3 ms·

Not arguing, but at 1M per station, that sounds like a self contained automated lab with various kinds of spectrometers. What could one do with 10k per station?

Not arguing, but at 1M per station, that sounds like a self contained automated lab with various kinds of spectrometers. What could one do with 10k per station? 1k? This looks pretty bad ass, https://ichrom.com/mid-4500-mass-spectrometer/ https://ichrom.com/mid-4500-mass-spectrometer/ On the low end, there are MEMs VOC sensors like https://www.bosch-sensortec.com/products/environmental-sensors/gas-sensors/bme688/ https://www.bosch-sensortec.com/products/environmental-senso... It would be trivial to build a solar powered, voc sensor node that talked lora. Easily less than 30$ BOM in qty 10.

reaperman 3y ago
http://www.state.wv.us/admin/purchase/bids/FY2014/B_DEP16574_01.pdf http://www.state.wv.us/admin/purchase/bids/FY2014/B_DEP16574... See here a public bid where a particularly low-cost standalone CEMS unit (will still need a good deal of engineering and construction above this cost) for a single basic chemical species (SO2) costs $15,000. In an area like Deer Park, TX, you might need 8-35 of these per monitoring station (highest variety of chemicals will be near 3rd party tank farms like ITC, along major chemical pipeline networks, at chemical railyard terminals, and ocean shipping terminals). Engineering costs to get a building put in place, telecommunications, power delivery, and SCADA control integration would be expected to run at least another $100,000 not counting any in-house project management hours. They'll need regular expensive maintenance and some installed redundancy. To give you a sense of the scale, here are just energy pipelines (natural gas, crude oil, etc). These maps lack thousands of chemical pipelines. https://www.prnewswire.com/news-releases/energy-web-atlas-world-pipeline-solution-enables-faster-decisions-300871054.html https://www.prnewswire.com/news-releases/energy-web-atlas-wo... https://pvnpms.phmsa.dot.gov/PublicViewer/ https://pvnpms.phmsa.dot.gov/PublicViewer/ Here are some maps of very basic commodity chemical pipelines, but cost $2,000 each to purchase. https://www.petrochemwire.com/maps-and-data/ https://www.petrochemwire.com/maps-and-data/ This one facility here is about 380 ft x 350 ft. There is a single pipeline going to this tiny facility which carries around 7 million pounds of chlorine per day. But that pipeline is just serving a tiny 3 acres of the western hemisphere's largest industrial complex in Freeport, TX which spans well over 13,000 acres. Just extracting the chlorine from salt water for this complex requires enough electricity to power 400,000 homes. https://www.google.com/maps/place/Freeport,+TX+77541/@28.9612012,-95.3235122,146m/data=!3m1!1e3!4m6!3m5!1s0x8640595961a3e7ad:0x42e685217bd37cba!8m2!3d28.9541368!4d-95.3596617!16zL20vMF95bjk?entry=ttu https://www.google.com/maps/place/Freeport,+TX+77541/@28.961...
reaperman 3y ago
You cant do anything at all with $10k per station. Literally nothing. Minimum $250k “Total VOC” is useless, because without identifying the specific chemical sub-family (e.g. differentiation between ethyl acrylate vs butyl acrylate) you wont be able to tell which company is responsible for the leak or what the leak is. > self-contained automated lab Is roughly correct and is the level generally needed. It may be possible to do a bit cheaper, but probably not less than $250,000. Go too cheap and everything will be clogged and broken all the time and you wont get any actionable data for the money you spent. A spectrometer or chromatograph is just one piece of the overall system, what you’ll want to search or build is a “gas analyzer” for the specific compounds. Take a look here for what those look like: https://www.emerson.com/en-us/automation/measurement-instrumentation/gas-analysis https://www.emerson.com/en-us/automation/measurement-instrum... They all also need pretty constant skilled maintenance from expensive technicians. It may be possible to design and manufacture cheaper continuous gas samplers and analyzers but I wouldn’t do that just for this project, that would be its own startup and would need to compete on its own in the industrial market, which is heavily reliant on incumbent relationships and customers are often willing to spend a lot more so they get reliability and are generally unwilling to bet even one hour of potential production time. Because each hour of production can be millions of dollars of revenue, it justifies spending an extra $100k to prevent that downtime. There’s not actually a lot of cost pressure on these gas analyzers at the moment. But if every plant was required to install 10 of them by law, then maybe buyers would care when they need to buy them for several plants at the same time.

reaperman 3y ago

http://www.state.wv.us/admin/purchase/bids/FY2014/B_DEP16574_01.pdf http://www.state.wv.us/admin/purchase/bids/FY2014/B_DEP16574... See here a public bid where a particularly low-cost standalone CEMS unit (will still need a good deal of engineering and construction above this cost) for a single basic chemical species (SO2) costs $15,000. In an area like Deer Park, TX, you might need 8-35 of these per monitoring station (highest variety of chemicals will be near 3rd party tank farms like ITC, along major chemical pipeline networks, at chemical railyard terminals, and ocean shipping terminals). Engineering costs to get a building put in place, telecommunications, power delivery, and SCADA control integration would be expected to run at least another $100,000 not counting any in-house project management hours. They'll need regular expensive maintenance and some installed redundancy. To give you a sense of the scale, here are just energy pipelines (natural gas, crude oil, etc). These maps lack thousands of chemical pipelines. https://www.prnewswire.com/news-releases/energy-web-atlas-world-pipeline-solution-enables-faster-decisions-300871054.html https://www.prnewswire.com/news-releases/energy-web-atlas-wo... https://pvnpms.phmsa.dot.gov/PublicViewer/ https://pvnpms.phmsa.dot.gov/PublicViewer/ Here are some maps of very basic commodity chemical pipelines, but cost $2,000 each to purchase. https://www.petrochemwire.com/maps-and-data/ https://www.petrochemwire.com/maps-and-data/ This one facility here is about 380 ft x 350 ft. There is a single pipeline going to this tiny facility which carries around 7 million pounds of chlorine per day. But that pipeline is just serving a tiny 3 acres of the western hemisphere's largest industrial complex in Freeport, TX which spans well over 13,000 acres. Just extracting the chlorine from salt water for this complex requires enough electricity to power 400,000 homes. https://www.google.com/maps/place/Freeport,+TX+77541/@28.9612012,-95.3235122,146m/data=!3m1!1e3!4m6!3m5!1s0x8640595961a3e7ad:0x42e685217bd37cba!8m2!3d28.9541368!4d-95.3596617!16zL20vMF95bjk?entry=ttu https://www.google.com/maps/place/Freeport,+TX+77541/@28.961...

reaperman 3y ago

You cant do anything at all with $10k per station. Literally nothing. Minimum $250k “Total VOC” is useless, because without identifying the specific chemical sub-family (e.g. differentiation between ethyl acrylate vs butyl acrylate) you wont be able to tell which company is responsible for the leak or what the leak is. > self-contained automated lab Is roughly correct and is the level generally needed. It may be possible to do a bit cheaper, but probably not less than $250,000. Go too cheap and everything will be clogged and broken all the time and you wont get any actionable data for the money you spent. A spectrometer or chromatograph is just one piece of the overall system, what you’ll want to search or build is a “gas analyzer” for the specific compounds. Take a look here for what those look like: https://www.emerson.com/en-us/automation/measurement-instrumentation/gas-analysis https://www.emerson.com/en-us/automation/measurement-instrum... They all also need pretty constant skilled maintenance from expensive technicians. It may be possible to design and manufacture cheaper continuous gas samplers and analyzers but I wouldn’t do that just for this project, that would be its own startup and would need to compete on its own in the industrial market, which is heavily reliant on incumbent relationships and customers are often willing to spend a lot more so they get reliability and are generally unwilling to bet even one hour of potential production time. Because each hour of production can be millions of dollars of revenue, it justifies spending an extra $100k to prevent that downtime. There’s not actually a lot of cost pressure on these gas analyzers at the moment. But if every plant was required to install 10 of them by law, then maybe buyers would care when they need to buy them for several plants at the same time.