Microbial Hydrogen From Depleted Oil Wells: Scaling, Costs & Challenges – CleanTechnica

---

Support CleanTechnica’s work through a Substack subscription or on Stripe.

---

Someone asked me recently about a Houston-based startup called Gold H2. That was enough to send me down the rabbit hole. After a few hours of reading, checking some chemistry, and doing a few back-of-the-envelope calculations, I had enough material that it was worth pulling together in one place. This is not the Australian company with the same name that is drilling for naturally occurring hydrogen. This is a private spin-out from Cemvita Factory that is trying to make hydrogen underground in depleted oilfields by feeding the right microbes and letting them do their work.

Gold H2 calls its approach Black 2 Gold. The process involves injecting nutrients and hydrogen-producing microbes into old oil wells, letting the reservoir sit for a period, and then producing gas back through the same well. The microbes are meant to break down residual hydrocarbons in the reservoir into hydrogen and other gases. The company says this avoids the need for new drilling and makes use of existing oilfield infrastructure. In concept, it is similar to microbially enhanced oil recovery, but instead of trying to get more oil out, the aim is to get hydrogen.

Microbially enhanced oil recovery is a set of techniques that use microorganisms to improve the flow of oil from a reservoir. The microbes can be injected into the reservoir or stimulated in place with nutrients, and they act in several ways. Some species produce gases such as carbon dioxide or methane that help push oil toward the wellbore. Others create biosurfactants that reduce the surface tension between oil and water, allowing trapped oil droplets to flow more easily. Certain microbes can change the viscosity of crude oil by breaking down heavier hydrocarbons. These methods have been tested in various oilfields since the 1980s, often with mixed results because microbial activity depends on reservoir temperature, salinity, pH, and nutrient availability.

In practice, microbially enhanced oil recovery has remained a niche tool applied in specific reservoirs where conventional secondary recovery methods have been exhausted or are uneconomic. That serves as a red flag for promises of using similar technologies to extract hydrogen instead of oil.

Only a fraction of depleted oil wells are likely to have the combination of conditions needed for microbial hydrogen production. Many old reservoirs are either too hot or too cold for the target microbes, or have salinity levels in the formation water that exceed microbial tolerance. Others have low residual oil saturation after decades of production or waterflooding, leaving little feedstock for conversion. Sulfate contamination from earlier water injection programs is also common and can lead to hydrogen loss through unwanted chemical pathways. When these factors are combined, the pool of viable candidates probably narrows to a minority of depleted wells, with the most promising found in reservoirs that still hold moderate residual oil, have formation water of moderate salinity, sit in the right temperature range, and have not been heavily altered by incompatible injection fluids.

In 2025, Gold H2 ran a field trial in California’s San Joaquin Basin. They reported that the produced gas stream contained about 400,000 ppm hydrogen, which is 40% by volume. The rest was not broken out in detail, but likely included methane, carbon dioxide, nitrogen, and possibly trace hydrogen sulfide, a very nasty gas that has to be carefully managed. There was no disclosure of the total gas flow rate, the absolute amount of hydrogen produced, or whether any of it was purified and stored. Without that information, it is difficult to assess how close they are to their stated target of producing hydrogen for under $0.50 per kilogram.

Hydrogen content in gas streams is often reported in volume percent. For gases at the same temperature and pressure, volume percent is the same as mole fraction. That is useful for pipeline and combustion considerations, but it can be misleading when thinking in terms of mass. Hydrogen is very light. A mix that is 40% hydrogen and 60% methane by volume works out to only about 7.7% hydrogen by mass. If the other 60% were carbon dioxide, the mass fraction of hydrogen would drop to about 3.6%.

Because Gold H2 did not release any flow rates, I bracketed some reasonable values based on what is common for end-of-life wells. In the United States, a marginal gas well, often called a stripper well, produces 90 thousand cubic feet per day or less. Many produce far less. I used three cases: 50 thousand cubic feet per day, 200 thousand, and 1,000 thousand. For a mid-case of 200 thousand cubic feet per day total gas and 40% hydrogen by volume, that works out to about 204 kg of hydrogen per day. At 300 operating days per year that is roughly 61 tons of hydrogen per year.

For depleted oil wells producing microbial hydrogen, 50 and 200 Mcf/d are sensible central cases. A rate of 1,000 Mcf/d is possible in an unusually favorable well but should not be the basis for planning. These wells are not high-pressure gas reservoirs. They have already lost most of their drive energy and often have high water saturation, with oil-wet rock limiting gas movement. The huff-and-puff cycle further reduces average output because production only happens between injection and soak periods. Even if the well briefly flows in the high hundreds of Mcf/d during production, the cycle average will be lower.

Many repurposed oil wells with gas output end up at or below the U.S. stripper threshold of 90 Mcf/d, which makes 50 Mcf/d a reasonable low case. A rate of 200 Mcf/d would indicate a well with decent permeability, a short lift to surface, low backpressure, and a productive microbial reaction. Achieving 1,000 Mcf/d would take unusually good rock, a high-quality completion, short flowlines, little water blockage, and strong transient drawdown, making it more suitable for a stress-test scenario than a base case.

As with my assessment of the potential of white hydrogen, i.e. naturally occurring hydrogen that is mined, extraction and purification is only half of the cost challenge, with compressing or liquefaction and use being costly as well. That’s why I articulated the value proposition for any actually recoverable white hydrogen reserves being for industrial feedstocks used in plants close enough to the hydrogen extraction site to allow short pipelines at most. We use 85% of hydrogen at the point we make it today because of that, and there’s no reason to believe that’s going to change. Hydrogen isn’t a useful energy carrier and it’s not a useful transportation fuel, so we’ll continue to make it or potentially extract it at point of use and turn it into something useful like ammonia or use it to hydrotreat biofuels.

To put the potential volumes into market context, ammonia plants in Louisiana, Oklahoma and Texas produce about 7.7 million tons of ammonia per year. Making that ammonia requires about 1.37 million tons of hydrogen annually. At 61 tons of hydrogen per well per year, it would take around 224 wells to meet 1% of that regional demand. Meeting 10% would require over 2,200 wells. Those figures are before any separation losses, and they assume steady output, which is unlikely for a huff-and-puff process that alternates injection, soaking and production.

Purification is unavoidable if the goal is to supply hydrogen for fuel cells or ammonia synthesis. Pressure swing adsorption and membrane systems have costs that rise as the inlet concentration falls. An 80% hydrogen inlet might add about $0.30/kg to the cost. At 40% hydrogen, separation might add $0.60/kg. At 20% hydrogen it could be $1.20/kg. If Gold H2 is starting at 40% by volume, purification alone could push their total production cost over the $0.50/kg target they have stated. That is before considering capital recovery, operating expenses for injection and handling, and any financing costs.

The composition of the other 60% of the gas matters for economics and climate impact. If it is mostly methane, there is potential value in selling it or using it to power operations. Of course, that’s perpetuation of fossil fuels, which isn’t exactly a recommended activity. If it is vented or leaks, the climate impact could be significant. If it is mostly carbon dioxide, there is little market value and it will still need to be managed. Both cases make purification more complicated. An actually sustainable practice would be to pump either gas back under ground in a different well, adding to the cost case.

Gold H2’s concept has some strengths. It aims to reuse existing oil and gas wells, which reduces the capital cost compared to building new infrastructure. It uses biological processes that work at reservoir conditions rather than energy-intensive surface equipment. The main questions are about scale, stability and economics. Without knowing how much hydrogen a well can produce over months or years, and what it costs to get that hydrogen to market at the required purity, it is hard to judge whether it is a dead end, is a niche solution for a few reservoirs or a pathway to significant volumes. My math suggests the first option.

As a reminder, while I’m a broad spectrum nerd who has spent a lot of time working to understand domains most haven’t, I’m not a subsurface oil and gas engineer and do make mistakes. It’s possible that my reading of the data is wrong, but I’m comfortable I have enough right with my assessment to consider Gold H2 to be deeply unlikely to ever be economically viable as a producer of hydrogen.