Experimental Investigation of In-Situ Hydrogen Segregation/Separation

Subsurface formations may play an important role in scaling up the hydrogen value chain. Hydrogen can be temporarily stored in depleted gas fields or produced in-situ by novel production methods such as Hydrogen Generation from Hydrocarbons Subterrain developed by Hydrogen Source AS or alternatives like microbiological conversion of biomass to hydrogen.

One of the key elements in subsurface hydrogen recovery is in-situ separation of gases. Due to the low density of hydrogen, the lightest of the gases, it is expected to accumulate at the top of the reservoir while other gases are distributed below. It is not expected that this separation will be complete, however the rate and extent of such separation has not been experimentally studied in detail.

For that purpose, a set of experiments were designed and performed at NORCE Norwegian Research Centre AS, as a part of the Hickory project owned by Hydrogen Source AS. The experimental set-up was built at a high-pressure laboratory at NORCE assigned to hydrogen experiments and comprised a core holder with a 30cm Berea core, pressure regulators and transducers, pumps, storage and accumulation tanks. To evaluate the rate and extent of gas (hydrogen) separation a series of experiments were run at different shut-in times ranging from a few hours to several days.

The experiments consisted of saturating a dry Berea core with a pre-mixed gas mixture (15% CO2, 50% H2, 35% methane) at the desired pressure and temperature (50 barg, room temperature). After a predefined shut-in the gas was produced, and effluent gas concentration profiles measured using gas chromatography. The shut-in times used were 1, 5, 17,41, 139 and 263 hours. To ensure repeatability several experiments were conducted twice.

The results show that:

1. The injected gas mixture separates in the core at the given conditions.

2. The in-situ separation process in the Berea core (∼ 300 mD) is very fast, with approximately 10% higher hydrogen concentration at the top of the core already after 1 hour.

3. Further separation is, however, undetectably slow, or stagnant, yielding similar concentration gradients for all experiments with longer shut-in times.

The experiment is currently being repeated with residual water present in the core to estimate the added effect of aqueous solubility on gas separation.

Further research is also needed, including effect of lithology, permeability and surface area as well as different gas mixtures to fully quantify and predict in-situ mixing and segregation of gases.