Though wind and solar continue to carve out larger and larger shares of world energy supply, the modern world still runs on petroleum, and will continue to do so for the foreseeable future.
A large chunk of my ongoing PhD work is on oil refineries so this was very exciting to see in my feed! This lines up well with my understanding of US refineries. Three comments below.
1. I had never seen that material flow diagram for the Richmond refinery—great find! One reason that operation is so damn complex is California’s ever-changing, uniquely restrictive regulations on air emissions from cars. As far as I’m aware, no or few other refineries in the country can produce fuels that meet Californian standards. Those that can face higher transport costs than Richmond. You can imagine there are downsides to this.
2. A side gig for several large oil refineries is actually exporting electricity. One way to describe a refinery is as a fossil-fuel burning steam generator, since that underlies many of the processes you describe. This means many refineries self-generate electricity. Further, most large refineries are grid-integrated and export their excess electricity as part of regular operations or just opportunistically. This quality of refineries can also lead to ugliness. I spoke to someone who ran the Richmond operation for a while and he described an episode where CAISO “compelled” them to shift electrons away from refining towards the grid during a generation crisis.
3. My PhD research is looking at how power outages at US oil refineries impact retail gas prices. IMO, that’s an underrated vulnerability that results from a general increase in the complexity of these refineries, particularly the post-distillation processes like mixing that typically rely on electricity to operate. Yes, many refineries cogenerate electricity but they are still reliant on the grid to some degree, sometimes to a large degree. This creates a complex policy problem: how can we get ERCOT operators to invest in infrastructure near refineries when the costs of such events are diffused amongst drivers from across the country?
What a great explanation! I know a fair amount about refining and this really nails the fundamentals.
A large chunk of my ongoing PhD work is on oil refineries so this was very exciting to see in my feed! This lines up well with my understanding of US refineries. Three comments below.
1. I had never seen that material flow diagram for the Richmond refinery—great find! One reason that operation is so damn complex is California’s ever-changing, uniquely restrictive regulations on air emissions from cars. As far as I’m aware, no or few other refineries in the country can produce fuels that meet Californian standards. Those that can face higher transport costs than Richmond. You can imagine there are downsides to this.
2. A side gig for several large oil refineries is actually exporting electricity. One way to describe a refinery is as a fossil-fuel burning steam generator, since that underlies many of the processes you describe. This means many refineries self-generate electricity. Further, most large refineries are grid-integrated and export their excess electricity as part of regular operations or just opportunistically. This quality of refineries can also lead to ugliness. I spoke to someone who ran the Richmond operation for a while and he described an episode where CAISO “compelled” them to shift electrons away from refining towards the grid during a generation crisis.
3. My PhD research is looking at how power outages at US oil refineries impact retail gas prices. IMO, that’s an underrated vulnerability that results from a general increase in the complexity of these refineries, particularly the post-distillation processes like mixing that typically rely on electricity to operate. Yes, many refineries cogenerate electricity but they are still reliant on the grid to some degree, sometimes to a large degree. This creates a complex policy problem: how can we get ERCOT operators to invest in infrastructure near refineries when the costs of such events are diffused amongst drivers from across the country?
Excellent analysis!