The chemicals industry today is largely downstream of the oil and gas industry, which provides the overwhelming majority of both its feedstocks and energy. As such, there’s a lot of operational overlap between oil and gas companies and chemicals companies: They’re both specialists in long-term deployment of capital, focused on finding low-cost resources, and subject to long-term commodity market fluctuations that they have to manage with various business and technology strategies. However, this linkage is becoming weaker over time as the chemicals industry is looking to decarbonize. One potential version of that is a chemicals industry that’s far more plugged into the electrical utilities sector, consuming electricity as a primary feedstock for electrochemical production of both hydrogen and hydrocarbons through CO2 electrolysis. This is not exactly a fringe theory: Groups like the International Renewable Energy Agency have forecast that by 2050, electrochemical methanol production will reach a few hundred million metric tons. What would it mean for the chemicals industry to be downstream of the utilities industry, in this case? The implications would be huge, and I can’t cover them all in this article. Instead, I want to propose a concept that’s been bouncing around my head for a few weeks: virtual chemical plants.
The utilities industry has been slowly adopting virtual power plant models over the last decade. The idea is fairly simple: There are a lot of small-scale assets on the grid (boilers, generators, batteries, etc.), which add up to a lot of aggregate power supply and/or demand. With digital tools, you can coordinate hundreds or thousands of these assets easily and use them in sync with each other to perform grid functions. This flexibility is enabled by the fact that electrons are highly fungible, and reducing consumption is essentially interchangeable with increasing production in several grid scenarios. If the chemicals industry of the future is more like the utilities industry, it stands to reason that there will be virtual chemical plants: lots of smaller assets acting in a coordinated manner to achieve some sort of technical or market function.
Does this sound crazy? I actually think we’re a lot closer to this then you might expect:
- The chemicals market is already a kind of virtual chemical plant — as any industry responding to market forces is — just operating over a much longer time frame and with sloppier control. Consider the current oversupply of olefins, which has crackers in Europe running at about 60% capacity. The virtual signal here is not digital control, but rather market forces like price. This is quite inefficient — more olefin supply is coming to the market over the next decade, which is not what the “control” signals would suggest.
- The chemicals industry would be subject to significantly more volatility when it connects to the grid. Oil and natural gas prices tend to move on a seasonal basis, while electricity prices fluctuate hour by hour, or even minute by minute. If chemicals companies build many wind farms or solar panels to power hydrogen or electrochemical production, they’ll have the option of selling electricity to the market. They’ll need some way to economically optimize their electricity and chemicals production and a virtual chemical plant can help with that.
- Production of chemicals is already getting more flexible and distributed. Conventional, large-scale cracking units really like to run at 100% uptime, due to their highly capital-intensive nature and the costs associated with turning them on and off. Many emerging production methods, including electrochemical production, are significantly more flexible: You can’t necessarily turn them on and off like a light switch, but they have a lot more agility to ramp up and down, and some approaches like fermentation can be switched from one type of production to another in a matter of days instead of years.
- In addition, there’s more flexibility in demand today. One key reason is recycling puts new supply in the market — producers of plastic products have the flexibility to choose between primary and recycled plastic. In addition, there are increasingly competitive plastic alternatives, like paper, for many applications.
So, how would a virtual chemical plant actually function? There are two basic functions: One is aggregation of demand from small actors into a single platform, so that those actors can be coordinated. The second is a mechanism for coordinating those actors to change behavior that’s separate from just a pricing signal — in no small part because you can’t use pricing to pay someone to not buy something. On a certain level, it’s simple: Centralized digital sales platforms already aggregate many small-scale sources of demand for chemicals and have the infrastructure to pay those companies to modulate their behavior. The digital technology, for the most part, already exists. The real questions are what are these small-scale sources of production and demand that would be aggregated, and what kind of economics would make this work? I have a few thoughts:
- The most straightforward is modulating the production of electrochemicals to sell electricity to the grid in a way that’s profitable for the chemicals company. The question is how flexible or responsive electrochemical production can be: Many electrolyzers take hours to warm up and cool down, so minute-to-minute flexibility for applications like grid stabilization is basically off the table. We are seeing progress on this front, however, as some of the U.S. Department of Energy’s recent projects will explore the use of green hydrogen for long-term load balancing.
- The really interesting stuff comes on the demand-reduction side. For utilities, this is easy: text people to turn off their lights, adjust their thermostats, or control when they charge their electric vehicles. Chemicals companies do have the power to modulate their customers’ behavior. If chemicals companies sell materials through digital platforms, they actually have a lot of freedom to dynamically price their materials to manipulate demand or just outright limit who can buy what and at what time. Obviously, big purchasers won’t be happy with this, but if chemicals companies start selling direct to many small buyers, they will have a fair bit of latitude to enforce their will on purchasers.
- Will chemicals companies pay customers not to buy plastics? Maybe not, but there are other forms of behavior modification that could be valuable. Of particular interest will be getting small actors to not manufacture unrecyclable plastic products — which may have significant costs for the chemicals company as extended producer responsibility (EPR) schemes come into effect. As recyclability is often a function of form factor, the plastic product manufacturer is in the driver’s seat. It may be cheaper for the chemicals company to pay producers to change their form factor than pay EPR fees, especially since any small producer may be except from EPR (and thus happy to continue pushing out negative-cost plastic products).
- Can chemicals companies get paid to not produce chemicals? It’s not as ridiculous as it sounds. There’s already a booming business in the forestry sector of companies getting paid not to cut down trees. Even in 2050, the chemicals industry will likely have a large volume of existing assets that it could run profitably but would have net-negative sustainability impacts due to carbon emissions, the production of harmful chemicals, increased landfilling of plastics, and the like. Why not try to find a purchaser for the act of not producing those problematic materials. This sounds kooky, but it may be the case that it’s easier to incentivize chemicals companies not to produce a problematic product like an unrecyclable plastic than it is to convince downstream users not to adopt it. For example, the relative value of plastic versus paper is quite high, but the relative profitability of plastic production could be pretty low — so why not just pay the company not to produce it rather than paying a packaging company to switch. I’m not convinced this is a good idea — the travails of carbon credit markets to not cut trees is not promising — but it is a possibility.
If there’s one point you should take away from all this, it’s that the dynamics and drivers of producing chemicals are likely to change significantly. The chemicals industry has been operating in a straightforward paradigm for the last 80–100 years: Demand growth is strong, so it must find the cheapest feedstocks and operate production as cost effectively as possible. That paradigm is likely to change significantly over the next 25 years, which could have much bigger consequences for how the industry functions and coordinates than you might think.