folks: One of the readers had some questions about my calculations, and definitions on some of the terms I had used, which required an answer to clarify. So, for your recreational reading...
-----------------------------------------------------------------------------------------------------------------------
Blue board readers questions:
Not so, unless you assume that all natural gas pipelines are within 200 feet of a school and all of the other schools are occupied 24 hours per day. There were plenty of incidents not counted in this graph because they occured in areas where noone was around.
And what do you mean if "a child is in school?" If there is an incident there may be 3000 children in school plus teachers and support staff. You seem to correlate one incident to mean one injury but an incident that involves multiple injuries is still just considered one incident. If one pipeline where to explode and cause death or serious injury to everyone in the school it would only be counted on your graph as one incident divided by all the miles of pipeline in the United States. So is one incident significant or isn't it?
Also you seem to forget that there are three pipelines not one. Additionally there there is a crude oil pipeline nearby.
Furthermore I would like to see a report on the incidence of accidents that occur in pipelines that are as old, as large and as pressurized as the ones that are at AME.
-----------------------------------------------------------------------------------------------------------------------
My response:
This is precisely why I do NOT use the "Serious" category in my estimations for the rest of the posts, and switch to "Significant" which makes no assumptions about population density along the pipelines.
All the incidents that were not counted in the "Serious" graph are included in the "Significant" incident data. Even all the incidents you refer to as "no one was around" are included in that category. And, the graph isn't mine - it is from the PHMSA. Please refer to the reference and similar graphs and raw data for Significant incident data.
And no, I am not correlating one child per incident. I am looking at the effect of an incident on a child. Perhaps the term "significant" in my post is the cause of some of your concern. Let me clarify. The term "Significant" has a specific meaning for the PHSMA. I dont call anything significant or insignificant, the PHSMA does.
"PHMSA defines Significant Incidents as those incidents reported by pipeline operators when any of the following conditions are met:
1. fatality or injury requiring in-patient hospitalization
2. $50,000 or more in total costs, measured in 1984 dollars
3. highly volatile liquid releases of 5 barrels or more or other liquid releases of 50 barrels or more
4. liquid releases resulting in an unintentional fire or explosion "
primis.phmsa.dot.gov/comm/reports/safety/SigPSI.htmlThe term is not open to colloquial interpretation in this context, and this is the meaning my posts use. It has no bearing on relative importance of 1 fatality/injury vs several per incident.
The measure of human impact is not done by incident probabilities, but by f/N curves, which map the number of fatalities per incident. This accounts for multiple fatalities per incident. Please see post 5/5.
For the record, in the past 20 years, the maximum number of fatalities in a given year from "Significant" gas transmission incidents is 22(1989), next highest is 15(2000), and all the rest of the years is under 3 per year. The 20 year average is 3 fatalities per year, and the average for the last 5 years is 1 fatality/year.
For injuries, the corresponding statistics are: maximum in a year is 28(1989) and the next highest is 22(1994). The average over 20 years is 10 injuries/yr and the 5-year average is 5. injuries/year.
Note that the total number of injuries+fatalities=204+59=263 which is about that 1041 incidents in this time period. I have conservatively used the larger number, 1041, instead of 263 so if anything my estimate is an overestimate.
So while the possibility of 3000 humans being affected by an incident is daunting, I would submit that 3000 fatalities/injuries is more that 10 times the TOTAL number of fatalities/injuries in the last 20 years in the entire United States!
The example I give in post 2/5 has 3 pipelines: 1 36" and 2 30" lines. The operation of the 30" pipelines was unaffected by the explosion of the 36" pipeline. While I suppose it is possible that the others could be affected, there is documented evidence that this is not something that can be assumed. If it does, the PIRs would overlap, and the larger one, the 36" pipeline PIR will dominate.
I was not aware of an oil pipeline - could you please point me to a reference as to its location and characteristics?
As for the statistics for 36" pipelines, the data is in public domain, in the hands of NTSB and PHSMA OCP. Feel free to add to our body of knowledge by doing the analysis and presenting the results.
I personally have been advocating the replacement of the 36" pipes, as has Arch, due to their age.
Cheers.