Public Lab Research note

Advice for real-time monitoring

by ottinger | August 30, 2018 02:18 30 Aug 02:18 | #17021 | #17021

In the "Meaning from Monitoring" project, we tried to take real-time measurements of toxic gases (mostly) and PM2.5 from Richmond, California (near the Chevron oil refinery) and Rodeo, California (at the fenceline of the Phillips 66 oil refinery) and make the data more accessible and useful to community members advocating for pollution reductions.

Among the strategies we tried:

  • developing a new website to enable people to explore historical data
  • creating a "user report" function to allow residents to annotate the data with smell reports and photos to document flares, etc
  • using FitBits and pulse oximeters to track health metrics alongside air quality measurements.

Our key findings are summarized in "Making the Most of Monitoring." The most important things we found were:

  • the more data you have, the more infrastructure you need to store and maintain it.
  • the more data you have, the more creative you need to be about how to interpret it--and you may want to get data experts involved (we found we needed to).
  • monitoring can divert attention from the more important issue of pollution prevention per se. It's worthwhile to recognize that potential trade-off before moving in the direction of monitoring.

These findings apply not only to the relatively high-tech, high-cost monitors being used in Richmond and Rodeo (optical open-path sensors, including a portable version called a Hound). They apply to PurpleAir and Awair devices, as well--any monitoring strategy that's going to produce a lot of data is going to come with comparable problems of data management and interpretation.


@LESBreathe, this post and the report linked in it might be helpful to check out

Reply to this comment...

This might be some useful information for you @sarage Check out the report linked " Making most of monitoring".

Reply to this comment...

For the most part, my experience was on the manufacturing end. And the answer would be very similar, but stated quite differently " the cost benefit analysis must justify additional instrumentation/personnel/and testing".There would have been things like GMP,GLP, etc. That also would have helped the issue along. But in the end, for individual batches, yes to testing. Any backup data (say the equivalent of PM2.5) always had to be justified. Repeatedly.

Reply to this comment...

Login to comment.