THE MOST TROUBLESOME ESCAPE MECHANISM for deep waste disposal is dissolution of the radioisotopes in local brines, followed by transport flow to the surface. In deep rocks, brines typically fill a few percent of the rock by volume. The brines are in cracks and in pores, but we know they communicate with each other since the pressure found in these brines tends to be the hydrostatic pressure, that is, approximately one bar (about one atmospheres) for every ten metres of depth. The rock (lithostatic) pressure is typically two to three times greater.

Oil and gas geologists will tell you that these brines are ‘stagnant’, but that only means that any movement is slow on a decade time scale. But the time scales for human safety are longer. Buried nuclear waste must be isolated for up to a million years. The tiny flow velocity of 1cm per year, undetectable by standard methods, could transport waste 10km in that time.

Traditional safety analysis for deep geology uses measurements of the rock permeability, and estimates of the ‘driving’ temperature, pressure and salinity gradients, to calculate escape times. The models include the possibility of undiscovered earthquake faults and future new ones, which could create zones along with higher permeability paths to the surface.

Nothing beats measurement: is there any hope of observing super-small brine flow rates? Methods have been developed over the past few years that can provide excellent estimates for water stagnation and geologists have shown that there are sites around the world that have had little movement for deep brines — many for the past 1.5 million years and some for hundreds of millions of years.

Strong Isolation

We use the term ‘strong isolation’ for a formation that has held its entrained water stagnant for over 100,000 years. Strong isolation is not necessary for a waste repository, but it offers a compelling argument that the waste will not be transported to the surface by brine flow.

Upward flow of deep brines is typically driven by convection, which is driven by temperature gradients, topography and stress changes. Deeper water heated from below becomes less dense than the water above it. It rises, like a balloon, until contact with the upper rock cools it. Then it flows around and down, to be reheated and rise again. This could, in principle, carry dissolved radioactive waste up to aquifers and the biosphere. But such convection cells can be stable — in strong isolation the density increase due to salinity must be greater than the density decrease due to temperature.

Measurement

New methods allow us to measure the stagnation of underground waters and how long it has been stagnant. One of the best indicators is chlorine-36, a radioisotope with a 300,000-year half-life. Cl-36 is produced underground from natural uranium and thorium in the rock formation. Because of their long half-lives (4.5 billion and 14 billion years for U-238 and Th-232, respectively), levels of uranium are constant over 100,000 to several million years. Alpha decay causes neutron emissions from other elements due to collisions. Neutrons are absorbed on chlorine Cl-35 to produce Cl-36 at a constant rate, so the Cl-36 concentration increases with time. After hundreds of thousands of years the rate of Cl-36 decay matches its rate of production and the stable state (‘secular equilibrium) can be calculated and is distinct for the rock mix. If measured Cl-36 levels match secular equilibrium, the Cl-36 has not been mixed with surface water for at least 1.5 million years.

Secular equilibrium in a basement rock can be shown to be isolated from an overlying layer, because if the brine is moving the Cl-36 will not reach its secular equilibrium concentration. There is a simple rule: if the level of Cl-36 is at half of the secular equilibrium level, then the water has been stagnant for about one half-life, that is, about 0.3 million years. If it is 75% of the secular equilibrium level, then it has been stagnant for two half-lives, or around 0.6 million years.

The most sensitive and precise method for measuring the Cl-36/Cl-35 ratio is accelerator mass spectrometry.

Here are some other ways to measure the age of the water that derive from the presence of uranium and thorium:

  • He-4. Most alpha particles from uranium and thorium decay come to rest in the rock and become helium atoms, which can escape as a gas or be carried away by brine flow. If, instead, it accumulates that gives a measure for the isolation of the formation. Measurements of helium in Germany have given isolation ages over 100 million years.
  • Ne-21. This rare isotope of neon is produced in the rock by alpha particles hitting nuclei of oxygen and fluorine. It is stable, so levels above the natural level (0.27%) indicates very long ages. In Canada it has shown brine in a basin is over a billion years old.
  • I-129. This radioactive isotope of iodine has a half-life of 16 million years, and is produced by spontaneous fission of U-238. From the known rate of fission, we can use the abundance of I-129 to estimate stagnation age. As with Cl-36, the I-129 in the deep subsurface reaches a balance between production and decay, and its concentration eventually reaches secular equilibrium. Because it has a much longer half-life than Cl-36, I-129 can provide information on the stagnation of brines for tens to hundreds of millions of years.

For short time periods C-14 (5730 year half-life) can be useful. Kr-81 (229,000 year half-life) is produced in the atmosphere, and measuring it at depth can indicate that surface water is moving downward. The levels of the other isotopes, including argon and xenon, can provide more information.

Another concern is unknown earthquake faults. Strong isolation offers evidence that their effect is not important. If such faults have not caused the groundwater to mix in a million years, then it is unlikely they will cause transport in the future. If the water has reached or is close to secular equilibrium, then no mechanism has transported the water away in 1.5 million years or more — not just earthquake faults, but also changes in climate that could change water pressure distributions and even trigger glacial scouring of the surface. (The last ice age ended only 20,000 years ago.)

It is true this method only measures the isolation of Cl-36, not of the water that carries it. If the chlorine interacts with the local rocks, its migration can be slowed. In fact, such interactions are thought to be small. But more importantly, the migration of Cl-36 that is more relevant for nuclear waste disposal than the rate of water flow. Cl-36, along with I-129, are two relatively long-lived isotopes in nuclear waste that have sufficient concentration in the waste, and long enough half-lives, that they present a radiation threat to future generations. Water may be the transport vehicle, but it is the chlorine itself whose upward movement through the rock poses the threat to safety.

Final thoughts

An important aspect of the strong isolation criterion is that good sites can be evaluated without depending solely on computer modelling of fluid flow and radionuclide transport. Such a model is valuable, but a measurement showing strong isolation can be a good confirmation of safety case.

Based on the strong isolation criterion, it is likely that there will be many safe sites around the country and around the world. Should we pick the best of these sites and bring all the waste there? Not necessarily. ‘Best’ is more than geology; it includes community support and transportation. There may be many sites, perhaps close to nuclear locations where the waste is currently stored.

Deep horizontal boreholes can also be beneficial for temporary storage of waste — an alternative to existing surface water pools and dry casks. Future retrieval from a storage borehole is not difficult; the drilling industry recovers objects from holes with ease. Deep Isolation in early 2019, demonstrated the recovery of a prototype canister. From an economic perspective, if the waste is to be stored for longer than 15 years, we estimate the borehole storage option to be cheaper than the surface, due to reduced security costs. After 25 years, it is half the cost of surface storage.

The basic concept of strong isolation is relatively simple to understand. We have found it has an appeal to the public, who are often distrustful of complex computer models and the assumptions that must be made in computer simulations. Just as radiocarbon dating has been successfully used to measure the ages of old bones, radio-chlorine and other methods can be used to measure the age of deep brines. If the brines have been stagnant the last million years, then they are unlikely to carry any waste to the surface for the next million years. Strong isolation may be the best indicator for the safety of nuclear waste that we have.

Finding a site

A geologic formation well-isolated from the surface and 1-2km deep offers an ideal location for the disposal of high-level nuclear waste and other hazardous materials. Such formations are typically too deep to be accessed by traditional mined repositories, but are easily reached by vertical access boreholes followed by horizontal storage sections.

Suppose a community is seeking a long-term solution to waste stored at a nuclear power plant.

To begin, we work with the local stakeholders to consider community interests. What is their perspective on where the waste should go? Do they want it transported out of their region, or do they prefer to minimise transportation? Would they be open to an option for disposal underground but within the perimeter of the nuclear plant itself? Is deep borehole burial a possible option?

If the community looks on horizontal borehole disposal favourably, then we study nearby drill-hole logs (these are, by law, publicly available in all US states). We might create a seismic profile of the site to see if the underground structure is similar to that at the closest drill holes. If it is similar, we drill a pilot hole, 15cm in diameter and 2km deep. In this hole we measure temperature and salinity. If the brine density increases sufficiently rapidly to suggest stability, we analyse the brine samples for isotopes that can indicate strong isolation. If we find secular equilibrium, and overlying formation with different levels of the isotopes, we have evidence for strong isolation. We would then run computer simulations to confirm our understanding.

If the samples do not show strong isolation, then there are several options for the community:

  • Do nothing. Keep the waste in surface storage until a better solution is found.
  • Drill deeper, in hopes of finding a strong isolation formation
  • Try a different site
  • Examine the results. Local mixing with a nearby formation, strongly isolated from the surface, may still be suitable for waste disposal.

Author information: Richard A. Muller, CTO, Deep Isolation Inc. Professor of Physics, UC Berkeley, emeritus