Radium- and Thorium-Contaminated Waste at Fernald

The Fernald site, originally called the Feed Materials Production Center, is located approximately 20 miles northwest of Cincinnati, Ohio. Its main mission was to produce uranium metal for use in the US nuclear weapons program. From 1952 until 1989, nine plants on the site processed a wide variety of uranium-containing materials, such as ore concentrates and recycle materials, and produced large quantities of radioactive and toxic wastes. Wastes were dumped in pits or scrap piles or stored in drums or silos. Production at Fernald ended in July 1989. Cost estimates for total Environmental Management activities at the site have ranged from $3 billion to $5.4 billion, and activities could stretch to the year 2030.

The most dangerous emissions have historically been in the form of radon-222 from Silos 1 and 2, which are tanks that contain large quantities of waste containing radium-226 from the processing of uranium ore. Silo 3 also contains radium-bearing wastes though at lower concentrations. Official studies have noted that there is concern about the structural integrity of the silos and the threat of roof collapse. This and the threat of radon gas emissions makes the remediation of these silos crucial to protection of the health of the communities around Fernald and to limiting worker exposure. A structural failure of the silos and discharge of their contents into the soil could also threaten the groundwater of the region over the long-term. (see table below for estimated radionuclide content of silos).

Actions taken so far have been, at best, temporary palliatives. At worst, they have been complete failures that have increased risks due to delays. For instance, in 1991 a layer of clay was added to the top of the material in the silos to try to reduce radon emissions. This succeeded temporarily, but the clay proved not to be an effective cap over the waste, and emissions are now back up. Furthermore, the clay will considerably complicate the job of actually emptying and decommissioning the tanks. We believe that a better approach would have been to install a tornado-resistant enclosure, estimated to cost $5 million and require 10 months to implement. This would have reduced short-term emissions without complicating long-term remediation. But adding a layer of clay was cheaper in the short-term.

During 1998, the DOE came up with yet another plan. Since the silos are deteriorating, it now wants to build a new set of tanks so that the wastes can be transferred to them. If successful, this would create new "temporary" storage that would eliminate the risk of short-term and medium-term large radon releases. However, the transfer of wastes could prove to be technically difficult, as it has in past attempts, due in part to the nature of the wastes. Thus, DOE is pursuing another untested approach on a large scale without having done sufficient preliminary work. Moreover, the problems with waste handling in the failed pilot plant project should have made the DOE more cautious about launching into a massive project on silo waste transfer without more technology testing.

For long-term remediation the DOE chose, in a December 1994 Record of Decision, to "vitrify" the silo wastes (though by that time design of a pilot vitrification plant was already underway). DOE unfortunately uses the term "vitrification" in two quite different ways. The first refers to mixing a relatively small quantity of radioactive material into a large volume of molten glass and making real glass logs laced with radioactive materials. The second is to take a large volume of radioactive waste, consisting principally of various kinds of soil, and convert the mix into a glass-like substance. In the former case, the glass-making is well-understood. Only the technologies to prepare and mix the radioactive materials with the molten glass need to be developed (and in some cases they already have been). In the latter case the composition of the "glass" cannot be controlled, and hence the "vitrification" technology itself needs to be developed. DOE's plans at Fernald involved the latter, much more uncertain, type of vitrification. (In this article we use the term in its latter meaning - conversion of radioactive soil into a glassy material.)

This project failed completely, largely as a result of serious technical mistakes by DOE and its contractor, Fluor Daniel Fernald.

Despite the fact that the waste in the silos was not fully characterized and a novel vitrification technology was being proposed, DOE and the contractor decided to "fast-track" the pilot plant project by proceeding with simultaneous design and construction. This led to significant problems. For example, the melter delivered by a subcontractor did not match the preliminary designs that Fluor Daniel Fernald had used in its construction of the rest of the pilot plant.

The technical failures at Fernald have been as bad as the managerial failures. Materials used in the melter, particularly molybdenum disilicide "bubbler tubes," were incompatible with the high-lead content of the waste. As a result, the melter was destroyed part-way through the first of two phases of pilot plant testing. This dramatic failure is of even greater concern because project personnel identified the exact issue that led to destruction of the melter during technical reviews, yet it was not resolved.

Cost and schedule increases

Contractor and DOE failures led to significant cost increases for the Vitrification Pilot Plant. In February 1994 the pilot plant effort was estimated to cost $15.8 million. By June 1996, the cost estimate for completion of all Pilot Plant testing was $66 million - a four-fold increase. Through November 1996, $50 million had been spent. In December 1996, during Phase I of testing (which only involved non-radioactive simulants of the waste in the silos), the accident that destroyed the melter rendered the pilot plant useless for future work.

Had the melter not failed, the $66 million estimate in June 1996 would surely have been exceeded because major modifications would have been necessary to prepare for tests involving actual radioactive waste from the silos. The plant, as built, could not have handled radioactive materials without high levels of worker exposure.

As costs mounted during design and construction of the Pilot Plant, DOE and Fluor Daniel Fernald began to revise their estimates for the full-scale vitrification facility. In January 1996, cost estimates for the whole project had more than tripled from $92 million to over $300 million. In April 1997, Fluor Daniel Fernald estimated the total cost to range between $376 and $563 million (This estimate involved substitution of cementation for vitrification as the treatment method for Silo 3 waste.) Additionally, the estimated completion (including decontamination and decommissioning) had slipped by nine years - from 2002 to 2011.

Technical, managerial, and financial shortcomings early on in the Pilot Plant project led to attempts to abandon the vitrification treatment selected in the Record of Decision. Changes from vitrification to cementation for all or part of the waste have been proposed even though there seems to be no established, essential technical obstacle to proceeding with a vitrification program for wastes in all three silos. Vitrification, if successful, would likely provide for better waste isolation and smaller final waste volumes.

These changes to the remediation program are being pursued in large part due to supposed cost savings, yet DOE has not made a proper comparison of the alternatives, nor has it adequately explained why treatment cost estimates have changed drastically from those cited in the Record of Decision.

IEER believes that DOE should take the following steps to get its program for treatment of the radium- and thorium-contaminated wastes contained in the Fernald silos on the proper track:

1. The entire remediation program for the silos needs to be put on a sound financial and technical footing. Given prior egregious cost misestimation and escalation and the fact that the project now is estimated to involve hundreds of millions of dollars, a thorough independent review of both the accounting and engineering aspects needs to be carried out before any cost increases are granted.
2. The waste in all three silos should be more thoroughly characterized. Development of vitrification techniques for the waste in Silos 1 and 2 should proceed along a focused, targeted effort in a one- to two-year time frame.
3. DOE should not rush into alternative treatments, such as cementation for Silo 3, given DOE's own evaluation of problems and difficulties with such technologies. Vitrification should still be given top priority.
4. A modular approach to vitrification, which would allow for operating flexibility in order to treat a potentially heterogeneous waste feed, is advisable.
5. DOE should more carefully consider building a tornado-proof roof over the existing silos and constructing a single new tank to establish the feasibility of waste transfer as an alternative to its current plan of building a new set of tanks for holding wastes.