. I've spoken to the folks at TSI and FTDI (makers of the USB/Serial chip) and there seems no way for the Windows drivers to handle multiple SidePaks. But Linux has no problem :-).

"SidePak Buddy" is software for the Linux and Windows platforms that plots and saves real-time aerosol levels measured by a TSI SidePak Aerosol Monitor. This software is meant to be a user-friendly way for both technical and non-technical people to see the immediate impact of a source of particulate matter on air quality.

I have recently used this software to demonstrate the levels of real-time particles that occur in an automobile with an active smoker present (see http://latimes.com/smoking).

This updated version of "SidePak Buddy" (Version 1.2) has the following features:

http://BurningIssues.org
For a period of more than 10 years from 1994 to the present, Dr. Wayne Ott of the Statistics Department, Stanford University measured indoor and outdoor particulate polycyclic aromatic hydrocarbons (PPAH) levels in his residential neighborhood in Redwood City, CA.

To predict the proximity effect, I have applied the "Markov I" stochastic model of Nicas (2001) to characterize the dispersion of air pollution from a "puff" release in a room that has dimensions of 10.7 x 7 x 2.6 meters (L x W x H). The model treats the dispersion of emitted particles as a Markov Chain process, where each particle moves by a series of "random walks" due to turbulent diffusion. The value for the turbulent diffusion coefficient, D, was set at 0.04 m2/sec.

In applying the model, I used 6 receptor positions located 0.5, 1, 1.5, 2, 2.5, and 3 meters away from the source in the horizontal direction. The air exchange rate for the simulation was 0.5 ach, which is typical for a residential location.

Attached to this post are two plots showing the results of the simulation:

1. A "Concentration versus Time" plot showing the concentration time series at each of the receptor points (note the y axis is on a log scale) for a period of 15 minutes after the release.

2. A "Average Concentration versus Distance" plot showing the proximity effect of 1-minute average concentration as a function of distance from the source. The 1-min average was take during the minute just after the release occurred.

These plots are in broad agreement with prior work showing that (A) air pollution from a "puff" release generally mixes thoroughly in the room of release within 5 to 10 minutes after release, and (B) that the proximity effect is generally an f(x) = 1/x function of Concentration vs. Distance.

This modeling approach can be expanded to include advection (air currents), obstructions, inlet and outlet flows, particle deposition, and reflection.

Note: Also attached to this post is the R source code used to run the Nicas model.

I've come across two webpages that contain online calculators, one for estimating the natural or mixed-mode ventilation rate of an arbitrary building:

http://phpaida.veetech.org.uk/phpaida.php

And one for estimating the ventilation rate of a building from steady-state CO2 measurements and occupancy level:

http://phpco2.veetech.org.uk/phpco2.php

In my research into prior work on indoor air pollutant dispersion, I have come across two paper that present valuable data relevant to our own work on the proximity effect and room mixing.

A paper by Drivas et al (Indoor Air. 1996; 6(4):271–277) applies a form of the atmospheric diffusion equation:

Modeling Indoor Air Exposure from Short-Term Point Source Releases. We have developed a simplified analytical indoor air model that describes the concentrations as a function of position and time in a room following a short-term release of airborne particles or gases. The indoor dispersion model considers the two main physical processes of (1) point-source dispersion with reflection from all walls and (2) the general concentration decay in a room due to room ventilation and surface deposition of pollutants. Comparison of model predictions with experimental indoor measurements conducted by other researchers showed excellent agreement. This model should prove useful for human-health risk estimations in which the inhalation dose resulting from an indoor, short-term release of a contaminant needs to be calculated.

Nicas builds from the work of Drivas et al. and uses a "Markov" probabilistic approach, incorporating the effects of advection (AIHAJ. 2001 Mar-Apr;62(2):149-58):

Modeling turbulent diffusion and advection of indoor air contaminants by Markov chains. Turbulent eddy diffusion models are used to describe a continuous concentration gradient with distance from an in-room contaminant emission source. A refined diffusion model termed the Drivas model also accounts for contaminant reflection by wall surfaces and partially accounts for removal by exhaust air. This article develops two models based on Markov chains to describe indoor air contaminant dispersion by turbulent diffusion and advection, and removal by the exhaust airflow. Markov model I is equivalent to the Drivas model and is computationally simple. Markov model II can provide more realism by accounting for the locations of air inlets and outlets, advective flow patterns, in-room reflective surfaces, and contaminant removal mechanisms at specific room positions. The price paid for this greater realism is greater computational complexity. Both Markov models are explicitly probabilistic and estimate the expected concentration values at given room positions.