Journal of Nuclear Energy Science & Power Generation TechnologyISSN: 2325-9809

All submissions of the EM system will be redirected to Online Manuscript Submission System. Authors are requested to submit articles directly to Online Manuscript Submission System of respective journal.
kurtkoy escort

Review Article, J Nucl Ene Sci Power Generat Technol S Vol: 0 Issue: 1

Decommissioning the University of Illinois Nuclear Research Laboratory

William R Roy* and Richard L Holm
Department of Nuclear, Plasma, and Radiological Engineering, University of Illinois at Urbana-Champaign, USA
Corresponding author : William R Roy
Department of Nuclear, Plasma, and Radiological Engineering, 104 South Wright Street, Urbana, Illinois 61801-2983, USA
Tel: 217-840-9769
E-mail: [email protected]
Received: September 26, 2013 Accepted: November 10, 2013 Published: November 20, 2013
Citation: Roy WR, Holm RL (2013) Decommissioning the University of Illinois Nuclear Research Laboratory. J Nucl Ene Sci Power Generat Technol S1. doi:10.4172/2325-9809.S1-004

Abstract

Decommissioning the University of Illinois Nuclear Research Laboratory

The Nuclear Research Laboratory (NRL) at the University of Illinois at Urbana-Champaign was decommissioned in about 10 months after being in SAFSTOR for 12 years. A detailed site characterization study was conducted in 2005, and the removal of low-level radioactive wastes was begun in 2011. The first phase of the project was the removal and packaging of wastes in the building that were created from the operation or experiments conducted within the reactor. This activity was followed by the removal of interference from the upper part of the bioshield, and from the reactor tank that held the TRIGA reactor. The upper part of the bioshield was then sawed into blocks for disposal.

Keywords: TRIGA reactor; SAFSTOR; Stellite; Low-level radioactive wastes; Brokk; Asbestos; Soil samples

Keywords

TRIGA reactor; SAFSTOR; Stellite; Low-level radioactive wastes; Brokk; Asbestos; Soil samples

Introduction

The Nuclear Research Laboratory (NRL) on the campus of the University of Illinois at Urbana-Champaign (Figure 1) was in operation for about 30 years since the reactor first went critical in 1960. The reactor was a TRIGA (Training, Research, Isotopic production, General Atomic) Mark II training and research nuclear reactor that was manufactured by the General Atomic Division of the General Dynamics Corporation. The NRL was a steel frame, concrete building that was 13.7 by 24.4 m (Figure 2). The floor of the reactor room was a 15-cm concrete slab that was laid over undisturbed soil. The clearance from the reactor floor to the roof supports was 10.7 m. There was a tunnel 4.6 m below grade that connected the reactor tank with the mechanical equipment room. The reactor was initially operated with a maximum power rating of 100 kW using 20% enriched uranium. By 1967, upgrades and license amendments allowed for the operating limit to be increased to 250 kW. Additional details were given in Taylor and Holm [1].
Figure 1: The Nuclear Research Laboratory on the campus of the University of Illinois at Urbana-Champaign prior to complete decontamination and dismantling.
Figure 2: The floor plan of the Nuclear Research Laboratory. The building was 13.7 by 24.4 m.
After 11,567 mega-watt hours, the reactor was shut down permanently in 1998. In 1999, the reactor was officially placed in SAFSTOR (a time period in which the fuel has been removed from the plant, and the plant is being monitored prior to the complete decontamination and dismantling of the site). The Bulk Shielding Tank (Figure 3) was used for storage of the used nuclear fuel following shutdown. In 2004, the spent fuel was removed and shipped to the U.S. Department of Energy’s Idaho National Laboratory. In 2005, a Historical Site Assessment (HSA) [2] and a Site Characterization Report [3] were prepared to identify the various isotopes and estimate their activities at the NRL. The University planned to remove all radioactive materials from the NRL, demolish the building, and release the property for unrestricted use.
Figure 3: Cross section of the bioshield of the TRIGA Mark II nuclear reactor, and sources of activated concrete, graphite, and steel.

Characterization Results from Previous Studies

As given in detail by Charters and Aggarwal [4] and Taylor and Holm [1] the major isotopes of concern were activation products, tritium, and fission products that might have been released during an experiment with uranium-235-coated tubes. Once identified, the radioactive materials needed to be segregated from non-radiological components prior to disposal as low-level radioactive wastes.
As expected, the heavy concrete of the bioshield contained activated products (Table 1). Thirty samples of the concrete revealed that there was asymmetrical “radius of activation” around the reactor orifice. The principle activation products were cobalt-60 (60Co) most likely derived from trace amounts of stable 59Co in the concrete, and europium-152 (152Eu) and 154, likely from the activation of trace amounts of stable Eu. Stable cobalt and europium can be found in almost any type of concrete [5]. Iron-55 (55Fe) was also present which was derived from the activation of stable iron. The concrete contained steel rebar, shadow shields, angle irons and supports, and magnetite (FeO4) aggregate. The volume of the activated concrete was about 20.3 m3, or about 71,700 kg. The total activity of the activated bioshield concrete was 84 mCi (Table 1). Activation products were not detected in samples collected from the concrete floor [4].
Table 1: Activities of the primary activation products in the activated bioshield concrete [3].
A thermal column for generating thermal neutrons penetrated the bioshield and core reflector assembly. All of the reactor-grade graphite that was removed from the column was activated. The estimated total activity of the graphite removed was 7.03 mCi. A smaller thermal column (a thermalizing column) also contained activated graphite with 2.56 mCi of activity. The major activation products in the graphite were carbon-14 (14C) and 152Eu (Table 2). The total mass of graphite removed from the two thermal columns was about 4,490 kg. There was also a graphite reflector surrounding the core, but it was not sampled during site characterization.
Table 2: Activities of the primary activation products in graphite removed from the two thermal columns [3].
Many of the reactor components were constructed of aluminum alloy. For example, the bioshield enclosed the aluminum reactor tank and the tank contained the reactor core assembly. The mass of the activated portion of the tank was estimated as 212 kg [3]. The major activation products of concern were 60Co, 152Eu, 55Fe, and nickel-63 (63Ni). A rotary specimen rack was used to irradiate materials and geologic samples for research applications. Although the specimen rack was composed of aluminum, it contained a chain drive with Stellite bearings. Stellite is a cobalt-chromium alloy designed for wear resistance. As discussed in Taylor and Holm [1], the otherwise stable 59Co was activated into 60Co. It was estimated that the total 60Co activity of the rotary specimen rack was about 4 Ci (148,000 MBq), making it the most problematic component during the decommissioning of the NRL. Similarly, during the decommissioning of a TRIGA reactor at the University of Arizona, it was determined that a rotary specimen rack was the most activated component of the reactor. Composed of aluminum and Stellite, the total activity of the rack was 0.74 Ci which accounted for about 22% of the anticipated activity of wastes created during the project [6].
The presence of tritium was anticipated because there had been four primary coolant water leaks that occurred in 1971, 1974, 1976, and 1978 from either the reactor tank or the primary cooling system [1]. Tritium was widespread; it was detected in the bioshield, the containment floor, subterranean walls and floors, and the subsurface soil [4] for example, when present in samples at activities greater than the minimum detectable activity (between 132 to 297 pCi/g), the tritium activity in the bioshield samples varied from 1,321 to 59,906 pCi/g. Concrete and soil samples collected from the tunnel beneath the NRL ranged from 1,650 to 53,763 pCi/g.

The SAFSTOR Period Ends

After being in SAFSTOR for about 12 years, the decision was made in the fall of 2011 to proceed with the decontamination and decommissioning of the NRL. After the evaluation of four proposals submitted to an open request for proposals, a contractor was selected: LVI Environmental Services whom subcontracted with Enercon Services, Inc., and the Cutting Edge Services Corporation. LVI was responsible for the removal of all radiologically contaminated materials, equipment, reactor components, and the on-site soil.
Prior to mobilization of the LVI personnel, a public meeting was held in the Engineering Sciences Building. In addition to LVI and the University of Illinois, representatives of the Nuclear Regulatory Commission (NRC) were present. The purpose of the meeting was to explain how the decommissioning operations would proceed, and what people working in the nearby buildings could expect to see in the coming months. The major concern from the audience was not related to radiological concerns, but more about the potential impact of equipment vibrations on laboratory instruments.
The progress of the project was monitored by the Reactor Committee which was composed of faculty and staff of the University of Illinois. One of the responsibilities of the Reactor Committee was to review and approve of several projects plans. With the onset of LVI mobilization, the Reactor Committee began to meet weekly.
The Illinois Department of Nuclear Safety provided four Air Particulate Detectors (APD). Each APD was placed outside of the NRL, and collected air samples. The air samples were filtered, and the particulates on the filter paper were placed in a counter onsite for gross alpha and gross beta. During the monitoring period, no detections were attributed to the decommissioning activities at the NRL.
The NRL was prepared for decommissioning by first removing and packaging wastes in the building that were created from the operation or experiments conducted within the reactor. These wastes included radiological debris, lead, and other hazardous materials [7]. One of the large radioactive items removed from the early decommissioning activities was a glove box located in the reactor room level. Contamination by fission products was suspected, and composite smear samples were taken to assess the extent of removable contamination (Table 3). The sum of fractions derived from plutonium and nickel-63 was about 83% allowed for Class A low-level radioactive waste (LLRW), and the glove box was placed in a Sealand El-087 container for shipping.
Table 3: Contamination of the glove box and waste classification.
As the glove box was removed, the process of removing the interference (items such as piping, ductwork that may hinder major decontamination or demolition activities) from the reactor tank was begun. The reactor components removed were the control rod drives, guide tubes, the reactor bridge, fuel racks, and cooling lines [7]. The pool water in the tank was left to provide shielding from the rotary specimen rack. About 5,440 kg of activated aluminum accumulated from the reactor components and the secondary cooling piping. The summation of contact doses of the activated aluminum items was 578 mrem/hour. The average 1-meter dose per component was 0.39 mrem/hour. As anticipated, the major activation products were cobalt-60 (60Co) and iron-55 (55Fe), but the sum of fractions for all the activation products (1.07 x 10-7) was considerably less than Class A LLRW limits (Table 4). The wastes were shipped in a Sealand EL-087 container.
Table 4: Activated aluminum from all sources and waste classification.
When the dry active wastes and metal debris were combined, the net weight was 3,402 kg when shipped. The average 1-meter dose per item was 0.12 mrem/hour. While cesium-137 (137Cs) was present in the combined wastes, the sum of fractions was only 0.11% of the limit for Class A LLRW (Table 5). These wastes were also shipped in a Sealand EL-087 container.
Table 5: Dry active waste and metal debris from all sources and waste classification.
Following the removal of the metal interference from the reactor tank, the team from Cutting Edge Inc. began preparations to remove the cantilever and the upper portion of the bioshield. The purpose of this effort was to make the reactor and Bulk Shielding Tank more accessible to support the demolition of the lower activated concrete in the bioshield. However, before making the initial cuts in the cantilever, the top surface of the bioshield had to be decontaminated. Contamination levels of about 10,000 dpm beta/gamma were reduced to non-detectable levels by scrubbing the surface with a surfactantwater mixture.
Using air-cooled diamond core drills, a water-cooled track saw, and diamond wire saws, the Cutting Edge crew cut the cantilever into pieces, each weighing no more than three tons because of the limitations of the overhead crane that was available to move each piece down to the floor. The upper bioshield and the Bulk Shielding Tank were then cut into blocks and removed (Figure 4). The concrete pieces were shipped to Brickyard Disposal and Landfill south of Danville, Illinois, as non-radiological waste.
Figure 4: The upper bioshield being cut into blocks with a water-cooled concrete saw.
In January of 2012, the process of removing the rotary specimen rack (the “Lazy Susan”) was begun. Because of the dose given off by the rack (60 rem/hour contact), there was the need to hoist the rack up from reactor tank using the overhead crane, then lower it into a shipping container as quickly as possible without delays to minimize exposure to the LVI team. To accomplish this goal, a dimensional mockup was fabricated out of wood. The mockup was moved four times during dry runs, yielding valuable experience to the crane operator and floor crew. When the rotary specimen rack was lifted from the reactor tank (Figure 5), and lowered into the shipping container, the process was accomplished in 86 seconds, which included a brief pause to measure dose while suspended above the bioshield (180 mRem/hour at a distance of 3 meters). The shipping container was located on the southeastern corner of reactor floor of the NRL, and was covered with lead blocks and pieces.
Figure 5: Lifting the rotary specimen rack from the reactor tank. A shipping container is to the right on the floor of the NRL.
Because of elevated dose rate of the Lazy Susan, a robust Type A transportation cask was required to transport the rack. The cask used was a top-loading 6-80-2-4 Transportation Cask provided by EnergySolutions, LLC. After the cask was closed and sealed, a side contact of 0.1 mRem/hour was measured. The cask was then driven to EnergySolution’s Containerized Waste Facility in Clive, Utah.
After the removal of the rotary specimen rack, a remote-controlled demolition machine (a Brokk 330 with a demolition hammer attachment) was moved to the NRL to rubblize the activated concrete within the bioshield. An excavation was initially created in the south side of the bioshield (Figure 6). As the excavation progressed, activated metal components that were imbedded in the concrete such as the shadow shields were also removed. This phase of the project was completed in about one month. Approximately 43,137 kg of concrete rubble and rebar were placed in 15 Super Sacs, and shipped to Clive, Utah as a Class A low-level waste.
Figure 6: The excavation created in the south side of the bioshield using the Brokk demolition machine to remove activated concrete.
As the demolition of the activated portion of the bioshield was nearing completion, an unrelated and unfortunate discovery was made. A hole made in a concrete block of an inside wall of the NRL revealed the presence of asbestos-containing vermiculite that had apparently been blown into the concrete blocks. The presence of the insulation was not documented in the blueprints of the NRL or in any historical documents. From 1919 to 1990, 70% of all vermiculite used in the U.S. was from a mine in Montana, and asbestos veins occurred with the vermiculite [8]. Because the asbestos-containing vermiculite was not anticipated, neither time or resources had been budgeted for its removal. The removal of the insulation was, however, accomplished in about a month by removing the concrete blocks that made up the walls, entombing the NRL with opaque sheets of plastic, and applying a slight negative pressure to reduce particle emissions while an LVI Asbestos Abatement crew removed and disposed the material.
Following the resolution of the asbestos discovery, the final phase commenced with the demolition of the remaining infrastructure. Using a large track hoe with various demolition attachments, the remaining building was demolished in two weeks [9]. As the concrete pylons were separated, the roof collapsed downward more quickly than anticipated. It was then discovered that there was a 15-cm layer of concrete on the roof that added weight to the collapsing structure. Like the asbestos, the presence of the concrete layer had not been recorded in the building blueprints or in any historical documents.
About 227 m3 of concrete waste were generated from the floor of the NRL, and the bioshield demolition [9]. Following the removal of scrap metal and debris, the remaining bioshield was reduced to rubble, and the debris was removed, leaving an excavation that was about 3 m below grade (Figure 7).
Figure 7: The excavation created after the removal of the rubble and debris. After the site was released without restrictions, the excavation was filled with soil and graded.
A Final Status Survey (FSS) written to comply with MARSSIM protocols was then conducted to document that the NRL site compiled with the radiological release criteria that were approved by the NRC for the termination of NRC License R-115. As part of the FSS, soil samples were collected on-site for comparison with the Derived Concentration Guideline Levels (DCGL) given in Table 6. A sampling grid was imposed and divided into two areas. The excavated area was designated as a Class 1 Survey Area because it was anticipated that it was more likely that residual contamination would be in the excavation. The perimeter of the excavation was designated as a Class 2 Survey Area, the area in which contamination was not as likely to be present. Fifty-six surface soil samples were collected in the excavation, and 28 samples were collected in the Class 2 area. The soil samples were submitted to Teledyne Brown Inc. for gamma spectroscopic analysis.
Table 6: Summary of radionuclide concentrations in soil samples collected for the Final Status Survey.
Of the eight radionuclides chosen for the FSS, only Cs137 was present in concentrations that were greater than the Minimum Detectable Concentration (MDC) (0.05 pCi/g), and in only two of 56 samples collected in the Class 1 area (Table 6). Four of the soil samples collected in the Class 2 area contained detectable Cs137. The largest concentration (0.25 pCi/g) represented only 2.3% of the release criterion. Although the Cs137 could have been derived from the NRL, it was more likely from fallout during the above-ground testing of nuclear weapons from 1945 to 1962. The concentration of Cs137 in the FSS were the same order of magnitude as those detected in undisturbed soils impacted by radioactive fallout [10]. Moreover, Cs137 has been detected in undisturbed soils in Champaign County, Illinois, at depths of 0 to 15 cm in concentrations of 0.32 to 0.64 pCi/g [Roy, unpublished data].
Posssium-40 and thorium-228 were detected in every soil sample, yielding mean concentrations of 11.04 and 0.49 pCi/g, respectively. Thorium-232 was detected in 91% of the soil samples, yielding a mean of 0.46 pCi/g. Radium-226 was quantified in 41% of the samples, yielding a mean of 1.70 pCi/g. The concentrations of 140Ba, 7Be, 141Ce, 144Ce, 134Cs, 58Co, 59Fe, 131I, 54Mn, 103Ru, 106Ru, 235U, 238U, 65Zn, and 95Zr were less than their respective MDC.

Lessons Learned

• A detailed site characterization avoids surprises: no unknown areas of contamination or radioactivity were found because of the comprehensive radiological site characterization.
• Good communication help: a weekly newsletter was circulated to the NRC, University management, and other interested parties. This kept all parties informed on current status and short-term goals. The NRC later used this newsletter as an example to other facilities.
• Practice potential high-dose evolutions to reduce dose commitment: the removal of the rotary specimen rack was practiced with a mockup. The first attempt took more than five minutes to complete, while the actual move was performed in 86 seconds.
• Allow for contingency funds: the discovery of asbestoscontaining vermiculite was an undocumented surprise that added $400,000 to the project, predominately covered by the contingency funds.

Summary

The Nuclear Research Laboratory (NRL) at the University of Illinois was decommissioned in about 10 months after being in SAFSTOR for about 12 years. The demolition of the activated portion of the bioshield, activated aluminum, graphite, dry active waste, and metal debris resulted in 125,802 kg (198 m3) of Class A radioactive waste. The total activity of the waste was 4.316 Ci of which the majority of the activity was from a rotary specimen rack that contained Stellite bearings. There were 260 m3 of asbestos waste resulting from the discovery of asbestos-containing vermiculite insulation in the walls of the NRL. The overall occupational dose was 2.568 man-rem which was less than that anticipated in the decommissioning plan. The overall cost of the project was about $4.3 million which was slightly more than that initially budgeted. The site was released without restrictions after the approval of the Final Status Survey. Among the lessons learned about decommissioning a university reactor laboratory is that historical documents used to plan and budget a project may be incomplete. Unknown site features may require revisions of plans, and additional time and expenses to meet project goals.

Acknowledgment

We thank Dan Jordan, Enercon Waste Manager, for all his on-site training and enthusiasm about low-level waste classification and shipping.

References











Track Your Manuscript

Media Partners