Every few years, a new “breakthrough” arrives promising to upend agriculture as we know it. Lab-grown meat, insect protein, vertical farming—each comes with bold claims about saving the planet, feeding billions, and, conveniently, generating tidy profits along the way. Now we’re told that bacteria fed on methane—yes, the same gas routinely framed as a climate villain—can be turned into a superior protein source that outcompetes soybeans and fish meal on both environmental and economic grounds.
A recent press release tied to a peer-reviewed paper in Carbon Research makes precisely that claim. Engineers from Beijing University of Chemical Technology, led by Yanping Liu and Ziyi Yang, present a life-cycle assessment (LCA) comparing three protein supply chains: soybean meal, fish meal, and microbial protein derived from methane-oxidizing bacteria (MOB). Their conclusion is unambiguous: the microbial route wins—by a lot.
According to the release:
“Shifting to MOB protein shrinks overall ecosystem damage by 88% relative to standard soybean farming.”
“The microbial route drops negative human health impacts by 41% when compared to the emissions and processing burdens of the fish meal industry.”
“The MOB protein system generated the highest net present value ($3.40 million) and secured a dominant 51% return on investment.”
On the surface, it sounds like a silver bullet: take a greenhouse gas, feed it to microbes in controlled vats, and produce high-quality protein with minimal land, water, or ecological footprint—while turning a handsome profit.
That’s the narrative. Let’s take a closer look at what’s underneath it.
The Allure of the Bioreactor
There’s no denying the conceptual elegance here. Methane, often portrayed as a potent climate forcing agent, becomes a feedstock. Bacteria consume it, grow rapidly, and produce protein-rich biomass. No farmland. No fishing fleets. No fertilizer runoff. No deforestation.
The press release leans heavily on this contrast:
“Because the bacteria grow in controlled vats, the method virtually eliminates the need for arable land and fresh water, effectively halting the deforestation and marine depletion associated with standard protein sourcing.”
That’s a powerful image—sterile steel tanks replacing vast agricultural landscapes. It appeals to a certain technocratic vision of the future where messy, variable natural systems are replaced by controlled industrial processes.
But this is where the first layer of skepticism is warranted: life-cycle assessments are only as good as the assumptions built into them.
Life-cycle assessments have become the go-to tool for evaluating environmental impacts. They can be useful—but they are also highly sensitive to system boundaries, data inputs, and methodological choices.
The study compares three systems:
- Soybean meal production
- Fish meal production
- Methane-based microbial protein
Each of these has very different characteristics, and how you define the “system” can dramatically alter the outcome.
For example:
- Soybean farming can be modeled using worst-case assumptions (deforestation in tropical regions, heavy fertilizer use) or more moderate ones (established farmland, improved practices).
- Fish meal can be framed as environmentally destructive or as part of managed fisheries with varying levels of sustainability.
- Microbial protein, meanwhile, is often modeled under idealized, optimized industrial conditions that may not yet exist at scale.
This asymmetry matters.
The press release notes that microbial protein production is “energy-intensive,” but quickly dismisses that concern:
“While producing microbial protein is an energy-intensive process, the trade-offs are incredibly favorable.”
That statement hinges entirely on how the energy is sourced and accounted for. If the process relies on low-cost, low-emission energy (as many models assume), the environmental footprint looks attractive. If not, the picture changes quickly.
Energy: The Missing Piece in the Narrative
Let’s focus on that “energy-intensive” qualifier, because it’s doing a lot of work here.
Growing bacteria in bioreactors at scale requires:
- Continuous gas supply (methane, oxygen)
- Mixing and agitation
- Temperature control
- Downstream processing (harvesting, drying, refining)
All of these steps consume energy—often significant amounts.
The study highlights a particular methane purification method:
“Pressure Swing Adsorption (PSA) proved to be the most robust method, cutting resource depletion by over 140% compared to alternative membrane technologies.”
That’s an impressive-sounding number, but it raises questions. A “140% reduction” suggests a relative comparison between modeled scenarios, not an absolute measure. It tells us which option is better within the model—not whether the overall system is truly low-impact in real-world conditions.
More importantly, none of this eliminates the fundamental dependence on energy inputs. If those inputs come from fossil fuels, the environmental benefits shrink. If they come from intermittent sources like wind or solar, the system must deal with variability—something continuous bioprocesses don’t handle well.
This is a recurring theme in many techno-optimistic proposals: the energy question is acknowledged, then quietly set aside.
There’s also an interesting rhetorical twist in using methane as the input.
Methane is frequently highlighted in climate discussions due to its higher short-term radiative forcing compared to CO₂. The idea of converting methane into something useful—protein, in this case—fits neatly into a “waste-to-value” narrative.
But where is this methane coming from?
The study doesn’t appear to focus on capturing fugitive emissions from landfills or agriculture (which would be a more compelling environmental case). Instead, it treats methane as a feedstock—something to be sourced, purified, and delivered.
That raises practical questions:
- Is the methane captured from existing emissions, or produced specifically for this process?
- What are the costs and losses associated with capture, transport, and purification?
- How scalable is this supply chain?
If methane must be produced or extensively processed to feed these systems, the environmental advantage becomes less clear.
The press release emphasizes strong financial performance:
“The MOB protein system generated the highest net present value ($3.40 million) and secured a dominant 51% return on investment.”
Those numbers are attention-grabbing, but they come from techno-economic modeling, not operational facilities at scale.
Anyone who has followed the history of alternative energy or agricultural technologies will recognize the pattern:
- Model a system under favorable assumptions
- Optimize parameters (efficiency, scale, costs)
- Present the resulting economics as evidence of viability
Real-world deployment often tells a different story.
Costs that are frequently underestimated or omitted in early models include:
- Capital expenditures for large-scale facilities
- Maintenance and operational complexity
- Supply chain logistics
- Regulatory compliance
- Market acceptance
And then there’s the question of competition. Soybean meal and fish meal are not static targets—they benefit from decades of optimization, global infrastructure, and economies of scale.
Replacing them is not just a technical challenge; it’s an economic and logistical one.
One of the strongest claims in favor of microbial protein is the reduction in land use:
“The method virtually eliminates the need for arable land and fresh water.”
That’s true in a narrow sense—bioreactors don’t require fields. But this framing assumes that land use is inherently problematic and that industrial processes are inherently preferable.
There’s a trade-off here:
- Agriculture uses land, but it is distributed, adaptable, and often integrated into ecosystems.
- Industrial bioreactors concentrate production into centralized facilities that depend on continuous inputs and infrastructure.
Neither system is impact-free.
Replacing agricultural land use with industrial production shifts the burden—it doesn’t eliminate it.
This isn’t the first time we’ve seen sweeping claims about replacing traditional agriculture.
Consider:
- Biofuels: once promoted as a clean alternative, later criticized for land use and food price impacts
- Lab-grown meat: still struggling with scale and cost despite years of hype
- Insect protein: niche adoption, but far from replacing conventional sources
Each of these began with strong modeling results and confident projections.
The microbial protein concept may well find a role—particularly in specialized applications or regions with limited agricultural capacity. But the leap from “promising technology” to “cornerstone of the global feed market,” as the press release suggests, is a large one.
The tone of the press release is worth examining:
“A greener, more profitable alternative to farming”
“The verdict leans heavily in favor of the bioreactor”
“A highly lucrative, environmentally superior reality”
This is not the language of cautious scientific reporting. It reads more like a pitch—one that emphasizes certainty and downplays uncertainty.
That doesn’t invalidate the research, but it should prompt a more careful reading.
Scientific progress rarely arrives in the form of clean, decisive victories. It’s incremental, messy, and often constrained by practical realities that don’t show up in models.
If methane-based microbial protein is as transformative as claimed, we should expect to see:
- Pilot plants operating at scale, with transparent data on energy use, costs, and outputs
- Independent replication of results across different regions and conditions
- Market adoption, with feed producers choosing this product over established alternatives
- Full accounting of inputs, including methane sourcing and energy supply
Until then, the technology remains in the category of “promising but unproven.”
Final Thoughts
There’s nothing inherently wrong with exploring new ways to produce protein. Innovation in agriculture and food systems is both necessary and inevitable.
But claims of sweeping superiority—environmental and economic—should be treated with caution, especially when they rely heavily on modeled scenarios and optimistic assumptions.
The idea of brewing protein from methane is intriguing. It may even prove useful in certain contexts. But the current presentation leans more toward aspiration than demonstration. As always, the real test won’t be in the pages of a journal or the lines of a press release—it will be in the messy, unforgiving world of large-scale implementation.
And that’s where many elegant ideas run into inconvenient realities.
Ya but can it replace a rib eye?
I agree a good ribeye is hard to beat, but I’ll go for a T-bone or fillet mignon occasionally.
So methane is a energy source for bacteria who wold have thunk it !
Looks like smoke and mirrors to me .
Comparing soy meal or fish meal against a dried bacterial mess WTF
High capital and energy costs in processing
How many batches do you loose because another organism gets in ?
Look where it comes from a country that has been trying to destroy western civilization for decades.
Good points. I bet Listeria loves that shit, although it appears they employ Pasteurization.
I bet does and lots of others
Maybe the greenies can feel a little virtuous knowing that the govt has allowable insect parts in commercial baked goods. 75 insect parts per 50 grams of baked goods (oh and one rodent hair). For chocolate it’s 60 parts per 100 grams and one rodent hair.
Just eat a lot to feel more virtuous I suppose:))
My first question was what’s the nitrogen source? Second question was why methane as a carbon source?
Anyway, from the paper:
“The main energy input costs in the microbial protein production process were electricity and water. The electricity consumption was mainly divided into two parts: (a) the electricity required for controlling the temperature during microbial fermentation; (b) the electricity used for stirring, pasteurization, separation, and drying during the fermentation process. Water consumption was mainly used in the fermentation process to maintain a constant hydraulic retention time, with an assumption of 10 employees working at the plant.
Resource input costs in System 3 primarily included methane, oxygen, nitrogen, and phosphorus. According to Pikaar et al. (2018), the production of 1000 kg of MOB-derived microbial protein required 300 $ worth of methane, 109 $ worth of oxygen, and 149 $ worth of nitrogen and phosphorus nutrients. When including electricity costs, the total came to 1162 $/t of microbial protein. These values were adjusted according to the scaled material and energy inventories derived from the proportional upscaling procedure and the downstream processing parameters taken from similar industrial microbial protein study (Järviö et al. 2021), used as baseline.”
So they are adding fertilizer to the vats. And for some metabolic reason oxygen must be supplied. I guess that makes sense if you need to do something with the 4 H’s in CH4.
Seems like a really hard way to make food. And you’re diverting a nice clean fuel as a carbon source. Why?
Let’s stick with natural sources of protein.
That is all.
Good points. Andy why not just oxidize the methane (burn it) in a power plant? You get the benefit (?) of converting the high GHG potential methane into CO2. You get the extra electricity and you use it in existing machines at no extra capital cost (except for the gathering technology).
But that sounds too practical and boring. No one gets rich.
And just exactly how much global warming is this so-called
potent short term radiative forcing agent going produce by
the end of the current century?
If anyone wants to claim that it’s more than a tenth of a degree,
they should pipe up and show their source and their work.
How do they plan to remove methane, measured in ppb, from the atmosphere? that is the question. 😉
No, the question and issue is how is Climate Science
going be convinced throw to in the towel? Part of that
scenario is making it painfully clear the warming from
methane is virtually immeasurable. People need to
understand that methane is on track to cause about
0.08 C°.of warming by 2100. I’ve been asking for
several years now for some one to pipe up and tell
me I’m wrong about that and show their source and
work. So far that hasn’t happened.
If we lit the cow farts, would that solve this supposed problem?
WHY ??
I can’t of anyone sane that would eat that sort of crap.
“operational facilities at scale”
This is the obvious problem with these ideas. What does “scale” mean in the context of making a dent in the production of protein? What building size, shape, and number are need to replace the protein in a field of cattle?
Here’s a link to a great idea that, enountering reality, ended badly:
Like the inquisition, no-one expects Soylent Green to occur
and how long until we find out it’s people. (joking…. hopefully)
In the abstract of the paper they say “PSA emerged as the favorable method, offering the lowest environmental impact and the highest operational robustness, reducing resource depletion by over 140% compared to Membrane Technique.”
What the paper actually says is “In contrast, membrane technology showed the highest environmental impact across all three endpoint categories, with its resource consumption impact even 140% higher than that of PSA, making it the least preferred option.”
A 100% reduction would eliminate resource consumption. A 140% reduction does not make sense.
If PSA is at 100, then membranes are 140% higher, at 240. That is a 58% reduction (140/240).
The paper is tagged as “Data/statistical analysis” but it is written by folks who are so innumerate they do not understand percentages.
Looking at table 5, I suspect they have cherry-picked optimistic assumptions for build-out, equipment, maintenance, labor, and operations to get the 51% ROI. If someone promised you an investment that had a 51% interest rate would you give them your life savings?
In the 1970s ICI had a plant in the NE of England producing bacterial protein from methane (North Sees gasoline
Finger trouble.
North Sea gas. The protein looked a bit like milk powder and one of its intended uses was as a replacement for fishmeal in animal diets.
IIRC it was very important to prevent contamination of the good bacteria by other bacteria such as E.Coli.
On reflection the methane might have been converted to methanol as the feedstock.
Turned 80 today so memory needs updating.
If you find a way to update it, please let us know. I imagine I am not the only one who needs it.
I believe you are correct that it was actually methanol which was used to make the protein.
We’ve had enough of things that come out of Chinese laboratories and when did the Chinese ever care about the environment.
One cow fart shake to go, please, with extra booger sprinkles on top.
Naw, I’ll just eat meat, eggs and dairy. Oh, and don’t forget the beans!
Wow, my mouth waters just thinking of that delicious bacteria! And I’ll feel virtuous too, helping to save the planet.
Bacteria. It’s what’s for dinner.
By Jove! They’ve invented Marmite.
If you burn the methane to produce CO2, and use that to fertilize plants, and feed the plants to cattle, doesn’t that accomplish the same thing? And the energy costs of processing the bacteria is eliminated, and you can do useful things from the energy of burning the methane.
Or, just burn the methane for energy and release the CO2 to allow plants to breathe more freely and produce still more energy.
It is not protein that is critical.
It is the mix of amino acids that comprise the proteins that are critical.
This is not addressed.