Problems And Solutions Relative To Subsampling And Analysis Of Tritium In Concrete

1
Problems And Solutions Relative To Subsampling
And Analysis Of Tritium In Concrete

• Thomas L. Rucker, Ph.D., SAIC
• Claude F. Laney, CHP, IES, Corp.
• C. Martin Johnson, Jr., U.S. Air Force, and
• Mike R. McDougall, Eberline Services

2
The Site Background

• The Carolinas-Virginia Nuclear Power Associates,
  Inc. (CVNPA) experimental nuclear power plant in
  Jenkinsville, South Carolina (a.k.a the Parr
  Site) is being decommissioned and demolished.
• Includes verifying that site soils, groundwater,
  remaining structures of the Reactor Building
  (RB), comply with the State of South Carolina
  dose limit of 15 mrem/yr.
• Included sampling and analysis of remaining
  structures of the Reactor Building, the Fuel
  Handling Building, and the Auxiliary Building.

3
Reactor Building, Fuel Handling Building, and
Auxiliary Building
4
Inside Reactor Building
5
Inside Reactor Building
6
Inside Auxiliary Building
7
Inside Fuel Handling Building
8
Building Concrete Characterization

• All concrete samples were analyzed for Co-60 and
  Cs-137.
• At least 10 were originally were analyzed for
  H-3, I129, C-14, Sr-90, and Tc-99.
• Initial results identified H-3 as the primary
  structural contaminant impacting status of the
  consolidated mass.
• The frequency of H-3 analysis was raised to at
  least 20 percent in some survey units and 100
  in other survey units to supplement the earlier
  data.

9
Concrete Sample Collection

• Three methods were used to collect concrete
  samples
• Hammer drill for poured concrete when secondary
  COC analyses were not required,
• Coring machine for poured concrete when secondary
  COCs were required, and
• Hand-held hammer for cinder blocks (primary and
  secondary COCs).
• Concrete coring was more labor-intensive but
  necessary to minimize H-3 (and possibly Tc-99)
  volatilization.
• The coring bit was connected to a water supply
  for cooling.
• A 2-inch diameter rotating bit was advanced
  approximately 11 inches using a turning wheel.

10
Concrete Core Drilling
11
Original Method Used for Tritium in Concrete

• The laboratory crushed the concrete core then
  mixed the crushate and randomly selected
  fragments for analysis.
• A high temperature oxidizer was used to
  super-heat the concrete fragments, and the
  condensate was collected.
• Tritium determined on condensate by liquid
  scintillation beta spectroscopy.

12
Sample Preparation Details

• The laboratory crushed the concrete core using a
  jaw crusher to pieces between 0.125-to-0.5 inches
  in size.
• Concrete fragments were mixed using the cone and
  quartering technique.
• The procedure was repeated several times taking
  care to minimize the loss of volatile compounds
  (e.g., H-3) during the process.
• Approximately 2 grams of concrete fragments were
  placed into a ladle of the high temperature
  oxidizer.

13
Crushed Concrete Sample
14
Crusher Debris Sub-samples
15
High Temperature Oxidizer
16
Initial Method Validation Results

• No blank contamination was identified.
• LCS recovery was acceptable.
• Excessive variability of both the laboratory and
  field duplicate results
• DER of less than 1.29 (99 C.L.) and/or RPD of
  less than 35 were used as the evaluation
  criteria.
• The DER criterion was exceeded in 4 of the 5
  laboratory duplicates 54.2, 24.9, 25.2, and
  104.2. The RPD criterion was also exceeded on the
  same three packages 85.9, 142, 102, and 184.
  
• Field duplicates sample results showed DERs at
  292 and 12.6 and RPDs at 199 and 41.

17
(No Transcript)
18
Problems Identified

• Excessive variance of duplicate results was
  attributed to sub-sampling errors due to the
  large particle size as compared to the small
  aliquot size and the heterogeneity of the
  contamination in the sample matrix (across 6 in.
  core).
• Due to the high variability of H-3 duplicate
  results, the original data were assumed to be not
  representative of the average concentration in
  concrete cores.

19
Theory of Particulate Subsampling from MARLAP
Chapter 12

• A subsample is guaranteed to be unbiased only if
  every particle in the sample has the same
  probability of being selected for the subsample.
• The weight of the subsample should be many times
  greater than the weight of the largest particle
  in the sample.
• The variance associated with subsampling may be
  reduced either by increasing the size of the
  subsample or by reducing the particle sizes
  before subsampling.

20
Concerns for a Modified Method

• Representative Subsampling
• Client wanted to use a larger aliquot size for
  analysis.
• Smaller particle size for subsampling.
• Volatility of 3H during crushing and grinding
  process.
• Heat generated while grinding to fine particle
  size could volatilize 3H.
• Extraction Efficiency
• Efficient recovery of 3H from the matrix

21
Revised Sub-sampling Method

• Place a 900g sample in a freezer at 0C (32F) or
  less overnight to cool the sample in preparation
  for crushing.
• Crush the sample using a jaw crusher to achieve a
  particle size of less than 1 cm.
• Cool the samples overnight at -78C.
• Reduce the sample to a mesh size of 5-10 (3/16
  in) using a plate grinder.
• Mix each entire individual ground sample
  thoroughly to achieve a homogenous sample mixture
  followed by coning and quartering to obtain
  approximately 100g of sample.

22
Plate Grinder
23
Ground Sample Cone and Quartering
24
Extraction and Analysis Method

• Place 100 g of sample in a 1 liter flask
  containing 500 ml of dead water.
• Shake / water equilibrate for 24 hours
• Distill the sample and/or water.
• Place an aliquot of the distillate in
  scintillation cocktail.
• Perform analysis of tritium in a liquid
  scintillation counter.

25
Shaking andDistillation Flask
26
Distillation Apparatus and Distilled Samples
27
(No Transcript)
28
(No Transcript)
29
(No Transcript)
30
(No Transcript)
31
Old Method Verses New Method
32
Old Method Verses New Method
33
Old Method Verses New Method
34
Old Method Verses New Method
35
Conclusions

• The weight of the subsample must be many times
  greater than the weight of the largest particle
  in the sample. (MARLAP Chpt. 12)
• A subsample size of 100g was successfully
  analyzed by Water Equilibrium and Distillation
  Method.
• Volatilization of H-3 during sample processing
  was minimized by freezing.

36
Conclusions (Cont.)

• Matrix spike recoveries between 99 and 106 were
  achieved.
• DERs of less than 1.29 (99 C.L.) and/or RPDs of
  less than 14 were achieved.
• Minimum Detectable Amounts (MDAs) of 3 pCi/g
  were obtained.

37
Questions?