The Refinery of the Future

photo credit: Google Gemini

TL&DR: we will continue to refine petroleum to make the chemicals and materials we need in a decarbonized future. We’ll just do it without the burning. It will be complex, expensive, and will take a new kind of refinery which looks quite different than today’s refinery.

You’ve no doubt heard the nirvana fallacy argument common among fossil fuel lovers, climate change minimalists and climate doomers alike: that when we give up burning fossils as fuels, we need to be prepared to give up the tens of thousands of materials and chemicals we make from fossil petroleum, gas and coal, because producing them is impossible without making fuels.

Those people are either talking out of an orifice best reserved for something other than talking, or they’re really saying they don’t know how we’d keep using petroleum to make chemicals and materials without making fuels. You shouldn’t listen to them. They’re wrong.

I’ve spent decades helping people develop and test alternative ways of making chemicals and materials from both petroleum and biomass, so I know a thing or two about this topic. However, this is a huge area of chemical process technology and no one living human knows it all. I’ve got a pretty good overview though, so I thought I’d take a stab at imagining what a future petroleum refinery would look like.

No, We Won’t Use Biomass Instead

There are some who see no path forward other than transitioning the entire carbon economy to start with biomass. As someone intensely familiar with making chemicals and fuels and materials from biomass, I can say that this will be a fairly small fraction of the future refinery.

Biomass, every sort of which was recently CO2 in the atmosphere, is “carbon neutral” insofar as we don’t make any fossil GHG emissions along the value chain. Today’s agriculture is very fossil based, with nitrogen fertilizer and farm fuels, grain drying and transport all more or less 100% fossil based. We’ll need to transition substantially all of that away from fossil fuels use and that’s a huge separate task. Agriculture also generates new, un-natural emissions of methane and N2O, both potent GHGs. We’ll need to cut that out as well, by more strategic fertilizer use, genetic engineering of soil organisms, and much, much better nutrient management.

There are a handful of materials of the chemical variety that make sense to make from biomass. Some of them we make already. Others are virtuous substitutes for fossil derived chemicals. But a mass shift from fossil petroleum to biomass sources for chemicals and materials is extremely unlikely in my view.

Why is that? Simple. Biomass has an average general chemical formula of C6 H10 O5. There are exceptions- food oils being one example- but the greatest mass of biomass is cellulose and lignin, not vegetable oil. It is hydrogen deficient, and worse still, there’s nearly one oxygen atom for every carbon atom. To make most useful chemicals, those oxygens need to be removed by reacting them with hydrogen to produce water, or burned off to produce CO2. Both represent a huge loss of energy and mass.

What will we do with biomass? We’ll make shipping and aviation fuels. We may need some green hydrogen, made from water and renewable electricity, to satisfy our hydrogenolysis needs. But all the CO2 that goes back into the atmosphere from these applications, which need liquid fuels, will have been in the atmosphere in the recent past. Making fuels from biomass is far easier than making anything else of value.

We Won’t Use Direct Air Capture (DAC)

DAC is thermodynamically and economically FUBAR. IT isn’t a real technology, it’s a predatory delay strategy, intended to keep us burning fossils for longer, guilt free. No, we won’t be snatching CO2 out of the atmosphere and reducing it chemically using green hydrogen to make chemicals or materials. There are far better sources of CO2 for the handful of chemicals it makes sense to make from CO2, like formic acid, formaldehyde and perhaps methanol. Just get “chemical carbon recycling” out of your head- it’s not going to be a thing.

The Refinery of the Present

The original oil refineries were simple batch retorts where oil was put into a container and heated, the vapours taken off the top and condensed, heating continued until no more vapours came off, and then the pot was cooled and the remaining tar or char was dug out with picks and shovels. And the original high value product was kerosene for lighting. The lighter fractions were thrown away, often into local rivers and streams.

Gradually, the heavier oil fractions became useful as a replacement for coal in steam engines. Tar had uses in roofing and waterproofing. But over time, as engines developed, the lighter fractions found use too, in gasoline and diesel engines. The lightest fractions were still burned as fuel gas to run the distillations.

Eventually the lighter fractions came to dominate demand as the internal combustion engine took over transport from the horse and the steam engine. And large sources of heavier oil were found. Crackers were soon invented, and cokers too, to break big molecules into more valuable smaller ones.

Soon a host of refinery operations were invented. New fluid catalytic crackers which efficiently made olefins (molecules with C=C double bonds) useful for plastics and to make larger molecules of the right type and size, but which also converted heavy oils to lighter fractions . Alkylators to put branches on chemicals. Plat-formers to turn straight chains into rings and hydrogen. Hydrotreaters and hydrocrackers to remove sulphur and nitrogen, using hydrogen generally made from natural gas. Cokers and extinction hydrocrackers to convert the very bottom of the barrel- the tarry junk called residuum or “resid” which doesn’t boil at 300 C and 1/760th of an atmosphere pressure- into petroleum coke and a little more of the good stuff. And every fraction along the way gets split off to another unit, reacted, converted, or sold off to someone else for further processing. Small molecules get bigger, and big ones get smaller all over the place. The refinery can change its suite of products by adjusting both feedstock blends and turning various knobs in the process.

Refineries are highly integrated, with major petrochemical producers having their plants nearby. Many are specialized around certain feedstocks- some take light oils, some specialize in heavy oils. Byproducts are recycled or burned as fuel gas. Heat energy is recovered from one step to use in another, or made into steam to run turbines to run pumps, compressors or to make electricity. The scale is giant, and most of the time, margins are quite thin. Big refineries therefore find it easier to consistently make money than little ones, because they have lower capital intensity due to greater economy of scale.

How Much Do We NOT Burn?

It varies a lot, but only between 15% and 25% of a petroleum refinery’s output is used to produce things that are not fuels. 75-85% ends up being burned at some point. When you take into account the fact that petroleum discovery, production, refining and distribution is about 81-83% efficient, meaning that 17-19% of each barrel is used between the well and your gasoline tank, the fraction used for truly high value uses seems even smaller still.

The Refinery of the Decarbonized Future

As mentioned already, petroleum refineries are incredible machines which make small molecules bigger, make big molecules smaller, and change the shape of molecules in profound ways as a routine part of making the suite of products we need today. And petroleum refineries have evolved to make a differing suite of products over time, as market demand and oil supply sources have both changed profoundly.

How on earth could we cope with making only 15-25% of the current products, and not making the rest? To some that seems just ridiculous and impossible, hence the nirvana fallacy argument mentioned at the start of this piece. But not to me. Not to anybody who understands the unit operations and chemistry.

All that would change is two things:

1) Energy sources: whereas current refineries burn byproducts and natural gas to make the heat necessary to run the various processes, the refinery of the future will do this with electricity.

2) Garbage disposal: fire (boilers and fireboxes on various units) and certain low value fuels, are both the current catch-all for mixtures of molecules that the refinery doesn’t need and which don’t fit the product specification of higher value products. That will need to end.

Wow, those are NOT small things that need to change! But they’re a far cry from impossible.

Getting Rid of Fire for Heating

Electric heating is by and large a no brainer. The only reason we use fire today is that we’ve been making all the heat we need via fire for the past 800,000 years- and in past we also made most of our electricity from fire. But electricity is pure thermodynamic work- it can easily be used to either pump heat or to make heat at any temperature or conditions desired. All we need is to stop treating the atmosphere as if it were a free and limitless public sewer, and all of a sudden there’s a value proposition for smarter ways to make and use heat.

Many unit operations such as distillation are a prime example of how we can be smart about this. In a refinery, the waste heat from one step is often used to provide heat for another, either directly by heat exchange or indirectly via steam generation and use- we’ll keep doing that in the future too, but we have another knob to turn. In a distillation unit, nearly every joule that goes into the reboiler, comes back out of the condenser, just at a lower temperature. By compressing the vapour to raise its condensation temperature, the reboiler heat exchanger can become the condenser. The coefficient of performance of this type of heat pumping, called mechanical vapour recompression, can be very high indeed. We just don’t do it often today because fuels and cooling water are both really cheap.

But where heat pumping schemes aren’t possible, electric heating of various sorts, from resistance heaters to inductive, microwaves and plasma arcs, can substitute for fire. And they are easier to control than fire, and usually generate no flue gas which needs heat recovery before discharge.

I designed and built dozens of pilot and demonstration scale plants for a dizzying array of refinery, chemicals and materials processes. Only one of those units ever deliberately used fire to make heat. All the rest was done electrically, because at small scale, fire is dangerous, difficult to control, and not worth the energy cost savings. As time goes on, scaling up electric heating will require some brainpower, but it’s not rocket science, or brain surgery, much less rocket surgery!

The Alternatives to Fire for Waste Disposal

An old movie quote from whoknowswhere rings in my head: someone asks “how will we sort the guilty from the innocent?”, to which the madman, soldier or action hero replies, “Kill ’em all and let God sort ’em out!” That’s more or less what we do with complex light hydrocarbon mixtures in the refinery, with God replaced by fire. Everybody, the valuable and the valueless alike, are reduced to their fuel value.

Numerous unit operations generate continuous streams of light gas molecules. Others generate such streams intermittently, from cleaning, de-coking or catalyst regeneration steps. Often, these gas mixtures contain water vapour, nitrogen, carbon monoxide, CO2, hydrogen, and a mess of light hydrocarbons with low economic value. The logical solution is to just burn the mixture to satisfy the enormous energy thirst of the refinery.

That too must end in the refinery of the future.

And it must end in society in general. We have to stop burning fossils, and that doesn’t just mean petroleum, natural gas and coal- it means everything made from them. Europe, for instance, will need to stop burning municipal solid waste because of its waste plastic content. Waste plastics will need to be recycled or landfilled instead.

Why? Because the product fossil CO2 from burning has only two possible destinations: carbon capture and storage, or “un-burning” (reversing combustion) using hydrogen, green or byproduct of the various refinery operations, to produce water and reduced hydrocarbons like CO or methanol- and then only for uses which are themselves not burned at end of life. Both are unpleasant and expensive.

Some CO2 generation in the refinery will be inevitable, so some “un-burning” or CCS will be inevitable. But making it worse by burning stuff instead of sorting the molecules would be a decidedly bad idea.

So: we’ll need to get into the molecule sorting business, big time.

Some companies are already doing this. They’re like urban racoons, sifting through the garbage disposal called the fuel gas system, hunting for valuable molecules like hydrogen or ethylene, burning the rest. Membranes, absorbents and other technologies can be used to do this, though it takes energy and money.

And we’ll also need to feed those light molecule mixtures to reformers, and then use, ***gasp- I never thought I’d say this!***, a total pile of steaming sh*t of a process called Fischer Tropsch, or F-T for short. F-T can also be understood as standing for “f*cking terrible”. It’s a process by which CO is hydrogenated to make hydrocarbon molecules and water:

CO + 2H2 ==> -(CH2)- + H2O

The CO and H2 are made from fossils by reforming:

nCH4 + mH2O + heat (i.e. more CO2 emissions) ===> xCO + yCO2 + zH2

F-T has been around for a long, long time. It was made popular by the Nazis when their oil supplies were short, as a way to make gasoline and diesel from coal, the coal being gasified to produce CO and H2 and CO2. And then another bunch of people who found themselves embargoed by most of the world for their unsavoury treatment of their own population on the basis of the colour of their skin, did the same thing in South Africa. You’ll note the theme here: ain’t nobody does F-T unless they’re f*cked otherwise, i.e. they have no other choice. Fascism, racism, genocide etc. are fortunately optional.

Why is F-T so terrible? Lots of reasons. But the biggies are its very poor energy efficiency, super-high carbon (GHG) intensity, lack of selectivity to molecules of the size you want, and ridiculously high capital intensity- which gets worse and worse the smaller you make the plant.

As an aside: anybody thinking they’re going to run F-T on syngas made from either CO2 and green hydrogen, or from biomass and green hydrogen, is just an idiot. Don’t listen to them and make sure you keep your wallet securely away from their hands’ reach.

Other than the Nazis and the South Africans, F-T has been deployed a few times successfully-ish. The key plant is Shell Pearl, a mammoth $24 billion (over a decade ago) project in Qatar. This plant eats basically free natural gas for power and feedstock, that the Qataris have in giant quantity with no other use for. It spits out high quality waxes and base oils for lubricants, plus the usual range of fossil fuels (LPG, gasoline and diesel fractions) that a refinery puts out- just nice clean ones with no sulphur or nitrogen or metals in them. Those either aren’t found in the natural gas feedstock or are removed in the syngas production process. That said, despite free feed and energy, mammoth vertical scale, and a giant atmosphere to dump the waste CO2 into for free, if petroleum isn’t selling above $40/barrel, Pearl reportedly cannot make money. That’s probably $60 in today’s dollars too, and will climb as carbon taxes climb.

F-T fundamentally makes a range of molecular weight products. If you don’t want waxes (because their market is small, and you don’t want to pay to hydrocrack waxes back to molecules of a size you really want), you end up making syngas (which you made from methane usually), back into methane and light molecules of the LPG range. The process basically reverses the process used to make its feedstock- and that simply cannot be stopped. (as an aside- the Sabatier reaction is basically the result of totally breaking a F-T catalyst so it makes nothing but methane). Those light molecules, if derived from fossil CO or CO2, are nearly useless in a decarbonized future, so you’ll need to recycle them to the reformer for another go-round, wasting energy and capital. And if you don’t want methane and LPG, you’re out of luck- you’re always going to make some, even if you turn the molecular weight knob up to 11 and force it to make as much wax as possible.

Yes, people claim to have fancy F-T processes and catalysts which make a narrower molecular weight distribution, and it’s true. But they all come with extra capital cost or other problems- and none of them avoid both LPG and waxes.

So- why would we reach for this fundamentally broken piece of tech? One reason and one reason only: it sucks less than burning plus CCS, especially if you can make hydrogen cheaply enough. It allows you to make molecules that you can recycle back into the refinery and convert to products.

You would probably put a methanol plant on there too. Methanol is the one process which can take CO2 (plus H2) as a feed and not just become economically craptastic as a result. The reason is that the methanol synthesis catalyst is a water-gas shift catalyst, catalyzing the reactions that interconvert CO and CO2 with H2 and H2O:

CO + H2O <====> CO2 + H2 + heat

So while some CO2 reacts directly with H2 to make methanol, the rest gets converted to CO which also reacts readily with H2 to make methanol. The trick, however, will be to ensure that none of the product methanol made from fossil feedstocks in your refinery, ends up being burned at its end of life. You’d need to keep the fossil methanol only for applications which remain unburned. Certain plastics use methanol as a monomer/reagent and fit that bill if we landfill them at end of life, after maximal recycling, rather than burning them.

What About Natural Gas?

Natural gas processing is much simpler than petroleum processing, and the vast majority of the gas is methane. Fossil methane is nearly worthless for anything other than two things: its heat energy content, which is useless post decarbonization due to the cost of CCS, and making syngas- to make fuels which themselves would be useless post decarbonization.

What does that mean? The fossil methane business is dead post decarbonization. We simply won’t produce much gas any more. We might store some for emergency power generation, and make a bit of it into hydrogen and valuable carbon products by pyrolysis, but that’s about it.

Ethane and propane are also found in natural gas and are more useful to make ethylene and propylene and hydrogen. But without a use for the methane, separating these gases out and re-injecting the methane without leaking any is going to look pretty unappealing. My feeling is that we’ll meet most of our ethylene and propylene needs by naphtha cracking in a decarbonized future instead.

So if you’re in the gas industry, as an employee or investor, best to bail now. No, hydrogen won’t rescue you either. You’re going out of business. It’ll take a long while though- longer than it really should, if we were truly serious about decarbonization.

What About Coal?

We do make some materials from coal tar, but none of them can’t be made by other means. And without a need for coke- which we can’t use post decarbonization for anything- there’s no reason to make coal tar. I suppose you could drive off the coal tar and just bury the coke again, but where’s the fun in that? Doubt it will make sense, or money. So we can also bid coal a fond good riddance, at long last.

What Will All This Cost?

Don’t kid yourself: petrochemicals refining will be a whole lot more difficult post decarbonization. That means the processing will cost more. And the scale of petroleum refining will drop, and new investments will be needed. We’re talking about a great deal fewer refineries, purpose built from scratch or almost totally re-built from existing. Huge amounts of money will be needed.

However, demand for petroleum will plummet. Distillates demand is already being affected by electric vehicles- more the 2 and 3 wheeled variety than EV cars and trucks, but they’re coming too. When demand drops to 15-25% of current, only low cost producers will remain in the market. That will drop costs of raw material, because its major market is going bye-bye.

So: petrochemicals and plastics will be more expensive to make, from cheaper feedstocks. Will that mean cheaper, or more expensive stuff? My bet is that it’s more expensive on the whole. And that’s good- expensive means we’ll be motivated to conserve and recycle them better.

Will This Really Happen?

We’d better hope it does.

A growing economy is nice and all, but we need a stable climate. And we won’t ever have one if we keep burning fossils. And while we can keep making materials and chemicals from biomass that the earth deoxygenated for us over millions of years, we need to do that without the burning. We know how. Not a single new invention would be needed- and there would be LOTS of them if we decided to go that way which will help out, a lot. Will we be wise enough to actually follow through on it?

For my kids’ sake, and theirs, I certainly hope so.

Don’t let anybody tell you that we have to do without the chemicals and materials we make from petroleum today. They’re selling you a bill of goods, likely with the intent to make you feel good about burning fossils for longer- or they’re selling you some other ideological point of view. Don’t worry- if we choose to make the economics work via carbon taxes, or to force the emissions out of the system via regulation, we chemical engineers have got your back. We’ll adapt. We have the technology to do it. But without the economic drivers to make it pay, the investments required will not be made, and it simply will not happen.

Disclaimer: this article was written by a human, and humans are known to make mistakes. If I’ve stuffed something up, and you can show me that I’ve done so via good references, I’ll correct my work here with gratitude. I care more about getting the issue right than about being right personally.

If however you don’t like what I’ve written because I’ve taken a dump on your precious idea- F-T running on CO2 and water and electricity for instance- then feel free to contact my employer, Spitfire Research Inc., who will be very happy to tell you to piss off and write your own article.

More Reading

…or watching, in this case- I talk about these ideas with david borlace on his brilliant YouTube channel:

Why e-fuels, and their epitome, e-methane, are dumbass:

How to make high value products from methane:

…and everything else: