Thank you for your time in reviewing our January 2013 monthly analysis. We appreciate your attention to the industry problem of chemical accidents, though we wish your response had included concern about the people who have reported health harms. One worried mother in Chalmette posted on January 25th that her family’s chronic sickness including nausea, throwing up, and dizziness) may be due from her home’s proximity to Chalmette chemical plants. http://www.oilspill.labucketbrigade.org/reports/view/8978
Your article highlighted challenges that some may have in understanding the monthly analysis. Thus we are working to clarify the format of the monthly report for lay people’s understanding.
In this letter, we have addressed the claims made in your article. We hope that you will accept our offer to meet (extended via letter and e-mail to President Chris John on 2.15.13) so that we can avoid future misunderstandings.
Louisiana Bucket Brigade
- LMOGA: “Answers to a number of very basic questions – such as: how ‘average citizens’ can be ‘eye witnesses’ to so-called accidents occurring inside a refining facility – are unfortunately beyond the scope of these reports.”
LABB Response: “Average citizens” are invaluable resources because the petrochemical companies, Bucket Brigade, the LDEQ, and even the EPA are not able to observe every refinery accident in Louisiana as it occurs. As a result, people we already trust – our parents, teachers, neighbors, and friends – use the iWitness Pollution Map to report their experiences seeing, smelling, and feeling the effects of chemical accidents.
Let’s say there was a fire in your neighbor’s house and you saw the smoke and flames from your window. You probably wouldn’t wait until the firefighters showed up to know if there was a fire or not. Your knowledge as an eyewitness is an important alert for authorities to put out the fire. You may not know what caused the fire—maybe it was burned toast or a candle—and that’s okay because it’s the police’s job to figure out why the fire started. Through the iWitness Pollution Map, LABB encourages people to be proactive about their safety; after all, it doesn’t take a firefighter to know what fire looks like.
By reporting accidents like fires, “average citizens” are anything but average; they are extraordinary people who look out for the well-being of themselves and the people around them.
- LMOGA: “Tallying the individual figures listed under ‘All Reports’ yields a total of 185. But the figure listed as the total number at the top of the report is 273. So what’s with this discrepancy? Is one wrong to assume that ‘All Reports’ and ‘Total Reports’ are the same thing? Regrettably, it’s impossible to tell. Giving the group the benefit of the doubt, 185 is closer to the NRC Report total of 183, so perhaps those two data sets should be synonymous. Even so, the numbers still don’t add up.”
LABB Response: To identify the number of Total Reports, defined as the number of National Response Center (NRC) and citizen reports, we subtract the total number of rows in the data set (279) –Row 1 for headings= 278 rows of data, or 278 Total Reports in the data set.
To identify the number of NRC reports, we use the FIND feature in the column Incident Title and search for the letters “NRC.” This letter combination shows up only once in the report title if it is sourced from the NRC. There are 183 instances where “NRC” came up in the search results, thus there are 183 NRC reports in the data set.
To identify report categories: each time a NRC or citizen report is submitted, the report is tagged with key terms called Report Categories. Each report may be tagged with one or more key terms. Thus it is not appropriate to sum the number of reports associated with all key terms. The monthly accident report seeks to analyze the frequency of 6 key terms (Onshore, Vessel, Pipeline, Platform, Storage Tank, Unknown) that are most relevant to citizens living next to onshore, active chemical plants.
To search the number of NRC and citizen reports which describe an accident that occurred “Onshore”, we search for “Onshore” in the Category column (44 reports). We can calculate the percentage of reports tagged with a specific key term by dividing by the total number of reports that was analyzed. For example, 44 “Onshore” reports / 278 NRC and citizen reports (“All Reports”) = 16% of NRC and citizen reports are tagged with the word “Onshore”. Thus 16% of all reports submitted in January are regarding accidents at onshore facilities.
- LMOGA: “What about utilizing the group’s own search function on the online iWitness Pollution Map, setting a date of Jan. 1, 2013-Jan. 31, 2013, and searching for the total number of reports? That search generates 278 reports, not 273. How is one version of the January 2013 report undercounting by five accidents, or vice versa?”
LABB Response: LABB typically analyzes data with the goal to publish the monthly analysis in the first week of the following month. For example, to study Jan 1, 2013- Jan 31, 2013, LABB downloads data in the last week on January, typically on the last day of January during business hours. As a result of the reporting timetable, reports submitted retroactively (reports submitted in February for observations made in January) may not be featured in the monthly report. Recognizing this limitation, we download a new data set every time we analyze reports within a specific time frame.
There were five citizen reports submitted retroactively for January 2013 that were not featured in the monthly report. Thus there are 183 NRC reports + 90 citizen reports + 5 retroactive citizen reports = 278 total reports, and 90+5=95 citizen reports. These 5 retroactive reports were submitted by Chalmette residents who have experienced foul odors around their homes. LABB would like to share 2 of these reports:
- LMOGA: “Leaving aside unanswered questions about what is needed for a citizen report to be considered ‘verified’ as opposed to ‘unverified’ ”
LABB Response: LABB utilizes the APPROVE feature to confirm that a reliable source submitted the report. All reports submitted through the National Response Center are approved. Citizen reports are approved when the LABB administrator has identified the source and content of the report. This is one method in which LABB prevents spam reports from showing on the iWitness Pollution Map.
LABB uses the VERIFY feature to confirm the content of citizen reports. Citizen reports are verified when:
- 3 or more citizen reports are received on the same day, and all 3 reports describe similar observations from the same location. For example, on December 7, 2012, 23 reports were submitted to the iWitness Pollution Map with observations of large, ongoing flares causing black smoke at Shell Motiva refinery in Norco, LA.
- A citizen report corresponds with the date, location, and observations displayed in a NRC report.
- A citizen report corresponds with a media story reporting similar observations on the same date and location.
- A citizen report includes a time-stamped, geographically verifiable photo or video
- LMOGA: “Whether or not text messages and email reports, for example, are filtered to eliminate double counting”
LABB Response: Our citizens report to the iWitness Pollution Map in four ways: text, voicemail, email or web submission.
Each citizen report submitted by text has a tagline at the end of the report, which includes the phrase “SMS-to-email.” Each citizen report submitted by voicemail has a different tagline at the end of the report, which includes the phrase “Play Message” with a link to the actual voicemail. Citizen reports submitted via email or web do not have such a tagline. Because the capacity to tell the difference between email and web submissions does not exist, reports submitted via email and web are counted in one category.
Searching for the phrase “SMS-to-email” in the Report Description column of a data set yields the number of reports submitted via text (45 for January 2013). Searching for the phrase “Play message” in the Report Description column yields the number of reports submitted via voicemail (23 for January 2013). To identify the number of citizen reports submitted via email or web, we subtract the total number of citizen reports from number of texts and from the number of voicemail submissions (95-45-23 = 27 reports submitted via web or email in January 2013).
In short, LABB’s method of filtering reports via taglines avoids double-counting between reports with different ID numbers. Double-submission of a report is possible. For example, a user could report “Smells like rotten eggs, feeling dizziness at Midway and Jewella Ave, Shreveport” through text message and through voicemail. The text report and voicemail report have different report ID numbers, and would be filtered by the method described above. Double-submission does not occur frequently in the iWitness Pollution Map.
- LMOGA: “How can the figures in the individual categories add up to 94 reports when the top of the document indicates that there were 90?”
LABB Response: To identify photos or videos in reports, any user can go to the map website at http://www.oilspill.labucketbrigade.org/reports, change the date range to Jan 1 2013 to Jan 31 2013, and under MEDIA select “Photos” or “Video.” In January 2013, 2 reports had photos. One report has 1 photo, the other has 3 photos (see http://www.oilspill.labucketbrigade.org/reports/view/8950 and http://www.oilspill.labucketbrigade.org/reports/view/8792). As a result, “4 photos” are reported on the iWitness monthly summary report. Each photo is not its own separate report. The 4 photos are submitted in 2 reports, which are already accounted for in the 95 citizen reports for January 2013. (The 95 citizen reports includes from 90 reports as shown in the monthly report + 5 retroactive citizen reports.)
- LMOGA: “The Bucket Brigade doesn’t share how it arrived at any of its figures, including the total of 183 NRC reports.”
LABB Response: To identify the number of National Response Center (NRC) reports, we use the FIND feature in the column Incident Title and search for the letters “NRC.” This letter combination shows up only once in the report title if it is sourced from the NRC. There are 183 instances where “NRC” came up in the search results, thus there are 183 NRC reports in the data set.
- LMOGA: “…the [NRC] reports for both the ExxonMobil Baton Rouge and Chalmette refineries are inflated by at least a factor of two in both instances.”
LABB Response: To identify accidents that may be most likely to be linked to activities at the ExxonMobil refineries, for example, we look through the entire list of 278 reports to find NRC and citizen reports that indicate close location and downstream wind direction to the refinery. Some of these reports explicitly mention “ExxonMobil” and some of them don’t. We recognize that researchers may use different conditions to link reports to specific chemical plants. LABB also knows that chemical plants besides ExxonMobil may also influence foul odors smelled in Baton Rouge. As a result, the monthly report does not identify chemical plants as definite sources, but as “Suspected Parties.”
LABB reviewed its monthly analysis for January 2013 and found 2 NRC reports were overreported for suspected party ExxonMobil in Baton Rouge. One NRC report was duplicated and the second NRC report listed Lion Copolymer as a Suspected Party. The NRC reports that may be linked to ExxonMobil and Lion CoPolymer (4 reports) and various Chalmette chemical plants (2 reports) are shown in the table below. Please note that reports can be tagged with multiple keywords, e.g. 1 report can be tagged for Odor and for Health Effects.
- LMOGA: “Unfortunately, as the group continues to show us: the ends (doing whatever they can to attack job-creators all around the state) always seem to justify the means (distorting the data, and sometimes just plain making it up).”
LABB Response: The petrochemical industry should hire more people to prevent accidents. This is why the United Steelworkers partner with LABB; they realize that a common sense approach of regular maintenance and accident prevention would help our economy.
Unfortunately, the numbers reflected in the iWitness Pollution Map monthly analysis underrepresent the current frequency of industry accidents. This is because many facilities may not report to the National Response Center, and because LABB is not active in many Louisiana communities that neighbor chemical plants. As the U.S. Coast Guard independently reaches out to more non-reporting facilities about the requirement to report to the NRC, and as LABB reaches out to more affected communities, LABB expects the numbers seen in the monthly analysis to rise and to more accurately reflect the human impact of chemical emergencies.